Pre Stress

Prestressed Concrete Structures Dr. Amlan K Sengupta and Prof.
Devdas Menon
1.1 Introduction
This section covers the following topics.
Basic Concept
Early Attempts of Prestressing
Brief History
Development of Building Materials
1.1.1 Basic Concept
A prestressed concrete structure is different from a conventional reinforced concrete

structure due to the application of an initial load on the structure prior to its use.
The initial load or prestress is applied to enable the structure to counteract the stresses
arising during its service period.
The prestressing of a structure is not the only instance of prestressing. The concept of
prestressing existed before the applications in concrete. Two examples of prestressing
before the development of prestressed concrete are provided.
Force-fitting of metal bands on wooden barrels

The metal bands induce a state of initial hoop compression, to counteract the hoop
tension caused by filling of liquid in the barrels.
Metal bands
Figure 1-1.1 Force-fitting of metal bands on wooden barrels
Pre-tensioning the spokes in a bicycle wheel

The pre-tension of a spoke in a bicycle wheel is applied to such an extent that there will
always be a residual tension in the spoke.
Indian Institute of Technology Madras

Prestressed Concrete Structures Dr. Amlan K Sengupta and Prof. Devdas Menon
Spokes
Figure 1-1.2 Pre-tensioning the spokes in a bicycle wheel
For concrete, internal stresses are induced (usually, by means of tensioned steel) for
the following reasons.
The tensile strength of concrete is only about 8% to 14% of its compressive
strength.
Cracks tend to develop at early stages of loading in flexural members such as
beams and slabs.
To prevent such cracks, compressive force can be suitably applied in the
perpendicular direction.
Prestressing enhances the bending, shear and torsional capacities of the flexural
members.
In pipes and liquid storage tanks, the hoop tensile stresses can be effectively
counteracted by circular prestressing.
1.1.2 Early Attempts of Prestressing
Prestressing of structures was introduced in late nineteenth century. The following

sketch explains the application of prestress.
Place and stretch mild steel rods, prior to concreting
Release the tension and cut the rods after concreting

Figure 1-1.3 Prestressing of concrete beams by mild steel rods

Mild steel rods are stretched and concrete is poured around them. After hardening of
concrete, the tension in the rods is released. The rods will try to regain their original
length, but this is prevented by the surrounding concrete to which the steel is bonded.
Thus, the concrete is now effectively in a state of pre-compression. It is capable of
counteracting tensile stress, such as arising from the load shown in the following sketch.
Figure 1-1.4 A prestressed beam under an external load
But, the early attempts of prestressing were not completely successful. It was observed
that the effect of prestress reduced with time. The load resisting capacities of the
members were limited. Under sustained loads, the members were found to fail. This
was due to the following reason.
Concrete shrinks with time. Moreover under sustained load, the strain in concrete
increases with increase in time. This is known as creep strain. The reduction in length
due to creep and shrinkage is also applicable to the embedded steel, resulting in
significant loss in the tensile strain.
In the early applications, the strength of the mild steel and the strain during prestressing
were less. The residual strain and hence, the residual prestress was only about 10% of
the initial value. The following sketches explain the phenomena.

Original length of steel rod (L1)
Original length of concrete beam (L2)

a) Beam before applying prestress
Reduced length of concrete beam (L3)
b) Beam at transfer of prestress
Final length of prestressed beam (L4)
c) Beam after long-term losses of prestress

Figure 1-1.5 Variation of length in a prestressed beam
The residual strain in steel = original tensile strain in steel compressive strains
corresponding to short-term and long-term losses.
Original tensile strain in steel = (L2 L1)/L1

Compressive strain due to elastic shortening of beam = (L2 L3)/L1
(short-term loss in prestress)
Compressive strain due to creep and shrinkage = (L3 L4)/L1
(long-term losses in prestress)
Therefore, residual strain in steel = (L4 L1)/L1
The maximum original tensile strain in mild steel = Allowable stress / elastic
modulus
= 140 MPa / 2105 MPa
= 0.0007

The total loss in strain due to elastic shortening, creep and shrinkage was also close to
0.0007. Thus, the residual strain was negligible.
The solution to increase the residual strain and the effective prestress are as follows.
Adopt high strength steel with much higher original strain. This leads to the
scope of high prestressing force.
Adopt high strength concrete to withstand the high prestressing force.
1.1.3 Brief History
Before the development of prestressed concrete, two significant developments of

reinforced concrete are the invention of Portland cement and introduction of steel in
concrete. These are also mentioned as the part of the history. The key developments
are mentioned next to the corresponding year.
1824 Aspdin, J., (England)

Obtained a patent for the manufacture of Portland cement.
1857 Monier, J., (France)

Introduced steel wires in concrete to make flower pots, pipes, arches and slabs.
The following events were significant in the development of prestressed concrete.

1886 Jackson, P. H., (USA)
Introduced the concept of tightening steel tie rods in artificial stone and concrete
arches.
Figure 1-1.6 Steel tie rods in arches
1888 Doehring, C. E. W., (Germany)

Manufactured concrete slabs and small beams with embedded tensioned steel.

1908 Stainer, C. R., (USA)

Recognised losses due to shrinkage and creep, and suggested retightening the
rods to recover lost prestress.
1923 Emperger, F., (Austria)

Developed a method of winding and pre- tensioning high tensile steel wires
around concrete pipes.
1924 Hewett, W. H., (USA)

Introduced hoop-stressed horizontal reinforcement around walls of concrete
tanks through the use of turnbuckles.
Thousands of liquid storage tanks and concrete pipes were built in the two decades to
follow.
1925 Dill, R. H., (USA)
Used high strength unbonded steel rods. The rods were tensioned and anchored
after hardening of the concrete.
Figure 1-1.7 Portrait of Eugene Freyssinet

(Reference: Collins, M. P. and Mitchell, D.,Prestressed Concrete Structures)
1926 Eugene Freyssinet (France)

Used high tensile steel wires, with ultimate strength as high as 1725 MPa and
yield stress over 1240 MPa. In 1939, he developed conical wedges for end
anchorages for post-tensioning and developed double-acting jacks. He is often
referred to as the Father of Prestressed concrete.

1938 Hoyer, E., (Germany)

Developed long line pre-tensioning method.
1940 Magnel, G., (Belgium)

Developed an anchoring system for post-tensioning, using flat wedges.
During the Second World War, applications of prestressed and precast concrete
increased rapidly. The names of a few persons involved in developing prestressed
concrete are mentioned. Guyon, Y., (France) built numerous prestressed concrete
bridges in western and central Europe. Abeles, P. W., (England) introduced the
concept of partial prestressing. Leonhardt, F., (Germany), Mikhailor, V., (Russia) and
Lin, T. Y., (USA) are famous in the field of prestressed concrete.
The International Federation for Prestressing (FIP), a professional organisation in

Europe was established in 1952. The Precast/Prestressed Concrete Institute (PCI) was
established in USA in 1954.
Prestressed concrete was started to be used in building frames, parking structures,

stadiums, railway sleepers, transmission line poles and other types of structures and
elements.
In India, the applications of prestressed concrete diversified over the years. The first
prestressed concrete bridge was built in 1948 under the Assam Rail Link Project.
Among bridges, the Pamban Road Bridge at Rameshwaram, Tamilnadu, remains a
classic example of the use of prestressed concrete girders.

Figure 1-1.8 Pamban Road Bridge at Rameshwaram, Tamilnadu

(Reference: http://www.ramnad.tn.nic.in)
1.1.4 Development of Building Materials
The development of prestressed concrete can be studied in the perspective of

traditional building materials. In the ancient period, stones and bricks were extensively
used. These materials are strong in compression, but weak in tension. For tension,
bamboos and coir ropes were used in bridges. Subsequently iron and steel bars were
used to resist tension. These members tend to buckle under compression. Wood and
structural steel members were effective both in tension and compression.
In reinforced concrete, concrete and steel are combined such that concrete resists
compression and steel resists tension. This is a passive combination of the two
materials. In prestressed concrete high strength concrete and high strength steel are
combined such that the full section is effective in resisting tension and compression.
This is an active combination of the two materials. The following sketch shows the use
of the different materials with the progress of time.

Compression (C) Tension (T) C and T
Stones, Bricks Bamboo, Ropes Timber
Concrete Steel bars, wires Structural steel
Passive combination Reinforced

Concrete
High Strength High Strength
Concrete Steel
Active combination Prestressed

Concrete
Figure 1-1.9 Development of building materials

(Reference: Lin, T. Y. and Burns, N. H.,
Design of Prestressed Concrete Structures)

1.2 Advantages and Types of Prestressing

Definitions
Advantages of Prestressing
Limitations of Prestressing
Types of Prestressing
1.2.1 Definitions
The terms commonly used in prestressed concrete are explained. The terms are placed
in groups as per usage.
Forms of Prestressing Steel

Wires
Prestressing wire is a single unit made of steel.
Strands
Two, three or seven wires are wound to form a prestressing strand.
Tendon
A group of strands or wires are wound to form a prestressing tendon.
Cable
A group of tendons form a prestressing cable.
Bars
A tendon can be made up of a single steel bar. The diameter of a bar is much
larger than that of a wire.
The different types of prestressing steel are further explained in Section 1.7,
Prestressing Steel.
Nature of Concrete-Steel Interface

Bonded tendon
When there is adequate bond between the prestressing tendon and concrete, it is called
a bonded tendon. Pre-tensioned and grouted post-tensioned tendons are bonded
tendons.

Unbonded tendon
When there is no bond between the prestressing tendon and concrete, it is called
unbonded tendon. When grout is not applied after post-tensioning, the tendon is an
unbonded tendon.
Stages of Loading
The analysis of prestressed members can be different for the different stages of loading.
The stages of loading are as follows.
1) Initial : It can be subdivided into two stages.

a) During tensioning of steel
b) At transfer of prestress to concrete.
2) Intermediate : This includes the loads during transportation of the
prestressed members.
3) Final : It can be subdivided into two stages.
a) At service, during operation.
b) At ultimate, during extreme events.
1.2.2 Advantages of Prestressing
The prestressing of concrete has several advantages as compared to traditional

reinforced concrete (RC) without prestressing. A fully prestressed concrete member is
usually subjected to compression during service life. This rectifies several deficiencies
of concrete.
The following text broadly mentions the advantages of a prestressed concrete member
with an equivalent RC member. For each effect, the benefits are listed.
1) Section remains uncracked under service loads

Reduction of steel corrosion
Increase in durability.
Full section is utilised
Higher moment of inertia (higher stiffness)
Less deformations (improved serviceability).

Increase in shear capacity.

Suitable for use in pressure vessels, liquid retaining structures.
Improved performance (resilience) under dynamic and fatigue loading.
2) High span-to-depth ratios
Larger spans possible with prestressing (bridges, buildings with large column-free
spaces)
Typical values of span-to-depth ratios in slabs are given below.
Non-prestressed slab 28:1
Prestressed slab 45:1
For the same span, less depth compared to RC member.

Reduction in self weight
More aesthetic appeal due to slender sections
More economical sections.
3) Suitable for precast construction

The advantages of precast construction are as follows.
Rapid construction
Better quality control
Reduced maintenance
Suitable for repetitive construction
Multiple use of formwork
Reduction of formwork
Availability of standard shapes.
The following figure shows the common types of precast sections.

T-section Double T-section
Hollow core Piles
L-section Inverted T-section I-girders

Figure 1-2.1 Typical precast members
1.2.3 Limitations of Prestressing
Although prestressing has advantages, some aspects need to be carefully addressed.

Prestressing needs skilled technology. Hence, it is not as common as reinforced
concrete.
The use of high strength materials is costly.
There is additional cost in auxiliary equipments.
There is need for quality control and inspection.
1.2.4 Types of Prestressing
Prestressing of concrete can be classified in several ways. The following classifications

are discussed.
Source of prestressing force

This classification is based on the method by which the prestressing force is generated.
There are four sources of prestressing force: Mechanical, hydraulic, electrical and
chemical.

External or internal prestressing

This classification is based on the location of the prestressing tendon with respect to the
concrete section.
Pre-tensioning or post-tensioning
This is the most important classification and is based on the sequence of casting the
concrete and applying tension to the tendons.
Linear or circular prestressing

This classification is based on the shape of the member prestressed.
Full, limited or partial prestressing

Based on the amount of prestressing force, three types of prestressing are defined.
Uniaxial, biaxial or multi-axial prestressing

As the names suggest, the classification is based on the directions of prestressing a
member.
The individual types of prestressing are explained next.
Source of Prestressing Force

Hydraulic Prestressing
This is the simplest type of prestressing, producing large prestressing forces. The
hydraulic jack used for the tensioning of tendons, comprises of calibrated pressure
gauges which directly indicate the magnitude of force developed during the tensioning.
Mechanical Prestressing
In this type of prestressing, the devices includes weights with or without lever
transmission, geared transmission in conjunction with pulley blocks, screw jacks with or
without gear drives and wire-winding machines. This type of prestressing is adopted for
mass scale production.

Electrical Prestressing
In this type of prestressing, the steel wires are electrically heated and anchored before
placing concrete in the moulds. This type of prestressing is also known as thermo-
electric prestressing.
External or Internal Prestressing

External Prestressing
When the prestressing is achieved by elements located outside the concrete, it is called
external prestressing. The tendons can lie outside the member (for example in I-girders
or walls) or inside the hollow space of a box girder. This technique is adopted in
bridges and strengthening of buildings. In the following figure, the box girder of a bridge
is prestressed with tendons that lie outside the concrete.
Figure 1-2.2 External prestressing of a box girder

(Reference: VSL International Ltd.)
Internal Prestressing
When the prestressing is achieved by elements located inside the concrete member
(commonly, by embedded tendons), it is called internal prestressing. Most of the
applications of prestressing are internal prestressing. In the following figure, concrete
will be cast around the ducts for placing the tendons.

Figure 1-2.3 Internal prestressing of a box girder

(Courtesy: Cochin Port Trust, Kerala)
Pre-tensioning or Post-tensioning
Pre-tensioning
The tension is applied to the tendons before casting of the concrete. The pre-
compression is transmitted from steel to concrete through bond over the transmission
length near the ends. The following figure shows manufactured pre-tensioned electric
poles.
Figure 1-2.4 Pre-tensioned electric poles

(Courtesy: The Concrete Products and Construction Company, COPCO, Chennai)

Post-tensioning
The tension is applied to the tendons (located in a duct) after hardening of the concrete.
The pre-compression is transmitted from steel to concrete by the anchorage device (at
the end blocks). The following figure shows a post-tensioned box girder of a bridge.
Figure 1-2.5 Post-tensioning of a box girder

The details of pre-tensioning and post-tensioning are covered under Section 1.3, Pre-
tensioning Systems and Devices, and Section 1.4, Post-tensioning Systems and
Devices, respectively.
Linear or Circular Prestressing

Linear Prestressing
When the prestressed members are straight or flat, in the direction of prestressing, the
prestressing is called linear prestressing. For example, prestressing of beams, piles,
poles and slabs. The profile of the prestressing tendon may be curved. The following
figure shows linearly prestressed railway sleepers.

Figure 1-2.6 Linearly prestressed railway sleepers

Circular Prestressing
When the prestressed members are curved, in the direction of prestressing, the
prestressing is called circular prestressing. For example, circumferential prestressing of
tanks, silos, pipes and similar structures. The following figure shows the containment
structure for a nuclear reactor which is circularly prestressed.
Figure 1-2.7 Circularly prestressed containment structure, Kaiga Atomic Power

Station, Karnataka
(Reference: Larsen & Toubro Ltd, ECC Division, 60 Landmark Years)

Full, Limited or Partial Prestressing

Full Prestressing
When the level of prestressing is such that no tensile stress is allowed in concrete under
service loads, it is called Full Prestressing (Type 1, as per IS:1343 - 1980).
Limited Prestressing
When the level of prestressing is such that the tensile stress under service loads is
within the cracking stress of concrete, it is called Limited Prestressing (Type 2).
Partial Prestressing
When the level of prestressing is such that under tensile stresses due to service loads,
the crack width is within the allowable limit, it is called Partial Prestressing (Type 3).
Uniaxial, Biaxial or Multiaxial Prestressing

Uniaxial Prestressing
When the prestressing tendons are parallel to one axis, it is called Uniaxial Prestressing.
For example, longitudinal prestressing of beams.
Biaxial Prestressing
When there are prestressing tendons parallel to two axes, it is called Biaxial
Prestressing. The following figure shows the biaxial prestressing of slabs.
Non-prestressed reinforcement
Duct for
prestressing
tendon
Figure 1-2.8 Biaxial prestressing of a slab
(Courtesy: VSL India Pvt. Ltd., Chennai)

Multiaxial Prestressing
When the prestressing tendons are parallel to more than two axes, it is called Multiaxial
Prestressing. For example, prestressing of domes.

1.3 Pre-tensioning Systems and Devices

Introduction
Stages of Pre-tensioning
Advantages of Pre-tensioning
Disadvantages of Pre-tensioning
Devices
Manufacturing of Pre-tensioned Railway Sleepers
1.3.1 Introduction
Prestressing systems have developed over the years and various companies have
patented their products. Detailed information of the systems is given in the product
catalogues and brochures published by companies. There are general guidelines of
prestressing in Section 12 of IS:1343 - 1980. The information given in this section is
introductory in nature, with emphasis on the basic concepts of the systems.
The prestressing systems and devices are described for the two types of prestressing,
pre-tensioning and post-tensioning, separately. This section covers pre-tensioning.
Section 1.4, Post-tensioning Systems and Devices, covers post-tensioning. In pre-
tensioning, the tension is applied to the tendons before casting of the concrete. The
stages of pre-tensioning are described next.
1.3.2 Stages of Pre-tensioning
In pre-tensioning system, the high-strength steel tendons are pulled between two end
abutments (also called bulkheads) prior to the casting of concrete. The abutments are
fixed at the ends of a prestressing bed.
Once the concrete attains the desired strength for prestressing, the tendons are cut
loose from the abutments.

The prestress is transferred to the concrete from the tendons, due to the bond between
them. During the transfer of prestress, the member undergoes elastic shortening. If the
tendons are located eccentrically, the member is likely to bend and deflect (camber).
The various stages of the pre-tensioning operation are summarised as follows.
1) Anchoring of tendons against the end abutments
2) Placing of jacks
3) Applying tension to the tendons
4) Casting of concrete
5) Cutting of the tendons.
During the cutting of the tendons, the prestress is transferred to the concrete with elastic
shortening and camber of the member.
The stages are shown schematically in the following figures.

End
abutment Steel tendon
Jack
Prestressing bed
(a) Applying tension to tendons
(b) Casting of concrete

Cutting of tendon
(c) Transferring of prestress

Figure1-3.1 Stages of pre-tensioning
1.3.3 Advantages of Pre-tensioning
The relative advantages of pre-tensioning as compared to post-tensioning are as

follows.
Pre-tensioning is suitable for precast members produced in bulk.

In pre-tensioning large anchorage device is not present.
1.3.4 Disadvantages of Pre-tensioning
The relative disadvantages are as follows.

A prestressing bed is required for the pre-tensioning operation.
There is a waiting period in the prestressing bed, before the concrete attains
sufficient strength.
There should be good bond between concrete and steel over the transmission
length.
1.3.5 Devices
The essential devices for pre-tensioning are as follows.

Prestressing bed
End abutments
Shuttering / mould
Jack
Anchoring device
Harping device (optional)
Prestressing Bed, End Abutments and Mould

The following figure shows the devices.
End
abutment Mould
Jack
Prestressing bed
Anchoring
device
Figure1-3.2 Prestressing bed, end abutment and mould
An extension of the previous system is the Hoyer system. This system is generally
used for mass production. The end abutments are kept sufficient distance apart, and
several members are cast in a single line. The shuttering is provided at the sides and

between the members. This system is also called the Long Line Method. The
following figure is a schematic representation of the Hoyer system
A series of moulds
Prestressing bed
Figure 1-3.3 Schematic representation of Hoyer system
The end abutments have to be sufficiently stiff and have good foundations. This is
usually an expensive proposition, particularly when large prestressing forces are
required. The necessity of stiff and strong foundation can be bypassed by a simpler
solution which can also be a cheaper option. It is possible to avoid transmitting the
heavy loads to foundations, by adopting self-equilibrating systems. This is a common
solution in load-testing. Typically, this is done by means of a tension frame. The
following figure shows the basic components of a tension frame. The jack and the
specimen tend to push the end members. But the end members are kept in place by
members under tension such as high strength steel rods.
Loading
jack
Test
High specimen P
strength
steel rods
Plan or Elevation Free bodies

Figure 1-3.4 A tension frame
The frame that is generally adopted in a pre-tensioning system is called a stress bench.
The concrete mould is placed within the frame and the tendons are stretched and
anchored on the booms of the frame. The following figures show the components of a
stress bench.

Jack
Elevation Threaded rod
Mould Strands
Plan
Figure 1-3.5 Stress bench Self straining frame
The following figure shows the free body diagram by replacing the jacks with the applied
forces.
Load by jack
Plan
Tension in
strands
Figure 1-3.6 Free body diagram of stress bench
The following figure shows the stress bench after casting of the concrete.
Elevation
Plan
Figure 1-3.7 The stress bench after casting concrete

Jacks
The jacks are used to apply tension to the tendons. Hydraulic jacks are commonly used.
These jacks work on oil pressure generated by a pump. The principle behind the design
of jacks is Pascals law. The load applied by a jack is measured by the pressure
reading from a gauge attached to the oil inflow or by a separate load cell. The following
figure shows a double acting hydraulic jack with a load cell.
Figure 1-3.8 A double acting hydraulic jack with a load cell
Anchoring Devices
Anchoring devices are often made on the wedge and friction principle. In pre-tensioned
members, the tendons are to be held in tension during the casting and hardening of
concrete. Here simple and cheap quick-release grips are generally adopted. The
following figure provides some examples of anchoring devices.

Figure 1-3.9 Chuck assembly for anchoring tendons

(Reference: Lin, T. Y. and Burns, N. H.,
Design of Prestressed Concrete Structures)
Harping Devices
The tendons are frequently bent, except in cases of slabs-on-grade, poles, piles etc.
The tendons are bent (harped) in between the supports with a shallow sag as shown
below.
Harping point Hold up device
a) Before casting of concrete
b) After casting of concrete

Figure 1-3.10 Harping of tendons
The tendons are harped using special hold-down devices as shown in the following
figure.

Figure 1-3.11 Hold-down anchor for harping of tendons

(Reference: Nawy, E. G., Prestressed Concrete: A Fundamental Approach)
1.3.6 Manufacturing of Pre-tensioned Railway Sleepers
The following photos show the sequence of manufacturing of pre-tensioned railway

sleepers (Courtesy: The Concrete Products and Construction Company, COPCO,
Chennai). The steel strands are stretched in a stress bench that can be moved on
rollers. The stress bench can hold four moulds in a line. The anchoring device holds
the strands at one end of the stress bench. In the other end, two hydraulic jacks push a
plate where the strands are anchored. The movement of the rams of the jacks and the
oil pressure are monitored by a scale and gauges, respectively. Note that after the
extension of the rams, the gap between the end plate and the adjacent mould has
increased. This shows the stretching of the strands.

Meanwhile the coarse and fine aggregates are batched, mixed with cement, water and
additives in a concrete mixer. The stress bench is moved beneath the concrete mixer.
The concrete is poured through a hopper and the moulds are vibrated. After the
finishing of the surface, the stress bench is placed in a steam curing chamber for a few
hours till the concrete attains a minimum strength.
The stress bench is taken out from the chamber and the strands are cut. The sleepers
are removed from the moulds and stacked for curing in water. After the complete curing,
the sleepers are ready for dispatching.
(a) Travelling pre-tensioning stress bench

Wedge and
cylinder
assembly at
the dead end
(b) Anchoring of strands
Initial gap
End
plate
Hydraulic jack at
stretching end
(c) Stretching of strands

Final gap
Threaded
rod
Extension of ram
(d) Stretching of strands
Fine aggregate
Coarse aggregate
(e) Material storage

Automated
batching
by weight
(f) Batching of materials
Hopper below
concrete mixer
(g) Pouring of concrete

(h) Concrete after vibration of mould
(i) Steam curing chamber

(j) Cutting of strands
(k) Demoulding of sleeper

(l) Stacking of sleeper
(m) Water curing

(n) Storage and dispatching of sleepers

Figure 1-3.12 Manufacturing of pre-tensioned railway sleepers

1.4 Post-tensioning Systems and Devices

This section covers the following topics
Introduction
Stages of Post-tensioning
Advantages of Post-tensioning
Disadvantages of Post-tensioning
Devices
Manufacturing of a Post-tensioned Bridge Girder
1.4.1 Introduction
Prestressing systems have developed over the years and various companies have
patented their products. Detailed information of the systems is given in the product
catalogues and brochures published by companies. There are general guidelines of
prestressing in Section 12 of IS 1343: 1980. The information given in this section is
introductory in nature, with emphasis on the basic concepts of the systems.
The prestressing systems and devices are described for the two types of prestressing,
pre-tensioning and post-tensioning, separately. This section covers post-tensioning.
Section 1.3, Pre-tensioning Systems and Devices, covers pre-tensioning. In post-
tensioning, the tension is applied to the tendons after hardening of the concrete. The
stages of post-tensioning are described next.
1.4.2 Stages of Post-tensioning
In post-tensioning systems, the ducts for the tendons (or strands) are placed along with
the reinforcement before the casting of concrete. The tendons are placed in the ducts
after the casting of concrete. The duct prevents contact between concrete and the
tendons during the tensioning operation.
Unlike pre-tensioning, the tendons are pulled with the reaction acting against the
hardened concrete.

If the ducts are filled with grout, then it is known as bonded post-tensioning. The grout
is a neat cement paste or a sand-cement mortar containing suitable admixture. The
grouting operation is discussed later in the section. The properties of grout are
discussed in Section 1.6, Concrete (Part-II).
In unbonded post-tensioning, as the name suggests, the ducts are never grouted and
the tendon is held in tension solely by the end anchorages. The following sketch shows
a schematic representation of a grouted post-tensioned member. The profile of the duct
depends on the support conditions. For a simply supported member, the duct has a
sagging profile between the ends. For a continuous member, the duct sags in the span
and hogs over the support.
Figure 1-4.1 Post-tensioning (Reference: VSL International Ltd.)
Among the following figures, the first photograph shows the placement of ducts in a box
girder of a simply supported bridge. The second photograph shows the end of the box
girder after the post-tensioning of some tendons.
Figure 1-4.2 Post-tensioning ducts in a box girder


Figure 1-4.3 Post-tensioning of a box girder

The various stages of the post-tensioning operation are summarised as follows.

1) Casting of concrete.
2) Placement of the tendons.
3) Placement of the anchorage block and jack.
4) Applying tension to the tendons.
5) Seating of the wedges.
6) Cutting of the tendons.
The stages are shown schematically in the following figures. After anchoring a tendon
at one end, the tension is applied at the other end by a jack. The tensioning of tendons
and pre-compression of concrete occur simultaneously. A system of self-equilibrating
forces develops after the stretching of the tendons.

Duct
Casting bed Side view

(a) Casting of concrete
Jack
(b) Tensioning of tendons
Anchor
(c) Anchoring the tendon at the stretching end

Figure 1-4.4 Stages of post-tensioning (shown in elevation)
1.4.3 Advantages of Post-tensioning
The relative advantages of post-tensioning as compared to pre-tensioning are as

follows.
Post-tensioning is suitable for heavy cast-in-place members.
The waiting period in the casting bed is less.
The transfer of prestress is independent of transmission length.
1.4.4 Disadvantage of Post-tensioning
The relative disadvantage of post-tensioning as compared to pre-tensioning is the

requirement of anchorage device and grouting equipment.
1.4.5 Devices
The essential devices for post-tensioning are as follows.

1) Casting bed
2) Mould/Shuttering
3) Ducts

4) Anchoring devices
5) Jacks
6) Couplers (optional)
7) Grouting equipment (optional).
Casting Bed, Mould and Ducts

The following figure shows the devices.
Mould
Duct
Casting bed
Figure 1-4.5 Casting bed, mould and duct
Anchoring Devices
In post-tensioned members the anchoring devices transfer the prestress to the concrete.
The devices are based on the following principles of anchoring the tendons.
1) Wedge action
2) Direct bearing
3) Looping the wires
Wedge action
The anchoring device based on wedge action consists of an anchorage block and
wedges. The strands are held by frictional grip of the wedges in the anchorage block.
Some examples of systems based on the wedge-action are Freyssinet, Gifford-Udall,
Anderson and Magnel-Blaton anchorages. The following figures show some patented
anchoring devices.

Figure 1-4.6 Freyssinet T system anchorage cones

(Reference: Lin, T. Y. and Burns, N. H., Design of Prestressed Concrete Structures)
Figure 1-4.7 Anchoring devices

(Reference: Collins, M. P. and Mitchell, D., Prestressed Concrete Structures)
Figure 1-4.8 Anchoring devices (Reference: VSL International Ltd)

Direct bearing
The rivet or bolt heads or button heads formed at the end of the wires directly bear
against a block. The B.B.R.V post-tensioning system and the Prescon system are
based on this principle. The following figure shows the anchoring by direct bearing.
Figure 1-4.9 Anchoring with button heads

Looping the wires

The Baur-Leonhardt system, Leoba system and also the Dwidag single-bar anchorage
system, work on this principle where the wires are looped around the concrete. The
wires are looped to make a bulb. The following photo shows the anchorage by looping
of the wires in a post-tensioned slab.
Figure 1-4.10 Anchorage by looping the wires in a slab

(Courtesy : VSL India Pvt. Ltd.)
The anchoring devices are tested to calculate their strength. The following photo shows
the testing of an anchorage block.

Figure 1-4.11 Testing of an anchorage device
Sequence of Anchoring
The following figures show the sequence of stressing and anchoring the strands. The
photo of an anchoring device is also provided.

Figure 1-4.12 Sequence of anchoring

Figure 1-4.13 Final form of an anchoring device

(Reference: VSL International Ltd)
Jacks
The working of a jack and measuring the load were discussed in Section 1.3, Pre-
tensioning Systems and Devices. The following figure shows an extruded sketch of the
anchoring devices.

Figure 1-4.14 Jacking and anchoring with wedges

Couplers
The couplers are used to connect strands or bars. They are located at the junction of
the members, for example at or near columns in post-tensioned slabs, on piers in post-
tensioned bridge decks.
The couplers are tested to transmit the full capacity of the strands or bars. A few types
of couplers are shown.
Figure 1-4.15 Coupler for strands

(Reference: VSL International Ltd)

Figure 1-4.16 Couplers for strands

(Reference: Dywidag Systems International)
Figure 1-4.17 Couplers for strands


Grouting
Grouting can be defined as the filling of duct, with a material that provides an anti-
corrosive alkaline environment to the prestressing steel and also a strong bond between
the tendon and the surrounding grout.
The major part of grout comprises of water and cement, with a water-to-cement ratio of
about 0.5, together with some water-reducing admixtures, expansion agent and
pozzolans. The properties of grout are discussed in Section 1.6, Concrete (Part-II).
The following figure shows a grouting equipment, where the ingredients are mixed and
the grout is pumped.
Figure 1-4.18 Grouting equipment

(Reference: Williams Form Engineering Corp.)
1.4.6 Manufacturing of Post-tensioned Bridge Girders
The following photographs show some steps in the manufacturing of a post-tensioned I-

girder for a bridge (Courtesy: Larsen & Toubro). The first photo shows the fabricated
steel reinforcement with the ducts for the tendons placed inside. Note the parabolic
profiles of the duct for the simply supported girder. After the concrete is cast and cured
to gain sufficient strength, the tendons are passed through the ducts, as shown in the
second photo. The tendons are anchored at one end and stretched at the other end by
a hydraulic jack. This can be observed from the third photo.

(a) Fabrication of reinforcement
(b) Placement of tendons

(c) Stretching and anchoring of tendons

Figure 1-4.19 Manufacturing of a post-tensioned bridge I-girder
The following photos show the construction of post-tensioned box girders for a bridge
(Courtesy: Cochin Port Trust). The first photo shows the fabricated steel reinforcement
with the ducts for the tendons placed inside. The top flange will be constructed later.
The second photo shows the formwork in the pre-casting yard. The formwork for the
inner sides of the webs and the flanges is yet to be placed. In the third photo a girder is
being post-tensioned after adequate curing. The next photo shows a crane on a barge
that transports a girder to the bridge site. The completed bridge can be seen in the last
photo.
(a) Reinforcement cage for box girder

(b) Formwork for box girder
(c) Post-tensioning of box girder

(d) Transporting of box girder
(e) Completed bridge

Figure 1-4.20 Manufacturing of post-tensioned bridge box girders

1.5 Concrete (Part I)

Constituents of Concrete
Properties of Hardened Concrete (Part I)
1.5.1 Constituents of Concrete
Introduction
Concrete is a composite material composed of gravels or crushed stones (coarse
aggregate), sand (fine aggregate) and hydrated cement (binder). It is expected that the
student of this course is familiar with the basics of concrete technology. Only the
information pertinent to prestressed concrete design is presented here.
The following figure shows a petrographic section of concrete. Note the scattered
coarse aggregates and the matrix surrounding them. The matrix consists of sand,
hydrated cement and tiny voids.
Figure 1-5.1 Petrographic section of hardened concrete

(Reference: Portland Cement Association, Design and Control of Concrete Mixtures)
Aggregate
The coarse aggregate are granular materials obtained from rocks and crushed stones.
They may be also obtained from synthetic material like slag, shale, fly ash and clay for
use in light-weight concrete.

The sand obtained from river beds or quarries is used as fine aggregate. The fine
aggregate along with the hydrated cement paste fill the space between the coarse
aggregate.
The important properties of aggregate are as follows.

1) Shape and texture
2) Size gradation
3) Moisture content
4) Specific gravity
5) Unit weight
6) Durability and absence of deleterious materials.
The requirements of aggregate is covered in Section 4.2 of IS:1343 - 1980.
The nominal maximum coarse aggregate size is limited by the lowest of the following
quantities.
1) 1/4 times the minimum thickness of the member
2) Spacing between the tendons/strands minus 5 mm
3) 40 mm.
The deleterious substances that should be limited in aggregate are clay lumps, wood,
coal, chert, silt, rock dust (material finer than 75 microns), organic material, unsound
and friable particles.
Cement
In present day concrete, cement is a mixture of lime stone and clay heated in a kiln to
1400 - 1600C. The types of cement permitted by IS:1343 - 1980 (Clause 4.1) for
prestressed applications are the following. The information is revised as per IS:456 -
2000, Plain and Reinforced Concrete Code of Practice.
1) Ordinary Portland cement confirming to IS:269 - 1989, Ordinary Portland Cement,
33 Grade Specification.
2) Portland slag cement confirming to IS:455 - 1989, Portland Slag Cement
Specification, but with not more than 50% slag content.
3) Rapid-hardening Portland cement confirming to IS:8041 - 1990, Rapid Hardening
Portland Cement Specification.

Water
The water should satisfy the requirements of Section 5.4 of IS:456 - 2000.
Water used for mixing and curing shall be clean and free from injurious amounts of oils,
acids, alkalis, salts, sugar, organic materials or other substances that may be
deleterious to concrete and steel.
Admixtures
IS:1343 - 1980 allows to use admixtures that conform to IS:9103 - 1999, Concrete
Admixtures Specification. The admixtures can be broadly divided into two types:
chemical admixtures and mineral admixtures. The common chemical admixtures are as
follows.
1) Air-entraining admixtures
2) Water reducing admixtures
3) Set retarding admixtures
4) Set accelerating admixtures
5) Water reducing and set retarding admixtures
6) Water reducing and set accelerating admixtures.
The common mineral admixtures are as follows.

1) Fly ash
2) Ground granulated blast-furnace slag
3) Silica fumes
4) Rice husk ash
5) Metakoline
These are cementitious and pozzolanic materials.
1.5.2 Properties of Hardened Concrete (Part I)
The concrete in prestressed applications has to be of good quality. It requires the

following attributes.
1) High strength with low water-to-cement ratio
2) Durability with low permeability, minimum cement content and proper mixing,
compaction and curing

3) Minimum shrinkage and creep by limiting the cement content.

The following topics are discussed.
1) Strength of concrete
2) Stiffness of concrete
3) Durability of concrete
4) High performance concrete
5) Allowable stresses in concrete.
Strength of Concrete
The following sections describe the properties with reference to IS:1343 - 1980. The
strength of concrete is required to calculate the strength of the members. For
prestressed concrete applications, high strength concrete is required for the following
reasons.
1) To sustain the high stresses at anchorage regions.
2) To have higher resistance in compression, tension, shear and bond.
3) To have higher stiffness for reduced deflection.
4) To have reduced shrinkage cracks.
Compressive Strength
The compressive strength of concrete is given in terms of the characteristic
compressive strength of 150 mm size cubes tested at 28 days (fck). The characteristic
strength is defined as the strength of the concrete below which not more than 5% of the
test results are expected to fall. This concept assumes a normal distribution of the
strengths of the samples of concrete.
The following sketch shows an idealised distribution of the values of compressive

strength for a sizeable number of test cubes. The horizontal axis represents the values
of compressive strength. The vertical axis represents the number of test samples for a
particular compressive strength. This is also termed as frequency. The average of the
values of compressive strength (mean strength) is represented as fcm. The characteristic
strength (fck) is the value in the x-axis below which 5% of the total area under the curve
falls. The value of fck is lower than fcm by 1.65, where is the standard deviation of the
normal distribution.

Frequency
1.65
5% area fck fcm

28 day cube compressive strength
Figure 1-5.2 Idealised normal distribution of concrete strength
(Reference: Pillai, S. U., and Menon, D., Reinforced Concrete Design)
The sampling and strength test of concrete are as per Section 15 of IS:1343 - 1980.
The grades of concrete are explained in Table 1 of the Code.
The minimum grades of concrete for prestressed applications are as follows.

30 MPa for post-tensioned members
40 MPa for pre-tensioned members.
The maximum grade of concrete is 60 MPa.
Since at the time of publication of IS:1343 in 1980, the properties of higher strength
concrete were not adequately documented, a limit was imposed on the maximum
strength. It is expected that higher strength concrete may be used after proper testing.
The increase in strength with age as given in IS:1343 - 1980, is not observed in present
day concrete that gains substantial strength in 28 days. Hence, the age factor given in
Clause 5.2.1 should not be used. It has been removed from IS:456 - 2000.
Tensile Strength
The tensile strength of concrete can be expressed as follows.
1) Flexural tensile strength: It is measured by testing beams under 2 point loading
(also called 4 point loading including the reactions).
2) Splitting tensile strength: It is measured by testing cylinders under diametral
compression.

3) Direct tensile strength: It is measured by testing rectangular specimens under

direct tension.
In absence of test results, the Code recommends to use an estimate of the flexural
tensile strength from the compressive strength by the following equation.
fcr = 0.7 fck (1-5.1)
Here,
fcr = flexural tensile strength in N/mm2
fck = characteristic compressive strength of cubes in N/mm2.
Stiffness of Concrete
The stiffness of concrete is required to estimate the deflection of members. The
stiffness is given by the modulus of elasticity. For a non-linear stress (fc) versus strain
(c) behaviour of concrete the modulus can be initial, tangential or secant modulus.
IS:1343 - 1980 recommends a secant modulus at a stress level of about 0.3fck. The
modulus is expressed in terms of the characteristic compressive strength and not the
design compressive strength. The following figure shows the secant modulus in the
compressive stress-strain curve for concrete.
fc
fc
fck
Ec
c
Figure 1-5.3 a) Concrete cube under compression, b) Compressive stress-strain
curve for concrete
The modulus of elasticity for short term loading (neglecting the effect of creep) is given
by the following equation.
Ec = 5000 fck (1-5.2)
Here,
Ec = short-term static modulus of elasticity in N/mm2
fck = characteristic compressive strength of cubes in N/mm2.

The above expression is updated as per IS:456 - 2000.
Durability of Concrete
The durability of concrete is of vital importance regarding the life cycle cost of a
structure. The life cycle cost includes not only the initial cost of the materials and labour,
but also the cost of maintenance and repair.
In recent years emphasis has been laid on the durability issues of concrete. This is
reflected in the enhanced section on durability (Section 8) in IS:456 - 2000. It is
expected that the revised version of IS:1343 will also have similar importance on
durability.
The durability of concrete is defined as its ability to resist weathering action, chemical
attack, abrasion, or any other process of deterioration. The common durability
problems in concrete are as follows.
1) Sulphate and other chemical attacks of concrete.
2) Alkali-aggregate reaction.
3) Freezing and thawing damage in cold regions.
4) Corrosion of steel bars or tendons.
The durability of concrete is intrinsically related to its water tightness or permeability.

Hence, the concrete should have low permeability and there should be adequate cover
to reinforcing bars. The selection of proper materials and good quality control are
essential for durability of concrete.
The durability is addressed in IS:1343 - 1980 in Section 7. In Appendix A there are

guidelines on durability. Table 9 specifies the maximum water-to-cement (w-c) ratio
and the minimum cement content for different exposure conditions. The values for
moderate exposure condition are reproduced below.
Table 1-5.1 Maximum water-to-cement (w-c) ratio and the minimum cement content
for moderate exposure conditions (IS:1343 - 1980).
Min. cement content : 300 kg per m3 of concrete
Max w-c ratio* : 0.50
(*The value is updated as per Table 5 of IS:456 - 2000.)

Table 10 provides the values for the above quantities for concrete exposed to sulphate
attack.
To limit the creep and shrinkage, IS:1343 - 1980 specifies a maximum cement content
of 530 kg per m3 of concrete (Clause 8.1.1).
High Performance Concrete

With the advancement of concrete technology, high performance concrete is getting
popular in prestressed applications. The attributes of high performance concrete are as
follows.
1) High strength
2) Minimum shrinkage and creep
3) High durability
4) Easy to cast
5) Cost effective.
Traditionally high performance concrete implied high strength concrete with higher
cement content and low water-to-cement ratio. But higher cement content leads to
autogenous and plastic shrinkage cracking and thermal cracking. At present durability
is also given importance along with strength.
Some special types of high performance concrete are as follows.

1) High strength concrete
2) High workability concrete
3) Self-compacting concrete
4) Reactive powder concrete
5) High volume fly ash concrete
6) Fibre reinforced concrete
In a post-tensioned member, the concrete next to the anchorage blocks (referred to as

end block) is subjected to high stress concentration. The type of concrete at the end
blocks may be different from that at the rest of the member. Fibre reinforced concrete is
used to check the cracking due to the bursting forces.
The following photo shows that the end blocks were cast separately with high strength
concrete.

Figure 1-5.4 End-blocks in a bridge deck

Allowable Stresses in Concrete

The allowable stresses are used to analyse and design members under service loads.
IS:1343 - 1980 specifies the maximum allowable compressive stresses for different
grades of concrete under different loading conditions in Section 22.8.
Allowable Compressive Stresses under Flexure

The following sketch shows the variation of allowable compressive stresses for different
grades of concrete at transfer. The cube strength at transfer is denoted as fci.
0.54fci 0.51fci
0.37fci 0.44fci
M30 M60 M40 M60

Post-tension Pre-tension
Figure 1-5.5 Variation of allowable compressive stresses at transfer
The following sketch shows the variation of allowable compressive stresses for different
grades of concrete at service loads.

0.41fck Zone I
0.35fck
0.34fck
Zone II 0.27fck
M 30 M 60
Figure 1-5.6 Variation of allowable compressive stresses at service loads
Here, Zone I represents the locations where the compressive stresses are not likely to
increase. Zone II represents the locations where the compressive stresses are likely to
increase, such as due to transient loads from vehicles in bridge decks.
Allowable Compressive Stresses under Direct Compression

For direct compression, except in the parts immediately behind the anchorage, the
maximum stress is equal to 0.8 times the maximum compressive stress under flexure.
Allowable Tensile Stresses under Flexure

The prestressed members are classified into three different types based on the
allowable tensile stresses. The amount of prestressing varies in the three types. The
allowable tensile stresses for the three types of members are specified in Section 22.7.
The values are reproduced below.
Table 1-5.2 Allowable tensile stresses (IS:1343 - 1980)
Type 1 No tensile stress
3 N/mm2.
Type 2
This value can be increased to 4.5 N/mm2 for temporary loads.
Type 3 Table 8 provides hypothetical values of allowable tensile stresses.
The purpose of providing hypothetical values is to use the elastic analysis method for
Type 3 members even after cracking of concrete.

1.6 Concrete (Part II)

Properties of Hardened Concrete (Part II)
Properties of Grout
Codal Provisions of Concrete
1.6.1 Properties of Hardened Concrete (Part II)
The properties that are discussed are as follows.

1) Stress-strain curves for concrete
2) Creep of concrete
3) Shrinkage of concrete
Stress-strain Curves for Concrete

Curve under uniaxial compression
The stress versus strain behaviour of concrete under uniaxial compression is initially
linear (stress is proportional to strain) and elastic (strain is recovered at unloading). With
the generation of micro-cracks, the behaviour becomes nonlinear and inelastic. After the
specimen reaches the peak stress, the resisting stress decreases with increase in strain.
IS:1343 - 1980 recommends a parabolic characteristic stress-strain curve, proposed by

Hognestad, for concrete under uniaxial compression (Figure 3 in the Code).
fc
fc
fck
0 cu c
Figure 1-6.1 a) Concrete cube under compression, b) Design stress-strain curve for
concrete under compression due to flexure

The equation for the design curve under compression due to flexure is as follows.
For c 0
2
fck = fck 2 c - c (1-6.1)
0 0
For c < c cu
fc = fck (1-6.2)
Here,
fc = compressive stress
fck = characteristic compressive strength of cubes
c = compressive strain
0 = strain corresponding to fck = 0.002
cu = ultimate compressive strain = 0.0035
For concrete under compression due to axial load, the ultimate strain is restricted to
0.002. From the characteristic curve, the design curve is defined by multiplying the
stress with a size factor of 0.67 and dividing the stress by a material safety factor of m =
1.5. The design curve is used in the calculation of ultimate strength. The following
sketch shows the two curves.
fc
fck Characteristic curve
0.447 fck Design curve
0 cu c
Figure 1-6.2 Stress-strain curves for concrete under compression due to flexure
In the calculation of deflection at service loads, a linear stress-strain curve is assumed

up to the allowable stress. This curve is given by the following equation.
fc = Ecc (1-6.3)
Note that, the size factor and the material safety factor are not used in the elastic
modulus Ec.

For high strength concrete (say M100 grade of concrete and above) under uniaxial
compression, the ascending and descending branches are steep.
fc
fck
Eci
Es
0 c
Figure 1-6.3 Stress-strain curves for high strength concrete under compression
The equation proposed by Thorenfeldt, Tomaxzewicz and Jensen is appropriate for high
strength concrete.

n c
fc = fck 0 (1-6.4)
nk
c
n - 1+
0
The variables in the previous equation are as follows.

fc = compressive stress
fck = characteristic compressive strength of cubes in N/mm2
c = compressive strain
0 = strain corresponding to fck
k = 1 for c 0
= 0.67 + (fck / 77.5) for c > 0. The value of k should be greater than 1.
n = Eci / (Eci Es)
Eci = initial modulus
Es = secant modulus at fck = fck / 0.
The previous equation is applicable for both the ascending and descending branches of
the curve. Also, the parameter k models the slope of the descending branch, which
increases with the characteristic strength fck. To be precise, the value of 0 can be
considered to vary with the compressive strength of concrete.

Curve under uniaxial tension

The stress versus strain behaviour of concrete under uniaxial tension is linear elastic
initially. Close to cracking nonlinear behaviour is observed.
fc
fc
(a) (b)
Figure 1-6.4 a) Concrete panel under tension, b) Stress-strain curve for concrete
under tension
In calculation of deflections of flexural members at service loads, the nonlinearity is

neglected and a linear elastic behaviour fc = Ecc is assumed. In the analysis of ultimate
strength, the tensile strength of concrete is usually neglected.
Creep of Concrete
Creep of concrete is defined as the increase in deformation with time under constant
load. Due to the creep of concrete, the prestress in the tendon is reduced with time.
Hence, the study of creep is important in prestressed concrete to calculate the loss in
prestress.
The creep occurs due to two causes.

1. Rearrangement of hydrated cement paste (especially the layered products)
2. Expulsion of water from voids under load
If a concrete specimen is subjected to slow compressive loading, the stress versus

strain curve is elongated along the strain axis as compared to the curve for fast loading.
This can be explained in terms of creep. If the load is sustained at a level, the increase
in strain due to creep will lead to a shift from the fast loading curve to the slow loading
curve (Figure 1-6.5).

fc Fast loading
Slow loading
Effect of creep
c
Figure 1-6.5 Stress-strain curves for concrete under compression
Creep is quantified in terms of the strain that occurs in addition to the elastic strain due
to the applied loads. If the applied loads are close to the service loads, the creep strain
increases at a decreasing rate with time. The ultimate creep strain is found to be
proportional to the elastic strain. The ratio of the ultimate creep strain to the elastic
strain is called the creep coefficient .
For stress in concrete less than about one-third of the characteristic strength, the
ultimate creep strain is given as follows.
cr,ult = el (1-6.5)
The variation of strain with time, under constant axial compressive stress, is
represented in the following figure.
cr, ult = ultimate creep strain

strain
el = elastic strain
Time (linear scale)

Figure 1-6.6 Variation of strain with time for concrete under compression
If the load is removed, the elastic strain is immediately recovered. However the
recovered elastic strain is less than the initial elastic strain, as the elastic modulus
increases with age.
There is reduction of strain due to creep recovery which is less than the creep strain.
There is some residual strain which cannot be recovered (Figure 1-6.7).

Elastic recovery
Creep recovery
strain
Unloading
Residual strain
Time (linear scale)

Figure 1-6.7 Variation of strain with time showing the effect of unloading
The creep strain depends on several factors. It increases with the increase in the
following variables.
1) Cement content (cement paste to aggregate ratio)
2) Water-to-cement ratio
3) Air entrainment
4) Ambient temperature.
The creep strain decreases with the increase in the following variables.
1) Age of concrete at the time of loading.
2) Relative humidity
3) Volume to surface area ratio.
The creep strain also depends on the type of aggregate.
IS:1343 - 1980 gives guidelines to estimate the ultimate creep strain in Section 5.2.5. It
is a simplified estimate where only one factor has been considered. The factor is age of
loading of the prestressed concrete structure. The creep coefficient is provided for
three values of age of loading.
Table 1-6.1 Creep coefficient for three values of age of loading
Age of Loading Creep Coefficient
7 days 2.2
28 days 1.6
1 year 1.1

It can be observed that if the structure is loaded at 7 days, the creep coefficient is 2.2.
This means that the creep strain is 2.2 times the elastic strain. Thus, the total strain is
more than thrice the elastic strain. Hence, it is necessary to study the effect of creep in
the loss of prestress and deflection of prestressed flexural members. Even if the
structure is loaded at 28 days, the creep strain is substantial. This implies higher loss of
prestress and higher deflection.
Curing the concrete adequately and delaying the application of load provide long term
benefits with regards to durability, loss of prestress and deflection.
In special situations detailed calculations may be necessary to monitor creep strain with
time. Specialised literature or international codes can provide guidelines for such
calculations.
Shrinkage of Concrete
Shrinkage of concrete is defined as the contraction due to loss of moisture. The study of
shrinkage is also important in prestressed concrete to calculate the loss in prestress.
The shrinkage occurs due to two causes.

1. Loss of water from voids
2. Reduction of volume during carbonation
The following figure shows the variation of shrinkage strain with time. Here, t0 is the time
at commencement of drying. The shrinkage strain increases at a decreasing rate with
time. The ultimate shrinkage strain (sh) is estimated to calculate the loss in prestress.
sh
Shrinkage strain
t0 Time (linear scale)

Figure 1-6.8 Variation of shrinkage strain with time

Like creep, shrinkage also depends on several factors. The shrinkage strain increases
with the increase in the following variables.
1) Ambient temperature
2) Temperature gradient in the members
3) Water-to-cement ratio
4) Cement content.
The shrinkage strain decreases with the increase in the following variables.
1) Age of concrete at commencement of drying
2) Relative humidity
3) Volume to surface area ratio.
The shrinkage strain also depends on the type of aggregate.
IS:1343 - 1980 gives guidelines to estimate the shrinkage strain in Section 5.2.4. It is a
simplified estimate of the ultimate shrinkage strain (sh).
For pre-tension
sh = 0.0003 (1-6.6)
For post-tension
0.0002 (1-6.7)
sh =
log10 (t + 2 )
Here, t is the age at transfer in days. Note that for post-tension, t is the age at transfer
in days which approximates the curing time.
It can be observed that with increasing age at transfer, the shrinkage strain reduces. As
mentioned before, curing the concrete adequately and delaying the application of load
provide long term benefits with regards to durability and loss of prestress.
In special situations detailed calculations may be necessary to monitor shrinkage strain

with time. Specialised literature or international codes can provide guidelines for such
calculations.

1.6.2 Properties of Grout
Grout is a mixture of water, cement and optional materials like sand, water-reducing
admixtures, expansion agent and pozzolans. The water-to-cement ratio is around 0.5.
Fine sand is used to avoid segregation.
The desirable properties of grout are as follows.

1) Fluidity
2) Minimum bleeding and segregation
3) Low shrinkage
4) Adequate strength after hardening
5) No detrimental compounds
6) Durable.
IS:1343 - 1980 specifies the properties of grout in Sections 12.3.1 and Section 12.3.2.
The following specifications are important.
1) The sand should pass 150 m Indian Standard sieve.
2) The compressive strength of 100 mm cubes of the grout shall not be less than 17
N/mm2 at 7 days.

1.6.5 Codal Provisions of Concrete
The following topics are covered in IS:1343 - 1980 under the respective sections. These
provisions are not duplicated here.
Table 1-6.2 Topics and sections
Workability of concrete Section 6
Concrete mix proportioning Section 8
Production and control of concrete Section 9
Formwork Section 10
Transporting, placing, compacting Section 13
Concrete under special conditions Section 14
Sampling and strength test of concrete Section 15
Acceptance criteria Section 16
Inspection and testing of structures Section 17

1.7 Prestressing Steel

Forms of Prestressing Steel
Types of Prestressing Steel
Properties of Prestressing Steel
Codal Provisions of Steel
1.7.1 Forms of Prestressing Steel
The development of prestressed concrete was influenced by the invention of high

strength steel. It is an alloy of iron, carbon, manganese and optional materials. The
following material describes the types and properties of prestressing steel.
In addition to prestressing steel, conventional non-prestressed reinforcement is used for

flexural capacity (optional), shear capacity, temperature and shrinkage requirements.
The properties of steel for non-prestressed reinforcement are not covered in this section.
It is expected that the student of this course is familiar with the conventional
reinforcement.
Wires
A prestressing wire is a single unit made of steel. The nominal diameters of the wires
are 2.5, 3.0, 4.0, 5.0, 7.0 and 8.0 mm. The different types of wires are as follows.
1) Plain wire: No indentations on the surface.
2) Indented wire: There are circular or elliptical indentations on the surface.
Strands
A few wires are spun together in a helical form to form a prestressing strand. The
different types of strands are as follows.
1) Two-wire strand: Two wires are spun together to form the strand.
2) Three-wire strand: Three wires are spun together to form the strand.
3) Seven-wire strand: In this type of strand, six wires are spun around a central wire.
The central wire is larger than the other wires.

