MECHANICAL PROPERTIES OF HIGHLY FILLED ELASTOIARS VI.

Influence of filler content and temporature on* ultimate tensile properties.

'if

Reported by

C.J. Nederveen and H.R . Bre

U

S

S

CENTRAAL* LABORATORIUM

, .N.o. D D C"RECEIVED U'. AUG 7 1967

AU AUG8 1967 B

CFSTI

•Lk

BestAvailable

Copy

mcmN

WCHANICAL PROPERTIES OF HIGHLY FILLED ELAWOMERS VI,

Influence of filler content and temperature onultimate tensile properties,

Reported by ..

C.J. Nederveen and H4. Bree

132 JULIANALAAN - DELFT - TELEPHONE 01730-37000 - THE NET',ERLANOS

CENTRAL LABORATORY TNO

ONR Technical Report No. 6

Mechanical Properties of H=v Fitled Elastomers VI.

Influence of filler content and temperature on

ultimate tensile properties

by

C.J. Nederveen and H.W. Bree

Contribution from the Centrai Laboratory TNO

Delft, Netherlands.

Report No. CL 67/34

Ccntract No. 1 62558-4375

Distribution of this document

is unlimited

Delft, May 1967

CENTRAAL LiBORATCRIWU TNO - DELFT, Holland, Report No. CL 67/34 page 2

MLCHANICAL PROP-RTIa.- OF HIGHLY FILM EiSJTORS VI.

Influence of filler content and temperatu-e on ultimate tensile properties.

Reported by

C.J. Nederveen and H.U. Bree

The work reported here was carried out by:

H.W. Bree, assisted by Preparation and characterization of filled

IMiss I'LP. van Puykeren. polyurethanes.

C.J. Nederveen, assisted by Neasliremeut of tonsiJe properties of filled

J. Wesselius. rubbers.

C.A. Schirippert and Digital read-out and processing of measuring

C.dI. v.d. %'al. data.

R. Nauta, assisted by Eachining of specimens.

G.J. Kap.

B. de Goede (Central Technical Classification of filler substance.

Institute TNO).

CENTRAAL IdLBORATRCIUTI TNO - DELFl, Holl-and, Report Noý CL 67/3A page 3.

SD]MHARY.

1 I. INTRODUCTION.

2. WhiAt UL S-.li3. EUTPFUMANFTAL T' X-MITj

4. DISCUSSION OF RiES•j.ifS.

4.1. Stress-strain data.

1.. 4.2. Modulus of elasticity.

4.3. Failure envelope.

I 4.4. Dewetting.

4,5. Comparison vwith results of tensile creep measurements.

5. CONCLUSIONS.

6. REFERICES.

1. TABLD].CAPTIONS TO FIGUIRES.

FIGURES.

DISTR7TBUTION LIST

V

CENTRAAL LAB•AtTORI24 TNO - D1)FI, Hoiland. Report No. CL 67/34 page 4,

Results are presented of stress-strain tests rerformed at various temperatures

and strain rates on materials filled with various amounts of sodium chloride

-.- * particles with a mean size of 0.1 mam. Stress and straih at rupture decreased with

increasing filler content. For each material investigated it was possible to

construct the well-known failure envelope by means of which the results could be

described adequately.

The dewetting of the particles in the rubbery matrix, sometimes resulting in

a maximum in the stress-strain curve, is ascribed to failure of the rubber between

the particles and not to failure of the rubber-salt bond, because the dewetting

maximum was dependent on temperature and time in the same way as the rupture

properties of the unfilled material.

The results are compared with earlier investigations, e.g. high speed tensile

tests and creep experiments, and a good agreoment is found.

A

4

M-NTRAAL IABORAMRIUM TO - DELT#, Holland, Report No. CL 67/34 page 5.

"In the design of solid rocket motors, a thorough knowledge of the mechanical

Sb h behaviour of solid propellant grains is very important. We especially mention the

formation of cracks in the propellrnt, which laads to malfunction of the motor.

Cracks may be formed during storago; they are then due to shrinkage or theimal

stresses. Or during combustion, when they are due to combustion pressure or

acceleration forces*

The composite ýropellnnt under study consiets of a polyurethane rubber filied

with amaor.Ium perchlorate particles. PFr ease of handling# we have so far used

model substances consisting of the same rubber but filled with an inert mn.terialp

viz. sodium chloride. Vrious fractions of NaC1 were used. The influence of filler

size, shape, concentration and surface treatment on the mechanical behaviour ofthose materiais was studied rnd reported earlier (1 - 8).

The present report describes fnilure properties of the model ,xiterials filled

with s•at particles of one particulnr size (about 0.1 mn): tensile tests were

performed on dumb-bell-shaped specincns at various tUmperutures.

4

CMTRAAL I s•RAADEIUM TN - DIELT, Hollnd, Report No. CL 67/34 poge 6.

2. M

All smplos of filled and iifilled polyurethare rubbers investigated were

based on r. linetr (polyprop)lefe othor)glyool (Demophn 3600, Forben ?abriken

"Bayer, leverkusen, Germany) with a molecular weight of about 2000. The moleculee

"of this polyether were lengthened with toluene diisocyenae and croselinked by

means of trimethylol propane in the presence of z catalyst. Full details conce.

ing the preparation of these filled and unfilled rubbers wore given previaely(0 t2,7).

A series of samples waz prepared containing various nmounts (0, 15, 30 an

4% by volume) of sodium chloride particles of about 100 tAm [fraction no* 4 (2),

labelled 90-105 jm, qwith largest dimensions between 50 amd 175 jAm am uirleat

dimensions between 25 rnd 100 .ml. No surfactant was used in the preparation of

these materials. In order to provide sufficient material for this study, all jcompositions were made in sixfold,

Det,-ils of e.'ch batch are given in Table 1 *. This table contains, for each

batch, the chemical composition of the prepolymor; the content of filler, as

calculated from the weights of the ingredients used, and as calculated from the

density; the fraction number and the size of the filler; the equilibrium wvelling

ratio (W" vol. increas.) in two different solvents (chlorofor- and trichloroethyl-

oev) ; the density, d. Details of those routine control mezurements were given

exlio( 2 ) Tho last column of Tnblc I gives the temperatures at which stress-

strain toots were performod.

From Table I it is seen thýt the chomical composition of the rubbery binder

was approximately thx same for all s-mploo prepared (columns 2, 3, 4). Columns 5

A 6 show thnt the difforences botwoon the volmo oon- atration as calculated

from th- weights of the ingromionts used and that calcul.tted from the density are

inaller th.-n 1% 0 vol.'.

As there exists n distitet rel-tionship hatween croselinkLing density and

equilibrium •ellik, a constancy of welling data (column 10 & 11) indicatte

a eooati-acy of the crjso-lini..ng deouty.

The a'%torials woro stored at 20oC and 65ý relative huidity for at least 14

dry prior to the mnc,.inig of test specimenso

4 .) . page 20.

I +

CWYPRAL LABCRAk-'CLIU TNO - DELtT, Hollanu; Report No. CL 67/34 page 7

3. E PoM-IMF TM1 1M-

Dumb-bell-31aped spccimenc were machined: on a modified Maximat Standard ,

r from th,4 materials de.-Oribed abo.,it Shapc and dimensions arV given in Fig. 1

r Stress-strain measiurnents -.:,re performcd on n tensile testing machine at

*+ vario-3 cross-head snceds (1O, 20, 40 and 60 cm/min), P.. at various temperatures(-45t -40, -735t -30: -20, 0 and "•C).

j|+ A small tensile test'r, developed by the Rubbew Research Institute TNO, was

kinraly plnced at our disposal b6 the sad institute. A cabinet was placed &round

the specimens, including the grips, and thermostated fas was blown :,ito the c:.'mber.

For temperatures above arbicnt, .lectiically heated air .;3s ured. For temperatures

below ambient, air was blown alonr solid C 2 . In the cabinet, a temperature 3cisor

operated an electrinaes ieatirn deovi o nl t'nus regrulated the temperature. The

sensor element ~ calibrated -zitli , tneiv.cucter placed in the cabinet.

The tensilo force w.- measured .-.th ,i force transducer :t tha upper clamp

T' (the lower cizmp vzý.s the drivlen clamp). The elUctrical output of this transducer

was fed to a str.p cha ,et trccrder. The voltage of tine signal was printed digitally

at fixed Aime inteý'vals o:' ? ccc. These data wer'e used for evvluaticn of the

results, usinr a digital computer, !2t3r.r:ards.

The tensile stress, c, was c lcu!atea with respect to the oripinal cros-qsec-

'tion of the prism-tic part of the opecim¶.n.

* The tensile a+rain, t: of tle .•ri-rmatic nmrt of the specimere depends cn thE

displacement, .%1, of the c'Aar- . Ao the cr(,s-section is not .jiifovr 1,nd tho

material behaves :Ln-dook' z, tn exact v'aluu o-' e cannot be+ obtained from the

clamp di. 1,lacemon:t. C:i•.P+a;... i+ found by as=W.mir the material to be

Hookenn mvirr'hn, an,4 ft r'nkXiM'-' C by :as'uaz thAt the heada of the n;eci-

meol do not det'I.AU ut ')1l:

In this x.tsi~ ' 1i k!)v 'In..ten th~e lot),,U. W~t-ean the rv-t b1 den 0

the ,trllcrt -.A V.L'-_ 'th of .'ýhc; pcmon. ior 11,h Uw..on Valua

+' It

cf 1' 'u in f ; i or Vi-. prim+tic 3Lt

of the .ipccimttr.

I 1Maier &Co. Ha.UAll "

CEIltRaAL LABORATORIUM TNO - DMEE, Holland, Report No. CL 67/34 page 8.

!

Shorrtly after the start of the experiment, deformations are small, and c will

. be close to cmin; near the end of the test the hods show little or no dewettimg,

so c will be closer to cmax" For convenience, tne mean value of cmax and cmin was

chosen as representative for e. From the dimensions of the dumb-bell according to

Fig. 1 it followa that

c= 1.48 (l/1 ) (2)

However, the real strain msy differ up to e5Op from this value.

Using this result, it was fouid that tie four cross-4head peeds mentioned-1

earlier correspond to strain rntes of respectively 0.025, 0.05, 0.1 and 0.15 s .

These figures are much smaller than the strain rates at which similar materials,

prepared in our labcratory, were tested at Columbia Univer'sity, New York (9):

-10.37, 2.77, 7.40 and 2-.1 s- . The results of "Columbia" Aill be compared trith our

results in the present report. -

ii

CENTRMAL LA. IH TUNO - DEIFT, Holland, Report No, CL 67/34 pag 9.

r ~4. DIS3CUSION OF MMSUS.

4.1. Stress-strain data.

A considerable spread L+ 30o-) was nbserved between the rupture properties of

I. individual specimens from the same batch. An investigation concerning the ultimate

stress at +300C of a number of specimens from various batches revealed that the

j1 mean *slues of about 5 measurements on specimens from one batch were rather close

to those of tests of other batches. From this finding it was concluded that the

batches are prectically identical uith respect to their ultimate piroperties.

Sometimes rupture was observed at one of the ends of the prismatic part of

the specimen, indicating that machining of specimens, or the shape of the specimen,

needed improvement. Results obtained irith these specimens were not rejected,

however, because no deviations from the geaeral behaviour were observed.

Representative results of stress-strain measurements at -various temperatures

are shown Ln Figs 2, 3, 4 & 5; eacii diagram relates to one particular filler

content. The curves presented in these pictures give rerresentatively selected

mean values from about 5 tests. Fig. 2 shoi:s results for the unfilled material.

I.Tith decreasing temperature, the strain-at-break, , reaches a maximum, whereas

the stress-at-break, ah, increases continuously. The maximum observed in the

stress-strain-curve at -45°C h0a. not yet been ex'plained.

Figs 3, 4 and 5 shoew results for tne filled rubbers. Eb and a depend in the

same way on temperature as do the rupture _-oI;er+ies of the unfilled material. A

definite maximum in the stress occurs at low temperatures for all filled materials.

For the 45, filled material, this also shows up at room temperature. This maximum

can be attributed to a tearing of the rubber bridges between particles, the so-

called dewettirn. Because, in these regions, the rubber is strained comparatively

more than the specimen as a whole, the da::ettiln takes place 9t a comparatively

low over-all deformation. This dewetting ',henomenon wll be discussed in more

detail in Section 4.4.

Each of Figures 6 to 12 rivos the stress-strain diagrams -t one constai-t

temperature for various amounts of filler. Observe that stress as well as deforma-

tion at rupture appear to decrease irith increasing filler content.

I

-T

CERAAL LABORATORIUh TNO - DELFT, Holland, Report No. CL 67/34 page 10,

* 4.2. Modulus of elasticity.

The modulus of elasticity, Young's ,Lodulus E, was calculated from the initial

slope of the stress-strain curve.

The tae-ile tester used in our experiments was not entirely suitable for this

type of tests: one of the troubles was that perfectly straight clamping was impossi-

ble. It was es 4 imated that, because of this inconvenience, the initial slope could

be in error up to a factor of two. Despite this the modulus was calculated from

the initial slope. This was done according to the formula:

E a(1+c)/e (3)

In Fig. 13, Young's modulus E is plotted foe the four different filler concentra-

tions as P function of temperature (open symbols). The filled symbols are results

obtained by tensile creep measurements on the same materials (8). Also inserted

are results (crosses) of tensile tests at high strain rates carried out at

Columbia University on similar materials [(9), Figs 1 and 91. MIoduli calculated

from these results in general appear to be 50 to I 0• higher than our results. In

vietr of all the possible errors, and the fact of the differences in the testing

speed, this range of agreement is satisfactory.

4.3, Failure envelope.

It is knoim from the literature (10-15) on ruptu-e properties of filled and

unfilled rubbers that a relationship exists between Eb' Wb' time to break, tb?

and absolute temperature, T. This relationship is described by a curve in a three-

dimensional space. Along the three orthagonal axes are plotted respectively

(273 %b/T), eb and the reduced time-to-break, tb/a Quantity a, is the well-

knov•n time-temperature shift according to the `.L.F.-equation:

log aT = -c 1(T-To)A(c2 +T-To) (4)

The values for aT were obtained from torsional pendulum and torsional creep

measurements at small deformations. The shift factors for the various temperatures

were p&atly published ( 6), and partly oltained from measurements not yet reported.

CENTRAAL LABOUATCIUM TNO - DELFT, Holland, Report No. CL 67/34 page 11.

Unfilled rubber and materials filled with various amounts of sodium chloride were

N investigated, as is indicated in Fig. 14 by different symbols. A curve according

to Eq.(4) was drawn through the individual points. By trial and error it wasIfound that the best fit was obtained through use of the following constants:

T 0 = -350C, c1 = 27.6, c2 = 38°C. T was arbitrarily chosen as a reference

00temperature. The shift can be referred to 1nother temperature, To, by replacingSthe constants in E(4) by new constants EI and 2(16)

c62 = c 2 + 'f -T 0 ()[~=o

When the glas-transition is taken as a reference, we find that T = T = -520 C ( )•0 g

" a, = 21, Z2 = 500C. Above A°C no values for the shift factor aT were available;

according to Williams, Landel and Ferry (10), the equation for the shift factor

holds up to 100°C above T . As a consequence, an extrapolation up to +-40°C should

be justified.

'&hen the curve mentioned before, representing the relationship betimen Lb,

T b tb and T, is projected onto the b, Eb -plane we obtain the so-called failure-

envelope. This is shown for the four materials under investigation in Fig. 15. A

rather large spread is observed, but this is not unusual for Tnis type of investi-

gations (1). Fig. 16 gives the upper part of the failure-envelope, in which

different symbol fillirgs are used for the four different temperatures in order

to show each temperature dependence. The impression is that measurements belonging

to one temperature lie on straight lines nuiking a fixcd slope, around unity, with

the axis. This may be explained as follows. Each stress-stra•i cuirve makes a slope

of around unity near the rupture point and the stress-strain curves of all

experiments are very much alike. Rupture eccurs at iarious positions along this

curve, resulting in the observed scatter. For ir,.retsing filler content, tie

stress-strrir cur_,-es become more horizontal at their end and the rupture points

lie more on a horizontal line.

