RUBBER PHYSICS

ALEXANDER DUBCEK UNIVERSITY OF TRENCIN

J. Gleick: Chaos: Making a New Science. /in Slovak translation/, copyright Ando publishing 1996.
Computer simulation of different fractal plants creation as a model of a polymer structure building
 SUMMARY

The module ‘Physics of polymers’ is the field of physics associated to the study of
polymers, understanding of the mechanical, physical, electrical and thermodynamic
properties of polymeric materials. Current areas of focus include structural and
mechanical behavior of networks, segmental relaxation and the glass transition,
miscible polymer blends, and polymer-based composites. Polymer physics is part of
the wider field of polymer science.

All of these aspects are discussed in this module, which consists therefore of the
following seven parts:

• Structure of polymers & Physical and phase states of polymers

• Mechanical properties of solid state polymers
• Payene effect & Viscosity and mechanical properties of viscous and
viscoelastic materials & Fracture properties of polymers
• Models of viscoelastic behavior of materials
• Selected physical properties of polymeric materials
• Electrical properties of polymers
• Physical Processes Influencing Surface Contact of Two Materials

The first sub-module discusses polymer materials; their basic structure units, that is
good for better understanding of distinctiveness of polymers as materials, it is
purposeful to consider the conception of the hierarchic disposition of
macromolecular materials. We have tried to explain the constitutional unit, there are
shown here some types of polymer chains. Density of cohesive energy is explained
here and we introduce for polymers derived quantity, namely the parameter of
solubility and geometry of polymer chain.
The aim of the second part of the first sub-module that is called physical and phase
states of polymers explains basic states for polymers as gaseous, liquid and solid
state, structural and phase transformation that takes place in them. Readers can
obtain iformation about 14 different crystal lattices, called Bravais Lattices. It
describes also phenomenological description of glassy state. Next there is phase
transitions described as transition of state, transformation of amorphous substance
into crystalline and vice versa, or transformation of one crystalline system into
another. Glass transition temperature defined as temperature, at which bend or
discontinuity occurs on dependence of specific volume on temperature. There are
deffined several methods for the determination of glass transition temperature.

The second sub-model deals with a description of deformation of solid elastic

materials as well as to set up phenomenological values describing mechanical
properties of these substances. We are going to learn about thermodynamic and
microstructural aspects of the process of elastic deformation. Later on these
concepts and knowledge will be generalized on polymers and rubber.
The third sub-module has three important parts. First part deals with a description of
Payne effect and also with its physical description. The second one introduces the
viscosity terms and viscoelastic behaviour of materials under loading. There is
presented the terminology of complex physical parameters (modules), WLF
transformations and results of rubber mixtures measurements and temperature
dependence of viscosity. And the third one is devoted to the issue of the fracture
attributes of polymeric materials, where fracture mechanics provides a methodology
evaluating the structural integrity of components containing such defects, and
demonstrating whether they are capable of continued, safe operation. The basic
criterion in any fracture mechanics analysis is to prevent failure. You can find there
information about the historical overview where are presented some approaches.

In the fourth sub-model are introduced basic theoretical approaches describing

viscoelastic behavior of materials. Readers will be familiared with basic models:
Maxwell, Voigt and their combinations.

The fifth sub-model provide readers with a schematic review of results of

measurements of selected physical quantities (especially mechanic and thermal)
gained for various polymeric materials and rubber as well as their connection with
theoretical background knowledge presented in previous parts. At the same time we
are going to point out differences between physical properties of elastomers and
polymers.

The sixth sub-model introduces the specific resistivity and conductivity, dielectric
properties of polymers, electrical stress of polymers and finally percolation
threshold.

The last one sub-model presents hysteretic attributes of viscoelastic materials in the
process of their deformation. Some theoretical approaches of hysteresis explanation
from the point of view of solids' contact are discussed. It also deals with gluing
questions and adhesion at mutual contact of materials problems, as well as with
theory of friction.

©Alexander Dubcek University of Trencin 2007 3

 TABLE OF CONTENTS
1. Structure of polymers............................................................................................. 5
1.1. Basic information............................................................................................ 5
1.2. The density of the cohesive energy................................................................. 7
1.3. Geometry of polymer chains........................................................................... 9
1.4. Physical and phase states of polymers .......................................................... 13
1.4.1. Glassy state ............................................................................................ 13
1.4.2. Phase transitions .................................................................................... 15
1.4.3. Glass transition temperature .................................................................. 18
References:........................................................................................................... 24
2. Mechanical properties of solid state polymers..................................................... 25
2.1. Description of deformation of solid elastic materials ................................... 25
2.2. Thermodynamic aspects of deformation....................................................... 35
References............................................................................................................ 37
3.Viscosity and mechanical properties of viscous and viscoelastic materials ......... 38
3.1. Viscosity ....................................................................................................... 38
3.2. Time dependence of deformation ............................................................... 46
3.3. The temperature dependence of viscosity – micro structural view .......... 54
3.4. Thermodynamic aspects of viscous elastic and rubber deformation ....... 56
3.5. Payne effect................................................................................................... 60
3.6. Fracture properties of polymers .................................................................... 65
References........................................................................................................... 75
4. Models of viscoelastic behavior of materials....................................................... 77
References:........................................................................................................... 83
5. Selected physical properties of polymeric materials ........................................... 85
References:......................................................................................................... 102
6. Electrical properties of polymers ....................................................................... 103
6.1. Electrically Conductive Polymers............................................................... 103
6.2. Electrically Conductive Composites........................................................... 104
6.3. Influence of Force Field on Polymer Behaviour......................................... 106
6.4. Electrical Strength of Polymers .................................................................. 107
6.5. Dielectric Properties of Polymers ............................................................... 108
References......................................................................................................... 110
7. Physical Processes Influencing Surface Contact of Two Materials................... 112
7.1. Hysteresis.................................................................................................... 112
7.1.1. Theories of Hysteresis ......................................................................... 113
7.1.2. Unified Theory..................................................................................... 113
7.1.3. Relaxation Theory................................................................................ 115
7.2. Gluing and adhesion ................................................................................... 115
7.2.1. Adhesion as a surface problem ............................................................ 117
7.2.2. The Role of Adhesion at Dynamic Contact of Two Materials –
Macroscopic and Molecular Understanding .................................................. 120
7.2.3. Ratio Theory ........................................................................................ 121
7.2.4. Mixed Theory ...................................................................................... 121
7.3. Friction........................................................................................................ 122
7.3.1. Friction as Dynamic Problem of Two Surfaces Contact ..................... 124
References......................................................................................................... 139

©Alexander Dubcek University of Trencin 2007 4

 CHAPTER 1
Structure of polymers

Objectives to achieve
In this chapter are explained polymer materials; their basic structure units, density
of cohesive energy and geometry of polymer chain and so on. The aim of this
section is also to explain basic states for polymers, structural and phase
transformation that takes place in them. It describes also phenomenological
description of glassy state.

1.1. Basic information

For better understanding of distinctiveness of polymers as materials, it is purposeful

to consider the conception of the hierarchic disposition of macromolecular
materials. Every condition, behaviour, or individual property is possible to be
considered at least from three points of view:
- from the view of the structure and dynamic behaviour of an individual
isolated molecule
- from the view of the change in behaviour of a macromolecule whose
movements are limited by the closeness of other macromolecules
- and finally from the view of properties of polymeric material formed by
large number of macromolecules which not only influence each other, but
they can make various structural formations distinguishing from each other
by thermodynamic parameters and certainly by numerous properties despite
the identical chemical composition.

A polymer is formed by a large number of mutually connected “mers”, units, which

are chemically identical. The repetition moment is typical for a polymer, when a
defined grouping of atoms is repeated many times in an identical constellation.

This repeating structure is called a structural, or a building unit, or according to the

new nomenclature a constitutional unit that is defined as the smallest unit, whose
repetition describes the regular polymer. In this process the constitutional unit
doesn’t need to be necessarily identical with the structure coming out from a
monomer. For example, the polymer for polyethyleneterephtalat is formed by poly-
condensation of ethylene glycol and terephtalic acid, but the constitutional unit, that
is, the repeating structure is:

– CH2 – CH2 – O – CO – C6H5 – CO – O –

that is, a condensation product of two different individual reagents.

A polymer is formed this way. From this point of view, every polymer is
simultaneously a macromolecule, i.e. a molecule with a very large mass
significantly exceeding a substance of so called low molecular materials. Molar

©Alexander Dubcek University of Trencin 2007 5

masses range between a few hundreds and millions, and some biological molecules
reach billions. There is not a clear border between low-mol. and macro-mol.
materials, the transition are made by so called oligomers.

At the same time the difference between terms “macromolecule” and “polymer”
may be pointed out. A large molecular weight is a determined parameter in the first
case, whereas the repetition moment in the second one. Therefore every polymer, as
it always has relatively large molecular weight, is simultaneously a macromolecule,
but not every macromolecule is a polymer. This case occurs when the repeating
structure is absent, as it, for example, goes for enzymes, which are formed by a
large number of many amino acids, but whose mutual connection occurs in specific
non-repeating combinations.

On the next picture are shown types of polymer chains:

- homopolymer chain
- copolymer – alternating, random and blockwise

Then the chains can be linear, branched or crosslinked.

Figure 1.1. Schematic representation of polymer chains

The primary connection of constitutional units in the macromolecule is provided

mostly by covalent bonds. Their length depends on atoms – which are bound by a
particular bond, and that is given for every pair of atoms in non-deformed state in
the table 1.1.

Table 1.1. The lengths of covalent bonds (nm) in non-deformed state for different
pairs
C–C 0.154 nm C=N 0,12 nm
C–N, C–O 0.14nm C≡N 0,117 nm
C–H,N–H,O–H 0.11nm to 0.096 nm C–Si 0,19 nm

©Alexander Dubcek University of Trencin 2007 6

 C=C 0.13 nm Si–O 0,18 nm

The solidness of the bond is formulated by dissipation bond energy, that is, by
energy needed for disruption of the given bond. The dissipation energy of simple
bonds is 250 - 400 kJ / mol, for multiple bonds it is 400 - 600 kJ / mol. It might be
seen from the comparison that the dissipation energy of, for example, the second
bond is lower than of a simple bond. It follows from this that a double bond is in
comparison with a simple one more reactive.

Also the fact, that the connection of atoms with a bond is sterically constant and that
it might be characterized by so called a valence angle, is important for
understanding of the structure of the macromolecular chain. These valence angles
are given for individual types of bonds in the table 1.2.

Table 1.2. Valence angles for different types of covalent bonds

H-C-H, C-C-C, C-O-C, C-N-C 105 - 113°
C-C=C 125°
C≡C 180°
Si-O-Si 134°

It is important to realize for a discussion about geometry of polymeric chains that a

free rotation of substituents around simple bonds is possible (producing different
conformations), whereas a rotation of substituents around multiple bonds is not
possible (configurational structure).

1.2. The density of the cohesive energy

Apart from relatively solid primary bonds, that provide cohesion of each molecule
in every material consisting of many molecules, there are much weaker secondary
bonds. These bonds stand for cohesive power that affects molecules consisting of
covalently bound atoms (van der Waals’s power). The molecules are 0.3 - 1 nm
distant from each other and the energy needed for their disruption is less than 40 -
50 kJ/mol. The secondary bonds provide cohesion of whole material and present a
power which molecules in a material are attracted mutually by. Their sum can be
then considered as the rate of intermolecular cohesion. This energy significantly
depends on the distance between molecules because the extent of intermolecular
cohesion is proportional to the power of six of the distance between molecules.

The solidness in low molecular materials is much lower than for primary bonds, and
they come into high values only if they are polar, for example, acid-alkaline
interactions, hydrogen bond, etc. To the contrary, the secondary bonds can be very
solid for macromolecules, where the chain contains several thousands carbons,
because the number of contacts is very large. Therefore, for example, butane is gas,
whereas octane is already liquid, hydrocarbons with a chain of 20 – 30 carbons have
a character of waxes with a low melting point, but, for example, linear polyethylene
is resilient material where the secondary intermolecular bonds are more solid than
the primary ones between atoms of carbons. Due to this fact the polymer is not

©Alexander Dubcek University of Trencin 2007 7

possible to be got into the gaseous state, because for disruption of the secondary
bonds, it is necessary to heat it on the temperature which causes quick breaking of
primary covalent bonds and the material starts burning.

The intermolecular cohesive energy can be then defined as the energy of mutual
attractive force of molecules. It can be determined simply as an energy needed for
the transition of a molecule from the liquid phase, where molecules are very close
to each other and interactions are strong, to the gaseous phase, where molecules are
so far from one another, that considering the fall of the cohesive energy with 6
power of the distance, the interactions can be consider zero. The amount of energy
needed for the transition of a mol of a material from the liquid to the gaseous phase
serves then as a measure of the mutual attractiveness of molecules. This parameter
is called the density of the cohesive energy that is defined as an evaporative
internal energy of a volume unit of liquid.

HKE= Uv/Vm= Uv/Mρ-1 (1.1)

Where individual symbols present: molar evaporative energy, molar volume, weight
and density. It is possible also to write:

Uv = Hv - P∆V (1.2)

Thus molar evaporative internal energy = molar evaporative enthalpy – expansion

work.

Uv and Hv rise with the molecule size, but at the same time also the molar volume
rises, so HKE goes up only moderately in the homologous series and then it
stabilizes on nearly constant value 250 MJm-3. HKE measures in fact the
intermolecular cohesion of sections of the chain and not the cohesion of whole
molecules. Therefore it is suitable also for characterization of polymers as a
measure of the intermolecular cohesion. It cannot be determined as the evaporative
energy, but it is necessary to use indirect methods.

We introduce for polymers therefore derived quantity, namely the parameter of

solubility defined as the square root of the density of the cohesive energy.

δ = HKE 1 / 2 (1.3)

We use an empirical finding for its determining, that is, liquids with similar values
of HKE can be mutually well mixable.

The following technique is then applied to determine the parameter of solubility

itself: The polymer is netted to become insoluble; subsequently it is left to swell in
series of liquids with different δ. Good dissolving agents try to melt the polymer,
what is not possible for netted material. It comes up to swelling where the effort of
the dissolving agent to penetrate into the polymer (increase of the entropy of the
system polymer – dissolving agent) prevents from gradual stretching of chains
(decrease of the entropy of the polymer chains). These two processes finally lead to
a balance, and the higher affinity of the dissolving agent against the polymer is, the

©Alexander Dubcek University of Trencin 2007 8

 higher degree of the swelling is. The polymer has δ identical with δ of the liquid in
which it reached the highest degree of the equilibrium swelling.

The importance of the parameter of solubility as a measure of the cohesive energy

might be demonstrated on the qualitative order of polymers in the direction of rising
HKE. Non-polar – saturated hydrocarbon polymers (PE, PP, and EPM) have the
lowest values, polarizable – unsaturated and aromatic hydrocarbon polymers (BR,
IR, SBR, PS), then polymers with polar groups as ether, ester, nitrile, Cl, F have
higher values.

Highly polar polymers with larger content amide, urethane, urea, aminic,
carboxylic, hydroxylic, possibly ion groups have the highest HKE. It is interesting
to mention here, for instance, perfluorit polymers might be expected to have the
high HKE considering highly polar fluorit atoms in a molecule, but on the contrary
they have this value low and they are very water-repellent. The explanation lies in
mechanism when the charges of dipoles C-F compensate mutually. Similar situation
arises for polysiloxanes, where CH3 groups gradate the polarity Si-O of the
skeleton.

In the table 1.3 are displayed some compounds and their densities of cohesive
energy.

Table 1.3. Densities of the cohesive energy of choose compounds

HKE [cal/cm3] HKE [cal/cm3]
lineardimetylsiloxán 35 – 24,6 ethanethiol 85,0
Aliphatic prefluorované 30 – 38,5 chloroform 86,0
uhľovod.
Heptane 53,1 dimetylsulpfide 88,0
1-hexylene 55,0 cyclohexanone 98,0
Dieethylether 55,0 ethylamine 100
diisobutylene 59,0 acetone 100
cyclohexane 67,0 Acetic acid 102
tetrachlorometane 74,0 aniline 106
Xylene 77,0 ethylene oxide 122
bromoethane 80,0 cyclohexanone 130
etyhylacetate 83,0 butanol 130
Benzene 85,0 acetonitrile 136
Water 585 ethanol 161
polytetrafluoroetyhylene 38,5 SBR (23 % styrene) 65,7
polydimethylsiloxane 53,0 NBR (25 % acrylonitrile) 79,0
PE 62,0 polychloroprene 74,0
IIR 61,0 fluorinecaoutchouc(Viton 44 – 69
A)
NR 65,7 polybutadiene - sodium 74,0
polybutadiene - emulsive 70,6
*/ for SI unit is relationship ratio: cal/cm3 = 4.1868.103 J/m3

1.3. Geometry of polymer chains

©Alexander Dubcek University of Trencin 2007 9

The appearance of a concrete polymer linear chain may be derived in the first
approach from the length of bonds and valence angles between atoms that form the
basic chain. A further parameter is a number and a form of substituents and of side
chains as well as the polarity of functional groups that can occur in main or side
chains. When judging the form of a polymer chain, we speak about its conformation
that is defined as an arrangement that might be changed by an internal rotation
round simple bonds. The change of the conformation is a physical action.

Conformation changes and factors that affect these changes can be demonstrated in
the simplest way on a segment, which consists of four carbons (for hydrocarbon
chains, for example polyethylene), it means of three bonds. An outside carbon and a
middle bond out of them are considered to be still within given space, and we watch
changes of positions of the second outside carbon at free rotation of the middle
bond. The individual positions of the carbon are at this movement energy unequal,
and in this process the parameter depends on the fact, to what degree the two
outside carbons of the watched segment prevent each other from moving. From this
point of view, the most favourable position is the one where both carbons are in a
plane, but on opposite sides from the reference middle bond (so called trans-
position), whereas both carbons in a plane, but both oriented to the same side from
the reference bond, present the most energy demanding position, so called cis –
position. If one carbon is out of the plane and moreover put into the space where it
is not covered with any hydrogen of an opposite carbon, we talk about gauche
positions (right or left), and finally the carbons can be put in the space, where their
projection is covered with a hydrogen of an opposite carbon, and we talk about
positions 60 or 240 o. Energy contents for described basic positions of the
conformation round a simple bond are given in the table 1.4. Of course, there can be
positions in any point of a 360 o circle at the rotation round a bond. These positions
are energy between the outside values, given in the table.

Table 1.4. Conformations round a simple bond

angle, ˚ conformation kJ mol-1
0 trans 0
60 12
120 gauche right 3,4
180 cis 15
240 gauche left 3,4
300 12

It also follows from the table that at the temperature of absolute zero, when the
energy content of the system is zero, all conformations have to be with zero energy,
so trans and chains present ideally straight sticks. The higher temperature is, the
larger proportion of the conformations with higher energy contents is. The whole is
to be considered here as a dynamic system, where the conformation of each
individual segment changes constantly.

Note:
The cis configuration arises when substituent groups are on the same side of a
carbon-carbon double bond. Trans refers to the substituents on opposite sides of the
double bond (see the picture 1.2).

©Alexander Dubcek University of Trencin 2007 10

 Figure 1.2. Schematic representation of cis and trans structures

Stereoregularity is the term used to describe the configuration of polymer chains.

Three distinct structures can be obtained. Isotactic is an arrangement where all
substituents are on the same side of the polymer chain. A syndiotactic polymer
chain is composed of alternating groups and atactic is a random combination of the
groups. The following pictures 1.3 and 1.4 show two of the three stereoisomers of
polymer chain.

Figure 1.3. Isotactic polymer Figure 1.4. Syndiotactic polymers

The energy criterion is an important, but not the only one parameter that determine
proportions of individual conformations. Some syndiotactic polymers (PVC, 1,2-
polybutadiene), which have substituents alternately on both sides and they do not
hinder each other in the space, have the planar conformation. Further polymers with
the large proportion of the trans conformation at relatively high temperatures are
polyvinylalcohol and polycaprolactam, where hydrogen bonds appear significantly,
or polyethyleneterephtalat, where a certain departing in the ester group determines.
The alternation of trans and gauche right is typical for isotactic polymers
(polypropylene), where the spiral conformation is formed. This makes better
completeness of the space, if the spirals are put into pairs in the lattice, one in the
direction and the other against the direction. However, also in such case, the
completeness is less ideal than for polyethylene, therefore the density of PE crystal
is higher than PP.

The distance of ends of the chain is an important quantity for the characterization of
the chain arrangement of the polymer in the space. For the PE chain with the planar
zigzag configuration the maximal distance of the ends of the chain for the chain
with the number of bonds n is:

©Alexander Dubcek University of Trencin 2007 11

rmax = nlo sin α/2, (1.4)

where the valence angle is α = 109,5 o and the length of the bond is lo = 0,154 nm.

The distance of the ends of the chain at the free rotation of bonds is between 0 and
rmax (see the picture 1.5).

An important term in this context is the statistic knot that is defined as the most
probable form of a macromolecule in the conditions when no forces from its
surrounding affect it. The macromolecule tries to get into this form at disruption by
outside force. The statistic knot presents in given conditions the largest sloppiness
of the chain, thus maximal entropy.

Figure 1.5. Schematic representation of macromolecule form. a – in the most

probably conformation (rp – the most probably distance of the ends of the chain);
b – in the elongated form by activity of external force ( r – distance of the ends of
the chain)

The distribution of distances of the ends of the chain for a chain of a given length
may be calculated with help of statistic methods. The quantity characterizing the
linear size of the knot is the root of the quadratic distance between the ends of the
chain, and this value is proportional to the root of the number of bonds in the chain.

(r2)1/2 = aZ1/2 (the length of the bond and the number) (1.5)

The distance of the ends of the chain has importance only for linear molecules. If
we want to describe branched molecules that have more ends than only two, we use
other parameter that is the gyration radius. This is defined as

(s2)1/2

the root of the middle quadratic distance of individual parts of the macromolecule
from its centre of mass. The relation between the gyration radius and the distance of
the ends is for linear macromolecules in direct proportion with the constant of the
proportion 1/61/2. The branched chains have a smaller gyration radius, what is
indicated, for example, by determining of the molecular weight, when the viscosity
of the solution of the branched molecules is lower than of linear with the same
molecular weight.

©Alexander Dubcek University of Trencin 2007 12

1.4. Physical and phase states of polymers

1.4.1. Glassy state

As for all substances, we can define three basic states for polymers. That is
gaseous, liquid and solid state. Despite general similarities, there are certain
peculiarities in polymers and that makes it different from low-molecular
substances.

Gaseous state in general is characterized by intense particle motion where

distances between molecules are big. While mutual particle process in the sense of
attractive forces is declining with the sixth power of distance, we can practically
ignore mutual attractive forces with respect to the molecule distance for molecules
in gaseous state.

Thermal molecule motion in liquid state is still very intense, but molecule
distances are in contrast to gas substantially smaller and molecules influence
themselves greatly.

Intensity of thermal motion in solid state is so much decreased, that it is

insufficient for breakage of intermolecular contacts. Molecules assume stable and
defined position in space and they only perform vibrating motion with frequency
of 1013-14 Hz. Molecule distances are not very different than those in liquid body,
what can be easily proved by comparing of substance consistency in all three
states. Difference in consistency between gas and liquid is a number of orders, but
on the other hand difference in consistency between liquid and solid substance is
small. Yet, we can say that distances between particles are approximately on the
level of molecule size. The differences of molecule mobility imply substantial
differences in quality as well. Intense translatory motion of molecules, besides
rotary and vibratory, is typical for liquid state. Molecules in solid state assume
rather constant positions and translatory motion is limited to minimum. Whilst
liquid changes its shape easily by the process of minimal force, e.g. even by its
own weight (gravitational force), we usually need to exert quite heavy force for
deformation of hard solid.

When discussing arrangement in solid state we have to point at two possible

arrangements. On the one hand it is accidental contingent similar to liquid, with
substantially lower thermal motion of particles, and on the other hand it is
crystalline, where molecules are arranged regularly with visible symmetry
alongside spatial axes. We can assert that materials in solid state can exist in
various structural phases differing in the way of molecule arrangement, while it is
not always possible to detach these phases. In this case, we differentiate two
definitions of phase, and that is according to structural or thermodynamic point of
view. In the second case, the phases differ in thermodynamic parameters, they are
separated by clearly distinguishable boundary line, and they are separable.
Representative example of material containing various structure phases is
coexistence of crystalline state, characterized by regular particle arrangement into
crystalline grid, and hypothermic liquid state, where mobility of molecules is low

©Alexander Dubcek University of Trencin 2007 13

but their arrangement in space is accidental without any signs of symmetry. There
are 14 different crystal lattices, called Bravais Lattices. (3 different cubic types, 2
different tetragonal types, 4 different orthorhombic types, 2 different monoclinic
types, 1 rhombohedral, 1 hexagonal, 1 triclinic). See Figure 1.6.

Figure1.6. Bravais Lattices

.
At accidental arrangement in space we talk about so-called amorphous state (from
Greek „morphe = shape“, „amorphous = shapeless“). Phases of this material are
chemically identical but they significantly differ in arrangement. As the example
for phases defined from thermodynamical point of view we can give composite
consisting of two various materials, e.g. slight suspension of clay in water. Below
melting temperature of water is whole system in solid state with two
thermodynamically different components that are separated by strictly defined
boundary line and are mutually separable (e.g. filtration after ice melting). We can
give an example for polymers, too – composite of plastic material and inorganic
filler.

When discussing phase states, where various substances occur at various

circumstances, we do not take gaseous state into consideration, because
intermolecular coherence is much higher than covalent bond strength due to the
length of strings and resultant number of contacts. That is why string destruction
by thermal degradation happens sooner than particular molecules could be released
into gaseous state. Character of phase transition is in liquid and solid state more
difficult than it is for low-molecular substances. For this discussion, it is useful if
we do not consider whole macromolecules from the mobility point of view but we
divide them into string elements – segments, typical example is e.g. 12 to 60
carbons in the main string. Equilibrium position of segments is at low temperatures
constant and motion is restricted only to vibratory or rotary-vibratory oscillations
around equilibrium positions. Polymer acts in the same way as low-molecular

©Alexander Dubcek University of Trencin 2007 14

substance in solid state. At deformation application the whole system is following
Hook’s law, Young’s modulus of elasticity is high and material is fragile. While
behaviour of polymer is similar to glass qualities, we call it glassy state. Glassy
state is observed at low temperatures up to so-called temperature of glassy
transition. In the sphere of temperature of glassy transition Tg occurs qualitative
change of segment motion that changes to rotary in the sphere above Tg. The string
can acquire high number of various conformational shapes; material has lower
modulus of elasticity and behaves as highly elastic solid.

