Advanced injection molding mold and molding process for improvement of weld line strengths and isotropy of glass fiber filled aromatic polyester LCP.

INTRODUCTION

Liquid crystalline polymers (LCPs) are exceptional plastics in that they maintain crystalline order of molecular chains of solid plastic also in melt state during injection molding process. The consequences are sharp melting point, low melting or solidification enthalpy, and low melt viscosity. Theoretically this also means low shrinkage, low warping, low coefficients of expansion, and high chemical and physical stability if the chemical structure of LCP is fully aromatic polyester, for example. An inherent problem anisotropy and weak weld lines of molded products was to be studied.

In "LCPHITEC" project at VTT between the years 2000 and 2002 there was a goal to develop a methodology for injection molding of thin wall parts of LCP composites and nanocomposites. Thin wall molding, high speed injection molding, new more isotropic LCP compounds, and better weld line strengths were the main issues in processing point of view in that project. Further research of processing LCP-parts from new LCP nanocomposites has been continuing in VTT at the project Nano LCP. This work is done in co-operation with Technical University of Chemnitz and Tampere University of Technology and it is mainly concentrating on injection molding simulations and certain injection molding difficulties of these new materials. After a certain length/thickness, ratio filling of the thin wall part is demanding substantially more force [1]. Increasing the flow rate is normally decreasing the viscosity and in that manner pressure is required for filling the part. According to Beaumont [2], above a certain flow rate level the pressure required for filling starts to increase again. Typically in thin wall injection molding the thickness of the part is clearly less than 1 mm.

In the first part of this study, an advanced injection molding tool was developed for accurate measurement of changes in strength properties at different geometrical positions of complex thin-walled molded parts. Using this optimized tool, the effect of flow rate was explored by systematic injection molding experiments at four industrial and research institute laboratories. We at VTT were using one of the fastest injection molding machines in the market. The Demag Ergotec machine can reach 1000 mm/s injection screw speed. With 22-mm cylinder this means 380 [cm.sup.3] flow rate. An Engel X-melt machine was another equipment for extra high speed injection molding. Pressure controlled X-melt is an injection molding technology developed by Engel company. This technology differs from conventional molding in that sense that the plastic melt is compressed before molding to a high pressure typically between 200 and 260 MPa. After compression phase valve gate is opened and the melt flows to the cavity by the force of decompression. The screw of the molding machine was not moved during filling. After filling, there is a packing phase like in conventional molding. Packing pressure in X-melt is quite low, typically 30-60 MPa.

[FIGURE 1 OMITTED]

One of the test series was made with a fast hydraulic molding machine capable of 320 mm/s with 25-mm diameter screw. These experiments are also made for finding out if there is a possibility to effect the weld line strength via flow rate or pressure settings. Weld line means the joining of the flow fronts together again after they have separated because of an obstacle in flow field. There is typically quite a big difference in LCP material properties in weld line area. This is due to fast solidification of LCP when two separate flow fronts get together in a cavity. For example, tensile strength tends to be remarkably [3] lower in weld line area than in other areas in the part. Orientation is usually not optimal in the weld line area. Beside this injection, molded LCP materials are noticed to have fluctuating multilayer structure [4, 5]. This trial work was done in several companies. Beside this flow rate study and Enqvist study [6], further information of numeral LCP-compounds that were made for finding out the effect of different kind of fillers to the flow and strength properties of LCP will be published later.

[FIGURE 2 OMITTED]

FABRICATION OF ANISOTROPY MEASUREMENT TOOL

A LCPHITEC test mould for determination of anisotropic and weld line tensile properties of LCP composites was designed and fabricated at VTT in LCPHITEC-project in 2000-2002. Figures 1-3 describe parts and test bars produced by this tool. Figure 1 shows the general geometrical structure of molded part with dimensions in millimeters. From the injection point, the velocity of LCP melt accelerates in small runners and it flows with a very high speed up to 50 m/s into the two separate sections of the mold. The first section at the upper part of the mould cavity produces solid LCP plate from which the flow direction oriented test bar (3) and the transverse-to-flow direction bar (4) are water-jet cut as shown in Fig. 2. The lower part of the mold cavity in Fig. 1 produces the knit line (2) bar behind a round hole and the butt weld line (1) bar behind a rectangular hole. The dimensions of the mini test bars are shown also in Fig. 2.