Tendons
A group of strands or wires are placed together to form a prestressing tendon. The
tendons are used in post-tensioned members. The following figure shows the cross
section of a typical tendon. The strands are placed in a duct which may be filled with
grout after the post-tensioning operation is completed (Figure 1-7.1).
Grout
Duct
Figure 1-7.1 Cross-section of a typical tendon
Cables
A group of tendons form a prestressing cable. The cables are used in bridges.
Bars
A tendon can be made up of a single steel bar. The diameter of a bar is much larger
than that of a wire. Bars are available in the following sizes: 10, 12, 16, 20, 22, 25, 28
and 32 mm.
The following figure shows the different forms of prestressing steel.
Prestressing wires,
Reinforcing bars strands and bars
Figure 1-7.2 Forms of reinforcing and prestressing steel

1.7.2 Types of Prestressing Steel
The steel is treated to achieve the desired properties. The following are the treatment
processes.
Cold working (cold drawing)

The cold working is done by rolling the bars through a series of dyes. It re-aligns the
crystals and increases the strength.
Stress relieving
The stress relieving is done by heating the strand to about 350 C and cooling slowly.
This reduces the plastic deformation of the steel after the onset of yielding.
Strain tempering for low relaxation

This process is done by heating the strand to about 350 C while it is under tension.
This also improves the stress-strain behaviour of the steel by reducing the plastic
deformation after the onset of yielding. In addition, the relaxation is reduced. The
relaxation is described later.
IS:1343 - 1980 specifies the material properties of steel in Section 4.5. The following
types of steel are allowed.
1) Plain cold drawn stress relieved wire conforming to IS:1785, Part 1, Specification
for Plain Hard Drawn Steel Wire for Prestressed Concrete, Part I Cold Drawn
Stress Relieved Wire.
2) Plain as-drawn wire conforming to IS:1785, Part 2, Specification for Plain Hard
Drawn Steel Wire for Prestressed Concrete, Part II As Drawn Wire.
3) Indented cold drawn wire conforming to IS:6003, Specification for Indented Wire
for Prestressed Concrete.
4) High tensile steel bar conforming to IS:2090, Specification for High Tensile Steel
Bars used in Prestressed Concrete.
5) Uncoated stress relieved strand conforming to IS:6006. Specification for
Uncoated Stress Relieved Strand for Prestressed Concrete.

1.7.3 Properties of Prestressing Steel
The steel in prestressed applications has to be of good quality. It requires the following
attributes.
1) High strength
2) Adequate ductility
3) Bendability, which is required at the harping points and near the anchorage
4) High bond, required for pre-tensioned members
5) Low relaxation to reduce losses
6) Minimum corrosion.
Strength of Prestressing Steel

The tensile strength of prestressing steel is given in terms of the characteristic tensile
strength (fpk).
The characteristic strength is defined as the ultimate tensile strength of the coupon
specimens below which not more than 5% of the test results are expected to fall.
The ultimate tensile strength of a coupon specimen is determined by a testing machine

according to IS:1521 - 1972, Method for Tensile Testing of Steel Wire. The following
figure shows a test setup.
Wedge grips
Coupon specimen
Extensometer
(a) Test set-up

(b) Failure of a strand

Figure 1-7.3 Testing of tensile strength of prestressing strand
The minimum tensile strengths for different types of wires as specified by the codes are
reproduced.
Table 1-7.1 Cold Drawn Stress-Relieved Wires (IS: 1785 Part 1)
Nominal Diameter (mm) 2.50 3.00 4.00 5.00 7.00 8.00
Minimum Tensile Strength fpk 2010 1865 1715 1570 1470 1375
2
(N/mm )
The proof stress (defined later) should not be less than 85% of the specified tensile
strength.
Table 1-7.2 As-Drawn wire (IS: 1785 Part 2)
Nominal Diameter (mm) 3.00 4.00 5.00
Minimum Tensile Strength fpk (N/mm2) 1765 1715 1570
The proof stress should not be less than 75% of the specified tensile strength.
Table 1-7.3 Indented wire (IS: 6003)
Nominal Diameter (mm) 3.00 4.00 5.00
Minimum Tensile Strength fpk (N/mm2) 1865 1715 1570
The proof stress should not be less than 85% of the specified tensile strength.
For high tensile steel bars (IS: 2090), the minimum tensile strength is 980 N/mm2. The
proof stress should not be less than 80% of the specified tensile strength.

Stiffness of Prestressing Steel

The stiffness of prestressing steel is given by the initial modulus of elasticity. The
modulus of elasticity depends on the form of prestressing steel (wires or strands or
bars).
IS:1343 - 1980 provides the following guidelines which can be used in absence of test
data.
Table 1-7.4 Modulus of elasticity (IS: 1343 - 1980)
Type of steel Modulus of elasticity
Cold-drawn wires 210 kN/mm2
High tensile steel bars 200 kN/mm2
Strands 195 kN/mm2
Allowable Stress in Prestressing Steel

As per Clause 18.5.1, the maximum tensile stress during prestressing (fpi) shall not
exceed 80% of the characteristic strength.
fpi 0.8fpk (1-7.1)
There is no upper limit for the stress at transfer (after short term losses) or for the
effective prestress (after long term losses).
Stress-Strain Curves for Prestressing Steel

The stress versus strain behaviour of prestressing steel under uniaxial tension is initially
linear (stress is proportional to strain) and elastic (strain is recovered at unloading).
Beyond about 70% of the ultimate strength the behaviour becomes nonlinear and
inelastic. There is no defined yield point.
The yield point is defined in terms of the proof stress or a specified yield strain. IS:1343
- 1980 recommends the yield point at 0.2% proof stress. This stress corresponds to an
inelastic strain of 0.002. This is shown in the following figure.

fp
Proof
stress
0.002 p
Figure 1-7.4 Proof stress corresponds to inelastic strain of 0.002
The characteristic stress-strain curves are given in Figure 5 of IS:1343 - 1980. The
stress corresponding to a strain can be found out by using these curves as shown next.
fp fp
0.95fpk 0.95fpk
0.9fpk 0.85fpk
0.002 0.005 p 0.002 0.005 p

Stress relieved wires, As-drawn wires
strands and bars
Figure 1-7.5 Characteristic stress-strain curves for prestressing steel
(Figure 5, IS:1343 - 1980)
The stress-strain curves are influenced by the treatment processes. The following figure
shows the variation in the 0.2% proof stress for wires under different treatment
processes.
low relaxation
fp
stress relieved
as-drawn
p
Figure 1-7.6 Variation in the 0.2% proof stress for wires under different treatment
processes

The design stress-strain curves are calculated by dividing the stress beyond 0.8fpk by a
material safety factor m =1.15. The following figure shows the characteristic and design
stress-strain curves.
fp
Characteristic curve
0.8fpk
Design curve
p
Figure 1-7.7 Characteristic and design stress-strain curves for
prestressing steel
Relaxation of Steel
Relaxation of steel is defined as the decrease in stress with time under constant strain.
Due to the relaxation of steel, the prestress in the tendon is reduced with time. Hence,
the study of relaxation is important in prestressed concrete to calculate the loss in
prestress.
The relaxation depends on the type of steel, initial prestress and the temperature. The
following figure shows the effect of relaxation due to different types of loading conditions.
fp
Fast loading
Effect of relaxation With sustained loading
p
Figure 1-7.8 Effect of relaxation due to different types of loading conditions
The following figure shows the variation of stress with time for different levels of
prestressing. Here, the instantaneous stress (fp) is normalised with respect to the initial
prestressing (fpi) in the ordinate. The curves are for different values of fpi/fpy, where fpy is
the yield stress.

100
90 f pi
fp = 0.6
80 f py 0.7
fpi 70 0.8
0.9
60
50
10 100 1000 10,000 100,000
Time (hours)
Figure 1-7.9 Variation of stress with time for different levels of prestressing
It can be observed that there is significant relaxation loss when the applied stress is
more than 70% of the yield stress.
The following photos show the test set-up for relaxation test.
Specimen
Load cell
(a) Test of a single wire strand

Specimen
(b) Test of a seven-wire strand

Figure 1-7.10 Set-up for relaxation test
The upper limits of relaxation loss are specified as follows.

Table 1-7.5 Relaxation losses at 1000 hours (IS:1785, IS:6003, IS:6006, IS:2090)
Cold drawn stress-relieved wires 5% of initial prestress
Indented wires 5% of initial prestress
Stress-relieved strand 5% of initial prestress
Bars 49 N/mm2
In absence of test data, IS:1343 - 1980 recommends the following estimates of

relaxation losses.
Table 1-7.6 Relaxation losses at 1000 hours at 27C
Initial Stress Relaxation Loss (N/mm2)
0.5fpk 0
0.6fpk 35
0.7fpk 70
0.8fpk 90
Fatigue
Under repeated dynamic loads the strength of a member may reduce with the number
of cycles of applied load. The reduction in strength is referred to as fatigue.

In prestressed applications, the fatigue is negligible in members that do not crack under
service loads. If a member cracks, fatigue may be a concern due to high stress in the
steel at the location of cracks.
Specimens are tested under 2 x 106 cycles of load to observe the fatigue. For steel,
fatigue tests are conducted to develop the stress versus number of cycles for failure (S-
N) diagram. Under a limiting value of stress, the specimen can withstand infinite number
of cycles. This limit is known as the endurance limit.
The prestressed member is designed such that the stress in the steel due to service
loads remains under the endurance limit. The following photo shows a set-up for
fatigue testing of strands.
Figure 1-7.11 Set-up for fatigue testing of strands
Durability
Prestressing steel is susceptible to stress corrosion and hydrogen embrittlement in
aggressive environments. Hence, prestressing steel needs to be adequately protected.
For bonded tendons, the alkaline environment of the grout provides adequate protection.
For unbonded tendons, corrosion protection is provided by one or more of the following
methods.

1) Epoxy coating
2) Mastic wrap (grease impregnated tape)
3) Galvanized bars
4) Encasing in tubes.
1.7.4 Codal Provisions of Steel
The following topics are covered in IS:1343 - 1980 under the respective sections. These
provisions are not duplicated here.
Table 1-7.7 Topics and sections
Assembly of prestressing and reinforcing steel Section 11
Prestressing Section 12

2.1 Losses in Prestress (Part I)

Introduction
Elastic Shortening
The relevant notations are explained first.
Notations
Geometric Properties
The commonly used geometric properties of a prestressed member are defined as
follows.
Ac = Area of concrete section
= Net cross-sectional area of concrete excluding the area of
prestressing steel.
Ap = Area of prestressing steel
= Total cross-sectional area of the tendons.
A = Area of prestressed member
= Gross cross-sectional area of prestressed member.
= Ac + Ap
At = Transformed area of prestressed member
= Area of the member when steel is substituted by an equivalent
area of concrete.
= Ac + mAp
= A + (m 1)Ap
Here,
m = the modular ratio = Ep/Ec
Ec = short-term elastic modulus of concrete
Ep = elastic modulus of steel.
The following figure shows the commonly used areas of the prestressed members.

= +
A Ac Ap At
Figure 2-1.1 Areas for prestressed members
CGC = Centroid of concrete
= Centroid of the gross section. The CGC may lie outside the
concrete (Figure 2-1.2).
CGS = Centroid of prestressing steel
= Centroid of the tendons. The CGS may lie outside the tendons or
the concrete (Figure 2-1.2).
I = Moment of inertia of prestressed member
= Second moment of area of the gross section about the CGC.
It = Moment of inertia of transformed section
= Second moment of area of the transformed section about the
centroid of the transformed section.
e = Eccentricity of CGS with respect to CGC

= Vertical distance between CGC and CGS. If CGS lies below CGC,
e will be considered positive and vice versa (Figure 2-1.2).
e CGC
CGC
CGS
e
CGS
Figure 2-1.2 CGC, CGS and eccentricity of typical prestressed members
Load Variables
Pi = Initial prestressing force
= The force which is applied to the tendons by the jack.

P0 = Prestressing force after immediate losses

= The reduced value of prestressing force after elastic shortening,
anchorage slip and loss due to friction.
Pe = Effective prestressing force after time-dependent losses
= The final value of prestressing force after the occurrence of creep,
shrinkage and relaxation.
2.1.1 Introduction
In prestressed concrete applications, the most important variable is the prestressing

force. In the early days, it was observed that the prestressing force does not stay
constant, but reduces with time. Even during prestressing of the tendons and the
transfer of prestress to the concrete member, there is a drop of the prestressing force
from the recorded value in the jack gauge. The various reductions of the prestressing
force are termed as the losses in prestress.
The losses are broadly classified into two groups, immediate and time-dependent. The
immediate losses occur during prestressing of the tendons and the transfer of prestress
to the concrete member. The time-dependent losses occur during the service life of the
prestressed member. The losses due to elastic shortening of the member, friction at the
tendon-concrete interface and slip of the anchorage are the immediate losses. The
losses due to the shrinkage and creep of the concrete and relaxation of the steel are the
time-dependent losses. The causes of the various losses in prestress are shown in the
following chart.
Losses
Immediate Time dependent
Elastic Friction Anchorage Creep Shrinkage Relaxation

shortening slip
Figure 2-1.3 Causes of the various losses in prestress

2.1.2 Elastic Shortening
Pre-tensioned Members
When the tendons are cut and the prestressing force is transferred to the member, the
concrete undergoes immediate shortening due to the prestress. The tendon also
shortens by the same amount, which leads to the loss of prestress.
Post-tensioned Members
If there is only one tendon, there is no loss because the applied prestress is recorded
after the elastic shortening of the member. For more than one tendon, if the tendons
are stretched sequentially, there is loss in a tendon during subsequent stretching of the
other tendons.
The elastic shortening loss is quantified by the drop in prestress (fp) in a tendon due to
the change in strain in the tendon (p). It is assumed that the change in strain in the
tendon is equal to the strain in concrete (c) at the level of the tendon due to the
prestressing force. This assumption is called strain compatibility between concrete
and steel. The strain in concrete at the level of the tendon is calculated from the stress
in concrete (fc) at the same level due to the prestressing force. A linear elastic
relationship is used to calculate the strain from the stress.
The quantification of the losses is explained below.

fp = Ep p
= Epc
f
= Ep c
Ec
fp = mfc (2-1.1)
For simplicity, the loss in all the tendons can be calculated based on the stress in
concrete at the level of CGS. This simplification cannot be used when tendons are
stretched sequentially in a post-tensioned member. The calculation is illustrated for the
following types of members separately.
Pre-tensioned Axial Members
Pre-tensioned Bending Members
Post-tensioned Axial Members

Post-tensioned Bending Members
Pre-tensioned Axial Members

The following figure shows the changes in length and the prestressing force due to
elastic shortening of a pre-tensioned axial member.
Original length of member at transfer of prestress
Pi
Length after elastic shortening

P0
Figure 2-1.4 Elastic shortening of a pre-tensioned axial member
The loss can be calculated as per Eqn. (2-1.1) by expressing the stress in concrete in
terms of the prestressing force and area of the section as follows.
fp = mfc
P
= m 0
Ac
P Pi
fp = m i m A (2-1.2)
At
Note that the stress in concrete due to the prestressing force after immediate losses
(P0/Ac) can be equated to the stress in the transformed section due to the initial
prestress (Pi /At). This is derived below. Further, the transformed area At of the
prestressed member can be approximated to the gross area A.
The following figure shows that the strain in concrete due to elastic shortening (c) is the
difference between the initial strain in steel (pi) and the residual strain in steel (p0).

Length of tendon before stretching

pi
Pi
p 0 c
P0
Figure 2-1.5 Strain variables in elastic shortening
The following equation relates the strain variables.

c = pi - p0 (2-1.3)
The strains can be expressed in terms of the prestressing forces as follows.
P0 (2-1.4)
c =
Ac Ec
Pi (2-1.5)
pi =
Ap E p
P0 (2-1.6)
p0 =
Ap E p
Substituting the expressions of the strains in Eqn. (2-1.3)
P0 P P
= i - 0
Ac Ec Ap E p Ap E p
1 1 Pi
or , P0 + =
Ac Ec Ap E p Ap E p
m 1 Pi
or, P0 + =
Ac Ap Ap
P0 Pi
or, =
Ac mAp + Ac
P0 Pi (2-1.7)
or =
Ac At
Thus, the stress in concrete due to the prestressing force after immediate losses (P0/Ac)
can be equated to the stress in the transformed section due to the initial prestress (Pi
/At).

The following problem illustrates the calculation of loss due to elastic shortening in an
idealised pre-tensioned railway sleeper.
Example 2-1.1
A prestressed concrete sleeper produced by pre-tensioning method has a

rectangular cross-section of 300mm 250 mm (b h). It is prestressed with 9
numbers of straight 7mm diameter wires at 0.8 times the ultimate strength of 1570
N/mm2. Estimate the percentage loss of stress due to elastic shortening of
concrete. Consider m = 6.
40
250
40
300
Solution
a) Approximate solution considering gross section
The sectional properties are calculated as follows.

Area of a single wire, Aw = /4 72
= 38.48 mm2
Area of total prestressing steel, Ap = 9 38.48

= 346.32 mm2
Area of concrete section, A = 300 250

= 75 103 mm2
Moment of inertia of section, I = 300 2503/12

= 3.91 108 mm4
Distance of centroid of steel area (CGS) from the soffit,

4 38.48 ( 250 - 40 ) + 5 38.48 40

y=
938.48
= 115.5 mm
115.5
Prestressing force, Pi = 0.8 1570 346.32 N

= 435 kN
Eccentricity of prestressing force,

e = (250/2) 115.5
= 9.5 mm
The stress diagrams due to Pi are shown.
+ =
Pi Pi .e Pi Pi .e
- y - y
A I A I
Since the wires are distributed above and below the CGC, the losses are calculated for
the top and bottom wires separately.
Stress at level of top wires (y = yt = 125 40)
P P .e
( fc )t = - i + i y t
A I
435103 435 103 9.5
=- + (125 - 40 )
75103 3.91108
= -5.8 + 0.9
= -4.9 N/mm2

Stress at level of bottom wires (y = yb = 125 40),

P P .e
( fc )b = - i - i y b
A I
435103 435103 9.5
=- - (125 - 40 )
75103 3.91108
= -5.8 - 0.9
= -6.7 N/mm2
Loss of prestress in top wires = mfcAp

(in terms of force) = 6 4.9 (4 38.48)
= 4525.25 N
Loss of prestress in bottom wires = 6 6.7 (5 38.48)

= 7734.48 N
Total loss of prestress = 4525 + 7735

= 12259.73 N
12.3 kN
Percentage loss = (12.3 / 435) 100%

= 2.83%
b) Accurate solution considering transformed section.
Transformed area of top steel,

A1 = (6 1) 4 38.48
= 769.6 mm2
Transformed area of bottom steel,

A2 = (6 1) 5 38.48
= 962.0 mm2
Total area of transformed section,

AT = A + A1 + A2
= 75000.0 + 769.6 + 962.0
= 76731.6 mm2

Centroid of the section (CGC)

A 125 + A1 (250 - 40) + A2 40
y=
A
= 124.8 mm from soffit of beam
Moment of inertia of transformed section,

IT = Ig + A(0.2)2 + A1(210 124.8)2 + A2(124.8 40)2
= 4.02 108mm4
Eccentricity of prestressing force,

e = 124.8 115.5
= 9.3 mm
Stress at the level of bottom wires,

435 103 (435 103 9.3)84.8
(fc)b = - -
76.73 103 4.02108
= -5.67 - 0.85
= -6.52 N/mm2
Stress at the level of top wires,

435 103 (435 103 9.3)85.2
(fc)t = - +
76.73 103 4.02108
= -5.67 + 0.86
= -4.81N/mm2
Loss of prestress in top wires = 6 4.81 (4 38.48)

= 4442 N
Loss of prestress in bottom wires = 6 6.52 (5 38.48)

= 7527 N
Total loss = 4442 + 7527

= 11969 N
12 kN
Percentage loss = (12 / 435) 100%

= 2.75 %

It can be observed that the accurate and approximate solutions are close. Hence, the
simpler calculations based on A and I is acceptable.
Pre-tensioned Bending Members

The following figure shows the changes in length and the prestressing force due to
elastic shortening of a pre-tensioned bending member.
Pi
wsw (self-weight)
Figure 2-1.6 Elastic shortening of a pre-tensioned bending member
Due to the effect of self-weight, the stress in concrete varies along length (Figure 2-1.6).
The loss can be calculated by Eqn. (2-1.1) with a suitable evaluation of the stress in
concrete. To have a conservative estimate of the loss, the maximum stress at the level
of CGS at the mid-span is considered.
Pi Pe.e M e (2-1.8)
fc = - - i + sw
A I I
Here, Msw is the moment at mid-span due to self-weight. Precise result using At and It in
place of A and I, respectively, is not computationally warranted. In the above
expression, the eccentricity of the CGS (e) was assumed to be constant.
For a large member, the calculation of the loss can be refined by evaluating the strain in
concrete at the level of the CGS accurately from the definition of strain. This is
demonstrated later for post-tensioned bending members.
Post-tensioned Axial Members

For more than one tendon, if the tendons are stretched sequentially, there is loss in a
tendon during subsequent stretching of the other tendons. The loss in each tendon can

be calculated in progressive sequence. Else, an approximation can be used to

calculate the losses.
The loss in the first tendon is evaluated precisely and half of that value is used as an
average loss for all the tendons.
1
fp = fp1
2
1
= mfc1
2
1 n P
= m i,j (2-1.9)
2 j=2 A
Here,
Pi,j = initial prestressing force in tendon j
n = number of tendons
The eccentricity of individual tendon is neglected.
Post-tensioned Bending Members

The calculation of loss for tendons stretched sequentially, is similar to post-tensioned
axial members. For curved profiles, the eccentricity of the CGS and hence, the stress in
concrete at the level of CGS vary along the length. An average stress in concrete can
be considered.
For a parabolic tendon, the average stress (fc,avg) is given by the following equation.
2 (2-1.10)
fc,avg = fc1 + ( fc 2 - fc1 )
3
Here,
fc1 = stress in concrete at the end of the member
fc2 = stress in concrete at the mid-span of the member.
A more rigorous analysis of the loss can be done by evaluating the strain in concrete at
the level of the CGS accurately from the definition of strain. This is demonstrated for a
beam with two parabolic tendons post-tensioned sequentially. In Figure 2-1.7, Tendon
B is stretched after Tendon A. The loss in Tendon A due to elastic shortening during
tensioning of Tendon B is given as follows.
fp = E p c
= E p [c 1 + c 2 ] (2-1.11)

Here, c is the strain at the level of Tendon A. The component of c due to pure
compression is represented as c1. The component of c due to bending is represented
as c2. The two components are calculated as follows.
PB
c 1 =
AEc
L
c 2 =
L
1 L P .e (x).eA (x)
= B B dx
L0 IEc
PB L
= eB (x).eA (x) dx (2-1.12)
Ec LI 0
Here,
A = cross-sectional area of beam
PB = prestressing force in Tendon B
Ec = modulus of concrete
L = length of beam
eA(x), eB(x) = eccentricities of Tendons A and B, respectively, at distance x
from left end
I = moment of inertia of beam
L = change in length of beam
The variations of the eccentricities of the tendons can be expressed as follows.
x x
eA (x) = eA1 + 4 eA 1
(2-1.13)
L L
x x
eB (x) = eB1 + 4 eB 1 (2-1.14)
L L
Where , eA = eA 2 eA1
eB = eB 2 eB1
eA1, eA2 = eccentricities of Tendon A at 1 (end) and 2 (centre), respectively.

eB1, eB2 = eccentricities of Tendon B at 1 and 2, respectively.

Substituting the expressions of the eccentricities in Eqn. (2-1.12), the second

component of the strain is given as follows.
PB 1 2 8
= eA1eB1 + ( eA1eB 2 + eA2eB1 ) + eA2eB 2 (2-1.15)
Ec I 5 15 15

2.2 Losses in Prestress (Part II)

Friction
Anchorage Slip
Force Variation Diagram
2.2.1 Friction
The friction generated at the interface of concrete and steel during the stretching of a
curved tendon in a post-tensioned member, leads to a drop in the prestress along the
member from the stretching end. The loss due to friction does not occur in pre-
tensioned members because there is no concrete during the stretching of the tendons.
The friction is generated due to the curvature of the tendon and the vertical component
of the prestressing force. The following figure shows a typical profile (laying pattern) of
the tendon in a continuous beam.
Figure 2-2.1 A typical continuous post-tensioned member

In addition to friction, the stretching has to overcome the wobble of the tendon. The
wobble refers to the change in position of the tendon along the duct. The losses due to
friction and wobble are grouped together under friction.
The formulation of the loss due to friction is similar to the problem of belt friction. The
sketch below (Figure 2-2.2) shows the forces acting on the tendon of infinitesimal length
dx.

d
P d/2
R
N
P + dP
N P + dP
Force triangle
P
dx
Figure 2-2.2 Force acting in a tendon of infinitesimal length
In the above sketch,

P = prestressing force at a distance x from the stretching end
R = radius of curvature
d = subtended angle.
The derivation of the expression of P is based on a circular profile. Although a cable

profile is parabolic based on the bending moment diagram, the error induced is
insignificant.
The friction is proportional to the following variables.

Coefficient of friction () between concrete and steel.
The resultant of the vertical reaction from the concrete on the tendon (N)
generated due to curvature.
From the equilibrium of forces in the force triangle, N is given as follows.

d
N = 2Psin
2
d
2P = Pd (2-2.1)
2
The friction over the length dx is equal to N = Pd.
Thus the friction (dP) depends on the following variables.

Coefficient of friction ()
Curvature of the tendon (d)
The amount of prestressing force (P)

The wobble in the tendon is effected by the following variables.

Rigidity of sheathing
Diameter of sheathing
Spacing of sheath supports
Type of tendon
Type of construction
The friction due to wobble is assumed to be proportional to the following.

Length of the tendon
Prestressing force
For a tendon of length dx, the friction due to wobble is expressed as kPdx, where k is
the wobble coefficient or coefficient for wave effect.
Based on the equilibrium of forces in the tendon for the horizontal direction, the
following equation can be written.
P = P + dP + (Pd + kPdx)
or, dP = (Pd + kPdx) (2-2.2)
Thus, the total drop in prestress (dP) over length dx is equal to (Pd + kPdx). The
above differential equation can be solved to express P in terms of x.
Px
dP
= - d + k dx
x

P P
0
0 0
or, lnP P = - ( + kx )
Px
0
Px
or, ln = - ( + kx )
P0
or, Px = P0e -( +kx ) (2-2.3)
Here,
P0 = the prestress at the stretching end after any loss due to elastic shortening.
For small values of + kx, the above expression can be simplified by the Taylor series
expansion.
Px = P0 (1 kx) (2-2.4)
Thus, for a tendon with single curvature, the variation of the prestressing force is linear
with the distance from the stretching end. The following figure shows the variation of

prestressing force after stretching. The left side is the stretching end and the right side
is the anchored end.
P0 Px
Figure 2-2.3 Variation of prestressing force after stretching
In the absence of test data, IS:1343 - 1980 provides guidelines for the values of and k.
Table 2-2.1 Values of coefficient of friction
Type of interface
For steel moving on smooth concrete 0.55.
For steel moving on steel fixed to duct 0.30.
For steel moving on lead 0.25.
The value of k varies from 0.0015 to 0.0050 per meter length of the tendon depending
on the type of tendon. The following problem illustrates the calculation of the loss due
to friction in a post-tensioned beam.
Example 2-2.1
A post-tensioned beam 100 mm 300 mm (b h) spanning over 10 m is stressed

by successive tensioning and anchoring of 3 cables A, B, and C respectively as
shown in figure. Each cable has cross section area of 200 mm2 and has initial
stress of 1200 MPa. If the cables are tensioned from one end, estimate the
percentage loss in each cable due to friction at the anchored end. Assume =
0.35, k = 0.0015 / m.

CL
50
50 CGC
Cable A
Cable B
Cable C
Solution
Prestress in each tendon at stretching end = 1200 200

= 240 kN.
To know the value of (L), the equation for a parabolic profile is required.
dy 4ym
= (L - 2 x )
dx L2
ym x
y (L)
Here,
ym = displacement of the CGS at the centre of the beam from the ends
L = length of the beam
x = distance from the stretching end
y = displacement of the CGS at distance x from the ends.
An expression of (x) can be derived from the change in slope of the profile. The slope
of the profile is given as follows.
dy 4ym
= (L - 2 x )
dx L2
At x = 0, the slope dy/dx = 4ym/L. The change in slope (x) is proportional to x.
The expression of (x) can be written in terms of x as (x) = .x,

where, = 8ym/L2. The variation is shown in the following sketch.


8ym/L
4ym/L

0 L/2 L x
The total subtended angle over the length L is 8ym/L.
The prestressing force Px at a distance x is given by
Px = P0e( + kx) = P0ex

where, x = + kx
For cable A, ym = 0.1 m.

For cable B, ym = 0.05 m.
For cable C, ym = 0.0 m.
For all the cables, L = 10 m.
Substituting the values of ym and L
0.0043x for cable A

x = 0.0029x for cable B
0.0015x for cable C

The maximum loss for all the cables is at x = L = 10, the anchored end.
0.958 for cable A

-L
e = 0.971 for cable B
0.985 for cable C

Percentage loss due to friction = (1 eL) 100%
4.2% for cable A

= 2.9% for cable B
1.5% for cable C


CL
CGC
Cable A
Cable B
Cable C
240 kN
Variation of prestressing forces
The loss due to friction can be considerable for long tendons in continuous beams with
changes in curvature. The drop in the prestress is higher around the intermediate
supports where the curvature is high. The remedy to reduce the loss is to apply the
stretching force from both ends of the member in stages.
2-2.2 Anchorage Slip
In a post-tensioned member, when the prestress is transferred to the concrete, the

wedges slip through a little distance before they get properly seated in the conical space.
The anchorage block also moves before it settles on the concrete. There is loss of
prestress due to the consequent reduction in the length of the tendon.
The total anchorage slip depends on the type of anchorage system. In absence of
manufacturers data, the following typical values for some systems can be used.
Table 2-2.2 Typical values of anchorage slip
Anchorage System Anchorage Slip (s)
Freyssinet system
12 - 5mm strands 4 mm
12 - 8mm strands 6 mm
Magnel system 8 mm
Dywidag system 1 mm
(Reference: Rajagopalan, N., Prestressed Concrete)

Due to the setting of the anchorage block, as the tendon shortens, there is a reverse
friction. Hence, the effect of anchorage slip is present up to a certain length (Figure 2-
2.4). Beyond this setting length, the effect is absent. This length is denoted as lset.
P0
Px
Figure 2-2.4 Variation of prestressing force after anchorage slip
2.2.3 Force Variation Diagram
The magnitude of the prestressing force varies along the length of a post-tensioned
member due to friction losses and setting of the anchorage block. The diagram
representing the variation of prestressing force is called the force variation diagram.
Considering the effect of friction, the magnitude of the prestressing force at a distance x
from the stretching end is given as follows.
Px = P0e -x (2-2.5)
Here, x = + kx denotes the total effect of friction and wobble. The plot of Px gives
the force variation diagram.
The initial part of the force variation diagram, up to length lset is influenced by the setting
of the anchorage block. Let the drop in the prestressing force at the stretching end be
P. The determination of P and lset are necessary to plot the force variation diagram
including the effect of the setting of the anchorage block.
Considering the drop in the prestressing force and the effect of reverse friction, the
magnitude of the prestressing force at a distance x from the stretching end is given as
follows.
Px' = ( P0 - P ) e 'x (2-2.6)

Here, for reverse friction is analogous to for friction and wobble.

At the end of the setting length (x = lset), Px = Px
Px
P0
P
Px after stretching
Px Px Px after setting
Px beyond lset
lset x
Figure 2-2.5 Force variation diagram near the stretching end
Substituting the expressions of Px and Px for x = lset
Since it is difficult to measure separately, is taken equal to . The expression of

P simplifies to the following.
P0e -lset = ( P0 - P ) e 'lset

P0e - ( +' )lset = P0 - P
P0 1- ( + ' ) l set = P0 - P
'
P = P0 ( + ' ) l set = P0 l set 1+ (2-2.7)

P = 2P0lset (2-2.8)
The following equation relates lset with the anchorage slip s.

1 P
s = lset
2 Ap E p
1 lset ' (2-2.9)
s = P0lset 1+
2 Ap E p
Transposing the terms,

2 2 Ap E p
lset = s
'
P0 1+

s Ap E p
= for ' =
P0

Therefore,
s Ap E p
lset = (2-2.10)
P0
The term P0 represents the loss of prestress per unit length due to friction.
The force variation diagram is used when stretching is done from both the ends. The
tendons are overstressed to counter the drop due to anchorage slip. The stretching from
both the ends can be done simultaneously or in stages. The final force variation is more
uniform than the first stretching.
The following sketch explains the change in the force variation diagram due to
stretching from both the ends in stages.
a) After stretching from right end
b) After anchorage slip at right end
c) After stretching from left end
d) After anchorage slip at left end

Figure 2-2.6 Force variation diagrams for stretching in stages

The force variation diagrams for the various stages are explained.
a) The initial tension at the right end is high to compensate for the anchorage
slip. It corresponds to about 0.8 fpk initial prestress. The force variation
diagram (FVD) is linear.
b) After the anchorage slip, the FVD drops near the right end till the length lset.
c) The initial tension at the left end also corresponds to about 0.8 fpk initial prestress.
The FVD is linear up to the centre line of the beam.
d) After the anchorage slip, the FVD drops near the left end till the length lset. It is
observed that after two stages, the variation of the prestressing force over the length
of the beam is less than after the first stage.
Example 2-2.2
A four span continuous bridge girder is post-tensioned with a tendon consisting

of twenty strands with fpk = 1860 MPa. Half of the girder is shown in the figure
below. The symmetrical tendon is simultaneously stressed up to 75% fpk from
both ends and then anchored. The tendon properties are Ap = 2800 mm2, Ep =
195,000 MPa, = 0.20, K = 0.0020/m. The anchorage slip s = 6 mm.
Calculate
a) The expected elongation of the tendon after stretching,
b) The force variation diagrams along the tendon before and after anchorage.
CL
0.76 0.6 0.76 0.6
13.7 13.7 3 3.7 15.2 15.2 3.7

All dimensions are in metres
Inflection points

Solution
Initial force at stretching end

0.75fpk = 1395 MPa
P0 = 0.75fpk Ap
= 3906 kN
The continuous tendon is analysed as segments of parabola. The segments are

identified between the points of maximum eccentricity and inflection points. The
inflection points are those where the curvature of the tendon reverses. The different
segments are as follows: 1-2, 2-3, 3-4, 4-5, 5-6, 6-7 and 7-8.
CL
3 5 7
1 2 4 6 8
The following properties of parabolas are used. For segment 1-2, the parabola in the
sketch below is used.
L
y
0 x
The change in slope from the origin to the end of the parabola is same as the slope at
the end of the tendon which is = 2e/L, where
L = length of the segment
e = vertical shift from the origin.
For segments 2-3 and 3-4 and subsequent pairs of segments, the following property is
used.

e2
e1
For the two parabolic segments joined at the inflection point as shown in the sketch
above, the slope at the inflection point = 2(e1 + e2)/L.
Here,
e1, e2 = eccentricities of the CGS at the span and support respectively
L = length of the span
L = fractional length between the points of maximum eccentricity
The change in slope between a point of maximum eccentricity and inflection point is
also equal to .
The change in slope () for each segment of the tendon is calculated using the above
expressions. Next the value of + kx for each segment is calculated using the given
values of , k and x, the horizontal length of the segment. Since the loss in prestress
accrues with each segment, the force at a certain segment is given as follows.
Px = P0 e -(+kx)
The summation is for the segments from the stretching end up to the point in the
segment under consideration. Hence, the value of ( + kx) at the end of each
segment is calculated to evaluate the prestressing force at that point (Px, where x
denotes the point).

0.163
0.111 0.163 0.144 0.144 0.144 0.144
+kx 0.039
0.050 0.060 0.036 0.059 0.059 0.036
( +kx) 0.149
k ) 0.050 0.110 0.185 0.244 0.303 0.339
e-( + kx) 0.712

0.861
1.000 0.952 0.896 0.831 0.783 0.738
Px (kN) 3363 2781
3906 3718 3500 3246 3058 2883
The force variation diagram before anchorage can be plotted with the above values of
Px. A linear variation of the force can be assumed for each segment. Since the
stretching is done at both the ends simultaneously, the diagram is symmetric about the
central line.
a) The expected elongation of the tendon after stretching

First the product of the average force and the length of each segment is summed up to
the centre line.
1 1
PavL = [3906 + 3718] 13.7 + [3718 + 3500] 13.7
2 2
1 1
+ [3500 + 3363] 3 + [3363 + 3246] 3.7
2 2
1 1
+ [3246 + 3058] 15.2 + [3058 + 2883] 15.2
2 2
1
+ [ 2883 + 2718] 3.7
2
= 227612.2 kN

The elongation () at each stretching end is calculated as follows.
Pav L
=
AP EP
227612103
=
2800195000
= 0.417 m
b) The force variation diagrams along the tendon before and after anchorage
After anchorage, the effect of anchorage slip is present up to the setting length lset. The
value of lset due to an anchorage slip s = 6 mm is calculated as follows.
s AP EP
lset =
P0
6 2800195000
=
13.7
= 15.46 m
The quantity P0 is calculated from the loss of prestress per unit length in the first
segment. P0 = (3906 3718) kN /13.7 m = 13.7 N/mm. The drop in the prestressing
force (p) at each stretching end is calculated as follows.
p = 2P0 lset
= 213.7 15464
= 423.7 kN
Thus the value of the prestressing force at each stretching end after anchorage slip is
3906 424 = 3482 kN. The force variation diagram for lset = 15.46 m is altered to show
the drop due to anchorage slip.
The force variation diagrams before and after anchorage are shown below. Note that
the drop of force per unit length is more over the supports due to change in curvature
over a small distance.

Prestressing force (kN)

4000
3500
3000
2500
0 20 40 60 80
Distance from end (m)
After anchorage Before anchorage

2.3 Losses in Prestress (Part III)

Creep of Concrete
Shrinkage of Concrete
Relaxation of Steel
Total Time Dependent Losses
2.3.1 Creep of Concrete
Creep of concrete is defined as the increase in deformation with time under constant
load. Due to the creep of concrete, the prestress in the tendon is reduced with time.
The creep of concrete is explained in Section 1.6, Concrete (Part II). Here, the
information is summarised. For stress in concrete less than one-third of the
characteristic strength, the ultimate creep strain (cr,ult) is found to be proportional to the
elastic strain (el). The ratio of the ultimate creep strain to the elastic strain is defined as
the ultimate creep coefficient or simply creep coefficient .
The ultimate creep strain is then given as follows.

cr,ult = el (2-3.1)
IS:1343 - 1980 gives guidelines to estimate the ultimate creep strain in Section 5.2.5. It
is a simplified estimate where only one factor has been considered. The factor is age of
loading of the prestressed concrete structure. The creep coefficient is provided for
three values of age of loading.
Curing the concrete adequately and delaying the application of load provide long term
benefits with regards to durability, loss of prestress and deflection.
In special situations detailed calculations may be necessary to monitor creep strain with
time. Specialised literature or international codes can provide guidelines for such
calculations.
The loss in prestress (fp ) due to creep is given as follows.

fp = Ep cr, ult (2-3.2)

Here, Ep is the modulus of the prestressing steel.
The following considerations are applicable for calculating the loss of prestress due to
creep.
1) The creep is due to the sustained (permanently applied) loads only.
Temporary loads are not considered in the calculation of creep.
2) Since the prestress may vary along the length of the member, an average value
of the prestress can be considered.
3) The prestress changes due to creep and the creep is related to the
instantaneous prestress. To consider this interaction, the calculation of creep can
be iterated over small time steps.
2.3.2 Shrinkage of Concrete
Shrinkage of concrete is defined as the contraction due to loss of moisture. Due to the
shrinkage of concrete, the prestress in the tendon is reduced with time.
The shrinkage of concrete was explained in details in the Section 1.6, Concrete (Part II).
IS:1343 - 1980 gives guidelines to estimate the shrinkage strain in Section 5.2.4. It is a
simplified estimate of the ultimate shrinkage strain (sh). Curing the concrete adequately
and delaying the application of load provide long term benefits with regards to durability
and loss of prestress. In special situations detailed calculations may be necessary to
monitor shrinkage strain with time. Specialised literature or international codes can
provide guidelines for such calculations.
The loss in prestress (fp ) due to shrinkage is given as follows.

fp = Ep sh (2-3.3)
Here, Ep is the modulus of the prestressing steel.
2.3.3 Relaxation of Steel
Relaxation of steel is defined as the decrease in stress with time under constant strain.
Due to the relaxation of steel, the prestress in the tendon is reduced with time. The
relaxation depends on the type of steel, initial prestress (fpi) and the temperature. To

calculate the drop (or loss) in prestress (fp), the recommendations of IS:1343 - 1980
can be followed in absence of test data.
Example 2-3.1
A concrete beam of dimension 100 mm 300 mm is post-tensioned with 5

straight wires of 7mm diameter. The average prestress after short-term losses is
0.7fpk = 1200 N/mm2 and the age of loading is given as 28 days. Given that Ep =
200 103 MPa, Ec = 35000 MPa, find out the losses of prestress due to creep,
shrinkage and relaxation. Neglect the weight of the beam in the computation of
the stresses.
100
300
50 CGS
Solution
Area of concrete A = 100 300

= 30000 mm2
Moment of inertia of beam section

I = 100 3003 / 12
= 225 106 mm4
Area of prestressing wires Ap = 5 (/4) 72

= 192.42 mm2
Prestressing force after short-term losses

P0 = Ap.fp0
= 192.4 1200
= 230880 N

Modular ratio m = Ep / Ec
= 2 105 / 35 103
= 5.71
Stress in concrete at the level of CGS
P0 P0e
fc = - - e
A I
230880 230880
=- 4
- 6
502
310 22510
= 7.69 2.56
= 10.25 N/mm2
Loss of prestress due to creep

(fp)cr = Ep cr, ult
= Ep el
= Ep (fc/Ec)
= m fc
= 5.71 10.25 1.6
= 93.64 N / mm2
Here, = 1.6 for loading at 28 days, from Table 2c-1 (Clause 5.2.5.1, IS:1343 - 1980).
Shrinkage strain from Clause 5.2.4.1, IS:1343 - 1980

sh = 0.0002 / log10(t + 2)
= 0.0002 / log10 (28 + 2)
= 1.354 10-4
Loss of prestress due to shrinkage

(fp)sh = Epsh
= 2 105 1.354 10-4
= 27.08 N/mm2

From Table 2c-2 (Table 4, IS:1343 - 1980)

Loss of prestress due to relaxation
(fp)rl = 70.0 N/mm2
Loss of prestressing force = fp Ap
Therefore,
Loss of prestressing force due to creep = 93.64 192.42
= 18018 N
Loss of prestressing force due to shrinkage

= 27.08 192.42
= 5211 N
Loss of prestressing force due to relaxation

= 70 192.42
= 13469 N
Total long-term loss of prestressing force (neglecting the interaction of the losses and
prestressing force)
= 18018 + 5211 + 13469
= 36698 N
Percentage loss of prestress = 36698 / 230880 100%

= 15.9 %
2.3.4 Total Time-dependent Loss
The losses of prestress due to creep and shrinkage of concrete and the relaxation of the
steel are all time-dependent and inter-related to each other. If the losses are calculated
separately and added, the calculated total time-dependent loss is over-estimated. To
consider the inter-relationship of the cause and effect, the calculation can be done for
discrete time steps. The results at the end of each time step are used for the next time
step. This step-by-step procedure was suggested by the Precast / Prestressed

Concrete Institute (PCI) committee and is called the General method (Reference: PCI
Committee, Recommendations for Estimating Prestress Losses, PCI Journal, PCI, Vol.
20, No. 4, July-August 1975, pp. 43-75).
In the PCI step-by-step procedure, a minimum of four time steps are considered in the
service life of a prestressed member. The following table provides the definitions of the
time steps (Table 2-3.3).
Table 2-3.3 Time steps in the step-by-step procedure
Step Beginning End
1 Pre-tension: Anchorage of steel Age of prestressing
Post-tension: End of curing
2 End of Step 1 30 days after prestressing or
when subjected to superimposed
load
3 End of Step 2 1 year of service
4 End of Step 3 End of service life
The step-by-step procedure can be implemented by a computer program, where the

number of time steps can be increased.
There are also approximate methods to calculate lump sum estimates of the total loss.
Since these estimates are not given in IS:1343 - 1980, they are not mentioned here.

3.1 Analysis of Members under Axial Load

Introduction
Analysis at Transfer
Analysis at Service Loads
Analysis of Ultimate Strength
Analysis of Behaviour
Notations
Geometric Properties
A prestressed axial member may also have non-prestressed reinforcement to carry the
axial force. This type of members is called partially prestressed members. The
commonly used geometric properties of a prestressed member with non-prestressed
reinforcement are defined as follows.
A = gross cross-sectional area
Ac = area of concrete
As = area of non-prestressed reinforcement
Ap = area of prestressing tendons
At = transformed area of the section
= Ac + (Es/ Ec) As + (Ep/ Ec) Ap
The following figure shows the commonly used areas of a prestressed member with
non-prestressed reinforcement.

= + +
A Ac As Ap
A
At
Figure 3-1.1 Areas for a prestressed member with non-prestressed reinforcement
3.1.1 Introduction
The study of members under axial load gives an insight of the behaviour of a
prestressed member as compared to an equivalent non-prestressed reinforced concrete
member. Prestressed members under axial loads only, are uncommon. Members such
as hangers and ties are subjected to axial tension. Members such as piles may have
bending moment along with axial compression or tension. In this section, no
eccentricity of the CGS with respect to CGC is considered. The definitions of CGS and
CGC are provided in Section 2.1, Losses in Prestress (Part I). The following figure
shows members under axial loads.

Hangers Piles
Figure 3-1.2 Members under axial load
The analysis of members refers to the evaluation of the following.

1) Permissible prestress based on allowable stresses at transfer.
2) Stresses under service loads. These are compared with allowable stresses
under service conditions.
3) Ultimate strength. This is compared with the demand under factored loads.
4) The entire axial load versus deformation behaviour.
The stages for loading are explained in Section 1.2, Advantages and Types of
Prestressing
3.1.2 Analysis at Transfer
The stress in the concrete (fc) in a member without non-prestressed reinforcement can
be calculated as follows.
P0
fc = -
Ac (3-1.1)
Here,
P0 = prestress at transfer after short-term losses.
In presence of non-prestressed reinforcement, the stress in the concrete can be

calculated as follows.

P0
fc = - (3-1.2)
Ac + (Es /Ec )As
The permissible prestress is determined based on fc to be within the allowable stress at

transfer.
3.1.3 Analysis at Service Loads
The stresses in concrete in a member without non-prestressed reinforcement can be

Pe P (3-1.3)
fc = -
Ac At
Here,
P = external axial force (In the equation, + for tensile force and vice
versa.)
Pe = effective prestress.
If there is non-prestressed reinforcement, Ac is to be substituted by (Ac + (Es/Ec) As) and

At is to be calculated including As.
The value of fc should be within the allowable stress under service conditions.
3.1.4 Analysis of Ultimate Strength
The ultimate tensile strength of a section (PuR) can be calculated as per Clause 22.3,
IS:1343 - 1980.
In absence of non-prestressed reinforcement,
PuR = 0.87fPk Ap (3-1.4a)
In presence of non-prestressed reinforcement,

PuR = 0.87fy As + 0.87fPk Ap (3-1.4b)

In the previous equations,

fy = characteristic yield stress for non-prestressed reinforcement with mild
steel bars
= characteristic 0.2% proof stress for non-prestressed reinforcement
with high yield strength deformed bars.
fpk = characteristic tensile strength of prestressing tendons.
The ultimate tensile strength should be greater than the demand due to factored loads.
The ultimate compressive strength of a section (PuR) can be calculated in presence of

moments by the use of interaction diagrams. For a member under compression with
minimum eccentricity, the ultimate strength is given as follows. Here, the contribution of
prestressing steel is neglected.
PuR = 0.4 fckAc + 0.67 fy As (3-1.5)
3.1.5 Analysis of Behaviour
The analysis of behaviour refers to the determination of the complete axial load versus
deformation behaviour. The analyses at transfer, under service loads and for ultimate
strength correspond to three instants in the above behaviour.
The analysis involves three principles of mechanics (Reference: Collins, M. P. and
Mitchell, D., Prestressed Concrete Structures, Prentice-Hall, Inc., 1991).
1) Equilibrium of internal forces with the external loads at any point of the load
versus deformation behaviour. The internal forces in concrete and steel are
evaluated based on the respective strains, cross-sectional areas and the
constitutive relationships.
2) Compatibility of the strains in concrete and in steel for bonded tendons. This
assumes a perfect bond between the two materials. For unbonded tendons, the
compatibility is in terms of total deformation.
3) Constitutive relationships relating the stresses and the strains in the materials.
The relationships are developed based on the material properties.

Equilibrium Equation
At any instant, the equilibrium is given by the following equation.
P = Acfc + Asfs + Apfp (3-1.6)
Here,
fc = stress in concrete
fs = stress in non-prestressed reinforcement
fp = stress in prestressed tendons
P = axial force.
Compatibility Equations
For non-prestressed reinforcement
s = c (3-1.7)
For prestressed tendons
p = c + p (3-1.8)
Here,
c = strain in concrete at the level of the steel
s = strain in non-prestressed reinforcement
p = strain in prestressed tendons
p = strain difference in prestressed tendons with adjacent concrete
The strain difference (p) is the strain in the prestressed tendons when the concrete
has zero strain (c = 0). This occurs when the strain due to the external tensile axial
load balances the compressive strain due to prestress. At any load stage,
p = pe ce (3-1.9)
Here,
pe = strain in tendons due to Pe, the prestress at service
ce = strain in concrete due to Pe.
The strain difference is further explained in Section 3.4, Analysis of Member under
Flexure (Part III).
Constitutive Relationships
The constitutive relationships can be expressed in the following forms based on the
material stress-strain curves shown in Section 1.6, Concrete (Part II), and Section 1.7,
Prestressing Steel.

For concrete under compression

fc = F1 (c) (3-1.10)
For prestressing steel

fp = F2 (p) (3-1.11)
For reinforcing steel

fs = F3 (s) (3-1.12)
The stress versus strain curve for concrete is shown below. The first and third
quadrants represent the behaviour under tension and compression, respectively.
fc
c
Figure 3-1.3 Stress versus strain for concrete
The stress versus strain curve for prestressing steel is as shown below.
fp
p
Figure 3-1.4 Stress versus strain for prestressing steel
The following stress versus strain curve is for reinforcing steel.

fs
Figure 3-1.5 Stress versus strain for reinforcing steel
The equilibrium and compatibility equations and the constitutive relationships can be
solved to develop the axial force versus deformation curve. The deformation can be
calculated as cL, where L is the length of the member.
The following plot shows the axial force versus deformation curves for prestressed and
non-prestressed sections. The two sections are equivalent in their ultimate tensile
strengths.
Axial force
Cracking Tensile strengths
Deformation
Compressive strengths
Prestressed section
Non-prestressed section
Figure 3-1.6 Axial force versus deformation curves
From the previous plot, the following can be inferred.

1) Prestressing increases the cracking load.
2) Prestressing shifts the curve from the origin.
For the prestressed member, there is a compressive deformation in absence
of external axial force.

A certain amount of external force is required to decompress the member.

3) For a given tensile load, the deformation of the prestressed member is

smaller.
Prestressing reduces deformation at service loads.
4) For a given compressive load, the deformation of the prestressed member is

larger.
Prestressing is detrimental for the response under compression.
5) The compressive strength of the prestressed member is lower.

Prestressing is detrimental for the compressive strength.
6) For a partially prestressed section with the same ultimate strength, the axial load
versus deformation curve will lie in between the curves for prestressed and non-
prestressed sections.
The above conclusions are generic for prestressed members.

3.2 Analysis of Members under Flexure (Part I)

Introduction
Analyses at Transfer and at Service
3.2.1 Introduction
Similar to members under axial load, the analysis of members under flexure refers to
the evaluation of the following.
1) Permissible prestress based on allowable stresses at transfer.
2) Stresses under service loads. These are compared with allowable stresses under
service conditions.
3) Ultimate strength. This is compared with the demand under factored loads.
4) The entire load versus deformation behaviour.
The analyses at transfer and under service loads are presented in this section. The
analysis for the ultimate strength is presented separately in Section 3.4, Analysis of
Member under Flexure (Part III). The evaluation of the load versus deformation
behaviour is required in special type of analysis. This analysis will not be covered in
this section.
Assumptions
The analysis of members under flexure considers the following.
1) Plane sections remain plane till failure (known as Bernoullis hypothesis).
2) Perfect bond between concrete and prestressing steel for bonded tendons.
Principles of Mechanics
The analysis involves three principles of mechanics.
1) Equilibrium of internal forces with the external loads. The compression in
concrete (C) is equal to the tension in the tendon (T). The couple of C and T are
equal to the moment due to external loads.
2) Compatibility of the strains in concrete and in steel for bonded tendons. The
formulation also involves the first assumption of plane section remaining plane
after bending. For unbonded tendons, the compatibility is in terms of deformation.

Variation of Internal Forces

In reinforced concrete members under flexure, the values of compression in concrete
(C) and tension in the steel (T) increase with increasing external load. The change in
the lever arm (z) is not large.
In prestressed concrete members under flexure, at transfer of prestress C is located

close to T. The couple of C and T balance only the self weight. At service loads, C
shifts up and the lever arm (z) gets large. The variation of C or T is not appreciable.
The following figure explains this difference schematically for a simply supported beam
under uniform load.
w1 w1
C1 z1
z1 C1
T1 T1
w2 w 2 > w1 w2
C2 C2
z2 z2
T2 T2
Reinforced concrete Prestressed concrete

C2 > C1, z2 z1 C2 C1, z2 > z1
Figure 3-2.1 Variations of internal forces and lever arms
In the above figure,

C1, T1 = compression and tension at transfer due to self weight
C2, T2 = compression and tension under service loads
w1 = self weight
w2 = service loads
z1 = lever arm at transfer
z2 = lever arm under service loads.

For the reinforced concrete member C2 is substantially large than C1, but z2 is close to
z1. For the prestressed concrete member C2 is close to C1, but z2 is substantially large
than z1.
3.2.2 Analyses at Transfer and at Service
The analyses at transfer and under service loads are similar. Hence, they are presented
together. A prestressed member usually remains uncracked under service loads. The
concrete and steel are treated as elastic materials. The principle of superposition is
applied. The increase in stress in the prestressing steel due to bending is neglected.
There are three approaches to analyse a prestressed member at transfer and under
service loads. These approaches are based on the following concepts.
a) Based on stress concept.
b) Based on force concept.
c) Based on load balancing concept.
The following material explains the three concepts.
Based on Stress Concept

In the approach based on stress concept, the stresses at the edges of the section under
the internal forces in concrete are calculated. The stress concept is used to compare
the calculated stresses with the allowable stresses.
The following figure shows a simply supported beam under a uniformly distributed load
(UDL) and prestressed with constant eccentricity (e) along its length.
CGC
e
CGS
Figure 3-2.2 A simply supported beam under UDL
The following sketch shows the internal forces in concrete at a section and the
corresponding stress profiles. The first stress profile is due to the compression P. The
second profile is due to the eccentricity of the compression. The third profile is due to
the moment. At transfer, the moment is due to self weight. At service the moment is
due to service loads.

