Deformation and Fracture Toughness in High-Performance Polymers

5

Deformation and Fracture Toughness in High-Performance Polymers Comparative Study of Crystallinity and Cross-Linking Effects Ruth H . Pater Mark D. Soucek , and Bor Z. Jang 1,

1

2

NASA Langley Research Center, Mail Stop 226, Hampton, VA 23665-5225 Materials Science and Engineering, 201 Ross Hall, Auburn University, Auburn, AL 36849 1 2

A systematic study was made of 10 principal thermoplastics

and

two

semiinterpenetrating

high-performance polymer

networks

(semi-IPNs). The fundamental tendency to undergo localized crazing or shear banding, as opposed to a more diffuse homogeneous shear-yielding deformation, was evaluated. Amorphous thermoplastics exhibited crazing as the primary mode of deformation. In contrast, semi crystalline materials displayed both crazing and shear banding. Increasing the crystallinity increased diffuse shear yielding at the expense of craze growth. Another effect was an enlargement of the deformation zone. Some ordered polymers showed only diffuse shear yielding, whereas others displayed a combination of weak crazes and diffuse shear yielding.

For a semi-IPN, increasing the degree of

cross-linking decreased crazing, deformation zone size, and fracture toughness of an amorphous thermoplastic. Thus, crystallinity acts like cross-linking in reducing crazing, but, exerts the opposite effect on changing the size of the deformation zone. These results suggest that the reduction in fracture toughness by crystallinity is mainly due to decreased crazing, whereas reduction by cross-linking arises from both decreased crazing and diminished deformation zone.

HIGH-PERFORMANCE THERMOPLASTICS ARE EMERGING as an important class of engineering materials. Thermoplastics are being considered for use as 0065-2393/93/0233-0105$ 10.25/0 © 1993 American Chemical Society

106

RUBBER-TOUGHENED PLASTICS

composite matrices, adhesives, molded articles, films, and coatings for a wide variety of aerospace structural, electronic-electrical, and automotive applications. Most of these applications utilize the materials' excellent resistance to impact and microcrack. The majority of these polymers are semicrystalline in nature. It is of theoretical and technological importance to understand the relation between fracture toughness and crystallinity in these polymers. Extensive studies of polyether ether ketone ( P E E K ) have shown that fracture toughness strongly depends on crystalhnity. Lee et al. (1) showed that the mode-I fracture toughness (K ) of P E E K 150P decreased by almost a factor of 3 with a crystallinity increase from 27 to 43 wt % . Similarly, Friedrich, et. al. (2) showed a 40% reduction in K for a crystalhnity change from 15 to 35% in P E E K 450G. C h u s study (3) related decreased molecular weight, larger spherulites, and more perfect crystal lamellae to a reduction in fracture toughness. Currently, P E E K is the only high-performance thermoplastic that has received considerable attention. Whether the same relation holds for other high performance semicrystalline polymers remains to be seen. Why does an inverse relation exist between fracture toughness and crystalhnity? Despite a theme variation, earlier studies offer no explanation for the relation. Evaluation of crystalhnity and deformation may answer the question and provide additional knowledge that would have important implications for designers, users, and manufacturers of these polymers. Like crystalhnity, cross-linking significantly affects fracture toughness of a toughened polymer, such as a semiinterpenetrating polymer network (semiIPN). The origin of the inverse relationship between fracture toughness and cross-linking is unclear. Thus, there also exists a need to understand the role of cross-linking i n deformation and fracture mechanisms i n a toughened polymer. Also of great interest is the comparison of crystalhnity and cross-linking effects on the fracture mechanisms, because such a comparative study under well-controlled conditions has not been undertaken for high-performance polymers. Crazing and fracture toughness are closely related. Crazing and shear banding are the most important deformation mechanisms in glassy thermoplastics (4, 5). Another deformation mode is diffuse shear yielding (4, 5). Whether crazing or shear banding can be a primary mode of plastic deformation in a highly cross-linked thermoset resin, such as epoxies, remains controversial. However, both crazing and shear banding have been suggested as the dominant energy-dissipating mechanisms in rubber-toughened thermoset resins (4, 6-8), although doubts about these claims have been raised (5, 9-12). Other proposed mechanisms include particle deformation and crack bridging (9, 10, 13), cavitation-induced shear deformation or stress relief (12, 14-21), crack pinning (22-25), and crack-tip blunting (26). Several review articles have been published on this subject (4, 5, 9, 12, 16, 26, 27). lc

I c

5.

PATER ET AL.

107

Deformation and Toughness in Polymers

The materials selected for this study include 10 principal highperformance thermoplastics that are under experimental and developmental evaluation and are commercially available. The chemical structures of the thermoplastics are shown in Chart I, except for the structure of New TPI, which remains undisclosed for proprietary reasons. Table I lists the neat resin thermal and fracture-toughness properties of the thermoplastics as reported in the cited literature. The materials chosen for study cover a broad area of thermoplastic chemistry varying from polyCarylene ether ketone) to polysulfone to polyimide, with variations in their connecting groups. Some of the materials are related to each other as structural isomers. A wide spectrum of crystalline morphology is represented, ranging from a very high degree of crystalhnity to a totally amorphous structure. Thus, these materials present a good opportunity for a systematic study of crystalhnity and deformation. Included in this study are two newly developed semi-IPNs called LaRC-RP41 and LaRC-RP40. Schemes I and II show their syntheses. By combining easy-to-process but brittle P M R - 1 5 with tough but difficult-to-process L a R C TPI or N R - 1 5 0 B , the resultant semi-IPNs exhibit an attractive combination of properties including significantly improved fracture toughness over the unmodified thermoset resin (fracture energy G 476 vs. 87 J / m in L a R C RP41, for example) (28-30). Changing the composition allows the crosslinking and fracture toughness to be varied and, thus provides an opportunity to study the role of cross-linking in plastic deformation and fracture tough ness in a semi-IPN environment. 2

2

I c

Experimental Details Materials. P E E K (100 μπι), New T P I (20 μπι), polyaryl sulfone (PAS, 80 μπι), polyether sulfone (PES, 140 μπι), and Upilex (80 μπι) films were commercially obtained and used as received. The film thickness is indicated in the parentheses. L a R C - C P I (30 μπι), LaRC-I-TPI (40 μπι), and polyimide sulfone ( P I S 0 , 80 μπι) films were previously prepared and are reported in references 31, 32, and 33, respectively. The films of L a R C - T P I , N R - 1 5 0 B , and their semi-IPNs were prepared in this study as follows: The L a R C - T P I polyamic acid solution, which had a 30-wt % solids concentration in N, IV-dimethylacetamide ( D M A c ) , was purchased from Mitsui Toatsu Chemicals, Chiyoda-Ku, Tokyo, Japan. The NR-150B monomer precursor solution with 54-wt % solids in iV-methylpyrrolidone ( N M P ) was supplied by D u Pont. The P M R - 1 5 molding powder used to cast films was obtained from the monomethyl ester of 5-norbornene-2,3-dicarboxyhc acid (NE), 4,4'-methylenedianihne ( M D A ) , and dimethyl ester of 3,3',4,4'-benzophenonetetracarboxylic acid ( B T D E ) . A 50-wt % methanol solution of the monomer mixture at a molar ratio of N E : M D A : B T D E = 2.00:3.09:2.09 was stirred at room temperature for 0.5 h to give a clear dark-brown solution. This solution 2

2

2


m jor°iaojc

1

t!

Ο

Polyetheretherketone (PEEK)

LaRC-CPi

•^?SLfi^ isriSr" M r

c

LaRC-TPI ο

ο

il

ii

Ο

ο

ii

ii

Ν-

LaRC-1-TPI

Ο ii

I

CF

ii ο

3

3

NR-150B2

Polysulfone (PS)

Ο Polyethersulfone (PES)

M

4 ^ï^|>TSrî-©f Polyimidesulfone (PIS0 ) 2

Ο

Ο

c

c Upilex S

Chart I. Chemical structures of Thermoplastics.

5.

PATER ET AL.


109

Table I. Thermal and Fracture-Toughness Properties of High-Performance Thermoplastics Polymer PEEK LaRC-CPI LaRC-TPI LaRC-I-TPI New TPI NR-150B PES (RadelA400) PES(Vietrex4100G) 2

PISO

2

Upilex S

Supplier ICI NASA-Langley Mitsui Toatsu NASA-Langley Mitsui Toatsu Du Pont Amoco ICI HiTech Services ICI/Ube Chem.

G (j/m )

T (°C) 143 222 252 259 250 360 220 230 273 328

2

Ic

g

343 350 348 none 388 none none none none

—

4025 6620 1768

— —

2398 3500 1925 1400

—

Ref. 37 31 41 32 42 42 37 37 37 43

NOTE: The dashes indicate that data were unavailable.

was concentrated at 80 °C in a nitrogen atmosphere for 2 h and further dried at 150 °C in air for 1.5 h. A resin solution with 20-wt % solids concentration i n Ν M P was centrifuged and the decantate was cast onto a glass plate by using an 80 μ m doctor blade and dried in a well-ventilated and dust-free chamber for 48 h to remove excess solvent. The film was thermally converted into the polyimide by heating at 80, 120, 150, 210, 250, 300, and 350 °C for 1 h at each temperature in air and then cooled to room temperature at a very slow rate (2-3 °C/min) to reduce residual thermal stress. The film was removed from the plate by soaking in water overnight. The thickness of the obtained film varied from 20 to 50 μπι.

Film Characterization. Rectangular 13-mm-wide strips were cut from each material with a pair of sharp scissors. Each strip specimen was supported on fight-weight cardboard and cut with a razor blade to create the edge notch. This procedure gave a crack that appeared sharp under the microscope (34). A plane-stress fracture toughness test was conducted on the rectangular specimens with a single-edge-notch (SEN) geometry. A microscope stage-top miniature testing system equipped with a 25-lb (100-N) load cell and a variable speed motor was used to deform the specimen in a tensile mode. The load-displacement curve of each test was recorded on a chart recorder. The crack-tip region was observed at 110 X or 220 X , and the onset of the crack propagation was noted and marked on the chart record. This visual determination technique gave an acceptable degree of consistency (34). The deformation zone near a crack tip could be observed i n situ continuously or intermittently at various stages of loading. Photomicrographs were obtained by stopping the cross-head at various extents of plastic

ο

HO

0

OH

i|

Scheme I. Synthesis of semi-IPN LaRC-RP4L

Linear LaRC-TPI

jj

Crosslinking PMR-15 +

LaRC-RP41 Semi-2 IPN

II

υ n

/ -i087

Ί0"ΝΗ 2

Scheme IL Synthesis of semi-IPN LaRC-RP40.

Linear NR-150B2

95 mole % para, 5 mole % meta

2

HN

Crosslinking PMR-15 +

0

Simultaneous Seml-IPN

LaRC-RP40

112


deformation. Deformation zones of select samples were also further exam ined on a scanning electron microscope ( S E M ) . The plane-stress fracture toughness (or stress-intensity factor, K ) was obtained from eq 1 in reference 35: I c

Y =

[3.94(2w/Tta)tsm(>na/2w)]

1/2

S = F/[(w

-

a)h]

where w is the specimen width, b is the specimen thickness, a is the notch depth, F is the load at which crack propagation begins, and Y is a geometric factor. The differential scanning calorimetry (DSC) analysis was carried out with 2-10-mg film samples on a D u Pont Model 940 calorimeter at a heating rate of 5 °C/min. From the D S C thermogram, the apparent glass-transition temperature (T ), crystalhzation temperature (T ), and melting temperature of a crystalline phase ( T ) were determined. The wide-angle X-ray diffraction was measured on films with the X-ray generator operated at 45 k V and 40 m H . The intensity of counts taken every 0.01° (2Θ) was recorded on hard disk for the angular range 10-40° (2Θ). Typical intensities ranged between 600 to 2500 counts/s. g

c

m

Results and Discussion Crystallinity Effects. PEEK. Figure 1 illustrates the D S C curves for the three samples of P E E K . The as-received material exhibited a T at 149 °C, a T at 178 °C associated with crystalhzation, and a T at 336 °C corresponding to the melting of the crystals. The heat of fusion ( Δ Η ) was 32.9 J/g. Annealing this sample at a temperature (200 °C) above its T for 1.5 h enhanced AH to a value of 37.2 J/g. However, annealing at a temperature (325 °C) slightly below its T for 2 h had a more pronounced effect on Δ H as well as on the crystalline structure: the AH value increased to 45.0 J/g and the crystals became more perfect, as indicated by the increased sharpness of the T peak. Blundell and Osborn (36) reported a value of 130 J/g for AH for the 100% crystalline P E E K . By using this value, the degree of crystalhnity for these three samples was estimated to be 25, 29, and 35%, respectively. The X-ray diffraction data shown in Figure 2 confirm this trend. Figure 3 shows the deformation behavior of the three samples viewed through a polarizing light microscope. When loaded i n a tensile mode, the as-received specimen showed crazes at the early stage of loading (Figure 3a). Continued loading turned the deformation mode into one of a more localized nature. Both crazes and shear bands dominated the crack-tip deformation zone. Crazes appeared to grow at a slightly faster rate than shear bands and g

c

m

£

g

{

m

{

{

m

{

5.

PATER ET AL.


113

178°C As received 200°C/1.5 hrs annealed 325°C/2 hrs annealed

EXO A

-^220°Cf

Relative heat flow

A

— - - J 3 6 J O \ jf _45._0_J/g

j

1

u

φ

158°C

336°C \i!

PI

i! Endo

332°C 100

200

300

J

400

Temperature, °C Figure 1. DSC scans of PEEK could be observed at the leading edge of the growing deformation zone. At a higher magnification (Figure 4), two packets of shear bands intersecting each other at an angle of 30-60° were observed. Comparison of Figure 3a with Figures 3b and 3c readily demonstrates that increasing the crystalhnity dramatically decreases crazing and shear banding but increases diffuse shear yielding. Thus, crystalhnity caused a transition in the deformation mechanism from a localized to a more homogeneous mode. Another effect is an enlarge ment of the deformation zone as shown in Figure 5. The deformation zone ahead of the crack tip as well as the fracture surfaces near the crack-tip regions were affected by the crystallinity as illustrated in Figure 6. LaRC-CPI. Compared to P E E K , L a R C - C P I has a higher T (226 °C vs. 149 °C), a higher T (350 °C vs. 336 °C), and, more importantly, a higher g

m

114


5

12

19 26 2Θ, degree

33

40

Figure 2. X-ray diffractograms of PEEK

Figure 3. Polarized optical micrographs of the crack-tip deformation zones in PEEK

5.