1,L-

CENTRAAL LABCATORIUM TNO - DaaTL, Holland, Report No. CL 67/34 page 12.

The two other projections of the curve are shown in Pigs 17 and 18. According

to Eq.(4), time and temperature are interrelatr'd. Therefore, the horizontal scale

in Figs 17 and 18 is at the same time a scale for a reduced temperature. The time-

temperature relationship is not a simple expression; for reasons of simplicity we

only indicated at the horizontal axis the temperature for a rupture-time of 60

seconds. As in our measurements the variations in strain rate were rather small,

points referring to the same temperature tend to cluster.

It was mentioned earlier that shape and level of the curves show a spread of

about 3CP. As can be seen from Rig. 18, materials with 0 and 45j filler differ a

factor of 2 to 10 in rupture strain. The magnitude of this effect is well beymnd

the experimental spread. It may be concluded that the deformation at rupture

decreases with increasing filler content.

The stress-at-break, q0, is not subject to geometrical errors. Comparing the

W; and 4%. filled materials, we observe a decrease of a factor of 5 at low and a

factor of 2.5 at high temperatures. The highest value for the unfilled material

is 5 x 107 N/m2. This value is in accordance irith corresponding values given in7~ 2

the literature: ob = 4.4 x 107 N/m2 for a natural rubber vulcanisate (11) and

ab = 3 x 10 7 N/m 2 for SBR rubber (1 3).

At Columbia University, New York, tensile tests were performed on similar

materials (9). Up to dewetting the curves are similac, but thereafter rupture was

found at much lower strains, obviously because specimens of another shape were

used which gave more stress concentrations in the clamps than the type of dumb-

bell used in our investigation. Therefore, a comparison of these results with ours

had to be restricted to the dewetting phenomenon and was not extended to the

ultimate rupture properties.

4.4. Dewetting.

The maximum in the stress-strain curve is ascribed to dewetting; it coincides

with bl~uwching of the specimen. Dewetting i' the failure of the rubber around the

particlos, or breakage of bonds between rubber and particle, Vacuoles are formed

and the volume of the specimen increases suddenly (3). If this process in due to a

failttre of the rubLer, dewettijV, should have the sane time and temperature

dependence as failure of uz'illed ublir. Therefore, w.o investigated the relation-

The ratio of s=illest to l':revst brbdth, b2/b 1 , I a I5 in our, and 1.5 in Columbia'ein.venig •at iono.

|1

CENTRAAL LABORAT'ORIUM TNO - DELUT, Holland, Report No. ( L 67/34 page 13.

ship between dewetting time t deformation at dewetting d , and stress-at-dewetting

S ad in the same way as was done for the rupture properties. For practical reasons

it was assumed that the maximum in the stress-strain curve represents the breakager of this intergranular rubber, although this breaking goes on beyond the yield

point. Because the most pronounced maxima were found for the highest filler

concentrations, the investigation was restricted to these materials. Results are

indicated b, crosses in Figs 19 and 20. In the same diagram, broken lines, copied

frcu Figs 16 and 17, are drawm, corresponding to specimen rupture. Also inserted

are results from the literature, namely: dewetting maxima from tensile tests at

high strain rates (9), and rupture stresses from tensile creep experiments (8).

Upon studying Figs 19 and 20 it will be evident that the dewetting stress, adt

depends in the same way on the reduced rupture time as the breaking stress, ab,

for an unfilled material; apart from the dewetting stresses of the 45, filled

materials being a factor of 4 lower than the bre~ing stresses of the unfilled

material. The same holds for the deformation which appears to be a factor of 20

lower than for the unfilled rubber. These differences can easily be understood,

"because the strain in the rubber between the particles is much greater than the

"macroscopic strain. T.L. Smith (17) derived an expression for the ratio of yield

and rupture strain:

Ed/Cb = 1 - 1.105 c (6)

in vhich c denotes the volume fraction of the filler; at c = 45%, thdi equation

predicts a factor of about 0.15, vhich iz not ia contradiction ,.ith our findings.

We conclude: becai~se the dewetting phenomenon deponds on time and tempera-

tur*e in the same wny as do the ultim,%te proierties of the rubber matrix, we are

very probably concerned wiith a urocess of failure in the ribber itself rather

than a faihlu'e nf Zondo betwetn rbbcr and particle.

Inae-endent evidenoe for t-i, vi:e:point was obtained from mi.croscopic ob-

serv. tions• otr ,a e ditetted salt-ivirticles fdle,,n from the broken specimen.

'Jhen weter ":as adkiod, it vis soo, ti- t the salt particle dissolved but that some

inrsolull'e tieces %t,:tirved, v r', U.y pirces of rubber, which orig/inally adhered

to the : o:e of iv zi cryodl. Onlyv part o0' the surface appeared to be

covered -tth it ur'H'tr. Thih irnve:,tigsitiun is still in progress.1!1

COMTRAA LAOCRATORIUM TI0 DEISTp Hell-alp Report; No. 67/34 pg 4

SOberth and Bruonr(r , came to the same conclusior, when they subjected

polyurethane specimens in which a single steel ball of 3 mm diameter was imbedded

to tensile tests: the rubber is the deterimining factor in the deiretting phenomenon.

4.5. Comparison with cr -p co:P:.ments performed previously.

Tensile creep experiments, up to failu.xe, on filled rubbers were reported in

Technical Report no. 5 (8), In a creep test, the specimen is bubjected to a

constant tensile force, which means that the stress is constant or slightly

increasing during the test. in a stress-strain test, the strain rate is practi-

cally constant so that the stress may vary and exhibit the maximum in the stress

mentioned earlier. For a creep test, the cc.2igurwtion after deiretting is more or

less unstable, resulting in a shz--p i-nroase in strain rate after dewetting.

From Fig. 5 it follows that the largest stress for a highly fil.led material

is the dewetting stress; a creep test with this material consequently measures

the dewetting stress. For lowrer filled materials the highest stress is the stress-

at-break, as can be seen from Figs ?. 3 & 4o This is the same stress which is

found in thi± creep experiment. This distinction is not of much practical impor-

tance since (see Fig. 19) it is obvious that, after the time-reduction, ad and

fall on the same curve.

The results obtained from the creep measurements as reported ea• ier (8) are

inserted in Figs 19 and 20. The agreement is -ather good dcspite the fact that

the type of deformation in both cxperiments is different. Moreover it should be

noted that specimen dimensions and shapc•; were diffcrent, Because the particle

size was not small with respect to the cross section, shape influences are likely

to appear.

C9E1TRAAL LABORATORIUM TNO - DELFT, Holland, Report No. CL 67/34 page 15.

5. CONCLUIONS.

I* When the filler content of the filled polyurethane rubber increases from 0 to

S45%1 by volume, the rupture stress decreases by a factor of 5 at low and by a

factor of 2.5 at high temperatures; the rupture strain decreases by a factor

[ of 2 to 10.

2. The theory of the failure envelope could be applied to the rupture properties.

3. The stress-strain curves of materials with 45/ filler show a maximum at

strains of about 2Q,,. The maximum is explained by assuming dewetting of the

particles in the matrix.

1 4. The dewetting maximum is temperature and strain-rate dependent, in the same

way as the ultimate properties of the unfilled rubber. This lends support to

the view that dew'etting is a failure of the rubber between the filler particles,

and not a failure of the bond between rubber and filler particles, nor a

I failure of the filler particles themselves.

II

I

I

I

CENTRA"i LABORATORIUM TNO - DELFT, Holland, Report No. CL 67/34 page 16.

6. REFERoCES.

1. F.R. Schwarzl Mechanical Properties of Highly Filled Elastomers,

Technical Report No. 1, Central Laboratory TNO,

Delft, July 1962.

2. F.R. Schwarzl Mechanical Properties of Highly Filled Elastolers

II, Influence of particle size and content of

filler on tensile propertles and shear moduli,

Tech•ical Report No. 2, Central Laboratory TNO,

Delft, April 1963.

3. P.R. Schwarzl Mechanical Properties of Highly Filled Elastomers

III, Influence of particle size and content of

filler on thermal expansion and bulk moduli,

Technical Report No. 3, Central Laboratory TNO,

Delft, June !964.

4. F.R. Schwarzl, H.W. Bree and Mechanical Properties of Highly Filled Elastomers

C.J. •ederveen I, Proc. 4th Int. CongrT. Rheology (Providence,

1963), E.H. Lee (Ed.), Interscience/wiley, N.Y.

(1965) Vol. 3, 241-263.

5. C.11. van der Wal, Mechanical Properties of Highly Filled Elastmers

H.W. Bree and F.R. Schwarzl II, J. Appl. Polymer Sci. .2 (1965) 2143-2166.

6. F.R. SchvTarzl et al. On Mechanical Properties of Unfilled and Filled

Elastomers, Proc. Fourth Symposium on Naval

Structural Iecimnics, A.C. iringen, H. Liebowitz,

S.L. Kok, J.K. Crowley, Editors, Pergaman Press

1967, Neir York pp 503-538.

7. H.W. Bree, F.R. Schwarzl and Mechanical Properties of Iihly' Filled Elastomers

L.C.E. Struik IV, Influence of particle size Wd content of

filler on tensile ('reep at lar"ve deformations,

Technical Report No. 4, Central Laboratory TNO,

Delft, July 1q65.

CELTRAAL LO3ALTORIUM TNO - DME1'i, Holland, Report No. CL 67/34 page 17.

8. L.C.E. Struik, H.W. Bree and Mechanical Properties of Highly Filled

F.R. Schwarzl Elastomers V, Influence of filler characteris-

tics on tensile creen at large deformations,

on rupture properties and on tensile strain

recovery, Technical Report No. 5, CentralrI Laboratory TNO Delft, April 1966.

9. T. Nicholas and The effect of Filler on the Mechanical

A.M. Freudenthal Proporties of an Elatitwer at high strain

rates. Tecliical Report -To. 36, Columbia Univeisity,

Departient of Civil jn,•ineering and Engineering

Iechanics, Neir York, November 1 366.

10. M.L. Williams, R.F. Landel and J. Amer. Chem. Soc. JZ (1955) 3701-3707.

J.D. Ferry

11. T.L. Smith J. Applied Plhvsics M (1964) 27-36.

12. T.L. Smith J. Polamer Science Al (1963) 3597-3615.

13. J.C. Halpin J. Applied Physics 55 (1964) 3133-3141.

14. J.C. Halpin and F. Bueche J. Applied Physics 35 (1964) 3142-3149.

15. R.F. Landel and R.F. Fedors Rupture Amorphc'is Unfilled Polymers in

"Fracture Processes in Polymeric Solids",

Ed. B. Rosen Interscience, New York, 1964,

pp 3&1-485.

16. F.R. Schc-arzl in Houvink/Staverman, "Chenie und Technologio

der Kunststoffe I, Akademische Verlasgesell-

schaft Oieest & Fortig K.-G. Leipzig, 4. Aufl

17 , 7 4'. . am,• th ? . u .S • livol . 3 (1,'];5 •) 11 ")- 136 .

18. A.E. O~ertýý ttnd :{:.':,•,: •:::.30•. 40,o0l. 1i(' 6%) 5-1650

eL

I

CENTRAAL LABORATORIUI4 TNO - DELFT, Holland, Report No. CL 67/34 page 18.

CATONNT FIGRES.

Fig. 1 Shape and dimensions, in mm, of the specimens used for stresd-strain testing.

Fig. 2 Stress-strain diagrams of unfilled polyurethanes at different

temperatures and strain rates as indicated.

Fig. 3 Stress-strain diagrams of polyurethane rubber filled with 15 vol%

of sodium chloride particles of 0.1 mm at different temperatures

"and strain rates as indicated.

Fig. 4 Stress-strain diagrams of polyurethane rubber filled Aith 30 vol0

of sodium chloride particles of 0.1 mm at different temperatures

and strain rates as indicated.

Fig. 5 Stress-strain diagrams of polyurethane rubber filled with 45 vol}i

of sodium chloride particles of 0.1 mm at different temperatures t

and strain rates as indicated.

Fig. 6 Stress-strain diagrams of a polyurethane rubber filled with

various amounts of sodium chloride particles of 0.1 mm at +30 C.

Fig. 7 Stress-strain diagrams of a polyurethane rubber filled with

various amounts of sodium chloride particles of 0.1 mm at OC.

Fig. 8 Stress-strain diagrams of a polyurethane rubber filled with

various amounts of sodium chloride particles of 0.1 mm at -20°C.

Fig. 9 Stress.-strain diagrams of a polyurethane rubber filled with

various ioiomts of sodiua chloride particles of 0.1 mm at -30oC.

Fig. 10 StresL-strJain diagrams of a polyurethane rubber filled itith

various amounts of soditun chloride particles of 0,1 mm at -35 C.

Fig. 11 Stress-strain diagrams of a polyurethnite rubber filled with

various maounts of sodium chloride particles of 0.1 :-n at -400 C.

Fig. 12 Stress-strain diagrams of a polywetthaio rubber filled with

various amounts of sodium chloride particles of 0.1 mm at -45 0 C.

CETMUU IABOBATORIUM TNO - DELT, Holland, Report No. CL 67/34 page 19.

Fig. 13 Young's modulus E as a function of temperature as found fromvarious tests.

r Fig. 14 Time-temperature shift function a_ vs temperature for polyurethane

rubbers filled with various amounts of NaCi. Reference temperature

is -35 0C. The drawn line is a curve calculated according to Eq.(4)

with c = 27.6 andc = 38C.1 2=

"Fig. 15 Failur-e envelopes for polyurethane rubbers filled with 0, 15, 30

and 45 vol5/ of sodium chloride particles of 0.1 mm.

Fig. 16 Same as Fig. 15, upper part of the curves.

Fig. 17 Reduced rupture stress vs reduced rupture time for polyurethane

rabbers filled writh 0, 15, 30 and 45 vo1o of sodium chloride

particles of 0.1 mm.

Fig. 18 Elongation at rupture vs. reduced rupture time for polyurethane

rubbers, filled with O, 15, 30 and 45 vol5U of sodium chloride

particles of 0.1 mm.

Fig. 19 Diagram of reduced stresses as a function of reduced rupture time.

Plotted are rupture stresses obtained from Fig. 17, and from creep

experiments, as well as detetting stresses.

Fig. 20 Diagram of elongation vs. reduced rupture time. Plotted are

elongation at rupture obtained from Fig. 18, and from creep

experiments, as well as elongation at dewetting.

Delft, 12 juni 1)t)7Nei IV

TAOE I , COId08ITION AnD PROIPTIES OW NaCl-FILLD SAMPLES PREPARED._ _.. ,• ,-_ ,. __ -- :

1 2 3 4 5 6 7 8 9

batch no. Compcsition prepolymer filler coutont fraction NaCl-fiLer eS/i00 g polyether vol % Nami no. particle (o vo

Deamophen 3600 from from size ahloroS.I Tn D- d-JAM d form

dientr

3600/283 19.6 4.0 4.0 400

3600/284 19.5 4.0 4.0 402

3600/286 19.6 4.0 2.5 404

3600/288 19 6 4.0 2.5 400

3600/289 19.6 4.0 2.5 401

W600/290 19.5 4.0 2.5 - 401

3600/315 19.6 4C0 3A4 29.3 29.8 4 90-105 365

3600/316 19.6 4.0 3.4 29.3 29.7 376

3600/317 1).6 4.C 3.4 29.3 29.8 375

3600/318 19 5 4.0 3-4 29.3 29.9 370

3600/319 19.6 4.0 3.4 29.3 29.7 390

3600/320 19.6 4.0 4.6 14.5 14.8 383

3600/321 19.5 4 0 4.6 14.5 14.8 379

3600/322 19.6 4.0 4.6 14.5 14.8 391

3600/323 19.5 4.0 4.6 14.5 14.8 385

3600/324 19.6 4.0 4.6 14.4 14.8 395

3600/325 ' 19.5 4.0 4.6 14.5 14.9 400

3600/326 19.6 4.0 3.4 29.3 29.7 383

3600/27 19.6 4.0 2.6 44.4 45.0 38646•0/328 19.5 4.0 2.6 44.4 45.2 393

3(00/3n 19,6 4.0 2.6 4.4 45,C 396WSC,,13!0 19.6 4.0 2.6 44.4 45.2 392

3600/331 19.5 4.0 2.6 44.3 45.1 394

3600, 3 19.5 4.0 2.6 44.3 45o2 . 38

•im. pW 20.

7 8 9 10 1 12

-action XaCl-filler swe.ling 230 C d/"A temperatureno. particle (, vol.increase) /cmI) of testsize chloro- trichloro- 200C OCJm form etI.ylene

400 306 1,067 +30

402 306 1.067 +30

404 305 1.067 +300310 1.067 0, +30

401 305 1.067 -45, -40, -35, -30, -20, +30

401 307 1.067 +30

4 90-105 365 287 1.398 -30, -20, +30

376 289 1.397 -40, O, +30

375 285 1.398 -45, -35, +30

370 285 1.399 +30

390 290 1.397 +30

383 295 1.235 +30

379 289 1.235 +30

391 300 1.236 +30

385 299 1.235 -45, -35, +30

395 302 1.236 -40, O, +30

400 304 1.237 -30, -20, +30

383 297 1.397 +30

386 290 1ý563 +30

393 291 I 565 +30

396 292 1.563 +30

392 237 1.566 -45, -35, +30394 297 1.564 -45, -40, -30, -20, O +30

398 299 1.566 -20, +30

II

;0.2

L CLAMPI R clA

6_"0.1 *1

to ca Of

n u

II nLAIX?

Fig. 1

IE IrU Ln o)

c in 0

I- I- m ) m a 0 r

-o Go o I o Go I o 40 c t0

t l Z) Q.0 0 0 4 4 a x

C),

Ez

Co00oL 0 Co0V CV

U--0C LOaLAD DC

.U- C D AM to E 00 00

61ý: z I I +____ -4 M~ LA LO .4 LA C__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ )

U- rU W 0 4 U

LC) o.-

(NN

CNE

bLn__ _

0C00D

0 0

.......... O 9 0

LGLJ

410 (D :ý 0 D C

toOfl 000

4~ C V C V) C -4 V

C)

Cf)

C...)

Ez

UO0 D

Cl Cý 0U) 0L0O

LOl Lo.

EnLnt

0 0 0 o0LOO 0to00 00

Wn (D to c mw 0 CcJ (W '4c)

UL1) OL -

. 6 . 1 0A0 *O~LCO

it-

____ ______ ___ __

LO

C.'4Ez

0 0o 0to C)to 0

V LC)

En +

* n 0LoC

U)OucO

o () Ln MN C)

MILL~fO.m litCI

______~~~~~~ __ _ __ _ _ c13___ _ - __ _ _ _ 6

ICxU

tIf

_ _ _ _ _ _ _ D

L~a c -

V)N

otL~04

CV1 0

Ln 0wan

Ln U1)

I CDI - _ __ _ _ __ _ _ _ _ Lr

LE tn 2tn c

iqj0) C S

I o

_ _ -- 7 _5

CNN

EOK\I S~== *%%

Cy) CN 9

r _ _ _

r 9.-Ii

.q4

44

4O C Ca) f)a

LL +C +

C14E_ _ _ _ _ _4-

z4* ~ 4

CD C4LO C) U4

'CDNC

E 0~

c 0v)

;_ E ) L r 7C L 0 **)* - -"O

06 O - O m fnlO(,cn C) C

~ ~cc~;ocC.

C) \CV I 4

LC

__ _ _____ _ __ _ _____ _ _____ _ __ _ ___

L.c 0 C)S

~ I tC)OLc)0

~c~L oOl U) 00 V)

>110

II

c 00

Q \

0 In

NCi

_ 0

__O 0

- ~ - -~ . --- ~~----~ \---\7-- - _C_

xC_ _ _ I _

_ n _ _ CD_

0' (V v -)

0 (VV) CIfLZ .0 0 00

--- C)

LJ CrcwO)

4) H 1 U-

c1 IV cc

Cl X4 4 & JU)x I . I L fni i

4Q.Ql0 CI!

IA U) I -- Y %

£1)l

I. o(Df

w.I~li________ _______

- IC>

C>j

* t......... ........ . ... .... .. ...

..... . .. .. . "-

-------..-- --

... ... ~ ." ... . .. ...1 . . b. . ,J . ". .

............ io ... .... .. . .. .•... .t.

.

.- i r -4 4 -K -. •• -. •.. . ' " -'

b t ! , : - - •4I t i h i l I l " -

*11

- --- - ----- --

2.

I II".-' •- -- ' ' "

-;e-0 LO

-T--

-

S. ... ; -;----•- • F"'• -- . . ... . • --- '- --- [ T- • -- -- -•--- -- "- - ! . .- , ---.----- - ---- -r -:- F

-• - - . i . .. ' ' -.-, -- --- -• ---.-- -. .-- - • I . . . . - • : ! : I .. .! -] 1 . '

S! i L .. .. II .. . . '" -~~~~~~~ -: -----------. .. . . . . . ..• - : . . .. . . : . . . . . . .. -,

"---------- -... *'. .. .... - '- " .... .".... •....'--------• t '-T-'-1- ' : . . 1.

I.IF. I

...... .. ....-. . : .i.

--•--... . .... .. ..÷ ....i... • ......F. .• - - - •. . -.. .A',T - • -I T T .,, I

V7' 7

~. .......... . . . . . . . . . .

*. .. .. .- ..- ..

I

.. .. .....

7I 4

A.Ii

----- -- --- -*

--- --- ---

7 777 7.I..... ..I

tt4

1T_ .I

4-77

-~i ~ ~ ..~..i..

.......... 4. -~ --- ...

!

1000 {- --- A - I-

I i

; 30%.

4,5 °/o ~t , T

V5/ I

.0- -- - __ .. . . . _

r ! - it

Nn V_______ _____._.____ ______

I Failure envelopespolyurethane rubber

- 1- -- NatCl 90-105AMm.0 %/ NaCt

2 00 ,

i00

10_ 10 1000. al9015um - t

• ,.-t---Fig 15" 15 , . .

* It I IllII _ D

-Li-

Ia _

'4' -±D

-.----- go -

- 4 V7 ___

Lc)

CCl-S0W L

2- __ ______ 0

__ __ __1

'W-0

-o La U -)C

LL.QLZ 0 @0 go-- -w- -

In.

C->

.C W' I

Lf) U' to so. --

4 to- '444

J.L4~O ~c~oO "/

* ~ LO * '-N'4D 4j

low/ I

a~' c ,- ZA/ '

//Ng (. I/ ..- -

C> CJ

I0 _ O

n I / :

o 4/ 1 --- -

o _ _ t- .ffi_ _ _ _

--- ,---.--t -_ _ __ _ _ __ _ _

1> !j... f 14.

,c LC)i~0W /OD I

IlL Z 0 -.-- -410

A /-

IN,,

0 to-) V_ Ll0) 0

* ~~~~ruo.ture . ' kr p...* ....: .. . . ... .

.. .~. .. ......

.\L~~~ a C; .1 0Q/* NaIujO

Iqoj Nat~ I 2W 10-00 im,-

. r -. 5 f .a .o o . ....I d

N~~~ t* LIT_12 -Mo

I. .. .. ...

a.7 1

:10

10 I0 101

4 71 -:1

.i~t~i~4li~dtIIi7IL.~.yr.I ... ..

ýJw ; - .. .......

~ . ~ .... ... .. .....

jA ic 101U

~ !kjint ~a'~SkaIonDOCUMENT CONTROL DATA - R & D

i$e.-irir1 0a$1,JI1-rfatiQn it tirmt, _body o( obtfUat and indv~ng4 .on~ogatijnn flifit -fir entered when the overall report Is reMxilWfcd

-}I. oRMGNATP.G ACFi@TY (Corporaro, author) 2.. RILPOAT SECURITY CLASSIFICA110H

Ceta lbrtry~O-Delft Unclassified2b. GROUP

!0Ooh~roal- Propartieo3 of Highly Yill~d Elastoirors VI. LhlfUence Of fillor contont

=.d tetporrtturo onltm te tnsilo properti~s.

4. O9SCAIWtIiV NOTES (Trp of report *ftditlusivfe dotes) TCJC1~p~ o

I j AU T94OR431 (rfr8Ue na , Midddt Itmttal, 102!< not"e)

_Yeddivoon, Cornelius Jo

*.REPORT OATE ?a. TOTAL. NO. OF PAGES 7b. NO. OF REFS

Juno 196720141a. CONTRACT OR GRANT NO. 98. ORIGINATOR'S REPORT NUMBERISI

NT 62558 ~cnc1~pr ob. PROJECT NO

e"mclRprno6

4375

C. 9b. OTHER REPORT NOISI (Ainy other numbers that may be aissignedthog report)

CL 67/34

10, OIStRIOUTION STATEMAENT

Distribution list for iinclacssirio I., ncl eors

II. SUPPLEMENTARY NOTES 12. SPONSORING MILITARY ACTIVITY

Office of .. vlal ýv. .joarc'.:

13I. A9SAIRACT

Rcmuit's zr presentecd of -tr.ostrlk1; 13 ~)or:,.xr,.ed at vcarioua~t~r~c n

* strain r'ater, on materials, filed -zdth v:.-xiLou =,Uunts Of Socilu'. , 12.oride --.tic"ez on t

arAcx _110'1. 011001 0.1 L;z!. Stros, ;ýad st~nat ra.i.urc o erun~" widthiu :o'Coaoktcent. F'or oach ,1~.(!tr~lia.1.~xt it 1was possiblo to oormtzac'ý Lhc~i l-(zafbonvolopa by -e-nim n2 vwhicki thc ~sut -could be doocrdbcd adcquav'X`L,y6

-CŽ dmo ettT' of. the ps)ati.Olco .Lr, lclx aabbe.-y %..aturixq o.t~L. in The strion-on'rain. alzmo,p I o c'.Cril-bz" to fai2.urc '"'.0 ~ .e''

~~~C "z -.d no'- t~ al.o o2r 0., i 1r-It buccnu-c "10 dcwttin . ."La:w

de~i Oil.t on ,.pertU1r 11U1 tl:.O inl tlQ -,=10 way' .13 hao rulajtwuro o:Tt.

lv- re-outo aar, ý,onp.acd w-Ith ctxarliciv xoo',i,:tionng o,,-. hi~1h a~'oend tom. . tvc.

u.. ~opoxperiment..;p and a g~ocd camumant '~oundo

FORM (AGE 1

D ,Novi J 47

3/ 10-01611Nrm1 Ut~w~

Secutitv csaqnIfirsat on

LNOC A LN S LINKN A

KEKY W0405 L 114 K ALIROLE~.ife 7 111061 ROLE *T

Polyureth.no ru~bberDW~rjy propollnnt"Filler content].Rpture strecoEloneation at ruptureFailure envelope

ReinforcementVacuolesDowetting

Di

. " I

! 0

j!I9 I I I. .