Fracture behavior of glass bead-filled poly(oxymethylene) injection moldings.
Glass bead-filled thermoplastics materials are being used in a wide variety of industrial applications. While unique properties, such as dimensional stability or increased modulus, are the usual motivation for exploiting glass bead-filled composites, special attention must be paid to other mechanical properties, such as yield stress and fracture toughness, since they are often degraded by the presence of glass beads. Poor adhesion between bead and the matrix, is a primary cause of low strength, especially at high bead volume fractions. In addition injection molding of thermoplastics and their composites often involves circumstances leading to weldlines in the injection molded parts. In the injection molding process, weld-lines are formed by the merging or impingement of separate melt flow-fronts. They are a potential source of mechanical weakness and poor appearance. With the ever expanding use of composite materials in demanding load-bearing applications the understanding of the weldline phenomenon and its effect on the in-use performance of fabricated parts is becoming increasingly important.
In the present study, deformation and fracture of glass bead filled polyacetal were studied as a function of glass bead content and glass bead diameter. Deformation of weld and unweld samples was studied through elastic modulus, strength, and the energy to break the samples. Fracture studies involved measurements of the fracture toughness at a low rate using single-edge-notched specimens. These specimens were also used to measure strain energy release rate under an instrumented Charpy impact test. In addition effects of notch tip radius and specimen orientation on fracture toughness were also investigated.
The polymer investigated in the present study is a thermoplastic copolymer of poly(oxymethylene) (POM) produced by Hoechst under the trade name Hostaform C 9021. The copolymer is made from trioxane with small amounts of comonomers. It has a density of 1.41 g/[cm.sup.3], the melt flow index of 9.5 g/10 min, and the crystalline melting point in the range of 164 to 167 [degrees] C.
Glass Bead Content
Three composites containing different amounts of glass bead were also received under the trade name Hostaform. The glass bead-filled grades contained 10, 20, and 30% by weight glass beads. Volume fractions, [[Phi].sub.f], of the glass beads were calculated using the respective densities for each material and the following relationship;
[[Phi].sub.f] = [[Rho].sub.c]/[[Rho].sub.f] [w.sub.f] (1)
where [[Phi].sub.f] is volume fraction of the glass beads, [[Rho].sub.c] is the density of composite, [[Rho].sub.f] is the density of glass, and [w.sub.f] is the weight fraction of the glass beads in the composite.
The densities of the composites were obtained from the data published by Hoechst and density of glass was taken as 2.54 g/[cm.sup.3]. Table 1 lists the exact compositions and material codes.
Glass Bead Diameter
To study the effect of glass bead diameter on mechanical properties, solid glass beads of varying diameters were received from Potters Industries. All three grades were coated with a silane coupling agent of type CPO3. The reported mean diameter range, d, are delineated in Table 2.
Composites were prepared by first mixing pre-weighed quantities of POM pellets and glass beads manually (each of which nominally contained 40% by weight glass beads) followed by compounding in a corotating twin screw extruder (Brabender 330) fitted with a rod die. The melt temperature was kept constant at 210 [+ or -] 10 [degrees] C. The extrudates were then granulated for injection molding of test pieces.
After compounding the exact weight fraction of glass beads was determined for each composite by burning off the polymer in a furnace and weighing the residue. The weight fractions were then calculated and found to be in the range 35 to 38% for all the composites giving volume fractions in the range of 24 to 25% (see Table 2).
Each of the composites and the unfilled resin were injection molded into tensile and flexural bars and square plaques as shown in Figs. 1 through 3. The materials were dried in an oven before molding as per [TABULAR DATA FOR TABLE 1 OMITTED] [TABULAR DATA FOR TABLE 2 OMITTED] manufacturer's recommendations. All the specimens were molded after the machine had attained steady state with respect to the preset melt and mold temperature.