Figure 1.7. Stiffness vs. temperature dependence

We call it so-called highly elastic or rubberlike state. This state is typical for linear
polymers and beside polymers is unknown for any other materials. Rotary motion
of segments becomes more intense with higher temperature and finally it also
allows wandering of segments and later of whole macromolecules, too. When so-
called flowing temperature is achieved, polymer is in visco-elastic state and
irreversible flow occurs. Polymers are thermally plastically shaped above this
temperature.

These reasoning apply to disordered amorphous solid phase. However, many

polymers form crystalline phase whereby it almost without exception coexist along
with amorphous phase in the form of so-called semi crystalline materials. Yet,
crystalline phase acts as solid and above melting temperature is transformed into
liquid, visco-plastic phase when breakage of crystallites occurs. In terms of
temperatures when substantial transformations occur, we differentiate temperature
of
• glassy transition - Tg
• melting - Tm
• flowing – Tf,
whereby Tg < Tm ~ Tf.

1.4.2. Phase transitions

Transition of state, transformation of amorphous substance into crystalline and

vice versa, or transformation of one crystalline system into another, is called as
phase or state transition. Moreover, several phases can coexist next to each other in
thermodynamic balance. They are separated by clearly identifiable boundary line
and single-phase spheres of final dimensions. Internal energy, specific heat and
coefficient of thermal expansivity change on the boundary line. According to the

©Alexander Dubcek University of Trencin 2007 15

character of this transformation we talk of phase transitions of the first or second
order.

We determine phase transitions from dependencies of internal energy change and

temperature growth. They can be experimentally simplest determined from thermal
dependency of specific volume. Discontinuity can be observed at phase transition
of the first order at both dependencies. Typical phase transition of the first order is
melting or reversible process – solidification at melting/solidification point. At this
point, e.g. at melting, we provide heat to the system and internal energy is
increasing, temperature however does not change. After all the crystalline part is
melted, the temperature starts to rise again (but it is temperature of liquid, now).
We can also observe similar phenomenon when watching transformation of
specific volume with temperature, where substantial change of this parameter
without change of temperature occurs at melting point.

• Total enthalpy method to evaluate the degree of crystallinity:

It is assumed, that only two components exist and each of them has a
certain enthalpy, ha (amorphous) and hc (crystalline)

We can write
∆h21 − ∆ha 21
ω cl = (1.6)
∆h 0fl

h2 = h a2

ha 1
Enthal py

∆h21

∆ha2 1 ∆hf1

h1
hc2
hc1

T1 T2

Figure 1.8. Dependency enthalpy vs. temperature - total enthalpy method to

evaluate the degree of crystallinity

At these dependencies, we observe disruption instead of discontinuity at phase

transition of second order. Dependency of e.g. volume on temperature is linear,
when slope of a straight-line dV/dT is coefficient of thermal expansivity. At phase
transition is this slope changed, but dependency is continual, uninterrupted. In fact,
there are many cases when we observe bigger or smaller bend in thermal interval
of few degrees (mostly 5-15oC) instead of discontinuity when dealing with
polymers. Then, temperature of glassy transition can be specified by linear
extrapolation as the intersection point of two straight lines.

©Alexander Dubcek University of Trencin 2007 16

 Figure 1.9. Volume as a function of temperature.

The jump in volume is illustrated in Figure 1.9.; the slope of the line FC is the
thermal expansion coefficient of the crystalline phase; at the melting point the
volume jumps from C to B, and the higher slope of BA denotes the expansion
coefficient of the liquid phase. Some substances are, however, not able to
crystallize, for instance normal glass, as a result of a too irregular molecular
structure. When such a substance is cooled down from the liquid state, and follows
the line AB, then from B to D it still remains a fluid, which solidifies at D without
showing a jump in volume. The line then continues as DE, with about the same
slope as CF; the matter is, however, not in a crystalline condition, but in an
unordered, amorphous, glassy state, and has, therefore, a greater volume. The
transition at D is called the glass transition, occurring at the glass transition
temperature, Tg. It follows that Tg is always lower than the melting point, Tm. It is
very important to distinguish very carefully between Tg and Tm !!

Polymers are sometimes wholly amorphous and non-crystallisable; they then follow
the line ABDE. However, when such a polymer is heated up to above its Tg it is not
immediately transferred into a liquid state, but first into a rubbery state, which, upon
further heating, gradually passes into a fluid. Tg is, therefore, called the glass-
rubber transition temperature.

A more realistic representation of the phases is given by the stiffness of the

material, which falls down to zero when the liquid state is reached. This is
illustrated in Figure 1.9; for a low-molecular matter the E-modulus (a measure of
the stiffness) decreases to zero at the glass transition temperature (though more
gradually than for a crystalline substance at Tm). A polymer shows, above Tg, after
a decrease in E by a factor of 1,000 to 10,000, a rubbery region, which, on the
temperature scale, is longer as the chains are longer.

©Alexander Dubcek University of Trencin 2007 17

 Figure 1.10. Rubbery region with polymers.

The way, how the melt achieves solid state depends on structure of polymer string
and on the cooling rate. If the substance has asymmetrical molecules, then regular
arrangement could not be achieved and material will not crystallize. Amorphous
solid phase will originate at all terms. On the other hand, whole volume of
hypothermic melt will practically never crystallize even at maximal regularity of
polymer structure. After reduction of temperature on the level of crystallization
temperature, polymer segments will start to arrange into crystalline grid. Segments
that are already built into the crystalline grid have strictly defined positions and
their translation motion is stopped. Mobility of neighbouring segments is at the
same time decreased, they are crystallizing slower and possibly do not have any
chance to arrange into regular shape. They create amorphous phase, i.e. frozen solid
but non-crystalline structure. From this point of view is superposition of
crystallization rate and mobility of segments along with decreasing crystallization
temperature very interesting. The lower is crystallization temperature Tm the bigger
is driving force that makes segments arrange themselves into the crystalline grid.
Viscosity of hypothermic melt is increasing at the same time, but mobility of
segments is decreasing. Therefore, dependency of crystallization rate on Tm gets
beyond maximum, when driving force of crystallization is small at small
hypothermia, while it is high at high hypothermia, but mobility of segments
allowing them to arrange is low at the time. Finally, translation motion of segments
in Tg sphere is practically stopped and the part of material that did not crystallize is
transformed into glassy state. This idea implies that at sufficiently high
crystallization rate we can theoretically get any substance in the form of amorphous
glass. Another consequence of this process is the fact, that when cooling is not
infinitely slow, there are always areas thermodynamically balanced in the material.
This leads us to phenomenon of so-called secondary crystallization, when motion of
segments can be increased after temperature rise in certain area (deeply below
melting temperature) so much that it would allow additional crystallization.

1.4.3. Glass transition temperature

As we have already mentioned, temperature of glassy transition Tg is defined as

temperature, at which bend or discontinuity occurs on dependence of specific
volume on temperature. Disruption of thermal dependencies of thermodynamic

©Alexander Dubcek University of Trencin 2007 18

functions, such as enthalpy and entropy, occurs in the same thermal sphere. First
derivation of basic thermodynamic functions, thermal coefficient of volume dV/dT
and thermal coefficient of enthalpy dH/dT = Cp (thermal capacity) are changed
discontinuously. Thermal coefficients of transport qualities, such as viscosity,
diffusion of gases and stress relaxation are changed discontinuously, too, whereby
modulus of elasticity increases by few orders. Absorption of mechanical and
electrical energy reaches maximum.

There are several methods for the determination of glass transition temperature, in
addition to the use of changes in quality in Tg sphere. Tg can be determined e.g.
dilatometrically, by watching thermal expansivity. Other commonly used method is
calorimetric method, mostly differential scanning calorimetry, DSC. Dynamical-
mechanical analysis is another important method that is used for direct detection of
certain motion releases at gradual temperature increase, whereby the most
significant response is observed at motion release of main string segments that
correspond with temperature Tg.

Methods of Tg determination are related to important parameter, so-called free

volume. We can explain the conception of free volume by following reasoning.
Molecules in hard solid are not arranged totally tight, there are free vacancies
between them. These relate with freezing or restriction of motion in real time in
certain state. Thermal volume expansion occurs at temperature increase. Yet, single
volume of molecule almost never changes with temperature. Practically, only the
area between molecules is changing and mobility of molecules increases at the same
time. According to Cohen and Turnbull theory, redistribution of size of vacancies in
liquid occurs relentlessly, whereby molecule can skip only to the space with at least
particular minimal volume. Skip frequency is determined not by energetic factors,
but by probabilistic factors, where critical volume is very important quantity.
Probability that certain molecule will skip to the other place is determined by
probability that there is large enough vacancy in its vicinity.

Thermal volume expansion is caused by the growth of free volume. Probability can
be expressed by quantity f as the quotient of occupied and unoccupied space.
Dependence of f on temperature is linear

f = αf (T - T∞) (1.7)

where T and T∞ is system temperature and reference temperature and αf is

coefficient of thermal expansivity. On basis of these presumptions we can infer
theorems for diffusion coefficient D

D = D∞ exp (-1 / αf (T - T∞) (1.8)

or other quantities, e.g. viscosity η

η = η∞ exp (-1 / αf (T - T∞) (1.9)

Alternative to probabilistic theories are energetic theories. Skip of liquid particle is

according to them possible only when the particle has certain superfluous energy

©Alexander Dubcek University of Trencin 2007 19

that will allow it to overcome the energetic barrier. Thermal dependencies are two-
parameter and have the form of Arrhenius equation

D = D∞ e-Ea/RT (1.10)

where Ea can be considered as activation energy or thermal coefficient of particular

process. At the same time we have to realize that total energy of system depends on
temperature, but energy of every particular element (or moving segment in the case
of polymers) is not identical with other particles and it is neither constant. Particles
(segments) collide with each other and they transfer energy. Accidentally, some
particle can gain energy that would suffice to break energetic barrier created by
neighbouring particles. Rearrangement – translation motion will happen in such
case.

At temperature T = 0 K, diffusion coefficient decreases to zero and viscosity gains

infinite value.

In 1921, Vogel, Fulcher and Tamman inferred important empiric equation during
study of inorganic glass viscosity. It allows direct calculation of melt viscosity η for
any temperature T, if constant B and viscosity η∞ was experimentally determined at
reference temperature T∞

logη = logη∞ + B / (T - T∞) (1.11)

In confrontation of theorems (1.7) and (1.11) we can see that

αf = log e / B. (1.12)

Williams - Landel, – Ferry in 1953 proposed another crucial semiempirical

equation. They have chosen Tg for reference temperature and ηg is viscosity at Tg.
This equation has form

log η / ηg = -c1g (T - Tg) / c2g + T - Tg (1.13)

and is so important that it is called WLF equation.

WLF equation is a law for determining frequency temperature equivalence (which

holds true for a given range). To have an approximate idea, it can be considered
that, in low frequencies (from 10 to 105 Hz), an increase in frequency by a factor of
10 has the same effect on the behaviour of the rubber as a 7 to 8°C drop in
temperature. For example, an elastomer with a glass transition temperature of -20°C
at 10 Hz will have a glass transition temperature of about +10°C at 105 Hz.

©Alexander Dubcek University of Trencin 2007 20

 Modulus

3200

1600

800
Glassy Rubbery
state state
400

200

100

-40 -30 -20 -10 0 10 20 30 40 o

Temperature C
Glass transition temperature (Tg)

Figure 1.11. Modulus vs. temperature dependency for 10 Hz

The above graph is plotted for a frequency of 10 Hz. Using the WLF equation, the
graph can be calculated for other stress frequencies (see below).
Modulus

3200

1600

Glassy
800 state
Rubbery
state
400

200

100

-40 -30 -20 -10 0 10 20 30 40 o

Temperature C
Glass transition temperature (Tg)
Figure 1.12. Modulus vs. temperature dependency for 105 Hz

For any given rubber, the glass transition temperature increases with the stress
frequency, which moves the vitreous state towards higher temperatures.

©Alexander Dubcek University of Trencin 2007 21

 Frequency
log

8 Glassy zone

Tg
of
line
6

Rubbery zone
0

o
-50 0 50 100 150 T in C
Figure 1.13. Frequency in log vs. temperature dependency

According to WLF equation, we can calculate viscosity η at temperature T on the

basis of known viscosity ηg at temperature Tg, where c1g and c2g are empirical
parameters.

WLF equation can be inferred from Vogel’s equation in such a way that we
substitute coordinates of reference point of ηg a Tg for basic parameters. After
substitution

c1g = B / Tg - T∞ (1.14)

and

c2g = Tg - T∞ (1.15)

we get WLF equation. Constants c1g and c2g are in this case almost truly
independent of temperature if we use Tg as reference temperature. Universal
(average) values of the constants are c1g = 17.4 and c2g = 52 K.

According to previous equations (1.14) and (1.15) we can express

αf = 1/2.3 c1gc2g (1.16)

and

fg = 1/2.3c1g (1.17)

When applying universal constants c1g and c2g we get universal values αf = 4.8 x 10-
4
K and fg = 0.025. At the same time we watch good agreement of universal αf with

©Alexander Dubcek University of Trencin 2007 22

experimental values measured for many polymers. From definition of fg results that
free volume at Tg is 2.5 % from total volume of the solid, that is 97.5 % is occupied
by the mass of molecules. When we raise temperature 100 oC over Tg, parameter fg
will get value of 0.075, thus free volume will raise to 7.5 %.

We can interpret this result in such a way that under Tg the free volume is
practically constant. In such case, measured thermal expansivity of material in
glassy state will be only rate of single molecule expansivity. On the basis of
Simha’s and Boyer’s reasoning the molecule volume Vm at absolute zero can be
achieved by the means of extrapolation from the liquid sphere.

As we have said before, Tg is formally the transition of II. order. However, this
statement is not fully accurate from thermodynamic point of view, because
measuring of thermodynamic quantities around Tg are not balanced, but time factor
plays important role here (rate of temperature change) and history of specimen
(preparation method). In this regard, Tg cannot be considered as thermodynamically
defined transition. Result is the fact that experimental determination of Tg of
particular polymer depends on used method and to a certain extent on experiment
conditions. That is why values of Tg in professional literature sometimes differ by
tens of degrees.

Thermodynamic theory presented by Gibbs and Di Marzio states that balanced

transition really exists and it lies somewhere around Vogel’s temperature T∞.
Though, these measurements are practically unrealisable, because they would
require infinitely slow temperature changes. Quite reliable picture could be got by
the use of lattice model. Conformational energy of the system and conformational
entropy can be calculated with its help. Number of realizable macro conformations
is decreasing along with temperature decrease and energetically low conformations
are understandably dominant. All macro conformations at limiting temperature of
absolute zero in perfectly frozen state would be trans, number of possible states
would be Figure 1.17 and conformational entropy would be zero. Real systems
freeze some tens of degrees above this temperature, when number of
conformational states is low and transitions are rare.

Figure 1.14. The conformational forms of n-butane

©Alexander Dubcek University of Trencin 2007 23

 Figure 1.15. The conformation energies along the middle carbon bond

References:

[1] http://plc.cwru.edu/tutorial/enhanced/files/polymers/apps/apps.htm
[2] Holzmüller, W., Altenberg, K.: Fyzika polyméru, SNTL, Praque 1966
[3] http://pslc.ws/macrog//tg.htm
[4] http://www.missouri.edu/~crrwww/katti/Thermal%20Behavior%20of%20Polymers.pdf
[5] http://www.udel.edu/mse/class/Opila/804/Class%20O.ppt

Fundamental questions from present part:

1. Explain the difference between terms “macromolecule” and “polymer”?

2. Write down law for density of cohesive energy and explain what present
the individual symbols present. Explain physical significance of the used
symbols.
3. Explain the basic types of polymer chains.
4. Explain the difference in cis and trans structure of polymer chains.
5. Describe the molecule motion in every state, liquid, gas and solid phase.
6. Describe the amorphous and crystalline state of the material.
7. What is characteristic for the melting point?
8. Explain the term: free volume.
9. Explain molecular kinetic on the base of free volume theory.
10. Explain phase transitions of the first and second order.
11. Describe the stiffness vs. temperature dependency for polymeric
material.
12. What is characteristic for glass transitions in polymers?

©Alexander Dubcek University of Trencin 2007 24

 CHAPTER 2
Mechanical properties of solid state polymers

Objectives to achieve
In this part we are going to deal with a description of deformation of solid elastic
materials as well as to set up phenomenological values describing mechanical
properties of these substances. We are going to learn about thermodynamic and
microstructural aspects of the process of elastic deformation. Later on these
concepts and knowledge will be generalized on polymers and rubber.

2.1. Description of deformation of solid elastic materials

Deformation of solid elastic materials can be found in the change of relative

positions of their atoms and molecules as a result of external forces. For this
reason, appropriate description of this phenomenon should issue from the
calculation of displacement of individual atoms compared to state without the
effect of external forces. On the base of practical reasoning, however, we imagine
the substance as a continuum, we describe particular deformations with the help of
appropriately applied elastic constants, modulus of elasticity and until then we try
to put these constants into continuity with atomic structure of the substance and
with bonding between individual atoms. Concerning “direction” properties of
bonds we can expect that crystals are in general deformed anisotropically and for
that reason their deformation properties will be described by tensors (strain tensor,
stress tensor, and so on).

However we are going to focus our attention on a description of deformation of

isotropic substances because - in majority of cases regarding rubber and polymers -
we are able to manage with this simplified description.

Let’s analyze deformation curve well known from the study of mechanical
properties of metallic materials. To consider deformability of solid elastic materials
we use the term stress ( σ ), that is the force affecting the unit surface of solid
elastic materials and strain ( ε ) modified by elongation proportion ( ∆l ) and primary
length (l0). The most commonly observed stress - strain dependence has a shape of
curve shown in figure 2.1.

©Alexander Dubcek University of Trencin 2007 25

 σ4

σ3
σ2
σ1

ε
Figure 2.1. The stress - strain dependence for elastic material

Gedde:

Potential energy is
U = C (r − r0 ) 2 , (2.1)

where C = constant and r0 = equilibrium bond length

Elastic force is
(2.2)
∂U
f= = 2C(r − r0 ) .
∂r
The stress is (Hookes law)
2C (2.3)
σ= (r − r0 ) = Eε ,
A

©Alexander Dubcek University of Trencin 2007 26

where A = cross-sectional area, ε = strain

This dependence has few characteristic areas. First area is one of linear dependence
where strain is directly proportional to stress. The prolongation itself is trivially
small (<0,01%). When the action of external force is brought to an end,
material returns to its previous state. The borderline of this area is limit of
proportionality ( σ1 ). Within the influence of higher stress than the limit of
elasticity, the material still deforms elastically, even though not linearly.

The initial state is renewed (resiled) when deformation is over. The borderline of
this area is called elastic limit ( σ 2). With the further stress increase an interesting
state occurs where deformation continues spontaneously even though the stress does
not increase any more. This phenomenon is called ductility and the beginning of
this area in graph is called the yield point ( σ 3). During ductility of deformed material
deformation caused by the stress needs to be raised. Finally when the ultimate strength
( σ 4) is reached the material tears apart. It is not difficult to explain qualitatively
presented behavior in substances on the base of their microscopic properties. Each
external force affecting crystal has a tendency to move against each other the whole
atomic plane (Figure 2.3), that is to eject each separate atom from the acting area of one
adjacent atom into the area of the other one.

Figure 2.3. The shift of atomic planes by deformation

27

Figure 2.4. The beginning of „slide“

First area in Figure 2.1 (stress below σ1) is characterized by the situation in
which atoms behave as harmonic oscillators – the force necessary to eject
atoms from state equilibrium is directly proportional to deviation because in
the environment of maximum of total potential energy the relativity U = U(r) is
approximately parabolic. When higher stress linearity stops to be valid but

©Alexander Dubcek University of Trencin 2007 27

deformation is still elastic up to the area equivalent to the situation shown in Figure
2.3. Corresponding angle is approximately 27°. Then it is not necessary to increase
the stress deformation itself happens by „slide" of individual planes. That is how the
sphere of ductility originates. Related stress ( σ 3) can be considered the measured
ultimate strength of material. On the base of well known results we can try to obtain
quantitative estimation of ductility stress. This stress is approximately determined
by the force, needed for releasing, all atoms of unit plane from bonds (Np) by their
displacement (d). Pursuit of this force ( σ 3d) has to be equal to the bond of energy Np
of atoms (Np Ev), so for ductility stress we will reach a simple relation of

Np E v
σ3 = (2.4)
d

If we enter appropriate data (N p ~10 1 9 m -2 , E v =3 eV and d~0,2.10 - 9 m) we will

reach the value of σ3 = 102 N m-2 (more accurate calculation e.g. for iron has the
value of 7 ⋅ 107 N m-2). However, the fact is that measured values of ductility
stress are at least three orders smaller. Which phenomenon causes that the
material starts to “leak” by such low stress? It is definitely proved that reasons
for this are dislocations.

Figure 2.5. The edge dislocation

We can imagine the edge dislocations as finalized line of atoms inside the crystal.
Viewing this defect direction of axle the situation (in plane) is displayed in Figure
2.5. Star (*) indicates line of the edge dislocation. It can be noticed that by the
influence of the external force the whole line of atoms close to dislocation line can
move this way by interatomic distance by which dislocation itself transfers via
opposite direction. If dislocation flows this way throughout the whole crystal cross-
section, one whole atomic plane concerning the other one „slides“ the interatomic
distance. Instead of simultaneous interruption of the atomic bonds in the whole
plane, in this case, it was enough to interrupt the bonds just in single lines of atoms.

That is why in order to create ductility when dislocation is present it is sufficient to

use lower stress. In this case we talk about dislocations that, by their movements,
enable light deformability of material, its plasticity and considerable decrease of
ductility limit regarding to theoretically calculated value.

©Alexander Dubcek University of Trencin 2007 28

If the listed reflections are true then it must be possible to reach the increase of
metal strength by preparation without dislocation. These crystals can be produced
only through high technology and only in very small contents.

Crystals produced in this manner are truly reaching theoretically calculated strength
(iron wire with diameter of 1 mm can sustain the weight up to 1 000 kg).

Real crystals always contain dislocations and by their straining more dislocations
are generated (by Frank—Read mechanism).

But when their concentration reaches the value of ~102m-3, they start to interfere
when moving. This way the possibility of light strain decreases. The crystal is
strengthened as a consequence of redundancy of dislocations. There through the area
between σ 3 and σ 4 in figure 2.1 can be naturally explained. The knowledge - that
through the restrain of dislocation movement it is possible to strengthen the crystal -
is practically used. In polycrystalline structure the dislocation mobility is obviously
smaller than in crystalline structure and for this reason these substances should have
higher strength. It has been really approved. But in practice the most commonly
used possibility is a different obstruction of dislocation movement - deliberate
application of small amounts of suitable elements into crystals. These atoms
obstruct dislocations movement very effectively and that is why for example iron with
little amounts of carbon, chrome, magnesium and wolfram have vastly higher strength.

In the following text we will impose parameters, which will be further used. Force
acting on steady based body can be always distributed in two components (see
Figure 2.6) normal (Fn) and tangential (Ft).

Fn F

Ft

Figure 2.6. The force decomposition

©Alexander Dubcek University of Trencin 2007 29

We define relevant strains by the following relations
F
σ⊥ = n , (2.5)
S

Ft
T = σ tan = . (2.6)
S

Normal force components cause compression deformation, tangential force

components (strain) cause shear deformation. To characterize them we define
strain ε by the following relation

∆l l − l 0
ε= = . (2.7)
l0 l0

Where l0 is the material length before the strain effect and l is the material length
after strain effect (Figure 2.7. a).

a)
u

d
γ

b)
Figure 2.7. Definition of the longitudinal a or shear deformation b

Relative change of primarily right angle ∆γ by shear deformation (concerning the

validity

tg γ = γ for small angles) defines the relation (Figure 2.7. b).

u
γ= . (2.8)
d

Where u is a slide of upper base considering the lower one by shear deformation
and d is a distance between bases = material thickness. According to Hooke’s law
parameters ε and σ⊥ or γ and σtan are mutually proportional hence if we use
relations for particular constants E and G we will gain relations for new length

©Alexander Dubcek University of Trencin 2007 30

calculation in tension (compression) of stressed rod, eventually slide of upper base
considering the lower one by shear deformation

⎛ σ ⎞
l = l 0 ⎜1 + ⊥ ⎟ , (2.9)
⎝ E ⎠

σ tan
u=d (2.10)
G

Cross reduction eventually cross extension is related to prolongation eventually

shortening of the rod. Relative cross reduction η is defined the following
proportion

a0 − a
η= , (2.11)
a0

where a0 is initial rod thickness and a is final rod length. Because a is proportional
to relative prolongation, the following relation for new thickness of deformed rod
is valid

⎛ σ ⎞
a = a 0 (1 − η) = a 0 ⎜1 − ⎟. (2.12)
⎝ mE ⎠
In isotropic substances within the area of elastic deformation, the relative length
change is directly proportional to stress, which evokes it, in accordance to fixed
terminology and nomenclature. On the base of Hooke’s law for strain ε and for
rectangular stress σ⊥ the following relation is valid

σ ⊥ = Eε . (2.13)

Where E is Young’s modulus of elasticity in tension.

By shear deformation for relative change of initially right angle γ and tangential
strain σtan the similar relation is valid

T = Gγ . (2.14)

Where G is shear modulus of elasticity in shear.

Relative cross reduction η is directly proportional to relative extension

ε = mη . (2.15)

Where m is Poisson modulus.

©Alexander Dubcek University of Trencin 2007 31

Poisson’s number ν is inverted value of Poisson modulus so we use

1
ν= . (2.16)
m

We usually choose E and ν as constants characterizing elastic properties of

isotropic material. There is a mutual relation among three elastic constants E, G and
m expressed as follows

m E
G= (2.17)
2 (m + 1)

To understand relation of phenomenological values with material microstructure we

have to clarify the coherency of elastic constants with microprocesses that take
place in substances during their deformation. The exact calculation is very
complicated and therefore it is sufficient to use only informative estimation based
on the following reflection.

We take in account diatomic layer of atoms placed on unit plane distant from each
other about interatomic distance a. The amount of atoms in one line is N = 1 / a2.
Force F acting on one of the sets (equal to numeric strain ν) causes relative
extension of our two-layer crystal λ = x / a , hence modulus of elasticity in tension
is defined by relation

ν a
E= =ν . (2.18)
λ x

In the area of elastic deformation the force deflecting atom from equilibrium
positioning is proportional to deviation i.e. fa = k′ x, therefore stress ν is designated
by N-multiple of this force and consequently

a k ′a k ′
E = k ′xN = = . (2.19)
x a2 a

The biggest problem is the estimation of k′ constant. We can use the knowledge that
the potential energy of atom oscillating in line corresponding to mean amplitude xs
(i.e. k′ xs 2 /2) is equal to kT, where k is Boltzmann´s constant.

