Driving Out Variation:

6 Test Methods to Get Rubber Raw Materials Under Control

There’s a lot riding on our testing.™

Introduction:
Forewarned is Forearmed.

This is not new news - rubber and polymer processing can sometimes appear Alpha Technologies’ global network of applications engineers routinely work
to be a crap shoot. From batch to batch a lot can happen to cause variations in with suppliers and users of raw elastomers to show how various test methods
your product. Things as simple as humidity in the plant or the mixing speed of can predict their properties and processability. In this ebook they provide
a Banbury can change key properties in a batch, which is why manufacturers detailed information about how Premier RPA can be used to drive variability out
routinely test each batch that comes out of a mixer using an RPA. of your process by testing raw materials.

But when do you first measure your materials? If it’s at the point right after it “Measurement is the first step that leads to control and eventually to improvement.
drops from the mixer, that’s too late. If you can’t measure something, you can’t understand it. If you can’t understand it,

An RPA should be your first line of defense against variability in the raw you can’t control it. If you can’t control it, you can’t improve it.”

materials – either coming into your plant, or, if you are a supplier, before leaving - H. James Harrington, Author & Management Mentor
your plant for a customer’s manufacturing facility.

Rubber product manufacturers need to test their raw materials when they show
up on the loading dock, before they are put into production. And of equal or
greater importance… suppliers of raw elastomers should be providing more
than Mooney Viscosity data when they ship out their materials.

2
How Do You Know If Your Raw
Materials Are Good or Bad… and Why?
Material Viscosity (Pa*s)

Tire Compound 250000

EPDM 100000

HCR 50000

The answer flows from viscosity.

Toothpaste 600

Caulk 300
Viscosity (Pa·s)
Optical LSR 50
A supplier will suggest, “My Mooney gives me the viscosity of my silicone Honey 10
250
rubber.” Which leads to the question: “So… what is the Viscosity of this silicone?” Oil 1

200 Water 0.001

The answer: It depends. For three important reasons.

VISCOSITY (PA*S)
150
Reason 1: Viscosity is shear rate dependent.
100
When a supplier ships a material and tells you, “ This is a 70 Mooney elastomer,”
that’s probably right at one point of shear rate. Most data sheets give viscosity
50
at 10 s-1 (reciprocal seconds), at room temperature, with alpha (state of cure) =
0. So whether you test on a Mooney viscometer or an MDR, you’re only getting 0
1 10 100 1,000 10,000
one singular data point – shown by the cross-hair in Figure 1.
SHEAR RATE (1/S)

Viscosity is shear rate dependent. Right: references for viscosities of common materials
1
and certain elastomers.

3
But during production the material will experience a wide range of processing Cox-Merz Rule
conditions. It’s going to be mixed and then extruded or molded. “Mixing
An empirical relationship that
will take place around 50 reciprocal seconds,” notes Richard Hanzlik, Product
is commonly used in science
Manager, Americas. “Compression molding will be at about 20 reciprocal
and industry to determine shear
seconds with injection molding at maybe 1,000 reciprocal seconds, and
rate viscosity on the basis of an
extrusion around 500 reciprocal seconds.” (A reciprocal second is the unit used
oscillatory test such as an RPA. It
to express shear rates. It’s basically how fast the material is being worked,
works for many common polymers,
expressed in reciprocal seconds.)
but not necessarily for all
Shear rate has a MAJOR impact on viscosity. So, suddenly the 70 Mooney polymers. It states that the steady
elastomer at 10 reciprocal seconds becomes a 20 Mooney at 20 reciprocal state shear viscosity at a given
seconds. That’s why you need to test your raw materials through the whole shear rate is equal to the dynamic
range of potential shear rates. With Premier RPA we can physically go up to 48 viscosity at the same frequency.
reciprocal seconds, and then use the Cox-Merz correlation to estimate out to
about 550 reciprocal seconds.

4
 Reason 2: Viscosity is Temperature Dependent “And the big idea is that no other instruments are giving you that full range,”
says Hanzlik. “Especially with temperature because it’s very hard to change
Temperature has a moderate impact on viscosity. In the real world, you’re
temperature on anything but an RPA. So, with your RPA you can change your
usually processing at 100 degrees to 150 degrees Celsius. In Figure 2, we show
shear rate, change your temperature, and measure your state of cure, all with
that our viscosity at 100 degrees Celsius is about 35 Pa-s. But in reality, we’re
one sample, running the same test on one instrument.”
going to be processing at 125ºC. That means our viscosity is actually 5 Pa-s.
That’s a big difference. Depending on the temperature range you want to explore, and the number of
shear rates you want, the test can be done as quickly as five minutes, or take up
to an hour, it’s all based on what degree of sensitivity you want.