M
CGC + + =
P
Internal P/A Pey/ I My/ I Resultant

forces in (stress (stress (stress stress
concrete due to P) due to P.e) due to M) profile
Figure 3-2.3 Stress profiles at a section due to internal forces
The resultant stress at a distance y from the CGC is given by the principle of
superposition as follows.
P Pey My
f =- (3-2.1)
A I I
For a curved tendon, P can be substituted by its horizontal component. But the effect of
the refinement is negligible.
Based on Force Concept

The approach based on force concept is analogous to the study of reinforced concrete.
The tension in prestressing steel (T) and the resultant compression in concrete (C) are
considered to balance the external loads. This approach is used to determine the
dimensions of a section and to check the service load capacity. Of course, the stresses
in concrete calculated by this approach are same as those calculated based on stress
concept. The stresses at the extreme edges are compared with the allowable stresses.
The following figures show the internal forces in the section.

C
ec
z
e T
T C
Internal forces at prestressing Internal forces after

(neglecting self-weight) loading
Figure 3-2.4 Internal forces at a section
The equilibrium equations are as follows.
C =T

(3-2.2)
M = C.z
M = C(ec + e) (3-2.3)
The resultant stress in concrete at distance y from the CGC is given as follows.
C Cec y
f =- (3-2.4)
A I
Substituting C = P and Cec = M Pe, the expression of stress becomes same as that
given by the stress concept.
(3-2.5)
P Pey My
f =-
A I I
Based on Load Balancing Concept

The approach based on load balancing concept is used for a member with curved or
harped tendons and in the analysis of indeterminate continuous beams. The moment,
upward thrust and upward deflection (camber) due to the prestress in the tendons are
calculated. The upward thrust balances part of the superimposed load.
The expressions for three profiles of tendons in simply supported beams are given.
a) For a Parabolic Tendon

e P
wup
L
Free body diagram of concrete
M
Bending moment diagram
Figure 3-2.5 Simply supported beam with parabolic tendon
The moment at the centre due to the uniform upward thrust (wup) is given by the
following equation.
wup L2
M=
8

(3-2.6)
The moment at the centre from the prestressing force is given as M = Pe.
The expression of wup is calculated by equating the two expressions of M. The upward
deflection () can be calculated from wup based on elastic analysis.
8Pe
w up =
L2
5wup L4
= (3-2.7)
384EI
b) For Singly Harped Tendon

P P
Wup
M
Figure 3-2.6 Simply supported beam with singly harped tendon
The moment at the centre due to the upward thrust (Wup) is given by the following
equation. It is equated to the moment due to the eccentricity of the tendon. As before,
the upward thrust and the deflection can be calculated.
Wup L
M= = Pe
4
4Pe
Wup =
L
Wup L3
= (3-2.8)
48EI

c) For Doubly Harped Tendon

P P
aL W Wup
up
M
Figure 3-2.7 Simply supported beam with doubly harped tendon
The moment at the centre due to the upward thrusts (Wup) is given by the following
equation. It is equated to the moment due to the eccentricity of the tendon. As before,
the upward thrust and the deflection can be calculated.
M = Wup aL = Pe
Pe
Wup =
aL
a ( 3 - 4a 2 ) Wup L3
= (3-2.9)
24EI
Example 3-2.1
A concrete beam prestressed with a parabolic tendon is shown in the figure. The
prestressing force applied is 1620 kN. The uniformly distributed load includes
the self weight. Compute the extreme fibre stress at the mid-span by applying
the three concepts. Draw the stress distribution across the section at mid-span.
45 kN/m 500
CGC
750
145
At end At mid-span
7.3m

Solution
a) Stress concept
Area of concrete, A = 500 750

= 375,000 mm2
Moment of inertia, I = (500 7503) / 12

= 1.758 1010 mm4
Bending moment at mid-span, M = (45 7.32) / 8

= 299.7 kNm
Top fibre stress

P Pe M
( fc )t = - + y top - y top
A I I
1620 10 1620 103 145
3
299.7 106
=- + 375 - 375
375 103 1.758 1010 1.758 1010
= - 4.32 + 5.01- 6.39
= - 5.7 N/mm2
Bottom fibre stress
P Pe M
( fc )b = - - y bot + y bot
A I I
162010 1620 103 145
3
299.7106
=- - 375 + 375
375103 1.7581010 1.7581010
= -4.32 - 5.01+ 6.39
= -2.9 N/mm2
b) Force concept
C
ec
z
e P
P C
Applied moment M = 299.7 kN-m

Lever arm z =M/P

= 299.7 103 / 1620
= 185 mm
Eccentricity of C ec =ze
= 185 145
= 40 mm
Top fibre stress
Cec
C
( fc )t = - - y top
A I
1620 103 1620 103 40
=- - 375
375 103 1.758 1010
= -4.32 -1.38
= -5.7 N/mm2
Bottom fibre stress

C Cec
( fc )b = - + y bot
A I
1620 103 1620 103 40
=- + 375
375 103 1.758 1010
= -4.32 +1.38
= -2.9 N/mm2
c) Load balancing method
Effective upward load, wup = 8Pe / L2

= 8 1620 103 145 / 73002
= 35.3 kN/m
Residual load wres = 45 35.3

= 9.7 kN/m
Residual bending moment Mres = 9.7 7.32 / 8
= 64.6 kNm

Residual bending stress (fc)res = 64.6 106 375 / 1.7581010

= 1.38 N/mm2
Total top fibre stress (fc)t = P/A (fc)res
= 4.32 1.38
= 5.7 N/mm2
Total bottom fibre stress (fc)b = P/A + (fc)res

= 4.32 + 1.38
= 2.9 N/mm2
The resultant stress distribution at mid-span is shown below.

5.7 N/mm2
2.9 N/mm2

3.3 Analysis of Members under Flexure (Part II)

Cracking Moment
Kern Points
Pressure Line
Introduction
The analysis of flexural members under service loads involves the calculation of the
following quantities.
a) Cracking moment.
b) Location of kern points.
c) Location of pressure line.
The following material explains each one of them.
3.2.1 Cracking Moment
The cracking moment (Mcr) is defined as the moment due to external loads at which the
first crack occurs in a prestressed flexural member. Considering the variability in stress
at the occurrence of the first crack, the evaluated cracking moment is an estimate.
Nevertheless, the evaluation of cracking moment is important in the analysis of
prestressed members.
Based on the allowable tensile stress the prestress members are classified into three
types as per IS:1343 - 1980. The types are explained in Section 1.2, Advantages and
Types of Prestressing. For Type 1 (full prestressing) and Type 2 (limited prestressing)
members, cracking is not allowed under service loads. Hence, it is imperative to check
that the cracking moment is greater than the moment due to service loads. This is
satisfied when the stress at the edge due to service loads is less than the modulus of
rupture.
The modulus of rupture is the stress at the bottom edge of a simply supported beam
corresponding to the cracking moment (Mcr). The modulus of rupture is a measure of
the flexural tensile strength of concrete. It is measured by testing beams under 2 point

loading (also called 4 point loading including the reactions or middle third loading). The
modulus of rupture (fcr) is expressed in terms of the characteristic compressive strength
(fck) of concrete by the following equation (IS:456 - 2000). Here, fcr and fck are in N/mm2.
fcr = 0.7 fck (3-3.1)
The following sketch shows the internal forces and the resultant stress profile at the
instant of cracking.
Mcr
CGC
yb Pe
fcr
Internal forces in Resultant stress
concrete profile
Figure 3-3.1 Internal forces and resultant stress profile at cracking
The stress at the edge can be calculated based on the stress concept as follows. The
cracking moment (Mcr) can be evaluated by transposing the terms.
Pe Pe ey b Mcr y b
- - + = fcr
A I I
Mcr y b P P ey
or, = fcr + e + e b
I A I
f I PI
or, Mcr = cr + e + Pe e (3-3.2)
y b Ay b
The above equation expresses Mcr in terms of the section and material properties and
prestressing variables.
3.2.2 Kern Points
When the resultant compression (C) is located within a specific zone of a section of a
beam, tensile stresses are not generated. This zone is called the kern zone of a section.
For a section symmetric about a vertical axis, the kern zone is within the levels of the
upper and lower kern points. When the resultant compression (C) under service loads
is located at the upper kern point, the stress at the bottom edge is zero. Similarly, when
C at transfer of prestress is located at the bottom kern point, the stress at the upper

edge is zero. The levels of the upper and lower kern points from CGC are denoted as kt
and kb, respectively.
Based on the stress concept, the stress at the bottom edge corresponding to C at the
upper kern point, is equated to zero. The following sketch shows the location of C and
the resultant stress profile.
CGC C kt
yb
Location of Resultant stress

resultant profile
compression
Figure 3-3.2 Resultant stress profile when compression is at upper kern point
The value of kt can be calculated by equating the stress at the bottom to zero as follows.
C Ckt y b
- + =0
A I
C Ck y
or, - + t 2 b = 0
A Ar
r2
or, kt = (3-3.3)
yb
The above equation expresses the location of upper kern point in terms of the section
properties. Here, r is the radius of gyration and yb is the distance of the bottom edge
from CGC.
Similar to the calculation of kt, the location of the bottom kern point can be calculated by
equating the stress at the top edge to zero. The following sketch shows the location of
C and the resultant stress profile.
yt
CGC C kb
Location of Resultant stress

resultant profile
compression
Figure 3-3.3 Resultant stress profile when compression is at lower kern point

C Ck b y t
- + =0
A I
C Ck y
or, - + b 2 t = 0
A Ar
r2 (3-3.4)
or, k b =
yt
Here, yt is the distance of the top edge from CGC.
Cracking Moment using Kern Points

The kern points can be used to determine the cracking moment (Mcr). The cracking
moment is slightly greater than the moment causing zero stress at the bottom. C is
located above kt to cause a tensile stress fcr at the bottom. The incremental moment is
fcr I/yb. The following sketch shows the shift in C outside the kern to cause cracking and
the corresponding stress profiles.
z
C
ec kt
CGC + =
e
CGS
fcr fcr
Location of
Resultant stress
resultant
profile
compression
Figure 3-3.4 Resultant stress profile at cracking of the bottom edge
The cracking moment can be expressed as the product of the compression and the
lever arm. The lever arm is the sum of the eccentricity of the CGS (e) and the
eccentricity of the compression (ec). The later is the sum of kt and z, the shift of C
outside the kern.
Mcr = C ( e + ec )
= C ( e + kt + z )
fcr I (3-3.5)
or, Mcr = C ( e + kt ) +
yb
Substituting C = Pe, kt = r2/yb and r2 = I/A, the above equation becomes same as the
previous expression of Mcr.

r 2 fcr I
Mcr =Pe +e+
yb yb
f I PI
or, Mcr = cr + e +Pee (3-3.6)
yb Ayb
3.2.3 Pressure Line
The pressure line in a beam is the locus of the resultant compression (C) along the
length. It is also called the thrust line or C-line. It is used to check whether C at
transfer and under service loads is falling within the kern zone of the section. The
eccentricity of the pressure line (ec) from CGC should be less than kb or kt to ensure C
in the kern zone.
The pressure line can be located from the lever arm (z) and eccentricity of CGS (e) as
follows. The lever arm is the distance by which C shifts away from T due to the moment.
Subtracting e from z provides the eccentricity of C (ec) with respect to CGC. The
variation of ec along length of the beam provides the pressure line.
M
z=
C
ec = z - e (3-3.7)
A positive value of ec implies that C acts above the CGC and vice-versa. If ec is
negative and the numerical value is greater than kb (that is |ec| > kb), C lies below the
lower kern point and tension is generated at the top of the member.
If ec> kt, then C lies above the upper kern point and tension is generated at the bottom
of the member.
Pressure Line at Transfer

The pressure line is calculated from the moment due to the self weight. The following
sketch shows that the pressure line for a simply supported beam gets shifted from the
CGS with increasing moment towards the centre of the span.

Pressure line CL
CGC kt
CGS kb
Figure 3-3.4 Pressure line at transfer
Pressure Line under Service Loads

The pressure line is calculated from the moment due to the service loads. The following
sketch shows that the pressure line for a simply supported beam gets further shifted
from the CGS at the centre of the span with increased moment under service condition.
Pressure line
CL
CGC kt
CGS kb
Figure 3-3.4 Pressure line under service loads
Limiting Zone
For fully prestressed members (Type 1), tension is not allowed under service conditions.
If tension is also not allowed at transfer, C always lies within the kern zone. The limiting
zone is defined as the zone for placing the CGS of the tendons such that C always lies
within the kern zone.
For limited prestressed members (Type 2 and Type 3), tension is allowed at transfer
and under service conditions. The limiting zone is defined as the zone for placing the
CGS such that the tensile stresses in the extreme edges are within the allowable values.
The following figure shows the limiting zone (as the shaded region) for a simply
supported beam subjected to uniformly distributed load.

CL
Locus of emin
CGC
Locus of emax
Figure 3-3.4 Limiting zone for a simply supported beam
The determination of limiting zone is given in Section 4.4, Design of Sections for Flexure
(Part III).
Example 3-3.1
For the post-tensioned beam with a flanged section as shown, the profile of the
CGS is parabolic, with no eccentricity at the ends. The live load moment due to
service loads at mid-span (MLL) is 648 kNm. The prestress after transfer (P0) is
1600 kN. Assume 15% loss at service. Grade of concrete is M30.
CGC
18.0m
500
200
150
1000 Values in mm.
CGS
150 200
250
Cross-section at mid-span

Evaluate the following quantities.

a) Kern levels
b) Cracking moment
c) Location of pressure line at mid-span at transfer and at service.
d) The stresses at the top and bottom fibres at transfer and at service.
Compare the stresses with the following allowable stresses at transfer and at
service.
For compression, fcc,all = 18.0 N/mm2
For tension, fct,all = 1.5 N/mm2.
Solution
Calculation of geometric properties
The section is divided into three rectangles for the computation of the geometric
properties. The centroid of each rectangle is located from the soffit.
1
yt
+ CGC
2
900 Values in mm.
y yb
500
3
100
Area of the section
Area of 1 = A1 = 500 200

= 100,000 mm2
Area of 2 = A2 = 600 150

= 90,000 mm2
Area of 3 = A3 = 250 200

= 50,000 mm2

A = A1 + A2 + A3
= 240,000 mm2
Location of CGC from the soffit
A1 900 + A2 500 + A3 100

y=
A
= 583.3 mm
Therefore,
y b = 583.3 mm
y t = 1000.0 - 583.3
= 416.7 mm
Eccentricity of CGS at mid-span
e = y -150
= 583.3 -150
= 433.3 mm
Moment of inertia of 1 about axis through CGC
1
I1 = 500 2003 + A1 (900 - 583.3)2
12
= 1.0361010 mm4

1
I2 = 150 6003 + A2 (583.3 - 500)2
12
= 3.32109 mm4
1
I3 = 250 2003 + A3 (583.3 -100)2
12
= 1.1841010 mm4

Moment of inertia of the section
I = I1 + I2 + I3
= (1.036 + 0.336 +1.184)1010
= 2.5521010 mm4
Square of the radius of gyration
I
r2 =
A
2.5521010
=
240,000
= 1.063105 mm2
a) Kern levels of the section
r2 r2
kt = kb =
yb yt
1.063105 1.063 105
= =
583.3 416.7
= 182.2mm = 255.1mm
CGC 182.2
+
255.1 Kern zone
CGS
Values in mm.
Calculation of moment due to self weight (MDL).

2
m2
w DL = 24.0 kN/m3 240,000 mm2 3
1
10 mm
2
= 5.76 kN/m
w DL L2
MDL =
8
5.76 18.02
=
8
= 233.3 kNm

b) Calculation of cracking moment
Modulus of rupture fcr = 0.7 fck

= 0.7 30
= 3.83kN/mm2
fcr I PeI
Mcr = + + Pe e
y b Ay b
3.83 2.5521010 0.851600103 2.5521010
= +
583.3 240103 583.3
+ 0.81600 103 433.3 Nmm
= 167.6 + 247.9 + 554.6
= 970.1 kNm
Live load moment corresponding to cracking
MLL cr = 970.1- 233.3

= 736.8 kNm
Since the given live load moment (648.0 kNm) is less than the above value, the section
is uncracked.
The moment of inertia of the gross section can be used for computation of
stresses.
c) Calculation of location of pressure line at mid-span
At transfer
MDL
z=
C
233.3103 ec = z - e
=
1600 = 145.8 - 433.3
= 145.8 mm = - 287.5mm
Since ec is negative, the pressure line is below CGC.
Since the magnitude of ec is greater than kb, there is tension at the top.

CGC
+
287.5 Kern zone
x
Location of pressure line
CGS
Value in mm.
At service
MDL+LL
z=
C
ec = z - e
(233.3 + 648.0)103
= = 648.0 - 433.3
0.851600
= 214.7mm
= 648.0 mm
Since ec is positive, the pressure line is above CGC.
Since the magnitude of ec is greater than kt, there is tension at the bottom.
x
Location of pressure line
214.7
CGC
+
Kern zone
CGS
Value in mm.
d) Calculation of stresses
The stress is given as follows.
P Pey My
f =-
A I I
M
CGC + + =
P
P/A Pey/ I My/ I Resultant

stress
profile
Calculation of stresses at transfer (P = P0)

P0 1600 103
=-
A 240 103
= -6.67 N/mm2
Stress at the top fibre
P0ey t 1600103 433.3 416.7

=
I 2.5521010
= 11.32 N/mm2
MDL y t 233.3106 416.7

=-
I 2.5521010
= -3.81N/mm2
fc = - 6.67 +11.32 - 3.81

t
= 0.84 N/mm2
Stress at the bottom fibre
P0ey b 1600103 433.3 583.3

=-
I 2.5521010
= -15.85 N/mm2
MDL y b 233.3106 583.3

=
I 2.5521010
= 5.33 N/mm2
fc = - 6.67 -15.85 + 5.33

b
= - 17.19 N/mm2
Calculation of stresses at service (P = Pe)

Pe P
= 0.85 0
A A
= - 5.67 N/mm2
Stress at the top fibre

Pf ey t
= 0.85 11.32
I
= 9.62
MLL y t 648.0106 416.7

=-
I 2.5521010
= - 10.58 N/mm2
fc = - 5.67 + 9.62 - 3.81-10.58

t
= -10.44 N/mm2
Stress at the bottom fibre

Pf ey b
= - 0.85 15.85
I
= - 13.47 N/mm2
MLL y b 648.0106 583.3

=
I 2.5521010
= 14.81N/mm2
fc = - 5.67 - 13.47 + 5.33 +14.81

b
= 1.0 N/mm2
The stress profiles are shown.

0.84 10.44
Numeric values
in N/mm2.
17.19 1.0
At transfer At service
The allowable stresses are as follows.

For compression, fc,comp = 18.0 N/mm2
For tension, fc,tens = 1.5 N/mm2.
Thus, the stresses are within the allowable limits.

3.4 Analysis of Members under Flexure (Part III)

Analysis for Ultimate Strength
Analysis of a Rectangular Section
3.4.1 Analysis for Ultimate Strength
Introduction
A prestressed member usually remains uncracked under service loads. The analysis
under service loads assumes the material to be linear elastic.
After cracking, the behaviour of a prestressed member is similar to a non-prestressed
reinforced concrete member. With increasing load, the stress versus strain behaviour of
concrete becomes non-linear. Close to the yielding of the prestressing steel, the stress
versus strain behaviour of steel also becomes non-linear.
The analysis of a prestressed member for ultimate strength is similar to that of a

reinforced concrete member. The analysis aims to calculate the ultimate moment
capacity (ultimate moment of resistance). The capacity is compared with the demand at
ultimate loads.
There is an inconsistency in the traditional analysis at the ultimate state. The force
demand is calculated based on elastic analysis, with superposition for the different load
cases using the load factors. But the capacity is calculated based on the non-linear limit
state analysis. The inconsistency is justified by the following arguments.
1) The moment versus curvature relationship is almost linear till the yielding of the
steel. The moment versus curvature relationship is also referred to as the
behaviour and is explained in Section 3.6, Analysis of Member under Flexure
(Part V).
2) The moment at yield is only slightly lower than the ultimate moment capacity.
Hence the behaviour is practically linear for most of the range of the moment.
3) The calculated moment demand for a load case based on elastic analysis is well
within the moment at yield. Hence, superposition for the load cases is applied to
find out the moment demand under combined loads.

Of course, superposition cannot be used to calculate the deflection under combined

loads.
Variation of Stress in Prestressing Steel

In non-prestressed reinforced concrete members, the tension and consequently the
stress in steel increase almost proportionately with increasing moment till yielding. The
lever arm between the resultant compression and tension remains almost constant. In
prestressed concrete members, the tension and consequently the stress in prestressing
steel increase slightly with increasing moment till cracking of concrete. The increase in
moment changes the lever arm significantly. This is explained in Section 3.2, Analysis of
Member under Flexure (Part I). After cracking, the stress in prestressing steel
increases rapidly with moment.
The following sketch explains the variations of the stress in prestressing steel (fp) with
increasing load. The variations are shown for bonded and unbonded tendons. After the
prestress is transferred while the member is supported at the ends, the stress will tend
to increase from the value after losses (fp0) due to the moment under self weight.
Subsequently the stress will tend to drop due to the time dependent losses such as from
creep, shrinkage and relaxation. The losses of prestress are covered in Section 2.3,
Losses in Prestress (Part III). The effective prestress after time dependent losses is
denoted as fpe.
Due to the moment under service loads, the stress in the prestressing steel will slightly
increase from fpe. The increase is more at the section of maximum moment in a bonded
tendon as compared to the increase in average stress for an unbonded tendon. The
stress in a bonded tendon is not uniform along the length. Usually the increase in
stress is neglected in the calculations under service loads. If the loads are further
increased, the stress increases slightly till cracking.
After cracking, there is a jump of the stress in the prestressing steel. Beyond that, the
stress increases rapidly with moment till the ultimate load. At ultimate, the stress is
represented as fpu. Similar to the observation for pre-cracking, the average stress in an
unbonded tendon is less than the stress at the section of maximum moment for a
bonded tendon.

fp
fpu
Bonded
fp0
Losses
fpe Unbonded
Load
Self Cracking
weight load
Service Ultimate
load load
Figure 3-4.1 Variation of stress in prestressing steel
The above sketch assumes that the section is failing in flexure. Other types of failure
are not considered.
Conditions at Ultimate Limit State

In the limit states method of analysis, the limit state of collapse (ultimate state) of a
member under flexure is defined as the state when the extreme concrete compressive
strain reaches a value of 0.0035. At ultimate, let the extreme concrete compressive
strain be denoted as cu. Thus, cu = 0.0035.
Depending on the amount of prestressing steel, a section can be under-reinforced or

over-reinforced. For an under- reinforced section, the amount of prestressing steel is
less and the steel yields before the extreme concrete strain reaches 0.0035. For an
over-reinforced section, the amount of steel is high and the steel does not yield at
ultimate. The transition situation is called a balanced condition. The strain profiles
across the depths of prestressed flexural members (up to the depth of CGS) for the
three situations are shown below.

cu cu cu
p p p
pu > pu,bal pu = pu,bal pu < pu,bal

Under-reinforced Balanced Over-reinforced
Figure 3-4.2 Strain profiles along the depths of three prestressed members

pu = strain in the prestressing steel at the level of CGS at ultimate
condition
pu,bal = strain in the steel for a balanced section.
The strain difference (p) is the strain in the prestressed tendons when the adjacent
concrete has zero strain (c = 0). The strain difference gets locked during the transfer of
prestress. The value can be determined as follows.
For pre-tensioned members, the strain difference gets locked when the tendons are cut.
The strain difference at that instant is given as follows.
p = pi 0 (3-4.1a)
Here,
pi = strain in tendons just before transfer
c = strain in concrete is zero.
For post-tensioned members, the strain difference gets locked when the tendons are
anchored. The strain difference at that instant is given as follows.
p = p0 c0 (3-4.1b)
Here,
p0 = strain in tendons due to P0, the prestress after transfer
c0 = strain in concrete due to P0.
In general at any load stage,

p = pe ce (3-4.1c)

Here,
pe = strain in tendons due to Pe, the prestress at service
ce = strain in concrete due to Pe.
As mentioned under material properties, the prestressing steel does not have a definite
yield point. The 0.2% proof stress is defined when the steel reaches an inelastic strain
of 0.2%. Hence, unlike reinforced concrete, the transition from under-reinforced to over-
reinforced section is gradual and there is no definite balanced condition. IS:1343 - 1980
does not explicitly enforce an under-reinforced section. But the IRS Concrete Bridge
Code requires that the strain in the outermost tendon should not be less than the
following.
0.87fpk
+ 0.005
Ep
The above value can be considered to be the strain in the steel at balanced condition.
Assumptions for Analysis

The analysis of members under flexure for ultimate strength considers the following.
1) Plane sections perpendicular to the axis of the member remain plane till the
ultimate state.
2) Perfect bond is retained between concrete and prestressing steel for bonded
tendons.
3) Tension in concrete is neglected.
4) The design stress versus strain curves of concrete and steel are considered.
The methods of analysis will be presented for three types of sections.

1) Rectangular section: A rectangular section is easy to cast, but it is not an efficient
section.
2) Flanged section: A precast flanged section, with flanges either at top or bottom
needs costlier formwork. But the section is efficient in flexure. A flanged section
can also be made of precast web and cast-in-place slab.
3) Partially prestressed section: A section in a member containing both prestressed
and non-prestressed reinforcement.

3.4.4 Analysis of a Rectangular Section
The following sketch shows the beam cross section, strain profile, stress diagram and
force couple at the ultimate state.
b cu=0.0035 0.447fck
0.42xu
xu Cu
d
Ap
Tu
p fpu
pu
Cross-section Strain Stress Force
Figure 3-4.3 Sketches for analysis of a rectangular section
The variables in the above figure are explained.

b = breadth of the section
d = depth of the centroid of prestressing steel (CGS)
Ap = area of the prestressing steel
p = strain difference
xu = depth of the neutral axis at ultimate
pu = strain in prestressing steel at the level of CGS at ultimate
fpu = stress in prestressing steel at ultimate
The stress block in concrete is derived from the constitutive relationship for concrete.
The relationship is explained in Section 1.6, Concrete (Part II). The compressive force
in concrete can be calculated by integrating the stress block along the depth. The
stress in the tendon is calculated from the constitutive relationship for prestressing steel.
The relationship is explained in Section 1.7, Prestressing Steel.
In the force diagram,

Cu = 0.36fck xu b (3-4.2)
Tu = Ap fpu (3-4.3)
The strengths of the materials are denoted by the following symbols.
fck = characteristic compressive strength of concrete
fpk = characteristic tensile strength of prestressing steel

For analysis of a prestressed section, three principles of mechanics are used. First, the
equilibrium relates the external applied forces with the internal forces. Second, the
compatibility condition relates the strain in the prestressing steel with the strain in
concrete at the level of CGS. This also considers the first two assumptions given in the
previous section. The third principle involves the constitutive relationships of the
materials.
Based on the above principles of mechanics, the following equations are derived.
1) Equations of equilibrium
The first equation states that the resultant axial force is zero. This means that the
compression and the tension in the force couple balance each other.
F = 0
Tu = Cu
Ap fpu = 0.36fck xu b (3-4.4)
The second equation relates the ultimate moment capacity (MuR) with the internal
couple in the force diagram.
MuR = Tu (d - 0.42 xu )
= Ap fpu ( d - 0.42 xu )
(3-4.5)
2) Equation of compatibility
The depth of the neutral axis is related to the depth of CGS by the similarity of the
triangles in the strain diagram.
xu 0.0035
= (3-4.6)
d 0.0035 + pu - p
3) Constitutive relationships
a) Concrete
The constitutive relationship for concrete is considered in the expression Cu =
0.36fckxub. This is based on the area under the design stress-strain curve for concrete
under compression.

b) Prestressing steel
fpu = F ( pu )
(3-4.7)
The function F(pu) represents the design stress-strain curve for prestressing steel under
tension.
The known variables in an analysis are: b, d, Ap, p, fck, fpk.

The unknown quantities are: xu, MuR, pu, fpu.
The objective of the analysis is to find out MuR, the ultimate moment capacity.
The simultaneous equations 3-4.1 to 3-4.7 can be solved iteratively. This procedure of
analysis is called the strain compatibility method. The steps are as follows.
1) Assume xu .
2) Calculate pu by rearranging the terms of Eqn. 3-4.6.
3) Calculate fpu from Eqn. 3-4.7.
4) Calculate Tu from Eqn. 3-4.3.
5) Calculate Cu from Eqn. 3-4.2.
If Eqn. 3-4.4 (Tu = Cu) is not satisfied, change xu.
If Tu < Cu decrease xu. If Tu > Cu increase xu.
6) Calculate MuR from Eqn. 3-4.5.
The capacity MuR can be compared with the demand under ultimate loads.
In the strain compatibility method, the difficult step is to calculate xu and fpu. IS:1343 -
1980 allows to calculate these variables approximately from Table 11, Appendix B,
based on the amount of prestressing steel. The later is expressed as a prestressed
reinforcement index p.
Ap fpk
p = (3-4.8)
bdfck
Table 11 is reproduced as Table 3-4.1 which is applicable for pre-tensioned and bonded
post-tensioned beams. The values of fpu and xu are given as fpu/(0.87fpk) and xu/d,
respectively.

Table 3-4.1 Values of xu and fpu for pre-tensioned and bonded post-tensioned
rectangular beams (Table 11, IS:1343 - 1980)
fpu/(0.87fpk) xu/d
p Pre-tensioned Bonded post- Pre-tensioned Bonded post-
tensioned tensioned
0.025 1.0 1.0 0.054 0.054
0.05 1.0 1.0 0.109 0.109
0.10 1.0 1.0 0.217 0.217
0.15 1.0 1.0 0.326 0.316
0.20 1.0 0.95 0.435 0.414
0.25 1.0 0.90 0.542 0.488
0.30 1.0 0.85 0.655 0.558
0.40 0.9 0.75 0.783 0.653
The values of fpu/(0.87fpk) and xu/d from Table 3-4.1 are plotted in Figures 3-4.4 and 3-
4.5, respectively. It is observed that with increase in p, fpu reduces (beyond certain
values of p) and xu increases. This is expected because with increase in the amount
and strength of the steel, the stress in steel drops and the depth of the neutral axis
increases to maintain equilibrium.
1.2
0.8
fpu / 0.87 fpk
0.6
0.4
0.2
0
0 0.1 0.2 0.3 0.4
p
Pre-tensioned Post-tensioned (bonded)
Figure 3-4.4 Variation of fpu with respect to p (Table 3-4.1)

0.9
0.6
xu / d
0.3
0
0 0.1 0.2 0.3 0.4
Pre-tensioned Post-tensioned (bonded)
Figure 3-4.5 Variation of xu with respect to p (Table 3-4.1)
Thus given the value of p for a section, the values of fpu and xu can be approximately
calculated from the above tables.
Example 3-4.1
A prestressed concrete beam produced by pre-tensioning method has a

rectangular cross-section of 100 mm 160 mm (b h). It is prestressed with 10
numbers of straight 2.5 mm diameter wires. Each wire is stressed up to a load of
6.8 kN. The design load versus strain curve for each wire is given in a tabular
form. The grade of concrete is M 40. The value of p is 0.0073.
Estimate the ultimate flexural strength of the member by the strain compatibility
method.

100
160
CGC
40
Values in mm
Cross-section of member
Design load (P) versus strain (p) values for the prestressing wire are given for
the range under consideration.
p P (kN)
0.006 5.4
0.008 7.6
0.010 9.0
0.012 10.0
0.014 10.7
Solution
Strain difference
p = 0.0073
The effective depth of the CGS (d ) is 120 mm.
The strain compatibility method is shown in a tabular form. Here,

Pu = load in a single wire obtained from the table
Tu = 10 Pu , for the ten wires.

xu xu/d pup pu Pu Tu Cu Checking

(mm) (kN) (kN) (kN)
(3-4.6) (Table) (3-4.2) (3-4.4)
60 0.5 0.0035 0.0108 9.4 94.0 86.4 Tu > Cu
65 0.54 0.0030 0.0103 9.1 91.0 93.6 Tu < Cu
63.5 0.53 0.0031 0.0104 9.15 91.5 91.4 Tu Cu
The ultimate flexural strength is given as follows.
MuR = Tu (d - 0.42xu )
= 91.5 (120.0- 0.4263.5) kNmm
= 8.5 kNm

3.5 Analysis of Members under Flexure (Part IV)

Analysis of a Flanged Section
3.5.1 Analysis of a Flanged Section
Introduction
A beam can have flanges for flexural efficiency. There can be several types of flanged
section.
1) A precast or cast-in-place flanged section, with flanges either at top or bottom or
at both top and bottom.
2) A composite flanged section is made of precast web and cast-in-place slab.
The following figures show different types of flanged sections.
T-section Double T-section
Single box section Double box section
L-section Inverted T-section I-girder

Figure 3-5.1 Examples of precast flanged sections

T T T T T T
Box section Composite beam-slab T-section

Figure 3-5.2 Examples of composite flanged sections
The analysis of a flanged section for ultimate strength is different from a rectangular
section when the flange is in compression. If the depth of the neutral axis from the edge
under compression is greater than the depth of the flange, then the section is treated as
a flanged section. In the following figure, the first strain profile shows that the depth of
the neutral axis (xu) is greater than the depth of the flange (Df). The section is treated
as a flanged section.
The second strain profile shows that xu is less than Df. In this situation, the section can
be treated as a rectangular section.
bf
Df xu
xu
d
Ap
bw
Cross-section Strain profile Strain profile
(xu > Df) (xu < Df)
Figure 3-5.3 Two possibilities of strain profile in a flanged section
The effective width or breadth of the flange (bf) is determined from the span of the beam,
breadth of the web (bw) and depth of the flange (Df) as per Clause 23.1.2, IS:456 - 2000.
Analysis of a Flanged Section

The following sketch shows the beam cross-section, strain profile, stress diagram and
force couples at the ultimate state. The following conditions are considered.
1) xu > Df : This requires an analysis for a flanged section.

2) Df (3/7) xu: This ensures that the compressive stress is constant at 0.447fck
along the depth of the flange.
bf 0.0035 0.447fck 0.5Df

Df 0.42xu Cuf
xu Cuw +
d
Ap
Tuw Tuf
fpu
p
bw pu
Figure 3-5.4 Sketches for analysis of a flanged section

bf = breadth of the flange
bw = breadth of the web
Df = depth of the flange
p = strain difference in the prestressing steel when strain in concrete is zero
The strain difference (p) is further explained in Section 3.4, Analysis of Member under
Flexure (Part III).
In the sketch, the tensile force is decomposed into two components. The first
component (Tuw) balances the compressive force carried by the web, including the
portion of the flange above web (Cuw). Thus Tuw= Cuw. The second component (Tuf)
balances the compressive force carried by the outstanding portion of the flange (Cuf).
Thus Tuf = Cuf.

The expressions of the forces are as follows.

Cuw = 0.36fck xu bw (3-5.1)
Cuf = 0.447fck (bf - bw )Df (3-5.2)
Tuw = Apw fpu (3-5.3)
Tuf = Apf fpu (3-5.4)

Apf = part of Ap that balances compression in the outstanding flanges
Apw = part of Ap that balances compression in the web
Based on the principles of mechanics (as explained under the Analysis of a Rectangular
Section in Section 3.4, Analysis of Member Under Flexure (Part III)), the following
equations are derived.
F = 0
Tu = Cu
Tuw +Tuf = Cuw + Cuf
( Apw + Apf ) fpu = 0.36 fck xu bw + 0.447 fck ( bf - bw ) Df (3-5.5)
MuR = Tuw ( d - 0.42 xu ) +Tuf ( d - 0.5Df )

= Apw fpu ( d - 0.42 xu ) + Apf fpu ( d - 0.5Df ) (3-5.6)
From Tuf = Cuf and Eqns. (3-5.2) and (3-5.4), Apf is given as follows. The calculation of
Apw from Ap and Apf is also shown.

0.447fck ( bf - bw ) Df (3-5.7)
Apf =
fpu
Apw = Ap - Apf (3-5.8)
2) Equation of compatibility
The depth of the neutral axis is related to the depth of CGS by the similarity of the
triangles in the strain diagram.
xu 0.0035 (3-5.9)
=
d 0.0035 + pu - p
a) Concrete
The constitutive relationship for concrete is considered in the expressions of Cuw and
Cuf. This is based on the area under the design stress-strain curve for concrete under
compression.
fpu = F ( pu ) (3-5.10)
The function F(pu) represents the design stress-strain curve for the type of prestressing
steel used.
The known variables in an analysis are: bf, bw, Df, d, Ap, p, fck and fpk.
The unknown quantities are: Apf, Apw, MuR, xu, pu and fpu.
The objective of the analysis is to find out MuR , the ultimate moment capacity. The
simultaneous equations 3-5.1 to 3-5.10 can be solved iteratively.
The steps of the strain compatibility method are as follows.
1) Assume xu = Df.
2) The calculations are similar to a rectangular section, with b = bf.
3) If Tu > Cu, increase xu. Treat the section as a flanged section.
4) Calculate pu from Eqn. (3-5.9).
5) Calculate fpu from Eqn. (3-5.10).
6) Calculate Apf and Apw from Eqn. (3-5.7) and Eqn. (3-5.8), respectively.

7) Calculate Cuw, Cuf, Tuw and Tuf from Eqns. (3-5.1) to (3-5.4). If Eqn. (3-5.5)
(Tu = Cu) is not satisfied, iterate with a new value of xu, till convergence.
8) Calculate MuR from Eqn. (3-5.6).
In the strain compatibility method, the difficult step is to calculate xu and fpu. Similar to
the rectangular section, an approximate analysis can be done based on Table 11 and
Table 12, Appendix B, IS:1343-1980. The tables are reproduced in Table 3-4.1 and
Table 3-4.2, respectively, in Section 3.4, Analysis of Member under Flexure (Part III).
The values of xu and fpu are available in terms of a reinforcement index pw.
A f (3-5.11)
pw = pw pk
bw dfck
Note that the index is calculated based on Apw instead of Ap. The calculation of Apw is
from Eqn. (3-5.8). But Apf depends on fpu, which is unknown. Hence, an iterative
procedure is required.
The steps are as follows.

1) Assume fpu = 0.87fpk.
2) Calculate Apf and Apw from Eqn. (3-5.7) and Eqn. (3-5.8), respectively.
3) Calculate pw.
4) Calculate fpu from Table 11 or Table 12.
Compare the calculated value of fpu with the assumed value. Repeat steps 1 to 4 till
convergence.
5) Calculate MuR.
If Df > (3/7) xu, the flange depth is larger than the depth of constant compressive stress.
An equivalent depth of the flange is defined as follows.
yf = 0.15xu + 0.65Df (3-5.12)
The equivalent depth yf is substituted for Df in the expression of MuR.

Example 3-5.1
A bonded post-tensioned concrete beam has a flanged cross-section as shown.

It is prestressed with tendons of area 1750 mm2 and effective prestress of 1100
N/mm2. The tensile strength of the tendon is 1860 N/mm2. The grade of concrete
is M60.
Estimate the ultimate flexural strength of the member by the approximate method
of IS:1343 - 1980.
460
175
140
550
900
175 115
460
Values are in mm.
Cross-section at mid-span
Solution
Effective depth d = 900 115

= 785 mm
Assume xu = Df = 175 mm. Treat as a rectangular section, with b = bf = 460 mm.
Reinforcement index
AP fPk
P =
bdfck
1750 1860
=
460 785 60
= 0.15

Tu = Apfpu
= 1750 1618
= 2831.5 kN
Cu = 0.36fckxubf
= 0.36 60 175 460
= 1738.8 kN
Tu > Cu. Hence xu > Df

Treat as a flanged section
Assume fpu = 0.87fpk

= 1618 N/mm2
Calculate Apf and Apw

0.447fck (bf - bw )Df
Apf =
fpu
0.447 60 (460 -140)175
=
1618
2
= 934 mm
Apw = 1750 - 934

= 816 mm2
Reinforcement index
Apw fpk
pw =
bw dfck
8161860
=
140 785 60
= 0.23
From Table 11,
fpu
= 0.92
0.87fpk
fpu = 0.920.87 1860
= 1489 N/mm2

2nd iteration
fpu = 1489 N/mm2

0.447 60 (460 - 140)175
Apf =
1489
2
= 1015 mm
Apw = 1750 -1015

= 735 mm2
Reinforcement index
735 1860
pw =
140 785 60
= 0.21
From Table 11,

fpu
= 0.94
0.87fpk
fpu = 0.94 0.87 1860
= 1521N/mm2
3rd iteration
fpu = 1521N/mm2

0.447 60 (460 - 140)175
Apf =
1521
2
= 994 mm
Apw = 1750 - 994

= 756 mm2
Reinforcement index
756 1860
pw =
140 785 60
= 0.21
The value of wpw is same as after 2nd iteration. Hence, the values of fpu, Apf and Apw
have converged.

Ultimate flexural strength
MuR = Tuw ( d - 0.42 xu ) + Tuf ( d - 0.5Df )
Tuw (d - 0.42 xu ) = Apw fpu (d - 0.42 xu )

= 7561521(785 - 0.42337)
= 739.9 kNm
Tuf (d - 0.5Df ) = Apf fpu (d - 0.5Df )

= 9941521(785 - 0.5175)
= 1054.5 kNm
The ultimate flexural strength is given as follows.
MuR = 1054.5 + 739.9

= 1794.4 kNm

3.6 Analysis of Members under Flexure (Part V)

Analysis of Partially Prestressed Section
Analysis of Unbonded Post-tensioned Beam
Analysis of Behaviour
3.6.1 Analysis of Partially Prestressed Section
Introduction
The analyses that are presented in the earlier sections, are for members which do not
have any conventional non-prestressed reinforcement. Usually conventional
reinforcement is provided in addition to the prestressing steel. When this reinforcement
is considered in the flexural capacity, the section is termed as a partially prestressed
section.
The reasons for using a partially prestressed section are as follows.

1) The section is economical.
2) The cambering is less compared to an equivalent section without conventional
reinforcement.
3) The ductility is more in a partially prestressed section.
4) Any reversal of moments (for example, due to earthquake) is not detrimental as
compared to an equivalent section without conventional reinforcement.
Analysis
A partially prestressed section can be either rectangular or flanged. A section can be
doubly reinforced with reinforcement near the compression face.
Here, the equations for a doubly reinforced rectangular section are given.
The following sketch shows the beam cross section, strain profile, stress diagram and
force couples at the ultimate state.

b 0.0035 0.447fck
As d 0.42xu Cs
s fs
dp xu Cc
d
Ap
fpu Tp
As s
p fs Ts
pu

Figure 3-6.1 Sketches for analysis of a partially prestressed section

d = depth of the centroid of the reinforcing steel (tension side)
d = depth of the centroid of the reinforcing steel (compression side)
dp = depth of the centroid of prestressing steel (CGS)
As = area of the reinforcing steel (tension side)
As = area of the reinforcing steel (compression side)
p = strain difference in the prestressing steel when strain in concrete is zero
s = strain in reinforcing steel (tension side) at ultimate
s = strain in reinforcing steel (compression side) at ultimate
fs = stress in reinforcing steel (tension side) at ultimate
fs = stress in reinforcing steel (compression side) at ultimate
fpu = stress in prestressing steel at ultimate.
The strain difference (p) is further explained in Section 3.4, Analysis of Member under
Flexure (Part III).


Cs = Asfs (3-6.1)
Cc = 0.36fckxub (3-6.2)
Tp = Apfpu (3-6.3)
Ts = Asfs (3-6.4)

fy = characteristic yield stress of reinforcing steel
Based on the principles of mechanics (as explained under the Analysis of a Rectangular
Section in Section 3.4, Analysis of Member Under Flexure (Part III)), the following
equations are derived.
F = 0
Tu = Cu
Tp +Ts = Cc + C' s
Ap fpu + As fs = 0.36fck xu b + As' fs' (3-6.5)
MuR Ap
= Ts ( d - d p ) + Cc ( d p - 0.42 xu ) + C' s ( d p - d' )
= As fs ( d - d p ) + 0.36fck xu b ( d p - 0.42 xu ) + A' s fs' ( d p - d' ) (3-6.6)
2) Equations of compatibility
For each layer of steel there is a compatibility equation. If there are distributed
reinforcing bars in several layers and the spacing between the layers is large, then the
use of compatibility equation for each layer is more accurate than the use of one
compatibility equation for the centroid of the layers. The following equations are
developed based on the similarity of the triangles in the strain diagram.

xu 0.0035 (3-6.7)
=
d p 0.0035 + pu - p
d - xu s (3-6.8)
=
xu 0.0035
xu - d' s'
= (3-6.9)
xu 0.0035
a) Concrete
The constitutive relationship for concrete is considered in the expressions of Cc. This is
based on the area under the design stress-strain curve for concrete under compression.
fpu = F1 ( pu ) (3-6.10)
c) Reinforcing steel
fs = F2 ( s ) (3-6.11)
f = F3 (
'
s
'
s ) (3-6.12)
For mild steel
fs' = 0.87fy (3-6.13)
fs = 0.87fy (3-6.14)
The known variables in an analysis are: b, d, d, dp, As, As, Ap, p, fck, fy and fpk.
The unknown quantities are: MuR, xu, s, s, pu, fs, fs and fpu.
The objective of the analysis is to find out MuR, the ultimate moment capacity.
The previous equations can be solved by the strain compatibility method as
discussed for the fully prestressed rectangular section.
1) Assume xu.
2) Calculate pu from Eqn. 3-6.7.
3) Calculate fpu from Eqn. 3-6.10.
4) Calculate Tp from Eqn. 3-6.3.
5) Calculate s from Eqn. 3-6.8.
6) Calculate fs from Eqn. 3-6.11.
7) Calculate Ts from Eqn. 3-6.4.
8) Calculate s from Eqn. 3-6.9.

9) Calculate fs from Eqn. 3-6.12.

10) Calculate Cs from Eqn. 3-6.1.
11) Calculate Cc from Eqn. 3-6.2.
If Eqn. 3-6.5 (Tu = Cu) is not satisfied, change xu.
If Tu < Cu decrease xu. If Tu > Cu increase xu.
12) Calculate MuR from Eqn. 3-6.6.
3.6.2 Analysis of Unbonded Post-tensioned Beam
In an unbonded post-tensioned beam, the ducts are not grouted. Hence, there is no
strain compatibility between the steel of the tendons and the concrete at a section. The
compatibility is in terms of deformation over the length of the member.
A sectional analysis is not possible. The analysis involves integrating the strain in
concrete to calculate the deformation over the length of the member.
The equation of compatibility is given as follows.
p = cp (3-6.15)
Here,
p = deformation of the tendon
cp = deformation of the concrete at the level of prestressing steel (CGS).
The change in stress in steel (fp) at ultimate is determined from p. The stress in steel
at ultimate is given by the sum of the effective prestress (fpe) and fp.
fpu = fpe + fp (3-6.16)
The value of fpu is less than that for a bonded tendon. The ultimate moment is given by
the following equation.
MuR = Ap fpu ( d - 0.42 xu ) (3-6.17)
The rigorous method of evaluating fpu, based on deformation compatibility, is difficult.

IS:1343 - 1980 allows to calculate fpu and xu approximately from Table 12, Appendix B,
based on the amount of prestressing steel. The later is expressed as the reinforcement
index p = Apfpk / bdfck. Table 12 is reproduced as Table 3-6.1 which is applicable for

unbonded post-tensioned beams. The values of fpu and xu are given as fpu/fpe and Xu/d,
respectively. The effective prestress (after the losses) in a tendon is represented as fpe.
Table 3-6.1 Values of xu and fpu for unbonded post-tensioned rectangular beams
(Table 12, IS:1343 - 1980)
fpu/fpe xu/d
p For values of L/d For values of L/d
30 20 10 30 20 10
0.025 1.23 1.34 1.45 0.10 0.10 0.10
0.05 1.21 1.32 1.45 0.16 0.16 0.18
0.10 1.18 1.26 1.45 0.30 0.32 0.36
0.15 1.14 1.20 1.36 0.44 0.46 0.52
0.20 1.11 1.16 1.27 0.56 0.58 0.64
The values of fpu/fpe and xu/d from Table 3-6.1 are plotted in Figures 3-6.2 and 3-6.3,
respectively. It is observed that with increase in p, fpu reduces and xu increases. This
is expected because with increase in the amount and strength in the steel, the stress in
steel drops and the depth of the neutral axis increases to maintain equilibrium.
1.5
l /d
30
fpu/fpe
1.25 20
10
1
0 0.05 0.1 0.15 0.2
Figure 3-6.2 Variation of fpu with respect to wp (Table 3-6.1)

0.6
l /d
0.4
30
xu / d
20
10
0.2
0
0 0.05 0.1 0.15 0.2
p
Figure 3-6.3 Variation of xu/d with respect to wp (Table 3-6.1)
Thus given the value of p for a section, the values of fpu and xu can be approximately
calculated from the above tables.
3.6.3 Analysis of Behaviour
The analysis of behaviour refers to the determination of the complete moment versus
curvature behaviour of the section. The analyses at transfer, under service loads and
for ultimate strength correspond to three instants in the above behaviour.
The curvature () is defined as the gradient of the strain profile.
c + cp
= (3-6.18)
d
Here,
c = extreme concrete compressive strain
cp = strain in concrete at the level of prestressing steel (CGS)
d = depth of the CGS.
The following sketch shows the curvature () in the strain profile.

b c

d
Ap
cp
Cross-section Strain profile

Figure 3-6.4 Definition of curvature
The analysis of behaviour involves the following three principles of mechanics.

1) Equilibrium of internal forces with the external loads at any point of the
behaviour. There are two equilibrium equations.
a) Force equilibrium equation
b) Moment equilibrium equation.
The internal forces in concrete and steel are evaluated based on the respective
strains, cross-sectional areas and the constitutive relationships.
2) Compatibility of the strains in concrete and in steel for bonded tendons. This
assumes a perfect bond between the two materials. For unbonded tendons, the
compatibility is in terms of deformation.
The relationships are developed based on the material properties.
The equilibrium and compatibility equations and the constitutive relationships can be
solved to develop the moment versus curvature curve for a section.
The following plot shows the curves for a prestressed section and a non-prestressed
section. The two sections are equivalent in their ultimate flexural strengths.

Moment
Ultimate strength
Service load level
Cracking load levels
Curvature
Prestressed section Non-prestressed section

Figure 3-6.5 Moment versus curvature curves
From the previous plot, the following can be inferred.

1) Prestressing increases the cracking load. This leads to the following benefits.
Reduction of steel corrosion
Increase in durability.
Full section is utilised
Higher moment of inertia (higher stiffness)
Less deformations (improved serviceability).
Increase in shear capacity.
2) Prestressing shifts the curve from the origin.
For the prestressed member, there is a negative curvature causing camber in
absence of external moment.
A certain amount of external moment is required to straighten the member.
3) For a given moment, the curvature of the prestressed member is smaller.
Prestressing reduces curvature at service loads.
4) For a given reverse moment, the curvature of the prestressed member is
larger.
Prestressing is detrimental for the response under reverse moment.
5) The ultimate strength of the prestressed member is lower.
Prestressing is detrimental under reverse moment.
6) For a partially prestressed section with the same ultimate strength, the
moment versus curvature curve will lie in between the curves for
prestressed and non-prestressed sections.

Ductility
The ductility is a measure of energy absorption. For beams, the curvature ductility () is
defined as
u (3-6.19)
=
y
Here,
u = curvature at ultimate
y = curvature at yield.
For prestressed beams, y can be defined corresponding to a plastic strain of 0.002 in
the prestressing tendons. It has been observed that the ductility of prestressed beams
is less than that in reinforced concrete beams.
In design of members for seismic forces, ductility is an important requirement. In

addition, seismic forces lead to reversal of moments near the supports of beams in a
moment resisting frame. Hence, prestressing of beams in a moment resisting frame is
not recommended in seismic areas.
Experimental Investigation
The behaviour of a beam and its ultimate strength can be determined by testing
prototype specimens. The tests can be conducted under static or dynamic loads.
Testing also helps to check the performance of the anchorage units.
The following photo shows the set-up for testing a prototype bridge girder.
Figure 3-6.6 Set-up for testing a bridge girder

4.1 Design of Members

Calculation of Demand
Design of Sections for Axial Tension
Introduction
The design of prestressed concrete members can be done by the limit states method as
given in Section 4 of IS:1343 - 1980.
First, the force demand in a member under the design loads is determined from a
structural analysis. A preliminary size of the member is assumed for analysis. Next, the
member is designed to meet the demand. If necessary, another cycle of analysis and
design is performed.
The following material explains the calculation of the demand in a member under the
design loads.
4.1.1 Calculation of Demand
In the limit states method, the design loads are calculated from the characteristics loads
by multiplying them with load factors (f). Several types of loads are considered to act
together under the selected load combinations. The load factors are included in the
load combinations as weightage factors.
The demand in a member for a particular type of load is obtained from the analysis of
the structure subjected to the characteristic value of the load.
The demands for the several load types are then combined under the load combinations,
based on the principle of superposition.
Characteristics Loads
For dead loads, a characteristic load is defined as the value which has a 95%
probability of not being exceeded during the life of the structure. This concept assumes
a normal distribution of the values of a particular dead load. In the following figure, the

shaded area above the characteristic value represents 5% probability of exceedance of

the load in the design life of the structure.
Frequency
5% probability
of exceedence
Mean Characteristic
value
Values of load
Figure 4-1.1 Idealised normal distribution for a dead load
For live load, wind load and earthquake load, a characteristic load is defined based on
an extreme value distribution. For example, the characteristic wind load is defined as
the value which has a 98% probability of not being exceeded during a year.
Frequency
2% probability
of exceedence
Characteristic
value
Annual maximum mean wind speed
Figure 4-1.2 Extreme value distribution
The characteristics loads can be obtained from IS:875 - 1987 (Code of Practice for
Design Loads for Buildings and Structures) and IS:1893 - 2002 (Criteria for Earthquake
Resistant Design of Structures) as follows.

Table 4-1.1 Codes covering information of loads

Type of load Code
Dead load (DL) IS:875 - 1987 Part 1
Live (imposed) load (LL) IS:875 - 1987 Part 2
Wind load (WL) IS:875 - 1987 Part 3
Snow Load (SL) IS:875 - 1987 Part 4
Earthquake load (EL) IS:1893 - 2002 Part 1
For special loads, there are some guidelines in IS: 875 - 1987, Part 5. In addition,
specialised literature may be referred to for these loads. The special loads are listed
below.
Temperature
Hydrostatic
Soil pressure
Fatigue
Accidental load
Impact and collision
Explosions
Fire
For special situations, the loads are determined from testing of prototype specimens.
Dynamic load tests, wind tunnel tests, shake table tests are some types of tests to
determine the loads on a structure. Finite element analysis is used to determine the
stresses due to concentrated forces and dynamic loads.
Load Factors and Load Combinations

The load factors and the combinations of the various types of loads are given in Table 5
of IS:1343 - 1980. The following are the combinations for the ultimate condition.
1.5 (DL + LL)
1.2 (DL + LL WL)
1.2 (DL + LL EL)
1.5 (DL EL)

1.5 (DL WL)

0.9 DL 1.5 EL
The load combinations for service conditions are as follows.

DL + LL
DL + 0.8 (LL EL)
DL EL
DL WL
Analysis of Structures
Regarding analysis of structures, IS:1343 - 1980 recommends the same procedure as
stated in IS:456 - 2000. A structure can be analysed by the linear elastic theory to
calculate the internal forces in a member subjected to a particular type of load.
Design of Members
There can be more than one way to design a member. In design, the number of
unknown quantities is larger than the number of available equations. Hence, some
quantities need to be assumed at the beginning. These quantities are subsequently
checked.
The member can be designed either for the service loads or, for the ultimate loads. The
procedure given here is one of the possible procedures. The design is based on
satisfying the allowable stresses under service loads and at transfer. Initially, a
lumpsum estimate of the losses is considered under service loads. After the first round
of design, detailed computations are done to check the conditions of allowable stresses.
Precise values of the losses are computed at this stage. The section is then analysed
for the ultimate capacity. The capacity should be greater than the demand under
ultimate loads to satisfy the limit state of collapse.