2k T
k′ = . (2.20)
x s2

When the average amplitude xs is expressed as the definite multiple of interatomis

distance (α a) we will obtain the relation

2k T
E= . (2.21)
α 2a3

©Alexander Dubcek University of Trencin 2007 32

By room temperature is k T ≈ 0.03 eV and from the other measurements it is known
that mean amplitude of oscillations at this temperature is from 2 to 4% of
interatomic distance. For this reason for modulus of elasticity in tension we will
reach the value of approximately 1011 N m-2 according to table 2.1. This
corresponds to the measured values. To illustrate we present values of modulus and
interatomic distances for three metals in the following figure 2.1.

Table 2.1. Interatomic distances (a) and Young’s modulus (E) [2]

Substancea [nm]E[N m-2]

Aluminum 0,404 0,73 * 109

Copper 0,361 1,29 * 109
Iron 0,286 2,16 * 109

Sometimes we also proceed the opposite way. From measured value of modulus E
the amplitude of thermal oscillations is deduced.

Note
Elastic properties of isotropic and homogeneous materials are designated by two
Lame’s constants λ and µ, which are related to elastic constants cαβ by relations
c − c 12
c 11 = λ + 2µ, c 12 = λ, c 44 = 11 = µ . A significant parameter that characterizes
2
behavior of solid body is also compressibility stated by
∆ 3(1 - 2 ν )
relation χ = - = s iikk = , where ∆ is a change of volume caused by
p E
hydrostatic pressure p and siikk is the components of yielding tensor (matrix of
yielding tensor values is inverse to matrix of elastic constants’ tensor c αβ ). Inverse
E
value of compressibility is called modulus of volumic elasticity Χ = .
3(1 - 2 ν )

In Hooke's law (with the elastic modulus tensor Cijkl we sum over k and l, but, due
to the constraint, the only strain component which is non-zero is ε11.
σij = Cijkl εkl = C1111ε11 + C1122ε22 + C1133ε33 = C1111ε11,
so the effective stiffness for constrained compression is C1111.

Let us find the physical significance of that tensor element in terms of engineering
constants.

One may also work with the elementary isotropic form for Hooke's law.

εxx = (1/E) {σxx - νσyy - νσzz}

εyy =(1/E) {σyy - νσxx - νσzz}
εzz = (1/E) {σzz - νσxx - νσyy}

©Alexander Dubcek University of Trencin 2007 33

For simple tension or compression in the x direction, the Poisson effect is free to
occur. There is stress in only one direction but there can be strain in three
directions. σxx ≠ 0, σyy = 0, σzz = 0. Then (σxx / εxx) = E.
So Young's modulus E is the stiffness for simple tension, with the Poisson effect
free to occur.

Consider constrained compression, with εyy = 0, εzz = 0. Then

σyy = νσxx + νσzz.
σzz = νσxx + νσyy.

Substituting,
σyy = σzz = σxx ( ν(1 + ν)/(1 - ν2)) .

So, substituting into Hooke's law, the stress-strain ratio for constrained
compression, which by definition is the constrained modulus C1111, is (σxx/εxx) =
C1111 = E ((1 - ν) / (1 + ν) (1 - 2ν)).

The physical meaning of C1111 is the stiffness for tension or compression in the x (or
1) direction, when strain in the y and z directions is constrained to be zero. The
reason is that for such a constraint the sum in the tensorial equation for Hooke's law
collapses into a single term containing only C1111. The constraint could be applied
by a rigid mold, or if the material is compressed in a thin layer between rigid
platens. C1111 also governs the propagation of longitudinal waves in an extended
medium, since the waves undergo a similar constraint on transversedisplacement.

Rubbery materials have Poisson's ratios very close to 1/2, shear moduli on the order
of a MPa, and bulk moduli on the order of a GPa. Therefore the constrained
modulus C1111 is comparable to the bulk modulus and is much larger than the shear
or Young's modulus of rubber.

Practicalexample -corkinabottle.
An example of the practical application of a particular value of Poisson's ratio is the
cork of a wine bottle. The cork must be easily inserted and removed, yet it also must
withstand the pressure from within the bottle. Rubber, with a Poisson's ratio of 0.5,
could not be used for this purpose because it would expand when compressed into
the neck of the bottle and would jam. Cork, by contrast, with a Poisson's ratio of
nearly zero, is ideal in this application.

Practical example - design of rubber buffers.

How does three-dimensional deformation influence the use of viscoelastic rubber in
such applications as shoe insoles to reduce impact force in running, or wrestling
mats to reduce impact force in falls?

Solution
Refer to the above analysis, in which deformation under transverse constraint is
analyzed. Rubbery materials are much stiffer when compressed in a thin layer
geometry than they are in shear or in simple tension; they are too stiff to perform
the function of reducing impact. Compliant layers can be formed by corrugating the
rubber to provide room for lateral expansion or by using elastomeric foam, which
typically has a Poisson's ratio near 0.3, in contrast to rubber for which Poisson's

©Alexander Dubcek University of Trencin 2007 34

ratio can exceed 0.49. Corrugated rubber is used in shoe (sneaker) insoles and in
vibration isolators for machinery. Foam is used in shoes and in wrestling mats.

Practical example - aircraft sandwich panels.

The honeycomb shown above is used in composite sandwich panels for aircraft.
The honeycomb is a core between face-sheets of graphite-epoxy composite. Such
panels are usually flat. If curved panels are desired, the honeycomb cell shape must
be changed from the usual regular hexagon shape, otherwise the cells will be
crushed during bending. Several alternative cell shapes are known, including those,
which result in a negative Poisson's ratio.

2.2. Thermodynamic aspects of deformation

Let’s look at the problematic of deformation from thermodynamic point of view.

Every thermodynamic system is described by state values that are typical by the
values that depend only on initial or closing state of the scale and they don’t
depend on the approach through which the scale got from initial to closing states.
For this reason, for example, the work isn’t state value because it can be realized in
the system isobarically (∆p=0,), isothermally (∆T=0), etc. Some state values are:
absolute temperature (T, [K]), pressure (p, [Pa]), volume (V, [m3]), entropy (S,
[J/K]) and internal energy (U, [J]). The meaning of the first three values is obvious.
Entropy as a state value characterizes orderliness of the system and it is a non-
decreasing function of time in insulated systems. Consequently it can only increase
or be constant what can be interpreted by the following words: closed systems are
spontaneously degenerated because they lead to chaos. Inner energy, concerning
ideal gas, is determined by kinetic energy of the molecules because the potential
energy is neglected.
Except of these basic state values it is sometimes convenient to impose so-called
thermodynamic potentials that have their specific meaning when describing
thermodynamic actions. These thermodynamic potentials contain the basic state
values and so they are state values themselves. These are: enthalpy H(p,S),[J], (
H=U+pV) that is suitable for action’s description by constant pressure, free energy
F(V,T),[J], (F=U-TS) that is suitable also for action’s description by constant
temperature and sometimes it is called Helmholtz´s potential. To determine the
state of equilibrium we use Gibson’s free energy G(p,T),[J], (G=U+pV-TS) which
is a minimal in the state of thermodynamic equilibrium.

Thermodynamic processes follow two basic equations

dQ = dU + pdV,
dU + pdV (2.22)
dS = .
T

©Alexander Dubcek University of Trencin 2007 35

First equation is called “the first thermodynamic law” and we can understand it as
law of thermal energy conservation. The second equation is a thermodynamic form
for entropy.

On the base of these definition relations for state values we can get formulations
for values that characterize process of deformation or let us say formulations of
state values by means of thermodynamic potentions through simple mathematical
modifications that won’t be done here.
Equations for compression and entropy

∂ F
= -p,
∂ V T (2.23)
∂ F
= − S .
∂ t V

Here after small modifications we obtain useful relation

∂ S ∂ p
= . (2.24)
∂ V T ∂ T V

We can express the coefficient of thermal expansion α as follows

1 dV 1 ⎛ ∂ G2
⎞
α = . P = . ⎜⎜ ⎟⎟ . (2.25)
V dT V ⎝ ∂ p ∂ T ⎠
Coefficient of compressibility is determined by

1 dV 1 ⎛ ∂ 2
G ⎞
χ = - . = - .⎜ ⎟. (2.26)
V dP V ⎜ ∂ 2
p ⎟
T ⎝ T ⎠
Specific heat by constant pressure is
dH
cp = . (2.27)
dT p

Above mentioned values are connected into mutual relation by Ehrenfest´s

equation
∆αVT dT
= , (2.28)
∆c p dp
or
∆χ dT
= . (2.29)
∆α dP

©Alexander Dubcek University of Trencin 2007 36

In conclusion we would like to mention one useful equation to describe
compression. This equation is reached by partial derivation of the first
thermodynamic law on the base of V and with the use of relation (2.23) and (2.24).
∂U ∂p
p=- + .T . (2.30)
∂V T ∂T V
This equation will be later used to derive the basic mechanical properties of rubber.
From the equations mentioned above results that the process of deformation is
closely related to thermodynamic state of mechanically stressed system and
parameters that described it. As we will see later, for polymer materials the
situation describing deformation is more complicated because above mentioned
equations are valid for the state of thermodynamic equilibrium.

References

[1] http://silver.neep.wisc.edu/~lakes/PoissonIntro.html
[2] Koštial, Pavol: Fyzikálne základy materiálového inžinierstva I - Žilina: ZUSI,
2000, ISBN 80-968278-7-1

Fundamental questions from present part:

13. What are the components of crystalline materials deformation? How

would you describe the cause of elastic-reversible deformation from the
point of view of elastic substance microstructure?
14. What causes plastic - irreversible deformation in crystals?
15. Write down Hook´s law for isotropic homogeneous solid deformed in
tension or compression. Explain physical significance of the used
symbols.
16. Write down law for shear deformation of isotropic homogeneous solid.
Explain physical significance of the used symbols.
17. In which SI scale units are modules E and G measured and what do they
determine?
18. Explain meaning of Poisson’s number and Poisson’s modulus according
to physics. What are their mathematical formulations?
19. What is physically determined non-dimensional deformation γ
presented in relation 2.5.?

©Alexander Dubcek University of Trencin 2007 37

 CHAPTER 3
Viscosity and mechanical properties of viscous and
viscoelastic materials

Objective to achieve
This chapter introduces the viscosity terms and viscoelastic behaviour of
materials under loading. There is presented the terminology of complex
physical parameters (modules), WLF transformations and results of rubber
mixtures measurements and temperature dependence of viscosity.

3.1. Viscosity

We are going to deal with peculiarities of viscoelastic materials, to which both

rubber and polymers belong. We will base on the terminology and the indication of
quantities established in the part 3.1.

The term viscosity is mostly associated with the liquid and gaseous state. The
viscosity is a physical phenomenon caused by Van der Waals’s forces acting among
particles of liquid and gas while they are moving. If the movement is only of
“sliding” character, then, as we already know from the basic course of physics, the
Newton’s viscous law is applied in the following form:

dν
T=η , (3.1)
dr

where T is a tangential tension, η is a dynamic viscosity and v is a flow speed. So

we can see that in the material there must be a gradient of the speed of the flow of
the consisting particles for using viscous processes. The liquids, whose behaviour is
determined by the equation 3.1, we call the Newton’s. However, the real liquids are
different from the Newton’s. Also water as relatively less viscous liquid behaves
“elastic”, if we, for example, apply high tensions on it in short impulses. We know
this phenomenon very well, for example from “unsuccessful jumps” into water,
when we fall on our belly (elastic reaction of the water is sufficiently perceptible).
The time development of mechanically loaded viscous environment also depends on
the duration of acting of the load.

In previous considerations, we have dealt only with properties of elastic materials.

Real materials, however, always appear within defined conditions as viscous as
well. These properties arise from the very character of materials, which are in
metastable state (gas, some high-polymeric substance). This phenomenon of the
flowing is possible to watch for glass on windows of old churches that run
spontaneously under influence of external conditions, what is proved in that way,
that glasses are thicker at frames than in the middle.

Deformation caused by impressed mechanical tension to steel cause a change of

inter-nuclear distances, but in plastic materials there is mainly a change of the
valence angle.

©Alexander Dubcek University of Trencin 2007 38

We call the changes in the form of plastic materials as flow. These processes (flow)
include the late elastic deformation, the constant flow, the gradual return into the
previous form, the returnable deformational process after heating, and so on.

We always watch the flow of material within tensile tests. We watch the constant
dγ
speed of the flow especially for big loads. In such cases after interrupting the
dt
load of material, a part of deformation remains, therefore we talk about the non-
returnable flow. If such process is running under low tensions, we say that the
material is not plastic or it does not have malleable solidity. However, if some
tension is needed for the induction of the flow, we say, that the material is plastic.

Figure 3.1. Dependence of speed of shear deformation on the impressed

tangential tension for the Newton’s liquid (a), the non–Newton’s liquid (b),
material with malleable solidity (c)

For better clearness, we have plotted the speed of a change of originally right angle
γ of a prism, which is affected by tangential tension as a function of impressed
mechanical tension for various types of material, on the Figure 4.1. The Newton’s
materials have the deformation speed described with the relation:

dγ T
= . (3.2)
dt η

Also glassy and hard plastic materials often behave as the Newton’s liquids.

For materials that have non-linear behaviour of the deformation speed from the
tension, it is applied:

dγ T n
= , (3.3)
dt η
or
dγ T
= (3.4)
dt η(T )

and we say that materials have the structural viscosity.

©Alexander Dubcek University of Trencin 2007 39

The following relations for speed are used for plastic materials like crystals, clays,
suspensions and dough:

dγ ⎛ T - T0 ⎞
=⎜ ⎟, (3.5)
dt ⎜⎝ η ⎟⎠
or
dγ (T - T0 )
n
= . (3.6)
dt η

It follows from above mentioned interpretation that the classical theory of elasticity
is in the case of polymers applicable only for low and short-term acting tensions.
According to above mentioned facts, we can draw the following comparison
between ideally elastic and ideally viscous materials (T, σ are tensions, ε is relative
dγ
deformation, S' = is the deformation speed, E is Young’s model, η is dynamic
dt
viscosity).

The Hook’s law Newton’s law

σ = εE , T = S'η (3.7)

In order to understand better the differences between metals, plastic materials and
rubber, we analyse graphs in the Figures 3.2 - 3.4.

Elastic material responds to the step change of tension in the way that after its
finishing the value of relative deformation returns to the previous value, when the
material had not been deformed. The residual deformation remains in viscous
material also after interrupting the acting of the tension pulse.

Let’s explain closer the particularities of viscoelastic behaviour of plastic materials

and rubber. We will follow from behaviour of the ideally elastic and ideally viscous
materials.

We will choose a simple example for this purpose. The tested body is put to the test
under constant shear tension σ of the size σ0 and we measure the shear deformation
γ within isothermal conditions (see the Figure 2.7).

©Alexander Dubcek University of Trencin 2007 40

 σ
σ0

Figure 3.2. The course of the shear deformation ε for ideally elastic materials
under constant tension σ

We can see that the shear deformation equivalent to ideally elastic material is
settled immediately after inducing the shear tension (Figure 3.2). Instead of the
shear module G0, we use further malleability in the skid I0:

1
I0 = , (3.8)
G0
1
and so instead of the equation γ = T0 we get the relation
G0

γ = l 0 T0 . (3.9)

For ideally viscous materials, as for example thick oils are, the same experiment
does not lead to the constant shear deformation, but to constant speed of the shear
deformation γ ' proportional to T0, so γ rises linearly with time.

1
γ= t T0 , (3.10)
η0

where η 0 is viscosity (see the Figure 3.3)

©Alexander Dubcek University of Trencin 2007 41

 σ
σ0

Figure 3.3. The course of the shear deformation γ for ideally viscous materials
under constant shear tension T

The individual layers of condensed systems do not rub against each other without a
friction. On the contrary, if we make these materials flow, it means move the
individual layers against each other, tangential tensions arise among them.

Now we are going to demonstrate, how the shear deformation affects plastic
materials and rubber. If constant shear tension T influences the sample of polymer
or rubber, the result is the shear deformationγ, which is dependent on time by more
complicated way than ideally elastic and ideally viscous materials are. It usually has
a course drawn in the following Figure 3.4a.

The exact mathematical analysis of recorded results shows that the shear
deformation, as the one drawn in the Figure 3.4b, is possible to divide into three
parts. Apart from the parts of the ideally elastic γe,

γ e = J S' T0 (3.11)

©Alexander Dubcek University of Trencin 2007 42

 γp
σ

σ0
t
t γv

γ
t

γr

a) b)
Figure 3.4. The experimental time dependence of the shear deformation a)
Divide of the deformation into elastic, viscous and relaxation parts b)

and the ideally viscous γv,

1
γV = t σ0 , (3.12)
η0

also the relaxation part γr comes out. It presents a transitional function, and in the
simplest case, we can describe it with the relation

γr = αk 1 − e ( t / τk
)T ,
0 (3.13)

where τ is the relaxation period and α determines the size of the relaxation part. J'S
is the malleability defined as the inverse value of the module G (see details below).
γ is usually not given only by one relaxation time, but by the total N, which stands
for the sum of relaxation terms and the relaxation period τk and the size αk. Then we
talk about the formation of the relaxation spectrum

( )
N
γ r = ∑ α k 1 − e − t / τk σ 0 . (3.14)
k =1

We get for the total shear deformation for plastic materials and rubber the following
relation

γ = J (t ) σ 0 , (3.15)

©Alexander Dubcek University of Trencin 2007 43

where the time dependence in the skid J(t ) is given by the relation

( ).
N
1
J (t ) = J S' + t + ∑ αk 1− e - t / τk
(3.16)
η0 k =1

The constants I0, η0, αk, τ k performing in the equation are dependent for a certain
polymer just only on temperature and on previous heat processing. So, for example,
1/η0 rises with temperature, whereas relaxation periods fall with temperature. In
certain thermal areas there are also 1/η0 and the individual αk zero, where, for
example, 1/η0 =0 means for ideally viscous materials the condition of the supple
solidity (the supple solidity can be watched for example in clays). If big shear
tension acts on plastic materials and rubber, the dependence on T0 joins that.

In term of the elastic part of the deformation of rubber it is necessary to point

out one more interest that is specific for these materials. The measured elastic
deformation of rubber consists of two parts. This deformation leads partly to
fall of entropy as a result of the increase of the system order and this process is
proved also by the fall of the heat capacity of the material. The module E1
arising from the fall of the entropy is directly proportional to the absolute
temperature. In non-ideal vulcanized rubber a part of deformational work is
bound in form of internal energy, in consequence of which, it comes up to
changes of inter-atomic distances, and especially to changes of valence angles
and corresponding change of the module Eo. Then the following relation is
applied for the whole the Young’s module:

1 1 1
= + , (3.17)
E E 0 E1
or
1 1 const
= + . (3.18)
E E 0 temperature

As it is visible in the Figure 3.5, the Young’s module of rubber rises at higher
temperatures with increasing temperature (the line a), what is consequence of
entropic character of the elasticity of rubber, whereas it falls for plastic
materials (the line b), because the portion of contributions of the returnable time
dependent deformations rises. We will explain these facts closer at the end of
this chapter.

©Alexander Dubcek University of Trencin 2007 44

Figure 3.5. Dependence of the module E from temperature. The change of the
module between the points A and B depends on the transition to ideally elastic
behaviour of materials and the temperature of the glass transition. The rubber
materials (a), viscous - elastic materials (b)

The module of rubber materials is substantially lower than it is for other

materials. The elastic deformation of rubber exceeds the original length of
material in many cases. It is therefore not possible to plot tension and
deformation for rubber, plastic material and steel on one diagram. Such
comparison is possible only in logarithmic scale, as it is visible in the Figure
3.6.

%
a
100

b c

ε 1

0,01

0,01 1 100
σ
Figure 3.6. Dependence of relative extension on tension for rubber (a),
viscoelastic material (b) and steel (c)

©Alexander Dubcek University of Trencin 2007 45

3.2. Time dependence of deformation

Let’s use an impulse of mechanical tension on a viscoelastic material (see the

Figure 3.7). We observe an elastic response of the material to this impulse in
parts of the line AB and CD. Metal materials, like monocrystals, behave
similarly. The arisen deformation is caused by extension of inter-atomic
distances. In parts BC and DE, we observe the relaxation of the deformation,
which is characteristic for viscoelastic materials, as, for example, plastic
materials are, and the deformation is caused by the changes of the valence
angles. There are also materials with memory of forms, when after deformation
it comes up to changes of forms which belong to reversible processes.

ε,σ σ = konst.

C
ε
D

B ε
E
σ=0
A
Figure 3.7. Time dependence of the deformation of viscoelastic material

There is also another possible approach that lies in the fact, that we evoke
certain, within time constant, deformation, and we see changes of tension in a
sample that falls with rising time. We call this process the relaxation of tension.
If the fall of tension is proportional to the difference σ - σ ∞ with the constant of
1
the proportion , then the following equation determines the fall of tension in
τ
the sample:
t
−
σ = σ - (σ 0 - σ ∞ )e τ . (3.19)

If is σ ∞ zero, then the relaxation time is time within which tension falls e-
times.

If we expose the viscoelastic material periodically in pull or in bend, a phase

difference arises between mechanical tension and deformation and the elastic
module will be dependent on the frequency (see the Figure 3.8.) For deformation
and tension, the following relations are applied:

ε * = ε 0 e iωt , σ * = σ 0 e i ( ωt + δ ) , (3.20)

from here, we get for the complex Young’s modulus and the tangent of the loss
angle this relation

©Alexander Dubcek University of Trencin 2007 46

 σ* σ 0 iδ E' '
E = * =
*
e = E 0 (cos δ + i sin δ) = E'+iE' '; tgδ = . (3.21)
ε ε0 E'

Figure 3.8. The illustration of the phase shift between deformation and tension

The relation 3.21 is possible to be drawn in the complex Gauss’s plane as it is seen
in the Figure 3.9.

Apart from the complex module, the tangent of the loss angle δ stands for a further
significant material parameter characterizing periodic exposure of samples.
Although we will come back to this problem, we can state so far, that the real part
of the complex module represents the elastic properties of material and the
imaginary part the viscous (loss) properties of material. tgδ determines the phase
shift between tension and deformation.

Figure 3.9. The transform of the relation 3.21 in the Gauss’s plane

If we expose viscously elastic material periodically in a skid, through similar

reasoning we get the complex module in the shear loading in this form

G * = G'+iG' ' . (3.21)

Under periodical exposure in the pressure, it is applied

B * = B'+iB' ' . (3.22)

For some cases, we can find in literature a general formulation of the complex
module regardless the type of deformation in the form

©Alexander Dubcek University of Trencin 2007 47

K * = K '+iK' ' . (3.23)

It is often suitable to introduce inverse values of modules that we call malleabilities

for a relevant type of periodically repeating deformation and we get for longitudinal
compliance

J * = J'- J' ' . (3.24)

We can write for the shear deformation similar

*
JS = J'S - J'S' . (3.25)

It is necessary to emphasize for all above introduced modules that they are the
functions of temperature and frequency. Components of the module K* and tgδ
have in real materials maxima at certain frequencies as we can see in the Figure
3.10.
ll
K

CONSTANT
TEMPERATURE, T
tan δ
or

l
K
ll
K, K
l

tan δ

FREQENCY, ω
Figure 3.10. The frequency dependence of the components of the complex
module K’ and K’’, or tan δ

The lines drawn in the Figure 3.8 were gained at constant temperature. As it
comes out from the experiment, the rise of temperature shifts the named lines to
the right, we can then say that maxima are observed at higher temperature with
higher frequency. The Williams-Landel-Ferry’s equation (also WLF equation as
abbreviation) describes this process of the shift of maxima in the form

- 8.86 (T - T0 )
log10 α T = . (3.26)
101.5 + (T - T0 )

The reference temperature T0 is characteristic for an every polymer and is

usually opted fifty degrees higher than its point of the glass transformation is. In
the Figure 3.11, we can see characteristic frequency and thermal dependencies
K’ in 3D diagrams. In the Figure 3.12, the mathematically counted temperature
dependence of the WLF – factor is drawn. We can see from the mentioned
dependence the high sensitivity of the WLF shift factor in the surrounding of
the reference temperature.

©Alexander Dubcek University of Trencin 2007 48

 l
STORAGE MODULUS, K
FREQUENCY, ω

LOSS TANGENT, tan δ

TEMPERATURE, T

FREQUENCY, ω

TEMPERATURE, T
Figure 3.11. Frequency and thermal dependence of the modules K’ and tan δ

25
- 8⋅86 [ T - To ]
LOG FREQUENCY SHIFT FACTOR, LOG10aT

LOG10aT=

To ∼ Τg + 50 °C

10

(T - To)°C
0 -100 0 100 200

-10
Figure 3.12. The frequency shift factor from WLF transformation

Note:
In order to understand better the meaning of the quantity tanδ , we bring out some
experimental results that present the influence of the value tanδ on utility and
physical properties of vulcanizates.
According to the character of acting, we can divide factors that influence dynamic-
mechanical properties of vulcanizates into these groups:

©Alexander Dubcek University of Trencin 2007 49

o chemical
− content and type of used rubber
− the degree of reticulation and dispersion of filler
− proportional representation of components in a blend
− type of used dopes
− the character of interactions filler – filler and rubber – filler
− the degree of filling of the blend
− distribution of phases in the system
− the type of used filler

o physical
− viscoelastic properties of vulcanizates
− temperature of glass transition of the blend

o conditions of measuring
− frequency under which the measurement is done
− temperature of measuring
− size of deformation of the sample during the measuring
− application
− conditions of exploitation of vulcanizate (operation temperature,
environment, degree of loading).

The characteristic dependence (the Figure 3.13) is gained by measuring

dependences tan δ = f (T ) of filled vulcanizates. The named dependence can be
divided into three areas, according to which the final properties of the blend can be
predicted:
∆TADM – the area characterizing the acting of adhesive forces; the area of the phase
shift,
∆THYS – the area characterizing an abrasion in consequence of a hysteretic
component of a friction; the area characterising so called “anti-blocking system”,
∆T >∆THYS – the area characterising the values of the rolling resistance of
vulcanizates.

Figure 3.13. Characteristic areas of graphic dependence tan tanδ = f (T ) for

filled vulcanizates

©Alexander Dubcek University of Trencin 2007 50

In the Figure 3.13., the characteristic areas of adherence, hysteresis and rolling
resistance are possible to be seen.