Reason 3: Viscosity is State-of-Cure Dependent

Viscosity (Pa·s) @10 (1/S) 0 Alpha
As a (state of cure) increases, viscosity also increases. But raw rubber doesn’t
90 cure until you put in the additives. So when you’re testing raw elastomers, you
80
don’t need to worry about cure, just yet. Alpha Application Specialists are always
70
available (worldwide) to answer questions about viscosity changes with the state
60
VISCOSITY (PA*S)

of cure. For more information about Alpha’s Support and Consulting Services go
50
to alpha-technologies.com/application-consulting-services.
40
30
20
10
0
0 20 40 60 80 100 120 140 160 180 200

TEMPERATURE DEPENDENCE VISCOSITY

Viscosity is temperature dependent.

2

5
Test Methods

More than Mooney… Frequency Sweeps to Characterize Raw Elastomers

We’ve established that Mooney Viscosity is the common data point provided have different processing properties. Typically, rubber product manufacturers
by raw material suppliers to their customers. But we have also established buy different rubbers from different suppliers. And maybe the modulus of
that it is a spectacularly limited view of the rubber properties because it is the rubber is the same… maybe the Mooney Viscosity is the same… but they
measuring at only one shear rate and one temperature condition. Additionally, have different properties. We can use ASTM D6204 to solve this question in
some companies think Mooney Viscosity is a very time-consuming test… for comparing different grades of rubber.”
all that effort you get just one data point. Another important testing efficiency
consideration is that the Mooney viscometer requires manual sample loading
whereas automation is available on the RPA. RPA 2000 Frequency Sweep of 2 Different Grades of A Polymer

10
Problem

A common scenario seen time and time again in the field is two materials that

S’ TORQUE (IN. LB.)

show the same value on a Mooney but when you go to process end up with
two different results. One may extrude nicely and the other will come out with a 1
poor surface finish or improper mold flow. 1 10 100 1,000 10,000

Case Study: Nitrile

In Figure 3, you can see how two elastomers with similar Mooney Viscosity 0.1

values have the same S’ value at one set of test conditions. “Sometimes there’s
a processability problem,” says Li Wan, Field Application Specialist, China. FREQUENCY (CPM)

“Maybe there are two materials with the same Mooney Viscosity, but they may
A B

Both polymers look the same on a Mooney viscometer. In this case the Blue Nitrile is
3
going to be much more difficult to injection mold than the green material.
6
 Molecular Weight Distribution

1,000

VISCOUS SHEAR STRESS G” (KPA)

100

10
1 10 100 1,000 10,000

FREQUENCY (CPM)

A B

RMA Nitrile Comparison of Molecular Weight Distribution using a frequency sweep.

4

Another comparison in Figure 4, this time of two nitrile rubbers with similar
Mooney values, show very different molecular weight distribution (MWD) – an
indication of how homogenous the chain lengths are in the polymer. “If you
have a bunch of long chains, short chains, medium chains – a big old mix of
“If you have a bunch of long
different chain lengths, you have a broad molecular weight distribution (a high
number),” says Hanzlik. “If you have a small molecular weight distribution that
chains, short chains, medium
means your chains are more consistent in their size – a material that’s more
homogenous. Typically, it’s going to have better end properties, but it might
chains – a big old mix of
also be more difficult to process.”

Nitrile rubber is a copolymer of two monomers: acrylonitrile and butadiene.

different chain lengths, you
The properties of these rubber compounds are determined by the acrylonitrile
(ACN) content. A high ACN content can result in better resistance to
have a broad molecular weight
hydrocarbon oils. A low ACN content offers better flexibility in low temperature
applications. In Figure 5, we used a frequency sweep to compare elastic shear
distribution (a high number).”
stress of two nitrile rubber samples with known ACN content.

A more detailed description of ASTM D6204 can be found here.

7
 Suppliers can use ASTM ACN Content

frequency sweeps as an
1,000

ELASTIC SHEAR STRESS G’ (KPA)

exploratory method to
determine the cause of an issue.
100

This is actionable information

that can be used to fix your
10
1 10 100 1,000 10,000

problem, so you don’t have

FREQUENCY (CPM)

scrap the next time around. A B

RMA Nitrile Comparison of Acrylonitrile (ACN) content using a frequency sweep.

5

While the previous examples have involved nitrile rubber, the characterization
of Styrene-butadiene rubber (SBR) can also be done using ASTM D6204 on a
Premier RPA. In Figure 5, we tested two different SBRs with a frequency sweep
at 7% strain. By comparing the slopes you can find out about the relative
molecular weight distributions. A steeper slope means a broader MWD, while a
shallower slope indicates a more narrow MWD.

“ The customer can use this information to define a specification – if G* at 10Hz

is above or below a certain value, that can be used as an acceptance/rejection/
exception criteria,” says Hanzlik. “Suppliers can also use ASTM frequency sweeps
as an exploratory method to determine the cause of an issue. This is actionable
information that can be used to fix your problem, so you don’t have scrap the
next time around.”

8
 Molecular Weight ASTM D6204

1,000 1
0.9
0.8
0.7

TAN (DELTA)
0.6

G* (KPA)
100
0.5
0.4
0.3
0.2
10 0.1
1 10 100 1,000 10,000 0.1 1 10 100 1,000 10,000

FREQUENCY (CPM) FREQUENCY (CPM)

A B A B

C D C D

ASTM D6204 frequency sweep at 7% strain for four different SBRs. Note: the slope may ASTM D6204 frequency sweep at 7% strain for 4 different SBRs.
6 7
indicate changes in molecular weight distribution. Note: the slope may indicate changes in molecular weight distribution.

