Pulp and Paper Chemistry and Technology Volume 4

Paper Products Physics and Technology

Edited by Monica Ek, Göran Gellerstedt, Gunnar Henriksson
Pulp and Paper Chemistry and Technology
Volume 4

This project was supported by a generous grant

by the Ljungberg Foundation (Stiftelsen Erik
Johan Ljungbergs Utbildningsfond) and
originally published by the KTH Royal Institute of
Technology as the “Ljungberg Textbook”.
Paper Products Physics and Technology

Edited by Monica Ek, Göran Gellerstedt,

Gunnar Henriksson
Editors
Dr. Monica Ek
Professor (em.) Dr. Göran Gellerstedt
Professor Dr. Gunnar Henriksson
Wood Chemistry and Pulp Technology
Fibre and Polymer Technology
School of Chemical Science and Engineering
KTH ⫺ Royal Institute of Technology
100 44 Stockholm
Sweden

ISBN 978-3-11-021345-4

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed
bibliographic data are available in the Internet at http://dnb.d-nb.de.

” Copyright 2009 by Walter de Gruyter GmbH & Co. KG, 10785 Berlin.
All rights reserved, including those of translation into foreign languages. No part of this book may
be reproduced or transmitted in any form or by any means, electronic or mechanic, including
photocopy, recording, or any information storage retrieval system, without permission in writing
from the publisher. Printed in Germany.
Typesetting: WGV Verlagsdienstleistungen GmbH, Weinheim, Germany.
Printing and binding: Hubert & Co. GmbH & Co. KG, Göttingen, Germany.
Cover design: Martin Zech, Bremen, Germany.
Foreword

The production of pulp and paper is of major importance in Sweden and the forestry industry has
a profound influence on the economy of the country. The technical development of the industry
and its ability to compete globally is closely connected with the combination of high-class education,
research and development that has taken place at universities, institutes and industry over many
years. In many cases, Swedish companies have been regarded as the initiator of new technology
which has started here and successively found a general world-wide acceptance. This leadership in
knowledge and technology must continue and be developed around the globe in order for the
pulp and paper industry to compete with high value-added forestry products adopted to a modern
sustainable society.
The production of forestry products is based on a complex chain of knowledge in which the biologi-
cal material wood with all its natural variability is converted into a variety of fibre-based products,
each one with its detailed and specific quality requirements. In order to make such products, knowl-
edge about the starting material, as well as the processes and products including the market demands
must constitute an integrated base. The possibilities of satisfying the demand of knowledge require-
ments from the industry are intimately associated with the ability of the universities to attract
students and to provide them with a modern and progressive education of high quality.
In 2000, a generous grant was awarded the Department of Fibre and Polymer Technology at KTH
Royal Institute of Technology from the Ljungberg Foundation (Stiftelsen Erik Johan Ljungbergs
Utbildningsfond), located at StoraEnso in Falun. A major share of the grant was devoted to the
development of a series of modern books covering the whole knowledge-chain from tree to paper
and converted products. This challenge has been accomplished as a national four-year project in-
volving a total of 30 authors from universities, Innventia and industry and resulting in a four
volume set covering wood chemistry and biotechnology, pulping and paper chemistry and paper
physics. The target reader is a graduate level university student or researcher in chemistry / renew-
able resources / biotechnology with no prior knowledge in the fields of pulp and paper. For the
benefit of pulp and paper engineers and other people with an interest in this fascinating industry,
we hope that the availability of this material as printed books will provide an understanding of all
the fundamentals involved in pulp and paper-making.
For continuous and encouraging support during the course of this project, we are much indebted
to Yngve Stade, Sr Ex Vice President StoraEnso, and to Börje Steen and Jan Moritz, Stiftelsen Erik
Johan Ljungbergs Utbildningsfond.

Stockholm, August 2009 Göran Gellerstedt, Monica Ek, Gunnar Henriksson

List of Contributing Authors

Bo Andreasson Christer Söremark

Expancel AB Christer Söremark AB
Box 13000 Hagav. 10
850 13 Sundsvall, Sweden 944 72 Piteå, Sweden
bo.andreasson@akzonobel.com christer@soremark.se

Anthony Bristow Göran Ström

Bristow Consulting AB Innventia AB
146 38 Tullinge, Sweden Drottning Kristinas väg 61
bristow@beta.telenordia.se Stockholm, Sweden
goran.strom@innventia.com
Christer Fellers
Johan Tryding
Innventia AB
Tetra Pak Packing Solutions AB
Drottning Kristinas väg 61
Ruben Rausingsgata
Stockholm, Sweden
221 86 Lund, Sweden
christer.fellers@innventia.com
johan.tryding@tetrapak.com

Mikael Nygårds Lars Wågberg

Innventia AB KTH Royal Institute of Technology
Box 5604 Chemical Science and Engineering
114 86 Stockholm, Sweden Fibre and Polymer Technology
mikael.nygards@innventia.com 100 44 Stockholm, Sweden
wagberg@pmt.kth.se
Sören Östlund
KTH Royal Institute of Technology Torbjörn Wahlström
School of Engineering Sciences The Packaging Greenhouse
Department of Solid Mechanics Axel Johnsons v. 6
100 44 Stockholm, Sweden 652 21 Karlstad, Sweden
soren@kth.se torbjorn@thepackaginggreenhouse.com
 ix

`çåíÉåíë

1 The Structure of Paper and its Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Christer Fellers
2 Paper Physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Christer Fellers
3 Development of Paper Properties during Drying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Torbjörn Wahlström
4 The Interaction of Paper with Water Vapour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Christer Fellers
5 Optical Properties of Pulp and Paper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Anthony Bristow
6 On the Mechanisms behind the Action of Dry Strength and Dry Strength Agents . . . . . 169
Lars Wågberg
7 On the Mechanisms Behind the Action of Wet Strength and Wet Strength Agents . . . . 185
Bo Andreasson and Lars Wågberg
8 The Surface of Paper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Anthony Bristow
9 Paper and Printing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Göran Ström
10 Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Christer Söremark and Johan Tryding
11 Laminate Theory for Papermakers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Christer Fellers
12 Mathematical Modelling and Analysis of Converting and Enduse . . . . . . . . . . . . . . . . . 315
Mikael Nygårds and Sören Östlund

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
100
 1

N==qÜÉ=píêìÅíìêÉ=çÑ=m~éÉê=~åÇ=áíë=jçÇÉääáåÖ

Christer Fellers
STFI-Packforsk AB

1.1 Functional Paper Properties 1

1.2 The Complexity of a Model 2

1.3 Modelling at Different Structural Levels 2

1.4 The Physical Nature of Paper 3

1.5 Factors Affecting the Mechanical Properties of Paper 4

1.6 The Nature of Paper 5

1.7 The Paper Structure 6

1.8 Paper Divided into Categories 8

1.9 The Structure of some Paper Products 9

1.10 Properties of Different Materials Compared to Paper 11

1.10.1 General Considerations 11
1.10.2 Textiles 12
1.10.3 Fibre Textiles 12
1.10.4 Polyethylene Film 12

1.11 Standardization 13

NKN==cìåÅíáçå~ä=m~éÉê=mêçéÉêíáÉë

During manufacturing, converting and end-use of paper there are many demands for its func-
tional mechanical properties. A few illustrative examples of this are given below.
• Example 1) Bending stiffness paper is probably the most important mechanical property of
paper and carton board. Figure 1.1 illustrates the obvious importance of bending stiffness
for a milk carton.
2

Figure 1.1. The bending stiffness is important for milk cartons as well as most paper grades.

• Example 2) During storage of corrugated boxes and cartons, the structure may collapse due
to creep in compression forces after a certain time of loading, Figure 1.2. The term lifetime
is sometimes used to describe the time to break.
It was shown already half a century ago that the lifetime decreases linearly in a semi-loga-
rithmic plot as shown in Figure 1.3.
• Example 3) During manufacturing, rewinding and printing, thin paper grades may experi-
ence web breaks. The breaks may be caused by high loads, defects or inferior fracture prop-
erties. Fig. 1.4 shows a web break in a paper machine.
• Example 4) Sacks are expected not to break during filling, transportation and end-use. In
the worst case the sack breaks which is shown in the following Figure 1.5.
• Example 5) During certain printing operations, for example as offset printing, the material
is exposed to high stresses in the thickness direction which may lead to delamination, Fig-
ure 1.6.
Many functional properties may be difficult to evaluate for the purpose of trade and pulp
evaluation. However efforts are made to standardise important methods for property evaluation
in ISO and national standards.
 3

Figure 1.2. Compressive collapse of corrugated boxes after a long time of loading.
4

100

80
(% of the short time strength)

60
force on the box

40

20

0
0.0001 0.001 0.01 0.1 1 10 100
days of loading

Figure 1.3. The lifetime, the time to failure of corrugated boxes as a function of the time of loading.

Figure 1.4. A web break in a paper machine. (Nordiskafilt).

 5

Figure 1.5. Break of a sack filled with cement.

Figure 1.6. Delamination of a carton board in offset printing.

NKO==qÜÉ=`çãéäÉñáíó=çÑ=~=jçÇÉä

In research and development of paper grades it is often useful to model the paper behaviour. De-
pending on the actual loading situation, the time scale involved and the environmental condi-
tions, different models must be used. For example, if the time scale is short, the model need not
include time as a variable. If the deformation is small, the material may be regarded as elastic.
The more complicated the loading situation is the more complex the model needs to be.

NKP==jçÇÉääáåÖ=~í=aáÑÑÉêÉåí=píêìÅíìê~ä=iÉîÉäë

During product development the paper may be modelled at different structural levels, depend-
ing on the purpose of the modelling. The descriptions of the structure may be expressed at dif-
ferent structural levels, the structure hierarchy, Figure 1.7.
6

structure

sheet
ZD

CD

MD
network

fiber

S1
S2
S3

fibril

cellulose
OH
OH
OH

OH OH

OH

Figure 1.7. The descriptions of the structure may be expressed at different structural levels, the structure hierar-
chy.

Below are listed some concepts to consider related to modelling:

• At the level of modelling, the properties of the material are considered to be constant in
small volumes (homogeneous), although the entire body can be heterogeneous, for example
made of several materials or have density variations.
• The structure of the material at smaller scales is neglected
• At the level of modelling, quantities of interest, for example the deformation of the mate-
rial, are viewed as varying in a continuous matter.
• Continuum models typically predict the “average” response of the material at the level
under consideration
Figure 1.8 illustrates that paper often is modelled as a continuum, which means that the
structure characteristics are disregarded. In other instances, the relation between the paper prop-
erties and the constituents are of interest.
 7

Figure 1.8. Paper is often modelled as a continuum, which means that the structure characteristics are disre-
garded. In other instances, the relation between the paper properties and the constituents are of interest.

For paper materials, each modelling level has particular advantages and disadvantages.
Typical uses:
• Macro level: predict behaviour of paper structure, for example boxes
• Meso level: Formation studies
• Micro level: Predict the influence of fibre and bond properties on sheet (macroscopic) prop-
erties
• Sub-micro level: Understand the influence of hydrogen bonding on the sheet (macroscopic)
properties
Even if the material is modelled as a continuum, different degrees of complexity may be nec-
essary to introduce in order to describe the material behaviour. The more levels of structure hi-
erarchy that are introduced, the more complex will the model be if all features shall be captured.
In reality one has to make shortcuts in the modelling. The following degrees of complexity may
be introduced, in order from simple to difficult.
• Elastic: The material is like a spring -all deformations are recovered when the loading is
removed
• Plastic: Permanent deformations occur which are non-recoverable
• Elastic-Plastic: The material is first elastic, and then plastic
• Viscoelastic: The material experiences elastic and viscous (time dependent) deformations
• Viscoplastic: Combines viscoelastic and plastic behaviour
For paper (or any material), you pick the most simple constitutive model that can accurately
simulate the material behaviour under consideration.
Examples:
• Elastic: Short term bending of corrugated board
• Elastic-Plastic: Embossing
• Viscoelastic/Viscoplastic: Long term stacking of containers
8

NKQ==qÜÉ=mÜóëáÅ~ä=k~íìêÉ=çÑ=m~éÉê

The physical properties of paper are evaluated by many different methods. This chapter treats
the most common of these methods. Testing of paper is carried out for different reasons, for ex-
ample for production control, as a measure of the properties of the paper in trade between seller
and buyer, and for purely functional reasons, where the paper shall function in a printing press,
in boxes etc.
Paper belongs to the field of polymeric materials. Its mechanical properties are thus much
more similar to different plastic materials than to for example metals. Characteristic of polymer-
ic construction materials is that the reaction of the material to mechanical stresses is time-de-
pendent. And it is not possible, as it is for many metals, to distinguish a purely elastic region at
small stresses followed by a well-defined yield point. Already under very small deformations of
paper, viscoelastic effects occur, i.e. the properties depend on how quickly the test is carried
out. A striking example of this is that the deflection of a unilaterally clamped horizontal paper
strip under the influence of gravity increases with time. Under large deformations, paper is also
permanently plasticized.
Technological tests therefore require a defined time schedule. It is essential to choose test
methods where the time scale is best adapted to the intended practical use of the paper.
As an indication of the magnitude of the rate dependence, it can be mentioned that the
strength increases by 5–10 percent when the rate of elongation is increased 10 times. For this
reason, all standard methods specify the rate of elongation.
The viscoelastic properties of paper are quantified, for example, by testing at a constant rate
of elongation, by a constant stress increase per unit time or by exposing the sample to sinusoidal
tensile force cycles at small deformations.
If a paper is deformed and is kept deformed at a certain elongation, the force decreases with
time. This phenomenon is called stress relaxation. An example when relaxation phenomena are
important in the use of paper is that a sheet of paper, which is left in a typewriter, adopts a tubu-
lar appearance.
If a paper is held under a constant force, the elongation increases with time. This phenome-
non is called creep. When corrugated cardboard boxes or cartons are stacked in a storeroom,
they creep under the compression load. The creep can lead to a collapse of the boxes after a cer-
tain time. The concept of lifespan is used sometimes to characterize this phenomenon.
Paper is a hygroscopic material and its properties are dependent on the relative humidity
(RH) and on the temperature in the surrounding air. Most paper tests are now carried out at a
standardized humidity of 50 per cent RH and at a temperature of 23 °C. In the choice of test
method, it must be decided whether this environment is representative for the technical use of
the paper. For example, paper at high temperatures and low relative humidity can show brittle
failure, while a paper creeps more rapidly in a moist environment.

NKR==c~Åíçêë=^ÑÑÉÅíáåÖ=íÜÉ=jÉÅÜ~åáÅ~ä=mêçéÉêíáÉë=çÑ=m~éÉê

The mechanical properties of a paper depend on a number of factors, the most important of
which are:
 9

• Botanical factors: The structure, length distribution, fibre wall thickness and proportion of
lumen in the fibres.
• Chemical factors: The degree of delignification, the degree of polymerisation of cellulose
DP, the content and type of hemicellulose.
• Papermaking factors: Beating, forming, pressing, drying and calendering.
• Chemical environment: The presence of electrolytes, polyelectrolytes and surfactant sub-
stances.
A theoretical analysis of the mechanical properties of paper is made more difficult by the fact
that the heterogeneous fibre material cannot be characterized uniformly. Proposed theories use
strongly idealized models.
Most theories for mechanical properties are based on the assumption that these properties
can be changed if the bonded area and strength between the fibres are changed. Many methods
have been tested in order to measure the bonded area, for example by nitrogen adsorption, mea-
surement of the bonded area between crossing fibres in a microscope, measurement of the light
scattering coefficient and determination of the density of the whole sheet.
Optical and mechanical properties of paper are always related, but in different ways depend-
ing on the type of pulp and on papermaking unit-operations.
For instance, the light scattering coefficient, which is a measure of the light scattering sur-
face area in the sheet, can for chemical pulps under very special conditions be used to suggest
the relative degree of bonding.
On the other hand, for mechanical pulp, the light scattering coefficient increases with in-
creasing energy input in the refining as a result of a large number of light scattering particles are
formed.
Density of a paper is increased by beating or pressing and will increase the chance for bond-
ing since the fibres to come closer to each other.
It is also shown in practice that the density is a good way of characterizing the way in which
the structure of the paper influences mechanical properties.
It is possible to find empirically certain relationships between the mechanical properties and
density of the sheet as below, Equation (1.1).

property k ˜ Ua (1.1)

The exponent a, varies with fibre source and all papermaking conditions.
Although paper fibres have a considerable strength in both the dry and the wet states, paper
has a very low strength when it is wet. Thus, it is evident that the strength of paper cannot be re-
lated in a simple way to the strength properties of the fibres.

NKS==qÜÉ=k~íìêÉ=çÑ=m~éÉê

There is no strict definition of paper. The main component of paper has always been fibrous
material. The most common are plant fibres from wood, grasses and cotton. Animal fibres are
also used, e.g. wool in roof board and leather in leather board. Synthetic fibres have a certain
importance and can be both inorganic such as glass fibres and organic such as synthetic poly-
mers from petroleum products.
10

Paper can be described as a layered structure consisting of fibres, which are more or less flat-
tened. A large amount of fibre fragments, so-called fine material, is usually present. The fibres
are pre-treated through mechanical treatment, beating, and are bonded to each other without the
need for other binding substances.
A large flexibility in the fibres and the presence of fine material increases the contact area
and improves the bonding ability between the fibres.
Paper contains about one million fibres per gram. Besides the fibrous material, many types of
paper contain considerable amounts of filler, usually a natural mineral, e.g. kaolin, finely
ground marble, chalk or talcum. The fillers, which are cheaper than the fibres, improve the opti-
cal properties and printability of the product. A large number of chemicals are also used as aids
in the paper making process.
Besides being our most important information-bearer, paper is an important and unique con-
struction material.
Paper is a cheap product due to the relatively cheap raw material, the fact that the manufac-
turing process is in principle simple and the fact that no binders need to be added to bind the fi-
bres together. The manufacturing process is also highly automated.
Paper can be re-utilized, it is combustible, it contains no heavy metals or agents toxic to the
environment, and it is environment-friendly.

NKT==qÜÉ=m~éÉê=píêìÅíìêÉ

Paper is formed through filtration of a fibre suspension on a wire, after which the wet product is
pressed and dried. Due to the forming process and the fact that fibres are much longer than their
thickness, the fibres become oriented more or less in the plane of the paper and paper becomes a
layered structure. Usually three directions in paper are of interest (Figure 1.9), the machine di-
rection MD, the cross-machine direction CD and the thickness direction ZD.

MD
CD

ZD

Figure 1.9. Definition of the three directions in paper.

Figure 1.10a, shows schematically the thickness structure of paper formed at a low consis-
tency. A more three-dimensional structure of paper can be achieved if the forming is performed
at higher consistency, Figure 1.10b.
 11

a) LC-structure

b) HC-structure

Figure 1.10. The structure of paper in the thickness direction. a) LC, low consistency, b) HC, high consistency

Since the fibres are much stronger in the longitudinal direction than in the transverse direc-
tion paper is much stronger in the plane of the paper than in the thickness direction.
By controlling the sheet forming, different types of more or less marked fibre orientations
can be produced. The anisotropy, i.e. the relationship between a property in the machine direc-
tion and in the cross-direction, is influenced by the fibre orientation and by the deformation of
the paper during the manufacturing process.
The forming consistency, the hydrodynamic conditions during forming and the fibre dimen-
sions and properties determines how well the fibres are distributed in the plane of the paper. The
term formation is used for this. Figure 1.11 shows a paper with bad formation and a good for-
mation. The formation affect the appearance, the printing properties and the strength of the pa-
per. Controlling the formation is an issue every day for the papermaker.

Figure 1.11. A paper with bad and good formation.

The fibre orientation can be controlled in the forming process and influences all physical
properties of the paper. In reality, paper never has a completely uniform distribution of the fi-
12

bres in the plane of the paper, but the fibres flock together in a spectrum of sizes. The lower the
forming concentration, the more even is the grammage distribution obtained. The task of the pa-
permaker is to produce as even a sheet as possible. The more evenly formed the sheet is, the bet-
ter are the properties of the sheet in general, both from a mechanical viewpoint and with regard
to surface smoothness and printability.
Different properties have different sensitivities to the forming concentration. Most properties
deteriorates with increasing forming concentration. The most sensitive property is the tensile
energy absorption index, tensile index, bursting index and strain at break show the same trend
but are not quite as sensitive. Tensile stiffness index and compression index are relatively insen-
sitive while the strength in the thickness direction increases with increasing concentration.
Thick paper is called carton board or board. The boundary between paper and carton board is
indistinct, except in the customs duties imposed in certain countries. Carton board is the raw
material for the production of packages of different kinds and boxes.
Fibre boards, so-called wallboard or hardboard, can be regarded as thick, stiff paper and they
are manufactured according to paper technological methods.
Paper need not be a plane product. Egg-packages are manufactured according to normal pa-
per technology, but are shaped in a special mould, which gives the product its final appearance.
Paper is nevertheless normally smooth, feels stiff to the hand and is slightly extensible. It is
easy to tear paper.
It is also easy to fold, and the crease becomes permanent. Figure 1.12 shows a fold failure in
a paper. As a consequence of the layered structure of paper, the folding ends up in a delamina-
tion between the fibres in the thickness direction and has resulted in a buckling on the compres-
sion side of the paper.

Figure 1.12. Folding failure in paper.

Paper can be folded and reopened many times like a hinge. In the folding of a package, fold-
ing marks are made in the carton board, so-called creases, without reducing the tensile strength
noticeably, Figure 1.13 and Figure 1.14.
 13

a b

c d

Figure 1.13. Schematic illustration of creasing operation and the bending of the creased carton board.

Figure 1.14. The bending of the creased carton board.

If the paper is wrinkled, permanent folds are formed with a rustling sound. It is not possible
to bend paper to a double-curved surface without a fold or dent arising.
During drying, the fibres become stiff and are locked into position. The high strength is es-
sentially due to the fact that surface tension forces during the drying draw the fibres together so
that so-called hydrogen bonds can be formed between the fibres. These chemical forces are
14

weaker than the forces, which hold the cellulose molecules together in the fibres, but are stron-
ger than most other physical attraction forces.
Most of the deformation under stress comes from the fibres themselves, with relatively small
contributions due to a reorganization of the fibres.
Delamination may be a problem during offset printing, plastic coating, corner gluing of box-
es and several other converting operations. Figure 1.15 shows one example, the delamination
between the surface layer and the layer below after offset printing. Naturally the product is sub-
ject to complaint from the printer.

a)

b)

Figure 1.15. Delamination of a paper after offset printing. A) The surface, B) The thickness direction.

Paper loses its strength on wetting since the hydrogen bonds between the fibres are broken
and the fibres become soft. Wet strength can be developed with special chemicals.
Although the papermaking process is simple, the mechanical properties of paper can be var-
ied strongly and adapted to end-use requirements.
Pulps of different kinds of fibre and different production procedures give papers with differ-
ent properties.
 15

If rigid papers are desired, they can be built-up of several layers. According to the I-beam
principle, the surface layers should be non-elastic and consist of highly beaten pulps whilst the
central layer should be made from pulps with the lowest possible density.
By controlling the drying shrinkage, the paper properties may be changed dramatically.

NKU==m~éÉê=aáîáÇÉÇ=áåíç=`~íÉÖçêáÉë

There is no simple way of dividing papers into different categories. A division of papers can be
made according to use as in Table 1.1.

q~ÄäÉ=NKNK=aáîáëáçå=çÑ=ÇáÑÑÉêÉåí=íóéÉë=çÑ=é~éÉêë=~ÅÅçêÇáåÖ=íç=~êÉ~ë=çÑ=ìëÉK

qÜáå=ïê~ééáåÖ=é~éÉêë péÉÅá~äíó=é~éÉêë
_~Ö=é~éÉê mÜçíçÖê~éÜáÅ=é~éÉêë
p~Åâ=é~éÉê bäÉÅíêáÅ=áåëìä~íáåÖ=é~éÉêë
båîÉäçéÉ=é~éÉê Å~ÄäÉ=é~éÉêëI
éêÉëëé~å=Ñçê=ãçíçê=ïáåÇáåÖI=
qÜáÅâ=ïê~ééáåÖ=é~éÉêë
íê~åëÑçêãÉêë
`~êíçå=Äç~êÇ=
cáäíÉê=é~éÉêë=Ñçê=~áêI=ï~íÉê=~åÇ=çáä
iáèìáÇ=Å~êíçå=Äç~êÇ
`áÖ~êÉííÉ=é~éÉê
mêáåíáåÖ=é~éÉêë dóéëìã=Äç~êÇ=Å~êíçå=Äç~êÇ
kÉïëéêáåí=EÇ~áäó=åÉïëé~éÉêëF t~ääé~éÉê
j~Ö~òáåÉ=é~éÉê=Eã~Ö~òáåÉëI=Å~í~äçÖìÉëI=~ÇîÉêJ _ççâJÄáåÇáåÖ=é~éÉê
íáëÉãÉåíëF `çêÉ=Å~êÇÄç~êÇ
råÅç~íÉÇ=Ep`F qÉÅÜåáÅ~ä=Çê~ïáåÖ=é~éÉêë
`ç~íÉÇ=Eit`F bãÉêó=é~éÉê
cáåÉ=é~éÉêë pÉÅìêáíó=é~éÉê=Ñçê=Ä~åâåçíÉë=~åÇ=ÅÜÉèìÉë
råÅç~íÉÇ=EÉKÖK=Å~êíçå=Äç~êÇ=é~éÉê=~åÇ=Åçéó= dêÉ~ëÉéêççÑ=é~éÉê
é~éÉêF pÉäÑJÅçéóáåÖ=é~éÉê
`ç~íÉÇ=EÉKÖK=~ÇîÉêíáëáåÖ=é~éÉêF
eóÖáÉåáÅ=é~éÉêë
pçÑí=ÅêÆéÉ=é~éÉê
qçáäÉí=é~éÉê
háíÅÜÉå=é~éÉê

NKV==qÜÉ=píêìÅíìêÉ=çÑ=ëçãÉ=m~éÉê=mêçÇìÅíë

Figure 1.16–1.24 shows the structure of some paper products, chosen to illustrate the large vari-
ety of properties of paper. By changing features of the pulp- and papermaking processes, the
structure of paper can be varied across wide limits. The structure is adapted to give suitable
properties to a given paper. In some of the figures we see a corner of a paper, showing both the
surface and the thickness direction, in other cases the surface structure.
16

Figure 1.16. Linerboard, the surface-layer in corrugated cardboard, is manufactured mainly from bleached pine
Kraft pulp in a yield interval of 47–52 %. The pulp is only slightly beaten. The pulp and structure is chosen
mainly to give high compression strength, creep resistance, toughness in converting and delamination resistance.

Figure 1.17. Newsprint, may be manufactured from different fibre sources. The most recent pulp is thermome-
chanical pulp of softwood Still newsprint is manufactured from groundwood and from recycled newspapers. To
improve strength properties often the paper is reinforced with an addition, in the order of 5–10 % of chemical
softwood pulp. The pulp and structure is chosen mainly to give good runnability during manufacturing and end-
use, good opacity and printability.
 17

Figure 1.18. Fine paper, consists of chemical, undamaged fibres, which are often a mixture of hardwood and soft-
wood. Note the filler particles. The pulp and structure is chosen mainly to give good opacity and surface proper-
ties.

Figure 1.19. Tissue. Creped paper with low grammages, 14–25 g/m2. The type of fibre can vary. Note the creped
structure in the picture. (The magnification is slightly lower than in the other pictures).
18

a)

b)

Figure 1.20. Greaseproof paper. The fibres in the greaseproof paper are so well beaten that the fibre structure in
the paper has been "erased". The paper becomes transparent, has a low opacity. Greaseproof paper is used e.g. for
drawing, baking and sandwich paper.
 19

Figure 1.21. Copy paper, often consists of chemical, undamaged fibres, often a mixture of softwood and hard-
wood. Note the open paper structure. The fibres and structure are chosen mainly to give flatness and, good sur-
face properties.

Figure 1.22. LWC, Light Weight Coated paper, consists of mechanical, fibres, reinforced with chemical fibres
and with a coating layer. The fibres and structure are chosen mainly to give good runnability in printing presses
and to provide a good base paper for the coating.
20

Figure 1.23. Coated carton board. Carton board is manufactured by a multi-layer technology, e.g. with a surface
layer of bleached kraft pulp and mechanical pulp in the middle, or with several layers of the same type, e.g.
bleached Kraft pulp. The coating layer makes the carton board surface more even and more suitable for high-class
print. The fibres and structure is chosen mainly to give bending stiffness, good converting properties and surface
properties.

Figure 1.24. The coating layer of a coated carton board seen in the thickness direction. Note the cross-sections of
the fibres.
 21

NKNM==mêçéÉêíáÉë=çÑ=aáÑÑÉêÉåí=j~íÉêá~äë=`çãé~êÉÇ=íç=m~éÉê

NKNMKN==dÉåÉê~ä=`çåëáÇÉê~íáçåë

When materials are chosen for instance in airplanes, the important material property often is
tensile stiffness per weight, called specific elastic modulus.
For paper this property is called tensile stiffness index. If we then calculate the bending stiff-
ness for different materials based on equal weight (Bending stiffness index), the following re-
sults are obtained, Table 1.2. It is apparent that paper belongs to the stiffest materials in this
respect.

q~ÄäÉ=NKOK=j~íÉêá~ä=éêçéÉêíáÉëK

j~íÉêá~ä bä~ëíáÅ=ãçÇìJ aÉåJ qÉåëáäÉ=ëíáÑÑJ _ÉåÇáåÖ=ëíáÑÑåÉëë=

äìë ëáíó= åÉëë=áåÇÉñ áåÇÉñ
b ê w E
E = w Sb Ew
ȡ S = =
w3 12˜ ȡ2
jkLãO âÖLãP j=kãLâÖ kãTLâÖP
píÉÉä ONMMMM TUMM OR MIMP
qáí~åáìã NOMMMM QRMM OR MINM
^äìãáåáìã TPMMM OUMM OR MIPM
j~ÖåÉëáìã QOMMM NTMM OR MITM
dä~ëë TPMMM OQMM OR MIQM
`çåÅêÉíÉ NRMMM ORMM S MIMU
`~êÄçå=ÑáÄêÉ=ÅçãéçëJ OMMMMM OMMM NMM OIMM
áíÉë
tççÇ=áå=Öê~áå=ÇáêÉÅíáçå NQMMM RMM OR UIPM
m~éÉêI=äáåÉêÄç~êÇ=áå=ja NRSMM TMM OO PITM

Paper differs from stamped felt, usually of cotton, through the manufacturing process. The
strength of a stamped felt, e.g. a felt hat, is produced by mechanical treatment of the formed
product through stamping and needling in which the fibres get tangled into each other and are
bonded by frictional forces. Particle boards are made from chips of wood, which can be ran-
domly distributed or ordered. They are held together with glue.
A piece of textile, a polyethene film and a paper can be separated even with the eyes closed.
All these products are essentially two-dimensional and formed by linear macromolecules.
These are also more or less organized into crystalline regions. If the textile material is made of
cotton, flax or ramie, it consists of the same material, cellulose, as bleached paper. Why do
these materials have different properties?
The answer lies in the construction of the materials, which leads to different types of interac-
tion between the components.
We shall here compare paper with three other types of material made from fibres: textiles, fi-
bre cloth and polyethene film.
22

NKNMKO==qÉñíáäÉë

A textile material is built-up of fibres which are first organized through carding and spinning
into a yarn which is thereafter used in a loom or knitting machine. The structures obtained allow
a certain freedom of movement to both the yarn and the fibres. The structure therefore reacts to
stresses in the first place through rearrangement of the fibres relative to each other, and rear-
rangement of the fibres in the yarn, and only in a second stage through elastic deformation of
the fibres themselves. The range of movement of the fibres is determined mainly by how
strongly the yarn is twisted. The mechanical properties can further be varied by letting the yarns
run through the textile in different ways. A woven textile can be torn relatively easily in the
thread direction, but not in between. Textile materials range from stretchable fibres of the Span-
dex type to very stiff carbon fibres.
A woven textile has very long threads of twisted fibres, which lie perpendicular to each oth-
er. Its strength comes from the fact that these threads run above and below each other so that
they are bent. The textile has essentially different mechanical properties in different directions.
It is slightly stretchable and strong in the thread directions, but considerably more stretchable
and weak in directions in between. Characteristic for a fabric is that it drapes itself, i.e. adapts to
the underlying structure. The product is held together by normal forces between the fibres, in
the twisted thread and between the crooked threads in the fabric.
A knitted textile is produced from a single thread which runs through more or less complicat-
ed loops. The product is almost isotropic, i.e. equally strong and stretchable in all directions.
Also in knitted goods, normal forces are the cause of the strength properties of the structure.
The textile feels flexible and soft. If it is wrinkled together and released, it regains its shape.
It can be draped over double-curved surfaces without noticeable stretching. If the textile is knit-
ted, it can be easily stretched in all directions, but after a long stretching it stiffens. If it is wrin-
kled together, no sharp creases are formed. The knitted product is very difficult to tear. If the
textile is wetted, it behaves in approximately the same way as when it is dry. Textiles can there-
fore be washed.

NKNMKP==cáÄêÉ=qÉñíáäÉë

Fibre (non-woven) textiles are manufactured from natural or synthetic fibres which randomly
form an essentially two-dimensional product. The fibres can be dispersed in air or water. The
normal forces between these randomly ordered fibres are negligible. The product can be given
strength by gluing the fibres together in the contact surfaces. This can be achieved either by
adding a suitable adhesive or, if the fibres are thermoplastic, by heating so that the fibres stick
together. The mechanical properties are determined by the fibre orientation and by the content
and type of adhesive. An extreme case occurs if the product contains so much binder that it can
be regarded as a fibre-reinforced composite.

NKNMKQ==mçäóÉíÜóäÉåÉ=cáäã

A polyethylene film is usually very smooth, it feels flabby and can easily be deformed by
stretching so that it conforms to a double-curved surface. If the film is released soon after the
 23

deformation, it contracts and returns essentially to its original shape. If the film is folded, no
permanent creases are formed. If an attempt is made to tear the film, it is difficult to start the
tearing but, once a tear has started, it is very easy to continue tearing. Water has no influence on
the mechanical properties.
There is nothing corresponding to the fibres in a plastic film. The structure lacks discrete
structural elements. Stresses lead to a distortion of the molecular structure or a sliding of the
molecules relative to each other, possibly leading to failure in the long tangled polymer mole-
cules. Ordered, crystalline areas act as stiffening reinforcements. Polymer technology explains
how the properties of the products can be modified, e.g. through the choice of polymer, molar
weight, degree of crystallization and orientation.

NKNN==pí~åÇ~êÇáò~íáçå

In order to facilitate trade, several standard methods exist for the determination of different pa-
per properties. These standards are issued by organizations, which have members representing
manufacturers, buyers, wholesale dealers and researchers. The most important standards in
Sweden are listed in Table 1.3.

q~ÄäÉ=NKPK=fãéçêí~åí=ëí~åÇ~êÇë=Ñçê=íÜÉ=pïÉÇáëÜ=áåÇìëíêóK

fpl fåíÉêå~íáçå~ä=lêÖ~åáò~íáçå=çÑ=pí~åÇ~êÇáò~íáçå
^pqj ^ãÉêáÅ~å=pçÅáÉíó=Ñçê=qÉëíáåÖ=~åÇ=j~íÉêá~äë
q^mmf qÉÅÜåáÅ~ä=^ëëçÅá~íáçå=çÑ=íÜÉ=mìäé=~åÇ=m~éÉê=fåÇìëíêó
`bk `çãáí¨=bìêçé¨Éå=ÇÉ=kçêã~äáë~íáçå
afk aÉìíëÅÜÉë=fåëíáíìí=Ѳê=kçêãìåÖ
pfp pí~åÇ~êÇáëÉêáåÖëâçããáëëáçåÉå=á=pîÉêáÖÉ
p`^kJqÉëí pÅ~åÇáå~îá~å=mìäé=~åÇ=m~éÉê=qÉëíáåÖ=jÉíÜçÇë

Buyers and sellers reach agreements about which different standards shall be applied.
Of interest are ISO Standard 9000 and the corresponding national standard, which constitute
a well-established system for quality control during manufacture. The method requires a couple
of years' personnel education and the introduction of new routines. Certification is followed by
supervision from the certifying organization. More and more buyers turn only to paper mills
which meet this standard.
The standardization organ CEN has the task of developing EN-standards, i.e. European stan-
dards. In the first place, it seeks to ensure that existing ISO-standards become EN-standards,
and thereby automatically national standards within the EU's member countries. The primary
interest within CEN is to remove technical trade obstacles in order to reinforce European har-
monization. This means that CEN is interested only in standard methods with a commercial in-
terest, i.e. methods which are of interest in the quality assessment of products (e.g. brightness
and spots of a pulp but not analyses of white liquor and mesa).
The main secretariat for CEN is in Brussels, whilst the secretariats for the different technical
committees (in all more than 150) are in different cities in Europe. A technical committee for
pulp and paper has been formed, TC 172, with its secretariat at DIN in Berlin.
24

CEN has begun to issue EN-standards relating to the pulp and paper industry. When an EN-
standard is issued, it automatically becomes a Swedish standard and is issued by SIS, since SIS
is Sweden's national standardization organ.
 25

O==m~éÉê=mÜóëáÅë=

Christer Fellers
STFI-Packforsk AB

2.1 Grammage 26

2.2 Thickness 26

2.3 Density 29

2.4 Stress and Specific Stress 30

2.5 Strain and Poisson´s Ratio 30

2.6 Tensile Strength Properties 32

2.7 Tensile Strength of Fibres Measured by Zero Span Technique 34

2.7.1 Zero-Span and Short-Span Tests 34

2.8 Page´s Theory for Tensile Strength of Paper 36

2.9 Strain at Break and Tensile Energy Absorption 37

2.10 Elastic Properties 37

2.11 Tensile Stiffness – Theoretical Approach 38

2.12 Ultrasonic Techniques for Stiffness Determination 40

2.13 Plastic Stress-Strain Properties in Tension 42

2.14 Creep Properties in Tension 43

2.15 Compression Properties 45

2.16 Viscoelastic Properties in Compression 49

2.17 Tearing Resistance 51

2.18 Theory for Tearing Resistance 52

2.19 Bursting Strength 53

26

2.20 Bending Stiffness 55

2.21 Bending Resistance 56

2.22 Fold Number and Folding Endurance 57

2.23 Thickness Direction Strength and Delamination Resistance 58

2.24 Methods that Simulate Converting Operations 60

2.25 Methods for Production Control 61

2.26 Fracture Mechanics 63

2.27 Air Permeance 66

OKN==dê~ãã~ÖÉ

Grammage, w is the mass per unit area. Basis weight is another term sometimes used.
The grammage includes the amount of water in the paper under the existing climatic condi-
tions, which shall be well specified.
"Dry" or "bone-dry" grammage, wd is sometimes indicated. This means the amount of dry
substance per square metre after a standardized drying, which takes place in an oven at 105 °C.
The relationship between the moisture ratio mr (mass of moisture per mass of paper), condi-
tioned grammage w and dry grammage wd is given by Equation 2.1.

w wd (1 mr ) (2.1)

OKO==qÜáÅâåÉëë

Thickness is a fundamental property of paper for several reasons. It has a great impact on the
bending stiffness of the paper and an even thickness profile important so that a paper reel has an
even cross profile, which in turn influences the runnability of the paper in the printing press and
in other converting situations. A paper thickness with narrow tolerances is required in book
manufacture where thickness errors from hundreds of pages would otherwise be cumulative.
According to SCAN-P 7, the thickness of the paper can be measured in two ways, as the
thickness of a single sheet or as the thickness in a stack, Figure 2.1. In both cases, the thickness
is measured as the distance between two parallel plates having an area of 200 mm2 and under a
pressure of 100 kPa.
In measurements of the thickness, especially for single sheets, the average thickness of the
paper will overrated since the surface unevenness of the paper are included in the measurement.
In the measurement of thickness in a stack, the idea is that errors of this kind should be reduced.
However, experience shows that surface unevenness effects cannot be completely avoided in
 27

this way. The measurement in a stack is, however, relevant for example in the estimation of
how thick a book will be.

single sheet stack

Figure 2.1. Thickness measurement according to SCAN-P 7.

SCAN P-88:01 describes an improved thickness determination for use in bending stiffness
calculations and studies of thickness changes in calendering and other applications.
According to this method, the thickness of a paper is scanned between two opposite mea-
surement points so that a thickness profile of the paper is obtained, Figure 2.2.

paper

Figure 2.2. Structural thickness measurement according to SCAN P-88:01. Paper and board. Structural thickness
and structural density.

Using this method, a thickness scan is performed. Figure 2.3 shows how the thickness varies
along a line across a paper sheet. The mean value is called the structural thickness. The
SCAN-P 7 thickness corresponds to a value close to the maximum value of the peaks.
The thickness of some papers can vary substantially and a measurement of thickness as the
distance between two flat plates sometimes becomes less relevant.
By adapting the force and diameter of the measurement points, a true mean value of the
thickness can be obtained. In measurements on papers manufactured in the same way but with
different grammage, a thickness value is thus obtained which is proportional to the grammage.
The method prescribed in SCAN-P 7 overestimates the thickness. In Figure 2.4 the thickness
as a function of grammage is performed on laboratory-made papers with equal surface charac-
28

teristics. If the thickness of the papers is extrapolated to zero grammage, a measure of the sur-
face unevenness of the paper is obtained.

Figure 2.3. Thickness variations in a paper measured with SCAN P:88:01. The mean value is called the structural
thickness.

0.4

apparent thickness
0.3 (SCAN standard)
thickness (mm)

0.2

0.1
intrinsic thickness

0
0 100 200 300
grammage (g/m2)

Figure 2.4. Apparent thickness of unbleached kraft paper determined according to the SCAN standard. The
intrinsic thickness is given by a parallel shift of the SCAN thickness values to yield zero intercept when extrapo-
lated to zero grammage.

Other methods have been proposed to measure structural thickness. For example, a piece of
paper has been submerged in mercury and the volume has been measured and a thickness gauge
with rubber-coated surfaces, which adapt to the unevenness of the paper surface has been used.
Table 2.1 shows a comparison of thickness values according to different methods.
The above examples show that a true thickness measurement is a prerequisite for a meaning-
ful density calculation to be used as a description of the structure of paper.
 29

q~ÄäÉ=OKN==mÉêÅÉåí~ÖÉ=ÇÉîá~íáçå=áå=íÜáÅâåÉëë=Ñçê=íÜÉ=p`^kJm=T=ãÉíÜçÇë=Åçãé~êÉÇ=ïáíÜ=p`^k=mJUUWMK=
qÜÉ=Öê~ãã~ÖÉ=çÑ=íÜÉ=ëÜÉÉíë=áë=SMÖLãOK=

mìäé cçìê=ëÜÉÉíë=áå=~=ëí~Åâ páåÖäÉ=ëÜÉÉíë

SCAN P 7 SCAN P 7
SCAN P88 SCAN P88
råÄäÉ~ÅÜÉÇ=éáåÉ=âê~ÑíK=UMM=mcfJãáää=êÉîK HNPB HONB
råÄäÉ~ÅÜÉÇ=ÄáêÅÜ=âê~ÑíK=NMMM=mcfJãáää=êÉîK HNUB HQOB

OKP==aÉåëáíó

Many constructional materials such as metals and plastics have a homogeneous structure, which
means that the density of the material is constant. There are many materials however where the
proportion of air can be varied, for example foamed plastics and wood where the density varies
between the species of wood. For paper, the density of the material can be varied between ca
300 and 1000 kg/m3. Pure cellulose has a density of ca 1500 kg/m3. The differences in paper
properties in this interval are considerable. The possibility of paper technology to control the
density means that there is a unique potential to steer the properties of the product to the desired
quality. Density is consequently a very important structural parameter of the sheet.
The density of paper is influenced by the type of fibre and all paper technological operations
such as fibre separation, beating, pressing, drying and calendering. The density U is calculated
as the grammage w divided by the thickness t (Equation 2.2).

w
U (kg/m3) (2.2)
t

The accuracy of the density determination is consequently dependent on the thickness mea-
surement which is difficult to measure unambiguously for a rough and compressible material
such as paper, see Section 2.2.
SCAN-P 7 specifies that density of paper shall be evaluated from the thickness measured in a
stack, but that for carton board, it shall be evaluated from the thickness of a single sheet. As
mentioned earlier, this measurement can give a large error depending on the surface roughness
of the paper. Structural density is usually for a meaningful density determination.
Within paper technology, the inverted density is often used, which is called the bulk (Equa-
tion 2.3).

1
Bulk (2.3)
U

The bulk is usually given in cm3/g, which gives a value with a magnitude of 1–2.
30

OKQ==píêÉëë=~åÇ=péÉÅáÑáÅ=píêÉëë

In solid mechanics, the concept stress, i.e. force per unit area is a fundamental, accepted way to
describe the intensity of loading. For paper, which is a porous material and a material where the
pore volume and density can be changed, the concept off stress for paper material ranking and
development can lead us in the wrong direction as exemplified in the following example.
The example illustrates a case where the stress is unity in the left figure, Figure 2.5. If the pa-
per is calendered and the thickness reduced to half its value the stress becomes twice as high. It
is quite obvious that the loading intensity of the fibres has not changed. If the stress is the failure
stress, the material to the right is judged as twice as good as the left material, which of course is
wrong.
A more realistic way is to use a specific stress or indexed stress, which simply means a divi-
sion of the force by the width and grammage, Figure 2.6. In this way the stress is equal regard-
less of the pore volume.
Different symbols will be used to distinguish between the different stress concepts.

grammage, w grammage, w

width, b
thickness, t width, b

force, F force, F

F 1 F 1
s  1 s  2
bt 1·1 bt 1 · 0.5

Figure 2.5. The change in stress with calendering.

grammage, w grammage, w

width, b
thickness, t width, b

force, F force, F

F 1 F 1
sw   1 sw   1
bw 1·1 bw 1·1
Figure 2.6. The change in specific, indexed stress, with calendering.

OKR==píê~áå=~åÇ=mçáëëçå…ë=o~íáç

For paper the useful definition of strain, H, is the elongation divided by the original length. A
piece of material is strained in the X-direction and gets the strain Hx. The strain in the transverse
 31

direction becomes Hy (Figure 2.7). For most material the material contracts (exceptions such as
cork exist).

ey

x ex
y

Figure 2.7. Strains in paper.

Poisson´s ratio vxy is uses to tell the relation between these strains, Equation 2.4

Hx = – vxy · Hx (2.4)

Tensile testing of paper is carried out by clamping a paper strip, usually with a width
b = 15 mm and a length l = 100 mm at each end, and straining the material until it breaks, as in-
dicated in Figure 2.8.
A large number of devices are available for tensile testing with both vertical and horizontal
clamping of the paper.

force F

length l

width b length change Dl

grammage w
force F

Figure 2.8. Tensile test of paper strip with definitions.

If the stress ıw is plotted against strain İ (with the definitions given in Equation 2.5) the
curve has the appearance shown in Figure 2.9. The curve is linear at first and then deviates to
reach a maximum which defines the failure point.

F 'l
Vw H (2.5)
b˜w l

A number of useful properties can be derived from the curve.

32

'V w
• Tensile stiffness index, E w is the slope of the linear part of the curve.
'H
• Tensile index, V Tw is the maximum stress.
• Strain at break, HT is the maximum strain.

• Tensile energy absorption index WTw is the area under the curve.

These properties will be further discussed below.

tensile index
stress

tensile absorbtion
index

D stress

D strain
strain at break

strain

Figure 2.9. A typical stress-strain curve for paper.

OKS==qÉåëáäÉ=píêÉåÖíÜ=mêçéÉêíáÉë

The strength of paper due to tensile loading may be expressed in various ways.
• Tensile index is the maximum tensile force per unit width and unit grammage.

FT
V Tw (Nm/kg) (2.6)
bw

• Tensile strength is the maximum force related to the width of the test piece.

FT (N/m)
V Tb (2.7)
b
 33

• The tensile failure stress is the maximum tensile force per unit width and unit thickness.

FT (Pa or N/m2) (2.8)

VT
bt

In older literature, the concept of breaking length is used. This designates the maximum
length which a hanging sheet of paper can have without breaking. The expression can be de-
rived as follows:

FT
V Tl (m) (2.9)
bt U g

The relation between breaking length and tensile index is derived

w
U and we get
t

V Tw
V Tl (Nm/kg) (2.10)
g

Note that

VT
V Tw (2.11)
U

VT Maximum stress, Pa (N/m2)

V Tb Tensile strength N/m

V Tw Tensile index Nm/kg

V Tl Breaking length m
FT Maximum force N
t Thickness m
b Width m
w Grammage kg/m2
U Density kg/m3
g Acceleration due to gravity m/s2

For paper, the tensile index is of the order of 10–100 kNm/kg, Figure 2.10. If, for example, a
paper roll is rolled out of an air balloon, it can thus hang down with a length of 1–10 km before
the paper breaks because of its own weight.
34

100
TMP = thermomechanical pulp
SBK
RMP = refiner pulp
80 CTMP = chemimechanical pulp
tensile index (kNm/kg)

UBS = unbleached kraft pulp

CTMP GWD = groundwood pulp
60
UBS SBK = semi-bleached kraft pulp
TMP
MP
40

20
GWD

0
0 200 400 600 800 1000
density (kg/m3)

Figure 2.10. Tensile index for different fibre raw materials plotted against density. (Restrained drying). Different
density levels are achieved by varying the beating. (Isotropic laboratory papers).

Figure 2.11 shows typical tensile failures in a paper. The failure zone is characterised by
elongated and broken fibres. Depending on whether the degree of bonding is low or high in the
paper, the character of the failure surface will change from a structure with few broken fibres
into a structure where nearly all the fibres are broken.

a) low degree of bonding b) high degree of bonding

Figure 2.11. The tensile failure in paper is characterized by elongated and broken fibres.

OKT==qÉåëáäÉ=píêÉåÖíÜ=çÑ=cáÄêÉë=jÉ~ëìêÉÇ=Äó=wÉêç=pé~å=qÉÅÜåáèìÉ

OKTKN==wÉêçJpé~å=~åÇ=pÜçêíJpé~å=qÉëíë

Sheet tensile properties are developed by the use of adequately strong fibres and the enhance-
ment of the fibre-fibre bonding properties. Good (or poor) fibre-fibre bonding is achieved pri-
marily by the pulping, bleaching, refining and drying systems. If one could eliminate the
influence of bonding in sheet tests, then presumably the fibre mechanical properties could be
 35

deduced without resorting to tests on individual fibres. In theory, shortening the span between
the two tensile grips to the point where the fibre bonds are insignificant – the ideal “zero” span
– means that all forces between grips are transmitted by those fibres, lying at various orienta-
tion, that are held within both grips simultaneously. For an in-plane statistically isotropic sheet,
the mean fibre strength, expressed as breaking force, is related to the zero-span tensile strength.
The zero-span test has become widely used, and several standards organizations have pub-
lished test procedures for it; for example, TAPPI issued T 231 cm-96 for the dry zero-span test
and T 273 pm-95 for the wet zero-span test. Equipment is readily available as a fixture for the
universal testing machine or as a stand-alone system. The zero-span tensile test was introduced
to distinguish the difference between the strength of the network and the strength of the fibres
themselves. For softwood pulps, zero-span sheet strengths are typically 1.5–3 times as great as
the strength values achieved in a conventional tensile test. For hardwood pulps, the ratios tend
to be greater.
Figure 2.12 shows a typical curve for the tensile index versus the free span between the
clamps in a tensile test. Special clamps are used to grip the paper without slipping. Still, a true
“zero span” is difficult to obtain and the values are sometimes extrapolated to a “true” zero span
value.
Figure 2.13 shows a schematic drawing of the clamps and the fibre network during a tensile
loading.

120

beaten, dry
tensile index (kNm/kg)

80

unbeaten, dry
40
unbeaten, wet

0
0 0.4 0.8 1.2 1.6 2.0
span between the clamps (mm)

Figure 2.12. A typical curve for the tensile index versus the free span between the clamps.
36

clamps

Figure 2.13. A schematic drawing of the clamps and the fibre network during a tensile loading.

OKU==m~ÖÉ…ë=qÜÉçêó=Ñçê=qÉåëáäÉ=píêÉåÖíÜ=çÑ=m~éÉê

Several researchers have attempted to calculate the tensile strength of paper from the fibre
strength and bonding ability. The most accepted theory was formulated by Page:

1 9 1 12
˜  (2.12)
V Tw 8 V zsw W s lD

where
V Tw = Tensile index
V ZS
w
= Zero span tensile index
WS = Shear stress at break of a fibre-to-fibre bond
D = Bonded area between the fibres per kg of fibres
l = Average fibre length

Due to the extremely complicated interaction between fibres of different dimensions and the
influence of different methods for sheet forming, pressing and drying the theory only describes
the strength in idealised cases. Equation (2.12) is reported to give good agreement between
properties of straight fibres and tensile index of papers dried under restraint.
The value of the equation lies in that it shows the relationship between some important bond-
ing and fibre properties. However, there are limitations for the use of the equation. Many impor-
tant factors are not taken into account, examples are listed below.
• Formation
• Fibre shape and fibre length distribution
• The true area and failure stress of the bonds
 37

OKV==píê~áå=~í=_êÉ~â=~åÇ=qÉåëáäÉ=båÉêÖó=^Äëçêéíáçå

Strain at break is of the order of 1–5 % for sheets dried under restraint. For freely dried sheets,
strain at break of up to 20 per cent can be obtained.
Tensile Energy Absorption (TEA) index

WTb
WTw (J/kg) (2.13)
w

where WTb is the TEA, the area under a stress-strain (force /width v.s strain) curve (J/m2).
The tensile energy absorption index is of great importance particularly for sack paper and de-
pends to a great degree on the shrinkage of the paper during drying and it can vary within wide
limits, in the range of 1–5 kJ/kg.

OKNM==bä~ëíáÅ=mêçéÉêíáÉë

The in-plane stiffness of paper due to tensile loading may be expressed in a few various ways.
For a homogeneous material, the linear part of the stress-strain curve is described by Hooke's
law:

V = E · H (N/m2) (2.14)

The elastic modulus E is thus a measure of the in-plane stiffness properties of the material.
Because paper has a porous structure, it is in many cases misleading to use the concept of
modulus of elasticity for paper as discussed in relation to Figure 2.5 and Figure 2.6.
If both sides of the equation representing Hooke's law are multiplied by the thickness t, we
obtained the following equation in which the tensile stiffness is included.

Vb = Eb · H (N/m) (2.15)

If both sides of the equation are divided by the grammage, Hooke's law has the following
form.

Vw = Ew · H (Nm/kg) (2.16)

In some literature, tensile stiffness is often found given as the elasticity modulus expressed
in metres. This is analogous to the expression for breaking length. Hooke´s law in these units is
given by the expression:
Vl = El · H (m) (2.17)

Ew
El (m) (2.18)
g
Note that
38

E
Ew (2.19)
U
V=E·H Hooke´s law for stress Pa (N/m2)
Vb = Eb · H Hooke´s law for force per unit width N/m
V =E ·H w w
Hooke´s law for force per unit width and grammage Nm/kg
Vl = El · H Hooke´s law for metres m
E Elastic modulus Pa (N/m2)
Eb Tensile stiffness N/m
w
E Tensile stiffness index Nm/kg
g Acceleration due to gravity m/s2
U Density kg/m3

The tensile stiffness index for paper is in the range of 1–10 MNm/kg, Figure 2.14.
TMP = Thermomechanical pulp
CTMP = Chemimechanical pulp
UBS = Unbleached kraft pulp
USS = Unbleached softwood sulphate
OCC = Recycled corrugated board

10 TMP = thermomechanical pulp

CTMP = chemimechanical pulp
8 UBS
NSSC UBS = unbleached kraft pulp
USS = unbleached softwood sulphate
tensile stiffness index

6 OCC = recycled corrugated board

OCC
CTMP
4
(MNm/kg)

TMP USS
2

0
0 200 400 600 800 1000
density (kg/m3)

Figure 2.14. Tensile stiffness index of a paper dried under restraint plotted against density for different fibre raw
materials. Different density levels have been attained by varying the beating.

OKNN==qÉåëáäÉ=píáÑÑåÉëë=J=qÜÉçêÉíáÅ~ä=^ééêç~ÅÜ

Tensile stiffness index may be related to the degree of bonding and fibre properties by Seth and
Page. Figure 2.15 shows the tensile stiffness index plotted versus the light scattering coefficient
 39

(indicating the state of bonding) for a pulp at different degrees of beating. For each degree of
beating, the wet pressing pressure has been increased.

tensile stiffness index

1/3 straight fibres

fibre
tensile
stiffness
index
fibres with dislocations

increasing degree of bonding

scattering coefficient

Figure 2.15. Tensile stiffness index versus light scattering coefficient for a bleached kraft pulp of Southern Pine.
The beating times are from 0 min for the bottom curve and 50 min for the upper.

The results show that the tensile stiffness index, for a given degree of beating, increases with
increasing press pressure up to a plateau value. The plateau value increases with increasing de-
gree of beating. The effects can be explained as follows: Figure 2.16 shows schematic pictures
of the state of bonding of a fibre at two degrees of bonding. At low press pressures, a given fibre
is bonded to only a few fibres. At fibre ends, the formed network is not able to transfer forces.
At high press pressures, more bonding points are obtained and a better force transfer is ob-
tained.

stress low degree of bonding stress high degree of bonding

I I I I
L L

Figure 2.16. Schematic picture of the state of bonding of a fibre at two degrees of bonding.

The influence of the beating can be explained with the help of Figure 2.16. Lightly beaten
pulps always contain fibres with a certain degree of dislocations, crookedness and micro-com-
pressions. During beating, the fibres swell and these faults have a tendency to be straightened
out, and this leads to a higher modulus. It can be shown theoretically that the elasticity modulus
in an isotropic sheet can at most be 1/3 of the modulus of the fibre.
40

For straight well-beaten fibres, the theoretically possible value of 1/3 of the tensile stiffness
index of the fibres is attained in an isotropic sheet. When the fibres have dislocations of differ-
ent kinds, the elasticity modulus is reduced.

OKNO==räíê~ëçåáÅ=qÉÅÜåáèìÉë=Ñçê=píáÑÑåÉëë=aÉíÉêãáå~íáçå

Ultrasound may be used to evaluate the in-plane stiffness of paper. The method is non-destruc-
tive and fast. The principle is given below.

Figure 2. 17. The in-plane stiffness may be evaluated by ultrasonic technique. Transmitters and receivers are
placed 100 mm apart in a circle. Ultrasonic pulses are sent from the transmitter and the time needed to reach the
receiver is recorded.
w
The ultrasonic stiffness index measured by ultrasonic technique EUS may be calculated from
the sound velocity v as follows.
The value will not be identical to the tensile stiffness index evaluated from tensile tests due
to the viscoelastic, time dependence of paper, and the boundary condition,
Depending on the fibre properties and orientation and moisture content, the ratio between the
two ways of evaluating stiffness will vary. In all cases the ultrasonic value will be higher.
A typical example is given in Figure 2.18 for a kraft paper. Properties in MD and CD is giv-
en for the tensile stiffness index and for the ultrasonic stiffness index versus moisture ratio.
A very useful application for ultrasonic stiffness is the determination of non-symmetry in pa-
per. Figure 2.19 shows a paper with slight off-axis symmetry. The method may be used for
cross profile studies and product control of different mechanical properties.
 41

20 kraft

15
stiffness index (MNm/kg)

10

MD USI

5 MD TSI
CD USI
CD TSI
0
0 5 10 15 20 25
moisture ratio (%)

Figure 2.18. Ultrasonic stiffness index for a kraft paper. Properties in MD and CD is giver for the tensile stiffness
index and for the ultrasonic stiffness index versus moisture ratio.

CD

MD

Figure 2.19. A paper and with a slight off-axis symmetry.

42

OKNP==mä~ëíáÅ=píêÉëëJpíê~áå=mêçéÉêíáÉë=áå=qÉåëáçå

Figure 2.20 shows stress strain curves for a kraft paper under tension at different strain rates.
Both tensile stiffness and tensile strength are influenced by the strain rate.

infinite rate
0.0022 (m/m/min)
0.0011
0.00046
stress

0
strain

Figure 2.20. Stress-strain diagram for a kraft paper at different strain rates.

If a paper is loaded and unloaded under a tensile load at a constant strain rate, the stress-
strain diagram gets a typical appearance as illustrated in Figure 2.21.
stress

sy

0 1 2 3 4 5
strain (%)

Figure 2.21. The appearance of the stress-strain diagram when a paper is loaded and unloaded.

The curve begins with a straight part. When the direction of strain is reversed at a given
force, the force decreases.
On unloading, the strain does not return to zero. The paper exhibits a permanent strain. When
the paper is again strained, the force increases up to the force to which the paper was subjected
earlier. Thereafter, the stress-strain curve proceeds along the original path.
 43

5 5

4 4

permanent strain (%)

3 3
stress

2 2

1 1
P a m
0 0
0 1 2 3 4 0 1 2 3 4 5
strain (%) maximum strain, m (%)

Figure 2.22. Left-hand picture: Force-elongation diagram up to a certain strain (m) and unloading. P indicates the per-
manent strain after a certain time at zero force. Right-hand picture: Permanent strain as a function of maximum strain.

Figure 2.22a shows a cyclic deformation of a paper at a constant strain rate. The paper is
loaded up to a certain maximum elongation, m, and is unloaded. At a force of zero at point a, the
paper is left unloaded. The elongation will therefore wish to return slowly to zero but, after a
long time, it exhibits a permanent elongation, p. As shown in Figure 2.22b, the permanent
change increases with increasing maximum elongation.

OKNQ==`êÉÉé=mêçéÉêíáÉë=áå=qÉåëáçå

If a material is subjected to a constant load, the elongation increases with time, Fig. 2.23. The
phenomenon is a consequence of the viscoelasticity of the material. The same behaviour applies
also for paper as shown in Figure 2.24, which show the strain as a function of time for a paper is
subjected to different stresses.

tertiary

secondary A
primary
elongation

time

Figure 2.23. Creep of a material.

44

stress (MPa)

3 60.7

55.9
strain (%)

2
45.3

1 35.3

20.2
0
10 102 103 104 105 106
time (s)

Figure 2.24. Creep of paper. The strain versus time at different stress.

In Figure 2.25 the data are expressed as stress strain curves at different times. We obtain iso-
chronous curves. The concept of isochronous curves will be further discussed later in relation to
Section 2.16, viscoelastic properties in compression.

75

10 s 102 s 103 s

104 s
50
105 s
106 s
stress (MPa)

25

0
0 1.0 2.0 3.0
strain (%)

Figure 2.25. Stress strain curves at different times. We obtain isochronous curves.
 45

e0
strain

L
0 time, t

s
s0
strain
stress

0 time, t

Figure 2.26. Principle of stress relaxation

If paper is strained to a given strain and kept at that strain, the stress decreases with time. The
phenomenon is called stress relaxation. The principal behaviour is shown in Figure 2.26.

OKNR==`çãéêÉëëáçå=mêçéÉêíáÉë

When paper is compressed in the plane of the paper, the paper usually buckles, as shown in Fig-
ure 2.27.

Figure 2.27. Buckling of a paper strip under compression with a long clamping length.

The failure force per unit width and unit grammage in compression increases when the
clamping length is decreased as illustrated for paper with different grammage in Figure 2.28. At
46

clamping lengths of less than circa 0.7 mm, a plateau value is attained independent of the gram-
mage of the paper.

40

30
compression (kNm/kg)

300 g/m2
stress at break in

20

100 g/m2
10

0
0 1 2 3 4 5 6 7 8
span (mm)

Figure 2.28. Stress at break in compression versus clamping length for paper with different grammage.

If buckling is prevented, for example by testing with a short clamping length, the pure com-
pression properties and strength of the paper can be evaluated. According to SCAN-P 46, the
compression strength can be evaluated by fixing the paper between two clamps with a clamping
length of 0.7 mm, Figure 2.29a.
The paper can also be prevented from buckling by using columns which rest against the pa-
per, Figure 2.29b. The columns must be used if the strain process is to be recorded, for example
in studies of creep of paper in compression.

a) b)

Figure 2.29. Two ways of evaluating the compression properties of paper. a) The SCAN-P 46 method with a
short clamping length, short span compression test. b) Long clamping length with columns support.
 47

Stress-strain curves for paper in compression are shown in Figure 2.30, where a comparison
is also made with tensile properties. In a comparison of tensile and compression curves in MD
and CD, the following can be observed:
• Paper is normally stronger in MD than in CD, both in tension and in compression.
• The strain at break is higher for MD than CD
• The slope of the curve is the same for small compressions in both tension and compression.
This means that tensile stiffness = compressive stiffness.
• The strain at break in compression is lower than that in tension.

140
120
MD
100 tension
stress (kNm/kg)

80
60
compression
40
MD CD
20
CD
0
0 1 2 3 4
strain (%)

Figure 2.30. Stress-strain curves for paper. Comparison between tensile and compression properties in MD and
CD. Positive stress and strain values are shown.

The compression strength of paper is measured primarily at high grammage, above ca.
100 g/m2, and it is important primarily for the ability of corrugated board boxes and carton
board packages to be stacked in storage.
Compression strength is the maximum stress which paper can support in compression.
Compression strength V Cb is the maximum compression force per unit width.

FC
V Cb (N/m) (2.21)
b
40
unbleached
softwood kraft
30
compression index

20 OCC
unbleached
hardwood kraft
(kNm/kg)

10

0
400 600 800 1000
density (kg/m3)

Figure 2.31. Compression index of paper dried under restraint plotted against density for a number of pulps. Different
densities have been attained by varying the beating. OCC = recycled Old Corrugated Containers.
48

Figure 2.32. Electron micrographs of fracture zones of paper, subjected to compression loading.
 49

The compression failure stress VT is the maximum compression force per unit width and unit
thickness.

FT
VT (N/m2) (2.22)
bt
Compression index V Cw is the maximum compression force per unit width and unit gram-
mage.

FC
V Cw (Nm/kg) (2.23)
bw
Compression index varies between 20 and 40 kNm/kg for liner pulps, see Figure 2.31.
Figure 2.32 shows three electron micrographs of paper, one seen from the edge of the paper
and two seen from above after compressive failure. The reason for the compressive failure is
that the fibre wall collapses. For this reason, the compression strength is relatively insensitive to
the length of the fibres.
There are many other methods of testing for the compression strength of paper. The most
common is the ring crush test, where a paper strip is first formed into a ring and is then crushed
edgeways in a special test apparatus, see Figure 2.33, where a flat plate is pressed against the
ring-shaped sample. Technically, this type of measurement involves great difficulties such as
buckling and difficulty to obtain sufficient parallelism of the loading plates.

Figure 2.33. Ring Crush Test (RCT).

OKNS==sáëÅçÉä~ëíáÅ=mêçéÉêíáÉë=áå=`çãéêÉëëáçå

The main use of carton board, liner and fluting is for packages of different kinds. An important
property of these materials is the ability to protect the contents in the packages against external
influences, especially in the long-term storage of the packages in stacks in a moist environment
or in an environment where moisture or temperature vary with time. The material properties,
which are important, are the creep properties of the materials in compressive loading.
50

In order to be able accurately to study the properties of paper in compressive loading, STFI
has constructed the creep apparatus where stress, strain and time are recorded with a computer.
The following figures show in principle how the creep properties of paper can be recorded
and evaluated. In the creep test, a constant stress is applied to the test piece, and the resulting
strain is recorded as a function of time, Figure 2.34. It is possible to adapt the data in Figure
2.34 to Equation 2.24 below, which relates stress, strain and time to each other.

s3
strain ( e)

s2

s1

t0 = 1 second t1 t2
log time

Figure 2.34. Creep curves at different stresses.

If the curves in Figure 2.34 are intersected, corresponding stress and strain values are ob-
tained. Figure 2.35 shows such isochronous (from Greek iso = equal and chronos = time) curves
at different times. Isochrones are often used for polymeric materials in strength dimensioning
for structures, which must stand for a long time.
At small strains, the curves are linear and then deviate at high strains. The linear part of the
curve corresponds to a linear viscoelastic behaviour of the paper while the non-linear part corre-
sponds to a non-linear viscoelastic and plastic behaviour.

t0

t1
stress ( s w)

t2
s3
s2
s1
E cr,w (t2)

strain ( e)

Figure 2.35. Isochronous stress-strain curves constructed from the creep curves in Figure 2.34.
 51

To describe the isochronous stress - strain curves from creep curves, we use the following
equation:
p
ª § D2 · § D2 2 ·º § t ·
Vw «D1 ˜ tanh ¨ ˜ H ¸  D 3 ˜ H ˜ tanh ¨ ˜ 100 ˜ H  ¸ » ˜ ¨ ¸ (2.24)
¬ © D1 ¹ © D3 ¹ ¼ © t0 ¹
where Vw is the stress, D1, D2, D3, and p are material parameters, H is the strain at time t, and t0
is the reference time, equal to one second.
The equation consists of two hyperbolic tangent functions to describe the shape and magni-
tude of a reference stress-strain curve and a power function to show the effect of time on the
magnitude of the curve.
In tension, the stress-strain curve may become approximately linear at higher strains, which
makes the second term in this equation necessary.

OKNT==qÉ~êáåÖ=oÉëáëí~åÅÉ

The tearing resistance, which is a way of evaluating the crack sensitivity of the paper, is a
strength dimension of central importance for paper.
A small cut is made in a pack of papers. In a pendulum apparatus, tearing is completed. The
work is divided by the tearing distance and the average force is obtained, defined as tearing re-
sistance.
Besides tensile index, tear index is the most important property in the characterization of
pulp for paper. The tear apparatus that has been standardised internationally measures the tear-
ing work with a pendulum tester of the Elmendorf type. Figure 2.36 shows the principle. The
tearing work is often so small that several layers of paper must be torn simultaneously. Most
standards prescribe the tearing of four sheets.
Just for notification, there are other tearing methods in use. One that occurs in Europe is the
Brecht-Imset method.

Figure 2.36. Tearing resistance according to Elmendorf.

52

Tearing resistance Tb is the mean force (N) during the tearing divided by the number of pa-
pers and the tear index Tw is calculated by dividing the tearing resistance by the grammage.
T b (Nm2/kg) (2.25)
Tw
w
Tear index lies in the range 10–30 (Nm2/kg).

OKNU==qÜÉçêó=Ñçê=qÉ~êáåÖ=oÉëáëí~åÅÉ

With increasing beating of chemical pulps, the tearing resistance usually passes through a max-
imum, Figure 2.37. The tear index of paper made from short-fibre hardwood pulp is consider-
ably lower than that of softwood pulp, but it often increases even after a severe beating.
Technical beating of hardwood pulps is seldom driven so far that the falling part of the curve is
entered.
In spite of the central importance of the tearing resistance, the theory is weak. For unbeaten
pulps with a low degree of bonding between the fibres, it is imagined that the tearing takes place
by drawing the fibres out of the structure without breaking them, roughly as in Figure 2.11A.
The tearing energy is then assumed to be determined by the length of the fibre and the number
of bonding points. With increased beating, the fibres are bonded together in more and more
places and the tearing energy increases. Finally, a situation is reached where the fibre breaks in-
stead of being drawn out of the network, as in Figure 2.11B. The tearing energy required to cre-
ate a fibre failure is then lower than if the fibre is pulled out of the network. The optimum
tearing energy is obtained at a suitable distribution between pulled-out and broken fibres.
The tearing of machine-made papers and multi-layer structures follows complicated paths
and delaminating phenomena are common, Figure 2.38. Different results are sometimes also
obtained depending on the number of paper sheets tested.

20
unbleached
kraft
15
softwood
tear index (Nm2/kg)

sulphite
10
mechanical
pulp
5
hardwood
hardwood sulphate
sulphite
0
0 30 60 90 120
tensile index (kNm/kg)

Figure 2.37. Tear index as a function of tensile index for some different pulps.
 53

Figure 2.38. Tearing of thicker papers and highly oriented papers in the cross direction may be difficult to evalu-
ate by the traditional tear method.

OKNV==_ìêëíáåÖ=píêÉåÖíÜ

The bursting strength test was invented by Mullen about 100 years ago and is therefore often
called the Mullen method. The Mullen test is thus one of the oldest methods for testing the
strength of paper. In the testing, the paper is clamped in a ring with a centrally recessed rubber
membrane, see Figure 2.39.
The membrane is pressed upwards by an increasing oil pressure and finally bursts the paper.
The oil pressure at the moment of failure is read on a manometer and this defines the bursting
strength. Figure 2.40 shows the fracture appearance in the Mullen test.

Figure 2.39. The Mullen test.

54

Figure 2.40. The fracture appearance in the Mullen test.

Bursting strength B is the pressure at break (Pa = N/m2). And the burst index is obtained by
dividing the bursting strength by the grammage w.
Burst index is

B
Bw (N/kg) (2.26)
w

The burst index is in the range of 1–10 MN/kg. In spite of its age, and doubtful relevance,
bursting strength is still used to characterize the properties of many paper grades. The bursting
strength has been derived from the tensile properties as follows:

§ §H · ·
0,5

k ¨ V Tb , MD  V Tb , CD ¨ MD ¸ ¸  H MD 
0,5
Bursting strength (2.27)
¨ © H ¹ ¸
© CD ¹

where
V Tb i the tensile strength in MD and CD respectively
H is the strain at break in MD and CD respectively

The bursting strength of the paper can thus be increased either by increasing the orientation
of the fibres in the machine direction, which leads to an increase in the tensile strength in the
machine direction, or by taking measures which increase the strain at break in the machine di-
rection, for example by reducing the draws in the machine direction. Another measure is to re-
duce the strain at break in CD by preventing the shrinkage in the cross direction of the paper
machine.
 55

OKOM==_ÉåÇáåÖ=píáÑÑåÉëë

The resistance of paper to bending, i.e. its bending stiffness, is one of the most important paper
properties. High bending stiffness reduces the tendency for buckling of boxes when the contents
press against the walls, prevents newspapers which we read from falling together under their
own weight and gives runnability for example sack paper in sack manufacture.
Bending stiffness, per unit width b, of a strip of paper can be obtained if we know the bend-
ing moment and the curvature (1/R) where R is the radius of curvature.

M
Sb ( Nm)
§1· (2.28)
¨ ¸ b
©R¹

It is often preferable to measure the bending indirectly by using mathematical relationships

between downwards bending and curvature in the two-point method, which is most common for
paper and carton board. The two-point method is shown in Figure 2.41. The bending stiffness is
calculated by measuring the force required to bend a strip with a certain width and length
through a certain angle.

length l
clamp

q
angle

Figure 2.41. Bending stiffness determination by beam bending.

Bending stiffness can then be evaluated by bending a strip of paper, as in Figure 2.41 in the
elastic region. Bending stiffness per unit width b is as follows

F 60 l 2
Sb (N/m) (2.29)
bT S
Bending stiffness may be calculated from elastic modulus and thickness as follows

Et 3
Sb (2.30)
12
As discussed previously, tensile stiffness index Ew is a better way of expressing in-plane
stiffness of paper than the elastic modulus E.

E w
Since E w and U we get
U t
56

1 3 w 1
Sb w E 2 (2.31)
12 U

To be able to make quantitative calculations according to this equation, a true measure of the
density of the paper is required, as mentioned earlier. We see that bending stiffness is propor-
tional to grammage raised to the power 3 and we can define a property bending stiffness index.

Sb
Sw (Nm7/kg3) (2.32)
w3

Furthermore a low density, raised to the power 2, and a high tensile stiffness index are re-
quired.
A resonance method for the determination of bending stiffness based on the determination of
the length at which a paper strip will come into resonance at a frequency of 25 Hz. The method
is shown schematically in Figure 2.42.

Figure 2.42. The bending stiffness for paper can be evaluated by determining the resonance length of the paper.

The bending stiffness is determined from Equation 2.33.

S b 2 ˜ 103 l 4 w (Nm) (2.33)

where
l = resonance length
w = grammage
The bending stiffness index for paper lies in the range of 0.5–2 Nm/kg3.
Paper and carton board are often manufactured in several layers. Through the use of laminate
theory, it is possible to calculate and optimise the bending stiffness of multi-layer sheets theo-
retically. Usually pulps with a high tensile stiffness index are used in the outer layers and pulps
with low density in the middle layers. Laminate theory will be discussed in another chapter.

OKON==_ÉåÇáåÖ=oÉëáëí~åÅÉ

In certain standards the bending resistance in terms of the maximum force, under standardised
conditions, is measured with the two-point method according to Figure 2.43. An angle of 15°, a
strip length of 10 or 50 mm and a strip width of 38 mm is often used in the standards. A bending
curve for paper is shown schematically in Figure 2.43. One problem associated with bending to
15q is that the paper might be plasticised which has the consequence that the bending stiffness is
underestimated.
 57

true bending FB
stiffness
force

measured
bending
stiffness

15°
angle

Figure 2.43. Schematic bending curve for paper. The force is plotted against the angle as indicated in Figure
2.41.

The relationship between bending stiffness and bending resistance can be calculated approx-
imately according to the following equations.
For a strip length of 50 mm:

S b FB ˜ 0,0837 (mNm)

For a strip length of 10 mm:

S b FB ˜ 0,00335 (mNm) (2.34)

Bending resistance can also be normalized approximately with respect to grammage and is
then called bending resistance index.

FB
FBw ( Nm6 / kg 3 ) (2.35)
w3

OKOO==cçäÇ=kìãÄÉê=~åÇ=cçäÇáåÖ=båÇìê~åÅÉ

According to the fold number method (SCAN-P:17), a paper strip is folded forwards and back-
wards at the same time as it is exposed to a constant tensile load. The folding endurance is the
common logarithm of the number of double folds which a paper can endure before the strip
breaks. There are some different measurement principles for fold number, for example Köhler-
Molin, Schopper and MIT which all give slightly different results. The principle for the Köhler-
Molin apparatus is shown in Figure 2.44.
Since the fold number can vary within very wide limits, from a few folds to above 10 000, it
is often practical to work with the concept of folding endurance, the common logarithm for the
fold number, instead of the number itself. A doubling of the common logarithm for the fold
number is thus interpreted as a doubling of the folding endurance. It can be worth mentioning
that folding endurance is the only existing fatigue test for paper.
58

The folding endurance is especially important for certain paper grades such as banknote pa-
per, map paper, files and telephone catalogue covers. The folding endurance is also often used
to evaluate ageing stability of paper used for archive purposes.
Paper can be folded and reopened many times like a hinge. In the folding of a package, fold-
ing marks are made in the carton board, so-called creases, without reducing the tensile strength
noticeably. If the paper is wrinkled, permanent folds are formed with a rustling sound. It is not
possible to bend paper to a double-curved surface without a fold or dent arising.

44°

156°
156°
7
1

Figure 2.44. The principle for fold number determination according to Köhler-Molin.

OKOP==qÜáÅâåÉëë=aáêÉÅíáçå=píêÉåÖíÜ=~åÇ=aÉä~ãáå~íáçå=oÉëáëí~åÅÉ

In many contexts, the paper or board must have sufficient strength in the thickness direction of
the sheet. A large number of methods are used in practice.
The Z-direction tensile test is the oldest method for measuring the internal bond strength. In
this test, a paper is fastened between two specimen holders, and then pulled apart by the appli-
cation of a tensile load perpendicular to the plane of the paper, Figure 2.45.
 59

Adhesive Specimen holder

tape Paper

Specimen holder

Figure 2.45. Z-directional tensile strength corresponds to the stress at failure of a test piece when pulling two
solid blocks apart.

The Z-directional tensile strength is defined as the force required to produce unit area of fracture
(kPa). A close relationship exists between Z-strength and apparent density for fully bleached
chemical pulps, Figure 2.46.

800
pulp
softwood sulphate (A)
600 softwood sulphite (I)
hardwood sulphate (M)
hardwood sulphite (O)
ZD-strength (kPa)

400

200

0
400 500 600 700
density (kg/m3)

Figure 2.46. Z-directional tensile strength is closely related the density of the papers made from different chemi-
cal fibres.

The Z-toughness method is a rather new method for measuring the delamination resistance of
paper materials It is based on fracture mechanics and it describes the crack propagation process
in a material for the opening of a crack in mode I, normal loading. The Z-toughness value ex-
presses the energy liberated per unit crack area formed, in fracture mechanics called the critical
strain energy release rate.
60

In this method, the test piece is attached between two polycarbonate beams with double-sid-
ed adhesive tape to create a double cantilever beam (DCB). The shape of the beams is designed
to give a constant force when the beams are pulled apart and the crack propagates through the
test piece. If the force and the shape and bending stiffness of the plastic beam are known, the Z-
toughness value can be calculated.
In this method, the paper samples were laminated on both sides with double-sided adhesive
tape, and then cut into test pieces (25×150 mm) in CD. The test pieces were then fastened be-
tween the polycarbonate beams as indicated in Figure 2.47. After a pressing procedure, a force
was applied so that a crack was initiated at the top end and propagated downwards. During the
splitting of the paper, the force required for delamination was recorded as a function of beam
deflection. The beams, with a constant width, were contoured to give a constant force during
crack propagation. For practical reasons, the force at the maximum point in the delamination
curve was used for the evaluation of Z-toughness, according to the equation:

GIc c˜ F2 (J/m2) (2.36)

where
c = constant depending on the elastic modulus of the polycarbonate beams
F = the force at the maximum point in the delamination curve

F F

Figure 2.47. Shape of the contoured double cantilever beams. The position of the paper is shown and during the
delamination the beams are in a vertical position. SCAN-P 90:03. Paper and board. Z-directional toughness.

OKOQ==jÉíÜçÇë=íÜ~í=páãìä~íÉ=`çåîÉêíáåÖ=léÉê~íáçåë

By simulating a printing or a converting process, it is possible to predict how a material will re-
spond to the real situation, but this provides no knowledge or understanding of how the inner
strength of the paper relates to the deformation or loading situation.
 61

A number of methods for assessing the ability of paper to withstand forces involved in the
splitting of ink films in printing processes have been developed. The IGT tester measures the re-
sistance to picking and delamination of paper and board (International Standard 1980).
An oil or ink film of controlled thickness and controlled viscosity is applied on a printing
disc. The printing disc is placed to form a nip with a semi-circular sector on which the test strip
of paper or board is mounted Figure 2.48. The sector rotates with increasing velocity with a
controlled constant printing force. The delamination resistance is defined as the minimum ve-
locity at which delamination occurs.

paper or board strip

printing disc
on the semi-circular
sector

Figure 2.48. Principle of the IGT tester.

OKOR==jÉíÜçÇë=Ñçê=mêçÇìÅíáçå=`çåíêçä

The Scott bond test is a dynamic, multi-fracture test, which is commonly used for paperboard.
The test specimen is laminated with double-sided adhesive tape and fastened between a rigid
foundation and an angular-shaped piece of metal, as shown in Figure 2.49. A pendulum is re-
leased and hits the angle causing the test piece to split. A calibrated scale indicates the resis-
tance to splitting in terms of the loss in potential energy during the swing of the pendulum. The
Scott bond value is therefore expressed in energy units for a unit surface produced in this test
(TAPPI 1971).
This test corresponds to a mixed mode loading situation which is hard to describe.
62

Figure 2.49. The Scott bond test gives the energy absorbed by the failure of the test piece, measured by the loss in
potential energy when a pendulum hits an angle attached to the test piece by adhesive tape.

Different configurations of the peel test are shown in Figure 2.50. The T-peel test is com-
monly used for plastic materials.

peel angle q 90° peel

90°
q

180° peel T-peel

180°

Figure 2.50. Several typical peel configurations.

The Brecht-Knittweis method (DIN1985) is similar to the Z-direction tensile test except that
the loading is in the edge of the test piece instead of in the centre, Figure 2.51. A test piece
(30x30 mm) is mounted between two blocks, the blocks are pulled apart, and the force at failure
of the test piece is referred to as the delamination strength.
 63

specimen holder
adhesive tape paper

specimen holder

Figure 2.51. The Brecht-Knittweis method is similar to the Z-direction tensile test, but the force is applied at the
edge of the test piece.

OKOS==cê~ÅíìêÉ=jÉÅÜ~åáÅë

The fracture mechanics of paper are of great importance for the paper industry. The most obvi-
ous application of fracture mechanics to paper is to reduce the number of web breaks in convert-
ing operations. Fracture mechanics is further a powerful tool in the process of developing
tougher, more fracture-resistant, paper qualities.
The fracture toughness of paper is a material property expressing the ability of paper to sus-
tain mechanical loading in the presence of defects. The fracture toughness of paper can be eval-
uated by material testing.
The reason for using fracture mechanics in the paper industry is that paper very seldom is
completely free from defects. In fact, paper may contain many different types of defects or
cracks. The problem is that these defects occur very seldom but may cause a web break when
the conditions are unfavourable, high force, large crack and low fracture toughness. Figure 2.52
illustrates a few possible defects in paper.

shives

edge tears slime holes calender wrinkles

Figure 2.52. A few possible defects in paper.

The purpose of fracture mechanics theory is for instance to calculate the critical force
Fcrit (N) and the critical elongation Gcrit (m) of a paper web with a defect of known size. The
fracture toughness of the paper material is a necessary input in such calculations.
64

A method for predictions of in-plane fracture of paper has been developed, based on non-lin-
ear fracture mechanics and the J-integral theory. This method, which has been standardized as
SCAN-P77, is described further below.
Two different material tests are required for making predictions of fracture in paper possible,
Figure 2.53. The first experiment is a tensile test on a rectangular paper test piece with the di-
mensions 100 by 15 mm. This material test is used to determine the mechanical behaviour of
the paper material. The second experiment is a tensile test on a 100 by 50 mm rectangular test
piece with a manufactured 20 mm centre crack. This second test, together with the known mate-
rial behaviour from the first test, is used to determine the fracture toughness (JIc) of the paper
material.

1.
bulkmaterial 3.
parameters

2. fracture
toughness critical force
and
critical elongation

Figure 2.53. The evaluation principle for calculation the fracture toughness and the critical force and critical elon-
gation for a given paper geometry and crack length.

The material behaviour and the fracture toughness of the material constitute the necessary
data for making predictions of fracture possible. Furthermore, such predictions generally re-
quire the adoption of numerical methods, such as the finite element method (FEM). In order to
make user-friendly and accurate fracture mechanics predictions in paper possible, a fracture me-
chanics software program, FractureLab see Figure 2.54, has been developed based on pre-cal-
culated FEM reference solutions. FractureLab can be used for prediction of the critical force
and the critical elongation of a full-scale paper web, based on a given crack size and crack posi-
tion. FractureLab can also be used for predictions of the critical crack size in a full-scale paper
web under a given loading situation.
 65

Figure 2.54. Fracture Lab calculates critical force and critical elongation for a given paper geometry and crack length.

Note: It is stressed that the fracture toughness should be used as a parameter in the fracture
mechanics predictions of critical force and critical elongation only. It is not to be used for rank-
ing of papers.
The following notations are used.
b (J/m) (2.37)
Fracture toughness J IC
b
w J IC (Jm/kg) (2.38)
Fracture toughness index J IC
w
The advantage with a predictive tool is that full-scale situations can be analysed at the labo-
ratory. The fibre treatment and papermaking conditions can be optimised and expensive pre-tri-
als minimised. Figure 2.55 shows results from a trial where predicted and measured values for
critical force are compared. The test piece dimensions are 100 mm long and 500 mm wide.

2.0
measurements
prediction
1.5
h = 1000 mm
critical force (kN)

1.0 a

0.5

0.0
0 10 20 30 40 50
crack length (mm) W = 500 mm

Figure 2.55. Results from a trial where predicted and measured values for critical force are compared. The crack
is made at the edge, an edge crack.
66

OKOT==^áê=mÉêãÉ~åÅÉ

A paper property which is closely related to its structure is the ability of the paper to allow air to
pass through, i.e. the air permeance. This property is especially important for example in the
filling of cement in paper sacks. With increasing beating, the air permeance decreases.
In practice, air permeance measurements are carried out in the following way: Air is allowed
to penetrate through a paper sample with given dimensions under standardized pressure, tem-
perature and humidity, and the amount of air which passes through the paper per time unit is re-
corded. The air permeance Pa is calculated according to Equation 2.39:

Q
Pa (m/s Pa) (2.39)
A 'p

where
A = area, m2
'p = pressure difference across the paper, Pa
Q = the air flow through the paper, m3/s
Air permeance according to Gurley (SCAN-P 19:78) is the most common method, Figure
2.56. In the Gurley-apparatus, the pressure difference is 1.21 kPa. The air permeance Pa can be
evaluated as:

128 (Pm /s)/Pa (2.40)

Pa
t

Where t = the time in seconds required for 100 ml of air to pass through the test piece
Besides the Gurley-apparatus, there are also a number of other instruments for the measure-
ment of air permeance, such as the Bendtsen, Bekk and Sheffield instruments.

internal cylinder

external cylinder

oil
clamping plate
and packing
hole for air discharge
 28.7 mm

Figure 2.56. Sketch of Gurley-apparatus.

 67

In the Sheffied-instrument, certain parts can be exchanged so that the measurement area of
the paper can be changed. The openings are 1/2, 1, 2 and 4 inches in diameter and in theory; this
means that the ratio of the areas of two successive openings is 4. This again means that the
amount of air, which passes per unit time should vary in the same ratio, but a certain caution is
advised here. An increase in the amount of air flowing through the apparatus can lead to a great-
er pressure drop in other parts of the apparatus and the real pressure drop across the paper is
thus less. Leakage around the edge can also influence the measurement result. This may mean
that the amount of air becomes slightly smaller than expected and the opening should not there-
fore be changed when a set of samples is being compared.
Both the Gurley-method and the method described in SCAN-9 26 for dense paper grades and
carton board only consider penetrating pores and, with regard to gas transport, these are the
most interesting. If it is also desirable to include the closed pores and to measure the total pore
volume, other methods exist.
The air permeance is closely related to the pore structure of the paper. An illustration of this
is given in Figure 2.57.

5000
trial 1 RH (%)
unfilled 19
4000
66
 40 % clay 19
air permeance (ml/min)

3000 66
trial 2
unfilled 19
2000 66
 25 % clay 19
66
1000

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
void volume fraction

Air permeance versus void volume fraction for trial 1 and 2.

Trial 1. Different densities obtained by wet pressing.
Trial 2. Different shrinkage levels during drying.

Figure 2.57. Air permeance as a function of the void volume of the paper.
100
 69

P==aÉîÉäçéãÉåí=çÑ=m~éÉê=mêçéÉêíáÉë=ÇìêáåÖ=aêóáåÖ

Torbjörn Wahlström
The Packaging Greenhouse AB

3.1 Introduction 69
3.1.1 Scope 69
3.1.2 Level of Observation 70
3.1.3 Drying of Paper 71

3.2 Nano Level – Forces Within or Between Fibres 73

3.2.1 Bonding Mechanisms 73
3.2.2 Network Forces 75
3.2.3 Capillary Forces 75
3.2.4 Chemical Forces 77

3.3 Micro Level – Fibre and Network 79

3.3.1 Single Fibre 79
3.3.2 Fibre Network 81

3.4 Macro Level – Paper Sheet 84

3.4.1 Shrinkage, Stretch and Paper Properties 84
3.4.2 Multi-Ply Board and Drying 89
3.4.3 Stock Preparation and Drying 90
3.4.4 Forming and Drying 92
3.4.5 Pressing and Drying 97

3.5 Machine Level – Moving Wide Web 97

3.5.1 Cylinder Drying 97
3.5.2 Drying Concepts 103

3.6 References 107

PKN==fåíêçÇìÅíáçå

PKNKN==pÅçéÉ

This chapter deals mainly with the development of paper properties during drying and how they
depend on the chosen drying concept. The influence of the unit processes preceding drying will
also be dealt with to show how they influence the paper‘s behaviour during and after drying.
The chapter aims at giving a conceptual understanding of fibre and network level and a ground
70

for calculations and simulations on a homogeneous sheet level. Due to its usability in mechani-
cal modelling and high sensibility to shrinkage and stretch tensile stiffness is used to demon-
strate the drying related effects, but later in the chapter also influences on other paper properties
are given.

PKNKO==iÉîÉä=çÑ=lÄëÉêî~íáçå

Depending of background of the researchers and their purposes drying has been studied on ob-
servation levels ranging from a nano to a macro level according to Figure 3.1.

length (m) level of observation

nano 10–9 m bonding mechanisms

micro 10–5 m fibre and network

ZD
CD
MD macro 10–2 m paper

machine 100 m moving web

Figure 3.1. Different levels of observation of drying in papermaking.

The description in this chapter will start on the nano level with the bonding mechanisms
within and between the fibres. On the micro level the behaviour of a single fibre and fibres
forming a network is described. On the macro scale we treat the paper as a homogeneous mate-
rial and describe its behaviour from different perspectives. Most emphasis is put on the macro
scale since bonding and fibre behaviour are also dealt with in other chapters of this teaching ma-
terial. Also the homogeneous perspective makes it possible to formulate models that can be di-
rectly applied on the next level of observation, the paper machine level. On this last level a wide
 71

and moving web is dealt with. Depending on paper machine concept the behaviour described on
the macro level will develop in different ways.

PKNKP==aêóáåÖ=çÑ=m~éÉê

Figure 3.2, shows schematically the amount of water in the different unit processes of paper
making. Paper is normally formed from a suspension of fibres at a concentration of 0.1–1.0 %
and in special cases at a concentration of up to 4 per cent. In the outflow on the wire, the suspen-
sion is air free until the so called dry line, where air begins to be sucked through the sheet. The
dry content, or dryness, at the dry line is around 3–5 per cent. With air suction across dry suc-
tion boxes and the couch, the dry content can be increased to circa 20 per cent before the web
enters the press section. In the press section, a further increase in dry content takes place by
stages until a dry content of the order of 40 per cent has been attained. After pressing, water is
present in the fibres and as free water between the fibres and in the lumen. The water-fibre sys-
tem is now located in an atmosphere where the relative humidity is less than 100 per cent,
which means that we are concerned with a drying process.
The chosen drying concept will have a large influence on the final paper properties but since
paper is mainly dried by cylinder drying most emphasis will be put on this concept. In cylinder
drying, the web is drawn through the dryer in the machine direction and heated on steam-heated
drying cylinders. Evaporation of the water takes place mainly in the free draws between the cyl-
inders. The paper properties develop with and is very sensitive to decreasing water content in
the paper. The paper properties are also very dependent on the amount of shrinkage that the pa-
per is allowed to develop during drying. In cylinder drying the paper can shrink in the cross di-
rection in the free draws between dryer cylinders.

forming pressing drying

dryness (%) 0.5 20 40 95

mr (kg/kg) 200 4 1.5 0.05

Figure 3.2. Water content in the papermaking process.

As a measure of water content in a paper sample the dry content, dryness or solids content,
according to Equation 3.1, is close to an industry standard. However, dryness is not a linear
measure of the water content in paper. Due to the close relationship between the water content
and paper mechanical properties, a linear measure such as the moisture ratio, mr, defined ac-
cording to Equation 3.2 is more useful. Figure 3.3 shows the relation between moisture ratio
and dryness. For example a 10 % increase in dryness from 40 to 50 % is equal to remove 0,5 kg/
kg water, whereas an increase from 80 to 90 % equals only 0,15 kg/kg water.
72

Weight of dry material

Dryness (3.1)
Total weight

Weight of water
mr (3.2)
Weight of dry material

2.0

1.5
moisture ratio (kg/kg)

1.0

0.5

0
0 20 40 60 80 100
dryness (%)
Figure 3.3. Relation between moisture ratio and dryness.

In the literature the water in wet pulps is classified in different ways relating to its location
within or between the fibres but also to its properties. „Free water“ has properties like normal
bulk water and is found between the fibres and in large pores. Water that interacts and is bound
to different degrees with the pulp has been named „Pore water“. In Figure 3.4 the „Pore water“
is divided into semi bound or „Freezing“ and bound or „Non-freezing“ water, where the Non-
freezing bound water is understood as directly bound water. First only free water is removed but
below a water content of around 1.0 kg/kg all fractions are removed simultaneously until all
Free and semi bound water has been removed. Finally only bound water remains. The different
water fractions were measured with DCS, differential scanning calorimetry. Deodhar and Luner
(1980) assumed that the largest pore that can carry only directly bounded water has a radius of
40 Å.

0.6 100

Total Free Bound

Water fraction, kg/kg

0.5
80
Water fraction, %

0.4 Semi
60
bound Free
0.3
Bound Semi-
40
0.2 bound

0.1 20

0.0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Moisture ratio, kg/kg Moisture ratio, kg/kg

Figure 3.4. Absolute and relative water fractions in birch kraft pulp during drying (Maloney 2000).
 73

PKO==k~åç=iÉîÉä=Ó=cçêÅÉë=táíÜáå=çê=_ÉíïÉÉå=cáÄêÉë

PKOKN==_çåÇáåÖ=jÉÅÜ~åáëãë

In papermaking the water is removed gradually and the fibre surfaces are forced into contact
with each other. Figure 3.5 shows schematically how bonds are established between two fibre
surfaces in forming, pressing and drying. Fibril threads, fines and hemicellulose form a swollen
gel like layer which facilitates the creation of attraction forces. In a wet sheet, these bonds give
a certain wet strength to the sheet.

forming pressing drying

Figure 3.5. Schematic illustration of how bonds are formed between two fibres during papermaking. (Nanko and
Wu, 1995).

The beating is very important for the size of the bonds. The beating makes the fibre more
flexible and this facilitates bonding. The beating also creates a fine material which fills the cav-
ities between the fibres. Figiure 3.6 shows schematically the bonding structure between two fi-
bre surfaces with increasing degree of beating from left to right. Figure 3.7 shows an electron
micrograph of a fibre structure which corresponds to the highly beaten structure in Figure 3.6.
Note that the fines have filled the cavity between the fibres, and that this creates a denser sheet.

S2 S1 S2 S2 S1 S2 S1 B S1 S2 S1 B S1

Figure 3.6. Schematic illustration of the bonding structure between two fibres with increasing degree of beating
from left to right. S1 and S2 are layers in the secondary wall of the fibre, B is fines. (Nanko and Wu, 1995).
74

Water is a necessary component to create strong paper. Attempts have been made to form pa-
per in a dry state but these papers have considerably lower strength than the wet-formed materi-
al. Through the different processes of forming, pressing and drying, the properties of the paper
can be influenced within wide limits. Figure 3.8 shows in principle how the tensile index of a
paper increases as the dry content is increased. Three types of forces can be distinguished which
act in different solids content ranges: mechanical or network forces, capillary forces (so-called
Campbell forces) and chemical forces.

Figure 3.7. Bonds between two well-beaten fibres in a paper. (Nanko and Wu, 1995).

100.00
Tensile index, kNm/kg

10.00 Chemical

1.00 Capillary

Network
0.10

0.01
0 2 4 6 8 10 12 14 16
Moisture ratio, kg/kg

Figure 3.8. Development of tensile index for paper of sulphite pulp with decreasing moisture ratio. (Robertson).
 75

PKOKO==kÉíïçêâ=cçêÅÉë

A fibre in free rotation in a diluted fibre suspension covers a spherical volume with a diameter
equal to the fibre length. The maximum concentration of totally free fibres with the average
length L can then be represented by closely packed spheres with the diameter L. The fibre sus-
pension flows almost like water and there are no network forces.
The sediment concentration is the lowest concentration where a network with an inherent
strength is formed. If a known weight of fibres are carefully added to water in a beaker and al-
lowed to sink by sedimentation, the height of the sediment gives the weight concentration. For a
sulphate pine pulp the weight or sediment concentration is 2–3 g/l or 0,2–0,3 %.
If an amount of fibres exceeding the sediment concentration is carefully added to water the
suspension does not automatically achieve network strength. It may be necessary to add turbu-
lent energy by stirring. During the stirring operation the fibres are deformed from their natural
shapes. After stirring the fibres tries to retain their original shapes but may be hindered and
locked by other fibres, thus network strength has been created. As is shown in Figure 3.9 three
contact points are needed to lock a fibre in a given position.
If the mechanical locking of the fibres are crucial for the network strength it is required that
the network strength is increased with increasing fibre concentration. Wahren has shown with
measurements in a rheometer that the network strength is dependent on the fibre concentration.
According to his findings the network strength is zero at the sediment concentration and there-
after increasing.

Figure 3.9. Three contact points are needed to lock a fibre in a given position.

PKOKP==`~éáää~êó=cçêÅÉë

Surface physics can be applied to the liquid meniscus in a fibre contact between parallel fibres
according to Figure 3.10. The smaller the radius of curvature of the liquid meniscus during
evaporation, the greater is the attraction between the fibres. The attractive force is proportional
to the underpressure in the water, which is proportional to the surface tension and inversely pro-
portional to the radius of curvature of the liquid meniscus. The underpressure and thus the at-
tractive force theoretically approach infinity when the volume of water approaches zero. In
reality, the increase in the attractive force gives rise to a local deformation of the fibres and the
line contact is transformed to a surface contact, which limits the size of the force.
76

The existence of low pressures in a liquid with a small negative radius of curvature is the rea-
son why fibres and fines attract each other when air is present, i.e. after the „dry line“ in the pa-
per machine. Capillary forces or „Campbell forces“ is usually used as the collective name for
these attraction forces. The fines created in the beating increase the capillary forces in the wet
sheet, which is one of the reasons why a higher fines content gives a higher wet strength. Exam-
ples of capillary forces between parallel fibres in everyday life are paint brushes, hair locks and
the well-known procedure of wetting a sewing thread before threading it through the needle. A
wet thread is approximately 100 times stiffer than a dry thread. The attractive forces between
perpendicular fibres are considerably lower than those between parallel fibres.

r
attractive force

drying
process water

water content

Figure 3.10. The negative pressure between two fibres increases with decreasing amount of water meaning that
the attractive forces between the fibres increase up to an equilibrium level (Skelton).

Consider a paper after it has been beaten, formed and wet-pressed. The swollen fibres are
then surrounded by thin water films. Gradually the water film around the fibres breaks, and the
water is then collected where fibres have the closest contact. The capillary forces now attract the
fibres to each other, as illustrated in Figure 3.10. The more softened and swelled the fibres are,
from for example beating, the easier the fibres form themselves after each other when they are
contracted by the surface tension forces. The fibres are also influenced by capillary forces
which act in the lumen of the fibres, and this contributes to a flatter shape of the fibres after dry-
ing. The importance of the capillary forces for the development of fibre bonding can be illustrat-
ed in a comparison between laboratory sheets dried by freeze-drying and laboratory sheets dried
in an oven, Figure 3.11. Freeze-drying means that the sheet is first frozen so that the water
forms ice. The ice is then removed by sublimation (frozen water evaporates as vapour directly
from the solid state). The freeze-dried sheet has much fewer collapsed fibres and free fibril
threads can be seen in the structure. Nor is there any optical contact between the fibre walls, so
that the light-scattering ability of the freeze-dried sheet is greater than that of the heat-dried
sheet. The freeze-dried sheet also has a much lower strength than the heat-dried sheet.
 77

Figure 3.11. Left picture; Freeze-dried laboratory sheet of kraft pulp. Right picture; Heat-dried laboratory sheet
of kraft pulp (Lobben).

PKOKQ==`ÜÉãáÅ~ä=cçêÅÉë

When the water below the liquid menisci decreases and contracts further into smaller pores be-
tween and in the fibres, the radii of the menisci becomes smaller and the under pressure in the
liquid becomes greater. The surface tension now very efficiently pulls the fibres into such inti-
mate contact that chemical bonds can start to form between the fibres.
The atoms comprising the water molecule have a V-configuration with the oxygen atom at
the angle and a hydrogen atom at the end of each arm according to Figure 3.12. The oxygen at-
om, which is strongly electro negative, attracts electrons away from the hydrogen atoms; this
leaves it with a net negative charge and both hydrogen atoms with a net positive charge. This re-
sults in electrostatic attraction between the hydrogen atoms of one molecule and the oxygen at-
oms of a neighbouring molecule, an example of the so-called hydrogen bond.

hydrogen bond 105°

0.99 A 1.77 A

Figure 3.12. Hydrogen bonding between two water molecules. (From Rance, Handbook of paper science).

The unique properties of water can be derived from the hydrogen bonds between the mole-
cules. Hydrogen bonds are present in water to an unusually large extent, giving rise to a high de-
gree of order. From the heat of fusion, it can be calculated that some 85 % of the hydrogen
bonds present in ice remain unbroken in liquid water at 0 °C and even at 25 °C more than 80 %
of the original number are still intact. Water, that locally may become almost crystalline, differs
78

from a true crystal in that the hydrogen bonds are constantly broken and reformed. The resulting
intermolecular attraction is acting through many hydrogen bonds and, although individually rel-
atively weak, collectively requires considerable energy to overcome. This gives rise to the high
heats of fusion and vaporisation, the high specific heat, the high surface tension and the fact that
water is a liquid, and not a gas, at normal temperatures.
The strength of a hydrogen bond can vary slightly but it is approximately one tenth of that of
the strong covalent forces, which hold organic molecules together. At short distances, the hy-
drogen bond is stronger than the dispersion forces, which lead to attraction between molecules.
The condition for hydrogen bonds is that OH- or NH2-groups exist in a molecule. If such groups
lie close to each other, hydrogen bonds can be formed within the molecule, i.e. intramolecular
hydrogen bonds. These can be found in the cellulose molecule and contribute to its stiffness.
Hydrogen bonds can also be formed between different molecules, so-called intermolecular hy-
drogen bonds. In crystalline cellulose, such bonds can be identified between adjacent mole-
cules. It is generally believed that the connecting force between cellulose chains is essentially
due to the many hydrogen bonds.
The same mechanism of hydrogen bonding results in the attraction of water molecules to the
hydroxyl, or other polar functional groups, of polysaccharides and of these latter groups one to
another. Figure 3.13A shows a representation of two cellulose molecule’s hydrogen bond to wa-
ter molecules and, through these, in effect to one another. For convenience the V-configuration
of the water molecules is shown in linear form. Considerable thermal effort is required to drive
off these residual water molecules since their bonds with cellulose, even when ruptured, readily
reform elsewhere. If carried far enough, however, the drying process reaches a stage at which
the two cellulose molecules are bound through a monolayer of water molecules, as in Figure
3.13B, and ultimately directly one to another (Figure 3.13C) as in the crystalline zones of the
cellulose micro fibril.

a) b)
cellulose molecule

c)

cellulose molecule

Figure 3.13. Hydrogen bonding between two cellulose molecules: (A) loosely through water molecules; (B) more
tightly through a monolayer of water molecules; and (C) directly. (From Rance, Handbook of paper science).

Since bonds between cellulose chains within a fibril can be explained essentially by hydro-
gen bonds, it is not unreasonable to assume that hydrogen bonds can play an important role for
the bonds between fibres. However, it should be pointed out immediately that the dispersion
forces which exist between all objects of course also act between cellulose fibres. The fibres
must come very close to each other for direct hydrogen bonds to be formed between the fibres.
 79

Well beaten and swollen fibres develop good contact between the fibre surfaces, through the in-
fluence of capillary forces in the consolidation of the sheet, and this encourages the creation of
hydrogen bonds. This is favoured for example by internal and external fibrillation and fines for-
mation in beating and by the occurrence of hemicellulose on the fibre surface. Mechanical fi-
bres are relatively stiff and do not give such well-developed contact surfaces as chemical fibres.
Since the surfaces on mechanical pulp fibres consist partly of lignin with fewer OH-groups,
there are fewer possibilities for hydrogen bonds to develop than on cellulose and hemicellulose.
This is one of the reasons why mechanical pulps give a low tensile index. However, it should be
remembered that the dispersion forces between the fibres can play an important role for the fi-
bre bonds. For the dispersion forces, the geometrical requirements are not as strict as for hydro-
gen bonds. Also the dispersion forces do not decrease as rapidly with increasing distance.
The strength of a wet paper sheet is only a few per cent of the strength of the dry sheet. This
can be understood as follows. When two fibres are to be separated by mechanical forces, hydro-
gen bonds must be broken and dispersion forces must be overcome. In a wet sheet, water mole-
cules are always available which can penetrate the bonding zones and form hydrogen bonds
between the exposed fibre surfaces. The dispersion forces are also much weaker when there is a
medium between the objects which interacts with the dispersion forces.

PKP==jáÅêç=iÉîÉä=Ó=cáÄêÉ=~åÇ=kÉíïçêâ

PKPKN==páåÖäÉ=cáÄêÉ

Shrinkage. Seen in a simplified way, crystalline micro fibrils lie at a small angle (5–15°) to the
fibre axis. Between the highly crystalline cellulose fibrils lies a matrix consisting of an amor-
phous material, mainly hemicellulose and lignin. When the swollen fibres begin to dry, the ma-
trix wishes to shrink. Since the crystalline micro fibrils lie in the length direction of the fibre,
shrinkage is prevented in this direction. It is natural that the fibre instead shrinks most in its
transverse direction. Figure 3.14 shows how the cross-sectional area of a fibre decreases from
the swollen to the dry state during drying. The authors reported cross section shrinkage of about
20 %, whereas the length or axial shrinkage of the fibre is in the order of only a few percent.
The transverse shrinkage of the fibres in Figure 3.14 is around 15 %, it will be shown later in
this chapter that the shrinkage of the fibre has a large influence on the paper properties.

Figure 3.14. Example of how the cross-sectional area of a fibre is changed from a swollen state (black) to a dry
state (lines). (Page and Tydeman 1966).
80

The left picture in Figure 3.15 shows the development of the transverse shrinkage for a
bleached kraft pulp fibre during drying. The transverse shrinkage was evaluated as the cross
section width, Dmax, reduction defined in Figure 3.16. The measurements were made with a con-
focal microscope on a surface fibre in a hand sheet. It is evident that the shrinkage of the fibre
happens very late in the drying process. As a measure of the structural changes in the fibres the
water retention value, WRV, was measured on the same fibres to characterise the hornification
of the pores in the cell wall (right picture in Figure 3.15). The hornification takes place during
the whole drying process. Down to a moisture ratio of 2.0 kg/kg (dryness up to 33 %) no horni-
fication took place. Based on the results in Figure 3.15 and visual observations a model of the
shrinkage progress according to Figure 3.16 has been proposed. The top row shows the shrink-
age of a fibre cross section and the bottom row an element of the fibre fine structure. Phase A-C
represents the large part of the drying process where almost no shrinkage of the cross section
width occurs. Water is removed from large fibre pores and voids between the lamellae of the fi-
bre wall are collapsed and causes a reduction in the WRV. The fact that almost no cross section
width shrinkage can be observed in this region means that the shrinkage of the cross section
area and the closure of large pores happen perpendicular to the lamellar layers. Note that phase
A–C is the inverse of Scallan’s well known swelling model. The late shrinkage observed in Fig-
ure 3.15 happens during phase D in Figure 3.16, which shows the effect of drying of the fine
structure. This last fraction of water is directly bound water dealt with in Figure 3.4.

40 2.00
35 1.75
30 1.50
Shrinkage, %

WRV, kg/kg

25 1.25
20 1.00
15 0.75
10 0.50
5 0.25
0 0.00
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Moisture ratio, kg/kg Moisture ratio, kg/kg

Figure 3.15. Transverse shrinkage and WRV development for bleached kraft pulp fibres during drying. (Weise
et.al., 1996).

Properties and drying history. The left hand part of Figure 3.17 shows the tensile stiffness
and strain at break versus a large compression of a single fibre during drying. The fibre species
was a holocellulose springwood fibre. The stiffness was reduced and the strain at break in-
creased by the compression. The right hand part shows the elastic modulus and strain at break
versus an applied load in grams on a single fibre during drying. The fibre species was a long
leaf pine holocellulose springwood fibre. The stiffness was increased and the strain at break was
reduced by the load. Note that applying a load on the fibre is equal to actually stretching the fi-
bre, the amount of stretch that the load corresponded to, was however not measured. The effect
of the applied load on the fibre properties is sometimes referred to as the „Jentzen effect“.
 81

a) b) c) d)

Dmax

Figure 3.16. Shrinkage progress for a fibre cross section (top) and an element of the fibre fine structure (bottom).
(Weise et al., 1995).

4.0 4.0
Tensile stiffness
3.5
Normalised value, -
Normalised value, -

3.0 3.0
Strain at break
2.5
2.0 2.0
1.5
1.0 1.0 Strain at break
Tensile stiffness
0.5
0.0 0.0
0.0 5.0 10.0 15.0 0.0 1.0 2.0 3.0 4.0
Compression, % Load, g

Figure 3.17. Tensile stiffness and strain at break of a single fibre during drying versus compression (Dumbleton)
and load (Jentzen).

PKPKO==cáÄêÉ=kÉíïçêâ

Previous section dealt with the single fibres behaviour during drying. The question is now how
their strength and the structural changes they undergo during drying is transferred to the net-
work of fibres that we call a paper.
Since the fibres in their wet, swollen state have low compressive strength, the width shrink-
age of the fibres is able to cause corresponding length shrinkage of each crossing fibre at the
bonding sites. This shrinkage at the fibre bonding zones has been named „micro compressions“
in the literature (Page and Tydeman). Note that this term is also used for a beating effect involv-
ing local damage of the fibre wall. The occurrence of micro compressions at the bonding sites
requires that the bond develop first and thereafter the shrinkage. Considering the very late de-
veloped transverse shrinkage in Figure 3.15 this seams to be reasonable.
82

It has also been shown experimentally that the shrinkage actually happens at the bonding
sites. Figure 3.18 shows the average longitudinal shrinkage of a sheet, an individual fibre in the
sheet and the segments (bonding sites and free segments) of a fibre in a fibre network (sheet).
The paper sheet was dried both free and restrained. Free drying means that there are no external
forces which prevent the shrinkage. Note that the shrinkage for the free dried case happened
only at the bonding sites. In the restrained dried case the sheet is clamped and the fibre as a
whole cannot shrink. But the bonding sites were still shrinking and they were actually stretching
the free segments. In the literature this is often referred to as activation of the fibre network. To
the right in Figure 3.18 is a schematic illustration of the measured results. Figure 3.19 shows
the distribution of the average segment shrinkage presented in Fig. 46.18. The distribution
shows that also for free drying some of the free segments are stretched but the average result is
zero.

10.0 Bonded
segment
8.0
Fibre
Free Drying
Shrinkage, %

6.0 Sheet Restrained

4.0

2.0
Free
segment
0.0

-2.0

Figure 3.18. Shrinkage of the sheet, fibre and segments during free and restrained drying. (Nanko and Wu, 1995).

35 35
Bonded segments
30 30 Free segments
Free drying
25 25 Free drying
Frequency, %
Frequency, %

Restrained
20 20 Restrained

15 15

10 10

5 5

0 0
-6 -4 -2 0 2 4 6 8 10 12 14 16 18 -6 -4 -2 0 2 4 6 8 10 12 14 16 18
Shrinkage, % Shrinkage, %

Figure 3.19. Distribution of shrinkage of bonded and free segments during free and restrained drying. (Nanko and
Wu, 1995).

Figure 3.20 shows electron micrographs of paper after free and restrained drying. The micro-
graphs confirm that the sheet dried under restraint has straighter fibres and fewer microcom-
pressions than the freely dried sheet.
 83

Figure 3.20. (Left)Electron micrographs of freely dried paper and (Right) paper dried under restraint. (Nanko and
Wu, 1995).

The shrinkage is greater in CD than in MD for a free dried paper which is explained in the
following way; in most papers, there is a larger number of fibres in MD than in CD. Each fibre
crossing causes shrinkage with the development of micro compressions and each fibre has more
bonding points per fibre in CD. Therefore, the sheet shrinks more in CD. Let’s assume we have
a unit cell of a schematic isotropic fibre network with ten fibres according to Figure 3.21. Iso-
tropic means without direction and is in this case demonstrated with five fibres in each in-plane
direction, MD and CD. The width of the unit cell is 300 μm and the fibre width is 30 μm. We
know that the width shrinkage of a fibre is in the order of 20 % or 6 μm in this schematic case.
The crossing fibre will however restrain a part of the fibres shrinkage potential. With reference
to Figure 3.18 it is assumed that the actual shrinkage of the bonding sites is 10 % or 3 μm in
both directions of the bonding site. With these assumptions the shrinkage of the unit cell will be
5 % in both in-plane directions. If we now redistribute the ten fibres in the unit cell to achieve
an anisotropic paper (a paper that has a direction or a MD and CD direction), we can for exam-
ple get the networks and anisotropies according to Figure 3.21. The anisotropy, A, is calculated
as the number of fibres in MD divided by the number of fibres in CD, giving the anisotropies
1.5; 2.3 and 4. The shrinkage in MD and CD for the four networks calculated with this simple
model is presented in Figure 3.22.

A = nMD /nCD = 5/5 = 1.0 A = 6/4 = 1.5 A = 7/3 = 1.3 A = 8/2 = 4

Figure 3.21. Schematic pictures of fibre networks with different anisotropies.

84

12.0
Network Model
10.0
CD
Shrinkage, %

8.0

6.0

4.0

2.0
MD
0.0
1.0 2.0 3.0 4.0 5.0
Anisotropy (nMD /nCD), -

Figure 3.22. Free shrinkage in MD and CD for the fibre networks in Figure 3.21.

PKQ==j~Åêç=iÉîÉä=Ó=m~éÉê=pÜÉÉí

PKQKN==pÜêáåâ~ÖÉI=píêÉíÅÜ=~åÇ=m~éÉê=mêçéÉêíáÉë

Total strain accumulated during drying. If a paper is deformed, either stretched or allowed to
shrink, during drying it will have a great effect on most of its paper properties. This deformation
can be characterized with the relative elongation or strain, H, according to Equation 3.3 and Fig-
ure 3.23 where G is the elongation and L the original length of the sample. Note that a shrinkage
or reduction in length from now on will be characterized by a negative strain even if it earlier in
this chapter was defined as positive.

G
H (3.3)
L

L
before drying

d
after drying – negative strain or shrinkage
d

after drying – positive strain or stretch

Figure 3.23. Illustration of shrinkage and stretch.

According to Wahlström et al (1999) the total strain that accumulates in a paper during dry-
ing can be decomposed into a sum of free shrinkage strain and mechanical strain. The free
shrinkage strain occurs when moisture is added to or removed from the paper without any exter-
nal forces acting on it. When moisture is added the paper expands and the free shrinkage strain
is positive. Removal of water makes the paper shrink and the free shrinkage strain is negative.
 85

Note that the definition is equal to the hygro expansion of paper with the difference that the
main focus here is on removing, not adding, water.
The mechanical strain is induced in the paper by applying external forces. The mechanical
strain is mainly a positive strain since paper is hard to compress in its in-plane directions. An
exception with negative strain is however the so-called lateral contraction (Poisson’s ratio) dealt
with later in this chapter. A negative total strain is often referred to as shrinkage, a positive as
stretch and a zero total strain after drying as restrained drying. Note that a standard hand sheet is
dried restrained, which means that a mechanical strain equal to the free shrinkage strain is in-
duced in the paper giving a zero total strain. Equation 3.4 shows the decomposition of total
strain, İ, into free shrinkage strain, H fs, and mechanical strain, H m, for an isotropic hand sheet.
Here, the paper will be treated as isotropic, i.e. without direction. In Paragraph 3.4.4, Forming
and Drying, fibre anisotropy will be introduced. An isotropic paper has the same properties in
MD and CD, which is typical for a laboratory made hand sheet.

H H fs  H m (3.4)

Free Shrinkage Strain (Free drying). Drying without any external forces acting on the pa-
per is often referred to as the drying strategy Free drying. The free shrinkage strain that occurs
when a paper is dried with only free drying is often referred to in short as the „free shrinkage“.
The moisture range that the paper goes through during drying must then be defined. The starting
point should always be before the shrinkage starts, which means around press dryness
(mr = 1,5 kg/kg, dryness = 40 %), to some reference point in the dry end, for example bone dry
or complete drying in a conditioned room at 23 °C, 50 % RH.
The free shrinkage strain for an isotropic hand sheet develops exponentially versus moisture
ratio, mr, as shown in Figure 3.24. The solid line represents Equation 3.5 with the free shrink-
age, H fs(mr=0) = –4,6 % (the reference point in the dry end is mr = 0)and k = –2,6.. Note that the
free shrinkage strain does not develop as rapid as the shrinkage of the individual fibre did in
Figure 3.15. The explanation for the difference may depend on the moisture gradient that devel-
ops in the thickness direction of a paper during drying. When the surface layer dries first it will
cause some shrinkage of the sheet also at a higher moisture ratio of the whole paper.

H fs H fs( mr 0)e k ˜mr (3.5)

0.0
Free shrinkage strain ( e ), %
fs

-1.0

-2.0

-3.0

-4.0

-5.0
0.0 0.5 1.0 1.5 2.0
Moisture ratio (mr ), kg/kg
Figure 3.24. Development of the free shrinkage strain during drying for an isotropic hand sheet.
86

Drying stress (restrained drying). If the total strain according to Equation 3.4 is kept at
zero during the entire drying process, a stress will develop in the paper. The drying strategy
where no shrinkage or stretch is allowed during drying is often referred to as restrained drying.
The drying stress for an isotropic hand sheet develops exponentially versus the moisture ratio,
mr, The solid line in Figure 3.25 represents Equation 3.6 with the final drying stress,
V mr = 0 = 4 N/m2 and k = –4. The final drying stress is defined as the final point in the diagram at
a defined final moisture content, 0 kg/kg in Figure 3.25.
There is no standard method for characterisation of drying stresses. Normally heated clamps
or pre drying of the sample at the clamping area is used to avoid water leakage next to the
clamps. The excess water causes a reduction of the maximum stress measured. Stress relaxation
will take place during and after the drying process which makes the measurement depend on
drying time and temperature. Restrained drying with the same temperature and time gives a lin-
ear relation between stiffness and final drying stress before relaxation (Htun and Fellers,1982).
Considering an anisotropic paper the drying stresses are greatest in MD due to the larger
number of fibres in this direction, compare with Figure 3.21. A final drying stress in the sheet
of the order of 10 kNm/kg corresponds to the stress obtained in tensile testing at an elongation
of circa 0,1 per cent. 0,1 per cent is in the elastic range, failure occurs at 60–100 kNm/kg.

V V mr 0e k ˜mr (3.6)

4.5
4.0
2
Drying stress, N/m

3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0.0 0.5 1.0 1.5 2.0
Moisture ratio, kg/kg

Figure 3.25. Development of drying stress during drying for an isotropic hand sheet.

Mechanical strain. When a mechanical strain, or stretch, is applied in one in-plane direction
of a paper, for example MD as in Figure 3.26, it will contract in CD. Figure 3.26 shows this
contraction for straining of a wet strip of paper. The moisture ratio was 1,3 kg/kg which is about
equal to a press dryness.

H CD
m
Q ˜ H MD
m
(3.7)
 87

eMD
m

0.0
), %

-1.0
m(CD)

-2.0
Mechanical strain (e

-3.0
-4.0 eCD
m

-5.0
-6.0
-7.0
-8.0
0.0 1.0 2.0 3.0 4.0

Mechanical strain ( e
m(MD)
), %

Figure 3.26. Contraction in CD due to a stretch in MD.

Development of paper properties during drying. Also the tensile stiffness index for an
isotropic hand sheet develops exponentially versus moisture ratio. as shown in Figure 3.27. The
solid line represents Equation 3.8 with the stiffness for a dry paper, Emr = 0 = 3,1 MNm/kg for the
free dried paper and 7,3 MNm/kg for the restrained dried paper. The exponential constant,
k = –4,8 for both the restrained and free dried paper. Compared to the development of free
shrinkage strain and drying stress an extra constant is needed to describe that the stiffness is
separated from zero in the wet end of the relation. This constant, c, is equal to 0,13 for both the
restrained and free dried paper in this case.

E E mr 0e k ˜mr  c (3.8)
Tensile stiffness index, MNm/kg

8.0

6.0

4.0
Restrained
2.0

Free
0.0
0.0 0.5 1.0 1.5 2.0
Moisture ratio, kg/kg

Figure 3.27. Development of tensile stiffness index during free and restrained drying for an isotropic hand sheet.

Relations between total strain and paper properties. There is a linear relation between
tensile stiffness index and total strain accumulated during drying for isotropic hand sheets
(Wahlström and Fellers 2000). With knowledge of the free shrinkage and the stiffness for a free
and a restrained dried isotropic hand sheet the stiffness can be calculated for a given total strain
88

with Equation 3.9. Er represents the tensile stiffness index for a restrained dried paper and Efs is
measured on a free dried paper. H fs is the free shrinkage and H is the total strain accumulated dur-
ing drying. H r is by definition zero. The solid line in the left figure in Figure 3.28 shows this re-
lation for isotropic hand sheets made of a bleached sulphate pulp with Er = 5,5 MNm/kg,
Efs = 2 MNm/kg, H fs = –6 %. Note however that this relation is not valid for high positive strains
where the stiffness is known to reach a maximum and thereafter decline.
There is also a linear relation between the other standard in-plane tensile properties and total
strain accumulated during drying. The right picture in Figure 3.28 shows relations between total
strain and normalised stiffness, strain at break (stretch) and tensile index. Normalisation was
made, with stiffness as an example, by dividing all values with the stiffness for restrained dry-
ing, Enorm = E/Er. For restrained drying (H = 0) the stiffness was 5,5 MNm/kg, stretch 3,3 % and
tensile 44 kNm/kg. If Equation 3.9 is applied also here k is 0,11 for the stiffness, 0,03 for tensile
and –0,25 for the stretch. Meaning that in this case stretch was most, and tensile least, sensitive
to a change in total strain.

E r  E fs
E kH  E r where k (3.9)
H r  H fs
Tensile stiffness index ( E ), MNm/kg

8.0 3.0
7.0
Stretch 2.5
Normalised value, -

6.0
r 2.0
5.0 (E , er )
4.0 1.5
Tensile
3.0
1.0
2.0
(E , e )
fs fs
0.5
1.0 Stiffness
0.0 0.0
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4

Total strain (e), % Total strain (e) , %

Figure 3.28. Left: Relation between tensile stiffness index and strain accumulated during drying for an isotropic
hand sheet made of bleached sulphate. Right: Normalised results for stiffness, tensile and stretch (same paper).

Some paper properties that are not tested in a specified direction can be dependent on the
change of the area strain rather than the strain in a certain direction. Figure 3.29 shows an ex-
ample for Bendtsen roughness and z-toughness plotted versus the total area strain accumulated
during drying. The paper used were isotropic hand sheets made of a bleached sulphate pulp
dried with a number of different drying strategies. The total area strain is the sum of the total
strain in the two in-plane directions of the paper. The decrease in both roughness and z-tough-
ness can be understood from Figure 3.20, which shows a more even surface but also more open
structure for the restrained dried paper compared to the free dried.
 89

2500 100
Bendtsen roughness, ml/min

80

2
2000

Z- toughness, J/m
1500 60

1000 40

500 20

0 0
-6.0 -4.0 -2.0 0.0 2.0 4.0 -6.0 -4.0 -2.0 0.0 2.0 4.0

Total area strain ( eA ) , % Total area strain (eA ) , %

Figure 3.29. Roughness and z-toughness versus total area strain accumulated during drying for an isotropic hand
sheet made of bleached sulphate.

PKQKO==jìäíáJmäó=_ç~êÇ=~åÇ=aêóáåÖ

Volume model. For modelling the free shrinkage strain, H fs, for a multi-ply laminate a volume
model can be formulated according to Equation 3.10. The product of the grammage, wr, stiff-
ness, Er, and free shrinkage strain for the laminate is assumed to be equal to the sum of the same
products for the individual plies.
The stiffness for a restrained dried laminate, Er, is calculated according to Equation 3.11,
where n is the number of plies, Er, is the restrained tensile stiffness index, w the grammage and
i denotes the individual plies. The grammage of the laminate, wr, is calculated as the sum of the
grammages of the individual plies.
n
r
wLam r
ELamH Lam
fs
¦ wir EirH ifs (3.10)
i 1

r
¦ wir Eir
i 1
ELam n (3.11)
¦ wir
i 1

Total strain and paper properties. Equation 3.9 can be applied also on the laminate and
give a relation between the stiffness of the laminate, E, and the total strain accumulated during
drying, H, giving Equation 3.12. The stiffness for a freely dried laminate, Efs, can be calculated
according to Equation 3.13, where n, is the number of plies, wi , the grammage and Ei , the stiff-
ness that the individual ply has with the total strain equal to the free shrinkage of the laminate.
Ei , is calculated for each ply using Equation 3.9 with H = H fs.
r
ELam  ELam
fs
ELam r
ELam  H (3.12)
H Lam
fs
90

fs
¦ wiH Lam
EiH Lam
i 1
E Lam n (3.13)
¦ wiH Lam

i 1

Figure 3.30 shows relations between stiffness and total strain for a typical liquid packaging
top, middle and bottom ply pulp. The top ply pulp is a bleached sulphate, the bottom ply an un-
bleached sulphate and the middle ply a sulphate and CTMP mix. The relations for each ply is
calculated with Equation 3.9 and the input data in Table 3.1. Based on this the stiffness- strain
relation is simulated with Equation 3.10–3.13.

q~ÄäÉ=PKNK=fåéìí=ãÉ~ëìêÉÇ=çå=Ü~åÇ=ëÜÉÉíëK
píáÑÑåÉëë=oÉëíê~áåÉÇI= píáÑÑåÉëë==cêÉÉI= cêÉÉ=ëÜêáåâ~ÖÉ= dê~ãã~ÖÉ=
jkãLâÖ jkãLâÖ ëíê~áåI=B oÉëíê~áåÉÇI=ÖLãO
qçé=éäó=EqmF TIUQ QIMP JRIRS NOS
jáÇÇäÉ=éäó=EjmF RISO OIUO JOITS NNV
_çííçã=éäó=E_mF UIRN PITT JSINM NOQ

10
Tensile stiffness index, MNm/kg

9
8
7
6
Bottom
5
Laminate 4
Top 3
Middle
2
1
0
-7 -6 -5 -4 -3 -2 -1 0 1 2
Total strain ( e) ,%

Figure 3.30. Measured input data on the plies and simulated laminate.

PKQKP==píçÅâ=mêÉé~ê~íáçå=~åÇ=aêóáåÖ

Raw material and beating. Figure 3.31 shows a comparison in stiffness between paper made
by an unbleached kraft pulp and an unbleached CTMP pulp. The pulps are unbeaten and beaten
to two higher degrees and thereafter evaluated for restrained and free drying. The maximum
beating energy was 200 kWh for both pulps. Generally it can be seen that the kraft paper has
higher stiffness values at restrained drying compared to the CTMP paper. On the other hand the
kraft paper shows a larger reduction with decreasing total strain. The kraft paper made of the
most beaten pulp actually gets a lower stiffness than the CTMP when free dried. An interesting
observation is what happens with the free dried papers when the beating is increased. For the
CTMP both the stiffness and the free shrinkage strain is increased. For the kraft the stiffness
first increases with increasing beating, but thereafter it is reduced again. Obviously the reduc-
 91

tion in stiffness, caused by the free drying, overcomes the stiffness increase, caused by the beat-
ing, for high high beating levels.

Tensile stiffness index, MNm/kg

Tensile stiffness index, MNm/kg

10.0 5.0
Kraft CTMP
8.0 4.0

6.0 3.0 540

370
4.0 2.0 200 CSF

2.0 27 1.0
76 16 SR
0.0 0.0
-12 -10 -8 -6 -4 -2 0 2 4 -4 -3 -2 -1 0 1 2

Total strain (e) , % Total strain (e) , %

Figure 3.31. Stiffness for an unbleached kraft and an unbleached CTMP pulp, both unbeaten and beaten to differ-
ent degrees versus total strain accumulated during drying.

Chemical addition. A dry strength additive with cationic charge functionality (Redibond
4300) was added to the 27 °SR kraft pulp in Figure 3.31. The stiffness was not affected by the
dry strength addition. Figure 3.32 shows the effect on tensile index versus total strain. The free
shrinkage strain did not change as a result of the chemical addition, but the tensile index in-
creased both for the restrained and free dried paper.

100
0.6%
Tensile index, kNm/kg

90

0.3%
80

Ref
70

60

50
-8 -6 -4 -2 0 2 4
Total strain (e) , %

Figure 3.32. Tensile index for an unbleached kraft pulp (27 °SR) with an addition of a dry strength additive ver-
sus total strain accumulated during drying.

Fillers. A large particle size rhombohedral PCC filler was added to the 27 °SR unbleached kraft
pulp used in Figure 3.31 and Figure 3.32. Both the tensile stiffness and the free shrinkage are
decreasing with increased filler amount (Figure 3.33). This is most likely due to that the interac-
tions between the fibres are disturbed by the filler particles.
92
Tensile stiffness index, MNm/kg

9.0
8.0
Ref
7.0
6.0
5.0 12%
4.0
3.0 14%
2.0
37%
1.0
0.0
-8 -6 -4 -2 0 2

Total strain ( e) , %

Figure 3.33. Stiffness for an unbleached kraft pulp (27 °SR) with different degrees of fillers added versus total
strain accumulated during drying.

PKQKQ==cçêãáåÖ=~åÇ=aêóáåÖ

Fibre orientation and anisotropy. Effects at the headbox slice and forming section principally
control fibre orientation. As the stock suspension is accelerated toward the slice, shear stresses
in the flow cause the fibres to orient preferentially in the MD. When the jet impinges on the
wire, a difference in the velocity between the fibres and the wire also creates shear stresses that
cause an alignment. This effect is at a minimum when the jet and wire velocities are the same
and becomes more significant as the velocity differential increases in either direction.
The amount of fibre orientation can be characterised by the anisotropy of the paper. An
anisotropy is often defined as the ratio of some property in the two principal in-plane directions
of paper, the machine or manufacturing direction, MD, and the cross machine direction, CD.
Equation 3.14 defines the fibre anisotropy, or fibre orientation, where n is the relative number of
fibres in the respective in-plane direction. The stiffness anisotropy for restrained dried paper is
defined in Equation 3.15 and for freely dried in Equation 3.16. The anisotropy in free shrinkage
strain is given in Equation 3.17. The anisotropy measured on machine made papers, as MD/CD
for some property, will of course depend both on the fibre orientation and the shrinkage in CD
and stretch in MD. This will be discussed in Section 3.5.

nMD
AFiber (3.14)
nCD

r
EMD
AE r r (3.15)
ECD

fs
EMD
AE fs fs (3.16)
ECD
 93

H MD
fs
AH fs (3.17)
H CD
fs

Free shrinkage strain – anisotropic. Figure 3.24 showed the development of the free
shrinkage strain during drying for an isotropic hand sheet. In Figure 3.34, the same relation is
given for anisotropic hand sheets made of the same pulp. It shows that the exponential constant,
k in Equation 3.5, has close to the same value regardless of anisotropy. Only the free shrinkage
potential, Hfs(mr = 0), varies. Consequently the development of the free shrinkage strain for aniso-
tropic papers can be calculated with Equation 3.17 and Equation 3.23.
Paper properties – anisotropic. Figure 3.27 showed the development of tensile stiffness in-
dex during free and restrained drying for an isotropic hand sheet. In Figure 3.35, the same type
of relation is given for anisotropic hand sheets made of the same pulp. It shows that, as for free
shrinkage strain, the exponential constant, k in Equation 3.8, has the same value regardless of
anisotropy. The stiffness for a dry paper, Emr = 0, and the constant c varies according to the geo-
metric mean assumption. Based on Figure 3.27 the relations in Figure 3.35 are given by Equa-
tion 3.15 and Equation 3.21 for restrained drying and Equation 3.16 and Equation 3.22 for free
drying.

0.0
Free shrinkage strain ( e ), %
fs

-2.0

-4.0
r
A (E )
-6.0 x 1.0
1.8
-8.0 3.3

-10.0
0.0 0.5 1.0 1.5 2.0
Moisture ratio (mr ), kg/kg

Figure 3.34. Development of the free shrinkage strain during drying.

Tensile stiffness index, MNm/kg
Tensile stiffness index, MNm/kg

12.0 12.0
Restrained Free
10.0 10.0

8.0 8.0
MD
6.0 6.0
MD
4.0 4.0

2.0 2.0
CD
0.0 0.0 CD
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Moisture ratio, kg/kg Moisture ratio, kg/kg

Figure 3.35. Development of tensile stiffness index during free and restrained drying for anisotropic hand sheets.
94

Geometric mean assumption. The fibre anisotropy is controlling the other defined anisotro-
pies according to Equation 3.18–3.20. It is common practice to use the stiffness anisotropy for
restrained dried paper as the fibre anisotropy according to Equation 3.18. The reason for this is
that the fibre orientation is time consuming and not so straight forward to measure. The fibre
anisotropy in machine made papers can have a value from close to 1 for certain board grades up
to around 4 in newsprint made on high speed machines.

AE r AFiber (3.18)

AE fs 2 AFiber  1 (3.19)

1
AH fs (3.20)
AFiber

Schrier and Verseput (1967) found that the geometric mean of MD and CD taber stiffness
was independent of changes in strength orientation. de Ruvo et al (1976) showed that the geo-
metric mean of the elastic modulus in MD and CD of an oriented sheet coincided with the elstic
modulus of an isotropic hand sheet made of the same raw material and pressed to the same den-
sity. The same relation was also true for the hygroexpansion. Htun and Fellers (1982) separated
the effect of drying restraint and fibre orientation. They showed that, for a given drying condi-
tion, the geometric mean value of MD and CD properties is possible to use as an invariant mea-
sure of mechanical properties of paper. However if the value is to be regarded as an isotropic
quality, the drying conditions must be the same in all sheet directions. Consequently the isotro-
pic value can be assumed to be equal to the geometric mean of MD and CD according to Equa-
tion 3.21 and Equation 3.22 and also constant for different fibre anisotropies. It has lately been
shown that the geometric mean assumption is valid also for the free shrinkage strain according
to Equation 3.23.

r
EIso r
EMD ˜ ECD
r
(3.21)

fs
EIso fs
EMD ˜ ECD
fs (3.22)

H Iso
fs
H MD
fs
˜ H CD
fs (3.23)

In Figure 3.36 –3.38, the validity of the geometric mean assumption is shown. Measure-
ments on isotropic hand sheets (same as in Figure 3.28) were used as input to the simulations
(solid lines). The restrained dried stiffness, Er, was 5,5 MNm/kg, the free dried stiffness, Efs,
2,0 MNm/kg and the free shrinkage strain, H fs, –6,0 %. The fibre anisotropy was varied between
1 and 4.
Figure 3.36 shows that the anisotropic behaviour of the tensile stiffness index for restrained
dried sheets can be simulated based on the isotropic hand sheet data. The solid line represents
the MD and CD tensile stiffness index, according to Equation 3.15 and Equation 3.21 with the
anisotropy varied from 1 to 4. The x-mark represents the isotropic input and the open and closed
 95

circles the experimental results in MD and CD for anisotropic hand sheets made of the same
pulp as the isotropic hand sheet.
Figure 3.37 shows that the anisotropic behaviour of the tensile stiffness index for free dried
sheets can be simulated based on the isotropic hand sheet. The solid line represents the MD and
CD tensile stiffness index, according to Equation 3.16, Equation 3.19 and Equation 3.22 with
the anisotropy varied from 1 to 4. The anisotropy is the stiffness anisotropy for restrained dried
sheets according to Equation 3.15. The x-mark represents the isotropic input and the open and
closed circles the experimental results in MD and CD for anisotropic hand sheets made of the
same pulp as the isotropic hand sheet.

12.0
Tensile stiffness index (E ),

Bleached sulphate MD
r

10.0

8.0
MNm/kg

6.0

4.0 CD

2.0

0.0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
Anisotropy (A Fiber ), -

Figure 3.36. Experimental and simulated stiffness relations between restrained dried isotropic and anisotropic
hand sheets.

12.0
Tensile stiffness index (E ),

Bleached sulphate
fs

10.0

8.0
MNm/kg

6.0
MD
4.0

2.0
CD
0.0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
Anisotropy (A Fiber ), -

Figure 3.37. Experimental and simulated stiffness relations between free dried isotropic and anisotropic hand
sheets.

Figure 3.38 shows that the anisotropic behaviour of the free shrinkage strain can be simulated
based on the isotropic hand sheet data. The solid line represents the MD and CD free shrinkage
strain, according to Equation 3.17 and Equation 3.23 with the anisotropy varied from 1 to 4. The
anisotropy is the stiffness anisotropy for restrained dried sheets according to Equation 3.15. The
 96

x-mark represents the isotropic input and the open and closed circles the experimental results in
MD and CD for anisotropic hand sheets made of the same pulp as the isotropic hand sheet.
Predicting anisotropic behaviour from isotropic hand-sheets. With isotropic hand-sheet
data and Equations 3.9–3.23, the anisotropic behaviour for an arbitrary total strain accumulated
during drying and anisotropy can be simulated based on isotropic hand sheet data. This means
that the relations between stiffness and total strain accumulated during drying presented in many
places in this chapter can be extended to the anisotropic case. Equation 3.9 was shown in Figure
3.28 to be valid for isotropic hand sheets. Wahlström and Fellers (1999) and Mäkelä (2003) has
shown a linear relation between total strain accumulated during drying and tensile stiffness index
also for anisotropic paper, which makes Equation 3.9 valid also for MD and CD data.

0.0
Free shrinkage strain ( e ), %

Bleached sulphate
fs

-2.0
MD
-4.0

-6.0

-8.0
CD
-10.0

-12.0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
Anisotropy ( A Fiber ), -

Figure 3.38. Experimental and simulated free shrinkage strain relations for isotropic and anisotropic hand sheets.

Figure 3.39 shows the simulated relations, using Equations 3.9–3.23, between tensile stiff-
ness index in MD and CD versus total strain in the respective direction for the anisotropy 1, 2, 3
and 4. Input to the simulation was the isotropic data for the bleached sulphate hand- sheets that
also was used as input to Figure 3.28. The restrained dried stiffness was 5.5 MNm/kg, the free
Tensile stiffness index, MNm/kg

14.0
MD, A=4
12.0 MD, A=3
10.0 MD, A=2
8.0
Iso, A=1
6.0
CD, A=2
4.0 CD, A=3
2.0 CD, A=4

0.0
-12 -10 -8 -6 -4 -2 0 2
Total strain, %

Figure 3.39. Simulated relations between tensile stiffness index in MD and CD and total strain in MD or CD for the
anisotropy 1, 2, 3 and 4.
 97

dried stiffness 2.0 MNm/kg and the free shrinkage strain –6.0 %, this data is also x-marked in
Figure 3.39. The same result could also be retrieved from the relations in Figures 3.36–3.38 by
taking the stiffnesses and free shrinkage for the chosen anisotropies.