4.1.2 Design of Sections for Axial Tension
Introduction
Prestressed members under axial loads only, are uncommon. Members such as
hangers and ties are subjected to axial tension. Members such as piles may have
bending moment along with axial compression or tension.
Design of Prestressing Force

First, a preliminary dimension of the member is selected based on the architectural
requirement. The prestressing force at transfer (P0) should be such that the
compressive stress in concrete is limited to the allowable value. At service, the
designed prestressing force (Pe) should be such that the tensile stress in concrete
should be within the allowable value. The amount of prestressing steel (Ap) is
determined from the designed prestressing force based on the allowable stress in steel.
At transfer, in absence of non-prestressed reinforcement, the stress in concrete (fc) is

given as follows.
P0 (4-1.1)
fc = -
Ac
Here,
Ac = net area of concrete
In presence of non-prestressed reinforcement, the stress in the concrete (fc) can be

P0
fc = - (4-1.2)
Ac + (Es /Ec )As
Here,
Es = modulus of elasticity of steel
Ec = modulus of elasticity of concrete.
At service, the stress in concrete (fc) can be calculated as follows.
Pe P
fc = - (4-1.3)
Ac At

Here,
At = transformed area of section
P = external axial force
The external axial force is considered positive if it is tension and negative if it is
compression. In the above expression, non-prestressed reinforcement is not
considered. If there is non-prestressed reinforcement, Ac is to be substituted by (Ac +
(Es/Ec) As) and At is to be calculated including As.
Analysis of Ultimate Strength

The ultimate tensile strength of a section (PuR) is calculated as per Clause 22.3,
IS:1343 - 1980. The ultimate strength should be greater than the demand due to
factored loads.
In absence of non-prestressed reinforcement, the ultimate tensile strength of a section

(PuR) is given as follows.
(4-1.4)
PuR = 0.87fpk Ap
In presence of non-prestressed reinforcement,
PuR = 0.87fy As + 0.87fpk Ap (4-1.5)
In the previous equations,

fy = characteristic yield stress for non-prestressed reinforcement
with mild steel bars
= characteristic 0.2% proof stress for non-prestressed reinforcement
with high yield strength deformed bars.
fpk = characteristic tensile strength of prestressing tendons.
The following example shows the design of a post-tensioned hanger for tension.

Example 4-1.1
Design a post-tensioned hanger to carry an axial tension of PDL = 300 kN (dead

load including self-weight) and PLL = 130 kN. The dimension of the hanger is 250
250 mm2.
Design the section without considering non-prestressed reinforcement. Tension

is not allowed under service loads.
The grade of concrete is M 35. The age at transfer is 28 days. Assume 15% long
term losses in the prestress.
The following properties of the prestressing strands are available from tests.
Type of prestressing tendon : 7 wire strand
Nominal diameter = 12.8 mm
Nominal area = 99.3 mm2
Tensile strength fpk = 1860 N/mm2
Modulus of elasticity = 195 kN/mm2.
Solution
Preliminary calculations at transfer
Ac A = 250 250
= 62,500 mm2
Allowable stress for M35 concrete under direct compression at transfer

fcc,all = 0.8 0.51fci
= 0.8 0.5135
= 14.3 N/mm2
Maximum prestressing force at transfer
P0 max = fc Ac
= 14.3 62,500
= 892,500 N
Preliminary calculations at service

At A = 250 250
= 62,500 mm2
Stress in concrete
Pe P
fc = - +
Ac At
Allowable stress at service
fct,all = 0 N/mm2
Considering 15% loss
Pe = 0.85P0
Substituting the values

0.85P0 P
0= - +
A A
Preliminary calculations at service (continued)
Solving,
0.85P0 = P
300 +130
P0 =
0.85
= 506 kN
Allowable prestress in tendon
fp 0 = 0.8fpk
= 0.81860
= 1488 N/mm2
Required area of tendon

506,000
Ap =
1488
= 340 mm2
Select 4 strands with Ap = 499.3

= 397.2 mm2
Prestress at transfer P0 = 397.21488 N

= 591kN
Final calculations at transfer
Ac = 62,500 - 397
= 62103 mm2
P0
Stress in concrete fc = -
Ac
591,000
=-
62,103
= - 9.5 N/mm2 OK
fc < fcc,all
Final calculations at service
E p = 195 kN/mm2
Ec = 5,000 35
= 29,580 N/mm2
195
At = 62,103 + 397.2
29.6
= 64,720 mm2
Stress in concrete
Pe P
fc = - +
Ac At
0.85 591,000 (300 +130)103

fc = - +
62,103 64,720
= - 1.4 N/mm2
No tensile stress in concrete. OK.
Final calculations for ultimate strength
PuR = 0.87fpk Ap
= 0.871860 397.2 N
= 643.0 kN

Demand under factored loads
Pu = 1.5(300 +130)
= 645.0 kN
PuR Pu OK
Designed cross-section
250
250 (4) 7-wire strands

with P0 = 591 kN
Nominal non-prestressed reinforcement is provided for resisting thermal and shrinkage

cracks.

4.2 Design of Sections for Flexure

Preliminary Design
Final Design for Type 1 Members
Special Case
Calculation of Moment Demand

For simply supported prestressed beams, the maximum moment at the span is given by
the beam theory. For continuous prestressed beams, the analysis can be done by
moment distribution method. The moment coefficients in Table 12 of IS:456 - 2000 can
be used under conditions of uniform cross-section of the beams, uniform loads and
similar lengths of span.
The design is done for the critical section. For a simply supported beam under uniform
loads, the critical section is at the mid span. For a continuous beam, there are critical
sections at the supports and at the spans.
For design under service loads, the following quantities are known.
MDL = moment due to dead load (excluding self-weight)
MLL = moment due to live load.
The material properties are selected before the design.
The following quantities are unknown.

The member cross-section and its geometric properties,
MSW = moment due to self-weight,
Ap = amount of prestressing steel,
Pe = the effective prestress,
e = the eccentricity.
There are two stages of design.

1) Preliminary: In this stage the cross-section is defined and Pe and Ap are
estimated.

2) Final: The values of e (at the critical section), Pe, Ap and the stresses in concrete
at transfer and under service loads are calculated. The stresses are checked with
the allowable values. The section is modified if required.
4.2.1 Preliminary Design
The steps of preliminary design are as follows.

1) Select the material properties fck and fpk.
2) Determine the total depth of beam (h).
The total depth can be based on architectural requirement or, the following empirical
equation can be used.
h = 0.03 M to 0.04 M (4-2.1)
Here, h is in meters and M is in kNm. M is the total moment excluding self-weight.

3) Select the type of section. For a rectangular section, assume the breadth
b = h/2.
4) Calculate the self-weight or, estimate the self-weight to be 10% to 20% of
the load carried.
5) Calculate the total moment MT including self-weight. The moment due to
self-weight is denoted as Msw.
6) Estimate the lever arm (z).
z 0.65h, if Msw is large (Msw > 0.3MT).
z 0.5h, if Msw is small.
7) Estimate the effective prestress (Pe)
Pe = MT / z, if Msw is large.
Pe = MI L / z, if Msw is small.
If Msw is small, the design is governed by the moment due to imposed load
(MI L = MT MSW).
8) Considering fpe = 0.7fpk , calculate area of prestressing steel Ap = Pe / fpe.
9) Check the area of the cross-section (A).
The average stress in concrete at service C/A (= Pe /A) should not be too high as
compared to 50% of the allowable compressive stress fcc,all . If it is so, increase the area
of the section to A = Pe /(0.5fcc,all).

4.2.2 Final Design for Type 1 Members
The code IS:1343 - 1980 defines three types of prestressed members.

Type 1: In this type of members, no tensile stress is allowed in concrete at transfer
or under service loads.
Type 2: In these members, tensile stress is within the cracking stress of concrete.
Type 3: Here, the tensile stress is such that the crack width is within the allowable
limit.
The final design involves the checking of the stresses in concrete at transfer and under
service loads with respect to the allowable stresses. Since the allowable stresses
depend on the type of member (Type 1, Type 2 or Type 3), the equations vary for the
different types. Here, the steps of final design are explained for Type 1 members. The
steps for Type 2 members are explained in Section 4.3, Design of Sections for Flexure
(Part II). The steps for Type 3 members are similar to Type 2, the only difference being
the value of the allowable tensile stress in concrete.
For small moment due to self-weight (Msw 0.3MT), the steps are as follows.
1) Calculate eccentricity e to locate the centroid of the prestressing steel (CGS).

With increasing load, the compression (C) moves upward from the location of the
tension (T) at CGS. At transfer, under the self-weight, C should lie within the kern zone
to avoid tensile stress at the top. The kern points and kern zone are explained in
Section 3.3, Analysis of Member under Flexure (Part II).
The lowest permissible location of C due to self-weight is at the bottom kern point (at a
depth kb below CGC) to avoid tensile stress at the top. The design procedure based on
the extreme location of C gives an economical section.
The following sketch explains the lowest permissible location of C due to self-weight
moment (Msw) at transfer.

Msw ct C/A =
kt P0/A
h + CGC
C kb C e
cb
CGS
T
fb
Internal force Stress in
in concrete concrete
Figure 4-2.1 Stress in concrete due to compression at bottom kern point

A = gross area of cross section
fb = maximum compressive stress in concrete at bottom edge
h = total height of the section
kt, kb = distances of upper and lower kern points, respectively, from CGC
ct, cb = distances of upper and lower edges, respectively, from CGC
P0 = prestress at transfer after initial losses.
The shift of C due to self-weight gives an expression of e.

e = (Msw / P0) + kb (4-2.2)
Here, the magnitude of C or T is equal to P0. The value of P0 can be estimated as

follows.
a) 90% of the initial applied prestress (Pi) for pre-tensioned members.
b) Equal to Pi for post-tensioned members.
The value of Pi can be estimated from the amount of prestressing steel determined in
the preliminary design.
Pi = Ap(0.8fpk) (4-2.3)
Here, the permissible prestress in the steel is 0.8fpk, where fpk is the characteristic
tensile strength.
2) Recompute the effective prestress Pe and the area of prestressing steel Ap.
With increasing load, C further moves up. Under the service loads, C should lie within
the kern zone to avoid tensile stress at the bottom. The highest permissible location of

C due to total load is at the top kern point (at a height kt above CGC) to avoid tensile
stress at the bottom.
The following sketch explains the highest possible location of C due to the total moment
(MT).
ft
h MT C ct
kt
+ CGC
C kb e cb C/A =
CGS
T Pe/A
0
Figure 4-2.2 Stress in concrete due to compression at top kern point
ft = maximum compressive stress in concrete at top edge.
The shift of C due to the total moment gives an expression of Pe.

Pe = MT /(e + kt) (4-2.4)
Considering fpe = 0.7fpk , the area of prestressing steel is recomputed as follows.

Ap = Pe / fpe (4-2.5)
3) Recompute eccentricity e
First the value of P0 is updated. The eccentricity e is recomputed with the updated value
of P0.
If the variation of e from the previous value is large, another cycle of computation of the
prestressing variables can be undertaken.
4) Check the compressive stresses in concrete.

The maximum compressive stress in concrete should be limited to the allowable values.
At transfer, the stress at the bottom should be limited to fcc,all , where fcc,all is the
allowable compressive stress in concrete at transfer (available from Figure 8 of IS:1343

- 1980). At service, the stress at the top should be limited to fcc,all , where fcc,all is the
allowable compressive stress in concrete under service loads (available from Figure 7
of IS:1343 - 1980).
a) At Transfer
The stress at the bottom can be calculated from the average stress P0/A.
P0 h (4-2.6)
fb = -
A ct
To satisfy |fb| fcc,all, the area of the section (A) is checked as follows.
P0h
A
fcc,all ct (4-2.7)
If A is not adequate then the section has to be redesigned.
b) At Service
The stress at the top can be calculated from the average stress Pe/A.
P h (4-2.8)
f =- e
t A cb
To satisfy |ft| fcc,all, the area of the section (A) is checked as follows.
Pe h
A (4-2.9)
fcc,all cb
4.2.3 Special Case
For large moment due to self-weight (Msw > 0.3 MT), the eccentricity e according to
e = (Mw / P0) + kb may violate the cover requirements or, may even lie outside the beam.
In such cases, locate e as per cover requirements. The location of C at transfer will be
within the kern zone without zero stress at the top. The expression of stress at the
bottom is different from that given earlier. The other steps are same as before.

At transfer, the stress at the bottom is calculated using the following stress profile.
e MSW/P0 C/A =P0/A

CGC
C
fb
Figure 4-2.3 Stress in concrete due to compression above bottom kern point
M
P0 e - sw cb
fb = - 0 -
P P0
(4-2.10)
A I
Substituting I = Ar2 and r2/cb = kt
e - Msw
P P0 (4-2.11)
fb = - 0 1+
A kt

e - Msw
P P0 (4-2.12)
A 0 1+
fcc,all kt

The following example shows the design of a Type 1 prestressed member.
Example 4-2.1
Design a simply supported Type 1 prestressed beam with MT = 435 kNm

(including an estimated MSW = 55 kNm). The height of the beam is restricted to
920 mm. The prestress at transfer fp0 = 1035 N/mm2 and the prestress at service
fpe = 860 N/mm2.
Based on the grade of concrete, the allowable compressive stresses are 12.5
N/mm2 at transfer and 11.0 N/mm2 at service.

The properties of the prestressing strands are given below.

Type of prestressing tendon : 7-wire strand
Solution
A) Preliminary design
The values of h and MSW are given.
1) Estimate lever arm z. Msw 55

=
MT 435
= 12.5 %
Since MSW < 0.3 MT ,

Use z = 0.5h
= 0.5 920
= 460 mm
2) Estimate the effective prestress.

Moment due to imposed loads
MIL = MT - Msw
= 435 - 55
= 380 kNm
Effective prestress
380 103
Pe =
460
= 826 kN
3) Estimate the area of the prestressing steel.
Pe
Ap =
fpe
826103
=
860
= 960 mm2

4) Estimate the area of the section to have average stress in concrete equal to 0.5 fcc,all .
Pe
A=
0.5fcc,all
826 103
=
0.511.0
= 150 103 mm2
The following trial section has the required depth and area.
Trial cross-section
390
100
100
920
100
Values in mm.
B) Calculation of geometric properties
The section is symmetric about the horizontal axis. Hence, the CGC lies at mid depth.
properties.
1
410 ct = 460
CGC 2
3
Values in mm.
Check area of the section
A = 2 A1 + A2
= 2(390100) + (720100)
= 150,000 mm2

Moment of inertia of the section about axis through CGC
I = 2I1 + I2
1 1
= 2 390 1003 + (390 100) 4102 + 100 7203
12 12
= 1.6287 1010 mm4

I
r2 =
A
1.6287 1010
=
150,000
= 108,580 mm2
Kern levels of the section

r2
kt = k b =
ct
108,580
=
460
= 236 mm
Summary after preliminary design
Properties of section
A = 150,000 mm2
I = 1.62871010 mm4
ct = cb = 460 mm
kt = kb = 236 mm
Properties of prestressing steel
Ap = 960 mm2
Pe = 826 kN

C) Final design
1) Calculate eccentricity e
P0 = Ap fp 0 Msw
e= + kb
P0
= 9601035
55.0 103
= 993.6 kN = + 236
993.6
290 mm
2) Recompute the effective prestress and associated variables.
MT
Pe =
e+ kt
435 103
=
(290 + 236)
= 827 kN
Since Pe is very close to the previous estimate of 826 kN, Ap, P0 and e remain same.
The tendons are placed in two ducts. The outer diameter of each duct is 54 mm.
Select (10) 7-wire strands with
Ap = 10 99.3
= 993.0 mm2

a) At transfer
P0h
A
fcc,all ct
993.6 920
=
12.5 460
= 158,976 mm2
b) At service
Pe h
A
fcc,all cb
827 920
=
11.0 460
= 150,364 mm2

The governing value of A is 158,976 mm2. The section needs to be revised. The width
of the flange is increased to 435 mm. The area of the revised section is 159,000 mm2.
Another set of calculations can be done to calculate the geometric properties precisely.
Designed cross-section at mid-span

435
100
100
CGC
290 920
CGS

with P0 = 994 kN

4.3 Design of Sections for Flexure (Part II)

Final Design for Type 2 Members
The steps for Type 1 members are explained in Section 4.2, Design of Sections for
Flexure (Part I).
4.3.1 Final Design for Type 2 Members
For Type 2 members, the tensile stress under service loads is within the cracking stress
of concrete. The allowable tensile stress in concrete (fct,all) as per IS:1343 - 1980 is
same for transfer and service load conditions. The value is 3.0 N/mm2, which can be
increased to 4.5 N/mm2 for temporary loads.
The following material provides the steps for sections with small self-weight moment.
For sections with large self- weight moment, the eccentricity e may need to be
determined based on the cover requirements.
1) Calculate eccentricity e to locate the centroid of the prestressing steel (CGS).

Under the self-weight, C may lie outside the kern region. The lowest possible location of
C due to self-weight is determined by the allowable tensile stress at the top.
The following sketch explains the extreme location of C due to self-weight moment (Msw)
at transfer.
fct,all
h MSW ct

kt
+ CGC
kb
C e1 e C cb
CGS 2
T
fb
Figure 4-3.1 Stress in concrete due to compression outside bottom kern point


fb = maximum compressive stress in concrete at bottom edge
fct,all = allowable tensile stress in concrete at top edge
h = total height of the section
kt, kb = distances of upper and lower kern points, respectively, from CGC
ct, cb = distances of upper and lower edges, respectively, from CGC
e1 = distance between the bottom kern point and the location of
compression
e2 = distance by which the compression travels from CGS due to self
weight
P0 = prestress at transfer after initial losses.
From the previous figure, the shift of C due to self-weight gives an expression of e2. It is
evident that if C is further shifted upwards by a distance e1 to the bottom kern point,
there will be no tensile stress at the top.
Msw
e2 = (4-3.1)
P0
The value of e1 is calculated from the expression of stress corresponding to the moment
due to the shift in C by e1.
P0e1ct
= fct,all
I
f I
e1 = ct,all
P0ct (4-3.2)
fct,all Ak b
e1 =
P0
Substituting I = Ar2 and r2/ct = kb
The distance of the CGS below the bottom kern point is given as follows.
Msw + fct,all Ak b
e1 + e2 = (4-3.3)
P0

The eccentricity e is calculated from the following equation.

e = e1 + e2 + k b
Msw + fct,all Ak b (4-3.4)
e= + kb
P0
The above expression can be compared with the expression of Type 1 member e =
(Msw / P0) + kb. Note that the eccentricity has increased for a Type 2 member due to the
allowable tensile stress fct,all.
2) Recompute the effective prestress Pe and the area of prestressing steel Ap.
Under the total load, C may lie outside the kern region. The highest permissible location
of C due to total load is determined by the allowable tensile stress at the bottom.
The following sketch explains the highest possible location of C due to the total moment
(MT).
ft
C
MT e3 ct
kt
h + CGC
kb
C cb
CGS
T
fct,all
Figure 4-3.2 Stress in concrete due to compression outside top kern point
From the previous figure, the expression of e3 is obtained by the tensile stress
generated due to the shift of C beyond the upper kern point.
Pe e3cb
= fct,all
I
f I
e3 = ct,all
Pe cb
fct,all Akt
e3 = (4-3.5)
Pe
Substituting I = Ar2 and r2/cb = kt

The shift of C due to the total moment gives an expression of Pe.
MT = Pe ( e+kt +e3 )
= Pe ( e+kt ) +fct,all Akt
MT - fct,all Akt
Pe =
e+kt (4-3.6)
The above expression can be compared with the expression of Type 1 member Pe = MT
/(e + kt). Note that the prestressing force has decreased for a Type 2 member due to
the allowable tensile stress fct,all. This will lead to a decrease in the area of prestressing
steel (Ap). Considering fpe = 0.7fpk, Ap is recomputed as follows.
Ap = Pe/ fpe (4-3.7)
First the value of P0 is updated. The eccentricity e is recomputed with the updated value
of P0.
4) Check the compressive stresses in concrete

The maximum compressive stress in concrete should be limited to the allowable values.
At transfer, the stress at the bottom should be limited to fcc,all , where fcc,all is the
allowable compressive stress in concrete at transfer (available from Figure 8 of IS:1343
- 1980). At service, the stress at the top should be limited to fcc,all , where fcc,all is the
allowable compressive stress in concrete under service loads (available from Figure 7
of IS:1343 - 1980).
a) At Transfer
The stress at the bottom can be calculated from the stress diagram.
C C ( kb + e1 ) cb
fb = - -
A I
fb = - 1+ b 2 b - 1 b
C k c Ce c (4-3.8)
A r I

From fct,all = Ce1ct / I, substituting Ce1 / I = fct,all /ct
C cb fct,all
fb = - 1+ - cb
A ct ct
C h fct,all
fb = - - cb (4-3.9)
A ct ct
C h fct,all
+ cb fcc,all
A ct ct
P0 h
fcc,all ct - fct,all cb
A
P0 h
A (4-3.10)
b) At Service
The stress at the top can be calculated from the stress diagram.
C C ( kt + e3 ) ct
ft = - -
A I
ft = - 1+ t 2 t - 3 t
C kc Ce c (4-3.11)
A r I
From fct,all = Ce3cb / I, substituting Ce3 / I = fct,all /cb
C ct fct,all ct
ft = - 1+ -
A cb cb
C h fct,all ct
ft = - -
A cb cb (4-3.12)
To satisfy |ft| fcc,all, the area of the section (A) is checked as follows.
C h fct,all ct
+ fcc,all
A cb cb
Pe h
fcc,all cb - fct,all ct
A
Pe h
A (4-3.13)

The following table shows a comparison of equations for Type 1 and Type 2 members.
Table 4-3.1 Comparison of equations for Type 1 and Type 2 members
Type 1 Type 2
Eccentricity Msw Msw + fct,all Ak b
e= + kb e= + kb
P0 P0
Effective prestress MT MT - fct,all Akt

Pe = Pe =
e + kt e + kt
Minimum area based on P0 h P0 h

A A
stress at bottom at transfer fcc,all ct fcc,all ct - fct,all cb
Minimum area based on Pe h Pe h

A A
stress at top at service fcc,all cb fcc,all cb - fct,all ct
The following example shows the design of a Type 2 prestressed member. The same
section was designed as a Type 1 member in Section 4.2, Design of Sections for
Flexure (Part I). The solutions of the two examples are compared at the end.
Example 4-3.1
Design a simply supported Type 2 prestressed beam with MT = 435 kNm

(including an estimated MSW = 55 kNm). The height of the beam is restricted to
920 mm. The prestress at transfer fp0 = 1035 N/mm2 and the prestress at service
fpe = 860 N/mm2.
N/mm2 at transfer and 11.0 N/mm2 at service. The allowable tensile stresses are
2.1 N/mm2 at transfer and 1.6 N/mm2 at service.


Solution
A) Preliminary design
The values of h and MSW are given.
1) Estimate lever arm z. Msw 55

=
MT 435
= 12.5 %
Since MSW < 0.3 MT ,
use z = 0.5h
= 0.5 920
= 460 mm
2) Estimate the effective prestress.

Moment due to imposed loads
MIL = MT - Msw
= 435 - 55
= 380 kNm
Effective prestress
380 103
Pe =
460
= 826 kN
3) Estimate the area of the prestressing steel.

Pe
Ap =
fpe
826103
=
860
= 960 mm2
4) Estimate the area of the section to have average stress in concrete equal to 0.5 fcc,all .
Pe
A=
0.5fcc,all
826 103
=
0.511.0
= 150 103 mm2

The following trial section has the required depth and area.
Trial cross-section
390
100
100
920
100
Values in mm.
B) Calculation of geometric properties
properties.
1
410 ct = 460
CGC 2
3
Values in mm.
Check area of the section
A = 2(390 100) + (720 100)

= 150,000 mm2
I = 2I1 + I2
= 2 390 1003 + (390 100) 4102 + 100 7203

1 1
12 12
= 1.6287 1010 mm4


I
r2 =
A
1.62871010
=
150,000
= 108,580 mm2
Kern levels of the section
r2
kt = k b =
ct
108,580
=
460
= 236 mm
Summary after preliminary design
Properties of section
A = 150,000 mm2
I = 1.62871010 mm4
ct = cb = 460 mm
kt = kb = 236 mm
Values of prestressing variables

Ap = 960 mm2
Pe = 826 kN
C) Final design
1) Calculate eccentricity e
P0 = Ap fp 0
= 960 1035
= 993.6 kN
Msw + fct,all Ak b
e1 + e2 =
P0
2.1
55 103 + 150,000 236
= 103
993.6
= 130mm

e = e1 + e 2 + k b
= 130 + 236
= 366 m m
2) Recompute the effective prestress and the area of prestressing steel Ap.
MT - fct,all Akt
Pe =
e+ kt
1.65
435103 - 150,000 236
= 103
336 + 236
= 625.6 kN
Since Pe is substantially lower than the previous estimate of 826 kN, Ap, P0 and e need
to be recalculated.
Pe
Ap =
fpe
625.6103
=
860
= 727 mm2
P0 = Ap fp 0
= 7271035
= 752.4kN
Msw + fct,all Akb

e= + kb
P0
2.1
55 103 + 150,000 236
= 103 + 236
752.4
= 172 + 236
= 408 400 mm
Check the cover requirement
Assuming the outer diameter of duct equal to 54 mm

Clear cover for the duct 1
= 460 - 400 - 54
2
= 33 mm

The clear cover at the bottom is greater than 30 mm (Clause 11.1.6.2, IS: 1343 - 1980),
which is satisfactory. The side cover in the web is slightly less than 30 mm. The
thickness of the web can be increased to satisfy the requirement.
Since the value of e has changed from 366 mm to 400 mm, prestressing variables are
recomputed.
MT - fct,all Akt
Pe =
e + kt
1.65
435 103 - 150,000 236
= 103
400 + 236
= 592.0 kN
Pe has further reduced from 625.6 kN. Ap and P0 are recalculated.
592103
Ap =
860
= 688.5 mm2
Select (7) 7-wire strands with
Ap = 799.3
= 695.1mm2
The tendons can be placed in one duct. The outer diameter of the duct is 54 mm.
P0 = 695.11035
= 719.4kN
Since the maximum possible eccentricity is based on cover requirement, the value of e
is not updated.

At transfer
P0h
A
719.4103 920
=
12.5 460 - 2.1 460
= 138,352 mm2

At service
Pe h
A
592103 920
=
11 460 -1.65 460
= 126,631 mm2
The governing value of A is 138,352 mm2. The section can be revised. The width of
the flange is reduced to 335 mm. The area of the revised section is 139,000 mm2.
Another set of calculations can be done to calculate the geometric properties precisely.
Design cross-section at mid-span

335
100
100
CGC
920
400
CGS
100
(7) 7-wire strands

with P0 = 719 kN
Comparison of Type 1 and Type 2 sections
The solutions from the examples of Type 1 and Type 2 members are placed together in
the next figure for comparison.
435 335
100 100
100 100
CGC CGC
290 920 920
CGS 400
CGS
100 100
(10) 7-wire strands (7) 7-wire strands

with P0 = 994 kN with P0 = 719 kN
Type 1 Type 2
Figure 4-3.3 Sections designed as Type 1 and Type 2 members

The following observations can be made.

1) In Type 2 section, the amount of prestressing steel and the prestressing force
are less than those in a Type 1 section. The area of cross-section is less for Type 2
section.
Type 2 section is relatively economical.
2) The eccentricity in Type 2 section is larger than in Type 1 section. For unit
prestressing force, the prestressing is more effective in Type 2 section.

4.4 Design of Sections for Flexure (Part III)

Choice of Sections
Determination of Limiting Zone
Post-tensioning in Stages
4.4.1 Choice of Sections
The type of section is selected based on the use of the structure, architectural
requirements, casting and fabrication options, available technology and skilled work
force. Here, a few comments are given for the available types of sections.
1) The section should have large depth below the CGC, so as to have the provision of
large eccentricity. The prestressing force can then be reduced.
2) There should be adequate concrete at the top and bottom to satisfy the allowable
stresses.
3) The end section is usually solid to increase the shear capacity and prevent
anchorage zone failure.
Here, the sections are broadly grouped under rectangular section, T-section, I-section
and inverted T-section. Some variations of each type are shown under the
corresponding broad groups. The sections in each group have similar analysis
procedure. The sections shown are not exclusive.

Table 4-4.1 Types of sections

Rectangular T-section I-section Inverted T-
section
Broad groups of
sections
Variations
(a) (b)
(a) More room for
tendons in lower
Remarks on the Torsionally stiff
flange.
sections and strong.
(b) Better stability
during erection.
Expensive form
Fabrication Easy Easy Difficult
work
Space for Less than

Adequate Good Good
reinforcement adequate
1) Very
inefficient.
Effeciency for
Very good, 2) Small
non-composite Poor, z 0.4h Good, z 0.5h
z 0.7h ultimate
sections
moment
capacity
1) Especially 1) Good for

Application of 1) Light load and good for long long span and
non-composite short span. span roofs, when heavy loads. Msw/MT large
sections 2) Msw/MT large. LL DL. 2) Msw/MT
2) Msw/MT large. small.
Effeciency for Very good Increases load Very good

composite particularly when capacity only Very good with cast-in-
sections section is shored. slightly. place flange.

1) Topping
serves to tie all
Application of sections Long span
Good for building
composite together. buildings and Bridges
construction
sections 2) No form bridges.
required for
composite pour.
The different types of sections can be compared by a measure of flexural efficiency .

The flexural efficiency is defined in terms of the radius of gyration r as follows.
r2
=
ct cb
r 2 ct + cb
=
ct cb h
r2 r2
+
c b ct
=
h
k + kb (4-4.1)
= t
h Thus for a given value of the depth h, if the
kern zone (kt+kb) is large then the section is efficient.

For a rectangular section, = 0.33.
For an I-section, > 0.33.
4.4.2 Determination of Limiting Zone
For full prestressed members (Type 1), tension is not allowed under service conditions.
If tension is also not allowed at transfer, C always lies within the kern zone. The
limiting zone is defined as the zone for placing the CGS of the tendons such that C
always lies within the kern zone. Also, the maximum compressive stresses at transfer
and service should be within the allowable values.
Also, the maximum compressive stresses at transfer and service should be within the

allowable values. Note that the limiting zone is a restriction for the CGS. The individual
tendons may lie outside the limiting zone.
The following figure shows the limiting zone (as the shaded region) for a simply
supported beam subjected to uniformly distributed load.
CL
Locus of emin
CGC
Locus of emax
Figure 4-4.1 Limiting zone for a simply supported beam
The limiting zone is determined from the maximum or minimum eccentricities of the
CGS along the beam corresponding to the extreme positions of C.
The maximum eccentricity (emax) at any section corresponds to the lowest possible
location of C at transfer, that generates allowable tensile stress at the top of the section.
Also, the maximum compressive stress at the bottom should be within the allowable
value.
The minimum eccentricity (emin) at any section corresponds to the highest possible
location of C at service, that generates allowable tensile stress at the bottom of the
section. Also, the maximum compressive stress at the top should be within the
allowable value.
The following material gives the expressions of emax and emin for Type 1 and Type 2
sections. The zone between the loci of emax and emin is the limiting zone of the section
for placing the CGS. The values of emax and emin can be determined by equating the
stresses at the edges of concrete with the allowable values. Else, explicit expressions of
emax and emin can be used. Here, the expressions of emax and emin based on allowable
tensile stress are given.
Type 1 Section
At Transfer

The following sketch shows the stress profile in concrete when C is at the lowest
permissible location due to self-weight moment (Msw) at transfer.
0
ct
kt
CGC
kb C emax c
CGS b
T
fb
Internal forces
Stress in
concrete
Figure 4-4.2 Stress in concrete due to compression at bottom kern point
From the shift of C due to self-weight, the following expression can be derived.
Msw
emax - k b =
P0
Msw (4-4.2)
or, emax = + kb
P0
Note that since MSW varies, emax varies along the length of the beam. Also, the stress at
the bottom needs to be checked to satisfy the condition |fb| fcc,all.
At Service
The following sketch shows the stress profile in concrete when C is at the highest
possible location due to the total moment (MT).
ft
C ct
kt
CGC
kb emin cb
CGS
T
0
Internal forces Stress in
concrete
Figure 4-4.3 Stress in concrete due to compression at top kern point
From the shift of C due to total moment, the following expression can be derived.
MT
emin + kt =
Pe
MT
or, emin = - kt
Pe
(4-4.3)
Note that since MT varies, emin varies along the length of the beam. Also, the stress at
the top needs to be checked to satisfy the condition |ft| fcc,all.
If for a particular section emin is negative, it implies that the CGS can be placed above
CGC. This happens near the supports.
Type 2 Section
At Transfer
The following sketch shows the stress profile in concrete when C is at the lowest
permissible location due to self-weight moment (Msw) at transfer.
fct,all
ct
CGC
kt
k
e1 b C emax cb
CGS e2
T
fb
concrete
Figure 4-4.4 Stress in concrete due to compression outside bottom kern point
Msw + fct,all Ak b
emax - k b =
P0
Msw + fct,all Ak b
or, emax = + kb (4-4.4)
P0
Note that emax for a Type 2 section is larger than that for a Type 1 section due to the
term fct,all Akb in the numerator. The stress at the bottom needs to be checked to satisfy
the condition |fb| fcc,all.
At Service
The following sketch shows the stress profile in concrete when C is at the highest
possible location due to the total moment (MT).

ft
C
e3 ct
CGC
kt
kb emin c
b
CGS
T
fct,all
concrete
Figure 4-4.5 Stress in concrete due to compression outside top kern point
MT - fct,all Akt
emin + kt =
Pe
MT - fct,all Akt
or, emin = - kt (4-4.5)
Pe
Note that emin for a Type 2 section is smaller than that for a Type 1 section due to the
term fct,all Akt in the numerator. The stress at the bottom needs be checked to satisfy the
condition |ft| fcc,all.
The zone between emax and emin is the limiting zone of the section for placing the CGS
for a given loading condition. The values of emax and emin for several sections can be
determined at regular intervals along the length of the beam to get their loci. Note that
the limiting zone for a Type 2 member is larger than the limiting zone for a Type 1
member. The following table shows a comparison of equations for Type 1 and Type 2
members.
Table 4-4.2 Comparison of equations for Type 1 and Type 2 members
Type 1 Type 2
Maximum eccentricity Msw M sw + fct,all Ak b
e= + kb e= + kb
P0 P0
Minimum eccentricity MT MT - fct,all Akt

e= - kt e= - kt
Pe Pe
The following example shows the calculation of limiting zone based on equating the
stresses at the edges of concrete with the allowable values.

Example 4-4.1
For the Type 2 post-tensioned beam with a flanged section as shown, the span is
18 m. For uniform loads, the profile of the CGS is parabolic. The live load
moment at mid-span (MLL) is 648 kNm. The prestress after transfer (P0) is 1600 kN.
Assume 15% loss at service.
Evaluate the limiting zone of CGS, if the allowable stresses at transfer and at
service are as follows.
For compression, fcc,all = 18.0 N/mm2

For tension, fct,all = 1.5 N/mm2.
500
200
150
1000
CGS
150 200
250
Values in mm.
Solution
A) Calculation of geometric properties
properties. The centroid of each rectangle is located from the soffit.

1 ct
+ CGC
900 2
y cb
500
100 3
Values in mm.
Area of the section
Area of 1 = A1 = 500 200 = 100,000 mm2

Area of 2 = A2 = 600 150 = 90,000 mm2
Area of 3 = A3 = 250 200 = 50,000 mm2
A = A1 + A2 + A3
= 240,000 mm2
Distance of CGC from the soffit
A1 900 + A2 500 + A3 100

y=
A
= 583.3 mm
Therefore,
cb = 583.3 mm
ct = 1000.0 - 583.3
= 416.7 mm
1
I1 = 500 2003 + A1 (900 - 583.3)2
12
= 1.036 1010 mm4
Moment of inertia of 2
1
I2 = 150 6003 + A 2 (583.3 - 500)2
12
= 3.32109 mm4

Moment of inertia of 3
1
I3 = 250 2003 + A3 (583.3 -100)2
12
= 1.184 1010 mm4
Moment of inertia of the section
I = I1 + I2 + I3
= (1.036 + 0.336 +1.184)1010
= 2.5521010 mm4
Calculation of moment due to self weight.

2
1 m2
wSW = 24 kN/m3 240,000 mm2 3 2
10 mm
= 5.76 kN/m
wSW L2
MSW =
8
5.76182
=
8
= 233.3 kNm
B) Determination of limiting zone
The values of emax and emin are determined by equating the stresses at the edges of
concrete with the allowable values. The expression of stress is given below.
P Pey My
f =-
A I I
M
CGC + + =
P
P/A Pey/ I My/ I Resultant

stress
profile
Limiting position at mid-span
For emax, consider the load stage at transfer.
i) Calculate e based on fb = 18.0 N/mm2.

P0 1600103
- =-
A 240103
= -6.67 N/mm2
P0ecb 1600 103 e 583.3

=-
I 2.5521010
= - 0.0366e
MSW cb 233.3106 583.3

=
I 2.5521010
= 5.33 N/mm2
fb = - 6.67 - 0.0366e + 5.33

= - 18.0 N/mm2
18.0 - 6.67 + 5.33

Solving, e =
0.0366
= 455.2 mm
ii) Calculate e based on ft = 1.5 N/mm2.
P0ect 1600 103 e 416.7

=
I 2.5521010
= 0.0261 e
MSW ct 233.3106 416.7

=-
I 2.5521010
= -3.81N/mm2
ft = - 6.67 + 0.0261e - 3.81

= 1.5 N/mm2
1.5 + 6.67 + 3.81

Solving, e =
0.0261
= 460.8 mm
Out of the two values of e, the lower value 455.2 mm governs.
emax = 455.2 mm

For emin, consider the load stage at service.
i) Calculate e based on ft = 18.0 N/mm2.
Pe P
= 0.85 0
A A
= - 5.67 N/mm2
Pe ect 0.85 1600103 e 416.7

=
I 2.5521010
= 0.022 e
MLLct 648.0106 416.7

=-
I 2.5521010
= - 10.58 N/mm2
ft = - 5.67 + 0.022e - 3.81- 10.58

= -18.0 N/mm2
-18.0 + 5.67 + 3.81+10.58

Solving, e =
0.022
= 93.6 mm
ii) Calculate e based on fb = 1.5 N/mm2.
Pe ecb 0.85 1600 103 e 583.3

=-
I 2.5521010
= - 0.031e
MLL cb 648.0106 583.3

=
I 2.5521010
= 14.81N/mm2
fb = - 5.67 - 0.031e + 5.33 +14.81

= 1.5 N/mm2
-1.5 - 5.67 + 5.33 +14.81

Solving, e =
0.031
= 418.4 mm
Out of the two values of e, the higher value 418.4 mm governs.
emin = 418.4 mm

Limiting position at end
MSW = MLL = 0.0

At transfer
For fb = 18.0 N/mm2
fb = - 6.67 - 0.0366e
= - 18.0 N/mm2
18.0 - 6.67
Solving, e =
0.0366
= 309.6 mm
For ft = 1.5 N/mm2
ft = - 6.67 + 0.026e
= 1.5 N/mm2
1.5 + 6.67
Solving, e =
0.026
= 314.2 mm
Selecting the lower value
emax = 309.6 mm
At service
For fb = 1.5 N/mm2
fb = - 6.67 - 0.0366e
= 1.5 N/mm2
Solving e = - 223.0 mm
For ft = 18.0 N/mm2
ft = - 6.67 + 0.0261e
= - 18.0 N/mm2
Solving e = - 436.0 mm
Since the values of e are negative the CGS lies above CGC. The position of CGS closer
to the CGC is selected.
emin = - 223.0 mm
Similarly, the values of emax and emin can be determined at regular intervals along the
span. The limiting zone is available by joining the points by straight lines.

In the following sketch the limiting zone is shown shaded.

CL
223.0 CGC
309.6 418.4
455.2
Values in mm.
From the sketch of the limiting zone, it is evident that the tendons can be spread out at
the ends. This is necessary to anchor the tendons and reduce the stress concentration
at the ends.
4.4.3 Post-tensioning in Stages
In the previous expressions of emax and emin, the values of P0 and Pe can be for different
levels of prestressing for post-tensioned members. At transfer the member can be
partially prestressed in the casting yard, from which P0 is calculated. After the member
is placed in its permanent location, it can be further prestressed before it is put into
service. The application of prestress in different stages is termed as post-tensioning in
stages. The value of Pe is calculated from the revised prestressing force.
With reduced P0 at transfer, emax is increased. Thus, the limiting zone for placing the
CGS and the available zone for the shift in C under service loads are also increased.

4.5 Design of Sections for Flexure (Part IV)

This section covers the following topic.
Magnels Graphical Method
Notations
The variables used in this section are as follows.
A = area of cross section of member
ct = distance of the top of the section from CGC
cb = distance of the bottom of the section from CGC
e = eccentricity of CGS with respect to CGC
ft = stress at the top of the section
fb = stress at the bottom of the section.
fcc,all = allowable compressive stress in concrete
fct,all = allowable tensile stress in concrete
I = moment of inertia of cross section of member
kt = distance of top kern point from CGC
kb = distance of bottom kern point from CGC
MSW = moment due too self weight
MT = total moment
P0 = prestress at transfer after immediate losses
Pe = prestress at service after long term losses
r = radius of gyration, r2 = I/A
Zt = section modulus corresponding to top of the section = I/ct
Zb = section modulus corresponding to bottom of the section = I/cb
= ratio of prestressing forces = Pe /P0
4.5.1 Magnels Graphical Method
The determination of maximum and minimum eccentricities at the critical section helps
in placing the CGS. But with different types of possible sections, the computations
increase. The graphical method proposed by G. Magnel gives a visual interpretation of
the equations involved.

There are essentially four stress conditions to be checked. These conditions are as
follows.
At transfer: ft fct,all and fb fcc,all
At service: ft fcc,all and fb fct,all
The above expressions are algebraic inequalities where the stresses ft and fb are
positive if tensile and negative if compressive. The allowable tensile stress fct,all is
assigned a positive value and the allowable compressive stress fcc,all is assigned a
negative value. The allowable stresses are explained in the Section 1.5, Concrete (Part
I).
It is to be noted that the values of fcc,all at transfer and at service are different. They are
calculated based on the strength of concrete at transfer and at service, respectively.
Similarly, the values of fct,all at transfer and at service can be different. As per IS:1343 -
1980, the values of fct,all at transfer and service are of course same.
The stresses ft and fb in the four inequalities are expressed in terms of the initial
prestressing force P0, the eccentricity e at the critical section of the member, the section
properties A, Zt, Zb, kt, kb and the load variables Msw and MT.
After transposition, 1/P0 is expressed in terms of e by linear inequality relationships.
For a selected section, these relationships are plotted in the 1/P0 versus e plane. The
acceptable zone shows the possible combinations of 1/P0 and e that satisfy all the four
inequality relationships. A combination of P0 and e can be readily calculated from the
acceptable zone.
The method is explained in a general form. For Type 1, Type 2 and Type 3 members,
the value of allowable tensile stress (fct,all) is properly substituted. For Type 1 members,
fct,all = 0 N/mm2.

At Transfer
The following sketch shows the variation of stress in concrete after the transfer of
prestress and due to the self weight.
ft
MSW ct
e P0 cb
fb
Figure 4-5.1 Stress profile in concrete at transfer
The stress at the top is calculated from P0, e, Msw as follows.
P0 P0ect Msw ct
ft = - + -
A I I
P ec M
= 0 -1+ 2t - sw
A r Zt
P0 e M
= -1+ - sw
A k b Zt (4-5.1)
The inequality relationship satisfying the stress at the top is expressed in terms of 1/P0
and e as follows.
ft fct,all
P0 e M
-1+ - sw fct,all
A k b Zt
1 (-1+ e/k b )
or,
P0 Msw (4-5.2)
fct,all + Z A
t
The following sketch shows the plot of inequality relationship. The straight line given by
the above inequality is plotted in the 1/P0 versus e plane and the acceptable zone is
shaded.

1
P0 Acceptable
kb e
Figure 4-5.2 Plot based on stress at the top at transfer
The following expression relates the stress at the bottom with the load and section
variables.
P0 P0ecb Msw cb
fb = - - +
A I I
P ec M
= - 0 1+ 2b + sw
A r Zb
P0 e Msw
=- 1+ + Z (4-5.3)
A kt b
The inequality relationship satisfying the stress at the bottom is expressed as follows.
fb fcc,all
P0 e Msw
- 1+
A kt + Z fcc,all
b
1 ( 1+ e/kt )
or, (4-5.4)
P0 Msw
-fcc,all + Z A
b
The following sketch shows the plot of the inequality relationship.
1
P0
Acceptable
e
- kt
Figure 4-5.3 Plot based on stress at the bottom at transfer

At Service
The following sketch shows the variation of stress in concrete at service and due to the
total moment.
ft
MT ct
e P0 cb
fb
Figure 4-5.4 Stress profile in concrete at service
Here, Pe is expressed as P0, where is the ratio of effective prestress (Pe) and
prestress at transfer (P0).
The expression of the stress at the top is given below.

P0 P0ect MT ct
ft = - + -
A I I
P ec M
= 0 -1+ 2t - T
A r Zt
P0 e M
= -1+ - T (4-5.5)
A k b Zt
The inequality relationship satisfying the stress at the top is expressed as follows.
ft fcc,all
P0 e M
-1+ - T fcc,all
A k b Zt
e
-1+
kb

1 (4-5.6)
or,
P0 MT
fcc,all + Z A
t
The following sketch shows the plot of inequality relationship. The straight line given by
the above inequality is again plotted in the 1/P0 versus e plane and the acceptable zone
is shaded.

1
P0
Acceptable
e
kb
Figure 4-5.5 Plot based on stress at the top at service
The following expression relates the stress at the bottom with the load and section
variables.
P0 P0ecb MT cb
fb = - - +
A I I
P ec M
= - 0 1+ 2b + T
A r Zb
P0 e MT
=- 1+ + Z
A kt
(4-5.7)
b
The inequality relationship is expressed as follows.

fb fct,all
P0 e MT
- 1+
A kt + Z fct,all
b
e
1+ k
t
1 (4-5.8)
or,
P0 MT
-fct,all + Z A
b
The following sketch shows the plot of the inequality relationship.

1
P0
Acceptable
- kt
e
Figure 4-5.6 Plot based on stress at the bottom at service

Next, the four lines are plotted simultaneously. The common region is the acceptable
zone.

(4-5.6) (4-5.2)
1
P0 Acceptable zone (4-5.8)
(4-5.4)
- kt kb
e
Figure 4-5.7 Acceptable zone
A combination of a trial section, prestressing force (P0) and eccentricity (e) at the critical
section, can be plotted in the form of the above graph. If the point lies within the
acceptable zone, then the combination is valid.
The following problem illustrates the use of Magnels graphical method.
Example 4-5.1
The section shown is designed as a Type 1 member with MT = 435 kNm (including
an estimated MSW = 55 kNm). The height of the beam is restricted to 920 mm. The
prestress at transfer fp0 = 1035 N/mm2 and the prestress at service fpe = 860 N/mm2.
N/mm2 at transfer and 11.0 N/mm2 at service.

For the section, find the acceptable zone by Magnels graphical method.
Compare the designed values of eccentricity (e) and the inverse of prestressing
force at transfer (1/P0) with the acceptable zone.

435
100
100
CGC 920
e = 290
CGS
with P0 = 994 kN
Solution
A) Calculation of geometric properties
properties.
1
410 ct = 460
CGC 2
3
Values in mm.
Area of the section
A = 2 A1 + A2
= 2(435100) + (720100)
= 159,000 mm2

I = 2I1 + I2
= 2 435 1003 + (435 100) 4102 + 100 7203

1 1
12 12
= 1.78 1010 mm4

I
r2 =
A
1.78081010
=
159,000
= 112,000 mm2
Section moduli
I
Zb = Zt = = 38,712,174 mm3
ct
Kern levels
r2
k b = kt = = 243.5 mm
ct
B) Calculation of the inequality relationships of Magnels graphical method
Ratio of effective prestress and prestress at transfer
Pe
=
P0
fpe 860
= =
fp 0 1035
= 0.83
At Transfer
ft fct,all 1 (-1+e/kb )

P0 Msw
fct,all + Z A
t
1 -1+ e 243.5

P0 55106
0 + 159,000
38,712,174
1 e
-1+
225,897.9 243.5

The relationship is plotted in the following graph.
0.0015
1/P 0 (1/kN)
0.001
0.0005
0
-300 -200 -100 0 100 200 300
e (mm)
At Transfer
fb fcc,all 1 ( 1+ e/kt )

P0 Msw
-fcc,all + Z A
b
1 1+ e 243.5

P0 55106
12.5 + 159,000
38,712,174

1 1+ e

2,213,397.9 243.5
0.0015
1/P 0 (1/kN)
0.001
0.0005
0
-300 -200 -100 0 100 200 300
e (mm)
At Service
e
-1+
ft fcc,all kb

1
P0 MT
fcc,all + Z A
t
1 (-1+ e 243.5)0.83

P0 435106
-11.0 + 159,000
38,712,174
1 -1+ e

45,358.0 243.5

0.0015
1/P 0 (1/kN)
0.001
0.0005
0
-300 -200 -100 0 100 200 300
e (mm)
At Service
fb fct,all e
1+ k
t
1
P0 MT
-fct,all + Z A
b
1 (1+ e 243.5)0.83

P0 435 106
0.0 + 159,000
38,712,174

1 1+ e

2,152,587.1 243.5
0.0015
1/P 0 (1/kN)
0.001
0.0005
0
-300 -200 -100 0 100 200 300
e (mm)
The four relationships are plotted in the following graph. The acceptable zone is shown.
The zone is zoomed in the next graph.

0.0015
1/P 0 (1/kN)
0.001
0.0005
0
-300 -200 -100 0 100 200 300
e (mm)
1/P 0 (1/kN) 0.0012
0.001
0.0008
225 250 275 300 325
e (mm)
The calculated values of e and 1/P0 for the Type 1 section are as follows.
e = 290 mm
1/P0 = 1/(994 kN) = 0.001kN1.
The solution of the design is shown in the graphs. It lies in the acceptable zone.

4.6 Detailing Requirements for Flexure

Tendon Profile
Minimum Amount of Reinforcement
Miscellaneous Requirements
Introduction
The detailing of the prestressing tendons and the reinforcing bars is important to satisfy
the assumptions in the analysis, proper placement of concrete and durability. After the
design calculations, drawings are prepared for construction. These drawings are
referred to as the design drawings which become a part of the construction documents.
The steel fabricator may prepare another set of drawings which are called shop
drawings. These drawings are similar to the design drawings but they contain additional
information such as the bar designations and bar bending schedule. It is essential to
show the detailing in the design drawings so that there is no ambiguity during
construction. It is also necessary to check the details in the shop drawings.
IS:1343 - 1980 specifies some minimum requirements. Here, these requirements are
briefly mentioned. There are requirements for the non-prestressed reinforcement as per
IS:456 - 2000 which are not covered here. The detailing requirements for shear and
torsion are covered in Section 5.3, Design for Shear (Part II) and Section 5.6, Design for
Torsion (Part II), respectively. Of course the detailing is best learned by preparing
drawings for construction projects.
4.6.1 Tendon Profile
For a simply supported post-tensioned beam with high uniformly distributed load, a
parabolic profile is selected. The equation of the profile is given as follows.
4ym (4-6.1)
y = ( ) (L - x )
L2
Here,
L = span of the beam
x = distance from one end
Y = vertical displacement of the tendon (from the level at the ends) at distance x

Ym = vertical displacement of the tendon (from the level at the ends) at the
middle of the beam
The following sketch shows the plot of the equation.
ym x
y
Figure 4-6.1 Profile of a parabolic tendon
Note that an individual tendon may be displaced from the CGS. Hence, the tendon
need not pass through the CGC at the ends. The figure below shows the parabolic
profiles of the ducts for placing tendon in a simply supported bridge girder.
Figure 4-6.2 Tendon profiles in a simply supported bridge girder

(Courtesy: Larsen & Toubro Rambll)
For continuous beams or slabs, parabolic profiles at the spans and at the supports are
connected to get the continuous profile of a tendon. The following sketch shows the
profile of the CGS in a continuous beam. The eccentricities of the CGS at the end span,
first interior support and first interior span are represented as e1, e2 and e3 respectively.

Points of contraflexure
(inflection points)
CGC e2
e1 e3
Points of maximum
eccentricity
CGC

Parabola 1 2 3 4 5 6 7
Figure 4-6.3 Profile of CGS in a continuous beam
A parabolic segment connects a point of maximum eccentricity with a point of

contraflexure. A point of contraflexure is the location where the curvature of the profile
reverses. It is also known as the inflection point. For varying spans and loading, the
segments on two sides of a point of maximum eccentricity, may not be symmetric. In
the second sketch of the above figure, the different parabolas between the points of
maximum eccentricity and the points of contraflexure are numbered.
The convex segment over a support is required to avoid a kink in the tendon. The length
of a convex segment is determined based on the minimum radius of curvature for the
type of tendon.
A parabolic segment satisfies two conditions.

1) It has zero slope at the point of maximum eccentricity.
2) At a point of contraflexure, the slopes of the parabolic segments on both sides
should match.
The equation of a parabolic segment is given below.
2
x (4-6.2)
y = ym
l

y
ym
x
l
Figure 4-6.4 Plot of a parabolic segment
In the previous equation,
l = length of the parabolic segment
x = distance from the point of maximum eccentricity
y = vertical displacement of the profile at distance x
ym = displacement of the point of contraflexure from the point of maximum
eccentricity.
The origin is selected at the point of maximum eccentricity at a critical section. The
equation satisfies the first boundary condition of zero slope at the point of maximum
eccentricity. The length (l) is determined from the requirement of minimum radius of
curvature at the support. The displacement ym is determined from the boundary
condition that at the point of contraflexure, the slopes of the segments on both sides
should match.
The following photo shows the profiles of the tendons in a continuous bridge girder.
Figure 4-6.5 Tendon profiles in a continuous bridge girder

(Courtesy: VSL International Ltd.)
The profile is implemented by the use of hangers or cross bars or chairs of varying
depth at regular intervals. In beams, the duct is supported by hangers from the top bars
or by cross bars attached to the stirrups. The depth of the hanger or cross bar at a

location can be calculated from the equation of the profile. In slabs, the duct is
supported on chairs resting on the form work.
The CGS of the tendon shifts from the centre line of the duct after stretching. The
following sketches show the shifts at the low and high points of the tendon. The shift in
the CGS is available from the type of tendon used and can be accounted for in precise
calculations.
CGS
Centre line
of duct
CGS
Shift
At low point At high point
Figure 4-6.6 Shift in the CGS of a tendon from the centreline of duct
4.6.2 Minimum Amounts of Reinforcement
Minimum Longitudinal Reinforcement

A minimum amount of longitudinal reinforcement should be provided to have sufficient
strength after the cracking of concrete.
According to Section 18.6.3.3-a, the minimum amount is as follows.
Minimum (As+ AP) = 0.2% A (4-6.3)

Here,
As = area of steel without prestressing,
Ap = area of prestressing steel,
A = total area of cross-section.