The tangent of delta (AKA tan delta, AKA tand ), or the loss factor, is a measure
Tand guidelines for behavior of material:
of hysteresis and is the ratio of the loss of modulus to the storage modulus.
More simply, the tan delta quantifies the way in which a material absorbs and „ >3: Liquid Behavior

disperses energy. In Figure 6, we measured tan delta for four SBRs. A lower tan „ 0.4 to 2.0: Uncured Rubber Behavior
delta indicates material that will be more difficult in processing (producing
„ 1: G’ = G” Crossover Point
rougher sheets).
„ 0.010 to 0.300: Cured Rubber Behavior

„ <0.04: Very Solid Behavior

9
RPA frequency sweep testing provides a more realistic overview of shear rates
the material will encounter allowing you to predict actual behavior during
Comparison for 23 Polymers
processing. While all of these advanced capabilities of the RPA are great, there
120
are still many companies that are stuck with specifications that call out for a
Mooney value. Fortunately, you can still get a Mooney reading with your RPA to 100

fulfill some of these, let’s say…. outdated requirements. 80

ML1+4@100
To show a correlation of ASTM D6204 with Mooney Viscosity, we compared it to 60
23 different types of polymers and found a correlation coefficient (R value) of
40
.97 – a very strong correlation. “ Typically, these correlations hold true as long
as you are at the same temperature and within the same family of polymers,” 20

says Hanzlik. “So if you’re doing EPDM it will stay true for most EDPMs, or SBRs or
0
butyls, etc.” 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

RPA S*, @100C 7% STRAIN, 0.1HZ

Example of a correlation with Mooney Viscosity.

8

“If you’re doing EPDM it will

stay true for most EDPMs,
or SBRs or butyls, etc.”

10
Ramp Up the Stress… Using Strain
Sweeps to Characterize Raw
Elastomers
To better understand the molecular characteristics of
your material, it is necessary to take it outside of the
Linear Viscoelastic Region (LVE).

Problem

When creating a new formulation or replacing an obsoleted component of an

existing formulation it can be challenging to decide what type of elastomer
to use as well as what grade. Many things can affect how the material will
behave during processing and in the final application such as molecular weight,
entanglements, long chain branching, and gel content. Each one of these
can also vary somewhat during polymer synthesis batch to batch in synthetic
rubbers or tree to tree in natural rubber.

In the past, people have used more qualitative methods to try to identify these
characteristics. For example, extruding a sample through a Garvey die – if it’s
bumpy you know it’s bad and will be difficult to process. “But it’s hard to assign

Premier RPA gives

a number to that,” says Hanzlik. “ There’s a lot of subjectivity. Also, it’s just one
shear rate, so you’re not getting the full range of potential forces the material

you the most

can experience. Premier RPA gives you the most complete picture of raw
elastomers and uncured compounds.”

The benefit to the raw material supplier or user is that ASTM D6204 frequency
sweeps, and high strain sweeps, can confirm that their material is consistent.
complete picture of
Raw material variability is an even bigger concern for suppliers who compound
with natural rubber. Where did it come from – Malaysia… Cambodia… Vietnam?
raw elastomers and
Was it a humid year or a dry year? Many things can impact the molecular weight
and molecular weight distribution, and that means consistency is very hard to
uncured compounds.
control. It’s up to the supplier to make their product in a consistent fashion.
And, if the customer has defined specifications for certain values, that data can
be collected and included in the traveler that is shipped to the customer. All of
which helps confirm consistency and drive out variability.

11
 Case Study: Nitrile Rubber & High Strain Sweeps In a high strain sweep, the strain must be greater than the Linear Viscoelastic

In a strain sweep of an elastomer, frequency and temperature are constant. The Region – which is often greater than 50%. The data indicates the materials

oscillation amplitude varies – low to high. Usually S’ and S” values increase with ability to be mixed and processed (a high peak in S’ may predict a problem in

increasing strain. G’ and G” values tend to be constant at low strains, because mixing). Higher Tan ∂ values are easier to process. In addition, Fourier transform

when you double the strain the torque will double. (FT ) rheology using Large Amplitude Oscillatory Shear (LAOS) allows the relative
measure of long chain branching.

Linear Viscoelastic Region

50

40

30
G’(KPA)

20

10

0
0.5 5 50

ANGLE DEGREE

Strain sweep on a raw elastomer shows G’ is constant over a wide strain range.
9
This is called the Linear Viscoelastic Region. The linear region varies with
molecular weight distribution.

12
 Figure 10 illustrates different types of S’ curve shapes at high strains. “ The
first thing we look at is the magnitude of the S’ – the higher the S’ the more
difficult the material will be to process,” says Hanzlik. “ The second thing you
look for is the presence – or absence – of peaks. In one example we see a very
sharp peak that shows entanglements or long chain branching and indicates
very difficult processing. Broad peaks are the second most difficult to process
(second example). A continuous increase at high strain (third example) is the
third most difficult to process. The fourth example shows the most ideal case for
processing, with a continuous increase that flattens off.”

