Printer Friendly

Load-Carrying Ability of Polystyrene Products With Molded-in Holes.

The combined influence of stress concentration, weld line, and local distribution of molecular orientation on injection molded polystyrene (PS) plates with a molded in hole is described. Birefringence measurements were used to investigate the molecular orientation, and three-point flexure was used to investigate the mechanical properties. Both blunt notch tests and transverse bending tests were carried out. Blunt notch tests showed a strong variation of the material resistance to fracture around the hole. The molecular orientation increased the material resistance not only at the hole equator, but also at the flow split area. In a large domain, weld line fracture was the dominant fracture mechanism. Transverse bending tests also showed a strong influence of the molecular orientation on the load-carrying ability. The weld line was definitely the weakest region, but the reduction caused by the weld line itself appeared to be small.


Almost any designed product has one or more holes, which can reduce the load-carrying ability, first, because it goes together with stress concentrations (1), and second, in the case of molded-in holes, because of the potentially weak hot weld line (2). The term "hot weld line" is used when melt flows rejoin after an obstacle in the mold. Hot weld lines may also occur after jetting and because of differences in product wall thickness, when there is additional flow after rejoining. The term "cold weld line" is used when the flow stops after a collision of two melt flows. This can be caused by the application of more than one mold gate.

According to linear-elastic theory, the maximum stress at a hole can be increased three times because of the stress concentration (1). Fortunately, there is often enough ductility in the material to reduce the negative effect of the stress concentration even in the case of a brittle material as polymethylmethacrylate (PMMA) (3). With respect to the strength of weld lines, a distinction is often made between brittle amorphous, ductile amorphous and semi-crystalline plastics (2). The strength reduction appears to be the highest for the brittle amorphous polymers. For example, a strength reduction of 25% and more was found for hot weld lines in PS tensile specimens (4, 5).

The strength reduction of hot weld lines is less than for cold weld lines. Partly, this is because the hot weld line strength is always compared to the relatively low bulk material strength perpendicular to the flow direction, i.e., perpendicular to the molecular orientation. Cold weld lines are oriented perpendicular to the flow direction and reduce the relatively high strength in the flow direction. This explains why more attention has been given to cold weld lines than to hot weld lines, although the latter is more frequently encountered.

Injection molded products always have some molecular orientation, particularly at the surface (6, 7). The orientation is caused by extension in the fountain flow, the high shear rates in the shear flow and further inwards possibly by the packing stage. As a result, an angle of 90[degrees] between flow direction and tensile stress direction is always unfavorable. Particularly, this is of importance at a molded-in hole, because the stresses are then increased by the stress concentration and act perpendicular to the weld line.

The strength reduction of hot weld lines is mainly caused by insufficient molecular entanglement at the outer layer where the chain diffusion across the interface is limited by the fast cooling rate. According to Tomari et al. (8), this poorly bonded outer layer serves as a V-notch. They conclude for cold weld lines in PS that this V-notch effect is more important than the real V-notch on the weld line surface, which might arise because of entrapped air or contaminants. Another cause of strength reduction, the unfavorable orientation at the weld line due to the fountain flow (9), is typically related to cold weld lines where the weld line is formed at the end of the flow path. The recovery of the entanglement density essentially depends upon the contact time above the solidification temperature. As a result, the weld line strength is influenced by the process parameters. The melt temperature is indicated as the most important parameter [2].

Design engineers will try to locate holes in areas with low stresses, but often this will not be practical. Subsequently, it can be tried to influence the flow direction such that the highest stresses do not act perpendicular at the weld line. Change of gate position or local modifications of the product offer such possibilities, in combination with modem flow analysis software. The aim of the current research is to quantify the effect of the weld line position on the load-carrying ability of products.



PS plates (PS Dow type Styron 638) with central molded-in holes of 2 mm diameter were injection molded (Fig. 1a). Four sets of plates were molded using different molding parameters as shown in Table 1. The melt temperature was chosen as main variable because of its practical importance for cycle time and product cost, and because of its generally accepted importance for weld line strength. The melt temperature was not measured directly, but the nozzle temperature is used as an approximation. The manufacturer recommends a melt temperature between 200[degrees]C and 250[degrees]C, therefore the applied nozzle temperatures are realistic.

