Influence of processing on quality of injection-compression-molded disks.
Precision molding is the trend of the injection molding technology. The focuses in the precision molding are on dimension, shrinkage, warpage, frozen stress, and orientation distribution. These qualities are determined by the comparatively long stages after filling, namely the post-filling process. Experiments have revealed that, though post-filling is critical to part quality, the control over post-filling through pressure and temperature is limited and ineffective. To reduce shrinkage and control local shrinkage, the only way is to raise the peak cavity pressure. However, packing flow is coupled with serious molecular orientation. Furthermore, the control of post-filling through pressure is lost after gates freeze-off. These limitations hinder the capacity of a conventional molder for precision molding. These inherent limitations are due to the nature of the packing-based densification mechanism whereby melt is forced to flow from small gates into a filled cavity to density the molded part.
Compression molding uses another densification mechanism. Pressures applied to the whole area of the molding from the cavity wall. Compression molding thus retains the advantage of modest deformation and no region of sharp stress gradient. However, as compared by Tucker (1), compression molding does not have the process advantage offered by injection molding in material handling and process automation. The fundamentals of injection and compression molding have also been investigated recently by several researchers (2). Wu and White (3) studied the birefringence in the parts shaped by injection and compression molding processes. They found that the principal direction of the refractive index tensor in injection-molded parts is different from the flow-thickness direction. Also, there is a birefringence peak in the core zone in injection-molded parts. On the other hand, they found that the principal directions of refractive index tensor in compression-molded parts coincide with the natural Cartesian coordinates. The birefringence changes smoothly from center to wall. They concluded that thermal stress is the main cause of birefringence in compression-molded parts.
Compression has been added to the post-filling process of injection molding for super precision molding. The so-called injection compression molding (ICM) has been used to produce precision parts such as lenses and compact disks (Klepek (4), Matsuda (5)). Injection compression molding incorporates both the packing-based and compression-based densification mechanisms in sequence. With aid of compression, super precision parts can be produced in an injection-based cycle.
Ke (6) experimentally studied the effects of adding compression to conventional injection molding process. It was found that the thermomechanical history during injection compression molding (ICM) process forms a quadrilateral-shaped path with the atmosphere isobar in the P-v-T diagram, with all three sides controllable by pressure. On the conventional injection molding (IM) process, however, the triangular-shaped path has only one side controllable through pressure (as illustrated in [ILLUSTRATION FOR FIGURE 1 OMITTED]). Injection compression molding can thus produce parts with Better dimensional control than those from conventional molding. Compression also improves conformity to the cavity profile and reduces lateral shrinkage significantly without adding much residual strain (7).
It is well known that processing parameters in molding determine the thermomechanical process experienced by the melt inside the cavity. Liu and Manzione (8) performed the process studies in precision injection molding. They pointed out that micro-level precision has been largely beyond the capacity of the conventional molding process. With high performance engineering plastics and new precision injection molding machines, they designed experiments to determine the optimal setting for micro-level tolerance molding. The four chosen injection molding parameters are barrel temperature, mold temperature, injection speed, and cut-off pressure.
Injection compression molding enhances capacity for precision molding. In addition to geometry and temperature, ICM involves compression-related processing parameters such as compression start-up time, and compression force, etc. The present study is devoted to investigate the effect of some relevant parameters on the quality of injection-compression-molded disks.
A mold maintained at temperature Tm has a cavity with characteristic length D and thickness h. The polymer melt at temperature To was injected and packed into the cavity. The thermal diffusivity (k/[Rho]Cp) of the melt at To is [Alpha]. Its compressibility (-1/v[(dv/dP).sub.T]) at To is [Beta]. The glass transition temperature of the polymeric material is [T.sub.g]. At the very start of filling, the melt was compressed. The compression is released at time [t.sub.e] after the start of filling. The compression pressure is P. The final part thickness is h - [Delta]h. The lateral dimension is D - [Delta]D. The compression effects in terms of dimensional shrinkage such as [Delta]h/h, [Delta]D/D, and other qualities can be expressed as functions of the following dimensionless variables for melt temperature, compression start-up time, compression pressure, and part thickness respectively:
[T.sup.*] = To - Tm/Tg - Tm,
[P.sup.*] = [Beta]P,
[t.sup.* ] = [Alpha]to/[h.sup.2],
[h.sup.*] = h/[([Alpha][t.sub.e]).sup.0.5].
Injection compression molding operations were conducted with a modified injection molding machine (Cheng-Shong SM-80, Taiwan), a machine with 80-ton clamping force powered by a hydraulic pump with 10 hp capacity. The screw diameter was 31 mm with a 22:1 length/diameter ratio. It has all the basic components of a typical injection molding machine. The additional compression cylinder was mounted on the moving platen, with 10 mm stroke and a maximum force of 25 tons (6).
A mold for disk moldings was constructed for this study. The disk cavity is 51.2 mm in diameter with a central gate of 6.50 mm. The thickness can be varied with spacer rings. The mold was constructed from a typical three-plate mold unit. It has two design features adapted for the injection compression process: a sliding mold bottom-plate and a shut-off mechanism. The sliding bottom-plate is connected to the ram head of the compression cylinder via a connecting rod-and-disk to transmit the compression. On the other side of the cavity is the shut-off mechanism to prevent back flow during compression. A shut-off pin on an orientation interlock can slide up-and-down along a channel in the floating plate of the three-plate mold. During mold filling and packing, a hole in the shut-off pin well matches the runner system so that the melt can flow into the cavity (6). The flow path was blocked as the pin is hammered down to shut the passage. A schematic diagram of the mold in gate-open and gate-shut positions is shown in Fig. 2.
To measure the cavity pressure, a piezo-electric pressure transducer (Kistler 6159) was flush-mounted at radius of 20.6 mm in the bottom-plate of the cavity. The pressure-caused charge was converted to voltage through a charge amplifier (Kistler 5041). The voltage was collected with a PC-based data acquisition system with a 12-bit A/D converter. An injection grade polystyrene, Polyrex PG-79 (CHI-MEI, Taiwan), was used in this study.
The aim of process optimization in precision molding is to produce parts with acceptable dimensional accuracy and stability, with surfaces geometrically conformable to that of cavity, and with low stress and orientation level to avoid subsequent distortion. To evaluate the quality of moldings, the following two aspects are focused upon: 1) dimensional accuracy, stability, and geometrical conformability as indicated by deviation from mold dimensions (shrinkage) and by part flatness, and 2) residual stress and molecular orientation as indicated by birefringence. The former were evaluated through measurement using micrometers and dial indicators, with micrometer accuracy; the later was evaluated using a circular polariscope.
INJECTION COMPRESSION MOLDING PROCEDURE
During cavity filling and the first stage of packing, the process is exactly the same as in conventional injection molding. When some delay after gates are mechanically shut-off, the compression mechanism pressed the slidable mold wall to compress the resin in the cavity, which was injected and partially packed in the cavity previously. The cooling continues to take place as the resin is compressed up to the instant the mold is open. The injection compression process can be divided into four stages: filling, intermediate packing, compression, and cooling.
RESULTS AND DISCUSSION
The Effect of Compression Start-Up Time
The typical pressure-time traces of process with the same filling and packing, but with different compression start-up times are shown by solid lines in Fig. 3. The maximum compression pressure drops with compression delay. The resultant part quality as shown in flatness and diameter deviation from the cavity dimension are shown in Fig. 4. The flatness and shrinkage are improved as the compression delays are shorten.
It is worth noting that the effect of compression is still significant, even when the compression is activated after the detachment point (point D as shown in [ILLUSTRATION FOR FIGURE 3 OMITTED]). The detachment point is defined as the instant at which the cavity pressure drops to atmospheric. In conventional injection molding, the part is no longer pressed in contact with the mold cavity after the detachment point; it cools and shrinks without the constraint form the cavity wall. According to Greener (9), in conventional injection molding, the temperature distribution in the part at the detachment point determines the conformability of the part to the cavity. Upon detachment, if some portion inside the part is still above the glass transition temperature, the part will not be vitrified and shaped under the constraint of the cavity wall. The surface of such parts will not conform well to that of cavity. This causes premature shrinkage. Well controlled shrinkage can be obtained if the temperature inside the part is uniform at the detachment point. In conventional molding, the only way to delay the detachment point for precise conformability is to raise the peak cavity pressure. However, the packing pressure is limited by the machine capacity, and packing flow induces orientation. In injection compression molding, however, the attachment of the part to the cavity can be delayed or resumed with compression. Compression thus can effectively control the timing and pressure in part-to-cavity contact during the post-filling process.
Polymer chains possess different polarizabilities, [[Alpha].sub.1] and [[Alpha].sub.2], along the perpendicular to the molecular backbone. The refractive index thus has a similar directional dependence. This results in optical anisotropy, known as birefringence. The birefringence patterns and values of the disks with various compression start-up times are shown in Figs. 5 and 6 respectively. The compression delay causes significant increase in birefringence. Birefringence displays one of the most significant effects of compression start-up timing.
Interpretation of the Birefringence Difference in Disks Molded With Various Compression Start-Up Times
Birefringence in an injection-molded part has been recognized to be the result of two distinct phenomena: orientation-caused and thermal-stress-induced birefringence (10, 11). First, molecular orientation is developed as melt is injected and packed into the cavity; the orientation is "frozen-in" during vitrification (12, 13). This frozen-in or residual alignment of molecular induces orientation-caused birefringence. This can be further distinguished as filling-induced and packing-induced orientation (14). A second maximum inside the core region can be observed only on parts molded with packing (15). Spencer and Gilmore (16) have shown that increasing packing time increases residual orientation. Secondly, thermal stresses are developed during the densification associated with inhomogeneous cooling following vitrification. The thermal stresses, which are tensile in the core and compressive near the surface, also induce birefringence. Wu and White (3) examined this phenomenon by studying the birefringence distributions in a compression-molded part. They found that the principal axis is normal to the mold wall. The stresses are compressive near the mold wall and tens fie at the core. This is what one would expect from the theory of thermal stresses. They confirmed that densification compression causes little flow, and birefringence in a compression-molded part is mainly due to thermal stresses.
The attempt to decouple orientation-caused and thermal-stress-induced birefringence in an injection-molded part has been done by Isayev (17) and Wu et al. (3). They assumed that the contribution of thermal stresses to birefringence can be approximated by the birefringence measured from a compression-molded part quenched at a cooling rate similar to that in injection molding. The birefringence was subtracted from the birefringence of injection-molded parts. The remaining birefringence is counted as orientation-induced and is expressed by an orientation factor.
Compression start-up timing in injection compression molding does not affect the cooling rate significantly, neglecting the compression heating. Therefore, a compression delay does not change the thermal stress developed- and thermal-stress-induced-birefringence. The primary birefringence difference in parts molded with different compression delays must come from the molecular orientation induced with intermediate packing. A late gate shut-off and longer compression delay allows more packing flow, and thus induces more orientation-induced birefringence. This has been confirmed with interrupted-packing experiments (18). Parts with less packing due to early gate shut-off have less packing-induced orientation and less birefringence.
These findings provide guidelines for gate shut-off and compression start-up timing control. Densification can be performed by either packing or compression. Packing brings in melt and induces orientation-caused birefringence. To avoid excessive packing-induced orientation, the compression should start as early as possible. However, the cavity must be filled and partially packed with enough material for desired part thickness. According to Friesenbichler et at (19) and Yang et al. (18), the melt packed in after peak cavity pressure brings in negligible mass but significant orientation, and the compression definitely should start before peak cavity pressure. For a system without a mechanical gate shut-off mechanism, the gate size should be carefully designed to ensure in-time gate freeze-off before compression start-up as well as enough material flow and packing in before gate freeze-off. With a gate shut-off mechanism available, precise timing for flow termination can be ensured. Large gates can be used for fast filling and effective packing. A mechanical shut-off mechanism would be an essential device for injection compression molding system.
Effect of Compression Pressure
With the same setting for melt filling and compression start-up, the compression force was varied to study the effects of compression pressure. The maximum cavity pressure measured during compression is used as an indication of compression force.
The part quality shown as flatness and deviation from the cavity diameter are shown in Fig. 7. Both the flatness and shrinkage reduction are improved with the compression force. Little birefringence difference can be seen in parts molded with different compression forces. The deviation from original cavity thickness increases as the compression force increases, as shown in Fig. 8. With the same amount of melt, higher compression pressure results in less thickness. This phenomenon is of primary concern to ICM molders since dimensional control in the thickness direction is strickly demanded. Prediction of the final part thickness call for rigorous simulation of the molding history along with precise description of thermodynamic states. Based on current P-v-T models, the two parts ICM molded with the different compression pressures as shown by dashed lines in Fig. 3 yields the same thickness. The P-v-T model failed to predict the pressure-induced densification.
On The Selection of State Models For Predicting Pressure-Induced Densification
As pointed out by Huilier (20), most state models are based on quasi-equilibrium assumptions, i.e., the density is only a function of pressure and temperature. Current P-v-T models failed to predict the pressure-induced densification. Microstructure-related effects have to be taken into account.
It is recognized that the properties of glassy polymers are controlled to some extent by conditions that prevail in the glass transition zone (21-24). Parts passed the glass transition zone at high pressure resulted at higher density. Greener (21) used a simplified phenomenological model to study the pressure-induced densification in injection molding and explained the observed density profiles in the gap direction. Another approach uses the multi-parameter models. A multi-parameter model proposes that the state of glassy polymers is determined not only by T and P, but also by the nonequilibrium structural state of the glass, indicated by a set of structure-related parameters. A representative one is the KAHR model (25). Hon (26) has simulated the ICM post-filled process with state models described by the KAHR model. Pressure-induced densification can be predicted.
Effects of Melt Temperature And Cavity Thickness
Temperature and thickness are important parameters common to both IM and ICM processes. But their effects may not be of the same trends for both processes.
With fixed compression start-up, melt temperatures were varied and experiments were repeated using different compression forces. The resultant part quality with respect to lateral shrinkage is shown in Fig. 9. With the same compression start-up and termination, the effect of temperature is not as straightforward as in conventional molding.
In both IM and ICM process, the influence of melt temperature on part dimension mainly comes from (1) the melt flow-in during filling and intermediate packing, and (2) cooling-induced shrinkage. The melt temperature has a third effect on ICM, namely compression-induced shrinkage (densification). The interaction between melt temperature, compression pressure, compression start-up and end-up in ICM makes the influence of melt temperature not as obvious as that observed in IM. Rigorous simulation is needed to predict the quality of final parts.
Greener (9) has pointed out that precision parts with thick sections are difficult to mold using injection molding. Insufficient packing may result in poor quality, even sink marks in surface or voids inside. To investigate the effect of thickness with compression added to post-filling, parts with three original cavity thickness (2.68, 4.70, and 7.10 mm) were injection-compression molded with various compression pressures. The quality shown as deviation from cavity diameter for several compression forces are shown in Fig. 10. With the same compression start-up and termination for each melt temperature, the optimal part thickness which yields the best quality may not be the thinnest. This is the result of two conflict factors: Thicker parts have more mass to densify, but they can be compression-densified at higher bulk temperature than thinner ones.
The effects of ICM process parameters, compression start-up, compression pressure, melt temperature, and part thickness on quality of injection-compression-molded disks have been investigated. The findings are summarized as follows:
(1) The improvement of quality with compression is significant, even when the compression is activated after the detachment point. Compression can preserve or even resume contact of the molding with the cavity surface to improve the conformability of the part surface to the cavity profile.
(2) Compression start-up time determines the transition of the densification mechanism from packing to compression. This transition significantly affects part mass and birefringence, because packing induces mass and orientation. Compression should start as soon as enough mass has filled and packed in; it should start before the peak cavity pressure.
(3) High compression force improves flatness and reduces lateral shrinkage. At the same time, part thickness is reduced. The control of final thickness becomes one of primary concern to ICM molders.
(4) Adding of compression to post-filling changes the effects of melt temperature and part thickness on the part quality. The effect of temperature is not as straightforward as in conventional molding. There is an optimal thickness which will give best quality for ICM process.
(5) Theoretical simulation of the ICM process is a necessity, considering the tight tolerance demanded and the complicated interaction between parameters introduced by compression. To predict the final thickness, a model describing thermodynamic states should include not only P and T, but also the nonequilibrium structural states of the polymeric glass.
The authors would like to express their thanks to the National Science Council of Taiwan, ROC for financial support, National Taiwan University for support in laboratory set-up, to the coworkers in the Grace Laboratory for Polymer Processing for stimulating discussion and experimental assistance, and also to Cheng-Shong Machinery Company for help in machine acquisition and modification.
Cp = Specific heat of the polymer melt at To (2.08. J/g- [degrees] C (26)).
D = Representative lateral dimension of the part (disk diameter 6.80 mm).
h = Part thickness of the compressed area (disk thickness [2.68 mm]).
k = Thermal conductivity of the polymer melt at To (0.151 J/mm-s- [degrees] C (26)).
P = Compression pressure (bar).
[t.sub.e] = The time interval from the start of filling to the end of compression (s).
to = The time interval from the start of filling to the start of compression (s).
To = The melt temperature (210 [degrees] C).
Tm = The mold temperature (25 [degrees] C).
Tg = The glass transition temperature of the polymer (100 [degrees] C).
[Alpha] = Thermal diffusivity of the polymer melt at To (70.4 [mm.sup.2]/s).
[Beta] = The compressibility of the polymer melt at To (8.54 x [10.sup.-5]/bar (26)).
[Rho] = Density of the polymer melt at To (1.03/g/[cm.sup.3] (26)).
[Delta]D = Deviation of the lateral dimension of the part from the cavity (mm).
[Delta]h = Deviation of the thickness of the part from original cavity thickness (ram).
[Delta]w = Flatness of the part (maximum height difference on the surface).
1. C. L. Tucker, in Injection and Compression Molding Fundamentals, Isayev, ed., Marcel Dekker Inc. (1987).
2. A. I. Isayev, ed., Injection and Compression Molding Fundamentals, Marcel Dekker Inc. (1987).
3. J. P. Wu and J. L. White, Polym. Eng. Sci., 31, 652 (1991).
4. G. Klepek, Kunststoffe, 77, 13 (1987).
5. S. Matsuda, A. Itoh, and T. Tamura, National Technical Report, 31, 5, 641 (1985).
6. M. Ke, Master's Thesis, National Taiwan University (1992).
7. S. Yang and M. Ke, SPE ANTEC Tech. Papers, 39, 2182 (1993).
8. C. Liu and L. T. Manzione, SPE ANTEC Tech. Papers, 39, 1085 (1993).
9. J. Greener, Polym. Eng. Sci., 26, 886 (1986).
10. J. L. White and W. Dietz, Polym. Eng. Sci., 19, 1081 (1979).
11. A. I. Isayev, Polym. Eng. Sci, 23, 271 (1983).
12. M. A. Kamal and V. Tan, Polym. Eng. Sci., 19, 558 (1979).
13. W. Dietz, J. L. White, and E. Clark, Polym. Eng. Sci., 18, 273 (1978).
14. K. Gissing and W. Knappe, Kunststoffe, 73, 241 (1983).
15. A. I. Isayev and T. Hariharan, Polym. Eng. Sci., 25, 1271 (1985).
16. R. S. Spencer, and G. D. Gilmore, Mod. Plastics, 27, 97 (1950).
17. A. I. Isayev, Polym. Eng. Sci., 23, 271 (1983).
18. S. Yang, and M. Ke, in preparation.
19. W. Friesenbichler, W. Knappe, and R. Rabitsch, "Injection Molding Without Packing," Proceedings of 3rd Ann. Meet. of Int. Polym. Proc. (1988).
20. D. G. F. Huilier, J. of Polym. Proc., 9, 238 (1990).
21. J. Greener, Polym. Eng. Sci., 26, 534 (1986).
22. A. J. Kovacs, J. M. Hutchinson, and J. J. Aklonis, The Structure of Non-crystalline Materials, Gaskell P. H., ed., Taylor and Francis (1977).
23. A. J. Kovacs, J. Polym. Sci, 30, 131 (1958).
24. J. Greener, J. M. O'Really, and K. C. Ng, in Material Research Society Symposium and Proceedings, R. J. Roe and J. M. O'Reilly, eds. (1990).
25. A. J., Kovacs, J. J. Aklonis, J. M. Hutchinson, and J. Ramos, Appl. Sci., 17, 1077 (1979).
26. M. Y. Hon, Master's Thesis, National Taiwan University (1993).
|Printer friendly Cite/link Email Feedback|
|Author:||Yang, S.Y.; Ke, M.Z.|
|Publication:||Polymer Engineering and Science|
|Date:||Aug 15, 1995|
|Previous Article:||Esterification in reactive extrusion.|
|Next Article:||Numerical simulation of fluid flow and heat transfer in twin-screw extruders for non-Newtonian materials.|