Mold is equipped with shut-off hot runner. Diameter of the cold runner is decreased from 1.4-0.5 mm from the beginning of the runner to the gate location (Fig. 2). Cross section of the runner is circular. There is no specific gate area. Draft angle in cavity is more than 1[degrees]. Radius of the bottom corner in cavity is 0.2-0.4 mm. Bottom of the cavity is arc machined. Fixed plate of the mold is straight except the runner area. Three pressure sensors are located in the fixed plate. Mould plates are 46 mm thick. Air removal is made via channels milled to the end of flow pattern, Fig. 3. These channels are 2-3 mm wide and the gap is 0.02 mm in the beginning of the channel. This gap area is working as "overflow gate" when pressure of the melt is high enough.

[FIGURE 3 OMITTED]

Tensile strength test was the one of ISO 527 standard with the exception of smaller dimensions of the test bar and the draw speed of the tests which was different. The thickness of the test bars was 0.5 mm which is smaller than the minimum thickness specified in ISO. Dimensions of the test bar were otherwise due to ISO 527 specifications. Speed for draw was 1 mm/min. Extensiometer was used for measuring the strain. Test bars were cut by water-jet cutting. Care was taken so that samples would not overstress during cutting and that the cutting event would be equal for every sample. Samples were conditioned at least 40 h before tests in temperature of 23[degrees]C and in humidity of 50%. Number of test specimens was five or more.

INSTRUMENTATION AND EXPERIMENTAL APPROACHES

Experiments for this study were made at four different laboratories. Vectra A130 was used in every trial. Processing parameters are not naturally exactly the same. Anyhow the values were set near to each other so comparison of these results is reasonable. The LCPHITEC mold described earlier was tested first with injection molding machine that is capable of extremely high filling rates. This machine was Demag Ergotech 100/420-120 El-EXIS S. The trial was made at Eimo company in Lahti in February 2002. Temperature of the hot runner in this test was 310[degrees]C and temperature of the mold was 90[degrees]C. Theoretical volume rates are used here. Reported filling rates are 30 and 380 [cm.sup.3]/s.

Injection molding trial with the LCPHITEC test mould was then made at Nokia research center in April 2002. ENGEL ES200H/80W/80L 90HL 3F injection molding machine was used. Injection fill rates 54, 69, 97, 117 [cm.sup.3] and maximum rate (130 [cm.sup.3]) were tested. Filling rates are calculated by the measured maximum speed of the screw in every speed setting. The injection molding machine had a screw diameter 25 mm. Temperature of the hot runner was 305[degrees]C except with maximum flow rate it was 320[degrees]C. Temperature of the mould was again 90[degrees]C.

The LCPHITEC mould was then exported to Austria in June 2002. X-melt trial was made in the factory of Engel in Schwertberg with Engel E-motion EM440-100 injection molding machine. The screw diameter of the machine was 30 mm. Three compression pressure values were tested. These values were 2200, 2400, and 2600 bar. Temperature of the hot runner in test was 325[degrees]C. Temperature of the mould was 90[degrees]C.

VTT hired the Demag Ergotech 100/420-120 El-EXIS S after experiments in Eimo. This molding described here was made in May 2002. Temperature of the hot runner in test was 310[degrees]C. Temperature of the mould was 100[degrees]C. Machine was set so that free overflow through air venting channels did occur. In that sense, parts having more flash could be approved by the quality assurance. This experiment was done out of scientific interest because it had been previously noticed that free overflow improves weld line strength remarkably. In production there would be material wastage and extra handwork because of this kind of overflow. Reported fill rate is 380 [cm.sup.3]/s. This is theoretical value. It could not be achieved.

COMPUTATIONAL COMPARISON OF DIFFERENT MOLDING PROCESSES

Some injection molding simulations were made to understand the difference of X-melt and conventional molding. X-melt was simulated with Moldflow. This was done by modeling hot runner and barrel of the injection molding machine with beam elements (Fig. 4). Process parameters were set so that there was an extreme high flow rate at the beginning of molding for producing the compression of the melt and after certain point there was smallest possible flow rate through the injection point. Part was then filled by the effect of decompression of the melt in the beam elements. As in X-melt the decompression of the melt is also in this simulation the driving force for filling. The main difference here is compared with the experimental reality that in X-melt the flow is starting after compression when the shut off valve is opened but here the "valve" was open during compression. About 25-35% of the volume of the cavity was filled during this compression phase in simulations. Process parameters were set so that in the end of filling there was a pressure about 40 MPa in the injection point. Packing pressure was 42 MPa.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

There were 7915 elements in the fusion model of X-melt simulation. The heat transfer coefficient has been changed in these comparison simulations from value 25,000 W/[m.sup.2]K (as used in Moldflow packages) to 5000 W/[m.sup.2]K (other sources like Blum [7] and Brunotte et al. [8] report values of 3000 or 500, respectively.). There is a discussion going on about the proper value of this parameter. The greater value means that the skin layer of the plastics freezes immediately after it has contacted the surface of the cavity. Thin wall part fills much easier when this value is smaller. This is due to higher temperature of plastics near the surface of the mould. Calculation took about 1 h with PC with 3 GHz processor.

