Physical properties of Fluon

unfilled and filled PTFE

Technical Service Note F12/13

2
Contents

Page Page
Summary 6 Refractive index 28
Melt viscosity 28
Introduction 6 High energy radiation 29
Chemical resistance 29
SECTION 1. THE STRUCTURE OF PTFE 7 Velocity of sound 29
Molecular conformation and crystal structure 7
Melting behaviour 7 SECTION 8. GENERAL - FILLED PTFE 30
Crystallinity and texture 8 Function of filler 30
Polymer particle structure 9 Choice of filler 30
Void formation in mouldings 10 Choice of Fluon grade 31
The Fluon range 31
SECTION 2. EFFECTS OF PROCESSING ON 11
PROPERTIES SECTION 9. MECHANICAL PROPERTIES 32
Molecular weight 11 - FILLED PTFE
Crystallinity 11 Compressive deformation 32
Orientation 12 Impact strength 37
Voids 12
SECTION 10. ELECTRICAL PROPERTIES 38
SECTION 3. MECHANICAL PROPERTIES 13 - FILLED PTFE
- UNFILLED PTFE Insulating materials 38
Compressive stress-strain relationships 13 Conducting materials 38
Tensile properties 18
Impact behaviour 18 SECTION 11. THERMAL PROPERTIES 40
- FILLED PTFE
SECTION 4. ELECTRICAL PROPERTIES 21 Thermal expansion 40
- UNFILLED PTFE Thermal conductivity 40
Permittivity and dielectric loss 21
Experimental techniques 21 SECTION 12. SURFACE PROPERTIES 41
D.C. conduction behaviour 21 - FILLED PTFE
High voltage uses of PTFE 24 Friction 41
Wear 43
SECTION 5. THERMAL PROPERTIES 25
- UNFILLED PTFE SECTION 13. OTHER PHYSICO-CHEMICAL 45
Melting point 25 PROPERTIES - FILLED PTFE
Thermal expansion 25 Permeability 45
Thermal conductivity 25 Chemical resistance 45
Specific heat and heat of fusion 26
Thermal stability 26 SECTION 14. TYPICAL PROPERTIES OF 46
UNFILLED AND FILLED MOULDED PTFE
SECTION 6. SURFACE PROPERTIES 27
- UNFILLED PTFE SECTION 15. SPECIFICATIONS RELATING 47
Coefficient of friction 27 TO PTFE
Angle of contact 27
SECTION 16. HANDLING PRECAUTIONS 48
SECTION 7. OTHER PHYSICO-CHEMICAL 28
PROPERTIES - UNFILLED PTFE SECTION 17. FURTHER INFORMATION 49
Permeability 28
Infra-red transmission 28 SECTION 18. REFERENCES 50

3
4
List of figures

Page Page
1. Molecular conformation of PTFE 7 18. Loss angle versus temperature for 22
2. The decay of birefringence in the melt 8 sintered granular PTFE (60 % crystallinity)
3. Particles of dispersion polymer 9 using evaporated gold electrodes
4. Aggregates of particles of dispersion 9 19. Loss angle versus frequency for 22
polymer PTFE at room temperature
5. Longitudinal section through 10 20. Plot of log10 apparent volume 23
unsintered extrudate resistivity versus log10 time for
6. Longitudinal section through 10 sintered granular PTFE
sintered extrudate 21. Effect of different electrode materials 23
7. Internal and external texture of a 10 and preconditioning for sintered
granular polymer granular PTFE
8. Variation of relative density 12 22. Variation of specific volume 26
with crystallinity with temperature
9. Basic creep testing equipment 13 23. Infra-red transmission spectrum 29
10. Compressive creep 14 (specimen thickness 0.05 mm (0.002 inch))
11. Isochronous stress-strain relationship 14 24. Creep in compression at 20C (68F), 33
in compression Fluon G163 and Fluon FC100-25 1000
12. Isochronous stress-strain curves in 15 25. Creep in compression at 20C (68F), 34
compression, 25C (77F), Fluon G163 at various stress levels,
13. Creep in compression, at 25C (77F), 16 Fluon FC100-25 1000
Fluon G163 26. Isometric stress-time curves in 35
14. Isometric stress-strain curves 17 compression at 20C (68F), and various
in compression, at 25C (77F) and strain levels: Fluon FC100-25 1000
various strain levels, Fluon G163 27. Creep in compression at 20C (68F): 36
15. Recovery from creep in compression 1 various Fluon grades
at 25C (77F), Fluon G163 28. Recovery from creep in compression: 36
16. Effect of temperature upon tensile 19 Fluon FC100-25 1000 loaded
stress-strain curves for PTFE at various stress levels
17. Effect of temperature upon tensile 20 29. Loss tangent versus temperature 38
yield stress of PTFE at 1 kHz: glass-filled PTFE

5
 Summary Introduction

This Note describes the physical properties of PTFE. Only rarely can the physical properties of PTFE
It should be used in conjunction with other Asahi Glass (polytetrafluoroethylene) be expressed in simple terms.
Fluoropolymers Technical Service Notes. Because of its high melt viscosity PTFE is not processed
by those techniques associated with most thermoplastic
polymers. The techniques used with PTFE consist
essentially of a cold shaping operation followed by
sintering, during which the polymer particles fuse and
coalesce; finally the fused polymer is cooled. One result
of the fabrication methods used with PTFE is that, in
many respects, the quality of the fabricated polymer is
unusually dependent on the skill with which the
fabrication is carried out. For this reason any table
purporting to express the physical properties of PTFE as
a list of figures - including such tables which appear in
this Note - must be regarded as a very simple summary
of a very complex subject.

Three general surveys of PTFE are:

Fluorine-containing polymers, Part 2:

Polytetrafluoroethylene, C A Sperati and H W
Starkweather, Fortschr. Jochpolym.-Forsch., 2, 1961,
465-495

Polytetrafluoroethylene, S Sherratt, Kirk-Othmer,

Encyclopaedia of Chemical Technology, 9, 2nd edition,
Interscience Publishers Inc., New York, 1966, 805-831.

Fluoroplastics, Vol 1: Non-Melt Processible

Fluoroplastics, S. Ebnesajjad, Published by Plastics
Design library, 2000. ISBN1-884207-84-7

6
Section 1. The structure of PTFE

MOLECULAR CONFORMATION AND CRYSTAL

STRUCTURE

PTFE is a linear chain polymer of great molecular length.

The linearity is indicated by an analysis of the infra-red
spectrum and by the fact that the powder as produced in
the polymerisation reaction is very highly crystalline,
with crystalline weight fractions of 0.90 to 0.95 being
indicated by density, infra-red and X-ray diffraction
measurements. Energy considerations also suggest that
branching by chain transfer is unlikely (ref. 1).

The crystal structure and chain conformation have been

discussed by Bunn and Howells (ref. 2) and later by
others (refs. 3, 4, 5). The crystalline melting point of
sintered PTFE is about 327C (620F) and of unsintered
material 332-346C (630-655F) but there are two
reversible first order transitions at lower
temperatures,19C and 30C (66F and 86F) (ref. 6),
which together involve a 1% change in density (ref. 7).
Three crystalline phases are observed at atmospheric
pressure: phase I (< 19C; 66F), phase ll (19-30C; 66-
86F) and phase lll ( >30C; 86F).

Below the 19C (66F) transition (phase l), the chain

repeat distance is 16.8 and the CF2 groups are equally
spaced along the chain which is twisted to form a helix
on which successive carbon atoms lie, thirteen carbon
atoms being involved in a twist of 180, (Figure 1).
Between 19 and 30C (66 and 86F) (phase ll) the repeat
distance increases to 19.5 (refs. 3,8) corresponding to a Figure 1. Molecular conformation of PTFE
twist of 15 carbon atoms in 180. Above 30C (86F),
(phase lll), further disorder sets in and although the
molecular conformation prevailing at lower
temperatures is maintained, the chains are displaced or MELTING BEHAVIOUR
rotated along their long axes by variable amounts which
increase as the temperature is raised further. The reason PTFE has a crystalline melting point at atmospheric
for the helical structure is the necessity to accommodate pressure of about 332-346C (630-655F) for unsintered
the large fluorine atoms (van der Waals radius 1.35 ). and about 327C (620F) for sintered material. The heat of
The rotation at each chain bond, with the slight opening melting has been estimated by Lupton (ref. 18) as about
up of the bond angles to 116, relieves the overcrowding 685 cal/mole of CF2 groups and leads to an estimate of
and permits the shortest F-F distance to be 2.7 (ref. 9). the entropy of melting of 1.14 cal/degK mole. This value
is low by comparison with polyethylene (2.29 cal/degK
Further studies by various authors (refs.10,11,12,13,14) mole) and arises from the fact that the stiff long
have examined the effect of pressure on the room molecules retain a good deal of side-by-side order in the
temperature transitions and the melting point. A study of melt. Indeed, if the depolarisation of light is used to
pressures above-atmospheric revealed a 2% increase detect the course of melting, a finite amount of
in density below 19C (66F). This fourth crystalline depolarisation is observed to remain above the melting
phase has been labelled phase lll by Weir (ref. 10). A point (ref. 19) (Figure 2). The decay of this residual
triple point exists at about 70C (158F) and 4.5 kilobars. depolarisation is sensitive to molecular weight, time and
The heats of transition were also determined by Yasuda temperature. Bunn (ref. 20) has ascribed the high melting
and Araki (ref. 11); dilatometric and calorimetric studies point to the rigidity of the fluorocarbon chain, small
have in addition been reported by other workers (refs. rotations being hindered by the slightly overcrowded
6,15, 16,17). fluorine atoms.

7
Figure 2. The decay of birefringence in the melt

CRYSTALLINITY AND TEXTURE fused and rapidly quenched samples. The initial high
crystallinity and melting point can never be completely
The crystallinity and texture of polytetrafluoroethylene recovered after fusion because of the complete
have received a good deal of study. Estimations of the reorganisation of the molecular arrangement upon
degree of crystallinity have been made by X-ray (ref. 21), sintering.
infra-red (refs. 22, 23) and density methods. For 100%
crystalline PTFE relative densities of 2.347 at 0C (32F) Crystallisation from the melt in bulk produces long bands
and 2.302 at 25C (77F) can be calculated from the X-ray which can be seen both in fracture surfaces produced by
crystallographic data (ref. 5). The density decrease of breakage at liquid nitrogen temperatures, and in sections
about 2% between these temperatures includes the of the material when viewed microscopically in polarised
decrease of about 1% arising from the transition at 19C light (ref. 26). Their dimensions vary from 100m x 1m
(66F). The relative density of amorphous PTFE is not to 10m x 0.2m depending on the crystallisation
affected by this transition and values around 2.00 are conditions and/or molecular weight, the band width
obtained by measuring the density at room temperature being inversely related to molecular weight (ref. 27).
(preferably 23C; 73F) for a number of samples of
varying crystallinity, and then extrapolating to zero The bands show fine parallel striations perpendicular to
crystallinity (refs. 8, 24, 25). the band length and it is thought that the bands are the
broken edges of thick lamellar crystals. The
If the degree of crystallinity is estimated as a result of X- interpretation of the striations is still uncertain and
ray or infra-red methods it is then possible to calculate a remains a subject for further investigation (ref.19).
theoretical density. Differences between this theoretical Birefringence measurements show the molecules to lie
density and the experimentally measured value can be perpendicular to the bands and since the average length
used to assess the void content of a sample. of the molecules is greater than the band thickness chain
folding is implied. As the fusion temperature is raised or
Neither the X-ray nor the infra-red method measures an a lower molecular weight polymer is used (both of which
absolute crystallinity. Each measures disorder by a tend to reduce the persistence of order in the melt) the
different criterion and, in addition to this, each is subject band growths become irregular and tend to agglomerate
to purely instrumental limitations, so that even for one of into spherulitic structures. The use of very high fusion
the methods the same value would not necessarily be temperatures of around 450C (842F) results in well
determined on two different instruments. Tests show the formed spherulites, presumably due to molecular weight
degree of crystallinity measured by the infra-red method reduction as a consequence of thermal degradation.
is systematically higher than the X-ray crystallinity, the
discrepancy being of the order of 5% at 90% increasing to Speerschneider and Li have investigated the role of the
10% at 50%. large bands in uniaxial deformation and have shown that
distortion of the bands occurs at low temperatures
For instance, the crystallinities estimated by the X-ray (ref. 28).
method range from about 90% for unsintered material,
75% for fused and slowly cooled material, to 50% for

8
POLYMER PARTICLE STRUCTURE aggregate to form a larger particle, made up of a loose
structure of agglomerates of the primary particles
The particles of polymer produced in the dispersion (Figure 4). During cold, lubricated extrusion the
polymerisation process are of the order of 0.2m in size, agglomerated particles are highly distorted, with their
whilst those from a granular polymerisation are primary particles becoming aligned and also drawn into
hundreds of m in size, built up from smaller particles. fibrous material. This effect is illustrated in Figure 5.
They are both highly crystalline - about 90 to 95%.
If an extrudate such as that depicted in Figure 5 is
The dispersion particles can be studied directly in a sintered and cooled slowly the extrudate will appear as
conventional transmission electron microscope, in Figure 6. In Figure 6 the direction of extrusion was
providing the electron intensity is kept low (ref. 29). On from left to right.
raising the electron beam intensity the particles change
rapidly in appearance to become transparent with a Granular Polymers
crumpled texture. At this stage the crystallinity has During the early stages of polymerisation granular
disappeared and the particles probably consist of a particles form as aggregates of smaller particles. This
shell of carbon. Electron diffraction patterns and dark process continues and large irregular fibrous structures
field micrographs suggest that the particles are are produced.
composed of a pile of small single crystals (ref. 30) with
the molecular axis along the axis of the brick-shaped This material is then modified mechanically to reduce it
particles (refs. 30, 31). The particles also appear to have to the familiar form suitable for processing. For example,
a striated surface structure generally parallel to the long Figure 7 shows the structure of a fine particle granular
axis. A replica of some dispersion particles is shown in powder.
Figure 3. During coagulation, the dispersion particles

Figure 3. Particles of dispersion polymer Figure 4. Aggregates of particles of dispersion polymer

1m 100m

9
 Direction of extrusion

1m
Figure 6. Longitudinal section through sintered
extrudate

0.5m 10m
Figure 5. Longitudinal section through unsintered Figure 7. Internal and external texture of a granular
extrudate polymer

VOID FORMATION IN MOULDINGS Rhines as reported by Goetzel (ref. 32) the important
factors affecting powdered metals are the number of
When PTFE is moulded the basis of the process is that particle-to-particle contacts and their area. Experiments
the polymer powder is preformed and then sintered. It is (ref. 33) suggest that, even where no voids exist,
therefore not surprising that mouldings can be produced variations in internal pressure within a preform result in
which contain microscopic voids and fissures arising variations in crystallite size presumably due to the
from the porous nature of the unsintered particles and variations in melting point with pressure, for which
the molecular re-arrangement caused by sintering. McGeer and Duus (ref. 34) have given the relationship:
Correct choice of polymer grade and careful use of
fabrication techniques will help to minimise the Tm = a + bP
formation of voids. The homogeneity of the final
moulding depends a great deal on the compressibility of Where T = K, P = atmospheres, a = 597K and
the particles and their surface structure. b = 0.154C/atm so that 16 MN/m2 (160 kgf/cm2; 1 tonf/in2)
difference in internal pressure can lead to a melting point
In this respect the behaviour of the particles closely difference of the order of 25C (45F). Such variations in
resembles that of powdered metals. Softer particles internal pressure will result either from poor packing of
(generally more porous ones) compact more easily than the powder in the mould due to poor powder flow
harder ones and will flow more easily to fill properties or from a lack of compressibility in the
interparticulate voids. This leads to a higher green individual particles. The variation in crystallite size
strength (of the unsintered preform) and in turn to a resulting from poor packing leads to variation in
higher sintered strength. As with powdered metals the translucency from area to area and lack of
strength both of the preform and of the final sintered compressibility gives an overall white and opaque
moulding depends upon the mean particle size, with appearance in thin machined sections such as 0.25mm
finer particles giving higher strengths. According to (0.01 inch) skived tape.

10
Section 2. Effects of processing on properties

The exact manner in which the fabrication of the CRYSTALLINITY

polymer is carried out affects certain intrinsic qualities in
the polymer which, in turn, influence some properties, As polymerised, PTFE is of 90-95% crystallinity but when
notably tensile strength, elongation and dielectric it is heated above its crystalline melting point the
strength. The intrinsic qualities of PTFE which can be polymer becomes amorphous. In the melt state PTFE
affected by fabrication conditions are molecular weight, does retain some two-dimensional order but, as the
crystallinity, orientation and the presence or otherwise of three dimensional crystal structure has been destroyed,
voids and interfaces. These factors will now be molten PTFE can properly be thought of as amorphous.
considered in turn.
On cooling, the extent to which crystallisation occurs is
MOLECULAR WEIGHT dependent both on the molecular weight of the polymer
and on the rate of cooling, particularly at temperatures
All commercial samples of PTFE (with the exception of just below the melting point. Slow cooling naturally
lubricant powders) are of extremely high molecular tends to increase the extent of crystallisation. Almost all
weight; granular polymers are of somewhat higher fabricated PTFE displays crystallinities within the range
molecular weight than dispersion polymers. Because of 50 to 75% depending on the rate of cooling.
its insolubility, the molecular weight of PTFE can be
measured only with difficulty. The crystallinity of PTFE is best measured by X-ray
diffraction or by infra-red absorption methods These
The best method of assessing the molecular weight of respectively assess order and disorder within the
PTFE is undoubtedly by measuring the viscosity of the specimen. However as the relative density of 100%
polymer at some temperature above its melting point. crystalline PTFE at 23C (73F) is 2.30 while that of 0%
This approach demands, however, some fairly crystalline (100% amorphous) PTFE under the same
sophisticated equipment if it is to be successful and less conditions is 2.00, the relative density of PTFE can be
direct methods of arriving at an idea of molecular weight used to give a simple index of crystallinity. Clearly it is
are often favoured. essential that the specimens used for relative density
measurements be free from voids and, since such voids
When PTFE is cooled from a temperature above its are less likely to occur in specimens produced from
melting point at a carefully controlled rate, the extent to coagulated dispersion polymers, the technique is
which the polymer crystallises will be inversely probably of greater value in this field than in that of
dependent on the molecular weight. Now, assuming the granular polymers. The relationship between relative
moulding contains no voids, its crystallinity can be density and crystallinity for void-free specimens of PTFE
inferred from its relative density, so the relative density is shown in Figure 8.
of a moulding prepared under precisely defined
conditions (this quantity is often called the standard The level of crystallinity in fabricated PTFE is a matter of
relative density) can be used as an index of molecular considerable importance. In general, fabricators attempt
weight. The relationship: to keep the crystallinity to a minimum as the higher the
_ amorphous content the greater the ultimate tensile
standard relative density = 2.612-0.058 log10 Mn strength and the longer the flex life will be. In some
_ instances, however, such as where gas permeability
has been given where weight Mn is the number average must be reduced, higher crystallinities will be desirable.
molecular weight .

The thermal stability of PTFE is so great that, if

processed correctly, the polymer should suffer no
significant degradation and the molecular weight of the
finished product will be determined almost entirely by
that of the raw polymer.

11
 Figure 8. Variation of relative density with crystallinity

ORIENTATION The problem of voids in granular mouldings can almost

always be overcome by careful choice of the correct
The orientation of PTFE is a measure of the extent to granular polymer for a particular duty and by use of the
which the polymer chains are aligned in any particular appropriate fabrication technique.
direction. This alignment, because of their methods of
processing, is more likely to occur with the coagulated For practical purposes the absence of voids in fabricated
dispersion polymers than with the granular or aqueous PTFE is denoted by good tensile strength, high
dispersion materials. elongation at break, low gas permeability, prolonged flex
life and high dielectric strength. It is often possible to
VOIDS detect voids by examining a thin section - preferably less
than 0.25 mm (0.01 inch) - by transmitted light when a
During fabrication, PTFE does not pass through the state sample free from voids will be translucent with a faintly
of being a low viscosity liquid and, as a result, articles blue colour. On the other hand a sample containing many
made from PTFE may possess voids and interfaces in the voids will be white and opaque with a chalky appearance.
mass of the polymer. These faults are more likely to If the sample is first treated with a penetrant dye then
occur in large articles made from granular polymers than the detection of voids is made a great deal easier.
in articles of thin cross-section made from either
aqueous or coagulated dispersion polymers.
Such as Ardrox 996P supplied in the UK by Chemetall plc, 65
Denbigh Road, Bletchley, Milton Keynes, MK1 1PB (UK)
Tel. +44 (0) 1908 649333 Fax +44 (0) 1908 361872
www.aerospace.chemetall.com
in mid-Europe by Chemetall GmbH, Frankfurt a.M. Tel. +49 (0) 697165-0
and in the USA by Chemetall Oakite, 50 Valley Road, N.J. 07922,
Berkeley Heights
Tel. +1 908 508 2214 Fax +1 908 464 7914 Toll-free 800 526 4473
www.oakite.com
12
Section 3. Mechanical properties - unfilled PTFE

In some respects PTFE is a typical thermoplastic relationship at a constant time (isochronous stress-strain
polymer; in others it is far from typical. Thus the curve) (Figure 11). The isochronous curve is derived by
mechanical properties of PTFE vary with changes in taking a constant time section through a family of creep
time, temperature and crystallinity in the way that one curves and replotting the stress and strain values of the
would expect of a thermoplastic. On the other hand, the intersections to give the isochronous curve. The
fabrication methods used with PTFE can have a very derivation is shown schematically in Figures 10 and 11.
large influence on the properties, particularly if
unsatisfactory processing allows a particulate structure Alternatively the isochronous curve may be obtained
to persist into the fabricated article. PTFE is used only experimentally on a single specimen by the application
infrequently in tension so that it is appropriate to make of a series of stresses (1 to 6 Figure 11) in
measurements on samples in compression. The successively increasing steps and measuring the strain
properties of design interest will be considered in some
produced ( 1 to 6 Figure 11) after the section time, t,
detail. has elapsed, allowing a recovery period of 4t by
complete removal of the stress on the specimen between
COMPRESSIVE STRESS-STRAIN RELATIONSHIPS each successive increase (ref. 36). The isochronous
stress-strain curves presented here have been obtained
Although the classical concept of modulus, which in this way. Isometric curves (stress-time relationships at
implies a linear proportionality between stress and constant strain) may also be obtained by taking constant
strain, is not strictly applicable to most plastics, the term strain sections through a family of creep curves.
is widely used and the resulting implications should be
considered. At the termination of a creep experiment the
phenomenon of recovery may be studied by removing
The Youngs Modulus of a metal such as steel is the ratio the load on the specimen and observing the decrease of
of stress to strain in the elastic region, and is constant. strain with time. It is convenient to present recovery data
For most plastics such a region does not exist and the on a fractional recovered strain versus reduced time
ratio of stress to strain will not be constant but will graph as an aid to comparison of data obtained on
depend both on the time for which the stress is applied specimens which have either attained different
and the resulting strain. The time-dependence of strain maximum strains at the termination of the creep
may be defined as the creep behaviour and a study of experiment or for which the times under load have not
creep, together with the equally important phenomenon been identical. Fractional recovered strain is defined as
of recovery, is essential for a full understanding of the the ratio of the strain recovered to the creep strain at the
mechanical properties. An apparatus has been specially start of recovery and reduced time as the ratio of the
developed for studying the compressive creep of PTFE: recovered time to the creep time (ref. 37). Thus a
full details of this equipment have been published fractional recovered strain of unity signifies complete
elsewhere (ref. 35) but a general impression is given in recovery and a reduced time of unity denotes a recovery
Figure 9. With this equipment the stress-strain-time time equal to the preceding creep time.
relationship at a constant temperature may be obtained
by observing either the strain-time relationship at a
constant stress (creep) (Figure 10), or the stress-strain

Spherical seating

Figure 9. Basic creep testing equipment

Fixed
platen

Iris diaphragm Specimen

device
Moving
platen

Pivot
Beam
Automatic
Load loading
device

13
 6 6
5 4
6
3
5
5
6
4

5 4
2
3
4 3

3 2

1 2

2
1
1
1
Stress
Strain

t Strain at time (t)

Time - (log scale)

Figure 10. Compressive creep Figure 11. Isochronous stress-strain

relationship in compression

The following information is the result of work done on Creep behaviour

behalf of AG Fluoropolymers. A complete picture of the
behaviour of PTFE has not yet been obtained, and in Figure 13 shows a family of creep curves at various
particular, work on effects of temperature is not stress levels, while Figure 14 shows the same
complete. Nevertheless, enough data are now available information plotted as stress against time for various
to provide some basic information. The data given are for strain levels. It should be noted that the latter are not true
Fluon G163 preformed at a pressure of about 16 MN/m2 stress relaxation curves, though the curves should give a
(160 kgf/cm2; 1 tonf/in2), and sintered at 380C (716F). very approximate indication of the decay of stress with
time in a component maintained at a constant strain
Isochronous stress-strain behaviour level.
Figure 12 shows the effect of time on the stress-strain
relationship of Fluon G163. The non-linearity of the Recovery
curves, even at quite low strains, shows how the Figure 15 shows the effect of four different stress levels
apparent modulus decreases with increasing strain. on the rate of recovery of strain after removal of the
applied compressive load. It can be seen that the higher
the stress the slower is the recovery.

14
 8

ds

s)
80 Kgf/cm 2

on

u te
ec

in
m
0s
1100

46
10

rs
ou
H
(2
s
)

nd
7 rs

co
u
ho 1000

se
4
70 Kgf/cm 2 16

10
ys
da
1
(1
s
nd
co
6 se 900
10
6

60 Kgf/cm 2

800

50 Kgf/cm 2 700

600
4
40 Kgf/cm 2

500

3
30
400

2 300
20

200

1
10
Stress - MN/m 2

100
Stress - Ibf/in 2

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4
Strain - %

o o
Figure 12. Isochronous stress-strain relationship in compression, at 25 C (77 F), Fluon G163

15
 5.0

)
in 2
lbf/
60
:11
4.5

m2
gf/c

)
/in 2
.6 k

lbf
(81

15
:10
N/m 2

2
4.0

m
f/c
8M

kg
.4
(71
m 2
N/
7M
3.5

)
n 2
f/i
lb
3.0

7 0
:8
2
m
f/c
kg
2
1.
(6
2
2.5 m
N/
M
6

2 )
f/in
lb
25
2 :7
2.0 m
f/c
kg
0
1.
2 (5
m
N/
M
5 2 )
bf /in
1.5 0l
2 :58
cm
kgf/
.8
2 (40
N/m
4M 2 )
1.0 l bf/in
2 :435
cm
6 kgf/
2 (30.
N/m
3M
2
f/in )
:290 lb 2

0.5 2 (20.4 kgf/cm

2 MN/m
2)
m
2 :145 lbf/in
(10.2 kgf/c 2
Strain - %

1 MN/m

0
1 10 10 2 103 104 105 106 107 108
6 Months
1 month
1 week

Time - seconds
1 hour

1 year
1 day

o o
Figure 13. Creep in compression, at 25 C (77 F), Fluon G163

16
 8

80 kgf/cm 2

1100

7
1000
70 kgf/cm 2

900
6

60 kgf/cm 2

800

50 kgf/cm 2 700

600
4
40 kgf/cm 2

500

3
30 kgf/cm 2
400

2 300
2
20 kgf/cm

200

1
10 kgf/cm 2
Stress - MN/m2

100
Stress - Ibf/in 2
6 Months
1 month
1 week
1 hour

1 year
1 day

0 0
2
0 10 10 103 104 105 106 107 108
Time - seconds

o o
Figure 14. Isochronous stress-strain curves in compression, at 25 C (77 F) and various strain
levels, Fluon G163
17
 0.8

2
0.6 0 lbf/in
cm ; 29
2
2 (20.4 kgf/ 2
2 MN/m 2 ; 580 lb
f/in
2
Fractional recovered strain

2 (40.8 k
gf/cm lbf/in
/ m 2 ; 870
4M N m 2
0.4
2 (61.2
kgf/c lbf/in
2 ; 1160
/m m
6 MN 2 1.6 kgf/c
/ m (8
8 MN
0.2

0
10-6 10-5 10-4 10-3 10-2 10-1 1 10 102
Reduced time
o o
Figure 15. Recovery from creep in compression, at 25 C (77 F), Fluon G163

TENSILE PROPERTIES Unnotched specimens of PTFE are resistant to fracture on

impact; even at temperatures as low as -196C (-
The tensile breaking stress and breaking strain are used 320F) well-fabricated specimens are tough. A test for
extensively for quality control purposes, but they are judging the quality of a sample from this point of view is
unsatisfactory quantities for design purposes for two to measure the flexural strength of specimens which
reasons: firstly, and most importantly, PTFE should never have been immersed for 15 minutes in liquid nitrogen
be used at strains beyond the yield point (the point at and then tested within a few seconds of removal. In this
which the load-deformation curve has a distinct change liquid nitrogen dip test which was carried out with three
of slope) and secondly, the point of fracture is dependent point loading, a span of 38mm (1.5 inches), a thickness of
on specimen shape and is therefore not useful for 3.2mm (0.125 inch) and a rate of test of 457mm/min (18
predicting behaviour in practice. inches/min) good specimens of PTFE do not break at the
maximum load, the apparent yield stress of such a
The tensile load-extension curves obtained with specimen being approximately 185 MN/m2 (1900 kgf/cm2;
specimens of PTFE depend on crystallinity, molecular 27 000 lbf/in2). However, less well-fabricated specimens
weight, the size, shape and perhaps the structure of the may be brittle with flexural strengths of approximately
original particles and the severity of faults remaining 135 MN/m2 (1400 kgf/cm2; 20 000 lbf/in2) in this test.
after fabrication. Furthermore they depend, as is usual
with thermoplastics, on test temperature and straining The behaviour of notched specimens typifies the reaction
rate. Because of these complications the data here can of PTFE components with built-in stress concentration
only be indicative of general behaviour. Figure 16 shows regions. This is shown by measurements of the Charpy
the general trends of behaviour in tension for PTFE as a impact strength: the test was carried out with three-point
function of temperature. These are typical curves from loading and an impact velocity of 2.44m/second (8
which the yield stress can be derived, though less ft/second) at temperatures between -35 and +23C (-31
precisely than is possible for most other plastics and +73F). One sample was cooled slowly at 25C/hour
materials. (45F/hour) and another cooled from the sintering
temperature of 380C (716F) to 20C (68F) in two hours.
The effect of temperature on the yield stress of PTFE is The notch tip radius of the specimens was varied
shown in Figure 17, which is for times to yield of between 0.25mm (0.01 inch) and 2.03mm (0.08 inch),
approximately one minute. If the material is to be under spans of 25mm (1 inch) and 38mm (1.5 inch) were used
load for any considerable length of time it should not be and the notch depth was held constant at 2.82mm (0.110
stressed beyond a small fraction of the yield stresses inch). There was no consistent difference between the
shown in Figure 17. impact strengths of the samples cooled at different rates.
At temperatures of -20C (-4F) and below all notched
IMPACT BEHAVIOUR specimens broke completely, [impact strength in the
range 6 to 10 kgf/cm2 (3 to 5 ft Ibf/in2), with an 0.25mm
The behaviour of plastics under impact conditions (0.010 inch) notch], whilst at -10C (+14F) and above
depends both on temperature and on the severity of the many specimens did not break completely - that is to say
applied stress, as well as on molecular parameters such hinge breaks occurred.
as molecular weight and fabrication effects. PTFE is no
exception to these generalisations and with the wide
variation in fabrication procedures available for this
polymer it is impossible to give other than general data.

18
 40
400 kgf/cm 2

-40C (-40F)
-20C (-4F)
35
5000

-10C (-14F)
30
300 kgf/cm 2

4000

0C (-32F)

25
10C (-50F)

3000
20
200 kgf/cm 2

25C (77F)

15

2000

50C (122F)

10
100 kgf/cm 2

100C (212F)

1000

150C (302F)
5

200C (392F)
2
Stress - MNm

Stress - Ibfin 2

0 0
0 100 200 300 400
Strain - %

Figure 16. Effect of temperature upon tensile stress-strain curves for PTFE

19
 200

2000 kgf/cm 2

25000

150
1500 kgf/cm 2

20000

15000
100
1000 kgf/cm 2

10000

50
500 kgf/cm 2

5000
Yield stress - MN/m 2

Yield stress - lbf/in 2

0 0
-220 -140 -60 20 100 180 260C
-364 -220 -76 68 212 356 500F
Temperature

Figure 17. Effect of temperature upon tensile yield stress of PTFE

20
Section 4. Electrical properties - unfilled PTFE

PERMITTIVITY AND DIELECTRIC LOSS

EXPERIMENTAL TECHNIQUES
It has been known since 1946 (ref. 38) that because of its
non-polar nature, the dielectric properties of PTFE were Measurements of permittivity and dielectric loss in the
of an ideal character. In 1953 a careful study by Ehrlich in audio frequency range (178 Hz - 31.6 kHz) were made
the USA (ref. 39) showed that the fall in permittivity from using a fully-shielded, three-terminal conjugate Schering
2.0 to 1.8 in the temperature range 24 to 314C (75 to bridge; a resonance substitution method, based on that
597F) could be accounted for entirely in terms of density published by Hartshorn and Ward (ref. 43) but suitably
changes by the Clausius-Mossotti formula. No changes modified to give a higher resolution (ref. 44), was used
of permittivity with frequency were detected and for the radio frequency region (105-108 Hz). A modified
scarcely resolved loss angle values less than 200 version of the re-entrant cavity resonator method of
radians were recorded. Parry (ref. 45) was used for the 108 - 109 Hz range; an H01
cavity resonator was used at 9 x 109 Hz: this made use of
In 1955 Mikhailov and co-workers in the USSR (ref. 40) the Bleaney, Loubser and Penrose (ref. 46) method of
found a loss peak in the -80 to -40C (-112 to -40F) range avoiding unwanted modes.
at audio and radio frequencies which was correlated in
its temperature/frequency location with dynamic D.C. CONDUCTION BEHAVIOUR
mechanical loss behaviour. From studies of the effect of
changes of crystallinity by quenching and slow cooling In attempting to study the d.c. conduction behaviour of
they concluded that the relaxation losses were PTFE, the current measured was that which occurs after
attributable to amorphous regions of the polymer. In the application of a d.c. voltage step. In Figure 20 the
1959 Krum and Muller (ref. 41) (of Marburg) reported results for a typical sintered sample, using evaporated
higher dielectric loss values than those found by earlier gold electrodes, are expressed as log (apparent volume
workers and found more detailed correlation with resistivity) as a function of time of polarisation. It will be
mechanical properties and effects of crystallinity seen that steady state conduction was not established
changes. Eby and Sinnott in the USA (ref. 42) have clearly in the time of the experiments (which was 15
suggested, however, that these higher loss values must minutes). On the diagram are shown lines of constant
be due to polar impurities. loss angle which can be calculated by means of a Fourier
transform assuming a constant permittivity of 2.0. The
The results of measurement done on behalf of AG short time values are consistent with the low frequency
Fluoropolymers are presented in Figure 18 where the values (= 20 radians) measured by a.c. methods. In fact
variation of the loss angle with temperature is given for it is considered that such d.c. step response results are
the range -140 to +240C (-220 to +464F) and in Figure 19 equivalent to a low frequency extension of the a.c.
as loss angle versus frequency at room temperature. frequency range (refs. 47, 48, 49). The apparent
These data, which were obtained using the experimental resistivity is to be thought of as a very low frequency
techniques described in the next column of this page, relaxation loss phenomenon rather than a steady state
confirm and extend the findings of Ehrlich and Mikhailov charge transport phenomenon, although the onset of
and support Eby and Sinnotts contention that Krum and conduction may be apparent above 160C (320F).
Mllers higher loss values must be due to polar
impurities. Figure 21 shows that such currents may be removed by
heat treatment in the presence of electrodes; apparent
The dielectric loss of PTFE is sufficiently low to allow the resistivity values of > 1018 ohm m have been obtained
permittivity to be calculated with an accuracy of better from such experiments without showing evidence of
than 0.5% using the Clausius-Mossotti formula: steady state conduction (ref. 47). However, much lower
values are often obtained if unsatisfactory electrodes are

-1 used or if temperature stability or polarising voltage
= Kd
+2 stability are not good enough to exclude
where
= permittivity dC dV
V or C terms for the current.
d = relative density dT dT

K = constant, 0.119
From which it can been seen that:

1 + 2 Kd
=
1 - Kd

thus, at a relative density of 2.174 the permittivity is

2.05.

Loss angle in radians is very nearly equal to loss tangent (tan ) x l06, ie 100 radians = loss tangent of 0.0001)

21
 240

200

31600 Hz
5620 Hz
160
1000 Hz
178 Hz
32 Hz

120

80
Loss angle - radians

40

0
-140 -140 -80 -40 0 40 80 120 160 200 240 C
-220 -184 -112 -40 32 104 176 248 320 392 464 F

Temperature

Figure 18. Loss angle versus temperature for sintered granular PTFE (60% crystallinity) using
evaporated gold electrodes

300

Sintered granular polymer, 60% crystallinity

Unsintered coagulated dispersion polymer
93% crystallinity

200

100
Loss angle - radians

0
1 2 3 4 5 6 7 8 9 10
log10 Frequency - HZ

Figure 19. Loss angle versus frequency for PTFE at room temperature

22
 18.5

150C
302F

18.0
160C

na s
320F

ar id
141C
m mho - )j/E( ytivitsiser emulov tnerappA 01gol

286F

5
=
171C

el g
n
340F 19.6C

as
17.5 67.3F

ol s

s
sn

na
17.0

id
id a

ar
r

00
05

5
=

=
elg

elg
n

n
as

as
so

ol s
16.5
l

0 1 2 3
log10 Time - seconds

Figure 20. Plot of log10 apparent volume resistivity versus log10 time for sintered granular PTFE

18

17

16
Evaporated gold electrodes
(Heated 40 hours at 280C;536F)
15
m mho - )j/E( ytivitsiser emulov tnerappA 01gol

14
Tinfoil electrodes
(as received)
13

12

11

10

-6 -4 -2 0 2 4
log10 Time - seconds

Figure 21. Effect of different electrode materials and preconditioning for sintered granular PTFE

23
HIGH VOLTAGE USES OF PTFE

With regard to high voltage applications it has been

known for a long time that in the presence of surface
discharges failure occurs by erosion, as PTFE is a non-
tracking material. Parr and Scarisbrick (ref. 50) have
compared the behaviour of a wide range of polymeric
insulators by the IEE tracking test using electrolyte, and
by an ASTM dust-fog test (D21 32-62T). They found that
PTFE was one of the erosion class which showed a long
life, i.e. >1000 hours in the dust-fog test. Thus PTFE has
useful surface characteristics for exploitation in outdoor
applications.

For bulk insulation high quality fabrication will be

required in order to produce structures with the very low
level of porosity and internal voiding demanded by high
voltage applications (ref. 51). Tests by means of
electronic discharge detectors (ref. 52) can be made to
ensure freedom from the damaging discharges which
may occur in voids. Alternatively it is possible to reduce
the discharges by impregnating the PTFE with dielectric
liquids or with a high pressure gas so as to fill, at least
partially, any voids in the polymer. In consequence,
values for dielectric strength obtained from tests
conducted in oil may be misleadingly high for poorly
fabricated PTFE due to impregnation of any voids present
by the oil.

24
Section 5. Thermal properties - unfilled PTFE

MELTING POINT 9.5 x 10-5 / C. Work by Kirby (ref. 53) indicates that this
coefficient is approximately constant down to -190C.
The melting point of as polymerised PTFE increases Above 15C the coefficient of expansion increases with
with increasing molecular weight and Wunderlich (ref. temperature. Therefore a more useful way of indicating
52a) has shown that PTFE also superheats, i.e. the thermal expansion is to express it as a percentage
apparent melting point increases with increasing heating increase in length between two temperatures. Values
rate. obtained parallel to and at right angles to the direction of
the moulding pressure are quoted separately as they
Melting points determined by Differential Scanning were found to be slightly different.
Calorimetry* on as polymerised powders at a heating
rate of 16C/minute (28.8F/minute) vary from about
332C (630F) for low molecular weight coagulated Temperature Parallel Perpendicular
dispersion polymer to about 346C (655F) for high range to direction to direction
molecular weight granular material. Measurements of moulding of moulding
made at different heating rates indicate that, owing to pressure pressure
the superheating effect, these values may be up to 10C C F % %
(18F) higher than would be obtained at infinitely slow
heating rates. 15 to 30 59 to 86 0.4 0.4
30 to 50 86 to 122 0.3 0.3
The influence of molecular weight on melting point is 30 to 100 86 to 212 0.8 0.8
much reduced after the polymer has been sintered (once 30 to 150 86 to 302 1.5 1.5
melted). Most sintered polymers melt in the range 325- 30 to 200 86 to 392 2.4 2.3
330C (617-626F) when reheated at 16C/minute 30 to 250 86 to 482 3.4 3.6
(28.8F/minute).

The way in which the melting point of sintered PTFE THERMAL CONDUCTIVITY
varies with applied pressure was studied by McGeer and
Duus (ref.34) who reported the following values: Over the temperature range 20-35C (68-95F) the
thermal conductivity of PTFE is 6 x 10-4 cal/cm s C. This
1 atmosphere 324C (615F) result may be expressed in a variety of units:
69 atmospheres 335C (635F)
207 atmospheres 356C (673F) 6 x 10-4 cal/cm s C
615 atmospheres 419C (786F) 2.2 x 10-1 kcal/m h C
2.6 x 10-3 joule/cm s C
These latter workers used their data to calculate the 1.7 Btu in/ft2 h F
latent heat of fusion of PTFE as 14 cal/g at 69
atmospheres and 8.4 cal/g at 207 atmospheres. The Kline (ref. 54) measured the thermal conductivity of PTFE
corresponding entropies of fusion are 0.0240 cal/g deg K at 0, 20, 50 and 70C (32, 68, 122, 158F). He reports the
and 0.0134 cal/g deg K. conductivity to be fairly constant, with a slight tendency
to rise at the higher temperatures. His value is about 5.1
THERMAL EXPANSION x 10-4 cal/cm s C.

If a graphical representation is made of the specific Eiermann and Hellwege (ref. 55) studied this property
volume/temperature relationship for highly crystalline over a much wider temperature range of -180 to +90C (-
fabricated PTFE the form of the graph is shown in Figure 292 to +194F). All their values fell within the range 5.4 to
22. This graph clearly reveals the presence of the 6.1 x 10-4 cal/cm s C. It was confirmed that the
transition point which occurs at 19C (66F) with PTFE conductivity tends to rise with temperature though a
(refs. 7, 53). sharp fall occurred at 20C (68 F), approximately the
temperature at which it has already been noted that a
Work done on behalf of AG Fluoropolymers showed that lattice transformation of the crystalline component of the
from -60C to +15C expansion is approximately linear at polymer occurs.

* The melting point of a polymer, as measured by DSC, is taken as the temperature at which the peak of the melting endotherm occurs. This peak is
reached when the rate of melting is maximal and indicates the melting point of the bulk of the polymer. The final melting point will be slightly higher
than this.
25
 0.66

0.64

0.62

0.60

0.58

0.56

0.54

0.52

0.50
Specific volume

0.48

0.46

0.44
-50 0 50 100 150 200 250 300 350C
-58 32 122 212 302 392 482 572 662F
Temperature

Figure 22. Variation of specific volume with temperature

SPECIFIC HEAT AND HEAT OF FUSION Lewis and Naylor (ref. 59) that when PTFE was
decomposed at temperatures between 600 and 700C
The specific heat of PTFE has been determined by Marx (1112 and 1292F) under pressures of 5 to 760mm Hg the
and Dole (ref. 56). For temperatures above 40C(104F) products were C2F4, C3F6 and C4F8 and that the proportion
they give the relationship: of tetrafluoroethylene among the products increased
with decrease in pressure and tetrafluoroethylene was
Cp = 0.227 + (2.50 x 10-4) T cal/g C. the sole product at very low pressures.

The heat capacity, enthalpy and entropy of PTFE have Cox et al. (refs. 60, 61) have studied the thermal
been studied and results are reported in two papers (refs. degradation of PTFE with particular reference to the
16, 57). differences observed between degradation in a vacuum
and in oxygen. They found that the temperature
THERMAL STABILITY necessary to achieve a 25% weight loss in two hours was
494C (921F) in a vacuum and 482C (900F) in oxygen;
Within its normal range of working temperatures, the they concluded, therefore, that the thermal degradation
upper limit of which is generally quoted as 260C (500F), of PTFE was relatively little affected by oxidising
PTFE suffers no degradation. Indeed, weight losses conditions.
observed between 260 and 360C (500 and 680F) will be
exceedingly small and due to the loss of minute amounts Siegle et al. (ref. 62) have evaluated the mechanism of
of moisture or gas absorbed in the polymer. the depolymerisation reaction from research work done
on heating thin PTFE films in a vacuum and Jellinek (ref.
At processing temperatures of about 380C (716F) the 63) reached similar conclusions. In the case of thicker
rate of decomposition of PTFE is still very low and it is sections, which are more likely to be met in practice, the
only at temperatures in excess of 400C (752F) that rate of pyrolysis is controlled by diffusion of monomer as
thermal decomposition of pure PTFE becomes pointed out by Siegle and Muus (ref. 64).
significant.
Paciorek et al. (ref. 65) studied the auto ignition of PTFE
Madorsky et al. (ref. 58) studied the pyrolysis of PTFE in in oxygen and in air. The respective auto-ignition
a vacuum at temperatures from 423.5 to 513C (794 to temperatures were 512C (954F) and 575C (1067F). In
955F) The decomposition rates which they report at oxygen only COF2, CO2 and CF4 were formed, while in
these temperatures are respectively 0.00152% per air, saturated fluorocarbons, COF2 and CO were the most
minute and 1.264% per minute. They further reported abundant species.
that tetrafluoroethylene was virtually the only product of
decomposition. This confirmed earlier reports of

26
Section 6. Surface properties - unfilled PTFE

COEFFICIENT OF FRICTION Makinson and Tabor (ref. 74) have also examined the
effect of sliding velocity and substantially agree with the
Although the low coefficient of friction of PTFE is widely variation in coefficient of friction with velocity given
known, it is interesting to reflect that no reference was above. They have found that whereas at low velocities a
made to this characteristic until about ten years after the thin continuous film of PTFE is laid down on the other
discovery of the polymer. Credit for this first publication slide surface (in this case glass), at higher velocities the
goes to Shooter and Thomas (ref. 66) who measured the PTFE is torn off in discrete fragments.
coefficient of friction using a Bowden-Leben machine
with loads of between 1 and 4 kg (2.2 and 8.8 Ib) and Of less general importance than the dependence on load,
sliding velocities from 0.1 to 10mm/s (0.02 to 2 ft/min). velocity and temperature, but still of interest is the
They reported that the coefficient was 0.04. observation by Tabor and Williams (ref. 75) that the
coefficient is influenced by the orientation of the
Other workers (refs 67, 68, 69) report that while polymer, the coefficient being about 30% higher when
Amontons law is fairly well obeyed at moderate loads sliding was across the chains than when it was along
the coefficient of friction rises steeply at very light loads, them.
say below 100g (31/2 oz). Thompson et al. (ref. 70) who
studied the coefficient at high loads found the extremely ANGLE OF CONTACT
low figure of 0.016 at a load of 1360kg (3000 Ib) .
Zisman and his co-workers have studied the contact
The coefficient of friction is dependent also on the sliding angles made with PTFE by a wide range of liquids (refs.
velocity, a high speed resulting in a high coefficient. By 76, 77, 78,79). A few of their results with common liquids
combining a low load and a high sliding velocity of are summarised:
1.89m/s (370 ft/min), Flom and Porile (ref. 71) found the
high value of 0.36 for the coefficient.
Liquid Contact angle
In their pioneer paper Shooter and Thomas (ref. 66)
claimed that, at the very low speeds they used, the Water 108
coefficient was independent of temperature over the n-Hexadecane 46
range 20 to 200C (68 to 392F). However, later work has Toluene 43
shown that temperature has some effect. King and Tabor Benzene 46
(ref. 72) report that the coefficient remains steady at Methylene iodide 88, 83
about 0.1 over the range 100 to -45C (212 to -49F). On Carbon tetrachloride 46
further cooling the coefficient rises to about 0.2 but does Mercury 150
not alter further even when the polymer is cooled to - Glycerol 100
80C (-112F). For the behaviour at elevated
temperatures the best guide is the work of McLaren and
Tabor (ref. 73) who demonstrated that the coefficient of Fox and Zisman (ref. 76) found that there was a critical
friction fell with increase in temperature. surface tension (ca 17.5 to 20.5 dynes/cm) below which
liquids would wet PTFE (i.e., would spread on a smooth
polymer surface).

27
Section 7. Other physico-chemical properties - unfilled PTFE

PERMEABILITY investigated the permeation of sulphur dioxide through a

range of polymers including PTFE.
Work using a sample of cast film produced from Fluon
GP1 led to the following test results for permeability: The permeability of PTFE to water vapour has been
studied by Konovalov (ref. 86) and by Korte-Falinski (ref.
Oxygen 10.5 x 10-10 87) who both found that PTFE has a lower permeability to
Nitrogen 4.0 x 10-10 water vapour than almost any other plastics material
Air 5.3 x 10-10 examined. For PTFE films in the thickness range 0.05-
0.20mm (0.002-0.008 inch), values were found (ref. 87)
The units are cm3 of gas at NTP x cm (thickness)/cm2 (area). equivalent to about 0.9-1.8 g/m2 per 24 hours, per
s. cm Hg measured at 23C 1C (73F 2F) 0.025mm (0.001 inch), at 20C (68F). Toren (ref. 88),
using a special electrolytic measuring technique,
Work published by Barton (ref. 80), who uses the same obtained a value equivalent to 2.7 g/m2 per 24 hours per
units, can be summarised as follows: 0.025mm (0.001 inch), for a PTFE film 0.08mm (0.003
inch) thick. A value of 5.4 g/m2 per 24 hours per 0.025mm
Hydrogen 2.4 x 10-9 (0.001 inch) at 30C (86F) has also been quoted (ref. 89).
Helium 7.0 x 10-8 These somewhat variable results for water vapour
Nitrogen 3.1 x 10-10 permeability of PTFE may, most probably, be explained
Oxygen 1.0 x 10-9 by differences in the film fabrication techniques used, as
Argon 5.8 x 10-10 well as by different methods of measurement.

Yasuda and Stone (ref, 81) obtained a substantially INFRA-RED TRANSMISSION

higher figure, 23.7 x 10-10 (units as above), for gaseous
oxygen, and an even greater value, 91.0 x 10-10, for Figure 23 shows the infra-red transmission spectrum for
dissolved oxygen, while Pasternak et al. (ref. 82) obtained PTFE.
a much lower value, 4.2 x 10-10, in experiments with
membranes in the thickness range 0.081-0.145mm REFRACTIVE INDEX
(0.0032-0.0057 inch). The latter authors also give values
for hydrogen (9.8 x 10-10), nitrogen (1.4 x 10-10), and Billmeyer (ref. 90), using sodium yellow light and a
carbon dioxide (11.7 x 10-10). Casper and Henley (ref. 83), sample of PTFE of density 2.12 reported a refractive
using PTFE film 0.094mm (0.0037 inch) thick found a index of:
value of 11.6 x 10-10 for hydrogen, and 0.65 x 10-10 for
nD= 1.376
ethane.
Using a far infra-red maser and monochromatic radiation
Work on the helium permeability of fabricated PTFE of wave-length 337m Chamberlain and Gebbie (ref. 91)
items has shown that the permeability is very dependent report a figure of 1.391 0.017.
on crystallinity (as indicated by relative density). For
isostatically-moulded granular PTFE discs the helium The refractive index of PTFE would be expected to vary
permeability at 25C (77F) varied from about 30-40 x 10- with density, or more strictly with crystallinity, in
15 mol m s-1 N-1 at a relative density of 2.08 to about 5-15
accordance with the equation:
x 10-15 moI m s-1 N-1 at a relative density of 2.15: the effect
of crystallinity was much greater than that of varying the n2 - 1 1
moulding pressure, or the type of PTFE polymer used. x=K
For tubing extruded from coagulated dispersion (CD) n2 + 2 d
polymers, a similar effect of crystallinity on permeability
where
was observed. At a relative density of 2.15 the
n = refractive index
permeability was about
d = density
15-25 x 10-15 mol m s-1 N-1 and this fell to about 5-10 x 10-
15 mol m s-1 N-1 at a relative density of 2.21. Again, no K = constant
effect of CD polymer type could be detected, even though
MELT VISCOSITY
a considerable number of both homopolymers and
copolymers were examined.
The melt viscosity of PTFE is extremely high by
comparison with other polymers. The observed value
Gerritse (ref. 84) has measured the permeability of PTFE
will depend somewhat on the experimental method
to oxygen and nitrogen as a function of temperature in
used, of which the parallel plate plastometer, the
the range 50-125C (122-257F); for both gases the
capillary rheometer and creep methods are the most
permeation rate at 125C was about 5-6 times greater
important.
than at 50C. Felder, Spence and Ferrell (ref. 85)

28
 100

80

60

40
Transmittance - %

20

0
3 4 5 6 7 8 9 10 11 12 13 14 15
Wavelength - microns

Figure 23. Infra-red transmission spectrum (specimen thickness 0.05mm [0.002 inch])

The melt viscosity of PTFE varies with the shear stress 1 Megarad has a measurable effect and 2-3 Megarads in
applied to the polymer and with the temperature of the air reduce strength by 40-75%. 4 Megarads reduce
polymer but, in general, commercial samples of granular tensile strength to 2% of the original.
polymer display viscosities of about 1011 poise in the
temperature range 360 to 380C (680 to 716F) and at A 1975 report from AERE, Harwell (ref. 99a) gives details
shear stresses of about 106 dynes/cm2 (refs. 92, 93, 94). of the effects of radiation on the mechanical and
electrical properties of PTFE used in cable for the GEOS
HIGH ENERGY RADIATION satellite.

The effect of high energy radiation on PTFE was first CHEMICAL RESISTANCE
noted by Liversage (ref. 95) who found that the electrical
resistance of the polymer fell on exposure to X-rays. As might be expected of a saturated aliphatic
fluorocarbon PTFE is almost completely inert chemically.
Harrington and Giberson (ref. 96), in a study of the Molten or dissolved alkali metals degrade PTFE by
decline in the tensile strength and elongation of PTFE abstracting fluorine from the molecule, while at elevated
when exposed to gamma radiation, showed that temperatures fluorine and compounds capable of
irradiation in a vacuum was less damaging than releasing fluorine can break the carbon skeleton and
irradiation in air. This point was confirmed by Wall and form low molecular weight fluorocarbons. Apart from
Florin (ref. 97) and a summary of their results is given these not very important exceptions, PTFE resists attack
below: by all the acids, bases and solvents that might be
encountered in industrial practice.

In addition to its remarkable chemical inertness, PTFE is

Radiation dose % retention of original not dissolved by any solvent within its normal range of
eV/g x 10-20 tensile strength working temperatures. Small quantities of solvents may
be absorbed by PTFE on prolonged exposure especially
Irradiation in 2.4 2 at elevated temperatures but this in no way impairs the
air 4.1 0 usefulness of the polymer.

Irradiation in 0.7 73 Rossa (ref. 100) has given details of the effect of 79
a vacuum 4.1 51 chemicals on PTFE with, in many cases, data on weight
32.0 43 gain.

Two general surveys of this subject have been made VELOCITY OF SOUND
(refs. 98, 99).
The velocity of sound in PTFE and the way in which the
The more recent work by Monnet and Bensa (ref. 99) velocity changes with changes in temperature has been
gives further data on the effect of radiation dose on studied by Kravtsov (ref. 101). He showed the velocity to
mechanical properties. They found that as little as 0.01 to pass through a maximum at 20C (68F) in the region of
0.1 Megarad dose can affect mechanical properties. the first-order transition.

29
Section 8. General-filled PTFE

FUNCTION OF FILLER CHOICE OF FILLER

The incorporation of fillers has the following general It is this choice which presents most problems to the end
effects: user. Unfortunately no simple answer is available, mainly
because the performance of different fillers cannot
(1) Wear resistance is increased to a very marked always be predicted reliably. Choice frequently depends
extent. upon the results of empirical testing rather than upon
(2) Resistance to creep or deformation under load is any understanding of the mechanism affecting
increased by a factor of 2 to 5. properties.
(3) Depending upon the filler used, the thermal
conductivity may be increased significantly. Many materials are candidates as fillers for PTFE
(4) Depending upon the filler used, thermal expansion provided they can be obtained in the appropriate particle
may be reduced by a factor approaching 5. size and will withstand the processing conditions
(5) By suitable choice of fillers some control over the necessary to incorporate them and subsequently to
electrical properties of PTFE can be achieved. enable the compound to be fabricated. Other
considerations such as availability, cost and
In addition to these advantages filled PTFE generally processability further restrict the number of potential
retains low coefficients of friction, the wide service fillers to a relatively small number, as indicated in Table 1.
temperature range and, depending upon the filler, the
chemical inertness of unfilled PTFE.

Table 1. Fillers used with PTFE

Filler Form Details

Glass Usually milled fibres A minimum aspect ratio

(length: dia) of 10 : 1 is
generally used

Carbon and graphite Usually in the form of a high purity powdered Particle size usually
coke, or natural or synthetic graphite. Particles less than about
are generally irregular in shape although fibres 60m
are known

Metals Notably bronze as irregular or spherical particles Particle size usually

less than about
60m

Others Various forms of ceramic material including mica are

used both in particulate and fibre form.
Molybdenum disulphide
Thermoplastics

Note: 1m = 10-6m ~
_ 40 x 10-6 inch

30
 CHOICE OF FLUON GRADE where lower filler contents of 5% or 15% are justifiable.
There is little case, however, for requiring intermediate
Having limited the number of potential fillers, the filler contents.
number of grades of filled PTFE which it is possible to
manufacture is still large, when combinations and Combinations of filler can give rise to improved wear
varying percentages of filler are taken into consideration. properties but this appears to be their only advantage.
However, there are reasons why these too may be
limited: THE FLUON RANGE

(1) The quantity of filler should effectively be limited The Fluon filled polymer range consists of moulding
to 40% by volume if reasonable mechanical grades denoted by a prefix FC1---
strengths are to be maintained.
(2) The method of fabricating the filled PTFE powder Free-flowing moulding grades (agglomerated) are
can have considerable effect upon its properties denoted by a prefix FC8---
and these can therefore be varied and controlled
during fabrication rather than by a proliferation of Development grades are denoted by a prefix XC e.g.
diverse formulations. XC1--- for moulding grades.

It is considered reasonable therefore to limit the basic Typical properties of the Fluon moulding grades are
range of filled compounds to filler additions of 25% and given in Table 2 and compared with those of unfilled
40% by volume although there may be exceptional cases PTFE.

Table 2. Properties of the Fluon range

Units FC168-63* FC100-15 1000 FC100-25 1000 FC160-60 FC150-25 FC140-15 Unfilled
PTFE

Fillers Bronze and Glass fibre Glass fibre Bronze Powdered Graphite None
graphite coke

FiIler by weight % 63 15 25 60 25 25 -

Filler by volume % 40 13.3 22.2 27 28 15 -

Relative density 3.2 2.25 2.25 3.8 2.1 2.2 2.17

Maximum tensile MN/m2 4.9-8.3 17.2-240 12.3-19.6 10.3-13.7 11.8-15.2 13.7-20.6 20.6-29.9
strength: moulded kgf/cm2 50-85 175-245 125-200 105-140 120-155 140-210 210-350
Ibf/in2 700-1200 2500-3500 1800-2800 1500-2000 1700-2200 2000-3000 3000-5000

Maximum tensile MN/m2 - 9.2-12.7 6.9-10.8 - 10.3-13.7 - 13.7-17.2

strength: extruded kgf/cm2 - 100-130 70-110 - 105-140 - 140-175
lbf/in2 - 1400-1900 1000-1600 - 1500-2000 - 2000-2500

Ultimate tensile % 10-15 300-400 200-300 80-160 100-150 200-300 250-400

elongation: moulded

Ultimate tensile % - 200-300 100-200 - 75-125 - 250-400

elongation: extruded

Hardness Shore D 55-65 55-60 55-70 55-65 60-65 55-65 50-55

Heat stability The weight loss at 300C (572F) should not normally exceed 0.1%

Porosity: All material can be made non-porous, depending upon the fabrication techniques used
dye immersion

*British Patent No 870117, 926718 and others

31
Section 9. Mechanical properties - filled PTFE

COMPRESSIVE DEFORMATION The room temperature creep behaviour of filled PTFE has
been studied and Figures 24 and 25 show how strain
The compressive creep behaviour of filled PTFE is one of varies with time and also at various stress levels for
the most significant properties to the designer and user. glass-filled Fluon FC100-25 1000 and, for comparison
purposes, for Fluon G163. Figure 26 illustrates the
In comparison with unfilled PTFE the compressive relationship between stress and strain in a way which
modulus of filled PTFE is greater and hence, for the same indicates the stress relaxation that can occur in PTFE with
stress, a lower deformation will occur. This difference time. Figure 27 shows the divergence in creep behaviour
diminishes, however, as temperatures rise and the between Fluon grades containing 25%, 22% and 40% of
compressive performance of unfilled and filled PTFE filler by volume. Figure 27 used in conjunction with
draws close together at 200C (392F). The reason is Figure 24 indicates the relative creep behaviour of Fluon
almost certainly that the general reduction in the filled grades.
compressive modulus of the PTFE matrix overrides the
support given by the random distribution of filler Individual design details of specific applications are
particles. Table 3 gives approximate values for the stress- bound, however, to significantly modify the data given in
strain relationship at different temperatures for Fluon Figures 24 and 27.
FC100-25 1000. These data are based on laboratory tests
and the configuration of the test specimens will have a Recovery
significant effect on the results obtained. For this reason After being subjected to a compressive load, filled PTFE
the figures should only be taken as a very approximate will recover some of the resulting strain when the stress
guide. is removed. This recovery is slow and may not be
complete. Figure 28 shows the recovery of Fluon FC100-
25 1000 after the removal of applied stress at various
levels.

Table 3. Stress-strain relationship for Fluon FC100-25 1000 at different temperatures and stresses to produce
stated compression strain after 24 hours.

Stress - kgf/cm2 at T : C
20 60 100 140 180 220 250

1% strain 63 37 27 17 10 8 6

2% strain 89 60 47 32 20 15 11

3% strain > 89 75 61 43 28 22 17

4% strain > 89 86 71 53 36 28 22

32
 10
9

6
Stress = 6.9 MN/m 2 = 70 kgf/cm 2 = 1000 lbf/in 2
5

63
G1
on
Flu

000
2 51
0 0-
F C1 Creep terminated
1
n
Fluo
0.9
0.8

0.7

0.6

0.5

0.4

0.3

Dotted lines show possible spread

as derived from 100 second isochronous
stress-strain relationship
0.2
6 Months
3 weeks
1 week
1 hour
Strain - %

1 day

0.1
1 10 10 2 103 104 10 5 106 10 7 10 8
Time - seconds

Figure 24. Creep in compression at 20C (68F) : Fluon G163 and Fluon FC100-25 1000

33
 10
9
8

3 2
f /in
lb
40
17
2 =
m
/c
kgf
2 2 2
12 /in
2 = l bf
5 0
m
N/ =
14
M 2
12 m
f/c
kg
2
10 2
=
2
/in
N /m lbf
M 60
10 11
2 =
1 m
f/c
0.9 kg
.6
81
0.8 2 =
m
N/ in
2
0.7 8M lbf
/
70
0.6 2 =5
cm
k gf/
1.2
0.5 2 =6
m
N/
6M
2
0.4
/in
lbf
80
2 =5
m
g f/c
0.3 8k
0.
2 =4
m
N/
4M

0.2 2

bf /in
9 0l
2 =2
f/cm
kg
.4
6 months

0
=2
1 month

2
1 week
1 hour

m
Strain - %

/
1 day

2 MN

0.1
1 10 10 2 103 104 10 5 106 10 7 10 8
Time - seconds

Figure 25. Creep in compression at 20C (68F) at various stress levels; Fluon FC100-25 1000

34
 11 1600

1500
10

1400

9 1300

1200
3%
8

1100
2%

7
1.5% 1000

900
6

1% 800

5 700

600
4

0.5% 500

3
400

2 0.2% 300

200

1
Stress - MN/m 2

6 months

Stress - ib/in 2
1 month

100
1 week
1 hour

1 day

0 0
1 10 10 2 103 104 10 5 106 10 7 10 8
Time - seconds

Figure 26. Isometric stress-time curves in compression at 20C (68F) , and various strain levels :
Fluon FC100-25 1000

35
 2.6

2.4

Stress: 10 MN/m 2 = 102 kgf/cm 2 = 1450 lbf/in 2

e
2.2 m
lu
vo
y
rb
le
2.0 fil
5%

e
m
-2

lu
00

vo
1.8 70

by
0
-3

er
0

fill
11

2%
1.6 FC

-2
me

00
olu
yv

10
1.4 rb
fille

5
-2
%

00
0
-4

1
-63

FC
1.2 1 68
FC

1.0

0.8

0.6

0.4

1 week
1 hour

1 day
Strain - %

0.2

0
1 10 102 103 104 105 106
Time - seconds

Figure 27. Isometric stress-time curves in compression at 20C (68F) , and various strain levels :
Fluon FC100-25 1000

1.0

6 MN/m 2
0.8 61.2 kgf/cm 2
870 lbf/in 2
Fractional recovered strain

0.6
2
2 :290 lbf/in )
2 MN/m
2 (20.4 kgf/cm
0.4
10 MN/m 2
0.2
2
f/in ) 102 kgf/cm 2
0 lb
cm :174
2
2 (122 kgf/ 1450 lbf/in 2
12 MN/m
0
10-5 10-4 10-3 10-2 10
-1
1 10
Reduced time

Figure 28. Recover from creep in compression : Fluon FC100-25 1000 loaded at various stress levels

36
IMPACT STRENGTH lower than at -20C (-4F) and again the lower
crystallinity material gave increased resistance to
Table 4 shows the results of Charpy impact tests on impact.
various Fluon grades, at different temperatures and
different crystallinities. All specimens were sharply There are minor differences between the various Fluon
notched (notch tip radius = 0.25mm; 0.01 inch) but even grades and, as expected, a low crystallinity gives
so none of the specimens broke completely, but gave a maximum impact strength. The conclusion is that
hinge break at + 20C (68F). At -20C (-4F) all the slow- unfilled and filled PTFE are tough in impact, even under
cooled (high crystallinity) samples broke completely, but severe test conditions e.g. with sharp notches at -20C
two of the six quenched samples (low crystallinity) gave (-4F) (see Table 4). They are better than most other
hinge breaks. In the liquid nitrogen dip test, all plastics.
specimens broke completely; strengths were always

Table 4. Charpy impact strength (notch depth 2.80 mm: 0.11 inch; notch tip radius 0.25 mm: 0.01 inch)

Median impact strength (cm kgf/cm2)

+ 20C -20C Liquid nitrogen (-180C approx)
(68F) (-4F) (-290F approx)

Slow cooled Quenched Slow cooled Quenched

Unfilled PTFE H* 6.8-9.0 8.6-9.6

Fluon FC100-15 1000 12.0 H 4.5 5.6

Fluon FC100-25 1000 H 9.4 H
Fluon FC160-60 10.9 12.8 4.1 4.5
Fluon FC150-25 7.3 7.9 3.2 3.2

Notes:
H = hinge break i.e. did not break completely.
H* = specimen almost broke completely.

37
Section 10. Electrical properties - filled PTFE
Although the primary objective in the development of for insulation. Tables 5, 6 and 7 give values of loss
Fluon grades has generally been to produce materials tangent, permittivity and volume and surface
with a range of mechanical properties, the addition of resistivities, measured at two levels of humidity, for
fillers does have a marked effect on the electrical Fluon glass-filled grades.
properties as well. In particular, graphite - and carbon
filled compounds may have a relatively high The effect of temperature on the loss tangent of some
conductivity, which can assist the dissipation of static glass-filled grades is shown in Figure 29.
charges in applications where these are a problem.
The level of dielectric loss tends to increase as glass
Filled PTFE is a mixture of materials, and voids - no content increases. However, the mechanisms giving rise
matter how small - are always present: as with all porous to dielectric behaviour of filled materials are complicated,
materials, the properties are dependent on the nature of and in general it must be assumed that these materials
the environment and of any inadvertent contamination. will exhibit Maxwell-Wagner loss processes of relatively
The electrical properties are no exception and are large amplitude at low frequencies (log10 frequency < 0)
markedly dependent on the environmental humidity. The in the dry state, moving progressively to higher
spillage of conducting liquids, electrolytes and greases frequencies with increasing moisture content. Rudner
on Fluon grades can affect the properties of the material (ref. 102) reports on, but does not interpret, the
even in otherwise dry conditions. properties of PTFE filled with titanium dioxide using
samples that had been kept at a constant low humidity
To some extent porosity (and therefore the effects of with silica gel crystals present to absorb moisture. It is
humidity and spillage) is dependent upon the method of suspected that these losses are due to a similar
fabrication, but even in a relatively non-porous part, the mechanism.
surface is liable to be absorbent.
CONDUCTING MATERIALS
INSULATING MATERIALS
No grade of filled PTFE is a good conductor when
Under dry conditions, the glass-filled grades are compared with, for example, copper or aluminium.
excellent electrical insulating materials over a wide range However, some have sufficiently low volume and surface
of temperature. Their insulating properties deteriorate resistivities to be considered for use in antistatic
with increasing humidity, but even at 95% r.h. the applications: see Table 8.
properties are comparable with those of plasticised PVC
and of some thermosetting compounds commonly used

Temperature F
50 100 150 200 250 300 350 400 450
0.030
0.028
0.026
0.024
0.022
00
70

0.020
30
0-

0.018
11
FC

0.016
0.014
0.012
0.010
0.008 000
-2 51
Loss tangent

0.006 100
FC
0.004
0.002 FC100-05
0
0 20 40 60 80 100 120 140 160 180 200 220 240
Temperature C

Figure 29. Loss tangent versus temperature at 1 kHz: glass-filled PTFE

Notes:
Form of specimen: 5 cm (2 inch) diameter disc of skived tape approx.1mm thick (0.04 inch). Electrode system: No electrodes, 3-terminal
fully shielded system. Test apparatus: General Radio Capacitance Measuring Assembly, type 1620A. Electric stress: Up to 120 V/mm (3
V/0.001 inch). Field direction: Perpendicular to the plane of the sample. Relative humidity: Dried for 150 h at 0,1 mm Hg at 116C (241F).
Accuracy of test result: Estimated to be 5%.

38
Table 5. Loss tangent at room temperature

Dry After 9 weeks at 95% r.h.

103 Hz 104 Hz 103 Hz 104 Hz 105 Hz 106 Hz

Unfilled PTFE 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001
Fluon FC100-05 0.00033 0 00032 0.14 0.122 0.058 0.0126
Fluon FC100-15 1000 0.00065 0.00063 0.32 0.36 0.275 0.068
Fluon FC100-25 1000 0 00107 0 00097 0.39 0.28 0.225 0.062

Dried for 150h at 0.1mm Hg at 116C (241F).

Note: Samples were skived from tape 1.0mm thick veneered (0.04 inch) from blocks preformed at 700 kgf/cm2; 10 000 Ibf/in2 freely sintered for 11/2
hours at 380C (716F) and cooled at approximately 30C per hour (54F per hour).

Table 6. Permittivity at 105 - 107 Hz

Ambient humidity 95% r.h.

Unfilled PTFE 2.02-2.09 2.02-2.09

Fluon FC100-15 1000 2.2-2.5 2.2-2.5
Fluon FC100-25 1000 2.2-2.5 2.2-2.5

Depending on density

Table 7. Volume and surface resistivity

Volume Surface
resistivity(1) resistivity(2)
(ohm cm) (ohm)

Dry(3) Ambient 95%

humidity r.h.

Unfilled PTFE >1 x 1018 1017 -

Fluon FC100-15 1000 >2 x 1015 1015 1015
Fluon FC100-25 1000 >2 x 1015 1015 1015

Notes
(1) Measured on discs of skived tape approx. 50.8mm (2 inch) in diameter and 10mm (0.04 inch) thick, using evaporated
aluminium electrodes. 1 min. value: 120 V/mm (3 V/0.001 inch).
(2) 1 min. value at equilibrium with the environment.
(3) Dried for 150 h at 0.1mm Hg at 116C (241F).

Table 8. Volume and surface resistivity of Fluon grades

Filler(1) Volume Surface

by weight resistivity(2) resistivity(3)
% (ohm cm) (ohm)

FC140-33 33% 102 104

FC150-25 25% 104 107
FC168-63 63% 104 104
FC140-15 15% 106 1014
FC160-60 60% 107 109

Notes
(1 ) Filler % by weight, the remainder being PTFE.
(2) Volume resistivity measured at 23C (73F) and 50% r.h. on tape specimens 0.25 to 1.0mm (0.01 to 0.04 inch) thick clamped between disc brass
electrodes. The values given are an indication only and may vary widely with fabrication methods.
(3) Surface resistivity measured at 23C (73) and 50% r.h. on tape specimens 0.25 to 1.0mm (0.01 to 0.04 inch) thick using disc and ring brass
electrodes applied by pressure only. No backing guard electrode used. With some materials the measured surface resistivity is very high even when
the volume resistivity is low. This is probably because the veneering method used to prepare tape specimens may have tended to smear a fine PTFE
layer on the surface.

39
 Section 11. Thermal properties - filled PTFE

THERMAL EXPANSION metals and metal oxides, have high thermal conductivity,
but in general they are effectively encapsulated by PTFE
The thermal expansion of most fillers is less than that of so that the conductivity of the compounds is still
PTFE and since the expansion of the composite will be relatively low.
somewhere between that of the two constituents, filled
PTFE compounds expand less than unfilled PTFE. During Typical figures for the Fluon range are given in Table 10,
fabrication certain fillers, notably fibres and platelets, together with those of some common metals and
become preferentially oriented (as do the particles of insulating materials.
PTFE to a much lesser degree). The result is that some
compounds have a markedly different expansion in the
Table 10. Thermal conductivities
directions parallel and perpendicular to the direction of
moulding.
Material Thermal conductivity
10-4 cal/cm Btu in/ft2
Measurements on the Fluon range are shown in Table s deg C h deg F
9. As thermal expansion is virtually linear below 15C
(59F) a coefficient is quoted; but since there is a Unfilled PTFE 6 1.7
transition point at about 19C (66F) and the expansion is Fluon FC100-15 1000 8 2.3
far from linear above 30C (86F) changes in dimension
Fluon FC100-25 1000 9 2.6
are quoted as a percentage for a given temperature
Fluon FC140-15 11 3.2
range. Figures for some common metals are also quoted
Fluon FC150-25 13 3.8
in Table 9 (Note 2).
Fluon FC160-60 19 5.5
It is very important to note that these figures are actual Fluon FC140-33 27 7.8
linear thermal expansions with virtually all stresses Fluon FC168-63 30 8.7
removed from the material. Changes in the dimensions Aluminium 4950 1400
of a specified part due to changes in temperature will Brass 2300 650
differ from these figures if stresses are present. Iron (cast) 1520 430
Steel (med. carbon) 1100 310
THERMAL CONDUCTIVITY Glass 18.4 5.2
Still air 0.64 0.18
The thermal conductivity of PTFE is very low, making it a Granulated cork 1.10 0.31
good thermal insulating material. Many fillers, notably Kapok 0.85 0.24

Table 9. Linear thermal expansion: Fluon range (in directions parallel and perpendicular to direction of moulding)

Units
FC100-15 1000 FC100-25 1000 FC150-25 FC160-60 FC140-15 FC168-63 Unfilled PTFE
Coefficient of Para Perp Para Perp Para Perp Para Perp Para Perp Para Perp Para Perp
expansion from
-60 to +15C x 10-5
(-76 to +59F) per C 9.2 6.6 9.2 4.4 6.9 6.4 6.6 6.4 8.6 6.4 5.4 5.2 9.9 9.6

Change from:
15-30C (5-86F) % 0.4 0.3 0.4 0.3 0.4 0.2 0.4 0.2 0.3 0.2 0.3 0.1 0.4 0.4

30-50C (86-122F) % 0.3 0.2 0.3 0.1 0.2 0.2 0.2 0.1 0.2 0.2 0.2 0.2 0.3 0.3

30-100C (86-212F) % 0.9 0.6 0.8 0.4 0.6 0.5 0.6 0.5 0.7 0.6 0.5 0.5 0.8 0.8

30-150C (86-302F) % 1.7 1.0 1.5 0.7 1.2 1.0 1.1 0.9 1.4 1.0 0.9 0.8 1.5 1.5

30-200C (86-392F) % 2.4 1.5 2.2 1.0 1.9 1.5 1.8 1.5 2.1 1.6 1.4 1.3 2.4 2.3

30-250C (86-482F) % 3.5 2.2 3.2 1.4 2.7 2.4 2.5 2.2 3.2 2.3 2.1 2.0 3.4 3.6

Notes:
1. All measurements made on samples taken from discs 7.30 cm dia x 1.27 cm thick (2 7/8 inch dia x 1/2 inch), preformed at 700 kgf/cm2 10 000 Ibf/in2 and
sintered at 380C (716F).
2. Coefficients of expansion (x 10-5 per C) of some other materials are: Aluminium 2.4
Brass 1.9
Glass 0.9
Iron (cast) 1.1
Steel (med, carbon) 1.2

40
Section 12. Surface properties - filled PTFE

It is not considered meaningful to tabulate results for when mating areas are large, friction is primarily due to
design properties in one large table and so each property adhesion.
is discussed separately and relevant values included in
their correct context. All values quoted are actual results, Orientation
and if applied to design calculation a relevant factor of It has been shown (see page 27) that the coefficient of
safety should be applied. friction can be affected by up to 30% depending upon the
orientation of the PTFE molecules.
FRICTION
Environment
Most studies on the friction of PTFE have been carried Steijn (ref.105) showed that prolonged and continuous
out with unfilled PTFE. Whilst the exact mechanisms running under dry nitrogen (5-10 parts per million of
involved are still not fully understood (ref.103) a picture water) gave rise to intermittently high coefficients of
emerges in which the dry coefficient of friction is friction but this was alleviated as soon as normally moist
dependent upon the pressure, the speed, the air (50% r.h.) was admitted. The short term tests at
temperature, the mating surface, the orientation of the temperatures from -1 to + 60C (30 to 140F) in helium,
PTFE, the environment and the time of running. oxygen, nitrogen and air showed no such effect and
Coefficients from 0.016 to 0.36 have been quoted and neither did tests in air at room temperature with relative
while this work is discussed in detail on page 27 it may humidities in the range 12 to 54%. The friction of PTFE in
be summarised (with some additional information) as vacuum (10-9 mm Hg) was studied by Buckley and
follows: Johnson (ref.106) who obtained coefficients of friction of
0.25 with a load of 1 kg. They also report the coefficient
Load and pressure to be constant over the speed range < 50mm-5 m/s (< 10-
The classical laws of dry friction state that the friction 1000 ft/min). This high figure could well be attributed to
force is independent of the apparent area of contact, the relatively small loads applied, but may be linked with
making the friction force proportional to load rather than Steijns observations regarding very dry atmospheres.
pressure. Many investigators (eg, refs. 66-70) quote the
coefficient of friction of PTFE as a function of load and Several investigators (refs.107, 108) have shown that the
show it to rise steeply at very light loads (below 5 Ib) and coefficient of friction is decreased dramatically by the
decrease with increasing load. R B Lewis (ref. 104) does addition of lubricants. This is not surprising since, if a full
not support this, but suggests the coefficient of friction film of oil is present, the friction is virtually independent
is proportional to the applied pressure P (Ibf/in2) of the mating surfaces.
according to the formulae:
Time of running
= CP-0,2 where C = 0.12 0.03 at velocities below The work of Steijn (ref. 105) shows that the coefficient of
10mm/s (2 ft/min) friction for PTFE on PTFE is influenced by the number of
C = 0.35 0.10 at 50mm/s (10 ft/min) traverses, the time lapse between runs, the nature
C = 0.45 0.15 at 500mm/s (100 ft/min) (especially velocity) of the preceding sliding and the
and above thermal history of the sliding components. Mitchell and
Pratt (ref. 109) demonstrated a similar increase in friction
Speed with time for PTFE on steel, up to a steady level (from
The coefficient of friction falls markedly at low speeds 0.05 to 0.20 in 4 hours), and showed this to be due to a
(below 50mm/s; 10 ft/min) and increases with increasing change in the surface of the PTFE rather than a change in
speed. the surface of the steel (i.e. the transfer of PTFE to the
steel).
Temperature
The coefficient of friction appears to be stable over the Filler type and volume
range -45 to 100C (-49 to 212F) but to rise at lower Thompson et al. (ref. 70) suggest that when using
temperatures and fall at higher temperatures (see page molybdenum disulphide (MoS2), asbestos, carbon,
25). graphite, and copper as fillers, as the volume of filler
increases the coefficient of friction increases from 0.016
Mating surface (no filler) to ~ 0.030 (30% of filler), but that there is little
Work by Steijn (ref.105) showed that sliding of PTFE difference in this effect between the various fillers.
against steel gave lower coefficients of friction than
sliding bulk PTFE against bulk PTFE. He suggests that

41
 For a similar range of fillers Milz and Sargent (ref.108) quoted show only a marginal decrease. Tests carried out
showed the coefficient to be independent both of the on behalf of AG Fluoropolymers have not confirmed this
type of filler and its volume addition. In particular, MoS2 and Buckley et al. (ref.106) found no improvement when
and graphite showed no advantage over glass fibre, working under vacuum. Similarly, practical tests (ref. 113)
asbestos and copper. Their results for all types ranged showed no advantages for adding MoS2 to glass
from 0.09 to 0.22 depending on velocity, load, etc. They although this combination was suggested (ref.115) as a
concluded that the filler was effectively encapsulated and possible means of reducing the scoring of shafts, and
the friction was that of PTFE only. (ref.111) for use in very dry gases. It is conceivable that
after prolonged continuous running under dry
ORourke (ref.110) originally came to the same conditions, the MoS2 is not subject to the rise in friction
conclusion but later (ref.111) states that friction is reported for PTFE.
dependent more upon the volume than the type of filler,
although cadmium oxide is claimed to be an exception. There is therefore conflicting evidence as to the effect of
At the very low temperatures of liquid oxygen and filler type and volume upon the coefficient of friction of PTFE.
nitrogen (ref.112) and under conditions of high vacuum
(ref.106) there is considerable variation in the coefficient, Filler particle size and shape
but this does not appear to be correlated with either filler It is difficult to separate the effects of particle size and
type or volume. In practical tests with the Wankel engine shape from those of filler type, since specific forms of
(ref.113) using various grades of PTFE as a seal, the particle tend to be used with specific types of filler (e.g.
coefficient was again found to be independent of the glass fibre, irregular particles of graphite and MoS2,
filler, whilst in a laboratory test, Ganz and Parkhomenko spherical-bronze, etc.). Moreover, in much of the
(ref.107) state that the type of filler is important; however, published work no details of filler particle are given.
they appear to quote the filler content as % by weight so
that filler type and filler volume are not separable. They The most explicit work in this field is that of
again found MoS2 and graphite fillers to give high Speerschneider and Li (ref.116) where, with the very hard
coefficients of 0.26 to 0.34. particles of alumina (Al2O3), they found spherical
particles gave coefficients of friction similar to that of
The evidence of Mitchell and Pratt (ref.109) is that filler unfilled PTFE (0.05-0.08) whereas irregular particles gave
type has a greater effect than filler volume, with MoS2 significantly higher results (0.14-0.15). They attributed
giving a lower coefficient than unfilled PTFE. They found this increase to cleavage of the irregular Al2O3 which
bronze had little effect and kieselguhr increased it by saturates the surface until the coefficient is that of Al2O3
25%. Work done on behalf of AG Fluoropolymers (see on steel. The abrasive nature of the filler also gives a
Table 11) suggests that volume of filler is not directly rough surface finish to the steel, thereby giving a
related to friction coefficient but fillers in general raise coefficient approximately double that of a smooth steel
the coefficient under these particular test conditions by a surface. This effect is less likely to occur with softer
factor of about two. fillers, and this has been found true with bronze, where
no difference in friction has been found between
It has also been suggested (refs.110,114) that the addition spherical and irregular particles, although Thompson et
of MoS2 and carbon to glass fibre compounds reduces al. (ref. 70) suggest that particle size can have an effect in
the coefficient of friction, although figures extreme cases.

Table 11. Dynamic coefficient of friction for various Fluon grades

Grade FC168-63 FC100-15 1000 FC100-25 1000 FC160-60 FC150-25 FC140-15 FC101-20

Coefficient 0.20 0.10 0.11 0.16 0.17 0.22 0.19

of friction

Test conditions
Mating surface: 420 S 37 steel in T condition (BS 970:Part 4:1970)
Surface finish: 0.3m R (CLA) BS 1134:1972
Pressure: 20 kgf/cm2 (300 Ibf/in2)
Speed: 0.25m/s (50 ft/min)

42
Other parameters (load, speed etc.) whilst there is wide agreement that too rough a mating
The statements made in the first part of this section for surface will cause rapid wear (refs. 111, 117, 119), one
the effects of load, speed, temperature, etc. upon the school of thought suggests that too smooth a surface
coefficient of friction of unfilled PTFE in general hold finish leads to high wear rates, (ref. 111 ) while others
good for filled PTFE, although ORourke (ref.111) shows suggest that is not so (refs. 117, 119). The answer may be
that the coefficient of glass-filled PTFE does not rise at that although too fine a finish may well inhibit good
low loads, whilst other investigations (refs.108, 112) transfer, many filled compounds are sufficiently abrasive
suggest that it does. to roughen the mating surface adequately. However, if
the filler or environmental conditions are too abrasive,
Environments rapid wear will occur through ploughing. The
Work with filled PTFE at low temperature and in contact entrapment of wear debris can have a similar affect,
with liquid oxygen and nitrogen (ref.112) shows the (ref. 109).
coefficient to rise with the passage of time (e.g. 0.18 to
0.43 in 23 hours), which tends to confirm the work of It has been suggested that chemical reactions at the
Steijn with unfilled PTFE (ref.105). interface may be important. Buckley and Johnson (ref.
106) consider that wear is related to the decomposition
This same effect at room temperature has been found by mechanism and hence to the temperature at the
work done on behalf of AG Fluoropolymers and by interface, while Hargreaves and Tantam (ref. 112)
Mitchell and Pratt (ref.109) although actual coefficients suggest lead oxide can be an oxygen carrier to other
are lower (0.07 to 0.20 in 20 hours). There is some metals, giving selective oxidation of roughnesses on the
evidence therefore that the coefficient of friction mating surface. Mitchell and Pratt (ref. 109) have noted
increases in the presence of liquid oxygen or nitrogen. the formation of copper fluoride at the interface of
High coefficients (0.2 to 0.4) were also found by Buckley bronze-filled PTFE, presumably caused by local
et al. (ref.106) for filled PTFE under high vacuum, but degradation of the PTFE and bronze. They do not,
some of the fillers, notably copper, silver and powdered however, attribute the reduction in wear accompanying
coke gave coefficients lower than for unfilled PTFE under the formation of copper fluoride to the chemical action,
the same conditions. The reasons for these effects are but rather to the fact that the area of contact at the
not known: the effects may be due to temperature or interface increases with time, which reduces the
environment, or the mechanisms may be similar to that interface temperature. Vinogradov (ref. 120) did however
experienced with graphite where the low coefficient of attribute a reduction in friction between copper and PTFE
friction is attributed to the presence of adsorbed gases at to the formation of the solid lubricant copper fluoride.
the crystallite interfaces where cleavage occurs (ref. 103).
Load and speed (Pressure x Velocity)
WEAR The most widely quoted formula for the wear of filled
PTFE is that of Archard and Hirst (refs. 121, 122) which
Wear mechanisms states that the volume wear (W) is proportional to the
The mechanisms responsible for the wear of PTFE are relative speed at the interface (V), the load supported
not fully understood, but it is generally thought that (M) and time run (T), ie,
adhesion and the freeing of transferred wear fragments,
either in terms of surface energy or by virtue of fatigue, W MVT
are of major importance (refs. 103, 104, 109, 117). It is
known that when PTFE is rubbed against other materials or using the wear factor K
a transfer takes place (refs. 116, 118) and it is believed
that the wear process involves the laying down and W= KMVT (Equation 1)
subsequent removal of such transferred layers. An ideal
situation is given (ref. 117) as having a highly oriented By dividing by an area (A), a linear wear (R) is obtained
mono-molecular layer of PTFE bonded to the metal
surface which then rubs against as smooth a mating W M
surface of PTFE as possible. such that R = = KVT;
A A
M
What is not clear is exactly how and why fillers and but is the pressure applied (P)
A
conditions affect both the initial laying down and .
. . R = KTPV (Equation 2)
subsequent removal of the PTFE particles. It is suggested
that a minimum temperature at the interface is required
Equation 2 suggests that the wear rates of materials can
to promote adequate bonding and that certain fillers
be classified in at least three ways:
function by causing frictional heat (ref. 117). It is also
clear that surface finish will affect this transfer, and

43
 (1) By quoting the maximum PV value the material can For example, although the difference in running
withstand (termed limiting PV value). This is found conditions between a thrust washer and a piston ring is
experimentally either by determining the maximum load mainly considered to be one of interface temperature, it
that can be applied at constant velocity while still is also true that wear debris is far less likely to become
maintaining temperature and/or frictional torque entrapped in the piston ring. It is also true that
equilibrium, or by determining the PV value at which the differences in wear rate can be attributed to differences
wear rate suddenly increases. in behaviour (abrasive or otherwise) when trapped wear
debris is present. Nevertheless, whatever the
(2) By quoting the PV value which gives a specified wear mechanisms, it is generally accepted (refs.104, 106, 113,
rate. This is generally quoted as the PV value to give 117) that an increase in interface temperature increases
0.127mm (0.005 inch) wear in 1000 hours, and is the wear rate.
determined from specific wear tests.
Mating surface and material
(3) By determining the constant K in Equation 2. This is At room temperatures and above it is generally agreed
known variously as the wear factor or K-value. Work that a hard, approximately 900 VPN (Vickers Pyramid
has been carried out to determine these factors for Number), mating surface is beneficial. Softer materials
various filled PTFE materials (refs. 109, 111). However, can be used providing the filler will not abrade them. The
during these and subsequent investigations (ref. 104) it materials with good dry bearing properties of their own
has been shown that there are two major errors in (e.g. bronze) are preferred to the softer more easily
accepting results derived from Equation 2: damaged materials (e.g. aluminium). There is some
(a) K is not necessarily a constant for a given material, divergence of opinion as to the suitability of chromium
but will vary with the load applied, the velocity, the plating. Pratt (ref.117) shows chromium plating to be
length of time run, the temperature and other factors advantageous whereas ORourke et al. (refs.111,124)
such as clearance and conditions at the interface. show it to give poor results. The answer might well be
(b) The method of determining the factors is that the fillers used by Pratt were less abrasive than those
invariably to test specimens. The conditions of test used by ORourke
have a considerable effect upon the results and
universal values cannot be obtained from one series The surface finish of a material is generally quoted as a
of tests. mean centre line average - CLA - or Ra (BS1134-2: 1990) of
the peaks and valleys of its surface as detected by
Temperature traversing a diamond stylus across it. This does not fully
R B Lewis (ref. 104) suggests that each material has two specify a surface however, since a turned and a ground
wear rates (metric), mild wear (K ~_ 7) and severe wear (K > surface of the same Ra value will be different. It is now
35) which is attributed to a rise in temperature at the generally accepted that a ground surface is superior to a
interface. The actual temperature at which the transition turned surface and that above 0.75m the wear rate of the
occurs is reported to depend upon the load. He concludes filled PTFE will increase. The existence of a lower limit is
that the PV value at which transition begins depends upon still in dispute and so the best compromise is to use a
the application geometry, ambient temperature, and ground surface finish of 0.2-0.4m.
manner and amount of cooling, whilst the slope of the
transition depends upon the application parameters and Lubrication
properties of the compositions. The mild wear is reported Lubricant is a very general term and it used to be stated
to be characterised by wear of the surface layers whilst that any liquid will act as a lubricant and be beneficial to
severe wear is characterised by bulk removal of material. PTFE. To some extent this is true in that, if hydrodynamic
Similar conclusions were drawn by Summers. conditions are established, no wear will take place, but
filled PTFE may run under conditions of boundary
Smith (ref.123) who considered the composite to be hard lubrication. Hydrocarbon oils are generally advantageous,
granules in a cement of softer materials suggests that with a significant reduction in wear rates. This is not so
the mild wear region corresponds to a gradual attrition of with water. ORourke (ref.111) confirmed that the wear
the hard granules whilst the change to severe wear occurs factor increased for unfilled and various filled PTFE
when the cement becomes softened by heat and the compounds when running against steel with water
granules are plucked bodily out of the matrix. From work boundary lubrication.
carried out by Mitchell and Pratt (ref.109) and work done
on behalf of AG Fluoropolymers, it is concluded that the Work done on behalf of AG Fluoropolymers has shown
entrapment of wear debris as well as surface temperature that boundary lubrication with water gave a reduction of
is a very important factor in determining whether severe wear life of 50% when filled PTFE ran against steel.
wear occurs or not.

44
Section 13. Other physico-chemical properties
- filled PTFE

PERMEABILITY CHEMICAL RESISTANCE

The permeability of filled PTFE is very dependent upon PTFE is one of the most chemically resistant materials
methods of fabrication. Since there is no chemical known, and fillers can only worsen this position. Table 12
bonding between PTFE and the fillers used, the gives a list of the probable suitability of the Fluon range
permeability is generally greater than that of unfilled for use with various common chemicals. As noted below
PTFE. As the filler content increases, particularly above this table, some of the information was obtained from
10% by volume, special care, or special techniques, have direct experimentation whilst other data were predicted
to be adopted to ensure minimum permeability. Specific from a knowledge of the chemistry of the individual
measurements have not been made, but from porosity constituents. As with unfilled PTFE, chemical attack on
tests it is concluded that permeabilities can be made the filled PTFE should not be confused with chemical
close to that of unfilled PTFE if the correct method of attack on adjacent parts resulting from permeability of
fabrication is used. the filled PTFE.

Table 12. Suitability of Fluon range for use with various chemicals at 50C (122F)

NB. This table is intended solely as a general guide to material selection. Because exposure conditions can vary considerably in different
applications, the user is strongly advised to undertake tests under conditions relevant to the specific application.

Fluon Grade
Chemical FC168-63 FC100-15 1000 FC150-25 FC160-60
FC100-25 1000 FC140- 15

50% H2SO4 U S* PS U
Conc. HCl U PU* PS U
Conc. HNO3 U PU* PS U
40% NaOH PS PU* PS PS
0.880 NH4OH U S* PS U
Benzene PS S* PS PS
Phenol PS S* PS PS
Trichloroethylene PS S* PS PS
Ethanol PS S* PS PS
Fluorine - PU PS -
Chlorine PS(dry) PS PS PS (dry)
Bromine - PS PS -
HF PS U PS PS
SO2 - PS PS -
Mercury U PS PS U

Notes
S= Satisfactory
U = Unsatisfactory
PS = Not tested, but probably satisfactory
PU = Possibly unsatisfactory
FC150-25 and FC140-15 will be attacked only by oxidation, although there may be a slight reduction in weight when in contact with other
chemicals, due to attack on the very small quantities of impurities (less than 2%) inherently present in the filler.
* The results of tests in which samples of tape 50.8mm (2 inches) square and 0.38mm (0.015 inch) thick were used in order to give the maximum
surface area to weight ratio. The samples were dried, weighed and then immersed in the various liquids at 50C (122F). After a period of four weeks
the samples were removed from the liquids, washed, dried and re-weighed to determine whether attack had occurred. Materials listed as
satisfactory showed no change in weight over the four week period and those showing some change are listed as possibly unsatisfactory. Over the
four week period the average change in weight of the possibly unsatisfactory materials was 7%. The tensile properties were measured before and
after the chemical resistance tests: no statistically significant changes in tensile properties were recorded in any of the tests, including those in which
weight losses occurred.
Up to 70C (160F).

45
 Section 14. Typical properties of Fluon unfilled and filled
moulded PTFE

Property Units Unfilled FC100-25 1000 FC160-60

Relative density - 2.1-2.2 2.25 3.8

Filler by weight and type - - 25% glass fibre 60% bronze

Maximum tensile strength MN/m2 20.6-34.3 12.3-19.6 10.3-13.7

kgf /cm2 210-350 125-200 105-140
Ibf/in2 3000-5000 1800-2800 1500-2000

Elongation at break % 250-400 200-300 80-160

Compressive modulus for MN/m2 392 588 736

1% deformation (20C,1 day) kgf/cm2 4000 6000 7500
1 bf /in2 55,000 85,000 105,000

Impact strength at -20C cm kg/cm2 8 9 11

(Charpy, notched) ft Ib/in2 4 45 55

Hardness Shore D 60-65 70-75 70-75

Volume resistivity ohm cm > 1018 1015 107

Surface resistivity ohms 1017 1015 109

Permittivity at 23C, lO5-107 Hz:

dry - 2.05 2.35 Conducting filler
95% r.h. - 2.05 2.35 Conducting filler

Loss tangent at 23C and 103 -104 Hz:

dry - 0.0001 0.001 Conducting filler
95% r.h. - 0.0001 0.3-0.4 Conducting filler

Linear thermal expansion 20-150C:

parallel to direction of moulding % 1.9 1.8 1.5
perpendicular to direction of % 1.8 1.0 1.1
moulding

Thermal conductivity 10-4 cal/cm s C 6 9 19

Btu in/ft2 h F 1.7 2.6 5.5

Continuous service temperature C -250 to +260 -250 to +260 -250 to +260

F -420 to +500 -420 to +500 -420 to +500

Coefficient of friction (dynamic) - 0.06 0.12 0.13

Resistance to weathering - Excellent Excellent Excellent

Chemical resistance - Fluon is chemically inert and unaffected by all

known chemicals except alkali metals, fluorine under
certain conditions, and some fluorine compounds at
elevated temperatures. Filled Fluon has inferior chemical
resistance because of the presence of fillers.

The data presented are based on experiments which are believed to be accurate and reliable. Additional data are available on request.
The Fluon range contains compositions of PTFE with fillers such as glass-fibre, bronze, powdered coke and graphite.

46
Section 15. Specifications relating to PTFE

Fluoropolymer raw materials can be designated and

specified using ISO 12086-1 and -2:1995.

Virgin PTFE semi-finished products can be designated

and specified using ISO 13000-1 and -2:1997.

In addition there are numerous national specifications

covering raw materials, semi-finished and finished
articles which change continually.

Please contact Asahi Glass Fluoropolymers UK Ltd for

details.

47
 Section 16. Handling precautions

Within its working temperature range PTFE is a

completely inert material, but when heated to its
sintering temperature it gives rise to gaseous
decomposition products or fumes which can produce
unpleasant effects if inhaled. Fumes can be produced
during processing: for example, when the material is
heated to sinter it, or when brazed connections are being
made to cable insulated with PTFE. The inhalation of
these fumes is easily prevented by applying local exhaust
ventilation to atmosphere as near to their source as
possible.

Smoking should be prohibited in workshops where PTFE

is handled because tobacco contaminated with PTFE will
during burning give rise to polymer fumes. It is therefore
important to avoid contamination of clothing, especially
the pockets, with PTFE and to maintain a reasonable
standard of personal cleanliness by washing hands and
removing any PTFE particles lodged under the
fingernails.

48
Section 17. Further information

The following is a comprehensive list of Technical Information contained in this publication (and otherwise
Service literature for Fluon PTFE available from the supplied to users) is based on our general experience
AG Fluoropolymers sales office. and is given in good faith, but we are unable to accept
responsibility in respect of factors which are outside our
F1 The Moulding of PTFE granular powders knowledge or control. All conditions, warranties and
liabilities of any kind relating to such information,
F2 The Extrusion of PTFE granular powders
expressed or implied, whether arising under statute, tort
F3/4/5 The Processing of PTFE coagulated or otherwise are excluded to the fullest extent
dispersion powders permissible in law. The user is reminded that his legal
responsibility may extend beyond compliance with the
F6 Impregnation with PTFE aqueous information provided. Freedom under patents, copyright
dispersions and registered designs cannot be assumed.
F8 Processing of filled PTFE powders
Fluon grades are general industrial grades. It is the
F9 Finishing processes for responsibility of the purchaser to check that the
polytetrafluoroethylene specification is appropriate for any individual application.
F11 Colouring of polytetrafluoroethylene Particular care is required for special applications such as
pharmaceutical, medical devices or food. Not all grades
F12/13 Physical properties of unfilled and filled are suitable for making finished materials and articles for
polytetrafluoroethylene use in contact when foodstuffs. It is advisable to contact
F14 Isostatic compaction of PTFE powders the AG Fluoropolymers sales office for the latest position.
Users of Fluon are advised to consult the relevant
F15 Cast Film from Fluon PTFE dispersion GP1 Health and Safety literature which is available from the
AG Fluoropolymers sales office.
FTI500 Fluon - A Guide to Applications, Properties
Users of any other materials mentioned in this
& Processing
publication are advised to obtain Health and Safety
information from the suppliers.
FTI800 Potential Material & Equipment Suppliers

This edition AGFP September 2002

49
 Section 18. References

1 Free energies of formation of fluorocarbons and 12 Transitions and melting of polytetrafluoroethylene

their radicals. Thermodynamics of formation and (Teflon) under pressure, C W F T Pistorius, Polymer,
depolymerisation of polytetrafluoroethylene, W M D 5, 1964, 315-317
Bryant, J.Poly. Sci., 56, 1962, 277-296
13 The pressures of some solid-solid transitions, G C
2 Structures of molecules and crystals of Kennedy and P N La Mori, J. Geophys. Res., 67, 1962,
fluorocarbons, C W Bunn and E R Howells, Nature, 851-856, Chem. Abstr., 57, 1962, 4119 a
174, 4429, 18th September, 1954, 549-551
14 Behaviour of polytetrafluoroethylene (Teflon) under
3 Crystal structure of polytetrafluoroethylene, R H H high pressures, R I Beecroft and C A Swenson, J.
Pierce, E S Clark, J F Whitney and W M D Bryant, App. Physics, 30, 11, November 1959, 1793-1798
Abstracts of 130th Amer. Chem. Soc. meeting,
Atlantic City, September, 1956, 9S 15 Thermal expansion of Ftoroplast IV (fluoroplast,
Teflon) between 190 and 325, l E Leksina and S I
4 Unusual features in the crystal structure of Novikova, Soviet Phys. Solid State, 1, 1959, 453-459,
polytetrafluoroethylene, E S Clark and L T Muus, Chem. Abstr., 54, 1960, 16914 f
Abstracts of 132nd Amer. Chem. Soc. meeting, New
York, September, 1957, 5T 16 Calorimetric properties of polytetrafluoroethylene
(Teflon) from 0 to 365K, G T Furukawa, R E
5 Partial disordering and crystal transitions in McCoskey and G J King, J. Res. Nat. Bur. Stand., 49,
polytetrafluoroethylene, E S Clark and L T Muus, 4, October 1952, 273-278
Zeit. Krist.,117, 1962, 119-127
17 Specific heat of synthetic high polymer, Part 5: A
6 Volume-temperature relationships for the room study of the order-disorder transition In
temperature transition in Teflon*, F A Quinn, D E polytetrafluoroethylene, P Marx and M Dole, J.
Roberts and R N Work, J. App. Phys., 22, 8, August Amer. Chem. Soc., 77, 1955, 4771 -4774
1951, 1085-1086
18 Effect of pressure on the specific volume of polymer
7 A room-temperature transition in melts, J M Lupton, Abstracts of 134th Amer. Chem.
polytetrafluoroethylene, H A Rigby and C W Bunn, Soc. meeting, Chicago, September, 1958, 37T-38T
Nature, 164, 4170, 1st October 1949, 583
19 R P Palmer, lCI Plastics Division, unpublished work
8 Fluorine-containing polymers, Part 2:
Polytetrafluoroethylene, C A Sperati and H W 20 The melting points of chain polymers, C W Bunn, J.
Starkweather, Fortschr. Hochpolym. - Forsch., 2, Poly. Sci., 16, 1955, 323-343
1961, 465-495
21 E R Howells, ICI Plastics Division, unpublished work
9 Chain configurations in crystals of simple linear
polymers, C W Bunn and D R Holmes, Disc. Faraday 22 An independent measurement of the amorphous
Soc., 25, 1958, 95-103 content of polymers, R G J Miller and H A Willis, J.
Poly. Sci., 19, 1956, 485-494
10 Transitions and phases of polytetrafluoroethylene
(Teflon). C E Weir, J. Res. Nat. Bur. Stand., 50, 2, 23 The molecular structure of perfluorocarbon
February 1953, 95-97 polymers. Infra-red studies on
polytetrafluoroethylene, R E Moynihan, J. Amer.
11 Effect of pressure on the room-temperature Chem. Soc., 81, 1959, 1045-1050
transition of polytetrafluoroethylene and its heat of
transition, T Yasuda and Y Araki, J App. Poly. Sci.,
5,15,1961, 331-336

*Trade mark of E I du Pont de Nemours and Company (Inc)

50
24 Physical properties of fluorocarbon plastics, Part 2: 306
Relations between crystallinity and room
temperature transition effects in PTFE, Toshihiko 38 Polytetrafluoroethylene,W E Hanford and R M Joyce,
Kuroda and Hiroshi Sakami - Nagoya Kogyo Gijutsu J. Amer. Chem. Soc., 68, 1946, 2082-2085
Shikensho Hokoku 7,1-8 (1958), Chem. Abstr., 57,
4828h 39 Dielectric properties of Teflon from room
temperature to 314C and from frequencies of 102 to
25 Structure of polytetrafluoroethylene, G Butenuth, 105 c/s, P Ehrlich, J. Res. Nat. Bur. Stand., 51, 4,
Verhandlungsber. Kolloid - Ges., 18, 1958,168-179, October 1953, 185-188
Chem. Abstr., 52, 14213 i
40 Dielectric losses of polytetrafluoroethylene, G P
26 The fine structure of polytetrafluoroethylene, C W Mikhailov, S P Kabin and A L Smolyanskii, Zhur. Tekh.
Bunn A J Cobbold and R P Palmer, J. Poly. Sci., 28, Fiz., 25, 1955, 2179-2182, Chem. Abstr., 50, 1956,
1958, 365-376 3026f

27 Some observations on the structure of 41 Vorbehandlung and dielektrisches Verhalten

polytetrafluoroethylene, C J Speerschneider and C H Hochpolymere, F Krum and F H Muller, Koll. Zeits.
Li, J. App. Physics, 33, 5, May 1962, 1871-1875 164, 1959, 81-107

28 A correlation of mechanical properties and 42 Transitions and relaxations in

microstructure of polytetrafluoroethylene at various polytetrafluoroethylene, R K Eby and K M Sinnott, J.
temperatures, C J Speerschneider and C H Li, J. App. App. Physics, 32, 9, September 1961, 1765-1771
Physics, 34,10, October 1963, 3004-3007
43 Hartshorn and Ward, J. Inst. Elec. Engrs, 1936, 79, 597
29 Dispersions de polytetrafluorethylne, E Grimaud, J.
Sanlaville and M Troussier, J. Poly. Sci., 31, 1958, 44 Barrie, Proc. Inst. Elec. Engrs, 112, 2, February 1965
525-527
45 Parry, Proc. Inst. Elec. Engrs, 98, Part 3, 54, July 1951,
30 A J Cobbold and R P Palmer, ICI Plastics Division, 303
unpublished work
46 Bleaney, Loubser and Penrose, Proc. Phys. Soc., 59,
31 Polymer single crystals, P H Geil, Interscience Part 2, March 1947, 185
Publishers, New York, 1963, 483
47 Reddish, l U P A C Symposium, Montreal, August
32 Treatise on powder metallurgy, 2, S. Goetzel 1961

33 W G Rodway, ICI Plastics Division, unpublished work 48 Reddish, Society of Chemical Industry Symposium,
April 1958, published proceedings
34 Effect of pressure on the melting point of Teflon
polytetrafluoroethylene resin. P L McGeer and H C 49 Reddish and Barrie, l U P A C Symposium,
Duus, J. Chem. Phys., 20, 1952, 1813-1814 Wiesbaden, 1958

35 Uniaxial compressive creep of 50 D J Parr and R M Scarisbrick, Proc. Inst. Elec. Engrs,
polytetrafluoroethylene, D A Thomas, Polymer 112, August 1965, 1625
Engineering and Science, 9, 1969, 415-422
51 Effects of fabrication on the properties of Teflon
36 Experimental technique in uniaxial creep testing, D A resins, P E Thomas. J F Lontz, C A Sperati and J L
Thomas and S Turner, Interscience Publishers Inc, McPherson, S P E Journal, 12, June 1956, 89-96
New York, Testing of polymers, Vol 4
52 G Mole, E R A reports V/T115,1952 and V/T149, 1962
37 S Turner, Polymer Engineering and Science, 6, 1966, 52a Superheating of linear high polymers: PTFE, E

51
 Helmuth, B Wunderlich and J H Rankin, Appl.
Polymer Symposium No 2,1966, 101-109 64 Pyrolysis of polytetrafluoroethylene, J C Siegle and L
T Muus, Abstracts of 130th Meeting of Amer. Chem.
53 Thermal expansion of polytetrafluoroethylene Soc., Atlantic City, 1956, 8S
(Teflon) from -190C to +300C, R K Kirby, J. Res.
Nat. Bur. Stand., 57, 2, 1956, 91-94 65 Oxidative thermal degradation of PTFE, K L Paciorek,
R H Kratzer: J Kaufman, J. Poly. Sci., 1973, 1465-1473
54 Thermal conductivity studies of polymers, D E Kline,
J. Poly. Sci., 50, 1961, 441-450 66 The frictional properties of some plastics, K V
Shooter and P H Thomas, Research, 2, 1949, 533-535
55 Thermal conductivity of high polymers from -180 to
+90C K.Eiermann and K H Hellwege J Poly Sci., 57, 67 The friction and deformation of polymers, M W
1962, 99-106 Pascoe and D Tabor, Proc. Roy. Soc., 235A, 1956, 210-
224
56 Specific heat of synthetic high polymers, Part 5: A
study of the order-disorder transition in 68 The frictional properties of plastics, K V Shooter and
polytetrafluoroethylene, P Marx and M Dole, J. D Tabor, Proc. Phys. Soc., 65B, 1952, 661-671
Amer. Chem. Soc., 77, 1955, 4771-4774
69 Frictional properties of plastics, K V Shooter, Proc.
57 Relative enthalpy of polytetrafluoroethylene from 0 Roy. Soc., 212A, 1952, 488-491
to 440C, T B Douglas and A W Harman, J. Res. Nat.
Bur. Stand., 69A, 2, 1965, 149-157 70 The sliding friction of Teflon, J B Thompson, G C
Turrell and B W Sandt, S P E Journal, 11, 4, April
58 Thermal degradation of tetrafluoroethylene and 1955, 13-14, 38
hydrofluoroethylene polymers, S L Madorsky, V E
Hart, S Straus and V A Sedlak, J. Res Nat. Bur. 71 Friction of Teflon sliding on Teflon, D G Flom and
Stand., 51, 1953, 327-333 N T Porile, J. App. Physics, 26, 1955, 1088-1092

59 PyroIysis of polytetrafluoroethylene, E E Lewis and 72 The effect of temperature on the mechanical

M A Naylor, J. Amer. Chem. Soc., 69, 1947, l968-l970 properties and the friction of plastics, R K King and D
Tabor, Proc. Phys. Soc., 66B, 1953, 728-736
60 Thermal degradation of fluorine containing
polymers, Part 1: Degradation in vacuum, J M Cox, B 73 Visco-elastic properties and the friction of solids, K G
A Wright and W W Wright, J. App. Poly. Sci., 8, 1964, McLaren and D Tabor, Nature, 197, 4870, 1963, 856-
2935-2950 859

61 Thermal degradation of fluorine containing polymers, 74 Friction and transfer of polytetrafluoroethylene, K R

Part 2: Degradation in oxygen, J M Cox, B A Wright Makinson and D Tabor, Nature, 201, 4918, 1964, 464-
and W W Wright, J. App. Poly. Sci., 8, 1964, 2951- 466
2961
75 Effect of orientation on the friction of PTFE, D Tabor
62 The molecular structure of perfluorocarbon and D E W Williams, Wear, 4, 5, 1961, 391-400,
polymers, Part 2: Pyrolysis of Rubber Abs. 40, 258
polytetrafluoroethylene, J C Siegle, L T Muus, Tung-
Po Lin and H A Larsen, J. Poly. Sci., A, 2, 1964, 391- 76 The spreading of liquids on low energy surfaces, Part
404 1: Polytetrafluoroethylene, H W Fox and W A Zisman,
J. Colloid Science, 5, 1950, 514-531
63 H G Jellinek, Paper No 16 of the 12th Canadian High 77 Wetting of fluorinated solids by hydrogen bonding
Polymer Forum, Candy Alpine Inn, Ste. Marguerite, liquids, A H Ellison, H W Fox and W A Zisman, J
Quebec Phys. Chem., 57, 1953, 622-627

52
78 Wettability of halogenated organic solid surfaces, A 92 Viscosity and plasticity of polytetrafluoroethylene
H Ellison and W A Zisman, J. Phys. Chem., 58, 1954, resin above the melting point, A Nishioka and M
260-265 Watanabe, J. Poly. Sci., 24,106, 1957, 298-300

79 Wetting of low energy solids by aqueous solutions 93 Molecular weight of polytetrafluoroethylene, R C

of highly fluorinated acids and salts, M K Bernett and Doban, A C Knight, J H Peterson and C A Sperati,
W A Zisman, J. Phys. Chem., 63 1959, 1911-1916 Abstracts of 130th Amer. Chem. Soc. Meeting,
Atlantic City, September 1956, 9S
80 The permeability of some plastic materials to H2, He,
N2, O2 and A, R S Barton, U K A E A Report M599, 94 Reactions of irradiated polytetrafluoroethylene
1960. Chem. Abstr., 54, 14750h resin, M I Bro, E R Lovejoy and G R McKay, J. App.
Poly. Sci., 7, 1963, 2121-2133
81 H Yasuda and W Stone, J. Poly. Sci., Al, 4, 1966,
1314-1316 95 The effects of X-rays on the insulation properties of
polytetrafluoroethylene, W E Liversage, Brit. J.
82 R A Pasternak, M V Christensen and J Heller, Radiol., 25, 1952, 434-436, Chem. Abstr., 46, 1952,
Macromolecules, 3, 1970, 366-371 10679g

83 G V Casper and E J Henley, Polymer Letters, 4, 1966, 96 Chemical and physical changes in gamma-irradiated
417-421 plastics, R Harrington and R Giberson, Modern
Plastics, 36, November 1958, 199-221, 314, 317
84 Gas chromatographic measurement of the
permeability of PTFE, PVC, polyethylene and nylon 97 Polytetrafluoroethylene - a radiation resistant
tubing towards oxygen and nitrogen. R G Gerritse. polymer, L A Wall and R E Florin, J. App Poly. Sci.,
Journal of Chromatography, 77, 1973, 406-409 2, 5, 1959, 251

85 Permeation of sulphur dioxide through polymers. R 98 The effect of radiation on PTFE, Fluon Engineering
M Felder, R D Spence, and J K Ferrell, Journal of News, ICI Plastics Division, May 1964, 4
Chemical and Engineering Data 20, 3, 1975
99 A Monnet and R Bensa, Energie Nucleaire, 13, 2,
86 Structure of polymers and properties of films, P G March/April 1971, 123-132
Konovalov, Sbornik Stateu Vsesoyuz, Zaochnogo
Politekh. Inst., Moscow, 23, 1960, 92-102, Chem. 99a The effects of radiation on the mechanical and
Abstr., 55, 1961, 20488b electrical properties of the GEOS satellite long
boom cable materials, D Verdin and P R Goggin,
87 Permabilit des matires plastiques la vapeur AERE, Harwell, July 1975
deau, M Korte-Falinski, J de Chimie Physique, 59,
1962, 27-35 100 Fluorocarbon polymers in the Chemical Industry, B
B Rossa, Praktische Chemie, 15, 2, 1964, 64-73
88 P E Toren, Anal. Chem., 37, 7, 1965, 922-923
101 Temperature dependence of sound velocity in
89 Journal of Teflon, (E I du Pont de Nemours & PTFE, V M Kravtsov, Akust. Zh., 11, 3, 1965, 400
Company Inc), European edition, 8, March/April 401, Chem. Abstr., 64, 3707b
1970; US edition, December, 7, 1964
102 Filled fluorocarbons - new component materials-,
90 Measurements of the refractive index of films, F W M A Rudner, Elec. Manuf. February 1955-80
Billmeyer, J. App. Physics, 18, 5, 1947, 431-434
91 Determination of the refractive index of a solid using 103 Friction, lubrication and wear: a survey of work
a far infra-red maser, J E Chamberlain and H A during the last decade-, F P Bowden and D Tabor,
Gebbie, Nature, 206, 4984, 1965, 602 Brit. J. Appl. Phys. 17, 1966, 1521-1544

53
 104 Predicting bearing performance of filled Teflon TFE 115 A comparison of materials for use as unlubricated
resins, R B Lewis, ASME Paper 66-WA/RP-1 journal bearings, F F Simpson, Proc. l. Mech. E. 175,
No.10 1961, 528-536
105 The effect of time, temperature, and environment
on the sliding behaviour of polytetrafluoroethylene, 116 The role of filler geometrical shape in wear and
R P Steijn, ASLE Trans. 9, 1966, 149-159 friction of filled PTFE, C J Speerschneider and C H
Li, Wear, 5, 1962, 392-399
106 Friction, wear and decomposition mechanisms for
various polymer compositions in vacuum to 10-9 117 Plastics as bearing materials, with particular
millimetre of mercury, D H Buckley and R L reference to PTFE, G C Pratt, Plast. Inst. Trans. and
Johnson, NASA Tech. Note D-2073, December 1963 Journal, 32, August 1964, 255-260

107 Anti-friction properties of PTFE filled with ground 118 The friction and transfer of PTFE, K Makinson E D
coke, S N Ganz and V D Parkhomenko, Soviet Tabor, Proc. Roy. Soc. 281A, 1964, 49-61
Plastics, January 1966, 42-43
119 The wear of PTFE impregnated metal bearing
108 Frictional characteristics of plastics, W C Milz and L materials, D C Mitchell, l. Mech. E. Conference on
B Sargent, Lubrication Engng.1955 Lubrication and Wear, London, 1-3 October 1957

109 Friction, wear and physical properties of some filled 120 A study of heavy metal to plastic friction duties and
PTFE bearing materials, D C Mitchell and G Pratt, I. of the wear of hardened steel in the presence of
Mech. E. Conference on Lubrication and Wear, polymer, G V Vinogradov, V A Mustafaev and Yu Ya
London, 1-3 October 1957 Podolsky, Wear, 8, 1965, 358-373

110 Design properties of filled-TFE plastics J T 121 Contact and rubbing of flat surfaces, J F Archard, J.
ORourke, Machine Design, 13 September 1962 Appl. Phys. 24,1953, 981-988

111 Fundamentals of friction PV, and wear of 122 The wear of metals under unlubricated conditions,
fluorocarbon resins, J T ORourke, Modern Plastics, J F Archard and W Hirst, Proc. Roy. Soc. 236A, 1956,
43, September 1965, 161-169 397-410

112 Performance of some plain bearing materials under 123 Performances of unlubricated piston rings, D
boundary conditions at low temperatures, R Summers-Smith, Lubrication and Wear Convention,
Hargreaves and D H Tantam, Proc. l. Mech. E. 175, I. Mech. E.1963, 280
No. 20 1961, 941-954
124 Internal report on DECHEMA (Germany),
113 Friction and wear tests at high rubbing speeds, VDI Colloquium 17, November 1967
Zeitschrift, 109, No.16 1967, 722-727
114 Non-lubricated bearings and piston rings of
tetrafluoroethylene, R B Fehr, SPE Journal, August
1960, 943-948

54
55
UK
For fluoropolymer & AFLAS enquiries
from EMEA (Europe, Middle East & Africa):
AGC CHEMICALS EUROPE, LTD.
PO Box 4
Thornton Cleveleys
Lancashire
FY5 4QD
UK
Telephone: +44 (0) 1253 861951
Fax: +44 (0) 1253 861950
email: info@agcce.com
web: www.agcce.com
AMSTERDAM
For fluorinated chemicals &
ETFE Film enquiries:
AGC CHEMICALS EUROPE
Commercial Centre
World Trade Center
Zuidplein 80, H Tower, Level 9
1077 XV Amsterdam, The Netherlands
Telephone: +31 (0) 20 880 4170
Fax: +31 (0) 20 880 4188
email: enquiries@agcce.com
web: www.agcce.com
JAPAN
ASAHI GLASS CO. LTD.
6th Floor Shin-Yurakucho Building
1-12-1 Yurakucho
Chiyoda-ku
Tokyo 100-8405
Japan
Telephone: +81-3218-5875
Fax: +81-3218-7856
email: kazuhiko-kameda@agc.co.jp
web: www.agc.co.jp
USA
For Lumiflon, AK-225, Fine Silica,
Fluor-Intermediates, Surflon, Asahi Guard,
Flemion & Selemion enquiries from
the Americas:
AGC CHEMICALS AMERICAS INC.
Corporate Office
229 E. 22nd Street
Bayonne, NJ 07002
United States of America
Telephone: +1 704-329-7614
Fax: +1 704-357-6308
email: kjones@agcchem.com
web: www.agcchem.com

For fluoropolymer & AFLAS enquiries
from the Americas:
AGC CHEMICALS AMERICAS INC. .
Corporate Office
255 S. Bailey Road
Downington, PA 19335
United States of America
Telephone: +1 610-423-4300
Toll Free Telephone (US only) 800-424-7833
Fax: +1 610-423-4305
email: fluon@agcchem.com
web: www.agcchem.com
SINGAPORE
ASAHI GLASS SINGAPORE CHEMICALS
PTE., LTD.
460 Alexandra Road
#17-03 PSA Building
Singapore 119963
Telephone: +65 6273 5656
Fax: +65 6276 8783
email: casey@agcsin.com.sg
CHINA
AGC CHEMICALS TRADING
(SHANGHAI) CO., LTD.
Room 6405, Rui Jin Business Center
118 Rui Jin (2) Road, Shanghai
China
Postcode: 200020
Telephone: +86 21 6415 165
Fax: +86 21 6415 9506
email: acs-suzu@uninet.cn