Specimens and Testing

Three-point flexural specimens were taken from the PS plates using a numerically controlled milling machine. The dimensions of these specimens are shown in Fig. lb. The upper specimen (type 1) was loaded using in-plane bending so that the molded-in hole at the bottom acted as a blunt notch. The geometry of this specimen corresponds rather well with the Charpy ISO 179/leB type specimen (10). Rather low displacement rates of 1 and 100 mm/s were applied, which allowed us to measure the force signal with less noise. Figure la shows where these specimens were machined from the plate. The specimens were cut at 22 different angles a to the flow direction (-90 [less than or equal to] [alpha] [less than or equal to] +90[degrees]).

The two lower specimens (type 2 and 3) in Fig. 1b were loaded in transverse bending at 100 mm/s. Specimens (type 2) with the hole in the middle were cut at seven different angles a to the flow direction (0[degrees] [less than or equal to] [alpha] [less than or equal to] +90[degrees]). The lowest specimen (type 3) was cut just along the hole. This type of specimen is shifted over 6 mm compared to the specimens with a hole. These specimens (type 3) were machined only for series B and only at angles [alpha] of -90[degrees], 0[degrees] and +90[degrees]. The specimen types 1 and 2 were machined for all four series.

The tests were performed on a specially designed hydraulic three-point bending set-up. Three samples were tested for each angle except for the specimens without a hole of which four specimens per angle were tested. All specimens were tested as molded.

Fracture Surface Morphology and Birefringence Measurements

The fracture surfaces have been studied using an optical microscope (Jenavert) and a scanning electron microscope (Philips). To investigate the molecular orientation, the optical birefringence has been measured using a specially equipped microscope with linearly polarized light (Jenapol) and an Ehringhaus E6 tilting compensator. PS is an appropriate material for this because the contribution of the residual stresses to the birefringence is hardly noticeable (11). Birefringence measurements were carried out on 0.3 mm thick layers milled out of the plate parallel to the plate surface.


Birefringence Measurements

A full three-dimensional description of the orientation distribution around the hole is very comprehensive. Therefore, we restrict ourselves to the general pattern and to the flow split area. Figure 2 shows a view of the orientation pattern at the hole. The arrows indicate the general direction of the molecular orientation. The skin layer of the hole was highly oriented except at the poles. Beyond the poles, the molecular orientation close to the hole surface was more or less in tangential direction. It is noted that stresses at the hole surface will also act in the tangential plane. The orientation was relatively high in the vicinity of the hole equator. Particularly, at the surface layer of series A there was a narrow zone along the sides of the weld line (almost up to the hole) where the orientation direction was quite precisely parallel to the weld line. The flow split area is very interesting. It appeared that at the plane of symmetry, the orientation was transverse to the flow direction at least betwee n the surface layers.

Quantitative birefringence results of the flow split area at the plane of symmetry are shown in Fig. 3. The birefringence has been measured in a plane parallel to the plate surface starting at 0.01 mm from the hole. The upper graph gives the average birefringence for two layers of series A of about 0.3 mm thick, namely the surface layer and the layer beneath. The lower graph indicates the angle a between orientation direction and main flow direction. By far the surface layer had the higher birefringence, because it included both the orientation peak of the fountain flow and the shear flow. This corresponds to the extensive birefringence measurements of (7) for the same type of PS. According to Nijman (7), the magnitude of both the fountain maximum and the shear maximum decreases with increasing melt temperature. This corresponded to our birefringence measurements of the series A. B. and D. Figure 3 shows that away from the hole, the molecular orientation was more or less in the flow direction. Closer to the hole, the orientation reduced and became zero at 2 mm from the hole in the subsurface layer and at about 0.4 mm in the surface layer. Subsequently, at the subsurface layer, the molecular orientation rotated over 90[degrees], which is highly profitable when the stress concentration acts in this area. Stress concentrations are usually only noticeable up to a distance equal to the hole radius. In this case it is not beyond the area of transverse orientation. At the surface layer, the orientation pattern is more random up to 0.4 mm from the hole. However, from the viewpoint of load-carrying ability, this is more profitable than the main orientation direction.