Most of authors evaluate the dynamic-mechanical properties according to the values

of the loss angle tan δ measured at the temperature of 60° C as a criterion for
evaluating the rolling resistance and tanδ measured at the temperature of 0° C as a
criterion for evaluating the traction properties. Lower values of the rolling
resistance indicate lower values of tan δ and better traction properties indicate
higher values of tan δ.

Differential mechanical thermal analysis DMTA studies molecular movements and

viscoelastic character of material by acting of tension on the material and measuring
its response. It is possible to determine two parameters: the module (shear G* and
bending E*) and the loss angel tanδ . The module contains the real and imaginary
components. The real part of the module G’ (E’) measures the elastic response of
material and the imaginary part G’’ (E’’) reflects the viscous response.

These data are best interpreted on a practical example (the Figure 3.14). It is
possible to define the temperature of the glass transition recorded as the top of tanδ
and also sub-tg transition that is not possible to be measured by DSC (differential
scanning calorimetry).

The data in the Figure 3.14 belong to the sample of polyurethane foam that was
exposed to the simple bend on a corbel. They show evidently the temperature of the
glass transition 159,9 °C with corresponding change in the bending module E’. The
used frequency was 1 Hz.

In the Figure 3.15 there are records of DMTA of natural rubber. These data show
the change in behaviour of fresh prepared and dating rubber, measured in shear
mode DMTA at the frequency 1 Hz. The old sample has the lower module than the
fresh one, and is weaker too. The top of tan δ is also much lower for the old
sample.

The influence of the value of tanδ on the utility properties of vulcanizates used in
tyre production is significant as well.

The area of temperatures ranging from -10°C to +5°C characterizes adherence in

the wet, whereas the temperatures ranging 60°C - 80 °C the rolling resistance. We
can see the use of different types of rubber blends mixed with soot and coagulated
SiO2 in the Figure 3.16.

The use of S-SBR (acarid butadiene styrene rubber S-SBR) leads to higher
values in the first area and lower values in the second area (it means improved
adherence in the wet and lowered rolling resistance) by both fillers. At the
same time the variations in the area of dosing of styrene and of content of the
vinyl form of butadiene component enable the improving of adherence in the
wet or the abrasion by preservation of the rolling resistance – see the Figure
3.17.

©Alexander Dubcek University of Trencin 2007 51

 Figure 3.14. Record DMTA (logE ′ = f (T ), tanδ = f (T )) of foam polyurethane

Figure 3.15. The record DMTA (logG ′ = f (T ), tanδ = f (T )) of the sample of fresh
prepared rubber blend and the same rubber exposed to the influence of ageing
HIGH Tg

©Alexander Dubcek University of Trencin 2007 52

Figure 3.16. The temperature dependence tanδ for emulsive butadiene styrene
rubbers E-SBR and for acarid butadiene styrene rubbers

Figure 3.17. The dependence of tanδ blend S-SBR with precipitated SiO2

©Alexander Dubcek University of Trencin 2007 53

3.3. The temperature dependence of viscosity – micro structural
view

It is well known that especially in summer month it can be seen so called rails
depressions caused by heavy trucks. Imprints of heels of women’s shoes in asphalt
in summer months are known as well. We know also that when we heat honey, it
becomes more liquid, and we could find many other similar examples. One fact
comes out from all of them, and that is, that viscosity of material is apart from the
acting time and the quantity of the impressed mechanical tension also a temperature
function.

In order to explain this phenomenon we will come out from so called hole model of
liquid. For this model it is presumed that liquid consists of molecules and of vacant
places so called holes, which can emerge in liquid by for example evaporating of
molecules. The jump of a molecule into new, energetically preferred position is
possible only if the neighbouring place is free and there is a hole. Concentration of
such holes (proportion of vacant places n to the total number N of knots of the
structural net) will be

n
= e − ∆H / RT , (3.27)
N

where ∆H is enthalpy needed for creating a hole (it is approximately equal to

2/5.∆Hevaporation), R is gas constant and T absolute temperature. Molecules oscillate
around balance positions, and twice in a second they attack surrounding potential
barriers. Then we get for frequency of possible jumps

1 n
= 2ν e − ∆H / RT = 2νe − ∆Q / RT ; ∆Q = ∆H A + ∆HB . (3.28)
τ N

The first enthalpy presents an activating enthalpy needed for creating a hole and the
second presents the height of a barrier that a molecule has to overcome by self-
diffusion. If we indicate the coordinating number of the molecule z (the number of
1
nearest neighbours is z), then for the frequency of jumps is ν = .
τz

If the external power F assists in the diffusion process, molecules will move then in
the direction of this power and the middle speed of the molecules will be
proportional to this power. The coefficient of proportionality will be called the
mobility of a molecule

ν =qF. (3.29)

If among liquid layers, which rub against each other, there is a distance δ equal to
the inter-atomic distance, then this relation is applied

1
η= (3.30)
qδ

©Alexander Dubcek University of Trencin 2007 54

and we get for the coefficient of dynamic viscosity (without writing further steps of
mathematical character)

zτ zkT Q / RT
η = kT = e , (3.31)
δ 3 2δ 3 ν

where k is the Boltzmann´s constant. We can see that the coefficient of dynamic
viscosity is falling with rising temperature, as we have already mentioned that
above. The given model is certainly qualitative and also other relations are used for
individual types of material and temperature range.

Now we are going to extend our knowledge about the physical interpretation of
viscous flow for polymers and rubbers. It is known that molecular jumps come up
mainly in amorphous areas, whereas they come up on their boundaries or on defect
places for crystals. It can cause twist of whole crystal as well. Kinetics of the flow is
described by two different processes.

Let’s analyse the first process. The jumps of molecules in the direction of the flow
lead to arrangement with smaller entropy without a change of internal energy; we
call this process entropy elasticity. However, it comes to occupying of these states
less often without acting of external forces. We can see the process in the Figure
3.18a.

Without a change of the form of potential lines, it may more often come to
occupying the states with higher energy, which is determined by existence of nodal
points and it is so called energy elasticity and the corresponding situation is in the
Figure 3.18b. Both these processes are called returnable flow.

In the second case, molecules or their parts are deflected from their surrounding and
the nodal points disappear. The division of potential energy is not changed, because
new potential beams are being established, whose energy distribution is the same as
the previous one. We call this flow the non-returnable flow and the Newton low for
viscosity is applied for it in the first approach.

©Alexander Dubcek University of Trencin 2007 55

 Figure 3.18. The dependence of the change of the potential from the distance.
The model of entropic elasticity (a), the model of energy elasticity (b). The full
lines correspondent the state before impressing tension and the dotted once
after applying tension.

3.4. Thermodynamic aspects of viscous elastic and rubber

deformation

The basic relations of phenomenological thermodynamics, as we have already said,

are applied also for thermodynamic balance. The establishing of this balance is
dependent on time and for polymers it may not be reached, according to
circumstances, during the observing period of this final state

In macromolecular solid materials there are not sharp boundaries between

individual phases. Amorphous and crystal areas are not separated with boundaries.
The crystal areas are mostly sub microscopically small; they consist of only a small
number of molecules. Polymers do not have a sharp point of melting, but an
extensive area of melting. This area remains preserved under random pressure, what
is against the phase rule for unary systems, if we admit that it is thermodynamic
balance. These reservations and findings, that it comes to significant changes of the
state within long-term observations, initiated us to waive the assumption of

©Alexander Dubcek University of Trencin 2007 56

multiphase in thermodynamic sense for partly high-molecular materials at all.
Discontinuous changes of the coefficient of expansion and changes in growth of
capacity in the area of the glass transition by heating and relatively fast observing,
we can interpret with help of long-term measurements in that way, that we call them
non-balanced state.

Firstly we are going to discuss the possibility of considering the phase transitions as
the changes of the second order. We observe a break in thermal dependence of
capacity and sudden discontinuity of the coefficient of thermal extension in the area
of melting. The thermal dependence of the coefficient of the thermal extension, the
capacity, specific heat and the module of elasticity in shear are visible in the Figure
3.12. The Figure 3.19 includes, at the same time, the theoretically counted course of
these quantities for the changes of the first and the second order according to
Ehrenfest’s equation. We get very good consensus, if we assume changes of the
second order in polymers.

Providing that they are thermodynamic balances, it is necessary to consider these

transitions:

The melting of crystals in the area of the rubber elasticity that begins within narrow
thermal interval, as experiments gained with the help of polarized light demonstrate.
Activation of the microbrown movement that is connected with the rise of the
specific heat and with the rise of the coefficient of the extension that comes up by
jumping.

The growth of specific heats can be explained with that fact, that the rotation of
molecular segments and side chains are beginning to be practiced.

In the “transformational point”, it is necessary to count with the fact, that

molecular jumps are beginning to be practised that are responsible for the flow,
resilient fracture, for dielectric and mechanical relaxations and for diffusion. If we
take the change of the coefficient of the extension that arises in the transformational
point by jumping, as a formation of unstable state, then we reach the fact, that the
state existing above the transformational point, it is a single-phase melt, must exist
in the thermodynamic balance also under this point of the glass transition, what we,
however, cannot realise during the final period of observing. Under the
transformational point, as we know, there is an unstable state “frozen” and liquid
with “fixed structure” comes to the existence.

©Alexander Dubcek University of Trencin 2007 57

 Figure 3.19. Thermal dependences of the specific heat under the constant
pressure cp, the capacity V, the coefficient of thermal extension α and the
complex module G* for phase changes of the first and second order as well as
the experimentally measured dependences for polymers.

After these initial notes that characterize the process of deformation in

macromolecular materials, we come to the formulation of their behaviour in exact
form. If we introduce instead of the capacity V the length l, instead of pressure we
will think of normal tension σ n , that we will consider to be minus, because it acts
reversely as the pressure, then the equation

∂U ∂p
p=- + .T (3.32)
∂V T ∂T V

will be transformed into a new form

©Alexander Dubcek University of Trencin 2007 58

 U σ
p=- + n .T (3.33)
l T T l

It comes out from the experiments that tension is proportional to temperature

ranging from -60°C to +60°C by the tensile deformation of vulcanized rubber,
∂σ
which does not exceed 300 percent. Therefore it must be applied n = constant ,
∂T
(3.34) from here, after putting into the equation 3.32, we get

U
= 0. (3.35)
l T

We can interpret the last equation as following: the potential energy does not rise at
deformation, what is diametrically different behaviour from a deformed steel string.
This relation can serve as definitional one for the ideal rubber material.

If we heat rubber, the entropy element changes according to the relation

σ n dl
dS = - . (3.36)
T

As dQ = TdS and σ n are positive, the length is falling in the process, and so the
rubber has a negative coefficient of the extension.

Instead of the state equation V = V (p, T ) it is applied for high – elastic materials the
equation l = l(σ n , T ) . We express the total differential of the length in the following
form

∂l ∂l
dl = .dσ n + dT , (3.37)
∂σ n T ∂T σn

from where we get this relation for the change of the system state with the constant
length (dl=0)

∂l
∂σ n ∂T σn
=- . (3.38)
∂T l ∂l
σn T

As the coefficient of the extension is negative, the denominator of the expression on

the left side of the equation is positive, the left part of the equation will be positive,
what means, that tension is rising while heating up rubber.

©Alexander Dubcek University of Trencin 2007 59

The deformation in pull leads to the state with higher molecular arrangement, and
thus with lower entropy. The thermal flow acts in the sense of renewal of the
dQ
previous disordered, that is non-deformed, state. As dS = , the system gives
T
heat to environment by tensile deformation.

If this is not possible, as a consequence, the increase of molecular movements

comes up. This adiabatic deformation of rubber results not only in increase of
temperature, but also in a bigger growth of tension against the deformation within
the constant temperature. This similarity with gas seems to be remarkable, however,
it come out from the basic laws of phenomenological thermodynamics.

3.5. Payne effect

Various interfering processes occur at cyclical application of stress on polymer

material, as was mentioned in chapter 3. The most important of these is
superposition of Hooke's deformation of polymer string segments at small
amplitude of cyclical deformation on one hand, and translation motion of molecules
on the other hand. Result is time delay of deformation after stress by a time period
that is at invariable conditions (temperature, frequency and amplitude of
deformation) constant. If we systematically increase temperature (or frequency)
during the experiment, we can determine conditions when release of certain motion
types of string parts occur and determine whole relaxation spectrum of polymer,
whereby the most important information is temperature of glassy transition.

If we modify conditions of experiment based on cyclical deformation, we can also

use the method for characterization of composites, e.g. rubber or plastic filled with
active filler. In this case we stabilize experiment conditions what regards
temperature and frequency and we change amplitude of deformation.

This procedure is based on the principle of so-called Payne effect. Payne effect is in
professional literature sometimes referred to as model of variable mesh density.
Mesh density is defined as number of cross-links in unit volume of cross-linked
polymer. Cross links can be in the polymer of clearly chemical, mostly covalent
character, e.g. sulphur bridges in the case of vulcanized rubber, or carbon-carbon
bonds at cross-linking of polyolefins, or they can be physical, if strings are in the
course of chemical grid creation interconnected e.g. into cycles and create so-called
tang lings. The situation is much more complicated when the filler is present,
because another space interconnection is created via stable macromolecule
attachments on the filler surface, or less stable interactions between polymer
segments and filler surface. Finally, even space formations only from filler elements
could be formed at filler volume increase. They form firmer or looser
interconnected filler elements.

Mesh density of vulcanizate is then given by total sum of all mentioned possibilities
and its relation to mechanical properties can be modelled by Mooney – Rivlin
equation

©Alexander Dubcek University of Trencin 2007 60

F
A0
( ) ( )
= 2C1 α − α −2 + 2C 2 1 − α −3 , (3.39)

⎛ C ⎞⎛ 1⎞
σ = 2⎜ C1 − 2 ⎟⎜ λ2 − ⎟ , (3.40)
⎝ λ ⎠⎝ λ⎠

F
where is mechanical stress (force F on unit cross-section of strained solid), α
A0
is lengthening rate given by the length of element after application of stress divided
by former length and C1 , C 2 are empirical constants, where first one is proportional
to the concentration of chemical cross-links and the second one is supposed to be
related to physical bonds.

In composite material, i.e. in the system consisting of polymer die and filler is total
number of cross-links that create spatial grid, sum of cross-links in polymer that
consist of chemical and physical bonds (tanglings), interactions of polymer and
filler surface that can be stable or temporary, and interactions between filler
elements alone, where all mentioned concentrations of cross-links must be
expressed in unit quantity of material.

We have to emphasize here that not all the created cross-links are included in
abovementioned theorem, but only part of them that is created by so-called
elastically active strings. Certain part of cross-links, not very numerous but
considerable, is offering no resistance to applied stress during deformation. Those
are relatively long strings and there are many shorter cross-links in their vicinity.
The shorter bonds are during deformation totally tightened sooner than longer string
between them takes at least part of stress. Of course, longer string in such case
contributes nothing to the value of dynamic modulus and it will not manifest itself
at cross-links number calculation form the parameters of Mooney-Rivlin equation.

If we deform cross-linked polymer with filler content, where all types of chemical
and physical cross-links are present, with regularly increasing force, growth of
deformation occurs and bonds are being gradually disrupting during it. Weak bonds
between polymer and filler surface are disrupted at first and then, step by step along
with growing deformation the stronger ones (for that you will have to make greater
and greater effort) and finally it comes to disruption of solid. However, if we bring
such system under cyclical strain with particular amplitude of deformation, the
bonds (disruption of which corresponds to particular deformation) are disrupted, but
regeneration of all or some disrupted bonds can occur subsequently during reverse
direction of stress in the frame of one cycle (decrease of stress and deformation). If
we progressively increase amplitude of deformation, still more and more physical
cross-links are being disrupted and their reversed regeneration is less probable.
Finally, in extreme case and at certain amplitude, all physical bonds forming
secondary physical spatial grid, i.e. grid that was formed by filler as well, are
disrupted in such extent that no regeneration will occur any more. Consequences of
this phenomenon will be described in following text.

At cyclical stress with variable amplitude of deformation is deformation oscillating

in relatively low figures. On the mostly legitimate assumption that this process is
running more or less in Hooke’s area of deformation, we can express modulus of

©Alexander Dubcek University of Trencin 2007 61

elasticity as the proportion of stress to amplitude of cyclical deformations. If we
regularly increase amplitude (i.e. deformation), we have to proportionally increase
stress as well until deformation is increased in such a way that disruption of some
grid element occurs. If we achieve total regeneration of disrupted grid at cyclical
stress release, the same modulus will be measured in next cycle as if the physical
grid has not been disrupted. However, another situation occurs when deformation is
so large that whole grid could not manage to regenerate reversibly to its original
state. Then, particular deformation (that is constant at cyclical stress) in next cycle
is achieved with lesser exerted force, what results in lower values of modulus of
elasticity. If we regularly increase the amplitude, then in extreme case we get the
state, when specified dynamic modulus will be identical to the modulus of unfilled
cross-linked rubber. In this case the whole physical grid is at specific amplitude so
much disrupted that its regeneration will practically never happen. We can get this
state by increasing the amplitude, while cyclical deformation frequency increase
leads to decrease of deformation, at which we observe decrease of dynamic
modulus of elasticity. Of course, increase in frequency leaves less time for grid
regeneration and it will not manage to regenerate, not even in the case, when its
disruption happens in reduced extent with lesser deformation.

These reasoning imply the interpretation for observed phenomena that we call
Payne effect. When we increase the grade of dynamic deformation, values of
modulus of elasticity (E’) for unfilled vulcanizates do not change significantly. But
there is modulus decrease in the case of filled vulcanizates, while this effect is more
significant for vulcanizates with high filler density. While difference of modulus
values is small for unfilled/filled vulcanizate at large deformations and it is
relatively high at small deformations, Payne effect is related mostly, if not
exclusively, to the grid created by elements of filler in polymer die. Polymer
trapped in such grid looses its identity and behaves as filler. Along with increase of
filler share in the system, there is increase in effective volume of filler and modulus
as well. If higher grade of deformation is applied, disruption of grid created by filler
elements occurs and is accompanied by release of polymer that was formerly bound
to filler. Effective filler volume (in dynamic process it represents real filler volume
increased by the volume of bound immobilized polymer) is by this action
decreased, released polymer gets back its former properties and decrease of elastic
modulus follows. This implies that Payne effect can be used for characterization of
grid formed by filler that comes from filler-filler and filler-polymer interaction.

In the case of loss modulus (E’’), like at modulus of elasticity, we observe increase
of its values along with rising degree of polymer filling in bundle of dynamic
deformations. However in dependence on deformation increase, no monotonous
decrease of values occurs (like in the case of modulus of elasticity), but we observe
maximum at particular deformation that is characteristic for specific pair of
polymer-filler. We assume that process of loss modulus is controlled by disruption
and re-creation of grid from filler elements. It implies that loss modulus is
dependent on rate of disruption and re-creation of the grid. Rate of disruption and
re-creation of the grid emerges from the phenomenon of dynamic deformation
alone. Disruption of the grid is considerable at large deformation and its reversed
re-creation practically does not happen. If deformation is large enough for the grid
not to re-create at specific frequency, influence of the grid on loss modulus will be
eliminated. Similarly, if deformation is small (or grid is firm enough) to cause

©Alexander Dubcek University of Trencin 2007 62

disruption of the grid, loss modulus will not be dependent on specific deformation,
but it will depend mostly on polymer die contribution during cyclical dynamical
deformation of composite.

Change of every modulus (elastic and loss) in dependence on extent of dynamic

deformation emerges from different mechanism. While modulus of elasticity
depends significantly on the physical grid presence (its gradual disruption at
dynamic deformation increase) that is formed by filler elements in polymer die, loss
modulus is influenced by repeating process of disruption and re-creation of specific
structure at cyclical stress. Loss factor (as their proportion) will thus reflect
combination of these two processes. When increasing the amplitude of elastic
deformation, we observe monotonous, but not linear decrease in the case of elastic
modulus, while dependence of loss modulus goes through maximum. Therefore at
dynamic deformation amplitude increase, development of loss factor, as the
quotient of loss and elastic modulus in dependence on the grade of dynamic
deformation for filled polymers, is much more influenced by changes of loss
modulus values. In dependence on deformation increase we observe increase of tan
δ values that go through maximum with following sharp decline at greater
amplitudes of deformation.

The way of creating physical grid depends on many factors, mostly on filler type
and parameters of its surface and shape. In the case of polymer with higher affinity
to hydrocarbon polymer, creation of the grid can be described by mechanism of
mutual interpenetration of filler polymer covers. Adsorption of polymer strings to
the filler surface happens by polymer-filler interaction and mobility of polymer
segments is thereby limited. Result of this is the creation of polymer cover on the
filler surface, where viscosity and polymer modulus will increase. Very high
modulus in the vicinity of filler element surface in polymer cover will decrease
along with increase of distance from the filler surface, until (in certain distance) it
will be the same as modulus for polymer die. If two or more filler elements or
aggregates are close enough to each other, they will create agglomerate by the
mechanism of mutual interpenetration of polymer filler element covers. The grid,
created by such a mechanism, will be less rigid than grid created by direct contact
between aggregates. Such a grid type can start to be disrupted at relatively small
deformations.

In the case of grid, created by direct contact of aggregates, process of repeated

disruption and re-creation of grid causes higher energy losses. This suggests that at
high temperatures internal friction between aggregates is dominant mechanism.
When we decrease temperature, polymer gets to the sphere of glassy transition,
where polymer die has the main participation on losses of inputted mechanical
energy.

If the grid is formed by mutual interpenetration of element covers, mechanism of

contribution to energy losses is different than it is for the abovementioned grid.
Polymer die is at higher temperatures in rubber state, but polymer adsorbed to the
filler surface is in transitional phase. That leads to fact that cover can absorb more
energy. Thickness of cover decreases along with rising temperature, that will
increase mobility of strings and that leads to lower hysteresis. Reversed filler
reagglomeration during cyclical dynamic deformation goes to higher level. If the

©Alexander Dubcek University of Trencin 2007 63

temperature is decreasing, polymer share in filler element cover is increasing,
mobility of segments is lowered and energy losses ratio in cover is increased
regarding energy losses in polymer die. If the temperature is so low that polymer in
cover is in glassy state, but polymer die is still in rubber state or possibly in
transitional phase, effective voluminous fragment of filler will be considerably
increased. Filler still has toughening effect and polymer in cover (as well as
polymer trapped in agglomerates) does not contribute to energy losses in particular
system.

Both mechanism types (method of direct contact and mutual interpenetration of

polymer element covers) play important roles at grid creation from filler elements
for particular polymer-filler system. However, with respect to polymer properties
and character of filler surface, which determine polymer-filler and filler-filler
interactions, one of the mechanisms can be much more evident than other one.

The most significant influence on change of dynamical properties of filled polymers

has morphology of filler elements; specifically it is the size of elements or their
specific surface and element structure. With the increase of specific surface of soots
at constant filling we observe significant decrease of elastic modulus with
increasing degree of dynamic deformation. Increase in Payne effect indicates great
ability of small elements (aggregates) of soots agglomeration in polymer die.

These reasonings imply that measuring of Payne effect can significantly contribute
to the characterization of interactions between elements of loaded vulcanizate in
dynamic conditions.

Figure 3.20. The influence of polymers and fillers on the dynamic properties
of elastomers, Payne effect

At the end we can conclude, that the principal characteristics of Payne effect consist
of a decrease in complex modulus (G*) with the amplitude of deformation.
Abundantly studied by Payne in carbon black (CB) filled elastomers and this
phenomenon named after him. Since Payne's studies, a large of investigations has
been performed with both experimental and theoretical approaches. The decrease in
G’ has been attributed to several local mechanisms, namely the destruction -
reformation of a percolating network of filler that can as well involve polymer

©Alexander Dubcek University of Trencin 2007 64

bounded to the filler adsorption - desorption of polymeric chains at the interface
desentanglement of bulk polymer from the rubber bounded to the surface.

3.6. Fracture properties of polymers

When cracks are identified in structures or components during service, they must be
evaluated to determine suitability for continued operation. Fracture mechanics
provides a methodology evaluating the structural integrity of components
containing such defects, and demonstrating whether they are capable of continued,
safe operation.
The basic criterion in any fracture mechanics analysis is to prevent failure. To do
so, the crack driving force must be less than the material resistance to cracking, as
illustrated in the next figure.

Fracture mechanics provides quantitative answers to such structural integrity

questions as:
- What is the critical crack size at service loads?
- How safe is the system if it contains a crack?
- How long might it take for a crack to grow from initial to critical size?
- How often should a particular structure be non destructively inspected?

The correct answer to these, and related questions, are of fundamental importance
for the safe operation and maintenance of plant and equipment, and enable
unexpected catastrophic failure to be almost completely prevented.

The fatigue life of a component is made up of initiation and propagation stages.

This is illustrated schematically in Figure 3.21.

©Alexander Dubcek University of Trencin 2007 65

 Figure 3.21. Initiation and propagation portions of fatigue life

The size of the crack at the transition from initiation to propagation is usually
unknown and often depends on the point of view of the analyst and the size of the
component being analyzed. For example, for a researcher equipped with
microscopic equipment it may be on the order of a crystal imperfection, dislocation,
or a 0,1 mm-crack, while to the inspector in the field it may be the smallest crack
that is readily detectable with nondestructive inspection equipment.

Nevertheless, the distinction between the initiation life and propagation life is
important. At low strain amplitudes up to 90% of the life may be taken up with
initiation, while at high amplitudes the majority of the fatigue life may be spent
propagating a crack. Fracture mechanics approaches are used to estimate the
propagation life.

Fracture mechanics approaches require that an initial crack size be known or

assumed. For components with imperfections or defects (such as welding porosities,
inclusions and casting defects, etc.) an initial crack size may be known.
Alternatively, for an estimate of the total fatigue life of a defect-free material,
fracture mechanics approaches can be used to determine propagation. Strain-life
approaches may then be used to determine initiation life, with the total life being the
sum of these two estimates.

Linear Elastic Fracture Mechanics Background

Linear elastic fracture mechanics (LEFM) principles are used to relate the stress
magnitude and distribution near the crack tip to:
• Remote stresses applied to the cracked component
• The crack size and shape
• The material properties of the cracked component

Historical Overview

In the 1920s, Griffith formulated the concept that a crack in a component will
propagate if the total energy of the system is lowered with crack propagation. That
is, if the change in elastic strain energy due to crack extension is larger than the
energy required creating new crack surfaces, crack propagation will occur.

©Alexander Dubcek University of Trencin 2007 66

Griffith’s theory was developed for brittle materials. In the 1940s, Irwin extended
the theory for ductile materials. He postulated that the energy due to plastic
deformation must be added to the surface energy associated with the creation of
new crack surfaces. He recognized that for ductile materials, the surface energy
term is often negligible compared to the energy associated with plastic deformation.
Further, he defined a quantity, G, the strain energy release rate or "crack driving
force," which is the total energy absorbed during cracking per unit increase in crack
length and per unit thickness.

In the mid-1950s, Irwin made another significant contribution. He showed that the
local stresses near the crack tip are of the general form

KI
σ ij = f ij (θ ) + ... (3.41)
2πr

where r and q are cylindrical coordinates of a point with respect to the crack tip (see
Figure 3.22.) and K is the stress intensity factor. He further showed that the energy
approach (the "G" approach above) is equivalent to the stress intensity approach and
that crack propagation occurs when a critical strain energy release rate, G, (or in
terms of a critical stress intensity, Kc) is achieved.

Figure 3.22. Location of local stresses near a crack tip in cylindrical

coordinates

LEFM Assumptions

Linear elastic fracture mechanics (LEFM) is based on the application of the theory
of elasticity to bodies containing cracks or defects. The assumptions used in
elasticity are also inherent in the theory of LEFM: small displacements and general
linearity between stresses and strains.

©Alexander Dubcek University of Trencin 2007 67

The general form of the LEFM equations is given in Eq. 3.41. As seen, a singularity
exists such that as r, the distance from the crack tip, tends toward zero, the stresses
go to infinity. Since materials plastically deform as the yield stress is exceeded, a
plastic zone will form near the crack tip. The basis of LEFM remains valid, though,
if this region of plasticity remains small in relation to the overall dimensions of the
crack and cracked body.

Loading Modes

There are generally three modes of loading, which involve different crack surface
displacements (see Figure 3.23.). The three modes are:
Mode 1: opening or tensile mode (the crack faces are pulled apart)
Mode 2: sliding or in-plane shear (the crack surfaces slide over each other)
Mode 3: tearing or anti-plane shear (the crack surfaces move parallel to the leading
edge of the crack and relative to each other)

The following discussion deals with Mode 1 since this is the predominant loading
mode in most engineering applications. Similar treatments can readily be extended
to Modes 2 and 3.

Figure 3.23. Three loading modes

Stress Intensity Factor

The stress intensity factor, K, which was introduced in Eq. 3.40, defines the
magnitude of the local stresses around the crack tip. This factor depends on loading,
crack size, crack shape, and geometric boundaries, with the general form given by

⎛a⎞
K = σ . π .a . f ⎜ ⎟
⎝ w⎠
(3.42)

where:
s=remote stress applied to component (not to be confused with the local stresses, sij,
in Eq.41)
a=crack length ,f (a/w) = correction factor that depends on specimen and crack
geometry

Figure 3.24. gives the stress intensity relationships for a few of the more common
loading conditions. Stress intensity factors for a single loading mode can be added
algebraically. Consequently, stress intensity factors for complex loading conditions

©Alexander Dubcek University of Trencin 2007 68

of the same mode can be determined from the superposition of simpler results, such
as those readily obtainable from handbooks.

Figure 3.24. The stress intensity relationships for a few of the more common
loading conditions

Figure 3.24. Stress intensity factor for

(a) Center-cracked plate loaded in tension,
(b) Edge-cracked plate loaded in tension,
(c) Double-edge-cracked plate loaded in tension
(d) Cracked beam in pure bending

Refraction occurs by big deformation tensions and energy needed for its realization
can be divided into three parts:
1.A part of energy is spent on a breach of main and lateral links and a breach of
a new free surface
2.Most of this energy is used for flow processes close to the fracture surface
3.Accumulated elastic energy is absorbed in a deformed body in a form of audio
as well as thermal energy.

©Alexander Dubcek University of Trencin 2007 69

We can characterize the fracture action with following processes:
a. The existence of submicroscopical flaws and indents
b. The creation of the maximum of the tension together with a fully elastic
deformation. These maxima arise on ends of microscopical flaws in indents with
small radiuses, on inhomogenities, eventually on the surface.
c. The creation of molecular skips in the strained area which enables a decrease of
critical tensions and a scrubbing of microscopical flaws.
d. The creation of a new surface during which not only molecules are split, but also
the lateral links.
e. Inertial weight nearby the strained place inhibits the spreading of big tensions to
longer distances.
f. Transformation of a part of elastic energy on the location of the fracture into
elastic waves spreading with the speed of sound that support the spread of the
fracture
g. Creation of ductile (also deformation) fracture in amorphous areas and fragile
fracture in crystalline areas. We distinguish slides in the direction of sliding
tensions from indents that are abeam toward the direction of acting tension. In the
first case the flow occurs nearby the fracture surface where it is possible to watch
the creation of characteristic formations that are typical for the ductile fracture. In
the case of the fragile fracture no flow occurs. The speed of deformation and the
possibility of closing up of molecular indents are characteristic for the kind of the
fracture.

We talk about so called pulse ductility with regards to fracture tests. The pulse
ductility is defined as work needed for breaking through a testing body calculated
on the surface unit. If the elastic tension is inhibited to be created in the large area
of material with creating of indents during the measurement of the pulse ductility
and if high accumulation of tension arises in an indent, then we talk about impact
ductility.

The hardness of material depends also on a temperature and the length of a testing.
Long-term testing gives lower values of the material hardness. Fatigue fractures are
given out by repetitive testing. We can attain enhancement of the material hardness
for instance by filling, lengthening, adding softeners, and as a result of these, the
temperature of vitrification slides to lower temperatures.

If we tried to quantify tension relations during a fracture testing on an example of

a bearer with the height h, width c, module E and radius of curvature r, we would
get for maximal tension with breaking of 1/a2 molecules on the surface unit (it
refers to quasi – cubic structure with middle intermolecular distance a) the value of
the maximal tension σ max ≈ 3.10 6 N / cm 2 , which is about a centuple higher value
than the measured one (calculated for energy C-C link equal to 4.8 10-19J and for
interatomic distance a = 0,2 10-9m) This difference can be explained with the fact
that not all molecules are strained at the same time. The most strained ones are
those whose axis is in the direction of the action of the tension. This re-orientation
of molecules into the direction of the acting tension cause extensive hardening of
the material. With given tension, this process is in progress in dependency on
temperature with delay. The relaxing period τ i = e ∆Ui kT / 2ν decreases quickly with
increasing temperature, which enables quicker closing up of indents. The tensions
therefore vary from a molecule to a molecule with maxima on places of indents.

©Alexander Dubcek University of Trencin 2007 70

The maximum tension on the edge of an indent of the length l and of radius r is
l
determined by the relation Tmax = 2T0 , where T0 is the middle tension. If the
r
maximum tension exceeds the molecular hardness, the indent proceeds.

The theory of indents explains below measured values of hardness only partially.
The main reason of these differences is the influence of the statistical position of
molecules on unequal allocation of tensions on individual links. We cannot talk
about any middle values of tensions for individual molecules. The material is
lengthened by molecular skips before breaking, which affects its hardening. The
pulse ductility as a function of temperature has significant maxima regarding to
plastic, which goes up to higher temperatures with increasing speed of impact.
40

30

20
kpcm

10

0
-60 -40 -20 0 20
o
temperature C
Figure 3.25. Dependence of pulse ductility of polyvinyl on temperature

The behavior of polymeric materials during breaking testing differs from similar
effects in inorganic glass and crystalline materials, especially by acting of main and
lateral links. In conclusion, we can summarize about fractures the following. The
breaking by a fragile fracture arisen from random indents occurs at low
temperatures. With an increasing temperature or a more gradual acting of a tension
the flow and re-orientation of molecules occur, which use significant part of impact
energy. While shaping plastic products, it is necessary to work with high cutting
speeds to prevent from internal strain. Processes like turning or milling correspond
to atermical fracture processes which are not accompanied with a flow. There is
a minimum regarding to expended work because at low temperatures energy needed
for fracture actions rises by increased acting of portative force of atoms. The
minimum of work expended for the breaking through the body is for polyvinyl at
minus twenty degrees Celsius (see Figure 3.25.).

Real fracture of PBT sample scanned by high speed camera is on Figure 3.26
(virgin (left) and toughened (right) impact curves).

©Alexander Dubcek University of Trencin 2007 71

 Figure 3.26. Virgin (left) and toughened (right) impact curves and high speed
camera picture

Some characteristics features of crack growth in rubber compounds is described in

Figures 3.29-3.30.

Note:
Fundamentals of crack growth in specific case are explained below:

Figure 3.27. Fatigue crack propagation

Tearing energy (G), independent of crack length (C)

G = Uh0 (3.43)

©Alexander Dubcek University of Trencin 2007 72

Power Law Dependency between Crack Growth Rate (dc / dn) and Tearing energy
(G)

dc
= AG α (3.44)
dn

where A & α : constants characteristic of rubber fatigue crack growth (independent

of specimen geometry).

Determination of crack growth rate:

Figure 3.28. Strain energy density for rubber compounds

©Alexander Dubcek University of Trencin 2007 73

 Figure 3.29. Dependence of crack length and cycle number for 35% strain, 1Hz
frequency, 40°C temperature

Figure 3.30. Dependence of crack growth rate vs. tearing energy for rubber
compounds

©Alexander Dubcek University of Trencin 2007 74

Electronic microscope pictures of NR are presented in Figure 3.32.

.
Figure 3.32. Fatigue-failure surfaces of various double - network NR

References

[1] http://search.epnet.com
[2] Winema, A.S., Rajagopal, K.R.: Mechanical response of polymers, 2000,
ISBN 0 521 64409 7.
[3] http://en.wikipedia.org/wiki/Polymer_physics
[4] Cochet, P.: Payne effect of silica vulcanizates. Silica in winter tire tread. NRC
conference, Finland, 2004.
[5] Frohlich, J., Niedermaier, W., Luginsland, H.D.: The effect of filler-filler and
filler elastomer interaction on rubber reinforcement. Composites: Part A, Vol.
36, 2005, P. (449-460).
[6] www.key-to-steel.com/articles/art45.htm
[7] Kaang, S., Jin, Y., Im, W.B.*, Kim, Y.-J.*, and Nah. C. **, An Advanced
Test Method to Evaluate Fatigue Crack Growth Characteristics of Rubbery
Materials, IRC 2004, Faculty of Applied Chemical Engineering, Chonnam
National University, Korea

©Alexander Dubcek University of Trencin 2007 75

Fundamental questions from presented part
1. What type of forces work at interpretation of viscous properties of
materials?
2. Describe the elastic, viscous and relaxation deformity.
3. What is the definition of tan δ and what its consequence? What identify
the components of complex Young’s modulus?
4. What is the temperature dependence of viscosity?
5. How does the elastic modulus of vulcanizate change in relation to
loading?
6. Explain the reason of Payne effect. What is the task of fillers in Payne
effect?
7. What are the basic models for fracture mechanics?
8. What is characteristic for fracture mechanics of polymers?
9. Name processes, which characterize the fracture action.
10. How does the behavior of polymeric materials during breaking testing
differ from similar effects in inorganic glass and crystalline materials,
especially by acting of main and lateral links?

©Alexander Dubcek University of Trencin 2007 76

 CHAPTER 4
Models of viscoelastic behavior of materials

Objectives to achieve
In this section are introduced basic theoretical approaches describing viscoelastic
behavior of materials. Readers will be familiared with basic models: Maxwell,
Voigt and their combinations.

Now let begin to look into the problem of viscous environments

phenomenologically. The Hook’s principle in the symbolic form σ = E ⋅ S ,
where σ is stress, S is strain, E is Young’s modulus, as we already know, is given
for an elastic environment. We can write the Newton’s principle of the viscous
flowing in the form T = η ⋅ S& , where S& stands for the speed of deformation, η is
dynamic viscosity is , T is tension.

The real environment, as we have already said, has viscous as well as elastic
attributes. Let us assume, that the tension Tv consists of “the elastic tension” σ
and “the viscous tension” T

Tv = σ + T = ES + ηS& . (4.1)

This equation is also called the Kelvin’s equation.

If we affect such an environment in time t = 0 with a rectangular impulse of the

mechanic tension T0, the deformation of the environment S will be described with
the following relation

η
S=
T0
E
( )
1 − e −t / τ ; τ=
E
. (4.2)

T0
Consequently, in sufficiently long time the constant deformation S ∞ =
E
independent from the time is formed and the environment is behaving as unviscous.
The response of the environment to the applied tension is not obviously immediate,
but it gradually reaches the maximum value. The term 4 creep has been employed
to describe this process. Likewise after interrupting of the performance of the
tension, the deformation decreases exponentially to zero, whereas the relation
η
determines the relaxation time τ = . We call this kind of deformation the late
E
residual deformation.

We will introduce the operator of the effective modulus 〈E〉 for time variable
processes of the deformation. For the operator of the effective model it is given

d ⎛ d⎞
〈E〉 = E + η = E⎜1 + τ ⎟. (4.3)
dt ⎝ dt ⎠

©Alexander Dubcek University of Trencin 2007 77

We get by performing of the harmonically changing tension the following:

〈E〉 = E + iωη = E(1 + iωτ ) , (4.4)

where ω is angular velocity.

The above mentioned model of the behavior of viscoelastic substances is called

Kelvin’s or also Kelvin – Voigt’s model.

Maxwell used different approach to the description of viscoelastic environments,

when he assumed that two types of deformations act in an environment, namely
elastic and viscous.

S = S E + Sη . (4.5)

By deriving the equation 4.5 and by substituting particular values we get equation

'σ' T
S = + . (4.6)
E η

If in time t = 0 the mechanic tension begins to perform in the form of rectangular

impulse of the height T0 in the Maxwell’s environment, which stays in the time
unchanged, the solution of the equation 4.6 will have the form (see the picture 4.1)

T0 T0
S= + t. (4.7)
E η

The equation 4.7 describes well attributes of many high polymeric materials like
colophony, asphalts and others. It is seen from the relation 4.7 that when the tension
impulse is short, material is behaving like elastic (the second element in the
equation is omissible). The fluxion of the material occurs only at sufficient length of
the performing of the tension impulse.

Now let in the time t = 0 the rectangular deformation impulse of the height S0
perform, which remains in the time unchanged, in this environment, then the process
of the tension dependent on time will be described with the following relation:

η
T = ES 0 e −t / τ ; τ= . (4.8)
E

We call the process of the exponential decrease of the mechanic tension in

dependency on time the relaxation of tension and the particular index τ is called the
relaxation time

©Alexander Dubcek University of Trencin 2007 78

f f

f0 f0

t t
s s

s∞
s0

t t
Figure 4.1. The response of Kelvin – Voigt´s and Maxwell’s environments to the
tension impulse

The operator of the effective modulus will have the following form in Maxwell’s
environment

d
η
〈E〉 = dt . (4.8)
d
1+ τ
dt

We get for the harmonic signal

iωη
〈E〉 = . (4.9)
1 + iωτ

If we split the expression for the effective modulus into real and imaginary parts,
we get

ω2 τ 2 η
〈E〉 = E + iω . (4.10)
1+ ω τ
2 2
1 + ω2 τ 2

The value

Eω 2 τ 2
Re〈E〉 = = E(ω )
1+ ω2 τ 2
(4.11)

is called the dynamic modulus

1 η
Im〈 E 〉 = = η (ω ) (4.12)
ω 1 + ω 2τ 2

is called the dynamic viscosity

It is seen from the relations 4.11 and 4.12 that at low frequencies (ωτ<<1) the
Maxwell’s environment is behaving like viscous liquid, because the dynamic
modulus is very small and the dynamic viscosity is independent on the frequency.
The dynamic viscosity begins to fall with the increasing frequency, the dynamic

©Alexander Dubcek University of Trencin 2007 79

modulus is growing and at high frequencies (ωτ>>1) the Maxwell’s environment is
behaving like stiff material. This process is called the relaxation of viscosity. The
flow of the dynamic viscosity and the dynamic modulus is in Figure 4.2.
E(ω),η(ω) E(ω),η (ω)

E∞ E∞
E(ω)
E(ω)
E0 E0
η η
η(ω)

η(ω)

0 1 2 3 4 5 -1,0 -0,5 0 0,5 1,0

ωτ lgωτ
Figure 4.2. The process of the dynamic modulus and the dynamic viscosity in the
Maxwell’s environment in the dependency on non-dimensional parameter ωτ

Some types of polymeric materials, like some types of chalcogen glasses are, have
the relaxation spectrum quasi interconnected, sums are substituted by integrals and
for instance we get for the complex Young’s modulus the following expression

H(τ )dτ
∞
E * = E(ω) + iωη(ω) = E 0 + iωη∞ + iω∫ . (4.13)
0 1+ ω τ
2 2

The value H(τ ) is called the density of the relaxation spectrum.

Now we are briefly going to mention mechanic models of above listed

environments, which enable to model real materials. The elastic environment is
presented in these models by a spring, and the viscous one by a cylinder moving in
a viscous liquid (Figure 4.3 a, b). These are basic elements.

a) b)
Figure 4.3. Models of the a) elastic and b) viscous environments

The Kelvin’s environment is possible to model by parallel-connected basic elements

herewith, that their junctor AB is fixed and the junctor CD moves along leads that
prevent the deformation of the system (Figure 4.4.) Then the extension of both basic
elements is equal and total power acting on the model is equal to the sum of powers
acting on both models. Thus both elastic and viscous tension in the Kelvin’s
environment is composed. We can see from the model that the immediate
deformation of such model would require infinitely big power, which is not possible.

©Alexander Dubcek University of Trencin 2007 80

With gradual deformation by a constant tension it is gradually achieved a constant
value of the deformation. The Maxwell’s environment is possible to model by series
connection of the basic elements in that way as it is seen in the Figure 4.5. The same
power acts on both elements in this case and their extensions are comprised. Likewise
we can see as in the previous case that if the constant mechanic tension acts in these
models, the material can flow.

C D

A B

Figure 4.4. The model of the Kelvin – Voigt´s environment

Figure 4.5. The model of the Maxwell’s environment

We often read in literature also about electric models of environments to which are
given the following assignments T → U, S → Q, S‘ → I, where U, Q, I are the
electric tension, electric charge and intensity of electric current consequently. The
elasticity is presented by the condensator with the capacity C = 1/E and the viscosity
at the resistance R (Figure 4.6.a, b).

R
R C

a) b)

Figure 4.6. The electric model of the Maxwell’s environment a) and the Kelvin –
Voigt´s environment b)

Of course the real viscoelastic environments with their attributes differ and thus it is
often necessary to use variety of series parallel connections of the spring and the
piston. As an example we are going to analyze therefore the series – parallel model
according to the Figure 4.7. Reological equation for this model has the following
form

Tv = ES + (Eτ + η )S ' . (4.14)

©Alexander Dubcek University of Trencin 2007 81

The effective modulus is expressed in the following form

d
η
dt
E = E0 + . (4.15)
d
1+ τ
dt

E1
E0

η1

Figure 4.7. The series – parallel model

S

S0

f0

f∞

Figure 4.8. The relaxation of the model tension according to the Figure 4.7.
f

f0

S
S∞

S0

t
Figure 4.9. The time dependence of the deformation in the model according to the
Figure 4.8

We get for the sinusoid deformation the equation

iωη
E = E0 + . (4.16)
1+ iωτ

As we can see from the equation 4.16, the mechanic tension in this environment is
comprised of the elastic tension commensurable to the deformation and the relaxing
tension. The time dependences of tensions and the deformation are illustrated for
this model in the Figures 4.8 and 4.9.

©Alexander Dubcek University of Trencin 2007 82

In conclusion we feature in the chart 4.1 the review of the most used models of
viscoelastic environments together with their typical parameters.

Table 4.1. The review of the chosen models of the viscoelastic environments
Model Scheme Components of complex compliance
Maxwell’s model 1
C′ =
E
E
1
C ′′ = ωτ
η E
Voigt´s model 1
C′ =
E[1 + ω 2τ 2 ]
ωτ
E

C ′′ =
η

E[1 + ω 2τ 2 ]

Modified ⎡ 1 ⎤⎧ E1 E 2 ⎫
Maxwell´s model C′ = ⎢ ⎥ ⎨1 + 2 ⎬
⎣ E1 + E 2 ⎦ ⎩ (E1 + E 2 ) ω τ 2 − E1 ⎭
2 2 2
E2

E1ωτ 2
E1

η0
C ′′ =
(E1 + E2 )2 ω 2τ 22 − E12
Modified 1 1 ⎡ 1 ⎤
Voigt´s model C′ = + ⎢ 2 2 ⎥
E1 E 2 ⎣1 + ω τ 2 ⎦
E1

E2
1 ⎡ 1 ⎤
C ′′ = +⎢
η

2 2 ⎥
E 2 ⎣1 + ω τ 2 ⎦
Maxwell´s 1 1 ⎡ 1 ⎤
Voigt´s model C′ = + ⎢ ⎥
E1
E1 E 2 ⎣1 + ω 2τ 22 ⎦
E2
1 1 ⎡ ωτ 2 ⎤
η
C ′′ = + ⎢ ⎥
E1ωτ 1 E 2 ⎣1 + ω 2τ 22 ⎦

strain
C* = = C '+iC ' ' ; τ i = ηi / E i [i = 1,2]
stress

References:

[1] Wineman, A. R., Rajagopal, K. R., Mechanical response of polymers,

Cambridge University Press 2000
[2] Rosen, S.L., Fundamental Principles of Polymeric Materials, Wiley, 1993
[3] Ferry,J.D., Viscoelastic Properties of Polymers, Wiley, 1980
[4] http://hcgl.eng.ohio-state.edu/~ce552/3rdMat06_handout.pdf

©Alexander Dubcek University of Trencin 2007 83

Fundamental questions from present part:

1. Describe basic physical assumptions of Maxwell model.

2. Describe basic physical assumptions of Voigt model.
3. Explain the map line of the dynamic modulus and dynamic viscosity in
dependency of frequency.

©Alexander Dubcek University of Trencin 2007 84

 CHAPTER 5
Selected physical properties of polymeric materials

Objectives to achieve
The aim of this part is to provide readers with a schematic review of results of
measurements of selected physical quantities (especially mechanic and thermal)
gained for various polymeric materials and rubber as well as their connection with
theoretical background knowledge presented in previous parts. At the same time we
are going to point out differences between physical properties of elastomers and
polymers.

In the Figure 5.1., the dependence of the mechanic tension on a deformation for
polyethylene foils is displayed. The stiffening of the material comes up by a relative
elongation over 300 percent, and this is characteristic for a lengthening. It is
inevitable action, which can only be eliminated by long-time tempering.

2,0
c

σ a
1,0

0 200 400 %
ε
Figure 5.1. The tension (stress-strain dependence) line of polyethylene foil.
The middle polymerising degree (a), 2050(b), 2600 (c) [1]

As we can see in Figure 5.2., the maximum of the relative elongation for the
dependence of the mechanic tension on the temperature can be observed at the
temperature of 50ºC for a polyethylene foil and at the same time the fall of solidness
σ pt . The above-described dependences for polystyrene and polyvinyl chloride and
composites on the basis of phenol-formaldehyde resins (Figure 5.6.) are displayed
in Figure 5.3. up to 5.5.

©Alexander Dubcek University of Trencin 2007 85

 3,5 600

%
ε
3,0 600

2,5 500

MPa

2,0 400
σpt

σpt ε
1,5 300

1,0 200

0,5 100

0 0
-30 0 30 60 °C 90
T

Figure 5.2.The temperature dependence of the relative elongation ( ε ) and the

stiffness ( σ ) in tension for polyethylene foil [1]
6 60
-7
Pa x 10

σp σ εε
t Κ

4 40

ε
σ
σptpt

2 20

ε
σΚ
σpt
0 0
-25 0 25 50 75 °C 100
T
Figure 5.3. The stiffness in tension σ Pt , slide limit σ K and relative elongation ε
as temperature functions for polystyrene [1]

©Alexander Dubcek University of Trencin 2007 86

 400

300

ε 200

100

0
-20 0 40 80 120 140
T °C
Figure 5.4. The temperature dependence of relative elongation and the stiffness in
tension for polyvinyl chloride [1]
5
-7

4
Pa x 10

3
b

σ
2

0 5 10 15 20 % 25
ε

Figure 5.5. The tension line for glassy polystyrene (a)

and polystyrene with rubber (b) [1]

©Alexander Dubcek University of Trencin 2007 87

 25

-7
20

Pa x 10
a b

15

σ
10
c1
c2
d
5

0 1 2 3 4 % 5
ε
Figure 5.6. The tension line of layered phenol – formaldehyde resins
Crosswise put paper reinforcement (a), crosswise put textile reinforcement (b),
crosswise put cotton reinforcement (c1 a c2), crosswise put asbestos (d) [1]

The all above mentioned substances showed in the process of heating under the
influence of a flow the declining of the modulus of elasticity, whereas this behavior
is for ideal rubber materials, especially within quick measuring of module, opposite.
The growth of the modulus of rubber materials at lengthening is visible in Figure
5.7. We can see that also tension rises quickly together with increasing deformation
and in the initial phase the line has a shape of the letter S. We watch a fall of the
module E with increasing temperature for real rubber at longer lasting tests that is
caused by molecular slides, which are counted with static setting of segments,
which are responsible for ideal rubber elasticity.

At measuring of the impact toughness in the dependence on temperature there is an

obvious influence of the frozen micro-brown movement and a transition to an
athermical, that is, fragile fracture (see Figure 5.8.). The impact toughness is not
possible to determine in many cases in the rubber area, because these substances are
leathery though or rubbery soft. The results of notch toughness are necessary to
distinguish from the impact toughness, because the maximum tension arises in 1-2
mm thick notch that leads to the start of a fracture. The considerable part of energy
is bound as elastic at setting of the impact toughness, what does not happen at the
notch toughness.

©Alexander Dubcek University of Trencin 2007 88

 60

50

40

σ 30

20

0 200 400 600 % 800

ε
Figure 5.7.The tension line of rubber materials [1]

220 a - metylester of polymetacryl acid

b - polystyren
5
Pa.m 10

200 c - nitrocellulose
d - polyvinyl PVC
180

160
a
140

120

d
100
specific strain energy

c b
80

60

20

0 20 40 60 80 100 120 °C 140

T
Figure 5.8. The impact toughness in dependence on the temperature [1]

It is possible to achieve relaxation spectra from the study of the frozen micro-brown
movement with the help of oscillation attenuations. As we can see in Figure 5.9, it
is possible to watch the maximum of logarithmic decrement for PVC at frequency
around 1Hz, what is caused by its softening that enables slides of molecules. It
leads to loss of rubber solidness, embitterment of material and the growth of the
module G at the fall of the temperature, when the micro-brown movement is frozen.

©Alexander Dubcek University of Trencin 2007 89

 G Λ

G
9
10
N
2
m 2,5

8
10
2,0

7
10
1,5

6 1,0
10

Λ Λ
0,5

0
-50 50 100 150°C
Figure 5.9. The shear module G and logarithmic decrement Λ as a function of
temperature for PVC [1]

Figures 5.10 and 5.11 gradually show the shear module and the logarithmic
decrement of fall off of slow torsional oscillations for different levels of rubber
polymerization. It comes to transition of a dispersive area to higher temperatures
with shortening of moving parts of chains by adding sulphur.

Figure 5.10. Temperature dependence of endo-heat for textile cords [unpublished

results]

©Alexander Dubcek University of Trencin 2007 90

 I
E
II
E
tan δ
100 0.30
0.28
70
0.26
50
40 0.24

30 0.22
0.20
20
0.18

tan δ
E /E [MPa]

0.16
10
0.14
I

7 0.12
II

5 0.10
4 0.08
3
0.06
2
1 0.04
0.02

1 0.00
0.5 0.9 1.4 1.8 2.3 2.8 3.2 3.7 4.1 4.6 5.0

Figure 5.11. Strain dependence of dynamic modulus components and tan δ ,

Dynamic sweep: static strain 8%, limit 400N, dynamic strain 0,5-5%, limit 100N,
Dyn. Incr.:0,5%, contact force 1N, frequency 10 Hz, temperature 60°C, sample
shape – prism [unpublished results],

9
10

N
2
m

8
10

G 7
10

6
10

5
10
-150 -100 -50 5 50 100 150 °C 200

Figure 5.12. The shear modulus G of rubber as a function of the temperature for
different levels of vulcanization. Non-vulcanized (a), 0,5 per cent sulphur (b), 5 per
cent sulphur (c) [1]

©Alexander Dubcek University of Trencin 2007 91

 4,0

3,5

3,0

logaritmic decrement
2,5

2,0
c

1,5

1,0
b

a
0,5

0
-150 -100 -50 0 50 100 150 °C 200
temperature
Figure 5.13.The logarithmic decrement of rubber as a function of the temperature
for different levels of vulcanization. Non-vulcanized (a), 0.5 percent sulphur (b), 5
percent sulphur (c) [1]

We can see results of measurement of mechanical and dielectric loss angle for
polyvinyl propionate in Figure 5.14. The dielectric losses have the maxim at
temperatures about twenty degrees higher than in mechanical tests, what is caused
by the fact that electric dipoles are less movable than individual structural segments
creating polymer.

-2
26.10 60
dB/cm
tgδ
falloff of ultrasonic waves α
20 50
6,0 α
m/s ε
cZ
16 40
2400
dielectric constant ε
d

5,0 cZ
dielectric losses factor tan

12 30
2000

8 4,0 20
1600

4 10
3,0 1200
ε tgδ
0
-200 -150 -100 -50 0 50 100 150 200
temperature
Figure 5.14. The temperature dependence of the speed and the fall off of ultrasonic
waves as well as dielectric constant and dielectric losses at frequencies 2⋅106 Hz for
polyvinyl propionate [1]

In Figure 5.15 there is a frequency dependence of the speed of longitudinal elastic

waves in poorly vulcanized rubber. The speed of waves is still relatively low also in
the frequency above 2000 Hz and we can say that it does not come to significant

©Alexander Dubcek University of Trencin 2007 92

increase of the dynamic module yet, what is caused by orientation of molecules
against periodically changing power.
110

m/s

100
speed of longitudinal sound waves

90

b
80

70

60
0 500 1000 1500 2000
frequency HZ
Figure 5.15. The speed of longitudinal sound waves in poorly vulcanized rubber
fibres in dependence on frequency for different samples. (a, b) [1]

The measurements of dynamic viscosity of polymeric materials show at linear

dependence ln η versus reciprocal absolute temperature for most polymers that is in
accord with theoretical prerequisites. The characteristic temperature dependence is
shown in Figure 5.16.
200 M=2400pcm
8
400 1000 cP.10
200
1600
150

100
80

60
50
40

30

20
η
15

10
8

6
5
4

1,5
-5 -5
310.10 322.10
1
62 58 54 50 46 42 38 34
1/Tabs

Figure 5.16. The viscosity of polyvinyl acetate in dependence on temperature [1]

In Figure 5.17 there are results of measurements of the expansivity on polystyrene

samples for different degrees of polymerization. The step change of the expansivity
arises at the temperature of the glassy transition. The temperature of the glassy

©Alexander Dubcek University of Trencin 2007 93

transition rises continuously with the polymerization degree. If we take reciprocal
value of the temperature of the glassy transition as a function of reciprocal value of
molecular weight, we get again a linear dependence, as we can see in Figure 5.18.

2
l p=1
kg

1,1
3
4,1

8,1
specific volume
24,6
898,6

1,0

glass transition temperature

below TG

0,9
-100 100 200 °C
temperature

Figure 5.17. The dependence of specific volume of polystyrene fractures on the

temperature, where parameter is polymerizing degree P

3
10
T
5

2 -5
0 100 200 300 400 500 10 600
M

Figure 5.18. The dependence of the inverse absolute temperature of the glassy
transition on molecular weight [1]

©Alexander Dubcek University of Trencin 2007 94

 1,12
l/kg

1,08

amorphous
specific volume

1,04

partly crystallised

1,00

-40 0

temperature
Figure 5.19. The dependence of specific volume on the temperature for amorphous
and partly crystallized rubber [1]

In Figures 5.20 and 5.21 there are temperature dependences of the specific volume
and of the coefficient of the temperature expansivity. As we can see in these graphs,
below the temperature of the glassy transition there is the coefficient of the
temperature expansivity equal for both structural modifications.

-4
8.10
1/K amorphous
coefficient of the thermal

-4 partly crystallised
4.10
expansivity

0 temperature
Figure 5.20. The temperature dependence of the coefficient of thermal expansivity
[1]

In Figure 5.21 there is the temperature dependence of the specific volume of

polyethylene for different structural modifications. We can see in these graphs, the
specific volume rises within the whole monitored temperature range from the
crystallized form, where it is minimal, to amorphous, where it is maximal. This
trend is in accord with the theory of free volume, which explains kinetics of
structural processes in polymeric materials.

©Alexander Dubcek University of Trencin 2007 95

 1,3
l
kg

1,2
s
hou
orp
am
lline
1,1
rtly crysta
specific volum

pa
s and
phou
glassy amor ne
crystalli
in e
1,0 rtly crystall
and pa
glassy
ine
crystall
0,9

0,8
-280 -240 -200 -160 -120 -80 TG -40 0 40 80 120 160

temperature

Figure 5.21. The temperature dependence of specific volume of polyethylene on the

temperature for different structural modifications [1]

1,0
kcal
kg.deg

0,8

0,6
specific heat

0,4

0,2

0
-250 -200 -150 -100 -50 0 °C 50
temperature
Figure 5.22. The dependence of specific heat of natural rubber on the temperature
for amorphous (full line) and partially crystallized material (dashed line) [1]

Further we are going to deal with temperature dependences of specific heat for
selected polymeric materials. In Figure 5.22 there is the dependence of rubber
specific heat for partially crystallized modification and for amorphous modification.
Specific heats of polymers generally rise with rising temperature. The specific heat
has a transition temperature changes by steps in the area of the glassy transition.
Melting of crystallized areas causes the steep maximum watched on the partially
crystallized modification of rubber. We can see the influence of adding of softener

©Alexander Dubcek University of Trencin 2007 96

on the steepness of the transition at the temperature of the glassy transformation.
The influence of the softener proves in the way that the transition becomes less
steep.
0,45
k.cal
kg.deg

0,40

specific heat 0,35

0,30

0,25

0,20
0 20 40 60 80 100
temperature °C
Figure 5.23. The dependence of the specific heat on the temperature of softened and
non-softened polyvinyl chloride. The pure polyvinyl chloride (black and white
circles), 10 percent trikresyl-phosphate (white circles), 30 percent trikresyl-
phosphate (black circles) [1]

In conclusion, we mention the characteristic dependence of the temperature

conductivity on the temperature for natural rubber. The triple increase of the value
of the temperature conductivity within temperature interval -80ºC up to -50ºC is
remarkable. Below and above this interval there is a direction of the measured
dependence regularly equal as for other polymers. The hysteretic behavior of the
temperature conductivity is also interesting. In the same picture there is also the
temperature dependence of the specific heat that has also the step change in the
above described critic interval of temperatures (Figure 5.24)
0,12 0,5
k.cal k.cal
m.h.deg kg.deg

0,10 0,4

0,08 0,3
temperature conductivity

specific heat

C
0,06 0,2

0,04 0,1

0,02 0

0
-180 -160 -120 -80 -40 0 20

Figure 5.24. The dependence of temperature conductivity and specific heat on the
temperature for natural rubber [1]

©Alexander Dubcek University of Trencin 2007 97

Other specific properties of rubber blends are presented in Figures 25 –33.

-20

-40
Tg C

-60
o

-80

-100

0 20 40 60 80 100

% 1,2 - SUBSTITUTION (vinyl)

Figure 5.25. Influence of 1, 2-vinyl substitution of butadiene rubber on the glass
transition temperature of BR rubber
0
10

-1
10
tan δ

-2
10

-3
10
0 2 4 6

Sulfur content in %
Figure 5.26. Influence of the cross-linking degree of rubber on the loos tangent tanδ

©Alexander Dubcek University of Trencin 2007 98

 Figure 5.27. Influence of CB grade on dynamic properties of rubber

Figure 5.28. Dynamic modulus G* and tan δ vs. temperature in E-SBR (0.1%
dynamic strain, 10 Hz, Gabo Explexor) [3]

©Alexander Dubcek University of Trencin 2007 99

 Figure 5.29. Carbon black and silica vs dynamic properties of E-SBR. RPA-
measurements/ J. Frohlich et al. (2005, same as above)

Figure 5.30. Dependence of vulcanizate property on crosslink density

©Alexander Dubcek University of Trencin 2007 100

A typical DSC curve of an elastomeric compound

Figure 5.32. Thermogravimetric analysis of elastomeric compound

penetration
∆L

expansion

Tg Ts

-43 -40
o
T( C )
Figure 5.33. Thermomechanical and dilatometric analysis of chloroprene rubber

©Alexander Dubcek University of Trencin 2007 101

References:

[1] Holzmüller, W., Altenburg, K., Physics of Polymer, SNTL Prag 1966, (czech
version)
[2]M. Schubnell, J. E. K. Schawe, Mettler Toledo GmbH, Sonnenbergstrasse 74,
CH-8603 Schwerzenbach, Switzerland
[3]J. Frohlich*, W. Niedermeier, H.D. Luginsland, Composites: Part A 36 (2005)
449–460

Fundamental questions from present part:

1. Explain main differences in behaviour of viscoelastic and elastomer

materials according to figures from this chapter.
2.Explain main differences of static and dynamic dependences of choosen
physical parameters on the base of study material from this chapter.
3. Explain characteristic points form dependence of specific volume vs.
Temperature for amorphous and partly crystalline polymer.
4.What is the influence of sulphur content in rubber compounds on tan δ ?
5.Explain the difference in dependence modulus vs. strain for fillers carbon
black and silica.

©Alexander Dubcek University of Trencin 2007 102

 CHAPTER 6
Electrical properties of polymers

Objective to achieve
Introducing the specific resistivity and conductivity. Dielectric properties of
polymers, electrical stress of polymers. Percolation threshold.

6.1. Electrically Conductive Polymers

Electrical conductivity of almost all polymers in pure state is very low. It is defined
from fundamental relation, which describes the dimension dependence of electrical
resistance in the form

ρ .l
R= (6.1)
s

where ρ is specific resistance, l is conductor length, s is cross-section of

conductor. From this relation we can obtain specific resistance in form

R.s
ρ= (6.2)
l

In some cases it is more convenient to introduce specific conductivity as an inverse

value of specific resistance in the form

σ=
1
ρ
[Ω −1
.m −1 ] (6.3)

Values of specific resistance range from 1010 to 1014 Ohm.m . From this point of
view we can regard polymer materials as very good insulants. Certain increase of
electrical conductivity is caused mostly by presence of dirt, that is why you have to
work in dustless environment e.g. when working on preparation of insulants with
extremely high specific resistance. Presence of microheterogeneities in structure
offers creation possibility of free charge bound onto its own structure. Such
microheterogeneities can be boundary lines between amorphous and crystalline
phase, where thermal activation of electric charge can occur. This increase in
conductivity has ionic character and occurs as result of present dirt dissociation.

Despite the fact that most polymers are dielectrics with very good insulating
properties, there are some inherently conductive polymers, too. Those are mostly
macromolecules with system of conjugated bonds in string. When we expose these
polymers in strong oxidizer or deoxidizer they become significantly more
electrically conductive. Most widely known conductive polymers are primarily
trans-polyacetylene, then polypyrrole, polythiophene (and its derivates on the basis
of poly-3-alkyltiophenes), polyaniline, poly-p-phenylene and some other. These
materials have found important applications, mostly if you need to use conductive

©Alexander Dubcek University of Trencin 2007 103

materials in corroding medium, where you cannot use metals. Problem is the fact
that conjugated structures are formed by polymer strings with low mobility and
therefore polymers with such structure are practically unworkable by common
technologies. Forming directly during synthesis (e.g. creation of conductive
polymer layer on electrode surface) is needed for their application, or application in
the form of colloids or composites, where conductive polymer acts as the filler.

In certain circumstances even polymer without conjugated bonds can be highly

conductive. E.g. when we affect polyisoprene with iodic vapours we can get
increase of his conductivity to as many as 10-1 S/cm. But opinions on charge-
transfer mechanism in such systems are still very divergent.

6.2. Electrically Conductive Composites

Other possibility how to get electrically conductive materials on the basis of

polymers is preparation of electrically conductive composites. The principle is very
simple. You have to mix electrically conductive filler into non-conductive
thermoplastic die, what will ensure significantly increased conductivity. In contrast
to electrically inherently conductive polymers, advantages of these materials lie in
the fact that they can be formed with technologies used for thermoplastics, even
though the high proportion of filler and high viscosity (that is associated with it)
make preparation of composite by mixing of components above melting
temperature of polymer die, as well as following thermal forming very difficult. For
fillers we can basically use any particles or fibres that are electrically conductive.
Most commonly used are various forms of hydrogen, mostly soots (carbon black)
and graphite, metal elements (we can use colloid elements of silver to get extreme
conductivity on the level of metals), as well as conductive polymers in the form of
powder. Elements of conventional polymers coated with layer of metal or
conductive polymer are another interesting possibility. These are advantageous,
because of their similar density to die, if we want to use the filler in a melt or
solution (e.g. for conductive adhesive) and we want to prevent filler sedimentation
while stocking.

In terms of conductivity we observe an interesting effect at conductive composites

and that is existence of so-called percolation threshold. It is manifested by intense
increase of conductivity by few orders in narrow interval of conductive filler
concentration. Explanation of this phenomenon lies in conception of creation of
conductive connections through whole geometrical section of observed body.
Conductivity is proportionally increased along with gradual increase of conductive
filler concentration, as shown in Figure 1, what is caused by quick skip of electrons
over bordered areas in material that are formed by clusters of conductive filler.
Number of these agglomerates and their size is gradually increasing along with
growth of filler concentration, while conductivity is growing more or less linearly
within the range of circa 2 – 3 orders.

©Alexander Dubcek University of Trencin 2007 104

Figure 6.1. Volume resistivity against conductive filler and carbon fibre loading for
NR and EVA based composites.

At specific concentration is first conductive connection formed through whole

section of the body. This fact produces completely new situation when electron can
run across the whole section without having to overcome discrete areas consisting
only of non-conducting die. The area of concentration where this phenomenon
occurs is called percolation threshold. Conductivity increases within addition limits
of circa 1 – 3 % of filler by at least four, but also by six orders. When we add more
filler, number of created conductive ways and conductivity is still growing, but
practically again relatively slowly and proportionally to the growth of filler
concentration. This consecutive increase presents another 2 – 3 orders within
concentration change of 15 – 30 % of filler volume.

Concentration of percolation threshold depends heavily on filler and die properties.

Generally we can say that percolation threshold is rapidly decreasing along with the
growth of conductive elements anisotropy. Inclination of filler to the creation of
agglomerates also leads to decline of percolation concentration. In this regard,
carbon-fibres (anisotropy) and soots (aggregate creation) are advantageous for low
percolation threshold of composites, mostly those with tendency to create
agglomerates of irregular or even chain shape. Graphite particles or metal regular
particle isotropic formations are less advantageous. E.g. you can reach percolation
threshold in polyethylene die with highly conductive soots Chezacarb already at
concentration around 2 hm %. On the other hand, we have to mention that
extraordinary tendency of these soots to agglomerate into anisotropic formations
leads to quick growth of melt viscosity and substantial degradation of composite
workability even at filler volume of only a few percents.

Composite conductivity and particularly percolation concentration significantly

depends even on die quality, and mostly on its crystalline portion. Regarding
amorphous polymers, whole filler is adequately dispersed into entire volume, while
at semicrystalline dies filler is arranged only in amorphous phase. Percolation
threshold thus can be reached with lower filler concentration than in the case, when
die is formed only by amorphous polymer. Percolation threshold for e.g. atactic
polystyrene is at concentration of circa 17 % of volume, for medium-crystalline

©Alexander Dubcek University of Trencin 2007 105

low-density polyethylene it is around 8 – 10 %, while for highly crystalline low-
density polyethylene it is only circa 5 %. These data are, of course, only for
orientation and they significantly depend on filler quality as well.

As we have mentioned before, sudden increase of conductivity in the area of

percolation threshold is caused by creation of conductive ways that consist of
conductive filler over entire section of experimental body. It can be easily
understood that during mechanical deformation of conductive composite with
rubber die and volume of conductive filler over percolation concentration comes to
decline of conductivity, because quite rigid conductive connections can be
relatively easily disrupted during uniaxial deformation. On the other hand,
consequential return into former state after stress release mostly does not lead to
restoration of former conductivity before deformation force application, but
proportions are much more complicated. Similar effects of conductivity change can
be also observed at the change of conductive composite dimensions in consequence
of thermal expansivity at warming or cooling of the material.

Another important feature of these materials is substantial embrittlement at filler

volume increase. It can be characterized e.g. by value of elongation at abruption. It
is interesting that area of concentration, where steepest decline of elongation occurs,
corresponds to percolation area with substantial increase of conductivity. The
explanation is based on reasoning, that creation of conductive ways from filler
elements presents also facilitated way for the growth of mechanic fissure at the
application of deformation force. Fissure is thus created at lower stress and its
growth is much quicker, if only there is a connection structure of body through
whole volume.

6.3. Influence of Force Field on Polymer Behaviour

Under the term force field we will in this chapter understand mostly the influence of
electric, eventually magnetic field. From this point of view we can define two types
of substances and that is

Electric current conductor that allows irreversible transport of electrically charged

elements of particle character.

Dielectric or insulator, where electric dislocation emerges that is caused by creation

of induced dipoles, or by orientation of dipoles yet present in material, while this
dislocation is accompanied by energy accumulation and is reversible after field
disposal.

Transition between these two groups of substances is fluent and external conditions
such as temperature are also influencing the properties. Divisions according to the
electric conductivity value σ is approximately as follows: metals with conductivity
of 102 to 106 S/m (Siemens per meter), semiconductors, which have conductivity
within the interval of 10-8 to 102 S/m and insulators with lower conductivity than 10-
8
S/m, while common insulators reach values of 10-12 to 10-15 with the limit
somewhere around 10-30 S/m. From practical point of view is group of antistatic
substances an important one. It has sufficient surface conductance for dispersion of

©Alexander Dubcek University of Trencin 2007 106

electrostatic charge that is created e.g. at friction. Conductivity of these substances
is in the sphere of less conductive semiconductors, i.e. around 10-6 S/m. Change of
conductivity along with temperature is also worthy of note. Metal conductivity is
decreasing along with growing temperature, while conductivity for substances from
next two groups is with temperature growing. For the growth of conductivity with
temperature we can use Arrhenius type formula

σ = σ∞ exp (-Ea / RT), (6.4)

where σ∞ is anteexponential factor with conductivity function at very high

temperature and Ea is activation energy of conductivity. However equation 10.1
suits for polymers only in narrow range of temperatures.

6.4. Electrical Strength of Polymers

As we have mentioned before, substantial quality of dielectric is its insulating

power, i.e. the ability to resist electric field influence. But this quality depends on
external conditions; in this case it is mostly intensity of influencing electric field.
Absolute intensity is reached at gradual increase of field intensity and dielectric
conductivity will arise to values that are characteristic for conductors. This state is
from outside manifested by dielectric breakdown. Electric voltage, corresponding to
field intensity is called breakdown voltage Up. When it is applied to thickness, it
characterizes electric strength Ep = Up / d, where Ep has parameters of electric field,
i.e. V /m.

Breakdown mechanism is difficult and sometimes it does not depend on thickness

and time. Most commonly you get relative datum Ep by direct measuring and this
datum depends on sample geometry and measuring conditions. Electric strength
cannot be for this reason considered as material characteristics and we have to use
standardized experiments for its determination.

Electric strength of polymers depends besides external conditions (temperature,

body geometry) on polymer structure. Generally we can say that electric strength is
growing along with decrease of string elements mobility and along with the growth
of molecule cohesion. From this point of view it decreases in following order: e.g.
polyvinyl-alcohol > polymethyl-methacrylate > polystyrene. It is lowest for non-
polar polymers with high mobility of segments, e.g. polyolefins and rubber.

On the basis of what has been said before, the observed decrease of Ep in the area
above temperature of glassy transition where mobility of strings differs substantially
appears to be logical. Ep grows with the grade of arrangement as well, e.g.
crystallinity. From the molecular point of view we can stress out the indirect
proportion to free volume.

©Alexander Dubcek University of Trencin 2007 107

6.5. Dielectric Properties of Polymers

Polymers are typical dielectrics with low electric conductivity and relatively high
electric strength. Every dielectric is outwardly electrically neutral, because it
contains equal number of positive and negative charges in dipoles. However dipole
orientation in the direction of influencing field occurs in electric field, it is so-called
polarization of dielectric.

If we connect voltage supply U to the plate condenser, it will be practically at once

charged with charge +Q or –Q, while the quantity of charge Qo = Co U is directly
proportional to the electric voltage U and constant of proportionality Co with the
size of Farad, F = C x V-1, is called vacuum capacity of condenser. Surface density
of electric charge is thus directly proportional to the field intensity between plates E
= U / d.

If we fill the space between plates by dielectric, polarization occurs by the action of
electric field. Capability rate of dielectric to polarize itself is quantity called electric
polarization P that is given by formula

P = D − ε0E (6.5)

where D is electric induction (D = εE, while ε presents dielectric permittivity) and

εo is vacuum permittivity. These quantities are dielectric’s capability rate to polarize
itself, to increase condenser capacity and to decrease force effect of electric field.

Generally, dielectric polarization is function of time. This quantity consists of more

components that differ in time reaction on the field change. Electric polarization Pe
occurs on the level of particular atoms and is caused by transition of external
electrons from electric shell. Dipoles in time interval of 10-15 s order are created in
this way. Atomic polarization runs slower, in time 10-13 s, and it is created by
influence of electric field on polar bond, whereby mutual distance of atoms in bond
and bond angle are changed. Electronic and atomic polarization present so-called
deformational polarization Pd. This can be considered as prompt at frequency of
field polarity change lower than 1012 and origin of such induced dipoles will be in
phase with changing electric field. The most important of these polarizations is so-
called elastic dipole polarization that is caused by elastic dipole deflection in
electric field from their equilibrium positions with lowest energy, which dipoles
occupy in the environment of adjacent molecules out of field influence. These three
polarization types are included in the common term of elastic polarization and it is
characterized by quick, practically immediate adoption of particular state after
electric field application and practically immediate return into its former state after
the field is disrupted.

But there are permanent dipoles in many substances, too, that are formed by polar
bonds between two atoms with different electroaffinity. Bonding electrons e.g. in
the carbon – chlorine bond are attracted more to chlorine atom with heavier
positively charged nucleus. Generally this bond is permanent dipole with partial
positive charge on carbon. In contrast to elastic dipole polarization we assume that
permanent dipole can also have more equilibrium positions that are qualitatively

©Alexander Dubcek University of Trencin 2007 108

different. Change of these positions is conditioned by overcoming of potential
barrier. If such a permanent dipole gets into electric field, it will try to orientate
itself in the field direction and so-called orientation polarization Po occurs. But this
action is slowed down in consequence of interaction with adjacent groups and thus
it is time-dependent. Overall time flow can be at most substances, including
polymers, described by three-parameter function

P(t) = Pd + Po(1 – e-t / τ), (6.6)

where Po is limiting value of orientation polarization that will occur at t → ∞ and τ

is so-called relaxation time.

If condenser is in static electric field with constant intensity, electric induction will
be monotonous function of time. But if we change field intensity with certain
frequency, then polarization does not follow sudden field changes immediately, but
with particular delay. Development of field intensity can be written in the form

E(t) = Eamp cos ϖ t . (6.7)

Electric induction is also sine function with the same frequency, but delayed by
time period that can be expressed by phase angle

D(t) = Damp cos (ϖ t - δ). (6.8)

For every frequency ϖ are amplitudes in relation

Damp = εo ε(ϖ) Eamp, (6.9)

where ε(ϖ) is frequency dependent dielectric permittivity. (Permittivity is the

criterion of polarizability and equals to surface density of charge / field intensity.)
Electric induction can be divided into two vector segments,

D(t) = D’ cos ωt + D’’sin ωt, (6.10)

one of them is in phase with influencing field and other one is displaced by 90o.

While

D = εa ε (ω) E, (6.11)

we define vector components ε’ and ε’’ that are also frequency dependent and are
defined

ε’ = ε cos δ and ε’’= ε sin δ. (6.12)

First one characterizes real component of permittivity, i.e. its relative capacity and
other one represents loss properties of dielectric.