PKQKR==mêÉëëáåÖ=~åÇ=aêóáåÖ

Figure 3.40 shows a comparison in stiffness between hand sheets made by an unbleached kraft
pulp (27 °SR) and an unbleached CTMP pulp (250 CSF). The two pulps were pressed in a labo-
ratory platen press to different densities, as indicated in the figure, and dried free and restrained.
An interesting difference between the behaviour of the kraft and CTMP made paper is that the
free shrinkage of the kraft paper is unchanged, but increases for the CTMP paper. Probably the
bondings between the fibres were favoured by the densification of the wet CTMP paper and the
shrinkage of the fibres could better be transferred to the fibre network. The wet kraft paper al-
ready had bonding good enough for transfer of the fibre shrinkage, but still the stiffness was im-
proved as expected. Note that the kraft pulp is not the same as in Figure 3.31 although similar,
but the CTMP pulp is (250 CSF).
Tensile stiffness index, MNm/kg

Tensile stiffness index, MNm/kg

10.0 6.0
Kraft CTMP
8.0 5.0

4.0
6.0
900 3.0 515
355
4.0 800
2.0 3
600 kg/m
3 215 kg/m
2.0 1.0

0.0 0.0
-8 -6 -4 -2 0 2 -3 -2 -1 0 1

Total strain (e), % Total strain (e) , %

Figure 3.40. Stiffness for an unbleached kraft and an unbleached CTMP pulp, pressed to different density, versus
total strain accumulated during drying.

PKR==j~ÅÜáåÉ=iÉîÉä=Ó=jçîáåÖ=táÇÉ=tÉÄ

PKRKN==`óäáåÇÉê=aêóáåÖ

Drying phases. The drying of paper is a dynamic process. Repeatedly and with very short time
cycles, heat is transferred from the drying cylinder and water is evaporated in the free draw. The
dynamic nature and its importance for the paper properties explain the difficulties in making use
of data from static laboratory trials. The key to advancing the knowledge about drying, the last
unexploited unit process in papermaking, is to study separately the sub-processes, the different
drying phases.
98

A single felted, or single tier, dryer section with vacuum rolls, Valmet VacRolls, is here used
as an example, to investigate the influence of the separate drying phases on paper properties. Fig-
ure 3.41 shows one cycle of the dryer section divided into the different drying phases. From A to
B, the paper is located on the dryer cylinder under the dryer fabric. The evaporation has been
simulated with a physical model for heat and mass transfer developed by Wilhelmsson (1995).

DRYER
CYLINDER
A B E

Fabric C D
Paper VAC
ROLL

Figure 3.41. One cycle of a single felted dryer section divided into its different drying phases.

Figure 3.42 shows a simulation of how the temperature develops over the drying phases in
one cycle. On the dryer cylinder, the temperature increases, particularly on the side of the paper
towards the dryer cylinder. The temperature becomes high enough for evaporation to start, but
the evaporation is reduced by the dryer fabric. Figure 3.43 shows a simulation of the local evap-
oration rate, based on paper area, in one cycle. Evaporation is low but not negligible; about
20 % of the total evaporation occurs on the dryer cylinder. From B to C and from D to E the pa-
per runs between the cylinders in the free draw. When the paper reaches the free draw the ener-
gy built up over the dryer cylinder leaves the paper as vapour. Initially the evaporation rate is
very high but it decreases rapidly as the paper temperature decreases. About 40 % of the evapo-
ration can take place in this phase. The remaining part of the evaporation takes place from C to
D on the VacRoll. The dryer fabric is located between the paper and the VacRoll and does not
therefore reduce the evaporation as it does on the dryer cylinder.

110
Middle
Paper Temperature, ¡C

100
Dryer side
Fabric side

90

80

70
A
28 B C29 D E
Position

Figure 3.42. Simulated temperature in a paper passing through a drying cycle as in Figure 3.41.
 99

140
Evaporation rate, kg/m2h

120

100

80

60

40

20

0
A
28 B C
29 D E
Position

Figure 3.43. Simulated evaporation for a paper passing through a drying cycle as in Figure 3.41.

Phase 1; on the dryer cylinder. The first drying phase considered is when the paper is on
the dryer cylinder and under the dryer fabric. If the paper is exposed only to this phase during
drying, the relevance of this phase for the paper properties can be investigated. Figure 3.44 and
Figure 3.45 shows results from a trial where 100 and 300 g/m2 testliners were dried over a
1.83 m dryer cylinder in a single pass. The results for 0 kN/m were measured separately with
the drying strategy free shrinkage. The pressure produced on the paper by the dryer fabric in-
creases the heat transfer from cylinder to paper, but it also restrains the shrinkage of the paper.
When the fabric tension is increased, the shrinkage is reduced and the final stiffness of the liner
increases. Compared to free shrinkage, the impact on shrinkage and stiffness is high, but in the
region of fabric tensions used on a production machine, 2–3 kN/m, the changes in shrinkage and
stiffness were small. For grades that require high bulk, it is not possible to use a high fabric ten-
sion because of the densifying effect. Too low a fabric tension leads to a deterioration in surface
properties and reduces the heat transfer.

4.0
Shrinkage, %

3.0

2.0 2
300 g/m
2
1.0 100 g/m

0.0
0 2 4 6 8 10 12 14 16
Fabric tension, kN/m

Figure 3.44. The shrinkage decreases with increasing fabric tension.

100
Tensile stiffness index, MNm/kg

6.0
2
5.0 100 g/m

4.0
2
300 g/m
3.0

2.0

1.0

0.0
0 2 4 6 8 10 12 14 16
Fabric tension, kN/m

Figure 3.45. The stiffness increases with increasing fabric tension.

Phase 2; in the free draw. Hitherto, no attempts have been made to increase the understand-
ing of the physics behind the shrinkage profile. With the finite element method, it has been pos-
sible to simulate stresses and strains in a paper web passing through a dryer section. The
simulations have been aimed at investigating how the conditions in the free draw influence the
shrinkage profile. The model was developed for a general orthotropic behaviour of the paper
and the kinds of relations presented in Figure 3.24, Figure 3.27, Figure 3.28 were used as input.
The total strain in the paper was assumed to be the sum of a mechanical strain and a hygroscop-
ic strain due to water evaporation. A mechanical strain can for example be due to a speed in-
crease in MD between two dryer groups and the resulting contraction in CD. The hygroscopic
strain is identical to the free shrinkage. It was assumed that when the paper web is on the cylin-
ders it is restrained from shrinkage by the pressure from the dryer fabric. In the free draw be-
tween the dryer cylinders there are no forces acting on the edges of the web.
To validate the model, the shrinkage profile was measured on two paper machines and their
respective free draw geometries simulated. Figure 3.46 shows the measured shrinkage profiles.
The dryer section with the double-felted configuration was 6.5 m wide and had a free draw

CD Shrinkage (Measured), %
5.0

4.5

4.0

3.5 Double felted

3.0

2.5

2.0

1.5

1.0
Single felted
0.5

0.0
0 1 2 3 4 5 6 7 8 9
Web width, m

Figure 3.46. Measured shrinkage profiles for a single felted and a double felted dryer section.
 101

length of 2.3 m. The free draw in double felting was defined as the length where the web is not
in contact with the dryer fabric. In the single-felted case, the free draw was 0.9 m and the web
width 9.5 m. Figure 3.47 shows the results of the simulations of these dryer sections. There are
some deviations in the absolute level of shrinkage and too low a prediction in the middle of the
web. The model was, however, able to capture the general behaviour of the shrinkage and it
gives a qualitative understanding of the influence of the studied variables.

CD shrinkage (- e22), %
5.0
4.5
4.0
double felted
3.5
3.0
2.5
2.0 single felted
1.5
1.0
0.5
0.0
0 1 2 3 4 5 6 7 8 9
Web width, m

Figure 3.47. Simulated shrinkage profiles for a single felted and a double felted dryer section.

Figure 3.48 shows the results of a series of simulations where the length of the free draw was
varied. When the length of the free draw was reduced, the shrinkage in the middle of the paper
web decreased significantly. At the edges, the shrinkage also decreased but to a smaller extent.
The region in the middle of the web with low shrinkage also became wider when the free draw
was shortened, which means that the shrinkage gradient at the edges of the web became steeper.
A shorter free draw reduced the total web width shrinkage, but the difference in shrinkage be-
tween the edge and the middle increased. This also means that the difference in most paper
properties between the edge and the middle would have increased with a reduction in the length
of the free draw.
The origin of the shrinkage profile can be explained by a difference in stresses. In CD, the
stress became much higher in the middle of the web than at the edge where it was close to zero.
The explanation of the higher stress was the boundary conditions, free edges in the free draw
and a fixed situation when the web was under the fabric or on the VacRoll. The high stress in
the middle of the web restrains the shrinkage more in the middle than at the edge. Figure 3.49
shows the calculated CD stress fields in the final free draw for the simulated cases in Figure
3.19. The shortest free draw had the highest CD stresses and consequently also the lowest
shrinkage.
102

CD Shrinkage (-e22), %
5.0

4.5
Free Draw Length
2.30 m
4.0
1.15 m
3.5 0.58 m
3.0

2.5

2.0

1.5

1.0

0.5

0.0
0 1 2 3 4 5 6
Web width, m

Figure 3.48. Simulated shrinkage profiles for three different lengths of the free draw.
0.58

CD stress (N/m2)
1.15
free draw length (m)
2.30

0 1 2 3 4 5 6
web width (m)

Figure 3.49. Simulated CD stresses in the final free draw for three different free draw lengths.

Phase 3; on the VacRoll. On the VacRoll, the web is pressed against the fabric by the vacu-
um inside the roll. This restrains the shrinkage but, as will be shown, not as effectively as the re-
straining effect of the dryer fabric over the dryer cylinder. To increase the understanding of this
drying phase, a pilot trial was carried out. A pilot-machine-made 200 g/m2 testliner was dried
over a VacRoll in a single pass by blowing hot air onto the paper.
Figure 3.50 shows how the vacuum influences the development of the shrinkage during dry-
ing. The total web width shrinkage is highest without any vacuum in the VacRoll. The shrink-
age starts at the same moisture ratio, around 0.9 kg/kg, in both cases. In the trial with vacuum,
the paper was dried in a more restrained state and the final level of shrinkage was lower. The
free shrinkage for this grade was 3.8 %. Since the shrinkage was about 2.6 % when the vacuum
was zero, some restraint was obviously imposed on the paper even in this case.
If the restraint is caused by stress gradients, it can be suspected that this drying phase also
gives rise to a shrinkage profile. Figure 3.51 shows the shrinkage profiles measured on the dry
200 g/m2 liner for three different levels of vacuum, 0, 1.1 and 4.8 kPa. With the highest vacu-
um, the shrinkage in the middle of the web is close to zero but it is as high as 2.6 % at the edge.
 103

The profile for the paper dried without vacuum shows a more even profile than the high vacuum
case but the total web width shrinkage in higher. The difference in shrinkage between the mid-
dle and edge of the web is greatest for the trial with the highest vacuum. As is indicated by lab-
oratory data in this paper, this also means that the difference in paper properties between edge
and middle was largest in this case.

3.0

2.5
Shrinkage, %

2.0
0 kPa
1.5

1.0

0.5 4.8 kPa

0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Moisture ratio, kg/kg

Figure 3.50. Development of shrinkage during drying with and without vacuum.

3.5

3.0
0.0 kPa
Shrinkage, %

2.5

2.0

1.5
1.1 kPa
1.0

0.5 4.8 kPa

0.0
0 100 200 300 400 500
Web width, mm

Figure 3.51. Shrinkage profiles for different levels of vacuum in the VacRoll.

PKRKO==aêóáåÖ=`çåÅÉéíë

Conventional double-felted drying section. The effect of cylinder drying on the mechanical
properties of paper is very dependent on the shrinkage or stretch during the drying process. Fig-
ure 3.52 shows a conventional double felted dryer section. Figure 3.46 and Figure 3.47 show
the results from both a measurement and a simulation of the shrinkage in the cross machine di-
rection of another double-felted dryer section. The free draw length in this case was 2.3 m and
the web width 6.5 m. This and many other double-felted dryer sections in use have relatively
long free draws, compared to new installations and in relation to the width of the machine.
104

The shrinkage in the cross direction of the paper machine is greater at the edges than at the
middle of the web. This difference also creates a great difference in mechanical properties, as
shown in the laboratory data presented in this paper. The simulations shows that the shrinkage
profile is created by the situation with free edges of the web in the free draw and the restrained
conditions under the dryer fabric.
Single-felted dryer section. In a single-felted dryer section, all the rolls at the top are steam-
heated cylinders and all the rolls at the bottom are vacuum rolls (VacRolls), Figure 3.53. The
sheet run is totally closed and supported through the entire dryer section for improved runnabil-
ity. The total web width shrinkage is less than that in a conventional dryer section.
Historically, the speed and the width of the paper machine has increased to increase produc-
tion, and the free draws have become shorter to improve runnability. Figure 3.46 and Figure
3.47 show how the shrinkage profile changes when the web width increases and the length of
the free draw decreases. In the middle of the web, the shrinkage decreases significantly. Conse-
quently a wider machine and shorter free draws reduce the total web width shrinkage and there-
by increases the mean stiffness. The difference in shrinkage between the edges and the middle
of the web increases however, which means that the differences in most paper properties be-
tween the edges and the middle also increase. The results presented in Figure 3.50 show that a
part of the shrinkage profile is also an effect of the conditions on the VacRolls.

Figure 3.52. Conventional double-felted dryer section.

Figure 3.53. Single-felted dryer section.

Impingement drying. The most recently developed and commercialised drying process is
the OptiDry impingement drying concept shown in Figure 3.54. The special features of the con-
cept are the totally closed sheet transfer beginning in the press section, the use of two different
forms of energy for heating, and a new high efficiency paper drying unit. The benefits are im-
 105

proved runnability, fast grade changes and a shorter dryer section. The concept consists typical-
ly of three air impingement modules added to a SymRun-type dryer section. On the
impingement modules or OptiDry units the web is on top of the dryer fabric and is vacuum-sup-
ported by a large diameter VacRoll. The effect on paper properties does not differ significantly
from that of the SymRun dryer section.

Figure 3.54. OptiDry impingement drying concept.

Condebelt drying. There are currently two board mills utilising the Condebelt process, the
Pankakoski mill of Stora Enso in Finland, shown in Figure 3.55, and the Ansan Mill of Dong Il
in South-Korea. In the Condebelt process, the wet web is carried on two permeable wires and
fed between two long smooth steel belts. The web is dried by being under pressure (0.5 to 5 bar)
and in contact with the upper steam-heated (111–159 °C) belt. As the moisture in the web evap-
orates, the vapour gene-rated passes through the wires and condenses on the lower cooler steel
belt (Retulainen 1999).

Figure 3.55. Condebelt

The z-directional pressure and the contact with the glossy, hot metal belt has two major con-
sequences with regard to sheet properties. First, the pressure applied to the hot and moist web
acts to plasticize fibres, improve bonding, increase density and create a smooth surface. Second,
the pressure prevents all shrinkage during drying, which has earlier been shown to have a large
effect on the paper properties in CD. The totally prevented shrinkage in the Condebelt unit, of
course, also eliminates the shrinkage profile. Figure 3.29 showed that a reduction in the shrink-
age leads to a reduction in z-toughness. In the Condebelt process, this drawback is more than
compensated for by the increase in density of the sheet.
106

Air-borne drying. In an Airborne dryer, the paper is dried by hot air blown towards the pa-
per web by nozzles. Figure 3.56 shows an Airborne dryer. The very long free draws extend
from the left-hand to the right-hand side of the dryer. On each side, the web is turned on turning
rolls. In the free draws, the air blown by the nozzles creates an air cushion that carries the web.
Figure 3.48 shows how the length of the free draw influenced the shrinkage profile in the pa-
per. These results can be extrapolated towards very long free draws, which give a higher total
shrinkage and a more even shrinkage profile. Only a small force is needed to pull the web
through the dryer, and this gives minimum web tension. The very low web tension also allows
for shrinkage in the machine direction. In a conventional cylinder dryer section, a small stretch
is always necessary between the dryer groups to give sufficient web tension for good runnabili-
ty.
Figure 3.28 showed that the tensile stiffness decreases when the shrinkage increases. This
means that the strain at break, that is the extent to which it is possible to stretch the paper before
it breaks, increases. This makes the airborne dryer suitable for the production of sack paper that
requires a high strain at break. Most installations of this type of dryer section are however made
for pulp drying.

Figure 3.56. Airborne dryer.

Air Impingement Drying for Sack Paper. The increasing restraint in the dryer section due
to shorter free draws and wider machines makes it harder to produce sack paper of good quality
in a modern conventional dryer section. Sack paper requires high shrinkage, which is beneficial
for producing a stretchable material. There are ways to increase the shrinkage in a conventional
cylinder dryers section. Figure 3.57 shows an installation of impingement drying hoods in a
conventional dryer.
Hot air is blown onto the web when it passes over a dryer cylinder. The web is not restrained
from shrinkage by the fabric on these particular dryer cylinders, and this explains the increased
shrinkage. The length of the free draw can then be considered as the length where the web is not
restrained by a dryer fabric. Since the evaporation capacity is high in impingement drying, a
large part of the drying takes place in this long „free draw“.
 107

Figure 3.57. Air impingement drying for sack paper.

PKS==oÉÑÉêÉåÅÉë

Htun, M., and C. Fellers. (1982) The invariant mechanical properties of oriented hand sheets.
Tappi Journal, 65 (4): 113–117.
Maloney, T. C. (2000) How the Structure of the Cell Wall Influences Press and Dryer Dewater-
ing. Proceedings of the 12th International Drying Symposium IDS2000, Elsevier Science,
Amsterdam, Paper No.296.
Nanko, H., and J. Wu (1995). Mechanisms of paper shrinkage during drying. International
Paper Physics Conference, Niagara-on-the-Lake, Canada, pp. 103–113.
Wahlström, T., and C. Fellers. (1999) Biaxial straining of paper during drying, relations
between stresses, strains and properties. TAPPI Engineering Conference, September 12–
16, 1999, Anaheim, CA, pp. 705–720.
Wahlström, T., K. Adolfsson, S. Östlund, and C. Fellers (1999) Numerical modelling of the
cross direction shrinkage profile in a drying section, a first approach. 1999 International
Paper Physics Conference, September 26–30, San Diego, CA, pp. 517–531.
Wahlström, T., and C. Fellers (2000) Biaxial straining of hand sheets during drying- effect on
in-plane mechanical properties. Tappi Journal, 83 (8): 91.
Weise, U., T. Maloney and H. Paulapuro (1996) Quantification of water in different states of
interaction with wood pulp fibres. Cellulose 3(4): 189–202.
100
 109

Q==qÜÉ=fåíÉê~Åíáçå=çÑ=m~éÉê=ïáíÜ=t~íÉê=s~éçìê

Christer Fellers
STFI-Packforsk AB, Stockholm

4.1 Introduction 109

4.2 The Moisture Content of the Air 113

4.3 Climatology 114

4.3.1 Moisture Influences on Papermaking and Converting 115

4.4 Sorption Isotherms and Hysteresis 116

4.4.1 Desorption 119

4.5 Hygroexpansion 120

4.6 Permanent Dimensional Changes during Moisture Cycling 128

4.7 Measurement Methods 130

4.8 Curl and Twist 133

4.9 The Mechanical Properties at Different Moisture Contents 141

4.10 Mechanosorptive Creep 143

QKN==fåíêçÇìÅíáçå

The equilibrium moisture content of paper changes if the moisture content of the surrounding
air changes and as a result the dimensions of the paper changes. Moisture can also be added in
printing processes or other end-use situations. These changes can lead to great practical prob-
lems. For example, it cannot be tolerated that the paper changes its dimensions too much during
the printing of the paper. A so-called register error occurs in multi-colour printing, where the
colours are printed one at a time. Small errors can be adjusted during printing.
Another serious error is when a paper moves so that it deviates from the plane state. The pa-
per exhibits curl and twist. Curl and twist are often the reason for failure in the sheet feeding of
paper and carton board in printing presses and in the filling of packages in packing machines.
When adding moisture to paper, the result may also be thickness swelling, surface changes,
and change in paper stiffness and strength.
110

Figure 4.1. When adsorbing moisture, paper deform in different ways.

Figure 4.2. When adsorbing moisture, paper may experience thickness swelling and surface changes. Effect of
wetting on calendered paper that contains mechanical pulp. Cross-sections of SC magazine paper (super calen-
dered). Wetting by water 30 min and drying at room temperature.

When water vapour access to a paper-air system, the vapour diffuses into the structure. This
process is called sorption. Throughout the papermaking process, the interaction between the pa-
per fibres and water plays a major role. The amount of water vapour sorbed are expressed as
moisture ratio, moisture content or dry solids content.
• Fibre concentration C
• Dry solids content DC
• Moisture ratio MR
• Moisture content MC
The concept of fibre concentration is used to indicate the amount of fibre in a suspension. In
the paper machine, from the dry line in the wire section and forwards in the process, the quanti-
ty is usually called dry solids content (DSC).

kg dry solids
C (%) DSC (%) ˜100 (4.1)
kg dry solids  kg water
 111

The amount of moisture in the paper during its passage through the paper machine is often
given as the moisture ratio (MR), the relationship between the amount of water and the amount
of dry substance. The moisture ratio is a linear measure of the water content in the sheet.

kg water
MR (4.2)
kg dry solids

Moisture content (MC) is defined as the amount of water in relation to the amount of dry
substance plus water. The concept of moisture content is appropriate to a finished paper.

kg water
MC (%) ˜100 (4.3)
kg dry solids  kg water

The relations between the different terms are given below.

DSC (%) 100  MC (%) (4.4)

MC (%)
MR (4.5)
100  MC (%)

MR
MC (%) ˜ 100 (4.6)
MR  1

100  DSC (%)

MR (%) (4.7)
DSC (%)

1
DSC (%) ˜100 (4.8)
MR  1

In general terms, diffusion is the tendency for a concentration gradient to be evened out by
molecular movements. The rate of diffusion and the time to attain equilibrium depend on the
initial moisture content of the paper, on the relative humidity of the air, on the thickness of the
material and on the diffusion coefficient. Figure 4.3 shows how fast different papers adsorb
moisture.
If the relative humidity is cycled in a regular fashion as in Figure 4.4a, the corresponding
moisture content versus RH-curve shows hysteresis (Figure 4.4b). This is a partly a conse-
quence of the finite time for sorption to reach equilibrium.
In an ideal system, the properties of the mixtures are proportional to the concentrations; no
heat release occurs and the processes are reversible. None of this is true for water vapour and fi-
bres.
112

3.5

newsprint
3.0

wrapping paper (kraft)

2.5
bond paper
moisture content (%)

2.0
filter paper

1.5 blotting paper

1.0

0.5

0
0 5 10 15 20 25 30
time (hours)

Figure 4.3. Rate of moisture adsorption from 35 % RH to 65 % RH.

90

80
moisture content (%)
relative humidity (%)

70

60

50

40

30
0 2 4 6 8 10 30 50 70 90
a) time (hours) b) relative humidity (%)

Figure 4.4. If the relative humidity is cycled in a regular fashion as in a), the corresponding moisture content ver-
sus RH-curve shows hysteresis as in b).

When moisture diffuses into dry paper fibres from wood, three things mainly happen:
• heat is released
• the weight increases
• the fibres swell
In this chapter, we shall describe more closely the phenomena that occur when paper fibres
sorb water from air at different relative humidity.
 113

QKO===qÜÉ=jçáëíìêÉ=`çåíÉåí=çÑ=íÜÉ=^áê

The properties of paper depend to a large degree on the amount of moisture of the paper. The
thermodynamic quantity determining the moisture content of paper is the chemical activity of
the vapour, which is equal to the relative humidity of the air. The concept of Relative Humidity,
usually abbreviated as RH is explained below.
Changes in moist air are often calculated with the help of a Mollier-diagram. A Mollier dia-
gram has an angular coordinate system with axes representing enthalpy, h, and water content in
the form of water vapour.
In this chapter, calculations relating to the interaction of paper with water vapour, a simpli-
fied Mollier-diagram will be used, Figure 4.5. An additional scale below in the diagram indi-
cates the partial pressure of the vapour. Note that the temperature interval is chosen for room
conditions.

RH 20 %
40

35 40 %

30
60 %
25 80 %
35%
20 100 %

15
t (°C)

10

–5

0 5 10 15 20 25 30
vapour (g/kg dry air)
0 10 20 30 40

partial vapour pressure (mbar)

Figure 4.5. Mollier-diagram for moist air. (Ekroth, Granryd: Applied thermodynamics. Dep. for Energy Technol-
ogy, KTH).

Air usually contains a certain amount of water, in practice 0–30 g water per kg dry air. The
absolute amount of water at any temperature cannot exceed a maximum value, given by the sat-
uration curve, which gives the maximum vapour pressure at each temperature.
114

For each temperature and water content, the relative humidity RH is calculated strictly with
the help of Equation 4.9 as the partial vapour pressure divided by the maximum vapour pres-
sure.

partial vapour pressure

RH (4.9)
maximum vapour pressure

The RH can also be expressed approximately as the amount of water that the air contains in
relation to the maximum quantity that it could hold at the given temperature, according to the
saturation curve.

kg vapour / kg dry air

RH | (4.10)
kg vapour / kg saturated air

The following example shows the use of the Mollier-diagram:

Assume that air at 20 °C contains 5 g water per kg dry air. The partial pressure of the vapour
is then 8.1 mbar. From the diagram, the maximum is 14.8 g water per kg dry air at this tempera-
ture. The partial pressure of the vapour is then 23.37 mbar. The relative humidity, according to
Equation 4.11, is thus:

8,1
RH 34,6% (4.11)
23,37

The approximate Equation 4.10 gives:

5,0
RH | 33,8% (4.12)
14,8

If the temperature of the air containing 5 g water/kg dry air decreases to 4.1 °C, the limit for
the maximum possible atmospheric humidity is reached. If the temperature decreases further,
condensed water is formed. This point is called the dew point.

QKP==`äáã~íçäçÖó

The dimensions and physical properties of paper such as electric resistance, heat conductivity
and mechanical properties depend on the temperature and on the amount of moisture in the pa-
per. As paper is a hygroscopic material, its moisture content depends on the RH and the temper-
ature of the surrounding air according to theories of sorption isotherms and hysteresis, deeper
described in Section 4.4 below.
The RH varies greatly outdoors from place to place. Across the oceans, the RH is largely
constant, about 80 per cent, whereas the RH in desert areas seldom exceeds 40 %. Examples of
RH-variations during the year at different places are shown in Figure 4.6.
 115

100
Stockhom
Manaus, Amazon basin, 3°S, tropical rain forest 90
The Faroe Islands
80

relative humidity, RH (%)

70
Bangalore, India tropica monsoon climate
60
50
Aoulef, Sahara, desert climate
40
30
20
10
JFMAMJJASOND
month

Figure 4.6. The annual variation in the relative humidity in different climate areas. Note that the climate indoors
is completely unrelated to RH outdoors.

The relative atmospheric humidity also shows daily variations, due primarily to the tempera-
ture variations during the day. Note that RH indoors in each country is completely unrelated to
RH outdoors.

QKPKN==jçáëíìêÉ=fåÑäìÉåÅÉë=çå=m~éÉêã~âáåÖ=~åÇ=`çåîÉêíáåÖ

Paper is normally manufactured with moisture content within the range of 6–9 %. The moisture
content in paper is determined in practice by measuring the loss of weight of a paper when it is
dried to constant weight at a temperature of 105 °C, in accordance with ISO 287 international
standard. Usually papermakers want to supply paper with as high moisture content as possible
to save drying heat. Paper is also sold according to weight and therefore the producers get paid
even for the moisture in the paper.
However, it is easier to maintain an even dry content across the width of the paper machine if
the dry content is high. For that reason over-drying is sometimes employed and the paper is then
moistened again. As an example, over-drying is always done when producing qualities for su-
percalendering.
In printing houses, it is not desirable to have too high dryness in the paper, since the sheets
can easily become charged with static electricity and stick together. Nor is too high moisture
content in the paper beneficial, since the paper then becomes floppy and difficult to carry
through the printing machines.
Indoors in paper mills, in storerooms, in transport vehicles etc, the RH will vary a lot as a
consequence of changes in the humidity and temperature of the air. A printing house or a carton
board factory must in their premises have an RH and temperature which are constant and they
must work with paper that is in equilibrium with this atmosphere, therefore storage can be nec-
essary before converting operations so that equilibrium will be reached.
116

For testing and classification of paper there is vital to take the climate variable in consider-
ation. The standard conditions can vary from country to country. The practical problem, which
the laboratories sometimes meet, is how to make it easy to control the climate in the room. It is
expensive in warm countries to cool the air in the testing room. In countries with a warm and
humid climate, it can also be difficult to dry the air.

QKQ==pçêéíáçå=fëçíÜÉêãë=~åÇ=eóëíÉêÉëáë

The usual way of studying moisture absorption in a material is to plot moisture content against
RH at a given temperature, a sorption isotherm.
• The term adsorption and desorption designates water sorption on the surface of a material
when moisture enters and leaving the material respectively.
• The term absorption is used for a more homogeneous water sorption inside the material.
• The limit between these two definitions is not sharp.
• Sorption is used as a general concept.
To measure the true isotherm of a material, the change in RH should theoretically take place
in infinitely small steps and infinitely slowly over the whole interval of 0–100 % RH. This can-
not, of course, be done in practice, so that each isotherm is a compromise depending on the
choice of RH-interval and the time scale. It is therefore not immediately possible to compare
isotherms of different origin without careful consideration.
It should be pointed out that experimentally, it is difficult to keep the RH stable at a high RH,
so that sorption isotherms seldom include regions above 85–90 %.
When a porous material is exposed to water vapour, it can take a long time for the material to
reach equilibrium through sorption. Air mixing plays a great role in reducing this time. This is
probably the greatest reason why true isotherms are difficult to determine.
The diffusion of water into paper is an exothermal process, which means that heat is re-
leased. The differential heat of sorption is defined as the heat released when one gram of water
is combined with an already partly water-filled fibre. The differential heat of sorption decreases
with increasing moisture content.
For sorption of moisture to occur on cellulose fibres the free energy of sorption must be neg-
ative. The free energy is given by

'G 'H  T 'S (4.13)

where
H enthalpy (heat)
S entropy (order)
T absolute temperature.
The adsorption of moisture onto cellulose generally involves an ordering of the water mole-
cules when they adsorb. Hence 'S is positive and for adsorption to occur –'H must be larger
than T'S. This is indeed the case and sorption is therefore said to be enthalpy driven.
As mentioned earlier, the RH is a measure of the chemical activity of the water in the vapour
phase. It is therefore the relative humidity of the air that determines the equilibrium moisture
 117

content of the paper at a given temperature, not the absolute moisture content of the surrounding
air. This is shown with the adsorption isotherms for cotton in Figure 4.7.

20

15
moisture content (%)

50 °C
10
40 °C
30 °C
20 °C
5 10 °C

0
0 20 40 60 80 100
relative himidity, RH (%)

Figure 4.7. Adsorption isotherms at different temperatures for cotton based on relative humidity, RH, and based
on the absolute moisture content of the air.

Although the absolute moisture content of the air varies by a power of ten between 10 and
50 °C at the same RH, the moisture content of the paper varies very little.
As a consequence of the fact that the adsorption of moisture into the cellulose fibre is an exo-
thermal process, the cellulose will at a given RH have lower moisture content at a higher tem-
perature. The reason why the sorption decreases with increasing temperature is that sorption is
enthalpy driven. To maintain equilibrium it follows that if the temperature is higher the amount
of adsorbed moisture must be lower.

(dry paper + water œ paper in equilibrium + heat).

The curve shape in Figure 4.7 is typical for the sorption of moisture in cellulose materials. In
different parts of the curve, moisture sorption takes place by different mechanisms such as
mono-layer sorption, multi-layer sorption and capillary condensation. These phenomena are
treated below.
• At low RH, surface adsorption takes place as a mono-layer. The OH-groups of the cellulose
molecule and of the hemicellulose molecule are considered to bind water through hydrogen
bonds in a monomolecular layer on all available surfaces. As proof, it is usually stated that
the first adsorbed water releases a large amount of heat (21 kJ/mole) and that this heat cor-
responds to the bonding energy of hydrogen bridges.
118

• At a medium RH, surface adsorption takes place as multi-layers. The behaviour of cellulose
fibres with regard to sorption and swelling is similar to the behaviour of a gel. According to
the gel swelling theory, absorption takes place in disordered areas in the cellulose, in the
hemicellulose and in the lignin.
• At very high RH levels, moisture absorption takes place by capillary condensation, where
larger capillaries in the fibre and paper structure are filled with liquid. The average size of
the voids that are filled by water molecules can be described by the Kelvin Equation, Fig-
ure 4.8 below.

calculation based on the Kelvin equation at 25 °C

P 2 gM
ln = relative vapor pressure effective capillary
P0 r rRT
of water (p/p0 ) radius ( mm)

where g = surface tension of water (0.072 N/m) 0.99999 106

M = molecular mass of water (18 ·10–3 kg/mole) 0.9999 10.6
0.999 1.06
r = radius of capillary (m) 0.99 0.106
r = density of the water (1000 kg/m3) 0.98 0.052
0.95 0.020
0.50 0.0015

Figure 4.8. The Kelvin Equation.

It must be pointed out that thermodynamic observations such as sorption and desorption iso-
therms can never explain how water is bound to the fibres. For this, other experiments are re-
quired e.g. with marked molecules or through the use of different spectroscopic methods.
The moisture content obviously depends on the chemical composition of the fibres in the pa-
per, Figure 4.9.

20

chemical pulps
15 kraft and NSSC
based on moist sample (%)

mechanical pulp

10
moisture content

5 cotton

0
0 20 40 60 80 100
relative humidity (%)

Figure 4.9. Adorption isotherms at 23 °C for papers made from different pulps.
 119

QKQKN==aÉëçêéíáçå

So far, the sorption process has been discussed. We shall now also discuss the reverse process,
desorption. Figure 4.10 shows three moisture sorption isotherms for paper. Curve 1 shows the
desorption from a never-dried paper. Curve 2 shows the sorption of the paper up to 100 % RH.
In the subsequent desorption according to curve 3, the paper has lower moisture content than in
curve 1 at the same RH.

30

20
moisture content (%)

3
1
b 2
c a
10

0
0 20 40 60 80 100
relative humidity (%)

Figure 4.10. Typical moisture sorption isotherms for paper. The desorption curve (1) is for never-dried fibres, the
sorption curve (2) is for the first moistening and the desorption curve (3) is for once-dried fibres.

It has been shown that a desorption curve always lies higher than the adsorption curve. The
moisture absorption is thus not reversible. The process shows hysteresis.
The loop a-b-c-d shows approximately what happens if the adsorption stops at a, if the paper
is then dried to b and c and then adsorbed to d and c again.
The inclination of the curve also becomes flatter, and it can be seen that the material be-
comes less and less hygroscopic. The physicist says that the material has been hornified. The ef-
fect can be considerable in recycled paper where the fibres have been dried many times. Porous
fibreboard is made less sensitive to moisture and more dimensionally stable by being dried,
hornified at a high temperature and thereafter carefully moistened. In paper testing, it is an ad-
vantage if the paper sample is conditioned from the dry side to the desired RH. All samples will
then reach the given RH along the same isotherm.
The hysteresis effect occurs in other systems than cellulose-water, especially in those where
long, more or less mobile molecules interact with the swelling medium.
There may be several reasons for sorption hysteresis. One important aspect is that during
desorption the fibre gel is under tension and during adsorption the gel is under compression.
From thermodynamics, it can be shown that a gel under tension has a higher vapour sorption
and if the gel is under compression the sorption decreases.
120

QKR==eóÖêçÉñé~åëáçå

If the hygroexpansion is plotted against moisture ratio, moisture content or RH for a small
change in the climate, linear relationships are obtained. It is then useful to define a property
called hygroexpansion coefficient, which assumes that the increase in strain is linearly propor-
tional to the RH change or change in moisture content, Figure 4.11.

ZD

CD

MD

e = strain

e
b=
DH

DH = change in relative humidity

on moisture ratio

Figure 4.11. Definition of the hygroexpansion coefficient E

How can the moisture absorption affect the dimensions of a sheet? The reason is primarily
the fibre's own swelling and the manner in which this swelling influences the shrinkage of the
paper sheet on the paper machine. If the swelling of paper and wood during moistening is pre-
vented by external limitations, large swelling forces arise. The effect can be used to blast stone
through the use of wooden wedges that are moistened. When the paper shrinks, corresponding
shrinkage forces occur. We shall below discuss the reasons for these effects.
Different pulps contain different amount of polymer components, Table 4.1, and are expect-
ed to swell differently due to the chemical composition.

q~ÄäÉ=QKNK=oÉä~íáîÉ=ÅÜÉãáÅ~ä=Åçãéçëáíáçå=çÑ=ëçãÉ=éìäéëK=
oÉä~íáîÉ=Åçãéçëáíáçå
qóéÉ=çÑ=éìäé ÅÉääìäçëÉ ÜÉãáÅÉääìäçëÉ äáÖåáå
qjm QQKO OVKR OSKP
eáÖÜ=óáÉäÇ=âê~Ñí RVKR NVKN ONKQ
hê~Ñí TQKM NTKU UKO
_äÉ~ÅÜÉÇ=âê~Ñí TUKP ONKQ MKP
 121

Figure 4.12 shows a comparison of hygroexpansion between different papers. No immediate

conclusion if the chemical composition plays a role. It is however clear that drying conditions
are important. In addition, beating and refining conditions as well as the chemical composition
affect the shrinkage and consequently the fibre shape and the amount of fines.

hygroexpansion coefficient (% /mr)

0.14
restraind dried
0.12 freely dried

0.1

0.08

0.06

0.04

0.02

0
SGW TMP CTMP CMP LYS, LYS BKP,
low yield BEATEN bleached
sulphate kraft pulp,
BEATEN

Figure 4.12. A comparison of hygroexpansion between different papers.

Swollen fibres shrink during the drying. They shrink approximately one per cent in the
length direction while they shrink 20–30 per cent in their width and in the thickness direction.
When they shrink, the fibres will attract each other in their bonding regions. Figure 4.13 illus-
trates the changes in the sheet after restrained and free drying. If the paper is dried under re-
straint in a certain direction, fibre segments are stretched.
If the shrinkage takes place freely, the width shrinkage of the shrinking fibre contracts the
crossing lengthwise fibre and micro-compressions are created.
A schematic representation of the development of micro-compressions during drying is
shown in Figure 4.14.
122

a)

b)

Figure 4.13. a) The structure of a paper, dried under restraint, b) The structure of a paper, dried freely.
 123

paper dried
under restraint

paper dried
freely

hygroexpansion

Figure 4.14. The structure in the bonding zone between two crossing fibres after different drying conditions. The
hygroexpansion is higher for a freely dried paper than for a paper dried under restraint.

Let us assume that the dried sheet is exposed to a moisture increase so that the fibres swell.
In the sheet dried under restraint, the fibres are mainly straight with small number of micro-
compressions. The structure of the fibre now causes the swelling to take place in only a small
part of the length direction of the fibre and this gives the paper a small hygroexpansion. In the
freely dried sheet, the structure of the fibres is disturbed by many micro-compressions, and
swelling easily stretches the lengthwise, folded fibre.
The most important paper technical factor which influences the dimensional change is thus
the drying conditions. The more bonds formed and the greater the shrinkage during drying, the
more micro-compressions are formed and the greater are the dimensional changes during subse-
quent moistening.
A more detailed mechanism for hygroexpansion may be viewed as follows. When a fibre in a
paper adsorbs moisture, the thickness and width swells much more than the length. Any disloca-
tion in the fibre will change its angle to the fibre direction and the fibre will be longer, Figure
4.15.

dry wet

q q‘

q q‘
L L‘

a a‘

Figure 4.15. When a fibre in a paper adsorbs moisture, the thickness and width swells much more than the length.
Any dislocation in the fibre will change its angle to the fibre direction and the fibre will be longer.

Shrinkage during drying gives rise to dislocations and there is a close relation between the
hygroexpansion coefficient and the shrinkage during drying, Figure 4.16.
124

0.03
hygroexpansion coefficient (b RH)

0.02

0.01

free drying
different drying conditions

0
0 4 8 12 16
shrinkage during drying (%)

Figure 4.16. Shrinkage during drying gives rise to dislocations and there is a close relation between the hygroex-
pansion coefficient and the shrinkage during drying. The hygroexpansion coefficient E RH is defined as the strain
divided by the change in RH.

The hygroexpansion coefficient is dependent on the amount of fines and the density obtained
by wet pressing, Figure 4.17. The higher the density the greater chance for fibre-to-fibre bonds
and more micro-compressions for freely dried papers. For restrained dried papers a greater
bonding results in less hygroexpansion, probably due to more straightening of fibre segments.
More fines tend to increase the hygroexpansion.

0.016
B3
dried
freely
hygroexpansion coefficient (b RH)

0.012
B1

0.008
dried restrained

B3
B1
0.004
unbeaten long fibers, B1
with 5.2 % fines, B 3

0
200 400 600 800 1000
density (kg/m3)

Figure 4.17. The hygroexpansion coefficient is dependent on the amount of fines and the density obtained by wet
pressing.
 125

Also the fibre curl affects the hygroexpansion, Figure 4.18. For freely dried papers the hy-
groexpansion increases with curler fibres. For restrained dried papers the fibre shape does not
influence the expansion.

0.012
dried
freely
hygroexpansion coefficient (b RH)

0.008 dried
restrained

0.004
curl 1.37
curl 1.22
curl 1.17

0
200 400 600 800 1000
3
density (kg/m )

Figure 4.18. The effect of fibre shape on the hygroexpansion coefficient.

The more the fibres are oriented in the machine direction, the greater is the probability for
bonded regions in series with CD, Figure 4.19. In MD we assume five fibres in this unit area
and in CD three. In CD we have five chances for micro-compressions, in MD three. This naïve
picture may explain the fact that the shrinkage and hygroexpansion increases the more the fibres
are oriented in MD, in a paper machine. In Figure 4.20 the same principle applies for restrained
dried papers.

before drying after drying

1, MD
CD
2, MD
MD
3, MD

4, MD

5, MD
1, CD 2, CD 3, CD

Figure 4.19. Micro-compression development as a function of fibre orientation.

 126

Figure 4.20. Hygroexpansion coefficient for restrained dried papers as a function of the fibre orientation.

Referring to Figure 4.12 and Figure 4.16 again, we can investigate if there is a correlation
between hygroexpansion coefficient and shrinkage during drying for these pulps, Figure 4.21.
While there is a general trend, it is apparent that the hygroexpansion did not show a particularly
good correlation against shrinkage considering all kind of pulps.

0.15

0.13 CMP
hygroexpansion coefficient (b f)

LYS
0.11 (beaten)

0.09 SGW CTMP

BKP
(beaten)

0.07 TMP LYS

0.05
1.0 2.0 3.0 4.0 5.0
drying shrinkage (%)

Figure 4.21. Hygroexpansion coefficient as a function of drying shrinkage for the pulps in Figure 4.12.

For a given pulp, it is evident that a reduction of the drying shrinkage would decrease the hy-
groexpansion. In Figure 4.22 the effect of wet pressing on hygroexpansion is investigated for
 127

the pulps in Figure 4.12. In this figure it appears that the drying shrinkage is constant up to a
certain threshold in density, beyond which drying shrinkage decreased.

4.5 beaten LYS

CMP
LYS
SGW
3.5
drying shrinkage (%)

2.5

1.5

0.5

0
300 400 500 600 700
sheet density (kg/m3)
increased wet pressing

Figure 4.22. Hygroexpansion coefficient as a function of density obtained by wet-pressing, for the pulps in Fig-
ure 4.12.

The critical solids content at which the shrinkage starts falling is closely related to the WRV,
Water Retention Value, Figure 4.23.

80

SGW

70
critical solids content (%)

CMP
LYS
60
LYS
(beaten)

50

40
1.4 1.6 1.8 2.0 2.2 2.4 2.6
water-retention value

Figure 4.23. Critical solids content as a function of WRV, for the pulps in Figure 4.12.

The analysis so far has indicated that the hygroexpansion is related to the fibre orientation,
the interaction between the fibres at the fibre bonds, the fibre shape and shrinkage. Furthermore
the shrinkage is related to the WRV and the amount of moisture in the paper after wet pressing.
128

QKS==mÉêã~åÉåí=aáãÉåëáçå~ä=`Ü~åÖÉë=ÇìêáåÖ=jçáëíìêÉ=`óÅäáåÖ

If a paper network is dried freely the fibres have reached the natural equilibrium state of defor-
mation. If this network is adsorbing and desorbing moisture the network swells and contracts
without any permanent deformation since the natural shrinkage state was already obtained dur-
ing the drying phase.
If the network on the other hand is dried under restraint, the shrinkage of the network has
been permanented in an un-natural state. Every time this network is adsorbing moisture the net-
work swells but when the moisture enters the network it softens the fibres and makes the paper
try to shrink part of the way to a freely dried state. When desorbing moisture the network defor-
mation is permanented again. Every time the network passes a moisture cycle, it tries to shrink
further until an equilibrium state is achieved.
Different terms for this phenomenon have been used in the literature. The term „release of
built-in strains“ seems to describe the situation best.
To achieve release of strains, RH of 65–70 % must be exceeded. The phenomenon is more
marked in MD than in CD, due to the drying conditions.
The strain release phenomenon is illustrated in Figure 4.24 where the RH is varied in the cy-
cle 35-60-85-60-35 %. The upper curve shows the effect for a freely dried sheet, and it is evi-
dent that the dimensional changes are not changed with time.
The lower curve applies to sheets dried under restraint. The paper shrinks gradually with in-
creasing number of moisture cycles.

1.6

1.2
dried
freely
0.8
hygroexpansion (%)

0.4

0
dried
–0.4 under
restraint

–0.8
RH-variation (%)
85 85 85 85 85
60 60 60 60 60 60 60 60 60 60
35 35 35 35 35 35

Figure 4.24. Dimensional changes in freely dried sheets (upper curve) and in sheets dried under restraint (lower
curve).

The same phenomenon may be illustrated as strain versus moisture content as in Figure 4.25
and Figure 4.26 for a fine paper in CD where the paper has dried more or less freely and in MD
where the paper has been dried under restraint or even strained a few percent.
 129

Figure 4.25. Hygroexpansion as a function of moisture content of fine paper (CD).

Figure 4.26. Hygroexpansion as a function of moisture content of fine paper (MD).

The dimensional interval in hygroexpansion is large for freely dried sheets. During repeated
moisture cycling, on the other hand, the dimensional interval changes very little. In the case of
sheets dried under restraint, the dimensional interval is relatively small, especially in the first
cycle and increases with repeated moisture cycling. The sheet also shows a permanent shrink-
age.
In the paper machine, built-in strains are built up more easily in the MD because the shrink-
age is prevented. In CD, the sheet shrinks and the degree of built-in strains becomes small. The
sheet then assumes its equilibrium position more rapidly in the CD than in the MD during re-
peated moisture cycling. The way in which a given paper shall be dried to give it the desired hy-
groexpansion properties must consequently mean that there must be a balance between whether
a high hygroexpansion coefficient and a small change during moisture cycling are desired as in
the CD or a small hygroexpansion coefficient but a large change during moisture cycling as in
the MD.
The „release of built-in strains“ sometimes results in extreme effects such as those described
in the following two figures. An MD sample of newsprint was exposed to different amounts of
moisture contents. Figure 4.27 show that the paper first expands and then contacts again due to
130

the release of the built-in strains. The consequence could be that the paper in a printing press be-
comes stiffer as the moisture increases.
The time effects of hygroexpansion strain with time after moisture exposure is shown in Fig-
ure 4.28. The starting atmosphere was 50 % RH.

0.08
newsprint
hygroexpansion strain (%)

0.04

7 9 11 13 15 17 19
moisture content (%)

Figure 4.27. Hygroexpansion strain as a function of moisture content of newsprint in MD.

0.09
newsprint
0.07
hygroexpansion strain (%)

75
80
0.05 85
70
0.03
88

0.01
90
RH
–0.01
–10 0 10 20 30 40 50 60 70
time (min)

Figure 4.28. The time effects of hygroexpansion strain in MD with time after moisture exposure. The starting
atmosphere was 50 % RH.

QKT==jÉ~ëìêÉãÉåí=jÉíÜçÇë

Many types of devices exist for recording the dimensional changes of paper during sorption.
STFI uses a device shown in Figure 4.29. The feature of the principle is that the paper is flat-
 131

tened by a weight during measurement to avoid buckling of the test piece. The deformation is
recorded by a deformation-recording device, LVDT.

spring weight razor blade LVDT

springs
paper

Figure 4.29. STFI deformation-recording device.

Speckle photography is a modern way to record two-dimensional deformations in a paper. In

short, on the paper surface is applied small irregular dots. A camera detects the average move-
ment of dotted areas and a very precise deformation recording can be achieved, Figure 4.30.

undeformed deformed
image image

PC CCD
camera
512

on-line monitor object

fixture 512

32 +
32
The experimental set-up for electronic speckle
photography for the evaluation of the in-plane
displacements of the object. It consists of a
CCD camera with synchronous pixel clocking,
PC with a frame grabber card and an on-line
monitor.
y-displacement (pixels)

10

Principle of the analysis means cross- 0

correlation and the true displacement between
the subimages was x = y = 7.5 pixels.
–10

–10 0 10
x-displacement (pixels)

Figure 4.30. Principle of speckle photography.

132

The strains can be expressed in polar plots, Figure 4.31. In this case the symmetry of the hy-
groexpansion strain has an off-axis angle to MD. The same off-axis angle can also be obtained
by ultrasonic techniques, Figure 4.32.

MD

aH

CD

0.10 %

Figure 4.31. The strains can be expressed in polar plots. In this case the symmetry of the hygroexpansion strain
has an off-axis angle to MD.

MD
aus

CD

5 (km/s)2

Figure 4.32. The stiffness index in a polar diagram where o) are measurement points. Note the tensor scale for
stiffness index. The tensile stiffness orientation, Dus is defined. Copy paper 80 g/m2.
 133

QKU==`ìêä=~åÇ=qïáëí

Curl in paper or carton board is a situation where the sheet bends out of the plane because of a
mechanical influence or because the temperature or moisture has influenced the sheet. Twist is a
situation where the sheet is turned like a propeller.
Curl caused by moisture content changes depends on the fact that the paper has an asymmet-
rical structure in the thickness direction. One side then swells or shrinks more than the other and
this causes the sheet to bend. The basic reason for curl problems is thus to be found in differenc-
es in the hygroexpansion of the different layers of the paper.
Curl due to moisture exposure can also have other causes. One such cause is the case where
the sheet is moistened or dried from one side, Figure 4.33. The reasons are hygroexpansion and
release of built-in strains as discussed before.

blanket
water application drying

Figure 4.33. The sheet is moistened or dried from one side. The reason is hygroexpansion and release of built-in
strains as discussed before.

Cockling, Figure 4.34, Figure 4.35 and Figure 4.36, are out-of-plane movements locally
over a paper.

Figure 4.34. Cockling is out-of-plane movements locally over a paper.

134

cockling caused by non-uniform moisture content before drying

cockled areas = dryer areas flat areas = wetter areas
before drying before drying

Figure 4.35. Cockling experiments.

shrinkage in
less moist area
buckling

drying

dry sheet
more moist area

Figure 4.36. Cockling mechanism.

Cockling stems from its irregular distribution of strain over a paper surface. According to
that theory, cockling takes place if paper inhibited in its shrinking from point to point.
Brecht, Müller and Weiss have tried to establish the correctness of this theory. Moist sheets
of paper were prepared, which had an uneven distribution of moisture over its surface corre-
sponding to a certain pattern. When the sheets were subjected to contact drying, there resulted
the artificially produced cockling shown in Figure 4.35, which exactly reproduced the pattern
of moisture distribution just mentioned. There were circular patches on the sheets, which before
were less moist than the remaining regions in the two left-hand pictures and more moist in the
two right-hand pictures. The cockles always coincided with the initially less moist spots, where-
as the initially moister patches remained flat. The explanation is that the initially drier patches
 135

enter that stage sooner during the drying in which fairly large shrinkage is connected with rela-
tively small evaporation. These spots thus tend to shrink more than the others. Consequently,
there arise tensions on their margins that hamper shrinking. If the sheet was flat until then, the
other parts of the sheet shrink as a whole towards the sheet centre in the further course of dry-
ing. The patches already dried cannot shrink any more; they are warped out o the plane and
form cockles.
Since cockling comes from uneven shrinkage conditions, one may expect that it will mani-
fest itself more markedly, on the whole, where greater shrinking is to be anticipated than in the
case of lesser tendency to shrink.
An out-of-plane shape of a paper or corrugated board may be subdivided into three compo-
nents, Figure 4.37.

MD warp

CD warp

twist warp

Figure 4.37. An out-of-plane shape of a corrugated board may be subdivided into three components, curl in MD,
curl in CD and twist.

8
6
Ky
edge displacement (mm)

4
2
0
–2
–4
Kx
–6
–8
0 5 10 15 20 25
relative humidity change (RH-%)

Figure 4.38. Measured curl components, Kx and Ky (crosses and circles, respectively) of a two-ply, 0º/90º lami-
nate of two copy paper sheets vs. a change in relative humidity, 'RH. The lines show a finite element calculation.
The bifurcation point at 'RH = 7 % indicates the buckling transition. After that, only one curl component is non-
zero.

1 1 1
w( x, y ) K x ˜ x 2  K y ˜ y 2  K xy ˜ xy
2 2 2
136

Curl in two directions cannot exist at the same time, Figure 4.38. At a given RH or moisture
content one of the directions dominates and the other returns to zero. The point is called the bi-
furcation point.
Figure 4.39 and Figure 4.40 show typical behaviours for curl when the adsorption starts by
adsorption and desorption respectively.
It is possible by laminate theory to calculate curl and twist if the properties of the plies of the
paper are known. Simplified equations have been derived in special cases, when the paper is
modelled as two plies with equal thickness, Figure 4.41.

Figure 4.39. Bending curvature response of fine paper to cyclical humidity changes when the process starts from
the adsorption stage.

Figure 4.40. Bending curvature response of fine paper to cyclical humidity changes when the process starts from
the desorption stage.

1 3 B
curl
R 2t

E ˜ H B  ET ˜ HT  (4.14)
 137

E is the hygroexpansion coefficient for the bottom ply B and the top ply T respectively. H is
the change in moisture.

t/2
thickness = t top ply
t/2
bottom ply

Figure 4.41. Paper is modelled as two plies with equal thickness.

Curl thus depends on the difference in hygroexpansion, and on the strain on the bottom and
top plies. Curl decreases with increasing thickness. It can be shown that the tensile stiffness in-
dex of the layers plays a very small role.
The main requirement for a paper sheet to remain flat during moisture exposure is thus that
the sheet is symmetrical with regard to its middle line in the thickness direction in both MD and
CD. For the different layers, the symmetry includes both thickness and hygroexpansion.
Whether an anisotropic machine-made sheet curls in the MD or the CD depends on the di-
mensions of the sheet. A thin strip always curls in its length direction regardless of whether this
is MD or CD. Symmetrical and round sheets tend to be dominated by curl in CD. Even if mois-
ture diffuses into the sheet from both sides, curl can occur if the hygroexpansion is different in
different layers. Curl is always balanced in practice by adjustment of the steam pressure in the
upper and lower cylinders in the dryer section on the paper machine.
Reasons for unequal sidedness in hygroexpansion, which causes curl in a single-layer sheet,
can also be different fibre orientation and the content of fines on the upper and bottom side of
the sheet. For a multi-layer sheet, the layer structure itself can be asymmetrical.

y
␸T
X

B ␸B
1
2

Figure 4.42. Twist occurs, for a two-ply paper when one or two plies have an off-axis symmetry. The definition
of the off-axis angle is indicated.

Twist occurs if some layer in the paper is asymmetrical in relation to the MD, Figure 4.42.
During exposure to moisture, such crookedness leads to a turning moment which twists the
sheet.
138

Figure 4.43. In the instrument the 100 mm by 100 mm samples are scanned by a laser gage and the three curl and
twist components calculated.

traversing, non-contact
distance-meter

test piece holder rack

micro-processor

conditioning unit

Figure 4.44. The curl and twist tester.

Twist may be expressed in a simple equation as follows for the paper shown in Figure 4.42.

3ª T T T
k6
t¬
  
I H E y  E xT  I B H B E yB  E xB º¼  (4.15)
 139

E is the hygroexpansion coefficient for the bottom ply B and the top ply T respectively. H is
the change in moisture and x and y are two perpendicular directions in the paper where 1 is MD
and 2 is CD.
Twist may be measured in several ways. In the instrument shown in Figure 4.43 and Figure
4.44, the 100 mm by 100 mm samples are scanned by a laser gage and the three curl and twist
components calculated. Another technique involves scanning the shape of a circular test piece.
Some other techniques have been proposed. Figure 4.45 shows the cross-cut method.

Figure 4.45. The cross-cut method for curl measurements

Working with narrow strips, the curl is evaluated by comparing the curvature against known
curvatures or calculating the curvature by measuring the bending of the strip, Figure 4.46.

Figure 4.46. Curl measurements by using strips.

Any two-sidedness will influence curl and twist. The two-sidedness of some machine param-
eters have been investigated, Figure 4.47–Figure 4.49.
140

water removal
650
to wire side
to top side
600
density (kg/m3)

550

500

450
unpressed
400
wire side top side

Figure 4.47. Two-sidedness in density created by wet pressing.

50 50
40 increasing jet speed 40 increasing jet speed
0.986 1.041
30 30
20 1.001 20
curl (mm)

curl (mm)

1.015
10 10
0 0
1.028 1.028
–10 –10
A B
– 20 – 20
0 2 4 6 0 2 4 6
a) position (m) b) position (m)

Figure 4.48. Curl due to different jet/wire ratios.

40
fines content (%)

30

duo-F
base belbaie-II news
20 stock

0 20 40 60 80 100
felt side wire side

Figure 4.49. Two-sidedness in fines content.

 141

QKV==qÜÉ=jÉÅÜ~åáÅ~ä=mêçéÉêíáÉë=~í=aáÑÑÉêÉåí=jçáëíìêÉ=`çåíÉåíë

The in-plane stiffness index of paper may be measured by ultrasonic USI and regular tensile
testing TSI (Chapter 2).
The development of these different stiffnesses with increasing RH or moisture ratio is differ-
ent due to the time scale of testing. The TSI is more sensitive than USI. A representative exam-
ple is given in Figure 4.50.

Stiffness Index, MNm/kg

20 Paperboard

15

10
MD USI

5 MD TSI
CD USI
CD TSI
0
05 10 15 20 25
Moisture Ratio, %

Figure 4.50. The development of these different stiffnesses with increasing RH or moisture ratio is different due
to the time scale of testing. The TSI is more sensitive than USI.

The hygroscopicity of the paper means a lot for the mechanical properties of the paper and
all paper properties are influenced in different ways by the moisture content. Figure 4.51 shows
the principal change as a percentage for some paper properties at different RH with 50 % RH as
reference.

80
60
40
tensile index
20
relative values

0
tear index
– 20
– 40 strain at
– 60 break
folding
– 80 endurance

0 10 30 50 70
RH (%)

Figure 4.51. Level deviation as a percentage of the value at 50 per cent RH.
142

As shown earlier in this chapter, the built-in strains in the paper are often released during
moisture cycling, which results in a permanent shrinkage of the paper. This release of strains in-
fluences the mechanical properties very little, however.

10 10

9 9
kraftliner 8 kraftliner

compression strength (kN/m)

8 A
7
compression strength (kN/m)

7 MD 6
6 MD
5

5 4
CD
3
4 CD
A B 2
3
1
2 B 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
1
moisture content (%)
0
0 20 40 60 80 100
RH (%)

Figure 4.52. Compression index against RH and against moisture content.

12 drying under restraint to 50 % RH

increasing RH to given levels
again conditioning at 50 % RH
tensile stiffness index (MNm/kg)

10

8 drying under restraint

to given level

2
40 60 80
relative humidity (%)

Figure 4.53. Tensile stiffness index against RH.

Figure 4.52 shows an example of this. The compression index of the paper is here plotted
against the moisture content of the paper. Independently of how a given moisture content has
 143

been reached, whether from dry or from humid conditions, the paper has the same strength at a
given moisture content. The same is true for tensile index and tensile stiffness index.
If the corresponding strength is plotted against the RH of the air, the earlier mentioned hys-
teresis phenomenon is evident since a given RH does not unambiguously determine the mois-
ture content.
The same applies for tensile stiffness index TSI, Figure 4.53. The TSI drops with RH, but
TSI does not change if the paper is subjected to high RH levels and then conditioned again.

QKNM==jÉÅÜ~åçëçêéíáîÉ=`êÉÉé

Creep is the time dependent strain as a result of a constant force (Chapter 2). Figure 4.54 shows
a typical creep result for paper in tension at 80 % RH. If the humidity of the air is lowered to
30 % RH, after a given time, here 4000 and then cycled between these levels, seconds, the creep
process accelerates. The phenomenon is called the „mechano-sorptive effect“.

2
TMP samples
cyclic 80 – 30 % RH
25 % of the beaking load at 50 % RH
strain (%)

80 % RH

0
1 2 3 4 5
log time (seconds)

Figure 4.54. Creep in tension at 80 % RH and creep in tension due to a variation in RH between 30 and 80 % RH.
100
 145

R==léíáÅ~ä=mêçéÉêíáÉë=çÑ=mìäé=~åÇ=m~éÉê

Anthony Bristow
Bristow Consulting AB, Tullinge

5.1 Introduction 146

5.2 Light Absorption and Light Scattering 146

5.3 The Kubelka-Munk Theory 147

5.3.1 Derivation of the Kubelka-Munk Equations 147
5.3.2 Limitations of the Kubelka-Munk Analysis 151
5.3.3 The Absorption and Scattering Coefficients 151
5.3.4 Applications of the Kubelka-Munk Analysis 152

5.4 The Relationship between s, k, Rf and R0 153

5.5 Measurement of Rf, RM=etc 154

5.6 Standard Illuminants 155

5.7 Standard Observers 156

5.8 Standardized Measurement Conditions (ISO 2469) 156

5.8.1 Instruments Conforming to ISO 2469 157
5.8.2 Calibration of the Measurement Instruments 158
5.8.3 Calibration within the Paper Industry – the ISO-Hierarchy 158

5.9 Optical Measurements 159

5.10 Colour Perception 161

5.11 Tristimulus Values 161

5.12 Chromaticity Coordinates 162

5.13 Dominant Wavelength and Excitation Purity 163

5.14 CIELAB-Coordinates 165

5.15 Whiteness 165

5.16 Fluorescence 166

146

RKN==fåíêçÇìÅíáçå

Optics is the science of light and human vision, and the optical properties of a paper are the
properties connected with its appearance, i.e. its brightness, opacity and colour.
The appearance of a material is not, however, a simple, easily defined, physical property. It
is the psychophysical consequence of several different processes. It is the result of:
• an interaction between the material and the incident light, which defines the characteristics
of the light reflected or radiated from the material,
• a stimulation of cells in the retina, when the light reflected or radiated from the material
meets the human eye,
• a perception in the human brain, as a result of an interpretation of the signals transmitted
from the eye to the brain
This means that we must consider this total phenomenon in two different stages:
• From a papermaking point of view, we need to consider how light interacts with the mate-
rial, and how different factors in the raw materials, in the manufacturing process and in the
attained structure influence this interaction
• From a quality control point of view, we need to consider how the eye and brain react to the
stimulus of the light received from the material, so that we can construct a realistic instru-
ment to measure the radiation in a relevant manner and to compute relevant parameters to
characterise the appearance of the material.
This chapter includes
• a short presentation of a mathematical theory for the interaction between the different com-
ponents of light and paper,
• an account of how instruments are designed and calibrated to give measurement values
which correspond to the perceptual conception of the appearance of a material, and
• a description of different mathematical ways of treating the measurement data to obtain
parameters which give a meaningful representation of properties such as the colour and
whiteness of paper.

RKO==iáÖÜí=^Äëçêéíáçå=~åÇ=iáÖÜí=pÅ~ííÉêáåÖ

When a ray of light of a given wavelength strikes a paper sheet, several different things happen.
Part of the light is reflected in the specular direction from the surface of the paper (i.e. at an an-
gle to the normal equal to the incidence angle) while another part passes into the material. The
light which enters the paper is then partly absorbed and partly scattered in different directions
before being reflected back from the surface of the paper or leaving the paper through its under-
side. The appearance of the material is dependent on the extent to which light of different wave-
lengths is reflected and absorbed.
• The interaction between the light and the paper can be described with the help of an absorp-
tion coefficient, k, and a scattering coefficient, s.
 147

• The light absorption coefficient, k, is a measure of that portion of the light which is
absorbed. If the absorption is different in different wavelength ranges, the material appears
coloured. Unbleached grades have a relatively high absorption coefficient, especially in the
blue range, while bleached grades have a very low absorption coefficient. Fillers also have
a low absorption coefficient.
• The light scattering coefficient, s, is a measure of the free, unbonded surface area of fibres
and other particles within the paper. Light is scattered at the interface between two phases
with different refractive indices, so that a primary requirement if light is to be scattered in a
material such as paper is that the material is porous. The light is reflected at the surfaces of
the fibres, both at external surfaces and at internal cracks etc in the fibre.
An example of what a change in the free, unbonded, light-scattering surface does for the op-
tical properties can be seen in the difference between white, newly-fallen snow and a piece of
transparent ice. The scattering coefficient in the newly fallen snow is very high because of its
porous nature and because of the very large free, light-reflecting surface area in each snowflake.
In ice, on the other hand, there are no free, light-reflecting surfaces and the scattering coefficient
is close to zero.

RKP==qÜÉ=hìÄÉäâ~Jjìåâ=qÜÉçêó

In the 1940s, Kubelka and Munk derived equations to describe light reflection from thin layers
of paints and similar materials. These equations have since been successfully applied to paper.
The Kubelka-Munk equations are valid only when both illumination and measurement take
place with diffuse light. Within the field of paper technology, diffuse illumination is therefore
used in most optical measurements.
The Kubelka-Munk equations are here derived for a simplified case, where the ray path is
drawn perpendicular to the plane of the paper. This derivation gives the same equations as the
more complete analysis with diffuse illumination.

RKPKN==aÉêáî~íáçå=çÑ=íÜÉ=hìÄÉäâ~Jjìåâ=bèì~íáçåë

Consider a sheet of paper with grammage W and reflectance factor R placed over a background
with reflectance factor Rg. This sheet is illuminated with light having an intensity I, and light
having an intensity J is reflected, as indicated in Figure 5.1. The reflectance factor R of the
sheet is then given by Equation 5.1:

J
R (5.1)
I

Within this sheet of paper, consider an infinitesimal layer dx, illuminated with intensities i
(from above) and j (from below at a distance x from the base). In its passage through this layer,
part of the light is absorbed and part of the light is scattered, so that i and j are reduced respec-
tively by (s + k) idx and (s + k) jdx. It is here assumed that all scattered light is scattered at an
angle of 180° to its original direction so that, s · jdx and s · idx, which respectively represent the
148

light scattered when the light passes in one direction, are added to the intensities i and j when
the light travels in the opposite direction.

I J
R

I(x) dx
r(x)
J(x) x
Rg

Figure 5.1. Model for derivation of the Kubelka-Munk equations.

The total changes di and dj, calculated in the x-direction, are therefore:

 di   s  k  ˜ i dx  s ˜ j dx (5.2)

dj   s  k  ˜ j dx  s ˜ i dx (5.3)

Across the infinitesimal layer dx, r = j/i , and this gives:

§ j· i dj  j di
dr d¨ ¸ (5.4)
©i¹ i2

Thus, if the two differential equations, Equation 5.2–5.3, are multiplied by j and i, respec-
tively, and the resulting equations are added, we obtain:

 j di   s  k  ˜ i j dx  s ˜ j 2 dx
i dj   s  k  ˜ i j dx  s ˜ i 2 dx
i dj  j di 2  s  k  ˜ i j dx  s ˜ i 2 dx  s ˜ j 2 dx

and division by i2 gives:

dr ª¬ 2  s  k  r  s  sr 2 º¼ dx (5.5)

If, in this equation, we set:

a
s  k  1
k
(5.6)
s s

we obtain:
 149

dr r 2
 2ar  1 s dx  (5.7)

It is necessary to solve this equation by integration between the grammage limits x = 0 and
x = w and between the reflectance factor limits r = Rg and r = R:

w R
dr
³ s dx ³ r 2  2 ar  1 (5.8)
0 Rg

Integration then gives:

R
1 ª § 2 r  2 a  4 a2  4 ·º
sw «ln ¨ ¸» (5.9)
4 a 2  4 «¬ ¨© 2 r  2 a  4 a 2  4 ¸¹ »¼
Rg

This equation can be simplified by considering a special limiting case of Equation 5.5. As-
sume that the reflectance from the surface R is the same as the reflectance from the background
Rg, i.e. dr = 0 through the whole sheet. In this situation, i.e. when the paper is placed over a pa-
per pad which is so thick that the reflectance factor is not changed when a further sheet is added
to the pad, the reflectance factor is designated R = Rg = Rf. From Equation 5.5, we then obtain:

s r 2  2 s  k  ˜ r  s 0 (5.10)

Whence

k ª§ k · 2 § k ·º
r Rf 1   «¨ ¸  2 ¨ ¸ » (5.11)
s «¬© s ¹ © s ¹ »¼

This is often written in the form:

1  Rf 
2
k
(5.12)
s 2 Rf

So that

1§ 1 ·
a ¨  Rf ¸ (5.13)
2 © Rf ¹

The reflectance factor Rf of a thick pad of paper is called the reflectivity of the paper and is
thus a material property, which is dependent only on the ratio between k and s, and not on their
absolute values.
150

As shown in Figure 5.2, where the ratio s/k is plotted on a logarithmic scale, large relative
changes in this ratio are required to change the reflectivity, Rf.

1.0

0.8
R

0.6

0.4

0.2

0
0.01 0.1 1 10 100 1000 10000
s/k

Figure 5.2. The reflectivity Rf increases with increasing value of the ratio s/k.

Substitution of Equation 5.13 in Equation 5.9 leads to:

R
ª § 1 ·º
1 « ¨ 1  R ¸»
sw «ln ¨ f »
¸ (5.14)
1 « ¨ r  Rf ¸»
 Rf
Rf «¬ ¨© ¸»
¹ ¼ Rg

After inserting the integration limits, we finally obtain:

1 § 1  RRf   Rf  Rg  ·
sw ln ¨ ¸
 Rf © 1  Rg Rf   Rf  R  ¹
1 ¨ ¸ (5.15)
Rf

and

§ 1  Rf 2 ·
kw sw ¨ ¸ (5.16)
¨ 2 Rf ¸
© ¹

Both s and k have thus been expressed as functions of a number of reflectance values and
these equations enable both s and k to be calculated if reflectance measurements can be made
against two different backgrounds, i.e. provided the material is not completely opaque, and if
the grammage is known. To use the equations given here, one measurement is made over an
opaque pad to obtain Rf and a second measurement is made over an arbitrary backing Rg to ob-
tain R.
 151

Equation 5.15 is simpler to handle if one measurement is made with an opaque pad of the
material to obtain Rf, and a second is made against a black background with a reflectance factor
Rg = 0 to obtain the reflectance factor designated R0:

1 § 1  R0 Rf  Rf ·
sw ln ¨ ¸¸
¨
 Rf ©  Rf  R0  ¹
1 (5.17)
Rf

The product of the scattering coefficient and the grammage, sw, is called the scattering pow-
er of the sheet. The scattering power or covering ability, sw, is a very important property for
printing paper and is a measure of the ability of the paper to conceal print on the reverse side of
the paper. It can also be an important characteristic of a layer of a bleached furnish on a white-
top liner or of a coating layer where it is necessary to hide the brown colour of the underlying
plies or substrate.
The ratio 100 R0/Rf is called the opacity of the material. In later sections of this chapter, the
way in which R, R0, Rg and Rf can be measured is described.