Material Resistance Around the Hole

We have studied the material resistance around the hole using blunt notch specimens loaded with in-plane bending (Fig. I b). The peak stresses were strongly concentrated at the hole surface, because both the bending stress maximum and the stress concentration were present over the entire plate thickness at the hole. To simplify the mutual comparison of the different results, we convert the load during the test into a nominal stress [[sigma].sub.n]. The nominal stress is defined as the maximum bending stress based on the assumption of linear elastic behavior and neglecting the effect of stress concentrations. The nominal stress equals the ratio between net section moment and net section modulus (12). This results in a linear relation between the maximum force measured during the test From and the corresponding maximum nominal stress [[sigma].sub.n] max.

[[sigma].sub.n,max] = 0.25[F.sub.max]* L/((1/6)t[[].sup.2])= 93 * [F.sub.max]/(t * [[].sup.2]) (1)

In this relation: L is the span (L = 62 see Fig. 1b), t is the thickness (Table 1), and [] is the net width (8 mm for these specimens).

Figure 4 plots the maximum nominal stress during the test versus the angle between specimen axis and main flow direction for series A, B, C and D (cross-head speed is 1 mm/s). The weld line position is illustrated for three situations. Figure 4 is symmetrical at -90[degrees] and +90[degrees] and therefore describes the total range of 360[degrees]. The nominal stress reached a for [alpha] = +90[degrees], when the weld line was situated at the net cross section. There was a negligible effect of the applied injection molding conditions on the weld line strength. The Figure shows a fast increase of the maximum nominal stress when the angle a decreases, down to an angle a of about In this range of a for all specimens, the crack initiated at the weld line. That qualitatively explains the concave shape as was apparent from a finite element method analysis (FEA).

Beyond this range, the crack initiated in the bulk material outside the weld line. Close to a [alpha] = 0[degrees] when the bending stresses were parallel to the weld line, the maximum nominal stress reached a maximum. An additional advantage is that the molecular orientation at [alpha] = 0[degrees] is relatively high in the area of the stress concentration. There was a clear effect of the molding conditions when fracture initiated outside the weld line. The lower the melt temperature, the more molecular orientation and the higher the maximum nominal stress.

The maximum nominal stress decreased slowly in the area of negative a-values, until a minimum was reached for a = -90[degrees], when the stress concentration acted on the flow split area. Evidently, the maximum nominal stress in this area is still negatively correlated to the melt temperature. This can be explained by the transverse orientation in this area as mentioned above (Fig. 3). This also explains why this minimum is still 53% to 75% higher than the weld line minimum.

Figure 5 shows the relation between the angle [alpha] and the fracture energy, defined as the area under the force-displacement curve. This relation corresponds with Fig. 4, but the difference between the extremes is much greater.

In Fig. 6 and 7, the above-mentioned results of series A and D obtained at a crosshead speed of 1 mm/s are compared with the results obtained at 100 mm/s. Higher crosshead speed resulted in higher maximum load and higher fracture energy, especially for series D and even for the weld line strength. Apparently, there is hardly any effect of molding conditions at 100 mm/s in the total [alpha]-domain.

Load-Carrying Ability

Transverse bending is the dominant mode of loading for many products. Therefore, the three-point bending specimens loaded in transverse bending were used to study the load-carrying ability (Fig. lb). For these specimens, the relation between maximum nominal stress and maximum force is:

[[sigma].sub.n.max] = [F.sub.max] * 93/([t.sup.2] * [] (2)