Deformation polarization components are determined in time 10-13 - 10-15 s and they
manage to adapt to the field change. Relaxation time is higher at orientation

©Alexander Dubcek University of Trencin 2007 109

polarization, and that is for low-molecular substances 10-8 up to 10-12 and for
polymers even higher. At small frequencies, both polarization components
(deformation and orientation) manage to watch field changes, time delay is zero and
δ = 0. Orientation component at high frequencies practically cannot respond to field
change, ε’ (∞) = ε∞, and losses are zero again, ε” = 0. Between these extreme limits
lies area, where values are non-zero and it is possible to measure relaxation areas
characteristic for specific material. In contrast to elastic polarization that occurs
immediately, in the case of orientation dipole polarization we assume that dipole
has more equilibrium positions. It can change these positions after potential
energetic barrier overcoming, e.g. thermal energy.

If we gradually increase field frequency, then we get from the area, where both
components manage to watch changes and we observe maximum and constant
induction, into the area where loss element is delayed by constant time period and
finally at high frequencies we get back to the area where induction is in phase,
because loss element cannot by any chance respond to field change. Frequencies,
phase transitions are observed at (viz. demonstration of loss element) are
characteristic for specific structure, what means that motions of dipoles liable to
orientation polarization are influenced by material structure. If the structure changes
e.g. by temperature change, addition of free volume influencing additive, change of
strings arrangement (e.g. branching and such), then we also observe changes in
position of registered motions on frequency axis, as well as in the level of response
characterizing intensity of particular motions. In this way we can register relaxation
spectrum, interpretation of which is qualitatively equal to the spectrum acquired by
DMTA method, thus it reflects presence of segments with various structure. That is
why this spectrum can be used for observed material characterization. We have to
say that by this method we will get relaxation spectra that are usually much more
complex than spectra acquired by dynamic-mechanical analysis. Thus they offer
more information on studied material, but on the other hand their interpretation is
more difficult as well. This method is less spread, comparing to DMTA, and that is
also because of experiment requirements and advanced technical difficulties on
measuring devices.

References

[1] Job, A.E., Oliveira, F.A., Alves, N., Giacommetti, J.A., Mattoso, L.H.C.:
Conductive composites of natural rubber and carbon black for pressure sensors.
Synthetic metals, 135-136 (2003). P.(99-100).
[2] Bertetto, A.M, Ruggiu, M.: Low cost resistive based touch sensor. Mechanics
Research Communications 30 (2003) P.(101–107).
www.elsevier.com/locate/mechrescom
[3] N.C. Das, N.C, Khastgir, D., Chaki, T.K., Chakraborty, A.: Electromagnetic
interference shielding effectiveness of carbon black and carbon fibre filled EVA
and NR based composites. Received 10 September 1999; received in revised form 6
March 2000; accepted 23 March 2000.
[4] Sau, K.P., Chaki, T.K., Khastgir, D.: Carbon fibre filled conductive composites
based on nitrile rubber (NBR), ethylene propylene diene rubber (EPDM) and their
blend. Rubber Technology Centre, Indian Institute of Technology, Kharagpur,

©Alexander Dubcek University of Trencin 2007 110

India. (Received 12 May 1997, revised 4 August 1997, accepted 10 September
1997).
[5] Volf, J. – Holý, S. – Vlček, J.: Using of Tactile Transducer for Pressure-
Distribution Measurement on the Sole of the Foot. In: Sensors and Actuators (A),
1997, No. 62, P. ( 556-561).

Fundamental questions from presented part

1. What does cause that conductivity is proportionally increased along

with gradual increase of conductive filler concentration?
2. What does the composite conductivity depend and particularly
percolation concentration on?
3. What is polarization of dielectric?
4. What we can say in generally about electric strength?
5. What are inherently conductive polymers?

©Alexander Dubcek University of Trencin 2007 111

 CHAPTER 7
Physical Processes Influencing Surface Contact of Two
Materials

Objective to achieve
This chapter presents hysteretic attributes of viscoelastic materials in the
process of their deformation. Some theoretical approaches of hysteresis
explanation from the point of view of solids' contact are discussed. It also deals
with gluing questions and adhesion at mutual contact of materials problems, as
well as with theory of friction.

7.1. Hysteresis

Hysteretic behaviour is substantial phenomenon we are meeting with not only at

deformation of viscoelastic materials but also at other physical phenomena, such as
ferromagnetism and others. Hysteresis curve in all cases limits the area representing
the loss - dissipation of corresponding energy. Hysteresis is also significantly
influencing total friction at the contact of viscoelastic material with different
material, e.g. metal. But now let us deal with hysteretic attributes, as we can notice
them at deformation of viscoelastic materials.

If the solid is at tension test by action of stress σ elongated by ε, consumed work at

weight equals integral ∫ σdε.

Consumed work (horizontally hatched area on Figure 7.1) is partially regained

during releasing of weight. Two energies differ by surface area of hysteresis loop.
Energy of deformation consists of elastic part (sparsely hatched area on Figure 7.1)
and losses transformed to heat. There is reversible flow besides perfectly elastic
deformation (curve a, at return motion a´). Consumed work is during reversible
deformation stored in the solid, partially as potential energy (elastic deformation
associated with internal energy absorption) and is partially used for reduction of
entropy occurring at deformation. Energy absorption is especially important at
periodic deformations, when weight and releasing of weight occurs at every cycle.

©Alexander Dubcek University of Trencin 2007 112

 observed curve

perfectly elastic deformation

Figure 7.1. To the explanation of hysteresis. Relative elongation dependent on stress

[27]

7.1.1. Theories of Hysteresis

In this section we will discuss some theoretical approaches of hysteresis explanation

from the point of view of solids' contact. Hysteresis along with adhesion (we will
talk later about it) contributes to friction among sliding elastomer and hard
supporting surface. Adhesive component of friction is in most applications
dominant due to molecular-kinetic phenomenon of sliding and adhesion alternation.
Adhesive component of friction shows very small figures on lubricated surfaces
because it is effectively suppressed by interfacial lubricating film. In such cases we
have to rely on hysteretic component of friction to create tractive power.

Mechanism of hysteresis can be best described as voluminal phenomenon with

definitive effect on sliding interlayer. Hysteresis has its origin in viscoelastic
attributes of polymers and thus certain final motion speed of elastomer on the
surface has to be developed (in contrast to adhesion). The amount of hysteretic
component of friction depends on the type of surface macroasperity and sliding
speed along with viscoelastic material attributes and operating temperature.
Macroasperity can be characterized as the combination of spherical, cylindrical and
conical irregularities describing geometric irregularities of specific surface, whether
it is road or abrasive paper. In following sections we will analyze major results of
theoretical approaches related to hysteresis.

7.1.2. Unified Theory

This theory provides result in the form of hysteretic friction coefficient fH for
various types of irregularities,

©Alexander Dubcek University of Trencin 2007 113

 ⎛ pn ⎞
fH = 4 ⋅ C ⋅ γ ⋅ ⎜⎜ n ⎟⎟ ⋅ tan δ , (7.1)
⎝E ⎠

whereby C and n are constants (see Table 7.2), p is surface pressure, γ is less than
one and at most equal to one and represents factor including density of surface
irregularities; E is Young’s modulus of elasticity and tanδ is the loss factor [7].

Table 7.1. Figures of the constants C and n for spherical, cylindrical and conical
irregularities

Shape of C N
irregularity
Spherical 81π 5 (1 − ν 2 ) 4 3
2,048
Cylindrical 256 2 ⎛ L ⎞ 2
⎜ ⎟(1 − ν )
2 3

3π ⎝ ⎠
3
R
Conical 3 3 15 2
2 2 9 ⎛ W ⎞ 2 ⎛ ω0 ⎞ 2⎛ R⎞ 2
(1 − ν 2 ) 2 ⎜ 3 ⎟
⎜⎜ ⎟⎟ ⎜ ⎟
3 π ⎝ ER ⎠ ⎝ Z0 ⎠ ⎝ε ⎠

Figure 7.2. Basic characteristics of irregularities used in unified theory [7]

According to this theory index n is between 2 and 3 for accidental irregularities,

presupposing that 3 involves the shape of irregularities that are on the interval
between never-ending and zero sharpness of irregularities. If we knew the exact
speed of elastomer motion, we could substitute Young’s modulus of elasticity E, in
the denominator of Equation 7.1, for complex modulus E*, representing numeric
value for real motion speed. When calculating deformed volume of elastomer Q we

©Alexander Dubcek University of Trencin 2007 114

neglect contact asymmetry in sliding direction, hence this effect increases the value
of adhesive friction coefficient.

7.1.3. Relaxation Theory

This theory of hysteresis is based on energy method of analysis and suggests simple
Maxwell's model of viscoelastic behaviour. It regards movement of elongated and
hard spherical solid upon the surface of elastomeric material. Surface of rigid body
is considered as smooth and thermal effects as constant on the surface of contact
during sliding. Negligible effect of adhesive friction component is also assumed.

Result of these considerations is inferred formula:

⎛β ⎞ ⎛ p⎞
f H = ξ ⋅ A.⎜ ⎟ ⋅ ⎜ ⎟ ⋅ tan δ , (7.2)
⎝α ⎠ ⎝ E ⎠

where ξ is linear ratio, A is numerical factor contingent on the shape irregularity

and ratio β /α characterizes the shape of irregularity (β >1 , α >1).

Out of the form of formula 7.2 for fH we can see, that it is analogical to the formula
7.1. Result of this theory is similar equation as the one in previous theory, despite
different approach. This analysis seems to be more rigorous than previous one. Also
with respect to the index in ratio (p/E), that reaches higher values in unified theory
while here it is equal to one.

It is remarkable that in every theory of adhesion and hysteresis is friction coefficient

proportional to tanδ, adhesive component is dependent on ratio (E/p) and hysteretic
component of friction is dependent on its reciprocal. It would be adequate to replace
modulus E in previous equations with complex value E* because of the change
possibility in speed and sliding temperature.

7.2. Gluing and adhesion

Another phenomenon significantly influencing dynamic behaviour of materials at

their mutual contact is adhesion. However, this phenomenon has much complex
use. Development of varnishes and glues required not only the knowledge about
their internal structure but also about the processes causing their adhesion to the
surface. It starts to shape up that from coupling forces it is mostly dispersion forces,
dipole moments and hydrogen bonds that determine firmness of glued joints. At the
same time we have to mention that there are remarkable differences between
calculated force values and measured firmness of glued joints that is influenced
mostly by following factors:

Macroscopic and submicroscopic asperities are practically allowing only partial

contact of glue lines. Therefore glue has to have as low surface tension as possible
to penetrate within limits into most of cavities and cracks.

©Alexander Dubcek University of Trencin 2007 115

Aqueous films, thin layers of grease, dissolvent and softeners have negative impact
on adhesion of glued layers.

Atmospheric pressure causes increase in firmness of adhesion.

If we want to glue large areas, we have to use diluted liquid glues or low-molecular
glues to ensure adhesion of glue to the surface. Curing of the glue occurs at the
evaporation of dissolvent and polymerization, only rarely at its solidification.

If surface of the solid is harsh or porous (leather, wood, plaster), glues must diffuse
quickly into the pores of material.

Situation is different with glass and metals. Evaporation of dissolvent must happen
before pressing of glued materials, but glue areas still have soft consistence.

Plastics during gluing are swelling and etching by action of the glue.

Volume shrinking occurs during solidification of the glue and also creation of
cavities and cracks, eventually creation of irregular surface and big shearing stress
decreasing joint firmness.

Creation of interfering interlayers often depends on absorption quality of glued

surface and glue. Hygroscopic are mostly alkaline glass areas binding water or its
vapour that is destroying glued joint. All plastic materials exposed to water for quite
a long time will transmit water vapour. Metals also incline to oxidation and creation
of hydroxides. Material under the layer of glue or varnish will corrode, bubbles will
occur or varnish will tear.

Above mentioned facts show that surface tension has important role during
realization of quality adhesion. Under the term surface tension σF we understand the
stress affecting on unit of length of marginal curve limiting the material. It is
identical with the work needed for creation of square unit of the new surface and it
can be expressed by following formula

dA = σ F ⋅ l ⋅ dl , (7.3)

whereby l⋅ dl is areal element, occurred at surface enlargement.

During tension test of the solid we have to subtract elastic deformation energy Ap
affecting square unit (it will demonstrate in the form of sound and kinetic energy of
fragments) that we will express by formula

AP =
(∆ l) E
2
, (7.4)
2l

where ∆ l is deformation to athermal fracture, E is Young's modulus and l is the

length of tested solid.

©Alexander Dubcek University of Trencin 2007 116

Then, for surface tension of perfectly fragile solid we get formula

A - AP
σF = . (7.5)
2

Factor two in denominator represents the fact that two fracture areas are occurring.

Only secondary forces, causing attraction for approaching solid are decisive for
surface tension. Therefore this inequality holds for high-molecular materials

A - AP
σF . (7.6)
2

It is necessary to remark that we cannot conclude value of surface tension out of

adhesion alone, because during gluing diffusion of glue macromolecules into glue
surface may occur, what helps to firmness of the joint greatly.

Dipole moments play important role during adhesion, too. De Bruyne [6] supposed
that dipole moment plays important role during adhesion, whereby molecules with
their ends electrically charged are the ones mostly applied. Molecules are
connecting with the base by dipole moments. Especially creation of hydrogen bonds
leads to a very strong couplings that can be encountered mostly at epoxy resins and
polyurethane. Strong couplings are emerging at metal gluing by interaction of
negative charges of free electrons in metal and positive charges. Plastic materials
(such as polyethylene) without dipoles are only very difficult to be glued. Variants
of two joining surfaces interaction are shown on the Figure 7.3.

Figure 7.3. To the theory of gluing. Athermal fracture (a) and distribution of the
molecules on the surface (b). Limiting areas without mutual interconnection (c),
Limiting areas interconnected with diffusion (d) [27]

7.2.1. Adhesion as a surface problem

Real surfaces have macro as well as micro asperities and these are directly
influencing real surface of contact, where elastomer is affected by force. For the
specific affecting force we can infer formula for real affecting pressure p on rough
surface [7]:

©Alexander Dubcek University of Trencin 2007 117

p = (Aapp/Aact) * papp, (7.7)

where we can imagine apparent area Aapp as contact area of completely smooth
surface and Aact is actual area, that is summation of particular areas of asperity
apexes for the real surface.

Almost every theory describing adhesion claims, that fA is inversely proportional to

the affecting pressure, and thereby the existence of macro irregularities reduces
coefficient of adhesion. Generally, contact surface of tyres is needed to have certain
pattern in the place of contact with the road to get slip-resistant rate and that is
despite illogicality that contact surface will reduce. Solution of this evidently
contradictory fact resides, of course, in distribution of adhesion on the wet as well
as on the dry road. On the dry road we will get maximal adhesion on the smooth
surface of tread without design. But when the road is wet, there is continual film of
liquid on the contact surface suppressing adequate adhesion and it is declining to a
very low level. That is why common tyres must have tread design.

It is very complicated to define exact contact surface between elastomer and

accidentally coarsened surface by experimentation. In literature we can find a
summary of distribution dependencies of rubber elements on hard and smooth
surface. It is documented by data in Table 7.2. In practice, rubber is elastic and
generally contact surface A appears to be directly proportional to the expression
(p/E´)n.

Kragelskij [7] defined:

n = (2s/2s+1), (7.8)

where s involves characteristic values of rubber elements’ particular shapes and n

increases in dependence on distribution complexity of rubber elements. It was
ascertained that n is nearing to one for the model with the most complex
distribution, but also in that case is hard surface smooth what is far away from real
surfaces.

©Alexander Dubcek University of Trencin 2007 118

Table 7.2. Determination of contact surface for various distributions of rubber
elements on the smooth surface [7]

Following above mentioned considerations we can F = A ⋅ s, where A = (p / E´)n

2s
and n = .
2s + 1

Result of high friction forces is wear or abrasion. In the case of elastomers we can
consider three mechanisms of wear depending upon the texture of surface.

For very sharp surfaces abrasive wear causes considerable degradation of elastomer
surface, corresponding with high coefficient of friction.

If surface asperity is more rounded than sharp, wear by fatigue is dominating and
relatively slight form of surface degradation occurs.

Wear by rolling process is on fine surfaces characteristically accompanied with high

coefficient of friction.

In real application, (e.g. sliding of tyre upon the concrete) all three forms of wear
can coexist.

©Alexander Dubcek University of Trencin 2007 119

Perhaps most widely used index of wear is abrasiveness γ that is defined as follows:

γ=A/f, (7.9)

where A is abrasive factor and f is coefficient of friction at sliding.

7.2.2. The Role of Adhesion at Dynamic Contact of Two Materials –

Macroscopic and Molecular Understanding

Friction force generated between moving solids consists of adhesion and

deformation part. Adhesion part manifests itself on the surface in the areas that are
going deeper not more than molecule dimensions, whereas deformation is
dimensional phenomenon. Exact nature of adhesion is not clear although it is
generally supposed that it consists of creation and interruption of joints on the
molecular level.

In the case of elastomers we can suppose occurrence of adhesion by two

mechanisms, namely macroscopic or molecular mechanism.

From macroscopic point of view, for overall adhesive force FADH holds:

FADH = ∑ Fi = ∑ Ai . Si, (7.10)

where Ai is local macroscopic contact surface between elastomer and hard rough
base, Si is local shearing stress in mesophase.

On the molecular level, following the theory of Kummer [9], we can write for any
place of the joint

Fi = ni . ji, (7.11)

where ni is number of molecular joints and couplings between elastomer and base
in the place i, ji is effective joint force. Equations 7.10 and 7.11 result in formula

Si = (ni / Ai) . ji (7.12)

and thereby local shear stress of macroscopic model is function of the molecular
joint force ji.

In the course of last years several theories of elastomer adhesion on smooth and
rough surfaces were presented. Some of them were based on molecular-kinetic
behaviour of individual molecules. Other theories followed phenomenological
theory and are based on mechanical model of spring and absorber.

©Alexander Dubcek University of Trencin 2007 120

7.2.3. Ratio Theory

This theory is based on principle of various contact times of elastomer with surface.
It defines so-called ”variable time τ“; it is time from interruption of the joint and
elastomer to recreation of a new one [10]:

⎡ Ea ⎤
⎢ k ⋅T ⎥
τ =τ 0 ⋅ e ⎣ ⎦
, (7.13)

where Ea is activation energy, T is absolute temperature, k is Boltzmann’s constant,

τ0 for elastomers has value 10-12 s.

Speed of elastomer slide is defined by formula

⎡ Ea ⎤ ⎡ Ea ⎤
⎢ k ⋅T ⎥ ⎡ λ ⋅ γ ⋅ F ⎤ ⎢⎣ k ⋅T ⎥⎦
V = V1 ⋅ e ⎣ ⎦
=⎢ ⎥⋅e (7.14)
⎣ 2 ⋅τ 0 ⋅ k ⋅ T ⎦

V1 depends on the friction force F and absolute temperature T, λ is distance

between neighbouring elastomer couplings on surface of the base (approximately
10-7cm). F is friction force (the abrasiveness) and γ is then

γ = λ / 2 * Nk, (7.15)

Nk is number of elastomer elements in contact with the base, Ea is activation energy

within the range of values 42 – 84 kJ.mol-1.

7.2.4. Mixed Theory

This theory [11] describes viscoelastic character of elastomer behaviour during

friction. For description of behaviour we use simplified phenomenon of sliding and
adhesion on the molecular level, as well as information from the theory of
mechanical adhesion model.

Model defines friction force by the formula

F = K . σ0 . (W / H) . tan δ, (7.16)

where K is constant of proportionality, W is weight of surface, H is surface

hardness, σ0 is stress and tan δ is loss factor.

Hysteretic friction coefficient is determined by formula

fA = K* ⋅ (σ0 / H ) * tan δ. (7.17)

Equations 7.16 and 7.17 describe all the monitored facts during adhesive friction
measuring as follows:

©Alexander Dubcek University of Trencin 2007 121

Friction force is proportional to gravity
frictional coefficient fA decreases along with hardness increase
viscoelastic characteristics are presented in formula tan δ.

7.3. Friction

Leonardo da Vinci (1452-1519) is said to be the first one who developed basic
friction conceptions. French scientist Amontons was inspired by Leonardo’s
sketches, that led him to realization of experiments and consequential formulation
of well-known Principles of Friction.

Coulomb also did thorough research to set measure of friction during sliding
motion, he extended Amontons' Principle an expressed it with following algebraic
equation:

F = f ⋅ W, (7.18)

whereby F is friction force, f is friction coefficient and W is normal element of

weight force.

Coulomb’s contribution to friction is substantial. His research involved phenomena

of sliding speed, material characteristics, contact area characteristics and surface
treatment and he was also dealing with influence of lubrication on overall friction.
He also defined static and kinetic friction.

One of the first scientists who in eighteenth century claimed that surface of solid is
uneven and contains great number of hemispheric protuberances and holes was
Belidor. He inferred basic formulas for the force needed to overcome these
obstacles that enables mutual movement between surfaces. Belidor also examined
friction of tackle-fall and friction in bearing.

Swiss mathematician Euler developed exact analysis explaining Amontons’,

Parewnt’s and Belidor’s results and he inferred mathematical formula for friction
between pulley and rope. No substantial results have appeared after these great
discoveries of the old school.

Guembel in 1925 clarified that friction consists of two intersecting parts, namely
friction on smooth surface and friction on scratched surface. Old conception of
sliding surface as hills and holes led him to presumption that there are only three
contact points at pressure absence. Number of contact points is increasing along
with pressure until all protuberances are in real contact and elastic deformation
occurs (solid friction). Deformation at greater weight is partially elastic and
partially plastic, friction on scratched surface occurs due to plastic contact. This
work pointed out the role of contact welding during metal friction and also
importance of adsorption on surface layers.

Ernst and Merchant introduced following expression with two variables for metal
friction coefficient:

©Alexander Dubcek University of Trencin 2007 122

F=S/H+tanθ, (7.19)

where S is shear resistance of actual contact surface, H is hardness of the contact

surface according to Brinell and θ is average angle between actual contact surface
and direction of friction force. Whereas we can consider first expression in the
equation as expression of adhesion, second one represents brake effectiveness due
to the surface asperity.

Thermal effects at dry motion were analyzed in 1937 by Blok. He came to the
conclusion, that maximal temperature increase of two contact surfaces depends on
motion speed, intensity of thermal flow, thermal conductivity and intrinsic
temperature of surfaces. According to this theory, maximal achieved temperatures
are lower than melting temperature except the temperature at high pressure. Blok’s
theory substantially contributed to the understanding of metal friction mechanisms.
Russian scientist Kragelskij [12] after Second World War published detailed
description of Soviet research. Although in principle he agreed with results of
Western scientists, it is obvious that some divergences of opinion occurred,
primarily regarding generality of above mentioned Bowden’s friction conception
and thickness of adsorbed layers of oil.

Greenwood [13] proved that if area of real contacts between surfaces depends on
ideal plastic flow of micro-contacts, so then contact area is proportional to the
weight.

Situation is more complicated during viscoelastic, elastoplastic or plastic

deformation with strain hardening, but Greenwood considered statistical
distribution of uneven surface protuberances and proved that average size of
microcontacts is almost independent on the weight.

Greenwood clearly distinguished two different working methods at surface

topography and he used following theories:
• one is based on ideal group of equal irregularities,
• second is based directly on measured surface profiles.

Williamson and Hunt [14] analyzed profilometric data of solid surfaces with the
help of digital computer and found out that density of irregularities, protuberances
distribution and average radius of irregularities curvature are important parameters
determining type of surface contact. Their studies also pointed out that most
surfaces have Gauss' protuberances distribution.

Greenwood and Williamson [15] defined plasticity number, which can be

considered as general parameter of surface structure combining material and
geometric surface characteristics as the criterion for contact selection: elastic or
plastic.

Schallamach [20] studied relation between speed and temperature of rubber solids
and he came to conclusion that adhesion mechanism must be kinetic process. Roth
with colleagues [11] examined static coefficient at room temperature, but Bartenev
[26] in 1955 pronounced theory that rubber should not have static coefficient. This
opinion was questioned by Ratner [32], who stand up for the conception of static

©Alexander Dubcek University of Trencin 2007 123

coefficient pursuant to experimental data. Dispute however goes on up to the
present day.

7.3.1. Friction as Dynamic Problem of Two Surfaces Contact

As we have already seen in previous introductory part, friction force depends on

quite a number of parameters characterizing surface condition (asperity, hardness),
material characteristics of rubbed solids (moduli of elasticity, viscous and elastic
parameters), external conditions (temperature, normal compressive force, motion
speed etc.) and environment conditions between contact surfaces (dry friction, wet
friction, presence of pollutants in the water and so on).

Friction character and its patterns depend considerably on material type. Different
friction mechanisms are between metals, between polymers in crystalline or glassy
state and totally different are between polymers in highly elastic state and hard base.
Theory of polymers friction emerges from theory of metals friction. As it is known,
following basic mechanisms of friction force origin occur at two metal surfaces
friction (Figure 7.4) I - planing, II - plastic embossing, III - elastic embossing, IV –
tearing off the metal surface layer, V – tearing off the surface associated with deep
pulling out [7].

I II III IV V

Figure 7.4. Mechanisms of friction force origin at metals friction.

I - planing, II – plastic embossing, III - elastic embossing, IV - tearing off the metal
surface layer, V - tearing off the surface associated with deep pulling out [7]

Friction force FT, that is basic characteristics of the process at two surfaces friction,
is determined by their interaction on area S of real or factual contact. In general,
friction force depends on normal pressure W, motion speed v, temperature T, time
of contact τ and other parameters.

Practically, specific nominal friction force FN is often used, and it is determined as

quotient of friction force and area of nominal surface FN = F S N . Another
frequently used parameter is friction coefficient f, defined as quotient of friction
force and normal compressive force f = F N . Besides specific nominal friction
force we use the term pressure, therefore specific nominal pressure p = N SN . Area
of factual contact S is always considerably smaller than area of nominal surface SN.
Their ratio depends on treatment level of friction areas, hardness of materials and it
also essentially depends on amount of pressure p. Friction force is directly
proportional to the area of factual contact F = cS . Constant c, ergo specific factual
friction force, may depend on other friction parameters (e.g. T, p, and v).

At solid friction we often encounter motion in very small discontinuities. Nature of

this phenomenon (stick-slip) is analyzed in works [8, 9]. The process is

©Alexander Dubcek University of Trencin 2007 124

characterized by maximal FTmax, minimal FTmin and average FTs friction force and
also by average frequency of self-oscillations. Mechanisms of stick-slip origin are
different in the case of two metals friction than in the case of elastomer friction on
the hard base.

Friction force can be expressed in the form of adhesive and deformational

(hysteretic) friction force summation [2, 10].

Fdeformational = Fhysteretic (7.20)

F = Fadh + Fhyst (7.21)

Adhesive is caused by molecular-kinetic thermally activated stick-slip mechanism

and operates mostly on sliding surface. This mechanism is attributed to molecular
couplings between surface atoms of both solids, thereby stress increases in the point
of contact, after exceeding limiting value it breaks away and relaxation follows.
This cycle is repeated regularly. In contrast to hard materials, elastomer structure is
formed of elastic strings that are in constant thermal motion. Particular strings in
surface layer at relative motion between hard surface and elastomer are trying to
bind to molecules of hard base and thus create local couplings. Sliding causes
tightening of these couplings, their tearing and finally relaxation, until new
couplings are created with equilibrium position moved by molecular distance. So
adhesion on the molecular level is created by stick-slip process.

Hysteretic component of friction force relates to deformation of softer solid at

mutual contact. Surface of hard solid is even after treatment still imperfectly smooth
– there are irregularities on it (Figure 7.5a).

There comes to small deformation of both surfaces in the place of contact at two
hard solids contact (Figure 7.5b), whilst surface is practically not deforming at
contact of metal and soft solid, e.g. elastomer (Figure 7.5c), but deformation of
polymer in the points of contact is very strong.

During relative peace rest of both solids, distribution of pressure forces in softer one
is symmetrical – resultant of pressure forces is perpendicular to the direction of
motion, resultant component in direction of motion is zero − friction force does not
exist (Figure 7.6). If solids are in relative motion, distribution of pressure forces is
becoming asymmetrical in softer solid and resultant of pressure forces has nonzero
component oriented opposite to the direction of motion – that is hysteretic part of
friction force (Figure 7.6b). In general, approximately one third of total friction
force is caused by hysteretic deformational mechanism.

©Alexander Dubcek University of Trencin 2007 125

 w

Figure 7.5. Microscopic description of surface change of two solids at their mutual
contact a) surface of hard solid (metal); b) contact of two metals; c) contact of
metal and elastomer [2]

P P

Figure 7.6. Physical interpretation of friction force hysteretic component a)

distribution of pressure forces at mutual peace of solids b) at relative motion of the
solids [2]

Adhesion component of friction coefficient is for specific combination of elastomer

and hard surface determined by following formula

fA= K2 * (E´/pr) * tan δ, (7.22)

where p is nominal pressure, r is exponent with value near 0.2 and K2 is constant
depending on sliding ratio. This equation shows that even though adhesion is
molecular-kinetic process of sliding-adhesion alternation, it also mirrors
viscoelastic character of elastomer.

Hysteretic component of friction force is illustrated on Figure 7.7. It shows pressure

distribution on particular surface irregularities during elastomer sliding and also
without any motion.

©Alexander Dubcek University of Trencin 2007 126

 (a) Wi
P
pHOR. 0
NO RELATIVE pvert. Ai NO NET
MOTION FRICTION
SYMMETRICAL + FORCE
PRESSURE
Wi pvert. Ai
DISTRIBUTION

(b) Wi
P
pHOR. 0
RELATIVE V pvert. Ai NET
MOTION FRICTION
UNSYMMETRICAL
+ FORCE
PRESSURE Wi pvert. Ai FHYST.
DISTRIBUTION

Figure 7.7. Physical interpretation of hysteretic friction component [7]

In the case when elastomer is not in motion, hysteresis is manifested symmetrically

in the volume of elastomer. Pressure, occurring during following of irregularities’
shape, is divided into horizontal and vertical component. We can see on Figure 7.7,
that summation of vertical component of pressure occurring in elastomer must be
equiponderant to weight and horizontal component of pressure is diminishing.

When elastomer is moving at particular speed with regard to surface, distribution of

pressure occurring in elastomer is not regular and it moves opposite to the direction
of elastomer motion (unsymmetrical pressure distribution).

Then, we can write following equation for hysteretic component of elastomer

friction coefficient

fH= K3 ⋅ (p/E´)n ⋅ tan δ [n ≥ 1] , (7.23)

where p is nominal pressure, K3 is constant, n is index depending on shape of

irregularities.

It is obvious that both components of friction - coefficient fA and fH - directly

depend on modulus tangent, so both friction components can contribute to the same
viscoelastic mechanism. For total friction coefficient then holds

f = K2 ⋅ [(E´/pr) + K4 ⋅ ( p/E´)n] ⋅ tan δ, (7.24)

where K3 = K2 ⋅ K4.

Equation 7.24 shows, that pressure adjustment in real conditions is compromise

between contribution of adhesion and hysteretic component, and optimal pressure
setting partially depends on surface asperity and existence/absence of effective
interfacial lubricants.

Friction of polymers in highly elastic state depends on motion speed. Roth with his
colleagues [11] examined friction dependence of technical tread rubber on glass and
steel. Experimental values obtained in their work are on Figure 7.8.

©Alexander Dubcek University of Trencin 2007 127

 µ
5

1,0

3 0,01

2
0,001

1 0,0001 cm/sec

10 20 30 40
L (cm)

Figure 7.8. Dependence of friction coefficient on motion trajectory at various

motion speeds [11]

As we can see on Figure 7.8, friction coefficient is increasing greatly with higher
motion speed. Transition from static friction to dynamic is interesting as well
(friction coefficient is smaller at the motion beginning). This phenomenon is most
obvious at high speed, at speed 0.001 cm/s almost fades away and at even lower
speed (0.0001 cm/s) it has opposite character.

Schallamach [12] in his work examined dependence of friction force of rubber on

glass on temperature in areas of low pressure (about 1 kg/cm2, i.e. approximately
100 kPa). Motion speed v was measured at constant tractive force F in tangential
direction – (Figure 7.9a). Yet, size of tangential force is equal to the size of friction
force (it is its reaction).

With increasing temperature T, motion speed is increasing, and that indicates of

friction force decrease along with growing temperature. There is dependence of
motion speed logarithm on reciprocal of absolute temperature on Figure 7.9b.

Experimental values for particular tangential force sizes are lying almost in line and
that testifies exponential dependence of v on 1/T. By the analysis of experimental
results, Schallamach found empirical relation for motion speed

©Alexander Dubcek University of Trencin 2007 128

 ⎛ γF - U ⎞
v = v 0 exp⎜ ⎟, (7.25)
⎝ kT ⎠

where v0 is constant almost equal to the speed of sound in rubber below vitrification
point (1000 m/s), U is activation energy of friction process, k – Boltzmann’s
constant, γF is work of friction forces and F is tangential force, whose reaction is
friction force. Activation energy for rubber, U = 83 kJ/mol, was set with the help of
experimental values.

0 0 8,8
F = 4,8 kg
40 C
O
7,8

6,8
-1 -1
5,8
logυ (cm/sec)

logυ (cm/sec)
-2 20OC -2

-3 -3
F = 4,94 kg

-4 -4
4 5 8 10 3,4 3,2 3,0
F (kg) 3
10 /T ( OK)

Figure 7.9. Dependence of friction force on motion speed of rubber on glass

a) Dependence of motion speed on tangential force at T = 20oC and T = 40oC
b) Dependence of motion speed on reciprocal of temperature for various values of
tangential force [12]

Works of Schallamach [12, 13] became starting point for molecular-kinetic theory
of highly elastic polymer friction that was developed by Bartenev [1, 3]. Basic
presumptions of this theory come out of polymer model, consisting of linked strings
that are performing chaotic thermal motion (Figure 7.10).

Figure 7.10. Molecular contact model of polymer and solid body [3]

©Alexander Dubcek University of Trencin 2007 129

String ends are joined to particular points of base surface in contact with polymer,
whereby they perform accidental thermal transitions from one attached position to
other one. Average time between transitions is τ and we can express it as follows

⎛U ⎞
τ = τ 0 exp⎜ ⎟, (7.26)
⎝ kT ⎠

where τ0 ~ 10 -12 s is constant and U is energy barrier, determined by adhesion

forces between polymer and base, that must be smaller than cohesion force between
polymers. From these presumptions Bartenev inferred theoretical relation between
friction force and motion speed

λ kT ⎛ U − γF ⎞
v= exp⎜ − ⎟ (7.27)
τ0 2πγF ⎝ kT ⎠

identical with empirical formula of Schallamach. Relation 7.27 holds on the

assumption, that γF > kT , i.e. for big friction force or low temperature. Quantity λ
is middle distance between neighbouring equilibrium positions of string on hard
surface, γ = λ n where n is number of strings found in contact with unit surface area
of metal. By rearrangement of the last equation we will get formula for dependence
of friction force on temperature

U kT ⎛ v ⎞
F= + ln⎜⎜ ⎟⎟ . (7.28)
γ γ ⎝ v0 ⎠

Coefficient at temperature is negative, because v < v0, so F is decreasing linearly

with the growth of temperature. By means of extrapolation of this dependence on
temperature axis (F = 0 when T = T0) we will get activation energy
U = kT0 ln (v v 0 ) . By this practice was in work [14] found value U = 50.2 kJ/mol
for friction of 40 ShA Natural Rubber rubber SKN-40 on steel. Friction constant
(specific factual friction force, see page 15) is then expressed by formula

2U ⎛ T ⎞
c= ⎜⎜1 − ⎟⎟ , (7.29)
λS k ⎝ T0 ⎠

where Sk is area of polymer string elementary contact with hard surface.

There is elastomer friction examined in wide interval of temperature (from 20°C to

120°C) in work [15]. The authors examined friction of rubber SKN-40 on steel at
normal pressure of 30 kPa. Friction force was at 20 - 60 °C interval decreasing
linearly in accordance with theory, but with higher temperature it started to grow
anomalous with maximum at approximately 110°C (Figure 7a), what is probably
related to oxidation processes at rubber thermal degradation at such temperature. In
favour of this interpretation speaks the fact, that experiments in vacuum [16] shew
smaller diversion from linearity and it occurs only at considerably higher
temperature (Figure 7.11b).

©Alexander Dubcek University of Trencin 2007 130

 F (kg)

1
1,2 2

4
F (kg)
1,0 5
6 1,5

1
1,0

0,8
0,5 2

40 80 120 40 80 120 160

o
T( C) T(oC)

Figure 7.11. a) dependence of friction force of rubber SKN-40 on steel on

temperature. Motion speed 1 – 2.8; 2 – 1.0; 3 – 0.5; 4 – 0.1; 5 – 0.05; 6 – 0.01 (in
mm/min) [15] b) dependence of rubber friction force on temperature: 1 – in air; 2
– in vacuum [16]

Monitoring of friction force dependence on motion speed can be found in available

literature, but results are mostly referring to very small motion speed. It is caused by
the fact that it is difficult to keep constant temperature of contact areas at higher
speed, because noticeable rubber abrasion occurs there and it distorts results of
measuring, especially at higher normal weight.

It was experimentally proved in work [17], that diversion of linear dependence F

from ln (v/v0) occurs at higher motion speed and according to the authors’
presumption it is related to the mechanism of losses in rubber itself (Figure 7.12).
At high motion speed we have to take not only time τ1 (mean lifetime of string in
specific position) into consideration, but also time τ 2 , needed for transition (self
diffusion) of string from one contact area to another.

©Alexander Dubcek University of Trencin 2007 131

 F (kg)
60
2

1
40

20
-7 -5 -3 -1 1 3
lnυ (cm/min)
Figure 7.12. Dependence of friction force on motion speed logarithm: 1 - rubber
SKN-18, 2 - rubber SKN-40 [17]

Results of laboratory measurements show that curve of friction force dependence on

temperature and on motion speed have similar shape. From theoretical reasoning we
can infer, that principle of temperature-time superposition [18] applies to friction
force, what means that there is relation between temperature and speed (friction
force is maximal for speed vmax)

⎛v ⎞
ln ⎜⎜ max ⎟⎟ = α 1T , (7.30)
⎝ v2 ⎠

where v2 and α1 are constants. Out of that we can infer that to change speed vmax by
one, we have to change temperature by ∆T = 2.3 α1 as well. This value is 19 ÷ 26 K
for ordinary rubber.

Some theories were suggested to clarify experimental data. According to these data,
friction force has maximum at specific motion speeds [19] (see e.g. [3, 4]), but not
even one is completely explaining all process patterns. Figure 7.13 shows
comparison of experimental data acquired by Grosch [19] with theoretical curves
[1].

Molecular-mechanic friction theories were elaborated by Schallamach [20], Bulgin

[21] and Savkoor [22]. Bulgin’s theory worked on presumption, that polymer
surface consists of great number of microprotuberances in contact with hard
material. Every microprotuberance has dimensions of 1 nm order and creates
adhesive coupling with contact solid. This coupling can be broken by tangential
force fmax. We suppose that force Fmax does not depend on motion speed of polymer
on contact solid. These microprotuberances are being deformed during motion of
the polymer (Figure 7.14-2). If it comes to dislocation by such a length ℓ, that force
deforming element exceeds Fmax, then the coupling breaks and binding of element in
new position occurs (Figure 7.14-3). This cycle is repeated with period τ = l v .

©Alexander Dubcek University of Trencin 2007 132

 2

3,0
1

2,0

F.a
3
1,0

-6 -4 -2 0 2 4
logυ (cm/sec)

Figure 7.13. Comparison of experimental values of friction force dependence on

motion speed (curve 3) with theoretical values (curves 1, 2) [21]

From energetic reasoning we can infer, that friction forces work must be equal to
deformation energy of all polymer elements. Following these presumptions, Bulgin
inferred approximate formula for friction force

p E ′′
FT = const σ , (7.31)
H E′

where σ is maximum mechanical stress in deformed element, p is pressure (normal

weight), H is hardness, E" is imaginary and E' real component of complex stress
modulus at frequency τ -1. Dependence of friction force on temperature and motion
speed is determined by frequency and temperature dependence of loss factor E"/E'.
These results confirm also experimental data in above mentioned work (Figure
7.14). Even though the Bulgin’s theory is just approximate and neglects many
important factors, it provides illustrative image of friction forces mechanism in
polymers.

©Alexander Dubcek University of Trencin 2007 133

 U
K

U
K´ K

Figure 7.14. Frictional coupling disintegration [21]

F,kg

2,70

1,80

0,90

0
0,2 0,4 0,6 0,8 1,0
E´´/E´

Figure 7.15. Friction force as function of loss factor E"/E' according to [21]

More exact formula for friction force is inferred in monograph [23] that takes into
consideration adhesive fA and hysteretic fH component of friction force separately.
According to this theory, friction force depends not only on loss factor E"/E', but
also on normal pressure p and hardness in slide s of hard surface:

⎡ E′ ⎛ p ⎞ ⎤ E ′′
n

f = fA + fH = K 1 ⎢s r + K 2 ⎜ ⎟ ⎥ , (7.32)
⎢⎣ p ⎝ E ′ ⎠ ⎥⎦ E ′

where exponent r has approximate value of 0.2 and exponent n is more or equal to
one. Graphic representation of frequency dependence f, fA and fH is on Figure 7.16.
As we have already mentioned, the higher motion speed, the higher frequency, that
polymer is periodically deformed with. Because ω = (2π / l ) v , then

©Alexander Dubcek University of Trencin 2007 134

log ω = log v + const . So graphs on Figure 7.16 show dependence of friction force
on frequency as well as motion speed. Displacement of frequency is according
to equation of Williams-Landel-Ferry [18] equivalent to increase of polymer
temperature, and that is represented by the second curve, expressing friction force at
higher temperature.

ADHÉZNE
MAXIMUM
HYSTERÉZNE
MAXIMUM
fA, fH, or f (kg)

T+UT
T
T+UT
T
f
f
fA
fH

log v, ω

Figure 7.16. Theoretical dependence of adhesive, hysteretic and total friction force
on angular frequency logarithm or on motion speed logarithm at various
temperatures [23]

Another important quantity influencing frictional process is area of factual contact

S. In many works, there have been equivalent dependencies of surface factual
contact on contact time of specimen and base S(t), motion speed S(v), normal
pressure S(p) and temperature S(T) inferred with the help of phenomenological
theory and also molecular-kinetic theory.

According to Lavrentjev [24], S is directly proportional to the number of strings that

are in time t bound to the base surface. S(t) is asymptotically nearing the value S ∞
for t → ∞ at invariable value of pressure and temperature. Let value of S in time
t = 0 be S0, coupling breaking frequency of string and base ν 1 and frequency of
coupling creation ν 2 . Value ν 1 depends on tangential force and pressure, whilst ν 2
depends only on pressure. At adequate normal pressure ν 2 >> ν1 . Then holds

S (t ) = S ∞ [1 − exp(− ν 2 t )] + S 0 exp(− ν 2 t ) . (7.33)

Let us define relative area of contact by formula ϕ = S S ∞ . Its dependence on time

is expressed by formula

ln (1 − ϕ ) + ϕ = − k t , (7.34)

where k is constant proportional to tangential stress. Comparison of this theoretical

dependence with experimental values is on Figure 7.17a. Curves 1 to 4 express
theoretical values equivalent to various normal pressures at temperature of 21°C.
Experimental values correspond almost exactly with calculated ones.

©Alexander Dubcek University of Trencin 2007 135

 Area of factual contact dependence on motion speed is examined in work [14]. On
Figure 7-17b we can see experimental values for proportion of equilibrium value of
factual contact area Sv at speed v to its equilibrium value S0 in peace.

It is known that highly elastic polymers are elastically deformed in contact zone and
in consequence of that, area of factual contact is at first growing with increase of
normal pressure almost linearly. Binding of almost all free strings to the base occurs
at higher pressures, so that value of φ is asymptotically nearing one. This
dependence is theoretically as well as experimentally examined in work [25].
Dependence of φ on pressure is expressed by formula

⎛ βp ⎞
ϕ = 1 − (1 − ϕ 0 ) exp⎜ − ⎟, (7.35)
⎝ E ⎠

where E is stress modulus and β is constant, value of which was determined in

mentioned work. Experimental and theoretical values are shown on Figure 7.14.

-1 Sυ
ln ( 1-ϕ ) + ϕ

1 S0 1
-2
1,0
2
2
-3
4 3
1,5
0 5 10 -7 -5 -3 -1 1 3
t (min) logυ (cm/sec)

Figure 7.17. a) dependence of relative factual contact surface at friction of rubber

SKN-26 on time at temperature of 21°C and pressure in kPa: 1 – 70; 2 – 180; 3 –
320; 4 – 600 [24] b) Dependence of Sv /S0 ratio on motion speed logarithm at
pressure of 550 kPa and temperature 20°C for rubber: 1 – SKN – 18; 2 – SKN –
40. [14]
ϕ
1,0 0
1

2 -2
1 - ϕ0
1-ϕ

0,5 3
-4
ln

2
-6
1

0 40 80 20 40 60
p (kg/cm2) 2
p (kg/cm )
Figure 7.18 a) dependence of relative factual contact surface on pressure for
rubber: 1 – SKN-18; 2 – SKN-26; 3 – SKN-40. b) the same data in semilogarithmic
scale [25]

©Alexander Dubcek University of Trencin 2007 136

Dependence of friction force on normal weight N or on specific normal pressure p
is examined in works [26, 27]. The authors approximated this dependence by semi-
empirical formula

F = c1 S + c 2 ω N , (7.36)

where c1 and c2 are coefficients depending on motion speed, temperature and

molecular character of polymers and ω is angular frequency of polymer
deformation. The course of friction force in dependence on pressure is on Figure
7.19.

Within the pressure limits of 0.1 to 15÷20 MPa, constant c1 does not depend on
pressure and constant c 2 = 0 . That means that dependence of friction force on N in
this area is determined by dependence of S on N, and that is exponential. Thus we
can express Principle of Friction by formula

f = c[1 − (1 − ϕ 0 ) exp(− αp )]. (7.37)

Thirion [28] proposed empirical formula for mean values of pressures (scale from
0.5 to 5 MPa)

1
= A + Bp , (7.38)
µ

where A, B are constants. On next figure (Figure 7.18a) we can see experimentally
gained dependencies f on p for friction of unfilled rubber thin foil on steel,
published in work [29]. Figure 7.18b shows experimental values of friction of
rubber SKN-40 on steel in comparison with various theoretical prognoses.
F

0 PKP p

Figure 7.19. Dependence of friction force on normal pressure according to (13)

[26]

©Alexander Dubcek University of Trencin 2007 137

 75

60

f (kg/cm )
180

2
3
f (kg/cm )

45
2

140 1

100 30 2

60 15
20

400 600 1000 1400 400 800 1200

2 2
p (kg/cm ) p (kg/cm )

Figure 7.20. a) dependence of specific friction force on pressure for friction of

unfilled rubber thin foil on steel. According to [29]. b) dependence of specific
friction force on pressure for rubber SKN-40 on steel: 1 – experiment, 2 – low
pressure theory, 3 – high pressure theory [26]

Wet friction is complicated process, when there is thin film of liquid material –
water, oil and the like – of variable thickness between surface of elastomer and hard
solid (Figure 7.21). Providing that there is constant viscosity µ (for friction) η (for
viscosity) and density ρ of liquid film, we can write equation for its motion

∂ ⎡ 3 ∂ p⎤ ∂ ⎡ 3 ∂ p⎤ ∂h ∂h
⎢h ⎥+ ⎢h ⎥ = 6 µ (u1 − u 2 ) + 6 µ h (u1 + u 2 ) + 12 µ , (7.39)
∂x ⎣ ∂x⎦ ∂ y ⎣ ∂ y⎦ ∂x ∂t

where u1 and u2 are speeds of upper and lower surface of rubbed solids, p is
hydrodynamic pressure. In the case, that one of the solids is elastomer, situation
expressed on Figure 7.22 occurs. Elastomer is deformed on hard base protuberances
and liquid film is divided into specific parts where areas of increased and reduced
pressure alternate. Such a friction mechanism is analyzed in works [23, 30].
dh
V= u1
dt
z,h

h ρ, µ, p x
y
u2

Figure 7.21. Contact of two solids separated by a film of liquid. According to [2]

Resultant friction force greatly depends on surface relief (microscopic irregularities,

macroscopic irregularities), motion speed, temperature, pressure and liquid film
parameters. Theoretical description in full extent does not exist. There are only
some empirical formulas for specific cases and many experimental data.

©Alexander Dubcek University of Trencin 2007 138

 W

u ELASTOMER

(a)

THIN
LUBRICANT
ELASTO- FILM
HYDRO-
DYNAMIC
PRESSURE
p + + +
(b)
- -
Figure 7.22 Interaction of hard surface wetted by liquid film and elastomer)
elastomer sputtering on irregularities b) generated pressure alteration [2]

Influence of foreign substances in water (mainly its pH) on wet rubber friction is
examined in works [31-33]. The authors have tested several alkaline as well as acid
solutions for friction of rubber on rubber, glass, concrete and macadam. They stated
that friction coefficient is diving rapidly at pH 9. It is caused by dissolving of fatty
acids in rubber. On the other hand, acid medium (pH= 6) mildly increases friction
force.

References

[1] G.M. Bartenev, V.V. Lavrentjev: Trenie i iznos polimerov, Izd. Chimija,
Leningrad, 1972.
[2] D.F. Moore: The Friction of Pneumatic Tyres, Elsevier, Amsterdam, 1975.
[3] G.M. Bartenev, V.V. Lavrentjev: Friction and Wear of Polymers, Elsevier,
Amsterdam, 1981.
[4] I.V. Kragelsky, M.N. Dobychin, U.S. Kombalou: Friction and Wear
Calculation Methods, Pergamon Press, Oxford, 1982.
[5] H. Uetz, J. Wiedermayer: Tribologie der Polymere, Carl Hanser Verlag,
Munich, Vienna, 1985.
[6] N.A. De Bruine: Flight. 51, 1939.
[7] I.V. Krageľskij: Trenie i iznos, 2nd ed., Izd. Mašinostrojenie, Moskva, 1968.
[8] V.A. Kudinov: Issledovanie kolebanij metallorežuščich stankov pri trenii, Izd.
AN SSSR, Moskva 1958.
[9] Ju. I. Kosterin: Mechaničeskije avtokolebania pri suchom trenii, Izd. AN SSSR,
Moskva 1960.
[10] F.P. Bowden, D. Tabor: Friction and Lubrication of Solids, Clarendon Press,
Oxford, 1964.
[11] F. Roth, D. Driskoll, W. Holt: Rubb. Chem. & Technol. 16 (1943), 155.
[12] A. Schallamach: Proc. Phys. Soc., B66 (1953), 386.
[13] A. Schallamach: Wear, 1 (1958), 384.
[14] G.M. Bartenev, V.V. Lavrentjev, A.N. Konstantinova: Mechanika polimerov 7
(1967), 726.
[15] G.M. Bartenev, E. Styran: Vysokomol. sojedinenia, 1 (1959), 978.

©Alexander Dubcek University of Trencin 2007 139

[16] A.I. Jelkin, V.N. Nikolejev: DAN SSSR, 173 (1967), 1302.
[17] G.M. Bartenev, V.V. Lavrentjev, N.A. Konstantinova: Mechanika polimerov 3
(1967), 309.
[18] M.L. Williams, R.F.Landel, J.D.Ferry: J. Amer. Chem. Soc., 77 (1955), 3107.
[19] K. Grosch, Proc. Roy. Soc. 247A (1963), 21.
[20] A. Schallamach: Wear, 6 (1963).
[21] D. Bulgin: Rubb. Plast. Weekly, 143 (1962), 636.
[22] A. Savkoor: Wear, 8 (1965), 221.
[23] D.F. Moore: The Friction and Lubrication of Elastomers, Pergamon Press,
Oxford, 1972.
[24] V.V. Lavrentjev: DAN SSSR, 175, №1 (1967).
[25] N.A. Konstantinova: Dizertačná práca, MGPI, Moskva 1967.
[26] G.M. Bartenev: DAN SSSR, 103, (1955), 1017.
[27] J. Slabeycius: Internal Report for Matador, Puchov, 2000.
[28] P. Thirion: Rubb. Chem. and Technology, 21 (1948), 505.
[29] B.V. Derjagin, Yu.P. Toporov: DAN SSSR, 146, (1962), 1356.
[30] J. Hirama, H. Ishiwata: Paper No.15, Proc. Inst. Mech. Eng., 180 (1965/66),
187.
[31] A.D. Roberts: Rubb. Develop. 49 (1996), 18.
[32] A.D. Roberts, S.C. Richards: Gummi Fasern Kunststoffe 50 (1997), 978.
[33] A.D. Roberts, S.C. Richards: Rubb. Develop. 49 (1996), 54.

Fundamental questions from presented part

1. What is the origin of polymers hysteresis?
2. What factors determine the firmness of glued joints?
3. Define the surface tension.
4. What type of tread surface yields the maximal adhesion on the wet and on
the dry road?
5. Describe the stick-slip process in the case of elastomer friction on the hard
base.

©Alexander Dubcek University of Trencin 2007 140