The question is, do we High Strain Sweep @100° C, 1Hz

assume the red peaks 16.1

14.1

are not sharper than the 12.1

S’TORQUE (DNM)
blue? The tand results
10.1

8.1

confirm that the red

6.1

4.1

sample has the best

2.1

0.1
0.1 1 10 100

characteristics in terms STRAIN DEGREES

of processability. A B

C D

Types of RPA S’ curve shapes at high strains.

10

13
 In situations where you’re not sure if one material has sharper or broader peaks
than another, tand will be the tie-breaker. A higher tand is going to mean easier
processing. Figure 10 shows the S’ curve on the left side and the tand on the
right side. The question is, how do we know which peak is sharper? The tand
results confirm that the red sample has the best characteristics in terms of
processability.

High Strain Sweep @100° C, 1Hz

Nitrile Study
3
TAN (DELTA)

25
2

20
1

15

S’ (DNM)
0
0 200 400 600 800 1,000 1,200
10
STRAIN %
5

Sample: e-SBR 3330 Sample: s-SBR SLR 4630

0
0 20 40 60 80 100 120
Sample: e-SBR 1500 Sample: Nd-BR High cis PBD
STRAIN IN DEGREES

High strain RPA tests on natural rubber. The tand results confirm that the red sample
11 A B
has the best characteristics in terms of processability.

C D
Taking it a step further, we did a case study comparing nine different grades
of nitrile rubber. In Figure 12, you can see one sample with a high sharp E F

peak (A). the next sample (C) has more of a broad peak. Sample A has the
best curve and would be the most favored. The difference in the grades is S’ versus Strain from high strain sweeps on 6 different nitrile grades.
12
molecular weight – Sample F has the highest molecular weight; sample A has
the lowest molecular weight. There are several generalities that can be deduced from high strain sweep tests.
For example, higher S’ values at high strain may indicate difficulty in processing.
S’ peaks may indicate even more difficulty in processing. A higher Tan ∂ value
may indicate easier processing.
14
Cross-Over Frequency Test

Problem When you want to take a deeper dive into the molecular characteristics of raw
Typical: a rubber product manufacturer has three suppliers of the same raw elastomers you can take ASTM D6204 and expand upon it. For example, you can
elastomer. add more conditions and look for the crossover point in raw elastomers. The
crossover point (also known as the Gel Point) is essentially where the G’ equals
Also typical: All raw elastomers have the same Mooney Viscosity, meaning their
the G”. And, where the G’ vs. G” lands can tell you whether the raw material has
molecular weight is the same. All are accepted by the manufacturer.
a higher or lower molecular weight, and narrower or broader molecular weight
Unfortunately typical: The raw elastomer from one of the suppliers is difficult to distribution.
process and requires higher mixing times.
Therefore, the frequency point where G’ crosses over G” can become a data
“So, the manufacturer needs to modify their process for that supplier,” explains point that can be used by product manufacturers to define an accept/reject
Dilip Dhupia, Applications Specialist, India. “ That is not good since they cannot specification for raw elastomers. And, can be used by raw elastomer suppliers to
change the process from one supplier to another supplier. So, in order to ensure their material won’t be rejected.
address this issue, we have developed a test that shows the characterization of
molecular weight distribution.”

15
 Case Study: Developing a QC Test for Silicone

Crossover Point
120

100
the manufacturer has defined accept/reject specifications it would simply be
80 Narrower MWD

MODULUS (KPA)
information without corresponding action. So why bother?
60
Higher Lower Meanwhile, the shop floor people at the rubber product manufacturer are
MWD MWD
40 constantly doing firefighting. Every issue that arises demands immediate action.
“What I’m trying to coach them is that when your process is running good, then
20 Broader MWD
you should collect more data,” adds Dhupia. “But no one – especially those on
0 the shop floor – likes doing the tedious calculations. They don’t want to export
0 2 4 6 8 10 12 14 16
the data to Excel and then draw a graph. They want the result in one value – is
FREQUENCY (HZ)
it good or not good. That’s it.”

“Fortunately, we have this built into the software, so you don’t have to do
G’ Elastic Modulus G” Viscous Modulus any calculations,” adds Hanzlik. “We have the ability to take this complicated
crossover test data and simplify it. It’s already in the Enterprise software.”
Frequency sweep of a raw silicone elastomer plotting G’ and G” versus frequency. where One tire manufacturer recently asked Alpha Application Specialists to consult
13
G’=G” (log-log scale). It is at this crossover point when the molecular weight (MW) and/ on a problem. Two suppliers of the same raw rubber, but one of the supplier’s
or molecular weight distribution (MWD) changes in linear elastomers.
raw rubber always had processing issues – for example, inconsistent extrusion
behavior. “ The rubber compound mixed from that supplier produced wavy

The Crossover Frequency Test was developed for the Premier RPA to estimate extrusions – like the waves in the sea,” says Dhupia.

the molecular weight and molecular weight distribution based on the crossover “I investigated the case and advised them to begin with crossover frequency
point where G’=G”. In practice it is a test that takes a reasonable amount of testing followed by LAOS testing,” explains Dhupia. “Based on that testing we
time. demonstrated that the supplier’s raw rubber was not up to par for what the

That said, neither rubber product manufacturers nor raw elastomer suppliers manufacturer required for their process. When we showed the results to the

are jumping to run this test. “Even though the product manufacturers are likely supplier they agreed. But before that, the supplier did not have any idea about

to have RPAs, in order to create a go/no go specification that is part of their molecular weight distribution. They only knew about Mooney Viscosity – which

regular routine they have to do a lot of work to generate the specifications is all they were required to know by the manufacturer.

for what is acceptable and what is not,” says Dhupia. “But if I bring in an RPA “Narrow molecular weight distribution means the rubber is tough. But if I have
exclusively for crossover frequency testing, I can realize cost savings in terms of a broad molecular weight distribution then processing is easier and requires
mixing time and in terms of extrusion set-up time. If my set-up time is reduced, less time. We need to educate the raw polymer manufacturer so they can
my rework of the compound is reduced. That’s always a cost benefit.” improve their process,” says Dhupia. “ They need to synthesize the same quality

One reason that raw elastomer suppliers aren’t using crossover frequency of polymer in every batch.”

testing is simply because their customers are not asking for it in their
specifications. Suppliers could run the test and provide the data, but until
16
FT-Rheology via Large Amplitude
Oscillatory Shear (LAOS) - Advanced
Molecular Characterization

Problem: “So, a raw material is intended for use in a specific application,” explains Scacchi.
“If the material is more or less branched, it will have different behavior in mixing
One goal of raw material molecular characterization is to ensure that the
and in extrusion. If we talk about raw elastomers, such as SBR, NBR, EPDM, BR,
elastomer grade from the supplier is in spec, and to evaluate variability. But it
etc. that show extreme branching content, the mixing and the extruding will be
can also be used to take a deeper dive into new grades provided by suppliers
to determine what their processing and performance attributes might be. more difficult.”

While the influence of the molecular architecture of polymers on viscoelastic There are other testing techniques that give insights into long chain branching,
properties is usually investigated by linear viscoelastic characterizations, such but only as an aggregate of the three main parameters of the macrostructure –
as the Payne effect, this approach is not advanced enough to reveal information molecular weight, molecular weight distribution, and degree of branching. The
on complex branching problems and for fully understanding the processing other techniques cannot distinguish branching from the other contributions.
behavior of elastomers and thermoplastics materials. Additionally, FT-rheology via LAOS is a very fast test method, compared to
Fortunately, the application of Fourier Transformation (FT-Rheology via the alternative techniques. In a very short time – typically a few minutes -
LAOS) is recognized as a very sensitive test method for detecting long chain you can distinguish the linear macrostructure from the long chain branch
branching (LCB) or to distinguish different polymer topologies. LAOS is used to microstructure.
evaluate and quantify the degree of branching to predict the next steps – the
processability in the extrusion, for example, or in the mixing.

“ The degree of branching is one of the most important variables of the

microstructure that significantly affect the processability in different
applications of the elastomers,” states Michele Scacchi, Applications
Specialist, EMEA.

17
 LAOS non-linear response

80

SHEAR STRESS (KPA)

30

-20 0.1 2.1 4.1 6.1 8.1 10.1 12.1

In an Alpha Technologies blog post on this topic (FT-Rheology via LAOS Part 1)
we show that the evolution in the shape of the distorted stress waveforms can
-70
be related to systematic changes in the internal microstructure of the material
or polymer topologies (linear or branched chain). In the non-linear region the -120

response of viscoelastic material is no longer characterized by a sinusoidal wave.

The torque/stress signal is distorted: the higher the imposed deformation the more TIME (S)

pronounced the distortion of the torque response, as shown in Figure 14.

A B C
The only way to describe the real response of material tested in the nonlinear
region (and quantify the distortion) is to apply Fourier Transformation to
account for all odd higher harmonics of the torque signal. Distortion of the stress signal from small (linear region) to large deformation
14
LAOS response can be visualized as parametric curves, called Lissajous- (non-linear region).

Bowditch figures, of the oscillating shear stress versus strain. It is possible to

distinguish between linear and branched problems.

18
 Lissajou Figure Following this criterion, the
150
LCB Increase
Secondary Loop

greater the ratio the higher

100
is the level of long chain
50
branching in the polymer.
SHEAR STRESS (KPA)
0

-6.3 -4.3 -2.3 -0.3 1.7 3.7 5.7

-50

-100
Long Chain Branching

Secondary Loop -150

SHEAR RATE (S-1)

Linear Polymer Highly Branched Polymer

Lissajous-Bowditch figure comparison: linear polymer versus branched polymer.