Figure 8 shows the maximum nominal stress plotted versus the angle [alpha] between flow direction and specimen axis for the specimens with a hole in the middle. The range of 90[degrees] describes the entire a-domain, because of symmetry. Again the maximum was at [alpha] = 90[degrees]. The load-carrying ability remained more or less constant up to a was about 45[degrees] At higher [alpha]-values, the load-carrying ability was reduced to a minimum of 46% to 49% for [alpha] = 90[degrees]. From the viewpoint of the designer, the convex shape of the graph is more favorable than a concave shape as in Fig. 4. Only the specimens with [alpha] = 90[degrees] had a plane of fracture that completely followed the weld line. Often, in the transition area ([30[degrees] [less than or equal to] [alpha] [less than or equal to] 75[degrees]), the plane of fracture coincided over a small area with the weld line. However, different from the blunt notch tests, no indication was found that the crack initiation point was on the weld l ine. Sometimes, almost exclusively for specimens with [alpha] = 75[degrees], fracture occurred beside the hole. This was probably due to the high orientation in this area.

It is noteworthy that there was hardly any effect of the melt temperature or of omission of the packing phase. Only series A with the lowest melt temperature had a slightly higher load-carrying ability.

Figure 9 shows the fracture energy plotted versus the angle [alpha]. Small differences in wall thickness were corrected assuming a linear relation between the wall thickness and the fracture energy. The minimum fracture energy ([alpha] = 90[degrees]) was only 10% to 12% of the maximum value ([alpha] = 0[degrees]). which implied a much greater reduction than was found for the maximum nominal stress (Fig. 8). Also, Fig. 9 showed no special effect of the molding conditions.

Specimens without a hole were also tested in transverse bending. In Table 2 and 3 the results are compared to the above-mentioned results of the specimens with a hole in the middle. The fracture energy of the specimens without a hole was corrected for the difference in net cross section. Table 2 and 3 clearly show the dominant influence on the load-carrying ability of the angle [alpha] between flow direction and bending stress direction. The effect of the hole (based on the net cross section of the specimen) was much less. The first two rows of Table 3 show a fracture energy reduction of only 8% caused by the hole for [alpha] 0[degrees]. The three lower rows of Table 3 show only a 13% higher fracture energy for the specimen with no hole and with the flow split area in the mid section ([alpha] = -90[degrees]) as compared to the two other specimens. This difference was small because fracture of this specimen did not initiate in the area with transverse orientation, but more away from the flow split point where the molecular orientation occurred in the main flow direction.

With respect to the maximum nominal stress (Table 2), the disadvantageous effect of the stress concentration was not found. For [alpha] = 0[degrees] even a 26% higher maximum nominal stress was found for the specimen with a hole. This could be explained by the relatively high orientation near the equator of the hole.

Flow Split Area, Stress Concentration and Weld Line

The blunt notch tests showed a relatively high material resistance at the flow split area due to the transverse molecular orientation. However, from the transverse bending tests (type 3), we have learned that it did not result in a considerably higher load-carrying ability because of the small dimensions of this area. This is illustrated in Table 4, in which for series B the results of the blunt notch test (Fig. 4 and 5), and the transverse bending test of the specimens without a hole (Table 2 and 3), are compared. The first row shows for the blunt notch test a 58% higher maximum nominal stress for the flow split area compared to the weld line, while for transverse bending the difference is only 7%. The second row of Table 4 shows a high difference in maximum nominal stress for both loading cases. A similar conclusion is valid for the fracture energy as followed from the lower two rows in Table 4.

We have concluded that the stress concentration at the hole has only a small effect on the load-carrying ability. This corresponds to earlier tests in our laboratory on PMMA (3). Indeed PS is a rather brittle material, but apparently the material had enough ductility in the applied test situation, that could cause some redistribution of stresses. This might also hold for the crack-like weld line defect. Consequently, it explains its rather low effect on the load-carrying ability and perhaps also the negligible effect of melt temperature. With regard to this it is mentioned that a depth of only 0.2 to 0.3 mm was found by Tomari et al. (8) for the poorly bonded layer of cold weld lines PS.