Processing parameters of X-melt in the simulation were set as equal as possible to test run made in Engel's factory. Melt temperature was 325[degrees]C, mold temperature was 95[degrees]C, flow rate was 9899 [cm.sup.3]/s till 25% of the cavity is filled, after that flow rate was set to 0.001% of the maximum ram speed. Change from speed control to pressure control when 99% of the volume of the part was filled, packing pressure was set so that it was 0.3 s same as the pressure in switchover point and after that it was 1 s 80% of the pressure in switchover point. Pressure in switchover point in this simulation was 41 MPa. Machine hydraulic response time was set to 1E-5 s. Calculated filling time was 0.029 s.

In conventional molding the simulation melt temperature was 320[degrees]C, mold temperature 95[degrees]C and flow rate 69 [cm.sup.3]/s, change from speed control to pressure control was done when 95% of the volume of the part is filled, packing time was 1 s and the packing pressure was set to decrease from 95% of pressure in switchover point till 80% of the pressure in switchover point in that time, cooling time was 1 s. Calculated filling time was 0.043 s. Hot runner was also modeled totally in the calculations.

The temperature of the melt is increasing quite a lot in X-melt and it remains high till the end of filling. Maximum temperature in knit line area was 358[degrees]C with X-melt and 342[degrees]C with conventional molding (Fig. 5). Maximum temperature was 336[degrees]C with X-melt and 329 [degrees]C with conventional molding in the butt weld line area in the cavity. Maximum bulk temperatures in filling were with X-melt 366[degrees]C and with conventional molding 356[degrees]C. On the other hand, the pressure in the uttermost edge of the cavity was very low in X-melt. There was a pressure of maximum 65 MPa in the butt weld line area just after unification of the melt fronts in conventional molding. In X-melt, the equivalent pressure was only few megapascals (Fig. 6). In knit line area maximum pressure after unification of the flow fronts was about 88 MPa in X-melt and 55 MPa in conventional molding (Fig. 6). In X-melt pressure is quickly decreasing to the level of 15 MPa. In conventional molding the pressure is decreasing to 20 MPa and then it starts to increase again up to 65 MPa. There is also big difference in time when the melt is reaching the knit line area. In X-melt results start from 0.003 s. In conventional molding melt is reaching knit line area after 0.2 s.

It can be concluded that the high melt temperature in weld line area is very important for achieving good weld line strength properties. On the other hand, it can be said that the pressure does not need to be or may not be very high on the weld line area straight after melt fronts have unified. Speed of the melt just before reaching the weld line area was also simulated. In X-melt, the speed of the flow front is high in knit weld line area, 880 cm/s, and low in butt weld line area, 40 cm/s. In conventional molding the difference is not so big between different weld line locations, speed was 340 cm/s in weld line area and 180 cm/s in butt weld line area. Speed of the flow front does not seem to be very important factor for weld line strength.

[FIGURE 6 OMITTED]

RESULTS

Clear trends in tensile strength of Vectra A130 LCP can be seen when changing the flow rate in conventional injection molding or compression load in X-melt (Fig. 7). Strength in direction of flow is decreasing when filling speed or compression load is increased. By Ticona's product information [9] the (machine direction) tensile strength of Vectra A130 is 190 MPa. Our measurements gave somewhat smaller strength values. In X-melt rise of the compression load is improving all the other strength values. In conventional injection molding faster injection flow rate increases butt weld line strength. When increasing the flow rate transversal to flow strength is decreasing near to the value 40 MPa and butt weld line strength values are at the same time increasing over it. Weld line strength values should be near the transversal strength because as described by Nguyen-Chung [10] melt fronts are flowing transversal to main flow direction in end of weld line formation. The butt weld line strength affects also the fact that this weld line is located at the outermost edge of the part. With injection flow rates 117 and 130 [cm.sup.3]/s there was overflow of the melt through the air venting channels. There was smaller overflow with flow rate 117 compared with 130 [cm.sup.3]/s. This was mainly due to higher temperature used with the flow rate 130 [cm.sup.3]/s. This is affecting especially the butt weld line strength. Sizable overflow is also increasing the transversal to flow strength. Increase in fill rate is decreasing all the other but butt weld line strength values. Strength in direction of flow is decreasing with higher filling rates.

Figure 8 shows the best value achieved with different kind of injection strategies. Compared techniques are X-melt, filling with moderate speed, filling with high speed, and filling with high speed with overflow. Overflow means here flowing of the melt trough air removal channels. X-melt is giving the best strength properties is this test series. With X-melt one must know that the plastic melt is staying in hot barrel much longer than with conventional injection molding. LCP is not sensitive for this kind of thermal stress but for that reason X-melt is not suitable manufacturing technique for all plastics materials. Quick filling with overflow gives almost as good results in regards to weld lines than X-melt. Knit weld line strength is 56% and butt weld line strength is 38% higher with overflow than without it in conventional injection molding. Here, we must anyhow remember the limitations of overflow phenomenon. Moderate filling speed is giving slightly better properties than fast filling without overflow. Anisotropy is the highest with high flow rate. Overflow is partially releasing the situation. Degree of anisotropy is quite low with overflow. Overflow is clearly increasing transversal to flow direction strength.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

CONCLUSIONS

The developed LCPHITEC mold proved to be extremely useful in determination of detailed anisotropic strength properties of LCP composites. Increase of flow rate from moderate to high in conventional molding decreased the tensile strength in direction of flow and transversal to flow. Also knit weld line strength is decreasing slightly. Butt weld line is increasing when increasing flow rate in conventional molding. In X-melt strength in direction of flow is decreasing when compression load is increased. Rise of the compression load is improving all the other strength values. As a result isotropy of LCP is improved. X-melt technique seems to be good injection molding technique for manufacturing small thin-walled parts from materials that can withstand longer residence time in cylinder of injection molding machine. It gave the best or near to the best strength values in every sector with the tested LCP grade.

Strength values with high speed injection and with overflow are almost as good as with X-melt. Here we must remember the limitations and drawbacks of overflow. There is a material wastage and extra manual work when this kind of overflow is used. When making parts from LCP with traditional injection molding a rise of about 40% in butt weld line strength could be achieved with overflow. There was also a remarkable rise in transversal to flow strength by which strength isotropy is reduced. Simulation, on the other hand, indicated that the melt temperature is theoretically more important factor for achieving better weld line strength than pressure.

ACKNOWLEDGMENTS

The authors would like to thank Harri Lasarov, Harri Juhola, Mikko Jalonen, Anders Nyback, Peter Pokorny and Andreas Pottler for test runs in companies and Maarit Enqvist for discussions in material characterization aspects.

REFERENCES

1. T. Palmer, Plastics for Portable and Wireless Electronics, Society of Plastics Engineers, Phoenix (1996).

2. J.P. Beaumont, Runner and Gating Design Handbook, Tools for Successful Injection Molding, Hanser, Munich (2004).

3. K. Engberg, A. Knutsson, P.-E. Wermer, and U.W. Gedde, Polym. Eng. Sci., 30, 24 (1990).

4. E. Suokas, J. Sarlin, and P. Tormala, Mol. Cryst. Liq. Ciyst., 153, 515 (1987).

5. E. Suokas, Processing, Microstructure and Properties of Thermotropic Liquid Crystalline Polymers and their Carbon Fibre Composites, Tampere University of Technology, Tampere (1999).

6. M. Enqvist, Characterization and Optimization of Selected liquid Crystalline Polymer (LCP)-Based Micro- and Nanocomposites and Liquid Crystalline Thermoplastic-Thermoset Polymer Adhesive, Tampere University of Technology, Tampere (2003).

7. R. Blum, Verbesserte Temperaturkontrolle beim Kunststoffspritzgiessen, RWTH Aachen University, Aachen (1996).

8. R. Brunotte, G. Mennig, and J. Nagel, Zeitschrift Kunststofftechnik--J. Plast Technol, 4, 1 (2006).

9. Vectra liquid crystal polymer (LCP). VC-7, Product Brochure, Ticona GmbH, Frankfurt am Main (2001).

10. T. Nguyen-Chung, Rheo Acta, 43, 240 (2004).

Matti Koponen, (1) Jouni Enqvist, (1) Tham Nguyen-Chung, (2) Gunter Mennig (2)

(1) VTT Advanced Materials, Sinitaival 6, 33101 Tampere, Finland

(2) Department of Mechanical Engineering, Technical University of Chemnitz, D-09107 Chemnitz, Germany

Correspondence to: Matti Koponen; e-mail: [e-mail]

Contract grant sponsor: Tekes.