15

To quantify the level of long chain branching in the polymers, we defined a

proper criterion and introduced two LCB indexes. The first one is the simple
ratio between the first and fifth harmonic components of G’. Following this
criterion, the greater the ratio the higher is the level of long chain branching in
the polymer. The second index is mainly used to check the onset of secondary
loops. In fact, only polymers characterized by negative values of this LCB Index
show secondary loops.

19
 LCB index 1: (G’)1/(G’)5 ➔ LCB 1
LCB index 2: (G’)1/(G’)5-E3 ➔ LCB 2

where E3 is: [ 5 + 1 G’
4 4 G’5
3
2 1 G’3
2 G’5 ] ( )
“Some end product
manufacturers, like the LCB indices.
16

tire companies, are using

LAOS not only to test for “ This criterion to determine the degree of branching is purely empirical,”
notes Scacchi. “But compared to other rheological test methods it has the big

long chain branching, but advantage of being insensitive to the molecular weight and the molecular
weight distribution.”

also for filler dispersion FT-Rheology via LAOS is rapidly gaining acceptance throughout the European
Union as well as China. “Every time when I introduce the Premier RPA to

characterization.” customers, LAOS is always highlighted,” says Li Wan. “Some of our raw elastomer
suppliers are using LAOS for their long chain branching tests. And, some end
product manufacturers, like the tire companies, are using LAOS not only to test
for long chain branching, but also for filler dispersion characterization.”

“Just to tell you a story,” adds Scacchi, “we received some samples of synthetic
rubber from a manufacturer that had variations when put into production. “We
performed the measurement in front of the customer and were able to show
not only that the RPA was able to recognize variation in the different batches,
but also was able to detect the different variations in the branching.”

20
The Case of the Elastomer Gel Content

Gel content, or gelling, develops during cross-linking in elastomer

polymerization, where regions of cross-link density are so high that it becomes
complicated to process and mix with other ingredients. Gelling can cause
problems with processing, extruding, and even cause physical deformities in
final products, particularly in extruded profile applications such as automotive
weatherstripping, window channels, hose and tubing, as well as in roofing and
waterproofing membranes. EPDM and other elastomer polymer suppliers are
challenged to accurately measure gels and contaminants in their products, data
of which are now being required on some certificates of analysis.

During a standard proof-of-concept testing involving the Premier RPA, an

EPDM supplier had problems with the quality of their products, and traditional
methods of testing took too long for their processes. Knowing that gel was the
cause of their failing material, they requested testing – specifically ASTM D6204,
followed by a low frequency strain sweep – in part to generate internal support
for purchasing a Premier RPA.

21
 Peptizer – 15 min Raw – 15 min
Gel Study TanD
Peptizer – 10 min Raw – 10 min
2
1.8 Peptizer – 5 min Raw – 5 min
1.6
1.4 Peptizer – 2 min Raw – 2 min
1.2

TAND
1
Peptizer – 1 min Raw – 1 min
0.8
0.6
Control – 0 min
0.4
0.2
0
0 10 100 1,000 10,000

FREQUENCY
“ The initial results did find some difference in the tand
that could highlight the good versus the bad material, says
Jessica Gialamas, Application Engineer, Americas. “But, when
we sent the report to them they found the parameters were
Tan ∂ from ASTM D6204B.
17 very tight and some passing compounds were in the same
regions as the failing compounds.”

Peptizer – 15 min Raw – 15 min

G’ Gel Study
Peptizer – 10 min Raw – 10 min
1,000

Peptizer – 5 min Raw – 5 min

100 Peptizer – 2 min Raw – 2 min

Peptizer – 1 min Raw – 1 min

G’

10
Control – 0 min

1
1 10 100 1,000 10,000

FREQUENCY

G’ from ASTM 6204B.

18
22
 Boxplot of TanD5@500%

.12

.10

TAND5 @ 500%
.08

.06

.04

“The total test

.02
Fail Pass

time is less than

COMPOUND

10 minutes, and 19
5th harmonic tangent delta at 500% strain boxplot between good and bad samples.

sample prep only

includes five “ The data was repeatable,” notes Gialamas. “ The manufacturer implemented this
process to set standards on their difficult EPDM materials. What’s more, the total

grams, straight test time is less than 10 minutes, and sample prep only includes five grams,
straight from the bin.”

from the bin.” The application derived from LAOS testing opens the door for more rapid gel
analyses’ using the Premier RPA.

23
What’s the Story with
Non-Rubber Materials?

Raw elastomer is the predominant part of the rubber formulation. The next Problem
biggest item in the compound is filler. Without filler, rubber compounds have Unlike the raw elastomer, you can’t really test fillers and additives by
poor mechanical properties. themselves, with any meaningful result, on an RPA. It’s their effect on the
Typically, carbon black and precipitated silica – in different grades and types – other ingredients that shows their performance. But you can test them as
are the two main fillers used to reinforce rubber. And then, there will be smaller part of a rubber formulation, both in uncured and cured states. That makes
amounts of other ingredients such as curatives, processing aids, coupling using an RPA very effective for testing alternate fillers or a new polymer-filler
agents, anti-degradants, etc. coupling additive.

“ Typically, raw elastomer testing focuses on frequency sweeps, using methods

like those described in ASTM D6204” says Dr. Chris Robertson (Polymer
Technology Services LLC), a consultant for Alpha Technologies. “But with fillers
the key is using strain sweeps. And the RPA has that important capability.

24
 Kind of the “entry level” for evaluating a filler material or an additive that effects
filler networking behavior is looking at the Payne effect. If you add more filler
the Payne effect gets bigger, which is generally good for tire traction but bad
for rolling resistance in the world of tire tread technology. If you introduce more
polymer-filler (P-F) interactions through a coupling agent, for example, the
Heat buildup experiments give
Payne effect gets smaller, as shown in Figure 20.
important insights into filler
Payne Effect
network effects, as viscoelastic
More filler
break-up of the filler network
Or less P-F interactions is a main source of heat
buildup in rubber.
Less filler
STORAGE MODULUS, G’

Or more P-F interactions

TAN d
Unfilled rubber

0.01 0.1 1 10 100

STRAIN AMPLITUDE, g (%)

Typical Payne effect behavior of filled rubber compounds.

20

Heat buildup experiments also give important insights into filler network
effects, as viscoelastic break-up of the filler network is a main source of heat
buildup in rubber. For a closer look at heat buildup testing using an RPA,
check out this blog post by Dilip Dhupia.

25
Jump!

A deeper dive that yields more information is testing not just how big the filler
network is, but how fast does it form. Strain amplitude jump experiments on an
RPA are useful in characterizing the evolving filler network in rubber. The test
can compare the behavior of different fillers, coupling agents, functionalized
elastomers, etc. toward reducing (or increasing!) hysteresis for next-generation
rubber performance.

A basic strain down-jump experiment is a time-dependent oscillatory shear

experiment after suddenly reducing the strain amplitude with an RPA.

26
 “If you add something to the rubber formulation
to slow down the process of the filler particles
aggregating with each other to form a network,
you could easily see that in this test.”

“As strain is increased in a Payne effect experiment, by the time you get to a
Strain Jump certain point you’ve broken up the filler network – in most cases about 25% to
50% strain,” explains Robertson. “So, in the first part of the strain down-jump
0.30
experiment that’s your starting point – the filler network is broken up.”
0.25
T=160 °C; f=1.67 Hz “ Then suddenly… and you need a high-quality RPA to do this because you
STRAIN AMPLITUDE, g

0.20 have to do it quickly,” Robertson continues, “you down jump to a lower strain
right in the middle of the Payne effect and then measure the change in
0.15
rheology with time. The build-up of the filler network can be tracked by the
0.10
time dependence of dynamic viscosity (h*) or storage modulus (G’). So, if
0.05 you add something to the rubber formulation to slow down the process of
the filler particles aggregating with each other to form a network, you could
0.00
0 10 20 30 40 easily see that in this test.” The strain down jump experiment and example
results are highlighted in Figure 21. There are two characteristics that can be
TIME (MIN)
quantified from this testing for comparing different rubber formulations: (1)
the instantaneous increase in dynamic viscosity upon the strain down jump,
Dh*inst; and (2) the time-dependent increase in dynamic viscosity during the
Basic strain down-jump experiment. filler networking process, Dh*net. The impact of a new filler, coupling agent,
21
or functionalized polymer on these two filler networking parameters can
be used to validate their performance in a rubber formulation relative to a
control material.
27
Thermal Stability Testing of Elastomers
and Effects of Anti-Degradants.

An important part of quality analysis for raw materials in the tire and rubber The RPA test for looking at thermal oxidation of the polymer involves a
industry is assessing the thermal stability of elastomers (gum polymers). rheological time sweep at high temperature. We are looking for changes in
Different types of elastomers have varied stability against thermal oxidation. the material behavior as degradation occurs, as shown in Figure 21 for natural
Even the same type and grade of elastomer can have varied stability when rubber. The reductions in dynamic viscosity at the highest temperatures are
alternate suppliers are compared; polymer manufacturers can use different characteristic of chain scission (molecular weight breakdown) for NR. Other
types or amounts of antioxidant (AO), but these details are not shared in elastomers like butadiene rubber (BR) show increases in dynamic viscosity and
product specifications. storage modulus with aging due to branching/crosslinking behavior. Tracking
Thermal stability testing is also required for screening the efficacy of new the degradation of polymers with time-dependent dynamic rheology is an
antioxidants / stabilizers for rubber compounds, which has recently grown in established protocol in the plastics industry, but is not common in the rubber
importance given the ongoing search to find alternatives to the widely used industry due to a general lack of awareness.
multi-functional stabilizer, 6PPD, due to environmental concerns when used
in tires. The Rubber Process Analyzer (RPA) is an effective characterization tool
for these testing needs, as oscillatory shear rheology is sensitive to aging-
induced changes in polymer macrostructure. Also, the sealed cavity design of
the RPA provides a low oxygen testing environment comparable to the internal
components of a tire. A scientific rheometer doesn’t have this ability.