The small reduction of the load-carrying ability due to hot weld lines loaded in bending was earlier reported by Selden (13) for four out of five (filled and unfilled) materials. A much greater strength reduction of 25% and more is reported (4, 5) for PS tensile specimens with hot weld lines as is mentioned above. Perhaps this is because of the tensile loading, but the opposite effect appears to be obvious for surface defects. Criens et al. (5) applied a net cross section that was only twice the hole diameter; therefore, the negative effect of the stress concentration might be higher. Also, the PS grade we used has a relatively low average molecular mass, and this enables an increase in the degree of bonding across a weld line (14). For impact bending a reduction of 50% to 80% is reported by Criens et al. (5) for the load-carrying ability of hot weld lines in PS. This suggests a transition to a greater reduction of the hot weld line ductility than a reduction of the bulk material ductility perpendicular to t he molecular orientation.


The following conclusions can be drawn from the blunt notch tests:

* The material resistance around the hole varies strongly. The weld line is definitely the weakest region.

* The material resistance at the flow split area is favored by the molecular orientation.

* The weld line strength is not influenced by the applied variation of melt temperatures or by leaving out the packing phase.

* Beyond the domain of weld line failure there was an effect of the molding conditions on the material resistance. It is not clear why this effect occurred only at low loading rates.

From the transverse tests we concluded:

* The load-carrying ability is strongly influenced by the angle a between flow direction and stress direction. Therefore, it can be useful for designers to predict the flow direction aided by flow simulation analysis.

* The load-carrying ability hardly reduces if 0[degrees] [less than] [alpha] [less than] 45[degrees]. This is a safe [alpha]-domain, if the bending stresses of one direction are dominating. In fact this is not influenced by the presence of a hole. We might say that in this case the molding situation (i.e., the location of the gate) is more important than the geometry itself (i.e., the location of the hole).

* The weld line causes only a small additional reduction of the load-carrying ability in transverse bending (say about 7%).

* The flow split area with relatively high material resistance is too small to cause a noticeable increase of the transverse load-carrying ability.

* The stress concentration of the hole has only a small effect on transverse bending strength.

* Process conditions hardly influence the load-carrying ability in transverse bending.

The conclusions above are restricted to the field of investigation. For example, the small difference in Fig. 2 between series A and the other series might be the beginning of a transition to a greater effect for lower melt temperatures. Therefore, further research must be carried out on larger holes with larger areas of stress concentration. Also, the effect of loading rate has to be studied further. It is likely that at higher strain rates the effect of stress concentration and probably also the effect of weld line and melt temperature will be higher." This might influence the size of the "safe [alpha]-domain."


(1.) R. E. Peterson, Stress Concentration Factors, John Wiley & Sons, New York (1974).

(2.) S. Fellahi, A. Meddad, B. Fisa, and B. D. Favis, Adv. Polym. Technol. 14, 169 (1995).

(3.) M. J. M. van der Zwet and A. J. Heidweiller, J. Appl. Polym. Sci., 67, 1473 (1998).

(4.) J. L. Williams and K. J. Cleereman, Styrene--Its Polymers, Copolymers and Derivatives, 490, R. H. Boundy and R. F. Boyer, eds., Reinhold Publishing Corporation, New York (1952).

(5.) R. M. Criens and H. G. Mosle. in Failure of Plastics, W. Brostow and R. D. Corneliussen, eds., Hanser, New York (1986).

(6.) L. Hoare and D. Hull, Polym. Eng. Sci., 17, 204 (1977).

(7.) G. Nijman, On the Origin of Molecular Orientation in Injection Moulded Products, PhD thesis, University of Twente, The Netherlands (1990).

(8.) K. Tomari, S. Tonagai, T. Harada, H. Hamada, K. Lee, T. Morii, and Z. Mackawa, Polym. Eng. Sci., 30, 931 (1990).

(9.) E. M. Hagerman, Plast. Eng., 29, 67 (1973).

(10.) Int. Org. for Stand., ISO 179, Plastics-Determination of Charpy Impact Strength, Geneva (1993).

(11.) R. Wimberger-Friedl, Orientation, Stress and Density Distributions in Injection-Moulded Amorphous Polymers Determined by Optical Techniques, p. 8, PhD thesis, Eindhoven University of Technology, The Netherlands (1991).