For each material (unfilled and glass bead-filled), the tensile specimens with weldlines (double gated) and without weldlines (single gated) were molded under identical conditions. The mold used to prepare tensile bars consisted of two dumbbell-shaped cavities each 1.7 mm thick as shown in Fig. 1. One of which was filled with the molten material entering the mold from one end and the other was filled with molten material entering from both ends. In the latter case a weldline was formed at the center of the molding by two impinging flow fronts.
For each material (unfilled and glass bead-filled), the flexural specimens of dimensions 120 by 10 by 4 mm were molded using a rectangular cavity which was filled through an edge-gate located as shown in Fig. 2.
Square plaques of dimensions 88 by 88 by 1.5 mm were also injection molded using a single mold cavity. The molten material entered the mold through a single edge-gate or a twin-gate located on one side of the square as shown in Fig. 3. In the latter case, the molten material enters the mold cavity as two parallel flows forming a weld line in the center of the plaque.
To fabricate these test pieces for each grade of material, the processing conditions had to be altered slightly in order to produce complete moldings. The range of conditions employed to produce test specimens for all grades are delineated in Table 3.
Molded specimens were then used to study the deformation and fracture behavior of polyacetal and its composites.
Tensile and Flexural Tests
The molded tensile specimens were tested at room temperature on an Instron testing machine. The majority of tensile tests were performed at a crosshead displacement rate of 5 mm/min. The nominal tensile strength value at yield, [[Sigma].sub.t], was calculated using the peak load determined from the load-displacement diagram ([ILLUSTRATION FOR FIGURE 4 OMITTED]).
The molded flexural bars were tested at room temperature on an Instron testing machine fitted with a three-point bend rig with a span-width of 64 mm and [TABULAR DATA FOR TABLE 3 OMITTED] with the crosshead driven at a constant displacement rate of 5 mm/min. Initial flexural modulus, [E.sub.f], and the flexural strength, [[Sigma].sub.f], were computed from the following expressions
[E.sub.f] = 1/4B (P/[Delta]) [(L/D).sup.3] [[Sigma].sub.f] = 3[P.sub.f]L/B[D.sup.2] (2)
where P is the applied load, [P.sub.f], is the maximum load on the load-displacement diagram, d, is the central deflection, L is the span-width, D is the specimen depth (=4 mm), and B is the specimen thickness (= 10 mm).
Fracture Toughness, [K.sub.c], Tests
The fracture toughness of the materials was measured by the use of single-edge-notched tension (SENT) and single-edge-notched bend (SENB) specimens ([ILLUSTRATION FOR FIGURE 5 OMITTED]).
Plaque moldings of polyacetal matrix and its composites with and without weldlines were cut into a rectangular strip of dimensions 88 by l0 by 1.5 mm in order to produce a series of SENT specimens. The length of each rectangular strip was either parallel to the melt flow direction (MFD) or perpendicular to it ([ILLUSTRATION FOR FIGURE 6 OMITTED]). The initial notch ran either perpendicular or parallel to the MFD. These different directions of crack propagation relative to MFD are referred to as transverse direction (TD) and the flow direction (FD) in the following text. For double-feed moldings, the initial notch in the FD specimens was always inside the weld line. The edge notches were introduced mid-way along the length of the specimens by first forming a saw cut which was then sharpened using a razor blade with a tip of radius of approximately 6 [[micro]meter]. The notch-to-depth ratio, a/D, ranged from 0.1 to 0.6. The tests were performed on an Instron testing machine at a crosshead displacement rate of 5 mm/min. All the specimens were tested using pneumatic grips with sample gauge length of 30 mm.