As Ap
Figure 4-6.7 Cross-section of a beam showing longitudinal reinforcement
The minimum reinforcement can be reduced to 0.15% A, if high yield strength deformed
bars are used.
Minimum Longitudinal Reinforcement with Unbonded Tendon

In a post-tensioned member when the ducts are not grouted, beyond the cracking load,
the number of cracks is small and the crack width is large. To reduce the crack width, a
minimum amount of non-prestressed reinforcement should be provided.
Since the non-prestressed reinforcement is bonded to the concrete, there are several
cracks with small crack width.
Unbonded
tendon
Unbonded
Non-prestressed
tendon
reinforcement
Figure 4-6.8 Crack pattern with and without non-prestressed reinforcement for beams
with unbonded tendon
As per the code of the American Concrete Institute (ACI 318), the minimum amount of
such reinforcement (As) is 0.4% At , where At is the area under tension between the

centroid of the section (CGC) and the tension edge. The above reinforcement is not
intended to provide flexural strength.
CGC
At Unbonded tendon
As
Figure 4-6.9 Cross-section of a beam showing longitudinal reinforcement and area
under tension
Minimum Side Face Reinforcement

When the depth of the web exceeds 500 mm, a minimum amount of longitudinal
reinforcement should be placed at each face (side face) of the web to check thermal
and shrinkage cracks.
According to Section 18.6.3.3-b, the minimum amount of side face reinforcement (As,sf )
is given as follows.
Minimum As,sf = 0.05% Aw (4-6.4)
Here,
Aw = vertical area of the web.
The maximum spacing of the bars is 200 mm.
As,sf As,sf
Figure 4-6.10 Cross-section of a beam showing side face reinforcement

4.6.3 Miscellaneous Requirements
Minimum Cover Requirements

A minimum clear cover of concrete is necessary to protect the steel against corrosion
and to develop adequate bond between concrete and steel. The cover is implemented
by chairs or blocks.
Clear cover
Figure 4-6.11 Cross-section of a beam showing cover
According to Section 11.1.6, the minimum cover requirements are as follows.
For pre-tensioned members, minimum cover for tendons is 20 mm. For post-tensioned
members, minimum cover for sheathing (duct) is 30 mm or size of the tendon.
The minimum cover should be increased by 10 mm in aggressive environment.
Minimum Spacing Requirements

A minimum clear spacing of the tendons or reinforcing bars is necessary for the flow of
concrete during casting and for the bond between concrete and steel.
Clear spacing
Figure 4-6.12 Cross-section of a beam showing spacing between tendons
According to Section 11.1.7, the minimum spacing requirements are as follows.

For single wires in a pre-tensioned member,
Clear spacing 3 wire diameter
1 maximum aggregate size.

For large bars or tendons,

Clear spacing 40 mm
maximum size of tendon / bar
maximum aggregate size + 5 mm.
For grouped tendons (maximum four tendons per group), the requirement is for the
spacing between the groups of tendons.
Vertical spacing
Horizontal spacing
Figure 4-6.13 Cross-section of a beam showing spacing between groups of tendons
According to Section 11.1.8, for grouped tendons the spacing requirements are as
follows.
Horizontal spacing 40 mm
maximum aggregate size + 5 mm
Vertical spacing 50 mm.
Anchorage of Reinforcement
In a partially prestressed section, where the non-prestressed reinforcement contributes
to flexural strength, the development length of the bars needs to be checked at the
critical section. The bars should be anchored at the supports by hooks to avoid
anchorage failure.
The following photo shows the fabrication of the reinforcement for a post-tensioned box-
girder of a bridge.

Figure 4-6.14 Fabrication of reinforcement


5.1 Analysis for Shear

Stress in an Uncracked Beam
Types of Cracks
Components of Shear Resistance
Modes of Failure
Effect of Prestressing Force
Introduction
The analysis of reinforced concrete and prestressed concrete members for shear is
more difficult compared to the analyses for axial load or flexure.
The analysis for axial load and flexure are based on the following principles of
mechanics.
1) Equilibrium of internal and external forces
2) Compatibility of strains in concrete and steel
3) Constitutive relationships of materials.
The conventional analysis for shear is based on equilibrium of forces by a simple

equation. The compatibility of strains is not considered. The constitutive relationships
(relating stress and strain) of the materials, concrete or steel, are not used. The strength
of each material corresponds to the ultimate strength.
The strength of concrete under shear although based on test results, is empirical in
nature.
Shear stresses generate in beams due to bending or twisting. The two types of shear
stress are called flexural shear stress and torsional shear stress, respectively. In this
section, the analysis for shear refers to flexural shear stress. The torsional shear stress
is covered in Section 5.4, Analysis for Torsion.
To understand flexural shear stress, the behaviour of a simply supported beam under
uniformly distributed load, without prestressing, will be explained first. The effect of
prestressing force will be subsequently introduced. The presentation will be in the
following sequence.

1) Stresses in an uncracked (homogenous) beam.

2) Types of cracks generated due to the combination of flexure and shear.
3) Components of shear resistance and the modes of failure.
4) Effect of prestressing force.
5.1.1 Stresses in an Uncracked Beam
The following figure shows the variations of shear and moment along the span of a
simply supported beam under a uniformly distributed load. The variations of normal
stress and shear stress along the depth of a section of the beam are also shown.
b
h 1
2
Variation Variation
Shear of normal of shear
force V stress (f) stress (v)
diagram
Momentd
iagram
M
Figure 5-1.1 Variations of forces and stresses in a simply supported beam
Under a general loading, the shear force and the moment vary along the length. The
normal stress and the shear stress vary along the length, as well as along the depth.
The combination of the normal and shear stresses generate a two-dimensional stress
field at a point. At any point in the beam, the state of two-dimensional stresses can be
expressed in terms of the principal stresses. The Mohrs circle of stress is helpful to
understand the state of stress.
Before cracking, the stress carried by steel is negligible. When the principal tensile
stress exceeds the cracking stress, the concrete cracks and there is redistribution of
stresses between concrete and steel. For a point on the neutral axis (Element 1), the
shear stress is maximum and the normal stress is zero. The principal tensile stress (1)
is inclined at 45 to the neutral axis. The following figure shows the state of in-plane
stresses.

v
v 1
2
2
1 2 1
= 45o
State of Principal
pure shear stresses Mohrs circle
Figure 5-1.2 State of stresses at a point on the neutral axis of a beam
At the level of neutral axis, the normal stress is zero and the shear stress is maximum.
An element at that level is under pure shear. A state of pure shear can be conceived as
a state of biaxial tensile-compressive stresses. These principle stresses are inclined at
45 with respect to the axis of the beam. It is necessary to study the principle stresses
to understand the cracking of concrete. The Mohrs circle is a representation of the
state of in-plane stresses on surfaces of various inclinations passing through a point.
The horizontal and vertical axes represent the normal and shear stresses, respectively.
For a state of pure shear, the centre of the Mohrs circle coincides with the origin of the
axes. It is expected that the reader is familiar with these concepts from a course in
strength of materials.
Since the shear force is maximum near the supports, cracks due to shear occur near
the supports. The cracks are formed around the neutral axis and perpendicular to the
principal tensile stress (1). The cracks are thus inclined at 45 to the axis of the beam.
The following sketch shows the inclination of the cracks forming at the neutral axis.
v 1
2
1
Figure 5-1.3 Inclination of crack at the level of neutral axis
For a point near the bottom edge of the beam (Element 2), the normal stress is
maximum and the shear stress is close to zero. The principal tensile stress (1) is
almost parallel to the bottom edge. The angle of inclination of 1 with respect to the axis
of the beam () is much smaller than 45. The following figure shows the state of in-
plane stresses.

v 2 (f,v)
1
2 f 2 1
2
< 45o
Shear stress and Principal

normal stress stresses Mohrs circle
Figure 5-1.4 State of stresses at a point close to the edge under tension
Adjacent to the bottom edge (edge under tension), the tensile stress due to flexure is
maximum and the shear stress is zero. The state of stress is nearly uniaxial tensile
stress. The principal compressive stress is negligible. The Mohrs circle is shifted
towards the axis of principal tensile stress.
Since the moment is maximum at mid span, cracks due to flexure occur near mid span.
The cracks are formed at the bottom edge and perpendicular to 1. Since 1 is parallel
to the edge, the cracks are perpendicular to the edge.
v 2
1
2 f
Figure 5-1.5 Inclination of crack close to the edge under tension
The previous concepts can be used to develop the principal stress trajectories. The
following figure shows the trajectories for a simply supported beam under a uniformly
distributed load. The crack pattern can be predicted from these trajectories.
Figure 5-1.6 Principle stress trajectories

(Courtesy: Pillai, S. U., and Menon, D., Reinforced Concrete Design)

5.1.2 Types of Cracks
The types and formation of cracks depends on the span-to-depth ratio of the beam and
loading. These variables influence the moment and shear along the length of the beam.
For a simply supported beam under uniformly distributed load, without prestressing,
three types of cracks are identified.
1) Flexural cracks: These cracks form at the bottom near the midspan and
propagate upwards.
2) Web shear cracks: These cracks form near the neutral axis close to the support
and propagate inclined to the beam axis.
3) Flexure shear cracks: These cracks form at the bottom due to flexure and
propagate due to both flexure and shear.
In the following figure, the formation of cracks for a beam with large span-to-depth ratio
and uniformly distributed loading is shown.
a) Initiation of flexural cracks
b) Growth of flexural cracks and formation of flexure shear

and web shear cracks.
Web Flexure Flexural Flexure Web

shear shear cracks shear shear
cracks cracks cracks cracks
c) Cracks before failure

Figure 5-1.7 Formation of cracks in a reinforced concrete beam

5.1.3 Components of Shear Resistance
The components of shear resistance are studied based on the internal forces at a
flexure shear crack. The forces are shown in the following figure.
Vcz
Va Vs
VP
Vd
Figure 5-1.8 Internal forces at a flexure shear crack
The notations in the previous figure are as follows.

Vcz = Shear carried by uncracked concrete
Va = Shear resistance due to aggregate interlock
Vd = Shear resistance due to dowel action
Vs = Shear carried by stirrups
Vp = Vertical component of prestressing force in inclined tendons
The magnitude and the relative value of each component change with increasing load.
5.1.4 Modes of Failure
For beams with low span-to-depth ratio or inadequate shear reinforcement, the failure
can be due to shear. A failure due to shear is sudden as compared to a failure due to
flexure. The following five modes of failure due to shear are identified.
1) Diagonal tension failure
2) Shear compression failure
3) Shear tension failure
4) Web crushing failure
5) Arch rib failure
The occurrence of a mode of failure depends on the span-to-depth ratio, loading, cross-
section of the beam, amount and anchorage of reinforcement. The modes of failure are
explained next (Courtesy: Pillai, S. U., and Menon, D., Reinforced Concrete Design).

Diagonal tension failure

In this mode, an inclined crack propagates rapidly due to inadequate shear
reinforcement.
Figure 5-1.9 Diagonal tension failure
Shear compression failure

There is crushing of the concrete near the compression flange above the tip of the
inclined crack.
Figure 5-1.10 Shear compression failure
Shear tension failure

Due to inadequate anchorage of the longitudinal bars, the diagonal cracks propagate
horizontally along the bars.
Figure 5-1.11 Shear tension failure

Web crushing failure

The concrete in the web crushes due to inadequate web thickness.
Figure 5-1.12 Web crushing failure
Arch rib failure

For deep beams, the web may buckle and subsequently crush. There can be
anchorage failure or failure of the bearing.
Figure 5-1.13 Arch rib failure
The objective of design for shear is to avoid shear failure. The beam should fail in
flexure at its ultimate flexural strength. Hence, each mode of failure is addressed in the
design for shear. The design involves not only the design of the stirrups, but also
limiting the average shear stress in concrete, providing adequate thickness of the web
and adequate development length of the longitudinal bars.
5.1.5 Effect of Prestressing Force
In presence of prestressing force, the flexural cracking occurs at a higher load. For
Type 1 and Type 2 sections, there is no flexural crack under service loads. This is
evident from the typical moment versus curvature curve for a prestressed section (refer
to Section 3.6, Analysis of Member under Flexure (Part V)). In presence of prestressing
force, the web shear cracks also generate under higher load.

With increase in the load beyond the cracking load, the cracks generate in a similar
sequence. But, the inclinations of the flexure shear and web shear cracks are reduced
depending on the amount of prestressing and the profile of the tendon.
The effect of prestressing force is explained for a beam with a concentric effective
prestressing force (Pe).
Pe Pe
1
Figure 5-1.14 A simply supported beam under concentric prestress and uniformly
distributed loads
For a point at the neutral axis (Element 1), there is normal stress due to the prestressing
force (fpe). The principal tensile stress (1) is inclined to the neutral axis at an angle
greater than 45. With the combination of shear stress, the principal compressive stress
(2) is inclined to the neutral axis at an angle much smaller than 45. The following
figure shows the state of in-plane stresses.
v 1
(fpe,v)
2 2 > 90
fpe fpe
1 2 1
> 45o
Shear stress and Principal
prestress stresses Mohrs circle
Figure 5-1.15 State of stresses at a point on the neutral axis for a prestressed beam
In the following figure, the formation of cracks for a prestressed beam with large span-
to-depth ratio and uniformly distributed loading is shown. This figure can be compared
with that for a reinforced concrete beam.
Pe Pe
Figure 5-1.16 Formation of cracks in a prestressed beam

After cracking, in presence of prestressing force, the length and crack width of a
diagonal crack are low. Thus, the aggregate interlock and zone of concrete under
compression are larger as compared to a non-prestressed beam under the same load.
Hence, the shear strength of concrete (Vc) increases in presence of prestressing force.
This is accounted for in the expression of Vc.

5.2 Design for Shear (Part I)

General Comments
Limit State of Collapse for Shear
5.2.1 General Comments
Calculation of Shear Demand

The objective of design is to provide ultimate resistance for shear (VuR) greater than the
shear demand under ultimate loads (Vu). For simply supported prestressed beams, the
maximum shear near the support is given by the beam theory. For continuous
prestressed beams, a rigorous analysis can be done by the moment distribution method.
Else, the shear coefficients in Table 13 of IS:456 - 2000 can be used under conditions
of uniform cross-section of the beams, uniform loads and similar lengths of span.
Design of Stirrups
The design is done for the critical section. The critical section is defined in Clause
22.6.2 of IS:456 - 2000. In general cases, the face of the support is considered as the
critical section.
When the reaction at the support introduces compression at the end of the beam, the
critical section can be selected at a distance effective depth from the face of the support.
The effective depth is selected as the greater of dp or ds.

dp = depth of CGS from the extreme compression fiber
ds = depth of centroid of non-prestressed steel.
Since the CGS is at a higher location near the support, the effective depth will be equal
to ds.
To vary the spacing of stirrups along the span, other sections may be selected for
design. Usually the following scheme is selected for beams under uniform load.
1) Close spacing for quarter of the span adjacent to the supports.
2) Wide spacing for half of the span at the middle.

For large beams, more variation of spacing may be selected. The following sketch
shows the typical variation of spacing of stirrups. The span is represented by L.
L/4 L/2 L/4

Figure 5-2.1 Typical variation of spacing of stirrups
5.2.2 Limit State of Collapse for Shear
The shear is studied based on the capacity of a section which is the limit state of
collapse. The capacity (or ultimate resistance) of a section (VuR) consists of a concrete
contribution (Vc) and the stirrup contribution (VS).
VuR = VC + VS (5-2.1)
Vc includes Vcz (contribution from uncracked concrete), Va (aggregate interlock) and Vd

(dowel action).
The value of Vc depends on whether the section is cracked due to flexure. Section 22.4
of IS:1343 - 1980 gives two expressions of Vc, one for cracked section and the other for
uncracked section. Usually, the expression for the uncracked section will govern near
the support. The expression for the cracked section will govern near the mid span. Of
course, both the expressions need to be evaluated at a particular section. The lower
value obtained from the two expressions is selected.
For uncracked sections,
Vc = Vco
Vc = 0.67bD ft 2 + 0.8fcp ft (5-2.2)
Vco is the shear causing web shear cracking at CGC.
In the above expression,

= bw, breadth of the web for flanged sections

D = total depth of the section (h)

ft = tensile strength of concrete = 0.24fck
fcp = compressive stress in concrete at CGC due to the prestress
= Pe/A.
The value of fcp is taken as positive (numeric value). Note that, a reduced effective
prestress needs to be considered in the transmission length (explained in Section 7.1)
region of a pre-tensioned beam.
The previous equation can be derived based on the expression of the principal tensile
stress (1) at CGC.
v 1
(fcp,v)
fcp fcp 2
2 1
State of stress Principal stresses Mohrs circle

at CGC
Figure 5-2.2 State of stresses at a point on the neutral axis for a prestressed beam
The principal tensile stress is equated to the direct tensile strength of concrete (ft).
fcp fcp2
1 = - + + v2
2 4
fcp fcp2 V Q
2
=- + + c 0
2 4 Ib
= ft (5-2.3)

I = gross moment of inertia
Q = At y
At = area of section above CGC
y = vertical distance of centroid of At from CGC.

At
+
CGC y
Figure 5-2.3 Cross-section of a beam showing the variables for calculating shear
stress in the web
Transposing the terms,
Ib 2
Vc 0 = ft + fcp ft
Q
0.67bD ft 2 + 0.8fcp ft (5-2.4)
The term 0.67bD represents Ib/Q for the section. It is exact for a rectangular section and
conservative for other sections.
To be conservative, only 80% of the prestressing force is considered in the term 0.8fcp.
For a flanged section, when the CGC is in the flange, the intersection of web and flange
is considered to be the critical location. The expression of Vc0 is modified by substituting
0.8fcp with 0.8 (the stress in concrete at the level of the intersection of web and flange).
In presence of inclined tendons or vertical prestress, the vertical component of the
prestressing force (Vp) can be added to Vc0.
Vc Vc 0 +Vp
= 0.67bD ft 2 + 0.8fcp ft +Vp (5-2.5)
For cracked sections,

Vc = Vcr
f V
Vc = 1- 0.55 pe c bd + M0 u
fpk Mu
0.1bd fck (5-2.6)

Vcr is the shear corresponding to flexure shear cracking. The term (1 0.55fpe /fpk)cbd
is the additional shear that changes a flexural crack to a flexure shear crack.
The notations in the previous equation are as follows.

fpe = effective prestress in the tendon after all losses
0.6fpk
fpk = characteristic strength of prestressing steel
c = ultimate shear stress capacity of concrete, obtained from Table 6 of
IS:1343 - 1980. It is given for values of Ap / bd, where d is the depth of
CGS. The values are plotted in the next figure.
= bw , breadth of the web for flanged sections
d = distance from the extreme compression fibre to the centroid of the
tendons at the section considered
M0 = moment initiating a flexural crack
Mu = moment due to ultimate loads at the design section
Vu = shear due to ultimate loads at the design section.
1.2
c (N/mm2)
0.8
0.4
0
0 1 2 3 4
A p /bd x 100
M30 M40
Figure 5-2.4 Variation of shear strength of concrete
The term (M0/Mu)Vu is the shear corresponding to the moment M0, that decompresses
(nullifies the effect of prestress) the tension face and initiates a flexural crack. The
expression of M0 is given below.
I
M0 = 0.8fpt (5-2.7)
y


fpt = magnitude of the compressive stress in concrete at the level of CGS due to
prestress only.
An equal amount of tensile stress is required to decompress the concrete at the level of
CGS. The corresponding moment is fptI / y.
In the expression of M0,

I = gross moment of inertia
y = depth of the CGS from CGC.
The factor 0.8 implies that M0 is estimated to be 80% of the moment that decompresses
the concrete at the level of CGS. Since the concrete is cracked and the inclination of
tendon is small away from the supports, any vertical component of the prestressing
force is not added to Vcr.
Maximum Permissible Shear Stress

To check the crushing of concrete in shear compression failure, the shear stress is
limited to a maximum value (c,max). The value of c,max depends on the grade of
concrete and is given in Table 7 of IS:1343 - 1980.
Vu
c,max (5-2.8)
bdt
In the previous expression,

dt = greater of dp or ds
ds = depth of centroid of regular steel
Vu = shear force at a section due to ultimate loads.

c, max (N/mm )
2
4
0
30 40 50 60
2
f ck (N/mm )
Figure 5-2.5 Variation of maximum permissible shear stress in concrete

5.3 Design for Shear (Part II)

Design of Transverse Reinforcement
Detailing Requirements
Design Steps
5.3.1 Design of Transverse Reinforcement
When the shear demand (Vu) exceeds the shear capacity of concrete (Vc), transverse
reinforcements in the form of stirrups are required. The stirrups resist the propagation of
diagonal cracks, thus checking diagonal tension failure and shear tension failure.
The stirrups resist a failure due to shear by several ways. The functions of stirrups are
listed below.
1) Stirrups resist part of the applied shear.
2) They restrict the growth of diagonal cracks.
3) The stirrups counteract widening of the diagonal cracks, thus maintaining
aggregate interlock to a certain extent.
4) The splitting of concrete cover is restrained by the stirrups, by reducing dowel
forces in the longitudinal bars.
After cracking, the beam is viewed as a plane truss. The top chord and the diagonals
are made of concrete struts. The bottom chord and the verticals are made of steel
reinforcement ties. Based on this truss analogy, for the ultimate limit state, the total area
of the legs of the stirrups (Asv) is given as follows.
Asv V -V (5-3.1)
= u c
sv 0.87fy dt
The notations in the above equation are explained.

sv = spacing of the stirrups
ds = depth of centroid of non-prestressed steel
fy = yield stress of the stirrups

The grade of steel for stirrups should be restricted to Fe 415 or lower.
Design of Stirrups for Flanges

For flanged sections, although the web carries the vertical shear stress, there is shear
stress in the flanges due to the effect of shear lag. Horizontal reinforcement in the form
of single leg or closed stirrups is provided in the flanges. The following figure shows
the shear stress in the flange at the face of the web.
Figure 5-3.1 Shear stress in flange due to shear lag effect
The horizontal reinforcement is calculated based on the shear force in the flange. The
relevant quantities for the calculation based on an elastic analysis are as follows.
1) Shear flow (shear stress width)
2) Variation of shear stress in a flange (f)
3) Shear forces in flanges (Vf).
4) Ultimate vertical shear force (Vu)
The following sketch shows the above quantities for an I-section (with flanges of
constant widths).
f f max
bf
Df
Vf Vf
Vu
Vf Vf
Shear flow Shear forces

Figure 5-3.2 Shear flow and shear forces in an I-section

The design shear force in a flange is given as follows.
f max bf
Vf = Df (5-3.2)
2 2
Here,
f,max = maximum shear stress in the flange.
The maximum shear stress in the flange is given by an expression similar to that for the
shear stress in web.
Vu A1 y
f max = (5-3.3)
I Df
Here,
Vu = ultimate vertical shear force
I = moment of inertia of the section.
A1 = area of half of the flange
= distance of centroid of half of the flange from the neutral axis at CGC.
y
A1
Figure 5-3.3 Cross-section of a beam showing the variables for calculating shear
stress in the flange
The amount of horizontal reinforcement in the flange (Asvf) is calculated from Vf.
Vf
Asvf = (5-3.4)
0.87fy
The yield stress of the reinforcement is denoted as fy .

5.3.2 Detailing Requirements
The detailing requirements for the stirrups in IS:1343 - 1980 are briefly mentioned.
Maximum Spacing of Stirrups

The spacing of stirrups (sv) is restricted so that a diagonal crack is intercepted by at
least one stirrup. This is explained by the following sketch.
dp
bw ds h
sv
Cross-section Elevation
Figure 5-3.4 Cross-section and elevation of a beam showing stirrups
As per Clause 22.4.3.2, the maximum spacing is 0.75dt or 4bw , whichever is smaller.
When Vu is larger than 1.8Vc , the maximum spacing is 0.5dt.
The variables are as follows.
bw = breadth of web
ds = depth of centroid of non-prestressed steel
Vu = shear force at a section due to ultimate loads
Vc = shear capacity of concrete.
Minimum Amount of Stirrups

A minimum amount of stirrups is necessary to restrict the growth of diagonal cracks and
subsequent shear failure. For Vu < Vc, minimum amount of transverse reinforcement is
provided based on the following equation.
Asv 0.4
= (5-3.5)
bsv 0.87fy


= bw, breadth of the web for flanged sections.
If Vu < 0.5Vc and the member is of minor importance, stirrups may not be provided.
Another provision for minimum amount of stirrups (Asv,min) is given by Clause 18.6.3.2
for beams with thin webs. The minimum amount of stirrups is given in terms of Awh , the
horizontal sectional area of the web in plan. The area is shown in the following sketch.
A A
Elevation
Awh
Section A - A
Figure 5-3.5 Elevation and horizontal section of a beam showing stirrups
In presence of dynamic load,

Asv,min = 0.3% Awh
= 0.2% Awh , when h 4bw
With high strength bars,

Asv,min = 0.2% Awh
= 0.15% Awh , when h 4bw
In absence of dynamic load, when h > 4bw

Asv,min = 0.1% Awh
There is no specification for Asv,min when h 4bw .

Anchorage of Stirrups
The stirrups should be anchored to develop the yield stress in the vertical legs.
1) The stirrups should be bent close to the compression and tension surfaces,
satisfying the minimum cover.
2) Each bend of the stirrups should be around a longitudinal bar. The diameter of
the longitudinal bar should not be less than the diameter of stirrups.
3) The ends of the stirrups should be anchored by standard hooks.
4) There should not be any bend in a re-entrant corner. In a re-entrant corner, the
stirrup under tension has the possibility to straighten, thus breaking the cover
concrete.
The following sketches explain the requirement of avoiding the bend of a stirrup at a re-
entrant corner.
Re-entrant corner
Incorrect detailing Correct detailing

Figure 5-3.6 Cross-section of the bottom flange of a beam showing stirrups
Minimum Thickness (Breadth) of Web

To check web crushing failure, The Indian Roads Congress Code IRC:18 - 2000
specifies a minimum thickness of the web for T-sections (Clause 9.3.1.1). The minimum
thickness is 200 mm plus diameter of the duct hole.
5.3.3 Design Steps
The following quantities are known.

Vu = factored shear at ultimate loads. For gravity loads, this is calculated
from VDLand VLL.
VDL = shear due to dead load
VLL = shear due to live load.

After a member is designed for flexure, the self-weight is known. It is included as dead
load.
The grade of concrete is known from flexure design. The grade of steel for stirrups is
selected before the design for shear. As per IS:1343 - 1980, the grade of steel is
limited to Fe 415.
The following quantities are unknown.
Vc = shear carried by concrete
Asv = total area of the legs of stirrups within a distance sv
sv = spacing of stirrups.
The steps for designing stirrups along the length of a beam are given below.
1) Calculate the shear demand Vu at the critical location.
2) Check (Vu / bdt) < c,max. If it is not satisfied, increase the depth or breadth of the
section. Here, b is the breadth of the web (bw) and dt is larger of dp and ds.
3) Calculate the shear capacity of concrete Vc from the lower of Vco and Vcr. In
presence of inclined tendons or vertical prestress, the vertical component of the
prestressing force (Vp) can be added to Vc0.
4) Calculate the requirement of shear reinforcement through Asv / sv . Compare the
value with the minimum requirement.
5) Calculate the maximum spacing and round it off to a multiple of 5 mm.
6) Calculate the size and number of legs of the stirrups based on the amount
required, type of section and space to accommodate.
Repeat the calculations for other locations of the beam, if the spacing of stirrups needs
to be varied.

Example 5-3.1
Design the stirrups for the Type 1 prestressed beam with the following section
(location of tendons shown at mid span).
435
100
100
CGC
290 920
CGS

with Pe = 826 kN
Longitudinal reinforcement of 12 mm diameter is provided to hold the stirrups.
The properties of the sections are as follows.

A = 159,000 mm2,
I = 1.7808 1010 mm4
Ap = 960 mm2
The grade of concrete is M 35 and the characteristic strength of the prestressing

steel (fpk) is 1470 N/mm2. The effective prestress (fpe) is 860 N/mm2.
The uniformly distributed load including self weight, is wT = 30.2 kN/m.
The span of the beam (L) is 10.7 m. The width of the bearings is 400 mm. The
clear cover to longitudinal reinforcement is 30 mm.
Solution
1) Calculate Vu at the face of the support (neglecting the effect of compression in

concrete).
L
Vu = 1.5 wT - x
2
10.7
= 1.530.2 - 0.2
2
= 233.3 kN

Here, x denotes half of the width of bearing. x = 200 mm.
2) Check (Vu / bdt) < c,max .
Effective depth dt = total depth cover diameter of stirrups diameter of

longitudinal bar.
Assume the diameter of stirrups to be 8 mm.
1
dt = 920 - 30 - 8 - 12
2
= 876 mm
Vu 233.3 103
=
bw dt 100 876
= 2.7 N/mm2
c,max for M 35 is 3.7 N/mm2. Hence, (Vu / bdt) < c,max .
3) Calculate Vc from the lower of Vc0 and Vcr .
Vco = 0.67bD ft 2 + 0.8fcp ft
Here,
ft = 0.24 35 Pe
fcp =
= 1.42 N/mm2 A
826103
=
159,000
= 5.19 N/mm2
= 0.67 100 920 1.422 + 0.85.19 1.42

= 173.4 kN
The vertical component of the prestressing force can be found out from the equation of
the parabolic tendon.
4ym
y = x (L - x )
L2

ym x

y
The following is the expression of the slope of the parabolic tendon.
dy 4ym
ta n = = (L - 2 x )
dx L2
ym x

y
At x = 0.2 m, y = 20 mm, dy/dx = 0.105 and = 6.0.
Vp = Pe sin
= 826 0.104
= 86.0kN
Vco +Vp = 173.4 + 86.0

= 259.4 kN

f
Vp Pe
V
Vcr = 1- 0.55 pe c bd + M0 u
fpk Mu
Here, fpe 860

= = 0.58
fpk 1470
100 Ap 100 960

=
bd 100 480
= 2.0
d = 460 + y
= 460 + 20
= 480 mm

From Table 6, for M 35 concrete, c = 0.86 N/mm2.

I
M0 = 0.8fpt
y
Here, Pe Pe y
fpt = - - y
A I
826103 826 103 20
=- - 20
159,000 1.7808 1010
= -5.19 - 0.02
= -5.21N/mm2
1.7808 1010
M0 = 0.85.21
20
6
= 3711.210 Nmm
= 3711.2 kNm
At the critical section,
x
Mu = 1.5wT (L - x )
2
0.2
= 1.530.2 (10.7 - 0.2)
2
= 47.6 kNm
f V
Therefore, Vcr = 1- 0.55 pe c bd + M0 u
fpk Mu
0.86 233.3
Vcr = (1- 0.550.58) 3
100 480 + 3711.2
10 47.6
= 28.1+18204.8
= 18232.9 kN
The governing value of Vc is 259.4 kN.
Vu < Vc .
4) Calculate Asv / sv .
Provide minimum steel.
Asv 0.4
=
bw sv 0.87fy

5) Calculate maximum spacing
sv = 0.75 dt = 0.75 876 = 656 mm

sv = 4bw = 4 100 = 400 mm
Select sv = 400 mm.
6) Calculate the size and number of legs of the stirrups
Select fy = 250 N/mm2.

0.4
Asv = bw sv
0.87fy
0.4
= 100 400
0.87 250
= 73.6 mm2
Provide 2 legged stirrups of diameter 8 mm.
Asv,provided = 250.3
= 100.6 mm2
Check minimum amount of stirrups.
Asv,min = 0.1% Awh

0.1
= 100 400
100
= 40 mm2
Provided amount of stirrups is larger. OK.
Provide same spacing of stirrups throughout the span.
Design of stirrups for flange
1
A1 = bf Df
2
1
= 435 100
2
= 21750 mm2

y = 410 mm
Vu A1 y
f max =
I Df
233.3103 21750 410
=
1.7808 1010 100
= 1.17 N/mm2
f max bf
Vf = Df
2 2
1.17 435
= 100
2 2
= 12724 N
Vf
Asvf =
0.87fy
12724
=
0.87 250
= 59.0 mm2
0.4
For minimum steel Asvf = Df sv
0.87fy
0.4
= 100 400
0.87 250
= 73.6 mm2
Provide 2 legged stirrups of diameter 8 mm.
Designed section
8 mm diameter stirrups
@ 400 mm c/c

5.4 Analysis for Torsion

Stresses in an Uncracked Beam
Crack Pattern Under Pure Torsion
Components of Resistance for Pure Torsion
Modes of Failure
Introduction
The analysis of reinforced concrete and prestressed concrete members for torsion is
more difficult compared to the analyses for axial load or flexure.
The analysis for axial load and flexure are based on the following principles of
mechanics.
1) Equilibrium of internal and external forces
2) Compatibility of strains in concrete and steel
3) Constitutive relationships of materials.
The conventional analysis of reinforced concrete and prestressed concrete members for
torsion is based on equilibrium of forces by simple equations. The compatibility of
strains in concrete and steel reinforcement is not considered.
The strength of each material, concrete or steel, corresponds to the ultimate strength.
The constitutive relationship of each material, relating stress and strain, is not used.
Torsion generated in a member can be classified into two types based on the necessity
of analysis and design for torsion.
1) Equilibrium torsion: This is generated due to loading eccentric to the centroidal
axis. For example, a) in a beam supporting cantilever slab or precast slab or floor
joists on one side, b) in a (curved) bridge deck subjected to eccentric live load
and c) in an electric pole subjected to loads from wires on one side.
The torsion demand is determined by equilibrium condition only. The member needs to
be analysed and designed for torsion. The following figure shows the situations where
eccentric loads are acting on the members.

+ CGC
CGC +
a) Bridge deck b) L-beam

Figure 5-4.1 Examples of members under eccentric load
2) Compatibility torsion: This is generated by twisting, to maintain compatibility in

deformation with the connected member. This type of torsion generates in a
primary beam supporting secondary beams.
In compatibility torsion, the torsion demand is determined by both equilibrium and

compatibility conditions. Else, the torsion can be neglected. This implies that primary
beam need not be analysed and designed for torsion, if the secondary beams are
designed as pin supported.
In this section, the emphasis is laid on equilibrium torsion. To understand the behaviour
of a beam under torsion, the presentation will be in the following sequence.
1) Stresses in an uncracked (homogenous) rectangular beam without prestressing
due to pure torsion (in absence of flexure), with constant torque along the span.
2) Crack pattern under pure torsion.
3) Components of resistance for pure torsion.
4) Modes of failure under combined torsion and flexure.
5) Effect of prestressing force.
Although pure torsion is absent in structures, understanding the behaviour of a beam
under pure torsion helps to analyse a beam under combined torsion, flexure and shear.
5.4.1 Stresses in an Uncracked Beam
The following figure shows a beam of rectangular cross-section under pure torsion. The
variations of the torsional shear stress () along radial lines in the cross-section are
shown. It can be observed that the maximum shear stress (max) occurs at the middle of

the longer side. Hence, the subsequent explanation will refer to the stress condition at
the middle of the longer side.
max
1
T Variation of
torsional shear
stress () in the
cross-section
Figure 5-4.2 Beam subjected to pure torsion
At any point in the beam, the state of two-dimensional stresses can be expressed in
terms of the principal stresses. The Mohrs circle of stress is helpful to understand the
state of stress.
Before cracking, the stress carried by steel is negligible. When the principal tensile
stress exceeds the cracking stress, the concrete cracks and there is redistribution of
stresses between concrete and steel.
For a point at the middle of the longer side (Element 1), the torsional shear stress is
maximum. The principal tensile stress (1) is inclined at 45 to the beam axis.

1
2
2
1 2 1
= 45o
State of Principal
pure shear stresses Mohrs circle
Figure 5-4.3 State of stresses at the side of a beam
Since the torsion is maximum at middle of the longer side, cracks due to torsion occur
around that location and perpendicular to 1.

1
2
1
Crack inclination
Figure 5-4.4 Inclination of crack at the side of a beam
5.4.2 Crack Pattern under Pure Torsion
The cracks generated due to pure torsion follow the principal stress trajectories. The
first cracks are observed at the middle of the longer side. Next, cracks are observed at
the middle of the shorter side. After the cracks connect, they circulate along the
periphery of the beam.

a) Initiation of torsional cracks in longer side
b) Initiation of torsional cracks in shorter side
c) Spiral torsional cracks

Figure 5-4.5 Formation of cracks in a beam subjected to pure torsion
In structures, a beam is not subjected to pure torsion. Along with torsion it is also
subjected to flexure and shear. Hence, the stress condition and the crack pattern are
more complicated than shown before.
5.4.3 Components of Resistance for Pure Torsion
After cracking, the concrete forms struts carrying compression. The reinforcing bars act
as ties carrying tension. This forms a space truss. Since the shear stress is larger near
the sides, the compression in concrete is predominant in the peripheral zone. This is
called the thin-walled tube behaviour. The thickness of the wall is the shear flow zone,

where the shear flow is assumed to be constant. The portion of concrete inside the
shear flow zone can be neglected in calculating the capacity.
The components in vertical and horizontal sections of a beam are shown below.
Figure 5-4.6 Internal forces in a beam
The components can be denoted as below.

Tc = torsion resisted by concrete
Ts = torsion resisted by the longitudinal and transverse reinforcing bars.
The magnitude and the relative value of each component change with increasing torque.
5.4.4 Modes of Failure
For a homogenous beam made of brittle material, subjected to pure torsion, the
observed plane of failure is not perpendicular to the beam axis, but inclined at an angle.
This can be explained by theory of elasticity. A simple example is illustrated by applying
torque to a piece of chalk.

Failure
surface
Figure 5-4.7 Failure of a piece of chalk under torque
For a beam of rectangular section, the plane of failure is further influenced by warping.
Torsional warping is defined as the differential axial displacement of the points in a
section perpendicular to the axis, due to torque.
For a reinforced concrete beam, the length increases after cracking and after yielding of
the bars. For a beam subjected to flexure and torsion simultaneously, the modes of
failure are explained by the Skew Bending Theory. The observed plane of failure is
not perpendicular to the beam axis, but inclined at an angle. The curved plane of failure
is idealised as a planar surface inclined to the axis of the beam.
The skew bending theory explains that the flexural moment (Mu) and torsional moment
(Tu) combine to generate a resultant moment inclined to the axis of the beam. This
moment causes compression and tension in a planar surface inclined to the axis of the
beam. The following figure shows the resultant moment due to flexural moment and
torsion in a beam.
Mu Tu
Resultant moment
Figure 5-4.8 Beam subjected to flexural moment and torsion

The modes of failure are explained based on the relative magnitudes of the flexural
moment (Mu) and torsional moment (Tu) at ultimate. Three discrete modes of failure are
defined from a range of failure. The idealised pattern of failure with the plane of failure
and the resultant compression (Cu) and tension (Tu) are shown for each mode
(Courtesy: Pillai, S. U., and Menon, D., Reinforced Concrete Design).
1) Modified bending failure (Mode 1): This occurs when the effect of Mu is larger
than that of Tu.
Zone
under Cu
Tu
Figure 5-4.9 Idealised pattern for Mode 1 failure
2) Lateral bending failure (Mode 2): This is observed in beams with thin webs
when the effect of Mu and Tu are comparable.
Tu
Zone
under Cu
3) Negative bending failure (Mode 3): When the effect of Tu is large and the top
steel is less, this mode of failure occurs.

Tu
Zone
under Cu
In presence of prestressing force, the cracking occurs at higher load. This is evident
from the typical torque versus twist curves for sections under pure torsion.
With further increase in load, the crack pattern remains similar but the inclinations of the
cracks change with the amount of prestressing. The following figure shows the
difference in the torque versus twist curves for a non-prestressed beam and a
prestressed beam.
Ultimate strength
Torque
Cracking torque levels
Twist
Prestressed beam Non-prestressed beam

Figure 5-4.12 Torque versus twist curves
The effect of prestressing force is explained for a beam under pure torsion with a
concentric prestressing force (Pe). The following figure shows such a beam.

Pe
T
Figure 5-4.13 Beam subjected to pure torsion and prestressing force
For a point at the middle of the longer side (Element 1), there is normal stress due to
the prestressing force (fpe). The principal tensile stress (1) is inclined to the neutral
axis at an angle greater than 45.
1 (fpe,)
2 2 > 90
fpe 2
1 1
> 45o
Shear stress Principal
and prestress stresses Mohrs circle
Figure 5-4.14 State of stresses at the side of a prestressed beam
In the following figure, the formation of cracks for a prestressed beam under pure
torsion is shown. This figure can be compared with that for a reinforced concrete beam.
Figure 5-4.15 Formation of cracks in a prestressed beam
In presence of prestressing force, the cracking is at a higher torque. After cracking, the
crack width of a spiral crack is low. Thus, the aggregate interlock is larger as compared
to a non-prestressed beam under the same torque. Hence, the torsional strength of

concrete (Tc) increases in presence of prestressing force. This is accounted for in the
expression of Tc.

5.5 Design for Torsion (Part I)

General Comments
Limit State of Collapse for Torsion
Design of Longitudinal Reinforcement
5.5.1 General Comments
Calculation of Torsion Demand

The restraint to torsion is provided at the ends of a beam. For beams in a building frame,
the restraint is provided by the columns. Precast beams are connected at the ends by
additional elements like angles to generate the torsional restraint. In bridges,
transverse beams at the ends provide torsional restraint to the primary longitudinal
girders. Box girders are provided with diaphragms at the ends.
For equilibrium torsion in a straight beam with distributed torque (tu), the maximum
torsional moment (Tu) is near the restraint at the support. The following figure shows a
schematic representation of the distributed torque.
Tu
tu
L
Figure 5-6.1 Beam subjected to distributed torque
The torsional moment near the support is given by the following expression.
tu L
Tu = (5-5.1)
2
Here,
L = clear span of the beam
tu = distributed torque per unit length.

For a straight beam with a point torque, the maximum torsional moment (Tu) is near the
closer support. If the location of the point torque is variable, Tu is calculated for the
location closest to a support. For a curved beam, Tu is calculated based on structural
analysis.
Design of Torsion Reinforcement

The design is done for the critical section. The critical section is defined in Clause 41.2
of IS:456 - 2000. In general cases, the face of the support is considered as the critical
section. When the reaction at the support introduces compression at the end of the
beam, the critical section can be selected at a distance effective depth from the face of
the support.
To vary the amount of reinforcement along the span, other sections may be selected for
design. Usually the following scheme is selected for the stirrup spacing in beams under
uniformly distributed load.
1) Close spacing for quarter of the span adjacent to the supports.
2) Wide spacing for half of the span at the middle.
For large beams, more variation of spacing may be selected. The following sketch
shows the typical variation of spacing of stirrups. The span is represented by L.
L/4 L/2 L/4
Figure 5-6.2 Typical variation of spacing of stirrups
First, an equivalent flexural moment Mt is calculated from Tu. Second, for the design of
primary longitudinal reinforcement, including the prestressed tendon, the total
equivalent ultimate moment (Me1) is calculated from the flexural moment (Mu) and Mt.
Third, the design of longitudinal reinforcement for other faces based on equivalent
ultimate moments Me2 and Me3 is necessary when the equivalent moment Mt is larger
than Mu. The following sketch shows the equivalent ultimate moments for design.

Me3
Me1 Me2
Figure 5-6.3 Equivalent ultimate moments
The design for Me1 is similar to the design of a prestressed section for flexure.
The design for Me2 is similar to the design of a prestressed concrete or reinforced
concrete section. The design for Me3 is similar to the design of a reinforced concrete
section. The design of stirrups including torsion is similar to the design of stirrups in
absence of torsion.
5.5.2 Limit State of Collapse for Torsion
The design for the limit state of collapse for torsion is based on the Skew Bending
Theory. For a beam subjected to simultaneous flexure and torsion, an equivalent
ultimate bending moment at a section is calculated.
The design for torsion involves the design of longitudinal reinforcement as well as the
transverse reinforcement. The longitudinal reinforcement is designed based on the
equivalent ultimate bending moment.
The transverse reinforcement is designed based on the Skew Bending Theory and a
total shear requirement. For the capacity of concrete, to consider the simultaneous
occurrence of flexural and torsional shears, an interaction between the two is
considered.
The equations in IS:1343 - 1980 are applicable for beams of the following sections.
1) Solid rectangular, with D > b.
2) Hollow rectangular, with D > b and t b/4.
3) Flanged sections like T-beams and I-beams.
The sections are shown in the following sketch.

b b
bw
D
t
Solid rectangular Hollow rectangular Flanged
Figure 5-6.4 Different sections for torsion design
The variables are as follows.

= bw for flanged section
D = total depth of the section
t = thickness of the section.
The average prestress in a section at the level of CGC, is limited to 0.3fck.
5.5.3 Design of Longitudinal Reinforcement
For the design of the longitudinal reinforcement, there are three expressions of the
equivalent ultimate bending moment for the three modes of failure (Reference: Rangan,
B. V. and Hall, A. S., Design of Prestressed Concrete Beams Subjected to Combined
Bending, Shear and Torsion, ACI Journal, American Concrete Institute, March 1975,
Vol. 72, No. 3, pp. 89 93). The modes of failure are explained in Section 5.4, Analysis
for Torsion. The figures of the failure pattern are reproduced here for explanation.
Mode 1 Failure

Zone
under Cu
M e1
Tu
Figure 5-6.5 Idealised pattern and design moment for Mode 1 failure
The equivalent ultimate bending moment for Mode 1 failure (Me1) is given by the
following equation.
M e1 = M u + M t (5-5.2)
The equivalent bending moment for Tu is given as follows.
2D
Mt = Tu 1+ (5-5.3)
b

Mu = applied bending moment at ultimate.
Mt = additional equivalent bending moment for torsion.
Tu = applied torsion at ultimate.
Since, the torsion generates tension in the reinforcement irrespective of the sign, the
sign of Mt is same as that of Mu.
Mode 2 Failure

Tu Me 3
Zone
under Cu
The equivalent ultimate transverse bending moment for Mode 2 failure (Me3) is given as
follows.
2
1+ 2b
D
Me 3 = Mt 1+ 1
x (5-5.4)

2e 1+ 2D
b
In the previous expression

e = Tu/Vu, ratio of ultimate torsion and ultimate shear force at a section.
x1 = smaller dimension of a closed stirrup.
The larger dimension of a closed stirrup is represented as y1. the dimensions are
shown in the following sketch.
y1
x1
Figure 5-6.7 Dimensions of a closed stirrup
The transverse bending moment Me3 is considered when the numerical value of Mu is
less than Mt. Me3 acts about a vertical axis.
Mode 3 Failure

Tu
Me2
Zone
under Cu
The equivalent ultimate bending moment for Mode 3 failure (Me2) is given by the
following equation.
Me2 = Mt Mu (5-5.5)
The expression of Mt is same as for Mode 1 failure, given before.
Mode 3 failure is checked when the numerical value of Mu is less than that of Mt. Me2
acts in the opposite sense of that of Mu.
The longitudinal reinforcement is designed for Me1 similar to the flexural reinforcement
for a prestressed beam. The design of flexural reinforcement is covered in Section 4.2,
Design of Sections for Flexure (Part I) and Section 4.3 Design of Sections for Flexure
(Part II). When Me2 is considered, longitudinal reinforcement is designed similar to a
prestressed concrete or reinforced concrete beam. When Me3 is considered,
longitudinal reinforcement is designed similar to a reinforced concrete beam. For a
singly reinforced rectangular section, the amount of longitudinal reinforcement (As) is
solved from the following equation.
f A
0.87fy As d 1- y s = Mu (5-5.6)
fck bd

d = effective depth of longitudinal reinforcement
fy = characteristic yield stress of longitudinal reinforcement
Mu = one of Me2 and Me3.

5.6 Design for Torsion (Part II)

Design of Transverse Reinforcement
Design Steps
5.6.1 Design of Transverse Reinforcement
For the design of the transverse reinforcement, the capacities of concrete to resist the
torsion and shear need to be determined. To consider the simultaneous occurrence of
flexural and torsional shears, a linear interaction between the two is considered.
The capacity of concrete to resist torsion is reduced from Tc, the capacity under pure
torsion. Similarly, the capacity of concrete to resist shear is reduced from Vc, the
capacity in absence of torsion.
Capacity of Concrete under Pure Torsion

The capacity of concrete is determined based on the plastic theory for torsion. The
capacity is equal to the torque generating the first torsional crack (Tcr).
For a reinforced concrete beam, Tcr is estimated by equating the maximum torsional
shear stress (max) caused by Tcr to the tensile strength of concrete (0.2fck). The
estimated tensile strength is less than that under direct tension because the full section
does not plastify as assumed in the plastic theory.
The estimate of the cracking torque (Tcr) for a rectangular section is given below.
b 2D b
Tcr 0.2 fck 1-
2 3D
b
Tcr = 0.1 b 2D 1- fck
(5-6.1)
3D
For flanged sections, the section is treated as a compound section. A compound section
is a summation of rectangular sections.
The cracking torque is estimated as a summation of the capacities of the individual

rectangular sections. Since the interaction between the rectangular sections is

neglected in the summation, the estimate of the cracking torque is a lower bound
estimate.
The following flanged section is shown as a compound section of five rectangles. For an
individual rectangle, the short side is b and the long side is D.
1 3
4 5
Figure 5-6.1 Flanged section as a compound section
The estimate of the cracking torque (Tcr) for a compound section is as follows.
b
Tcr = 0.1 b 2D 1- fck (5-6.2)
3D
Here, the summation is for the individual rectangles.
For a prestressed concrete beam, the strength of concrete is multiplied by the factor p,
which is a function of the average effective prestress (fcp).
12fcp (5-6.3)
p = 1+
fck
The value of fcp is taken as positive (numeric value). It can be observed that the strength
increases with prestress. The cracking torque (Tcr) and the capacity of concrete to
resist torsion (Tc) for a prestressed concrete beam are thus estimated as follows.
Tc = Tcr
b
Tc = 0.15b 2D 1- p fck
(5-6.4)
3D


b = breadth of the individual rectangle
D = depth of the individual rectangle.
Interaction of Shear and Torsion

In presence of flexural shear, the torsional capacity of concrete reduces. Similarly, in
presence of torsion, the flexural shear capacity of concrete reduces. This is referred to
as interaction of shear and torsion. The capacity of concrete under shear is explained
in Section 5.2, Design for Shear (Part I). A linear interaction of the shear and torsion
capacities of concrete is considered as shown in the following figure. In the horizontal
axis, the shear demand is normalised with respect to the capacity of concrete under
flexural shear. In the vertical axis, the torsional demand is normalised with respect to
the capacity of concrete under pure torsion.
1.0
Tu/Tc
Vu/Vc 1.0
Figure 5-6.2 Interaction diagram for shear and torsion
The interaction equation is given as follows.
Tu Vu
+ =1
Tc Vc (5-6.5)
This is a linear interaction equation.
Tu = applied torsion at ultimate
Vu = applied shear at ultimate
Tc = capacity of concrete under pure torsion.
Vc = capacity of concrete under flexural shear.
Based on the interaction equation, the reduced capacity of concrete to resist torsion
(Tc1) is given below.

e
Tc1 = Tc (5-6.6)
e + ec
Tu / 2
Tc1 is limited to Tu/2 to restrict concrete reaching its capacity.
The parameter e is the ratio of torsion and shear demands at ultimate. The parameter
ec is the ratio of the corresponding concrete capacities.
e = Tu/ Vu (5-6.7)
ec = Tc/ Vc (5-6.8)
The reduced capacity of concrete to resist shear is given below.
ec
Vc1 = Vc (5-6.9)
e + ec
Calculation of Transverse Reinforcement

The transverse reinforcement is provided in the form of closed stirrups enclosing the
corner longitudinal bars. The amount (Asv) is equal to the higher value determined from
two expressions.
The first expression is based on the Skew Bending Theory.
Mt sv
Asv = (5-6.10)
1.5b1d1fy
The notations are as follows.

b1 = distance between the corner longitudinal bars along the short side
d1 = distance between the corner longitudinal bars along the long side.
Mt = additional bending moment from torsion.
sv = spacing of the stirrups
fy = characteristic yield stress of the stirrups.
The dimensions b1 and d1 are shown in the following sketch.

d1
b1
Figure 5-6.3 Dimensions between the corner bars
The second expression of Asv is based on the concept of total shear.
Asv = Av + 2 AT (5-6.11)
The first component Av corresponds to the flexural shear to be carried by the stirrups.
The second component At corresponds to the torsional shear to be carried by the
stirrups. The factor 2 considers that the torsional shear is additive to flexural shear in
both the legs.
The following sketch shows the addition of flexural and torsional shears for a hollow
section.
Due to Vu Due to Tu
Figure 5-6.4 Distribution of flexural and torsional shears for a hollow section
The two shears are additive in the left web, whereas they are subtractive in the right
web. Since, the stirrups have equal areas in the two legs, the torsional shear is
considered additive to flexural shear in both the legs.
In solid sections, the two shears are not additive throughout the web. The flexural shear
is distributed, whereas the torsional shear is restricted in the shear flow zone. Thus for

solid sections, the expression of Asv is conservative. The following sketch shows the
addition of flexural and torsional shears for a solid section.
Due to Vu Due to Tu
Figure 5-6.5 Distribution of flexural and torsional shears for a solid section
If the breadth of the web is large, the two shears can be designed separately. The
stirrups for flexural shear can be distributed throughout the interior of the web. For
torsional shear, closed stirrups can be provided in the peripheral shear flow zone.
The expressions of Av and At are derived from the truss analogy for the ultimate limit
state.
Av =
(Vu -Vc1 ) sv
(5-6.12)
0.87fy d1
AT =
(Tu -Tc1 ) sv
0.87fy b1d1 (5-6.13)
The minimum amount of transverse reinforcement is same as that for shear in absence
of torsion.
Asv 0.4
= (5-6.14)
bsv 0.87fy
5.6.2 Detailing Requirements
The detailing requirements for torsional reinforcement in Clause 22.5.5, IS: 1343 - 1980
are briefly mentioned.
1) There should be at least one longitudinal bar in each corner. The minimum diameter
of the longitudinal bars is 12 mm.

When any side is larger than 450 mm, provide side face reinforcement (As, sf), as per
the following.
Minimum amount As,sf,min = 0.1% bD
Maximum spacing smax = 300 mm or b, whichever is less.
This amount is sufficient to check thermal and shrinkage cracks.
2) The closed stirrups should be bent close to the tension and compression surfaces
satisfying the minimum cover. The stirrups should be perpendicular to the axis of the
beam. Closed stirrups should not be made of pairs of U-stirrups lapping one another.
This is clarified in the following sketch.
Incorrect detailing Correct detailing

Figure 5-6.6 Detailing of closed stirrups
3) The maximum spacing is (x1 + y1)/4 or 200 mm, whichever is smaller. Here x1 and y1
are the short and long dimensions of the stirrups respectively.
4) Proper anchorage of stirrups as mentioned under detailing requirements of shear

reinforcement. It is recommended to bend the ends of a stirrup by 135 and have 10
times the diameter of the bar (db) as extension beyond the bend. The following sketch
clarifies the detailing of end hooks.
135
10db
Figure 5-6.7 Detailing of end hooks for stirrups

5) The stirrups should be continued till a distance h + bw beyond the point at which it is
no longer required. Here, h is the overall depth and bw is the breadth of the web.
5.6.3 Design Steps
The following quantities are known at the selected section.

