8/6/2019 070510804_ftp sushil

1/8

Fracture Mechanics Investigation on the PP/ EPDM/lonomer Ternary Blends Using j-Integral by Locus Method

C H A N C -S I K H A ," Y O U N G K Y O O KIM, and WON-JEI CHODepartmentof Polymer Science an d Engineering, Pusan National University, Pusan 609-735, Sou th Korea

SYNOPSISThe fracture mechanics investigation of the polypropylene (PP ) /ethylene-propylene-dieneterpolymer ( E PDM ) ionomer ternary blends wa s performed in terms of the J-integral bymeasuring fracture energy via the locus method. Blends were prepared in a laboratoryinternal mixer. The composition of PP and EPDM was fixed at a 50/50 ratio by weight.Two kinds of poly(ethy1ene-co-methacrylic acid) ( EMA ) ionomers were used. The J-integral value at crack initiation, Jc, of the PP /EPDM /EMA ionomer ternary blendswere affected by the cation types (Na' or Zn2+)and contents (5-20 wt % ) of the addedEMA ionomers. The ternary blend having 5 wt % of Na-neutralized ionomer showed ahigher Jc value than that of any other ternary blends. The results were discussed withregard to the fracture topology by a scanning electron microscope ( S EM ) .0 994 JohnWiley & Sons, Inc.

INTRODUCTIONThe fracture toughness characterization of brittleor quasi-brittle polymers has usually followed thelead of meta ls characterization based on linear frac-tur e mechanics. Th e utility of linear fracture me-chanics, however, is inadequate for toughened poly-mers of multiphases, since the considerable amountof energy put into th e material is dissipated, creatingplastic deformation ahead of the crack tip.'-5

When linear elastic fracture mechanics could notpredict performance of ductile an d toughened poly-mers, interes t moved from fracture toughness char -acterization based on the stress intensity factor,K,6-8 o the J-integral concept introduced by Riceor Begley and Landes?-" Th e J-integral is an energyinput parameter th at can be used as a criterion forcrack initiation in the fracture of elastic-plasticmaterials. Little work, however, has been publishedto apply the J-integral analysis to ductile or tough-ened polymers.

There have been several efforts to determ ine theJ-integral value at crack initiation, J c , for ductileor impact-modified polymers, even though the re-

* To whom correspondence should be addressed.Journal of Applied Polymer Science, Vol. 51,1381-1388 (1994)0 1994 oh n Wiley & Sons, Inc. CC C 0021-8995/94/081381-08

ported methodologies have been very controver-~ i a l . ' ~ ? ~ ~ecently, a novel test technique was devel-oped by Kim and Joe, in which a fracture test wascarried out by using simple single edge-notched(SE N) tensile specimen^.'^^'^ The technique wasbased on the locus of crack initiation poin ts on load-displacement records. It was reported tha t the eval-uation method of J c was successfully applied tohighly deformable materials including a thermo-plastic elastomer like Santoprene without re-stricting the ratio of the initial crack length to thespecimen width ( a / w ) r the specimen length, pro-vided that the locus line could be located on theload-displacement record.

Th e object of th is work was to analyze the fracturemechanics investigation on the polypropylene (PP)ethylene-propylene-diene terpolymer (EPDM ) /ionomer ternary blends using the J-integral by thelocus method, along with the f racture topology. Th elocus method is believed to be suitable for our PP/EPDM / ionomer ternary blend systems, since thepresent work deals with material very similar to thatused for Kim and Joe's work. In our previous work,l6we revealed tha t th e addition of poly (ethylene-co-methacrylic acid) (EMA) ionomer, where the acidgroups are partially neutralized by Na +or Zn2+ ons,significantly affected the rheological properties andmorphology of a PP/EPDM binary blend. It may

1381

8/6/2019 070510804_ftp sushil

2/8

1382 HA, KIM, AND CHO

be thought that the present work will shed somelight on the toughness of the ternary blends as wellas of the P P/ EP DM binary blend, even though theJc by itself is not a widely accepted measure oftoughness.

In the present investigation, two kinds of EMAionomers containing different metal ions (Na' orZn2') were used for comparison, and th e effect ofthe metal component for neutralization in the io-nomer and the ionomer contents in the PP/ EP DM /ionomer ternary blends were discussed. The com-position of PP and EPDM in the blends was fixedat a 50/50 ratio by weight. The ionomer contentswere varied from 5 to 20 wt % based on the totalamount of PP and EPDM.

EXPERIMENTALMaterialsThe characteristics of the polymers used in thisstudy are summarized in Table I. The polypropylene(PP)was kindly obtained from Korea Petrochemical

Table I Materials and Their Characteristics

Table I1 Blended Materials andTheir Compositions

~

PP EPDMContents Contents Ionomer

Notation (wt %) (wt %) ContentsPP50-EP50 50.0 50.0 No ionomersPP50-EP50/IA5 47.5 47.5 5.0PP50-EP50/IA15 42.5 42.5 15.0PP50-EP50/IA10 45.0 45.0 10.0PP50-EP50/IA20 40.0 40.0 20.0

PP50-EP50/IB10 45.0 45.0 10.0PP50-EP50/IB20 40.0 40.0 20.0

PP50-EP50/IB5 47.5 47.5 5.0PP50-EP50/IB15 42.5 42.5 15.0

Co. (PP4017 grade). The ethylene-propylene-dieneterpolymer (EPDM ) with ethylidene-2-norbornene(E NB as a termonomer was supplied by Uniroyal(Royalene 521, weight-average molecular weight[M w ]= 1.80X lo5; thylene content [C , ] = 52% inmol). Two kinds of EMA ionomers were suppliedby Dupont. The polymers were all commerciallyavailable grades and used as received.

Material Properties Source Blend PreparationPP

EPDM

IonomerA

Ionome rB

M,, = 2.83 x lo4M , = 2.02 x l o 5MFI" = 6.0M JM , = 7.14vinh (dL/g)b = 1.22I.V.? 15.2PE/PP (mol %) d

= 52.0/48.0ENB typeCation type: Na+Ethylene/methacrylic

acid = 91/9% Neutralization: 50%Specific gravity = 0.94MFI" = 1.3Cation type: Zn2+Ethylene/methacrylic

acid = 91/9% Neutralization: 50%Specific gravity = 0.95MFI" = 1.1

KoreaPetrochemicalCo. (PP4017)

Uniroyal(Roy. 521)

DuPont(Surlyn 8528)

DuPont(Surlyn 9520)

a Melt flow index.' y ICl titration method.0.5 g/dL xylene solution a t 70C.By IR analysis.

Binary PP /EP DM or ternary PP/EPDM/ionomerblends were prepared in a Brabender Roller Mixer(Type w50H) a t 190C and 60 rpm for 15 min. Thecomposition of PP and EPDM was fixed at a 50/50ratio by weight. For the ternary blends, the ionomercontents were varied from 5 to 20 wt % based onthe total amount of PP and EPDM. The samplenotations of blends are summarized in Table 11.

Fracture Energy DeterminationMany different specimen configurations are accept-able for the measurement of fracture toughness. Weused single-edge-notched (SEN ) tensile specimenswith a constant length from compression-moldedsheets. The thickness and width of the specimenswere 1.0 and 20.0 mm, respectively. Th e length be-tween the grips were 60.0 mm. Razor blades wereused to create sharp initial cracks, which varied from0.20 to 0.80 in terms of th e ratio of the initial cracklength ( a ) o th e specimen width (w ) Tensile testswere performed at a crosshead speed of 50 mm/minwith an ambient temperature of 26C and relativehumidity of 35%. Load-displacement graphs were

8/6/2019 070510804_ftp sushil

3/8

FRACTURE MECHANICS OF PP/EPDM BLENDS 1383

recorded and initiation points were marked on eachloading line during the test. T he fracture toughnesswas interpreted in terms of the critical J-integralvalue, Jc , by the locus method developed by Kimand Joe.I4

It was reported that the locus method using atensile specimen was successfully applied to theSantoprene- like highly deformable materials whencompared to the conventional method using a three-point bending ~pec ime n.' ~, '~rack initiation pointswere easily observable since the crack opened widelybefore the crack propagated. The areas under theloading curves were calculated numerically from thedata points taken from the load-displacement rec-ords. The essential energy needed for crack propa-gation per unit thickness, AU,, vs. the initial cracklength, a , was plotted. The linear least-squares fittedslope was taken a s J c.

The method determines Jc based on-1 AU,J c = - -B Aa

where B is the thickness of the specimen; a , th einitial crack length; and U,, the enclosed area be-tween loading line and th e locus line. The tota l en-ergy consumed during fracture testing includes someamount of remote energy loss as well as the energysupplied for the crack propagation. Remote plasticenergy loss occurs during the loading process ifplastic deformation, remote from the crack tip, ex-ists. Several single-specimen test methods that usethe relation between th e essential energy U, and thetotal energy absorbed, UT,have been deve10ped.I~Th e value of 4(=U,/ U,) was known to be sensitiveto the remote energy absorption. With the Jc valueknown, 4 for each crack size can be determined fromeq. ( 2 ) :

If crack initiation has occurred at a constant dis-placement and the locus of the crack initiationpoints continued to be a vertical line down to thedisplacement axis, then all the 4 values should havebeen 1. It was reported that the remote energy ab-sorption can affect the accuracy of the experimen-tally determined Jc values and th at eliminating theremote energy term in the early stage of the pro-cedure can help to yield consistent Jc ~a1ues. l~e-tails of the locus method are described e l~ ewher e . '~ ,~~

Reproducibility of the fracture energy data wasmeasured by testing different sample lengths. How-

ever, all samples were tested with the initial samplelength of 100 mm unless otherwise specified. Ex-aminations of the fractured surfaces were made us-ing a scanning electron microscope (SEM; JEOLJSM-35CF).

RESULTS AND DISCUSSION/c ValuesFigure 1 shows typical load-displacement records forthe P PI EP DM binary blend. The bars denote thespread of observed crack initiation points. The locusline slightly deviates from a vertical line.

AU, per unit thickness is plotted against eachinitial crack size for the PP/E PDM /ion ome r ter-nary blends as well as for the PP IE PD M binaryblend a s shown in Figures 2 and 3. The slope of th eleast-square-f itted line yields Jc , and th e resultingvalues are replotted against ionomer content in Fig-ure 4.

Th e average correlation coefficient of th e Jc testwas 0.98. I t is seen that linear relationships betweenUc an d a have been obtained for the both binaryand ternary blends. The result implies that the Jvalue at the crack initiation point, Jc , is a constantfor a given thickness, which is the only assumptionmade in th e locus method and proves it s validity forour blend system.

In Figure 4, t is shown that the ionomer-addedternary blends exhibit higher Jc values than tha t ofthe P P/E PDM binary blend irrespective of the io-nomer types an d contents, meaning tha t the incor-

1 2 0 1

Disp lace ment ccmlFigure 1denote crack initiation points.

Load-displacement curves: the spread bars

8/6/2019 070510804_ftp sushil

4/8

1384 HA, KIM, AND CHO

I t

10c25-m2 5

0 0 0.5 1 1.5 2a (cm)

Figure 2 Variation of essential energy needed for crackinitiation per unit thickness (U JB )with initial crack size( a ) .The broken lines are the least-squared ones. PP50-EP50/IA: (A) wt %; ( 0 ) wt %; (0)0 wt %; ( A ) 15wt %; ( 20 w t %.

poration of ionomer enhanced the toughness of thebinary blend. In addition, it should be noted thatthe toughening effect by ionomer addition was mos tprominently observed when th e contents of the io-nomers were 5 wt %.

0 0.5 1 1.5 2a(cm)

Figure 3 Variation of essential energy needed for crackinitiation per unit thickness (U,/B)with initial crack size( a ) .The broken lines are the least-squared ones. PP50-EPSO/IB: (A) w t %; (0)w t %; ( 0 ) O w t %; ( A ) 15w t % ; ( D ) 2 0 w t % .