PATER ET AL.


Figure 4. Higher magnification view of Figure 3a.

Figure 5. SEM micrographs of the deformation zones in PEEK.

115

116


Figure 6. SEM micrographs of the fracture surfaces near the crack-tip regions in PEEK

5.

PATER ET AL.


117

G [6620 J / m (31) compared to 4025 J / m (37)]. The X-ray diffraction patterns shown in Figure 7 suggest that the as-received film as well as the annealed samples had a high degree of crystalhnity. The D S C curves for the three L a R C - C P I film samples are illustrated in Figure 8. Annealing the as-received sample at 200 °C for 1.5 h increased the crystalhnity somewhat ( Δ Η increased from 17.8 to 18.5 J/g) and induced more perfect crystals, as evident from the increased sharpness of the T peak. A pronounced increase in the crystalhnity was obtained when the as-received material was annealed at a temperature of 327 °C, close to its T . The as-received film showed crazes extending a considerable distance ahead of the crack tip (shown in Figure 9). Comparison of Figure 10a with Figure 10c shows that the crystalhnity in L a R C - C P I exerted the same effects seen in the previously discussed P E E K material. The apparent crazes were dramatically reduced, and the deformation zone enlarged as the crystalhnity increased. Diffuse shear yielding became the primary mode of deformation for the sample with the highest crystalhnity. 2

I c

2

£

m

m

LaRC-TPI. The crystalline behavior of the chemically imidized L a R C T P I 1500 series was extensively investigated by H o u and Bai (38), who showed, as indicated in Figure 11, that the area under the T peak increased as the area under the T peak decreased with increasing annealing time. H o u and Bai also indicated that, unlike P E E K , a L a R C - T P I polymer is not readily recrystallized after melting the initial crystalline phase. Annealing a L a R C - T P I m

g

J

5

12

1

1

1

ι

19 26 2Θ, degree

λ

ι 33

ι

ι 40

Figure 7. X-ray diffractograms of LaRC-CPI.

118


A s received 200°C/1.5 hrs annealed

100

200

300

Temperature, °C Figure 8. DSC scans of LaRC-CPI

specimen at temperatures above 320 °C resulted in a fully amorphous structure as determined by D S C . Similarly, thermally imidizing L a R C - T P I also results in a fully or nearly fully amorphous material, as evidenced from the following results. Three film samples of the thermally imidized L a R C - T P I were prepared in this study. One film was obtained by treating the 316 °C cured material at 200 °C for 1.5 h. The second film was slowly cured at 80, 120, 150, 210, 250, 300, and 350 °C for 1 h at each temperature. The details of the preparation were described in the Experimental Details section. The third sample, designated 316 °C cured material, was prepared by heating the film at 100, 200, and 316 °C for 1 h at each temperature. Figure 12 shows the D S C traces for these three samples, all of which exhibited a T at temperatures of 255-262 °C, but no T (327 and 341 °C from reference 38) could be found. The absence of a semicrystalline structure is also confirmed by the X-ray data shown in Figure 13. Because there is a variation in the area under the T peak (see ΔΗ values in Figure 12) and considering the results of the study by H o u and Bai (38) discussed earlier, it is possible that the three samples g

m

g

{

5.

PATER ET AL.


119

Figure 9. Polarized optical micrographs of the deformation zone in as-received LaRC-CPI contained crystals that were too small or too few in number to be detected by both the D S C and X-ray diffraction measurements, but that do exert effects on the deformation behavior. Figure 14 illustrates that the 350 °C cured sample has crazes and that shear bands dominate the crack-tip deformation zone. A closer examination of the right and left arm regions as indicated by arrows (Figures 14a and 14c) reveals the presence of two packets of shear bands intersecting at an angle of 30-60° to each other. The shear-banding phenomenon was quite pronounced in this sample. The SE M micrographs of Figure 15 compare the craze-growth behavior of the three films. The 200 °C annealed specimen has many crazes that grew to a great length without diminishing their width (Figure 15a). The 350 °C cured sample (Figure 15b) also shows well-developed crazes that display smaller width and length than in the previous sample. In comparison to the other two samples, the 316 °C cured material has the least amount of crazes but considerably more shear bands. LaRC-I-TPI. LaRC-I-TPI is a structural isomer of L a R C - T P I . A comparison of their chemical structures can be made from Chart I. Both materials behave alike in many aspects of their physical and mechanical properties that have been compared in a study by Pratt and St. Clair (32). The thermally imidized film of LaRC-I-TPI shows a T at 251 °C (Figure 16), which is lower than a reported value (32) of 259 °C. The absence of a semicrystalline structure is evident from the D S C (Figure 16) and the X-ray data (Figure 17). This amorphous polymer displays deformation behavior g

120


Figure 10. SEM micrographs of the crack-tip deformation zones in LaRC-CPI.

5.

PATER ET AL.

121


I 180

ι

ι

220

260

ι 300

.j_

ι

340

380

T(°C)

Figure 11. DSC scans of LaRC-TPI 1500 series samples annealed at 310 °C for various times. (Reproduced with permission from reference 38. Copynght 1990.) (Figure 18a) analogous to the 350 °C cured L a R C - T P I sample (Figure 14). Both crazes and shear bands dominate the crack-tip deformation zone as shown at a higher magnification in Figure 18b. New TPI . H o u and Reddy (39) reported that the commercially obtained New T P I powder had no detectable T but two T s at 354 and 384 °C with a A H value of 46-49 J/g. They also showed five reflections in the X-ray diffraction pattern. The present New TPI film (also obtained from a commercial source) shows a T at 260 °C, a T at 308 °C, and a T at 383 °C (Figure 19). A T of 250 °C and a T of 388 °C have been reported (Table I). The X-ray pattern shown in Figure 20 suggests an amorphous structure for 3

g

m

f

g

g

c

m

m

New is part of the name of a specific compound. For proprietary reasons the specific compound cannot be divulged.

3


Ο Χ

'

200°C/1.5 hrs annealed 350°C cured 316°C cured / I

K

—

—

2

-

4

J

/

g

4,

% c ω

y

2.2 J/g 1.8 J/g

Τ I 100

.

255°C

J 200

1 300

400

Temperature, °C Figure 12. DSC scans of LaRC-TPI.

I

5

·

1

12

1

1

1

I

19 26 2Θ, degree

ι

ι

33

•

ι

40

Figure 13. X-ray diffractograms of LaRC-TPI.

5.

PATER ET AL.


123

As received LaRC-TPI Figure 14. Polarized optical micrographs of 350 °C cured LaRC-TPI. Parts a and c are higher magnification views of the regions indicated by arrows. the present New TPI. The deformation and fracture behavior of this material (Figure 21) resembles that of L a R C - T P I and LaRC-I-TPI. A higher magnification view of one of the arms (Figure 22) illustrates the presence of well-developed localized crazes and shear bands that grew to a considerable length without diminishing their widths. Polyimide Sulfone. The polyimide sulfone ( P I S 0 ) film was fully amorphous as evidenced by the D S C data shown in Figure 23, and its T occurred at 271 °C, which is close to a value of 273 °C reported by St. Clair and Yamaki (33). Among the 10 thermoplastics studied, the present P I S 0 2

g

2

124


Figure 15. SEM micrographs of the crack-tip deformation zones in LaRC-TPI.

Ο χ w

χ CO

φ X

< 251 °C ο

"D C LU

JL

200

300 Temperature, °C

Figure 16. DSC scan of LaRC-I-TPI.

400

5.

PATER ET AL.


I

5

»

1

12

1

1

ι

I

19 26 2Θ, degree

ι

I

33

125

ι

40

Figure 17. X-ray diffractogram of LaRC-I-TPI.

Figure 18. Polarized optical micrographs of the crack-tip deformation zone in LaRC-I-TPI and b, crazes near the leading edge of a grown deformation zone

126


308°C

200

300 Temperature, °C Figure 19. DSC scan of New TPI.

I

5

1

1

1

1

12

.

I

19 26 2Θ, degree

ι

33

•

»

40

Figure 20. X-ray diffractogram of New TPI. exhibits the most pronounced crazes, which are very wide and extend to long distances, as can readily be seen in Figures 24 and 25. NR-1S0B . A n NR-150B polymer is generally thought of as an amor phous polymer, because of the presence of a bulky hexafluoroisopropylidene 2

2

5.

PATER ET AL.


127

Figure 21. Polarized optical micrographs of the crack-tip deformation zone New TPI.

Figure 22. SEM micrographs of the deformation zone showing shear bands in New TPI.

128


200

300

400

Temperature, °C Figure 23. DSC scan of polyimide sulfone.

Figure 24. Polarized optical micrographs of the crack-tip deformation zone in polyimide sulfone at different magnifications. group that tends to prevent the formation of a crystalline structure. The D S C scan of the prepared NR-150R film shows a sharp thermal transition peak at 353 °C (Figure 26). However, the X-ray diffraction pattern shown in Figure 27 clearly indicates that some ordered structure, if not a semicrystalline structure, is present. Possibly a thermal transition due to crystal melting occurs at a temperature above its decomposition temperature and, thus, 2

5.

PATER ET AL.

Déformation and Toughness in Polymers

129

Figure 25. SEM micrograph of the deformation zone in polyimide sulfone showing extended crazes.

200

300 Temperature, °C

400

Figure 26. DSC scan of NR-150B . 2

cannot be observed in the D S C scan. The peak at 353 °C is most likely due to a T , because a T of 360 °C has been reported (Table I). Figure 28 illustrates diffuse shear yielding as the principal mode of deformation. The deformation zone initiated with two primary arms, each at an angle of approximately 45° with respect to the crack-tip propagation direction. These two arms gradually reoriented almost parallel to the tip growth direction. At the root of the crack-tip, only a heavily deformed shear-yielding zone is observed. No crazes g

g

130


Figure 28. Polarized optical micrographs of the crack-tip deformation zone in NR-150B . 9

5.

PATER ET AL.


131

or shear bands are discernible even at a higher magnification (Figure 29). Most likely the NR-150B material had a high entanglement density, which is responsible for the diffuse shear yielding. 2

Polyaryl Sulfone (PAS). The commercial polyaryl sulfone film ex hibits multiple thermal transitions in the D S C curve shown in Figure 30. In conjunction with the X-ray data of Figure 31, the peaks at 194 and 230 °C may be associated with T s, whereas the peaks occurring at 331, 403, and 470 °C are associated with T s. A T of 220 °C was reported (Table I). The transition peak at 331 °C has a very small Δ iff value (0.4 J/g) and its assignment as a T , rather than a T , is consistent with the chemical structure shown in Chart 1. A combination of shear banding and diffuse shear yielding characterizes the deformation behavior as readily seen i n Figure 32, particu larly at higher magnification (Figure 32b). g

m

m

g

g

Polyeiher Sulfone (PES). Analogous to the behavior of PAS, the commercially obtained polyether sulfone (PES) film showed multiple thermal transitions in the D S C trace (Figure 33) and an ordered structure in X-ray

Figure 29. SEM micrograph of the defor mation zone in NR-150B . 9

132


\

5

1

1

12

1

ι 19 26 2Θ, degree i

ι

ι

ι

33

•

ι

40

Figure 31. X-ray diffraetogram of polyaryl sulfone showing ordered morphology. diffraction (Figure 34). The peak at 235 °C can be assigned to a T , whereas the peaks at temperatures above 400 °C are most hkely caused by different forms of the crystalline structure. As shown in Figure 35, the crack-tip deformation zone shows poorly developed crazes that are short and thin. A considerable amount of homogenous diffuse shear yielding is also observed. The inability to grow extended crazes is seen in Figure 36. One possible explanation for this behavior is that a large number of small crystallites act like "cross-links" to prevent molecules from becoming highly oriented, a condition necessary for crazing. g

PATER ET AL.


Figure 32. Polarized optical micrographs of the deformation zone in sulfone at different magnifications.

200

300 400 Temperature, °C

500

Figure 33. DSC scan of polyether sulfone.

134


I

5

I

ι

I

12

'

•

»

19 26 2Θ, degree

•

»

33

•

'

40

Figure 34. X-ray diffractogram of polyether sulfone showing ordered morphology. Upilex. The Upilex sample had two T s at 230 and 333 °C and a T at 403 °C (Figure 37). A T of 328 °C has been reported (Table I). The crystalline structure of Upilex is confirmed by the X-ray data that show at least six reflections (Figure 38). Upilex shows the same deformation behavior as the NR-150B specimen presented earlier. As shown in Figure 39, its deformation is dominated by diffuse shear yielding without any traces of crazing or shear banding. The similarity in the fracture behavior between the NR-150B and Upilex materials is unexpected because of the difference in the degree of rigidity in their molecular structures. The NR-150B contains a hexafluoroisopropylidene connecting group, which acts as a flexibilizing unit, whereas the Upilex has a stiff rigid biphenyl bridge between rigid imide rings. The Upilex has a greater probability of forming a crystalline phase. Further study is needed to clarify this point. g

m

g

2

2

2

Cross-Linking Effects. Figure 40 illustrates the deformation zone patterns for the series of semi-IPNs prepared from P M R - 1 5 and L a R C - T P I (LaRC-RP41). Another series of the semi-IPNs comprising PMR-15 and N R - 1 5 0 B (LaRC-RP40) was also studied. Because many variables, such as processing conditions, specimen thermal history, testing conditions, and temperature, affect the results, a comparative study was made under carefully controlled conditions. A l l the specimens were prepared and tested under identical conditions as described in the Experimental Details section. Increas ing the thermosetting component concentration probably increases the crossfink density in the semi-IPN system. Comparison of Figure 40a with Figures 40b-40f shows that an increase in cross-finking resulted in a size reduction of 2

5.

PATER ET AL.


135

Figure 35. Polarized optical micrographs of the deformation zone in polyethe sulfone at different magnifications.

Figure 36. SEM micrograph of the deformation zone in polyether sulfone.

136


200

300 400 Temperature, °C

500

Figure 37. DSC scan of Upilex.

I

5

1

1

12

ι

I

ι

I

19 26 2Θ, degree

ι

»

•

33

»

40

Figure 38. X-ray diffractogram of Upilex. the deformation zone as well as a shift from primarily crazing deformation to only diffuse shear-yielding deformation as the weight percentage of L a R C - T P I to PMR-15 varied from 100 to 0%. A dramatic decrease in the deformation zone size occurred with the composition containing 35 and 65 wt % of L a R C - T P I and PMR-15, respectively. With PMR-15 as the matrix and L a R C - T P I as the dispersed phase, diffuse shear yielding was the primary mode of plastic deformation as shown in Figures 40d-40f. Conversely, with L a R C - T P I as the matrix, both crazing and shear yielding were still the major mechanisms for energy dissipation in the semi-IPN (Figures 40a-40c). However, when L a R C - T P I was replaced by N R - 1 5 0 B , the resultant semi2

5.

PATER ET AL.

137


Figure 39. Polarized optical micrograph of the deformation zone in Upilex. I P N films exhibited only diffuse shear yielding. This diffuse shear-yielding behavior is understandable because both the pure N R - 1 5 0 B and P M R - 1 5 components showed only diffuse shear yielding. The K values for pure PMR-15, L a R C - T P I , and NR-150B and their semi-IPNs are plotted in Figures 41 and 42. The low K value for P M R - 1 5 (0.2 M P a · m ) shows that P M R - 1 5 is a very brittle material because of its high cross-link density. By incorporating 25 wt % of a thermoplastic component (either L a R C - T P I or NR-150B ), the K value of PMR-15 can be more than doubled. The pure thermoplastic films are considerably tougher than PMR-15 and their semi-IPN counterparts, as expected. Although L a R C TPI is tougher than NR-150B in thin film form, the NR-150B is more effective in toughening PMR-15. This situation may be due to the difference in phase morphology that facilitated operation of different toughening mechanisms or to the same mechanisms but to a different extent. 2

2

lc

I c

1 / 2

2

2

lc

2

Summary High-performance thermoplastics exhibit three modes of plastic deformation: crazing, shear banding, and diffuse shear yielding. Semicrystalline polymers, such as P E E K and L a R C - C P I , show a combination of crazing and shear banding. When crystalhnity is increased through annealing, diffuse shear yielding increases at the expense of craze growth. This process is accompa-

40. Polarized optical micrographs showing the deformation zone variation in LaRC-TPI with increasing PMR-15 concentration in semi-IPN LaRC-RP41.

5.

PATER ET AL.

139


Figure 42. Fracture toughness as a function of NR-150B concentration in Semi-IPN LaRC-RP40. 2

140


nied by enlargement of the deformation zone. For fully amorphous materials, like L a R C - T P I , LaRC-I-TPI, New TPI, and polyimide sulfone, crazing is the primary energy dissipation mechanism, but some shear bands are also found. Some of the ordered polymers, namely, NR-150B and Upilex, show only diffuse shear yielding, whereas others, such as polyaryl sulfone and polyether sulfone, display weak crazing (short and thin crazes) combined with diffuse shear yielding. Adding a small amount (25 wt %) of a thermoplastic component, either L a R C - T P I or N R - 1 5 0 B , significantly improves the fracture toughness of highly cross-finked and brittle P M R - 1 5 polyimide (40). Pure PMR-15 film shows only diffuse shear yielding and has a very small deformation zone. This behavior is consistent with the limited extensibility of a highly cross-finked network structure. Increasing the cross-finking by increasing the thermoset component concentration causes a reduction in the deformation zone, dimin ished craze growth, and a shift from crazing to diffuse shear yielding in semi-IPNs of L a R C - T P I . Variations in fracture toughness are correlated with changes in both deformation mode and zone in the semi-IPN systems. The cross-linking plays a role similar to that of crystalhnity in retarding craze growth and in promoting diffuse shear yielding. However, cross-linking and crystalhnity exert opposite effects on the size of the deformation zone. The inverse relation between the crystallinity and fracture toughness is most likely due to decreased crazing, whereas the reduction in fracture toughness by cross-finking may arise from both decreased crazing and a diminished defor mation zone. 2

2

References 1. Lee, W. I.; Talbott, M. F.; Springer, G. S.; Berglund, L. A. J. Reinf. Plast.

Compos. 1987, 6, 2-12.

2. Friedrich, Κ.; Carlsson, L. Α.; Gillespie, J. W., Jr.; Karger-Kocsis, J. In Thermo plastic Composite Materials; Carlsson, L. Α., Ed.; Elsevier Science Publishers, Β. V.: New York, 1991; p 242. 3. Chu, J.-N. M.Sc. Thesis, University of Delaware, Newark, D E , 1988. 4. Bucknall, C. B. Toughened Plastics; Applied Science Publishers: London, 1977. 5. Kinloch, A. J.; Young, R. J. Fracture Behavior of Polymers; Applied Science Publishers: New York, 1983.

6. Sultan, J. N.; McGarry, F. J. Appl. Polym. Symp. 1971, 16, 127.

7. Sultan, J. N.; McGarry, F. J. J. Polym. Sci. 1973, 13, 29. 8. Bucknall, C. B.; Yoshii, T. Br. Polym. J. 1978, 10, 53. 9. Kunz-Douglass, S.; Beaumont, P. W. R.; Ashby, M . F. J. Mater. Sci. 1980, 1109. 10. Kunz, S.; Beaumont, P. W. R. J. Mater. Sci. 1981, 16, 3141. 11. Kunz, S. Ph.D. Thesis, University of Cambridge, 1978.

15,

12. Yee, A. F.; Pearson, R. A. In Fractography and Failure Mechanisms; Roulin -Moloney, A. C., Ed.; Elsevier Applied Science Publishers: London, 1989; p 291.

13. Evans, A. G. et al. Acta Metall. 1986, 34, 79-87.

14. Bascom, W. D. et al. J. Mater. Sci. 1981, 16, 2657.

5.

PATER ET AL.

15. 16. 17. 18. 19.

Bitner, J. R. et. al. J. Adhes. 1982, 13, 3. Donald, A. M.; Kramer, E. J. J. Mater. Sci. 1982, 17, 1765. Hagerman, Ε. M . J. Appl. Polym. Sci. 1973, 17, 2203. Yee, A. F. J. Mater. Sci. 1977, 12(8), 757. Yee, A. F. et al. In Toughened and Brittleness of Plastics; Deanin, R. D.; Grugnola, A. M., Eds.; American Chemical Society: Washington, D C , 1976; p 97. Kinloch, A. J.; Williams, G. J. J. Mater. Sci. 1980, 15, 987. Bascom, W. D.; Cottington, R. L. J. Adhes. 1976, 7, 333. Lange, F. F. Phil. Mag. 1970, 22, 983. Evans, A. G. Phil. Mag. 1972, 26, 1327. Evans, A. G. J. Mater. Sci. 1974, 9, 1145. Green, D. J.; Nicholson, P. S. J. Mater. Sci. 1979, 14, 1413. Cartwell, W. J.; Roulin-Moloney, A. C. In Fractography and Failure Mechanisms; Roulin-Moloney, A. C., Ed.; Elsevier Applied Science Publishers: London, 1989; pp 233-290. Garg, A. G.; Mai, Y. W. Comp. Sci. Technol. 1988, 31(3), 179-223. Pater, R. H . Polym. Eng. Sci. 1991, 31(1), 20-27. Pater, R. H.; Morgan, C. D. SAMPE J. 1988, 24(5), 25-32. Pater, R. H . In Encyclopedia of Composites; Lee, S., Ed.; V C H Publishers: New York, 1990; Vol. 2, pp 377-401. Hergenrother, P. M.; Haven, S. J. SAMPE J. 1988, 24(4), 13-18. Pratt, J. R.; St. Clair, T. L. SAMPE J. 1990, 26(6), 29-38. St. Clair, T. L.; Yamaki, D. A. In Polyimides; Mittal, K. L., Ed.; Plenum Press: New York, 1984; pp 99-116. Hinkley, J. Α.; Mings, S. L. Polymer 1990, 31, 75-77. Williams, J. G. Fracture Mechanics of Polymers; Applied Science Publishers: London, 1987; p 64. Blundell, D. J.; Osborn, Β. N. Polymer 1983, 24(8), 953-958. Johnston, N. J.; Hergenrother, P. M . Proc. 32nd SAMPE Symp. 1987, 1400-1412. Hou, T. H.; Bai, J. M . Proc. 35th SAMPE Symp., 1990, 35, 1594-1608. Hou, T. H . ; Reddy, R. M . Report No. CR 187445, National Aeronautics and Space Administration: Washington, D C , 1990. Jang, B. Z.; Pater, R. H; Soucek, M . D.; Hinkley, J. A. J. Polym. Sci. Polym. Phys. Ed., 1992, 30. Johnston, N. J. Presented at the 4th Industry, Government Review of Thermo plastic Matrix Composites, San Diego, CA, February 9-12, 1987. Johnston, N. J. Presented at the Interdisciplinary Symposium on Recent Advances in Polyimides and Other High Performance Polymers, Sponsored by American Chemical Society, San Diego, CA, January 22-25, 1990. Kochi, M.; Horigome, T.; Mita, I.; Yokota, R. Proc. 2nd Internat. Conf. Poly imides 1985, 454-468.

20. 21. 22. 23. 24. 25. 26.

27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

43.


141

RECEIVED for review March 19, 1991 ACCEPTED revised manuscript March 13, 1992

Deformation and Fracture Toughness in High-Performance Polymers

Recommend Documents