Mu = factored flexural moment
Vu = factored shear
Tu = factored torsional moment.
For gravity loads, these are calculated from the dead load and live load.
The grades of concrete and steel are selected before design. As per IS: 1343 - 1980,
the grade of steel for stirrups is limited to Fe 415.
For the design of longitudinal reinforcement, the following quantities are unknown.
The member cross-section.
Me1, Me2, Me3 = total equivalent flexural moment
Ap = amount of prestressing steel,
Pe = the effective prestress,
e = the eccentricity
As = area of longitudinal reinforcement
As = area of longitudinal reinforcement in opposite face.
Prestressing steel Ap may be provided in the opposite face.

For the design of stirrups, the following quantities are unknown.
Vc1 = shear carried by concrete
Tc1 = torsion carried by concrete
Asv = total area of the legs of stirrups within a distance sv
sv = spacing of stirrups.
The steps for designing longitudinal and transverse reinforcements for beams subjected
to torsion are given.
1) Calculate Mu, Vu and Tu at a selected location. Select a suitable cross-section.
For high value of Tu, as in bridges, a box section is preferred.

For longitudinal reinforcement

2a) Calculate Me1.
2b) Design Ap and As. The design procedure involves preliminary design and final
design, which are explained in the Section 4.2, Design of Sections for Flexure (Part
I) and Section 4.3, Design of Sections for Flexure (Part II)
3a) Calculate Me2 if Mu < Mt.
3b) Design As. The design procedure is similar for a reinforced concrete section. If
Ap is provided, the design is similar to a prestressed concrete section.
4a) Calculate Me3 if Mu < Mt.
4b) Check the adequacy of transverse bending based on the corner bars. If
inadequate, design side face reinforcement (As,sf). As,sf includes the corner bars.
The design is similar to that for a reinforced concrete section.
For transverse reinforcement

5a) Calculate Tc, Eqn. (5-6.4).
5b) Calculate Vc from the lower of Vc0 and Vcr.
5c) Calculate e (if not calculated earlier) and ec.
5d) Calculate Tc1 and Vc1. Limit Tc1 to Tu/2.
6) Calculate Asv / sv from the greater of the values given by Eqns. (5-6.10),
(5-6.11), (5-6.12), and (5-6.13).
Compare the value with the minimum requirement Eqn. (5-6.14).
7) Calculate maximum spacing and round it off to a multiple of 5 mm.
8) Calculate the size of the stirrups based on the amount required.
Repeat the calculations for other locations of the beam if the spacing of stirrups needs
to be varied.

Example 5-6.1
Design a rectangular section to carry the following ultimate loads.

Tu = 44.5 kNm
Mu = 222.5 kNm (including an estimate of self-weight)
Vu = 89.0 kN.
The material properties are as follows.
fck = 35 N/mm2
fy = 250 N/mm2
fpk = 1720 N/mm2
The prestressing is fpe = 1035 N/mm2.
Solution
1) Calculate Me1.
Let D/b = 2
2D
Mt = Tu 1+
b
= 44.5 1+ 2 2
= 99.5 kNm
Me1 = Mu + MT
= 222.5 + 99.5
= 322.0 kNm
2) Select section. Design Ap and As.

Select
b = 250 mm
D = 500 mm
d = 450 mm.

Provide (2) 16 mm diameter corner bars. The flexural design results are as follows.
As = 2 201
= 402 mm2.
Required amount of prestressing steel with dp = d = 450 mm is Ap = 484 mm2.
Provide 11 mm diameter strands with area = 70 mm2.
Required number of strands = 484 / 70 = 6.8 7
Provided amount of prestressing steel
Ap,prov = 7 70
= 490 mm2
3) Calculate Me2 .
Since Mu > Mt , design for Me2 is not required.
4) Calculate Me3 .
Since Mu > Mt , design for Me3 is not required.
5a) Calculate Tc .
Pe
fcp =
A
fpe Ap
=
bD
1035 490
=
250500
= 4.06 N/mm2
12fcp
p = 1+
fck
12 4.06
= 1+
35
= 1.55

fcp < 0.3 fck . OK
b
Tc = 0.15b 2D 1- p fck
3D
1
= 0.15 2502 500 1- 1.55 35 Nmm
3 2
= 35.8 kNm
5b) Calculate Vc from the lower of Vco and Vcr .
100 Ap 100 490

=
bd 250 450
= 0.43
From Table 6, for M 35 concrete, c = 0.46 N/mm2.
Pe Pee 2
fpt = - -
A I
507,150 507,150 2002
=- -
125,000 2.604109
= -11.85 N/mm2
Here,
e = 450 500
= 200 mm
I = 250 5003 / 12
= 2.604 109 mm4.
I
M0 = 0.8fpt
y
2.604109
= 0.811.85
200
= 123.43 kNm
f V
Vcr = 1- 0.55 pe c bd + M0 u
fpk Mu
0.46 250 450 89
= (1- 0.550.6) 3
+123.43
10 222.5
= 84.0 kN
Here,
fpe/fpk = 1035 / 1720
= 0.6.

= 0.67 250500 1.422 + 0.8 4.061.42

= 215.6 kN
Here,
ft = 0.24 35
= 1.42 N/mm2
Vc = Vcr = 84.0 kN
5c) Calculate e and ec .
Tu Tc
e= ec =
Vu Vc
44.5 35.8
= =
89.0 84.0
= 0.50 m = 0.43 m
5d) Calculate Tc1 and Vc1.
e
Tc1 = Tc
e + ec
0.50
= 35.82
0.50 + 0.43
= 19.26 kNm
ec
Vc1 = Vc
e + ec
0.43
= 84.0
0.50 + 0.43
= 38.84 kN
Tu
Tc1 < OK.
2
6) Calculate Asv / sv
Asv Mt
=
sv 1.5b1d1fy
99.5 106
=
1.5 200 400 250
= 3.3 mm2 /mm

Estimated values
b1 = 250 50
= 200 mm
d1 = 500 100
= 400 mm.
Av Vu -Vc1
=
sv 0.87fy d1
(89.0 - 38.8)103
=
0.87 250 400
= 0.58 mm2 /mm
AT T -T
= u c1
sv 0.87fy b1d1
(44.5 - 19.26)106
=
0.87 250 200 400
= 1.45 mm2 /mm
Asv AV A
= + 2 T
sv sv sv
= 0.58 + 21.45
= 3.48 mm2 /mm
Minimum amount of stirrups
Asv 0.4
=
bsv 0.87fy
Asv 0.4 250

=
sv 0.87 250
= 0.46 mm2 /mm
Select
Asv / sv = 3.48 mm2/mm.

7) Calculate maximum spacing

x1 + y1
sv
4
204 + 422

4
156 mm
Estimated values
x1 = 250 46
= 204 mm
y1 = 500 78
= 422 mm.
The other values of sv do not govern.
8) Calculate the size of the stirrups
Select 2 legs of 12 mm diameter stirrups.

Asv = 2 113
= 226 mm2
226
sv =
3.48
= 65 mm
The spacing can be increased by bundling the stirrup bars.
Designed section
250
12 mm diameter stirrups
@ 65 mm c/c
500 CGC
(4) corner bars
200 16 mm diameter
(7) 11 mm diameter strands

Dimensions in mm. with Pe = 507.15 kN
As D > 450 mm, side face reinforcement is required.

6.1 Calculation of Deflection

Introduction
Deflection due to Gravity Loads
Deflection due to Prestressing Force
Total Deflection
Limits of Deflection
Determination of Moment of Inertia
Limits of Span-to-effective Depth Ratio
6.1.1 Introduction
The deflection of a flexural member is calculated to satisfy a limit state of serviceability.

Since a prestressed concrete member is smaller in depth than an equivalent reinforced
concrete member, the deflection of a prestressed concrete member tends to be larger.
The total deflection is a resultant of the upward deflection due to prestressing force and
downward deflection due to the gravity loads. Only the flexural deformation is
considered and any shear deformation is neglected in the calculation of deflection.
Shear deformation is included in members such as deep beams and wall type of
structures.
The deflection of a member is calculated at least for two cases.

1) Short term deflection at transfer
2) Long term deflection under service loads
The short term deflection at transfer is due to the prestressing force (before long term
losses) and self-weight. The effect of creep and shrinkage of concrete are not
considered. The long term deflection under service loads is due to the effective
prestressing force (after long term losses) and the gravity loads. The permanent
components of the gravity loads are considered in the effect of creep. These
components are dead load and sustained live load.

6.1.2 Deflection due to Gravity Loads
The methods of calculation of deflection are taught in a course on structural analysis. It

is expected that the students are familiar with the methods. The methods include the
following.
1) Double integration method
2) Moment-area method
3) Conjugate beam method
4) Principle of virtual work
Numerical solution schemes can be developed based on the above methods and
executed in a computer. For members with prismatic cross-sections, common support
conditions and subjected to conventional loading, the deflections are available in tables
of text books on structural analysis.
The expressions of deflection () for a few cases are provided. Here,

I = moment of inertia
E = modulus of elasticity of concrete
Simply supported beams

w 5 wL4
=
384 EI
P at centre
PL3
=
48EI

Cantilever beams
w wL4
=
8EI
P
PL3
=
3EI
6.1.3 Deflection due to Prestressing Force
The prestressing force causes a deflection only if the CGS is eccentric to the CGC. The
deflection due to prestressing force is calculated by the load-balancing method. This
method is explained in Section 3.2, Analysis of Member under Flexure (Part I). The
upward thrust (represented as wup for curved tendons and Wup for bent tendons) and
the upward deflection (also called camber and represented as P) due to the
prestressing forces in typical profiles of tendons are reproduced here.
a) For a Parabolic Tendon

P e P
wup
L
8Pe
wup =
L2
4
5 w up L
P = (6-1.1)
384 EI

b) For a Singly Harped Tendon

P P
Wup
4Pe
Wup =
L
Wup L3
P = (6-1.2)
48EI
b) For a Doubly Harped Tendon

P P
aL
Wup Wup
Pe
Wup =
aL
a ( 3 - 4a 2 ) Wup L3 (6-1.3)
P =
24EI
6.1.4 Total Deflection
The total deflection is calculated for the following two cases.

1) Short term deflection at transfer
2) Long term deflection under service loads
The short term deflection at transfer (st) is given as follows.

st = -P 0 + SW (6-1.4)
Here,
P0 = magnitude of deflection due to P0
sw = deflection due to self-weight
P0 = prestressing force before long term losses.
The long term deflection under service loads is difficult to calculate because the
prestressing force and creep strain influence each other. Creep of concrete is defined
as the increase in deformation with time under constant load. Due to the creep of

concrete, the prestress in the tendon is reduced with time. The ultimate creep strain is
found to be proportional to the elastic strain. The ratio of the ultimate creep strain to the
elastic strain is called the creep coefficient . The values of as per IS:1343 - 1980 are
given in Section 1.6, Concrete (Part II).
The following expression of the long term deflection under service loads (lt) is a
simplified form, where an average prestressing force is considered to generate the
creep strain. The effect of shrinkage on the prestressing force is neglected in the
expression.
+ Pe
lt = -Pe - P 0 + ( DL + SL ) (1+ ) + LL (6-1.5)
2
The notations in the previous equations are as follows.

P0 = magnitude of deflection due to P0
Pe = magnitude of deflection due to Pe
Pe = effective prestressing force after long term losses.
DL = deflection due to dead load (including self-weight)
SL = deflection due to sustained live load
LL = deflection due to additional live load
A more rigorous calculation of total deflection can be done using the incremental time-
step method. It is a step-by-step procedure, where the change in prestressing force
due to creep and shrinkage strains is calculated at the end of each time step. The
results at the end of each time step are used for the next time step. This procedure was
suggested by the Precast / Prestressed Concrete Institute (PCI) committee and is also
called the General method (Reference: PCI Committee, Recommendations for
Estimating Prestress Losses, PCI Journal, PCI, Vol. 20, No. 4, July-August 1975, pp.
43-75).
In the PCI step-by-step procedure, a minimum of four time steps are considered in the
service life of a prestressed member. The following table provides the definitions of the
time steps.

Table 6-1.1 Time steps in the step-by-step procedure

Step Beginning End
Pre-tension: Anchorage of steel

1. Age of prestressing
Post-tension: End of curing
30 days after prestressing or when

2. End of Step 1
subjected to superimposed load
3. End of Step 2 1 year of service
4. End of Step 3 End of service life
The step-by-step procedure can be implemented in a computer program, where the

number of time steps can be increased.
6.1.5 Limits of Deflection
Clause 19.3.1 of IS:1343 - 1980 specifies limits of deflection such that the efficiency of
the structural element and the appearance of the finishes or partitions are not adversely
affected. The limits of deflection are summarised next.
1) The total deflection due to all loads, including the effects of temperature, creep
and shrinkage, should not exceed span / 250.
2) The deflection after erection of partitions or application of finishes, including the
effects of temperature, creep and shrinkage, should not exceed span/350 or 20
mm, whichever is less.
3) If finishes are applied, total upward deflection due to prestressing force should
not exceed span / 300.
6.1.6 Determination of Moment of Inertia
Type 1 and Type 2 Members

These types of members are designed to be uncracked under service loads. The gross
moment of inertia (Ig) can be used to calculate the deflections.
Type 3 Members

This type of members is expected to be cracked under service loads. Strictly, the gross
moment of inertia (Ig) cannot be used in the calculations. IS:1343 - 1980, Clause 22.6.2,
recommends the following.
1) When the permanent load is less than or equal to 25% of the live load, the gross
moment of inertia can be used.
2) When the permanent load is greater than 25% of the live load, the span-to-
effective depth (L/d) ratio should be limited to bypass the calculation of deflection.
If the L/d ratio of a member exceeds the limit, the gross moment of inertia can still be
used if the tensile stress under service loads is within the allowable value. This
recommendation is suggested because the calculation of gross moment of inertia is
simpler as compared to an effective moment of inertia.
6.1.7 Limits of Span-to-Effective Depth Ratio
The calculation of deflection can be bypassed if the span-to-effective depth (L/d) ratio is
within the specified limit.
The limits of L/d ratios as per Clause 22.6.2, IS:1343 1980, are as follows.
For L 10 m
For cantilever beams L /d 7
For simply supported beams L /d 20
For continuous beams L /d 26
For L > 10 m
For simply supported beams L /d (20 10/ L)
For continuous beams L /d (26 10/ L)
Here, L is in metres. Deflection calculations are necessary for cantilevers with L > 10 m.

6.2 Calculation of Crack Width

Introduction
Method of Calculation
Limits of Crack Width
6.2.1 Introduction
The crack width of a flexural member is calculated to satisfy a limit state of serviceability.
Among prestressed concrete members, there is cracking under service loads only for
Type 3 members. Hence the calculation of crack width is relevant only for Type 3
members. The crack width is calculated for the cracks due to bending which occur at
the bottom or top surfaces of a flexural member.
The flexural cracks start from the tension face and propagate perpendicular to the axis
of the member. This type of cracks is illustrated in Section 5.1, Analysis for Shear. If
these cracks are wide, it leads to corrosion of the reinforcing bars and prestressed
tendons. Also, the cracks tend to widen under sustained load or cyclic load. To limit
the crack width, Type 3 members have regular reinforcing bars in the tension zone
close to the surface, in addition to the prestressed tendons.
The crack width of a flexural crack depends on the following quantities.

1) Amount of prestress
2) Tensile stress in the longitudinal bars
3) Thickness of the concrete cover
4) Diameter and spacing of longitudinal bars
5) Depth of member and location of neutral axis
6) Bond strength
7) Tensile strength of concrete.

6.2.2 Method of Calculation
IS:456 - 2000, Annex F, gives a procedure to determine flexural crack width. The
design crack width (Wcr) at a selected level on the surface of the section with maximum
moment is given as follows.
3acr m
Wcr = (6-2.1)
2 ( acr - Cmin )
1+
h- x
The notations in the previous equation are as follows.

acr = shortest distance from the selected level on the surface to a
longitudinal bar
Cmin = minimum clear cover to the longitudinal bar
h = total depth of the member
x = depth of the neutral axis
m = average strain at the selected level.
The values of Cmin and h are obtained from the section of the member. The evaluation
of the other variables is explained.
Evaluation of acr
The location of crack width calculation can be at the soffit or the sides of a beam. The
value of acr depends on the selected level. The following sketch shows the values of acr
at a bottom corner (A), at a point in the soffit (B) and at a point at the side (C).
Neutral axis
C acr3
acr1 acr2
A
B
Figure 6-2.1 Cross-section of a beam showing the distances from surface to the
nearest longitudinal bar

Usually the crack width is calculated at a point in the soffit, which is equidistant from two
longitudinal bars. This point is the location of maximum estimated crack width. The
following sketch shows the variables used in computing acr.
s
db
Cmin acr dc
Figure 6-2.2 Cross-section of a beam showing variables for calculation
Using geometry, the value of acr is obtained from the following equation.
2
s d
acr = + dc 2 - b (6-2.2)
2 2
Here,
db = diameter of longitudinal bar
dc = effective cover = Cmin + db/2
s = centre-to-centre spacing of longitudinal bars.
The values of db, dc and s are obtained from the section of the member.
Evaluation of x and m
The values of x and m are calculated based on a sectional analysis under service loads.
The sectional analysis should consider the tension carried by the uncracked concrete in
between two cracks. The stiffening of a member due to the tension carried by the
concrete is called the tension stiffening effect. The value of m is considered to be an
average value of the strain at the selected level over the span. The following sketch
illustrates the cracking and the uncracked concrete in a flexural member.

Uncracked
Cracked section
concrete
Figure 6-2.3 Elevation of a beam showing the cracking and uncracked concrete
The analysis of a Type 3 member should be based on strain compatibility of concrete

and prestressing steel. IS:456 - 2000 recommends two procedures for the sectional
analysis considering the tension stiffening effect.
1) Rigorous procedure with explicit calculation of tension carried by the concrete.
2) Simplified procedure based on the conventional analysis of a cracked section,
neglecting the tension carried by concrete. An approximate estimate of the
tension carried by the concrete is subsequently introduced.
Here, the simplified procedure is explained. For a rectangular zone under tension, the
simplified procedure gives the following expression of m.
b ( h - x )( a - x )
m = 1 - (6-2.3)
3Es As ( d - x )
For a prestressed member, (EpAp + EsAs) is substituted in place of EsAs.

The second term considers the tension carried by the concrete approximately by
reducing the strain (1) obtained from the analysis of a cracked section.

a = distance from the compression face to the level at which crack
width is calculated
= h, when the crack width is calculated at the soffit
b = width of the rectangular zone
d = effective depth of the longitudinal reinforcement
Ap = area of prestressing steel.
Es = modulus of elasticity of non-prestressed steel
Ep = modulus of elasticity of prestressed steel

1 = strain at the selected level based on a cracked sectional analysis

= s(a x)/(d x)
s = strain in the longitudinal reinforcement.
The depth of neutral axis (x) can be calculated by a trial and error procedure till the
equilibrium equations are satisfied. The following sketch shows the beam cross section,
strain profile, stress diagram and force couples under service loads. The contribution of
non-prestressed reinforcement is also included.
b c fc
0.33x
dp x C
d
Ap
fp Tp
As s
dec fs Ts
p

Figure 6-2.4 Sketches for analysis of a rectangular section

C = 0.5Eccxb (6-2.4)
Tp = ApEpp (6-2.5)
Ts = AsEss (6-2.6)
Based on the principles of mechanics, the following equations are derived.
F = 0
Tp +Ts = C
Ap E p p + As Es s = 0.5Ec c xb (6-2.7)
The second equation relates the moment under service loads (M) with the internal
M A = Ts ( d - d p ) + C ( d p - 0.33 x )
p
= As Es s ( d - d p ) + 0.5Ec c xb ( d p - 0.33 x )

(6-2.8)
The value of M should be equal to the moment due to service loads.
2) Equations of compatibility
The depth of the neutral axis is related to the depth of CGS and the depth of non-
prestressed reinforcement by the similarity of the triangles in the strain diagram.
x c
= (6-2.9)
d p c + p - dec
x c
=
d c + s (6-2.10)
Linear elastic constitutive relationships are used in the earlier expressions of C, Ts and
Tp.
The known variables in the analysis are: b, d, Ap, As, dec, Ec, Ep, Es, M.
The unknown quantities are: x, c, p, s.
The steps for solving the above equations are given below.
1) Assume c
2) Assume x.
3) Calculate p and s from Eqn. (6-2.9) and Eqn. (6-2.10), respectively.
4) Calculate C, Tp and Ts from Eqns. (6-2.4), (6-2.5), (6-2.6), respectively.
5) If Eqn. (6-2.7) is not satisfied, change x. If Tp + Ts < C, decrease x. If Tp
+ Ts > C, increase x.
6) Calculate M from Eqn. (6-2.8). If the value differs from the given value, change
c and repeat from Step 2.
6.2.3 Limits of Crack Width
Clause 19.3.2 of IS:1343 - 1980 specifies limits of crack width such that the
appearance and durability of the structural element are not affected.

The limits of crack width are as follows.

Crack width 0.2 mm for moderate and mild environments
0.1 mm for severe environment.
The types of environments are explained in Table 9, Appendix A of IS:1343 - 1980.

7.1 Transmission of Prestress (Part I)

Pre-tensioned Members
7.1.1 Pre-tensioned Members
The stretched tendons transfer the prestress to the concrete leading to a self
equilibrating system. The mechanism of the transfer of prestress is different in the pre-
tensioned and post-tensioned members. The transfer or transmission of prestress is
explained for the two types of members separately.
For a pre-tensioned member, usually there is no anchorage device at the ends. The
following photo shows that there is no anchorage device at the ends of the pre-
tensioned railway sleepers.
Figure 7-1.1 End of pre-tensioned railway sleepers

For a pre-tensioned member without any anchorage at the ends, the prestress is
transferred by the bond between the concrete and the tendons. There are three
mechanisms in the bond.
1) Adhesion between concrete and steel
2) Mechanical bond at the concrete and steel interface

3) Friction in presence of transverse compression.

The mechanical bond is the primary mechanism in the bond for indented wires, twisted
strands and deformed bars. The surface deformation enhances the bond. Each of the
type is illustrated below.
Pitch
Elliptical indentations
Circular indentations
Examples of indented wires
Twisted strand Deformed bar

Figure 7-1.2 Indented wires, twisted strands and deformed bars
The prestress is transferred over a certain length from each end of a member which is
called the transmission length or transfer length (Lt). The stress in the tendon is zero
at the ends of the members. It increases over the transmission length to the effective
prestress (fpe) under service loads and remains practically constant beyond it. The
following figure shows the variation of prestress in the tendon.

fpe
Lt
Figure 7-1.3 Variation of prestress in tendon along transmission length
Hoyer Effect
After stretching the tendon, the diameter reduces from the original value due to the
Poissons effect. When the prestress is transferred after the hardening of concrete, the
ends of the tendon sink in concrete. The prestress at the ends of the tendon is zero.
The diameter of the tendon regains its original value towards the end over the
transmission length. The change of diameter from the original value (at the end) to the
reduced value (after the transmission length), creates a wedge effect in concrete. This
helps in the transfer of prestress from the tendon to the concrete. This is known as the
Hoyer effect. The following figure shows the sequence of the development of Hoyer
effect.

Diameter
Original after
diameter stretching
a) Applying tension to tendon
Diameter
Original after
diameter stretching
b) Casting of concrete
Sinking of tendon
fp0 Diameter
Original after
diameter stretching
c) Transferring of prestress
Figure 7-1.4 Hoyer effect
Since there is no anchorage device, the tendon is free of stress at the end. The
concrete should be of good quality and adequate compaction for proper transfer of
prestress over the transmission length.
Transmission Length
There are several factors that influence the transmission length. These are as follows.
1) Type of tendon
wire, strand or bar
2) Size of tendon
3) Stress in tendon
4) Surface deformations of the tendon

Plain, indented, twisted or deformed

5) Strength of concrete at transfer
6) Pace of cutting of tendons
Abrupt flame cutting or slow release of jack
7) Presence of confining reinforcement
8) Effect of creep
9) Compaction of concrete
10) Amount of concrete cover.
The transmission length needs to be calculated to check the adequacy of prestress in

the tendon over the length. A section with high moment should be outside the
transmission length, so that the tendon attains at least the design effective prestress
(fpe) at the section. The shear capacity at the transmission length region has to be
based on a reduced effective prestress.
IS:1343 - 1980 recommends values of transmission length in absence of test data.

These values are applicable when the concrete is well compacted, its strength is not
less than 35 N/mm2 at transfer and the tendons are released gradually. The
recommended values of transmission length are as follows.
Table 7-1.1 Values of transmission length

For plain and intended wires Lt = 100
For crimped wire Lt = 65
For strands Lt = 30
Here, is the nominal diameter of the wire or strand.

To avoid the transmission length in the clear span of a beam, IS:1343 - 1980
recommends the following.
1) To have an overhang of a simply supported member beyond the support by a
distance of at least Lt.

Lt
Figure 7-1.5 End of a simply supported member
2) If the ends have fixity, then the length of fixity should be at least Lt.
Lt
Figure 7-1.6 End of a member with fixity
Development Length
The development length needs to be provided at the critical section, the location of
maximum moment. The length is required to develop the ultimate flexural strength of the
member. The development length is the minimum length over which the stress in
tendon can increase from zero to the ultimate prestress (fpu). The development length is
significant to achieve ultimate capacity.
If the bonding of one or more strands does not extend to the end of the member (de-
bonded strand), the sections for checking development of ultimate strength may not be
limited to the location of maximum moment.
The development length (Ld) is the sum of the transmission length (Lt) and the bond
length (Lb).
Ld = Lt + Lb (7-1.1)

The bond length is the minimum length over which, the stress in the tendon can
increase from the effective prestress (fpe) to the ultimate prestress (fpu) at the critical
location.
The following figure shows the variation of prestress in the tendon over the length of a
simply supported beam at ultimate capacity.
fpu
fpe
Lt
Figure 7-1.7 Variation of prestress in tendon at ultimate
The calculation of the bond length is based on an average design bond stress (bd). A
linear variation of the prestress in the tendon along the bond length is assumed. The
following sketch shows a free body diagram of a tendon along the bond length.
bd
fpeAp fpuAp
fpu
fpe
Lb
Figure 7-1.8 Assumed variation of prestress in tendon along the bond length
The bond length depends on the following factors.

1) Surface condition of the tendon
2) Size of tendon
3) Stress in tendon

4) Depth of concrete below tendon
From equilibrium of the forces in the above figure, the expression of the bond length is
derived.
Lb =
( fpu - fpe )
(7-1.2)
4 bd
Here, is the nominal diameter of the tendon.

The value of the design bond stress (bd) can be obtained from IS:456 - 2000, Clause
26.2.1.1. The table is reproduced below.
Table 7-1.2 Design bond stress for plain bars

Grade of concrete M30 M35 M40 and above
bd (N/mm2) 1.5 1.7 1.9
End Zone Reinforcement

The prestress and the Hoyer effect cause transverse tensile stress (t). This is largest
during the transfer of prestress. The following sketch shows the theoretical variation of
t.
Lt
Figure 7-1.9 Transverse stress in the end zone of a pre-tensioned beam
To restrict the splitting of concrete, transverse reinforcement (in addition to the

reinforcement for shear) needs to be provided at each end of a member along the

transmission length. This reinforcement is known as end zone reinforcement.
The generation of the transverse tensile stress can be explained by the free body
diagram of the following zone below crack, for a beam with an eccentric tendon.
Tension (T), compression (C) and shear (V) are generated due to the moment acting on
the horizontal plane at the level of the crack. The internal forces along the horizontal
plane are shown in (a) of the following figure. The variation of moment (due to the
couple of the normal forces) at horizontal plane along the depth is shown in (b).
Lt
T C
Lt
Variation of moment
Free body diagram of at horizontal plane
zone below crack along depth
Figure 7-1.10 Forces in the end zone
The end zone reinforcement is provided to carry the tension (T) which is generated due
to the moment (M). The value of M is calculated for the horizontal plane at the level of
CGC due to the compressive stress block from the normal stresses in a vertical plane
above CGC. The minimum amount of end zone reinforcement (Ast) is given in terms of
the moment (M) as follows.
2.5M
Ast = (7-1.3)
fs h


h = total depth of the section
M = moment at the horizontal plane at the level of CGC due to the
compressive stress block above CGC
fs = allowable stress in end zone reinforcement.
The lever arm for the internal moment is h/2.5. The value of fs is selected based on a
maximum strain.
The end zone reinforcement should be provided in the form of closed stirrups enclosing
all the tendons, to confine the concrete. The first stirrup should be placed as close as
possible to the end face, satisfying the cover requirements. About half the
reinforcement can be provided within a length equal to Lt from the end. The rest of the
reinforcement can be distributed in the remaining Lt.
References:
1) Krishnamurthy, D. A Method of Determining the Tensile Stresses in the End Zones

of Pre-tensioned Beams, Indian Concrete Journal, Vol. 45, No. 7, July 1971, pp. 286-
297.
2) Krishnamurthy, D. Design of End Zone Reinforcement to Control Horizontal

Cracking in Pre-tensioned Concrete Members at Transfer, Indian Concrete Journal, Vol.
47, No. 9, September 1973, pp. 346-349.

Example 7-1.1
Design the end zone reinforcement for the pre-tensioned beam shown in the
following figure.
The sectional properties of the beam are as follows.

A = 46,400 mm2
I = 8.47 108 mm4
Z = 4.23 105 mm3
There are 8 prestressing wires of 5 mm diameter.

Ap = 8 19.6 = 157 mm2
The initial prestressing is as follows.
fp0 = 1280 N/mm2.
Limit the stress in end zone reinforcement (fs) to 140 N/mm2.
200
60
80
400 CGC
90
CGS
Cross-section at end
Solution
1) Determination of stress block above CGC
Initial prestressing force

P0 = Ap.fpo
= 157 1280 N
= 201 kN

Stress in concrete at top
P0 P0e
ft = - +
A Z
201103 201103 90
=- +
46400 4.23 105
0 N/mm2
Stress at bottom
P0 P0e
fb = - -
A Z
201103 201103 90
=- -
46400 4.23105
= -8.60 N/mm2
200
60 C1
1.29 C2
80 140 C3
CGC 4.30
y3 y2 y1
CGS
8.60
Stress profile Components of

compression
block
2) Determination of components of compression block
C1 = 1.29 200 60
= 7.74 kN
y1 = 140 + 60
= 160 mm
C2 = 1.29 140 80
= 7.22 kN
y2 = 140
= 93.3 mm
C3 = 4.3 140 80
= 24.08 kN
y3 = 140
= 46.7 mm

200
60
1.29
80 140
4.30
8.60
3) Determination of moment
M = Ci.yi
= C1.y1 + C2.y2 + C3.y3
= (7.74 160) + (7.22 93.3) + (24.08 46.7)
= 3036.6 kN-mm
4) Determination of amount of end zone reinforcement
2.5M
Ast =
fs h
2.5M
=
140 400
2.5 3036.6103
=
140 400
= 135.6 mm2
With 6 mm diameter bars, required number of 2 legged closed stirrups

= 135.6 / (2 28.3) 3.
For plain wires, transmission length

Lt = 100
= 500 mm.
Provide 2 stirrups within distance 250 mm (Lt/2) from the end. The third stirrup is in the
next 250 mm.

Designed end zone reinforcement
(3) 6 mm diameter stirrups

7.2 Transmission of Prestress (Part II)

Post-tensioned Members
7.2.1 Post-tensioned Members
Unlike in a pre-tensioned member without anchorage, the stress in the tendon of a post-
tensioned member attains the prestress at the anchorage block. There is no
requirement of transmission length or development length.
The end zone (or end block) of a post-tensioned member is a flared region which is
subjected to high stress from the bearing plate next to the anchorage block. It needs
special design of transverse reinforcement. The design considerations are bursting
force and bearing stress.
The stress field in the end zone of a post-tensioned member is complicated. The
compressive stress trajectories are not parallel at the ends. The trajectories diverge
from the anchorage block till they become parallel. Based on Saint Venants principle,
it is assumed that the trajectories become parallel after a length equal to the larger
transverse dimension of the end zone. The following figure shows the external forces
and the trajectories of tensile and compressive stresses in the end zone.
Tensile stress Compressive stress

trajectories trajectories
t
yp0 y0 = larger
transverse
dimension of
end zone
y0
Bearing plate
Stress trajectories in the end zone

Figure 7-2.1 Stress trajectories in the end zone of a post-tensioned beam
The larger transverse dimension of the end zone is represented as y0. The
corresponding dimension of the bearing plate is represented as yp0. For analysis, the

end zone is divided into a local zone and a general zone as shown in the following
sketch.
Local zone General zone
Bearing plate y0
Figure 7-2.2 Local and general zones in the end zone
The local zone is the region behind the bearing plate and is subjected to high bearing
stress and internal stresses. The behaviour of the local zone is influenced by the
anchorage device and the additional confining spiral reinforcement. The general zone
is the end zone region which is subjected to spalling of concrete. The zone is
strengthened by end zone reinforcement.
The variation of the transverse stress (t) at the CGC along the length of the end zone
is shown in the next figure. The stress is compressive for a distance 0.1y0 from the end.
Beyond that it is tensile. The tensile stress increases and then drops down to zero
within a distance y0 from the end.
0.1y0 0.9y0
Fbst
t
Distance along
axis of beam
Figure 7-2.3 Transverse stress in the end zone
The transverse tensile stress is known as splitting tensile stress. The resultant of the
tensile stress in a transverse direction is known as the bursting force (Fbst). Compared

to pre-tensioned members, the transverse tensile stress in post-tensioned members is

much higher.
Besides the bursting force there is spalling forces in the general zone.
Spalling force Bursting force

Figure 7-2.4 Spalling and bursting forces in the end zone
IS:1343 - 1980, Clause 18.6.2.2, provides an expression of the bursting force (Fbst) for
an individual square end zone loaded by a symmetrically placed square bearing plate.
y
Fbst = Pk 0.32 - 0.3 p 0 (7-2.1)
y0
Here,
Pk = prestress in the tendon
yp0 = length of a side of bearing plate
y0 = transverse dimension of the end zone.
The following sketch shows the variation of the bursting force with the parameter yp0 / y0.
The parameter represents the fraction of the transverse dimension covered by the
bearing plate.
0.32
Fbst
Pk
0.02
y p0
1 y0
Figure 7-2.5 Variation of bursting force with size of bearing plate

It can be observed that with the increase in size of the bearing plate the bursting force
(Fbst) reduces. The following sketch explains the relative size of the bearing plate with
respect to the end zone.
(1) (2) (3)
Figure 7-2.6 End views of end zones with varying size of the bearing plate
In the above end views of end zones, the bursting force (Fbst) will be largest for Case (1)
and least for Case (3). For a rectangular end zone, Fbst is calculated from the previous
equation for each principle direction. For a circular bearing plate, an equivalent square
loaded area is considered in the calculation of Fbst. For more than one bearing plate, the
end zone is divided into symmetrically loaded prisms. Each prism is analysed by the
previous equation.
End Zone Reinforcement

Transverse reinforcement is provided in each principle direction based on the value of
Fbst. This reinforcement is called end zone reinforcement or anchorage zone
reinforcement or bursting links. The reinforcement is distributed within a length from
0.1y0 to y0 from an end of the member.
The amount of end zone reinforcement in each direction (Ast) can be calculated from the
following equation.
Fbst
Ast = (7-2.2)
fs
The stress in the transverse reinforcement (fs) is limited to 0.87fy. When the cover is
less than 50 mm, fs is limited to a value corresponding to a strain of 0.001.
The end zone reinforcement is provided in several forms, some of which are proprietary

of the construction firms. The forms are closed stirrups, mats or links with loops. A few
types of end zone reinforcement is shown in the following sketches.
Mat Links
Figure 7-2.7 Types of end zone reinforcement
The local zone is further strengthened by confining the concrete with spiral
reinforcement. The performance of the reinforcement is determined by testing end block
specimens. The following photo shows the spiral reinforcement around the guide of the
tendons.
Figure 7-2.8 Spiral reinforcement in the end zone

The end zone may be made of high strength concrete. The use of dispersed steel fibres
in the concrete (fibre reinforced concrete) reduces the cracking due to the bursting force.
Proper compaction of concrete is required at the end zone. Any honey-comb of the
concrete leads to settlement of the anchorage device. If the concrete in the end zone is
different from the rest of the member, then the end zone is cast separately.

Bearing Plate
High bearing stress is generated in the local zone behind the bearing plate. The
bearing stress (fbr) is calculated as follows.
Pk
fbr = (7-2.3)
Apun
Here,
Pk = prestress in the tendon with one bearing plate
Apun = Punching area
= Area of contact of bearing plate.
As per Clause 18.6.2.1, IS:1343 - 1980, the bearing stress in the local zone should be
limited to the following allowable bearing stress (fbr,all).
Abr
fbr,all = 0.48fci (7-2.4)
Apun
0.8fci
In the above equation,

Apun = Punching area
= Area of contact of bearing plate
Abr = Bearing area
= Maximum transverse area of end block that is geometrically similar
and concentric with punching area
fci = cube strength at transfer.
The expression of allowable bearing stress takes advantage of the dispersion of the
bearing stress in the concrete. The following sketch illustrates the dispersion of bearing
stress in concrete.

Apun
Abr
Apun Abr
End view
showing
bearing plate
Figure 7-2.9 End and isometric views of end zone
The performance of anchorage blocks and end zone reinforcement is critical during the
post-tensioning operation. The performance can be evaluated by testing end block
specimens under compression. The strength of an end block specimen should exceed
the design strength of the prestressing tendons.
The following photos show the manufacturing of an end block specimen.
(a) Fabrication of end zone reinforcement

(b) Anchorage block and guide
(c) End zone reinforcement with guide and duct

(d) End block after casting

Figure 7-2.10 Manufacturing of an end block specimen
Example 7-2.1
Design the bearing plate and the end zone reinforcement for the following bonded
post-tensioned beam.
The strength of concrete at transfer is 50 N/mm2.
A prestressing force of 1055 kN is applied by a single tendon. There is no

eccentricity of the tendon at the ends.
400 400
100
400 600
100
100
Section beyond end zone Section at end zone

Solution
1) Let the bearing plate be 200 mm 300 mm. The bearing stress is calculated below.
Pk
fbr =
Apun
1055103
=
200300
= 17.5 N/mm2
The allowable bearing stress is calculated.
Abr
fbr,all = 0.48fci
Apun
400 600
= 0.48 50
200 300
= 48 N/mm2
Limit fbr,all to 0.8 fci = 0.8 50 = 40 N/mm2. Bearing stress is less than fbr,all. Hence OK.
2) Calculate bursting force.
In the vertical direction
y
Fbst = Pk 0.32 - 0.3 p 0
y0
300
= 1055 0.32 - 0.3
600
= 179.3 kN
In the horizontal direction
y
Fbst = Pk 0.32 - 0.3 p 0
y0
200
= 1055 0.32 - 0.3
400
= 179.3 kN

3) Calculate end zone reinforcement.
Fbst
Ast =
0.87fy
179.3103
=
0.87 250
= 824.6 mm2
Provide Ast = 824.6 = 550 mm2 within 0.1 y0 = 60 mm and 0.5 y0 = 300 mm from
the end.
Select (6) 2 legged 8 mm diameter stirrups.
Provide Ast = 824.6 = 275 mm2 within 0.5 y0 = 300 mm and y0 = 600 mm from the
end.
Select (5) 2 legged 6 mm diameter stirrups.

200
300
End view

(5) 6 mm stirrups from 300 to 600
(6) 8 mm stirrups from 60 to 300
End zone reinforcement

8.1 Cantilever Beams

Analysis
Determination of Limiting Zone
Tendon Profile
Introduction
Prestressed cantilever beams are present in buildings and bridges. Usually, the
cantilever is provided with a back span (anchor span) to reduce the torsion in the
supporting member. In a building, the cantilever can be an extension of a continuous
beam. In a bridge, the cantilever is a part of the balanced cantilever girder. The
construction can be in-situ, where the concrete is cast in formwork that is temporarily
supported. Else, a segmental construction can be adopted, wherein slices of the girder
are placed and temporarily prestressed.
The following figure illustrates the cantilevers in buildings and bridges schematically.
Segment
Back span Cantilever span Cantilever spans

(a) Cantilever in a building frame (b) Cantilevers in a bridge girder
Figure 8-1.1 Cantilevers in buildings and bridges

Figure 8-1.2 Segmental construction, Bridge over Hooghly,

West Bengal (Courtesy: L & T Ramboll)

The analysis of a section of a prestressed cantilever for flexural stresses is similar to

that for a simply supported beam. The difference is that for gravity loads, the bending
moment in cantilever is negative, that is compression is generated at the bottom. Thus,
the CGS is placed above the CGC and the eccentricity e is negative.
The following aspects need to be considered in the analysis and design of a

prestressed cantilever beam.
1) Certain portions of the back span are subjected to both positive and negative
moments. Hence, there will be two design moments at service loads.
2) The beam may be subjected to partial loading and point loading.
3) The sequence of loading is important to design the prestressing force.
4) High values of moment and shear occur simultaneously near the support.
8.1.1 Analysis
The analysis of a cantilever beam with a back span, is illustrated to highlight the aspects
stated earlier. The bending moment diagrams for the following load cases are shown
schematically in the following figure.
1) Dead load (DL)
2) Live load (LL) only on the back span
3) Live load only on the cantilever span
4) Dead load and live load along full length.

wDL
Moment diagram for DL
wLL
Moment diagram for LL on back span
wLL
Moment diagram for LL on cantilever span
wDL + wLL
Moment diagram for DL + LL along full length

Figure 8-1.3 Moment diagrams for dead and live loads

The envelop moment diagrams are calculated from the analysis of each load case and
their combinations. In the following envelop moment diagrams, Mmax and Mmin represent
the highest and lowest values (algebraic values with sign) of the moments at a section,
respectively. Note that certain portions of the beam are subjected to both positive and
negative moments.
Mmax
Mmin
Figure 8-1.4 Envelop moment diagrams for dead and live loads
For moving point loads as in bridges, first the influence line diagram is drawn. The
influence line diagram shows the variation of the moment or shear for a particular
location in the girder, due to the variation of the position of a unit point load. The vehicle
load is placed based on the influence line diagram to get the worst effect.
8.1.2 Determination of Limiting Zone
The limiting zone of placing the CGS of the tendons is helpful in selecting a tendon
profile. Note that for a non-prismatic section, the section properties such as A, I, kt, kb, r
change with position along the length.
The limiting zone was explained for a simply supported beam in Section 4.4, Design of
Sections for Flexure (Part III). Here the concept and the equations are first reviewed for
a simply supported beam with positive moment.
For full prestressed members (Type 1), tension is not allowed under service conditions.
If tension is also not allowed at transfer, the compression in concrete (C) always lies
within the kern zone. The limiting zone is defined as the zone for placing the CGS of

the tendons such that C always lies within the kern zone. Also, the maximum
compressive stresses at transfer and service should be within the allowable values.
Also, the maximum compressive stresses at transfer and service should be within the
allowable values.
The limiting zone is determined from the maximum or minimum eccentricities of the
CGS along the beam corresponding to the extreme positions of C. Note that, the
limiting zone is related with the CGS of the tendons. Individual tendons may lie outside
the limiting zone.
For a simply supported beam, the maximum eccentricity (emax) at any section
corresponds to the lowest possible location of C at transfer, that generates allowable
tensile stress at the top of the section. The maximum compressive stress at the bottom
should be within the allowable value. The minimum eccentricity (emin) at any section
corresponds to the highest possible location of C at service, that generates allowable
tensile stress at the bottom of the section. The maximum compressive stress at the top
should be within the allowable value.
The values of emax and emin can be determined by equating the stresses at the edges of
concrete with the allowable values. Else, explicit expressions of emax and emin can be
developed. The following material gives the expressions of emax and emin for Type 1 and
Type 2 sections. The values of emax and emin can be determined at regular intervals
along the length of the beam. The zone between the loci of emax and emin is the limiting
zone of the section for placing the CGS. The equations are summarised
Type 1 Section
Based on the stress at the top at transfer, the following expression of emax is derived.
Msw
emax - k b =
P0
Msw
or, emax = + kb
P0 (8-1.1)

Also, the stress at the bottom should be less than the allowable value at transfer.
Based on the stress at the bottom at service, the following expression of emin is derived.
MT
emin + kt =
Pe
MT (8-1.2)
or, emin = - kt
Pe
Also, the stress at the top should be less than the allowable value at service. If for a
particular section emin is negative, it implies that the CGS can be placed above the CGC.
This happens near the supports.
Type 2 Section
The corresponding equations for a Type 2 section are given below.
Msw + fct,all Akb

emax - kb =
P0
Msw + fct,all Akb (8-1.3)
or, emax = + kb
P0
MT - fct,all Akt
emin + kt =
Pe
MT - fct,all Akt
or, emin = - kt (8-1.4)
Pe
In a simply supported beam, the external moments are always positive. The minimum
moment is due to self weight. The maximum moment is under service loads. For
cantilever beams, the minimum external moment need not be at transfer, when the
moment is due to self weight (MSW). Also, under service loads there are two moments
Mmin and Mmax at a location, obtained from the envelop moment diagrams.
The maximum and minimum eccentricities emax and emin at a particular location are first
determined for service loads from Mmin and Mmax , respectively, at that location and the
effective prestress Pe. Next, another set of emax and emin are calculated at transfer from
MSW and the prestress P0. The final emax is the lower of the two values calculated at
service and at transfer. Similarly, the final emin is the higher of the two values calculated
at service and at transfer.

The expressions of emax and emin for the simply supported beam were developed for
positive moments. For a cantilever, corresponding to a negative moment, the
eccentricity implies that the CGS is located above CGC. The expressions for a
cantilever are given below.
Type 1 section
At Service
Mmin (8-1.5)
emax = + kb
Pe
Mmax (8-1.6)
emin = - kt
Pe
At Transfer
Msw
emax = + kb (8-1.7)
P0
Msw
emin = - kt (8-1.8)
P0
Type 2 section
At Service
M min + fct,all Ak b (8-1.9)

emax = + kb
Pe
M max - fct,all Akt
emin = + kt (8-1.10)
Pe
At Transfer
MSW + fct,all Ak b (8-1.11)
emax = + kb
P0
MSW - fct,all Akt (8-1.12)
emin = - kt
P0

A typical limiting zone is shown in the following figure.

kt Limiting zone
kb
Figure 8-1.5 Limiting zone for a non-prismatic beam
8.1.3 Tendon Profile
The tendon profiles for a few beams with cantilever spans are shown schematically in
the following figures. The vertical scale is enlarged to show the location of the CGS
with respect to CGC.
a) Uniform cross section
b) Varying cross section in the cantilever span

Figure 8-1.6 Beams with single cantilever span

b) Varying level of soffit
c) Varying level of top of beam

Figure 8-1.7 Beams with double cantilever spans
For a prismatic beam with uniform cross section along the length, the tendon profile is
similar to the moment diagram under uniform load. Thus for regions of negative
moment, the CGS is located above the CGC. Since there cannot be a sharp kink in the
tendons and, the supports are not true point supports, the profile is shown curved at the
right support.
For a beam with varying depth, the tendon profile can be adjusted (within emax and emin)
to be straight for convenience of layout of the tendons.

8.2 Continuous Beams (Part I)

Analysis
Incorporation of Moment due to Reactions
Pressure Line due to Prestressing Force
Introduction
Beams are made continuous over the supports to increase structural integrity. A
continuous beam provides an alternate load path in the case of failure at a section. In
regions with high seismic risk, continuous beams and frames are preferred in buildings
and bridges. A continuous beam is a statically indeterminate structure.
The advantages of a continuous beam as compared to a simply supported beam are as

follows.
1) For the same span and section, vertical load capacity is more.
2) Mid span deflection is less.
3) The depth at a section can be less than a simply supported beam for the same
span. Else, for the same depth the span can be more than a simply supported
beam.
The continuous beam is economical in material.
4) There is redundancy in load path.
Possibility of formation of hinges in case of an extreme event.
5) Requires less number of anchorages of tendons.
6) For bridges, the number of deck joints and bearings are reduced.
Reduced maintenance
There are of course several disadvantages of a continuous beam as compared to a

simply supported beam.
1) Difficult analysis and design procedures.
2) Difficulties in construction, especially for precast members.
3) Increased frictional loss due to changes of curvature in the tendon profile.
4) Increased shortening of beam, leading to lateral force on the supporting columns.
5) Secondary stresses develop due to time dependent effects like creep and
shrinkage, settlement of support and variation of temperature.

6) The concurrence of maximum moment and shear near the supports needs
proper detailing of reinforcement.
7) Reversal of moments due to seismic force requires proper analysis and design.
End span Intermediate span

(a) Continuous beam in a building frame
End span Intermediate span

(b) Continuous beam in a bridge
Figure 8-2.1 Continuous beams in buildings and bridges
8.2.1 Analysis
The analysis of continuous beams is based on elastic theory. This is covered in text
books of structural analysis. For prestressed beams the following aspects are
important.
1) Certain portions of a span are subjected to both positive and negative moments.
These moments are obtained from the envelop moment diagram.
2) The beam may be subjected to partial loading and point loading. The envelop
moment diagrams are developed from pattern loading. The pattern loading
refers to the placement of live loads in patches only at the locations with positive
or negative values of the influence line diagram for a moment at a particular
location.
3) For continuous beams, prestressing generates reactions at the supports. These
reactions cause additional moments along the length of a beam.

The analysis of a continuous beam is illustrated to highlight the aspects stated earlier.
The bending moment diagrams for the following load cases are shown schematically in
the following figures.
1) Dead load (DL)
2) Live load (LL) on every span
3) Live load on a single span.
wDL
Moment diagram for DL
wLL
Moment diagram for LL on every span
wLL
Moment diagram for LL on one span

Figure 8-2.2 Moment diagrams for dead and live loads

For moving point loads as in bridges, first the influence line diagram is drawn. The
influence line diagram shows the variation of the moment or shear for a particular
location in the girder, due to the variation of the position of a unit point load. The vehicle
load is placed based on the influence line diagram to get the worst effect. An influence
line diagram is obtained by the Mller-Breslau Principle. This is covered in text books
of structural analysis.
IS:456 - 2000, Clause 22.4.1, recommends the placement of live load as follows.
1) LL in all the spans.
2) LL in adjacent spans of a support for the support moment. The effect of LL in the
alternate spans beyond is neglected.
3) LL in a span and in the alternate spans for the span moment.
The envelop moment diagrams are calculated from the analysis of each load case and
their combinations. The analysis can be done by moment distribution method or by
computer analysis.
In lieu of the analyses, the moment coefficients in Table 12 of IS:456 - 2000 can be
used under conditions of uniform cross-section of the beams in the several spans,
uniform loads and similar lengths of span.
The envelop moment diagrams provide the value of a moment due to the external
loads. It is to be noted that the effect of prestressing force is not included in the envelop
moment diagrams. The following figure shows typical envelop moment diagrams for a
continuous beam.
Mmin
Mmax
Figure 8-2.3 Envelop moment diagrams for DL + LL

In the above diagrams, Mmax and Mmin represent the highest and lowest values
(algebraic values with sign) of the moments at a section, respectively. Note that certain
portions of the beam are subjected to both positive and negative moments. The
moment from the envelop moment diagrams will be represented as the M0 diagram.
This diagram does not depend on whether the beam is prestressed or not.
8.2.2 Incorporation of Moment Due to Reactions
As mentioned before, for continuous beams prestressing generates reactions at the

supports. The reactions at the intermediate supports cause moment at a section of the
continuous beam. This moment is linear between the supports and is in addition to the
moment due to the eccentricity of the prestressing force. The concept is explained by a
simple hypothetical two-span beam in the following figure. The beam is prestressed
with a parabolic tendon in each span, with zero eccentricity of the CGS at the supports.
The moment diagram due to the eccentricity of the prestressing force and neglecting the
intermediate support is denoted as the M1 diagram. This diagram is obtained as M1 =
Pe, where, P is the prestressing force (P0 at transfer and Pe at service) and e is the
eccentricity of the CGS with respect to CGC. Neglecting the variation of P along the
length due to frictional losses, the value of M1 is proportional to e. Hence, the shape of
the M1 diagram is similar to the cable profile.

l l
Profile of the CGS
Pe
M1 diagram
3wl / 8 3wl / 8
w
10wl / 8
Simplified free body diagram
wl2/2 = 4Pe
5wl2/8 = 5Pe
=
Pe
M2 diagram
Figure 8-2.4 Moment diagram due to prestressing force for a two-span beam
Next, the moment diagram due to the prestressing force and including the effect of the
intermediate support is denoted as the M2 diagram. This is obtained by structural
analysis of the continuous beam subjected to the upward thrust. Since the profile of the
tendon is parabolic in each span, the upward thrust is uniform and is given as wup = w =
8Pe/l2. The downward thrust at the location of the central kink is not considered as it
directly goes to the intermediate support. The hold down force at the intermediate
support neglecting the downward thrust is 10wupl/8 = 10Pe/l. The downward forces at
the ends are from the anchorages. The moment diagram due to wup alone (without the

support) is added to that due to the hold down force. The resultant M2 diagram is
similar to the previous M1 diagram, but shifted linearly from an end support to the
intermediate support.
For a general case, the resultant moment (M2) at a location due to the prestressing
force can be written as follows.
M2 = M1 + M1/ (8-2.1)
In the above equation,

M1 = moment due to the eccentricity of the prestressing force neglecting the
intermediate supports
= Pee.
M 1/ = moment due to the reactions at intermediate supports.
Pe = effective prestress
e = eccentricity of CGS with respect to CGC.
M1 is the primary moment and M1/ is the secondary moment.
The moment due to the external loads (M0) that is obtained from the envelop moment
diagrams is added to M2 to get the resultant moment (M3) at a location.
M3 = M2 + M0
M3 = M1 + M1/ + M0 (8-2.2)
The variation of M3 along the length of the beam (M3 diagram) can be calculated as
follows.
1) The M0 diagram is available from the envelop moment diagram.
2) Plot M1 diagram which is similar to the profile of the CGS. The variation of Pe
along the length due to friction may be neglected.
3) Plot the shear force (V) diagram corresponding to the M1 diagram from the
relationship V = dM1/dx.
4) Plot the equivalent load (weq) diagram corresponding to the V diagram from the
relationship weq = dV/dx. Note, over the supports weq can be downwards. Also,
a singular moment needs to be included at an end when the eccentricity of the
CGS is not zero at the end.

5) Calculate the values of M2 for the continuous beam (with the intermediate
supports) subjected to weq using a method of elastic analysis (for example,
moment distribution or computer analysis). Plot the M2 diagram.
6) The M3 diagram can be calculated by adding the values of M2 and M0 diagrams
along the length of the beam.
The following figures explain the steps of developing the M2 diagram for a given profile
of the CGS and a value of Pe.
Given profile of the CGS
Step (2) Plotting of M1 diagram
Step (3) Plotting of V diagram
Step (4) Plotting of weq diagram
Step (5) Plotting of M2 diagram

Figure 8-2.5 Development of the moment diagram due to prestressing force

The important characteristics of the diagrams are as follows.

1) A positive eccentricity of the CGS creates a negative moment (M1) and an
upward thrust.
2) The M2 diagram has a similar shape to the M1 diagram, which is again similar to
the profile of the CGS. This is because the moment generated due to the
reactions (M1/) is linear between the supports.
8.2.3 Pressure Line due to Prestressing Force
The pressure line (thrust line or C-line) due to the prestressing force only can be
determined from the M2 diagram. It is to be noted that the external loads are not
considered in this pressure line. This is used to select the profile of the CGS.
The calculation of pressure line from the M2 diagram is based on the following
expression. The pressure line can be plotted for the different values of M2 along the
length.
ec= M2/Pe (8-2.3)
Here,
ec = distance of the pressure line from the CGC at a location. A positive
value of ec corresponds to a hogging value of M2 and implies that the
pressure line is beneath the CGC.
The following sketch shows the pressure line for a given profile of the CGS.
Pressure line
e ec CGC
Profile of the CGS

Figure 8-2.7 Pressure line for a continuous beam
The important characteristics of the pressure line are as follows.

