Mechanical behavior of injection-molded polystyrene/polyethylene blends: Fracture toughness vs. fatigue crack propagation

Mechanical Behavior of Injection-Molded Polystyrene/Polyethylene Blends: Fracture Toughness

vs. Fatigue Crack Propagation

M. N. BUREAU,’* E. DI FRANCESCO,’ J. DENAULT,2 and J. I. DICKSON’

1 Dkpartement de mktallurgie e t de genie des mteriuux

Montreal, Canada, H3C 3A7

2 znstitut des materiawc industriels Conseil national de recherche Canada

Boucherville, Canada, J4B 6Y4

&ole Polytechnique

Blends of polystyrene and polyethylene (PS/PE), including blends in which a styrene/ethylene-butylene/styrene (SEBS) terpolymer was employed as a compati- bilker, were studied. Their rheology showed that the effect of the addition of SEBS to PS/PE blends was strongly affected by the blend composition and the shear rates involved in the blending and post-forming processes. The addition of PE to PS led to a reduction of fracture toughness compared with that of PS. This effect was attributed to the fine minor phase morphology of the blends obtained after extrusion blending and injection molding. The fatigue crack propagation (FCP) results showed that the fatigue crack growth rates were significantly reduced at low and moderate range of stress intensity factor (AK) by the presence of PE. Performance was enhanced when SEBS was present. The results also showed that both the fracture toughness and the FCP behavior of the blends were strongly dependent on the loading direction, the minor phase morphology, the composition of the blend, and, to a lesser degree, the presence of a compatibilizer. This study demonstrates that the fracture toughness and the FCP performance of such polymer blends can vary inversely.

INTRODUCTION

he rheology (1-4), morphology (1, 4-8) and me- T chanical behavior (1, 5-14) of blends of low glass transition temperature thermoplastic polymers as the minor phase and glassy thermoplastic polymers as the matrix, such as polystyrene/polyethylene (PS/PE) blends, have been previously studied. Such blends were developed to obtain mechanical behavior similar to that of rubber-toughened systems in which rubber particles are incorporated in a thermoplastic matrix to obtain multiple crazing effects responsible for toughening (15). It has been shown that the morphology and distribution of the rubber phase play a critical role in improving the toughness of rubber-toughened systems (15-24). An improvement in toughness, generally indicated by increased impact resistance or ductility, was also reported in PS-rich PS/PE blends (6-8, 11, 12, 14) when the minor phase morphology and distribution were adequately controlled.

-~ ~~ ._

To whom correspondence should be addressed. e-mail: [email protected]

POLYMER ENGINEERING AND SCIENCE, JUNE 1999, Vol. 39, No. 6

While the morphology and distribution of the minor phase in conventional rubber-toughened systems such as acrylonitrile-butadiene-styrene (ASS) or high- impact polystyrene (HIPS) can usually be well controlled, that of the minor phase in thermoplastic polymer blends evolves continuously during the blending and post-forming processes. This morphological evo- lution of the minor phase is related to drop disintegra- tion, minor phase relaxation and phase coarsening (2, 3, 25). It is also greatly affected by the rheological behavior of the blend under the conditions of blending and post-forming (i.e., flow intensity, elongational or shear flow, temperature, phase interactions, blend composition, viscosity ratio, etc.). Thus, these conditions as well as the resulting blend morphology must be addressed when discussing the mechanical behavior of thermoplastic polymer blends.

Addition of diblock or triblock copolymers to such blends to promote the dispersion of the minor phase and to improve its morphological stability as well as to increase the adhesion at the minor phase-matrix interface has also been addressed (1, 4-10, 26, 27). In PS/PE blends, a fivefold reduction of the interfacial

1119

M. N. Bureau, E. Di FYancesco, J. Denault, and J. I . Dkkson

tension was recently reported when 10% to 20% of c styrene/ethylene-butylene/styrene (SEBS) terpolymer

by weight of the minor phase was added to the blend (4). The emulsification curve in PS/PE blends showed a rapid decrease of the average minor phase diameter reaching a plateau at 10 wt?? to 20 wtYo of the minor phase of SEBS. In the presence of SEBS, evidence of increased adhesion at the PS-PE interface and improvements of ductility and toughness were also reported (1, 5, 6, 10-14).

While the mechanical performance of these thermoplastic polymer blends is generally evaluated by monotonic loading (impact resistance or tensile behavior), their long term mechanical performance in fatigue has been little addressed in the literature and has mostly been studied for homogeneous polymers and rubber-toughened systems (28). For this reason, the fatigue crack propagation (FCP) behavior of twin- screw extrusion-blended and injection-molded PS/PE blends was recently studied (29, 30).

The objectives of this paper are to evaluate and compare the fracture toughness and the FCP behavior of injection-molded PS/PE blends. The effects of the loading direction and of the addition of SEBS as a compatibilizer are also addressed.

ExPERIMEnrTAL

IlllrterialPmpamtion

The materials employed were a HBL455A high density polyethylene (HDPE), an atactic GPPS103-00 PS, both supplied by Novacor Chemistry (Canada) Ltd., and a Kraton G1652 SEBS triblock copolymer supplied by Shell Development Corp. Two groups of PS-rich blends were obtained using a twin screw extruder. The first group contained 1oOh SEBS by weight of the minor phase added to this phase before the PS/HDPE blendmg to optimize the compatibilization. The second group contained no cornpatibilker. The composition of the blends without SEBS was 95/5,85/ 15 and 70/30 PS/HDPE, and the composition of the blends with SEBS was 95/(0.5/4.5), 85/( 1.5/ 13.5) and 70/(3/27) PS/(SEBS/HDPE). The materials were injection-molded into 5-mm-thick plates, with the exception of the 70/30 PS/HDPE blends which were injection-molded into 7-mm-thick plates. The material and specimen preparation conditions have been described previously (29).

Rheological Behavior

The rheological behavior of PS and HDPE as well as 90/10 PS/SEBS and 90/10 HDPE/SEBS was determined by dynamic shear flow measurements carried out using a Rheometric Mechanical Spectrometer (RMS) model 605 in the frequency sweep mode at 2OOOC under a dry nitrogen atmosphere.

IUorphobgid Characterization

Computer-assisted image analysis was used to evaluate the average diameter of the HDPE particles in

the blends on the cryogenically fractured specimens. The fracture surfaces were first coated with a thin layer of gold-palladium and then observed using a JEOL JSM-6100 scanning electron microscope (SEM).

Fracture Toughnem~

The fracture toughness was measured at 23°C and 50% relative humidity following the ASTM D-5045 and E-399 standard test methods. Compact tension (CT) specimens (Rg. 1) were used, with a thickness B equal to that of the injection-molded plates, a specimen width W of 50 mm, a notch depth of 15 mm and a notch opening width of 2.5 mm. A razor-blade cut was introduced manually at the notch tip just prior to the start of fatigue precracking. In order to obtain the sharpest crack tip possible and maximum reproduc-

t l -1.25w - r - - - - - - - -

I I I I

I I I I I I I L

I -1.2w - II 3.1 I

Fig. 1. Diag~m showing the psition and orientation of the CT 4 transverse and bJ bngitudinal specimens used in the injection-molded plates as a m n of the melt&, direction 0.

POLYMER ENGINEERING AND SCIENCE, JUNE 7 9 9 9 , Vol. 39, No. 6 1120

Mechanical Behavior of Injection Molded PS/PE Blends

ibility, every specimen for which numerical results are presented was precracked in fatigue, as specified in ASTM E-399, at an R-ratio (R = Pm/Pm where P- and P,, are, respectively, the minimum and maximum applied loads) of 0.1 and a loading frequency of 20 Hz. Fatigue precracking was continued until the crack length a (a is the sum of the fatigue precrack and the notch depth) to specimen width ratio, the normalized crack length a/ W, was approximately equal to 0.45. As specified in ASTM E-399, fatigue precracking was done at stress intensity factor variation (AK) values never exceeding 60% of the fracture toughness (Kg) value of the materials. For comparative purposes, some tests were also performed on specimens with razor-cut precracks and their Kg values were systematically about 10% higher than those obtained on fatigue precracked specimens, probably because the crack tip radius was sharper in the case of the fatigue precrack.

The precracked CT specimens were loaded until fracture at a constant crosshead speed of 10 mm/min. The precrack length was then accurately measured on the fracture surfaces using an optical microscope. The specimens were loaded for crack propagation either normal (transverse specimens) or parallel (longitudinal specimens) to the melt flow direction. A diagram of the position of these specimens in the injection-molded plates is presented in Fig. 1. The specimen thickness B, equal to that of the injection-molded plates (either 5 or 7 mm), did not respect the usual criteria

based on the minimum thickness B,, (see ASTM D- 5045) given by

B,, = 2.5 * (1)

in which Kg. the fracture toughness, is gwen by

where cry is the yield stress, W the specimen width, P the applied load and f ( a / W ) an empirical function of the normalized crack length, a/W, based on the specimen geometry. Accordingly, plain strain fracture toughness KIc values could not be obtained from the fracture toughness KQ measurements. The fracture toughness values reported are therefore the Kg values obtained according to ASTM D-5045. Five tests were performed for each Kg value reported. The experimen- tal error attributed to each of these values, mainly due to the crack length measurements and the crack front curvature, was estimated to be equal to or less than 0.1 Wadm.

Fatigue Crack Propagation

FCP tests at increasing AK with a constant AK-gra- dient (K-'.dK/da) of 0.09 mm-' were performed at 23°C and 50% relative humidity following the ASTM E-647 standard test method, using a computer-controlled servo-hydraulic Instron tester, a sinusoidal

1 o6

1 o5

I. o4

o3

1 o2

i 0' 1 0- lo-' 1 oo 10' 1 o2 1 o3 1 o4

Frequency (s-' ) Flg. 2. Complex viscosity of PS (O), HDPE (A). 90/10 HDPE/SE1Bs (A) and 90/10 PS/SEBS (0) as afundion of@equenq or shear rate. The c u ~ e extrapolated using the Carreau-modified equation (31) is drawn for each material tested.

POLYMER ENGINEERING AND SCIENCE, JUNE 1999, Vol. 39, No. 6 1121

M. N. Bureau E. Di Francesco, J. Denadt, and J. I. Dickson

waveform, a cycling frequency of 20 Hz and an R-ratio of 0.1. CT specimens similar to those for the fracture toughness evaluation were employed. A razor-blade cut was introduced manually at the notch tip just prior to the start of fatigue precracking. Every specimen tested was precracked over a distance of 3.5 mm at R = 0.1 at a loading frequency of 20 Hz following ASTM E-647. The specimens were loaded for propagation either n o d (transverse specimens) or parallel (longitudinal specimens) to the melt flow direction. A detailed description of the FCP test procedure has been presented (29).

RESULTS AND DISCUSSION

Rhealogical Behavior of P-PE Blend.

The rheological behavior of PS and HDPE and that of 90/10 PS/SEBS and 90/10 HDPE/SEBS at 200°C are shown in Rg. 2. This Fygure indicates that the low deformation rate viscosity of PS and HDPE is not very different and that their viscosity in the range of deformation rates corresponding to extrusion blending (- 102 s-,) is similar. The viscosity of HDPE and PS at high deformation rates, extrapolated from the Cmeau- modified equation (3 1) given by

q = T o s [ 1 + (T* . + ) m ’ ] - q (3)

(where q is the viscosity, qo the zero-shear viscosity, T* the primary relaxation time, 9 the shear rate and m, and m, are adjustable parameters), indicated that it is approximately 2.5 times higher for HDPE than for PS. The addition of SEBS to HDPE resulted in a sign& cant increase: whereas, its addition to PS resulted in a decrease of the complex viscosity at low frequencies.

The probable explanation for these effects is that, when SEBS was added to HDPE, large micelles of SEBS formed, with a PS core and a solvated HDPE

matrix shell, interacting with each other. When SEBS was added to PS, however, SEBS micelles must have been in the form of a compact ethylene-butylene (EB) core with short terminal blocks of PS engendering higher mobility of the PS macromolecules in the same way as a plasticizer would. A schematic representa- tion of these SEBS micelles in a PS matrix and a HDPE matrix is presented in Fig. 3. Recently, it was reported that the number average molecular weight M n of the terminal blocks of PS in the type of SEBS used was approximately 8,000 (5, 14). This is considerably less than the critical entanglement molecular weight of PS of approximately 30,000 (32). As a result, disentanglement and motion of these short PS blocks can occur, in agreement with the plastification effect of the PS terminal blocks. Figure 2 thus suggests that the effect of SEBS in

PS/HDPE blends strongly depends on the blend composition as well as on the shear rates involved in the blending and post-forming processes. Previous results (1 3) have shown that the presence of SEBS in PS-rich blends leads to increased minor phase coarsening during compression molding. This increased minor phase coarsening was attributed to enhanced mobility of the PS molecules at the interlayer of the HDPE domains, which favors diffision of the HDPE macromolecules. In HDPE-rich blends, the presence of SEBS leads to increased minor phase stability during compression molding. This increased stability is probably associated with increased local viscosity of the HDPE matrix at the interlayer of the PS domains, which results in a decreased diffusivity of the PS macromolecules. In such a case. SEBS acts as a barrier at the PS/HDPE interlayer and stabilizes the HDPE-rich blend morphology.

In the presence of high shear rates (- lo4 s-l), such as those in injection molding, the results presented in

SEB S I

Ips PE

m. 3. Schematic representdon proposed for the wn&uration of the SEBS micelles in a PS matriw and a HDPE mairk

1122 POLYMER ENGINEERING AND SCIENCE, JUNE 7999, Vol. 39, No. 6


Table 1. Morphological Characterization of the PS/HDPE blends.

Blend Composition fa "a Observation da L T Ud

95/5 PS/HDPE 0.054 0.34 normal 0.64 1.66 1.02 2.6 parallel 0.70 1.82 1.12 2.6

954 0 W4.5) PS/(SEBS/H DP E) 0.054 0.43 normal 0.60 1.56 0.96 2.6 parallel 0.64 1.66 1.02 2.6

85/15 PS/HDPE 0.161 0.52 normal 0.69 1.20 0.51 1.7

85/(1.5/13.5) PS/(SEBS/HDPE) 0.162 0.45 normal 0.75 1.30 0.55 1.7 parallel 1 .oo 1.73 0.73 1.7

70/(3/27) PSI( SEBS/H D P E) 0.319 0.73 normal 0.86 1.1 1 0.25 1.3 parallel 1 .oo 1.33 0.30 1.3

(Pm-? Plane (Pm) (Pm) (Pm)

~~ ~~~ ~~~ ~. . ~. ~ ~-

.. ~ ~~~~ __ - . . . .~ ~ ~ ~

parallel 0.98 1.70 0.72 1.7 _ _ _ ~ _ _ _ _ . _ ~ ~ ~ ~ ~ ~ -. ~~~

~ ~ ~~ -~

______ ~~~ .- ... ~ ~ ...... - - ~.

Notes The minor phase volume fraction, fa, the average particle diameter. d., the average number of particles per unit volume, n,, the average particle separation, L, the average interparticle distance, T and the average particle separatton to average particle diameter ratio, LM, in the planes normal and parallel to the melt flow direction

Fig. 2 indicate that the addition of SEBS to either PS or HDPE did not lead to any sigdicant variation of the viscosity when compared to that of PS and HDPE. This suggests that coarsening during injection molding is mainly related to the viscosity of the matrix and to the ability of the minor phase particles to come in contact with each other and coalesce. Whether SEBS was present or not, less minor phase coarsening was observed in HDPE-rich blends than in PS-rich blends (13) in agreement with the sigdicantly higher, extrapolated complex viscosity in HDPE-rich blends, as shown in Fig. 2.

Minor Pham IUorpholom

The morphology of the minor phase particles were analyzed in the planes normal and parallel to the melt flow direction in the 95/5 and 85/15 PS/HDPE injection-molded blends and in the 95/(0.5/4.5), 85/( 1.5/ 13.5) and 70/(3/27) PS/(SEBS/HDPE) injection-molded blends. A discussion of the morphology development during compounding and following molding has been presented previously (13). The resulting minor phase morphology is summarized in Table 1 for the two planes of observations. I t should be noted that the average minor phase diameter of the 70/30 PS/HDPE blend could not be measured since this blend exhibited a co-continuous morphology similar to that shown previously (13). The Occurrence of a co-continuous morphology in the 70/30 PS/HDPE blend but not in the 70/(3/27) PS/(SEBS/HDPE) blend is in agreement with the well-known effect of a compatibilizer in an immiscible blend to displace the occurrence of the transition from a co-continuous blend morphology to a dual-phase blend morphology to a higher volume fraction of the minor phase (2).

The results presented in Table 1 indicate that the average minor phase diameter increases with the minor phase volume fraction, that it is not si@icant- ly affected by the presence of SEBS, that the average particle diameter is in each case lower than or equal to 1 pm, and that the xytio of average particle separa-

tion to average particle diameter decreases with the minor phase content. Because the viscosity ratio of PS and HDPE approach unity in the range of deformation rates corresponding to extrusion blending (= 102 s-l),

a relatively high level of emulsification is obtained in these PS/HDPE blends without the addition of SEBS as a compatibilizer, thus limiting any additional emulsification effect associated with its presence. The results presented in Table 1 also indicate that the average particle diameter is higher in the plane parallel to the melt flow direction than in the plane normal to this direction, as a result of the important shear flow during injection molding in this direction.

Fracture Behavior

The fracture behavior of a material depends on the nature of the stress and on the stress intensity factor as well as on the material properties and on the mechanism(s1 by which the crack propagates. The behavior in fracture toughness testing, where increased crack tip plasticity generally tends to increase the toughness, is often very different than in fatigue crack propagation, where the cyclic crack tip plasticity generally is responsible for the fatigue crack propagation. The fracture behavior of PS/HDPE blends is therefore dealt with separately in monotonic (fracture toughness measurements) and cyclic (fatigue crack propagation results) loading.

Fracture Toughness

obtained for transverse and longitudinal specimens of PS and of PS/HDPE and PS/(SEBS/HDPE) blends of various composition. The longitudinal specimens show lower Kg values than the transverse specimens. These results indicate that lower fracture toughness is obtained when the crack propagation plane is parallel, instead of normal, to the melt flow direction. The average orientation of the matrix and the minor phase domains thus influences the fracture toughness of the blends studied. This is in agreement with the loading

Rgure 4 shows the fracture toughness values


M. N . Bureau, E. Di Rancesco, J. Denault, and J. I . Dickson

4 , , 1 1 1 1 , 1 , 1 , 1 1 , , 1 1 1 , , , , 1 , ~ , , 1 , , * , 1 , ~

3 -

n E x c 2 - W

1 - I

0 5 10 15 20 25 30 % minor phase

Flg. 4. Kg fracture bughness values plotted against the minor phase content in transverse [-) and longitudinal [- - -1 specimens of PS/HL3PE and PS/(%BS/€UIPE) blends of various compositions. Open circles represent specimens without SEBS, closed circles represent specimens with SEl3S.

direction effect reported previously on the tensile behavior of injection-molded PS and PS/HDPE specimens (13) and also with the well-known influence of the matrix and particle orientation in blends and com- posites characterized by high levels of anisotropy (33-36).

Figure 4 also indicates that 4 progressively decreases as the minor phase content is increased. This is in agreement with the results presented in Table I since the criteria (16-23) proposed for toughening due to multiple crazing during monotonic loading in a PS matrix are not satisfied, with an average particle diameter higher than 1 pm and an average particle separation to average particle diameter ratio lower than 1.45. The results presented in Flg. 4 are also in agreement with the effect of the minor phase particles, act- ing as stress concentrators in the PS matrix, lowering the stress at which crazing occurs, as illustrated by the significant number of minor phase domains on the fracture surfaces in the fast crack growth region of PS/HDPE specimens shown in Fig. 5. This type of fracture surface differs from that observed in specimens of PS (Fig. s), which shows the patch pattern generally observed in the fast crack growth region in monotonic tension associated with craze fracture in a homogeneous polymer (37).

Flgure 4 also indicates that the fracture toughness for a given blend composition is generally increased when SEBS is present, in agreement with previous studies (1, 5. 6, 10-14) on the ability of SEBS to increase the PS-HDPE interface adhesion. Figures 5b

and 5c illustrate this increase in the PS-HDPE interface adhesion in the presence of SEBS. In the non- compatibilized specimens (Fig. 5c), PS-HDPE interface decohesion was generally observed with relatively few elongated HDPE domains present on the fracture surfaces. In the compatibilized specimens (Flg. 5b), many of the HDPE domains on the fracture surfaces were greatly elongated, which indicates that the PS-HDPE interface adhesion is increased by the presence of SEBS. Despite the effect of SEBS to increase the fracture toughness, its presence did not lead to a transition of the PS/HDPE blend fracture behavior from brittle to tough, as previously shown (19). Flgwe 4, however, indicates that the 95/5 PS/HDPE longitudinal specimens did not show a significant variation of the & value when SEBS was added to the blend. The combined effects of the minor phase morphology, the crack propagation orientation and the particle-matrix adhesion in the 95/5 PS/HDPE and the 95/(0.5/4.5) PS/(SEBS/HDPE) longitudinal specimens probably account for the similar 4 values obtained. Neither the minor phase characterization nor the hctograph- ic observations performed suggested an explanation for this effect.

Fatigue Crack Propagation Valid FCP test results were obtained for longitudinal

and transverse specimens of PS, of 95/5 and 85/15 PS/HDPE blends and of 95/(0.5/4.5) and 85/(1.5/13.5) PS/(SEBS/HDPE) blends. Valid FCP curves could not be obtained for longitudinal and transverse specimens



Fig. 5 . Fracture surfaces of transversefracture toughness specimens of a) 95/(0.5/4.5) PS/(SEBS/HDPE), b) 70/(3/27) PS/(SEBS/HDpEI and c) 70/30 PS/HDPE obserued near the start of the fast fracture region The crack propagation direction isfrom bottom to top.

Rg. 6. Racture surface of a transversefracture toughness specimen of PS observed near the start of the fastfracture region The crack propagation direction is from bottom to top.

of 70/30 PS/HDPE blends and for transverse specimens of 70/(3/27) PS/(SEBS/HDPE), since a 30" to 45" crack deviation from the plane perpendicular to the load axis was systematically obtained for these specimens during precracking. Valid FCP tests were however obtained for longitudinal specimens of 70/(3/27) PS/(SEBS/HDPE). In this material, some crack deviation from the plane perpendicular to the load axis was also obtained but was always less than the 10" per- mitted by ASTM E-647. These results illustrate the important level of anisotropy encountered in injection- molded thermoplastic polymer blends and also imply improved resistance to fatigue crack propagation associated with such crack deviation.

The log-log curves of the fatigue crack growth rate (da/dnr) versus AK of the materials tested are presented in Fig. 7 as power law regressions of the Paris regime (38).

(4)

where C and n, which represent, respectively, the ordinate at the origin (i.e., at AK = 1 Wadm) and the slope of a log da/clN - log AK plot, are material con- stants for given test conditions. The C and n values calculated from Eq 4 are presented in Table 2. The re- gression factor is always higher than 0.9.

da/dN = C * AK"


M. N. Bureau, E. Di Fi-ancesco, J. Denault, and J. I . Dickson

Table 2. Average Paris Regime Parameters, Cand n, Obtained From the FCP Curves of Transverse and Longitudinal Specimens of PS and PSlHDPE Blends of Various Compositions.

Material Specimen orientation C n

PS transverse 6.6 X lo4 2.6 longitudinal 1.1 x 103 2.5

95/5 PS/HDPE transverse 5.1 X lo4 2.6 longitudinal 1.8 x la-3 2.5

-

95/(0.5/4.5) PS/(SEBS/HDPE) transverse longitudinal

3.5 x 10-4 1.4 X la-3

2.5 2.7

85/15 PS/HDPE transverse longitudinal

4.5 x 10-4 8.8 x 104

3.9 3.9

85/(1.5/13.5) PS/(SEBS/HDPE) transverse longitudinal

3.6 X 10-4 8.5 x 10-4

3.8 3.8

~ ~~

70/(3/27) PS/(SEBS/HDPE) longitudinal 2.0 x 10-4 3.4

The value of the slope n obtained on specimens of a given material is almost constant, regardless of the presence of SEBS or the specimen orientation. As ex- pected, the FCP results shown in Rg. 7 indicate that the transverse specimens show lower fatigue crack growth rates at a given AK than the longitudinal specimens, in agreement with the fracture toughness results presented in Rg. 4. The results presented in Table 2 also show that the ordinate at the origin C is reduced by an average factor of 2.7 when the specimens were loaded for propagation normal (transverse specimens) instead of parallel (longitudinal specimens) to the melt flow direction. These results indicate that lower fatigue crack growth rates are obtained when the crack propagation plane is normal to the orientation of the matrix and to the elongated direction of the HDPE minor phase particles.

The FCP curves presented in Rg. 7 and the ordinate at the origin C presented in Table 2 indicate that the incorporation of HDPE to PS generally led to a reduction of the fatigue crack growth rates, especially at low and intermediate AK. Figure 7, however, shows a small increase of the fatigue crack growth rates in the 95/5 PS/HDPE longitudinal specimens when compared to PS longitudinal specimens. When the HDPE content was increased to 15% and 30%, a considerable reduction of the fatigue crack growth mtes was obtained for both transverse and longitudinal specimens. The average n values obtained were 2.5-2.6 for the PS specimens as well as for the 95/5 PS/HDPE and 95/ (0.5/4.5) PS/(SEBS/HDPE) specimens, 3.8-3.9 for the 85/15 PS/HDPE and 85/(1.5/ 13.5) PS/(SEBS/HDPE) specimens and 3.4 for the 70/(3/27) PS/(SEBS/HDPE) specimens. These results indicate that the slope n and the ordinate at the origin C of the Paris regime obtained are considerably influenced by the minor phase content, once this content becomes sufficiently large. The n values obtained are consistent with reported n values for PS and rubber-toughened PS-rich blends of 2.8 to 4.8 (28).

Flgure 7 also indicates that the addition of SEBS to HDPE in PS-rich PS/HDPE specimens led to a signifi-

cant reduction of the fatigue crack growth rates, in agreement with the increa,sed adhesion obtained at the PS-HDPE interface in the presence of SEBS. The increased minor phase aspect ratio due to a reduction of the interfacial tension (4) in the presence of SEBS (13) probably accounts for at least part of this reduction of fatigue crack growth rates obtained.

The strong deviation of the precrack observed during the FCP tests in the 70/30 PS/HDPE longitudinal and transverse specimens can be attributed to the co- continuous morphology observed in these specimens. It should be noted that similar crack deviation was observed in the 70/(3/27) PS/(SEBS/HDPE) transverse specimens characterized by a dual-phase morphology instead of a co-continuous morphology. The observations of the fatigue fracture surfaces of these specimens, presented in Rg. 8, indicate that the crack deviates progressively to become more parallel to the length of the minor phase domains, which show large length-to-width ratios. These observations thus strongly suggest that the presence of aligned HDPE domains with significant length-to-width ratios in the direction normal to the intended orientation of the crack plane also constituted an effective barrier to FCP, at least at low AK.

In summary, the FCP results presented in Rg. 7 indicate that the FCP performance of the PS/HDPE blends strongly depends on the mechanical loading conditions. When tested at low levels of AK, a reduction of the fatigue crack growth rates by an approximate factor of 20 in longitudinal specimens of 70/(3/27) PS/(SEBS/HDPE) can be obtained compared to longitudinal specimens of PS. Since most of the life before failure during FCP is generally determined by the fatigue crack growth rates at low AK (391, the resistance to fatigue crack propagation in many practical applications would be considerably increased in a 70/(3/27) PS/(SEBS/HDPE) blend when compared to PS. Consequently, from these results, it appears evident that long term me-

lated directly from monotonic loading or impact testing, as often assumed. In the present study, the addition of

chanical PropeltiB of such materials cannot be extrapo-



1 0 3 - n W u h -

10-4 - z 2 a a 4’. . / ’ 1 0 5 4

./= /’ /‘

-

I

HDPE generally reduced the fatigue crack growth rates at low and moderate values of AK, but the Z-C,- value at which final fracture occurred in fatigue did not increase with the presence of HDPE, in agreement with the fracture toughness results. With their relatively low Young’s modulus [the modulus decreases

from 3.5 GPa for PS to 2.2 GPa for 70/30 PS/HDPE (13)], the fatigue performance of the PS/HDPE blends studied could be interesting in imposed displacement applications of low to moderate amplitudes, since the fatigue crack growth rates at low AK in the blends with sufficient HDPE are much lower than for PS.


M . N. Bureau, E. Di Rancesco, J. &nu&, and J. I . Dickson

Fig. 8. Fracture surface of a transverse 70/(3/27) P S / ISEBS/HDPEEJ specimen am ap pmdmately 6.2 mm of prerrack propagation [dariation = 45") showing a very oriented and elongated minor phase morphology parallel to the FCP direction The macroscopic FCP direction isfrom top to bottom

CONCLUSIONS

The effect of the addition of SEBS to PS/HDPE blends is strongly dependent on the blend composition and on the shear rates involved in the blending and post-forming processes.

The addition of HDPE to a PS matrix leads to a reduction of the fracture toughness when compared to that of PS specimens, probably owing to the very fine minor phase morphology obtained which does not respect the morphological criteria proposed for toughening.

Particle-matrix adhesion due to SEBS addition increases the fracture toughness somewhat, but the blend morphology appears to be the controlling pa- rameter for toughness improvements in PS/HDPE.

The fatigue crack growth rates are generally reduced by the presence of HDPE at low and moderate values of AK, especially when SEBS is added, but the K,.- value at which final fracture occurs in fatigue did not appear to increase with the presence of HDPE, in agreement with the fracture toughness results.

Both the fracture toughness and the FCP behavior of these materials are strongly dependent on the loading direction, the matrix and minor phase morphology, the composition of the blend and, to a lesser degree, on the presence of a compatibilizer.

The long term mechanical performance of such materials cannot be extrapolated from results obtained in monotonic loading (e.g., fracture toughness or impact resistance results) since the mechanical behavior of these materials strongly depends upon the loading conditions, which are considerably different in monotonic and cycling loading.

ACIWOWLEDGMENTS

Financial assistance from the NSERC (Canada) and K A R (Quebec) research support programs is gratefully acknowledged.

NOMENCLATURE

polystyrene polyethylene styrene/ethylene-butylene/styrene acrylonitrile-butadiene-styrene high-impact polystyrene high-density polyethylene ethylene-butylene number average molecular weight melt flow direction fatigue crack propagation stress intensity factor plane strain fracture toughness maximum stress intensity factor fracture toughness stress intensity factor variation load ratio load minimum load maximum load yield stress fatigue crack growth rate Paris regime ordinate at the origin Paris regime slope compact tension specimen thickness minimum thickness required for plane strain conditions specimen width crack length normalized crack length empirical function of the normalized crack length viscosity zero-shear viscosity primary relaxation time shear rate Carreau-modified equation adjustable parameters minor phase volume fraction average particle diameter average number of particles per unit volume average particle separation average interparticle distance average particle separation to interparticle distance ratio melt flow direction scanning electron microscopy

REFERENCES 1. D. Heikms and W. M. Barensten, Polymer, 18,69 (1977). 2. L. A. Utracld, pblymer Akqs and Blends:

andRhedogy. Hanser Publ.. Munich ( 1 9 7

1128 POLYMER ENGINEERING AND SCIENCE, JUNE 7 9 9 9 , Vol. 39, Iyo. 6


3. L. A. Utracki and 2. H. Shi. Polym Eng. Sci, 32, 1824 ( 1992).

4. N. Mekhilef, B. D. Favis, and P. J. Carreau, J. Polym Sci: Part B: Polym Phys., 38, 293 (1997).

5. M. Matos. P. Lomellini, and B. D. Favis, SPEANTEC Papers, 40, 1517 (1994).

6. R. Fayt, R. JerBme, and Ph. Teyssie, J. Polym Sci, P o l y r ~ ~

7. R. Fayt, R. JerBme, and Ph. Teyssie, J. Polym Sci, Polym

8. R. Fayt and Ph. Teyssie, Polym Eng. Sci , 30, 937 (1990). 9. H. El Kadi, J. Denault, D. Tapin, M. F. Champagne, L.

A. Utracki, and M. M. Dumoulin, SPEANTEC Papers, 41, 3143 (1995).

10. W. M. Barensten and D. Heikens, Polymer, 14, 579 (1973).

11. C. R. Lindsey, J. W. Barlow, and D. R. Paul, J. AppL Polym Sci, 26, 1 (1981).

12. M. Bousmina. P. Bataille, S . Sapieha, and H. P. Schreiber, Compos. Interfaces, 2, 171 (1994).

13. M. N. Bureau, H. El Kadi, J. Denault, and J. I. Dickson, Polym Eng. Sci, 37, 377 (1997).

14. T. Appleby, F. Cser, G. Moad, E. Rizzardo, and C. Stavropoulos. Polym Bull., 32, 479 (1994).

15. C. B. Bucknall, Toughened Plastics, L. Holliday and A. Kelly, eds., Applied Science Publishers. London (1977).

16. A. S. Argon and R. E. Cohen, Adu. Polym Sci, 91/92, 301 (1990).

17. S . Y. Hobbs. Polyym Eng. Sci, 26, 74 (1986). 18. M. Matsuo, T. T. W a g . and T. K. Kwei, J. Polym Sci A-

19. S. Wu, Polymer. 26, 1855 (1985). 20. C. B. Bucknall, Adu. Polym Sci, 27. 121 (1978). 21. A. S. Argon, R. E. Cohen, 0. S . Gebizlioghi, and G . E.

22. J . A. Sauer, J. Trent, and C. C. Chen, Polym Eng. Sci,

Let. m., is, 79 (1981). Phys. Ed.. 19, 1269 (1981).

2, 10, 1085 (1972).

Schwier, Adu. Polym Sci, 82/83,275 (1983).

29, 69 (1989).

23. G. Dagli, A. S. Argon, and R. E. Cohen, Polymer. 36,

24. J. Silberberg and C . D. Han, J. Appl. Polym Sci, 22,

25. B. D. Favis, Canadian J. Chem Eng., 69, 619 (1991). 26. R. Fayt, R. JerBme, and Ph. Teyssik, Polym Eng. Sci.

27, 328 (1987). 27. R. Fayt. R. JerBme, and Ph. Teyssie, J. Polym. Sci.

Polyym Phys. Ed, 20, 2209 (1982). 28. R W. Hertzberg and J. A. Manson. Fatigue of Engineering

Plastics, Academic Press, New York (1980). 29. M. N. Bureau, J. I. Dickson, and J. Denault, J. Makr.

Sci, 33, 1591 (1998). 30. M. N. Bureau, J. I. Dickson, and J. Denault, J. Mater.

Sci, 93. 1405 (1998). 31. L. A. Utracki, Current Topics in Polymer Science, R. M.

Ottenbrite, L. A. Utracki, and S . Inoue, eds., Hanser Publ., Munich (1987).

32. R. P. Wool, Polymer Interfaces: Structure and Strength, Hanser Publications, Munich (1995).

33. I. M. Ward, Mechanical Properties of Solid Polymers. 2nd Ed., John Wiley & Sons, New York (1983).

34. M. R. Kamal, M. A. Sahto, and L. A. Utracki. Polym Eng. Sci, 23, 637 (1983).

35. B. F. Bright and M. W. Darlington, Plast Rubber Proc. AppL, 1. 139 (1981).

36. J. Denault and T. Vu-Khanh, Polym Compos., 9. 360 ( 1988).

37. D. Hull, Polymeric Materials, ASM, Metals Park, Ohio (1975).

38. P. C. Paris and F. Erdogan, J. Basic Eng. lhms.: ASME Ser. D, 86. 528 (1963).

39. R W. Hertzberg, &formation and n a t u r e Mechanics of Engineering Materials. John Wiley & Sons, New York (1976).

2173 (1995).

599 (1978).


Documents

Mechanical behavior of injection-molded polystyrene/polyethylene blends: Fracture toughness vs. fatigue crack propagation