The critical stress intensity factor, [K.sub.c], the so-called fracture toughness for the SENT specimens was calculated at the peak load using the following equation (1);
[K.sub.c] = [P.sub.c]/BD Y [square root of a] (3)
where [P.sub.c] is the peak load on the load-displacement trace, B is the thickness of the specimen (= 1.5 mm), D is the width of the specimen (= 10 mm), and Y is the geometrical factor introduced to account for the finite width effect. It must be noted that since SENT specimens were tested with their ends clamped, the calculation of the geometry factor, Y, as given by Brown and Srawley (1) was inappropriate. They assume that the tensile force is uniformly distributed across the sample width which is consistent with pin-loading configuration provided that the distance between the pins is not less than three times that of the specimen width. Furthermore, when SENT specimens are pin-loaded, the crack tip experiences a bending effect which would be absent when the specimen ends are fixed. Harris (2) obtained the following expression for Y for the case in which the specimen ends are clamped;
Y = 5[square root of [Pi]]/[square root of 20 - 13(a/D) - 7[(a/D).sup.2]] (4)
The Y function given by Eq 4 does not vary as much with a/D as the Brown and Srawley's function ([ILLUSTRATION FOR FIGURE 7 OMITTED]) and was used throughout this study for analyzing the SENT data.
Series of single-edge notched bend specimens (SENB) having thickness, B, and depth, D, of 4 and 10 mm respectively ([ILLUSTRATION FOR FIGURE 5 OMITTED]) were also prepared using the flexural bars notched at the center by a sharp razor blade. The notch-to-depth ratio, a/D, ranged from 0.2 to 0.6. The specimen support span was adjusted to give a span-to-specimen depth ratio, L/D, of 4. Specimens were placed on the support with the notch edge on the lower side and directly in line with the point of loading. In addition, sets of SENB specimens with notch-to-depth ratio, a/D, ranging from 0.2 to 0.6 were prepared using two V-shaped cutters; one having a tip radius of 0.25 mm and the other 1 mm. All SENB specimens were tested on an Instron testing machine at a crosshead displacement rate of 5 mm/min.
Fracture toughness in bending mode was calculated at the peak load, [P.sub.c], using the following equation;
[K.sub.c] = 3[P.sub.c]L/2B[D.sup.2] Y [square root of a] (5)
where Y for the SENB specimen with L/D of 4 is given by (1);
Y = 1.93 - 3.07(a/D) + 14.53[(a/D).sup.2] - 25.11[(a/D).sup.3] + 25.8[(a/D).sup.4] (6)
Impact Strain Energy Release Rate, [G.sub.c], Measurements
One approach in assessing toughness is to use an impact pendulum and measure the energy, U, absorbed in fracturing a series of notched Charpy bars. For each bar, the depth D, width or thickness, B, and crack length a, ([ILLUSTRATION FOR FIGURE 5 OMITTED]) is recorded. A fracture toughness also known as the critical strain energy release rate, [G.sub.c], is then calculated from a straight line graph according to the equation (3):
U = BD[Phi][G.sub.c] (7)
where [Phi] is a known function of crack length given by (3);
[Phi] = L/18D[Y.sup.2](x).x + [integral of] [Y.sup.2](x).x. dx/[Y.sup.2](x).x (8)
where x = a/D and Y(x) is the geometrical factor which for SENB geometry with L/D of 4 is that given by Eq 6.
To determine [G.sub.c], of the unfilled resin and its composites, flexural bars having depth, D, and thickness, B, of 10 and 4 mm respectively were notched at the center by a V-shaped cutter of tip radius 0.25 mm to obtain notch-to-depth ratio, a/D, ranging from 0.1 to 0.6. The specimen support system was adjusted to give a span-to-specimen depth ratio, L/D, of 4:1.
A CEAST instrumented Charpy pendulum impact tester was employed to carry out the tests on the polyacetal and its composites using an average impact velocity of 0.83 m/s. The pendulum mass was 2.2 kg. This velocity was chosen to ensure that the dynamic behavior of the specimen and the machine did not interfere with the measurements and the subsequent analysis. After fracture, the notch depth of each specimen was measured using a traveling microscope. All tests were conducted at room temperature.
RESULTS AND DISCUSSION
Effect of Glass Bead Content
Tensile and Flexural Tests
The effect of glass bead addition on the tensile load-displacement behavior of the weld and unweld specimen is shown in Fig. 4. The load-displacement of pure POM and its composites showed yielding during which stress whitening took place. The stress whitening phenomenon is thought to be caused by vacuole formation around the glass beads as they separate from the matrix under load. With the addition of glass beads into POM, plastic deformation of the weld and unweld specimens was reduced and preceded by tensile failure. Nominal tensile strength, [[Sigma].sub.t], of the composite as a function of [[Phi].sub.f] for both weld and unweld specimens is shown in Fig. 8 in which each point represents the average value for eight specimens. As can be seen, introduction of glass beads into the POM copolymer resulted in a reduction of the tensile strength of the weld and unweld specimens. The effect of weld-lines on mechanical properties was quantitatively expressed in terms of a weldline factor, F, defined as;
F = Property value of specimen with weldline/Property value of specimen without weldline (9)
The lowest value of F based on strength at yield was 0.98 for these materials; thus implying that for the unfilled and the filled grades, the presence of the weld-lines does not have a significant influence on the tensile strength at yield. Observation of the specimens during the test showed that upon loading, beads were separated from the matrix at comparatively low macroscopical stress levels forming voids which were clearly visible on the specimen surfaces. It was noted that the failure of the majority of the specimens containing weldlines had taken place some distance away from the weldline but within the gauge length. Thus indicating that weldlines do not necessarily represent the weakest zone in the unfilled matrix or its composites. The reduction in tensile strength of the composite with or without weldline with increasing [[Phi].sub.f] suggests that the presence of glass bead has a weakening effect rather than a strengthening one. When the glass beads debonds, the volume fraction of composite carrying the load falls; worse than that, the matrix becomes full of holes analogous to that of Swiss cheese, i.e. it is weakened. These observations are consistent with those reported in literature (4-9) for glass bead filled polymers such as polypropylene, polystyrene, ABS, nylon 6,6, and poly(phenylene oxide).
Several empirical relationships have been proposed (10-12) to describe the tensile strength variation as a function of filler volume fraction, [[Phi].sub.f]. The most commonly used relationship is that of Nicolais and Narkis (10) who derived the following expression from simple geometric considerations for spherical shape fillers;
[Mathematical Expression Omitted]
where [[Sigma].sub.r] is the relative tensile strength defined as the ratio of the composite strength to the resin strength. Equation 10 is plotted in Fig. 9 as [[Sigma].sub.r] vs. [[Phi].sub.f]. The horizontal line in the Figure represents the upper bound relative strength for perfect adhesion and the curve represents the lower bound strength ratio in the absence of adhesion between glass beads and the matrix. The experimental data shows that the addition of the glass bead to POM tends to move the data closer to the lower boundary, thus indicating that there is little stress transfer between the polymer and the glass beads as [[Phi].sub.f] increases. Using the more general power law expression (11);
[Mathematical Expression Omitted]
where a and n are constants depending upon the assumed particle shape and arrangements in the model composite, the present experimental data gives a and n values of 1.52 and .91 respectively.
An equally good fit was obtained by using the theory of Piggott and Leinder (12) given by
[[Sigma].sub.c] = [Lambda][[Sigma].sub.m] - [Chi][[Phi].sub.f] (12)
where [Lambda] is the stress concentration factor and [Chi] is a constant dependent upon the particle-resin adhesion. With reference to Fig. 8, one may observe that the present data fits the above model with [Lambda] and [Chi] values of .99 and 100.4 respectively.
The variation of flexural strength with bead fraction is also shown in Fig. 8. Each point represents the average value for eight specimens. The comparison of the values of the tensile strength and flexural strength indicates that although the variation with [[Phi].sub.f] is comparable, the flexural values are considerably higher than the corresponding tensile values. There are three reasons for this. Firstly, in the tensile test, when particles separate from the matrix, the load thereafter is carried by the polymer matrix. Fracture therefore takes place when the matrix surrounding the vacuoles is under a stress which is equal to the ultimate strength of pure resin. In the flexural test, however, failure can take place by a crack propagating from the tension side to the compression side of the beam. The presence of the beads hinders crack propagation and this could lead to higher strengths in flexural specimens than that of tensile specimens. Secondly, and perhaps more significantly, since the load-displacement diagrams for the POM matrix and its composites were nonlinear before attaining the maximum load (none of the specimens were broken at the maximum load), the assumption of linear elastic deformation becomes invalid as the whole cross-section is yielded. Under such conditions Eq 2 overestimates the flexural strength of the material quite considerably. Finally, since a smaller volume of the material is loaded in bending, the probability of finding a large defect in bend specimen is smaller than that in tension. Consequently a higher strength in bending is expected. Nonetheless, variation of flexural strength with [[Phi].sub.f] is still linear giving [Lambda] and [Chi] values of 1.0 and 154.2, respectively.
The work to break, U, under the load-displacement diagrams of the weld and the unweld specimens was also measured and plotted as a function of the bead fraction, [[Phi].sub.f], as shown in Fig. 10. Whereas for a given volume fraction of filler, the work to break the specimens is not significantly affected by the weldline; it decreases with increasing bead fraction in the present range. This is due to the reduction in overall resin strain and resin strength with increasing [[Phi].sub.f].
The effect of [[Phi].sub.f] on the composite modulus relative to that of the unfilled POM is shown in Fig. 11. It can be seen that the relative modulus, [E.sub.r], increases linearly with [[Phi].sub.f]. These results are consistent with predictions for spherical filled systems, as presented by Einstein (13) in which the relative modulus, [E.sub.r], is expressed in terms of [[Phi].sub.f] as;
[E.sub.r] = 1 + [[Kappa].sub.E][[Phi].sub.f] (13)
where [[Kappa].sub.E] is the Einstein coefficient having a value of 2.5 for spherical shape fillers. The slope of the line in Fig. 11, gives [[Kappa].sub.E] values of 2.46 which is in good agreement with the value of 2.5 given by Einstein for spherical shape fillers.
Fracture Toughness Tests
Figures 12a-d show plots of [[Sigma].sup.2][Y.sup.2] vs. [a.sup.-1] for SENT specimens of POM and its composites. The linearity of the data in the plots implies the validity of applying LEFM to these materials for a/D ratios greater than 0.15. Plots also indicate that the notch orientation and the weldline have no measurable effect on the fracture data. It is apparent that for each composite the fracture toughness, [K.sub.c], as derived from the slope of these plots is not affected by the notch orientation nor is it weakened by the presence of the weldline in the specimen. The composite fracture toughness relative to that of the untilled matrix as a function of [[Phi].sub.f] is shown in Fig. 13. It is evident that the relative fracture toughness, [K.sub.r], decreases with increasing [[Phi].sub.f]. In general the presence of glass beads could either have a toughening effect or a weakening effect. On one hand since the localized stress (hence strain) is highest at the notch tip, particle debonding occurs. A notch tip damage zone is then formed which consists of a porous matrix (of debonded glass spheres). The modulus of this damage zone is obviously less than that of surrounding (undamaged) composite. Since the notch tip stress scales with modulus, the localized stress falls as the tip damage accumulates. This can be thought of as a strengthening mechanism. On the other hand, this damage zone is weaker than the surrounding undamaged composite. Competition exists, then, between the reduction in localized stresses (toughening effect) and the weakened matrix brought about by the creation of holes and voids around particles; hence the toughness of the composite depends upon the one that dominates. From the results presented here, it appears that it is later mechanism which controls the toughness of the glass bead filled POM systems rather than the former.
To study the effect of specimen geometry on fracture toughness a series of 3-point single edge notched specimens were also tested. Fracture toughness values obtained from these tests are given in Table 4 and compared with the [K.sub.c], values obtained by way of SENT specimens.
Although the results of the three-point bend tests further confirm that the incorporation of glass beads into polyacetal matrix reduces fracture toughness, they also imply that polyacetal and its composites are 13 to 18% tougher when tested in bending. Referring to Fig. 13, it is evident that the relative fracture toughness for the single edge notched bend specimens is higher than that of single edge notched tension specimens. It must be noted that in order for [K.sub.c] to be regarded as a material property the state of stress at the tip of the crack must be that of plane strain. To achieve this stress state, the specimen must satisfy certain size requirements as suggested by the ASTM (2). According to ASTM, plane strain conditions are achieved when the dimensions of the test specimen satisfy the following minimum values;
[B.sub.min] = 2.5 [([K.sub.c]/[[Sigma].sub.y]).sup.2] and [D.sub.min] = 2[B.sub.min] (14)
where [[Sigma].sub.y] is the tensile yield stress. These specimen size requirements suggest that for plane strain conditions to prevail, specimen thickness for polyacetal and its composites must be greater than 9 mm and the specimen depth must exceed 18 mm. Since the dimensions of the test pieces used in this study are considerably smaller than the above values, we may regard the measured [K.sub.c] values as the apparent fracture toughness rather than the true plane-strain values, hence the observed geometrical dependence of the [K.sub.c] values.
Table 4. Effect of Specimen Geometry on Fracture Toughness.
[K.sub.c] (MPa[square root of m]) [[Phi].sub.f]% SENT SENB
0 3.75 4.22 6 3.00 3.53 12 2.77 3.21 19 2.54 2.95
The effect of notch tip radius on fracture toughness was also studied using SENB specimens with the notch depth in the range between 0.1 and 0.5D. Figures 14a and b show the relationship between [[Sigma].sup.2][Y.sup.2] and [a.sup.-1] for the unfilled matrix and its composites for notch tip radius of 1 and 0.25 mm respectively. The relationship is linear indicating that the concept of linear elastic fracture mechanics is applicable. [K.sub.c] values obtained from the slope of these lines are plotted in Fig. 15 as a function of the notch tip radius. The fracture toughness increases as the notch tip radius increases. It is noteworthy, that fracture toughness of the untilled POM is affected more by the sharpness of the notch than its composites indicating that the addition of the glass beads reduces the notch sensitivity of the matrix material.
Fracture Toughness Tests Under Impact
The load-time and the load-displacement plots which form the basis of the analysis of the impact test were indicative of brittle fracture with no evidence of excessive yielding. The area under the load-displacement plots up to the point of fracture represents the fracture energy, U, required to fracture a specimen.
Results obtained from these tests for the unfilled resin and its composites are shown in Figs. 16a-d as U vs. BD[Phi]. The linearity of the data in the plots indicates that all materials fracture in a brittle manner, without any significant degree of plastic yielding. The measured values of [G.sub.c] as obtained from the slopes of these plots are shown in Fig. 17 as a function of [[Phi].sub.f]. It can be seen that incorporation of glass bead into the polyacetal matrix has led to less tougher composites, further demonstrating the significance of glass beads in the matrix material.
In addition to [G.sub.c], analysis of the peak load and how it varies with variation in notch depth, enables [K.sub.c] to be obtained using Eq 5. Figure 18 shows plots of [[Sigma].sup.2][Y.sup.2] vs. [a.sup.-1] for polyacetal and its composites under impact. The [K.sub.c] values obtained under impact conditions are given in Table 5 and compared with the slow rate SENB values obtained for the same notch tip radius. It is worth noting that the fracture toughness values are affected by the rate of testing; having lower values at the higher rate (impact).
Effect of Glass Bead Diameter
Tensile and Flexural Tests
The load-displacement behavior of the POM and its composites containing 35% by weight glass beads were studied using range of crosshead speeds. It was observed that the behavior is dependent on the speed at which the specimens were elongated. All the specimens exhibited stress whitening due to void formation around the glass beads. Increasing strain rate (or crosshead speed) decreased the deformation capability (ductility) of the POM and its composites. The relationship between strain rate, [e.sup.o], and tensile stress at yield is shown in Fig. 19. Each data point represents the mean value of five specimens. With these composites, tensile stress at yield increases as strain rate increases. The relationship between strain rate and tensile strength is linear for POM and its composites and could be expressed as;
[[Sigma].sub.t] = [K.sub.1] + [K.sub.2] log([e.sup.o]) (15)
where [K.sub.1] and [K.sub.2] are constants having the values given in Table 6.
The effect of glass bead diameter on tensile yield stress of the composite is shown in Fig. 20. It has been shown (14) that for constant volume of spherical glass beads, the tensile strength of a composite is inversely proportional to the square root of the sphere diameter;
[[Sigma].sub.c] = [[Sigma].sub.m] + [Gamma]/[square root of d] (16)
where [Gamma] is a constant. Although the range of glass bead sizes tested here are not diverse, nevertheless, results do suggest that the above relationship can be applied to glass bead filled POM composites. Results further suggest that the slope of the line, [Gamma], is somewhat invariant with testing rate.
The effect of particle size, d, on weldline strength is shown in Table 7 and compared with that of the unwelded samples tested at the same crosshead speed. Evidently, diameter of the glass beads has no significant effect on weldline strength giving weldline factors, F, in the range of 0.99 to 0.97.
Finally, there was no significant change in flexural modulus with glass bead diameter as indicated by the values given in Table 7.
Fracture Toughness Tests
The effect of glass bead diameter on fracture toughness of the unfilled POM and its composites containing 35% by weight glass beads were studied using SENB specimens notched to a/D ratios ranging from 0.1 to 0.6 using a cutter having a tip radius of 0.25 ram. Plots of [[Sigma].sup.2][Y.sup.2] vs. 1/a as shown in Fig. 21 were then constructed from which [K.sub.c] values were evaluated as before. Evidently, fracture toughness in not a sensitive function of glass bead diameter; it decreases from a value of 3.2 MPa [square root of m] for glass bead diameter of 7 [[micro]meter] to 2.93 MPa [square root of m] for glass bead diameters of 26 and 36 [[micro]meter].
Effect of Interparticle Spacing on Mechanical Properties
It is possible to express the measurements carried out in this study in terms of interparticle spacing, s, which appears in the Lang model (15) and is related to the volume fraction, [[Phi].sub.f], and mean particle spacing, d, by
S = 2d(1 - [[Phi].sub.f])/3[[Phi].sub.f] ((17)
Figure 22 shows the variation of flexural modulus, [E.sub.f], tensile yield stress, [[Sigma].sub.t], and the fracture toughness, [K.sub.c], vs. [s.sup.-1] for all six composites (for comparison the values for the unfilled POM serving as the matrix is plotted at [s.sup.-1] = 0). In addition, all the aforementioned properties were employed to estimate the crack opening displacement, C.O.D, of the unfilled POM and its composites by
[Mathematical Expression Omitted]
Table 5. Fracture Toughness Values for SENB With Notch Tip Radius of .25 mm.
[K.sub.c] (MPa [square root of m]) [[Phi].sub.f]% Impact (Eq 5) Slow rate
0 4.90 5.27 6 3.75 4.36 12 3.36 3.90 19 3.02 3.42
Results indicate that apart from the flexural modulus which shows an increase with decreasing interparticle spacing, s, tensile yield stress, fracture toughness, and the crack opening displacement all decrease with decreasing s.
The increase in modulus with decreasing, s, for small displacements which is the condition upon which our measurements were taken is expected because of the absence of debonded glass spheres at small displacements; large scale deflections would bring about debonding and the modulus would then fall. Indeed the reduction in tensile yield stress with decreasing, s, is attributed to the large displacement bringing debonding of the glass spheres from the matrix well before the yield strain is reached.
It is noteworthy, that the variation of fracture toughness with s is in contrast to what has been reported on glass filled brittle matrices such as epoxy resin (15-17). The crack-pinning mechanism is generally considered to be the major cause of increased toughness with decreasing interparticle spacing. During fracture, a moving crack is momentarily pinned at particle sites within the brittle matrix. This interaction' leads to the bowing out of the crack front between the pinning positions, thus increasing its total length [TABULAR DATA FOR TABLE 7 OMITTED] which in turn leads to an increase in the material's toughness. However, as shown here, the inclusion of the rigid glass spheres into tough polymer matrix such as POM polymer tends to reduce the toughness of the polymer. In such a material, high stresses at the crack tip extend well in advance of the crack tip and cause debonding of the glass spheres well ahead of the tip. These debonded particles become unable to effectively pin cracks and hence have a weakening effect rather than a toughening one. The closer the particle the smaller is the volume fraction of the polymer between the particles and the smaller is the crack opening at the crack tip.
It was the objective of this work to investigate the influence of glass beads on the fracture behavior of poly-acetal. Properties such as tensile strength, flexural modulus, fracture toughness, and the strain energy release rate were all found to be strongly influenced by the addition of the glass beads to polyacetal. The addition of glass beads reduced tensile strength, fracture toughness, and strain energy release rate. The only improvement in the material performance was that of the flexural modulus. Results also showed that the presence of the weldline in the specimen do not affect tensile strength or fracture toughness of polyacetal and its composites. Moreover, notch orientation had no measurable effect on fracture toughness, indicating the absence of anisotropy in the glass bead filled molded components.
The effect of glass bead size on the above mentioned properties was also studied. Results indicated the tensile strength of the composite is inversely proportional to the square root of the glass sphere diameter. The effect of bead size on flexural modulus and the weldline strength of the composite was almost insignificant. Fracture toughness of the glass filled POM composite showed slight decrease with increasing bead size.
1. W. F. Brown and J. E. Strawley, "Plane strain crack toughness testing of high strength metallic materials," ASTM STP 410, pp. 1-66 (1966).
2. D. O. Harris, J. Bas. Eng., 49, 89 (1967).
3. E. Plati and J. G. Williams, Polym. Eng. Sci., 15, 470 (1975).
4. N.J. Mills, J. Appl. Polym. Sci., 15, 2791 (1971).
5. D. L. Faulkner and L. R. Schmidt, Polym. Eng. Sci., 17, 657 (1977).
6. K. L. Trachte and A. T. Dibenedetto, Int. J. Polym. Mater, 1, 75 (1971).
7. H. Zhang and L. A. Berglund, Polym. Eng. Sci., 33, 100 (1993).
8. L. Nicolais, E. Drioli, and R. F. Landel, Polymer, 14, 21 (1973).
9. D. C. Phillips and B. Harris, In: Polymer Engineering Composites, p. 45, M. O. W. Richardson, ed., Applied Science Publishers Ltd., London (1977).
10. L. Nicolais and M. Narkis, Polym. Eng. Sci., 11, 194 (1971).
11. S. Ahmed and F. R. Jones, J. Mater. Sci., 25, 4933 (1990).
12. M. R. Piggott and J. Leidner, J. Appl. Polym. Sci., 18, 1619 (1974).
13. A. Einstein, Ann. der Phys., 19, 289 (1906).
14. H. Hojo, W. Tanura, and N. Kawanura, Polym. Eng. Sci., 140 604 (1974).
15. F. F. Lang and K. C. Radford, J. Mater. Sci., 6, 1197 (1971).
16. F. F. Lange, Phill. Mag., 22, 983 (1970).
17. R. J. Young and P. W. R. Beaumont, J. Mater. Sci., 12, 684 (1977).
|Printer friendly Cite/link Email Feedback|
|Author:||Hashemi, S.; Din, K.J.; Low, P.|
|Publication:||Polymer Engineering and Science|
|Date:||Jul 15, 1996|
|Previous Article:||Three-dimensional finite element analysis of polymer rotational extrusion.|
|Next Article:||Three-dimensional flow simulation of a film-casting process.|