1) The shift of the pressure line from the profile of the CGS is a linear
transformation. It is because M2 diagram has a similar shape to the profile of the
CGS.

The pressure line will have the same intrinsic shape as the profile of the CGS.
2) Since M2 is proportional to the prestressing force, the eccentricity of the pressure
line (ec) remains constant even when the prestressing force drops from the initial
value P0 to the effective value Pe.
The location of the pressure line for a given profile of the CGS is fixed,
irrespective of the drop in the prestressing force.
Example 8-2.1
The profile of the CGS for a post-tensioned beam is shown in the sketch. Plot the
pressure line due to a prestressing force Pe = 1112 kN.
0.06 0.24 0.12 0.27 CGC

0.08 rad
0.176 rad
A 9m D 6m B 7.5 m 7.5 m C
Values of eccentricity in metres.
Solution
1) Plot M1 diagram
The values of M1 are calculated from M1 = Pee.

e (m) M1 (kN m)
0.06 66.7
0.24 266.9
0.12 133.4
0.27 300.2

0.12
0.06 0.24 0.27
A D B C
Profile of the CGS
133.4
66.7
266.9 300.2
M1 diagram (kN m)
2) Plot V diagram
For AD,
dM1
V=
dx
-266.9 - (-66.7)
=
9
= -22.2 kN
For DB,
dM1
V=
dx
133.4 - (-266.9)
=
6
= 66.7 kN
For BC, to find dM1/dx, an approximate parabolic equation for the M1 diagram can be
used.
4Pe ex
M1 = - (L - x )
L2
dM1
V=
dx
4P e
= - 2e ( L - 2 x )
L
Pee x
M1

At B,
dM1
V=
dx x=0
4P e
=- e
L
4 (133.4 + 300.2)
=-
15
= -115.6 kN
The exact value of V at B is
V = - 107.0 kN
The difference of V between C and B is given from the change in slope of the M1
diagram.
V|C - V|B = 0.176 1112
= 195.7 kN
Therefore, value of V at C is given as follows.

V|C = 195.7 107.0
= 89.0 kN
133.4
66.7
266.9 300.2
M1 diagram (kN m)
66.7 89.0
22.2
107.0
V diagram (kN)
3) Plot equivalent load (weq) diagram
Include moment 66.7 kN-m at A.
Point load at D
W|D = 66.7- (- 22.2 )
= 88.9 kN

Since B is a reaction point, the downward load at B need not be considered.

Distributed load within B and C
89.0 - (-107.0)
w BC =
15
= 13.0 kN/m
89.0
66.7
22.2
107.0
V diagram (kN)
66.7 kN m 88.9 kN 13.0 kN/m
Equivalent load diagram
4) Plot the M2 diagram.

Calculate moment at supports by moment distribution
88.9 13.0
66.7
DF 0.5 0.5
88.9962 88.992 6 13.0152
FEM
152 152 12
=128 = -192 = 244 244
Bal 194.7 244
CO 97 122
Bal 38.5 38.5
Total 66.7 327.5 327.5 0
In the previous table,

Bal = Balanced
CO = Carry Over moment
DF = Distribution Factor
FEM = Fixed End Moment.

The moment at the spans can be determined from statics. But this is not necessary as
will be evident later.
66.7 kN m 88.9 kN 13.0 kN/m
327.0
0.0
66.7
M2 diagram (kN m)
5) Calculate values of ec at support.
The values of ec are calculated from ec= M2/Pe.
M2 (kN m) ec (m)
66.7 0.06
327.0 0.294
0.0 0.184
The deviations of the pressure line from the CGS at the spans can be calculated by
linear interpolation.
Pressure line
0.294
0.136 0.184
0.06
Profile of CGS

8.3 Continuous Beams (Part II)

Principle of Linear Transformation
Concordant Tendon Profile
Tendon Profiles
Partially Continuous Beams
Analysis for Ultimate Strength
Moment Redistribution
Introduction
Before the discussion on the tendon profile (profile of the CGS), the following concepts
are introduced.
1) Principle of linear transformation
2) Concordant tendon profile.
8.3.1 Principle of Linear Transformation
When the profile of the CGS is moved over the interior supports of a continuous beam
without changing the intrinsic shape of the profile within each individual span, the profile
is said to be linearly transformed. In a linear transformation, the curvatures remain
constant and the locations of bends remain unchanged.
The following sketch explains the concept of linear transformation of the profile of the
CGS.
Profile after linear transformation
CGC
Profile of the CGS

Figure 8-3.1 Linear transformation of the profile of the CGS
Linear transformation cannot involve the movement of the CGS at the ends of a beam
or at the support of a cantilever.

Theorem
In a continuous beam, a profile of the CGS can be linearly transformed without
changing the position of the resultant pressure line. This theorem can be proved based
on the requirement that the curvature of the profile of the CGS remains constant under
linear transformation. The following sketch explains that the pressure line remains
constant for linearly transformed profiles of the CGS.
Profile after linear transformation
Common pressure line
CGC
Profile of the CGS

Figure 8-3.2 Pressure line for linearly transformed profiles of the CGS
8.3.2 Concordant Tendon Profile
A concordant tendon profile in a continuous beam is a profile of the CGS which

produces a pressure line coincident with the profile itself. A concordant tendon profile
does not produce reactions at the supports or secondary moments in the spans. The
upward and downward equivalent loads balance each other.
The following sketch shows a concordant tendon profile which is coincident with the
pressure line.
Concordant profile at
the pressure line
CGC
Figure 8-3.3 Concordant profile
The advantage of a concordant cable profile is that the calculations become simpler.

1) There is no secondary moment in the spans due to the prestress. The M2

diagram coincides with the M1 diagram.
2) The pressure line due to the prestress coincides with the cable profile. The shift
of the pressure line due to external loads can be measured from the profile
directly.
A concordant profile can be developed from the moment diagram due to external loads
for a certain load combination using the following theorem.
Theorem
Every real moment diagram for a continuous beam on non-settling supports produced
by any combination of external loads, whether transverse loads or moments, plotted to
any scale, is one location for a concordant tendon in that beam.
The theorem can be proved based on the condition of no deflection at the supports due
to external loads. Also, for a concordant profile since there is no reaction at any
support, there is no possibility of deflections at the supports. Thus, it is easy to obtain a
concordant profile from the moment diagram of the external loads for a certain load
combination, drawn to a certain scale. The following figure shows the steps of the
development of concordant profile from the moment diagram.
(a) Loading on a continuous beam
(b) Moment diagram
CGC
(c) Concordant profile

Figure 8-3.4 Development of concordant profile

Discussion
The computation of the concordant profile helps in the layout of the tendon profile. The
tendon profile need not be designed to be a concordant profile. It should be such that
the stresses in concrete at transfer and at service are within the allowable values. If a
concordant profile is selected then the calculations become simpler.
8.3.3 Tendon Profiles
The steps of selecting a tendon profile (profile of the CGS) are based on trials. The
steps are as follows.
1) Assume the section of the beam for calculating self weight. For the preliminary
design, the type and depth (h) of the section can be selected based on
architectural requirement and deflection criteria.
2) Calculate the moment due to self weight (Msw) and the maximum moment (Mmax)
and minimum moment (Mmin) along the length of the beam (envelop moment
diagrams) due to the external loads, including self weight.
3) Compute the required Pe based on the values of Mmax and Mmin, at the critical
locations, similar to the calculations for a simply supported beam. Revise the
section if necessary. If Msw is large,
Pe = MT / z (8-3.1)
z 0.65h (8-3.2)
Here,
MT = Mmax or Mmin
z = estimated lever arm.
4) Considering fpe = 0.7fpk , calculate area of prestressing steel Ap = Pe / fpe.

5) Check the area of the cross-section (A) based on A = Pe /(0.5fcc,all).
6) Calculate the kern distances kb and kt, and the maximum and minimum
eccentricities (emax and emin) along the length. The zone between emax and emin
along the length of the beam is the limiting zone. The equations of emax and emin
are same as that for a simply supported beam.
The value of P0 can be estimated from Pi as follows.

a. 90% of the initial applied prestress (Pi) for pre-tensioned members.

b. Equal to Pi for post-tensioned members.

The value of Pi can be estimated as follows.
Pi = Ap (0.8fpk) (8-3.3)
Ap = Pe / 0.7fpk (8-3.4)
7) Select a trial profile of the CGS within the limiting zone. If the profile is a
concordant profile, the pressure line due to prestress coincides with the profile of
the CGS.
Calculate the shift in the pressure line due to external loads. For a Type 1 member, if
the final pressure line lies within the kern zone, then the solution is acceptable. If final
pressure line lies outside the kern zone, try another profile.
For Type 2 and Type 3 members, if the final pressure line lies within a zone such that
the stresses at the edges are within the allowable values, then the solution is
acceptable. If final pressure line lies outside the zone, try another profile.
8) Linearly transform the profile of the CGS to satisfy the cover requirements and
the convenience of prestressing.
For a prismatic beam with uniform cross section along the length, the tendon profile can
be selected similar to the moment diagram under uniform load. Since there cannot be a
sharp kink in the tendons and the supports are not true point supports, the profile needs
to be curved at an intermediate support. For a beam with varying depth, the tendon
profile can be adjusted (within the limiting zone) to be relatively straight for convenience
of layout of the tendons and reduction of losses due to friction. The tendons can be of
segments of single curvature to reduce frictional losses.
The following sketches show the profiles of the CGS for common continuous beams.

b) Varying cross section
c) Combination of a) and b)
d) Uniform cross section with overlapping tendons
Figure 8-3.5 Profiles of CGS for continuous beams
8.3.4 Partially Continuous Beams
Due to the difficulties in construction of continuous beams, an intermediate system

between simply supported beams and continuous beams is adopted. These are called
partially continuous beams.
First, the individual precast members are placed at the site. Next continuity is
introduced by additional prestressing tendons or coupling the existing tendons.
Continuity can also be introduced in a composite construction, where non prestressed
continuity reinforcement is introduced in the cast-in-place topping slab.
A few examples are given in the following sketches. Other innovative schemes are also
used.

Additional
tendon
individual
tendon
a) With additional tendon
Coupler
Jack
b) With coupling of tendons
Cast-in-place
Topping
c) Composite construction with continuity reinforcement

Figure 8-3.6 Partially continuous beams
8.3.5 Analysis for Ultimate Strength
The analysis of continuous beams for ultimate strength is difficult for the following
reasons.
1) Due to non-linear behaviour, superposition of stresses is not valid.
The concept of load balancing is not truly applicable.
2) The prestressing force varies at the location of cracks.
3) Neglect of the secondary moment due to prestressing is erroneous, unless full
moment redistribution is allowed.
Clause 18.6.4 of IS:1343 - 1980 insists on considering the secondary moment.

8.3.6 Moment Redistribution
It was mentioned in Section 3.4, Analysis of Member under Flexure (Part III), that there
is an inconsistency in the traditional analysis at the ultimate state. The demand is
calculated based on elastic analysis, whereas the capacity is calculated based on the
non-linear limit state analysis. Although the analysis for demand at ultimate is based on
an elastic analysis, IS:1343 - 1980 allows to take advantage of the post-yield
deformation of the highly stressed sections in a continuous beam. The underlying
concept is known as moment redistribution.
Moment redistribution means the transfer of additional moments to the less stressed
sections, as the highly stressed sections with peak moments yield on reaching their
ultimate moment capacities.
To apply moment redistribution, the highly stressed sections are designed for lower
moments and the less stressed sections are designed to carry higher moments than the
values obtained from an elastic analysis. This gives an economical solution.
IS:1343 - 1980, Clause 21.1.1 specifies the following conditions for moment
redistribution.
1) The redistributed moments must be in a state of static equilibrium with the
factored external loads.
2) For serviceability requirements, the ultimate moment of resistance at any section
(MUR) should not be less than 80% of the moment demand from an elastic
analysis (Mu).
3) To limit the demand on post-yield rotation, the reduction in moment at the highly
stressed sections is limited to 20% of the numerically largest moment anywhere
in the beam calculated by an elastic analysis.
4) To ensure ductile behaviour of the highly stressed sections, the following
relationship should be checked.
xu M
+ 0.5
d 100 (8-3.5)

Here,
xu = depth of neutral axis
d = the effective depth

M = the percentage reduction in moment.
Example 8-3.1
The prestressed concrete beam shown in the figure, is fixed at the left end and
roller supported at the right. It is post-tensioned with a single tendon with a
parabolic profile, with indicated eccentricities.
a). Locate the pressure line due to application of a prestress force of 1068 kN.
b). Find the primary, secondary and total moments due to prestressing force
at the face of the fixed support.
c). What is the magnitude and direction of the reaction produced at the roller
by prestressing force?
d). What minor adjustment can be made in the tendon profile to produce a
concordant profile?
250 150 300

300
6m 6m

Solution
a) Locate the pressure line.
1) Plot M1 diagram.
The values of M1 are calculated from M1 = Pee.
e (m) M1 (kN m)
0.250 267.0
0.150 160.2
0.0 0.0
250 150
Profile of the CGS
267.0
160.2
M1 diagram (kN m)
2) Plot V diagram.
The M1 diagram is made up of two parabolic segments.
267.0
160.2
M1 diagram (kN m)
For each segment,
4Pex
M1 = - (L - x )
L2

dM1
V=
dx
4Pe
= - 2 ( L - 2x )
L
Pe x
M1
4Pe 4Pe
V =- V =
x=0
L x =L
L
4 ( 267.0 +160.2 ) 4160.2

-
12 12
= -142.4 kN = 53.4 kN
V diagram (kN)
3) Plot equivalent load (weq) diagram.
dV
w eq =
dx
53.4 +142.4
=
12
= 16.3 kN/m
16.3 kN/m
4) Plot the M2 diagram.

Calculate moment at supports by moment distribution.

FEM 16.3122 195.8

12
=195.8
Bal 195.8
CO 97.9
Total 293.7 0
In the previous table,
Bal = Balanced
CO = Carry Over moment
FEM = Fixed End Moment.
The moment at the span can be determined from statics. But this is not necessary as
will be evident later.
16.3 kN/m
293.7
M2 diagram (kN m)
5) Calculate values of ec at support.
The values of ec are calculated from ec= M2/Pe.

M2 (kN m) ec (m)
293.7 0.275
The deviations of the pressure line from the CGS at the span can be calculated by
linear interpolation.

Pressure line
275 250
150 12.5 = 137.5
(275 250)
= 12.5
b) Calculation of primary, secondary and total moments.
M1 = 267.0 kN m primary
M2 = 293.7 kN m total
M1/ = M2 M1
= 293.7 267.0
= 26.7 kN m secondary
c) Calculation of reaction.
w eq L R1 R1
R1 =
2
16.3 12 weq
=
2
= 97.6 kN
M2
R2 R2
L
M2
R2 =
L
293.7
=
12
= 24.5 kN
R1 R2 = 73.1 kN
Resultant reaction at roller is downwards.
d) The tendon can be shifted to coincide with the pressure line to get a concordant
profile.

Concordant profile
275 137.5
Values in mm.

9.1 Composite Sections

Introduction
Analysis of Composite Sections
Design of Composite Sections
Analysis for Horizontal Shear Transfer
9.1.1 Introduction
A composite section in context of prestressed concrete members refers to a section with

a precast member and cast-in-place (CIP) concrete. There can be several types of
innovative composite sections. A few types are sketched below.
T T T T T T
Box section Composite beam-slab T-section

Figure 9-1.1 Examples of composite sections
The following photos show the reinforcement for the slab of a box girder bridge deck
with precast webs and bottom flange. The slab of the top flange is cast on a stay-in
formwork. The reinforcement of the slab is required for the transverse bending of the
slab. The reinforcement at the top of the web is required for the horizontal shear
transfer.

Figure 9-1.2 Box girder bridge deck with precast webs and bottom flange and CIP
slab; Top: Aerial view, Bottom: Close-up view (Courtesy: Jos Turmo)
The advantages of composite construction are as follows.

1) Savings in form work
2) Fast-track construction
3) Easy to connect the members and achieve continuity.

The prestressing of composite sections can be done in stages. The precast member
can be first pre-tensioned or post-tensioned at the casting site. After the cast-in-place
(cast-in-situ) concrete achieves strength, the section is further post-tensioned.
The grades of concrete for the precast member and the cast-in-place portion may be
different. In such a case, a transformed section is used to analyse the composite
section.
9.1.2 Analysis of Composite Sections
The analysis of a composite section depends upon the type of composite section, the
stages of prestressing, the type of construction and the loads. The type of construction
refers to whether the precast member is propped or unpropped during the casting of
the CIP portion. If the precast member is supported by props along its length during the
casting, it is considered to be propped. Else, if the precast member is supported only at
the ends during the casting, it is considered to be unpropped.
The following diagrams are for a composite section with precast web and cast-in-place
flange. The web is prestressed before the flange is cast. At transfer and after casting
of the flange (before the section behaves like a composite section), the following are the
stress profiles for the precast web.
At transfer After casting of flange

Section
due to due to
P0 + MSW Pe + MSW MCIP
Figure 9-1.3 Stress profiles for the precast web
Here,
P0 = Prestress at transfer after short term losses
Pe = Effective prestress during casting of flange after long term losses
MSW = Moment due to self weight of the precast web
MCIP = Moment due to weight of the CIP flange.
At transfer, the loads acting on the precast web are P0 and MSW. By the time the flange
is cast, the prestress reduces to Pe due to long term losses. In addition to Pe and MSW,

the web also carries MCIP. The width of the flange is calculated based on the concept of
effective flange width as per Clause 23.1.2, IS:456 - 2000.
At service (after the section behaves like a composite section) the following are the
stress profiles for the full depth of the composite section.
Unpropped Propped
+ or +
At service
Section due to
Pe + MSW MCIP MLL
Figure 9-1.4 Stress profiles for the composite section
Here, MLL is the moment due to live load. If the precast web is unpropped during
casting of the flange, the section does not behave like a composite section to carry the
prestress and self weight. Hence, the stress profile due to Pe + MSW + MCIP is
terminated at the top of the precast web. If the precast web is propped during casting
and hardening of the flange, the section behaves like a composite section to carry the
prestress and self weight after the props are removed. The stress profile is extended up
to the top of the flange. When the member is placed in service, the full section carries
MLL.
From the analyses at transfer and under service loads, the stresses at the extreme
fibres of the section for the various stages of loading are evaluated. These stresses are
compared with the respective allowable stresses.
Stress in precast web at transfer
P0 P0ec MSW c
f =- (9-1.1)
A I I
Stress in precast web after casting of flange
P P ec (M + MCIP )c (9-1.2)
f = - e e SW
A I I

Stress in precast web at service
(a) For unpropped construction
Pe Peec (MSW + MCIP )c MLLc / (9-1.3a)

f =-
A I I I/
(b) For propped construction
Pe Pe ec MSW c (MCIP + MLL )c /

f =- (9-1.3b)
A I I I/
Here,
A = area of the precast web
c = distance of edge from CGC of precast web
c/ = distance of edge from CGC of composite section
e = eccentricity of CGS
I = moment of inertia of the precast web
I / = moment of inertia of the composite section.
From the analysis for ultimate strength, the ultimate moment capacity is calculated.
This is compared with the demand under factored loads. The analysis at ultimate is
simplified by the following assumptions.
1) The small strain discontinuity at the interface of the precast and CIP portions is
ignored.
2) The stress discontinuity at the interface is also ignored.
3) If the CIP portion is of low grade concrete, the weaker CIP concrete is used for
calculating the stress block.
The strain and stress diagrams and the force couples at ultimate are shown below.

bf
0.0035 0.447 fck
Cuf
Df Cuw
d xu
Ap +
fpu
Tuw Tuf
bw p
pu
Section Strain Stress Force
Figure 9-1.5 Sketches for analysis at ultimate

bw = breadth of the web
p = strain difference for the prestressing steel
fck = characteristic compressive strength of the weaker concrete
Cuw = resultant compression in the web (includes portion of flange
above precast web)
Cuf = resultant compression in the outstanding portion of flange
Tuw = portion of tension in steel balancing Cuw.
Tuf = portion of tension balancing Cuf.

Cuw = 0.36fck xu bw (9-1.4)
Cuf = 0.447fck (bf - bw )Df (9-1.5)
Tuw = Apw fpu (9-1.6)
Tuf = Apf fpu (9-1.7)
The symbols for the areas of steel are as follows.

Apf = part of Ap that balances compression in the outstanding flanges

Apw = part of Ap that balances compression in the web
The equilibrium equations are given below. These equations are explained in Section
3.5, Analysis of Members under Flexure (Part IV). The ultimate moment capacity (MuR)
is calculated from the second equation.
F = 0
(A pw + Apf ) fpu = 0.36 fck xu bw + 0.447 fck ( bf - bw ) Df (9-1.8)
MuR = Apw fpu ( d - 0.42 xu ) + Apf fpu ( d - 0.5Df ) (9-1.9)
9.2.3 Design of Composite Sections
The design is based on satisfying the allowable stresses under service loads and at
transfer. The section is then analysed for ultimate loads to satisfy the limit state of
collapse. The member is also checked to satisfy the criteria of limit states of
serviceability, such as deflection and crack width (for Type 3 members only). Before the
calculation of the initial prestressing force (P0) and the eccentricity of the CGS (e) at the
critical section, the type of composite section and the stages of prestressing need to be
decided. Subsequently, a trial and error procedure is adopted for the design.
The following steps explain the design of a composite section with precast web and
cast-in-place flange. The precast web is prestressed before the casting of the flange.
The member is considered to be Type 1 member.
Step 1. Compute e.
With a trial section of the web, the CGS can be located at the maximum eccentricity
(emax). The maximum eccentricity is calculated based on zero stress at the top of the
precast web. This gives an economical solution. The following stress profile is used to
determine emax.

ct
CGC
CGS cb e
fb
Web section Stress profile
Figure 9-1.6 Stress profile for maximum eccentricity of CGS
Msw
emax = kb +
P0
Here,
CGC = Centroid of the precast web
kb = Distance of the bottom kern of the precast web from CGC
Msw = Moment due to self weight of the precast web.
P0 = A trial prestressing force at transfer.
Step 2. Compute equivalent moment for the precast web.

A moment acting on the composite section is transformed to an equivalent moment for
the precast web. This is done to compute the stresses in the precast web in terms of
the properties of the precast web itself and not of the composite section.
For a moment Mc which acts after the section behaves like a composite section, the
stresses in the extreme fibres of the precast web are determined from the following
stress profile.
ct
ft
CGC ct
cb
fb
bw
Composite section Stress profile
Figure 9-1.7 Stress profile for the composite section
Mc ct'
ft =
I'
Mc cb'
fb =
I'
Here,
CGC = centroid of the composite section

ct = Distance of the top of the precast web from the CGC

ct = Distance of the top of the composite section from the CGC.
cb = Distance of the bottom of the precast web (or composite section)
from the CGC
I = moment of inertia for the composite section.
The following quantities are defined as the ratios of the properties of the precast web
and composite section.
I
ct
mt =
I'
ct'
I
cb
mb =
I'
cb'
Then the stresses in the extreme fibres of the precast web can be expressed in terms of
mt and mb as follows.
mt Mc ct mt Mc (9-1.10)
ft = =
I Akb
mbMc cb mbMc
fb = = (9-1.11)
I Akt
Here,
A = Area of the precast web
kb = Distance of the bottom kern of the precast web from CGC
kt = Distance of the top kern of the precast web from CGC
The quantities mt Mc and mb Mc are the equivalent moments. Thus, the stresses in the
precast web due to Mc are expressed in terms of the properties of the precast web itself.
Step 3. Compute Pe
Let MP be the moment acting on the precast web prior to the section behaving like a
composite section. After Mc is applied on the composite section, the total moment for
the precast web is MP + mbMc.

The stress at the bottom for Type 1 member due to service loads is zero.
Therefore,
Pe Pee MP + mbMc
- - + =0
A Akt Akt
or, Pe =
(MP + mbMc ) (9-1.12)
e + kt
Note that the prestressing force is acting only on the precast web and hence, e is the
eccentricity of the CGS from the CGC of the precast web.
Step 4. Estimate P0 as follows.

a) 90% of the initial applied prestress (Pi) for pre-tensioned members.
b) Equal to Pi for post-tensioned members.
The value of Pi is estimated as follows.
Pi = Ap(0.8fpk) (9-1.13)
Ap = Pe / 0.7fpk (9-1.14)
Revise e, the location of CGS, as given in Step 1 based on the new value of P0.
Msw
emax = kb + (9-1.15)
P0
Step 5. Check for the compressive stresses in the precast web.
At transfer, the stress at the bottom is given as follows.
P0 P0e Msw
fb = - - + (9-1.16)
A Akt Akt
The stress fb should be limited to fcc,all, where fcc,all is the allowable compressive stress in
concrete at transfer (available from Figure 8 of IS:1343 - 1980).
At service,
Pe Pee (MP + mt Mc )
ft = - - + (9-1.17)
A Ak b Ak b

The stress ft should be limited to fcc,all , where fcc,all is the allowable compressive stress in
concrete under service loads (available from Figure 7 of IS:1343-1980). If the stress
conditions are not satisfied, increase A.
Step 6. Check for the compressive stress in the CIP flange.
Mc ct"
ft= (9-1.18)
I'
The stress ft/ should be limited to fcc,all , where fcc,all is the allowable compressive stress
in concrete under service loads.
9.1.4 Analysis for Horizontal Shear Transfer
With increase in the load, the bottom face of the CIP portion tends to slip horizontally
and move upwards with respect to the top face of the precast portion. To prevent this
and to develop the composite action, shear connectors in the form of shear friction
reinforcement is provided.
The required shear friction reinforcement (per metre span) is calculated as follows.
1000bv h
Asv = (9-1.19)
0.87fy
The minimum requirements of shear friction reinforcement and spacing are similar to
that for shear reinforcement in the web.

Asv = area of shear friction reinforcement in mm2/m
bv = width of the interface of precast and CIP portions
h = horizontal shear stress at the interface in N/mm2
fy = yield stress in N/mm2
= coefficient of friction
= 1.0 for intentionally roughened interface with normal weight concrete

The shear reinforcement in the web can be extended and anchored in the CIP portion to
act as shear friction reinforcement, as shown below.
bv
Intentionally roughened
Figure 9-1.8 Shear reinforcement used for shear transfer

The following example shows the analysis of a composite beam.
Example 9-1.1
The mid-span section of a composite beam is shown in the figure. The precast
web 300 mm 920 mm (depth) is post-tensioned with an initial force (P0) of 2450
kN. The effective prestress (Pe) is estimated as 2150 kN. Moment due to the self
weight of the precast web (MSW) is 270 kNm at mid-span.
After the web is erected in place, the top slab of 150 mm 920 mm (width) is
casted (unpropped) producing a moment (MCIP) of 135 kNm. After the slab
concrete has hardened, the composite section is to carry a maximum live load
moment (MLL) of 720 kNm.
Compute stresses in the section at various stages.

920
150
920
200
300
Section
Solution
1) Calculation of geometric properties.
Precast web
A = 2.76 105 mm2
I = 1.95 1010 mm2
Distance of CGC from bottom = 460 mm.
Composite section
A/ = 4.14 105 mm2
I/ = 4.62 1010 mm2
Distance of CGC/ from bottom = 638 mm.
CGC/
CGC
638
460
2) Calculation of stresses in web at transfer
P0 P0ec MSW c
f =-
A I I
245010 2450103 260 460 270106 460
3
=-
2.76105 1.951010 1.951010
= -0.22 N/mm2
= -17.54 N/mm2

3) Calculation of stresses in web after long term losses
Pe Peec MSW c
f =-
A I I
215010 2150103 260 460 270106 460
3
=-
2.76105 1.951010 1.951010
= -0.97 N/mm2
= -14.61N/mm2
4) Calculation of stresses in web after casting of flange
Pe Peec (MSW + MCIP )c

f =-
A I I
215010 2150103 260 460 (270 +135)106 460
3
=-
2.76105 1.951010 1.951010
= -4.16 N/mm2
= -11.42 N/mm2
5) Calculation of stresses in the composite section at service

Stress due to MLL
At top fibre At bottom fibre
MLLct// MLLcb/
/
ft = - fb =
I/ I/
750103 432 750103 638
=- =
4.621010 4.621010
= -7.01N/mm2 = 10.36 N/mm2
At top fibre of precast web, the stress due to MLL is calculated from proportionality of
triangles.
7.01 282
ft = -
432
= -4.57 N/mm2

Total stress in precast web

ft = -4.16 - 4.57 fb = -11.42 +10.36
2
= -8.73 N/mm = -1.06 N/mm2
Total stress in CIP slab

The total stress is due to MLL only.

/
f = - 4.57 N/mm
b
2 ft / = -7.01N/mm2
Stress profiles
7.01 7.01
0.22 0.97 4.16 4.57 4.57
8.73
17.54 14.61 11.42 10.36 1.06
At transfer After casting At service

After losses Due to MLL

9.2 One-way Slabs

Introduction
Analysis and Design
9.2.1 Introduction
Slabs are an important structural component where prestressing is applied. With

increase in the demand for fast track, economical and efficient construction, prestressed
slabs are becoming popular. The slabs are presented in two groups: one-way slabs
and two-way slabs. The two-way slabs are presented in details in Sections 9.3 and
9.4.
Rectangular slabs can be divided into the two groups based on the support conditions
and length-to-breadth ratios. The one-way slabs are identified as follows.
1) When a rectangular slab is supported only on two opposite edges, it is a one-way
slab spanning in the direction perpendicular to the edges. Precast planks fall in
this group.
2) When a rectangular slab is supported on all the four edges and the length-to-
breadth (L / B) ratio is equal to or greater than two, the slab is considered to be a
one-way slab. The slab spans predominantly in the direction parallel to the
shorter edge.
The following sketches show the plans of the two cases of one-way slabs. The
spanning direction in each case is shown by the double headed arrow.
L L
(a) (b)
(a) Supported on two opposite edges (b) Supported on all edges (L/B > 2)
Figure 9-2.1 Plans of one-way slabs

A slab in a framed building can be a one-way slab depending upon its length-to-breadth
ratio. A one-way slab is designed for the spanning direction only. For the transverse
direction, a minimum amount of reinforcement is provided. A hollow core slab is also an
example of a one-way slab. A ribbed floor (slab with joists) made of precast double tee
sections, is analysed as a flanged section for one-way bending.
Other types of rectangular slabs and non-rectangular slabs are considered to be two-
way slabs. If a rectangular slab is supported on all the four sides and the length-to-
breadth ratio is less than two, then it is a two-way slab. If a slab is supported on three
edges or two adjacent edges, then also it is a two-way slab. A slab in a framed building
can be a two-way slab depending upon its length-to-breadth ratio. A two-way slab is
designed for both the orthogonal directions.
A slab is prestressed for the following benefits.

1) Increased span-to-depth ratio
Typical values of span-to-depth ratios in slabs are given below.
Non-prestressed slab 28:1
Prestressed slab 45:1
2) Reduction in self-weight
Increased durability
4) Quick release of formwork
Fast construction
5) Reduction in fabrication of reinforcement
6) More flexibility in accommodating late design changes.
Precast planks are usually pre-tensioned. Cast-in-situ slabs are post-tensioned. Post-
tensioned slabs are becoming popular in office and commercial buildings and parking
structures, where large column-free spaces are desirable. The maximum length of a
post-tensioned slab is limited to 30 to 40 m to minimise the losses due to elastic
shortening and friction.
Slabs can be composite for the benefits of reduction in form work, cost and time of
construction and quality control. A precast plank can be prestressed and placed in the
final location. A topping slab is overlaid on the precast plank. The grades of concrete in

the two portions can be different. The following sketches show the sections of some
one-way slabs.
Cast-in-situ topping
Precast and prestressed

plank
Figure 9-2.2 Cross-section of a composite slab
9.2.2 Analysis and Design
One-way slabs are analysed and designed for the spanning direction similar to
rectangular beams. The analysis and design is carried out for the width of the plank or
a unit width (say 1 m) of the slab. For continuous slabs, the moment coefficients of
IS:456 - 2000 (Table 12) can be used.
The analysis and design procedures for simply supported rectangular beams are
covered in Sections 3.2 to 3.6 and Sections 4.2 to 4.6, respectively. These materials are
briefly reproduced here.
Preliminary Design
1) Select the material properties fck and fpk. Here, fck is the characteristic
compressive strength of concrete and fpk is the characteristic tensile strength of
prestressing steel.
2) Determine the total depth of slab (h), based on the span to effective depth ratio
(L / d), given in Clause 22.6 of IS:1343 - 1980. Consider d h 25 mm. Round
off h to a multiple of 10 mm.
3) Calculate the self weight.
4) Calculate the total moment (MT) including moment due to self weight (Msw).
5) Estimate the lever arm (z).
z = 0.65 h if Msw is large (say Msw > 0.3 MT)
z = 0.5 h if Msw is small.
6) Estimate the effective prestress (Pe).
Pe = MT /z if Msw is large
Pe = MI L/z if Msw is small.
Here, the moment due to imposed loads is given as MI L = MT Msw.
7) Considering fpe = 0.7 fpk , calculate area of prestressing steel Ap = Pe / fpe .

8) Check the area of cross section (A) A = 1000 mm h mm. The average stress
C/A should not be too high as compared to 50% fcc,all .
Final Design
The final design involves the checking of the stresses in concrete at transfer and under
service loads with respect to the allowable stresses. The allowable stresses depend on
the type of slab (Type 1, Type 2 or Type 3). Here, the steps of final design are
explained for Type 1 slabs only. For Type 1 slabs, no tensile stress is allowed at
transfer or under service loads.
For small moment due to self-weight (Msw 0.3 MT), the steps are as follows.
1) Calculate eccentricity (e) to locate the centroid of the prestressing steel (CGS).
The lowest permissible location of the compression (C) due to self-weight is at the
bottom kern point (at a depth kb below CGC) to avoid tensile stress at the top. The
design procedure based on the extreme location of C gives an economical section. For
this location of C, the following equation can be derived.
Msw (9-2.1)
e= + kb
P0
The magnitude of C or T is equal to P0, the prestress at transfer after initial losses.
The value of P0 can be estimated as follows.

a) P0 = 0.9 Pi for pre-tensioned slab
b) P0 = Pi for post-tensioned slab
Here, Pi is the initial applied prestress.

Pi = (0.8fpk) Ap (9-2.2)
The permissible prestress in the tendon is 0.8fpk.
2) Re-compute the effective prestress Pe and the area of prestressing steel Ap.
For the extreme top location of C under service load, the shift of C due to the total
moment gives an expression of Pe.

MT
Pe = (9-2.3)
e+ kT
For solid rectangular slab, kb = kt = h / 6.
Considering fpe = 0.7fpk , the area of prestressing steel is recomputed as follows.

Ap = Pe / fpe (9-2.4)
The number of tendons and their spacing is determined based on Ap. The value of P0 is
updated.
3) Re-compute e with the updated values of Ap and P0.

For large Msw if e violates the cover requirements, e is determined based on cover.
4) Check the compressive stresses in concrete

For the limiting no tension design at transfer, the stress at the bottom (fb) is given as
follows.
P0 h 2P
fb = - =- 0 (9-2.5)
A ct A
The stress should be less than fcc,all , where fcc,all is the allowable compressive stress in
concrete at transfer. The condition to satisfy can thus be written as | fb | fcc,all.
For the limiting no tension design at service, the stress at the top (ft) is given as follows.
P h 2P
f =- e =- e (9-2.6)
t A cb A
The stress should be less than fcc,all , where fcc,all is the allowable compressive stress in
concrete at service. The condition to satisfy can be written as | ft | fcc,all.
For Type 2 and Type 3 slabs, the tensile stress should be restricted to the allowable
values. For a continuous slab, a suitable profile of the tendons is selected similar to that

in continuous beams. The design of continuous beams is covered in Sections 8.2 and
8.3.
When the value of e is fixed (in either pre-tension or post-tension operations), the
design steps are simpler. If the tendons are placed at the CGC (e = 0), then the uniform
compressive stress due to prestress counteracts the tensile stress due to service loads.
To have zero stress at the bottom under service conditions, the value of Pe can be
directly calculated from the following equation.
Pe MT
=
A Zb
M
or, Pe = A T (9-2.7)
Zb
Zb is the section modulus. The above expression is same as Pe = MT / kt, with e = 0.

The stresses at transfer can be checked with an estimate of P0 from Pe.
5) Checking for shear capacity

The shear is analogous to that generates in a beam due to flexure. The calculations can
be for unit width of the slab. The critical section for checking the shear capacity is at a
distance effective depth d from the face of the beam, across the entire width of the
slab. The critical section is transverse to the spanning direction. The shear demand
(Vu) in the critical section generates from the gravity loads in the tributary area.
d
Tributary area
Critical section
Figure 9-2.3 Tributary area and critical section for shear
For adequate shear capacity, VuR Vu where, VuR = Vc, the shear capacity of uncracked
concrete of unit width of slab. The expression of Vc is given in Section 5.2, Design for
Shear (Part I). If this is not satisfied, it is preferred to increase the depth of the slab to
avoid shear reinforcement.

6) Provide transverse reinforcement based on temperature and shrinkage.

As per IS:456 - 2000, Clause 26.5.2.1, the minimum amount of transverse
reinforcement (Ast,min in mm2) for unit width of slab is given as follows.
Ast,min = 0.15% 1000 h for Fe 250 grade of steel

= 0.12% 1000 h for Fe 415 grade of steel.
Usually the transverse reinforcement is provided by non-prestressed reinforcement.

The minimum reinforcement is sufficient for the transverse moment due to Poissons
effect and small point loads. For a heavy point load, transverse reinforcement needs to
be computed explicitly.
The following example shows the design of a simply supported precast prestressed
composite slab.
Reference:
Santhakumar, A.R., Partially Precast Composite PSC Slab, Published by Building
Technology Centre, Anna University, Chennai.
Example 9-2.1
Design a simply supported precast prestressed (Type 1) composite slab for the
following data.
Width of the slab = 0.3 m
Clear span = 2.9 m
Effective span (L) = 3.1 m
Thickness of the precast plank = 50 mm
Thickness of the cast-in-situ topping slab = 50 mm
Grade of concrete in precast plank : M60
Grade of concrete in topping slab : M15
The pre-tensioned tendons are located at mid depth of the precast slab. During
the casting of the topping, planks are not propped.

Live load = 2.0 kN/m2

Load due to floor finish = 1.5 kN/m2.
Solution
1) Calculation of moments.
Load per unit area
Weight of precast plank = 1.25 kN/m2
Weight of topping slab = 1.25 kN/m2
Weight of floor finish = 1.50 kN/m2
Live load = 2.00 kN/m2
Total = 6.00 kN/m2
Total moment (MT) along the width of the slab is given as follows.
wBL2 6 0.3 3.12

=
8 8
= 2.16 kNm
The individual moments are calculated based on the proportionality of the loads.
MSW = moment due to self weight of precast plank
= 2.16 (1.25 / 6.00) = 0.45 kNm
Mtop = moment due to weight of topping slab

= 2.16 (1.25 / 6.00) = 0.45 kNm
Mfin = moment due to weight of floor finish

= 2.16 (1.50 / 6.00) = 0.54 kNm
MLL = moment due to live load

= 2.16 (2.00 / 6.00) = 0.72 kNm.
2) Calculation of geometric properties.

Precast section

Area
A1 = 300 50 = 15,000 mm2
Moment of inertia
50mm
300mm
1
I1 = 300 503
12
= 3,125,000 mm4
Distance to the extreme fibres
50
c b = ct =
2
= 25 mm
Section moduli
3,125,000
Zb = Zt =
25
= 125,000 mm3
Composite section
Since the grades of concrete are different for the precast- prestressed (PP) and cast-in-
situ (CIS) portions, an equivalent (transformed) area is calculated. The CIS portion is
assigned a reduced width based on the equivalent area factor (modular ratio).
Equivalent area factor = Modulus of CIS / Modulus of PP

= (Grade of CIS / Grade of PP)
=15/60
= 0.5

50mm
Composite section
50mm
150mm
Equivalent section
300mm
Location of CGC from bottom
Atop = 50 150 = 7,500 mm2
Abot = 50 300 = 15,000 mm2
A2 = Atop + Abot = 22,500 mm2
Atop 75 + Abot 25
y=
A
937,500
=
22,500
= 41.7 mm
CGC 58.3 mm
41.7 mm
Moment of inertia
1
Itop = (0.5300)503 + 7500(75 - 41.7)2
12
= 9,894,166.8 mm4
1
Ibot = 300503 +15000(41.7 - 25)2
12
= 7,293,333.5 mm4
I = 9,894,166.8 + 7,293,333.5
= 17,187,500
= 1.719 107 mm4

Distance to the extreme fibres

yb = 41.7 mm
yt = 58.3 mm
Section moduli
Zb = 17.19 106 /41.7

= 412,527 mm3
Zt = 17.19 106 /58.3
= 294,703 mm3
3) Calculation of prestress
The tendons are located at the mid depth of the precast plank. Hence, e = 0 for the
precast plank. The value of Pe is calculated directly from the following stress profiles.
+ +
Section Pe MSW + Mtop Mfin + MLL
Stress profiles
To avoid tensile stress at the bottom under service conditions, the resultant stress is
equated to zero.
Pe MSW + Mtop Mfin + MLL

- + + =0
A1 Z1b Z2b
M + Mtop Mfin + MLL
or, Pe = A1 SW +
Z1b Z2 b
In the above expression, the first term inside the bracket corresponds to the precast
section. The moments due to self weight and topping slab are resisted by the precast
section alone.
The second term inside the bracket corresponds to the equivalent section. The
moments due to weight of the floor finish and live load are resisted by the equivalent
section.
0.45 + 0.45 0.54 + 0.72
P = A + 106
125,000 412,527
= 50300 (7.2 + 3.0)
= 153,816 N

Assuming around 20% loss,

The prestress at transfer (P0) = 1.2 153,816
= 184,579 N.
Wires of diameter = 7 mm and ultimate strength (fpk) = 1500 MPa are selected for
prestressing.
Area of one wire (Ap) = 38.48 mm2.

The maximum allowable tension in one wire
= 0.8 fpk Ap
= 0.8 1500 38.48
= 46,176 N.
No. wires required = 184,579 / 46,176

= 3.99
4.
Required pull in each wire = 184,579 / 4

= 46,145 N.
Total prestressing force (P0) = 4 46,145

= 184,580 N.
Effective prestressing force (Pe) = 0.8 184,580

= 147,664 N.
4) Checking of stresses in concrete
a) At transfer
The compressive strength at 7 days (fci ) = 0.7 fck
= 0.7 60
= 42 MPa.
Allowable compressive stress (fcc,all) = 0.44 fci

= 0.44 42
= 18.5 MPa.
For Type 1 members, the allowable tensile stress (fct,all) is zero.
Stresses at the mid-span of the precast portion

fc = P0/A1 MSW/Z1
= 12.3 (0.45 106 / 125,000)
15.9
ft = 15.9 MPa
fb = 8.7 MPa
8.7
|ft| fcc,all OK
b) After casting of topping slab at 28 days
Allowable compressive stress (fcc,all) = 0.44 fck

= 0.44 60
= 26.4 MPa.
The allowable tensile stress (fct,all) is zero.
Stresses at the mid-span of the precast portion

fc = P0/A1 (MSW + Mtop)/Z1
= 12.31 ((0.45 + 0.45) 106 / 125,000) 19.5
ft = 19.5 MPa
fb = 5.1 MPa
|ft| fcc,all OK 5.1

c) At service
i) For the precast portion

= 0.35 60
= 21 MPa.
The allowable tensile stress (fct,all) is zero.
Stresses at the mid-span of the composite section for unpropped construction

fc = Pe /A1 (MSW + Mtop) / Z1 (Mfin + MLL) / Z2
fjunc = (147,664 / 15,000) ((0.45 + 0.45) 106 / 125,000)

((0.54 + 0.72) 106 / 2,063,625)
= 17.6 MPa
fb = (147,664 / 15,000) + ((0.45 + 0.45) 106 / 125,000)

+ ((0.54 + 0.72) 106 / 412,527)
= 0.4 MPa
0
|fjung| fcc,all OK
fb = fct,all OK
ii) For cast-in-situ portion

= 0.35 15 4.3
= 5.2 MPa.
Stresses at the mid-span of the composite section

0.4
ft = (0.54 + 0.72) 106 / 294,703)
17.6
= 4.3 MPa
fjunc = (0.54 + 0.72) 106 / 2,063,625)
Indian Institute of Technology Madras 0.4

= 0.6 MPa
ft fcc,all OK
Note that the critical stress at the junction is in the precast portion.
5) Check for shear

VuR = Vc
= Vc0
= 0.67bh(ft2 + 0.8fcpft)
= 0.67 300 50 (1.862 + 0.8 9.36 1.86)
= 41.9 kN
Vu = wuB L / 2
= 1.5 6 0.3 3.1 / 2 = 4.2 kN
VuR Vu
Therefore, the shear capacity is adequate.
6) Transverse reinforcement
Using Fe 415 grade of steel, for 1m width

Ast,min = 0.12% 1000 h
= 0.0012 1000 100
= 120 mm2.
Provide 8 mm diameter bars at 300 mm on centre.
7) Provide nominal reinforcement for shrinkage in the longitudinal direction of the

topping slab.
Using Fe 415 grade of steel, for 1m width

Ast,min = 0.12% 1000 h

= 0.0012 1000 50
= 60 mm2.
Provide 6 mm diameter bars at 300 mm on centre.

Reinforcement details
(4) 7 mm wires
(a) Plan of precast plank 8 mm rebar

6 mm rebar
@ 300 mm c/c @ 300 mm c/c
(b) Plan of topping slab
(c) Longitudinal Section of composite slab

: diameter

9.3 Two-way Slabs (Part I)

Introduction
Analysis and Design
Features in Modeling and Analysis
Distribution of Moments to Strips
9.3.1 Introduction
The slabs are presented in two groups: one-way slabs and two-way slabs. The one-
way slabs are presented in Section 9.2. When a rectangular slab is supported on all the
sides and the length-to-breadth ratio is less than two, it is considered to be a two-way
slab. The slab spans in both the orthogonal directions. A circular slab is a two-way
slab. In general, a slab which is not falling in the category of one-way slab, is
considered to be a two-way slab.
Rectangular two-way slabs can be divided into the following types.

1) Flat plates: These slabs do not have beams between the columns, drop panels
or column capitals. Usually, there are spandrel beams at the edges.
2) Flat slabs: These slabs do not have beams but have drop panels or column
capitals.
3) Two-way slabs with beams: There are beams between the columns. If the
beams are wide and shallow, they are termed as band beams.
For long span construction, there are ribs in both the spanning directions of the slab.
This type of slabs is called waffle slabs.
The slabs can be cast-in-situ (cast-in-place). Else, the slabs can be precast at ground
level and lifted to the final height. The later type of slabs is called lift slabs. A slab in a
framed building can be a two-way slab depending upon its length-to-breadth (L / B)
ratio. Two-way slabs are also present as mat (raft) foundation.
The following sketches show the plan of various cases of two-way slabs. The spanning
directions in each case are shown by the double headed arrows.

a) Flat plate b) Flat slab
L
c) Two-way slab d) Waffle slab
with beams
Figure 9-3.1 Plans of two-way slabs
The absence of beams in flat plates and flat slabs lead to the following advantages.
1) Formwork is simpler
2) Reduced obstruction to service conduits
3) More flexibility in interior layout and future refurbishment.
Two-way slabs can be post-tensioned. The main advantage of prestressing a slab is

the increased span-to-depth ratio. As per ACI 318-02 (Building Code Requirements for
Structural Concrete, American Concrete Institute), the limits of span-to-depth ratios are
as follows.
For floors 42
For roofs 48.
The values can be increased to 48 and 52, respectively, if the deflection, camber and
vibration are not objectionable. The following photographs show post-tensioned flat
plate and flat slab.

(a) Flat plate
(b) Flat slab

Figure 9-3.2 Post-tensioned two-way slabs
(Courtesy: VSL India Pvt. Ltd.)
9.3.2 Analysis and Design
Analysis
The analysis of two-way slabs is given in Section 31, IS:456 - 2000, under Flat Slabs.
The analysis is applicable to flat plates, flat slabs and two-way slabs with deflecting

beams. For two-way slabs with beams, if the beams are sufficiently stiff, then the
method (based on moment coefficients) given in Annex D, IS:456 2000, is applicable.
The direct design method of analysing a two-way slab is not recommended for
prestressed slabs. The equivalent frame method is recommended by ACI 318-02. It is
given in Subsection 31.5, IS:456 - 2000. This method is briefly covered in this section
for flat plates and flat slabs.
The slab system is represented by a series of two dimensional equivalent frames for
each spanning direction. An equivalent frame along a column line is a slice of the
building bound by the centre-lines of the bays adjacent to the column line.
The width of the equivalent frame is divided into a column strip and two middle strips.
The column strip (CS) is the central half of the equivalent frame. Each middle strip (MS)
consists of the remaining portions of two adjacent equivalent frames. The following
figure shows the division in to strips along one direction. The direction under
investigation is shown by the double headed arrow.
1 2 3
MS CS MS
l1
l2
Figure 9-3.3 Equivalent frame along Column Line 2
In the above figure,

l1 = span of the equivalent frame in a bay
l2 = width of the equivalent frame. This is the tributary width for calculating
the loads.
The following figure shows a typical elevation of an equivalent frame.

Figure 9-3.4 Elevation of an equivalent frame
The analysis is done for each typical equivalent frame. An equivalent frame is modelled
by slab-beam members and equivalent columns. The equivalent frame is analysed for
gravity load and lateral load (if required), by computer or simplified hand calculations.
Next, the negative and positive moments at the critical sections of the slab-beam
members are distributed along the transverse direction. This provides the design
moments per unit width of a slab.
If the analysis is restricted to gravity loads, each floor of the equivalent frame can be
analysed separately with the columns assumed to be fixed at their remote ends, as
shown in the following figure. The pattern loading is applied to calculate the moments
for the critical load cases. This is discussed later.
Figure 9-3.5 Simplified model of an equivalent frame
The steps of analysis of a two-way slab are as follows.

1) Determine the factored negative (Mu) and positive moment (Mu+) demands at the
critical sections in a slab-beam member from the analysis of an equivalent frame.
The values of Mu are calculated at the faces of the columns. The values of Mu+

are calculated at the spans. The following sketch shows a typical moment
diagram in a level of an equivalent frame due to gravity loads.
Mu
Mu +
Figure 9-3.6 Typical moment diagram due to gravity loads
2) Distribute Mu to the CS and the MS. These components are represented as Mu,
CS and Mu,MS, respectively. Distribute Mu+ to the CS and the MS. These
components are represented as Mu,+CS and Mu,+MS, respectively.
Mu Mu
Mu +
MShalf
Mu+MS
CS MuCS MuCS
Mu+CS
MShalf
MuMS
Figure 9-3.7 Distribution of moments to column strip and middle strips
3) If there is a beam in the column line in the spanning direction, distribute each of
Mu,CS and Mu,+CS between the beam and rest of the CS.
Mu Mu
Mu+
MShalf MuMS
Mu
MS Mu+MS
CS Mu
CS MuCS
Mu+CS Beam
MShalf width
Figure 9-3.8 Distribution of moments to beam, column strip and middle strips

4) Add the moments Mu,MS and Mu,+MS for the two portions of the MS (from adjacent
equivalent frames).
5) Calculate the design moments per unit width of the CS and MS.
Design
Once the design moments per unit width of the CS and MS are known, the steps of
design for prestressing steel are same as that for one-way slab. The profile of the
tendons is selected similar to that for continuous beams. The flexural capacity of
prestressed slab is controlled by total amount of prestressing steel and prestress rather
than by tendon distribution. But the tendon distribution effects the load balancing.
Some examples of tendon distribution are shown.
Normal reinforcement
100 % banded through 100 % banded in one direction

columns in both directions and distributed in other
direction
CS MS CS
75 % in column strip and 75 % in CS and 25 % in MS

25 % in middle strip in in one direction and
both directions distributed in other direction
Figure 9-3.9 Typical tendon layouts

Maximum spacing of tendons or groups of tendons should be limited to 8h or 1.5 m,

whichever is less. Here, h is the thickness of the slab. A minimum of two tendons shall
be provided in each direction through the critical section for punching shear around a
column. The critical section for punching shear is described in Section 9.4, Two-way
Slabs (Part II). Grouping of tendons is permitted in band beams.
A minimum amount of non-prestressed reinforcement is provided in each direction

based on temperature and shrinkage requirement. As per IS:456 - 2000, Clause
26.5.2.1, the minimum amount of reinforcement (Ast,min in mm2) for unit width of slab is
given as follows.
Ast,min = 0.15% 1000h for Fe 250 grade of steel

= 0.12% 1000h for Fe 415 grade of steel.
The ducts for placing the individual strands are oval shaped to maintain the eccentricity,
reduce frictional losses and convenient placement of crossing ducts. The ducts are not
commonly grouted as the use of unbonded tendon is not detrimental in buildings.
The following photo shows the ducts for the prestressing tendons and the non-
prestressed reinforcement in a two-way slab.
Non-prestressed reinforcement
Oval shaped
duct for
prestressing
tendon
Figure 9-3.10 Reinforcement in a two-way slab


9.3.3 Features in Modelling and Analysis
The features in modelling and analysing an equivalent frame are discussed.
Gross section versus cracked section

For determining the stiffness of a slab-beam member, the gross section can be
considered in place of the cracked section. This is to simplify the calculation of moment
of inertia of the section.
Equivalent column
The actual column needs to be replaced by an equivalent column to consider the
flexibility of the transverse beam in the rotation of the slab. The portions of the slab in
the MS rotate more than the portions in the CS because of the torsional deformation of
the transverse beam. In the following figure, the size of the arrows qualitatively
represents the rotation of the slab. Note that the rotation is higher away from the
column.
c2 c1
Upper column
l2
h
Transverse beam Lower column

Figure 9-3.11 Isometric view of a slab-column junction
The transverse beam need not be a visible beam, but a part of the slab in the
transverse direction, bounded by the edges of the column or column capital. In
presence of beam or column capital or in absence of beam, the cross-section of the
modelled transverse beam is taken as shown in the following sketches.

a) In presence of b) In presence of c) In absence of

transverse beam column capital transverse beam
Figure 9-3.12 Cross-section of modelled transverse beam
The flexibility of the equivalent column is equal to the sum of the flexibilities of the actual
column and the transverse beam.
1 1 1
= + (9-3.1)
K ec K c Kt
Here,
Kec = flexural stiffness of the equivalent column
Kc = Kc,upper + Kc,lower
Kc,upper = flexural stiffness of the upper column
Kc,lower = flexural stiffness of the lower column
Kt = torsional stiffness of the transverse beam.
An approximate expression for the flexural stiffness of a column (Kc) is given below.
4Ec Ic (9-3.2)
Kc =
L - 2h
Here,
Ec = modulus of concrete
L = length of the column
h = thickness of the slab
Ic = moment of inertia of the column.
An approximate expression for torsional stiffness of the transverse beam (Kt) is given
below.
9EcC
Kt =
c (9-3.3)
l 2 1- 2
l2

Here,
C = equivalent polar moment of inertia of transverse beam
c2 = dimension of column in the transverse direction
l2 = width of equivalent frame.
For a rectangular section, the expression of C is given below.
x x3 y
C = 1- 0.63 (9-3.4)
y 3
Here, x and y are the smaller and larger dimensions of the transverse beam. The
expression of C is a lower bound estimate, that is, the calculated value is always lower
than the actual moment of inertia of the transverse beam. For a transverse beam of
compound section, the value of C is the summation of the individual values of the
component rectangles. The splitting of the compound section into component
rectangles should be such, so as to maximise the value of C. For the following two
cases of splitting, select the larger value of C.
Figure 9-3.13 Component rectangles of a compound section
If there is a beam in the column strip in the spanning direction, then Kt is replaced by Kt
(Isb / Is).
Here,
Is = moment of inertia of slab without the projecting portion of the beam
(shaded area in Sketch (a) of the following figure)
Isb = moment of inertia of slab considering it as a T-section
(shaded area in Sketch (b) of the following figure).

l2 h
(a) (b)
Figure 9-3.14 Sections for determining moments of inertia
Slabbeam members
The variation of the flexural moment of inertia of a slab-beam member is considered as
follows.
The value of the moment of inertia (I) is constant (say equal to I1) in the prismatic
portion, that is, in between the faces of the columns or column capitals or drop panels.
It is also constant, with a different value (say equal to I2) in the region of a drop panel.
The value varies in the region from the face of the column or column capital to the
center line of the column. But it is approximated to a constant value equal to the
following.
I2
2
c2 (9-3.5)
1- l
2
Here,
I2 = moment of inertia at the face of the column or column capital
c2 = dimension of column in the transverse direction
l2 = width of equivalent frame.
The following figure shows the variation of the moment of inertia of the slab beam
member.

(a)
I2
I2
2
c2 I1 (b)
1- l
2
Figure 9-3.15 a) Elevation of equivalent frame, b) Variation of the moment of inertia of

the slab-beam member
Arrangement of live load

Since the factored live load (wu, LL) may not occur uniformly in all the spans in a floor, a
distribution is considered to generate the maximum values of the negative (Mu) and
positive moments (Mu+) at the critical sections. If the distribution of wu, LL is known, then
the load is applied accordingly. If the distribution is not known, then a pattern loading
is considered based on the value of wu, LL with respect to that of the factored dead load
(wu, DL). Of course, the load case with wu, LL on all the spans should be also analysed.
1) For wu, LL wu, DL

The possible variation in wu,LL in the different spans is neglected. wu,LL is applied
uniformly on all the spans.
wu,LL.l2
wu,DL.l2
A B C D E
Figure 9-3.16 Distribution of live load for wu, LL wu, DL
2) For wu LL > wu,DL

For maximum value of Mu+ in a span, wu,LL is applied on the span and the alternate
spans. For example, if the maximum value of Mu+ in Span BC of the frame below is to
be determined, then wu,LL is placed in Spans BC and DE. This distribution will also
give the maximum value of Mu+ in Span DE. For maximum value of Mu near the

support, wu LL is applied on the adjacent spans only. For example, if the maximum
value of Mu near Support B is to be determined, then wu,LL is placed in Spans AB
and BC.
wu,LL.l2
wu,DL.l2
A B C D E
Distribution of live load for maximum Mu+ in Spans BC and DE
wu,LL.l2
wu,DL.l2
A B C D E
Distribution of live load for maximum Mu- near Support B

Figure 9-3.17 Distribution of live load for wu LL > wu,DL
Critical section near a support

The critical section is determined as follows.
1) At interior support
At the face of support (column or column capital, if any), but not further than 0.175l1
from the center line of the column.
2) At exterior support
At a distance from the face of column not greater than half the projection of the column
capital (if any).
9.3.4 Distribution of Moments to Strips
In absence of a rigorous analysis (say finite element analysis), the procedure for
reinforced concrete slabs may be used to distribute the moments Mu+ and Mu to the
column strip (CS) and middle strip (MS). Of course, ACI 318-02 does not recommend
this procedure to be used for prestressed slabs.

Distribution of Mu at interior support

Mu,CS = 0.75 Mu (9-3.6)
Mu,MS = 0.25 Mu (9-3.7)
Here,
Mu,CS = negative moment in the CS
Mu,MS = total negative moment in the two MS at the sides.
Distribution of Mu at exterior support

If the width of the column or wall support is less than l2,
Mu,CS = Mu (9-3.8)
Mu,MS = 0. (9-3.9)
If the width of the column or wall support is greater than l2, then Mu is uniformly
distributed along the width l2.
Distribution of Mu+ at mid span

Mu,+CS = 0.60 Mu+ (9-3.10)
Mu,+MS = 0.40 Mu+ (9-3.11)
Here,
Mu,+CS = positive moment in the CS
Mu,+MS = total positive moment in the two MS at the sides.
The total moments in MS (Mu,MS and Mu,+MS) are distributed to the two middle strips at
the sides of the equivalent frame, proportional to their widths. The combined MS from
two adjacent equivalent frames is designed for the sum of the moments assigned to its
parts.

9.4 Two-way Slabs (Part II)

Checking for Shear Capacity
Spandrel Beams
Anchorage Devices
Additional Aspects
9.4.1 Checking for Shear Capacity
The checking of shear capacity of flat plates and flat slabs is of utmost importance. In
absence of beams, the shear is resisted by the slab near the slab-column junction.
The shear capacity of a slab should be adequate to resist the shear from two actions.
1) One-way (beam) shear
2) Two-way (punching) shear.
One-way shear
The one-way shear is analogous to that generates in a beam due to flexure. This is
checked in a two-way slab for each spanning direction separately.
The critical section for checking the shear capacity is at a distance effective depth d
from the face of the column, across the entire width of the frame. The critical section is
transverse to the spanning direction. For gravity loads, the shear demand in the critical
section generates from the loads in the tributary area shown in the next figure. For
lateral loads, the shear demand is calculated from the analysis of the equivalent frame.
In presence of a drop panel two critical sections need to be checked. The first section is
at a distance d1 from the face of the column, where d1 is the effective depth of the drop
panel. The second section is at a distance d2 from the face of the drop panel, where d2
is the effective depth of the slab.

CL
Critical section
Plan l2
d
ln
Elevation
(a) Slab without drop panel

CL
Critical sections
Plan l2
d1 ln
d2
Elevation
(b) Slab with drop panel

Figure 9-4.1 Critical sections for one-way shear
The calculations for shear can be for unit width of the slab. The shear demand due to
gravity loads per unit width is given as follows.
Vu = wu (0.5ln d) (9-4.1)
Here, ln is the clear span along the spanning direction.

The shear capacity per unit width is given as follows.
VuR = Vc (9-4.2)

Vc is the shear capacity of uncracked concrete of unit width of slab. The expression of
Vc is given in Section 5.2, Design for Shear (Part I).
For adequate shear capacity

VuR Vu (9-4.3)
If this is not satisfied, it is preferred to increase the depth of the slab to avoid shear
reinforcement along the width of the slab.
Two-way shear
The two-way shear is specific to two-way slabs. If the capacity is inadequate, the slab
may fail due to punching around a column. The punching occurs along a conical
frustum, whose base is geometrically similar and concentric to the column cross-
section. The following figure illustrates the punching shear failure.
Elevation Isometric view

Figure 9-4.2 Punching shear failure
Two-way shear is checked for the two spanning directions simultaneously. The critical
section for checking the shear capacity is geometrically similar to the column perimeter
and is at a distance d / 2 from the face of the column. The depth of the critical section is
equal to the average of the effective depths of the slab in the two directions. The
sketches below show the critical section. The tributary area of the column is the area
within the centre-lines of the spans minus the area within the critical section. It is shown
shaded in the third sketch.

b1
Critical section
b2 c2
c1
(a) Plana
b2
d
b1 r
(b) Isometric view
CL
CL
Critical section
CL
CL
(c) Tributary area for shear due to gravity loads

Figure 9-4.3 Critical section for two-way shear
The lengths of the sides of the critical section along axes 1-1 and 2-2 are denoted as b1
and b2, respectively.
b1 = c1 + d (9-4.4)
b2 = c2 + d (9-4.5)
Here,
c1 = dimension of the column or column capital along axis 1-1
c2 = dimension of the column or column capital along axis 2-2.
For a non-rectangular column, the critical section consists of the slab edges as per
Figure 13, IS:456 - 2000. For edge and corner columns, the critical section consists of
the slab edges as per Figure 14, IS:456 - 2000.

The demand in terms of shear stress is given as follows.
Muv b1 M b2
V 2-2 2 uv 1-1
2
v = u + + (9-4.6)
b0d J 2-2 J 1-1
Here,
Vu = shear due to gravity loads from the tributary area
Muv = fraction of moment transferred about an axis
b0 = perimeter of the critical section = 2(b1 + b2).
J = polar moment of inertia of the critical section about an axis
The second and third terms are due to transfer of moments from slab to column. The
moment about an axis is due to the unbalanced gravity loads for the two sides of the
column or due to lateral loads. It is transferred partly by the variation of shear stress in
the critical section and the rest by flexure. The fraction transferred by the variation of
shear stress about an axis is denoted as Muv.
Muv|2-2 = Fraction of moment transferred about axis 2-2

Muv|1-1 = Fraction of moment transferred about axis 1-1
The forces and stresses acting at the critical section are shown below.
2
1 1
2
Vu Muv|2-2 Muv|1-1
(a) Shear and moments acting at the critical section
1 1
2
Due to Vu Due to Muv|2-2 Due to Muv|1-1
(b) Stresses acting at the critical section

Due to Vu + Muv|2-2 + Muv|1-1

(c) Resultant shear stress diagram at the critical section
Figure 9-4.4 Forces and stresses at the critical section
The fraction of moment transferred by the variation of shear stress about an axis (Muv),
is given in terms of the total moment transferred (Mu) as follows.
Muv = (1- )Mu (9-4.7)
The value of Mu due to unbalanced gravity loads is calculated by placing live load on
one side of the column only. The value of Mu due to lateral loads is available from the
analysis of the equivalent frame. The parameter is based on the aspect ratio of the
critical section.
1
= (9-4.8)
2 b1
1+
3 b2
The polar moments of inertia of the critical section, about the axes are given as follows.
1 3 1 3 b2
2
(9-4.9)
J 1-1 = 2 b2d + db2 + b1d
12 12 2
1 3 1 3 b1
2
J 2-2 = 2 b1d + db1 + b2d (9-4.10)
12 12 2
For adequate shear capacity
v ks c (9-4.11)
The shear stress capacity of concrete for a square column is given as follows.
(9-4.12)
c = 0.25 fck

Here, fck is the characteristic strength of the concrete in the slab. The effect of prestress
is neglected. The factor ks accounts for the reduced shear capacity of non-square
columns.
ks = 0.5 + c (9-4.13)
The value of ks should be less than 1.0. c is a parameter based on the aspect ratio of
the column cross-section. It is the ratio of the short side to long side of the column or
column capital.
If v exceeds ksc, a drop panel or shear reinforcement needs to be provided at the slab-
to-column junction. The shear reinforcement can be in the form of stirrups or I section
(shear head) or based on shear studs. The reinforcement based on shear studs
reduces congestion for conduits and post-tensioning tendons. If v exceeds 1.5c, then
the depth of the slab needs to be increased in the form of drop panels.
The stirrups are designed based on the following equation. Asv is the area of the
vertical legs of stirrups.
Asv =
( v - 0.5 c )
(9-4.14)
0.87fy
The stirrups are provided along the perimeter of the critical section. The first row of
stirrups should be within a distance of 0.5d from the face of the column. They can be
continued in outer rows (concentric and geometrically similar to the critical section) at an
interval of 0.75d, till the section with shear stress v = 0.5 .
The different types of reinforcement at the slab-to-column junction are shown in the
following sketches.
References:
1. Bureau of Indian Standards,
Handbook on Concrete Reinforcement and Detailing
(SP 34 : 1987)

2. Khan, S. and Williams, M.

Post-tensioned Concrete Floors
Butterworth-Heinemann Ltd.
0.75d < 0. 5d
Elevation of Section A-A
A A
Plan at mid-depth of the slab

(a) Stirrups in geometrically similar rows
Elevation of Section A-A
A A

(b) Beam cage stirrups

0.5d
~ 2d Elevation of Section A-A
A A

(c) Beam cage stirrups with bent-up bars
(d) Shear head reinforcement
(e) Shear stud reinforcement

Figure 9-4.5 Shear reinforcement at slab-column junction

The following photo shows the ducts and reinforcement at the slab-column junction in a
slab with a drop panel.
Figure 9-3.6 Reinforcement at slab-column junction

The residual moment transferred by flexure (Muf), is given in terms of the total moment
transferred (Mu) as follows.
Muf = Mu (9-4.15)
Additional non-prestressed reinforcement is provided at the top of the slab over a width
c2 + 3h (centred with respect to the column) to transfer Muf.
9.4.2 Spandrel Beams
The flat plates are provided with spandrel beams at the edges. These beams stiffen the
edges against rotation. In turn the beams are subjected to torsion.
The maximum torsion is calculated by assuming a uniform torsional loading along the
width of the equivalent frame (ACI 318-02 recommends a triangular distribution). The
spandrel beams are provided with closed stirrups to resist the torsion. The design for
torsion is given in the Module of Analysis and Design for Shear and Torsion.

The following figure shows the distribution of the torsional loading on the spandrel
beam.
Variation of
torsion
l2
Tu,max
Figure 9-4.7 Torsion in spandrel beams
The maximum torsion (Tu,max) is given as follows. Here, Mu,e is the moment at the
exterior support of the equivalent frame.
-
l2 - c2 Mu,e
u,max = (9-4.16)
2 l2
9.4.3 Anchorage Devices
In post-tensioned slabs, the anchorage devices transfer the prestress to the concrete.
The device at the stretching end consists of an anchor block and wedges. At the dead
end, the wires are looped to provide the anchorage. Bursting links are provided in the
end zone to resist transverse tensile stresses in concrete. The following photos show
some anchorage devices.
(a) Anchorage device at stretching end

(b) Anchorage device for dead end with bulb
(c) Anchorage device for dead end with plate

Figure 9-4.8 Anchorage devices for slabs

The following photos show the anchorage devices, end zone reinforcement, spandrel
beam before casting and stretching and anchoring of the tendons after casting of
concrete in a slab.
Spandrel Beam
Bursting links
(a) End-zone reinforcement at stretching end
Bursting links
Dead end anchorage
(b) End-zone reinforcement and anchorage at dead end

(c) Stretching of tendons
(d) Anchorage block at stretching end

Figure 9-4.9 End-zone reinforcement and anchoring of tendons in a slab
(Reference: VSL India Pvt. Ltd.)

9.4.4 Additional Aspects
Restraint from vertical elements

Due to the restraint from monolithic columns or walls, the prestressing force in the slab
is reduced. Hence, the stiff columns or walls should be located in such a manner that
they offer least restraint. Alternatively, sliding joints can be introduced which are made
ineffective after post-tensioning of the slab.
Calculation of deflection
The deflection of a two-way slab can be approximately calculated by the equivalent
frame method. The deflection at a point is the summation of the deflections of the two
orthogonal strips passing through the point.
For an accurate evaluation, the following models can be adopted.

a) Grillage model
b) Finite element model.
Proportioning of drop panels and column capitals

Section 31 of IS:456 - 2000 provide guidelines for proportioning drop panels and
column capitals. A minimum length and a minimum depth (beyond the depth of the
slab) of a drop panel are specified. For column capitals it is preferred to have a conical
flaring at a subtended angle of 90. The critical sections are shown in Figure 12 of the
code.

9.5 Compression Members

Introduction
Analysis
Development of Interaction Diagram
9.5.1 Introduction
Prestressing is meaningful when the concrete in a member is in tension due to the

external loads. Hence, for a member subjected to compression with minor bending,
prestressing is not necessary. But, when a member is subjected to compression with
substantial moment under high lateral loads, prestressing is applied to counteract the
tensile stresses. Examples of such members are piles, towers and exterior columns of
framed structures.
As the seismic forces are reversible in nature, the prestressing of piles or columns is
concentric with the cross-section. Some typical cross sections are shown below.
Partially prestressed Prestressed circular and

column hexagonal piles
Figure 9-5.1 Examples of prestressed members subjected to compression

Figure 9-5.2 Stacked prestressed piles

(Reference: Industrial Concrete Products Berhad)
Precast prestressed piles have the following advantages.
1) Large bending and axial tension capacities.
2) Better quality control than bored and CIP piles.
Increased durability.
4) Use of high strength concrete gives reduced section.
The piles are prestressed by pre-tensioning. The procedure of pre-tensioning is

explained under Pretensioning Systems and Devices.
Since a prestressed member is under self equilibrium, there is no buckling of the

member due to internal prestressing with bonded tendons. In a deflected shape, there is
no internal moment due to prestressing.

The justification is explained in the next figure.

P
P
P C T

P P
a) Under external b) Under internal
compression prestressing
Figure 9-5.3 Internal forces at deflected configuration
In the first free body sketch of the above figure, the external compression P causes an
additional moment due to the deflection of the member. The value of the moment at
mid-height is P. This is known as the member stability effect, which is one type of P-
effect. If this deflection is not stable, then buckling of the member occurs. In the second
free body sketch, there is no moment due to the deflection of the member and the
prestressing force, since the compression in concrete (C) and the tension in the tendons
(T) balance each other.
When the additional moment due to deflection of the member is negligible, the member
is termed as short member. The additional moment needs to be considered when the
slenderness ratio (ratio of effective length and a lateral dimension) of the member is
high. The member is termed as slender member. In the analysis of a slender member,
the additional moment is calculated by an approximate expression or second order
analysis. In this module only short members will be considered.
9.5.2 Analysis
The stress in the section can be calculated as follows.
P0
fc = (9-5.1)
A

Here,
A = Area of concrete
In this equation, it is assumed that the prestressing force is concentric with the cross-
section. For members under compression, a compressive stress is considered to be
positive. The permissible prestress and the cross-section area are determined based on
the stress to be within the allowable stress at transfer (fcc,all).

The analysis is analogous to members under flexure. The stresses in the extreme fibres
can be calculated as follows.
Pe N Mc
fc = + (9-5.2)
A At It
In this equation, the external compression for a prestressed member is denoted as N

and is concentric with the cross section. The eccentricity is considered in the external
moment M.

A = area of concrete
At = area of the transformed section
c = distance of the extreme fibre from the centroid (CGC)
It = moment of inertia of the transformed section
The value of fc should be within the allowable stress under service conditions (fcc,all).
Analysis at Ultimate
When the average prestress in a member under axial compression and moment is less
than 2.5 N/mm2, Clause 22.2, IS:1343 - 1980, recommends to analyse the member as
a reinforced concrete member, neglecting the effect of prestress. For higher prestress,
the analysis of strength is done by the interaction diagrams.

At the ultimate limit state, an interaction diagram relates the axial force capacity (NuR)
and the moment capacity (MuR). It represents a failure envelop. Any combination of
factored external loads Nu and Mu that fall within the interaction diagram is safe. A
typical interaction diagram is shown below. The area shaded inside gives combinations
of Mu and Nu that are safe.
NuR
Compression failure
1 Balanced failure
eN
Tension failure
MuR
Figure 9-5.4 A typical interaction diagram for compression and bending
The radial line in the previous sketch represents the load path. Usually the external
loads increase proportionally. At any load stage, M and N are related as follows.
M = N eN (9-5.3)
Here, eN represents the eccentricity of N which generates the same moment M. The
slope of the radial line represents the inverse of the eccentricity (1/eN). At ultimate, the
values of M and N (Mu and Nu, respectively) correspond to the values on the interaction
diagram. For high values of N as compared to M, that is when eN is small, the concrete
in the compression fibre will crush before the steel on the other side yields in tension.
This is called the compression failure.
For high values of M as compared to N, that is when eN is large, the concrete will crush
after the steel yields in tension. This is called the tension failure.
The transition of these two cases is referred to as the balanced failure, when the
crushing of concrete and yielding of steel occur simultaneously. For a prestressed
compression member, since the prestressing steel does not have a definite yield point,
there is no explicit balanced failure.

9.5.3 Development of Interaction Diagram
An interaction diagram can be developed from the first principles using the non-linear
stress-strain curves of concrete under compression and steel under tension. Several
sets of NuR and MuR for given values of eN or xu are calculated. The distance of neutral
axis from the extreme compressive face is denoted as xu. Partial safety factors for
concrete and prestressing steel can be introduced when the interaction diagram is used
for design. Here, the procedure is illustrated for a rectangular section with prestressed
tendons placed at two opposite faces symmetrically, and without non-prestressed
reinforcement.
D
CGC
B +
Ap1 d1 d2 Ap2
Figure 9-5.5 A rectangular prestressed section
The notations used are as follows.
B = dimension of section transverse to bending
D = dimension of section in the direction of bending
Ap1 = area of prestressing tendons at the tension face
Ap2 = area of prestressing tendons at the compression face
d1, d2 = distances of centres of Ap1 and Ap2, respectively, from the centroid of the
section (CGC).
The strain compatibility equation is necessary to relate the strain in a prestressing

tendon with that of the adjacent concrete. Due to a concentric prestress, the concrete at
a section undergoes a uniform compressive strain. With time, the strain increases due
to the effects of creep and shrinkage. At service, after the long term losses, let the strain
be ce. Also, let the strain in the prestressing steel due to effective prestress be pe.

ce
pe pe
Figure 9-5.6 Strain profile due to effective prestress only
The strain compatibility equation for the prestressed tendons is given below.
p = c + p (9-5.4)
where, p = pe - ce
The strain difference of the strain in a prestressing tendon with that of the adjacent
concrete is denoted as p. The design stress-strain curve for concrete under
compression is used. This curve is described in Section 1.6, Concrete (Part II). The
design stress-strain curve for the prestressed tendon under tension is expressed as fp =
F(p).
The calculation of NuR and MuR for typical cases of eN or xu are illustrated. The typical
cases are as follows.
1) Pure compression (eN = 0, xu = )
2) Full section under varying compression (0.05D < eN eN xu = D , xu D)
3) Part of section under tension (eN xu = D < eN , xu < D)
4) Pure bending (eN = , xu = xu,min)
The above cases are illustrated in the following sketches.
e=0
e
Elevation
Strain
diagram
xu
Case 1 Case 2

e
Elevation
Strain
diagram
xu
xu
Case 3 Case 4
Figure 9-5.7 Typical cases of eccentricity and depth of neutral axis
In addition to the above cases, the case of pure axial tension is also calculated. The
straight line between the points of pure bending and pure axial tension provides the
interaction between the tensile force capacity and the moment capacity.
MuR
NuR
Figure 9-5.8 A typical interaction diagram for tension and bending
1. Pure compression (eN = 0, xu = ).

The following sketches represent the strain and stress profiles across the section and
the force diagram.
0.002 Strain profile
p1 p2
fp1 fp2
0.447 fck Stress profile
Force diagram
Tu1 Cu Tu2
Figure 9-5.9 Sketches for analysis at pure compression
The forces are as follows.

Cu = 0.447fck (Ag Ap) (9-5.5)

Tu1 = Tu2 = Ap1 fp1 (9-5.6)
= Ap1 Ep ( 0.002 + p)
The steel is in the elastic range. The total area of prestressing steel is Ap = Ap1 + Ap2.
The area of the gross-section Ag = BD. The moment and axial force capacities are as
follows.
MuR = 0 (9-5.7)
NuR = Cu Tu1 Tu2

= 0.447fck (Ag Ap) Ap Ep (pe 0.002 ce) (9-5.8)
In design, for simplification the interaction diagram is not used for eccentricities eN
0.05D. To approximate the effect of the corresponding moment, the axial force capacity
is reduced by 10%.
NuR = 0.4fck (Ag Ap) 0.9Ap Ep (pe 0.002 + ce) (9-5.9)
2. Full section under varying compression (0.05D < eN eN xu = D , xu D)

the force diagram.
xu = kD
3 / 7 D Strain profile
0.002
p1 p2
f p1 fp2
0.447 fck Stress profile
g Asector
Force diagram
Tu1 Cu Tu2
Figure 9-5.10 Sketches for analysis for section under varying compression
The limiting case for full section under compression corresponds to xu = D, when the
neutral axis lies at the left edge of the section. The strain diagram pivots about a value
of 0.002 at 3/7D from the extreme compression face. To calculate Cu, first the

reduction of the stress at the edge with lower compression (g) is evaluated. Based on
the second order parabolic curve for concrete under compression, the expression of g
is as follows.
2
4
7 D
g = 0.447fck
kD - 3 D
7
2
4
= 0.447fck (9-5.10)
7k - 3
The area of the complementary sector of the stress block is given as follows.
1 4
Asector = g D
3 7
4 (9-5.11)
= gD
21
Asector
x/
g +
4/7D
Figure 9-5.11 Complementary area of the stress block
Distance of centroid from apex (x /) = (3/4)(4/7)D = 3/7 D The forces are as follows.
Cu = [0.447fck D - Asector ] B
4
= 0.447fck D - gD B
21
4 4 2
= 0.447fck BD 1- (9-5.12)
21 7k - 3
Tu1 = Ap1fp1
= Ap1E p p1
= Ap1E p ( c1 + p )
D
xu - + d1
= Ap1E p -0.002 2 +
p (9-5.13)
3D
xu -
7

Tu 2 = Ap 2fp 2
= Ap 2E p p 2
= Ap 2E p ( c 2 + p )
D -d
x - 2
+
u
= Ap 2E p -0.002 2
p (9-5.14)
3D
xu -
7
The strains in the concrete at the level of the prestressing steels c1 and c2 are
determined from the similarity of triangles of the following strain profile.
c1
0.002
xu (D/2 + d1)
xu (3 / 7 D)
0.002 c2
xu (3 / 7 D)
xu ( D / 2 d 2 )
Figure 9-5.12 Strain profile across section
The moment and axial force capacities are as follows.

NuR = Cu Tu1 Tu2 (9-5.15)
MuR = Mc + Mp (9-5.16)
The expressions of Mc and Mp about the centroid are given below. Anticlockwise
moments are considered positive. The lever arms of the tensile forces are shown in the
following sketch.
3 D
Mc = 0.447fck DB 0 + Asector B x / + D -
7 2
10
= gD 2B
147 (9-5.17)
Mp = Tu1d1 Tu2d2 (9-5.18)

d1 d2
+
Tu1 Cu Tu2
Figure 9-5.13 Force diagram across the section
3. Part of section under tension (eN xu = D < , eN , xu < D)

the force diagram.
p 1 xu
p2 Strain profile
0.0035
fp1
fp2
0.447fck Stress profile
Force diagram
Tu1 Cu Tu2
Figure 9-5.14 Sketches for analysis for part of section under tension
The forces are as follows. The compression is the resultant of the stress block whose
expression can be derived similar to a reinforced concrete section.
Cu = 0.36fck xu B
Tu1 = Ap1 fp1

= Ap1 F (p1)
= Ap1 F (c1+ p)
Tu2 = Ap2 fp2
= Ap2 Ep p2
= Ap2 Ep (c2+ p)
The strains c1 and c2 are calculated from the similarity of triangles of the following
strain diagram.
c 1 0.0035
=
D xu (9-5.19)
+ d1 - xu
2

c 2 0.0035
=- (9-5.20)
D
xu - - d 2
xu
2
xu
c1
c2
0.0035
xu (D/2 d2)
(D/2 + d1) xu
Figure 9-5.15 Strain profile across section

NuR = Cu Tu1 Tu2 (9-5.21)
MuR = Mc + Mp (9-5.22)
The expressions of Mc and Mp about the centroid are as follows.

Mc = 0.36fck xu B [ (D / 2) 0.42 xu ] (9-5.23)
Mp = Tu1d1 Tu2d2 (9-5.24)
The lever arms of the forces are shown in the following sketch. The location of Cu is
similar to that of a reinforced concrete section.
d1 d2
+
Tu1 Cu Tu2
(D / 2) 0.42 xu
Figure 9-5.16 Force diagram across the section
4. Pure bending (eN = , xu = xu,min)

The value of xu is determined by trial and error from the condition that the sum of the
forces is zero.
Cu Tu1 Tu2 = 0
or, 0.36fck xu B Ap1 fp1 Ap2 fp2 = 0 (9-5.25)
Ap1fp1 + Ap 2E p p 2 (9-5.26)
or, xu =
0.36fck B

The strains p1 and p2 are calculated from the strain compatibility equations. The strain
p2 is within the elastic range, whereas p1 may be outside the elastic range. The
stresses fp1 and fp2 are calculated accordingly from the stress versus strain relationship
of prestressing steel.
The steps for solving xu are as follows.

1) Assume xu = 0.15 D (say).
2) Determine p1 and p2 from strain compatibility.
3) Determine fp1 and fp2 from stress versus strain relationship.
4) Calculate xu from Eqn. (9-5.26).
5) Compare xu with the assumed value. Iterate till convergence.

NuR = 0 (9-5.27)
MuR = Mc + Mp (9-5.28)
The expressions of Mc and Mp are same as the previous case.
5. Axial tension
The moment and axial force capacities are as follows. The cracked concrete is
neglected in calculating the axial force capacity.
NuR = 0.87fpk Ap (9-5.29)
MuR = 0 (9-5.30)
The above sets of NuR and MuR are joined to get the interaction diagram.
Example 9-5.1
Calculate the design interaction diagram for the member given below. The
member is prestressed using 8 strands of 10 mm diameter. The strands are
stress relieved with the following properties.
Tensile strength (fpk) = 1715 N/mm2.

Total area of strands = 8 51.6

= 413.0 mm2
Effective prestress (fpe) = 1034 N/mm2
Modulus (Ep) = 200 kN/mm2
Strain under fpe (pe) = 0.0042.
Grade of concrete = M40

Strain under fpe (ce) = 0.0005.
300
300
50 200 50
Solution
Calculation of geometric properties and strain compatibility relationship.
Ag = 300 300 = 90,000 mm2

Ap1 = Ap2 = 4 51.6 = 206 mm2
d1 = d2 = 100 mm
p = 0.0042 + 0.0005 = 0.0047
p = c + 0.0047
1. Pure compression (eN = 0, xu = )

MuR = 0 kNm
Cu = 0.447fck (Ag Ap)

= 0.447 40 (90,000 413)
= 1601.8 kN
Tu1 = Tu2 = Ap1 Ep ( 0.002 + p)

= 206.4 200 (0.0047 0.002)
= 111.5 kN

NuR = Cu Tu1 Tu2

= 1601.8 2 111.5
= 1378.8 kN
With 10% reduction, to bypass the use of interaction diagram for eccentricities
eN 0.05D
NuR = 1204.9 kN
2. Full section under compression (0.05D < eN eN xu = D , xu D)
400
Select xu = 400 mm
= (4 / 3) 300 mm
k =4/3
2
4
g = 0.447 fck
7k - 3
2
4
= 0.447 40
7(4/3) - 3
= 7.13 N/mm2
4 4
2
Cu = 0.447fck BD 1-
21 7k - 3

4 4
2
= 0.447 40300 1-
2

21 7(4/3) - 3

= 1486.9 kN
271.4 (3/7) 300 = 128.6
0.002
c1
150
Tu1 = Ap1E p ( c1 + p )
150
= 206.4 200 -0.002 + 0.0047
271.4
= 148.4 kN

271.4
0.002 c2
350
Tu 2 = Ap 2E p ( c 2 + p )
350
= 206.4 200 -0.002 + 0.0047
271.4
= 87.5 kN
NuR = Cu Tu1 Tu2

= 1486.9 148.4 87.5
= 1251.0 kN
Limit NuR to 1240.9 kN to bypass the use of interaction diagram for eccentricities eN
0.05D.
10
Mc = gD 2B
147
10
= 7.133002 300
147
= 13.1 kNm
M p = T u1d1 - Tu 2d 2
= 148.4100 - 87.5100
= 6.1 kNm
MuR = Mc + Mp
= 13.1 + 6.1
= 19.2 kNm 300
Select xu = 300 mm
k = 1
By similar calculations,
g = 17.9 N/mm2 NuR = 1060.6 kN
Cu = 1304.1 kN Mc = 32.9 kNm
Tu1 = 169.9 kN Mp = 9.6 kNm

Tu2 = 73.6 kN MuR = 42.5 kNm.
3. Part of section under tension (eN xu = D < eN , xu < D)

Select xu = 200 mm.
Cu = 0.36fck xu B
= 0.36 40 200 300
= 864.0 kN
0.0035
c 1 = 50
200
= 0.0009
p1 = 0.0009 + 0.0047
= 0.0056
c1 200
0.0035
50
Strain corresponding to elastic limit

py = 0.87 0.8fck / Ep
= 0.87 1715 / 200 103
= 0.0059.
p1 < py
fp1 = Ep p1
= 200 103 0.0055
= 1115 N/mm2
Tu1 = Ap1 fp1

= 206.4 1115
= 230.1 kN
0.0035
c 2 = - 150
200
= - 0.0026
p 2 = - 0.0026 + 0.0047
= 0.0021
200
c2 0.0035
150
fp2 = Ep p2
= 200 103 0.0021
= 416 N/mm2
Tu2 = Ap2 fp2 NuR = Cu Tu1 Tu2

= 206.4 416 = 864 230.1 85.9
= 85.9 kN = 548.0 kN
Mc = 0.36fck xu B [ (D / 2) 0.42xu ]
= 864 (150 0.42 200)
= 57.0 kNm
Mp = Tu1d1 Tu2d2 MuR = Mc + Mp

= 230.1 100 85.9 100 = 57.0 + 14.4
= 14.4 kNm = 71.4 kNm
4. Pure bending (eN = , xu = xu,min)

NuR = 0.0 kN
Try xu = 100 mm.
Cu = 0.36fck xu B
= 0.36 40 100 300
= 432.0 kN
0.0035
c 1 = 150
100
= 0.0052
p1 = 0.0052 + 0.0047
= 0.0099

c1
100
150 0.0035
From stress-strain curve

fp1 = 0.87fpk
= 1492 N/mm2
Tu1 = Ap1 fp1

= 206.4 1492
= 308.0 kN
100
c2 0.0035
50
0.0035
c 2 = - 50
100
= - 0.0017
p 2 = - 0.0017 + 0.0047
= 0.0029
fp2 = Ep p2
= 200 103 0.0029
= 580 N/mm2
Tu2 = Ap2 fp2

= 206.4 580
= 120.0 kN
Tu1 + Tu2 = 428.0 kN

This is close enough to Cu = 432.0 kN. Hence, the trial value of xu is satisfactory.

Mc = 0.36fck xu B [ (D / 2) 0.42xu ]
= 0.36 40 100 300 (150 0.42 100)
= 46.6 kNm
Mp = Tu1d1 Tu2d2
= 308.0 100 120.0 100
= 18.8 kNm
MuR = 46.6 + 18.8

= 65.4 kNm
5. Axial tension
MuR = 0.0 kNm
NuR = 0.87fpk Ap
= 0.87 1715 413.0
= 616.2 kN
The above sets of NuR and MuR are joined to get the following interaction diagram. The
limit on axial force capacity to consider the effect of eccentricity less than 0.05D, is not
shown.
1500
1000
N uR (kN)
500
0
0 20 40 60 80
-500
-1000
M uR (kNm)

Along with the interaction curve for the prestressed concrete (PC) section, the
interaction curves for two reinforced concrete (RC) sections are plotted. The section
denoted as RC 1 has the same moment capacity at zero axial force. The section
denoted as RC 2 has the same axial force capacity at zero moment. The gross section
of RC 1 is same as that of PC, but the section of RC 2 is smaller.
2500
2000
1500
NuR (kN)
PC
1000
RC 1
500
RC 2
0
-500 0 50 100 150
-1000
M uR (kNm)
Figure 9-5.17 Interaction diagrams for reinforced and prestressed sections
0 20 40 60 80
0
-250
NuR (kN)
-500
-750
-1000
M uR (kNm)
PC Strength RC 1 Strength
PC Cracking RC 1 Cracking
Figure 9-5.18 Interaction of moment and tension for cracking and strength
Comparing the curves for PC and RC 2, it is observed that if the moment demand is
small, then a smaller reinforced concrete section is adequate to carry the axial force. Of
course with increasing moment, the flexural capacity of the prestressed concrete
section is higher. Comparing the curves for PC and RC 1, it is inferred that for two

sections with same flexural capacities, the axial load capacity of a prestressed concrete
section is less. However if there is tension, the cracking load combination is higher for
PC as compared to RC 1.
Thus, prestressing is beneficial for strength when there is occurrence of:

a) Large moment in addition to compression
b) Moment along with tension.
Such situations arise in piles or columns subjected to seismic forces. In presence of

tension, prestressing is beneficial at service loads due to reduced cracking. Non-
prestressed reinforcement may be used for supplemental capacity.

9.6 Circular Prestressing

Introduction
General Analysis and Design
Prestressed Concrete Pipes
Liquid Storage Tanks
Ring Beams
Conclusion
9.6.1 Introduction
When the prestressed members are curved, in the direction of prestressing, the
prestressing is called circular prestressing. For example, circumferential prestressing in
pipes, tanks, silos, containment structures and similar structures is a type of circular
prestressing. In these structures, there can be prestressing in the longitudinal direction
(parallel to axis) as well. Circular prestressing is also applied in domes and shells.
The circumferential prestressing resists the hoop tension generated due to the internal
pressure. The prestressing is done by wires or tendons placed spirally, or over sectors
of the circumference of the member. The wires or tendons lay outside the concrete
core. Hence, the centre of the prestressing steel (CGS) is outside the core concrete
section.
The hoop compression generated is considered to be uniform across the thickness of a

thin shell. Hence, the pressure line (or C-line) lies at the centre of the core concrete
section (CGC). The following sketch shows the internal forces under service conditions.
The analysis is done for a slice of unit length along the longitudinal direction (parallel to
axis).

CGS
CGC
p
C C
Pe Pe T
R
a) Due to prestress b) Due to internal pressure
Figure 9-6.1 Internal forces under service conditions
To reduce the loss of prestress due to friction, the prestressing can be done over
sectors of the circumference. Buttresses are used for the anchorage of the tendons.
The following sketch shows the buttresses along the circumference.
Buttress
Figure 9-6.2 Use of buttress in circumferential prestressing
9.6.2 General Analysis and Design
Analysis
The basics of analysis and design for circumferential prestressing are provided for a
general understanding. Specific applications such as pipes, liquid storage tanks and
ring beams will be explained later.
The compressive stress can be calculated from the compression C. From equilibrium,
C = P0, where P0 is the prestress at transfer after short-term losses. The compressive
stress (fc) is given as follows
P0
fc = - (9-6.1)
A

Here,
A = area of the longitudinal section of the slice.
The permissible prestress is determined based on fc within the allowable stress at
transfer (fcc,all).

The tensile stress due to the internal pressure (p) can be calculated from the tension T.
From equilibrium of half of the slice, T = pR where, R is the radius of the mid-surface of
the cylinder. The resultant stress (fc) due to the effective prestress (Pe) and internal
pressure is given as follows.
Pe pR (9-6.2)
fc = - +
A At
Here,
At = area of the transformed longitudinal section of the slice.
The value of fc should be compressive and within the allowable stress at service loads
(fcc,all). In the previous equation, since Pe = pR and At is greater than A, fc is always
negative. Thus, the concrete will be under compression. To meet the safety standards,
a factor of safety can be further introduced.
Design
The internal pressure p and the radius R are given variables. It is assumed that the
prestressing steel alone carries the hoop tension due to internal pressure, that is Pe =
Apfpe = pR.
The steps of design are as follows.

1) Calculate the area of the prestressing steel from the equation Ap = pR / fpe.
2) Calculate the prestress at transfer from an estimate of the permissible initial
stress fp0 and using the equation
P0 = Ap fp0. (9-6.3)
3) Calculate the thickness of concrete shell from the following equation.

A = P0 / fcc,all (9-6.4)

Here, fcc,all is the allowable compressive stress at transfer.

4) Calculate the resultant stress fc at the service conditions using Eqn. (9-6.2). The
value of fc should be within fcc,all, the allowable stress at service conditions.
9.6.3 Prestressed Concrete Pipes
Prestressed concrete pipes are suitable when the internal pressure is within 0.5 to 2.0
N/mm2. There are two types of prestressed concrete pipes: cylinder type and the non-
cylinder type. A cylinder type pipe has a steel cylinder core, over which the concrete is
cast and prestressed. A non-cylinder type of pipe is made of prestressed concrete only.
IS:784 - 2001 (Prestressed Concrete Pipes (Including Specials) - Specification)
provides guidelines for the design of prestressed concrete pipes with the internal
diameter ranging from 200 mm to 2500 mm. The pipes are designed to withstand the
combined effect of internal pressure and external loads. The minimum grade of
concrete in the core should be M40 for non-cylinder type pipes.
First, the core is cast either by the centrifugal method or by the vertical casting method.
In the centrifugal method the mould is subjected to spinning till the concrete is
compacted to a uniform thickness throughout the length of the pipe. In the vertical
casting method, concrete is poured in layers up to a specified height.
After adequate curing of concrete, first the longitudinal wires are prestressed (the wires
can be pre-tensioned). Subsequently, the circumferential prestressing is done by the
wire wound around the core in a helical form. The wire is wound using a counter weight
or a die. Finally a coat of concrete or rich cement mortar is applied over the wire to
prevent from corrosion.
For cylinder type pipes, first the steel cylinder is fabricated and tested. Then the
concrete is cast around it.

Figure 9-6.3 Stacked prestressed pipes

(Courtesy: The Indian Hume Pipe Co. Ltd., Mumbai)
The analysis and design of prestressed concrete pipes consider the stresses due to the
different actions. A horizontal layout of the pipe is considered to illustrate them.
Analysis
The stresses in the longitudinal direction are due to the following actions.
1. Longitudinal prestressing (fl1)
2. Circumferential prestressing (fl2)
3. Self weight (fl3)
4. Transport and handling (fl4)
5. Weight of fluid (fl5)
6. Weight of soil above (fl6)
Longitudinal prestressing
The longitudinal prestressing generates a uniform compression.
Pe
fl 1 = -
Ac1 (9-6.5)
Here,

Pe = effective prestress
Ac1 = area of concrete in the core.
Circumferential prestressing
Due to the Poissons effect, the circumferential prestressing generates longitudinal
tensile stress.
Pe
fl 2 = 0.284 (9-6.6)
Ac
The above expression estimates the Poissons effect.
Self weight
If the pipe is not continuously supported, then a varying longitudinal stress generates
due to the moment due to self weight (Msw).
Msw
fl 3 = (9-6.7)
Zl
Here,
Zl = section modulus about the centroidal axis.
Transport and handling

A varying longitudinal stress generates due to the moment during transport and
handling (Mth).
Mth
fl 4 = (9-6.8)
Zl
Weight of fluid
Similar to self weight, the moment due to weight of the fluid inside (Mf) generates
varying longitudinal stress.
Mf
fl 5 = (9-6.9)
Zl

Weight of soil above

The weight of soil above for buried pipes is modelled as an equivalent distributed load.
The expression of stress (fl6) is similar to that for the weight of fluid.
The longitudinal stresses are combined based on the following diagram.
D + +
Section fl 1 fl 2 fl 3 + fl 4 + fl 5 + fl 6
of pipe
Figure 9-6.4 Stress profiles across section
The stresses in the circumferential direction are due to the following actions.
1. Circumferential prestressing (fh1)
2. Self weight (fh2)
3. Weight of fluid (fh3)
4. Weight of soil above (fh4)
5. Live load (fh5)
6. Internal pressure (fh6)
Circumferential prestressing
The compressive hoop stress (fh1) is given as follows.
Ps
fh 1 = -
Ac 2
Ps
=- (9-6.10)
1 tc
Here,
Ps = tensile force in spiral wire in unit length of pipe
Ac2 = area for longitudinal section of unit length
tc = thickness of the core.

Actions 2. to 5.
For each of these actions, first the vertical load per unit length (W) is calculated.
Moment (M) and thrust (T) develop across the thickness owing to distortion of the
section due to W, as shown in the following sketch.
W W
T M
Figure 9-6.5 Forces due to vertical load
The hoop stress at a point is calculated by the following equation.
M T
fh = + (9-6.11)
Zh A
The expressions of M and T due to W are as follows.

M = CM W R (9-6.12)
T = CT W (9-6.13)
Here,
CM = moment coefficient
CT = thrust coefficient
W = vertical load per unit length
R = mean radius of pipe
A = area of longitudinal section for unit length of pipe
Zh = section modulus for hoop stress for same length
= (1/6)t2 1000 mm3/m
t = total thickness of core and coat
Values of CM and CT are tabulated in IS:784 - 2001.

The internal pressure is as follows.
pR
fh 6 = (9-6.14)
At

The hoop stresses are combined based on the following diagram.
Coat
Core
fh 1
fh 2 + fh 3 + fh 4 + fh 5
fh 6
Figure 9-6.6 Stress profiles across the thickness
9.6.4 Liquid Storage Tanks
In the construction of concrete structures for the storage of liquids, the imperviousness
of concrete is an important basic requirement. Hence, the design of such construction
is based on avoidance of cracking in the concrete. The structures are prestressed to
avoid tension in the concrete. In addition, prestressed concrete tanks require low
maintenance. The resistance to seismic forces is also satisfactory.
Prestressed concrete tanks are used in water treatment and distribution systems, waste
water collection and treatment system and storm water management. Other
applications are liquefied natural gas (LNG) containment structures, large industrial
process tanks and bulk storage tanks.
The construction of the tanks is in the following sequence. First, the concrete core is
cast and cured. The surface is prepared by sand or hydro blasting. Next, the
circumferential prestressing is applied by strand wrapping machine. Shotcrete is
applied to provide a coat of concrete over the prestressing strands.

A few photographs are provided for illustration.
(b) Circumferential prestressing (close-up)
(a) Circumferential prestressing
(c) Shotcrete operation

Figure 9-6.7 Construction of a liquid storage tank
(Reference: DYK Incorporated)
IS:3370 - 1967 (Code of Practice for Concrete Structures for the Storage of Liquids)
provides guidelines for the analysis and design of liquid storage tanks. The four
sections of the code are titled as follows.
Part 1: General Requirement
Part 2: Reinforced Concrete Structures

Part 3: Prestressed Concrete Structures

Part 4: Design Tables.
Analysis
The analysis of liquid storage tanks can be done by IS:3370 - 1967, Part 4, or by the
finite element method. The Code provides coefficients for bending moment, shear and
hoop tension (for cylindrical tanks), which were developed from the theory of plates and
shells. In Part 4, both rectangular and cylindrical tanks are covered. Since circular
prestressing is applicable to cylindrical tanks, only this type of tank is covered in this
module.
The following types of boundary conditions are considered in the analysis of the
cylindrical wall.
a) For base: fixed or hinged
b) For top: free or hinged or framed.
The applicability of each boundary condition is explained next.
For base
Fixed: When the wall is built continuous with its footing, then the base can be
considered to be fixed as the first approximation.
Hinged: If the sub grade is susceptible to settlement, then a hinged base is a

conservative assumption. Since the actual rotational restraint from the footing is
somewhere in between fixed and hinged, a hinged base can be assumed.
The base can be made sliding with appropriate polyvinyl chloride (PVC) water-stops for
liquid tightness.
For top
Free: The top of the wall is considered free when there is no restraint in expansion.
Hinged: When the top is connected to the roof slab by dowels for shear transfer, the
boundary condition can be considered to be hinged.

Framed: When the top of the wall and the roof slab are made continuous with moment
transfer, the top is considered to be framed.
The hydrostatic pressure on the wall increases linearly from the top to the bottom of the
liquid of maximum possible depth. If the vapour pressure in the free board is negligible,
then the pressure at the top is zero. Else, it is added to the pressure of the liquid
throughout the depth. The forces generated in the tank due to circumferential prestress
are opposite in nature to that due to hydrostatic pressure. If the tank is built
underground, then the earth pressure needs to be considered.
The hoop tension in the wall, generated due to a triangular hydrostatic pressure is given
as follows.
T = CT w H Ri (9-6.15)
The bending moment in the vertical direction is given as follows.

M = CM w H3 (9-6.16)
The shear at the base is given by the following expression.

V = CV w H2 (9-6.17)
In the previous equations, the notations used are as follows.

CT = coefficient for hoop tension
CM = coefficient for bending moment
CV = coefficient for shear
w = unit weight of liquid
H = height of the liquid
Ri = inner radius of the wall.
The values of the coefficients are tabulated in IS:3370 - 1967, Part 4, for various values
of H2/Dt, at different depths of the liquid. D and t represent the inner diameter and the
thickness of the wall, respectively. The typical variations of CT and CM with depth, for
two sets of boundary conditions are illustrated.

CT CM
(a) Fixed base free top
t
CT CM
(b) Hinged base free top
Figure 9-6.8 Variations of coefficients for hoop tension and bending moment
The roof can be made of a dome supported at the edges on the cylindrical wall. Else,
the roof can be a flat slab supported on columns along with the edges. IS:3370 - 1967,
Part 4, provides coefficients for the analysis of the floor and roof slabs.
Design
IS:3370 - 1967, Part 3, provides design requirements for prestressed tanks. A few of
them are mentioned.
1) The computed stress in the concrete and steel, during transfer, handling and
construction, and under working loads, should be within the permissible values
as specified in IS:1343 - 1980.
2) The liquid retaining face should be checked against cracking with a load factor of
1.2.
CL/WL 1.2 (9-6.18)
Here,
CL = stress under cracking load
WL = stress under working load.

Values of limiting tensile strength of concrete for estimating the cracking load are
specified in the Code.
3) The ultimate load at failure should not be less than twice the working load.
4) When the tank is full, there should be compression in the concrete at all points of
at least 0.7 N/mm2. When the tank is empty, there should not be tensile stress
greater than 1.0 N/mm2. Thus, the tank should be analysed both for the full and
empty conditions.
5) There should be provisions to allow for elastic distortion of the structure during
prestressing. Any restraint that may lead to the reduction of the prestressing
force, should be considered.
IS:3370 - 1967, Part 3, also provides detailing requirements. The cover requirement is
as follows. The minimum cover to the prestressing wires should be 35 mm on the liquid
face. For faces away from the liquid, the cover requirements are as per IS:1343 - 1980.
Other requirements from IS:1343 - 1980 are also applicable.
9.6.5 Ring Beams
Ring beams support domes in buildings, tanks, silos and nuclear containment
structures.
Circular prestressing is applied on a dome by a grid of tendons. The cylindrical wall is

prestressed circumferentially and vertically. The ring beam is circumferentially
prestressed. The sketches below show schematic representation of the elements and
the prestressing tendons.

Dome
Ring beam
Cylindrical wall
Raft foundation
Figure 9-6.9 Cross-section of a nuclear containment structure
Tendon for vertical

prestressing of wall
Tendons for
prestressing of ring
beam Tendon for dome
prestressing
Tendons for
prestressing of wall
Figure 9-6.10 Typical layout of prestressing tendons at dome and ring beam junction

The following photo shows a prestressed nuclear containment structure.
Figure 9-6.11 Containment Structure, Kaiga Atomic Power Plant, Karnataka

(Reference: Larsen & Toubro)
Analysis
The analysis of a ring beam is based on a load symmetric about the vertical axis. Since
the dome is not supposed to carry any moment at the edge, the resultant reaction at the
ring beam is tangential. The following figure shows the forces at the base of dome.
H
Rd sin
V
Rd

Figure 9-6.12 Forces at the base of dome
Let the total vertical load from the dome be W. The vertical reaction per unit length (V)
is given as follows.
W
V= (9-6.19)
2Rd sin

Here,
Rd = radius of the dome
= half of the angle subtended by the dome.
The horizontal thrust (H) is calculated from the condition of the reaction to be tangential.
The value per unit length is given as follows.
H = V cot
W cot
= (9-6.20)
2Rd sin
The thrust is resisted by the effective prestressing force (Pe) in the ring beam. Pe can
be estimated from the equilibrium of half of the ring beam as shown in the following
sketch.
Pe
Rd sin
Plan of ring beam
Figure 9-6.13 Forces in the ring beam
Pe = H Rd sin
W cot
= (9-6.21)
2
9.6.6 Conclusion
Prestressing of concrete is observed in other types of structural elements, such as

bridge decks, shells and folded plates, offshore concrete gravity structures hydraulic
structures, pavements and raft foundations. The analysis of special structures is based
on advanced theory of structural analysis or the finite element method. After the
analysis, the design of such structures follows the basic principles of prestressed
concrete design. It is expected that in future, further innovations in structural form,
prestressing systems and construction technology will promote the application of
prestressed concrete.

A few photos of recent applications follows.
Figure 9-6.14 Cement silo, Jayanthipuram, Tamilnadu

(Reference: Larsen & Toubro)
Figure 9-6.15 Curved box-girder bridge, Jaipur-Kishangarh Highway, Rajasthan

(Reference: L & T Ramboll)

(a) Exterior view
(b) Interior view

Figure 9-6.16 Folded plate
(Department of Ocean Engineering, Indian Institute of Technology Madras)

Figure 9-6.17 Prestressed aqueduct, Gomti, Uttar Pradesh

(Courtesy: Hindustan Construction Company Ltd.)

Pre Stress

Uploaded by

Copyright:

Available Formats

Pre Stress

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pre Stress

Uploaded by

Copyright:

Available Formats

Prestressed Concrete Structures Dr. Amlan K Sengupta and Prof.

1.1.1 Basic Concept

A prestressed concrete structure is different from a conventional reinforced concrete

Force-fitting of metal bands on wooden barrels

Figure 1-1.1 Force-fitting of metal bands on wooden barrels

Pre-tensioning the spokes in a bicycle wheel

Indian Institute of Technology Madras

Figure 1-1.2 Pre-tensioning the spokes in a bicycle wheel

1.1.2 Early Attempts of Prestressing

Prestressing of structures was introduced in late nineteenth century. The following

Place and stretch mild steel rods, prior to concreting

Release the tension and cut the rods after concreting

Indian Institute of Technology Madras

Figure 1-1.4 A prestressed beam under an external load

Indian Institute of Technology Madras

Original length of steel rod (L1)

Original length of concrete beam (L2)

Reduced length of concrete beam (L3)

b) Beam at transfer of prestress

Final length of prestressed beam (L4)

c) Beam after long-term losses of prestress

Original tensile strain in steel = (L2 L1)/L1

Indian Institute of Technology Madras

1.1.3 Brief History

Before the development of prestressed concrete, two significant developments of

1824 Aspdin, J., (England)

1857 Monier, J., (France)

The following events were significant in the development of prestressed concrete.

Figure 1-1.6 Steel tie rods in arches

1888 Doehring, C. E. W., (Germany)

Indian Institute of Technology Madras

1908 Stainer, C. R., (USA)

1923 Emperger, F., (Austria)

1924 Hewett, W. H., (USA)

Figure 1-1.7 Portrait of Eugene Freyssinet

1926 Eugene Freyssinet (France)

Indian Institute of Technology Madras

1938 Hoyer, E., (Germany)

1940 Magnel, G., (Belgium)

The International Federation for Prestressing (FIP), a professional organisation in

Prestressed concrete was started to be used in building frames, parking structures,

Indian Institute of Technology Madras

Figure 1-1.8 Pamban Road Bridge at Rameshwaram, Tamilnadu

1.1.4 Development of Building Materials

The development of prestressed concrete can be studied in the perspective of

Indian Institute of Technology Madras

Compression (C) Tension (T) C and T

Stones, Bricks Bamboo, Ropes Timber

Concrete Steel bars, wires Structural steel

Passive combination Reinforced

Active combination Prestressed

Figure 1-1.9 Development of building materials

Indian Institute of Technology Madras

1.2 Advantages and Types of Prestressing

Forms of Prestressing Steel

Nature of Concrete-Steel Interface

Indian Institute of Technology Madras

1) Initial : It can be subdivided into two stages.

1.2.2 Advantages of Prestressing