0 1 I0 10 2 01 0nomer (wt.%)

Figure 4energy J, : ( A ) PP50-EP50/IA; ( B ) PP50-EP50/IB.

Effect of the added ionomer on the fracture

The result may be due to the compatibilizationof EPDM an d PP by the ionomers, as reported inour previous work.16 The difference for ionomertypes are within experimental error and did not af-fect the Jc values.

Remote Energy AbsorptionTh e total energy consumed during fracture testingincludes some amo unt of remote energy loss as wellas the energy supplied for the crack propagation.Th e remote energy absorption away from the crackti p is not negligible if the specimen length is not10ng.I~ n th is work, the specimen length is 100 mm.Th e remote energy absorption in determining theJc value was also evaluated.

Figure 5 shows the effect of the ionomer addit ionon the 6 alues. In this plot, the 4 value for theinitial crack in length ( u / w )of 0.5 was compared.The 6 value for the PP/EPDM binary blend wasless than 1.0, but that for the ternary blends wasequal to or slightly larger than 1.0, regardless of io-nomer types and contents. Th e result implies tha tthe effect of remote energy abso rption is negligiblefor the ionomer-added t ernary blends, whereas theeffect should be taken into account to estimate Jcvalues from our locus method for the PP/EPDMbinary blend. T he 4 values for the ternary blendshaving 10 an d 15 wt % of ionomer B are larger thanany other tern ary blends and are almost 1.5. Thistren d was because the crack initiation locus below

8/6/2019 070510804_ftp sushil

5/8

FRACTURE MECHANICS OF PP/EPDM BLENDS 138521B I, \.A -.'41.5

/

t

,.\,

//- -a-_A

01 I I0 1 0 2 0lonomer ( wt . %)

Figure 5 Effect of the added ionomer on the ratio ( 4 )of essential crack initiation energy to total energy neededfor crack initiation: ( A)PP50-EP50/IA, (B)PP50-EP50/IB.

the loading curve of the largest crack size has a neg-ative slope in the load-displacement record, meaningth at the critical displacements (crack initiation dis-placements) are larger at the same initial cracklength. Thus, th e Jc value determined on the locusmethod for the ionomer A-containing ternary blendis more accurate than that of the ionomer B-con-taining ternary blend. Note, however, that the sig-

1

0.5

0E0a

Lu"c3 - 0.5

- 1

\\\' \\ \

\ \ A\ - _- -l\ ,./- - - * - -\\ - t -.// '% _ - -B\ '0

0 10 20lonomer ( w t %)

Figure 6 Effect of the added ionomer on the remote orplastic deformation energy (U R) :A) PP50-EP50/IA; (B )PP50-EP50/IB.

Table I11or PP/EPDM/Ionomer Ternary BlendsFracture Energy of PP/EPDM Binary

Sample Fracture Energy, J c (kgf cm/cm')PPPP50-EP50PP50-EP50/IA5PPSO-EP50/IAlOPP50-EP50/IA20PP50-EP50/IA15PP50-EP50/IB5PP50-EP50/IBlOPP50-EP50/IB15PP50-EP50/IB20

2.902.087.134.883.496.828.625.306.226.91

nificance is not large and, thus, the Jc values forthe ionomer B-containing ternary blend are also ac-ceptable in our work.

The plot of remote absorption energy,UR,againstthe ionomer content, as shown in Figure 6 , alsoshows the same trend as referred to in Figure 5. Forthe PP/EPDM binary blend, the effect of remoteenergy absorption is significant (URk 1.0) ,whereasfor the ionomer-added ternary blends, the effect isnegligible regardless of ionomer contents and types

Table I11 shows the fracture energy data for allthe blend samples investigated. Of interest is thefact tha t the Jc value of the PP/ EP DM binary blendis lower than that of PP, whereas the Jc values ofthe ionomer-containing ternary blends are highertha n those of PP as well as of the P P/ EP DM binaryblend. The result implies that the incorporation ofionomer enhances the toughness of PP and the PP/EPDM binary blend regardless of the ionomer typesand contents.