RKPKO==iáãáí~íáçåë=çÑ=íÜÉ=hìÄÉäâ~Jjìåâ=^å~äóëáë

It is clear, by consideration of Equation 5.17, that the Kubelka-Munk analysis can yield values
for s and k only if it is possible to obtain values of R0 and Rf which are significantly different
from each other. In addition, the following conditions must be observed:
• a single sheet of the material must be optically thick, i.e. sufficiently thick to allow multiple
scattering to occur, with a reflectance factor over 50 % and a transmittance less than 20 %.
• there must be negligible surface reflection, i.e. the material must not be highly glossy and
direct surface reflection must be avoided with a gloss trap during measurement
• the structure must be uniform throughout the sheet (for laminated structures, special equa-
tions are necessary)
• the illumination must be diffuse
• the optical phenomena must be limited to absorption and scattering, which means that no
fluorescence is permitted.

RKPKP==qÜÉ=^Äëçêéíáçå=~åÇ=pÅ~ííÉêáåÖ=`çÉÑÑáÅáÉåíë

Although these absorption and scattering coefficients are mathematical parameters derived
from reflectance data by application of the Kubelka-Munk theory, the ability of a material to ab-
sorb or scatter light of different wavelengths is a physical reality.
Certain molecular groups have the ability to absorb light in certain wavelengths, and the
presence of these groups in e.g the lignin means that the paper or pulp appears off-white or
coloured, and such groups are called chromophores. It is the task of cooking and bleaching to
remove such chromophores and thus give a less yellow pulp, i.e. these operations reduce the k-
value. In some cases, dyestuffs are added to give a strongly coloured paper, i.e. to increase the
k-value in the complementary wavelength range. In other cases, e.g. in the manufacture of
152

newsprint, small quantities of a bluish dye may be added to counteract the residual yellowness
of the pulp and thus give a more neutral shade. In a pulp, the absorption power is essentially an
intrinsic property of the sheet and it is not affected by factors such as the degree of consolida-
tion of the material in the sheet.
In contrast, the scattering power is greatly affected by the sheet structure. Surfaces and inter-
faces scatter light only if the two phases have different refractive indices and if they are not in
„optical contact“, which means that the distance between two surfaces must be greater than half
the wavelength of light. The process of consolidation of the sheet and subsequent calendering
operations therefore tend to lower the scattering power.

RKPKQ==^ééäáÅ~íáçåë=çÑ=íÜÉ=hìÄÉäâ~Jjìåâ=^å~äóëáë

If the k and s values of various potential components of a furnish are known, the properties of
the furnish can in principle be calculated by simple additivity. For example, the absorption coef-
ficient of a mixed furnish is given by Equation 5.18 where pA, pB, pC etc are the gravimetric pro-
portions of components A, B and C in the stock.

k p A k A  pB k B  pC kC   (5.18)

Similarly, to a first approximation, the scattering coefficient of a mixed furnish is given by:

s | p A s A  pB sB  pC sC   (5.19)

This equation must be used with caution, since the manner in which the different compo-
nents bond together may affect the total free scattering surface in the resulting material. This is
particularly the case with fillers, where packing of the particles means that the full potential of
the filler in this respect is rarely achieved and it is necessary to talk of an „apparent“ scattering
coefficient. Figure 5.3 shows this effect for different mixtures of precipitated calcium carbonate
(PCC) and ground chalk, where there is a linear additive relationship between the absorption co-
efficients of the two pulps but where the scattering coefficients are not additive.
The two forms of calcium carbonate have different particle sizes and different size distribu-
tions and thus have different scattering coefficients.

0.40 60
scattering coefficient
coefficient (m2 /kg)
light absorption

0.30 50
(m2 /kg)

0.20 40

0 20 40 60 80 100 PCC (%) 0 20 40 60 80 100 PCC (%)

100 80 60 40 20 0 chalk (%) 100 80 60 40 20 0 chalk (%)

Figure 5.3. (a) The light absorption coefficient and (b) the light scattering coefficient of papers containing differ-
ent mixtures of precipitated calcium carbonate and ground chalk (after Larsson & Sikker, 1989).
 153

RKQ==qÜÉ=oÉä~íáçåëÜáé=ÄÉíïÉÉå=s, k, Rf=~åÇ=RM

Figure 5.4 shows, in an optical constitutional diagram, the relationships between some basic op-
tical quantities and how the different quantities are dependent on each other.
The diagram shows Rf and the opacity as functions of the light absorption and light scatter-
ing coefficients k and s, and clarifies the principles for the control of papermaking towards im-
proved optical properties. The lines relating to the reflectivity are independent of grammage.
The lines relating to opacity have been calculated assuming a grammage of 60 g/m2. The dia-
gram is valid independently of wavelength, which means in other words that the relationships
hold provided that all measurements are made at the same wavelength.

reflectivity
4
50
opacity
90
3 55
k (m2/kg)

60 85

2 65 80

75
70
70
75
60
1
80
85
90

0
10 20 30 40
s (m2/kg)

Figure 5.4. The relationships between the optical properties. The opacity data refer to a grammage of 60 g/m2.

The optical properties of a material are dependent on the light absorption coefficient and
light scattering coefficient of the material. It is thus important for the papermaker to know how
the different processes in the papermaking process influence these two material properties. In
general, it may be said that pulping and, in particular, bleaching processes seek to reduce the
light absorption coefficient, whereas the papermaking process involves efforts to increase the
light scattering coefficient.
The first step in the refinement of a pulp is a bleaching process, which lowers the absorption
coefficient. In this process, one moves downwards in the diagram and crosses the straight lines
for the Rf-value towards a higher reflectivity.
When paper is produced from the pulp, it must first be mechanically beaten. In the case of
mechanical pulps, the refining process leads to a greater number of small particles which give
an increase in the scattering coefficient, so that one moves horizontally to the right in the dia-
gram towards a higher Rf-value and a higher opacity.
154

In the case of chemical pulps, on the other hand, beating does indeed lead initially to a great-
er number of small particles, but in the papermaking process these bond together and come into
optical contact with each other and the scattering coefficient drops This means that one moves
horizontally to the left in the diagram towards a decreasing light scattering coefficient, decreas-
ing Rf-value and decreasing opacity.
If a filler which has a higher light scattering coefficient than the pulp is added, one moves to
the right in the diagram towards both a higher Rf-value and a higher opacity.
The addition of a dye to the paper always increases the absorption coefficient, so that one
thus moves upwards in the diagram towards a lower Rf-value, but towards a higher opacity.
The fact that the opacity may be raised by increasing k as well as by increasing s is often im-
portant. A black paper is more efficient than a white paper if a highly opaque material is re-
quired.
The dilemma of the papermaker is that any measure to increase the scattering coefficient nor-
mally tends to decrease the strength and vice versa. This can be seen in the effects of beating a
chemical pulp, of adding mechanical pulp to a chemical pulp furnish, of adding a filler, of wet
pressing etc.
Figure 5.5 shows the curves of Figure 5.4 divided into two groups that show reflectivity and
opacity separately, and approximate positions for a number of different paper grades are indi-
cated in relation to the different parameters.

20 20
grammage: 60 g/m2
20

30
16 16
unbleached kraft unbleached kraft

50 reflectivity
12 12
k (m2/kg)

k (m2/kg)

60

8 8 opacity mechanical pulp

mechanical pulp 70

90

4 bleached chemical pulp 80 4 80 bleached chemical pulp coated paper

coated paper 70 filler-containing fine paper
filler-containing fine paper
60
90 50

0 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
2
s (m /kg) s (m2/kg)

Figure 5.5. The relationship between k and s showing (a) reflectivity contours and (b) opacity contours, with an
indication of the approximate positions of different paper grades.

RKR==jÉ~ëìêÉãÉåí=çÑ=ofI=oM=ÉíÅ

The reflectivity Rf and reflectance factors R0, Rg etc can in principle be measured under arbi-
trary conditions, but measurements are best carried out under conditions which correspond to
the human perception of the appearance of the paper. Before an image is created in the human
brain, three components are required:
 155

• radiation from some light source must strike an object

• at least part of this radiation must be reflected from the object
• the reflected radiation must reach the eye and a signal must be sent to the brain.
The signal which reaches the brain and gives rise to a perception of brightness and colour of
a material is thus dependent on
• the spectral composition of the incident radiation
• the spectral reflectance characteristics of the material
• the spectral sensitivity of the human eye
• the geometrical conditions of the illumination and viewing situations.
To obtain a meaningful measurement of the optical properties of a material, it is thus not suf-
ficient merely to measure the reflectance of the material in some arbitrary wavelength range.
Firstly, it is necessary to select and standardise the type of illumination used. Secondly, it is
necessary to filter the radiation using a filter which corresponds to the sensitivity of the eye at
different wavelengths. Thirdly it is necessary to decide upon a suitable geometrical presenta-
tion.

RKS==pí~åÇ~êÇ=fääìãáå~åíë

To facilitate measurement under standardized conditions, CIE (The International Lighting

Commission) has standardized a number of standard illuminants. The most important of these
are:
• Standard illuminant A, which corresponds to an incandescent light bulb (Planck-radiator
with a temperature of 2854 K),
• Standard illuminant C, which corresponds to daylight on the Northern hemisphere, but
which is becoming accepted as representative of indoor daylight,
• Standard illuminant D65, which corresponds to average daylight (Planck radiator with a
temperature of 6500 K) which contains considerably more energy in the ultraviolet region
than the C illuminant.
The relative spectral power distributions of the standard illuminants A, C and D65 are given
in Figure 5.6.

Relative power distribution

C A

100

D65

0
300 400 500 600 700
Wavelength, nm

Figure 5.6. The spectral power distributions of CIE standard illuminants A, C and D65.
156

RKT==pí~åÇ~êÇ=lÄëÉêîÉêë

Experiments carried out to determine how humans react to light and how the human brain inter-
prets the stimulus of electromagnetic energy in a specific limited wavelength region in order to
perceive colour have led to the concept of the CIE Standard Observer.
In the first place, it has been shown that humans have a sensitivity to light which is a maxi-
mum at a wavelength of about 550 nm and drops according to a roughly bell-shaped curve to
zero at approximately 400 nm and 700 nm. This curve defines what is meant by visible radia-
tion, and is known as the V(O) curve. The region at wavelengths shorter than 400 nm is the ul-
tra-violet (UV) region and the region at wavelengths longer than 700 nm is the infra-red (IR)
region. It can thus be said that the eye functions in this respect as a broad-band filter with an ef-
fective wavelength of 557 nm.
In the second place it has been shown that the interpretation of colour can be described by a
combination of three separate sensitivities in the red, green and blue regions of the spectrum. In
fact, it has later been shown that the retina of the eye does indeed possess three different types
of cones sensitive to different spectral regions. It can thus be said that the eye functions in this
respect as a combination of three broad-band filters described by the x (O ) , y (O ) and z (O )
functions, where y (O ) is identical to the V(O) function. These functions are illustrated in Fig-
ure 5.7.

2
_ CIE 1931 (2¡)
z (l)

_
y (l)
1
_
x (l)

0
400 500 600 700
Wavelength, nm

Figure 5.7. The CIE Tristimulus functions representing the colour vision of the human eye

The 1931 Standard Observer was defined in relation to experiments where the half-angle of
the viewing cone was only 2°. Later, the 1964 Standard Observer was defined in relation to the
results of experiments carried out with a wider viewing cone having a half-angle of 10°. The
change in viewing angle makes a slight difference to the human perception of colour.

RKU==pí~åÇ~êÇáòÉÇ=jÉ~ëìêÉãÉåí=`çåÇáíáçåë=Efpl=OQSVF

For the measurement of the optical properties of pulps and paper, it has been agreed within the
International Standardization Organization (ISO) in a document with the number ISO 2469 that
measurement shall take place:
 157

• with diffuse illumination and normal observation (d/0°)

• with CIE illuminant C
• according to the CIE 1931 (2°) standard observer
• without the gloss component
In the measurement of whiteness (see 48.15), the CIE illuminant D65 is sometimes used to-
gether with the CIE 1964 (10°} standard observer

RKUKN==fåëíêìãÉåíë=`çåÑçêãáåÖ=íç=fpl=OQSV

The International Standard was originally written in relation to an instrument made by Zeiss and
marketed under the name Elrepho. During recent years, new instruments have appeared. The
first instruments to replace the Zeiss Elrepho were the Datacolor Elrepho 2000 and the Techni-
dyne microTB1C instruments. A third generation of instruments has now entered the market
with the Datacolor Elrepho 3000, the L&W Elrepho, the Technidyne ColorTouch and the Mi-
nolta CM-3630.
These instruments all use a sphere to provide a diffuse illumination of the test piece and the
test piece is viewed by a suitable detector through an opening in the sphere immediately oppo-
site to the test piece opening, Figure 5.8. The sphere is internally lined with a matt white mate-
rial such as barium sulphate, and the lamps are screened so that direct light from the lamps can
fall neither on the detector nor on the test piece. It is thus only diffuse light reflected from the
inside of the sphere which strikes the sample, and only the light reflected normal to the surface
which reaches the detector.
In the early instruments, glass filters corresponding to the different sensitivity functions of
the eye and the desired illuminant were used. The Datacolor E2000 introduced a new technolo-
gy with an array of diodes which simultaneously detect the reflected radiation at 16 different
wavelengths at intervals of 20 nm between 400 and 700 nm. Functions corresponding to the
chosen illuminant and to the sensitivity of the eye are then applied in the software where the re-
sults are computed. In the third generation instruments, measurements are made at intervals of
10 nm.
receptor

d/0° sphere
gloss trap

lamps

baffles

sample

Figure 5.8. Diagram showing the essential features of the instruments used within the pulp and paper industry for
optical measurements.
158

RKUKO==`~äáÄê~íáçå=çÑ=íÜÉ=jÉ~ëìêÉãÉåí=fåëíêìãÉåíë

It is an important feature of such instruments that they must be calibrated. A spectral (wave-
length) calibration and a photometric (reflectance factor scale) calibration are required.
The manufacturer is responsible for the spectral calibration, i.e. that the measurement takes
place at the indicated wavelengths, or that some type of correction is made for any wavelength
deviations.
The user is responsible for the photometric calibration, i.e. a decision has to be made as to
what is meant by 100 per cent and the instrument has to be adjusted on a daily basis so that it
gives a correct reading at 100 per cent. The instrument usually also has to be calibrated at zero
per cent. Sometimes, consideration must be given to the fact that the instrument does not neces-
sarily measure linearly across the whole range between 0 and 100 per cent.
The absolute photometric scale is based on the perfect reflecting diffuser. This is a concept
rather than a physical reality, but several laboratories are able to carry out measurements in rela-
tion to this concept and they have also been shown to be in reasonable agreement with each oth-
er. (Before this absolute scale could be determined, MgO was used as 100 percent.) As a zero
standard, a hollow cavity lined internally with black velvet is usually used.

RKUKP==`~äáÄê~íáçå=ïáíÜáå=íÜÉ=m~éÉê=fåÇìëíêó=Ó=íÜÉ=fplJeáÉê~êÅÜó

For the calibration of instruments for the measurement of paper and pulp, a hierarchy has been
established by the ISO organisation. This consists of (a) three Standardizing Laboratories
(NIST, PTB, NRCC) and (b) a number of Authorized Laboratories, which supply reference
standards of different levels.

qÜÉ=éÉêÑÉÅí=êÉÑäÉÅíáåÖ=ÇáÑÑìëÉê=EiÉîÉä=NF
p
pq^ka^oafwfkd=i^_lo^qlofbp
p
oÉÑÉêÉåÅÉ=ëí~åÇ~êÇë=çÑ=iÉîÉä=O=EfoOF
p
^rqelofwba=i^_lo^qlofbp
p
oÉÑÉêÉåÅÉ=ëí~åÇ~êÇë=çÑ=iÉîÉä=P=EfoPF
p
fkarpqof^i=i^_lo^qlofbp

Each Authorized Laboratory maintains a reference instrument which is calibrated with a ref-
erence standard from one of the Standardizing Laboratories. A special type of tablet is used as
IR2 reference standard. To achieve maximum agreement with each other, the different Autho-
rized Laboratories have now reached an agreement to use the same type of tablet for this cali-
bration transfer. 300 IR3 reference standards are sent each month from the Authorized
Laboratories to different pulp and paper mill laboratories. These reference standards consist of
an optically stable, non-fluorescent, cotton-based paper.
 159

RKV==léíáÅ~ä=jÉ~ëìêÉãÉåíë

The reflectometer does not measure the total reflectance, i.e. the proportion of the incident light
reflected from the sample. Instead, it detects the reflected light only within a narrow cone and it
measures the reflectance factor, which is defined as the ratio of the reflectance of the sample to
the reflectance of the perfect reflectance factor under the same conditions of illumination and
detection.
The reflectance factor from an optically infinitely thick pad of sheets is called the reflectivity
of the material and is designated Rf. In practical terms, this means that the pad must be chosen
to be sufficiently thick so that a further thickness increase does not influence the result.
The reflectivity can be measured at different wavelengths. In the modern instruments, the
light reflected from the sample passes through a monochromator which scatters the light across
a number of diodes which make measurements at 400, 420, …, 700 nm.
From the spectral measurement values, the tristimulus values, X, Y and Z, of the material can
be calculated. These three values correspond to the function of the eye in observing colours (see
Section 5.13).
The Y-value alone is a quantity which is adapted to the sensitivity of the eye to light and the
Y-value is used alone as a measure of the luminance of the material as it would be perceived by
an average observer without colour vision. The Y-function covers the whole visible range but it
has a maximum effective wavelength of 557 nm. This reflectance factor value is often designat-
ed Ry.
A property of particular interest for pulps, but which is also used as a characteristic of paper,
is the ISO brightness value. The origin of this property lies in the arbitrary choice of a filter to
permit an efficient control of the bleaching process. Since the natural colour of pulp lies in the
yellowish range, a measurement in the blue range is more sensitive than the Y-value to how the
natural colour of the pulp is removed during the bleaching. The brightness is therefore measured
with a filter or calculated with a mathematical function having an effective wavelength of
457 nm, and is often designated R457.
The essential differences between the ISO brightness filter and the Y-value filter (based on
the V(O) function) are shown in Figure 5.9. In this figure, the two filters are shown on relative
scales with the peaks at a value of 1.0. To show the advantages and limitations of these two fil-
ters, typical curves for a blue and a red paper sample are also shown in the figure. The bright-
ness filter is sensitive in the blue region of the spectrum and thus gives a high brightness value
to a blue sample. The Y-value filter, which stretches over and defines the full visual range, has
its maximum sensitivity in the mid-spectral range and is insensitive in both the blue and red re-
gions of the spectrum.
The reflectance factor from a single sheet against a black background is designated R0. An
important quantity for printing paper is the ratio between R0 and Rf, weighted with respect to
the Y-filter and the C illuminant, which is the opacity of the material:

R0
opacity (5.20)
Rf
The opacity of a material is a measure of the ability to prevent the reading of a text from be-
ing disturbed by the text on an underlying sheet. Note that Rf is a material property, whereas R0
is a single sheet property and is dependent on the grammage.
160

brightness V (l)
1.0
red

blue

0.0
400 500 600 700
Wavelength, nm

Figure 5.9. The characteristics of the ISO-brightness and Y-value (V(O)) filters. Typical reflectance factor curves
for a blue and a red sample are also shown to illustrate the advantages and limitations of the two filters

These various properties are listed in Table 5.1. The measurement and calculation of colour
coordinates and whiteness values are discussed in more detail in later sections.

q~ÄäÉ=RKNK=aáÑÑÉêÉåí=çéíáÅ~ä=éêçéÉêíáÉëK
mêçéÉêíó póãÄçä aÉëÅêáéíáçå
oÉÑäÉÅíáîáíó of êÉÑäÉÅí~åÅÉ=Ñ~Åíçê=çÑ=~=é~Ç=çÑ=íÜÉ=ã~íÉêá~äI=ëç=íÜáÅâ=íÜ~í=~å=
áåÅêÉ~ëÉ=áå=íÜáÅâåÉëë=ÇçÉë=åçí=~ÑÑÉÅí=íÜÉ=î~äìÉ
oÉÑäÉÅí~åÅÉ=Ñ~Åíçê=çîÉê= oM êÉÑäÉÅí~åÅÉ=Ñ~Åíçê=çÑ=~=ëáåÖäÉ=ëÜÉÉí=çîÉê=~=åçåJêÉÑäÉÅíáåÖ=Ää~Åâ=
Ää~Åâ Å~îáíó
lé~Åáíó t î~äìÉ=Å~äÅìä~íÉÇ=~ë=NMM=íáãÉë=íÜÉ=ê~íáç=çÑ=RM=ÇáîáÇÉÇ=Äó=Rf
fpl=ÄêáÖÜíåÉëë oQRT êÉÑäÉÅí~åÅÉ=Ñ~Åíçê=~í=~å=ÉÑÑÉÅíáîÉ=ï~îÉäÉåÖíÜ=çÑ=QRTåã
vJî~äìÉI=äìãáå~åÅÉ oó ~=ãÉ~ëìêÉ=çÑ=íÜÉ=äáÖÜíåÉëë=çÑ=íÜÉ=ã~íÉêá~ä=~í=~å=ÉÑÑÉÅíáîÉ=ï~îÉJ
äÉåÖíÜ=çÑ=RRTåã
qêáëíáãìäìë=î~äìÉë uI=vI=w êÉÑäÉÅí~åÅÉ=Ñ~Åíçêë=áå=íÜÉ=êÉÇI=ÖêÉÉå=~åÇ=ÄäìÉ=êÉÖáçåë=ïÜáÅÜ=
íçÖÉíÜÉê=ÇÉÑáåÉ=íÜÉ=Åçäçìê=çÑ=~=ã~íÉêá~ä
`Üêçã~íáÅáíó=ÅççêÇáJ ñI=ó ÅççêÇáå~íÉë=ÇÉêáîÉÇ=Ñêçã=XI=YI=Z=ïÜáÅÜ=íçÖÉíÜÉê=áåÇáÅ~íÉ=íÜÉ=
å~íÉë ÅÜêçã~íáÅáíó=EÅçäçìê=ÅçåíÉåíF=çÑ=íÜÉ=ã~íÉêá~ä
açãáå~åí=ï~îÉäÉåÖíÜ= OaI=éb î~äìÉë=ÇÉêáîÉÇ=Ñêçã=íÜÉ=ÅÜêçã~íáÅáíó=ÅççêÇáå~íÉëI=áåÇáÅ~íáåÖ=íÜÉ=
~åÇ=ëéÉÅíê~ä=éìêáíó ãçåçÅÜêçã~íáÅ=ï~îÉäÉåÖíÜ=ïÜáÅÜ=ãçëí=ÅäçëÉäó=ÅçêêÉëéçåÇë=íç=
íÜÉ=Åçäçìê=çÑ=íÜÉ=ã~íÉêá~ä=~åÇ=íÜÉ=Åçäçìê=ëíêÉåÖíÜ
`fbi^_=î~äìÉë iGI=~GI=ÄG íê~åëÑçêã~íáçåë=çÑ=íÜÉ=XI=YI=Z=î~äìÉë=íç=ÖáîÉ=ãÉ~åáåÖÑìä=î~äìÉë=áå=
~=íÜêÉÉJÇáãÉåëáçå~ä=Åçäçìê=ëé~ÅÉ
tÜáíÉåÉëë t íê~åëÑçêã~íáçå=çÑ=íÜÉ=XI=YI=Z=î~äìÉë=íç=ÖáîÉ=~=ãÉ~åáåÖÑìä=î~äìÉ=Ñçê=
íÜÉ=ïÜáíÉåÉëë=çÑ=íÜÉ=ã~íÉêá~ä
qáåí qï íáåí=î~äìÉ=~ëëçÅá~íÉÇ=ïáíÜ=íÜÉ=ïÜáíÉåÉëë=î~äìÉ=íç=áåÇáÅ~íÉ=ïÜÉíÜÉê=
íÜÉêÉ=áë=~åó=Åçäçìê=ÇÉîá~íáçå=íçï~êÇë=íÜÉ=ÖêÉÉå=çê=êÉÇ
 161

RKNM==`çäçìê=mÉêÅÉéíáçå

A colour percept arises in the brain through a complicated process, when electromagnetic radia-
tion meets the retina of the eye. Electromagnetic radiation having wavelengths between 400 and
700 nm, in daily speech called „light“, stimulates the light-sensitive cones of the retina.
Through a series of physiological processes in the cones and other cells of the retina, analogue
potentials arise which are transformed to electrical pulses in the ganglion cells of the retina and
these are transmitted via the optic nerve further to the optic centre of the brain. Through a psy-
chological process, a colour percept is created. Colour is thus the result of a physical-physiolog-
ical-psychological process initiated when light is reflected from the test material and stimulates
the colour-sensitive cones and light-sensitive rods of the retina.
We can describe the perceived colour sensation as blue, green, yellow, or red or as a mixture
of these hues. We can also characterise the colour as light or dark, as strongly or slightly chro-
matic etc. Any attempt to order colours thus requires a three-dimensional system involving attri-
butes such as hue, chromaticity and lightness.
It is not possible to quantify the physiological and psychological processes which relate a
colour percept to its physical stimulus through an explicit mathematical function. The physical
stimulus and the percept produced by it can, however, be linked by an empirical psychophysical
relationship. The CIE standard observer functions referred to in Sections 5.7 and 5.11 are exam-
ples of such relationships, but they are tabulated functions and not mathematical formulae.

RKNN==qêáëíáãìäìë=s~äìÉë

In 1852, Grassman established that any colour percept can be matched by a mixture of three pri-
mary colours in suitable proportions, on condition that they have been chosen so that a mixture
of two of these primary colours cannot give the third. The system adopted by CIE (Commission
Internationale de l'Eclairage, the International Lighting Commission) in 1931 for colorimetry
(colour measurement) is based on Grassman's law and this leads to the three spectral colour
matching functions x (O ) , y (O ) and z (O ) upon which the CIE system for colorimetry is
based. These functions are shown in Figure 5.7.
For each wavelength, O the products of these functions, the reflectance factor R(O of the
sample and energy distribution of the illuminant S(O at that wavelength give the proportions
with which the three functions contribute to the total appearance at the wavelength in question.
The whole contribution is given by the summations of these products over the visual range:

700
X k ¦ S ( O ) R (O ) x (O ) (5.21)
400

700
Y k ¦ S (O ) R (O ) y (O ) (5.22)
400

700
Z k ¦ S (O ) R (O ) z (O ) (5.23)
400
162

Where k is a normalization factor:

100
k 700
(5.24)
¦ S (O ) y (O )
400

This gives the Y-value = 100 for the perfect reflecting surface. The values X, Y and Z are
called the tristimulus values of the colour.
Since the functions included in Equations 5.21–5.24 lack an explicit mathematical form, a
summation procedure must be used, based on tabulated values published by CIE for these illu-
minant and colour matching functions.

RKNO==`Üêçã~íáÅáíó=`ççêÇáå~íÉë

The three tristimulus values give an unambiguous definition of the colour which they represent,
but they are difficult to interpret. E.g., the two following sets of figures represent the same hue
and the same chromaticness, since the three virtual colours are included in the same propor-
tions.

(X, Y, Z) = (10, 20, 30) (5.25)

and

(X, Y, Z) = (25, 50, 75) (5.26)

The second group apply to a much lighter colour than the first group, however. The informa-
tion is better displayed by a transformation of the data , such that:

X Y Z
x y z (5.27)
X Y  Z X Y  Z X Y  Z

Since x + y + z = 1, z is defined by the values x and y, and the three tristimulus values can
thus be replaced by two values, x and y.
If it is desired to represent colours graphically by their chromaticity coordinates, this is done
in the CIE xy-space. an orthogonal coordinate system with x and y as coordinates. In this two-di-
mensional coordinate system, the point (1,0) indicates the colour X, the point (0,1) indicates the
colour Y and the origin (0,0) indicates the colour Z. Figure 5.10 shows this coordinate system.
The horseshoe-shaped curve is the locus of all the pure spectral colours. The straight line join-
ing the extremities of this curve is the purple line, and all real colours are limited to the enclosed
area.
It appears here that the representation of colour is reduced to a two-dimensional plane, but
this is not the case, the diagram is completed by adding the vertical direction of lightness, the Y-
value, as shown in Figure 5.11. The dimensions x and y are called the chromaticity coordinates
 163

of the colour and are together a measure of the colour content of the material while the Y-value
indicates its luminance (lightness).

y 520
CIE 1931
0.8
510 540

0.6
500
580

0.4
600
490
760
0.2
480

470
380
0.0
0.0 0.2 0.4 0.6 0.8 x

Figure 5.10. The CIE chromaticity diagram.

E Y

100

.8
«
»C

y 0
.7

x
0

Figure 5.11. The CIE chromaticity diagram in three dimensions, showing the peak at Y = 100.

RKNP==açãáå~åí=t~îÉäÉåÖíÜ=~åÇ=bñÅáí~íáçå=mìêáíó

Even though the use of chromaticity coordinates facilitates the interpretation of colour values, it
is difficult to understand what colour is represented by any pair of values (x, y). To give a better
understanding of what these numbers mean, a transition can be made from the rectangular coor-
dinate system to a polar coordinate system. With the neutral point defined by the coordinates of
the illuminant (xn, yn), as origin, a straight line is drawn through the point (x, y) representing the
colour towards the locus for the spectral colours. All colours which lie on this straight line are
164

said to have the same dominant wavelength, which is given by the wavelength where the line in-
tersects the locus for the spectral colours. The ratio of the distance between the coordinates (xn,
yn) and (x, y) to the distance between (xn, yn) and the point of intersection of the line with the
spectral locus is called the excitation purity of the colour. A chromaticity diagram showing
these features is presented in Figure 5.12. In this figure the newsprint has a dominant wave-
length of 575 nm and an excitation purity of 10 %.
For various reasons, this representation is becoming obsolete. One reason is that the physical
quantities „dominant wavelength“ and „spectral purity“ do not correspond to a sufficiently ac-
ceptable extent to the visual characteristics of „hue“ and „chromaticness“. Colours having the
same dominant wavelength do not, for example, have the same visual hue.
Another disadvantage of this diagrammatic representation of colours is that the mathematical
distances corresponding to the same visual difference between two colours are different in dif-
ferent regions of the diagram. Although the system gives an unambiguous description which has
been valuable in e.g. the standardization of the shade of newsprint, the system is now obsoles-
cent and it is being replaced by the CIELAB coordinate system described in the next section.
The neutral point indicated as (xn, yn) is not the equi-energy point (x = 0.333, y = 0.333). It is
the point representing the colour characteristics of the illuminant/observer combination chosen.
In Figure 5.12, the neutral point represents the C/2° situation. Neutral point data for different il-
luminant/observer combinations are given in Table 5.2.

y 520
CIE 1931
0.8
510 540

0.6
dominant wavelength
500
580

excitation
0.4 purity
600
newsprint
490
C(x,y)
760
0.2
480

470
380
0.0




0.0 0.2 0.4 0.6 0.8 x
Figure 5.12. The CIE chromaticity diagram, showing for a sample of newsprint how the dominant wavelength
and excitation purity are defined.

q~ÄäÉ=RKOK=m~ê~ãÉíÉêë=Ñçê=íÜÉ=åÉìíê~ä=éçáåíë=çÑ=ÇáÑÑÉêÉåí=áääìãáå~åíLçÄëÉêîÉê=ÅçãÄáå~íáçåëK

fääìãáå~åíLçÄëÉêîÉê ^LOø `LOø aSRLNMø

Xå NMVIURM VUIMTQ VQIUNP
Yå NMMIM NMMIM NMMIM
Zå PRIRUR NNUIOPO NMTIPMQ
xå MIQQTS MIPNMN MIPNPU
yå MIQMTQ MIPNSN MIPPNM
 165

RKNQ==`fbi^_J`ççêÇáå~íÉë

In an attempt to establish a system which better mirrors the perceptual experience of colours
and colour differences, a new way of mathematically describing different colours has been stan-
dardized by CIE in the CIELAB-system. This system is a non-linear transformation of the X, Y,
Z-values into a three-dimensional system where L* corresponds to a lightness value, a* is the
red/green axis and b* is the yellow/blue axis, according to:
L* = 116 (Y/Yn)1/3 – 16 (5.28)

a* = 500 [(X/Xn)1/3 – (Y/Yn)1/3] (5.29)

b* = 200 [(Y/Yn)1/3 – (Z/Zn)1/3] (5.30)

Here, Xn, Yn, Zn are tristimulus values for the neutral point, which is dependent on the select-
ed illuminant and observer conditions. Tristimulus values of the neutral point for different illu-
minant/observer combinations are given in Table 5.2.

L*
lightness

yellow
b*

green –a* a* red

–b*
blue

Figure 5.13. A representation of the CIELAB colour space.

This transformation leads to a more or less uniform colour space, as shown in Figure 5.13.
To a first approximation, the colour difference ǻE between two different colours can be calcu-
lated as the distance between two points in this CIELAB-space, according to:

'E ' L 2  ' a 2  ' b 2 (5.31)

RKNR==tÜáíÉåÉëë

Many different attempts have been made to develop a way of mathematically indicating the
whiteness of a material. Variations in whiteness are colour differences within a limited range of
166

colours which are regarded as almost white. The brightness value is not sufficient as a measure
of whiteness since it takes into consideration only the reflectivity within the blue region.
It is clear that whiteness is in some way a perception of colour and that any attempt to mea-
sure the whiteness of a material must be based on a full colour measurement over the whole
spectral range, i.e. the tristimulus values. Many different mathematical expressions have been
proposed which relate a whiteness parameter to the tristimulus values of the material. Eventual-
ly, a committee established within CIE presented an equation for the so-called CIE-whiteness W
calculated as:

W = Y – 1700(y – yn) – 800 (x – xn) (5.32)

This is complemented by a calculation of a red/green tint value Tw according to:

Tw = 650(y – yn) – 900 (x – xn) (5.33)

The conditions for a material to be white in the application of these equations are:

40 < W < (5Y – 280) (5.34)

and

– 3 < Tw < +3 (5.35)

There is clearly a problem in trying to describe whiteness by a single value, since this con-
flicts with the understanding that the brain's interpretation of a colour sensation is always three-
dimensional, but within the region where the concept of whiteness is meaningful, the equation
is becoming accepted.
Methods have now been developed for measuring the whiteness of paper by applying Equa-
tion 5.32 to calculate a whiteness value from the measured tristimulus values.
An important feature of this equation is that it recognises that in the subjective human appre-
ciation of whiteness there is a preference for a bluish rather than a yellowish tint. Traditionally
in the textile industry, and also in the paper industry, the sensation of whiteness has often been
strengthened by the addition of a small amount of a blue dye. More recently, the emphasis has
been on the addition of a fluorescent, optical whitening agent, as is discussed in the next sec-
tion.

RKNS==cäìçêÉëÅÉåÅÉ

To produce a higher whiteness than can be attained simply by bleaching or by the addition of a
blue dye, a fluorescent whitening agent (FWA) is added to certain products, particularly certain
types of fine paper and graphic arts board. This material absorbs energy within the UV-range
and re-emits it within the blue visual range at 430–470 nm. This increase in the reflectance fac-
tor (radiance factor) in the blue range is observed by the eye as an increase in the whiteness of
the material. Figure 5.14 shows spectral curves for a non-fluorescent paper, for a paper contain-
ing a blue dye and for a paper containing a fluorescent whitening agent.
 167

Radiance factor, %
120

FWA

100

80
blue dye

60
400 500 600 700
Wavelength, nm

Figure 5.14. Spectral radiance factor curves for a near-white paper, a paper with added blue dye and a paper with
added FWA.

Table 5.3 shows calculated data for the curves shown in Figure 5.14.

q~ÄäÉ=RKPK=bÑÑÉÅíë=çå=çéíáÅ~ä=éêçéÉêíáÉë=çÑ=~ÇÇáåÖ=ÄäìÉ=ÇóÉ=çê=ct^K
mêçéÉêíó åçêã~ä ÄäìÉ=ÇóÉ ct^
YJî~äìÉ=E`LOøF UVIQ UOIR VMIO
fpl=ÄêáÖÜíåÉëë UUIN UTIP VUIQ
`fb=ïÜáíÉåÉëë=E`LOøF UQKN VUKP NNUIN
Here it can be seen that a blue dye lowers the Y-value and the ISO-brightness but increases
the CIE-whiteness value. The addition of FWA has a negligible effect on the Y-value but leads
to an increase in both the ISO-brightness and CIE-whiteness values
The use of a fluorescent additive gives rise to measurement problems, however, since the re-
sult of the measurement is to a great extent influenced by the UV-content of the illumination
provided by the instrument lamps. To solve this problem, modern instruments include an adjust-
able UV-cut-off filter, the position of which can be adjusted so that the relative UV-content in
the illumination falling on the sample can be made to match that of any desired illuminant.
The Authorized Laboratories referred to in Section 5.8.3 therefore also provide fluorescent
reference standards with assigned whiteness values corresponding to the desired illuminant to
enable industrial laboratories to adjust the filters in their instruments to the correct positions.
In the initial work when the CIE-whiteness equation was being developed, it was assumed
that the D65-illuminant was the most suitable since this has a large UV-content and greatly ex-
cites the fluorescent whitening agent, and CIE-whiteness (D65/10°) is indeed now an interna-
tionally standardised property.
However, the UV-content in this illuminant exceeds the level which exists in most indoor sit-
uations where white paper is used, so that an adjustment of the UV-content to match the C-illu-
minant is now recommended for indoor conditions, and the concept of indoor whiteness has
also been internationally approved, based on the measurement of CIE-whiteness (C/2°). These
are the values given in Table 5.3. Reference standards with assigned values for both these con-
ditions are now available.
100
 169

S==lå=íÜÉ=jÉÅÜ~åáëãë=ÄÉÜáåÇ=íÜÉ=^Åíáçå=çÑ=aêó=píêÉåÖíÜ
====~åÇ=aêó=píêÉåÖíÜ=^ÖÉåíë

Lars Wågberg
Department of Fibre and Polymer Technology, KTH, Stockholm

6.1 Background 169

6.2 Paper Strength as an Adhesion Problem 170

6.3 Influence of Dry Strength Additives on Paper Strength – The More

Traditional View 172
6.3.1 Tensile Properties 172
6.3.2 Compression Properties 178

6.4 References 182

SKN==_~ÅâÖêçìåÇ

The increasing use of recycled fibres and other inexpensive furnish components, such as fillers,
has led to a decrease in the strength properties of the paper produced from theses raw materials.
To overcome this problem it is very common to add different dry strength additives, such as cat-
ionic starch [1–15], modified polyacrylamide [16, 17], polyamideamine epichlorohydrine resins
(PAE) in combination with carboxymethyl cellulose [18, 19], cationic dialdehyde starches [20]
and chitosan [5, 20, 21]. This list can probably be made much more extensive and the references
given here can serve more as examples and an introduction to the use of dry strength additives
in different paper grades. However, in for example testliner manufacture a large amount of
starch is added both in the wet end and in size press applications in order to reach the desired
strength properties of the paper. If for example the size press could be avoided it should be pos-
sible to significantly increase the productivity of the machines. A general understanding of the
mechanism behind the dry strength additives would also be desirable since this would enable an
optimised use of these additives.
Wet strength additives are today mostly used for the production of different grades of hy-
giene papers, filter papers, special grades of board and sack paper [22, 23]. It is also well ac-
cepted that the different chemicals are efficient through different mechanisms but there is still a
debate regarding the molecular mechanism responsible for the wet strengthening action. The
development of new types of wet strength chemicals has been very limited over the last 30 years
but the development of new chemicals for preparation of wrinkle free clothing has spurred a de-
velopment of new chemicals also for the paper industry [24-28]. So far the application of these
latter additives has been limited to laboratory trials but very high relative wet strengths can be
achieved however so far at the cost of folding endurance of the produced papers.
170

SKO==m~éÉê=píêÉåÖíÜ=~ë=~å=^ÇÜÉëáçå=mêçÄäÉã

Regardless of the exact mechanism responsible for the interaction between two fibres it can be
concluded that that the interaction between the fibres during consolidation and drying of the pa-
per will be very important for the strength of the paper. During the consolidation and drying
process the highly swollen fibre surfaces will be pushed together by the capillary forces formed
between the fibres during water removal. This capillary force will deform the external surfaces
of the fibres and an intimate contact will be formed between the fibres in the dry sheet resulting
in the situation schematically described in Figure 6.1.

bonded joint

Figure 6.1. Schematic representation of the bonded joint formed between the fibres during consolidation and
drying.

The strength of this bonded joint will be determined by

1. The molecular contact area in the contact zone
2. The intermolecular forces
3. Mechanical entanglement between opposing surfaces
4. Possible existence of covalent linkages
In order to determine the relative influence of these factors it is necessary to perform model
experiments under well-controlled conditions but this information is unfortunately still lacking.
Traditionally the molecular contact area between the fibres has been determined with light
scattering measurements. Since the dimensions that can be determined with light scattering is
about half of the wavelength of the light it is easy to realise that the molecular contact area can
not exactly be determined with light scattering experiments. This is also one of the reasons why
the other factors in connection with Figure 6.1 not has been determined quantitatively.
A more correct description of the „true“ dimensions in the contact zone between the fibres is
given in Figure 6.2 where both the area that can be determined with light scattering has been
marked together with „true“ contact area. The enlarged picture in Figure 6.2 also demonstrates
that in the area in real contact there might still be areas in real contact and areas where there still
exists a small gap between the surfaces.
The specific joint strength in the partly joined area will be determined by several different
factors and these different factors are described schematically in Figure 6.3.
As was mentioned earlier the mechanical interlocking between the surfaces will contribute to
the joint strength in a type of Velcro® organisation of cellulose fibrils on the fibre surface. An-
other type of interlocking that might occur is the inter-diffusion of polymers across the inter-
face. This inter-diffusion will result in a situation where one part of the polymer resides in the
bulk of one phase and with the other part in the bulk of the opposing phase. The total contribu-
 171

tion to the joint strength from this inter-diffusion will be dependent on the number of molecules
that have crossed the interface and the length of the embedded polymer in each phase. However,
if the migration is such that the embedded length is much smaller in one phase than in the other
the shorter chain will be the limiting factor for the contribution to joint strength. In practical
terms it is likely that short chains of hemicelluloses might migrate across the interface between
the fibres and in this way contribute to the specific joint strength.

partly joined
area area in molecular contact
area not available for
light scattering (d < l /2)

Figure 6.2. Description of which areas that might be detected with light scattering and which areas between the
fibres in a fibre/fibre joint that are in real contact. The inset also shows that there still might be areas in poor con-
tact also in the partly joined area.

partly joined
area area in molecular contact interdiffusion
area not available for
light scattering (d < l /2)

mechanical
induced dopoles interlocking

OH
OH

Cellulose

Figure 6.3. Schematic description of the different types of interactions that will determine the specific joint
strength in the contact zone between the fibres.

The intermolecular forces that was mentioned in conjunction with Figure 6.1 has in Figure
6.3 been divided in three different groups:
1. Hydrogen bonding
2. Van der Waals interactions
3. Ionic bonding
172

There are no real proofs for which of these that will dominate the contribution from the inter-
molecular forces but most likely the hydrogen bonding and the van der Waals forces will be the
dominating factors. Non-specific van der Waal interaction energy, WvdW, between macroscopic
bodies are characterised by the Hamaker constant (A) of the material, are rather long-range and
scales as 1/distance between the surfaces according to Equation (6.1), which is valid for crossed
cylinders of radius R1 and R2.

A  R1 R2  (6.1)
WvdW
6D

Since the force between the crossed cylinders is the distance derivative of the interaction en-
ergy the force between the two cylinders will be

A  R1 R2  (6.2)
FvdW
6D2

Since the Hamaker constant for cellulose is in the range of 6 · 10–20 Nm the van der Waals at-
traction between two crossed cylinders can be calculated at least for the dry state. In order for a
better accuracy the solid cylinders should be exchanged for cylindrical shells.
The hydrogen bonds on the other hand are specific which in turns mean that they are very
sensitive to the distance between the hydrogen bonding groups on adjacent surfaces and also the
geometric orientation between these groups. This latter fact has during the last years led to
doubts about the importance of hydrogen bonding for specific joint strength and hence paper
strength since it is very unlikely that to rough fibre surfaces will come into such close proximity
that hydrogen bonds might develop between the surfaces. However, recently it has in several
publications been shown that the wet transverse elastic modulus of fibres is very low, i.e. in the
order of 1–5 Mpa and maybe even lower. This means that the fibre surface resembles the sur-
face of a soft gel and considering the large capillary forces that will pull the fibre together dur-
ing drying it is not at all unlikely that highly mobile molecules on the fibre-gel surface might
orient to form hydrogen bonds with hydrogen bonding groups on the adjacent fibre-gel surface.
This area is currently in focus for many research groups and new knowledge about the impor-
tance of the different factors composing the intermolecular forces will definitely be available in
the near future.

SKP==fåÑäìÉåÅÉ=çÑ=aêó=píêÉåÖíÜ=^ÇÇáíáîÉë=çå=m~éÉê=píêÉåÖíÜ=Ó=qÜÉ
=======jçêÉ=qê~Çáíáçå~ä=sáÉï

SKPKN==qÉåëáäÉ=mêçéÉêíáÉë

Despite the common use of the dry strength additives there is still no mechanism explaining
how these additives really work. Several attempts have been made over the years and Davison
[29] made a very thorough survey of different mechanisms behind the development of paper
strength in general and more specifically how chemicals may affect the strength development.
 173

However, in [29] no conclusions regarding any specific mechanism were given. In another
communication [30] Davison suggested that the weak link in paper strength was the fibre-to-fi-
bre bond since the fibre strength was at least twice as large as the dry strength of papers com-
posed of these fibres. This is illustrated in Figure 6.4, taken from [30], where data show that
despite the fact that a sheet was believed to be „fibre strength limited“ it was possible to pull out
a large amount of intact fibres from these sheets. The conclusion drawn by Davison [30] was
that the fibre-to-fibre bond should be increased in order to improve the paper strength.

A B C
T T T
50
6

E E E
29 fibers intact 35 fibers intact 56 fibers broken
fibermat test specimen
(arrows indicate application of load)

summary of results from fiber extraction experiments

from weak handsheets from strong handshee
exposed fiber length  1.1 mm (tensile strength = (tensile strength =
7–11 lb/in.) 15–21 lb/in.)
intact broken intact broken
number of fibers extracted 30 15 34 41
fiber length extracted (mm) 1.45 0.51 1.11 0.44
peak resistance to extraction (g) 5.0 5.6 6.9 7.8
effective extracted length (mm) 0.73 0.34 0.70 0.28
work involved (g-cm) 0.127 0.090 0.153 0.094
work-to-break (g-cm/cm length) – 0.626 – 0.702

Figure 6.4. Results taken from [30] showing that despite the fact that sheets are strong it is still possible to pull
out a large number of intact fibres from the sheets indicating that the joint strength between the fibres is the limit-
ing factor also in these sheets.

In later work [8, 31], where the „shear bond strength“ between fibres was determined with
the aid of the Page theory [32] which relates the paper strength to the fibre strength and the
„shear bond strength“ between the fibres according to Equation (6.3).

1 9 12 C
 (6.3)
T   
8 Z P ˜ L ˜ b ˜ RBA 

where
T = Tensile strength of the paper
Z = Zero span of fibres
C = Fibre coarseness
P = Fibre perimeter
b = Shear bond strength
RBA = Relatively Bonded Area
174

By pressing sheets to different densities with and without cationic starch and inserting rele-
vant values in equation (6.3) it was also found that cationic starch [5] and PAE (polyamideam-
ine epichlorohydrine) polymers increased the paper strength by increasing the „shear bond
strength“ between the fibres. This procedure demands that the RBA can be determined in order
to extract the shear bond strength from the slope in a plot of 1/T as a function of 1/RBA. This en-
tity is determined by measuring the light scattering from sheets with zero strength (S0) (extrapo-
lated values) and from sheets pressed to different densities (S). These values are then used to
calculate RBA according to Equation (6.4).

S0  S
RBA (6.4)
S0

Considering the distance over which the van der Waal forces are active according to equa-
tions (6.1–6.2) it is easy to realise that the light scattering can not be used to determine a molec-
ular contact area necessary to calculate a correct RBA and the Page theory for paper strength is
based on the assumption that there is a direct relationship between the real molecular contact
area and the area detected by the lightscattering experiments. The use of this theory to calculate
a „true“ shear bond strength between fibres has been and is being questioned and it is mentioned
here not to support or to question its applicability but more since it is the only semi-quantitative
theory available today to link paper strength to fibre strength and fibre properties.
Regarding the action of cationic starch on paper strength it has, on the contrary, also been
suggested [10] that the cationic starch is efficient through the creation of an increased number
of „bonds“ between the fibres. This conclusion was drawn from experiments where air-dried
and freeze-dried sheets with starch were compared. It was shown that the relative increase in
strength was larger for the freeze-dried sheets and this was taken as a „proof“ for the formation
of new „bonds“. This is illustrated in Figure 6.5 taken from [10].

80 300
70
burst factor increase (%)

60 air dried
200 freeze dried
50
burst factor

40
30 freeze dried 100 air dried
20
10
0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
percent cationic starch addition percent cationic starch addition

Figure 6.5. By adding cationic starch to fibres and forming sheets that were subsequently either air dried or
freeze dried it was found that since the increase in strength, in %, was higher for the freeze dried sheets it could be
concluded that the starch could form new joints between the fibres. From [10].

This formation of new „bonds“ was directly linked to the cationic nature of the starch since a
native starch did not give the same strength response. It was however also stated in [10] that the
starch can be efficient through reinforcement of already existing bonds.
 175

The influence of hydrogen bonding on paper strength was investigated in a pioneering work
by McKenzie [32]. By acetylating 20 % of the hydroxyl groups of the fibres it was found that
the strength of papers formed from water was decreased to 10 % of the strength of the unmodi-
fied fibres. If paper was formed from the modified fibres in acetone it was found that sheets as
strong as the sheets from unmodified fibres were formed in water, this is shown in Figure6.6.
McKenzie [33] concluded that hydrogen bonding was important for the formation of strong fi-
bre/fibre joints in water whereas in acetone the mechanism behind the creation of paper strength
could be related to“..mechanical interlacing of fibres and cellulose chains to an extent depend-
ing on the degree of swelling of the fibres.“

10 10

9 9

8 8

7 7
tenacity (106 dyn)

tenacity (106 dyn)

6 6

5 5
acetone
4 limited 4 full
restraint restraint
3 3

2 2
acetone
water water
1 1

0 0
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
hydroxyl groups substituted (%) hydroxyl groups substituted (%)

Figure 6.6. By acetylating 20 of the hydroxyl groups of the fibres McKenzie [33] found that the strength of paper
made from water decreased to 10 % of the value achieved for unmodified fibres whereas the strength of papers
from modified fibres from acetone was even stronger than the original paper made from water.

McKenzie also introduced the concept of looking at the fibre/fibre joint as an adhesion prob-
lem and interdiffusion of surface polymers of the fibres into adjacent fibres to increase the
strength of the fibre/fibre joint [34]. In a later work by Pelton et al [35] the compatibility of sur-
face polymers, adsorbed to the fibres, on the formation of strong fibre/fibre joints was shown. It
was found that when the compatibility of the surface polymers was low, despite full water solu-
bility of the polymers, there was a clear decrease in the joint strength between the fibres. The in-
fluence of the outermost layers of the fibres on the formation of strong fibre/fibre contacts has
also been shown in recent work [36, 37]. All these results show the large potential of elucidating
and optimising the molecular contact area and the bond strength in the molecular contact zone.
In Figure 6.7 the effect of the application of polelectrolyte multilayers to fibres on paper
strength is shown [36]. The fibres were prepared with layers of cationic polallyalmine and an-
ionic polyacrylic acid before sheet preparation and as can be seen the application of 4.5 bilayers
176

of polyelectrolytes was sufficient to increase the dry strength of the paper about three times. As
can be seen there is a dependence of which polymer that is fixed in the outer layer of the multi-
layer and this probably depends on the structure of the polyelectrolyte multilayer. There is also
an effect on the wet strength of the paper but this will be discussed in Part II of this chapter.

Influence of polymer multilayers, no heat

60

50 PAH in outer layer

Dry properties
Wet properties
Tensile index (kNm/kg)

40

30

20

PAA in outer layer

10

0
0 1 2 3 4 5 6 7 8 9 10
Layer number

Figure 6.7. Influence of treatments of fibres with polyelectrolyte multilayers on the achieved paper strength. The
cationic polymer was a polyallylamine and the anionic polymer was a polyacrylic acid [36].

For the use of starch in clay filled sheets Lindström et al. [1] introduced the stress concentra-
tion minimisation (SCM) theory in order to explain how starch will prove the strength. In this
theory it is first stated that the introduction of filler produces stress concentrations in the sheet,
which subsequently leads to breakage of the sheet. As described in [1] the lifetime of a paper
strip under load can be described by the Zhurkov theory for lifetime of materials and in a sim-
plified version the relationship between lifetime, load and the stress concentration factor can be
written as

ªU  j V º
« 0 U»
tf t0 ˜ exp « » (6.5)
RT
«¬ »¼

where
tf = Life time of the material
t0 = Empirical constant related to the time of an atomic vibration | 10–13 s
U0 = Activation energy for bond rupture
j = Stress concentration factor
V = Applied stress
U = Density of the sheet
 177

By plotting the ln(life time) as a function of applied stress the stress concentration factor can
be determined from slope of the curve in this diagram.
The introduction of starch in the contact point between the filler and the fibres leads to a lo-
cal yielding when the paper is subjected to a stress increase and a global failure of the sheet can
therefore be avoided. This is shown in Figure 6.8 where the stress concentration factor is pre-
sented as a function of addition of cationic starch at different filler concentrations [1].

0 % clay
20 % clay
40 % clay
mol · MN · m

(CD)
kJ · kg

20
stress concentration factor

10

1 2 3 4 5
C-starch add (%)

Figure 6.8. The influence of cationic starch and filler concentration on the stress concentration factor in papers
containing bleached chemical fibres and filler clay [1].

In an extension of this work [7] it was suggested, as a model, that the fibres were covered
with a film of filler particles and that the properties of this film was essential for the strength
upon addition of the starch, the experimental results in [1, 7] showed that the incorporation of
large amounts of starch in the paper could almost entirely overcome the negative effect of large
amounts of filler particles. It was also shown that a wet end addition was more efficient that a
size press addition and this was attributed to the need to have the starch between the fibres in or-
der to have full effect of the added starch.
In summary of this first part it may be concluded that for tensile properties and bending stiff-
ness properties it is fairly safe to suggest that the major mechanism behind the dry strength ad-
ditives is an increase in joint strength between the components in the furnish. It is not clear
though if the additives are able to create new joints between the furnish components sine the
techniques used so far have been too rough to distinguish what is happening in a size range of
150 Å, which is the range at which the surfaces start to interact with each other in deionised wa-
ter. In process waters this distance would be even smaller.
178

SKPKO==`çãéêÉëëáçå=mêçéÉêíáÉë

So far the discussion has almost entirely been focused on graphic papers but it is naturally also
valid for different types of packaging papers. The literature in this field is very limited and rath-
er few papers are published, which treat the combination of these type of papers and dry
strength additives [16–18, 38]. The most important property of these papers is the compression
strength but many of the papers written in this area have been focused on tensile properties or
burst strength and very little, if any, has been written on the mechanism of dry strength addi-
tives in these papers. Furthermore, it is very difficult to find a unified view on the mechanism
behind compression failure in additive free raw materials used for packaging products. Fellers
[39] has given an extensive review of the possible mechanisms and with reference to general
papers on composite materials he suggests that either of the following or combinations of sever-
al of the mechanisms can be responsible for the compressive failure in paper
1. Micro buckling of the fibres, matrix still elastic
2. Matrix yielding followed by micro buckling of the fibres
3. Debonding and interlaminar shear followed by microbuckling of the fibres
4. Shear failure (kink, shear band formation)
5. Separation by transverse tension through the thickness (= delamination)

When examining the compression failure zone in paper the situation in figure 6.9 is usually
found. In this figure, presented by C.Fellers STFI, Stockholm, Sweden, it can be seen that there
is a kink-band formation where the paper structure has been delaminated but it is not possible to
determine whether this caused by a fibre wall damage or a buckling of the segments before the
fibre wall delamination.

Fellers, C STFI, Stockholm, Sweden

Figure 6.9. Delamination zone of paper that has failed under compressive load. An obvious kink band is formed
but the exact mechanism behind the failure is not known.

Paper might be described as a composite material of the components constituting the fibre
wall but it has to be remembered that a paper with a density of 750 kg/m3 still has a void volume
of 50 % and a continuum does not exist. Therefore the similarities to composites have to be
treated with large caution. This fact has been considered, among others, by Seth [40] who con-
 179

cluded that fibre/fibre „bonding“ is very important at low paper densities but as the density in-
creases the influence of the „bonding“ in filler free sheets, on the compression strength,
decreases and at high densities the compression strength of the fibres becomes the limiting fac-
tor. Seth assumed that the „bonding“ was increased as the density increased. It should naturally
be stated that the „bonding“ is as important at high densities as at low densities. This discussion
hence suggests that as the papermaking conditions changes there might be a changeover from
one compression failure mechanism to another.
With reference to the discussion by Davison [30], mentioned above, this conclusion might be
very important for the understanding of the influence of dry strength additives on the compres-
sion strength of paper. The possible changeover between different mechanisms can also be seen
in Figure 6.10 where the compression strength of a paper made from a kraft liner pulp is shown
as a function of paper density with or without addition of different additives (Wågberg, L. Un-
published results). At low densities there is a large influence of the additives but at higher den-
sities all different papers tend to approach the same SCT (Shortspan Compression Strength)
value.
It might be speculated that the joint strength between the fibres has a large influence on pa-
per strength at lower densities whereas the inherent fibre properties will start to dominate at
higher densities. It must also be stressed that these questions are currently under intense investi-
gation since recycled linerboard has started to show such poor quality that new additives and
new insight into the compression failure mechanism and its dependence on additives is needed.

SCT Index Kraftliner

40

38

36
SCT Index (kNm/kg)

34

32

30

28 No additives
26 0.2 % PVAm
24 0.2 % G-PAM
1.5 % G-PAM
22

20
450 500 550 600 650 700 750 800
3
Density (kg/m )

Figure 6.10. Short span Compression Strength of papers made fromk a kraft liner pulp with or without the addi-
tion of different additives. PVAm=Polyvinylamine and G-PAM is glyoxal treated polyacrylamide. As the density
is increased the relative influence of the additives decreases. (Wågberg, L. Unpublished results).

Sachs and Kuster [41] linked the load deformation curve in compression to the behaviour of
cross-sections of the paper as observed in a special device mounted inside a scanning electron
microscope. They concluded that „the failure mechanism appears to involve the formation of
dislocations in the form of proturberances and fissures in the cell walls, the detaching of fibre
wall tissue and separation of microfibrillar bonds (particularly of the S1/S2)“. From these find-
ings they then concluded that it is this separation of S1 and S2 that leads to the delamination of
180

the fibre wall and subsequently to the compressive failure of the linerboard. A typical micro-
graph from their work is shown in Figure 6.11 where the fissures in the fibre wall of fibres in
the compressive failure zone can clearly be detected.

S2
kink band formation formed
by cracks/delamination in the
fibre wall

Figure 6.11. Scanning electron micrograph of a fibre that has failed under compressive load. Fissures in the fibre
wall were taken as an indication that it was the fibre wall delamination that was the initial process behind the
compressive failure of the paper [41].

On the other hand Perkins et al. claimed [42], with support from a theoretical approach and
experimental investigations, that the compressive failure is a localised buckling phenomena and
that the incremental transverse shear modulus is the most important variable for the compres-
sive strength of linerboard. This theory is however very difficult to apply since some of the pa-
rameters in the equations are virtually impossible to determine.
Uesaka [43] has also suggested that the compressive failure is a bending shear buckling phe-
nomena. This was concluded from, among other things, experiments where a relationship be-
tween the interlaminar shear modulus and compressive strength was found. It should though be
mentioned that Uesaka [43] used a buckling equation, which does not take the transverse shear
into consideration and obtained a finite value of the critical load when the slenderness ratio ap-
proached zero. However, a more rigorous analysis, including transverse shear, shows that the
critical load goes to infinity when the slenderness ratio approaches zero.
As is obvious from this short summary there is no clear single mechanism explaining the rea-
son to the compression failure in packaging papers. On the contrary it is the opinion of the au-
thor that it has been demonstrated that there might be several explanations depending on the
papermaking conditions, additives and papermaking raw materials.
Dry strength additives might affect several properties of the fibres that might have of large
influnec on the paper strength. First of all the additives can increase the number of active joints
in the paper. This will decrease average length of the free segments between fibre joints and nat-
urally this will increase the buckling resistance of the paper. Secondly the additives might form
a thin very stiff layer on the fibre surfaces that might increase the buckling resistance of the fi-
bre segment between two fibre/fibre joints. A third possibility is that the additives might block
crevices and inhomogeneities on the fibre wall and in this way remove possible initiation points
 181

cracks that might propagate throughout the fibre wall under compressive load. Finally the addi-
tives might cross link the entire fibre wall enabling the avoidance of a delamination of the fibre
wall when the paper is subjected to compressive loading.
All these mechanisms are schematically shown in Figure 6.12 and if it would be possible to
tailor chemicals that could be specialised to handle a single mechanism of those mentioned in
Figure 6.12 it would also be possible to determine the relative influence of the different mecha-
nisms causing the compressive failure in a paper under compression load.

a) b)

improvement in number of stiff polymer layer on the

active bonds, shorter segments fibre surface
between joints

c) d)

filling of crevices on the crosslinking of the fibre

fibre surface, elimination of wall, prevention of kink
starting points for cracks band formation

Figure 6.12. Different possible influences of additives on the improvement in compression strength of the paper
following addition of dry strength additives. a) Improvement in the number of active joints and hence a decrease
in the efficient segment length betweemn two joints.b) Formation of a stiff polymer layer on the fibre surface that
might improve the buckling load of a single segment. c) Blocking of crevices and inhomogeneities that might be
initiation points for cracks. d) Crosslinking of the fibre wall to prevent fibre wall delamination.

The reason for summarising the possible mechanisms for compression failure of packaging
papers is to draw the attention to the possibilities for improvements with an appropriate chemis-
try. Astonishingly little work has been conducted to investigate if it would be possible to im-
prove the compression strength of the paper through fibre wall strengthening, improved bending
stiffness of the fibre segments and increased joint strength between the fibres.
182

SKQ==oÉÑÉêÉåÅÉë

[1] Lindström, T., Kolseth, P., and Näslund, P.: In V. Punton (Ed.) Papermaking Raw Materi-
als. Proceedings of the Confernce held at Oxford 1985, Mech. Eng. Publ., London, 1985,
pp.589.
[2] Lindström, T., and Florén, T.: Sven. Papperstidn. 87(1984)R97.
[3] Björklund, M.: Lic Thesis, KTH, 1994 .
[4] Retulainen, E., Nieminen, K., and Nurminen, I.: Appita 46(1993)1,33.
[5] Ghosh, A.K.: Appita 47(1994)3,227.
[6] Formento, J.C., Maximino, M.G., Mina, L.R., Sray, M.I., and Martinez, M.J.: Appita
47(1994)4,305.
[7] Rigdahl, M., Lindström, T., and Kolseth, P., in H. Giesekus and M.F. Hibberd (Eds):
Progress and Trends in Rheology. Proceedings of the Second Conf. of European Rheolo-
gists, Prague, June 17–20, 1986, pp.240.
[8] Howard, R.C., and Jowsey, C.J.: J.Pulp Paper Sci. 15(1989)6, J225.
[9] Gaspar, L.A.: Proc. Tappi Annual Meeting, Atlanta, GA, 1982. Tappi Press, 1983,
pp.89–91.
[10] Moeller, H.W.: Tappi 49(1966)5,211.
[11] Lindström, T., Hallgren, H., and Wågberg, L.: Proceedings from the EUCEPA Conference
held in Florence 1986, pp.13:1.
[12] Reynolds, W.F. (Ed.) Dry strength additives. Tappi Press, Atlanta, 1980.
[13] Laleg, M., Pikulik, I.I., Ono, H., Barbe, M.G., and Seth, R.S.: Proceedings from the 77th
Annual Meeting of CPPA/TS, Montreal, 1989, B159.
[14] Laleg, M., Ono, H., Barbe, M.G., Pikulik, I.I., and Seth, R.S.: Paper Techn. Ind.
32(1991)5,24.
[15] Marton, J.: Tappi 63(1980)4,87.
[16] Smith, D.C.: Proceedings from Tappi Papermakers Conference 1992, Tappi Press Atlanta,
1992, pp.393.
[17] Kimura, Y., and Hamada, M.: Proceedings from Japan Tappi Conference 1993, Tappi
Press, Atlanta, 1993, pp.83.
[18] Stratton, R.A.: Nordic Pulp Paper Res.J. 4(1989)2,104.
[19] Stratton, R.A., and Colson, N.L.: Materials Research Society, Mat. Res. Soc. Symp. Proc.
Vol. 197(1990)173.
[20] Laleg, M., and Pikulik, I.I.: Nordic Pulp Paper Res. J. 8(1993)1,41.
[21] Allan, G.G., Fox, J.R., Crosby, G.D., and Sarkanen, K.V.: Fibre-Water interactions in
papermaking. Transactions of the symposium held at Oxford:Sept.1977, Technical Divi-
sion of the Paper and Board Industry, London, U.K., pp.765.
[22] Dunlop-Jones, N., in J.Roberts (Ed.): Paper Chemistry. Blackie Academic & Profes-
sional, ISBN 0 7514 0236 2, 1996, pp. 98.
[23] Bates, R., Beijer, P. and Podd, B., in L. Neimo (Ed.): Papermaking Chemistry. Fapet Oy,
ISBN 952 52160407,1999, pp. 289.
[24] Welch, C.M., and Kottes Andrews, B.A.: Textile Chemist and Colorist 21,2(1989)13.
[25] Welch, C.M.: Textile Res. J.58,8(1988)480.
[26] Caulfield, D.F.:Tappi J. 77,3 (1994) 205.
 183

[27] Luner, P., Zhou, Y.J., Caluwe,P., and Tekin, B., in C.F. Baker (Ed.): Products of Paper-
making. Transactions of the 10th Fundamental Research Symposium held at Oxford,
Sept.1993, Pira International, 1993, ISBN 1 858020549,pp.1045.
[28] Xu, Y., Yang, C.Q., and Chen, C-M.: Tappi J. 82,8(1999)150.
[29] Davison, R.W., in W.F. Reynolds (Ed.): Dry strength additives. Tappi Press, Atlanta,
1980, pp.1–31.
[30] Davison, R. W.: Tappi 55(1972)4, 567.
[31] Wågberg, L., and Björklund, M.: Nordic Pulp Pap. Res. J 8(1993)1, 53.
[32] Page, D.H.: J.Pulp Paper Sci. 115(1989)6,J229.
[33] McKenzie, A.W., and Higgins, H.G.: Australian J. Appl. Sci. 6,2(1955)208.
[34] McKenzie, A.W.: Appita 37(1984)580.
[35] Pelton, R., Zhang,J., Wågberg, L., and Rundlöf, M.: Nordic Pulp Paper Res. J.15,5
(2000) 440–445.
[36] Wågberg, L., Forsberg, S., and Juntti, P.: J. Pulp Paper Sci. 28,2(2002)222.
[37] Laine, J,. Lindström, T., and Glad-Nordmark, G.: Nordic Pulp Paper Res. J.
15,5(2000)520.
[38] Glittenberg, D.: Paper Tech. Vol(1992) 34.
[39] Fellers, C., in R. Mark (Ed.): Handbook of Physical and Mechanical testing of Paper and
Paperboard, Vol.1. Marcel Dekker, Inc., New York and Basel, 1984, pp.349–383.
[40] Seth, R. S., Soszynski, R.M., and Page, D.H.: Tappi 62(1976)10, 97.
[41] Sachs, I.B., and Kuster, T.A.: Tappi 63(1980)10,69.
[42] Perkins, R.W.Jr, and McEvoy, R.P.Jr.: Tappi 64(1981)2,99.
[43] Uesaka, T., and Perkins, R.W.Jr.: Sven.Papperstidn. (1983)R191.
100
 185

T==lå=íÜÉ=jÉÅÜ~åáëãë=_ÉÜáåÇ=íÜÉ=^Åíáçå=çÑ=tÉí=píêÉåÖíÜ
====~åÇ=tÉí=píêÉåÖíÜ=^ÖÉåíë

Bo Andreasson
SCA Graphic Research, Sundsvall
Lars Wågberg
Department of Fibre and Polymer Technology, KTH, Stockholm

7.1 Introduction 185

7.2 Wet-Strength Mechanisms 186

7.3 The Chemistry of Commercial Wet-Strength Resins 189

7.3.1 Urea-Formaldehyde 189
7.3.2 Melamine-Formaldehyde 190
7.3.3 Alkaline-Curing Resins 192
7.3.4 Glyoxalated Polyacrylamide Resin (G-PAM) 196
7.3.5 Starch 200

7.4 Functional Groups in Fibres 203

7.5 Future 204

7.5.1 Future Use of Existing Chemicals 204
7.5.2 Future Chemistry 205

7.6 References 206

7.7 Patents 207

TKN==fåíêçÇìÅíáçå

During World War II, the need for wet-strength papers initiated the development of wet-
strength resins. In Europe, the first wet-strength paper was produced in Germany in 1939 by
adding polyethyleneimine (PEI) to the pulp. The following years, intensive development work
led to wet-strength resins based on formaldehyde. The formaldehyde resins are more effective
and cheaper than PEI is. In 1960s alkaline curing wet-strength resins were developed and re-
placed the formaldehyde resins. Today, alkaline curing wet-strength resins are still dominant.
In 1988, about 65 million tons of paper, card and board were produced in Europe. About
5–6 % of this was paper accounted for wet-strength grade [1]. The most important types are
wet-strength sanitary papers like kitchen towels, paper handkerchiefs, serviettes, wipes, etc.
Consumption of tissue paper products has been growing at around 4 % per year since the begin-
186

ning of the 1980s [1]. Other important grades of wet-strength paper produced in Europe are:
sacks, packaging paper, wallpaper, map paper, bank note paper, etc.
This chapter reviews the chemistries of wet-strength additives starting with formaldehyde
resins. Many comprehensive reviews have been published in the field of wet-strength chemistry
and the mechanisms giving wet-strength in paper [2–7] and this is an attempt to summarise
these. In Section 7.2, wet-strength mechanisms, in general, will be described. In Section 7.3, the
chemistry of commercial wet-strength resins is described. Section 7.4 describes functional
groups in fibres, which are very important in this context. Finally, in Section 7.5, the future in
this area is discussed

TKO==tÉíJpíêÉåÖíÜ=jÉÅÜ~åáëãë

It seems generally accepted that the fibres that make up a paper are held together by intermolec-
ular forces. A widely spread opinion [4] is that the main parts of these forces are hydrogen
bonds between hydroxyl groups on adjacent fibres. However, hydrogen bond forces are very
short-ranged (< 5 Å) and will probably be responsible for holding the fibres together only when
the fibres are very close to contact. When we come to consider long-range interactions between
macroscopic particles and surfaces we shall find that the two most important forces are the van
der Waals- and electrostatic-forces [8]. Between molecules, van der Waals forces are fairly
weak, much weaker than coulombic or hydrogen bonding interactions.
However, between macroscopic bodies the van der Waals forces become large due to addi-
tivity. Contrary to hydrogen bonding forces, van der Waals forces between surfaces are always
present and have a range of at least 100 Å. A small fraction of covalent bondings between fibres
are probably also present [4]. When a paper is saturated with water, the fibres swell and all the
bonds are broken except the covalent bonds. However, the number of covalent bonds is small
resulting in a drop in strength leaving only a few percentages of the original dry strength. This
process is very quick; an untreated paper saturated with water loses all its strength within a few
seconds.
Since many products need wet-strength, chemical wet-strength resins have been developed.
Wet-strength is usually expressed relative its dry-strength as a percentage. It has been suggested
that papers with relative wet-strength more than 10–15 % should be considered wet-strength pa-
pers [9]. Wet-strength can be divided in temporary and permanent. Papers saturated with water
losing its wet-strength gradually within a couple of hours are said to have temporary wet-
strength whereas papers saturated with water keeping its strength for much longer times are said
to have permanent wet-strength. In Figure 7.1, temporary and permanent wet-strengths are
shown. The strength of paper with permanent wet-strength is almost constant upon soak time
whereas paper with temporary wet-strength loses its strength much faster.
The purpose of the wet-strength chemical is to [4, 9, 10]:
1. Protect existing bonds by preventing fibre swelling.
2. Form bonds that are insensitive to water i.e. covalent bonds.
In the literature [4, 9–11], at least two methods are invoked to explain the development of
wet-strength in paper:
 187

35

permanent wet strength paper

30
relative wet-strength (%)

25
temporary wet strength paper
20

15

10

5 untreated paper

0
0 10 20 30 40 50 60 70 80
soak time (hours)

Figure 7.1 . Relative wet-strength vs. soak time.

1. Protection mechanism: the wet-strength chemical diffuses into the fibres and cross-links
with itself to form an insoluble network around and through the fibre contacts. When the
treated paper is rewetted, this network inhibits fibre swelling in the fibre-fibre contact areas
and fibre separation. The existing bonds hold the fibres together and a part of the original
dry-strength is preserved. This is schematically shown in Figure 7.2.

fibre

wet-strength resin

Figure 7.2. Protection mechanism.

2. Reinforcement mechanism: The wet-strength chemicals react with the hydroxyl groups or
carboxyl groups on the fibres and form linkages of covalent bonds between fibres (fibre-
wet-strength chemical-fibre). The covalent bonds are insensitive to water. The linkages
reinforce the fibre-fibre contacts when the paper is rewetted as shown in Figure 7.3.
188

fibre

wet-strength resin

Figure 7.3. Reinforcement mechanism.

During the late 1930s and 1940s the first wet-strength resins were developed. From that time
up to now, there has been a rapid growth in the use of wet-strength resins and wet-strength pa-
pers. Commercial wet-strength resins share four attributes:
1. Water-soluble or water dispersible, thus allowing even dispersion and effective distribution
on the fibres. However, there is a class of chemicals that are not water-soluble but have
been used to give wet-strength to paper [4]. These are latexes such as the polyvinyls, acryl-
ics and styrene-butadiene’s. Important differences between these and the water soluble
products are the fact that they give rise to other properties in the paper, higher dosage is
required to give wet-strength and the mechanism by which they operate are different. This
is partly due to their different form (latex compared to dissolved polymers) and partly due
to their low reactivity to cellulose. Probably they form a continuous phase of polymers in
and around the fibre-fibre contacts.
2. Cationic thus facilitates adsorption onto anionic fibres. It seems generally accepted that the
initial attraction of the resin to fibres is primarily electrostatic. The cationic resin molecules
are attracted to the negatively charged surface of fibres and fines. Once the resin is attracted
to the pulp, its retention appears to be due to ion exchange with the counterions of the neg-
atively charged pulp [6]. A few anionic wet-strength resins are also mentioned in the litera-
ture. The retention of these requires the use of a cationic promoter.
3. Polymeric. With high molecular weight polymers being more completely adsorbed to the
surface of the fibres and forming stronger bonds at the same adsorbed amount.
4. Reactive with formation of cross-links with itself and/or with cellulose thus promoting net-
works.
Of the numerous patents describing compositions for wet-strength chemicals, relatively few
have been exploited and developed into products that have gained commercial acceptance. Ini-
tially Urea-formaldehyde- and melamine-formaldehyde-resins set the standard for performance
and cost. Later, resins were developed which cured under neutral or alkaline conditions. The
following sections review each of the major chemical groups of wet-strength resins in chrono-
logical order of their appearance in the patent literature.
 189

TKP==qÜÉ=`ÜÉãáëíêó=çÑ=`çããÉêÅá~ä=tÉíJpíêÉåÖíÜ=oÉëáåë

TKPKN==rêÉ~Jcçêã~äÇÉÜóÇÉ=

Early experiments showed that the impregnation of paper with formaldehyde gave rise to wet-
strength, but the offensive odour and the brittleness of the paper resulting from its low pH were
drawbacks [4]. Later on, the reaction product of urea and formaldehyde, dimethylolurea, was
found to produce wet-strength in paper. The structure of dimethylolurea is shown in Figure7.4.

O O O
pH = 7– 8
H2N C NH2 + 2 H CH HOCH2 NH C N H CH2OH

urea formaldehyde dimethylolurea

Figure 7.4. Formation of dimethylolurea.

Dimethylolurea self-cross link to a polymer when pH is lowered to 4–5 [5]. When this poly-
mer is added to the pulp it will further cross-link to an insoluble, three-dimensional network [4].
However, this product is not very substantive to fibres. This was solved by reacting dimeth-
ylolurea with water soluble polyfunctional amines such as ethylenediamine, diethylenetriamine
and triethylenetetramine. This resin is cationic and by that, substantive to pulp.
Since urea-formaldehyde is the earliest wet-strength resin developed for commercial use, it
has been used extensively in the study of wet-strength mechanism in paper. In a review article
by Chan [5], many studies over the years are presented. The main conclusion is that the insolu-
ble, three dimensional network can protect existing fibre-fibre bonds and restrict hydration and
swelling when the paper is rewet, i.e. the UF resin does not form covalent bonds with groups on
the fibres. Obviously, the UF-resin is supposed to produce wet-strength by the protection mech-
anism mentioned earlier (see Figure 7.2). The true nature of the physical form of UF resins in
solution is not totally clear. Helmer [12] showed that the higher molecular mass the more effi-
cient was the UF resins. In later work [13], Helmer also found that the higher molecular mass
detected could be an artefact of a change from a true solution to a colloidal form more than be-
ing due to a cross-linking to a high molecular mass single polymer.
UF resins are thermosetting and the pH in the system has a dramatic effect on the rate of
cure. For example, paper made at a pH of 5.5 will eventually reach the same wet-strength level
as paper made at a pH of 4.5 but will require much greater ageing time. UF resins may be used
in the pH range 3.8–4.5 [4] with an optimum pH of 4.0. At this pH the resin achieves most of its
wet-strength property after about two weeks of ageing. Addition as close to the head box as pos-
sible is recommended when UF resins are used. Typical resin addition levels vary between
0.5 % and 2.5 %, depending on the type of paper produced desired. The advantages and draw-
backs of UF resins are the following:
Advantages:
• UF resins have no organochlorine in the polymer system and will not contribute any
absorbable organic halides (AOX) to the paper making system.
• Low cost
• Easy repulping (compared to PAE)
190

• Permanent wet-strength
Drawbacks:
• The resin, produced in a conventional manner, contains approximately 6 % of free formal-
dehyde [5] that is complicated for health and safety reasons. However, using modifiers and
additives [5] the amount of free formaldehyde can be reduced to about 1 %. Acid resistant
handling system (i.e. machine parts) is required to deal with UF resins.
• UF resins are relatively low in cost. They are still used in making paper towels, tissue
paper, bag paper and wet-strength linerboard.

TKPKO==jÉä~ãáåÉJcçêã~äÇÉÜóÇÉ

Next development was melamine-formaldehyde (MF) resins that were first patented for use in
paper in 1944 [4]. Melamine reacts with formaldehyde at pH of 7–8 to form a series of methylol
melamines. In Figure 7.5, reaction between melamine and formaldehyde of 1:1 by molar ratio is
shown.

NH2 NH CH2 OH

C O C
N N N N
+ H CH
C C C C
H2N N NH2 H2N N NH2

melamine formaldehyde monomethylol

melamine

Figure 7.5. Reaction between melamine and formaldehyde.

In aqueous solution at acidic conditions the methylol melamine’s are cationic and very sub-
stantive to negatively charged fibres. It has been found that the use of 3:1 by molar ratio of
formaldehyde to melamine gives the best MF resins for wet-strength [5]. Like UF resin, MF res-
in has been found to not form any specific bonds, like covalent bonds, with groups on fibre [5].
It seems likely that the main reaction in wet-strength development is the polycondensation of
the resin itself. To a great extent the resin functioned to protect the existing bonds by reducing
the swelling in the bonded areas [4, 5, 9]. The polycondensation mentioned above is the result
of ether and methylene linkages resulting in melamine-formaldehyde colloids of about 20 meth-
ylol melamine monomers. The formations of ether and methylene linkages are shown in Figure
7.6a and Figure 7.6b.
The formations of the colloids takes place at pH lower than 5.5. It is recommended to add the
resin to a dilute stock near the head box. The rate of cure of the resin is increased by low pH and
high temperature. MF resin can be used in the pH range 4.0–5.5 with an optimum at 4.5. MF
resins have been used in combination with polymers [5]. For example, addition of polyvinyl al-
cohol to MF resin gives a product with better wet-strength performance [5]. Also a greater water
absorption rate for the treated paper is observed in this case.
 191

NH2 NH2
C C
N N N N

C C C C
H2N N NH CH2 OH H O CH2 HN N NH2

NH2 NH2
C C
N N N N
+ H2 O
C C C C
H2N N NH CH2 O CH2 HN N NH2

Figure 7.6a. Ether linkage between two monomethylol molecules.

NH2
NH2
C OH C
N N N N
CH 2
C C C C
H2N N NH CH2 OH H N N NH2

NH2 NH 2
C OH C
N N N N
CH 2 + H2 O
C C C C
H2N N NH CH2 N N NH2

Figure 7.6b. Methylene linkage between two monomethylol molecules.

Advantages:
• Also MF resins are cheaper than PAE but a bit more expensive than UF resins.
• Permanent wet-strength
• The broke handling is easy
Drawbacks:
• Commercial MF resins have a pH value of about 2 that makes it very corrosive.
• An acid-resistant handling system is required.
• Like the UF resins, MF resins contain about 2–5 % free formaldehyde, which is compli-
cated for health and safety reasons.
192

MF resins are still used, especially in products where high permanent wet-strength is needed
for example in currency paper, map paper and photographic paper.
The drawbacks are serious since it is very difficult to solve the problem with free formalde-
hyde in both UF and MF resins. As mentioned earlier, experiments have shown that it is possi-
ble to reduce the level of free formaldehyde to about 1 % [5]. The UF resin from Eka Chemicals
contains less than 0.5 % free formaldehyde [14] and the future will tell us if further reduction is
possible. Another drawback with using UF or MF resins in tissue manufacture is the need for a
low curing pH. This will largely disturb the coating on the yankee cylinder and it is definitely
not possible to run these chemicals at their optimum working pH.

TKPKP==^äâ~äáåÉJ`ìêáåÖ=oÉëáåë

As mentioned earlier, UF and MF resins require acid conditions to cure effectively in paper. In
the 1950s, alkaline-curing epichlorohydrin products were investigated [6]. The first alkaline-
curing wet-strength resins to become commercially practical were poly (aminoamide)-epichlo-
rohydrin (PAE) resins. These rapidly started to replace UF and MF resins in many applications.
They offered improved absorbency and reduced machine corrosion and other benefits. Later,
polyalkylenepolyamine-epichlorohydrin (PAPAE) and amine polymer-epichlorohydrin (APE)
were introduced commercially. These are, like PAE, cationic and thermosetting at near neutral
and alkaline pH conditions. The terms PAE, PAPAE and APE describe the backbone polymers
of the resins. Therefore PAPAE and APE resins are sometimes categorised together as
polyamine-epichlorohydrin resins [6]. The poly (aminoamide) backbone of the PAE resin is a
result from the reaction between adipic acid and diethylenetriamine, Figure 7.7.

O O
Heat
HOC(CH2)4COH + H 2NCH2CH2NHCH2CH2NH2

Adipic acid Diethylenetriamine Pressure

O O
{NHCH2CH2NHCH2CH2NHC(CH2)4C}n

Water soluble poly(aminoamide)

Figure 7.7. Formation of poly (aminoamide) from adipic acid and diethylenetriamine.

The resulting poly (aminoamide) is then reacted with epichlorohydrin. In the PAPAE resin, a
polyalkylenepolyamine is reacted directly with epichlorohydrin and in the APE resin an amine
polymer is reacted with epichlorohydrin. The structure of the final wet-strength resin will de-
pend on whether the reaction partner of epichlorohydrin is a primary, secondary or tertiary
amine [6]. Secondary amines react with epichlorohydrin to form tertiary aminochlorohydrins,
which form cyclic structures of the 3-hydroxy-azetidinium salt type, Figure 7.8. Since these
structures are fairly constrained they are also fairly reactive.
 193

H O
heat +
{R N R'}n + ClCH2 CHCH2 {R N R'}n {R N R'}n

secondary epichlorohydrin H2C CH2

CH2
amine CH
HCOH
Cl - OH
CH2Cl

aminochloro- azetidinium
hydrin chloride

Figure 7.8. Reaction of epichlorohydrin and secondary amines.

The most important PAE resins are derived from secondary amines and the 3-hydroxy-azeti-
dinium ring is their principal reactive group. Tertiary amines react with epichlorohydrin to form
a glycidyl (2,3-epoxypropyl) ammonium salt. Only a few wet-strength resins based on the reac-
tion product of primary amines and epichlorohydrin exist and they will not be described in this
report. The functional groups can occur independently of the category of backbone polymer.
The most important resins with 3-hydroxy-azetidinium as reactive groups are of PAE and PA-
PAE type and the most important resins with glycidyl (2,3-epoxypropyl) ammonium as reactive
groups are of PAE and APE type. On weight basis, PAPAE resins are less effective than PAE
resins, while many APE resins, especially of epoxide type, are more effective. When PAE is
mentioned in the literature [3, 4, 6, 9–11] it is often PAE with 3-hydroxy-azetidinium as reac-
tive group. PAE is the most important alkaline curing wet-strength resin. PAPAE and APE res-
ins that were later introduced to the paper industry have not reached the same volume of use. In
this report PAE with 3-hydroxy-azetidinium as reactive group will be in focus and from now it
will be denoted only PAE. As mentioned earlier, the groups formed when the poly (aminoam-
ide) is reacted with epichlorohydrin is aminochlorohydrin that self-alkylate to form 3-hydroxy-
azetidinium rings, Figure 7.8.
The chemistry of this PAE resin has been extensively studied and a number of comprehen-
sive reviews have been published [4, 6, 9, 10]. The 3-hydroxy-azetidinium ring confers both re-
activity and permanent (pH-independent) cationic charge [4]. As indicated in Figure 7.8 the
reaction between aminochlorohydrin and 3-hydroxy-azetidinium is reversible and at equilibri-
um approximately 80–85 % is 3-hydroxy-azetidinium and 15–20 % is aminochlorohydrin [15].
The 3-hydroxy-azetidinium ring may react in different manner:
1. Reaction with other PAE macromolecules (homo-cross linking)
2. Reaction with cellulose fibres (co-cross linking)
3. Reaction with water

The different reactions of the 3-hydroxy-azetidinium ring are shown in Figure 7.9a–Figure
7.9c.
194

+
a) {R N R'}n N
H2C CH2 + NH CH2CHOHCH2N
CH
Cl - OH

azetidinium
chloride

+
b) {R N R'}n N

H2C CH2 + cellulose COO- CH2CHOHCH2 OCO

CH
cellulose
Cl - OH

azetidinium
chloride

+
c) N
{R N R'}n
slow
H2C CH2 + H2O CH2CHOHCH2 OH
CH
Cl - OH

azetidinium
chloride

Figure 7.9a-c. Different reaction mechanisms of azetidinium chloride from PAE. Reaction with secondary
amines, cellulose carboxyl groups and water.

Model compound studies using sucrose or methylglucoside indicate that PAE resins do not
react with cellulose hydroxyl groups [10]. In contrast, direct and indirect evidence supports the
reaction of 3-hydroxy-azetidinium groups with the carboxylate groups of pulp [10]. For in-
stance, direct spectroscopic evidence for the reaction of 3-hydroxy-azetidinium groups with
pulp carboxylate to form ester groups has been reported [16]. In the final resin, about 35 % of
the active groups are involved in homo-cross linking, 50 % are free as 3-hydroxy-azetidinium
groups and 15 % are free as aminochlorohydrin [15]. A picture of crosslinked PAE is shown in
Figure 7.10.
 195

+ + + +
N N N N N
H2C CH2 CH2 H2C CH2 H2C CH2 H2C CH2
CH CH CH CH
HC OH
OH OH OH OH
CH2
+
N N N N
H2C CH2 CH2
CH2
CH
HCOH HC OH
OH
CH2 Cl CH2
+ + +
N N N N N N
H2C CH2 H2C CH2 H2C CH2
CH2 CH2
CH CH CH
HC OH HCOH
OH OH OH
CH2 CH2 Cl
+ +
N N N N N N
H2C CH2 CH2 CH2 H2C CH2
CH2
CH CH
HC OH HCOH HCOH
OH OH
CH2 CH2 Cl CH2 Cl
+ +
N N N N N

CH2 H2C CH2 CH2 H2C CH2

CH CH
HCOH HC OH
OH OH
CH2 Cl CH2
+ +
N N N N N
H2C CH2 H2C CH2 CH2 CH2
CH CH
HC OH HCOH
OH OH
CH2 CH2 Cl
+ +
N N N N
H2C CH2 H2C CH2
CH2
CH CH
HCOH
OH OH
CH2 Cl

Figure 7.10. The structure of crosslinked PAE.

Since the resin reacts with carboxylic groups but not with the hydroxylic groups on fibres,
the carboxyl content of the pulp affects performance of the resin, and typically a higher wet-
strength can be achieved with unbleached kraft than bleached kraft. Bleached sulphite pulps,
which have a lower carboxyl content are the most difficult to treat [4].
The presence of soluble lignosulphonates reduces the efficiency of PAE resins by reaction
with the resin at the expense of the fibres [17]. Poorly washed neutral sulphite pulp contains
high levels of sodium lignosulphonates (SLS) which reduces the efficiency of PAE. However,
by addition of alum or calcium chloride to the neutral sulphite pulp, SLS concentration may be
reduced [17]. The degree of pulp refining can effect the performance of PAE resins. A more
highly refined stock will develop wet-strength because the higher surface area of the fines, in
particular, allows it to retain more resin [6, 18]. This effect should also be pronounced in me-
chanical pulp where the fraction of fines is high.
The optimal pH for alkaline-curing resins in paper manufacture is 6 to 8. The resin itself is
very stable down to about pH = 5 when self-cross linking of the resin will be severe. Also, at
low pH the fraction of anionic carboxylate groups on the pulp is reduced, resulting in lower sur-
196

face charge on the fibres. The upper limit is about pH = 9 [4]. The residence time is also an im-
portant factor. Up to a point, increasing residence time improves the retention of the resin
molecules. However, PAE tend to lose efficiency at very long exposure times. The reason for
this behaviour is probably due to polymer migration away from the surface and into the pores of
the fibre. PAE may also interact with inorganic and organic substances that can affect its inter-
action with the cellulose fibres, for a review see reference 6. PAE can be used in combination
with other polymers. For example, PAE can be used in combination with cationic and neutral
polymers like starch, dialdehyde starch, polyacrylamide-glyoxal resin, etc. [6]. Even better per-
formance can be reached in combination with anionic polymers. Beneficial effects require the
use of controlled ratios of anionic to cationic polymer so that the formed complexes have a cat-
ionic net charge. Synergistic effects have been reported for complexes formed between PAE
and carboxymethyl cellulose (CMC), polyacrylates, anionic starch, etc. [6].
As mentioned earlier, PAE resins are widely used as wet-strength chemicals. It is frequently
used in tissue and towel, liquid packaging board, food packaging, currency, maps, etc. PAPAE
resins find somewhat more limited use but they are used in unbleached pulps, mostly in packag-
ing. Dosages between 0.5 %–1.5 % are recommended depending on the type of paper produced
desired.
Advantages:
• Neutral or slightly alkaline conditions which reduces machine corrosion (compared to UF
and MF).
• Permanent wet-strength
• High relative wet-strength
Drawbacks:
• Difficult to repulp
• Organic chlorine containing
• Less absorption capacity of PAE treated papers (compared to untreated paper and paper
treated with G-PAM)
Today the companies are trying to solve the three problems mentioned above. Paper contain-
ing PAE is very difficult to repulp. For example, for unbleached paper, a high pH (11–12 using
sodium hydroxide) and high temperature (77 °C) in addition to mechanical treatment, is usually
necessary. Today, Hercules and Eka Chemicals have PAE resin almost free from organic chlo-
rine. In a patent from Hercules [19], Espy presents much improved absorbency in tissue papers
containing PAE. Instead of using only PAE or PAE/CMC, a combination of PAE, CMC and a
tertiary-amino polyamide epichlorohydrin resin was used. The mechanisms giving the faster ab-
sorption are not outlined or at least not mentioned in the patent.

TKPKQ==däóçñ~ä~íÉÇ=mçäó~Åêóä~ãáÇÉ=oÉëáå=EdJm^jF

G-PAM was introduced on the market in the 1960s and has grown to be an important additive
for wet-strength in paper. Since G-PAM gives temporary wet-strength, the principal use is in
tissue grades. G-PAM can be prepared in different ways, as an example, preparation of G-PAM
in two steps [4, 9] is presented in Figure 7.11 and Figure 7.12. In the first step, shown in Figure
7.11, a cationic polyacrylamide (PAM) is formed by the reaction of acrylamide and quaternary
ammonium cationic monomers.
 197

Cl -
O
+
95 CH2 CH C NH2 + 5 CH2 CH CH2 N CH2 CH CH2
CH3 CH3

acrylamide diallyl-dimethylammonium
chloride

{CH2 CH}95 {CH2 CH CH CH2}5

C O CH2 CH2
+
NH2 N Cl -
CH3 CH3

cationic polyacrylamide

Figure 7.11. An example of the first step in the preparation of G-PAM. Reaction of acrylamide and diallyl-dime-
thylammonium chloride.

In the second step, shown in Figure 7.12, G-PAM is formed by the reaction of the cationic
polyacrylamide (PAM), formed in the first step, and glyoxal.

{CH2 CH}95 {CH2 CH CH CH2}5

O O
C O CH2 CH2
+ + HC CH
NH2 N Cl -
CH3 CH3

cationic polyacrylamide glyoxal

{CH2 CH}80 {CH2 CH}15 {CH2 CH CH CH2}5

C O C O CH2 CH2
+
NH2 NH N Cl -
CH3 CH3
HCOH

HC O

glyoxalated cationic polyacrylamide

Figure 7.12. The second step in the preparation of G-PAM. Reaction of cationic polyacrylamide and glyoxal,
resulting in glyoxalated cationic polyacrylamide.
198

As can be seen in Figure 7.12, the final G-PAM resin contains three active groups, namely:
unreacted amines, amides reacted with glyoxal and quaternary ammonium cations. The unreact-
ed amines are free to form hydrogen bonds with hydroxyl groups on cellulose resulting in in-
creased dry strength. The quaternary ammonium cations are important for interaction with
negatively charged fibres. The amides reacted with glyoxal are the groups that will form homo-
and co-cross links. Obviously, in the preparation of G-PAM the amount of glyoxal used can
control the reactivity of the final resin. Evidence [4, 7, 9] strongly suggests that G-PAM imparts
wet-strength to paper primarily through covalent bond (hemiacetal and acetal) formations be-
tween free aldehydes on its reactive group and hydroxyl groups on fibres, see Figure 7.13.

co-crosslinking

{CH2 CH}n + 2 cellulose OH {CH2 CH}n

C
H2 O C
O O

NH NH

HCOH HCOH

HC O HC O cellulose
OH

glyoxalated acrylamide cellulose hydroxyl hemiacetal bond

H+
{CH2 CH}n

C O

NH

HCOH

HC O cellulose
O
cellulose

acetal bond

Figure 7.13. Co-cross linking reactions between the active group on G-PAM and hydroxyl groups on cellulose.
Hemiacetal and acetal bonds may be formed depending on pH.

Also homo-cross links are formed between aldehydes and free amines as indicated in Figure
7.14.
Probably also fibre-resin-fibre bonds are formed. Obviously, G-PAM gives wet-strength
with the reinforcement mechanism. As indicated in Figures 7.13 and Figure 7.14, the reaction
of G-PAM with cellulose is helped by acidic conditions. The reaction is very rapid at pH around
5 but rapid enough at neutral conditions. Curing of the paper gives little or no additional wet-
strength. Also indicated in Figure 7.13 is the reversibility of the reaction of G-PAM with cellu-
 199

lose in the presence of water. As a consequence, paper treated with G-PAM gradually loses a
portion of its wet-strength on prolonged soaking in water; G-PAM hence gives temporary wet-
strength [7]. It has been shown that if a laboratory hand sheet is soaked for four hours it still re-
tains over 85 % of its wet tensile strength after drying [4]. However, the action of alkali is im-
mediate and complete. The wet-strength loss with alkali is not reversible, probably due to a
destruction of the aldehyde functionality [2, 7]. Sulphite or bisulphite ions can react to form an-
ionic bisulphite adduct which react with cationic charge of the resin, through internal „salt“ for-
mation, resulting in poor retention, see Figure 7.15.

homo-crosslinking

H+
{CH2 CH}n + {CH2 CH}n {CH2 CH}n

C O C O C O

NH2 NH NH

HCOH HCOH

HC O HC OH

NH

C O

{CH2 CH}n

acrylamide glyoxalated acrylamide amidol

Figure 7.14. Homo-cross linking reactions between different groups on G-PAM forming an amidol.

O O O OH

C NH CH CH + HSO 3- C NH CH C SO3-

OH OH

Figure 7.15. Reaction of bisulphite with the active group of G-PAM.

This can be avoided by addition of a high charge density cationic polymer. Also bicarbonate,
HCO3–, may have negative effect on the performance of G-PAM. HCO3–, decompose with heat
to CO2 and OH– G-PAM may then react with OH– instead of hydroxyl groups on cellulose. This
negative effect can be minimised or even eliminated by simply reducing pH. pH ” 6.7 elimi-
nates efficiency losses due to bicarbonate [7]. This is another reason to work at pH levels slight-
ly acidic. G-PAM, with its relatively low cationic charge, is susceptible to interference from
anionic contaminants. On storage, the resin continues to cross-link and can gel, therefore, in or-
der to achieve the desired stability, the resin should be stored much diluted.
Since paper treated with G-PAM has temporary wet-strength, it is repulped more readily
than paper made with PAE. Broke repulping normally requires no special treatment, but can be
200

accelerated by heat and alkali. As mentioned in the previous section, PAE have negative effect
on the absorbency of paper. This is not the case with G-PAM. The time to absorb a water drop is
almost the same in a paper treated with G-PAM as an untreated paper [7]. The tendency to re-
tain hydrophobic materials by G-PAM are reduced compared to PAE is one of the reasons for
this behaviour [7]. Papers treated with G-PAM get increased dry strength, which can have a
negative effect on sheet softness, which is important in tissue. However, dry strength can also
be an advantage, for example increased utilisation of weaker fibres [7]. Also sheets with low ba-
sis weight can be produced if the dry strength is high.
Manufacturers give the following application advises for best performance: Thick stock to
minimise resin adsorption to fines; Rapid mixing; Addition after all refining; Stock pH< 8; Con-
tact time less than five minutes.
G-PAM can be used at pH between 4.5 and 7.5 with optimum between 6.0 and 7.0 [4]. Far-
ley [7] recommend as low pH as possible, but not below pH = 4.0. At high concentrations of an-
ionic material a large portion of the resin is consumed in neutralising the anionic materials. A
good solution of this problem is to pre-treat the furnish with a relatively inexpensive highly
charged cationic polymer, for example a quaternary ammonium polymer.
Advantages:
• Easy repulping
• High absorption capacity of the treated paper (applicable for tissue grades)
• Dry strength, which allow the use of weaker fibres
Drawbacks:
• Temporary wet-strength
• Decomposition of paper at alkaline contact
As mentioned earlier, dry strength is sometimes listed as a drawback. Too much dry-strength
can give stiff papers, which is a drawback in tissue grades. Temporary wet-strength is not a se-
rious drawback in tissue products, wiping up spills or drying ones hand normally takes a few
seconds. Loss of strength in a couple of seconds which is the case when using alkaline house-
hold cleaners is probably more serious.

TKPKR==pí~êÅÜ

Starch is a natural polymer. It is present in roots, tubers and seeds. The main sources for starch-
es are corn, wheat, rice and potato. Normally a plant synthesises two types of starch molecules,
amylose and amylopectin. Amylose is a linear polysaccharide whereas amylopectin is a
branched polysaccharide. Amylose consists of 100–10000 glucose monomers whereas amylo-
pectin consists of approximately 100000 glucose units, so the degree of polymerisation is much
higher for amylopectin. Both amylose and amylopectin are built up by glucose monomers. Pota-
to starch consists of 79 % amylopectin and 21 % amylose and maize starch consists of 72 %
amylopectin and 28 % amylose. Waxy maize starch contains only amylopectin [20]. In order to
use starch as wet-strength agent, the natural starch must be modified [4, 21]. Starch has been
used as wet-strength additive since the 1960s [4]. By oxidation of the adjacent hydroxyl groups
with periodic acid, aldehyde groups will be formed Figure 7.16.
 201

CH2OH CH2OH

CH O CH O
H C OH CH + HIO4 HC CH + HIO3 + H2O
CH CH O CH CH O

OH O O
n n

starch dialdehyde starch

Figure 7.16. Oxidation of a glucose unit in starch with periodic acid resulting in formation of aldehyde groups.

This oxidation can be very selective and effective and in most cases 80–90 % conversion to
dialdehyde starch is found [4]. The reaction with the hydroxyl groups on cellulose is very simi-
lar to that of G-PAM mentioned in the previous part. The similarities are many, also the fact that
DAS give temporary wet-strength. In order to make DAS substantive to pulp, cationic aldehyde
starch can be formed by reacting DAS with betaine hydrazide hydrochloride in the presence of
water and heat [22]. This reaction is shown in Figure 7.17.

CH2OH (CH3)3 CH2OH

CH O N + CH O
Cl - water
HC CH + HC CH
CH2 heat
CH CH O CH CH O
C O
O O N O
n NH n
NH
NH2
O C
+
H2C N (CH3)3
dialdehyde starch betaine hydrazide Cl -
hydrochloride cationic dialdehyde starch

Figure 7.17. Cationic dialdehyde starch formed by reaction of dialdehyde starch and betaine hydrazide hydro-
chloride.

As indicated in Figure 7.18 both hemiacetal- and acetal-bonds may be formed.

Hemiacetal bonds, formed at high pH, are easily broken by water and give temporary wet-
strength. Acetal bonds, formed at low pH, are more stable to normal tap water but hydrolysed at
acidic conditions and give wet-strength that is more permanent [4]. Of course, starch macromol-
ecules can also bind to other starch molecules by hemiacetal bonds. Unreacted groups of starch
may form hydrogen bonds with free hydroxyl groups on cellulose, this contributes to increased
dry strength [21]. One of the problems with DAS is the self cross-linking of the polymers upon
storage. This has been solved by Solarek et al. [21]. The idea they introduced was to add the al-
dehyde groups to the starch backbone in a blocked and unreactive form. Then, at the time the
starch was cooked the blocked aldehyde groups could be unblocked and activated and available
for cross-linking reactions (homo- and/or co-crosslinkings). This resin is called blocked reactive
group starch (BRG-starch). In the example shown in Figure 7.19, dialdehyde starch has been
202

modified to cationic starch by addition of a quaternary ammonium group. The cationic starch is
then reacted with N-(2,2-dimethoxy-ethyl)-N-methyl-chloroacetamide (DMCA).

R CH + R' OH R CH O R'
H2O
OH
cellulose or
DAS DAS hydroxyl hemiacetal

H+
R CH O R' + R' OH R CH O R' + H 2O

OH O R'
cellulose or
DAS hydroxyl acetal

Figure 7.18. Hemiacetal and acetal formation of dialdehyde starch.

cationic starch OH + Cl CH2 C N CH2CH(OCH3)2

CH3

O
+
(OCH3)2HC CH2 N C CH2 O starch O CH2CH2N(CH2CH3)2

CH3 H Cl -

acetal modification of starch

Figure 7.19. Blocked reactive group (BRG) starch.

The starch monomers consist of a cationic group and the acetal substituent group. This is a
dry product delivered as a powder [21]. To activate the reactive aldehyde group a starch solu-
tion must be adjusted to pH = 2.5 and cooked at 95–100 °C for 20–30 minutes Figure 7.20.
It is supposed that the aldehyde groups in starch react with cellulose and other starch mole-
cules in the same manner as DAS and G-PAM. As already mentioned, DAS and BRG starches
give temporary wet-strength and it seems likely that a faster decay as for G-PAM is observed
[4, 21]. Relatively high wet-strengths have been reported [21, 23] but not as high as for PAE, G-
PAM and MF.
 203

O
+ acid cook
(OCH3)2HC CH2 N C CH2 O starch O CH2CH2N(CH2CH3)2
pH=2.5
CH3 H Cl -

O
+
HOC CH2 N C CH2 O starch O CH2CH2N(CH2CH3)2
CH3 H Cl -

Figure 7.20. Activation of the reactive aldehyde group by cooking at pH = 2.5.

Advantages:
• Easy repulping
• Biologically degradable
• Dry strength, which allow the use of weaker fibres
• From a renewable resource
Drawbacks:
• Less effective than PAE, G-PAM and MF
• Temporary wet-strength
• Decomposition of paper at alkaline contact
More research work must be spent to make starch more effective. Temporary wet-strength is
not a very dramatic drawback in tissue products. Since the chemistry behind the mechanisms is
the same as for G-PAM, paper treated with starch is decomposed by alkaline which is a serious
drawback. Starch can also be used in combination with other wet-strength chemicals and exhibit
synergistic improvements. The most studied combinations are probably with PAE, for a review,
see reference 6. Exciting results have also been presented by Laleg and co-workers [23]. Cat-
ionic starch in combination with cationic zirconium-oxychloride, ZrOCl2, gave synergistic ef-
fects on relative wet-strength.

TKQ==cìåÅíáçå~ä=dêçìéë=áå=cáÄêÉë

Roughly speaking wood consist of cellulose, hemicellulose and lignin in the proportions 40 %,
30 % and 30 % respectively [24]. Since the functional groups in these macromolecules are in-
volved in hydrogen bonds and covalent bonds they are of great importance. The functional
groups are also responsible for the negative charge in the fibres that is important for the attrac-
tion of cationic ligands.
In cellulose, the functional groups are hydroxyl groups (–OH and –CH2–OH). pKa values
for these groups are high (pKa § 15) and at normal pH the amount of ionised groups can be ne-
glected. Even at extremely high pH values, only a small fraction of ions are present (0.1 % at
pH = 12). However the hydroxyl groups may be involved in hydrogen bonds responsible for
holding fibres together. Like in cellulose, hemicelluloses like glucomannan and galactan have
the functional groups –OH and –CH2–OH and for the same reason, they will not be ionised at
normal pH.
204

The functional groups in the hemicellulose xylan (about 10 % in spruce) [25] are carboxyl
groups (–COOH). Carboxyl groups have pKa = 4–5 and are highly ionised at normal pH
(99.7 % –COO– at pH = 7).
The functional groups in lignin are methoxyl (–OCH3), phenol groups like C6H5–OH and
C6H5–CH2–OH and some terminal aldehyde groups. Among these, the phenol groups can con-
tribute to the charge (pKa § 10). At pH = 7, 0.1 % of these groups are ionised but at higher pH a
considerable fraction may be in a dissociated form. It should also be mentioned that many of the
phenolic groups are occupied through linkages to neighbouring groups [24]. After wood pulp-
ing and the following bleaching (ECF or TCF), the chemical composition may be dramatically
changed. In the sulphate process followed by bleaching, most of the lignin is removed. Also, a
considerable amount of hemicellulose is removed resulting in a large drop of charged carboxyl
groups. Obviously the total charge and the surface charge of the fibres are very dependent upon
pH which must be taken into account in interaction studies between fibres and charged ligands.
The total charge of bleached sulphate pulp has been determined at pH = 7, to about 30 Pekv/
g [26]. In the sulphite process, lignin is attacked by sulphite or bisulphite and form sulfonic ac-
ids (C6H5–HSO3). pKa for sulfonic acid is about –6.5 [27], so sulfonic acid is always ionised at
normal conditions. In the following bleaching step most of the lignin residues are removed leav-
ing only a small fraction of sulfonic acid. The total charge of bleached sulphite pulp has been
estimated at pH = 7 to about 35 Pekv/g [26]. Probably a small part of this total charge originates
from sulfonic acids.
In chemithermomechanical (CTMP) pulping, wood chips is pre-treated with alkaline solu-
tion of sodium sulphite prior to the mechanical defibration. The lignin is attacked by sulphite re-
sulting in a large amount of charged sulfonic acid groups. Since most of the xylan is preserved
in the CTMP process, the pulp is also rich of charged carboxyl groups. After bleaching with
H2O2, the total charge was estimated at pH = 7 to about 250 Pekv/g [28]. Also the important
surface charge (for ligand interaction) has been estimated by polyelectrolyte titration to about
32 Pekv/g (at pH = 8) [28]. Using conductometric titration, the amount of sulfonic acid groups
and carboxyl acid groups have been estimated in bleached (H2O2) CTMP. [26]:
• Sulfonic acid groups 40–70 Pmol/g
• Carboxyl acid groups 150–250 Pmol/g
As have been described earlier in this chapter, the functional groups can also form hydrogen
bonds and covalent bonds with charged or uncharged ligands.

TKR==cìíìêÉ

TKRKN==cìíìêÉ=rëÉ=çÑ=bñáëíáåÖ=`ÜÉãáÅ~äë

In the time of environmental awareness, some of the wet-strength chemicals mentioned earlier
in this report have some „bad“ properties. The chemical industry has reacted responsibly. UF
resins have been produced with lower „free“ formaldehyde. PAE resins with reduced quantity
of organic chlorine-compounds are available. Among the existing wet-strength resins, PAE and
G-PAM seem to be most widely spread. PAE is very established and paper with high relative
wet-strength can be produced with this chemical. As mentioned earlier, PAE give rise to paper
that is stiffer than untreated paper, which is a drawback in tissue grades. PAE treated paper does
 205

also have less absorption capacity and longer absorption times. Researchers seem to work a lot
in trying to minimise the bad paper properties caused from the use of PAE.
For instance Espy [19], presented (in a patent assigned to Hercules Inc.) a method to improve
the absorption time considerably. Using a combination of PAE/CMC together with a poly (ami-
no amide) polymer without azetidinium groups, much better absorption times were achieved.
Being highly charged, the azetidinium free polymer competes with PAE for the binding sites on
the negatively charged fibres. This probably led to a less bonded structure between the fibres,
which result in shorter absorption time. It is probable that the use of the azetidinium free poly-
mer also result in lower levels of relative wet-strength.
The absorption capacity is an important factor in tissue grades. By definition, the use of wet-
strength resins, whose purpose is to prevent fibre swelling, reduce the absorption capacity. A
general believe is that the use of CTMP- or HT-CTMP-pulp, which have high absorption capac-
ity, could solve this problem. However, lab studies definitely show that the use of HT-CTMP
and wet-strength resins alone is not the solution to this specific problem.
In a very interesting patent by Phan [29], assigned The Procter & Gamble Company, a com-
bination of many chemicals was used. 75 % of kraft pulp was treated with PAE (for wet-
strength), CMC (for dry-strength) and a mixture of dihydrogenated tallow dimethyl ammonium
methyl sulfate and polyethylene glycol, (DTDMAMS/PEG-400) (for better absorbency and
softness). 25 % CTMP was treated with non-ionic surfactant and DTDMAMS/PEG-400. The
two fractions were then mixed. No results were presented, but keeping in mind the excellent
products from the Company, this patent is very interesting and important. The use of the TAD
technique that seem to be standard in The Procter & Gamble Company does of course also have
an influence on their results, this technique is described in numbers of patents and will not be
further discussed.

TKRKO==cìíìêÉ=`ÜÉãáëíêó

Many patents in the field of new wet-strength resins do only include modifications of PAE or
other existing wet-strength resins. This accentuates how difficult this field is and how difficult it
will be to replace PAE and G-PAM. To find wet-strength resins that are biodegradable is an in-
teresting part of the future chemistry. As already mentioned, starch is an alternative.
ATO-DLO (Instituut voor Agrotechnologisch Onderzoek) in The Netherlands, have im-
proved the process for oxidation of starch to dialdehyde starch. Dialdehyde starch is an excel-
lent starting point for future modification of the starch molecule. An interesting part is the
possibility to repulp paper treated with starch by amylase and/or amylopectinase. This enzymat-
ic break down is very interesting and should of course give advantages when dealing with the
repulping problem. Anyway, much more research work must be spent to make starch more ef-
fective as a wet-strength resin.
Another alternative is chitosan which is an amino-polysaccharide and, as can be seen in Fig-
ure 7.21, the structure of chitosan is very similar to that of starch.
Chitosan is cationic without chemical modifications, however, the charge is dependent upon
pH. Chitosan is available from shellfish wastes and has been evaluated as a paper strength addi-
tive [30]. It has been suggested [31] that chitosan interact with cellulose fibres that is more or
less a requirement for a potential wet-strength resin. Anyway, chitosan is expensive and must be
much cheaper to be a realistic alternative in the future.
206

CH2OH

CH O
HC OH
CH
CH CH O

NH2
n

chitosan

Figure 7.21. The chemical structure of chitosan.

In the patent literature, an abundance of new wet-strength resins are continuously presented.
It is always interesting to read about new ideas and new chemistry but it is a very time consum-
ing work to map all new ideas and even more time consuming to make experiments.

TKS==oÉÑÉêÉåÅÉë

[1] Stange, A. M. W. (1994): in Wet-Strength Resins and Their Application. Tappi Press,
Atlanta. p. 101–108.
[2] Farley, C. E.: Glyoxalated Polyacrylamide wet strength resin. In 1988 Tappi Wet and Dry
Strength Seminar Notes. 1988: Tappi Press, Atlanta.
[3] Espy, H. H., and Rave, T. W.: The mechanism of wet-strength development by alkaline-
curing amino polymer-epichlorohydrin resins. Tappi Journal 71 (1988) 133–137.
[4] Dunlop-Jones, N. (1991): in Paper Chemistry. Blackie & Sons Ltd., Glasgow and Lon-
don. p. 76–96.
[5] Chan, L. L., and Lau, P. W. K. (1994): in Wet-Strength Resins and Their Application.
Tappi Press, Atlanta, Georgia. p. 1–11.
[6] Espy, H. H. (1994): in Wet-strength Resins and Their Application. Tappi Press, Atlanta. p.
13–44.
[7] Farley, C. E. (1994): in Wet-Strength Resins and Their Application. Tappi Press, Atlanta.
p. 45–61.
[8] Israelachvili, J. N. (1985): in Intermolecular and surface forces. With applications to col-
loidal and biological systems. Academic Press inc. (London) Ltd., London.
[9] Neal, C. W.: A review of the Chemistry of wet strength development. In Tappi wet & dry
strength short course. 1988. Chicago: Tappi Press, Atlanta.
[10] Espy, H. H.: The mechanism of wet-strength development in paper: a review. Tappi Jour-
nal 78 (1995) 90–99.
[11] Devore, D. I., and Fischer, S. A.: Wet-strength mechanism of polyaminoamide-epichloro-
hydrin resins. Tappi Journal 76 (1993) 121–128.
[12] Helmer, U.: Investigation of cellulose modifying chemicals. In Department of polymer
technology. 1985, The royal institute of technology.: Stockholm.
[13] Helmer, U.: Private Communication.
 207

[14] Wilde, M. (1996): Private Communication.

[15] Gorzynski, M. (1996): Private Communication.
[16] Wågberg, L., and Björklund, M.: On the mechanism behind wet strength development in
papers containing wet strength resins. Nordic Pulp & Paper Research Journal 8 (1993)
53–58.
[17] Merrett, K. J.: Factors affecting the performance of polyamide type wet strength resins.
Appita 37 (1983) 233–236.
[18] Espy, H. H.: The effects of pulp refining on wet-strength resin. Tappi Journal 70 (1987)
129–133.
[19] Espy, H. H.: Absorbance and permanent wet-strength in tissue and toweling paper. United
States Patent United States 5,316,623 (1994) Hercules Incorporated.
[20] Svegmark, K.: Stärkelse. Struktur och reologi. (1987) Svenska livsmedelsinstitutet (SIK).
547 (Swedish).
[21] Solarek, D., Tessler, M. M., Jobe, P., and Peek, L.: Cationic Starch Aldehydes. in Paper
Chemistry Symposium 1988. 1988. Stockholm: STFI.
[22] Mehltretter, C. L., Yeates, T. E., Hamerstrand, G. E., Hofreiter, B. T., and Rist, C. E.:
Preparation of cationic dialdehyde starches for wet strength paper. Tappi 45 (1962)
750–752.
[23] Laleg, M., and Pikulik, I. I.: Unconventional strength additives. Nordic Pulp and Paper
Research Journal (1993) 41–47.
[24] Sjöström, E. (1993): in Wood chemistry. Fundamentals and applications. Academic
Press, Inc., San Diego.
[25] Annergren, G.., and Wilhelmsson, K.: Sågtimmer/fiberråvaror-användning i SCAs
svenska fabriker. (1996) SCA Graphic Research. K1135
[26] Wågberg, L., and Annergren, G.: Physiochemical characterization of papermaking fibres.
(1997) SCA Research AB. F2325.
[27] Solomons, T. W. G. (1992): in Organic Chemistry. John Wiley & Sons, Inc., Toronto.
[28] Wågberg, L., and Eriksson, M.: Effekt av våtstyrkemedel på arkegenskaper i papper av
CTMP och blekt sulfatmassa. (1990) SCA Research AB. K961 (Swedish)
[29] Phan, D. V., and Trokhan, P. D.: Soft absorbent tissue paper with high permanent wet
strength. European Patent France EP 0 610 337 B1 (1996) The Procter & Gamble Com-
pany.
[30] Allan, G. G., Crosby, G. D., and Sarkanen, K. V.: Evaluation of chitosan as a strength addi-
tive for alpha-cellulose and unbleached sulphite papers. In International paper physics
conference, 1975, Atlanta.
[31] Domszy, J. G., Moore, G. K., and Roberts, G. A. F. (1985): in Cellulose and its derivatives.
Halstead Press England, p. 463–473.

TKT==m~íÉåíë

[32] Allen, A. J.: Azetidinium polymers for improving wet strength of paper. United States
Patent United States 5,510,004 (1996) Hercules Incorporated.
208

[33] Clungeon, N. S., Devore, D. I., Fischer, S. A., and Giordan, J. C.: Wet strength resin com-
position and methods of making the same. International Patent United States WO 95/
27008 (1995) Henkel Corporation.
[34] Darlington, W. B., and Lanier, W. G.: Repulpable wet strength paper. United States Patent
United States 5,427,652 (1995) The Mead Corporation.
[35] Dickerson, J. A., Goldy, H. J., Smith, D. C., and Staib, R. R.: Improving the strength of
paper made from pulp containig surface active carboxyl compounds. European Patent
Germany EP 0 723 047 A2 (1996) Hercules Incorporated.
[36] Espy, H. H., and Putnam, S. T. Cationic water soluble polymeric reaction product of
poly(diallylamine)-epihalohydrin and nitrogen compound. United States Patent United
States 3,966,694 (1976) Hercules Incorporated.
[37] Espy, H. H.: Absorbance and permanent wet-strength in tissue and toweling paper. United
States Patent United States 5,316,623 (1994) Hercules Incorporated.
[38] Fu, Y. L., Huang, S. Y., and Dexter, R. W.: High strength wet webs for the production of
paper and process for producing paper making fiber webs having wet web strength. Euro-
pean Patent Germany 0 289 823 (1988) American Cyanamid Company.
[39] Jansma, R. H., Begala, J., and Furman, G. S.: Strength resins for paper. United Stated Pat-
ent United States 5,490,904 (1996) Nalco Chemical Company.
[40] Phan, D. V.: Soft absorbent tissue paper with high temporary wet strength.
[41] European Patent France EP 0 610 340 B1 (1996) The Procter & Gamble Company.
[42] Phan, D. V., and Trokhan, P. D.: Soft absorbent tissue paper with high permanent wet
strength. European Patent France EP 0 610 337 B1 (1996) The Procter & Gamble Com-
pany.
[43] Ramakant, T. S.: Sulfonated cellulose and method of preparation. UK Patent United
Kingdom 2 289 695 (1995) Kimberly-Clark Corporation.
 209

U==qÜÉ=pìêÑ~ÅÉ=çÑ=m~éÉê

Anthony Bristow
Bristow Consulting AB

8.1 Introduction 210

8.1.1 Two-sidedness 210

8.2 Surface Roughness 211

8.2.1 Different Roughness Measures 211
8.2.2 Surface Contact Area 212
8.2.3 Surface Volume 212
8.2.4 Air-leak Methods: Bendtsen, PPS, Sheffield 213

8.3 Surface Compressibility 214

8.4 Surface Porosity 215

8.5 Surface Permeance 216

8.6 Surface Absorptivity 216

8.7 Surface Wettability 217

8.7.1 Surface Contact Angle 217
8.7.2 Wetting Delay 218

8.8 Surface Printability 218

8.8.1 Ink Transfer 218
8.8.2 Print Density 220
8.8.3 Print Through 221
8.8.4 Set-off 222
8.8.5 Mottle 223

8.9 Surface Strength, Linting 223

8.9.1 Pick Strength 224
8.9.2 Fibre Lifting 224

8.10 Surface Friction 225

8.11 Surface Gloss 225

8.11.1 Reflection from a Plane Surface 226
8.11.2 Reflectance from a Rough Surface 228
8.11.3 Measurement of Gloss 229
8.11.4 Gloss Variations 230

8.12 The Creation of the Surface 230

210

UKN==fåíêçÇìÅíáçå

A paper is a three-dimensional network consisting primarily of fibres but, although it can often
be considered to be an infinite network in two dimensions, the network is always finite in the
thickness direction. In the thickness direction, there is a clearly defined boundary between the
network and the air, and the nature and structure of the network at this boundary define the sur-
face of the paper.
The surface is naturally relatively flat because the fibres lie preferentially in the plane of the
sheet, but it is not intrinsically sufficiently flat to fulfil its purpose and to meet end-use require-
ments. Various steps must therefore be taken in the papermaking process to ensure that the sur-
face is indeed as flat and smooth as is desired.
The surface of the paper is an extremely important region, particularly in printing papers.
The function of the surface is to accept and carry a printed message and it is sometimes said in
this context that the main task of the fibre network is merely to act as a support for the surface.
In a mathematical sense, a surface is a plane with zero thickness and thus zero volume, but in
a paper technology context the surface of the paper refers to a region with a certain thickness
and a certain volume possessing well-defined and important properties. If we consider a paper
web to consist of fibres lying in the plane of the sheet, then the surface is a region which is one
and in some cases perhaps two fibres thick.
In this chapter, we consider surface properties such as surface roughness, surface porosity,
surface compressibility, surface strength, surface wettability and surface gloss, and how these
properties are defined and measured. In the final section of the chapter, we consider briefly
some of the important papermaking processes required to ensure that the surface meets the end-
use specifications. These processes are discussed in more detail in other chapters of this text-
book.

UKNKN==qïçJëáÇÉÇåÉëë

A sheet of paper has two sides and therefore two surfaces. For some purposes, e.g. in packag-
ing, papers are made which are intentionally different in furnish or in finish on their two sides,
but for other purposes, e.g. in printing, the two sides must be as similar as possible. It is in the
latter context that the term „two-sidedness“ is used, in a negative sense, indicating a problem to
be solved or preferably a problem to be avoided.
In the traditional fourdrinier machines, there is more than one cause of two-sidedness. In the
first place, the fact that the paper sheet is formed on a wire means that the side of the paper in
contact with the wire is embossed with the pattern of the woven wire. In the second place, the
unidirectional drainage of the water from the sheet through the wire means that there is a flow
through the sheet which tends to give a non-uniform distribution of fines, filler particles etc
through the sheet, with either an enrichment or a depletion on the wire side. In this case there is
a corresponding depletion or enrichment on the top side. The top side of the sheet often reflects
the texture of the felt in the press section and may be referred to as the felt side.
The desire to avoid this two-sidedness was one of the factors which led to the development
of twin-wire formers with a similar wire on each side of the wet paper web and simultaneous
dewatering in both directions.
 211

In some cases, a two-sidedness may be desired as for instance in wrapping papers made on a
machine with a yankee cylinder, where one surface is fairly rough while the other is highly
glazed.
In other cases, differences between the two sides can be introduced by differences in furnish.
A liner destined to be used on the outside of a corrugated board has an inner surface which must
glue efficiently to the corrugated medium and an outer surface which must resist weather and
mechanical handling and must also bear a printed message concerning the contents of the box.
This outer surface will normally have a higher quality kraft pulp furnish, but it may also be
bleached in so-called white-top board, and may even be coated.
Many paper grades are now coated and in these products there is usually a desire that the two
sides shall be the same, so that the coating must be applied to the two sides in a manner which
will avoid the development of two-sidedness.

UKO==pìêÑ~ÅÉ=oçìÖÜåÉëë

The surface of a paper is never perfectly smooth, but it is necessary to be able to specify and
control the smoothness. This means that it has been necessary to consider how to define the
smoothness or roughness of the surface and how to make reproducible measurements of this
property. The name given to this property depends on the measurement technique adopted. If
the value of the property increases as the surface becomes smoother, the method is referred to as
a smoothness measurement; if the value of the property increases as the surface becomes rough-
er, the method is referred to a roughness measurement

UKOKN==aáÑÑÉêÉåí=oçìÖÜåÉëë=jÉ~ëìêÉë

The roughness is instinctively defined in relation to deviations from a plane surface. The mea-
surement of the roughness is however usually related particularly to the behaviour of the paper
in a printing process. It is therefore natural to define the roughness in terms of the deviation
from a plane surface pressed against the surface under a pressure corresponding to the nip pres-
sure in the printing press.

a) contact fraction (%)

b) surface volume (cm3/m2)

c) surface pit distribution

d) mean separation (mm)

Figure 8.1. Different ways of characterising the surface roughness.

212

Figure 8.1 illustrates this situation, and shows that several different measures of the rough-
ness are possible. Parameters which have been suggested are
• the surface contact area, expressed as a percentage,
• the surface volume, expressed in cm3/m2
• the surface pit distribution
• the mean surface displacement, expressed in μm
Here it is interesting to note that measures (b) and (d) are equivalent, since 1 cm3/m2 is equal
to 1 μm.

UKOKO==pìêÑ~ÅÉ=`çåí~Åí=^êÉ~

Although it is not widely used, the method developed by Chapman is worth mention. In this
method, a glass prism is placed on top of the paper and the surface is illuminated in a manner
which leads to total internal reflection in the regions where the glass is in contact with air and
not with the paper. The result is a pattern of contact and non-contact regions. This pattern may
be visually assessed or photographed and analysed to give a mean contact area as a proportion
of the total area and also detailed information about the contact pattern.
Similar information can be obtained by printing under a low pressure in a laboratory press
with a very small amount of ink. Under these conditions, the ink is transferred only to the raised
regions of the paper surface, and the print obtained shows the nature of the surface in a manner
which can be quantified by image analysis.

UKOKP==pìêÑ~ÅÉ=sçäìãÉ

The surface volume is a useful concept, but it is usually difficult to measure because any liquid
used to assess this volume is usually absorbed into the paper. The surface volume may therefore
best be assessed as the extrapolation of an absorption curve to zero time. The Bristow absorp-
tion tester – commonly known as the Bristow wheel – is one method which provides such a
measure. The principle of the method is illustrated in Figure 8.2.

liquid container

track

sample

rotating wheel

Figure 8.2. The principle of the Bristow absorption method.

 213

In this method, the absorption is assessed as the length of a stain left by a given volume of
liquid when it is absorbed into the paper through a slit in a liquid container which passes over
the surface at a given speed. The absorption time is varied by altering the speed of rotation of
the wheel on which the test piece is fixed.
A plot of the amount of a non-swelling liquid absorbed as a function of the square root of the
absorption time is usually a straight line, as shown in Figure 8.3. The intercept of this line on
the ordinate at zero time is the amount of liquid transferred to the surface in zero time before the
absorption can begin, i.e. the surface volume, in cm3/m2. The significance of the horizontal sec-
tion of the curve for water in this diagram is discussed in section 8.6.2.

80

oil A
60
transferred liquid (cm3/m2)

oil B

40

water

20

square root scale

0
0.0 0.1 0.2 0.5 1.0 1.5 2.0
time (s)

Figure 8.3. A plot of transferred liquid versus time, on a square root scale, obtained using the Bristow absorption
tester.

UKOKQ==^áêJäÉ~â=jÉíÜçÇëW=_ÉåÇíëÉåI=mmpI=pÜÉÑÑáÉäÇ=

The most commonly used method of assessing the roughness of a paper surface is an indirect
method, involving an air-leak concept. The basic principle is that a cup is inverted over the sur-
face and air is introduced into the cup under pressure. If the surface is rough, the seal between
the edge of the cup and the surface is imperfect, and the air tends to leak out between the surface
and the edge of the cup. The rate at which the air flows out is thus a measure of the roughness.
In the Bendtsen method, the roughness of the paper is expressed simply as a flow rate in ml/
min.
Parker refined the measurement technique in an instrument known as the Parker-Print-Surf
(PPS) instrument. He developed a more sophisticated measuring head, shown in section in Fig-
ure 8.4, he used a clamping pressure in the measurement nip which was similar to that in the nip
in a printing press, and he developed an equation for converting the air-flow rate into a measure
of roughness expressed in meaningful units.
214

regulated low
pressure air to flowmeter

measuring land

paper

resilient backing

Figure 8.4. A section showing the main features of the measuring nip in the Print-surf instrument.

His equation was based on the concept that the flow through a gap is proportional to the cube
root of the thickness of the gap. According to the model adopted by Parker, the spaces through
which the air flows between the paper and the measuring land can be considered to be a uniform
gap, so that the mean deviation G of the surface from a plane is given by the expression:

G = [(12 K b Q)/(l 'p)] 1/3 (8.1)

where
K is the viscosity of air
b is the width of the measuring land
Q is the volume of air flowing in unit time
l is the length of the measuring land
'p is the pressure difference across the measuring land, corrected for the air compressibility

The Print-Surf instrument thus gives a roughness measure expressed in μm. The Sheffield in-
strument used in North America is similar in its concept to the Bendtsen instrument, but uses
different dimensions, different pressures etc..

UKP==pìêÑ~ÅÉ=`çãéêÉëëáÄáäáíó

The roughness of paper is of particular concern when the paper passes through a printing nip
and ink on the printing forme is expected to transfer from the forme to the paper, but the fact
that paper is a compressible material is also important. In the 1970s, the concept of „surface
compressibility“ was introduced when it was realised not only that the compression of the paper
during its passage through a printing nip may be important but that the surface region may com-
press differently from the bulk as a whole.
The surface compressibility can be assessed by measuring the change in surface roughness
when the paper is compressed under different pressures, as shown in Figure 8.5.
 215

8
uncalendered
calendered
6 supercalendered
roughness-pps (mm)

1 2 3 4 5 7 10
clamp pressure (MPa)

Figure 8.5. The decrease in surface roughness with increasing pressure is a measure of the surface
compressibility.

It is routinely assessed by measuring the roughness under two different clamping pressures
without moving the paper, and it is then expressed as

K = 100 (G1 – G2)/G1 (8.2)

where G1 and G2 are the roughness values at clamping pressures of 1 MPa and 2 MPa respec-
tively

UKQ==pìêÑ~ÅÉ=mçêçëáíó

The porosity of a paper is properly expressed as the fractional void volume. If a sheet has a den-
sity UP and if the density of the cellulosic solid phase is UC, then for a given mass m of material,
the volume of the paper is equal to m/UP and the volume of the cellulosic solid phase is equal to
m/UC. The difference between these two volumes is the volume of the void space and the poros-
ity is then given by expressing the void volume as a fraction of the total volume:

porosity = (m/UP – m/UC)/m/UP = 1 – UP/UC (8.3)

It is clear from a philosophical point of view that the porosity of the surface region of the pa-
per may differ from that of the bulk sheet, but it is difficult to make measurements to demon-
strate this. Attempts have been made to study thin cross-sections of a paper sheet by image
analysis and to determine the relative amounts of material and of voids in successive layers
through the cross-section, but there is always some doubt as to how much distortion may have
been introduced by the sectioning in the microtome.
A method for measuring the distribution of surface pores has been proposed which is essen-
tially the inverse of the Chapman method (section 8.1.2), where the non-contact areas are analy-
sed, but it is not possible in this method to distinguish between surface pores leading into the
interior of the paper and surface pits, which are merely indentations in the surface.
216

UKR==pìêÑ~ÅÉ=mÉêãÉ~åÅÉ

The term „porosity“ is often incorrectly used to refer not to the amount of air in the material, but
to the rate of air flow through the material. This property should however more correctly be re-
ferred to as the air permeance of the material. The word „permeance“ is used rather than „per-
meability“ to emphasise that this is a sheet property, and not a material property normalised
with respect to thickness.
The air permeance through a sheet is measured in a device such as the Gurley or Bendtsen
apparatus where a single sheet of the material is clamped between a metal edge and a rubber
gasket, a pressure difference is applied across the material, and the flow of air in unit time is de-
termined. Under steady-state conditions, the flow through the bulk of the sheet is the same as
the flow through the surface, and this means that it is not possible to make any distinction be-
tween the bulk permeance and the surface permeance.
Although the permeance is dependent on the thickness and structure of the sheet, it may pro-
vide some information about the density and tightness of sealing of the surface region.

UKS==pìêÑ~ÅÉ=^Äëçêéíáîáíó

In assessing the effect on the surface of a surface treatment such as a calendering of the paper, it
may not be sufficient merely to measure the change in roughness or the change in gloss. Some
means of assessing the extent to which the surface pores have been closed up, especially in the
case of a coated surface, may be required.
The K&N ink test is such a method. The K&N ink is a proprietary ink manufactured for the
purpose of assessing ink absorption and it consists of a pigment extender and a purple dye in an
oil. A quantity of ink is applied to the surface of the paper and the excess is removed after a giv-
en time (2 minutes). The extent to which the ink has been absorbed is assessed optically by de-
termining the intensity of the stain. Measurements are made using a spectrophometer
conforming to ISO 2469 and incorporating C/2° and the CIE-Y-function as illuminant/observer
conditions and the K&N-value is calculated as:

K&N = 100 (Rf – Rp)/Rp (8.4)

where Rf is the reflectivity of a pad of the unstained paper and Rp is the reflectance factor of
the stained area of a single sheet of paper, obtained by placing the stained sheet over an opaque
pad of the unstained paper.
This test has its limitations. The method was developed as a qualitative method, but it has
been found to be sufficiently useful to justify its acceptance as a quantitative procedure. The
value obtained is to some extent dependent on the batch of ink used and it is clear from equation
(8.4) that the value obtained is also dependent on the reflectivity Rf of the substrate. The meth-
od is, nevertheless, sensitive to changes in the surface structure, and a decrease in the K&N-val-
ue after, for example, a calendering operation is a clear indication that the surface pores have
been closed up so that the absorption through the surface has been reduced. Further valuable in-
formation can often be obtained by carrying out the test over a range of different contact times.
 217

UKT==pìêÑ~ÅÉ=tÉíí~Äáäáíó

The wettability of the surface by liquids is important in many situations. Some products are de-
signed to absorb water and others are required to repel water. An essential factor in any gluing
operation, for example, is that a liquid phase – whether it be a solution, an emulsion or a pure
liquid – must wet the substrate. Similarly, any printing operation requires that the paper be wet-
ted by the ink, and this may be oil-based or water-based. The lithographic offset process re-
quires a correct balance between an oil-based ink and a water-based solution which is used to
provide a demarcation of the non-image areas. The lithographic process is essentially an exer-
cise in surface chemistry, even though the process often operates under conditions where the
surface chemical aspects are not a critical problem.

UKTKN==pìêÑ~ÅÉ=`çåí~Åí=^åÖäÉ

A basic measure of the wettability of the surface is the contact angle between a liquid and the
surface. If a drop of liquid is placed on a surface, an equilibrium is established, as indicated in
Figure 8.6.

gL

liquid
gS q gSL

solid

Figure 8.6. The equilibrium of forces determining the contact angle of a liquid on a surface.

The contact angle is then defined by the Young equation:

J sv J lv cosT  J sl (8.5)

where Jsv is the surface free energy of the solid/vapour interface, Jlv is the surface tension of
the liquid/vapour interface and Jsl is the surface free energy of the solid/liquid interface. This
means that the contact angle is dependent on the relationship between the different surface ten-
sions according to:

T arccos > (J sv  J sl ) J bl @ (8.6)

It is not easy to measure the angle itself, but instruments have been devised which project an
image of the drop onto a screen so that the height and base of the drop can easily be measured.
Assuming that the projected image is a sector of a circle, the contact angle Tcan then be calcu-
lated as:
218

T 2arctan  2h b  (8.7)

where b is the base and h is the height of the image

Modern instruments are computerised so that this is done automatically. Such instruments
should not, however, be used uncritically. One problem is that the paper must be held sufficient-
ly flat to ensure that the base is properly visible on the screen. Another problem is that on a po-
rous substrate such as paper, the liquid may be progressively absorbed into the material, so that
it is necessary to define the measurement conditions if reproducible results are to be obtained.
A more advanced treatment of the surface wettability takes into consideration the fact that
the surface tension can be divided into polar and non-polar components.

UKTKO==tÉííáåÖ=aÉä~ó

The thermodynamic analysis of wettability in terms of contact angles and surface tensions is
based on equilibrium conditions, but printing, gluing or similar operations take place under dy-
namic conditions and consideration must in practice be given to the dynamic wettability. There
is some indication that there may sometimes be a significant wetting delay. In Figure 8.3, the
curve representing the uptake of water has a horizontal region, and this had been interpreted as
indicating that there is a wetting delay, with a wetting time characteristic of the surface, before
the absorption can begin. Such a wetting delay is not observed in the case of the oils shown in
the same diagram.

UKU==pìêÑ~ÅÉ=mêáåí~Äáäáíó

One of the important properties which a printer requires when he purchases a paper is good
printability. Normally a printer talks simply of a paper's printability but in this chapter we talk
of „surface printability“ to emphasise that it is essentially the nature of the surface which deter-
mines whether or not the paper can be said to have a good printability. The printability of the
surface can be summarized under three headings as the ability of the paper to contribute to
• a good transfer of ink from the printing forme to the paper
• a satisfactory interaction between the ink or its components and the paper
• a good appearance of the ink on the paper, i.e. good print quality

UKUKN==fåâ=qê~åëÑÉê

The importance of the surface is emphasised in the theory of ink transfer. The transfer of ink to
a paper in a printing nip is dependent on a number of press-related and ink-related factors, but it
is primarily dependent on the chemical and physical characteristics of the paper surface. Fetsko
and Walker developed a theory for the transfer of ink to a paper surface where they proposed
that a certain amount of ink b is immobilised by the paper at the moment of contact in the print-
ing nip. If the amount of ink on the printing forme is x (whether g/m2 or μm), then the free quan-
 219

tity between the forme and the paper when the paper leaves the nip is equal to (x – b). This
means that the total amount y transferred to the paper is given by:

y b  f ( x  b) (8.8)

where f is a splitting factor which depends on e.g. the nip geometry. This is shown diagram-
matically in Figure 8.7.

x
l

(l–f)(x–b)
x–b
b
f (x–b)

y y = l[b+ f (x–b)]

Figure 8.7. Model of ink transfer according to Fetsko and Walker.

This transfer can however only take place when there is contact between the ink-covered
forme and the paper. If the fraction of the paper surface in contact with the forme is l then the
amount transferred is, after some rearrangement, equal to:

y l > f x  b(1  f ) @ (8.9)

A further adjustment is required to take into account the fact that this relationship cannot ap-
ply for small amounts of ink when x < b. To take this into account, l and b are replaced respec-
tively by the exponential functions:

l 1  e  kx
(8.10)

and

§  ·
x
b b ¨1  e b ¸ (8.11)
¨ ¸
© ¹

so that the full Fetsko-Walker equation is:

ª § x
·º
1  e  « f x  b 1  f  ¨¨1  e

 kx
y b
¸¸ » (8.12)
¬« © ¹ ¼»
220

Many attempts have been made to improve on this equation, but it remains valid as a general
description of the process, and it emphasises that the transfer of ink is related to a surface-
roughness parameter k and a pore-structure or absorptivity factor b which governs the initial im-
mobilisation of the ink. In this treatment, the wettability of the paper is not considered to be a
problem.

UKUKO==mêáåí=aÉåëáíó

In studies of ink transfer to paper, it is usual to evaluate the result by considering the optical
density of the ink on the paper. This is done by measuring the reflectance factor of the printed
area Rp and the reflectance factor Rf of the paper and calculating the density D as

D log10  Rf R p  (8.13)

An important parameter is then the ink requirement (ink demand), defined as the amount of
ink needed on the paper to achieve a given print density. The density criterion depends on the
process and on the type of paper being printed.
Figure 8.8 shows a typical graph of print density against amount of ink on the paper. The fig-
ure shows how the ink requirement to achieve, in this case, a density of 1,2 with this particular
ink on this particular paper is determined.

1.5

1.0
print density

ink requirement
at D = 1.2
0.5

0.0
0.0 1.0 2.0 3.0 4.0
ink on paper (g/m2)

Figure 8.8. Print density plotted against the amount of ink on the paper.

When two papers are being compared with respect to ink requirement, it is important to note
that there may be other criteria to consider. In all half-tone printing, pictures and other images
are produced by breaking down the picture into a pattern of dots. Good print quality demands
that all the dots are transferred completely and as discrete dots to the paper. With increasing ink
quantity, however, the ink tends to spread out and flow together so that essential information is
lost. This sets an upper limit to the quantity of ink which may be used. An example is given in
Figure 8.9, which shows print density curves for four different types of paper. On each curve, a
limit is shown above which it is not possible to print without an unacceptable filling in of dots.
 221

2.5

2.0
A
C
D
print density

1.5 F

1.0

0.5

0.0
0.0 1.0 2.0 3.0
ink on paper (g/m2)

Figure 8.9. Limits on the print density curve.

UKUKP==mêáåí=qÜêçìÖÜ

An important type of interaction between ink and paper is that where the ink, or components of
the ink, penetrate into the paper. This leads to a so-called print-through effect, where the image
on the front face is more or less visible on the reverse side.
Print through, G, is quantified by determining the reflectance factor Rq on the reverse side of
the print, and is calculated by analogy with the calculation of print density as:

G log10  Rf Rq  (8.14)

This phenomenon can be divided into different components, as indicated in Figure 8.10.

GST GST + GPP GST + GPP + GVS

Figure 8.10. The different components of the print-through phenomenon.

• a show-through component, Gst, corresponding to the opacity of the paper, i.e. an unavoid-
able component unless the paper is totally opaque
• a pigment-penetration component, Gpp, due to the penetration of the ink into the paper, i.e. a
component which increases in magnitude the more the ink penetrates into the paper
• a vehicle-separation component, Gvs, due to the fact that the ink releases oil which pene-
trates further into the paper and reduces the opacity of the paper behind the print.
222

This means that the opacity can be mathematically represented as the sum of three terms

G Gst  G pp  Gvs (8.15)

and that the components can each be expressed by a term resembling equation (8.14) as fol-
lows:

G log10  Rf Ra   log10  Ra Rb   log10  Rb Rq  (8.16)

Techniques have been developed for determining meaningful values of Ra and Rb so that
these components can be isolated.
Figure 8.11 shows data corresponding to that shown in Figure 8.8. Here the print through on
the reverse side of the paper corresponding to the print density of 1,2 on the front surface is in-
dicated. The calculated show-through component Gst is also indicated. The difference between
the two curves is thus the print through due to the penetration of pigment and oil into the paper.

0.4

0.3
print through
print through

at D = 1.2
0.2

0.1 show through

0.0
0.0 1.0 2.0 3.0 4.0
ink on paper (g/m2)

Figure 8.11. Print through plotted against the amount of ink on the paper. The magnitude of the show through due
solely to the opacity of the sheet is also shown.

UKUKQ==pÉíJçÑÑ

The interaction between the ink and the paper also affects the drying of the ink. Some inks are
dried by the application of external energy in the form of heat or UV-radiation. Others dry slow-
ly by atmospheric oxidation. In the latter case, it is important that the ink passes rapidly through
a so-called initial setting stage. During the time before this setting is complete, the ink can easi-
ly transfer from the paper to any other surface into which it comes into contact.
Set-off is the name given to this transfer phenomenon when it takes place under direct con-
tact without any rubbing or shearing action. It is measured by pressing the print against a recip-
ient substrate a short time after printing. It is calculated by analogy with print density according
to the equation:

SO log10  Rf , SO R p , SO  (8.17)
 223

where SO is the set-off and Rf,SO and Rp,SO are respectively the reflectance factors determined
on the set-off substrate and on the stain on the set-off substrate. The values depend on the sub-
strate chosen. The substrates commonly used are either a smooth cast-coated paper or a second
sheet of the paper being tested.
In some cases, it is important to determine the set-off as a function of time after printing in
order to assess which ink/paper combination shows the fastest setting. In other cases, it may be
important to determine the magnitude of the set-off at a given time as a function of the amount
of ink on the original print. Figure 8.12 shows such set-off data corresponding to the print den-
sity data shown in Figure 8.8. The degree of set-off corresponding to a print density of 1,2 on
the front face of the paper is also shown.

1.5

1.0
set-off

0.5

set-off at D = 1.2

0.0
0.0 1.0 2.0 3.0 4.0
ink on paper (g/m2)

Figure 8.12. Set-off plotted against the amount of ink on the paper.

UKUKR==jçííäÉ

The calculated print density is a mean value corresponding to the area measured. The spectro-
photometer described in Chapter 5 has a circular measurement area with a diameter of 30 mm
so that, although the Fetsko-Walker equation considers the possibility of uneven transfer in a
physical sense, the measurement of print density provides a mean value and ignores the fact that
an uneven transfer or a subsequent uneven interaction can lead to an uneven appearance. In gen-
eral an uneven appearance is referred to as mottle, and the tendency for a paper to give a mot-
tled print is an important quality criterion. The mottle is assessed visually or by an image
analysis technique, which makes it possible to consider the mottle in different spatial wave-
lengths.

UKV==pìêÑ~ÅÉ=píêÉåÖíÜI=iáåíáåÖ

In the traditional letterpress and lithographic offset printing operations, the surface of the paper
is exposed to considerable outward tensile forces as the paper leaves the printing nip, due to the
high-viscosity and high-tack printing inks in use. This means that particles (lint) may be torn
from the paper and these then collect on the printing form or rubber blanket and eventually im-
224

pede ink transfer and lead to a deterioration in print quality. The printer must then stop and
clean the press and lose valuable productivity. This is a problem particularly in the multicolour
lithographic offset process where the application of water in the non-image areas in an early
printing station can lead to a weakening of the surface in areas to which ink is transferred in a
later printing station.
Special measures must be adopted by the papermaker in his choice of furnish, additives and
surface treatment to ensure that the surface strength of the paper meets the requirements of the
printing process.

UKVKN==máÅâ=píêÉåÖíÜ

The need to be able to measure a paper's propensity to leave quantities of lint in the press has led
to the development of a special test procedure involving the use of the IGT Printability tester.
The principle of this test is that a disc, 1 cm wide and carrying a thin oil film on its circum-
ference, runs over a strip of paper at an accelerating speed. The surface of the paper strip is then
examined and the position where there are the first signs that particles are being lifted from the
surface is recorded. The speed associated with this position is then determined. The force exert-
ed by the oil on the surface of the paper as it leaves the nip is proportional to the product of the
velocity and the viscosity. This product – or, in the case of a single standard oil, the speed – is a
measure of the strength of the surface, and is known as the IGT-pick strength. Different oils
with different known viscosities are available and these enable the strength to be determined
over different ranges. It is important that this test is carried out under standard conditions of
temperature and relative humidity, not only because the surface strength of the paper is mois-
ture-dependent but also because the viscosity of the oil is temperature-dependent.
In practice, any increase in printing speed leads to a greater tendency for fibres or other par-
ticles to be pulled out from the paper surface. The tendency within the printing industry towards
ever-increasing press speeds thus means that there is a continual demand for papers with higher
pick strength.

UKVKO==cáÄêÉ=iáÑíáåÖ=

In recent years, interest in the phenomenon of a partial separation of fibres known as fibre lift-
ing (fibre rising) has grown, especially in connection with heat-set offset printing. In this pro-
cess, the paper is exposed to both water and heat and this process weakens the anchoring of the
fibre in the surface so that the tendency for one end of the fibre to break away from the surface
may increase.
Extensive work has been done to clarify the nature of the problem, and a method has been
developed where a paper is treated with a defined amount of water and then heated and dried.
Thereafter it is bent over a curved rod and the surface is viewed with a video-camera. This en-
ables the number of fibres which rise from the surface as the paper is bent over the curved rod to
be determined.
 225

UKNM==pìêÑ~ÅÉ=cêáÅíáçå

When two surfaces rub against each other, forces of friction always oppose the movement.
These forces may be critical for the performance of a paper in any given situation. In some cas-
es, friction is a desirable and necessary property. In other cases, friction may be a disadvantage.
When paper sacks containing some product are stacked in a warehouse, it is important that there
is sufficient friction to prevent the sacks from sliding. When paper sheets are being fed through
a printing press, it is important that they pass through singly and do not block into pads contain-
ing several sheets.
Surface friction is measured in a device which measures the stress when a sled to which one
sheet of paper is fixed is drawn at constant speed over a second sheet of the same paper. Tests
have shown that it may be important in this test to distinguish between the initial stationary fric-
tion and the dynamic friction. The force needed to start the sled moving may be greater than the
force required to keep it moving at a constant speed over the surface. It has also been found that
the force required tends to decrease if the sled is drawn repeatedly over the same position on the
paper. To increase the precision of the test, it is common practice to record the frictional force
for the third traverse of the sled across the paper sheet.
It is known that paper friction is dependent on the roughness, that it can be affected by the
presence of lubricant substances on the surface and that it may be affected by the moisture con-
tent of the paper and thus indirectly by the atmospheric relative humidity.

UKNN==pìêÑ~ÅÉ=däçëë

The concept of gloss falls into the same category as colour (cf. Chapter 5) in that it is a psy-
chophysical attribute rather than a simple physical property. The human brain considers a mate-
rial to be glossy if there is a glare from the surface, a lustre in the structure of the surface or
highlights in the surface showing a direct reflection of a light. Gloss is, however, a multidimen-
sional phenomenon which cannot easily be measured.
Steps are taken to impart a high gloss to many grades of paper, but gloss is not always desir-
able, particularly in books for leisurely reading, as is clear perhaps in the following quotation:
„The demand of the reading public for a paper of less gloss, which shall cause less strain
on the eyes, has already been met by a number of magazine and textbook publishers, and it
is apparently only a question of time before the use of highly calendered paper will be a
rarity for ordinary reading matter.“
Although these words sound fairly modern, it is instructive to note that they were written by
L. R. Ingersoll as long ago as 1914 in a journal called Electrical World. In part, they represent a
true forecast. There are, for instance, product standards requiring low-gloss paper in school text-
books.
Here, we are more concerned to consider how the psychophysical phenomenon of gloss can
be understood and converted into an instrumental measurement. The gloss of a paper is assessed
by measuring the amount of light which is reflected from the surface of the paper in a narrow
cone at an angle (the specular angle) equal to the incidence angle.
226

UKNNKN==oÉÑäÉÅíáçå=Ñêçã=~=mä~åÉ=pìêÑ~ÅÉ

The reflectance R is defined as the ratio of the reflected light J to the incident light I :

J
R (8.17)
I

The amount of light reflected from a smooth flat surface of a material with a given refractive
index is dependent on the polarisation of the light, and is given by the Fresnel equations. When
a ray light strikes an optically flat non-metallic surface with an angle of incidence T, some light
is reflected from the surface and some is refracted into the material. The amount of light reflect-
ed at the specular angle is dependent on the refractive index. The refractive index defines the
angle of refraction I within the material according to Snell's law:

sin T
n (8.18)
sin M

The reflection at the surface is dependent on the angle of refraction and on the plane of polar-
isation of the light. For light oscillating parallel to the plane of the incident ray, the reflectance
Rp is given by:

sin 2 (T  M )
Rp (8.19)
sin 2 (T  M )

and for light oscillating perpendicular to the plane of the incident ray, the reflectance Rs is
given by:

tan 2 (T  M )
Rs (8.20)
tan 2 (T  M )

These equations can be transformed to show the reflectance as a function of the angle of inci-
dence T for a given refractive index n. For light polarised parallel to the incidence plane, the re-
flectance Rp is given by:

2

ªcosT  n 2  sin 2 T
 º
1/ 2

¬« ¼»
Rp (8.21)
2

ªcosT  n 2  sin 2 T
 º
1/ 2

«¬ »¼

and for light polarised perpendicular to the incidence plane, the reflectance Rs is given by:
 227

2

ª n 2 cosT  n 2  sin 2 T
 º
1/ 2

¬« ¼»
Rs (8.22)
2

ª n 2 cosT  n 2  sin 2 T
 º
1/ 2

«¬ »¼

The total reflectance in the case of a non-polarised light source is then given by:

R p  Rs
R (8.23)
2

i.e. by the expression:

ªª 2 2 º
  º
 
1/ 2
ª n 2 cosT  n 2  sin 2 T 1/ 2
º
« «cosT  n  sin T
2 2
»
1 «¬ ¼» ¬« ¼» »
R  (8.24)
2« 2 2 »

« ªcosT  n 2  sin 2 T  º

ª n 2 cosT  n 2  sin 2 T
 º
1/ 2 1/ 2
»
¬ ¬« ¼» ¬« ¼» ¼

1.0

0.8

0.6
reflectance

0.4

Rp
0.2
Rs

0.0
0 20 40 60 80
angle of incidence

Figure 8.13. The reflectance as a function of angle of incidence.

Equations (8.21) and (8.22) are shown graphically in Figure 8.13 for a material having a re-
fractive index of 1.5. The curve for light oscillating parallel to the incident ray shows a progres-
sively increasing reflectance, whereas the curve for light oscillating perpendicular to the plane
of incidence drops to zero at the so-called Brewster angle TB which is defined by the expression

S
TB  MB (8.25)
2
228

Note that the curves do not pass through the origin but that they give a reflectance of about
4 % for normal incidence (T = 0).
These relationships are important for the definition of any gloss scale and in all the instru-
ments used, the reflectometer value is converted into a gloss value by defining a scale such that
the reflectance from the surface of black glass of a given refractive index at a given wavelength
is defined as a gloss value of 100.

UKNNKO==oÉÑäÉÅí~åÅÉ=Ñêçã=~=oçìÖÜ=pìêÑ~ÅÉ

The Fresnel equations lead to interesting relationships describing the reflectance of an optically
flat surface for different angles of incidence, but they are not of primary concern when the gloss
of papers is being measured. The Rayleigh criterion for a surface to be considered optically
smooth is that the height h of any asperities in the surface should not exceed the distance given
by:

O
h (8.26)
8cosT

This means that for measurement at an angle of 75°, the limit is 250 nm, whereas values of
the order of several μm are commonly given by the PPS method for determining surface rough-
ness.
It has been suggested that the effect of the surface roughness on the gloss can be assessed by
the equation:

ln( RR / RF ) 16S 2 cos 2 T ( s / O ) 2 (8.27)

where RR and RF are respectively the reflectances of the rough and the corresponding flat sur-
face, T is the angle of incidence, O is the wavelength of incident light, and s is the mean rough-
ness of the surface in μm.
Gloss can also be studied with a gonioreflectometer, i.e. with an instrument which allows a
free choice of both incidence and measurement angle.
If a surface is not flat, then light is reflected not only in the specular direction but in all direc-
tions, and the light reflected from the surface is mixed with light reflected after multiple scatter-
ing from within the material. It is, however, extremely rare for the light to be reflected
uniformly in all directions. If the intensity of the light reflected from a surface is measured in
different directions when an incident light ray strikes a surface at an oblique angle, it is found
that this intensity is direction-dependent. The light reflected from the point of incidence of the
light can then be described with vectors, the length and direction of which show the directional
dependence of the intensity. The curve which connects the end points of such vectors is called
the indicatrix of the vectors.
For a completely matt surface, the indicatrix is a circle, whereas for a perfect mirror it con-
sists of a point, in accordance with the law of specular reflection. Figure 8.14 shows the appear-
ance of the indicatrices for surfaces with different gloss levels.
 229

Figure 8.14. The reflectance indicatrix for different types of surfaces. (1) a Lambertian surface, (2) a matt paper,
(3) a drawing paper (4) a calendered coated art paper.

• Indicatrix 1, is a polar curve for an ideal material for which reflection takes place according
to Lambert's cosine law, which says that a light beam which meets a completely matt sur-
face is reflected in each direction with an intensity which is proportional to the cosine of the
angle between this direction and the normal through the point of incidence:

I (M ) cos M (8.28)

• A surface which gives a perfectly diffuse reflection according to this law is known as a
Lambertian surface.
• Indicatrix 2, shows the directional dependence for a highly matt paper with only low gloss.
• Indicatrix 3, has been obtained for drawing paper. This almost elliptical shape is a typical
example of mixed reflectance.
• Indicatrix 4, is for a supercalendered coated art paper. The indicatrix is here spherical as for
a matt surface, but it is superposed with a peak relating to the specular reflectance. This
peak does not however lie exactly in the direction of specular reflection, but in a direction
which forms an angle to the normal which is slightly larger than the incident angle.

UKNNKP==jÉ~ëìêÉãÉåí=çÑ=däçëë

A conventional gloss meter is a reflectometer with incident illumination at a given angle and
with measurement in the direction of specular reflection. The reflectance is measured with the
use of a filter which corresponds to the sensitivity of the human eye to light, i.e. the V(O) func-
tion referred to in Chapter 5.
Measurements are commonly made with an angle of 75° between the incidenct ray and the
normal through the point of incidence, and with measurement at the corresponding specular an-
gle of 75°. Two methods are available as International Standards: These give different results
primarily because the the methods have different angles of acceptance around the specular an-
gle. The importance of this can be seen in Figure 8.14 (3). A wider angle of acceptance means
that more of the light scattered around the specular angle will reach the detector and will be in-
cluded in the measurement.
230

It is not however satisfactory to measure all kinds of paper at the same angle and measure-
ment at a smaller incidence angle is recommended for highly glossy papers. Two methods are
used, one having an incidence angle of 20° and another having an incidence angle of 45°.
The result of a measurement is given not in reflectance units, but in gloss units which are de-
fined in relation to the reflection by a smooth black glass reference with a defined refractive in-
dex, as indicated in Table 8.1. This glass is given a gloss value of 100 in each method. It can be
seen in the table that the reference refractive index is not the same in all cases. The reasons for
this are historical rather than optical.

q~ÄäÉ=UKNK=däçëë=ëÅ~äÉ=êÉÑÉêÉåÅÉë=áå=ÇáÑÑÉêÉåí=pí~åÇ~êÇëK
pí~åÇ~êÇ ^åÖäÉ oÉÑÉêÉåÅÉ oÉÑÉêÉåÅÉ=ï~îÉJ péÉÅíê~ä=äáåÉ
êÉÑê~ÅíáîÉ=áåÇÉñ äÉåÖíÜ=EåãF
fpl=UORQJN TRø NIRQM RUVIP k~=aJäáåÉ
fpl=UORQJO TRø NIRST RUTIS eÉ=aJäáåÉ
fpl=UORQJP OMø NIRQM RUTIS eÉ=aJäáåÉ
bk=NQMUS QRø NIRST RUTIS eÉ=aJäáåÉ

UKNNKQ==däçëë=s~êá~íáçåë

The standard type of gloss measurement gives a mean value for a fairly large area of paper, but
there is a growing interest in measuring not only the average gloss but also small-scale varia-
tions in gloss, since these can be a disturbing factor in the visual appreciation of a printed sur-
face. For this purpose, image analysis techniques are usually required, where the area studied
can be broken down into a large number of picture elements (pixels) and the matrix of pixels
can be analysed.
In the measurement of the colour, brightness of a paper etc, it is considered desirable that the
gloss effect be eliminated. As a result, the sphere in the instruments used for this purpose is fit-
ted with a so-called gloss trap, which prevents any direct specular reflection from the sample
from reaching the measurement cell. In a similar manner, it has been argued that diffuse reflec-
tion from the material can interfere with a proper assessment of gloss, and a method has been
proposed for the determination of what is called visual lustre. In this method, the paper is illu-
minated at an angle of 45° to the normal and measurements of the reflectance are made at both
the specular 45° angle and at 0°. The theory is that the latter measurement provides an assess-
ment of the diffuse reflection, as indicated in the indicatrix shown in Figure 8.14 (4). This value
is then subtracted from the value obtained at 45° to leave the true gloss value.

UKNO==qÜÉ=`êÉ~íáçå=çÑ=íÜÉ=pìêÑ~ÅÉ

The papermaker must devote considerable attention to the task of achieving a good surface. At
his disposal, he has a number of variables and processes. These are dealt with in detail in other
chapters. Here a few general comments are appended.
 231

`çãéçåÉåíë=çÑ=íÜÉ=ëíçÅâ

The furnish strongly influences the nature of the surface. A furnish consisting of long coarse fi-
bres is not likely to give a very smooth surface. Instead both shorter fibres (birch) and fillers are
added to improve the surface. Here also, sizing agents which affect the wettability and absorp-
tivity, and additives which improve the surface strength are also important.

pÜÉÉíJÑçêãáåÖ

The surface is born in the sheet-forming process. It has already been mentioned that, for in-
stance, a twin-wire forming system may be important way of minimising the risk of an asym-
metric two-sidedness in the paper.

tÉí=éêÉëëáåÖ

The wet pressing of the fibre web is usually discussed in relation to the consolidation of the
web, but its importance for the structure of the surface should not be forgotten. An important
consideration in the application of the extended shoe type of pressing is the extent to which this
improves the surface or may in fact harden the surface and make it less amenable to subsequent
calendering.

aêóáåÖ

A special type of drying which imparts a high gloss to one side of the paper but has no effect on
the other side is drying on a hot so-called „yankee“ cylinder. The paper web is held in contact
with the surface of this large rotating highly glazed cylinder for a part of its passage through the
paper machine.

pìêÑ~ÅÉ=ëáòáåÖ

A feature of many paper machines making paper or lithographic offset printing is a surface-siz-
ing unit after the main dryer section. Here a CMC or starch solution is applied to the surface to
impart a greater surface strength.

`~äÉåÇÉêáåÖ

A process designed especially to improve the surface, to reduce the roughness and to impart
gloss is a treatment between rollers known as calendering.
In cases where the paper is subjected to treatment in a supercalender, the choice of which
side of the paper is exposed to the metal roller in the first soft nip may determine the extent to
which the two-sidedness of the uncalendered sheet is reduced or enhanced.
232

`ç~íáåÖ

A coating is often applied to the coating to improve the surface. This is dealt with in another
chapter. The coating must be porous in order to absorb certain components of the ink in printing
process, and in order to provide a light-scattering layer which covers the substrate and gives a
white appearance. The coating must also have sufficient strength to resist linting (piling) in a
lithographic offset press. The strength of the coating and the strength with which it adheres to
the substrate are governed by the type and amount of binder used in the coating formulation.
This means that the concept of „binder requirement“ may be important property of a coating
formulation, i.e. the amount of binder required to resist a given stress. To achieve the highest
gloss, cast coating is adopted.
 233

V==m~éÉê=~åÇ=mêáåíáåÖ

Göran Ström
Innventia AB

9.1 Introduction 234

9.2 Conventional Printing Techniques 234

9.2.1 Offset Lithography 234
9.2.2 Flexography 235
9.2.3 Rotogravure 236

9.3 Digital Printing Techniques 238

9.3.1 Electrophotography 238
9.3.2 Inkjet 241

9.4 Short About Prepress 242

9.4.1 Gray Scale and Resolution 242
9.4.2 Colour Printing and Colour Separation 245

9.5 Printing Papers 245

9.5.1 Newsprint 245
9.5.2 Magazine Paper 246
9.5.3 Book Paper 246
9.5.4 Office Paper 246
9.5.5 Graphic papers 247
9.5.6 High Quality Inkjet Paper 247

9.6 Printing Inks and Ink Drying 248

9.6.1 Offset Inks 248
9.6.2 Rotogravure Inks 249
9.6.3 Flexographic Inks 249
9.6.4 Ultraviolet Curable Inks 250
9.6.5 Inkjet Inks 251

9.7 Print Quality 252

9.7.1 Print Density 252
9.7.2 Print Gloss 252
9.7.3 Unevenness in Print Density and Gloss 253

9.8 Ink Drying in Offset Printing 253

9.8.1 Physical Drying 254
9.8.2 Chemical Drying 255
234

VKN==fåíêçÇìÅíáçå

The production of a printed material can be divided into three main stages viz. prepress, print-
ing and finishing. Prepress involves all the operations needed before the material can be printed.
Today most of those operations are carried out with computers, but before the revolution of pre-
press most operations were done manually. Printing is accomplished in conventional and digital
processes. Print finishing operations depend on types of printed work. The most common oper-
ations are cutting, trimming, folding and binding. In cutting, the material is cut to enable other
finishing operations, while trimming is the term used for cutting the material into its finished
size.
Each printing process requires specially designed ink to enable transport in the press and ef-
ficient drying. Although the inks differ strongly in composition and properties, the principal
components are pigment, binder and solvent/oil. The most common printing media is paper and
board. General requirements on the media are proper physical appearance, good transport
through the press, a balanced interaction with the ink and an attractive print quality.

VKO==`çåîÉåíáçå~ä=mêáåíáåÖ=qÉÅÜåáèìÉë

Conventional printing processes all have a non-variable printing plate that can be made off-
press or in-press. The printing process is named after the characteristics of the plate.
The most important conventional techniques are offset, flexography and gravure. Other less
important are letterpress, since this is an old and outgoing technique and screen printing which
mainly uses other media than paper e.g. bottles, plastics, fabrics, etc.
The printing evolution started around 1440 when Johan Gutenberg invented the letterpress
process. Today this process has a small and decreasing market share mainly in production of
books. The market share of flexography increase while that of rotogravure is fairly constant.
Offset has a large market share, about 50.5 % but it decreases due to the expanding digital print-
ing business.

VKOKN==lÑÑëÉí=iáíÜçÖê~éÜó

Offset lithography or simply offset is a planographic process, meaning that the areas on the
printing plate that carry ink (image areas) are in the same plane as those that do not carry ink
(non-image areas). The non-image areas are processed in such a way that the adhesion between
ink and this area becomes very weak. This hinders transfer of ink from the form roller (an ink-
carrying roller) to the non-image areas.
In traditional offset the non-image areas are hydrophilic (water loving). They consist of po-
rous aluminium oxide onto which a hydrophilic polymer (e.g. gum arabic) is adsorbed. The area
is then coated by water (fountain solution) and this „water-surface“ does not pick up ink in the
form roller nip. In waterless offset the non-image areas consist of silicone rubber. The adhesion
between the ink and this non-image area is very low since silicone has a very low surface ener-
gy. Moreover, it is believed that inks for waterless offset contain small amounts of silicone oil,
which concentrates at the boundary between ink and non-image area, and this further decreases
the adhesion.
 235

Since heat is generated during printing and fountain solution cools the press, waterless offset
presses are equipped with a cooling system. The main advantages with waterless offset over
conventional offset is lower waste paper since there will be no need for adjustment of ink-water
balance, a process in conventional offset that can create considerably amounts of waste. Water-
less offset also eliminates one of the reasons for dot gain namely emulsification of dampening
solution and this sharpens the image. The disadvantage with waterless offset is the more expen-
sive and the shorter lifetime of the printing plate since the silicone rubber layer is easily worn
down.
The image area is always a hydrophobic polymer, in fact it is a photo polymer which po-
lymerises when it is exposed to light during development of the image. Areas which have not
been polymerised carry unpolymerised substance which is washed away and the underlying sur-
face becomes the non-image area. Figure 9.1 shows a traditional offset plate with ink and foun-
tain solution, and a plate for waterless offset with ink.
Figure 9.2 shows a schematic drawing of an offset printing unit. One unit is needed for each
colour. The dampening system with the fountain solution contacts the plate before the ink. Ink
will not adhere to the non-image area since the split in the exit side of the nip will take place in
the weakest layer which is the water layer. The image is then transferred to the paper via a cyl-
inder covered by a rubber blanket. This cylinder is called blanket cylinder or offset cylinder.
Since water is used in the process and this is transferred to various surfaces it is important that
the ink can incorporate water. This is accomplished by emulsification.
A sheet-fed offset press prints on individual sheets, while a web offset press prints on a con-
tinues web of paper. The printing speed increases as the presses develop. The impression per
hour is in the range of 15 000 for modern sheet-fed presses and 40 000 for modern web presses.
A web press uses cold-set inks when printing on newsprint and heat-set inks when printing
on higher quality media such as magazine and fine papers. If the press uses heat-set ink, a hot-
air dryer to remove ink solvents will be included after the last printing unit. Chill rolls are then
located after the dryer. A sheet-fed press uses ink which dries through absorption of ink solvents
into the paper surface followed by a polymerisation of binders. These presses may also utilize a
dryer in order to speed up the drying process. Other drying mechanisms besides evaporation and
absorption are involved in the process where the ink film becomes solidified. These will be dis-
cussed later.

VKOKO==cäÉñçÖê~éÜó

Flexography is a relief process since the image on the printing plate (or form) is raised over the
non-image area. The form can be manufactured from a flexible rubber material, which enable
printing on rough media such as packaging boards, or from a polymer to give high print quality
on media with smooth surfaces. A flexographic printing unit is illustrated in Figure 9.3
The image area receives ink from an anilox roller. Its outer surface most often consists of a
ceramic material with engraved small and uniform cells. Excess ink is removed from the roller
by a doctor blade or by the fountain roller, which applies the ink to the anilox roller. The
amount of ink transferred to the media is very much determined by cell geometry and volume.
The impression cylinder generates a light pressure, just enough to transfer the ink. The pressure
is much lower than that in offset printing.
236

ink silicone rubber

water ink water

polymer

Figure 9.1. Illustration of an offset plate with ink and fountain solution (left) and a plate for waterless printing
containing silicon rubber as non-image area (right).

ink fountain

water fountain
paper feed

Figure 9.2. Configuration of a print unit of a sheet-fed offset press.

In contrast to offset both flexography and gravure uses liquid inks, which have low viscosity,
somewhere between water and thin oil. Offset inks are termed past inks and are more viscous,
close to syrup. Flexo ink contains high amounts of solvents and the image is dried after each
printing unit in contrast to web offset where the drying unit is placed after the last printing units.

VKOKP==oçíçÖê~îìêÉ

Rotogravure is an intaglio process since the printing cylinder holds the ink in recessed areas
(see Figure 9.4). The ink is feed to the printing cylinder by various techniques e.g. spraying, by
immersing in the ink fountain, from an intermediate inking roller. Ink excess is removed by a
doctor blade and the scraped off areas become the non-image areas. A dryer is located after each
printing unit. Toluene is a common solvent in roto inks when printing on publication paper
where fast drying is strongly requested. The drier has an efficient system to take care of the
evaporated solvent. Water-alcohol based inks can be used when printing on packaging board
because the printing speed is quite low which puts less demand on fast ink drying. Rotogravure
presses are large and expensive. The most serious disadvantage with these presses is the expen-
 237

sive printing cylinders. However the print quality is superior to flexo and offset although offset
print quality of today is almost as good as rotogravure print quality.

paper feed

plate
cylinder

anilox
ink tray roller

Figure 9.3. A Conventional Flexographic Unit.

paper feed

doctor
blade

ink tray

Figure 9.4. Illustration of ink transfer in rotogravure.

238

VKP==aáÖáí~ä=mêáåíáåÖ=qÉÅÜåáèìÉë

Electronic or digital printing refers to printing where the complete workflow from manuscript to
final product is handled from the computer. Digital printing also requires a printing system
where the image produced can vary from sheet to sheet. This may be accomplished by a vari-
able printing cylinder as in electrophotography or by the software as in inkjet printing. Variable
printing enable production of personalized materials for direct mail, conferences, catalogues,
etc.
Digital printing enables „printing on demand“ (PoD), which means printing what you need,
when you need it and where the customer needs it. This is an advantage when very short lead-
time and/or small editions are required.
One important advantage of digital printing over conventional printing is the low cost per
copy for small editions. In offset, for instance, the start up (set-up time) of a print job is time
consuming. Most often it requires off-press manufacture of printing plates, mounting the plates
in the press and start up, where the first 30–50 copies are rejected. Digital printing requires no
printing plates and the first copy is as good as the rest. However, the traditional printing indus-
try has recognized this shortcoming and made strong improvements. For instance certain offset
presses (e.g. Heidelberg Quickmaster DI) have been developed where the workflow is handled
from the computer and the printing plate is developed on the press. These presses have not a
variable image but the work is done digital and therefore they are often referred to as digital off-
set.
The disadvantage of digital printing over conventional printing (e.g. offset) is lower print
speed, lower print quality and higher production costs at high editions. The brake even point is
at around 1000 copies.
The two most important digital printing processes are electrophotography and inkjet. Elec-
trophotography is sometimes called laser electrophotography or just laser printing. This origi-
nates from the earlier way to use laser to create an image on the photo conducting drum. Today
the laser, to a large extent, is replaced by light emitting diodes (LED).
When printing on plain, uncoated paper electrophotography gives better print quality than
inkjet, but the hardware of electrophotography is very complex, costly, unreliable and creates
high maintenance costs. When printing on expensive special paper, inkjet can reach photo qual-
ity although at a low print speed. Therefore inkjet is more common as home desktop printer
while electrophotography has a large market as office printers and industrial digital print press-
es. High-speed inkjet printing can be achieved using special print head techniques and will be
discussed below. In addition inkjet has a market in wide format printing of posters although
special papers are used and printing speed is low.

VKPKN==bäÉÅíêçéÜçíçÖê~éÜó

Electrophotography is today the most common process in digital printing. It uses dry or liquid
toner. The liquid toner is a suspension of toner particles. The size of liquid toner particles is
around 1–2 μm, which is much smaller than that of dry toner. The size of dry toner particles is
around 6–10 μm but development of smaller toner particles is in progress. The smaller the toner
particles the higher potential for print quality. The toner transfer process is illustrated in Figure
9.5 for dry toner.
 239

The image is created on a drum coated with a photoconductor, which may be selen based
such as As2Se3 or, what is more common nowadays, an organic photo conducting material. The
photoconductor is charged by a corona1 at stage 1. The image is then exposed to the photocon-
ductor by means of a tightly focused light beam generated from a laser or light emitting diodes
(LEDs). The light beam focuses on the non-image areas and dissipates the charge in this area.
The areas of the drum that still carry charge attract the toner in stage 3 developer and a toner im-
age is formed onto the drum. The toner has developed a tribo-electric charge opposite that of the
drum by means of friction during thorough mixing with magnetic carrier particles in the station
and is transferred to the drum as a so-called „magnetic brush“. At station 4 the toner is trans-
ferred to the paper by electrostatic interaction with the paper surface. The electrical charge on
the paper has been induced by a second corona unit located on the backside of the paper. In the
next stage the toner is fixed to the paper by a fuser. This is most often done by the combination
of heat and pressure. The next stage of the drum is to remove any remaining electrostatic charge
and to clean it from toner particles.

charger

erase (discharge) LED printhead

cleaner photoreceptor drum developer

paper
feed fuser
transfer

Figure 9.5. The principle of dry toner transfer in electrophotography.

During the end of 1993 Xeikon launched a web-based digital printing machine. This was the
first electrophotographic printer to print in full colour on both sides of the paper (duplex) in a
single pass. A scheme of the machine is shown in Figure 9.6. Eight electrophotographic print-
ing stations are included in the tower to give four colors (CMYK) on each side. The fusing unit
is located after the print tower meaning that the whole print on both sides is fused at the same
time. Station 5 is a gloss unit where the print is calendered in a heated polymer nip and then cut
into desirable size. Electrical and heat conductivity are important paper properties. A unit to dry
the paper to the right electrical properties is therefore placed before the tower. The press uses
dry toner and reaches a resolution of 600 dpi. The printing speed in A4 sheets per minute is 70
for Xeikon DCP 32 D and 100 for Xeikon DCP/50 D, a later member of this family.

1
A corona is a thin gold-coated tungsten wire which is subjected to a voltage of thousand volts. The wire can
impart electrostatic charge to objects close by.
240

4
3

1
5

prs tower fuser

gloss

Figure 9.6. Scheme of the Xeikon DCP 32D press.

While the Xeikon press work with a direct transfer of toner from the drum to the paper the
Xerox press is equipped with a belt to which the toner is first transferred. When the whole im-
age has been transferred to the belt it is deposited to the paper. This indirect toner transfer im-
proves print quality and put less demand on the paper. The Xerox machine is shown in Figure
9.7.The resolution is 600 dpi and print speed is 60 A 4 sheets per minutes which is similar to
that of the Xeikon press.

Figure 9.7. The Xerox DocuColor 2060 digital printing machine using indirect toner transfer.
 241

VKPKO==fåâàÉí

Inkjet technology is widely spread in personal desk printers for homes and offices. The printers
are inexpensive although the ink cartridges are quite expensive. Inexpensive plain papers e.g.
ordinary uncoated copy papers can be used if the print quality demand is moderate or if only
text is to be reproduced. Inkjet printing can yield very high print quality, in the same range as
photographs, but this requires special photo quality paper and those are quite expensive. Inkjet
has a unique potential for high print quality since the intensity of an ink spot on the paper can be
varied by adding several ink droplets on top of each other. Printers with very high resolution are
on the market. Epson Stylus C 80 has a resolution of 2880 dpi using variable size droplets from
a volume of 3 pl.
The market for inkjet technology is within small printers for homes and small offices, indus-
trial printing of wide format and industrial printing in general. The last segment is today quite
limited but is believed to grow. The wide format (typically 1 to 1.5 m) and super wide format
(> 2 m) color printing are used for short run poster work. The print quality is high and the nor-
mal resolution is 360–720 dpi, although high quality coated papers are required.
The ink droplets are generated as a continuous flow or as a „drop-on-demand“ (DOD) In the
continuous flow technique (see Figure 9.8) ink droplets are generated by piezo and the nozzle.
The drops are charged and can then be directed to the image. The unwanted drops are deflected
into a recycling reservoir. In DOD a drop is generated only when a dot on the paper is required.

image – printed by
deflected drops

h.v. deflection plates

charge electrode

drop generator
nozzle plate

gutter – to pick up
undeflected drops

Figure 9.8. Scheme of a continuous inkjet head.

In DOD, ink droplets are formed either by the piezo technology (piezo inkjet, PIJ) or by heat
pulses, which vaporizes some of the ink, which in turn induces a pressure. The latter is referred
to as thermal inkjet (TIJ) or bubblejet. It is general believed that continuous inkjet (CIJ) gives
higher print quality than DOD.
The first technical application of inkjet was disclosed by Siemens-Elema AB in Stockholm
1952. It was a chart recorder (Mingograf) based on a continuous jet with primary use in medical
instrumentation. The inkjet was formed in a glass capillary 3 cm long and a diameter of 0.1 mm
ending in a 15 μm nozzle opening. In 1976 IBM launched an inkjet system based on CIJ that
242

achieved near letter quality, 240 dpi. The first piezo DOD inkjet printer to reach the market was
introduced by Siemens in 1977.
The invention of thermal inkjet fundamentally changed the inkjet picture. This technique is
both cheaper and the print heads are much smaller compared to PIJ. The print head is equipped
with a transistor which heats the adjacent ink to > 290 °C for 2–3 μs. The ink vaporizes quickly
and creates a pressure pulse which ejects the ink drop. Also the condensation of vaporized ink is
quickly. The total cycle is 100 μs, which results in a drop rate of 10 thousand per second. It was
invented by Cannon in the late 1970th and given the name bubblejet.
So far only one printing press for industrial use has reached the market. It is manufactured by
Scitex and based on CIJ. The printing speed is 300 m/minutes or 2000 A4 colour sheets per
minutes at a resolution of 300 dpi. The speed is much higher (> 10 times) than in electrophotog-
raphy but the resolution is lower and more expensive papers are needed when high print quality
is requested.
Advantages and disadvantages of the different inkjet technologies can be summarized as:
• CIJ can generate high drop rate by manipulating the ink supply pressure (high printing
speed). CIJ is more expensive than TIJ and PIJ. Pigment based inks can not be used in CIJ
due to clogging during recirculation.
• TIJ has become the most commercially successful technology approach to DOD design due
to smaller size and lower production cost. This enables disposable print heads with a life-
time of 5 years. TIJ is limited to vaporizable inks. The bubble collapse process is hard to
control. PIJ can use all types of inks, also the non-vaporizable ones like oil and wax based
inks.
As guidelines we may say that inkjet drops have a volume of 40 to 80 pl (pico litres) al-
though volumes down to 3 pl are reality today. A drop of 3 pl has a diameter of about 18 μm
and the dot diameter on the paper becomes 30 to 50 μm depending on the paper. The drop ve-
locity is typically 2 m/s for DOD and 10–20 m/s for CIJ. The distance from print head to paper
is 1 mm or just above. A drop flight time of 50 μs is common.

VKQ==pÜçêí=^Äçìí=mêÉéêÉëë

Prepress is the common expression for operations that need to be done before the actual print-
ing (presswork). Examples of operations are design, layout, typesetting, graphic art photogra-
phy, colour separation, proofing, image assembly, filmmaking and plate making. In
conventional prepress most or at least some of these operations are done manually, while in
electronic prepress, which also is called desktop publishing these operations are done from the
desktop. The advantages of electronic prepress are lower production cost and decrease in pro-
duction time. For instance manual film image assembly and plate making may be eliminated.

VKQKN==dê~ó=pÅ~äÉ=~åÇ=oÉëçäìíáçå

Most printing systems deposit a certain amount of ink on a certain area of the paper. The printed
image consists of inked areas and non-inked areas, and the amount of ink per area unit is the
same for all inked area. Thus, it is either a certain amount of ink or nothing at all. A few excep-
 243

tions from this exist. In rotogravure the ink amount can be varied through the depth of the cells.
In inkjet certain systems can fire a few drops in a fast sequence and the drops merge to a large
one just before it contacts the paper. But we will consider the general situation where we have
ink in a fixed amount per area or no ink on the paper. The actual amount can be controlled by
the printer through measurements of print density on full tone areas. In fact the ink amount is
highly dependent on paper quality. A rough paper requires more ink than a smooth paper to
reach the same print density.
Pictures taken by a conventional camera have a continuous gray scale meaning that the
strength of the colour varies over the picture. In order to reproduce this with a fixed ink amount
per surface area, halftone dots are produced. The grey scale is constructed by printing small
dots, which vary in size. Small dots give light coloured areas and large dots give darker area. On
very dark areas the ink forms a network with dots of non-inked area. Thus, instead of varying
the ink film thickness, the area covered by ink is varied. Since the dots are very small the human
eye perceive the area as light or darker without observing the individual dots. The dots can take
different shapes like round, square, oval, etc.
Halftoning can be done with a special camera equipped with a halftone screen that is in con-
tact with the unexposed film. The screen is constructed in such a way that high light intensity
creates a larger halftone dot than lower light intensity. A more modern way to covert a continu-
ous tone photograph into a halftone picture is by the use of a scanner. The picture is then con-
verted into a digital form and can be stored and further treated in the computer.
Grey scale refers to the number of different levels of grey a picture can display. It goes from
completely white, i.e. no ink at all to an area completely covered by ink, i.e. a full tone area. In
offset, for instance, and by using a halftone screen, a very high number of grey levels can be ob-
tained since the halftone dot size can be varied almost continuously. Figure 9.9 shows an exam-
ple with ten grey levels from 0 % to 90 %. The rectangles show the grey levels and the right part
of the picture shows an enlargement to visualize the single halftone dots. When the grey levels
are produced by camera and films the number of grey levels is limitless since the size of the
halftone dot (or area that will carry ink) can be varied continuously. When using a scanner to re-
produce a picture, or in digital printing, the situation is different since the area that carries ink
only can be varied step wise. This will be discussed in the following section on halftone resolu-
tion. It is generally believed that the human eye can recognize 64 different grey levels but the
eye is more sensitive in certain ranges and thus it is normally recommended to use at least 100
levels.
Halftone resolution is a measure of the screen frequency i.e. how many lines there are per
length unit, i.e. one over raster width (see Figure 9.10). Most often this is given as lines per inch
(lpi) but sometimes also as lines per cm or mm. In digital printing using electrophotography or
inkjet the printer resolution is given in dots per inch (dpi). There is an important difference be-
tween halftone resolution given in lines per inch and printer resolution given in dots per inch.
The latter is used for scanners in reproduction of images and in digital printing where the dot or
more correctly the exposure dot is the smallest produced ink unit. Therefore exposure dots are
used for creating both grey scale and halftone resolution. A certain number of exposure dots are
needed to produce the grey scale. These dots fill up a raster cell and form a halftone dot. As
mentioned above, at least 64 levels of grey are needed and this is up to 64 exposure dots, i.e.
8x8 dots, see Figure 9.10. Then a 240 dpi printer with 64 grey levels would give a halftone res-
olution of 30 lpi (240/8) and a 600 dpi printer would yield 75 lpi. Newspaper prints usually are
screened at 65–85 lpi, while magazines are screened at 100–150 lpi and art prints at > 200 dpi.
244

To be more accurate, an 8×8 raster cell gives 65 grey levels since one adds the situation with no
exposure dot in the cell. Thus we always have nxm+1 grey levels.

0%

10 %

20 %

30 %

40 %

50 %

60 %

70 %

80 %

90 %

Figure 9.9. Grey scale obtained from changing the size of the half tone dots. Nine grey scale levels are shown. 64
levels plus the zero one are needed for an acceptable print.

Raster cell

Halftone
dot

Exposure
dot

Raster width

Figure 9.10. Generation of grey scale in a halftone (raster) cell. Raster cells that can house 8×8 exposure dots are
needed to obtain a grey scale of 1+64 levels where the first level is without any dots.
 245

VKQKO==`çäçìê=mêáåíáåÖ=~åÇ=`çäçìê=pÉé~ê~íáçå

Only three colours, cyan, magenta and yellow are needed to reproduce a print with all possible
colours. When two of these primary colours are mixed secondary colours are obtained. For in-
stance yellow and magenta gives red, while yellow and cyan results in green. Black is obtained
if all three primary colours are mixed in equal proportions. Therefore black is not needed but
used in practice to yield higher contrast. A full colour printing press uses four printing stations,
one for each primary colour and one for black. It is quite common that a printing press has more
than four printing units (and then more than four colours) in order to reach higher colour repro-
duction and special effects. The sequence in sheet-fed offset is generally to first print black and
then cyan, magenta and yellow. One printing plate is needed for each colour. A colour image
needs to be separated into four images, one for each primary colour and one for black. Today
this is done by a scanner. Colour reproduction involves a large number of operations which can
not be included in this text.

VKR==mêáåíáåÖ=m~éÉêë

Printing papers are best classified from the field of its use. Since quality and price go hand in
hand a very high number of paper grades exist on the market, and it is important to choose the
right paper grade for the product. Newspapers and magazines require paper of lower quality
than exclusive books with illustrating pictures like books about art, gardening, cooking etc. The
important paper quality features differ among the grades but smoothness, brightness, bulk and
printability are important for all. With printability we refer to the paper ability to facilitate ink
drying and yield a high print quality. Smoothness and bulk is most often a compromise.
Smoothness can be highly improved by calendering but this reduces bulk, which in turn reduces
stiffness.

VKRKN==kÉïëéêáåí

The largest end product for this paper grade is of course newspapers. The paper is either made
from deinked pulp or from a mechanical pulp (groundwood or thermomechanical) or a combi-
nation of deinked and mechanical pulp. Small amounts of kraft pulp may be added to improve
paper strength. Common grammages are 42, 45 and 49 g/m2, 45 g/m2 as the most common.
Important paper properties are:
• Strength. Newspapers are most often printed in web offset and the grammages is low in
order to keep cost down. Web brakes occur although they are not common, but very costly.
Surface strength is required in order to minimize dusting.
• Opacity. High opacity reduces print through i.e. an unwanted appearance of print from
images on the reverse side of the paper. Opacity is important for low grammages products.
• Surface roughness. High surface roughness reduces print quality such as print gloss, but
may also reduce smearing of ink from printed images to adjacent paper or the reader.
246

Improved newsprint is used in inserts, magazines, etc. This product may contain bleached
pulp and various special pigments to improve opacity. It is also produced in higher grammages.
Common grammages lie between 49 and 56 g/m2.

VKRKO==j~Ö~òáåÉ=m~éÉê

This paper is used in production of magazines, journals, catalogues, etc. Several different paper
grades are manufactured. The two most common are supercaledered (SC) and light weight
coated (LWC). SC paper is made from a furnish containing high amounts of mechanical
(groundwood, thermomechanical) or deinked pulps, some reinforcement pulp (sulphate) and
high amounts of filler, normally clay. The paper is heavily calendered to reach a smooth and
glossy surface. LWC paper is made from chemical pulp and mechanical pulp, typically in a 1 to
1 ratio. Filler is present from broke. The paper is coated with low coat weight 8–10 g/m2 and
side. The paper is finished by calendering. The grammages for SC and LWC papers is in the re-
gion 49 to 60 g/m2. The most common grammages are 52 and 56 g/m2 for SC and somewhat
higher for LWC.
Important paper properties are:
• Strength. Magazine paper is most often printed in web offset or rotogravure and web brakes
must be avoided. The grammages is low in order to keep production and mailing cost
down. These products are often mailed to the consumer.
• Surface strength is needed in offset printing due to the tackiness of the heat-set ink.
• Opacity. High opacity is requested to reduce print through for these thin papers.
• Surface roughness and paper gloss. Smoothness is an important feature since it increases
paper gloss and print quality such as print gloss and missing dots in rotogravure. This paper
is heavily calendered and as a consequence bulk and stiffness is reduced.

VKRKP==_ççâ=m~éÉê

Book paper is used in production of ordinary books with text only, pocket books, novels, etc. It
is made from mechanical pulp with some chemical pulp to improve strength. It is uncoated but
sometimes pigmented. Normal grammages are 60–90 g/m2 and the most important paper prop-
erty is bulk.

VKRKQ==lÑÑáÅÉ=m~éÉê

Normal papers for use in copy machines and office/home printers are uncoated and made from a
furnish containing high amounts of short fibres (e.g. birch), some long fibres (e.g. pine) and
high amounts of filler. Calcium carbonate is common due to its low price and high brightness.
Precipitated calcium carbonate is quite popular since it also improves bulk. The product is inter-
nal and surface sized to control wetting and absorbency of aqueous inks. It is important that
these papers can be used in many printing processes like offset, electrophotography, and inkjet.
Papers for documents and letters are often pre-printed in offset to mark it with logo, addresses,
etc. It is then used in electrophotography or inkjet. These papers are often termed multipurpose
 247

papers. They contain no or very low amount of mechanical pulp and are often referred to as
wood free uncoated fine papers. Grammages are in the range of 75 to 100 g/m2.
Higher quality papers, often coated, for colour copying is entering the market but used in
small amounts today.
Important paper properties are:
• Brightness, opacity. The optical appearance of the paper is important.
• Bulk. A high bulk is wanted in this product as in almost all paper products.
• Electrical conductivity. Charge transfer mechanisms are involved when toner is transferred
to the paper in electrophotography.
• Heat transfer. The fusing process in electrophotography involves heating the paper from
the backside. The heat is transferred through the paper to melt the toner.
• Ink absorptivity and wettability. These properties are important for ink drying and print
quality in inkjet printing.

VKRKR==dê~éÜáÅ=m~éÉêë

Graphic paper is used for products like annual reports, commercials, excusive magazines, il-
lustrated books like art books, etc. The furnish is quite similar to the one used for office paper,
i.e. chemical pulps of birch or other short fibre, some long fibres and calcium carbonate as filler.
These papers are coated in one single pass on line, or in several passes in an off-line coater.
Matt grades are often single-coated while multi-coated papers are used for silk and gloss grades
after appropriate calendering. The pigment coating composition is normally a mixture of calci-
um carbonate and clay with calcium carbonate being the dominant pigment due to cost and
brightness advantages. Plastic pigment are often used in the colour of glossy grades to improve
gloss which often is > 70 %. Common grammages are 100–170 g/m2, but higher exists.
Important paper properties are:
• Brightness, bulk. This is a very exclusive paper and must have a good appearance and stiff-
ness.
• Smoothness. This property is very important for print quality like print gloss. The paper
needs to be uniform in physical properties in order to yield evenness in print quality param-
eters like print gloss and print density.

VKRKS==eáÖÜ=nì~äáíó=fåâàÉí=m~éÉê

Uncoated office paper is used in inkjet printers when printing only text or the requirement on
print quality is moderate. These papers are multipurpose papers and also used in offset printing
presses and laser printer. When high quality inkjet printing is wanted, papers with special coat-
ings are required (Figure 9.11). These coatings may contain several specific layers, for instance
a dye fixation layer and an ink-sorbent layer. The dye fixation layer contains chemicals that in-
teract with the dye molecules and anchor the dye at the surface. This improves print density,
colour saturation, and water fastness. Chemicals for dye fixation are cationic polymers such as
polydadmac and polyvinyl pyrolidon. The ink-sorbent layer may consist of porous particles
with high specific area such as silica gels, or of a swellable polymer e.g. gelatine. The ink ab-
248

sorption into the micro porous layer is fast and governed by capillary forces (capillary absorp-
tion) while the absorption into the swellable layer is slow and controlled by molecular diffusion.
Glossy photo quality inkjet papers most often have a swellable layer and print quality is ex-
tremely high although the drying time is long.

Coated inkjet paper

Micro porous coating of special pigments (silicate, PCC)
Paper

Photo paper
Image fixation (dye fixation) layer (e.g. Al-sol+PVP).
Not always present
Ink-sorbent layer: Micro porous layer of special pigments
(silicate, PCC) or swellable layer (gelatin, cellulose deviate)
PE film filled with TiO2 as liquid barrier
Paper or polymer support

Figure 9.11. Structure of high-quality inkjet papers.

VKS==mêáåíáåÖ=fåâë=~åÇ=fåâ=aêóáåÖ

All printing inks consist of three main components, viz. colorant, binder and solvent/oil. The
colorant is most often a pigment but dyes occur, for instance in inkjet inks. The solvents range
from pure hydrocarbons through ketones, esters and alcohols to water. Water and alcohols are
frequent in flexo and inkjet inks. Type of binder depends very much on solvent. For instance
acrylate latex binder is frequent in water born inks, while oil born inks often uses alkyd and
hard resin binders. UV curable inks occur in offset, flexo and inkjet.

VKSKN==lÑÑëÉí=fåâë

Sheet-fed offset inks are manufactured by dispersing the pigment into a vehicle, which is a mix-
ture of binder and oil. Normal binders are alkyd and hard resin. The latter is solid and amor-
phous, while the alkyd is a liquid. Additional oils and various additives are added to the ink to
reach the right properties. A typical ink formulation is given in the Table 9.1.

q~ÄäÉ=VKNK=^=íóéáÅ~ä=ëÜÉÉíJÑÉÇ=çÑÑëÉí=áåâ=Ñçêãìä~íáçåK=
fåâ=ÅçåëíáíìÉåí mÜóëáÅ~ä=ëí~íÉ `çåíÉåí=EBF
máÖãÉåí=~åÇ=ÑáääÉê ÇáëéÉêëÉÇ=é~êíáÅäÉë NRÓOM
e~êÇ=êÉëáå=EÄáåÇÉêF ÇáëëçäîÉÇ=éçäóãÉê OMÓPM
^äâóÇ=êÉëáå=EÇêóáåÖ=çáäI=ÄáåÇÉêF äáèìáÇ=éçäóãÉê UÓNO
qêáÖäóÅÉêáÇÉ=EÇêóáåÖ=~åÇ=ëÉãáJÇêóáåÖ=çáäF ÜáÖÜ=îáëÅçìë=çáä NMÓOM
jáåÉê~ä=çê=îÉÖÉí~ÄäÉ=çáä=ÉëíÉêë=EåçåJÇêóáåÖ=çáäëF äçï=îáëÅçìë=çáä NRÓOR
^ÇÇáíáîÉë=ÉKÖK=ï~ñI=ÇêáÉêëI=~åíáçñáÇ~åíë PÓR
 249

Pigments are most often organic compounds but inorganic pigments occur. The particle size
is typically 0.1–0.5 μm. Carbon black is used as pigments for black inks. The primary particles
of carbon black are very small, in the range of ten to twenty manometers, but they form clusters
that are larger.
Hard resins used in lithographic inks are normally an ester of a polyhydric alcohol such as
glycerol or pentaerythritol and a rosin acid or a derivative of rosin acid. Such derivatives may
be dimerised or polymerized rosin acids, maleic or fumaric fortified rosin or a rosin modified
phenolic resin. Since the rosin derivative is a polybasic acid the hard resin normally has a high
molar mass. It is solid and amorphous with a softening point well above 100 °C often above
150 °C.
Alkyd resins are also esters of a polyhydric alcohol but the acids are a mixture of monobasic
and dibasic carboxylic acids. The mono basic acids are unsaturated C18 acids while the dibasic
acid is phatalic acid or isomers of phatalic acid. The alkyd resin is a liquid at room temperature.
Drying oils are triglycerides with high amounts of highly unsaturated fatty acids like linolic
and linoleic acid. They solidify (dries) through an oxygen induced polymerisation. Linseed oil
is a drying oil and quite often alkyd resin is referred to as a drying oil.
Sheet-fed offset inks are used to print on coated paper. It dries through absorption of ink oil
into the coating and polymerisation of the alkyd binder and drying oil, i.e. the part of the drying
oil that still remain in the ink film. Most of the oil deplete from the ink film and are absorbed by
the coating. Catalytic driers based on soaps of cobalt and manage are used to increase the drying
rate, while antioxidants are used to slow down drying rate on the press.
Heat-set offset inks are used in web offset printing, quite often on coated paper. This printing
press is equipped with a hot air drying unit to evaporate the oils from the print. The oils are de-
pleted from the ink film not only through evaporation but also through absorption into the coat-
ing. Oils with low boiling point are used to facilitate oil evaporation. Resins are specially
selected for their ability to release ink oil rapidly and completely upon heating. Drying oils are
added in small amounts, not as binders but to act as pigment wetting agents and as plasticizers.
Cold-set offset inks are used in web offset printing on newsprint. These inks contain only
small amount of drying oil, mainly to facilitate pigment wetting and dispersion. Consequently,
there is no oxidative (polymerisation) drying. This ink dries only through absorption of ink oil
into the paper and film formation of binder, which mainly is hard resin. This ink usually con-
tains somewhat more pigment than sheet-fed and heat-set offset inks.

VKSKO==oçíçÖê~îìêÉ=fåâë

Offset inks have a very high viscosity and are classified as past inks. Rotogravure inks, on the
other hand have a low viscosity and thus belongs to the group liquid inks. Roto inks dry through
evaporation of the solvent, which most often is toluene. Metal resenates are common binders in
these inks. Water born roto inks are used although they are much less common. Nitrocellulose is
a common binder in those inks.
250

VKSKP==cäÉñçÖê~éÜáÅ=fåâë

Flexo inks are also liquid inks. Solvent and water based as well as UV curable inks exist. Water
based inks are common for printing on paper and board while UV inks are used for printing on
plastic films. Nitrocellulose is the most common resin in solvent born inks. It has good pigment
wettability and wide compatibility with other resins. Acrylic resin is also commonly used both
in solvent and water born inks. When used in water it can either be formulated as a latex or be
used at a high pH. Acrylic resin has a good compatibility with nitrocellulose. Other resins are
rosin esters and polyamide resins both are used in alcohol born inks. Common solvents are ethyl
alcohol, propyl alcohol, propyl acetate, ethyl methyl ketone and water in combination with am-
monia to provide high pH for dissolution of acrylic resin. Tables 9.2 and 3 give some examples
of compositions of flexographic inks.

q~ÄäÉ=VKOK=bñ~ãéäÉ=çÑ=ëçäîÉåí=ÄçêåÉ=ÑäÉñçÖê~éÜáÅ=áåâëK
cçê=mb=Ñáäãë cçê=ÑçäÇáåÖ=Å~êíçåë
`çäçê~åíë NOB=çêÖ~åáÅ=éáÖãÉåí NQB=çêÖ~åáÅ=éáÖãÉåí=~åÇ=SB=íáí~åáìã=ÇáçñáÇÉ=~ë=ÑáääÉê
_áåÇÉê OOB=éçäó~ãáÇÉ NNKRB=åáíêçÅÉääìäçëÉ
QB=åáíêçÅÉääìäçëÉ UB=ã~äÉáÅ=êÉëáå
pçäîÉåíë PQB=åJéêçéóä=~äÅçÜçä ORB=ÉíÜóä=~äÅçÜçä
NPB=ÉíÜóä=~äÅçÜçä ORB=åJéêçéóä=~äÅçÜçä
NOB=åJéêçéóä=~ÅÉí~íÉ NMB=åJéêçéóä=~ÅÉí~íÉ
^ÇÇáíáîÉë OB=mb=ï~ñ RB=éä~ëíáÅáòÉê
NB=~ãáÇÉ=ï~ñ PKRB=mb=ï~ñ

q~ÄäÉ=VKPK=bñ~ãéäÉë=çÑ=ï~íÉê=Äçêå=ÑäÉñçÖê~éÜáÅ=áåâë=Ñçê=éêáåíáåÖ=çå=åçåJ~ÄëçêÄÉåí=~åÇ=~ÄëçêÄÉåí=
ëìÄëíê~íÉëK
`çåëíáíìÉåíë cçê=åçå=~ÄëçÄÉåí=ëìÄJ cçê=~ÄëçêÄÉåí=ëìÄëíê~íÉë
ëíê~íÉë
`çäçê~åí PRB=éáÖãÉåí=ÇáëéÉêëáçå RM QM
_áåÇÉêë ^ÅêóäáÅ=ëçäìÄäÉ=éçäóãÉê NM PM
^ÅêóäáÅ=Éãìäëáçå=Eä~íÉñF PM NOKR
pçäîÉåíë t~íÉê R NP
lêÖ~åáÅ=~ãáåÉ N N
^ÇÇáíáîÉë mçäóÉíÜÉåÉ=ï~ñ P P
lêÖ~åáÅ=~åíáÑç~ã=~ÖÉåí MKR MKR
pìêÑ~Åí~åí MKR M

VKSKQ==räíê~îáçäÉí=`ìê~ÄäÉ=fåâë

These inks are special designed to go from a liquid state to a solid state when exposed to ul-
traviolet (UV) radiation. UV curable inks are available for most printing techniques e.g. offset,
flexography, inkjet. Both non-aqueous and aqueous systems exist. The ink vehicle contains pre-
polymers, oligomers and monomers which polymerise in UV light by the aid of a photoinitiator
e.g. benzophenone. The prepolymers are used to provide the ink with its resin component and to
serve for proper rheological properties. The viscosity of an offset ink is much higher than that of
an ink for flexography or inkjet. The time required for the ink film to polymerise depends
 251

among other on ink film thickness and number of UV lamps. The polymerisation process is fast
and the film can be solidified within less than 10 ms.

VKSKR==fåâàÉí=fåâë

The first requirement on inkjet inks is low viscosity, preferably < 10 cp, but at least < 20 cp.
Many inkjet inks have a viscosity of only a few cp. The inks may me classified as:
• Oil based inks, where the oil is aliphatic hydrocarbons. The ink dries by absorption of oil
into the paper. The oil does not evaporate.
• Solvent based inks. Examples of solvents are methanol and methyl ethyl ketone. The ink
dries by absorption into the paper and evaporation.
• Aqueous inks. The solvent is water but co-solvents (e.g. alcohol, glycerol, diethylene glycol
monobutyl ether) and surfactants are required. The ink dries by absorption into the paper
and evaporation.
• Hot melt inks. Mixtures of waxes and amides are typical. The material has a sharp glass
transition temperature in the range 70 qC–120 qC and is jetted when it is in its liquid state.
It cools down on the paper and solidifies.
Printers used in homes and small offices are equipped with aqueous or hot melt inks. Oil and
solvent based inks are for industrial use. These inks must be handled correctly and care must be
taken to provide a safe indoor environment and to minimize air pollution. The colorant is either
a dye or a pigment. Dye colorants have high brightness and saturation when just printed but the
colour fades rapidly. Pigments are believed to dominate in the near future due to its higher light
stability and water fastness. The inks contain low amounts of binders, if any. Thus, fixation
agents are needed in the paper. Inkjet ink formulations are given in Tables 9.4 and 9.5

q~ÄäÉ=VKQK=^=Ñçêãìä~íáçå=çÑ=~=ëçäîÉåí=ÄçêåÉ=áåâ=Ñçê=`fg=ä~ÄÉääáåÖK
`çãéçåÉåí `çåíÉåíI=ïíKJB cìåÅíáçå
jÉíÜ~åçä QO ëçäîÉåí
jÉíÜóä=ÉíÜóä=âÉíçåÉ PM ÅçJëçäîÉåí
t~íÉê NKR ÅçJëçäîÉåí
bíÜóäÉåÉÖäóÅçä=ãÉíÜóä=ÉíÜÉê V Éî~éçê~íáçå=êÉí~êÇ~åí
jÉíÜóä=ÉëíÉê=çÑ=êçëáå NKQ ÄáåÇÉê
píóêÉåÉJ~ÅêóäáÅ=~ÅáÇ=ÅçJéçäóãÉê NP ÄáåÇÉê
aóÉ O Åçäçê~åí
kçåóäJéÜÉåçñéçäóÉíÜçñó=ÉíÜ~åçä Q ëìêÑ~Åí~åí

q~ÄäÉ=VKRK=^å=Éñ~ãéäÉ=çÑ=~=ï~íÉê=ÄçêåÉ=áåâàÉí=áåâK=fåâ=éêçéÉêíáÉë=~êÉW=ëìêÑ~ÅÉ=íÉåëáçåW=PSKRãgLãOI===
îáëÅçëáíóW=OKTÅé=~åÇ=ée=TKSK
`çãéçåÉåí `çåíÉåíI=ïíKJB
t~íÉê=EëçäîÉåíF TPKR
däóÅÉêáåÉ=EÅçJëçäîÉåíF NU
aóÉ OKU
_áçÅáÇÉ MKO
pìêÑ~Åí~åí RKR
252

VKT==mêáåí=nì~äáíó

Two of the most important general print quality parameters are print density and print gloss.
These are important in all printing processes. There are certain quality defects that are important
for a specific printing process, for instance missing dot in rotogravure and bleeding in inkjet.
Missing dot means that the paper did not pick up the ink from the cell of the roto cylinder. The
most frequent reason for this is the roughness and compressibility of the paper. If the paper has
a shallow pit at the location where the dot was to be placed, there will be no contact between ink
and paper, and the ink dot will not be transferred to the paper. Bleeding occurs in inkjet when
drops of different colours bleed into each other upon contact during absorption.

VKTKN==mêáåí=aÉåëáíó

Print density (D) is the optical density of the print. It is defined as the logarithm of the ratio of
the reflectivity of the paper (Rf) and the reflectance of the print (Rp) when this is placed on top
of a pad of unprinted paper sheets. The reflectance is the ratio between reflected light and inci-
dence light.

D = log(Rf/Rp)

Print densities as a function of ink amount of a few graphical papers (coated fine papers) are
given in Figure 9.12. It is clear that print density is strongly determined by ink amount.

2.4
2.2
2
Print density

1.8
1.6
1.4
Gloss Silk 1
1.2 Silk 2 Matt
1
0 1 2 3
ink amount (g/m2)

Figure 9.12. Print density is highly dependent on ink amount.

VKTKO==mêáåí=däçëë

Gloss is the ability of the surface to reflect light. Most often gloss is measured by illuminating
the surface at a certain angle to the normal to the surface, and measure the reflected light at the
same angle. Angles normally used are 20, 45 and 75 degrees. A glossy surface is measured at
small angles while a matt surface like newsprint is measured at a high angle. Print gloss is deter-
mined by surface roughness, ink levelling and refractive index. Ink levelling takes place directly
 253

after print transfer and it relates to the flowing of ink lumps to an smooth surface. The effects of
ink amount on print gloss for a few graphical papers are shown in Figure 9.13. The matt paper
is the roughest one and the gloss paper is the smoothest one. It is seen that the smooth paper
needs only a small amount to reach a high gloss while the matt paper needs higher amounts.
While print density is more affected by ink amount than by surface smoothness, the opposite
holds for print gloss.

100
90
Paper and print gloss (%)

80
70
60
50
40
30
20 Gloss Silk 1
10 Silk 2 Matt
0
0 1 2 3
ink amount (g/m2)

Figure 9.13. Print glosses very much depend on type of paper and the surface roughness is extremely important.

VKTKP==råÉîÉååÉëë=áå=mêáåí=aÉåëáíó=~åÇ=däçëë

The mean values of print density and print gloss are important print quality parameters but the
homogeneity in these features is as important or even more important. Nonuniformity in print
density is referred to as print mottle or only mottle. There are several origins of this print defect.
One is local variation in ink amount caused by local variation in ink setting rate, which in turn
depends on local variation in the paper pore structure.
Nonuniformity in print gloss i.e. print gloss variation is a severe print defect. It depends to a
large extent on surface roughness and is more pronounced on silk and matte coated papers than
on glossy ones. Evenness in print is a term often used to describe print quality. This is deter-
mined subjectively by viewing the paper at different angles. There is no general definition of
this quality property but print gloss variation plays an important roll.

VKU==fåâ=aêóáåÖ=áå=lÑÑëÉí=mêáåíáåÖ

The drying of offset prints on paper involves several phenomena which can be summarized into
two main categories, namely:
• Removal (depletion) of ink oil from the ink film and
254

• Chemical curing2 (mainly polymerisation) of residual drying oils e.g. alkyd resin and lin-
seed oil.
The depletion of ink oil can take place through evaporation as in heat-set offset or through
absorption into the substrate, which is the case for sheet-fed offset prints on coated substrates.
The depletion of ink oils is referred to as the physical drying whereas the chemical curing of al-
kyd resin is referred to as the chemical drying or oxidative drying. The latter expression comes
from the fact that the curing process is initiated by oxygen.

VKUKN==mÜóëáÅ~ä=aêóáåÖ

The initial stage of physical drying is referred to as ink setting. After the ink setting stage the
ink film does not smear when it is subjected to a light touch, which can be the hand of the print-
er or a paper surface in a set-off measuring device. It is quite obvious that ink setting time de-
pends on the technique used to measure it. For coated papers and boards ink setting time varies
from less than a minute to 10–20 minutes. Ink setting should be optimized. Too fast a setting re-
duces print gloss due to insufficient levelling of ink bomps from ink filaments and ink are build-
ing up on the rubber blanket while too slow setting may cause smearing in the stack. For coated
papers ink setting is strongly determined by the coating porosity and pore size. Higher porosity
and smaller pores gives faster setting. The effect of pore size on ink setting rate is shown in Fig-
ure 9.14. P&I slope is a measure of setting rate and a high value means fast setting.

16
14
Ink setting, P&I slope

12
10
8
6
4
2
0
50 100 150 200
Pore diameter (nm)

Figure 9.14. Ink setting rate decreases with increasing pore size of the coating.

The ink film is only partly dry when ink setting is completed. Once the ink is „set“, the prints
can be handled, although with care. At setting time roughly half of the oil is absorbed by the
coating. The absorption of printing ink oils by the paper coating continues but at a much slower
rate, and the mechanical stability of the ink film increases. Ink oil is not only absorbed by the
pores of the coating but also by the latex used as binder in the coating. The oil molecules diffuse
into the latex matrix which quite often is a co-polymer of butadiene and styrene.

2
Another type of chemical curing takes place in UV curable inks. This will not be discussed here.
 255

100 100
90 M-E
Remaining Oil in Ink Film (%)

90

Remaining Oil in Ink Film (%)

TG M-E
80 80
TG
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
0 5 10 15 20 25 0 100 200 300 400 500
Time (min) Time (min)

Figure 9.15. The initial absorption of ink oils into the coating (ink setting) is fast while the subsequent absorption
is slow. The data shows the remaining oil in the ink film as a function of time after printing for a vegetable-oil ink.
TG stands for triglycerides (mainly linseed oil) while M-E is the abbreviation for mono-ester oil.

Figure 9.15 shows the absorption of ink oils into the coating of a glossy coated paper.

VKUKO==`ÜÉãáÅ~ä=aêóáåÖ

The chemical drying of an offset inks (UV inks excluded) is an oxygen induced polymerization
of alkyd resins and drying oils (e.g. linseed oil). The process is also called oxidative drying. The
reactive part of the material is the unsaturation and the carbon next to the double bond of the fat-
ty acid pendant groups. Oxygen molecules adhere to this site and a hydroperoxide is formed.
The second stage is the decomposition of the hydroperoxide into free radicals. This is a fairly
slow process and catalysts based on soaps of cobalt and manganese is added to speed up the de-
composition. The catalysts are also referred to as driers or sicatives. As a comparison we may
stress that it may take 14 days for a film of linseed oil to dry in absence of driers but only a few
hours in presence of a moderate amount of drier.
The third stage is a radical polymerisation of the binder. Also formed are oxygen containing
functionalities like alcohol and aldehyde groups, as well as scissoring volatile compounds and
loss of unsaturation.
Antioxidants are used in order to reduce polymerisation on the printing press during short
stops or over-night stops. This material stabilizes the free radicals until it is consumed. Exam-
ples of antioxidants are hydroquinone, methyl ethyl ketoxime, cyklohexanone oxime and buty-
lated hydroxy toluene. With the proper balance of catalytic driers and antioxidants,
polymerisation will be delayed a certain time but once it starts it is completed within a relative
short period of time. The open time is the time it takes for a relative thick ink film laid on a glass
plate, to polymerise in open air. Inks with an open time of around 20 hours are referred to as
overnight inks or press stable inks.
Figure 9.16 shows drying curves after pile drying of a vegetable-oil ink printed on a coated
paper in sheet-fed offset. The three curves are for different temperatures and it is evident that
drying temperature has a strong impact on ink drying. Ink drying was characterized by rubbing
the print against an unprinted standard paper (uncoated) in one linear movement at a fairly high
pressure. The amount of ink transferred from the print to the recipient paper is given by the rub-
256

off value. The lower the value the dryer the ink film. The drying is very slow at low tempera-
tures. Follow for instance the curve for 6 qC. The initial decrease is due to physical drying, i.e.
depletion of ink oil from the ink film. This ends in a well-pronounced plateau. Then follows a
sharp decrease to the final level. The sharp decrease corresponds to the final chemical drying
i.e. polymerisation. The time for this to be completed is almost 20 days at 6 qC, about 4 days at
23 qC and only 16 h at 40 qC. The drying time is also very much dependent on excess to air.
Drying with free excess to air is about 5–10 times faster than pile drying.

90
80 40 C
70 23 C
60 6 6CC
Rub-off x 1000

50
40
30
20
10
0
0 100 200 300 400 500 600
Drying Time (h)

Figure 9.16. Ink drying after sheet-fed offset printing on coated paper using a vegetable-oil ink. The drying time
is strongly affected by temperature. The print was dried in the stack. Drying with free excess to air is 5–10 times
faster.
 257

NM==m~Åâ~ÖáåÖ

Christer Söremark and Johan Tryding

Christer Söremark AB, Tetra Pak

10.1 Introduction 257

10.2 Packaging Materials and Converting 258

10.2.1 Packaging Materials 258
10.2.2 Corrugated Board 259
10.2.3 Converting 261
10.2.4 Barrier Properties 269

10.3 Requirements on Packaging Performance 270

10.3.1 Definitions 270
10.3.2 The Supply Chain 270
10.3.3 Packaging Requirements 273

10.4 Design of Packaging Performance 276

10.4.1 Technical Performance 277
10.4.2 Marketing Performance 284

10.5 Future Trends 284

10.5.1 Digital Printing and E-print 284
10.5.2 E-tags 284
10.5.3 Smart Packages 285
10.5.4 New Packaging Materials 286

10.6 Literature 286

NMKN==fåíêçÇìÅíáçå

Packaging is a prerequisite for today’s society with its demand for distribution of food and other
products globally, nationally and regionally. From a business point of view the packaging sys-
tem is part of the business successes. The packaging system should contribute to:
• An efficient logistic flow
• Marketing and selling the product
• A reduction of the environmental load in the goods flow
Product development in for example the foodstuff area will lead to a wider range of products
and thus the development of new packaging systems. Changing consumer patterns, home shop-
ping, environmental issues etc. will result in new distribution channels.
258

Worldwide the packaging industry is a major part of industry. In the industrialised countries
it belongs to the top ten. The annual revenue is 550 billion € and the split between consumer and
industrial packaging is 70 to 30. About 5 million persons are employed in 100 000 companies.
By value paper and board is the largest material sector with 34 % followed by plastics with
30 %, metals with 25 % and glass with 6 %. Two-thirds of the world expenditure is in Europe,
North America and Japan combined.
In Sweden the annual packaging production is 2 billion € and the industry employs directly
more than 10 000 people.
To engineer packaging requires an understanding of all the requirements that exists and
knowledge on how to match these with the packaging performance.
Chapter 10 will focus on packaging materials, packaging requirements and packaging per-
formance for the paper and board sector.
First there will be a short introduction to packaging materials and converting. Then the pack-
aging chain will be discussed followed by its generated requirements, direct or indirect, on the
packaging. The process to create a packaging performance that meets the requirements will be
analysed next. Finally future trends in packaging will be discussed.

NMKO==m~Åâ~ÖáåÖ=j~íÉêá~äë=~åÇ=`çåîÉêíáåÖ

In this first section we will briefly introduce paper based packaging materials and the structure
and manufacturing of corrugated board. Finally the basics of converting packaging materials
into packaging will be presented.

NMKOKN==m~Åâ~ÖáåÖ=j~íÉêá~äë

Packaging can be divided into two groups, Flexible Packaging and Rigid Packaging. As seen
in Table 10.1 the material for flexible packaging is paper. For rigid packaging the material con-
sist of paper board or corrugated board, two materials that themselves have comparatively high
bending stiffness. Table 10.1 also gives the main applications for the two groups.

q~ÄäÉ=NMKNK=m~Åâ~ÖáåÖ=~åÇ=é~Åâ~ÖáåÖ=ã~íÉêá~äëK

cäÉñáÄäÉ=m~Åâ~ÖáåÖ oáÖáÇ=m~Åâ~ÖáåÖ
j~íÉêá~äë m~éÉê m~éÉê=_ç~êÇ `çêêìÖ~íÉÇ=_ç~êÇEäáåÉê=C=ÑäìíáåÖF
m~éÉê=tê~ééáåÖ
^ééäáÅ~íáçåë cçäÇáåÖ=`~êíçåë `çêêìÖ~íÉÇ=_çñÉë
m~éÉê=p~Åâë
m~éÉê=_~Öë=C
`~êêáÉê=_~Öë

For flexible packaging the largest application (by volume) is material for wrapping followed
by sacks and bags (including carrier bags). The papers are mostly made from unbleached kraft
pulp although the market share for bleached kraft pulp is increasing. To some extent also
bleached sulphite pulp is used, primarily for smaller bags and envelopes. The bleached products
can be coated with clay to improve the printing appearance. In addition special papers, e.g.
 259

greaseproof and glassine are used for applications related to their specific properties.
Greaseproof (grease-resistant) is for instance used for fatty food and industrial products protect-
ed by grease. Glassine (oil and grease-resistant) is used in much the same applications as
greaseproof, but where the demands for protection are higher. Also tissue paper is used in pack-
aging e.g. as soft wrapping for silverware, flowers etc.
The materials for rigid packaging consist of paper board or corrugated board. The manufac-
turing of the former and its different grades has been discussed elsewhere. The structure and
manufacturing of corrugated board will be discussed below. A typical market for paper board is
the food sector, e.g. cereals and snack foods, carriers for multi-packs (6-pack of beers), milk
containers etc. The corrugated board is mostly used for transport packaging due to higher re-
quirements on strength. However the development of corrugated board with lower basis weight
and thickness has opened a „grey“ market where both paper board and corrugated board com-
pete.

NMKOKO==`çêêìÖ~íÉÇ=_ç~êÇ

Corrugated board is a structure that consists of a core (medium) which has been fluted to a
wave-like form. To this core is glued one or two flat sheets of paper (facings) to form a single
face or single wall corrugated board, respectively. Additional cores and facings can be added to
give different structures, see Figure 10.1.

a) b)

Figure 10.1. Different corrugated board structures. (a) Single face corrugated board. (b) Single wall (top), double
wall (middle) and triple wall corrugated board (bottom).

The flat facing is called liner and the fluted medium is called fluting or corrugating medium.
Both types of papers can be produced from virgin or recycled fibres. The structure of corrugated
board is an adaptation of the engineering beam principle of flat load-bearing panels separated
by a rigid core. In this way much more bending stiffness can be gained at the same basis weight
compared to solid board.
The corrugated board is manufactured on a corrugator, Figure 10.2.
260

7 9

3 5 4 8 10
12 13 14

Figure 10.2. Manufacturing of corrugated board.

The process can be divided into six operations:

1. Unwinding and conditioning of the liners and medium
2. Corrugation of the medium
3. Gluing medium and liners together
4. Drying of the board
5. Cutting/slitting of the corrugated board. The board may also be creased in this process.
6. Stacking of sheets
The fluting is preheated (conditioned) and then passed through the hot (160 °C) corrugating
rolls where the flute profile is generated. Immediately after the glue is applied to the flute tips
and pressed together with the preheated liner. The introduction of heat is needed for the corru-
gation of the fluting, but also to achieve a fast cure of the glue. It is important that the glue de-
velop a sufficiently strong tack in a short time so that the single face board can be transported to
the bridge without separation of the two papers. This initial glue tack is called the green bond
and develops within milliseconds. Besides temperature the adsorption of the glue to the papers
governs this tack development. Normally the glue consists of 20 % modified starch dispersed in
80 % water.
The single face web is then glued to the second liner (double backer liner) after glue has been
applied to the flute tips. The single wall board is then dried under pressure on hot plates in the
dry end of the corrugator. Since the stiff corrugated board cannot be reeled it is slitted in length-
wise- and cut in crosswise direction to desired sheet dimensions before they are stacked at the
end of the corrugator. In connection with the slitting operation the board can be creased length-
wise to facilitate later folding of the board.
The corrugated board is made with different structures that are defined by the top-to-bottom
height of the fluting and the number of flute waves per unit length, see Table 10.2

q~ÄäÉ=NMKOK=aáÑÑÉêÉåí=ÑäìíÉ=íóéÉëI=ï~îÉ=ÜÉáÖÜí=~åÇ=åìãÄÉê=çÑ=ï~îÉë=éÉê=ãÉíÉêK

cäìíÉ=qóéÉ t~îÉ=ÜÉáÖÜíI=ãã kçK=çÑ=ï~îÉëLã

^ QIU NNM
` PIS NPM
_ OIQ NRM
b NIO OVM
c MIT PRM
d=C=k MIR RRM
 261

Beside the structure of the board, papers with a wide range of properties can be used. As-
sume that we have 20 different liners and 10 different flutings to make boards from. From these
papers it is possible to make 16 000 different boards of four different single wall flute types, or
1 600 000 different boards of two different types of double wall. These examples illustrate quite
clearly the possibilities to create boards that satisfy a wide variety of requirements.

NMKOKP==`çåîÉêíáåÖ

The runnability of packaging materials in packing lines is of greatest importance for the manu-
facturer. Knowledge and experience of the processes and issues affecting efficiency at each
stage of paperboard conversion is the key factor for runnability. Materials that not run well in
the converting machines will never come into the market.
The chain of conversion processes is shown in Figure 10.3.

print cut folding forming,

and and and filling and
varnish crease gluing closing

Figure 10.3. The chain of conversion processes.

In all operations the packaging materials should be consistent in moisture content, stiffness,
basis weight and thickness for best performance. They should also be free of debris and sheets
must be supplied flat.

NMKOKPKN==mêáåí=~åÇ=s~êåáëÜ=

Printing is an essential step in the conversion process. The requirement on the print quality is
high since consumers buying decision is influenced by the print quality.
The printing press consists schematically of the following units:
• Sheet or reel feed unit
• Printing units
• Sheet or reel collection unit
Sheet-to-sheet printing press is the most common used printing process for board and corru-
gated board. The reel-to-reel printing press is mostly used for liquid packaging board and pre-
printed liner. Reel-to-sheet printing press for board is usually used in combination with in-line
cutting and creasing equipment.
The main purpose of the printing units in the printing press is to transfer the ink to the board.
The ink has to be applied selectively to certain areas on the board in order to make an image.
The transfer of the ink to the board surface is done under pressure which is necessary for the ink
transfer to occur. The board is pressed between a cylinder with the ink and an impression cylin-
der. The printing units in a printing press consist typical of 4-6 units with vanishing.
The principles to transfer the ink to the board in the printing unit are:
262

• Flexography
• Offset lithography
• Rotogravure
In flexographic (flexo) printing ink transfer to the board is done by first transferring the ink
via an ink roll to a printing plate mounted on a cylinder. The plate distributes the ink onto the
board at the nip as schematically shown in Figure 10.4(a). The plate on the cylinder is flexible
and made of polymer or rubber. The plates are directly mounted to the cylinder by self-adhe-
sives or indirect by the use of sleeves.
In offset lithographic printing the ink is transferred to the board via a separate transfer cylin-
der covered with a rubber blanket as shown in Figure 10.4(b).

ink roll

board board

plate
ink roll cylinder

plate impression transfer impression

cylinder cylinder cylinder cylinder

a) b)

Figure 10.4. (a) Flexographic and (b) Offset lithographic principles.

In rotogravure printing the printing cylinder is made of chromed copper and the printing im-
age is a negative relief engraved or etched on the copper surface. Rotogravure printing is expen-
sive but can produce large number of prints at high speed.
The most common types of printing for paper board are sheet-fed offset (lithography) and ro-
togravure. Liquid packaging board is mostly printed in flexography. For general packaging the
use of flexography has been limited, however it is gaining increasing acceptance due to techni-
cal improvements that give better print quality. Lithographic printing gives in general a better
quality than flexographic printing. Varnishing of the printed board is done in order to protect
the print against rubbing that may occur in the packaging chain.
Runnability requirements in the printing and varnishing operations are flatness and dimen-
sional stability of the board, delamination strength in the thickness direction of the board and
dust and debris free board. Dust and debris cause spots in the image areas of the board. Prob-
lems with ink are relatively rare. If ink problems arise they are linked to taint and odours.
When printing corrugated board (post-print) the flexographic technique is by far the most
common. Also in pre-print (when the liner is printed separately before converted to corrugated
board) flexographic printing is the prevailing technique.
 263

NMKOKPKO==`ìííáåÖ=~åÇ=`êÉ~ëáåÖ

The purpose of cutting and creasing is to convert a sheet of packaging material into a so-called
blank, see Figure 10.5. After these operations, which are done at the same time, the blank has
the right dimensions and can be folded into a box. The creasing is necessary to get a god fold
and thus consistent dimensions of the boxes.

Figure 10.5. (a) Sheet of packaging material, i.e. blank and (b) blank folded into a box.

This process is the most critical when considering improvements in the line efficiency in the
folding, gluing and packing operations. The creasing rule must be of the correct specification
for the board type and quality. Changing the board specification without considering the crease
rule width, depth and profile will result in less satisfactory creases. The converter must be aware
of the correct cutter and creaser profiles and settings to use. This will affect the folding and glu-
ing operations and the final erection and sealing of the material. This is one of the major causes
of dimensional stability problems in board packaging.
Boards coated with polyethylene (PE), polypropylene (PP) and polyethyleneterephthalate
(PET) need special creasing and cutting forms, to ensure the cut is clean and the film is not
damaged.
There are two principal operations to cut and crease board and packaging materials,
• Flat bed and
• Rotary
The flat bed cutting and creasing principle is schematically shown in Figure 10.6. The cut
and crease tools are parallel to each other and consist of a male die with creasing rule and cut-
ting knife and a female die with scored channels. The sheet is placed between the dies. The male
die is punched into the sheet and female die. During the punch operation the sheet is creased
264

and cut by the ceasing rule and cutting knife, respectively. A flat bed cut and crease machine
can be place off-line after a sheet-fed printing press or in line with a reel-fed printing press.

cutting
die (1)

plywood
plate

board
sheet

counter
die (5)
rubber
creasing rule (5)
line (4)

creasing channel/groove (6)

cutting rule/ knife (2)

Figure 10.6. Principal sketch of flat bed cutting and creasing.

In the rotary cutting and creasing operation the male die and female die tools are mounted on
cylinders. The creasing and cutting of the board or packaging material occurs in the nip between
the cylinders. The benefits with rotary over flat bed cutting and creasing are the higher speed
and longer runs. The drawback is the cost for the rotary die tools. A rotary cut and crease ma-
chine is in-line with a reel-fed printing press.
To obtain a cutting that meets the requirements of distinct cut edges the cutting knifes on the
male die must be sharp. Over time the knife will wear out and blunt. This will lead to cuts that
produce dust and hairy cut edges.
During the punch operation the board is pre-stressed by tension acting in the plane of the sheet.
The male die is punched into the board and the board deforms under compression immediately
beneath the male die, see Figure 10.7 (a). Shear deformation occurs in the gap between the
male die and female die on either side. The shear deformation causes damage that is primarily
concentrated at the interfaces between the different layers in multi-ply board. The damage due
to the shear deformation causes delamination, which governs the quality of the crease, between
the layers. Upon unloading of the male die, most of the deformation is recovered. The remain-
ing deformation of the board is shown in Figure 10.7 (b).
Runnability requirements of the board in the cut and crease operations are dependent on the
moisture content of the board. For flatbed cutting and creasing the flatness of the board is an im-
portant requirement. Board properties that influence the results of the cutting, creasing and fold-
ing are the board thickness, and strength and elongation properties in the in-plane and out-of-
plane directions including fibre type and fibre orientation. Dust and debris also affects the run-
nability.
 265

a) b) c)

Figure 10.7. The behaviour of board during (a) creasing, (b) after creasing and (c) at folding.

For large carton blank formats it is especially important that the board is of the same consis-
tent stiffness and that creases are formed correctly.

NMKOKPKP==cçäÇáåÖ=~åÇ=däìáåÖ

Folding is done along the scored line as is shown in Figure 10.7(c). The folding bends the board
around the crease. The bending introduces compressive and tensile stresses in the creased zone.
Due to the shear damage along the interfaces obtained during the indentation, the compression
will cause the layers to buckle outwards separating the interfaces more and more at increasing
folding angle.
Badly formed creases along the score line can after folding cause a skewness that varies from
one side to another and gives a bad appearance of the folded scored line. Also surface cracks in
the sheet plane give bad appearance and functionality. A symmetric crease will on the other
hand give a scored line after folding with a neatly appearance and precisely located, and hence
more attractive. Figure 10.8 shows photographs of the appearance of the scored line after fold-
ing for one uncreased sample and four creased samples. The creased samples in Figure 10.8
vary from badly to excellently creased samples.

a) b) c) d) e)

Figure 10.8. Folding for one uncreased (a) and four creased samples (b)-(e). The creased samples vary from (b)
badly, (c) not well, (d) well and (e) excellently creased samples.

compression gluing prefolding carton

blanks

Figure 10.9. Side gluing.

266

Side gluing of carton blanks is the most frequently used pre-gluing of blanks before they are
shipped to the goods manufacturer. At side gluing one side of the blank is glued and the other
side is fold over and overlapping the glued side, as is shown in the gluing section in Figure
10.9.
The gluing machines used for side gluing consists of the operations:
• pre-folding
• application of adhesive
• folding
• sealing
• curing
In the gluing machine the blanks are fed into the pre-folding unit where the blanks are folded
90–180 degrees and thereafter folded back to the original position as shown in the pre-folding
section in Figure 10.9. The adhesive is applied at the length side of the blank and thereafter the
other side is folded to overlap the glued side and sealed under pressure, see the gluing section in
Figure 10.9. The pre-glued blanks are then stacked for curing and finally packed for delivery as
illustrated in the compression section in Figure 10.9.
The procedure of joining the two board sides together by gluing can be summarised in the
following steps:
• application of the glue on the board substrate
• wetting of the glue to spontaneously flow out on the board
• penetration of the glue in to the board
• curing the sealed board, so that the glue consolidate, followed by drying or cooling
The adhesives that are typically used for gluing are water based dispersion glue and hot melt
glue. The water based dispersion glue can be applied to the board directly at room temperature.
After gluing it has to be dried. The hot melt glue has to be heated up before it can be applied
onto the board.
If the board is coated with a thermoplastic polymer the sealing is done by melting the coating
and then joining the opposite board surfaces.
Important parameters for the performance in the gluing machine are:
• open time, i.e. the time from application of the glue to the sealing
• closing time, i.e. the time the sealed blank is under pressure
• pressure on the sealed zone
• amount of glue
• temperature
• speed of the gluing machine
It is important that the board surface where the glue will be applied shall not be varnished in
order to get good adhesion.
The consistency and quality of the print surface, crease profile, squareness, and the integrity
of adhesive joints are all-important factors when maximum efficiencies are being sought. Other
factors influencing improved efficiencies are the consistency of the board specification, espe-
cially stiffness and thickness, ambient conditions of the cartons – they must be stored at least
24 hrs in the packing area prior to use - and age of cartons - ideally not more than three months
old after gluing.
 267

NMKOKPKQ==cçêãáåÖI=cáääáåÖ=~åÇ=`äçëáåÖ

The basic function of a packaging machine system is to perform:

• Forming of the package
• Filling the package with a product
• Closing the package
There are two principal ways of feeding the packaging machine system with the packaging
material, i.e.
• Reel
• Blanks
At reel feeding of the packaging material into the packaging machine, the longitudinal form-
fill-seal is done by forming the packaging material from the reel into a tube over a forming col-
lar, see Figure 10.10. The next step is to form, fill and horizontally seal the package. This is
done during the downward movement of the packaging material. Horizontal knives cut the tube
to produce a bag-like package. Thereafter the package is either packed as it is or the package is
feed into a final forming unit where its final form is obtained. A well-known package filled with
this method is the Tetra Brik.

longitudinal seal

forming, filling and

horizontal seals

reel
folding unit

Figure 10.10. The packaging material in the packaging machine. The form-fill-seal is done by forming the pack-
aging material from the reel into a tube over a forming collar.

The blanks that are fed into a packaging machine system are either flat pre-glued blanks or
blanks. The blanks are erected to a rectangular or a square form. The product is filled into the
erected blanks and thereafter the package is closed and sealed, see Figure 10.11. There are two
ways of filling the container. One is the top-loading system were the container is bottom sealed
after the erecting and the top is opened for the product to be filled into the container. The other
268

is the end-loaded system, see Figure 10.11, were the product is first filled and then the ends are
closed and sealed.

Figure 10.11. The product is filled into the erected blanks and thereafter the package is closed and sealed.

The products that are filled into a container can be divided into three different groups of
goods:
1. Liquid and viscous products
2. Granulates and powder products
3. Solid products
The process of filling the package consists of three basic steps. The first step is transportation
of the product to the filling equipment, the second step is to control that the amount of product
that is to be filled in the package is correct and the final step is transfer of the right amount of
the product from the filling equipment to the package.
An additional process function in the packaging machine system is the condition under
which a product is filled into a package, i.e. if the product should be filled under vacuum, gas
injection, or sterilisation.
Factors to consider when selecting a packaging machine system are:
• What kind of product should be packed
• Which package material and package method should be used
• Which productivity demands do the market have on the product
• What package formats are demanded from the market
 269

NMKOKQ==_~êêáÉê=mêçéÉêíáÉë

Paper and board are porous and consists of fibres linked together to form a network. The voids
in the network make paper and board porous and hence sensible to gas and liquid mass trans-
port. A barrier on the paper or board is then used to protect and restrain the package material
from mass transport of gases and liquids. The requirement on boards with barriers is that the
barrier should fulfil the demands a product put on the package barrier properties.
Barriers on paper and board are applied on the surface of paper and board as extrusion coat-
ing, dispersion coating or lamination. At extrusion coating a molten high molar mass polymer
from an extruder is applied onto the surface of a moving paper or board web. Dispersion coating
is a technique, were the water-soluble barrier substrate is coated onto the surface of the board.
The dispersion coating yields a package material that is easy to re-pulp, thus favouring recy-
cling.
The most frequently used polymers at extrusion coating are low and high-density polyethyl-
ene (PE) and polypropylene (PP). Low density PE (LDPE) is the most used barrier for frozen
food and it resists all oils and greases in short term. High density PE (HDPE) is a more dens ma-
terial than LDPE and has better barrier properties than LDPE. At long term exposure the HDPE
has a higher resistance to cracking then LDPE. Higher demands on barrier properties of packed
products, than PE can offer, are fulfilled with PP. Typical end-use applications for PP-coating
are for example ready meals and pet food.
An overview of polymer coating, often PE, on both sides of the board and lamination of an
aluminium foil is schematically shown in Figure 10.12. At the left chill roll in the figure the
molten plastic is extrusion coated onto the surface of the board and an aluminium foil is lami-
nated onto the molten polymer. When the coated board leaves the chill roll the temperature in
the coated board is about 20–30 °C. At the left chill roll in Figure 10.12 the uncoated board side
is extrusion coated with a polymer. Thereafter the package material is rewound into a package
material roll. The procedure shown in Figure 10.12 is typical for converting of liquid board ma-
terials.
unwinding stand for
aluminium foil

molten molten
plastic plastic

rewind

chill roll chill roll

Figure 10.12. Extrusion coating and lamination of barriers on liquid board.

270

Lamination barriers that are used are aluminium foils, metallised polyester and metallised
oriented polypropylene. The metallised polyester and metallised oriented polypropylene are
laminated directly onto the surface of the board. Metallised foils have very high resistance
against gas mass transportation.

NMKP==oÉèìáêÉãÉåíë=çå=m~Åâ~ÖáåÖ=mÉêÑçêã~åÅÉ

NMKPKN==aÉÑáåáíáçåë

For the following a definition of the different packaging levels is needed. The definitions ac-
cording to the EU Directive on Packaging and Packaging waste define primary, secondary and
tertiary packaging as:

w p~äÉë=çê=éêáã~êó=é~Åâ~ÖáåÖW m~Åâ~ÖáåÖ=ÅçåÅÉáîÉÇ=ëç=~ë=íç=ÅçåëíáíìíÉ=~=ë~äÉë=ìåáí=íç=íÜÉ=
Ñáå~ä=ìëÉê=çê=ÅçåëìãÉê=~í=íÜÉ=éçáåí=çÑ=éìêÅÜ~ëÉK
w dêçìéÉÇ=çê=ëÉÅçåÇ~êó=é~Åâ~ÖáåÖW m~Åâ~ÖáåÖ=ÅçåÅÉáîÉÇ=ëç=~ë=íç=ÅçåëíáíìíÉ=~í=íÜÉ=éçáåí=çÑ=éìêJ
ÅÜ~ëÉ=~=ÖêçìéáåÖ=çÑ=~=ÅÉêí~áå=åìãÄÉê=çÑ=ë~äÉë=ìåáíë=ïÜÉíÜÉê=
íÜÉ=ä~ííÉê=~êÉ=ëçäÇ=~ë=ëìÅÜ=íç=íÜÉ=Ñáå~ä=ìëÉê=çê=ÅçåëìãÉê=çê=áí=
ëÉêîÉë=çåäó=~ë=~=ãÉ~åë=íç=êÉéäÉåáëÜ=íÜÉ=ëÜÉäîÉë=~í=íÜÉ=éçáåí=çÑ=
ë~äÉëK=fí=Å~å=ÄÉ=êÉãçîÉÇ=Ñêçã=íÜÉ=éêçÇìÅí=ïáíÜçìí=~ÑÑÉÅíáåÖ=áíë=
ÅÜ~ê~ÅíÉêáëíáÅëK
w qê~åëéçêí=çê=íÉêíá~êó=é~Åâ~ÖáåÖW m~Åâ~ÖáåÖ=ÅçåÅÉáîÉÇ=ëç=~ë=íç=Ñ~Åáäáí~íÉ=Ü~åÇäáåÖ=~åÇ=íê~åëéçêí=
çÑ=~=åìãÄÉê=çÑ=ë~äÉë=ìåáíë=çê=ÖêçìéÉÇ=é~Åâ~ÖÉë=áå=çêÇÉê=íç=éêÉJ
îÉåí=éÜóëáÅ~ä=Ü~åÇäáåÖ=~åÇ=íê~åëéçêí=Ç~ã~ÖÉK=qê~åëéçêí=é~ÅâJ
~ÖáåÖ=ÇçÉë=åçí=áåÅäìÇÉ=êç~ÇI=ê~áäI=ëÜáé=~åÇ=~áê=Åçåí~áåÉêëK
w m~Åâ~ÖáåÖ=ëóëíÉãW qÜÉ=ÅçãÄáå~íáçå=çÑ=éêáã~êóI=ëÉÅçåÇ~êó=~åÇLçê=íÉêíá~êó=é~Åâ~ÖJ
áåÖ=íÜ~í=áë=ìëÉÇ=áå=íÜÉ=ëìééäó=ÅÜ~áå=Ñçê=~=ÅÉêí~áå=éêçÇìÅíK

In addition to these definitions of packaging levels a load carrier is often used. The most
common example is here the standardised pallet. The pallet is made of either wood or plastic
materials. The pallets constitute the unit load and are easy to handle with for example a fork lift.
The packaging system has to fulfil a number of different requirements throughout the supply
chain. All different requirements shall be met by the performance of the packaging. The perfor-
mance, on the other hand, is related to one ore more properties of the packaging, e.g. stiffness.
Finally, the material and structural properties constituting the packaging give the packaging
properties.

NMKPKO==qÜÉ=pìééäó=`Ü~áå

To understand the packaging requirements, expressed by the actors in the distribution system,
calls for a description of the complete packaging flow. The right box in Figure 10.13 illustrates
the flow of products from the manufacturer/filler to the end users. In order to facilitate the dis-
tribution and protection of the products different aspects on the requirements are set by the dif-
ferent actors in this supply chain. The left box in Figure 10.13 illustrates the suppliers of
packaging materials that deliver engineered packaging with adapted performance. As an exam-
 271

ple a manufacturer of toothpaste has to contact suppliers of plastics, carton board and corrugat-
ed board.

Figure 10.13. The supply chain.

The operations in different parts of the supply chain that result in the requirements on the
packaging are described below.
• The manufacturer: The product is normally first packed into sales or primary packaging,
which thereafter is packed into a grouped or secondary packaging. The secondary packag-

manufacturer wholesaler retailer end-user

transport

transport

transport
storage
packaging levels

Figure 10.14. Illustration of material flow from manufacturer to end-user.

272

ing facilitates the handling at both wholesalers and retailers. To simplify the transport from
the manufacturer to the wholesalers the secondary packaging is packed into a transport/ter-
tiary packaging. Normally the tertiary packaging is stacked on a pallet (load carrier). The
packaging levels are schematically illustrated in Figure 10.14.
• The Wholesaler: The tertiary packaging is opened and the secondary packaging is reloaded
into new transport packaging (i.e. roll cages, Figure 10.15) for distribution to retailers, see
Figure 10.14. Waste handling of packaging materials is an important task for the wholesal-
ers. It should be noticed that most of these handling operations are performed manually.

Figure 10.15. The most common transport system from wholesaler to retailer.

• The retailer: The tertiary packaging and most often the secondary packaging is unpacked.
The primary or secondary packaging is then put on the shelves, see Figure 10.16. These

Figure 10.16. Point of sales at retailer.

 273

processes are time consuming since all the handling is done manually. Waste handling of
packaging materials is an important and time-consuming task.
• Transport and Storage: The packaging is transported between the manufacturer, wholesaler
and retailer by different types of vehicles (truck, train, boat, air). Reloading occur a number
of times along the transport chain and is mostly done mechanically by forklifts. During
transport the packaging is subjected to both static and dynamic loads as well as changes in
surrounding environment (humidity, temperature, light etc.) During intermediate storage
the impact on the packaging consists of static loads and changes in environment.
• The end user/consumer: The primary packaging must clearly inform about the product. It
must be easy to open the package and to get rid of the packaging material. The retailers
offer both primary and secondary packaging, e.g. one bottle of soda and a six-pack of soda.
Some products, especially industrial products, are delivered directly from the manufacturer
to the end-user, for example products from subcontractors to the car industry.

NMKPKP==m~Åâ~ÖáåÖ=oÉèìáêÉãÉåíë

The fundamentals for packaging is to transport, serve and sell a product. This should be done in
the most economically efficient way and at the same time taking environmental aspects into
consideration.
The product primarily gives the requirements on packaging. From the manufacturer’s point
of view the product should be given an optimal combination of primary, secondary and tertiary
packaging – a so-called packaging system – has to be designed. In doing this all different re-
quirements along the supply chain has to be considered.
A systematic illustration of the packaging requirements is shown in Figure 10.17. This sys-
tem is based on interviews with employees at the different actors in the supply chain for food.
The systematisation consists of five groups, i.e. product flow, protection, runnability, environ-
ment and communication. Each group has different requirements that are linked to the supply
chain.

1. product flow – handling

– transport
– storage
2. protection – mechanical
– barrier
3. runnability – stiffness
– flatness
4. environment – material savings
– environmental concern
5. communication – information
– promotion

Figure 10.17. Systematisation of packaging requirements during the supply chain.

The terms technical and marketing requirements will be introduced to differentiate the re-
quirements into two groups. Technical requirements refer to physical properties that protect and
274

facilitate the distribution of products. Marketing requirements are related to the customer end-
use and the marketing of the product. Marketing requirements are on the credit side of packag-
ing.

NMKPKPKN==mêçÇìÅí=cäçï

Product flow includes all physical aspects of taking a product from the manufacturer to the end-
user. During this process a number of handling operations take place at different packaging lev-
els. At the manufacturer this handling is mostly done automatically due to the fact that large
quantities of identical products packed in the same way are produced at one place. Most com-
mon is to use pallets for transportation to the wholesaler, where the pallets are stored before re-
loading and shipping of the packaging to retailers.
In order to make the transportation efficient the capacity of carriers has to be utilised to its
maximum. In the food industry most often the volume limits the amount of goods to be trans-
ported in a carrier, rather than the weight. Volume efficiency is a measure of the utilisation of
cubic capacity. Since the pallet (EU-pallet) is the most common load carrier, the dimensions of
transport packaging are crucial. The outer dimensions must be adjusted so that the footprint of
one or more packaging is the same as the dimension of the pallet. As a consequence, not only
the dimensions of secondary and primary packaging, but also the product itself has to be consid-
ered for highest total volume efficiency. Also the height of the transport packaging must be ad-
justed in such a way that two or more pallets stacked on top of each other utilise the full height
of a carrier. The dimensions at different packaging levels are not that crucial when the limiting
factor for the carrier is weight.
At the wholesaler the incoming unit loads (pallets) are broken down, sometimes repacked,
and reloaded onto a new load carrier. The new load carrier is adapted for the specific demands
of supply of the individual retailers. This is often a pallet or in the retail business a roll cage,
which is then shipped to the retailer. At the retailer the unit is broken down and tertiary and sec-
ondary packaging opened and the primary packaging (sales packaging) put on the shelves. At
the wholesaler an extensive labour intensive manual handling starts, which then ends at the end-
user who finally opens the primary packaging. All this handling introduces a number of differ-
ent aspects on packaging requirements, not only on handleability, but also on management of
packaging waste.
Handleability is a wide concept that includes a number of requirements on packaging such
as:
• to grip
• to stack
• to open and fold
• waste handling
• recycling of material or energy re-use
• packaging re-use
 275

NMKPKPKO==mêçíÉÅíáçå

Product protection includes the interaction between the product and the protective power of the
packaging towards different external impacts in the packaging system. These impacts can be
mechanical or related to the surrounding climate (temperature and humidity). Light and other
electromagnetic radiation fields are other example that could harm the product. For other prod-
ucts chemical and/or biochemical impacts (oxygen, mould) are crucial. Another aspect of pro-
tection is that the packaging itself should not contaminate the product e.g. chemically or
biochemically.
The protection of the product is essential in packaging design – both from an economical and
an environmental point of view. The consequences of insufficient barrier properties may result
in large waste of food (i.e. milk) – and the resources that were taken into account to produce the
food. If the mechanical protection fails the product may even contaminate land areas (e.g. toxic
chemicals).

NMKOKPKP==oìåå~Äáäáíó

The runnability of packaging materials in the packing lines is of greatest importance for the
manufacturer. Materials that do not run well in the machines will never enter the market.

NMKOKPKQ==båîáêçåãÉåí~ä

There are two aspects on environmental requirements on packaging. One is related to protection
and the environmental consequences if the products are destroyed or contaminate the environ-
ment in case of inadequate protection. The other aspect deals with the packaging itself and is re-
lated to both the production of packaging and the use of it. Environmentally friendly and
efficient use of processes, raw materials (including energy) and additives is expected in the pro-
duction of packaging materials and packaging. In the design of a packaging system the goal is
minimum use of packaging materials, which saves both resources and minimise the amount of
waste. Furthermore the packaging should support re-use or recycling.

NMKOKPKR==`çããìåáÅ~íáçå

The requirements on communication of packaging can be divided into two aspects. One is relat-
ed to the distribution and involves requirements on flow information. Here the packaging
should carry information so that the right products are sent to the right place and that they are
handled in the correct way. Different requirements prevail on different packaging levels.
The other aspect of communication requirements relates to marketing performance. These
requirements can be divided into three different groups:
• Product Information
• Selling Capacity
• Safety
276

The first group includes requirements on description of both the product and the packaging.
Examples are; declaration of content, how to use, last date of consumption, composition of
packaging etc. The requirements on selling capacity are much differentiated between products.
Most aspects are related to packaging design and graphic design with the purpose to increase
the attractiveness of the product. Safety requirements may include aspects on protection against
theft and childproof design and components.

NMKQ==aÉëáÖå=çÑ=m~Åâ~ÖáåÖ=mÉêÑçêã~åÅÉ

Packaging performance denotes the ability of a packaging to satisfactory manage the different
tasks as they are defined by the requirements. Packaging performance is thus the tool to match
all the requirements in the most efficient way.
Packaging performance depends on:
• The properties of a single packaging.
• The interplay between product and packaging and the interplay between packaging at dif-
ferent levels (primary, secondary, tertiary).
By properties it is understood the characteristics that contribute to the performance of a sin-
gle packaging. These properties can often be measured and thus quantified. There are only two
ways to influence the packaging properties and in the extension the performance of a packaging.
These are:
• Material properties
• Structural design
Consider as an example a bridge. The structural design of the framework and the materials
used (including joints) give the properties that govern the performance. Performances that shall
match requirements such as maximum load bearing capacity and deflection, torsional stiffness,
colour and external attractiveness. In packaging, material properties like stiffness, friction and
barrier are examples that mainly impact the technical performance and colour and print have
their main impact on marketing performance. A package must be assembled in such a way that
the joint itself is not the weak point. Structural design influences the volume efficiency but also
to a high degree marketing performance.
To optimise packaging performance is a fairly complex process, which calls for a holistic ap-
proach, i.e. an approach that involves both the product itself and the distribution chain. Some-
times different requirements result in contradictory demands for packaging properties.
Examples are: less packaging material vs. high stacking strength, excellent print vs. low materi-
al cost, easy-to-open vs. high packaging stiffness etc. From this it follows that the design in-
volves different knowledges and skills and that it involves a number of compromises in order to
achieve the best cost/performance for a packaging system.
In the following sections technical and marketing performances will be discussed separately.
 277

NMKQKN==qÉÅÜåáÅ~ä=mÉêÑçêã~åÅÉ

Technical performance is the combined packaging properties that sufficiently manage the tech-
nical requirements, see Figure 10.18. By changing the material properties and the structural de-
sign it is possible to influence the packaging properties and thus the performance of the
packaging. By considering the interaction between the product and the primary, secondary and
tertiary packages it is possible to design a packaging system that is optimal for the technical re-
quirements. A successful optimisation of a packaging system is dependent on how well the
technical requirements are defined quantitatively and how well the packaging performance can
match these requirements.

packaging requirements

packaging performance

packaging properties

material properties structural design

Figure 10.18. Packaging requirements versus material properties and structural design.

The basic procedure for design of the technical performances includes:

• Theoretical predictions (if possible)
• Verification of performance in laboratory
• Full-scale tests
The process is most often iterative and contains elements of „Design – Build – Test“. To
avoid sub-optimisation a final evaluation of the performance versus all requirements must be
made, with the purpose to find a packaging solution that satisfactory fulfils these requirements.
This step involves compromises between contradictory requirements.
These procedures for packaging design do not in principle differ from those used in aircraft
and car industries. It starts with identification of requirements and then the task is to find a solu-
tion for performance that satisfy these requirements. The engineering follows the same routes
and the design variables are material properties and structural design. Laboratory and full scale
testing follow classical methods for mechanical analysis, i.e. the finite element method. During
the last ten years an increasing amount of the engineering is done by the help of simulation
methods run on computers that speeds up the development phase considerably. This evolution is
due to better software and constitutive (material) models. The same trend can be followed in the
278

packaging business although this industry is some years behind the automotive industry primar-
ily due to the lack of advanced constitutive models for paper and board.
To illustrate the procedure for optimisation of the technical performance the stacking perfor-
mance during transport and storage will be discussed, i.e. the resistance against damage due to
compression loads. For this general discussion we assume a package on any level of packaging
(primary, secondary or tertiary). The packages are transported and stored stacked on a stan-
dardised pallet. The packages are staked on top of each other either as tower stacking or inter-
lock stacking, see Figure 10.19. The interlocked stack gives a more stable stack since the
individual boxes are constrained by their neighbours.

a) b)

Figure 10.19. (a) Tower and (b) Interlocked stack pattern.

During transport and storage the packaging will be subjected to static loads, vibrations,
shocks, and environmental changes. Static loads are given by the weight of the packages
stacked on top of the bottom package, and dynamic loads are given by a combination of vertical
and horizontal accelerations. The static and dynamic loads affect the choice of both the structur-
al design and material properties whereas changes in relative humidity affect primarily the ma-
terial properties.
To illustrate the procedure for packaging design only static compression loads will be dis-
cussed in the following. For a more complete design it is necessary to include the dynamic forc-
es during transport and a separate analysis must be made for these requirements. If the
magnitude of the shocks and vibrations are known a new estimation of the requirements are
done, which then should be matched by the stacking performance of the packaging.

NMKQKNKN==pí~íáÅ=`çãéêÉëëáçå=iç~Çë

The packaging must be engineered so that it can handle the static top-load compression during
transport and storage. The top-load compression is estimated by answering the question:
• What is the top-load compression of the most stressed package in the pallet?
The most stressed packages on a pallet are those in the bottom layer. The number of layers
and the weight of each packaging then give the top-load compression. To account for uncertain-
 279

ties due to impact of relative humidity, imperfections in stacking or shock loads a safety factor
is used. The safety factor, which is most often based on experience, is multiplied with the calcu-
lated load to give the final requirement on compressive strength.
From the estimated top-load compression strength value and the packaging geometry the ma-
terial needed can be calculated. The theoretical calculations can be divided into:
• Empirical analytical calculations
• Numerical simulations based on the finite element method
The empirical analytical calculation is often based on the McKee formulae. The McKee for-
mulae are calibrated to laboratory compression tests of packaging. In recent years analysis
based on the finite element method has been used more and more as an alternative to empirical
analytical calculations.
Investigations where results from finite element analysis are compared with experimental
observation show that use of formulae based on buckling theories, e.g. the McKee formulae, are
not able to predict neither the stiffness nor the maximum box compression strength of packag-
ing with short panels. Short panels are panels with a height (H) to width (W) ratio less than one,
i.e. H/W < 1, if the top-load compression is in the height direction.
Below we will give an example of the design procedure for a specific corrugated transport
packaging. For good reasons it is expected that the most severe requirements on compression
strength prevail during transport or storage. The assumed boxes including the product, with a
mass of say 10 kg, are stacked on a pallet with a height of six layers. This gives a load on the
bottom box of five boxes equal to 50 kg. During transport two pallets are stacked on top of each
other, which give a maximum load on the lowest box of 110 kg, i.e. 11 boxes multiplied with
the weight of the package 10 kg. (The weight of the pallets is here neglected.) During storage,
however, three pallets are stacked on top of each other, which correspond to a maximum load of
170 kg. This may lead to the conclusion that the compression performance should be 170 kg –
at minimum. However, during transport the boxes are subjected not only to a static load, as in
storage, but also to vibrations and shocks. These dynamic impacts in the vertical direction may
result in a requirement for compression strength that is higher than 170 kg although the static
load is only 110 kg during transport.
Next will follow a more general discussion on compression requirements on packaging dur-
ing transport and storage, see Figure 10.20. The solid line represents the stacking requirement
during different situations along the distribution chain. In position A the requirements represent
the compressive load of three pallets stacked on top of each other. Position B refers to two
stacked pallets in a high humidity storage. After some time they are rearranged and thereby sub-
jected to vertical vibration loads. Finally, in position C one pallet is handled carelessly and sub-
jected to a vertical shook. The dotted lines in Figure 10.20 (1, 2 and 3, respectively) represent
three different boxes engineered to have three different levels of strength.
280

2
A
1
B

Figure 10.20. Compression requirements (solid line) and performance (dotted lines). The (1), (2) and (3) dash-
dotted lines correspond to different boxes.

Let us first consider the corrugated box represented by the dash-dotted line (1) in Figure
10.20. In position A the impact of three stacked pallets exceeds the property of the boxes in the
lowest layer on the lowest pallet and a collapse of the box will occur. A new box with higher
box strength is then evaluated, i.e. the box denoted by the line (2). In a position after A on the
solid black line the relative humidity in the environment is temporarily increased, which results
in a decrease in the compressive properties of the box, the step down of the line (2). This will
cause a box collapse in position B. A box with even higher strength has to be considered to
manage the requirements in position B. The box denoted by the dash-dotted (3) line is engi-
neered. This box withstands the impacts in position B. Thereafter the relative humidity is de-
creased to its initial value, which results in an increase in the compression properties of the box.
The box recovers more or less its original strength. As a result of this recovery the box also
manages the requirements from the shock loading at position C.
The same design procedure shall be used for other requirements along the supply chain, i.e.
to find the dimensioning requirements for every specific property of the packaging. This way to
design the packaging solution is the fastest way but the requirements must be known quantita-
tively. In addition to this the different properties of the packaging must be possible to calculate
from material properties and structural design.
In most situations neither the requirements are known nor is it possible to predict all require-
ments from material properties and structural design. Consequently other procedures than the
one described above must be used. These involve both laboratory testing and full-scale trans-
portation tests.
To finally verify the design a full-scale transport test of box (3) in Figure 10.20 has to be per-
formed. This is the normal procedure because large values may be lost if the boxes fail. For this
reason „dummies“ of the same weight as the product sometimes substitute the product in these
tests.
In the manufacturing of packaging there is always a distribution in the different properties
due to variations in the process and the material. This is also true for the requirements. Consider
 281

for instance the impact of change in relative humidity or dynamic loads. These may vary due to
time of day/year or traffic situation. If these distributions are Gaussian (see Figure 10.21) it is
possible to estimate the percentage of box failures from statistical considerations. If for instance
the standard deviation is V % for both the requirement and property, and the mean of the prop-
erty exceeds the mean of the requirement by four, i.e. 5.2V see Figure 10.21, the frequency of
failure will be 2 % (+/– 2,6V includes 99 % of all values). If the mean value of the requirement
and the property is the same, 50 % of the packaging will by definition fail („underpacking“). If
the performance exceeds the requirement enough no failures will occur („overpacking“). Over-
packing, to use more packaging material than necessary, implies expensive packaging, in-
creased costs for transportation and higher environmental impact. Underpacking may result in
damaged or wasted products, which results in delayed deliveries, bad will and in the end disap-
pointed customers. Underpacking is most often very expensive for the producer. Also from an
environmental point of view underpacking may be harmful due to waste of production resourc-
es for the product, and eventually contamination of land (in case of harmful products).

mean requirement mean property

2.6 sr  sr  sr 2.6 sr
2.6 sp  sp  sp 2.6 sp

Figure 10.21. Gaussian distribution of possible box failures.

The balance between over- and underpacking is an important consideration in design of

packaging systems. These considerations can be summarised as shown in Figure 10.22, where
the relation between total cost and product protection (red line) is outlined. Increased protection
asks for more packaging material and thus higher costs (blue line). On the other hand less pro-
tection increases the cost for damaged products (black line). This increase is normally more pro-
nounced than the increase in cost for packaging.
Life Cycle Analysis (LCA) is a tool by which the environmental consequences systematical-
ly can be analysed. LCA takes into account the consumption of renewable and non-renewable
resources and emissions to air and water. It is used to compare the consequences of the produc-
tion of different packaging solutions or to evaluate the consequences of different packaging sys-
tems.
282
total cost

damage packaging cost

cost

product protection

Figure 10.22. Environmental exposures versus package performance.

NMKQKNKO==fåíÉê~Åíáçå=_ÉíïÉÉå=m~Åâ~ÖÉë

Let us consider two different products, canned tomatoes and potato chips, respectively. The can
is stiff and has high compression strength and can thus fulfil all compression requirements from
stacking. The packaging solution can then be an open try with a wrap-around in plastic that fa-
cilitates the handling of e.g. 12 cans. This packaging constitutes the grouped/secondary packag-
ing as well as the transport/tertiary packaging. At the retailer the wrap-around is easily removed
and the sales packaging can be put on the shelves.
In the other example, potato chips in a paper bag, no compressive load can be taken by the
sales/primary packaging. A secondary packaging is needed to make it easy to handle the bags
and satisfy the compression requirement that exists from producer to retailer. These demands
create a need for a separate tertiary packaging. In the first mentioned case the cans take 100 %
of the load and the transport packaging (wrap-around) 0 %. In the second case the reverse is val-
id, the transport packaging takes 100 % and the primary packaging 0 %. However in many situ-
ations packaging at more than one level contribute to the overall compression strength.
Figure 10.23 illustrates schematically a laboratory compression test where the compression
load versus the deformation of the package is recorded for some primary package and second-
ary corrugated package.
In Figure 10.24(a) the typical load-deformation curves for the primary and the secondary
packaging are shown in the same graph. The figure shows that the entire compression load-de-
formation curve for the primary package is completed before the secondary package carries any
substantial load. Thus, we have a situation where the primary packaging is damaged at a load
Qa, see Figure 10.24(a).
 283

a) b)

Figure 10.23. Load versus deformation during compression. (a) Primary and (b) secondary packaging.

Qb

Qa

a) b)

Figure 10.24. Influence of head space on box collapse loading. (a) Zero head space between the primary and sec-
ondary packages. (b) Head space between the primary and secondary packages.

However, if we during packing leave a space between the primary and the secondary packag-
ing this is equivalent to moving the load-deformation curve for the primary packaging in Figure
10.24(b) to the right. The space between the primary package and the secondary package is
called headspace. During loading the packaging system interacts, i.e. first the secondary pack-
284

age will carry the load and experience the deformation, and when the deformation has exceeded
the headspace the primary package will start to deform. In the example given in Figure 10.24(b)
a load Qb can be achieved, which is substantially higher than Qa, even before the primary pack-
aging is deformed at all.

NMKQKO==j~êâÉíáåÖ=mÉêÑçêã~åÅÉ

Marketing performance is the combined packaging properties that sufficiently meet the market-
ing requirements (c.f. Figure 10.17).
Packaging is often the only tangle point of difference between products. This is particularly
true for many food products, beverages and personal health-care products. Thus, product devel-
opment is often limited to a new design of packaging.
Packaging is a very cost-effective marketing tool. In a self-service environment the packag-
ing has a very strong impact on buying decisions. After purchase, the packaging can also be
used to inform about and promote other products and stimulate a repetitive purchase. The pack-
aging is the „Silent Salesman“!
The tools available to affect packaging properties and change the performance are packaging
design, material properties and graphic design. The process to create good marketing perfor-
mance is very much a dialogue between manufacturer and packaging supplier. The evaluation
of a new or changed marketing performance is not easy to perform; it takes time and is most of-
ten related to sales volumes.
The requirements for high quality print demand specific properties of the packaging surfac-
es. These properties are dependent on the printing method used.

NMKR==cìíìêÉ=qêÉåÇë

NMKRKN==aáÖáí~ä=mêáåíáåÖ=~åÇ=bJéêáåí

A future trend, in the development of the existing printing technologies with ink on paperboard,
is the use of computers in the process. By digital printing technologies the design image is print-
ed directly onto the package. The process is similar to the process of printing a document from a
PC to a desk-top printer. The trend in the market is that digital printing technologies increase
their market shares. This technology enables the possibility to make every printed package
unique.
New technologies to display images on packages emerge and today some very interesting
new technologies have been developed, for example metallic ink printed onto paper to form cir-
cuits and conductive polymers that are laminated onto paper.

NMKRKO==bJí~Öë

In the supply chain the products physically flows in the direction from the producer to the end-
user. The product information flow is in the reverse direction, i.e. from the end-user to the pro-
 285

ducer. The information of the products is done by uniquely identifying each product in the
packaging chain at each packaging level.
Today the market uses the EAN-UCC system for product identification. The development of
Internet has triggered standardisation organisations, companies and research teams to link the
product identification to the Internet. The research today is concentrating around the product
identity code that will link the product to stored information on the Internet and development of
a mark-up language for describing the physical objects on Internet.
The physical medium that can be used for identification is divided into the following groups
• Printed barcodes
• Radio Frequency Identification (RFID)
• Magnetic strips
The barcodes is printed on a label or directly onto the package, and the key feature of bar-
codes is that the barcode is a standardised and very inexpensive system. The drawback of the
printed barcode is that it is sensible to dirty or wet environments.
The RFID-system and magnetic strips are not so often used as barcodes in identification of
products. This is mainly due to the cost.
The RFID-system consists of a transponder unit and a read/write unit. The transponder unit
is often applied onto a label that is put onto the product. Both the transponder and the write/read
units have internal intelligence and an antenna that receives, sends and stores data.
The key features of RFID-technology are:
• Data can be stored on the object and changed.
• The transponder unit can be read through water, dirt and other non-metallic surfaces.
• Identification of several objects simultaneously is possible.
The drawback of the RFID-system is today that the cost for the transponder unit, with a chip
and an antenna, is much higher than the cost for a printed barcode.
Magnetic strips will in the future probably be replaced by the RFID-technology in combina-
tion with barcodes.

NMKRKP==pã~êí=m~Åâ~ÖÉë

A smart package is defined as packaging that has build-in intelligence. The new possibility,
with cheap computer chips that are placed on packages, allows a new range of consumer driven
applications. For example one can design chips that
• measure moisture and temperature and from this information can display in real time the
shelf life of the product
• adjust the microwave oven to automatically prepare the cocking
This form of interaction between the package and its surrounding will enhance the perfor-
mance of the package for the consumer.
286

NMKRKQ==kÉï=m~Åâ~ÖáåÖ=j~íÉêá~äë

The performance of packages depends on both the material properties and the formability of the
packaging material.
Today, paperboard is folded into a package. The sides of the package consist of flat panels.
There is an increasing market demand for a change in geometrical shape of packages. One way
to increase the flexibility for the designers of packaging systems is to form the sides into more
curved panels. To be able to do this with paper and board the permanent elongation of paper and
board, without damaging the material, needs to be increased.
A sandwich construction consists of two stiff outer skins that are separated by a more com-
pliant core. The stiffness of the board relies on the stiffness of the outer skins and the distance
between the skins. Sandwich constructions of board materials are multi-ply board and corrugat-
ed board.
In corrugated board the stiffness is primarily given by the distance between the two liners on
each side of the corrugated fluting. During the last hundred years a wave pattern of fluting has
separated the liners. A more suitable pattern consists of a honeycomb pattern. At the Catholic
University in Leuven, Belgium a continuous process to form the core into a honeycomb type of
pattern for corrugated board has been developed. By gluing liners on each side of the honey-
comb pattern a very stiff corrugated board is obtained, not exhibiting some of the drawbacks of
the wave pattern.

NMKS==iáíÉê~íìêÉ

For further studies of packaging material in the supply chain the following literature is recom-
mended.
Förpackningslogistik, andra utgåvan, Packforsk, 2000.
Fundamentals of Packaging, Ed. F. A. Paine, Institute of Packaging, Brookside Press Ltd, Leic-
ester, 1985.
Packaging Foresight, Packa Futura 2001, Carl Olsmats, Packforsk 2001.
Packaging in the 21st Century, A. Stirling-Roberts, J. Prebble and P. Page, Pira International
Ltd., 2001.
Papermaking Science and Technology, Book 12, Paper and Paperboard Converting. A. Salvol-
ainen (ed.), Fapet Oy, 1998.
Papermaking Science and Technology, Book 18, Paper and Board Grades. H. Paulapuro (ed.),
Fapet Oy, 2000.
 287

NN==i~ãáå~íÉ=qÜÉçêó=Ñçê=m~éÉêã~âÉêë

Christer Fellers
Innventia AB

11.1 Introduction 288

11.2 Definition of a Laminate 289

11.3 Stress and Strain for a Layer in a Laminate 292

11.4 Matrices 294

11.5 Comparison between Orthotropic and Isotropic Materials 295

11.6 Strain in the Plane due to Moisture Absorption 296

11.6.1 Strain in the Plane due to an External Mechanical Force 296
11.6.2 Total Strain in a Layer in the Plane due to both Moisture Sorption and Mechanical
Forces 297
11.6.3 Strain due to Bending 297

11.7 Total Strain in a given Layer due to Moisture Changes, Mechanical Influences and
Bending 299

11.8 Forces and Moments 299

11.9 Laminate Theory where the Layers are Oriented with an Arbitrary Angle in
Relation to the 1-Direction (MD) 301

11.10 Transformation of Stresses and Strains 301

11.11 Transformation of the Hygroexpansion Coefficient 302

11.12 Transformation of the Stiffness Matrix, Q 302

11.13 Forces and Moments 303

11.14 Bending Stiffness and Tensile Stiffness 303

11.15 Simplified Expression for Bending Stiffness 305

11.16 Calculation Example for Bending Stiffness 306

11.17 Curl and Twist, Strain, Shear, as a Result of a Change in Moisture 307
288

11.18 Curl and Twist 308

11.19 Leaning Stacks of Folded Printing Papers 308

11.20 Tensile Stiffness and Bending Stiffness Studied in a Polar Diagram 309

11.21 Hygroexpansion, Curl and Twist Studied in Polar Diagrams 309

11.22 Hygroexpansion Orientation and Tensile Stiffness Orientation 310

11.23 Shear-free Angle for Hygroexpansion 311

11.24 Twist Free Angle 311

11.25 Finding the Height at a Point on a Bent Paper 313

11.26 Stress and Strains in the Thickness Direction of the Laminate due to
Hygroexpansion 314

NNKN==fåíêçÇìÅíáçå

In order to be able to calculate the bending stiffness of a paper and its tendency to curl and twist,
knowledge of the paper's elastic and hygroscopic properties is required. These properties are
discussed in this chapter.
When the moisture content in paper or carton board changes, the sheet often does not remain
flat. The term curl is used if the sheet bends in the machine-direction or cross direction of the
sheet, MD or CD, and the term twist is used to indicate that the sheet becomes twisted like a
propeller. In corrugated fibreboard, the term warp is usually used to describe this type of bend-
ing. The reason why the sheet bends when the moisture content changes is that the hygroexpan-
sion is different in the different layers of the sheet. Twist or warp occurs if one or more of the
layers has an oblique symmetry with respect to the machine direction.
Curl and twist in paper, carton board and corrugated fibreboard can cause great problems in
the converting and use of these materials, and can lead to great costs as the result of complaints.
The most important mechanical property of a paper, board or corrugated fibreboard is the bend-
ing stiffness of the sheet.
With the help of laminate theory, it is possible to calculate how a sheet material such as pa-
per is deformed as a result of external forces and of the internal forces caused by hygroexpan-
sion in the different layers of the sheet. A paper is regarded as a layered structure with different
properties in the different layers.
With the help of laminate theory, it is possible to do several things:
• Make sensitivity investigations of the causes of curl and twist
• Calculate bending stiffness.
A separate computer program has been developed to make it possible to carry out the calcu-
lations rationally.
 289

Examples of problems where the moisture has an influence:

• Multi-colour offset printing
• Continuous stationery for computers
• The twist and curl of corrugated fibreboard
• Twist and curl and bending stiffness can be calculated with laminate programs.

NNKO==aÉÑáåáíáçå=çÑ=~=i~ãáå~íÉ

• A laminate consists of a number of uniformly thick layers which do not move in relation to
each other.
• The laminate is ordered in a coordinate system 1-2, with the 3-axis, the Z-direction,
directed downwards, Figure 11.1.
• Each layer is assumed to be orthotropic, i.e. to have different properties in any two mutu-
ally perpendicular directions.
• Each layer is ordered in a coordinate system x-y (on-axis), which lies parallel with the 1-2
system (off-axis).
• The coordinate system x-y of each layer can be displaced at an angle relative to the coordi-
nate system 1-2.
• The positive angle is measured clockwise from the 1-system to the x-system.

␸
x

y
2

Figure 11.1. Each layer lies in an x-y-coordinate system. The layer is placed in a 1-2 coordinate system according
to. In the case of a paper, the sheet is placed so that 1 = MD and 2 = CD.

Figure 11.2 shows a laminate consisting of n layers. The origin lies at the geometrical mid-
point of the laminate. The layers are numbered from the top and downwards. Note that the
thickness coordinates are calculated from the middle to the lower interface of the layer.
The thickness coordinates are calculated as follows.

n
t ¦ tk (11.1)
k 1
290

z0 2 t/ 2
z1
z2 3 t
zk t/ 2
k
zN

Z, 3

Figure 11.2. A laminate consisting of n layers. The origin lies at the geometrical mid-point of the laminate. The
layers are numbered from the top and downwards.

stress N1 and strain e10 in the 1-direction

2
N2

stress N2 and strain e20 in the 2-direction

1
shear stress N6 and shear strain e60

Figure 11.3. Coordinate system and directions.

 291

The coordinates of the layer are given by:

t
z0  (11.2)
2

zk zk 1  tk (11.3)

where
tk = the thickness of the layer k
k = 1, 2,…, n
Positive stress N and positive strain H are defined in Figure 11.3.
Any arbitrary curl of a paper sheet can be divided into three components, N1, N2 and N6.
Positive bendings (N) and positive moments (M) are shown in Figure 11.4.

3 w 2

2
M1 k1

M2

k2

Twist k2
M6

Figure 11.4. A curl of a paper sheet can be divided into three components, N1, N2 and N6.
292

Curl in direction 1

w 2w 1 ª1º
N1  N1 «m» (11.4)
wx12 R1 ¬ ¼

Curl in direction 2

w2w 1 ª1º
N2  N2 «m» (11.5)
wx22 R2 ¬ ¼

where R is the radius of curvature in the appropriate direction.

Twist:

N6 2
w 2w
N6 2
hb ª1º (11.6)
wx1 x2 l «m»
¬ ¼

The twist, N6, can be illustrated as twice the change in the inclination of the surface from
point A to point B in Figure 11.5.
1

l h
B

h
A
2
b

Figure 11.5. The twist, N6, can be illustrated as twice the change in the inclination of the surface from point A to
point B.

NNKP==píêÉëë=~åÇ=píê~áå=Ñçê=~=i~óÉê=áå=~=i~ãáå~íÉ
Consider one of the orthotropic layers in the laminate. The layer is oriented with its symmetry
directions in the x- and y-directions according to Figure 11.6. If the layer is strained Hx in the x-
direction, the layer contracts in the y-direction (Hy in the y-direction is negative).

ey

x ex
y

Figure 11.6. Consider one of the orthotropic layers in the laminate.

 293

Hooke's law gives:

Vx
Hx (11.7)
Ex

The contraction in the y-direction perpendicular to the loading direction is given by the equa-
tion:

Hy Q xy ˜ H x (11.8)

where
Vx = stress in the x-direction
Hx = strain in the x-direction
Hy = strain in the y-direction
Qxy = Poisson's ratio (strain in the y-direction because of loading in the x-direction)
Ex = Elastic modulus in the x-direction
If the layer is strained in the y-direction, we obtain by analogy:

Vy
Hy (11.9)
Ey

The contraction in the x-direction perpendicular to the loading direction is given by the equa-
tion:

Hx Q yx ˜ H y (11.10)

where
Vy = stress in the y-direction
Hx = strain in the x-direction
Hy = strain in the y-direction
Qxy = Poisson's ratio (strain in the x-direction because of loading in the y-direction)
Ey = Elastic modulus in the y-direction
If the layer is loaded in both the x- and y-directions, the strain in the x-direction is:

Vx Vy
Hx  Q yx (11.11)
Ex Ey

The corresponding strain in the y-direction is:

Vy Vx
Hy  Q xy (11.12)
Ey Ex

The shear strain is defined as:

294

wv wu
J xy  (11.13)
wx wy

The relationship between shear stress Wxy and shear strain Jxy is

W xy Gxy ˜ J xy (11.14)

where
Gxy is the shear modulus, N/m2

NNKQ==j~íêáÅÉë

The equations above can be expressed compactly in matrix form as:

ª Hx º ª S xx S xy 0 º ªV x º
« » «S « »
«Hy » « yx S yy 0 »» ˜ «V y » (11.15)
«¬J xy »¼ «¬ 0 0 S ss »¼ «¬W xy »¼

where [S] is the compliance matrix with the following components

1 Q yx
S xx S xy 
Ex Ey
Q xy 1
S yx  S yy
Ex Ey
1
S ss
Gxy

The matrix is symmetrical, Sxy = Syx, from which it follows that

Ex Q xy
(11.16)
Ey Q yx

If we assume that the material has its highest modulus of elasticity in the x-direction and if
we assume that we strain the material equally in both directions, the contraction will be greatest
in the y-direction when a load is applied in the x-direction, as shown in Figure 11.7.
 295

ey ey

x ex x ex
y y

Figure 11.7. Contraction in x-direction and y-direction as a function of applied equal strain.

The following equations indicates which strain is obtained as a function of the external
stress. If one wishes to know which stresses are obtained as a function of the external strain, the
compliance matrix is inverted and we obtain:

ªV x º ªQxx Qxy 0 º ªHx º

« » «Q « »
«V y » « yx Qyy 0 »» ˜ « H y » (11.17)
«¬W xy ¼» «¬ 0 0 Qss »¼ ¬«J xy ¼»

where [Q] is the stiffness matrix with the following components.

Ex Q xy ˜ E y
Qxx Qxy
1  Q xy ˜Q yx 1 Q xy ˜Q yx

Q yx ˜ Ex Ey
Qyx Qyy Qss Gxy
1  Q xy ˜Q yx 1  Q xy ˜Q yx

For reasons of symmetry, Qyx = Qxy.

The stiffness matrix given in these equations is the starting-point in calculations according to
laminate theory.

NNKR==`çãé~êáëçå=ÄÉíïÉÉå=lêíÜçíêçéáÅ=~åÇ=fëçíêçéáÅ=j~íÉêá~äë

In the case of isotropic materials, the following relationship applies between elasticity modulus;
shear modulus and Poisson's ratio.

E
G (11.18)
2(1  Q )

In the case of an orthotropic material such as paper, the following approximate relationship
applies, according to Baum:
296

Ex ˜ E y
Gxy (11.19)
2 ª¬1  Q xy ˜Q yx º¼

Q xy ˜Q yx 0,293 (11.20)

Gxy 0.387 Ex ˜ E y (11.21)

Previously equation (11.16) shows

Ex Q xy
Ey Q yx
Combination of equations (11.16) and (11.20) gives:

Ex
Q xy 0.293 (11.22)
Ey

NNKS==píê~áå=áå=íÜÉ=mä~åÉ=ÇìÉ=íç=jçáëíìêÉ=^Äëçêéíáçå

We assume that the layer also suffers a strain HH because of moisture absorption

HH E H ˜ 'H (11.23)

where EH = hygroexpansion coefficient

'H = the change in the moisture content of the paper
In the laminate equations, the index H is in general used, but unfortunately this usage covers
two different cases.
Index H = RH (Relative Humidity), and we obtain.

HH E RH ˜ 'RH (11.24)

Index H = mc (moisture content), and we obtain.

HH E mc ˜ 'mc (11.25)

NNKSKN==píê~áå=áå=íÜÉ=mä~åÉ=ÇìÉ=íç=~å=bñíÉêå~ä=jÉÅÜ~åáÅ~ä=cçêÅÉ

We assume that a layer is exposed to a strain HM due to the application of an external mechanical
force.
 297

NNKSKO==qçí~ä=píê~áå=áå=~=i~óÉê=áå=íÜÉ=mä~åÉ=ÇìÉ=íç=ÄçíÜ=jçáëíìêÉ=pçêéíáçå=~åÇ=
jÉÅÜ~åáÅ~ä=cçêÅÉë

H HM HH (11.26)

NNKSKP==píê~áå=ÇìÉ=íç=_ÉåÇáåÖ

If a beam is subjected to a bending moment, M, the beam is formed into an arc with a certain ra-
dius of curvature, R, Plane cross sections in the beam remain plane after bending. Longitudinal
elements on the convex side of the beam are stretched, and those on the concave side are com-
pressed. The region in the beam's cross-section, which is neither stretched nor compressed, is
called the neutral surface. It need not lie in the middle of the beam. Figure 11.8a shows the
strain distribution in a bent beam.

R
M M

compression

tension

neutral

Figure 11.8a. The strain distribution in a bent beam.

Let us consider a region of the bent beam in greater detail, as shown in Figure 11.8b. Note
that in laminate theory the origin of the z-coordinate is placed in the geometrical middle plane
of the laminate.
298

z
d

Figure 11.8b. Bent beam at a larger scale.

Similar triangles give:

G l
z R

The strain is given by:

G z
HB (11.27)
l R
or

HB z ˜N (11.28)

where

1
N N = the curvature, m–1
R

The strain of the mean surface is H°. The total strain in a given layer during bending is then:

H H q  HB (11.29)

Figure 11.9 illustrates the strains in a bent laminate, where z is the coordinate from the mid-
dle plane and N is the curvature.
 299

eo
eB = z k

Figure 11.9. The strains in a bent laminate.

NNKT==qçí~ä=píê~áå=áå=~=ÖáîÉå=i~óÉê=ÇìÉ=íç=jçáëíìêÉ=`Ü~åÖÉëI=
jÉÅÜ~åáÅ~ä=fåÑäìÉåÅÉë=~åÇ=_ÉåÇáåÖ

In a given layer, equations (11.26) and (11.29) apply:

HM HH Hq H B (11.30)

Equations (11.24), (11.28) and (11.30) give:

HM H q  z ˜ k  E H ˜ 'H (11.31)

NNKU==cçêÅÉë=~åÇ=jçãÉåíë

In the uniaxial case, where the layers are oriented symmetrically in the 1-direction (MD) and the
mechanical strain is H M, the stress in a given layer is, according to the equation equal to:

V Q ˜H M (11.32)

Equations (11.31) and (11.32) give

V 
Q ˜ H q  z ˜ N  E H ˜ 'H  (11.33)

The forces and moments in all the layers, with the z-axis directed downwards as in the previ-
ous figure, are integrated to give

N ³ V dz (11.34)

M ³ V ˜ z dz (11.35)
300

where
M = moment per unit width, Nm/m
N = force per unit width, N/m
Equations (11.33) – (11.35) give:

³ Q H q  z ˜ k  E 
˜ 'H dz
H
N (11.36)

³ Q ˜ z H q  z ˜ k  E 
˜ 'H dz
H
M (11.37)

These expressions are developed with the integration limits, where i goes from 1 to n

¦  Q ˜ H q  zi  zi 1   12 Q ˜ k  zi2  zi21   Q ˜ E H ˜ 'H  zi  zi 1  

n
N (11.38)
i 1

¦  12 Q ˜ H q  zi2  zi21   13 Q ˜ k  zi3  zi31   12 Q ˜ E H ˜ 'H  zi2  zi21  

n
M (11.39)
i 1

This equation can be written compactly as:

ªN º ª A B º ªH qº ª N H º
«M » «B D» « N »  « H » (11.40)
¬« ¼» ¬« »¼ ¬« ¼» ¬« M ¼»

or

ªN º ªNH º ª A B º ªH q º
«M »  « H » «B D» « N » (11.41)
¬« ¼» ¬« M ¼» ¬« »¼ ¬« ¼»

where

n
A ¦ Qi > zi  zi 1 @ (11.42)
i 1

1 n
B ¦ Qi ª zi2  zi21 º¼
2i 1 ¬
(11.43)

1 n
D ¦ Qi ª zi3 3i 1 º¼
3i 1 ¬
(11.44)

The internal forces and moments caused by the hygroexpansion are:

 301

n
NH ¦ 'H i ˜ EiH ˜ Qi > zi  zi 1 @ (11.45)
i 1

1 n
MH ¦ 'H i ˜ EiH ˜ Qi ª¬ zi2  zi21 º¼
2i 1
(11.46)

Each internal force and moment has three components.

ª N1H º
« H»
« N 2 » = internal forces per unit width, N/m (11.47)
« N 6H »
¬ ¼

ª M 1H º
« H»
« M 2 » = internal moments per unit width, Nm/m (11.48)
« M 6H »
¬ ¼

The same applies to external forces N, moments M, strains H ° and curvatures N.

NNKV==i~ãáå~íÉ=qÜÉçêó=ïÜÉêÉ=íÜÉ=i~óÉêë=~êÉ=lêáÉåíÉÇ=ïáíÜ=~å=
^êÄáíê~êó=^åÖäÉ=áå=oÉä~íáçå=íç=íÜÉ=NJaáêÉÅíáçå=EjaF

When a layer is oriented at a certain angle to the 1-2 system, the expressions above are in prin-
ciple the same, although more complicated. To be able to make calculations, the following
properties must be transformed to the 1-2 system (off-axis) from the x-y system (on-axis).
1. Stresses and strains
2. Stiffnesses Q
3. Hygroexpansions

NNKNM==qê~åëÑçêã~íáçå=çÑ=píêÉëëÉë=~åÇ=píê~áåë

Transform x-y strains (on-axis) to the 1-2 system (off-axis). The angle is defined in Figure 11.1.

ª H1 º ª Hx º
« H »
« 2 » >T @1 «« Hy »
»
(11.49)
¬« 2 J 12 ¼»
1 «¬ 1 2 J xy »¼

where
302

ªc 2 s 2 2cs º
« 2 » ­c cos I
>T @ 1
«s c 2
2cs » ® (11.50)
« cs cs c 2  s 2 » ¯ s sin I
¬ ¼

Transform stresses to the 1-2 system (off-axis)

ªV 1 º ªV x º
«V »
« 2» > @ «V y »»
T
1 «
(11.51)
«¬ W 6 »¼ «¬W xy »¼

NNKNN==qê~åëÑçêã~íáçå=çÑ=íÜÉ=eóÖêçÉñé~åëáçå=`çÉÑÑáÅáÉåí

The hygroexpansion coefficient in the x-y direction (ON-axis) is transformed to the 1-2 direc-
tion (OFF-axis) as follows:

ª E1H º ª E xH º
« H » « »
« E2 » >T @1 « E yH » (11.52)
« 1 2 E12H » « 1 2 E12H »
¬ ¼ ¬ ¼

NNKNO==qê~åëÑçêã~íáçå=çÑ=íÜÉ=píáÑÑåÉëë=j~íêáñI=n

ªV1 º ª Q11 Q12 Q16 º ª H1 º

«V » «Q Q26 »» «« H 2 »»
« 2» « 12 Q22 (11.53)
«¬W 12 »¼ «¬Q16 Q26 Q66 »¼ «¬J 12 »¼

The stiffness elements are.

ª c4 s4 2c 2 s 2 º
4c 2 s 2
ª Q11 º « 4 »
«Q » « s c4 2c 2 s 2 4c 2 s 2
» ªQ º
« 22 » «c 2 s 2 c2s 2 c4  s4 » « xx »
4c 2 s 2
« Q12 » « » «Qyy »
« »
«Q66 »
«c 2 s 2
«
c 2 s 2 2c 2 s 2 
c 2  s 2 » «Q »
» « xy »
 (11.54)

« Q16 »
« »
« c3s
«

cs 3 cs 3  c3s 2 cs 3  c3 s » ¬ Qss ¼
»

¬«Q26 ¼» « cs 3

c3 s c3s  cs 3 2 c3 s  cs 3 » 
¬ ¼
 303

NNKNP==cçêÅÉë=~åÇ=jçãÉåíë

The forces and moments become:

ª N1 º ª N1 º A16 º ªH10 º ª B11 B16 º ª N1 º

H
ª A11 A12 B12
«N »  «N H » «A A22
« »
A26 »» «H 20 »  «« B12 B22 B26 »» ««N 2 »»
« 2» « 2 » « 12 (11.55)
«¬ N 6 »¼ « N 6H » «¬ A16 A26 A66 »¼ «¬H 60 »¼ «¬ B16 B26 B66 »¼ «¬N 6 »¼
¬ ¼

ª M1 º ª M1 º B16 º ªH10 º ª D11 D16 º ª N1 º

H
ª B11 B12 D12
« » « H» « « »
B26 »» «H 20 »  «« D12 D26 »» ««N 2 »»
«M 2 »  «M 2 » « B12 B22 D22 (11.56)
«¬ M 6 »¼ « M 6H » «¬ B16 B26 B66 »¼ «¬H 60 »¼ «¬ D16 D26 D66 »¼ «¬N 6 »¼
¬ ¼

The equation is a more general form of the previous equations, see Equation (11.41)

ªN º ªNH º ª A B º ªH q º
«M »  « H » «B D» « N »
¬« ¼» ¬« M ¼» ¬« »¼ ¬« ¼»

NNKNQ==_ÉåÇáåÖ=píáÑÑåÉëë=~åÇ=qÉåëáäÉ=píáÑÑåÉëë

We now study the application of Equation (11.41). It is assumed that there is no hygroexpan-
sion.

NH = MH = 0.

Equation (11.41) has in principle three different forms. The calculations refer to these three
equations. The manner in which the inversion is carried out is not shown here.
A) Strains and curvatures give external forces and external moments

ªN º ª A B º ªH qº
«M » «B D» « N » (11.57)
¬« ¼» ¬« »¼ ¬« ¼»

B) Forces and curvatures give external strains and external moments

ªH q º ª A B º ª N º
«M » « »« » (11.58)
¬« ¼» ¬«C D ¼» ¬« N ¼»

C) Forces and moments give external strains and external curvatures

304

ªH q º ª Ac Bc º ª N º
«N » «C c Dc» « M » (11.59)
¬« ¼» ¬« »¼ ¬« ¼»

Equation (11.59) is expanded

ªH10 º c
ª A11 c
A12 c º ª N1 º ª B11
A16 c c
B12 c º ª M1 º
B16
« 0» « Ac
«H 2 » c
A22 c » « N 2 »  « B12
A26 c c
B22 c » «M 2 »
B26 (11.60)
« 12 »« » « »« »
«H 60 » ¬« A16
c c
A26 c ¼» ¬« N 6 ¼» ¬« B16
A66 c c
B26 c ¼» ¬« M 6 ¼»
B66
¬ ¼

ª N1 º c C12
ª C11 c c º ª N1 º ª D11
C16 c c
D12 c º ª M1 º
D16
«N » «C c C c » «
c N 2  D12
C26 » « c c
D22 c » «M 2 »
D26
« 2» « 12 22 »« » « »« » (11.61)
«¬N 6 »¼ «¬C16
c C26
c c » « » « c
C66 ¼ ¬ N 6 ¼ ¬ D16 c
D26 c »¼ «¬ M 6 »¼
D66

Equation (11.61) is applied to bending stiffness. The bending stiffness is defined as the mo-
ment per unit curvature, Figure 11.10.

M
k =1/R

M
Sb =
k

M
k

Figure 11.10. The definition of bending stiffness.

We add a moment in the 1-direction. No forces act on the laminate, i.e. N1 = N2 = N6 = 0.

From Equation (11.61) we obtain:

N c ˜ M1
D11 (11.62)

Through the definition of bending stiffness, we obtain the bending stiffness in the 1-direc-
tion, 2-direction and 6-direction (torsional stiffness).
 305

­ b M1 1
° S1 c
° N1 D11
° b M2 1
®S2 (11.63)
° N2 c
D22
° b M6 1
° S6
¯ N6 c
D66

Equation (8-48) applied to tensile stiffness. Tensile stiffness is defined as force per unit
width per unit strain.

N
Eb (11.64)
Hq

We apply a force per unit width N1. From Equation (11.60) we obtain:

H10 c ˜ N1
A11 (11.65)

and thus:

N1 1
E1b (11.66)
H1o c
A11

N2 1
E2b (11.67)
H 2o c
A22

N 61 1
E6b shear stiffness Equation (11.68)
H16o c
A66

NNKNR==páãéäáÑáÉÇ=bñéêÉëëáçå=Ñçê=_ÉåÇáåÖ=píáÑÑåÉëë

In the uniaxial situation, the bending stiffness can be calculated in the following way.
Consider the Equation (11.41)

ªN º ªNH º ª A B º ªH qº
«M »  « H » «B D» « N »
«¬ »¼ «¬ M »¼ «¬ »¼ «¬ »¼

If NH = MH = 0 there is no moisture change, and

N A˜H q  B ˜N (11.69)

M B ˜H q  D ˜N (11.70)
306

If there is no external force, N = 0, so that Hq is eliminated and we obtain:

ª B2 º
M «D  » ˜N (11.71)
¬ A¼

The definition of bending stiffness is the moment, M, per unit curvature, N.

M
Sb (11.72)
N

The final expression for bending stiffness is then:

B2
Sb D (11.73)
A

NNKNS==`~äÅìä~íáçå=bñ~ãéäÉ=Ñçê=_ÉåÇáåÖ=píáÑÑåÉëë

The following calculation example refers to a four-layer sheet with the properties given in the
following table. The starting values are the grammage, w, the tensile stiffness index, Ew, and the
density, U

mäó= w E U E Ew ˜ U w=
â ÖLã
O Ew P t
U = âÖLã O U
kLãã =Ejm~F
jkãLâÖ ãã
N UM TKR UMM SMMM MKN
O UM OKR QMM NMMM MKO
P RM QKM RMM OMMM MKN
Q TM TKNQ TMM RMMM MKN

The total thickness is calculated as:

t ¦ tk 0,1  0, 2  0,1  0,1 0,5 (11.74)

As indicated in Figure 11.2 the Equations (11.2) and (11.3) are:

t
z0 
2

zk zk 1  tk where tk is the thickness of the layer k = 1, 2, …, n

0.5
z0   0.25
2
 307

From Equation (11.3), we obtain:

z1 z0  t1 0.25  0.1 0.15

z2 0.05
z3 0.15
z4 0.25

The following table gives squares and cubes of the z-coordinates.

O P
i~óÉê=k zâ=ãã zk2 =ãã zk3 =ãã
M ÓMIOR MIMSOR ÓNRISORuNMÓP
N ÓMINR MIMOOR ÓPIPTRuNMÓP
O MIMR MIMMOR MINORuNMÓP
P MINR MIMOOR PIPTQuNMÓP
Q MIOR MIMSOR NRISORuNMÓP

Equations (11.42) – (11.44) are used to calculate A, B and D

i~óÉê=k Qk ( zk  zk 1 ) | Ek ( zk  zk 1 ) Qk ( zk2  zk21 ) | Ek ( zk2  zk21 ) Qk ( zk3  zk31 ) | Ek ( zk3  zk31 )

N SMMM=xÓMINRÓ=EÓMIORFz=Z=SMM SMMM=EMIMOORÓMIMSORF=Z= SMMM=xÓPKPTRÓENRISORFz=uNMÓP=Z=
ÓOQM TPIR
O NMMM=xMIMRÓ=EÓMINRFz=Z=OMM NMMM=EMIMMORÓMIMOORF=Z NMMM=xMKNORÓ=EÓPIPTRFz=uNMÓPZ=
ÓOM PIR
P OMMM=xMINRÓMIMRz=Z=OMM OMMM=EMIMOORÓMIMMOR=FZ= OMMM=xPIPTRÓMKNORz=uNMÓP=Z=
QM SIR
Q RMMM=xMIORÓMINRz=Z=RMM RMMM=EMIMSORÓMIMOORF=Z= RMMM=xNRISORÓPKPTRzuNMÓP=Z=
OMM SNIOR

A ¦ 1500 1 1
B ¦ 10 D ¦ 48,25
2 3
The bending stiffness is calculated according to Equation (11.73).

(10) 2
Sb 48, 25  48,18 mNm
1500

NNKNT==`ìêä=~åÇ=qïáëíI=píê~áåI=pÜÉ~êI=~ë=~=oÉëìäí=çÑ=~=`Ü~åÖÉ=áå
==jçáëíìêÉ

In this case there are no external forces applied, but the hygroexpansion causes internal forces
and moments. These can be very large. It is difficult to prevent expansion as a result of moisture
absorption.

N=M=0
308

We invert the matrices in the same way as equations (11.59) – (11.61). Internal forces and
moments give strains and curvatures

ªH10 º c
ª A11 c
A12 c º ª N1H º ª B11
A16 c c
B12 c º ª M 1H º
B16
« 0» « Ac « » « »
«H 2 » c
A22 c » « N 2H »  « B12
A26 c c
B22 c » « M 2H »
B26 (11.75)
« 12 » « »
«H 60 » «¬ A16
c c
A26 c »¼ «¬ N 6H »¼ «¬ B16
A66 c c
B26 c »¼ «¬ M 6H »¼
B66
¬ ¼

ª N1 º c C12
ª C11 c c º ª N1H º ª D11
C16 c c
D12 c º ª M 1H º
D16
«N » «C c C c » « H» « « »
c
C26 » « N 2 »  « D12c c
D22 c » « M 2H »
D26 (11.76)
« 2» « 12 22 »
«¬N 6 »¼ «¬C16
c C26
c c »¼ «¬ N 6H »¼ «¬ D16
C66 c c
D26 c »¼ «¬ M 6H »¼
D66

These can be represented in compact form as:

ªH qº ª Ac Bc º ª N H º
«N » «C c Dc» « H » (11.77)
«¬ »¼ «¬ »¼ «¬ M »¼

We can then extract the component, which we need from the equation.

NNKNU==`ìêä=~åÇ=qïáëí

With the help of Equation (11.76) Curl and Twist due to a moisture change can also be ob-
tained. We can extract the component, which we need from the equation.

>N1 @ >C11c @ ª¬ N1H º¼  >C12c @ ª¬ N 2H º¼  >C16c @ ª¬ N 6H º¼ 

(11.78)
> D11c @ ª¬ M1H º¼  > D12c @ ª¬ M 2H º¼  > D16c @ ª¬ M 6H º¼

>N 2 @ >C12c @ ¬ª N1H ¼º  >C22c @ ¬ª N 2H ¼º  >C26c @ ¬ª N6H ¼º 

, (11.79)
> D12c @ ª¬ M1H º¼  > D22c @ ª¬ M 2H º¼  > D26c @ ª¬ M 6H º¼

>N 6 @ >C16c @ ¬ª N1H º¼  >C26c @ ª¬ N 2H º¼  >C66c @ ª¬ N6H º¼ 

(11.80)
> D16c @ ª¬ M1H º¼  > D26c @ ª¬ M 2H º¼  > D66c @ ª¬ M 6H º¼

NNKNV==iÉ~åáåÖ=pí~Åâë=çÑ=cçäÇÉÇ=mêáåíáåÖ=m~éÉêë

Folded printed papers are stacked in bundles. If the layers have an oblique symmetry and if the
moisture content changes, each layer will shear slightly in relation to the 1-direction, MD. The
 309

shearing is then accumulated in the bundle and this develops a propeller-like appearance, as
shown in Figure 11.11.

+␸

1
2

Figure 11.11. Stack lean.

With the help of Equation (11.75) we obtain the shear due to moisture.

ª¬H 60 º¼ > A16c @ ª¬ N1H º¼  > A26c @ ª¬ N 2H º¼  > A66c @ ª¬ N 6H º¼ 

(11.81)
> B16c @ ª¬ M1H º¼  > B26c @ ª¬ M 2H º¼  > B66c @ ª¬ M 6H º¼

NNKOM==qÉåëáäÉ=píáÑÑåÉëë=~åÇ=_ÉåÇáåÖ=píáÑÑåÉëë=píìÇáÉÇ=áå=~=mçä~ê=
==aá~Öê~ã

By coordinate transformation it is possible to obtain plots of bending stiffness and tensile stiff-
ness in polar diagrams. The results may look as follows. By tensile stiffness orientation we
mean the angle at which there is symmetry, Figure 11.12.

j
MD

CD

Figure 11.12. Tensile stiffness and bending stiffness in a polar diagram.

NNKON==eóÖêçÉñé~åëáçåI=`ìêä=~åÇ=qïáëí=píìÇáÉÇ=áå=mçä~ê=aá~Öê~ãë

By coordinate transformation it is possible to obtain plots of hygroexpansion, curl and twist in

polar diagrams. The results may look as follows. By hygroexpansion orientation we mean the
angle at which there is symmetry, Figure 11.13.
310

j
MD

CD

Figure 11.13. Hygroexpansion, curl and twist in a polar diagram.

NNKOO==eóÖêçÉñé~åëáçå=lêáÉåí~íáçå=~åÇ=qÉåëáäÉ=píáÑÑåÉëë=
==lêáÉåí~íáçå

The tensile stiffness in different directions in the laminate can be measured using ultrasonic
technique. Measurement of the hygroexpansion orientation, can be obtained with speckle pho-
tography. Figure 11.14 shows the tensile stiffness orientation and hygroexpansion orientation
for a copy paper at different distances from the edge. The two orientations seem to correspond
well.

Orientation, degrees
12
a us
E H
a
8

-4
0 0,4 0,8 1,2 1,6 2
Position from the edge, m

Figure 11.14. Tensile stiffness orientation and hygroexpansion orientation for a copy paper at different distances
from the edge.
 311

NNKOP==pÜÉ~êJÑêÉÉ=^åÖäÉ=Ñçê=eóÖêçÉñé~åëáçå

The hygroexpansion orientation can be calculated analytically, i.e. the shear-free angle for
strain due to hygroexpansion
Transform 1-2 strains to the 1´-2´ systems.

unknown known
ª H1c º ª H1 º
« Hc » >T @1 « » (11.82)
« 2 » « H2 »
«¬ 1 2 H 6c »¼ «¬ 1 2 H 6 »¼

where

ªc 2 s 2 2cs º
« 2 » ­c cos I
>T @ 1
«s c 2
2cs » ® see (11.50)
« cs cs c 2  s 2 » ¯ s sin I
¬ ¼

We know the properties in the 1-2 system and require the properties in the 1´-2´ system. The
positive angle goes from the unknown to the known.

ª H1c$ º ª H1q º
« q » « q »
« H 2c » >T @
1
« H2 » (11.83)
« 1 2 H 6cq » « 1 2 H 6q »
¬ ¼ ¬ ¼

The shear-free angle is given by H 6o 0 The angle of symmetry then becomes

1 § Hq ·
I  tan 1 ¨ q 6 q ¸ (11.84)
2 © H1  H 2 ¹

In the case of hygroexpansion, it is desirable to indicate angles from the 1-axis to the symme-
try axis. The hygroexpansion angle is defined as D H = –M

NNKOQ==qïáëí=cêÉÉ=^åÖäÉ

Figure 11.15 shows a paper with curl in both MD and CD, and twist. The task is to find the
twist-free angle, i.e. the curl orientation Icurl ˜ N 6c 0 . The Curl angle, i.e. the twist-free angle
for strain due to hygroexpansion, can be calculated.
312

ª N1c º ª N1 º
« Nc »
« 2 » >T @1 «« N 2 »» (11.85)
¬« 2 N 6c ¼» ¬« 2 N 6 ¼»
1 1

ªc 2 s 2 2cs º
« 2 »
>T @ 1
«s c 2
2cs » (11.86)
« cs cs c 2  s 2 »
¬ ¼

Equation (11.85) gives ( N 2c NM ):

NI N1 ˜ sin 2 I  N 2 ˜ cos 2 I  N 6 ˜ sin I ˜ cos I (11.87)

1 § N6 ·
Itwist  ˜ tan 1 ¨ ¸ (11.88)
2 © N1  N 2 ¹

The curl orientation is indicated by the angle to axis 1. This gives the same sign of the angle
as N6 = the twist, according to earlier definitions.

k1 k2 jtwist
1‘

kj

2
w

Figure 11.15. Twist free angle.

We know the curl orientation Mcurl and the curl NM according to the above and wish to know
the curvatures in the 1-2 directions. Equation (11.87) gives:

N 2 NI  (11.89)

The angle is defined as being positive from the unknown to the known, as shown in Figure
11.16.
 313

jtwist
d

1‘

2
2‘

Figure 11.16. The angle is defined as being positive from the unknown to the known.

Insert the angle G = –M̓ curl

k1c kM sin 2 G
k2c kM cos 2 G
k6c 2 kM sin G cos G

NNKOR==cáåÇáåÖ=íÜÉ=eÉáÖÜí=~í=~=mçáåí=çå=~=_Éåí=m~éÉê

The task is to find the height at a point on a bent paper, Figure 11.17.

w (a1 , a2)
1
a2

a1

2
w

Figure 11.17. Finding the height at a point on a bent paper.

1 2
w 
2

a1 ˜ k1  a22 ˜ k2  a1 a2 k12  (11.90)

where
a1 and a2 are coordinates, (m)
N1, N2, N12 are curl and twist values, (m–1)
314

NNKOS==píêÉëë=~åÇ=píê~áåë=áå=íÜÉ=qÜáÅâåÉëë=aáêÉÅíáçå=çÑ=íÜÉ=
==i~ãáå~íÉ=ÇìÉ=íç=eóÖêçÉñé~åëáçå

For each interface in the laminate, the strains may be calculated according to Equation (11.31),
and the stresses according to Equation (11.33). An example of a strain distribution is given in
Figure 11.18.

Figure 11.18. Stress and strains in the thickness direction of the laminate due to hygroexpansion.
 315

NO==j~íÜÉã~íáÅ~ä=jçÇÉääáåÖ=~åÇ=^å~äóëáë=çÑ=`çåîÉêíáåÖ
======~åÇ=båÇìëÉ

Mikael Nygårds
Innventia AB
Sören Östlund
Royal Institute of Technology, KTH

12.1 Introduction 315

12.2 The Mechanical Modelling Process 317

12.3 Paper from a Mechanical Point of View 320

12.3.1 Length Scales 320
12.3.2 Continuum Models 321
12.3.3 Network Models 324

12.4 The Finite Element Method 325

12.4.1 Displacements 325
12.4.2 Stresses 327
12.4.3 Matrix Formulation 327
12.4.4 Principle of Virtual Work 328

12.5 Example: Creasing of Paperboard 329

12.5.1 Deformation Mechanisms 329
12.5.2 Model Generation 330
12.5.3 In-plane Model 330
12.5.4 Delamination Model 331
12.5.5 Modelling 333

12.6 References 333

NOKN==fåíêçÇìÅíáçå

Numerical methods are becoming increasingly used to study the behaviour of materials and
structures. For modelling of solids the finite element method (FEM) is the most popular ap-
proach. Among the reasons to do numerical modelling some can be mentioned:
• Different types of loadings can be investigated.
• The effect of different material properties can be investigated.
• More information about damage and deformation mechanisms can be gained.
• Material behaviour can be predicted.
316

• Properties that are important for the manufacturer can be linked to properties that are
important for the user.
If the investigations above would be performed by experiments it would be expensive and
time consuming, since trial materials and structures must be manufactured and tested. In a nu-
merical model the parameters only need to be changed and a new simulation can be started.
This can be used to better understand the problems before an experimental study is conducted.
Then it becomes possible to learn what kind of experiments that will be of interest.
Ideally numerical simulations should be used together with experiments to learn more about
the materials. Experimental investigations and visual inspections make it possible to detect the
macroscopic behaviour of the materials. Numerical methods also enable studies of the macro-
scopic behaviour under different types of loading, but in addition the method also offers the
possibility to study field properties such as stresses and strains within the material.
In the past experimental studies has been conducted to learn more about material behaviour.
There it was shown how different properties such as elastic, plastic, viscoplastic etc can be mea-
sured by different test methods. In all methods presented a property is measured, and it is possi-
ble to predict how the material behaves under this particular loading. But if the loading
condition change, it is not obvious how the mechanical response change. Therefore, a model is
needed. The knowledge gained from an experimental investigations can been used to formulate
theories that predict the material properties. It is however difficult to find analytical solutions to
predict material responses if the boundary conditions are different from the experimental setup.
With aid of numerical methods it is possible to solve systems of equations that are needed to do
better and more complicated predictions than was possible earlier.
In the finite element method field properties are studied. The field properties of interest for
paper are primarily;
• Stress,
• Strain,
• Temperature,
• Moisture.
Interpretations of these fields must be done in order to understand the problem solution. The
next task (which probably is the most difficult one) is to understand how process and manufac-
turing parameters together with different loading conditions interact and affect the paper to cre-
ate the field properties.
As an example consider the behaviour during converting and end-use. It is the process and
manufacturing parameters that creates the actual material, but in order to fully understand how
deformation and damage evolve knowledge of the underlying field properties, such as stress and
strain, is required. Another example is curl. Curl is created during in the papermaking process
since an unsymmetrical paper or paperboard is created. By changing process parameters it is
possible to redesign the material in the thickness direction, or change the drying strategy etc.
This will certainly change the curl behaviour of the material. It is however difficult to predict
the change in curl a priori if there is no knowledge of how these process parameter changes af-
fects the underlying stress field. It is an unbalanced stress field in the material that causes the
out-of-plane deflection, and this can not easily be controlled directly by process parameters.
Therefore it is important to gain knowledge about field properties in order to explain material
behaviour.
 317

NOKO==qÜÉ=jÉÅÜ~åáÅ~ä=jçÇÉääáåÖ=mêçÅÉëë

The aim with modelling is generally to create and simulate a condition that mimics reality.
However, assumptions need to be made, because modelling needs to focus on certain aspects
that are dominating in the analysis. If no assumptions would be made the models would become
very large, and thus computer demanding to solve. Moreover, if modelling focus on important
aspects it is easier to evaluate and understand the results. In general three types of assumptions
can be made.
• Geometry: A complicated structure is normally idealized to a computable structure. The
main features of the geometry need to be represented, but details such as e.g. surface rough-
ness are normally not considered, basically because it does not influence the overall
mechanical behaviour. The primary aspect that affects the phenomena of interest does how-
ever need to be incorporated.
• Material behaviour: Assumptions about material behaviour is always needed, because
there are hardly any material models that exactly resemble the behaviour of real materials
under all possible loading conditions. Assumption like this does however lead to satisfac-
tory results in most cases.
• Boundary conditions: In many applications the load is not known or even measurable; it is
then very difficult to get accurate boundary conditions for the model analysis. In other
cases it is possible to model the total load on a structure, but it is difficult to exactly know
the true distribution of the load, while the opposite is true in other situations.
Thus, what is created is a model of the real structure. It is important to be aware of the as-
sumptions you do since that should influence the way you evaluate the results. In certain cases a
coarse model is sufficient to detect principal behaviours of the structure, while other applica-
tions require large models with many details. Evaluation of what kind of model that is need for
different applications require experience.
There are certain steps that are needed in the formulation of a mathematical model; a sche-
matic description of the model generation is shown in Figure 12.1. To formulate a mechanical
model you need to use the laws of mechanics directly or indirectly. It should be emphasised that
it is the model properties that are evaluated when the problem is solved not the real behaviour.
The interpretation of the model results in terms of the real physical behaviour has to be done by
the user. If the criteria used to formulate the model are close to reality the interpretation of mod-
el result are straightforward. However, normally simplifying assumptions have been made.
They are motivated by reductions in both modelling time and analysis (computer) time. There-
fore, the accuracy of a model is often a tradeoff between available resources and time versus the
complexity of the model. It is however important to be aware of how the response would
change if the model had larger complexity.
318

laws of mechanics

solution of
real process /object mathematical problem
mathematical problem

experimental verification

Figure 12.1. A schematic representation of the mechanical modelling process.

When the mathematical problem following from the modelling has been formulated it needs
to be solved. Within mechanical modelling the resulting mathematical models often consist of
partial differential equations (or in simpler cases ordinary differential equations) and corre-
sponding boundary conditions. It is in general difficult to solve these differential equations ana-
lytically. Therefore efficient numerical methods that yield the result with engineering accuracy
have been developed.
One of the most important steps in the analysis is to check if the model behaviour resembles
that of the real phenomenon of interest. This needs to be done by experimental verification. If
the comparison is not satisfactory the mathematical model needs to be reformulated, since im-
portant features have been left out of the model. This verification step is many times performed
on a simplified problem because real problems might be too complex to study in detail experi-
mentally.
Example
To illustrate the modelling generation concept consider a 10mm wide paper strip that is
100 mm long, e.g. cut from this page. Load this strip by a 1.0 kg weight. How much does the pa-
per strip elongate due to this load?
If the paper has a thickness of 0.1mm, then stress in the paper strip can be calculated from

F 1.0 u 9.8
V N / m 2 10 MPa ,
A 10 u 103 u 0.1 u 103

In this regime copy paper is in general elastic, with a possible elastic modulus of E = 3 GPa.
Thus, Hooke´s law can be used to approximate the elongation,

V 10 u 106
'L L 0.1 u m 0.3mm .
E 3 u 109

In this example is it sufficient to consider paper as a continuum. This approach gives the to-
tal elongation that was needed in the problem. If more details were needed in the analysis it
would have been incorporated in the model. For example, the fibrous structure does create lo-
cally high stresses and non-homogenous strain fields. Moreover, it is assumed that the load due
to the weight is evenly distributed. Thus, even though approximations are done the problem has
been solved satisfactory. The elongation of the strip could be calculated without detailed knowl-
edge of the fibrous structure and load transfer between fibres.
 319

The level of complexity in a model should be different depending on what kind of modelling
that are performed. A normal modelling approach in the industry is to incorporate more details
if the scale is small. As an example of this approach it is possible to look at the different kind of
length scales that are of interest in modelling of a diary product box, as illustrated in Figure
12.2. First, by experience of how the product behaves in normal use, it is possible to specify the
loads that affect the box, i.e. the boundary conditions representing loads due to for example
stacking or gripping. A box model can then be formulated to investigate how the model behaves
under the specified load. From the results of this analysis it is possible to extract the loading

Figure 12.2. Modelling of different scales in a diary product box. Both the box model and studies of the creases
can give information about critical material behaviour.

Material critical
properties loads
analysis

detailed properties box properties use of

model of details model of box product

Figure 12.3. Modelling of different scales in a diary product box. Both the detailed model and the box model can
give information about critical material behaviour for the product.
320

conditions at a corner of the box. On both the detailed scale and the box model it is possible to
evaluate how the assumed material behaves, and from the analyses can critical loads be extract-
ed where risk for failure arise. Thus, the material can be improved if the weakest areas are iden-
tified. Moreover, the same approach can be used in the opposite direction. If the loads on certain
details are known a priori, then it is possible to calculate the criteria need needed on the product,
which is schematically shown in Figure 12.3.

NOKP==m~éÉê=Ñêçã=~=jÉÅÜ~åáÅ~ä=mçáåí=çÑ=sáÉï

Due to the complex structure of paper there are problems associated with modelling of paper.
The paper fibres form a two-dimensional network that is loosely bonded in the out-of-plane di-
rection. Therefore, the properties in the three principal directions MD (machine direction), CD
(cross machine direction) and ZD (out-of-plane direction) differ considerably, as seen in Figure
12.4. This means that the paper has an anisotropic behaviour. The stiffest direction is MD, since
this is the preferred direction of the fibres. Generally for machine made paper, the properties in
the MD and CD direction can differ up to about a factor 5, while the stiffness in ZD on the other
hand can be a factor 100 less than in the MD.

Figure 12.4. Tensile properties of paperboard in MD, CD and ZD.

NOKPKN==iÉåÖíÜ=pÅ~äÉë

Figure 12.5 shows examples of the different length scales that influence the mechanical behav-
iour of a paper product such as a box. Those are: box, web, fibre network and fibre with fibrils.
In all constitutive modelling an appropriate length scale must be chosen. In modelling of me-
chanical behaviour of paper a box can be used to illustrate the end use products. Modelling of
these require knowledge about the design as well as the material properties. On web or sheet
scale there are continuum models that are used to model the material. On the fibre network scale
details in the structure can be studied. Then appropriate models for the fibres and bonds must be
known as input in the models. Very few models on the fibre level and below exist to predict me-
chanical behaviour. The interested reader is referred to the PhD- thesis by Persson (2000).
 321

box web fibre network fibre fibrils

Figure 12.5. Mechanical properties of paper are present at different length scales in paper.

NOKPKO==`çåíáåììã=jçÇÉäë

Continuum modelling of paper is the most commonly used approach. Basically because in most
applications it is the sheet properties of paper that are important, and not the properties of the in-
dividual fibres in the network of connected fibres. It is therefore sufficient to study how paper
behaves as a continuum. In the continuum approach material properties of the paper are used as
input in the models. Thus, experiments are conducted on paper specimens to determine, e.g.
elastic and plastic properties.
Materials models for continuum mechanics modelling can in general be divided into four
types of models, linear rate-independent models, linear rate-dependent models, non-linear rate-
independent models and non-linear rate-dependent models.
In the rate dependent models stress and strain are functions of time, while they are not in rate
independent models. In some applications time is an important aspect of the mechanical proper-
ties. While paper in room temperature is normally assumed to be rate independent, the rate-de-
pendence might play an important role at increased moisture and temperature. If rate dependent
models are used, the paper needs to be characterized with respect to time, i.e. creep (constant
stress) or relaxation (constant strain) experiments need to be performed to gather the time de-
pendence.
The concepts of linearity and non-linearity here refer to deformation as a function of load. In
linear models the strain response is a linear function of stress, and an increase of the stress by a
factor of two will result in a doubling of the strain, while this is not the case for non-linear mod-
els.
322

Example
Simulation of web tension in a printing press where different material models have been used,
as shown in Figure 12.6. The material models used are from top to bottom: elastic, viscoelastic
and viscoelastic with hygroexpansion models.

500
450
400
350
300
web tension (N/m)

250
200
150
100
50
0
– 50
0 5 10 15 20 25 30
distance from roll stand (m)

Figure 12.6. The web tension in a printing press simulated with three different material models. From top to bot-
tom: elastic, viscoelastic and viscoelastic with hygroexpansion.

NOKPKNKN==bä~ëíáÅJéä~ëíáÅ=_ÉÜ~îáçìê

This section will highlight the elastic-plastic behaviour of paper, and will show the necessary
equations to formulate a theory. In a continuum model constitutive equations must be formulat-
ed. In Figure 12.4 above it can be observed that the initial part of the stress-strain curves can be
approximated as linear. If the paper in this regime upon unloading returns to the initial configu-
ration the material is said to be linear elastic. As seen in Figure 12.4 the initial slope of the MD,
CD and ZD curves differ considerably. This causes a problem when we later introduce the finite
element method. It is generally difficult to solve numerical models that have large differences in
the elastic constants in different principal directions. Therefore, problem can arise if constitu-
tive equations developed for other material classes are used, such as metals.
Since paper behaves differently in the three principal directions it is said to be orthotropic. In
the elastic regime the stresses and strain can be expressed with aid of 12 material constants. The
components of the symmetric strain tensor are in a vectorial notation related to the components
of the symmetric stress tensors as
 323

ª 1 Q CD  MD Q ZD  MD º
« E   0 0 0 »
ECD EZD
« MD
»
« Q MD CD 1 Q ZD  CD »
ª H MD º « E ECD

EZD
0 0 0 »
ª V º
« H » « MD » « MD »
« CD » « Q MD  ZD Q CD  ZD 1 » « V CD »
« H ZD » «  0 0 0 »
« »
« » « EMD ECD EZD » « V ZD » .(12.1)
«J MD CD » « 1 » «V MD CD »
« 0 0 0 0 0 »
« J CD  ZD » « »
« » « GMD CD » « V CD  ZD »
« » V
¬«J MD  ZD ¼» « 0 0 0 0
1
0 » ¬« ZD  MD ¼»
« GCD  ZD »
« 1 »
« 0 0 0 0 0 »
«¬ GZD  MD ¼»

The physical interpretation of components of the compliance matrix in Equation (12.1) fol-
lows in a straight-forward way by considering the strain response to each of the stress compo-
nents. From any given stress state it is then possible to calculate the corresponding strains.
Moreover, the equation above can be inverted if it is desired to calculate the stresses from a giv-
en strain state. Thermodynamic considerations require that the compliance matrix should be
symmetric. Thus, Equation (12.1) can be complemented with three symmetry conditions

Q MD  CD Q CD  MD Q ZD  MD Q MD  ZD Q ZD CD Q CD  ZD
, , (12.2)
EMD ECD EZD EMD EZD ECD

that reduces the number of elastic constants for an orthotropic linear elastic material to nine.
At some stress level the relations between stress and strain will no longer be linear, and the
material starts to deform non-linearly. This deformation is usually described as elastic-plastic,
and is not reversible. If a loaded specimen that has been deformed plastically is unloaded, a per-
manent deformation will remain upon unloading. For uniaxial loading plastic deformation (or
yielding) is initiated at a stress level, VY, known as the yield stress. For multiaxial loading the
stress state that initiates plastic deformation will in general be a function of all components of
the stress tensor. This function can be looked upon as a surface in stress space called the yield
surface. If the current stress state is inside the surface the deformation is purely elastic. This
yield surface needs to be defined. During the elastic-plastic deformation the stress components
always fulfil this function, i.e. the stress state is on the yield surface. The yield surface is written
on the form

f (V MD ,V CD ,V ZD ,W MD CD ,W MD  ZD ,W CD  ZD ) 0 , (12.3)

where Vi the normal are stress components in the normal direction and Wi are the shear stress
components. The yield surface is used to evaluate if yielding occurs, independently of loading
conditions. A yield criterion that is commonly used for paper is the Tsai-Wu criteria. The crite-
rion is anisotropic, and distinguishes tensile and compressive loads. The Tsai-Wu criterion is
expresses as
324

f D1V MD
2
 D2V CD
2
 D3V zD
2
 D4V MDV CD  D5V MDV ZD  D6V CDV ZD 
(12.4)
D7W MD CD  D8W MD  ZD  D9W CD  ZD  D10V MD  D11V CD  D12V ZD  1 0,

where the constants Di are functions of the yield stresses in the different loading directions.
After initial yielding the material deforms elastic-plastically, i.e. the deformation is build-up
of both elastic reversible and inelastic irreversible deformation. The stress-strain response is
then not as stiff as in the elastic regime, but will still show a hardening behaviour, i.e. stress in-
creases as strain increase. Hardening functions that depend on the plastic strain, H p captures this
behaviour. Thus each stress component is expresses as

V i V is  f (H p ) , (12.5)

where Vi is either the normal stress components of the shear components, and V is is the yield
stress in the corresponding direction.

NOKPKP==kÉíïçêâ=jçÇÉäë

Network models can be used to study certain mechanisms in paper materials in more detail, an
example of a network model by Heyden (2000) is shown in Figure 12. 7. In this approach a fi-
bre network is constructed by distribution of fibres, fines and other additives. Fibres that come
in contact will bond to each other, which require a material model for the bonding. Moreover,
material data on the fibre level also need to be submitted. There are two challenging tasks with
network modelling. First, one can predict paper properties, i.e. continuum properties from prop-
erties at the fibre level, by assigning material data for the fibres and bonds. Second, mechanisms
on the fibre level can be studied in detail; this can include mechanisms such as fibre bridging at
delaminations, creep, fibre deformation etc. It is often the available computer resources that

Figure 12.7. Example of three-dimensional network model, (Heyden, 2000).

 325

limit the type of problem that is possible to study with network models. Since in general many
fibre are incorporated in network models, in order to simulate a real paper, the models easily be-
come computer demanding.

NOKQ==qÜÉ=cáåáíÉ=bäÉãÉåí=jÉíÜçÇ

The finite element method is a mathematical tool for solving ordinary and partial differential
equations that frequently appear the modelling of continuum mechanics problems. There are
several commercially available finite element programs. One of the problems with those is that
there are normally no predefined models for paper material available in these programs. How-
ever, commercial programs in general have an option to define your own material models, al-
though this is not an easy task and requires large experience. Several aspects should be
considered when defining your own material, as discussed below.
The finite element method is based on the principle of virtual work. In the finite element
method boundary conditions are used to constrain the model. These can be tractions, tr, or dis-
placements u on the outer surface, or body forces fr or corresponding displacement fields u act-
ing on the whole body. The energy given to the model by the boundary conditions need to be
transformed to internal stresses, V, and strains H. An energy balance can then be formulated
through the principle of virtual work

³ V : H dV0 ³ t ˜ udS  ³ f T ˜ udV ,

T
(12.6)
V0 S V

where the integrals indicate that the model as a whole need to fulfil the energy balance. Thus
note that the formulation is weak since it does not require energy balance in each point separate-
ly, as long as the whole model fulfils energy balance.
In the finite element formulation a displacement field is proposed, from that information can
the corresponding strain field be calculated, and thereafter the corresponding stress field is cal-
culated. It can be shown that this will lead to an upper limit of energy needed, i.e. it is not an ex-
act solution since the displacement field is assumed. But it will be a conservative solution. If the
exact strain field is known, the solution will be exact. In general, an accurate strain field gives a
better stress field.

NOKQKN==aáëéä~ÅÉãÉåíë

To be able to do an initial guess on the displacement field, space is divided into a mesh, i.e. dis-
crete elements with nodes along their boundaries are created. The nodes are the points where
neighbouring elements are connected to each other. In Figure 12.8 are some common elements
shown. Different types of elements are used depending on the dimensionality and geometry of
the problem.
326

Figure 12.8. Common two and three-dimensional finite elements.

When the geometry is deformed the nodes in each element will move, as illustrated in Figure
12.9, where a two-dimensional continuum element with four nodes is shown.

Figure 12.9. A four-noded continuum element. When deformation progress in a model, the nodes in the element
move.

From the boundary conditions it is possible to assume how each node move. By assuming
that the displacement between the nodes can be described by piecewise polynomial approxima-
tions of displacement fields it is possible to get the displacement of the whole element. In Fig-
ure 12.9 it is shown how an initial square element deforms, since three of its nodes get new
positions. In the figure is the deformation of the element edges assumed to be linear. It is then
possible to calculate a displacement of an arbitrary point within the element with aid of the dis-
placements of the corner nodes uN. This is done by defining a tensor NN, that contain interpola-
tion functions that depend on some material coordinate system,

u NNuN , (12.7)

where u is the incremental displacement at an arbitrary point in the element. In the same manner
strain can be calculated from the displacements of the corner nodes. Hence,

H ENu N (12.8)

where EN is a matrix that depends on position of the point being considered.

 327

NOKQKO==píêÉëëÉë

Constitutive equations are needed to calculate stresses from strains, i.e.,

V F H  . (12.9)

If the constitutive equations are elastic Equation (12.9) reduces to Hooks generalized law,
and the stresses are directly calculated by putting in the strains in the equation. However, if the
constitutive equations are non-linear, e.g. elastic-plastic, the equation needs to be solved numer-
ically in an iterative manner. This is because we need to check whether the calculated stress-
state is elastic or elastic-plastic, and in the latter case also verify that the final stress-state fulfils
the yield criteria as expressed generally by Equation (12.3).

NOKQKP==j~íêáñ=cçêãìä~íáçå

Above it is described how strain and stress can be calculated at arbitrary positions in an ele-
ment. When practical models are constructed they consist of many elements. Then, strain is cal-
culated in a certain number of points within each element. These are called integration points of
Gauss points. As this is done for all elements in the model, a strain field is created. By aid of
this information strain contours can be displayed in the model. Moreover, this creates a system
of equations where displacements in the integration points, 1-M, can be calculated from the
node displacements

ª u1 º ª N11   N1M º ª u11 º

«u » «   « »
« 2» «  »» « u12 »
(12.10)
«  » «    »«  »
« » « »« 1 »
¬u M ¼ ¬ N N 1   N NM ¼ «¬uM »¼

In the same manner the strains can be calculated from the displacements. Hence,

ª H1 º ª E11   E1M º ª u11 º

«H » «   « »
« 2» «  »» « u12 »
. (12.11)
«  » «    »«  »
« » « » 1« »
¬H M ¼ ¬ E N 1   E NM ¼ ¬«u M ¼»

Thus a matrix formulation comes in handy to solve the problems, and the two systems of
equations above can be replaced with a matrix formulation,

u Nu N , (12.12)

H EuN . (12.13)
328

If the position for each integration point is used as input to formulate the matrix EN. In the
same manner the stresses in the integration points, 1 – M, can be calculated

ª V1 º ª F1  H1 ,, H M  º
« » « »
«  » «  », (12.14)
¬«V M ¼» « FM  H1 ,, H M  »
¬ ¼

or in matrix notation

V F H  (12.15)

Moreover the surface tractions and body forces can be expressed in vector notation. Hence,

ª t1 º ª f1 º
t «» f «  »
« » « » (12.16)
«¬t M »¼ «¬ f M »¼

NOKQKQ==mêáåÅáéäÉ=çÑ=sáêíì~ä=tçêâ

To minimize the energy in the model is in general an iterative approach used. The aim is to
find a solution to a given problem by minimizing the energy in the system. This can be written
as

³ E N : V dV0 ³ N N ˜ tdS  ³ N N ˜ fdV .

T T
(12.17)
V0 S V

This system of equations is the basis for the finite element method, and can be written on the
form

F N uM  0. (12.18)

The solution of this system of equations is straight forward if the equations are linear, i.e. if
the material is linear elastic and small deformations are assumed. Then the problem has an ana-
lytical solution that can be found with e.g. Gauss elimination. If the system of equations instead
is non-linear, e.g. the materials has been modelled as elastic-plastic, viscoelastic, viscoplastic or
finite deformations are considered, the system of equations needs to be solve by a numerical
solver that iterates until a solution has been found. This is done be repeating the steps described
above. The following steps are then considered
• Displacements are proposed.
• Strains are calculated
• Stresses are calculated.
 329

• If Equation (12.18) is fulfilled within given tolerances then the solution is accepted, other-
wise are new displacements are proposed and the procedure is repeated.

NOKR==bñ~ãéäÉW=`êÉ~ëáåÖ=çÑ=m~éÉêÄç~êÇ

NOKRKN==aÉÑçêã~íáçå=jÉÅÜ~åáëãë

Before packages are folded the edged of what will be the box are creased. In Figure 12.10 pho-
tographs of a creasing operation performed in a scanning electron microscope can be seen. The
paperboard is placed above a female die, and a rule is pressed down into the paperboard. During
this operation several mechanisms are activated in the paperboard. The most important mecha-
nism will be:
1. Elastic deformation during loading in MD, CD and ZD.
2. Initiation of plastic deformation.
3. Plastic deformation in MD, CD and ZD.
4. Delamination between paper plies due to out-of-plane shear in MD and CD during loading.
5. Elastic deformation in MD, CD and ZD during unloading.
6. Delamination between paper plies due to normal stresses in ZD during unloading.
In the creasing operation it is a controlled amount of delamination of the paperboard that is
desirable, since this creates edges and corners in the box that has better properties than paper-
board that has not been creased. It is thus advantageous to cause unrecoverable damage in the
material during the creasing operation. A numerical model of paperboard that is used to model
creasing must therefore incorporate the delamination effect, since it is a feature that greatly in-
fluences the overall material and structural behaviour.

Figure 12.10. Scanning electron microscope pictures of the creasing process, (a) punch step and (b) after removal
of punch. (Dunn, 2000).
330

NOKRKO==jçÇÉä=dÉåÉê~íáçå

In order to model the creasing operation a constitutive framework has been proposed by Xia
(2000). The model consists of two parts, an in-plane model and a delamination model. The in-
plane model accounts for elastic-plastic properties of each layer in a paperboard, while the de-
lamination model accounts for delamination between or within layers. The delamination model
accounts for the plastic deformation in the ZD. The theoretical framework will be outlined be-
low; it will closely follow the work by Xia (2000). The constitutive equations account for large
deformation. It is not possible to outline the whole large deformation formalism in this chapter;
instead the interested reader is referred to for example Ottosen and Ristinmaa (1998).

NOKRKP==fåJéä~åÉ=jçÇÉä

The in-plane model assumes that there is elastic-plastic deformation in the MD-CD plane, and
elastic deformation in the ZD.

NOKRKPKN==píêÉëëJëíê~áå=oÉä~íáçåëÜáé

In large displacement analysis the deformed state, x, is distinguished from the undeformed state,
X. A line element, dx, in the deformed state depends on a line element in the undeformed state,
dX, through the deformation gradient tensor, F. Hence,

dx F dX (12.19)

where

wx
F . (12.20)
wX
It is assumed that the total deformation gradient F at a material point can be multiplicatively
decomposed into elastic and plastic parts. Hence

F F eF p (12.21)

where Fp represents the accumulation of inelastic deformation. Evaluation of plastic strains is

governed by

F p Lp F p , (12.22)

where Lp is the plastic velocity gradient that is defined by the flow rule. In this case it assumed
that

Lp J K , (12.23)

where J is the equivalent plastic strain and K is the normalized flow directions.
 331

The elastic strain is obtained by using the elastic Green strain measure,

1 eT e
Ee
2
F F I ,  (12.24)

where I is the second order identity tensor. The second Piola-Kirchoff stress measure, T, is relat-
ed to Green strain using the linear relationship

T C : Ee , (12.25)

where C is the fourth-order elastic stiffness tensor, which is taken to be orthotropic.

To model the through-thickness nonlinear elastic stress-strain relationship under ZD com-
pression, the through-thickness elastic modulus is taken to be exponential functions of the com-
pression. Hence,
e
EZD 0  aE22
EZD e (12.26)

NOKRKPKO==váÉäÇ=`çåÇáíáçå

In the model it is assumed that the yield surface can be constructed from six sub-surfaces, where
NI is the normal to the Ith sub-surface. The sub-surfaces are related to loading directions where it
is possible to do experiments. In this case the sub-surfaces are: MD tension, CD tension, shear
(positive), MD compression, CD compression and shear (negative). The criterion was proposed
since no existing criteria successfully described the experimental paperboard data available in
the literature. The yield criteria is expressed as

2k
6
§T :NI ·
f (T ) ¦FI ¨ I ¸ 1 , (12.27)
I 1 © S ¹

where k is a constant, normally k = 2, and SI are the equivalent strengths corresponding to the
sub-surfaces. In general SI depends on the equivalent plastic strain to incorporate the hardening
behaviour. Lastly, the switching parameter FI has the properties

­°1 if T : N I ! 0
FI ® . (12.28)
°̄0 otherwise

NOKRKQ==aÉä~ãáå~íáçå=jçÇÉä

The delamination model describes how two interfaces delaminate during loading. It is a cohe-
sive model that has elastic and plastic parts. Delamination is assumed to initiate due to tensile
loading in the out-of-plane direction of the paperboard, and due to out-of-plane shear loading.
332

NOKRKQKN==píêÉëëJëíê~áå=oÉä~íáçåëÜáé

The relative displacement between two opposing surfaces can be divided into elastic and plastic
parts. Thus each displacement component can be written as

G i G ie  G i p , (12.29)

where i represent the three loading direction: ZD tension, MD-ZD shear and CD-ZD shear in a
local coordinate system. The total displacement increment during a time increment 't thus be-
comes

'G i 'G ie  'G i p (12.30)

The change in the traction vector across the interface due to incremental relative displace-
ments is governed by

'TD KD ( 'GD  'G Dp ) , (12.31)

where .a denotes the components of the instantaneous interface stiffness in the D-direction,
which normally depend on the effective plastic displacement, G p. This dependence can be used
to formulate a damage law that mimics the experimentally observed behaviour.
The evolution of plastic displacements is in the model formulated as

'G i p F M i 'G p , (12.32)

where Mi are the normalised components of the unit flow, and F behaves such that

°­1 if f 0 and Ti 'G i p ! 0

F ® . (12.33)
°̄0 if f  0 or f 0 and Ti 'G i  0
p

NOKRKQKO==váÉäÇ=`êáíÉêáçå

Yielding occurs at the interface when

S1TD2
f (T , G p ) ¦ SD2
 T1  S1 0 (12.34)

where SD is the instantaneous interface strength which in general depends on the equivalent
plastic displacement, G p , according to an expression that mimics the observed experimental be-
haviour.
 333

NOKRKR==jçÇÉääáåÖ

After the theoretical model has been implemented into the commercial finite element program
ABAQUS (2004) it can be used to do simulations. Since the theoretical framework is general it
can be used to study any combination of loading. Before a model like this is used it is important
to verify it against experiments. It can be both macroscopic experiments, such as a tensile test,
and microscopic experiments. From the macroscopic experiments the principal behaviour of the
model can be verified. From the microscopic experiments it can be clarified that the same defor-
mation mechanisms are present in the model as in the real loading case, e.g. it need to be veri-
fied that delamination occur in the same manner and at the same position in the model and the
experimental set-up. When the model is verified against well defined experiments, the model is
trust worthy, and it can be used to study new loading cases and configurations. In Figure 12.11
shows two simulations of creasing of a five-ply paperboard. These simulate the experiments
shown in Figure 12.10, both with respect to material properties and external loading.

a) b)

Figure 12.11. Finite element modelling of the creasing process, (a) punch step and (b) after removal of punch.

NOKS==oÉÑÉêÉåÅÉë

ABAQUS (2004) version 6.4, www.abaqus.com

Dunn, H. (2000) Micromechanisms of paperboard mechanics. Master Thesis, Department of
Mechanical engineering, MIT, USA.
Heyden, S. (2000) Network modelling for the evaluation of mechanical properties of cellulose
fibre fluff. Doctoral Thesis, Division of Structural Mechanics, Lund University, Sweden.
Lif, J. (2003) Analysis of the time and humidity-dependent mechanical behaviour of paper webs
at offset printing press conditions. Doctoral Thesis, KTH Solid Mechanics, Sweden.
Persson, K. (2000) Micromechanical modelling of wood and fibre properties. Doctoral Thesis,
Report, Division of Structural Mechanics, Lund University, Sweden.
Saabye Ottosen, N., and Ristinmaa, M. (1998) The Mechanics of Constitutive Modelling Volume
2 Numerical and thermodynamical concepts. Division of Solid Mechanics, Lund Univer-
sity, Sweden.
100
 335

fåÇÉñ

A
absorption, Bristow method 212 chemical forces, fibres 77–79
absorption coefficient 151–152 chromaticity coordinates 162–163
absorptivity, surface 216 CIE whiteness 167
adhesion, paper strength 170–172 CIELAB-coordinates 165
adipic acid, wet strength 192 climatology 114–116
adsorption, moisture, isotherms 117 coated carton board, structure 20
air, moisture content 113–114 coated paper, light weight, structure 19
air-borne drying 106 cockling 133–134
air-leak methods, surface roughness colour, separation 245
213–214 colour perception 161
air permeance 66–67 – tristimulus values 161
alkaline-curing resins 192–196 colour printing 245
amines, wet strength 193 colour space, CIELAB 165
anisotropy colour vision, human eye 156
– isotropic hand-sheets 96 compressibility, surface 214–215
– paper properties 92–93 compression index, moisture content 142
azetidinium chloride, wet strength 194 compressive collapse, paper boxes 3
condebelt drying 105
B consolidation, bonded joint 170
bending curvature, humidity 136 contact angle, surface 217
bending resistance 56–57 contact area, surface 212
bending stiffness continuum models, mechanical modelling
– laminate 303–305 321–324
– paper 2, 55–56 converting
board, corrugated 259–261 – mathematical modelling 315–333
bonded joint, drying 170 – moisture influences 115–116
bonding mechanisms, fibres 73–74 – packaging materials 258–270
book paper 246 converting operations, simulation 60–61
Brecht-Knittweis method 62 copy paper, structure 19
brightness corrugated board 259–261
– ISO 160, 167 creasing
Bristow absorption method 212 – deformation mechanisms 329
buckling, paper 45 – delamination model 331–332
bursting strength, paper 53–54 – in-plane model 330–331
– model generation 330
C – modelling 333
capillary forces, fibres 75–77 – packaging materials 263–265
carton board – paper 13
– coated, structure 20 – paperboard 329–333
– delamination 5 – stress-strain relationship 330–332
categories, paper 15 – yield condition 331
336

– yield criterion 332 E

creep, mechanosorptive 143 e-print, packaging 284
creep curves 50 e-tags, packaging 284–285
curl enduse, mathematical modelling 315–333
– laminate 308 epichlorohydrin, wet strength 193
– measurement 135, 139 excitation purity 163–164
– moisture influence 133–140 eye, colour vision 156
– polar diagram 309–310
cutting, packaging materials 263–265 F
cylinder drying 97–103 fibre lifting, surface 224
fibre network, micro level 81–84
D fibre textiles, properties compared to
delamination 14 paper 22
– carton board 5 fibres
delamination model, creasing 331–332 – bonding mechanisms 73–74
delamination resistance 58–60 – capillary forces 75–77
density, paper 29 – chemical forces 77–79
desorption, moisture 119 – functional groups 203–204
diethylenetriamine, wet strength 192 – micro level 79–84
digital printing 238–242 – nano level forces 73–79
– packaging 284 – network forces 75
dimensional changes – tensile strength 34–36
– measurement 130–132 fine paper, structure 17
– moisture cycling 126–130 finite element method 316, 325–329
dimethylolurea, wet strength 189 – matrix formulation 327–328
displacement, finite element method – virtual work 328
325–326 flexible packaging 258
dry strength additives 172–182 flexography 235–236
– compression properties 178–181 – inks 250
– tensile properties 172–177 fluorescence 166–167
dry strength agents, mechanisms 169–183 flute types 260
dry strength mechanisms 169–183 fold number 57–58
drying 69–107 folded printing papers, leaning stacks
– bonded joint 170 308–309
– concepts 103–107 folding, packaging materials 265–266
– cylinder 97–103 folding endurance 57–58
– hygroexpansion coefficient 123–127 folding failure 12
– impingement 104 forces
– multi-ply board 89–90 – capillary, fibres 75–77
– observation 70 – chemical, fibres 77–79
– paper sheet 92–97 – fibres, nano level 73–79
– phases 97–102 – laminate 299–301, 303
– pressing 97 – network, fibres 75
– stock preparation 90–92 formaldehyde, wet strength 189
drying stress 86 fracture mechanics, paper 63–65
friction, surface 225
 337

functional groups, fibres 203–204 inks

– flexography 250
G – printing 248–251
g-PAM see glyoxalated polyacrylamide ISO-brightness 160, 167
gloss ISO-hierarchy, optical measurements 158
– measurement 229–230 isotropic hand-sheets, anisotropy 96
– scale 230 isotropic materials, laminate 295–296
– surface 225–230
– variations 230 J
gluing, packaging materials 265–266 joint strength, adhesion 170
glyoxalated polyacrylamide (g-PAM)
196–200 K
grammage, paper 26 Kelvin equation, sorption 118
graphic papers 247 Kubelka-Munk theory, light reflection
gray scale, printing 242–244 147–152
greaseproof paper, structure 18 – applications 152
gurley-apparatus 66 – limitations 151

H L
hand-sheets, anisotropy 96 laminate
human eye, colour vision 156 – bending stiffness 303–305
humidity, bending curvature 136 – curl 308
hygroexpansion 120–127 – definition 289–292
– laminate, stress ans strain 314 – forces 299–301, 303
– moisture content 129 – hygroexpansion coefficient 302
– orientation 310 – isotropic materials 295–296
– polar diagram 310 – matrices 294–295
– shear-free angle 311 – moisture absorption, strain 296–299
– strain 130 – moments 299–301, 303
hygroexpansion coefficient 120 – orthotropic materials 295–296
– drying 123–127 – properties, moisture 307–308
– laminate 302 – stiffness matrix 302
hysteresis, sorption isotherms 116–120 – strain transformation 301
– strain, mechanical influences 299
– strain, moisture absorption 296–299
I – stress transformation 301
IGT tester 61 – tensile stiffness 303–305
illuminants, standard 155 – twist 308
impingement drying 104 – twist free angle 311–313
ink drying 248–251 laminate layer 289–292
– chemical 255 – angle 301
– offset printing 253–256 – strain 292–294
– physical 254 – stress 292–294
ink transfer 218–220 laminate theory, papermaker 287–314
inkjet inks 251 length scales, mechanical modelling 320
inkjet paper 247–248 level deviation, moisture ratio 141
338

lifetime, paper boxes 4 – twist 133–140

light absorption 146–147 moisture content 111, 141–143
light absorption coefficient 152 – air 113–114
light reflection, Kubelka-Munk theory – compression index 142
147–152 – hygroexpansion 129
light scattering 146–147 – tensile stiffness 142
light scattering coefficient 152 moisture cycling, dimensional changes
light weight coated paper, structure 19 126–130
linerboard, structure 16 moisture ratio 111
linting, surface 223–224 – stiffness 141
lithography, offset 234–235 mollier-diagram, moist air 113
monomethylol, wet strength 191
M mottle, surface 223
machine level, wide web 97–107 Mullen test, bursting strength 53
macro level, paper sheet 84–97 multiply board, drying 89–90
magazine paper 246 multilayers, polyelectrolyte, paper
materials, properties compared to paper strength 176
21–23
mathematical modelling N
– converting 315–333 nano level forces, fibres 73–79
– enduse 315–333 network, micro level 79–84
matrices, laminate 294–295 network forces, fibres 75
matrix formulation, finite element method network models 324
327–328 newsprint 245
mechanical modelling 317–320 – structure 16
– continuum models 321–324
– length scales 320 O
– network models 324 office paper 246
mechanical properties offset inks 248–249
– affecting factors 8–9 offset lithography 234–235
– moisture content 141–143 offset printing, ink drying 253–256
mechanical strain 86 optical measurements 159–160
mechanosorptive creep 143 – instrument calibration 158
melamine-formaldehyde, wet strength – ISO-hierarchy 158
190–192 optical properties 145–167
modelling – measurement conditions 156–158
– mechanical 317–320 – papermaking 146
– complex papers 5 – quality control 146
– paper structural levels 5–7 – standard observers 156
moisture optical quantities, relationships 153–154
– absorption, laminate 296–299 orthotropic materials, laminate 295–296
– absorption, paper 112
– curl 133–140 P
– laminate, properties 307–308 packages
– mollier-diagram 113 – filling 266–268
– sorption isotherms 116–120 – interaction 282–284
 339

packaging 257–286 – fracture mechanics 63–65

– compression requirements 278–282 – grammage 26
– deformation 283 – hygroexpansion 120–127
– design 276–284 – interaction with water vapour 109–143
– digital printing 284 – mechanical aspects 320–325
– e-print 284 – modelling of structure 1–24
– e-tags 284–285 – moisture adsorption 110, 112
– environmental exposures 282 – nature 9–10
– future trends 284–286 – physical nature 8
– marketing performance 284 – Poisson´s ratio 30–32
– performance, design 276–284 – printing 245–248
– performance, requirements 270–276 – specific stress 30
– performance, technical 277–284 – stiffness determination 40–41
– requirements 270–276 – strain 30–32
– smart packages 285 – strain at break 37
– supply chain 270–273 – stress 30
packaging materials 258–270 – stress at break 46
– barrier properties 269 – stress-strain curves 47
– converting 258–270 – structure 1–24
– creasing 263–265 – structure of products 15–20
– cutting 263–265 – tearing resistance 51
– folding 265–266 – tearing resistance, theory 52–53
– gluing 265–266 – tensile energy absorption 37
– new 286 – tensile stiffness 38–39
PAE see poly(aminoamide)-epichlorohy- – tensile strength 32–34
drin – tensile test 31
Page´s theory, tensile strength 36 – thickness 26–29
paper – thickness direction strength 58–60
– air permeance 66–67 – thickness measurement 27
– and printing 233–256 – twist 137
– bending resistance 56–57 paper machine, web break 4
– bending stiffness 55–56 paper physics 25–67
– buckling 45 paper products, structure 15–20
– bursting strength 53–54 paper properties
– categories 15 – affecting factors 8–9
– creasing 13 – anisotropy 92–93
– creep curves 50 – compared to materials 21–23
– delamination 14 – compression 45–49
– delamination resistance 58–60 – creep 43–45
– density 29 – drying 87
– directions 10 – during drying 69–107
– drying 71–72 – elastic 37–38
– elastic-plastic behaviour 322–324 – functional 1–5
– fold number 57–58 – macro level 84–89
– folding endurance 57–58 – modelling, complexity 5
– folding failure 12 – modelling, structural levels 5–7
340

– moisture content 141–143 printability, surface 218–223

– optical 145–167 printing
– standardization 23 – colour 245
– strain 87–89 – conventional techniques 234–237
– stress-strain 42–43 – digital 238–242
– viscoelastic 49–51 – gray scale 242–244
paper sheet – resolution 242–244
– drying 92–97 printing inks 248–251
– macro level 84–97 printing papers 233–256
paper strength – folded, leaning stacks 308–309
– adhesion 170–172 production control, methods 61–63
– dry strength additives 172–182 pulp, optical properties 145–167
– polyelectrolyte multilayers 176
paper surface see surface Q
paperboard, creasing 329–333 quality control, optical properties 146
papermaker, laminate theory 287–314
papermaking R
– moisture influences 115–116 radiance factor, fluorescence 167
– optical properties 146 reflectance, surface 228–229
– water content 71 reflectance factor, measurement 154–155
permeance, surface 216 reflection, surface 226–228
pick strength, surface 224 reflectivity 150
plastic behaviour, paper 322–324 reflectivity factor, measurement 154–155
Poisson´s ratio, paper 30–32 resins
polar diagram, mechanical properties – alkaline-curing 192–196
309–310 – wet-strength 189–203
poly(aminoamide)-epichlorohydrin (PAE) resolution, printing 242–244
– crosslinked 195 rigid packaging 258
– wet strength 192 rotogravure 236–237
polyacrylamide resin, glyoxalated – inks 249
196–200 roughness
polyelectrolyte multilayers, paper – measures 211–212
strength 176 – surface 211–214
polyethylene film, properties compared to
paper 22 S
porosity, surface 215 scattering coefficient 151–152
prepress 242–245 Scott bond test 62
pressing, drying 97 set-off, surface 222–223
print shear-free angle, hygroexpansion 311
– quality 152–253 short-span test 34–36
– unevenness 253 shrinkage, macro level 84–89
print density 220, 252 single fibre, network 79–81
– unevenness 253 smart packages, packaging 285
print gloss 252 soak time, wet strength 187
– unevenness 253 sorption
print through 221–222 – Kelvin equation 118
 341

– moisture, isotherms 116–120 – air leak methods 213–214

specific stress, paper 30 – compressibility 214–215
speckle photography 131 – contact angle 217
spectral power distribution, standard illumi- – contact area 212
nants 155 – creation 230–232
spectral radiance factor, fluorescence 167 – friction 225
stacks, folded printing papers 308–309 – gloss 225–230
standard illuminants 155 – gloss measurement 229–230
– spectral power distribution 155 – linting 223–224
standard observers, optical properties – permeance 216
156–158 – porosity 215
standardization, paper properties 23 – printability 218–223
standardized measurement conditions, opti- – reflectance 228–229
cal properties 156–158 – reflection 226–228
starch, wet strength 200–203 – roughness 211–214
STFI deformation 131 – strength 223–224
stiffness – two-sidedness 210–211
– moisture ratio 141 – volume 212–213
– polar diagram 309–310 – wettability 217–218
stiffness determination, ultrasonic tech- – wetting delay 218
niques 40–41
stiffness matrix, laminate 302 T
stock preparation, drying 90–92 tearing resistance 51
strain 30–32 – theory 52–53
– and paper properties 87–89 tensile energy absorption, paper 37
– laminate layer 292–294 tensile stiffness 38–39
– laminate, hygroexpansion 314 – laminate 303–305
– laminate, mechanical influences 299 – moisture content 142
– laminate, moisture absorption 296–299 – orientation 310
– mechanical 86 tensile strength 32–34
strain at break 37 – fibres 34–36
strain transformation, laminate 301 – Page´s theory 36
stress 30 tensile test 31
– drying 86 tension, creep properties 43–45
– finite element method 327 textiles, properties compared to paper 22
– laminate layer 292–294 thickness, paper 26–29
– laminate, hygroexpansion 314 – measurement 27
stress at break 46 thickness direction strength 58–60
stress-strain, and paper properties 42–43 tissue, structure 17
stress-strain curves 47 tristimulus values, colour perception 161
stress-strain relationship, creasing twist 137
330–332 – laminate 308
stress transformation, laminate 301 – moisture influence 133–140
stretch, macro level 84–89 – polar diagram 309–310
surface 209–232 twist free angle, laminate 311–313
– absorptivity 216 two-sidedness, surface 210–211
342

U
ultrasonic techniques, stiffness determina-
tion 40–41
ultraviolet curable inks 250
unevenness
– print density 253
– print gloss 253
urea-formaldehyde, wet strength 189–190

V
vapour, interaction with paper 109–143
virtual work, finite element method 328

W
water content, papermaking 71
water vapour, interaction with paper
109–143
wavelength, dominant 163–164
web break, paper machine 4
wet strength
– chemicals 204–205
– mechanisms 185–188
– protection mechanism 187
– reinforcement mechanism. 188
– soak time 187
wet strength agents, mechanisms 185–208
wet-strength resins, chemistry 189–203
wettability, surface 217–218
wetting delay, surface 218
whiteness 165–166
– CIE 167
wide web, machine level 97–107

Z
zero-span test 34–36
 343
DE GRUYTER

HOLZFORSCHUNG
International Journal of the Biology,
Chemistry, Physics, and Technology
of Wood
Editor-in-Chief: Oskar Faix, Germany
Publication frequency: bi-monthly (6 issues per year).
Approx. 700 pages per volume. 21 x 29.7 cm
ISSN (Print) 0018-3830
ISSN (Online) 1437-434X
CODEN HOLZAZ
Language: Englisch

Holzforschung is an international scholarly journal that publishes cutting-edge research on the biol-
ogy, chemistry, physics and technology of wood and wood components. High quality papers about
biotechnology and tree genetics are also welcome. Rated year after year as the number one scientific
journal in the category of Pulp and Paper (ISI Journal Citation Index), Holzforschung represents inno-
vative, high quality basic and applied research. The German title reflects the journal‘s origins in a long
scientific tradition, but all articles are published in English to stimulate and promote cooperation be-
tween experts all over the world. Ahead-of-print publishing ensures fastest possible knowledge transfer.

Indexed in: Academic OneFile (Gale/Cengage Learning) – Aerospace & High Technology Database –
Aluminium Industry Abstracts – CAB Abstracts – Ceramic Abstracts/World Ceramic Abstracts –
Chemical Abstracts and the CAS databases – Computer & Information Systems Abstracts – Copper
Data Center Database – Corrosion Abstracts – CSA Illustrata – Natural Sciences – CSA / ASCE Civil
Engineering Abstracts – Current Contents/Agriculture, Biology, and Environmental Sciences – Earth-
quake Engineering Abstracts – Electronics & Communications Abstracts – EMBiology – Engineered
Materials Abstracts – Engineering Information: Compendex – Engineering Information: PaperChem –
Journal Citation Reports/Science Edition – Materials Business File – Materials Science Citation Index –
Mechanical & Transportation Engineering Abstracts – METADEX – Paperbase – Science Citation
Index – Science Citation Index Expanded (SciSearch) – Scopus – Solid State & Superconductivity
Abstracts.

All de Gruyter journals are now hosted on Reference Global,

de Gruyter’s new and integrated platform.
Please visit www.reference-global.com for more information
and free TOC alerts.
Electronic sample copy at www.degruyter.com/journals/holz

www.degruyter.com