(12.) J. M. Gere and S. P. Timoshenko, Mechanics of Materials, Van Nostrand Reinhold (UK) Co. Ltd., Berkshire, England (1987).

(13.) R. Selden, Polym. Eng. Sci., 37, 205(1997).

(14.) R. P. Koster, J. Inj. Molding Technol., 3, 154 (1999).
Table 1.
Main Molding Conditions for Series A, B, C and D and Final
Plate Thickness.
 A B C D
 Nozzle temp. 205 215 215 230
Packing pressure 30 30 0 30
Plate thickness 2.02 1.95 1.93 1.94
Table 2.
Maximum Nominal Stress During Transverse Bending for Specimens With
and Without a Hole. Mean Value and Coefficient of Variation
(c.o.v.) of Series B is Indicated.
 Angle [alpha] [[sigma].sub.n, max] c.o.v.
Hole [alpha] = 0[degrees] 102 [MPa] 0.3%
No hole [alpha] = 0[degrees] 81 [MPa] 1.6%
Hole [alpha] = +/-90[degrees] 49 [MPa] 3.8%
No hole [alpha] = -90[degrees] 47 [MPa] 2%
No hole [alpha] = +90[degrees] 44 [MPa] 0.9%
Table 3.
Fracture Energy During Transverse Bending for Specimens With and
Without a Hole. Mean Value and Coefficient of Variation (c.o.v.) of
Series B is indicated.
 Angle [alpha] [[sigma].sub.n, max] c.o.v.
Hole [alpha] = 0[degrees] 268 [Nmm] 3.5%
No hole [alpha] = 0[degrees] 290 [Nmm] 12%
Hole [alpha] = +/-90[degrees] 31 [Nmm] 5.6%
No hole [alpha] = -90[degrees] 35 [Nmm] 3.9%
No hole [alpha] = +90[degrees] 31 [Nmm] 1.5%
Table 4.
Ratios of Maximum Nominal Stresses and Ratios of Fracture
Energies ([U.sub.f]) as Established for Two Loading Cases of Series B.
 Loading Case
 Blunt Bending
 Notch (No Hole)
[[sigma].sub.n], max] ([alpha] = 1.58 1.07
-90[degrees])/[[sigma].sub.n], max]
([alpha] = +90[degrees])
[[sigma].sub.n, max] ([alpha] = 2.42 1.85
0[degrees])/[[sigma].sub.n, max]
([alpha] = +90[degrees])
[U.sub.f] ([alpha = -90[degrees])/ 1.92 1.13
[U.sub.f] ([alpha] = +90[degrees])
[U.sub.f] ([alpha] = 0[degrees])/ 4.85 9.45
[U.sub.f] ([alpha] = +90[degrees])

[Graph omitted]

[Graph omitted]

[Graph omitted]

[Graph omitted]

[Graph omitted]

[Graph omitted]

[Graph omitted]

[Graph omitted]
COPYRIGHT 2001 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Polymer Engineering and Science
Article Type:Statistical Data Included
Date:Aug 1, 2001
Previous Article:A Uniform Phenomenological Constitutive Model for Glassy and Semicrystalline Polymers.
Next Article:Nonlinear Viscoelastic and Viscoplastic Response of Glassy Polymers.

Related Articles
Weld lines and mechanical properties of injection molded polyethylene/polystyrene/copolymer blends.
Impact behavior of microcellular foams of polystyrene and styrene-acrylonitrile copolymer, and single-edge-notched tensile toughness of microcellular...
Evaluation of collisional damage in polystyrene foam constructions using a dual hammer impact test.
The craze initiation response of a polystyrene and a styrene-acrylonitrile copolymer during physical aging.
A parametric study of the shock characteristics of expandable polystyrene foam protective packaging.
An Experimental Matrix Design to Optimize the Weldline Strength in Injection Molded Parts.
Effect of supercritical carbon dioxide on PMMA/rubber and polystyrene/rubber blending: Viscosity ratio and phase inversion.
Influence of molecular weight on strain-gradient yielding in polystyrene.
Optimal design and manufacture of thin-walled polystyrene structures.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters