Printer Friendly

Co-injection Molding: Effect of Processing on Material Distribution and Mechanical Properties of a Sandwich Molded Plate.

The effect of molding parameters on material distribution and mechanical properties of co-injection molded plates has been studied using experimental design. The plates were molded with a polyamide 6 (PA 6) as skin and a 20% glass fiber-reinforced polybutyleneterephtalate (PBTP) as core. Five molding parameters--injection velocity, mold temperature, skin and core temperature, and core content--were varied in two levels. The statistical analysis of the results showed that three parameters--injection velocity, core temperature, and core content--were the most significant in affecting skin/core distribution. A high core temperature was the most significant variable promoting a constant core thickness, while core content was the most significant factor influencing a breakthrough of the core. Mechanical properties, such as flexural and impact strength showed a high correlation with the skin/core distribution. The slight increase in falling weight impact strength of the sandwich molded plates, compared to similar plates molded from PBTP only, could be explained from the failure process, which initiates in the brittle core and propagates through the ductile skins.


In sandwich injection molding (co-injection) two different polymer melts are injected sequentially into a cold mold, forming a skin/core structure. The formation of a skin/core structure can be explained from the mold filling process [1,2]. The polymer injected first will cool and solidify as it comes into contact with the cold mold wall, while a flowing region of molten polymer still exists at the center of the flow channel. When the second polymer is injected it will force the first toward the flow front while flowing beneath the solid layer of the first polymer, until the mold is filled and a sandwich part has been formed.

The possibility of using recycled material in the core can be regarded as one of the main advantages with the process. Large volumes of recycled plastics can be reused in products without significant loss of properties, since the skin, which can be made of virgin material, dominates properties and appearance.

An increased use of the sandwich technique should be favored by a better understanding of the mold filling process and the relation between molding parameters, material distribution and mechanical properties. Mold filling studies of sandwich molding have been reported by some authors (e.g., 1, 3-15). The material distribution is influenced by a number of factors, such as the amount of core, viscosities of skin and core, molding parameters and mold geometry. A natural upper limit to the amount of core that can be injected is the possibility of a breakthrough, i.e., the core breaks through the skin and reaches the mold surface. For simple rotational symmetric parts, the upper limit is between 65% and 75%. For more complicated geometries this limit is reduced [2,6].

Viscosity ratio between core and skin ([[eta].sub.core]/[[eta]]) should be between 0.5 and 5 for optimum mold filling, i.e. the core should have a similar or slightly higher viscosity than the skin [2-5,7-13]. Lower viscosity ratios will lead to a breakthrough, while higher ratios will lead to poor mold filling and a variable core thickness at different parts of the mold.

Important molding parameters are injection velocity and melt temperature [6-10]. A high injection velocity will give the skin less time to solidify against the mold wall; instead it accumulates at the far end of the mold, resulting in a thin skin layer close to the gate and thick skin at the far end of the mold. Melt temperatures will affect the viscosities and thus the viscosity ratio as discussed above. Mold temperature can affect mold filling if the skin viscosity is highly temperature dependent [3].

Few investigations on sandwich mechanical properties have been reported in the literature. Donovan et al. [1] measured Gardner impact strength in plates containing various grades of ABS. including regrind, and showed that the skin had a major influence on impact properties. Tomari et al. [8] studied flexural properties of various polyamide/polycarbonate-sandwich specimens and found that flexural stiffness could be calculated from the composite beam theory assuming good bonding between the materials.

In this report layer thickness and mechanical properties of a sandwich-molded plate were measured as a function of molding parameters using experimental design. Five parameters--injection velocity, mold temperature, skin and core temperature, and core content--were varied in two levels. The statistical evaluation of the results was facilitated by a computer analysis, showing the effects of molding parameters and their interactions on layer thickness and mechanical properties.


Materials used in this study were unfilled polyamide 6, Durethan B 30 S, and glass-filled PBTP, Pocan B3225, containing 20% glass fibers (GF), both materials received from Bayer. The sandwich specimens were molded with PA as skin and PBTP as core. The PA 6 was colored black with 1% masterbatch prior to injection molding, to facilitate identification of the interface between the two materials.

Injection molding was done in a 110 ton K110 S2F injection molding machine from Ferromatik Milacron. It can be used both for conventional injection molding and standard sandwich or mono sandwich molding. In this study the mono-sandwich technique was used for producing specimens [2]. Experimental design was used to study the effect of molding parameters on layer thicknesses and mechanical properties. This means that five selected parameters were varied in two levels: injection velocity, mold temperature, skin and core temperature, and core volume ratio. Table 1 summarizes the experimental design matrix and the various parameter settings. A fractional factorial design was used, neglecting higher order interactions than two-parameter interactions [16], giving [2.sup.5-1] = 16 experiments. The computer program used for analyzing the results, recommends the use of center points; thus four additional experiments with center points were included as shown in Table 1 (experiment 1, 1820), giving a total of 20 experim ents. The low and high level for melt and mold temperatures were chosen close to the supplier's recommended maximum and minimum temperatures. The center points were chosen in between. Levels for injection velocity and core volume ratio were chosen by trial and error; several settings were tried and those leading to a overall satisfactory quality with regard to surface (visual) properties were finally chosen. A slight breakthrough of the core material at the far end of the mold was allowed in some cases, in order to expand the variation in parameter space. Viscosity- shear rate curves for the two materials at the relevant molding temperatures are shown in Fig. 1. Viscosity ratio between core and skin ([[eta].sub.core]/[[eta]]) at a shear rate of [10.sup.3] [s.sup.1], varied between 0.8 and 3.

The geometry of the mold is shown in Fig. 2a. It was a square plate, 100 x 100 mm, with a thickness of 3 mm. It was provided with a film gate, having an entrance width of 50 mm. Single polymer (control) plates were also produced along with the sandwich molded plates using only one cylinder, in order to compare material mechanical properties with sandwich properties. Mechanical properties are shown in Table 2.

Skin layer thickness were measured as a function of molding parameters using a Nikon SMZ-U optical microscope, equipped with a Sony Video graphic printer. The molded plates were cut in eight pieces, as shown in Fig. 2b, and skin layer thickness were measured in two perpendicular directions at thirteen positions. Because of the complete symmetrical material distribution, the number of recorded layer thickness could be reduced to seven, labeled Al to A4 and B1 to B3, as shown in Fig. 2b. Core layer thickness was calculated from actual plate dimensions by subtracting the skin thickness. Three specimens were used for each measurement of layer thickness and all data were fed into the computer.

Specimens for flexural testing, 100 X 12 x 3 mm, were cut from the molded plates both parallel and perpendicular to flow, as shown in Fig. 2a. Flexural testing in three-point bending were performed in a Monsanto tensile tester in accordance with ASTM-D 790, with a support span of 60 mm and a crosshead speed of 20 mm/min. All plates were dried in an oven at +80[degrees]C, 10% RH for 48 hours prior to testing. They were then stored in an excicator until testing, which was made within two weeks. Three specimens were tested for each direction and parameter setting and the results fed into the computer. Some of the fractured specimens were examined in a scanning electron microscope (SEM), type JEOL JSM-T100. Fractured specimens were cut in smaller pieces and gold was sputtered onto the fracture surfaces before inspection in the SEM.

Impact testing was performed in an instrumented falling weight impact tester, using a hemispherical head with a diameter of 10 mm and a drop height of 1 meter. Impact strength was calculated by integrating the force vs. time trace, where force was measured by a piezoelectric crystal. The absorbed energy after full penetration, when the load had returned to zero, was taken as the impact strength. The molded plates were laid flat without clamping, on a circular support containing a central hole with a diameter of 20 mm. The impact properties were measured on specimens not previously dried, but stored at room temperature at 50% relative humidity for approximately 6 months before testing. Five specimens were tested for each value of impact strength.

The PC-Windows program Modde 4.0(R), delivered by Umetri AB, Sweden, analyzed all data from measurements of layer thickness and mechanical properties. The program uses partial least squares together with analysis of variance (ANOVA) to compute regression coefficients for each response. The ANOVA takes into account different sources of variation into the process sensitivity analysis, including the effects of individual factors and interaction between factors and variability due to errors. The results are obtained as effect plots showing effects of molding parameters and two parameter interactions on layer thickness and mechanical properties. A two-parameter interaction means that the effect of a variation of a parameter A depends on the level of another parameter, B.


Material Distribution

Table 3 shows the average values, together with minimum and maximum values, of skin layer thickness, for the 20 different parameter settings. Average skin thickness at position B3 is about 40% higher than at position B1, and average skin thickness at position A3 is about 4 times as high as in position A1.

Skin thickness at position B2 was evaluated as a function of molding parameters and the result from the computer analysis is shown in Fig. 3 as an effect plot. The staple diagram shows the effect of the various parameters and interaction effects and the heights of the staples show the magnitude of the effects sorted in descending order. Positive effects are upward staples and negative effects are downward staples. The two thin horizontal lines above and below the X-axis correspond to a confidence level of 0.95. Smaller effects are not significant. The most significant effects on B2 are obtained from injection velocity and core content. The diagram shows that an increase in these parameters (from minimum to maximum condition) decreases the skin thickness at B2 by 0.09 mm and 0.07 mm, respectively (13% and 10%, compared to average value, see Table 3). Skin and core temperatures have less but opposite effects on B2, implying that an increased viscosity ratio ([[eta].sub.core]/[[eta]]) decreases B2.

Skin thickness at position B3 can be compared to the skin thickness at position B1 by calculating the difference B3-B1. This is an important response since an optimum core distribution would correspond to a minimum of B3-B1. The effect of molding parameters on the skin thickness difference B3-B1 was therefore evaluated; the results are shown in Fig. 4. The most significant effect on B3-B1 is obtained from core temperature, which decreases B3-B1 by 0.13 mm (57%). An increased injection velocity increases B3-B1 by 0.09 mm, while an increased core content decreases B3-B1 by 0.05 mm. Other parameters have less temperature, a low injection velocity, and a high core content.

The effect of molding parameters on skin thickness at position A2 is shown in Fig. 5. The most significant effect on A2 comes from an increased core content, which decreases A2 by 4 mm (63%). A positive effect comes from injection velocity, which increases A2 by 2.5 mm (40%). Other effects from mold- and core/skin- temperatures are smaller. The effect plots for A2 and 132 are similar but there is a notable difference with regard to injection velocity; injection velocity decreases B2 while it increases A2. This means that the core cross section, perpendicular to flow, becomes slightly more circular at high injection velocity [17].

Skin thickness difference A3-A1 was also evaluated. This quantity is also of interest since a low value corresponds to a more uniform core extension. The effect of molding parameters on A3-A1 is shown in Fig. 6. The results look similar to the results for B3-B1, except that the relative sizes of the effects are different.

The most significant parameters are now injection velocity, which increases A3-Al by 5 mm = 50%, and core content, which decreases A3-Al by 4 mm, while core temperature has comparatively little effect. In Fig. 6 can also be seen a significant interaction between core temperature and core content; this interaction is displayed in detail in Fig. 7 when InV, MT, and ST are set at minimum values. The lowest values of A3-Al. corresponding to the most uniform core extension under these optimum circumstances, are obtained with a high core content in combination with a low core temperature.

Skin thickness at position A4, i.e. at the far end of the mold, was also measured. In the case of a breakthrough at position A4, the skin does not reach the mold wall. In this case the distance between the mold wall and the skin flow front was measured and A4 was given this value, but with a negative sign. Figure 8 is photograph of a typical breakthrough.

An effect plot for A4 is shown in Fig. 9. Significant effects on A4 are obtained from core content and core temperature, which both decrease A4. A strong positive effect is obtained from injection velocity. These are all expected results. In Fig. 9 can also be seen an interaction effect between core temperature and core content. This interaction is displayed in detail in Fig. 10 using the same parameter setting as in Fig. 7. A combination of a high core temperature and a high core content gives smaller A4. Values less than zero, corresponding to breakthrough, can be found at the central and upper part of the Figure. The critical line for a breakthrough, corresponding to a skin thickness of A4 = 0, was inserted in Fig. 7. Many combinations of core content and core temperature now become excluded in Fig. 7 since they give a breakthrough. The smallest possible value of A3-A1 without a break-through becomes 8.0 mm.

Mechanical Properties

Flexural modulus and strength were measured in two directions, parallel and perpendicular to flow, as described earlier. Average values for flexural modulus and strength are given in Table 2. Flexural properties did not show large variations with molding parameters. An effect plot for flexural strength along flow is shown in Fig. 11. The most significant effects are obtained from core content and core temperature, which both increase flexural strength (by 4 MPa = 3%). Other parameters have less effect. Flexural strength parallel to flow could show a high correlation with skin thickness at position B3, as displayed in Fig. 12. The line in the Figure represents the least square fit to the data. Both flexural strength and skin thickness B3 correlate well with core content, and Fig. 12 is a reflection of this. Flexural modulus and strength across flow could be correlated with skin thickness at B2 or B1. A correlation between flexural modulus across flow and B2 is shown in Fig. 13.

For three-point bending tests on sandwich plates, the following relation holds for the flexural rigidity [8, 17]:

EI = [E.sub.s][I.sub.s] + [E.sub.c][I.sub.c] (1)

where [E.sub.s] and [E.sub.c] are the bending modulus of skin and core, respectively, and [I.sub.s] and [I.sub.c] are the second moments of area of the skin and core layers. If the bond strength between skin and core is high, [I.sub.s] and [I.sub.c] can be calculated as follows:

[I.sub.s] = b([[t.sup.3].sub.s]/6 + [t.sub.s] [([t.sub.s] + [t.sub.c]).sup.2]/2 [I.sub.c] = b([[t.sup.3].sub.c]/12)} (2)

Where b is specimen width and [t.sub.s] and [t.sub.c] the skin and core thickness, respectively. In the case of no bonding, [I.sub.s] and [I.sub.c] are obtained as:

[I.sub.s] = b([[t.sup.3].sub.s]/6) [I.sub.c] = b([[t.sup.3].sub.c]/12)} (3)

Taking modulus data from single polymer specimens tested in flexure, calculations according to Eqs 2 and 3 were performed. Layer thickness values were taken from position B2. Calculated values were compared with experimental as shown in Table 2. The experimental results corresponded to the case of good bonding and were even slightly (8%-14%) higher than calculated for good bonding.

Some of the fractured sandwich specimens were inspected in the SEM to study skin and core morphology. Typical SEM photographs of fracture surfaces are shown in Fig. 14. The core, consisting of glass fiber-reinforced PBTP, displayed a well-known layered structure, Fig. 14a [18-20]. The fiber orientation in the central part of such a layered structure is usually slightly random but mainly perpendicular to flow, while the fiber orientation in the two outer layers is mainly parallel to flow direction. [The layered structure is often referred to as a skin-core structure [18-20]. This term will not be used here, since it may cause confusion with the sandwich skin and core structure.] The photograph shows that the PBTP core had such a layered structure and the central layer is approximately 1/3 of the core thickness. Similar photographs taken from fractured single polymer plates containing only PBTP with 20% GF also displayed a layered structure with a central layer extending to approximately 1/3 of the thickness, Fig. 14b. Thus, differences in layer thickness cannot explain the difference between calculated and experimental modulus in Table 2. Details of fiber orientation in the various layers could explain the difference and can be evaluated by calculating various fiber orientation parameters [18,19]. Small variations in the moisture content of the PA 6 skin can also have an effect since moisture strongly influences the modulus of PA 6 [21].

Results from impact tests on single polymer plates and sandwich molded plates are shown in Table 4. PA 6 displayed a ductile behavior, giving a high impact strength, while PBTP(GF) displayed a brittle behavior and a low strength. The impact strength of the sandwich-molded specimens was higher than for single polymer specimens containing PBT. The impact strength increased 60% on average by adding a tough PA skin. This is less than expected, since the PA skin was expected to dominate the behavior [1]. Inspection of the fractured specimens by SEM showed that crack growth initiated at the tensile side of the PBT core. The cracks then spread to the tougher PA skins. The cracks formed in the brittle core thus seem to influence the way in which the tougher skins fail by providing areas with high strain and stress concentrations, promoting crack growth into the PA skins. (Adding a brittle paint layer to the surface of a tough polymer can give similar effects. Cracks formed in the paint layer can easily propagate in the polymer.) A printout from one of the impact tested sandwich specimens is shown in Fig. 15, showing typical three consecutive peaks in the force vs. deformation trace. It is possible to identify the first peak with crack initiation of the (brittle) PBTP core, and the following peaks with cracks propagating through the lower and upper skins. Impact strength in general showed a high correlation with skin thickness at position B2, which is explained from the fact that the falling dart impacts the specimens at this position: see Fig. 16.


The main conclusions from this investigation on layer thicknesses and mechanical properties of co-injected PA/PBTP(GF) plates are summarized below.

* A relative constant skin and core thickness along the flow direction is obtained by maximizing core temperature. Injection velocity should be low and core content high.

* A uniform core extension along the flow direction is obtained by adjusting the same three parameters but in different order, i.e.: Injection velocity should be low, core content high, and core temperature high.

* The most important parameter for controlling breakthrough is core content. The core content should be decreased and/or injection velocity increased in the case of a breakthrough.

* Flexural and impact properties show a high correlation with the skin/core distribution in the thickness direction. An increased core thickness increases flexural modulus and strength and decreases impact strength.

* The core, containing glass fiber-reinforced PBTP, displayed a well-known layered structure with a central layer consisting of fibers oriented predominantly perpendicular to the flow direction, surrounded by two outer layers with fibers more parallel to flow.

* Sandwich plates of PA/PBTP(GF) displayed a 60% increase in impact strength compared to control plates molded from PBTP(GF). This limited increase in impact strength can be explained from the failure process, which initiates in the brittle core and propagates through the ductile skins.


Thanks are due to IRECO AB, Institute for Research and Competence Holding, Sweden for financial support. Acknowledgments are due to C. Andersson, M. Eng., for assistance in making the statistical analysis and T. Sj[ddot{o}]holm for injection molding of the specimens.


(1.) R. C. Donovan, K. S. Rabe, W. K. Mammel, and H. A. Lord, Polym Eng. Sci., 15, 774 (1975).

(2.) R. Seld[acute{e}]n J. Inj. Mold. Technol., 1, 189 (1997).

(3.) S. S. Young, J. L. White, E. S. Clark, and Y. Oyanagi, Polym. Eng. Sci, 20. 798 (1980).

(4.) S. S. Young, J. L. White, E. S. Clark, and Y. Oyanagi, SPE ANTEC Tech. Papers. 26, 163 (1980).

(5.) D. J. Lee, A. I. Isayev, and J. L. White, SPE ANTEC Tech. Papers, 44, 346 (1998).

(6.) C. Jaroscheck, Kautschuk Gummi Kunstst., 47, 672 (1997).

(7.) G. Akay, Polym. Composites, 4, 256 (1983).

(8.) K. Tomari, H. Hamada, H. Takashima, and Z. Maekawa, SPE ANTEC Tech. Papers, 40, 2583 (1994).

(9.) G. Schlatter, A. Davidoff, J. F. Agassant, and M. Vincent. SPE ANTEC Tech. Papers, 41, 456 (1995).

(10.) P. Somnuk and G. F. Smith, SPE ANTEC Tech. Papers, 41, 760 (1995).

(11.) K. Kehlmann and G. W. Ehrenstein, SPE ANTEC Tech. Papers, 44, 372 (1998).

(12.) A. Lanvers, S. Galuschka, and W. Tietz, Engng. Plast., 7, 223 (1994).

(13.) F. A. Eigl and G. R. Langecker, SPE ANTEC Tech. Papers, 43, 456 (1997).

(14.) S. C. Chen, K. F. Hsu, K. S. Hsu, and M. C. Jeng, SPE ANTEC Tech. Papers, 39, 82 (1993).

(15.) S. C. Chen, K. F. Hsu, and W. R. Jung, SPE ANTEC Tech. Papers, 40, 671 (1994).

(16.) G. E. P. Box, W. G. Hunter, and J. S. Hunter, Statistics for Experimenters, John Wiley & Sons, New York (1978).

(17.) R. Seld[acute{e}]n, J. Inj. Mold. Technol., 2, 166 (1998).

(18.) K. Friedrich, Plast. Rubb. Proc. Appln., 3, 255 (1983).

(19.) R. S. Bailey, M. Davies, and D. R. Moore, Composites, 20, 453 (1989).

(20.) M. J. Folkes, Short Fibre Reinforced Thermoplastics, Research Studies Press, a division of John Wiley & Sons, New York (1982).

(21.) Durethan Polyamide 6 und 66, Bayer Company publication (1983).
COPYRIGHT 2000 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:SELDEN, R.
Publication:Polymer Engineering and Science
Geographic Code:1USA
Date:May 1, 2000
Previous Article:Spatial and Temporal Control of the Degree of Cure in Polymer Composite Structures.
Next Article:Phosphorylation of Phosphazenes and Its Effects on Thermal Properties and Fire Retardant Behavior.

Related Articles
Weld lines and mechanical properties of injection molded polyethylene/polystyrene/copolymer blends.
In-mold textile laminating gives boost to injection-compression.
Predicting mechanical properties of acrylonitrile-butadiene-styrene terpolymer in injection molded plaque and box.
Morphology and mechanical properties of injection molded articles with weld-lines.
Injection and compression molding of polystyrene/high-density polyethylene blends - phase morphology and tensile behavior.
Statistical Study of the Effects of Processing Conditions on the Impact Strength.
Numerical and Experimental Studies on the Ejection of Injection-Molded Plastic Products.
Nature of Contact Between Polymer and Mold in Injection Molding. Part II: Influence of Mold Deflection on Pressure History and Shrinkage.
Effect of machine compliance on mold deflection during injection and packing of thermoplastic parts.
Plastic shaping by means of IR heating and direct pellet molding.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters