"
NASA-CR-17361619840017699
A Reproduced CopyOF
Reproduced for NASA
by the
NASA Scientific and Technical Information Facility
LIBRARY COpyUr. r .. ? fllu..~
, ' IJ jjut'
LANGLEY RESEARCH CENTERLIBRARy, NASA
HAMPTON, VlrtGlNIA
FFNo 672 Aug 65
https://ntrs.nasa.gov/search.jsp?R=19840017699 2018-06-14T14:53:30+00:00Z
CCMS-84-·0l\lPI-E-8J..9
VIRGINIA IECH
CE TER FORCOMPOSa«fE !ViATEREJ\lSAND STRlJCTl~F(ES
240S1
Vh'ginia ~)ol,·te(:I'mk:
hutituteand
Scott Mii~n Mi~kcwkh
Carl T. ~aera»u)'likh
Temperature D~penden(e of Elastic and SirengthProperdesof T300/5208 (jraphi~e-EpoJt},'
«NASA-CR-17J616) TE~prRATOBG DEPENDENCE OF UBU-25167Lt;"STIC A!-W ~T.~El!G'fi.i PhUPEIl'I1£S UP nOO/52GS,,~"fi!IT;'~-E.P()XY nr.l.ryir..ia Pol yto;;;chnic Int>'i:."<lull Stdte (lIlt .. ..) 4.J P he A03/f'lf 1'.01 tJnclds
CSCL Hn GJ/2 1+ 13/:>11
College of Engineering
V1rgfnfJ Polytechnfc Institute and State University
Blacksburg, VA 24061
VPI-E-!33-9 April t 1984
~,~ii
Temperature Dependence of Elastic and StrengthPropert ies of T300/520B Graphite-Epoxy
Scott ~ilan Hilkovich 1
Carl T. Herdkov1ch 4
Department of Engine~rlng Science &Mechanics
Interim Report 41The r~SA-Virginia Tech Composites Program
NASA Cooperative Agreement NAG-1-343
Prepared for: Applied Materials 3ranchNational Aeronautics &Space Administrationlangley Research CenterHampton, VA 23665
IGroduate Student
2Professor of En9inee~1ng Science &Mechanics
i<',
ABSTHACT
Experimental results are prE.'Sent(~ for the elastic and strength
properties of T300/520U graphite-epoxy at room temperature. 116K
(w2tiOOf). and 394K (-t250°F). Results C2re presentl>d for unidirectional
UO. 90°. and 45° laminates. and ±30o, ±45°, and t6U IJ angle-ply lami··
nates. It is shown th,it the stress-strain behavior of the 0° and 9Uo
laminates is essentially linear for all three temperatures and that the
stress-strain behavior of all other laminates is l1ne·ar at 116K. A
second-order curve provides the best fit for the t~nperature dependence
of thf! elastic modulus of an laminates "nd for the principal $hf~i.H'·
modulus. Poisson's ratio appears to vary linf)arly with tempefatUI'~.
1\11 IllOduli decrease with increasing temperature except for £1 I'/hiC'l
exhlbits a small increase. The strength temperature depende'K.e 15 al'!,t
temperJture dependence. In many cases the tt:111perattH'E' dellendence of
properth..s is nearly linear.
2
1.0 INTRODUCTION
The use of fiber-r9inforced composite materials in the .1erospace
industry has greatly ificrec1scd in recent years. The gr'eat advantage of
these materials being their superior strenvth to weight and stiffness tJ
weight ratios. This rnal:es composite Hiat~r'als ideal for use in space~' :
app 11 cat1ons. In order to efficiently design structures for use 1n
space. it is necessary to determine the effect of the space-envir'onment
on these materials. The cold of space coupled w1tb radiant solar heat-
10g effects cao lead to a wide range of operating temperatures. Because
of this. it is necessary to understand how temperJture affects th~ basic
material properties of composite materials.
Obviously. strength and stiffne'!>s are two basic ffillt.erial pr(\pf~rties
whiCh are of fundclmenta 1 importance. Many proposed app1i cat 1ens of
fiber-reinforced material~ fOf' use in spac(!-er:vir'on~nts 'Involve gr'aph~'
fte-epoxy compos Hes. Tt,.: s paper presents data on the strength dnd
st1ffness propert i es of nOO/S20B graph ite-epoxy compos He over the
temperature range of 116K to 394K (-25l)e'F to +2S0°f). ;;1l:! des i gnati on
f300/5208 indicates that the graphite fibers are Thornel (Union Carbide)
nOD fibers in a matrix consisting of Marmeo 5208 epoxy r-esin. Tl1e
tel?'perJture range of 116K to 394K represents the temper'atw'e extremes
that llldy be cncount.:red ina spaee··envi ronment [l]. Di:tta 9 generated
from this study. arc presented in graphical form. (stiffness, strength
versus temper-ature and stress-strain curvps) as \icl13s in tabular
form. Test procedu:-es are di c;c\Jssed and comments reg3 ,'di ng materi a1
behavior are injected as warranted.
,~
i~_"''''~'_''__''_:_''''''~_AA"_'' ".~'--~'"-";-'~";~.--;'7 ';'''':;;;:;:;;;:'.;:''n.':-''''''·''''='''''''';;,-", «"""'" ...,_ .. , .. ~-'-''''''
2.0 PROC£ljURE
Test specimens were c:ut from panels of T300/5208 graphite-epoxy
that had been ultrasonfc1y C-scanned to insure their integrity. The
test specimens measured 1.27 centimeters by 25.40 centimeters (0.5
inches by 10.0 inches) and were 1.02 millimeters thick (0.040 inches).
This thickness corresponds to 8 lamina layers. Six different lailJinate
lay-up configurations were tested: (O]ti. (45]8' and [90J8 (all of which
are unidirectional lI'!atf.:ria1) as well as (+3°21-3°2]5' [+452/-452]5' and
(+60 2/-60 2]s' Tabs were not used for load introduction. After allowing
fOl' the gripping region. the aspect ratio (lengthh~ir.lth) of the test
weeks.
Prior to testing. each composite specimen \'las fitted with a r\~;:·
specimens \'1as twelve. Once the specimens had been cut to SLH~. they
WK-OO-120WR-350 (Hicro-Measure-
ments, !nc.) strain-gages Here used. Tht se gages \~ere choos.en bec.ausE;
were placed in a drying oven set at 373K (212 Q F) for a pedod of hio
tangular (45°) strain-gage rosette.
they are designed to ~"thstafld large tempe,'ature ranges. The gages wen~
mounted using M-Bond 6UO adhesive (r~icro-t'1easurements). fonowing the
suggested mounting proc~dure. This procedure includes lightly sanding
the composite surface to produce a smooth area were the stra'in-gage .is
to be located. Finally wide-temperature range lead wires were attached
to the strain-gage rosette (326-GJF. Micro~1easurements).
All t.ests \4ere pel'funned 1n an ATS environmental (;hamber that uses
res)stance element:; f('r l1eating and liquid nitrogen fOI' <;oo11ng. The
heat from the resistance elements is circtdated by an internal fan. The
liquid nitrogen evaporates as it enter's the chamber and is circulated by
4
a slight overpressure from the nitrogen source. Temperature is moni
tored throughout the chamber by thermocoupl cs pl aced at var10us loca
tions within the chamber. including one thermocouple attached directly
to the test specimen. Tests were conducted at three different tempera
tures: 116K. 301K (room temperature). and 394K. Soak times of foul" to
six hours were required to attain stable conditions at the low test
temperture. Two to four hours were requi red at the hi gil test tempera
ture.
The envi ronmental chamber was mounted on an MTS hydralic tensile
testing machine. The tensile machine was fitted with special "moffil::n1;-
free" grips which fit entirely within the chamber [2J. This fixture was
designed to provide accurate axi ..!l alignment 3r:d rotational fl'eedol11.
Load was measured by a resistance load cell located outside or the
chamber. During a test. stress and strain data \'4ere automatica1!j' .:'nd
periodically sampled and recorded by a computerized data acquisitiOi'l
syste.ll that is tied directly into the testing equipment. After each
test hatI been performed. ail test data is graphed. tabulated. and ana
lyzed by computel'.
3.0 RESULTS
Sample stress-strain curves produced for this work are shown in
Fig. 1 through 6. f(e~ults of the tests performed are presented in
Tables 1 to 4. Data is catagorized by stiffness (Table 1). strength
(Table 2). Poisson's :-atio (Table 3), and shear modulus (Tahle 4). Each
teble is then ~ub-divided according to laminate lay-up and test tempera-
ture. The data is plotted as a function of temperature in Figs. 7
'.~ ".... ~,-\
....-
5
through 21, and the coefficients of a least-squares fit of the data are
presented in Tables 5 and 6.
3.1 Stress-Strain Curves-. ..,~_.._...
The data presented 1n Figs. 1 and 2 shows that stress-strain behav-
ior is nearly linear for both the 0° and 90°-material at all test
temperatlll'es. The pri nr. i pa1 moduli increase at both 116K (-250") and
394K (+2S0°F), with the larger increase occurring at the lower test
temperature. In all cases, very little nonlinear behavior is noted
prior to failure. The DO-material does exhibit a stiffening behavior at
high strains for room and elevated temperature (3].
Significant nonlinear behavior is exhibited in the stress-strain
curves at room and elevated temperatures for the other laminates 't~!sted
(Figs. 3-6). All four sets of stress-stra"in behav"jor ([45]8' [+302/"
302]5' [+452/-452]5> and [+6021-602Js ) show similar trends. At 116K
(-250°F) the stress-strain curve is esseJ"jtially linear. As the test
temperature is increased, the elastic modulus decreases and the degree
of nonlinearity increases. In ail cases. the 1arges. degree ofooo11n-
earity is exhibited at the high test temperature. This behavior is as
expected in vi ew of the kno...m i nfluenc0 of temperature on the response
of the epoxy matrix flldterial. The nonlinearity is more pronounced in
the laminates ~/hose behavior is more dominated by matrix properties.
The nonlinearity is not evident in the [90]a laminate because of the low
fai lure strain.
6
ORIGIN.J\L PAC!! f9OF POOR QUP.Lrrf
240.00 r-----------'----------,
1500.00
1200.00
300.00T300/52OB )GrlEp j [OJ8
- ROOM TEMPERATURE_.-. t-250°F (394°K)---- -250°F (116°K) I
025 Q~50 OJg-----tOO 0.00
STRAIN) PERCEl\JT
Stre5s-Str~1n Curves as a Function of Temperature forthe (0]8 Ldmfnate.
160.00
200.00
40.00
§2.....~ 120.00W0.::
tn80.00
Fig. 1..
,:1~'.
.,#
7
Stress-Str~in Curves as a Function of Temperature forthe [90]a Laminate•
Fig. 2.
ORIGINAL PM~£ isOF POOR QUALlTV
6.00• -1
35.00
5.00
30.00
4.0025.00
d?~ ~
-- (f)~
~3.00 2QOO CJ)
LLlCL: O:'~.- I-if) .15.00 (f)
2.00
10.00
1.00 T300/5208) GrIEp) [90Ja5.00- ROOM TEMPERATURE
--- t250°F (394°10•. _-- -2t:,O of (116°K)
0.0°0.00I ! - I -_..- con
0.08 0.16 0.24 0.32 . w
STRAI~~) PEr-rCFNT
..
8
"
..I
ORIGINAL PAG\! t9OF POOR QUALITY
24.00 r----- ,__, --1
.;
'i,., " 150.00
20.00
30.00
cP90.00 ~.
ta0:::::
~6000
120.00
II
II
II
II
II
II
I /
,/ (II ,
/f4.00 T30015208)GrIEp) (45)8- RCA-1M TEMPERATURE_.- t250°F (384°K)---- -250°F (! I GOK )
o 00 ._-,-_._-'- i ~-_._- 0.00, 0.00 0.25 0.50 0.75 1.00
STr~AIN, PERCEr\JT
B.OO -
IG.OO
en~
'"~ 12.00w~I-en
Fig. 3.. Stres.s-Strain Curves as a Ftmc:t1on of Temperature for the[45]8 Laminate"
.-
9
50.00
10000
25O.CO cE~
-'"200.00 ~
LLl0:::l-
15000 (f)
Stress-St,,~l''in Curves as a Function of Temperature for the(+30z/··;~Q2)s lcil'ill nate.
350.00
ORIGINAL PtH:'E ;f,OF POOR QUALITY
/I
II
I .
" ~/ .'
/1'I '
//I"
I ~
I T300/520e, GriEp) (i-30d- 3O~)sI - ROOM TEM PEHATIJRE
I _._ .·250"F (394"'K)I. -2500F (116"'K)"
O.OOo.ol:...o---,·-.I.--'---.l----- I --- 0000.20 0.40 O.GO 0.80'STRAIN) PERCENT
10.00
50.00
GO.Oor,...._.-------'----'---,-----.
40.00
§2...(J) 3QOO(/)w0:::l-eI>
2000
Fig. 4..
10
80.00
n<:!~....(f)
GO.oo if)W0::'I(j)
40.00
20.00
120.00
100.00
ORIGINJ\.L pr\,.~E F.lOF r00R Q! l,l',l.,rrf
T300/5208) Gr/Ep) (t45z/-45Js--ROOM TEMPERATURE--- +250°F (394 OK)---~ -2.50°F UI6°K)
I I I --___ 0000.20 0.40 0.60 0.80'STRAIN PE!~CENT. ,
I /
I
;' /'I ,
/ YI '
/fI I
/ /
/ f/ f
II
II
II
Ir
II
l
3.00
0.0°0.00
15.00
18.00 I--~----
12.00
(j)~
"(J) 9.00U)W0::1--<.f)
6.00 ~
Fign 5. Stress-Strain Cur'lr's as a function of Temperatlwe for the[+45~/··452]s Laminate.
·.11
.~
Fig. 6•./ Stress-Strain Cur'ves as a Function of T~erature for the[+6021-60Z]s Laminate.
40.00
80.00
d?~(I}
60.00tf)LLJQ.':jif;
12QOO
iOOOO
20.00
1
1
OHlGINAL PAGE r9OF POOR QUALl TV
9.00
6.00
12.00
15.00
IIIIIIII
I3.00 - I
I
III
0.00 L. -l--------L:--.-.-- 0.000.00 0.20 0.40 0.60 0.80
/ STRA.lN. PEF<CENT'}
18.00
(/)~
(f)"(f)LL.l0::~(f)
.. , ".' \
The results for the transver'se I\lOdulus E2f
This indicates that the epoxy ma
Since fiber properties dre independent of temper'iltlJretooperaturc.
temperatLlI'e re~ults. iflr.rea:'ing the t~;f"perilture above room temperatur'z
test temperature act"; to r.:.~lieve r(?:;idu,ll stresses, I>le lila,}' conclude th,)t
temperature value.
the transverse !nodulus ,.it the low temperature compared to the f'O-':l)1
(Fig_ 8, Tables 1 and 5) indicate th'lt thel'E is a 30 p\~r'cel1t inc:re<l~e in
this type of behavior at low teo1lpt.'i'iltures.
residual stresses.
that its effect on the modulus of the composite overrides the effect of
lower temperature Is a result of an incr'}iiSe in thi'! stHfnes~ of tht:>:
12
fibers should result ~n a hi\jher modolus of tl1'i! cOflip.)slte due to t\1~
absence of nonlinl;>ar IMtiix behavior (S~9]. The higher ITIoduh:s ell: tht;~
in lower lllC:trix stl'esses. Lower strt'sses in the IMtrh Mel str"igl't.nr
believed to be due prfmorily to reduced i'esidual stresses which rt~5u1t
3.2 Elastic Pr~pe~ties
fit are given in Table 5. The modulus is 6 to 7 percent higher at both
(over this range), the temperdture dependence observ.:d here is a fvnc-
tion of the changing matrix propt:rties. fiber wavIness, and resit1\J-:.l
much stiffer ilt the lower tel:lperature. Epades are known to (~xhlblt
OO-lamf;j(lt~. Tht1 cxpel'fment,11 tesults for the ela!>tic moduhlS t
El' are plotted ,1S a function of temperatuc'e fn Fig_ 7. Moduli value..
are presented in TaDle 1 and the coefficients of a second-order curve
elevated aCId low temperatures, ~s c(lmp~l'ed to th~t measured at rool1l
has much less influence on the trdnSVerSf.' m0<lulus. Since incf'E:a$in~1 th,;:
stresses (4]. Th~ higher md!Jlus at the elevated tempel'attlre is------ -..
--
..
-_.
13
01?p:" ..n C', t r'" 'r~ '9UF P(l.M,;i~ f~,,)-"1_~TY
40.0 ---...--------240.0
35.0 .I
210.0
30.0180.0
Q~(n~ 25.0 l?U'J 15QO "',
-}- - (./)Cf) (f)llJ 20.0
i LdZ 1200 _Oil'
t:.c:..I..L LLl.L. 15.0 Ll-I-- 9QO 1-(f) Cl)
10.0 GO.OT300/5208, GrIEp, (O)a
5.0 30.0
I0.0 _L_..L _.L.- _J 0.0
100 150 200 250 300 350 400
TEMPER/..\TURE> K
L ' _.__..1 .1_._.1 .1.--.__.1
-280 -190 -100 -/0 +·80 . +17'0 +2GO
TEMIPE·· f::J /\ T!JOE CI I \;-\ I I \ ) I
•
,~.
v..
4.0
3.5
3.0
(f)
~ 2.5""(J)
tG 2.0
6:I..L 1.5l--(j)
1.0
0.5
0.0100
14
ORIGINAL PACE isOF pom~ QUALITY
24.0
21.0
18.0 a0...
15.0(.9(f).....
(f)
12DldZlLLLan '-........~.-(/)
1 6.0T300/5208; GrfEpI [SOJa J
. . 13.0--'.-_L._._-L-._ ,_----1... " Q.O
150 200 250 300 350 400
TEfV1PERl.\TURE) K
'\
L_l...__l ...l-._! J._"--l·2BO ~ 190 ~ I00 ~ '0 +80 "'/70 +260
TEI\1PER/~TURE) I-
Fig. 8. Modulus of Elasticity~ £2. V~rsu$ Temperature~
•
..,
15
residual stresses do not playas ldrge a role in affect log trMs,,~rse
modulus 3S compared to their influence in the fiber direction. This is
cons i stent with the r~ch lower ion uence of (i bel' wa'li ne'iS on the trans-
verse stiffness •
45°-1amlnate: qU. This laminate \'135 chosen as the b0St specimen
for determination of the shear modulus, G12 , based upon the findings in
I'efer~nce [2] •. AVerdje results for the shear modulus, G12 , are plotted
in Fig. 9 and values for modulus and a second-ordtr fit are given in
Tables I, 4. and 5. The resuits were obtdined using a 45-unidir'ectional
off-axis speclme l) and the following st.;1nclard relationship (10).
where Ex is the mt'!dsurE!d axial modulus of the 45'"'-speClrifln and the oth.;!'
quanti t t es cOf'respond to prapert 1cs oj n the materi d 1 pro; nc i pa1 c'.)ordi na te
.sy':>tem. The results 1ndi cate that d second-order curve fit prO'll des In
excellent correlation of the tEmperature dependence of GIZ " The higher
~lle at the low temperature is consistt~nt with the temperature depen
dent response of w~trix dominated properties.
The variation of the axial modulus. Ex. of the 45°-laminate wfth
temperature is shown in Fig. 10 and -/alues are given 'in Tables 1 ar.d
6. Comparison betl'Jeen experimental and theoretical values (Table 1)
indicates excellent agreement at room and elevated temperatures. but.
less than stisfactory agreement at the low temperature.
Poi SSO" I S rat i2_:_ \112. ThE: temperature dependf~n(.e of \112' (Fi g.
11) h essentLil1y linear, Clecreilsing \iith increasing temperature. The
16
200 ,-_.- --------12!X)
1.75
10.001.50
SK'>O&; ? (f)
1.25~ <..9,7.50 u5'~. '"
-~en
(f)t3 1.00GPO
LtJZ ZlL. 11..l!... 0.75 LL..I-
14.50 1--tf) UJ
0.&.1 3.00T300I5208) Gr-/Epl [4% I
0.25 J 1.50. ,
J~ 00 0 L __..L.-_.__J._.~--l.__ t 0.00, . 100 150 200 250 300 350 400~.
TEMPERATUI~E) 1<
•
L, 1 -----L.__......1--..1. j _ .,,_J.-280 -190 -100 -10 +SO tllO +260
TEMPE!~TURE) F
i
jI
17
ORIGINAL rAG\! ;9OF POOR QUAUT'i
•
2(100
10.00
30.00
4Qoo
50.00
,_300
T300/5208)Gr/Ep, [45]8
--_._---,,-------------1
L-_. L_·, L -l-280 -100 teO +260
TE~~PEFi!~TURt") F
1.00 -
0.00 1 ' ~-100 200
aoo[
7.00
6.00
if)5.002: -
(f)"\(f)
4.00WZLLlJ... 3pOl-(/)
2000
E.xFi9. lC. AX131 Modulus of the [45]a lami~ate Versus TemperJture.
..
..
18
ORiGINAL PAGE tSrOF POOR QUALITY
L-__ l. -"_~.. l
-280 -100 +80 t260
TErvlPERA\TUr~S F
0.80
0.70...
';(
0.60 ...0~
0.50<r0::
0.40(f)...Z0
0.30 en({)-0
0.20 o.~
'0.10
4000.00
,.-.--------
: L=:8,Gr~~p,(O:e
100 200 300
TE~J1PERATURE>K
· .l>~
~
;<'~
0.80,
,~,....0.70
,, .... ;\," 0,60.~
...0I-
0.50" «/,/ :~y 0:::ll·,
i (/)~~: 0040
f -z
i 0'",I
J,';';'l if) 030t if).·,~I:-
~, ~~~. 020t.":'~~;{j~
19
h1 gher Poi sson I s ratlo at the low tempel'atll re is sorn(!what sur Ilr1 s i n9 in
view of the elastic symmetry condition:
The modulus I'atio, El/f2~ is smaller at the low temperature compared to
the room tempel'Citu re rat f o. Thus a hf gher' v12 at the low ternperatu re
should be accompanied by a significa~tly higher \121 at the low tempera
ture. While this trend was evident from the experimf:t1tal results, thl:
values did not correspond to satIsfaction to the E·\I relationship. In
fact, the ~asured value of \1 21 was l'\Jch too high. Accurdte measurement
,of the small strains involved is very difficult due to errors intf'od~lced
by the trdnSlJerSe sensitivity of the strain gages. In the case Qf 90o~
material, transverse sensitivity became prohibitively large compat'ed tc
the small strains that were measured.
An91e-a!,f~minates. Results for the effect of temperat!Jre on tlw
modulus of the angle-ply laminates tested are shown in Figs. 12-14., and
Tables 1 and 6. As the temperatLlre increases. the modulus decrease:; for
all lamil'Jate~. The decrease in modulus with temperature increase is
epoxy matrix becomes stiffer resulting in increased lcrninilte stiff-
3 to 6). As discu5sed in previous sections. at low temperatul'es the
also 'il1ustrated in the stress-st."alr] curves for these laminates (F'igs.
The stteSs-5tf'dfn curves exhibit incre\~sed nonlinearity ''lithness.
incre:ising temperature. nl'is results in dt:cred$(~d ove:"all stiffness in
the laminate. It should be noted here that reduced 'esidual stresses
(at higher temperature) are not N:pected to influence the stiffness of
these laminates nearly as much as it influences the stiffness of unidi-
.'
.'
r:c:(9
iff'(j)LtJZIJ....Lt..
le/)l
J
75.00
45.00
30.00
t- 60.00
T
------'._---
20
L--.--__.l- L . .J
~280 - 100 t80 +260TE~J]PERATUF~E) F
4.00 .
6.00
8.00
2.00
aoo1o!.-o-·----"--'----..J-------- 0.00200 300400
TEMPERATURE, K
12.00 ------
10.00
(j)
~
(/)""(/)wZ1..1lLI(!)
21
c£<"9
~,
(J)(f)WZLLI..!-i-(f)
35.00
. 25.00
ORIGINAL PAm·: ISOf POOR QUALilV
45,00
. 15.00
55.00
.A..--- L. -l
··100 1'80 t2GOTEMPERt\TURE) F
5.00
---,-'- I ._-- 0.00200 300 400
TEI\1PERATURE, K
L-280
8.00
7.00
6.00
(f)5.00~
(J)(f)
4.00WZLLLL 3.00I-(f)
2.00
1.00
0°°100
1_.__.L ..l ._,---'
-280 -100 t80 .;.260
TEMPEHATURE, .F
55.00
5000
4500
4O!X)r£
35CO l?(j)....
3000 U)wZ
2EiOO LLIJ..-2000 l-(/)
15.00
•. iODO
5400
40000
7.00
ORIGINAL PAGE mOF POOR QUALITY
8.00
22
6.00
2.00 l '-~t--
1.00 L-=OO/:208,GrlEp,[rSOJ -SOJa
0·°,00 200--'~------
TE1\.1PERt4TUI~E) f-(
-~ 5.00
(f)'"
tG 4.00ZLLl.L. 3.00J=(f)
'"
..,
23
rectional material in the fiber direction. As with the 45~laminate,
{ompariSOil between experimental and theoretical values (Table 1) indi-
cates exce11 ent agreement .J.t room and elevated temperatu res, but 1ess
than satisfactory agreement at the low tC'mperature •
3.3 .~trel!9th Proe~rties
OO-lamin~l~.L.-!.T' Results for the temperature dependence of (.I"' ••
laminate strength, including the type of failure mode noted~ are shown
in Figs. 15 and 16. It is observed that the strength increases rleariy
linearly with increasing temperature. These results are apparently due
to changing mati Ix properties and residual stresses. At the lOI'JEH'
temperatur'e, the matri~ is brittle and stiff, .resulting in higher' resfd~
ual stresses and less efficient load transfer in regions of stress
concentr&tion such ~s at fiber broaks. The reverse is true at elev~tcd
temperatures.
The type of failure mode observed after the tensil e tests of [oJ8
laminates is indicated in Fig. 12. Th·~ room temperature dnd high tem-
perature tests both exhibit a "shatter-type" failure mode. ~jhen the
to the fibers into many small p'ieces. Very little of the original
composite test specimen fails in this mode, the specimen splits parallel
specimen is left intact. This phenomenom appears to ne more severe with
On the other hand, lm'l tel1lpel'ature tests
at lo\;~ test temperatures. Jhese results may be an indication that the
fiber/mattix bond is stl'cnger at tJ1e low temperature. This could be the
exhibit a "transverse-break" failure mode. Instead of splitting, the
compos'lte specimen fails by a single jagged break across its I'/idth.
This 1s further evidence that the matrix material becomes more brittle
increasing temperature.
I
~\
\\i
'. \'\.
..-- J~\iI
!I
~ 1II:,
O~O - J I ._L 1 --" ! 00100 /50 200 250 ?:IJO 3:0 4O'J
TEMPERATURE) K
oQ..~......(/)(f)LLJ0::I(f)
600.0
12000
2400.0
2100.0
1500.0
1800.0
T300/5208) GrIEp) [Ol a
24
, . L-.--L I --_-L__...l-280 -190 -/00 -10 ~80 t-170 .. 260
TEMPERATURE) F
ORIGINJ\L Pf.\G~ i~
OF. POOR QUALITY
350.0
300.0
400.0
500
100.0
(f)~ 250.0
.....(/)
~ 200.00:::I-(f) 150.0
Fi 9.. 15.. Ul tirnate Str~ngth_ XT• Versus Teflllleratul'e.
..
/
Fig. 16. Failure Modes of the [uJ8 Laminate as a Function ofTempera t.u re.
+250' F (394K)
ORIGINAL PAGE ~g
OF POOR QUA.LlTV25
Room Tempe ra tu re
26
result of Ii much tighter "shrink'·fit" due to the higher r~sidual
stresses in the brittle matrix· material.
gOo-laminate: Yr. The results in Fig. 17 tndlcote that the trans-."
verse strength is not influenced strongly hy temperature. As indicated
In the figure. there is conside-rable scatter in the data at all three
test temperatures. This is typical of tests Qn gOo-material.
45°-hminate. The temperature dependence of the strength of this
laminate is shovin in Fig. 18. Clearly. there is significant variation
in str'ength depending upon the temperature. This laminate fails 11'\ it
mixed mode of lnplane shear and tr'allsverse tension. Thus it Is not 3
good specimen for determinatioi1 of fundamental properties. btlt rather
useful for correlation of theory and experiment for failure under bfi1)!("
ial loading. Such compzlrison wi.ll not be made here. The results ari;
chosen for measure;nent of sheaj' roodu·lus e not shear str'ength.
~ngle··ply .lnmi!,dt~. Temperature deptmdent strl:ngth results for
the angle~ply laminates are presented in Figs. 19-21 and Tables 2 and
6. The data shows that laminate strength does not follow a consistent
trend as it function of temperature. \lbvivusly. these 1i'lminati?S are
affected in different \/3,YS and to di fferent degl'ees by the changes in
th~ cO'"l1os'lte inducHl by temp\~ratur-e. A complex interact10n occur's
ture). the increase in fiber-direction strel1gth (with incr'ed5ing tf!lilper-
atur-e). "nd the eff(~ct of residual stresses that mayor .nay not be
relieved hy combinations of th{~ above properties. All these intf:,rac:~
ti~H1S m,J;~l~ precise stre1lyth p:'i:,dict.ions difficult for' larninate'i.
.. -
• 2.7
Fi 9- 17_ Ul timate Strength. ¥To Vet'sus len\lerature.
&:<..:>u'f(j)!.JJ0::~U)
45.00
ORIGINAL Pi~Gt: fSOf POOR \)UALIn
------.,--,,'--..,-,---,-'--l' 6000
L__ _.- , -,_..1.......-....,__,_.._--1.-200 -100 ·:-60 , ~2GO
TErvlPEI~ATUR[> F
6.00
aoo
10.00
, ~----- .L ---....
4DO r' -=::O~:::OJ8 ::::L --'-_. -1--_... ODO
100 200 300 400
TEr\~PERl1.TURE.K.'
..
28
T30Ql5208) GrIEp) [45le
:: I_L__.~,t_'----l.__--l. t __"
100 !50 200 2f~O 300 350 400
TEfv1PERATURE) K
0.0
30.0
240.0
/80.00
0.:
/500 ~'"(/)
(/)
120.0Ld0::l-t/)
9QO
000
2/0.0
ORIGINAL PAGE (9OF POOR QUALITY
----_._----_._--_.....,
/5.0
20.0
40.0
10.0 .
30.0
35.0
(f)~ 25.0
uf(f)WC1::r(/)
\:
L--I.----L----I.-_---L--.L__J-280 -190 -100 - 10 t·OO ·/70 ~ 260
TEMPERAlURE, F"
Fig. 18. Ultill1ate Strength of the [45]8 la.",lnate 'ier.'sI.IS Tempe-rature.
?9
ORlGINJ\L PI"; r~E P1OF POOR Qu,:\~.n"{
80.00 ..---._----_._------.._-
70.00 500.00
60.00
0.00
400.00 d?2:
20000
.30000
100.00
400_._--.L--.-..__.....
300
---------+-~
J
QOO,oo ----- 2~O
TEfV1PER/.\TURE} K
2000 .
~ 50.00
~r 40.00LLJ0.::f-- 30.00(f)
' .."
I --1 .. _ L ..-l
-280 -100 t80 +260
TE~~PEf~ATURE) F
Fig.. 19.. Ultimate StrCl'l£lth of the [d02/-302]s laminate VersusTet~eratIH'e..
30
ORIGINAL PAGE isOF POOR QUALITY
40.0 ---"J240.0
35.0210.0
30.0180.0
&(f)~ 25.0 2"
15UO '"'l~ (()(()
C.f) 20.0if)
W 120.0 I.JJ0::
---+-~0.::
J- I-.Cf> 15.0
9"~OU1
10.0 GODT300/5208) GI"/Ep ) [v-45tf-45z!s
5.0 30.0
00 L-L._.....4.--._...1___L.__--L.-_._ OD. iOa 150 200 250 300 350 400
TEiVlPERATURE) K
l.-.--.L .! • . .• L ..L-.._...,_J. J
- 280 -190 - 100 - 10 t80 ~ 170 <~ 260TF\/iOER ""'IUDe F_I u' _ '..A. i l\L)
Fig. 20.. U1t~mate Streo9th of the [{'4S2!-452Js lamina.te Vcr'susTemperature.
"
"
"
, ,i
, '
31
40.0
240.035.0
T300/5208, Gr/EpJt60....I·roJ~ 210.0
30.0180.0
(f)25.0 &-~
150.0 ~'" ....(f)
(f)f{j 20.0 (J)
/20.0 w0::: cr::.-,- 1-el) 15.0900
if)
--~IOJ) - 60.0
5.0
~30.0
0.0,00i L- 1 •• 1 0.0
150 200 250 300 350 400
TEMPERATURE K.>
L .L__ I -L__ I J I
-280 -12-0 -100 -10 I-so ~170 l-200TE~J1PEI~ATURE > F
fig .. 21.. Ultimate SJ:.rength of the [+60ZI-60Z]s laminate Ver~u5Telilperatur~.
32
TABLE 1
Stiffness (Ex) Versus Temperature Data
La1..:!!Q. 116K (-250°F), 301K (H. ~.J. 394K, (+250~E.L
[0]8: El 157.9GPa (22.90msi) 147.7 (21.42) 156.0 (22.62)
[45]8 28.2 (4.089) 16.9 (2.452) 15.2 (2.211 )
<THEORETICAL>* 20.3 (2.950) 16.9 (2.450) 15.2 (2.210)
[90]8 14.5 (2.100) 11.1 (1.61.4) 11.7 (1.695)
[+302/-302]s 70.4 (10.217) 52.2 (7.578) 50.3 (7.291)
<THEORETICAl>* 6:.2 (9.160) 10.9 (8.420) 55.8 (8.090)
[+452/-452]s 31.2 (4.523) 20.7 (2.997) 18.2 (2 r;'~8)\ .0 ...
<THEORETICAL>* 27.0 (3.910) 23.8 (3.450) 20.3 (2.9&O)
[+602/-(;02]5 27.6 (4-.000) 12.5 (1.814) 14.6 (2.11~)
<HIEOREl lCAl>'" 17 .0 (2.470) 14.1 (2.050) 12.8 (1.860)
*C1uss;·:ai liHoil1atiofl theory
'-
.... -
//,.
33
TABLE 2
Strength (oult) Versus Temperature Data
Lay-up 116K (-250°F) 301K (r~. 'LJ. ,394K (+2S0"F)---- -
[O]a:Xr l,210MPa (175.7 ksi) 1,450 (209.7) 1.570 (228.0)
[45]8 82.9 (12.02) 120.9 (17.54) 78.9 (11.45 )
[90]a: Yr 33.2 (4.815) 28.3 (4.107) 32.6 (4.729)
[+3°2/-3°2]5 294.1 (42.66) 276.5 (40.10) 227.9 (33.05)
[+452/-452]s 101.9 (14.78) 112.5 (15.32) 104.4 (15.14)
[+602/-602]s 45.2 (6.56) 74.3 (l0.77) !i9.0 (8.55)
,/ .,
/I ',I'
34
TABLE 3
Poisson's Ratio (v12 ) versus Temperature Data
,/'
:~,
116K (-250°F)- '
0.4922 0.2979
TABLE 4
0.2098
Shear Modulus (G12) Versus remperature Data
7.85GPa (1.138105i) 6.94(1.007) 5.78 (0.8379)
.1 .,.. .
35
TABLE 5
-0.2753 x 10-4
-0.3993 r. 10-5
0.8674 x 10 n4
0.1258 x 10-4
0.2608 x 10-3
0.3782 x 10-4
'0.5184 x 10-3
0.7519 x 1.0-4
Coet'f1cietlts for Elastic and St.rength Temperatur~ Dependence
2p .. Co + C1, T + Ct T
where P Is the property of interest and T is in kelvin ••
..
Pro.Q.£!1l. ~.D.. h
£1 (Gr>a) 0.18~4 x 103 -0.2713 x 100
E1 (msf) 0.2645 x 102 -0.3935 x lO-1
£2 (GPa) 0.1961 x 102 -0.5429 x 10-1
E2(msi) 0.2844 x 101 -0.7874 x 10-2
"12 0.6092 x 100 -0.1021 x 10-2
i -I
0.7495 x 101 0.6599 x 10-2I Gi2 (GPa)II
I,' GI2 (ms1) 0.1087 x 101 0.9571 x 10-3
,I
X (MPa) 0.1060 x 104 0.1293 x 101T
Xl ("::ii ) 0.1538 x 103 0.1875 x 10°
YT (MPa) 0.4537 x 102 -0.1.'351 x 100
YI (ksi) 0.6580 x 101 -0.1960 x 10-1
, I
36
TABLE 6
f.,ropertl. .£0. h £4'
• 0.4069 x 102 -0.1257 x 10° 10,·3Ext [45J8 (GPa) 0.1552 x
Ex. [45]8 (msi) 0.5901 x 101 -0.1823 x 10-1 0.2251 x 10-4
EXt [+302/-302Js (GPa) 0.9143 x 102 -0.2140 x 10° 0.2773 x 10-3
Ex. [+3°2/-3°2]5 (ms;) 0.1326 x 10~ -0.3103 x 10-1 0.4021 x 10-4
Ex. [+452/-452J5 (GPa) 0.4158 x 102 -0.1023 x 10° 0.1089 x 10-3
Ex, [+452/-452J5 (ms;) 0.6031 x 101 -0.1483 x 10-1 0.1579 x 10-4
.EX' [+6°2/-602]5 (GPa) 0.5008 x 102 -0.2373 x 100 0.3734 x 10-3
Ex. [+6°2/-6°2]5 (msi) 0.7263 y. 101 -e.344J )( 1O~1 0.5416 x 10-4
[45J8
_ (j'lPa)')
0.1191 }{ lOl -0.2364 x 10-2au1t ' -0.2353 x 10(.
llu1 t' [45]8' (ksi) -0.3413 x 101 0.1728 x 10° -0.3429 x 10,·3
<1u1t • [+302/-3°2]5 (~1Pa) 0.2320 x 103 0.5455 x 10° -0.1537 x 10-2
(1ult~r4-"O 1 "0 ] (ks i ) 0.3648 x 102 0.7912 x 0-1 -0.2229 x' in,·3L'" 2 -j 2 s
O'ult • [+45 2/- 452]5 (t1Pa) 0.7709 x 102 0.2747 x 100 -0.5211 x 10-3
Cult' [+452/-4S2Js (ksi) O.llIS x 102 0.3984 x 10-1 ··0.7558 x 10-4
O'ult' [+60 2/-6J2]s (MPa) -0.1335 x 102 0.6392 x 100 -O.l1b6 x 10-2
°Ult' [+602/-602J 5 (ksi ) -0.1936 x 101 0.9270 x 10-1 -0.1677 x 10-3
...
"
II! -,-'
-.:;," _.. :-':'"
concentration, such as at fiber breaks. leading to decreased strength in
This disagreement
Straighter fibers tend to
becomes \'iorse ~lith increasing fiber-·angle, Possibly~ the increase in
brittle. A stiffer' matrix gillE:S rise to increases in stiffness in all
Comparison between experiment,l and theoretical values for laminate
At 10\'/ temperatures, the epoxy matrix material becomes stiff and
residual stresses and less efficient load transfer in regions of stress
orientatiQnsof the ,-;omposite. Strengtl is also increased in all cases
except the 0°-material. A brittle and stiff matrix re:..u1ts in higher
37
The results presented show that the elastic and strength properties
of nOO/5208 graphite·'epoxy com;msite va 1')' either linearly or quadra··
tically with tem?erature. The important properties of longitudinal and
transverse modulus (E1 2lnd E2) show slight increases at both high and
low temperatures. The remaining elastic properties exhibit a decrease
At elevated temperatures, I'educed residual stresses result in lower
4.0 SUMMARY
matrix stresses and str'aignter' fibel"s •
modulus. for the 4')o-lam"inate and the angle-ply laminates. indicates
excellent agreement at room and elevated tei1lperatl!res~ but less than
sati5factory agreerr;.::1t at the 1ow temper.)ture.
fl. and longitudinal strength, XT' are increased wh'l1e all other propel'
ties are affected very 'little.
the fiber direction.
improve the properties of the composite in the fiber direction, but have
little effect in othel' directions. Therefore. 'the lorig1tl1dinal modulus.
with increasing temperature. Strength in tbe fiber-direction is linear
with an increase in strength accompan"ied by an increase in tempera
. ture. All other strength I~asurements exhibit a second-order curve fit.
..:., .
/
•
38
modulus and brittleness in the matrix rr.aterial at lower temperatui'es has
an affect on angle-ply laminates that is not taken into consideration
with the classical iamination theory. t1ol'e work will have to be done in
this area to resolve this question.
ACKNOWLEDGEMENT
This work was supported by the NASA~V'i rg"inia Tech Composites Pro
gram under NASA grant CA-NCGl-15. The authoi"$ would also like to thank
Peter Stein and Doug Wesley for their many hour!: spent assisting in the
1aboratory.
.\
I. ...
t,',.
39
REFERENCES
1. Tenney, D. R•• Sykes, Go F•• and Bowles, O. £., "Space EnvironmentalEffects on Materials," AGARD Conference Proceedings No. 327,Environmental Effects on l-1aterials for Spac~ Applications, paperspresented at the 55th mef:ting of the AGMD Structures and NaterialsPanel in Toronto, Canada, on 19-24 September 1982.
2. Pindera, M. J. and Herakovich. C. T., "An Endochronic Theory forTransversely Isotropic Fiber'ou5 Composites," VPI-E-81-?7, Virginia
. Polytechnic Institute and State University. October 1981.
3. Pinder'a, M. J. and Ht~ral.:ovich, C. T., "An Elastic Potential for theNonlinear Response of Unidirectional Graphite Composites," Journal~elied Mechanic~ Vol. xx.
4. Hyer', H. W., Herakovich, C. T., Hilkovich, S. H., and Short, J. S.,"Temperature Dependence of Mechanical and lhe,'mal Expansion Properties of 1300/5203 Graphitl:!-Epoxy,1I _Gom!?.2.~jtes. Volume 14. July 1983,pp. 276 ..280.
5. Curtis, G. J •• Milne, .1. M., and Reynolds. \4. N.. "Non-HookeanBehavior of Strong Carbon Fiber'.;," Natu~-, Volum~ 220. December1968, po 1024•
6. Beetz, C. P., Jr., "Stro.in-Induced Stiffenlng or Carbon Fibres,"llbre ';;,£ien£.u.nd TecnnolQ.9Y.;. Volume 16, 198~.
7. Mansfield, Eo H. and Purslow, D., "The Influence of Fiber Wav~ness
on the Moduli of Unidirection Fib(;~i" Reinforced Composites." CP No.1339, December 1974.
8. Bazant. Z. P•• "Effect of Curvature of the Reinforci~g Fibers on theHoduli of Elasticity and Strength of Composites," PO·f..Lr!Ier Mechi0..1.s.?.,Volume 4, Number 2, 1968. pp. 251-258.
9. VanDreumel. W. ~l. M•• and Kamp, J. l. r1., "Non-Hookean Behavior inthe Fibre Direction of Carbon-Fibre Composttes and :he Influence ofFibre ~Javiness on the Tensile Properties." ,J. Composite Materials,VOlume ll. October 1977. p. 461. . --------.