(URk 0.0).

Fracture Surface TopologyIn our previous work, the blends showed very com-plicated multiphasic morphologies.16 Th e fracturesurface topology was observed by scanning electronmicroscopy.

The PP/EPDM binary blend showed somewhatquasi-cleavage fracture topology, 9 as shown in Fig-ure 7. However, all the ionomer-added ternaryblends showed slightly different fracture surface to-pologies. Figure 8 shows the SEM micrographs ofthe fractured surfaces taken around the crack-tip.In this case, the initial crack length was8 mm. Theionomer-added ternary blend shows the typicalfracture surface topology of tough materials. Careful

8/6/2019 070510804_ftp sushil

6/8

1386 HA, KIM, AND CHO

Figure 7nary blend. The initial crack length is 8 mm.

SEM microfractograph of the PP/EPDM bi-

inspection of Figures 7 and 8 shows tha t the fracturesurface of th e 5 wt % ionomer-added ternary blendshas clearer dimple fracture topology, exhibitingtougher characteristics than that of the 10 wt %ionomer-added ternary blends, regardlessof ionomertype.

Figure 9 SEM microfractograph of the PP50-EP501IA5 ternary blend at a lower magnification of 300X. Theinitial crack length is 8 mm. ( a ) s a notched region and( b ) is an initiation region.

Figure 9 shows a lower magnification of the frac-tured surface of the 5 wt % ionomer A-added ternaryblends. The micrograph clearly reveals dimple rup-ture topology, which is usually observed in tough

Figure 8 SEM microfractographs of th e PP/EPDM/ionomer ternary blends: ( a ) PP50-EP50/IA5; ( b ) PP50-EP50/IA10; ( c ) PP50-EP50/IB5; ( d ) PP50-EP50/IB10.The initialcrack length is 8 mm.

8/6/2019 070510804_ftp sushil

7/8

FRACTURE MECHANICS OF PP/EPDM BLENDS 1387

Figure 10 SEM microfractograph of the PP50-EP50binary blend at a lower magnification of 300X. The initialcrack length is 8 mm. ( a) -( c) are the same as referred t oin legend to Figure 11.

materials. When the samples with an initial crackare subjected to load, they initially resist crackpropagation, but on exposure to high load, they f i -nally undergo spontaneous failure. When a materialis tough enough to resist crack propagation, the ma-terial will stretch over long distances unti l it reachedfailure. We call the distances between the initialcrack region and th e end band region an initiationregion.

Th e SE M micrograph of Figure 10 shows a ma-jority of the fractured surface of the P P/ EP DM bi-nary blend. In these micrographs, A, B, an d C rep-resent the initial crack tip, initiation region, andend band region, respectively. The direction of cracktravel is from top t o bottom. The side view of a frac-tured sample is schematically shown in Figure 11.

Table IV shows the length of the initiation regionfor each sample. It is seen th at t he ternary blendscontaining ionomers showed larger initiation regionsthan did the binary blend without ionomers, re-gardless of ionomer types an d contents.

Figure 11 Schematic diagram of a fractured specimenfrom a sideview, where ( a) , (b ) and ( c ) are a notchedregion; an in itiat ion region, i.e., s tretched zone; and a nen d band region, i.e., rapid des troyed region, respectively.

Table IVBlends, d mm)

Crack Initiation Length of

Sample Crack Initiation Length, d (mm)PP50-EP50PP50-EP50/IA5PP50-EP50/IA10PP50-EP50/IB10PP50-EP50/IB5

0.1240.3630.1480.3790.152

CONCLUDING REMARKSSeveral conclusions can be drawn from the aboveresults concerning the fracture mechanics investi-gation of the PP/EPDM/ionomer ternary blendsby using the Jc integral via the locus method (1 )PP/EPDM blends containing ionomer showedhigher Jc values than those of the P P/ EP DM bi-nary blend as well as of PP itself. ( 2 ) The Jc valueswere highest for the ternary blends containing 5 wt% of ionomer, regardless of the ionomer types. ( 3 )Th e ternary blend containing ionomer showed neg-ligible remote absorption energy, which manifeststhe validity of the evaluation of Jc values by thelocus method.

Th e second conclusion will be discussed with ref-erence to our previous work.16 In our previouswork, l6we revealed tha t the ionomer enhanced themiscibility of PP and EPDM and the effect was mostprominent at a 5 wt % concentration. The rheolog-ical and morphological studies revealed that ionomerA showed more of a compatibilizing effect betweenPP and EPDM than did ionomer B at the sameconcentration.

In our present investigation, ternary blends con-taining ionomer B showed higher Jc values over theentire ionomer concentration investigated. This maybe due to the inherent higher toughness of the addedionomer B than that of the added ionomer A, ac-cording to the manufacturers data. Even though thecompatibilization by ionomer A is larger tha n tha tby ionomer B, the effect of ionomer addition on theJc values was very similar. The miscibility will surelyaffect the fracture mechanics behavior of polymerblends. More detailed research should be carried outto investigate the relationship of miscibility andfracture toughness, since there is some controversyas to whether Jc by itself is an accepted measure oftoughness.

It can be concluded, however, that the incorpo-ration of a small amount of ionomer increased Jc

8/6/2019 070510804_ftp sushil

8/8

1388

values, probably enhancing the fracture toughnessof the PP/EPDM blend and the enhancing effectis most prominently observed at a certain concen-tration (in this case, 5 wt % ) and the effect is ir-respective of the ionomer types. Note also that theevaluation of Jc integrals, using a simple single-edgenotched specimen, was successfully applicable toinvestigate the fracture behavior of the PP/EPDM/ionomer ternary blends.

REFERENCES1. I. Narisawa, Polym. Eng. Sci., 27( ) , 1 (1987).2. I. Narisawa and M. T. Takemori, Polym. Eng. Sci.,3. D. D. Huang, Polym. Eng. Sci., 30 (2 1) , 1341 (1990).4. H. J. Sue, J. Muter. Sci., 2 7 , 3098 (1992).5. H. Aglan, A. Chudnovsky, A. Moet, T. Fleischman,

and D. Stalnaker, Int. J. Fract., 4 4, 167 (1990).6. F. F. LangeandK. C. Radford, J. Muter. Sci., 6,1197

(1971).7. S. W. Tsui, R. A. Dukett, and I. M. Ward, J. Muter.

Sci., 2 7 , 2799 (1992).8. T. De Kalbermatten, R. Jaggi, P. Flueler, H. H.

29 (1 0) , 671 (1989) .

Kausch, and P. Davies, J. Muter. Sci. Lett., 11 , 543(1992).

9. J. R. Rice, J. Appl. Mech., 35, 379 (1968).10. J. A. Begley and J. D. Landes, ASTM Spec. Techn.11. J. D. Landes and J. A. Begley, ASTM Spec. Tech.12. Y. Huan gand A. J. Kinloch, J. Muter. Sci., 2 7 , 275313. Y. Huang an d A. J. Kinloch, J. Muter. Sci., 2 7, 276314. B. H. Kim and C. R. Joe, Polym. Test., 7,355 (1987).15. B. H. Kim and C. R. Joe, Eng. Fract. Mechan., 32 (Z) ,

225 (1989).16. Y. Kim, C. S. Ha, T. K. Kang, Y. Kim, and W. J. Cho,

to appear.17. H. J. Sue, E. I. Garcia-Meitin, B. L. Burton, and

C. C. Garrison, J. Polym. Sci. Part B Polym. Phys.,2 9 , 1623 (1991).

18. B. Y. Ni, J. C. M. Li, and V. K. Berry, Polymer,32 (1 5) , 2766 (1991).

19. G. H. Pittinato, V. Kerlins, A. Phillips, an d M. A.Russo, SEMITEM Fractography Handbook, Mc-Donne11 Douglas Astronautics Co., New York, 1975.

Publ., 5 1 4 , l - 2 0 ( 1 97 2 ).Publ., 514,24-39 (1972).(1992).(1992).

Received February 1 1 , 1993Accepted August 6, 1993