28
 Thermogravimetric Analysis ( TGA) is one way to look at the stability of rubber
materials, but is based on weight loss – it requires enough heat for a total “If someone has ten new
breakdown of the material so that it loses weight. Changes in rheology happen
at far lower temperatures than a TGA test. Typical TGA results for natural rubber antioxidants they’re
do not show any significant weight loss until temperature is increased above
about 300° C, yet clear degradation is diagnosed with the RPA at substantially investigating, they could
lower temperatures of 185 to 220° C (Figure 22).
quickly find out which
Measuring Rubber Degradation ones rise to the top.”
200

Dh*net
150
h* (KPA-S)

100 For many years, 6PPD has been the key antioxidant/antiozonant in tires. It helps
Dh*inst keep your tires together after long exposure to oxygen (air) and ozone. It’s an
EV tire tread compound (oil control)
50 essential ingredient for keeping us safe on the road. However, recent studies
100 phr silica
2nd pass compound (no curatives) have determined that 6PPD-quinone – the reaction product of ozone and 6PPD
160° C, 1.67 Hz
0 – is a threat to the environment, particularly to salmon and other aquatic life.
0 10 20 30 40
6PPD is currently on California’s watch list, with Europe coming along soon.
TIME (MIN)
There is thus renewed R&D interest in antioxidants and antiozonants given
the potential environmental concern about 6PPD, which is the mainstay
Strain = 25% Strain = 1.4%
stabilizer for natural rubber and synthetic diene elastomers. A rheological
time sweep procedure at high temperatures with the RPA can be a useful tool
Natural rubber degradation during RPA time sweep at the indicated temperatures. in the development and quality control process for anti-degradants and other
22
rubber additives.

“You wouldn’t even need to fully formulate a compound,” continues

Robertson. “You could make a simple blend of the raw elastomer and new
antioxidant. You could do a quick experiment by taking one or two parts (phr)
of the antioxidant, mix it into an elastomer using a two-roll mill or small-scale
internal mixer (lab prep mixer), and quickly generate a sample that you could
put into an RPA and see how it compares to a control material. And because
the RPA can be automated, you can do a rapid set of experiments. If someone
has ten new antioxidants they’re investigating, they could quickly find out
which ones rise to the top.”

29
Conclusions - Who’s on First?

As Strother Martin famously said in Cool Hand Luke, “what we have here is a At Alpha, we believe that everyone has to do their part to drive out variation
failure to communicate.” and make the safest most effective products possible. Raw material suppliers

Suppliers of raw materials are hesitant to do more advanced testing of their would benefit from collecting more data about their elastomers. For example,

elastomers because their customers – the rubber product manufacturers – ASTM D6204 strain sweeps would better characterize their material and validate

aren’t providing the specs to test to. As a result, the conventional wisdom is the consistency of their product.

“Mooney Viscosity must be good enough.” While Rubber manufacturers need to take a closer look at the ingredients of their

On the other hand, manufacturers are scrambling to meet the tighter process, and begin to define the specifications for their raw materials that will

specifications of their end customers. A lot of effort is being spent looking at help drive variability out. They need to run advanced characterizations of their

how their in-process rubber is affected by their mix methods, filler interactions polymers such as high strain sweeps, crossover frequency tests and FT-rheology

and cure cycles. via LAOS and collect the data and develop specs. And (and this is a big AND),
they need to communicate these new accept/reject specs to their suppliers.
Meanwhile the guys on the shop floor at both locations are busy fighting fires.
They should be the boots-on-the-ground… collecting more data from their raw And while rubber product manufacturers have been reticent (to say the least)

materials. But without guidance from QC managers, why should they care? to share their specification and process secrets, Alpha Technologies has the
scientific and engineering experience to fill in the gaps for elastomer suppliers
So, who’s on first?
and rubber manufacturers, alike.

30
References
Pawlowski, H., Hanzlik, R. Molecular characteristics of elastomers measured
with modern rheological tools, Rubber & Plastics News, May 18, 2020 issue,
pp. 14-18 (2020).

Giacomin, A.J., Dealy, J.M. Large-Amplitude Oscillatory Shear. In: Collyer, A.A.
(eds) Techniques in Rheological Measurement. Springer, Dordrecht (1993).
doi.org/10.1007/978-94-011-2114-9_4

Leblanc, J.L., Nonlinear Viscoelasticity of (Unvulcanized) Natural Rubber, Derived

Materials, and Compounds Through LAOS Testing, Rubber Chem. Technol. 83,
65–96 (2010). doi.org/10.5254/1.3548267

Henning, S.K., Robertson, C.G., EV Tread Development: Reducing the Payne

Effect with Sustainably Sourced Silane-terminated Resins, Rubber News,
in press (2023).

Robin, C., Robertson, C.G. Glass-like Signatures in the Dynamic Rheology of

Particle-Filled Polymers, Macromolecules 55, 2729–2738 (2022).
doi.org/10.1021/acs.macromol.2c00115

31