EJECTION FORCES AND STATIC FRICTION COEFFICIENTS FOR RAPID TOOLED INJECTION MOLD INSERTS DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Mary E. Kinsella, M.S. ***** The Ohio State University 2004 Dissertation Committee: Approved by Professor Blaine Lilly, Adviser Professor Jose Castro Professor Jerald Brevick ______________________________ Adviser Industrial and Systems Engineering Graduate Program ABSTRACT While manufacturing is typically considered a high-volume industry, the necessity for small quantities of products and components exists for aerospace customers and those producers wishing to mass customize their products. Because of the high cost of tooling, injection molding processes are seldom used to produce only small quantities of parts. This, however, can be remedied if cost effective tooling methods are implemented. Rapid prototyping processes show great potential for such tooling applications because they generally require shorter lead times, produce less waste, and, in some cases, use less expensive materials. The research presented herein studies the feasibility of using injection mold inserts produced with additive methods by investigating ejection and friction. Through experimentation, the application of P-20 steel, laser sintered LaserForm ST-100, and stereolithography SL 5170 tools to produce limited quantities of a thin-walled cylindrical part are explored. A substantial amount of data and statistical analysis are provided that reveal conditions during the actual injection molding process, and comparisons are made among the three insert types. Experimental ejection forces from each tool type are compared with model-based calculations, and apparent coefficients of static friction are calculated and compared to standard test results. Based on the data and analyses, the benefits and limitations of using rapid tooled injection mold inserts are presented. ii For Michael, Amelia, and Nathaniel iii ACKNOWLEDGMENTS “Gratitude is not only the greatest of virtues, but the parent of all the others.” --Cicero At the Materials and Manufacturing Directorate in the Air Force Research Laboratory, I am grateful to Charlie Browning, Bill Russell, and Chuck Wagner for providing the time and funding to complete this research; to John Jones for assembling and programming the data acquisition system; and to Neal Ontko, Nick Jacobs, and Ben Gardner for performing friction tests. At the NASA Marshall Space Flight Center, thanks to Ken cooper, who provided the laser sintered and stereolithography inserts for the experimental work. At The Ohio State University, many thanks to the following people: Brian Carpenter for finite element modeling of the inserts for thermal and deformation simulations, and for helping with experiments; Bob Miller for his machining and injection molding expertise; Mary Hartzler for providing machining services and design consultation; Barney Barnhart for providing equipment and expertise for the thermoplastic tensile tests; Mauricio Cabrera-Rios for design of experiments and statistical analysis consultation; Narayan Bhagavatula for helping with tensile tests and injection molding simulations; and especially my adviser, Dr. Blaine Lilly, and Drs. Jose Castro and Jerry Brevick for serving on my dissertation committee. Finally, I extend my gratitude to my parents, Robert and Carolyn Corbin, who deserve more than I can ever express. iv VITA June 5, 1961 Born – Longview, WA, USA 1983 1983 - 1986 1987 - Present Bachelor of Science, Applied Science Miami University, Oxford, OH, USA Production Supervisor, NCR Microelectronics Fort Collins, CO, USA Project Engineer, Materials and Manufacturing Directorate US Air Force Research Laboratory Wright-Patterson AFB, OH, USA 1991 Master of Science, Materials Engineering University of Dayton, Dayton, OH, USA PUBLICATIONS Kinsella, M. E., Heberling, M. E. 1997, “Applying Commercial Processes to Defense Acquisition,” National Contract Management Journal, vol. 28, issue 1, p. 11. Kinsella, M. E., Lilly, B. L., Bhagavatula, N., Cooper, K. G. 2002, “Application of Solid Freeform Fabrication Processes for Injection Molding Low Production Quantities: Process Parameters and Ejection Force Requirements for SLS Inserts,” Proceedings of the 13th Annual Solid Freeform Fabrication Symposium, Austin, TX, pp. 92-100. FIELDS OF STUDY Major Field: Industrial and Systems Engineering v TABLE OF CONTENTS Abstract ........................................................................................................... ii Acknowledgments.......................................................................................... iv Vita .................................................................................................................. v Table of Contents ........................................................................................... vi List of Figures ................................................................................................ xi List of Tables................................................................................................. xv Chapter 1 Introduction .................................................................................... 1 1.1 Background ........................................................................................................... 1 1.2 Problem Statement ................................................................................................ 6 1.3 Research Objective.............................................................................................. 10 1.4 Research Description........................................................................................... 11 1.5 Organization........................................................................................................ 12 Chapter 2 Literature Search .......................................................................... 14 2.1 Ejection Force ..................................................................................................... 14 2.1.1. Ejection Force Models............................................................................... 14 2.1.2 Shrinkage ................................................................................................... 22 vi 2.1.3 Friction and Adhesion ................................................................................ 25 2.2 Rapid Tooling ..................................................................................................... 34 2.2.1 Background ................................................................................................ 34 2.2.2 Stereolithography and Laser Sintering for Injection Molding Tools........... 44 2.2.3 Summary .................................................................................................... 48 Chapter 3 Theory........................................................................................... 50 3.1 Thermoplastic Materials...................................................................................... 50 3.1.1 High Impact Polystyrene ............................................................................ 51 3.1.2 High Density Polyethylene ......................................................................... 54 3.2 The Adhesion Component of Friction ................................................................. 55 3.3 Ejection Force Model Derivation ........................................................................ 60 3.3.1 Model derivation ........................................................................................ 60 3.3.2 Additional Consideration for Strain............................................................ 65 Chapter 4 Experimentation ........................................................................... 66 4.1 Friction Testing ................................................................................................... 66 4.1.1 Friction Test Apparatus .............................................................................. 68 4.1.2 Test Matrix and Procedure ......................................................................... 70 4.2 Measurement of Elastic Modulus ........................................................................ 72 4.3 Injection Molding................................................................................................ 75 4.3.1 Mold Design and Materials ........................................................................ 75 4.3.2 The Injection Molding Process................................................................... 79 4.3.3 Design of Experiments ............................................................................... 84 vii 4.3.4 Experimental Procedure ............................................................................. 87 4.4 Set-up and Data Acquisition................................................................................ 88 4.4.1 Temperature Measurement and Thermal Model ......................................... 89 4.4.2 Ejection Force Measurement ...................................................................... 98 4.4.3 Diameter and Thickness Measurement ..................................................... 100 4.4.4 Calculation of Static Friction Coefficient ................................................. 101 Chapter 5 Results and Analysis .................................................................. 102 5.1 Injection Molding Experiments ......................................................................... 103 5.1.1 Experimental Results and Discussion....................................................... 103 5.1.2 HDPE Experimental Ejection Force Results............................................. 105 5.1.3 HIPS Experimental Ejection Force Results .............................................. 107 5.1.4 Experimental Ejection Force Results from the P-20 and ST-100 Inserts .. 109 5.1.5 Experimental Ejection Force Results from the SL 5170 and SL 5170/P-20 Inserts................................................................................................................ 109 5.2 Statistical Analysis ............................................................................................ 112 5.2.1 DOE Results............................................................................................. 112 5.2.2 Main Effects and Interactions................................................................... 112 5.3 Standard Friction Testing Results...................................................................... 120 5.3.1 HDPE Standard Friction Results .............................................................. 120 5.3.2 HIPS Standard Friction Results ................................................................ 124 5.4 Reliability of the Data ....................................................................................... 128 5.5 Calculation of Ejection Force Using the Model................................................. 130 viii 5.5.1 Calculated Ejection Force for HDPE........................................................ 133 5.5.2 Calculated Ejection Force for HIPS.......................................................... 134 5.5.3 Possible Sources of Error ......................................................................... 135 5.6 Calculation of Apparent Friction Coefficients using the Menges Model ........... 138 5.6.1 HDPE Apparent Coefficient of Friction Results....................................... 140 5.6.2 HIPS Apparent Coefficient of Friction Results......................................... 141 5.6.3 Apparent Friction Coefficient Results from the P-20 and ST-100 Inserts. 143 5.6.4 Apparent Friction Coefficient Results from the SL 5170 and SL 5170/P-20 Inserts................................................................................................................ 144 5.6.5 Comparing Calculated Friction Results to Standard Friction Test Results 146 5.7 Other Observations of Rapid Tooled Inserts...................................................... 150 Chapter 6 Conclusions ................................................................................ 156 6.1 Molding HDPE and HIPS with ST-100 and SL 5170 Inserts ............................ 156 6.1.1 Benefits and Limitations of Using Rapid Tooled Injection Mold Inserts .. 156 6.1.2 Friction and Ejection Force Considerations.............................................. 158 6.2 Using a Model to Determine Ejection Force and the Coefficient of Friction..... 160 6.3 Implications and Future Work........................................................................... 162 6.4 Summary........................................................................................................... 165 LIST OF REFERENCES ............................................................................ 167 Appendix A Data Tables............................................................................. 174 A.1 Tensile Test Data Table.............................................................................. 176 A.2 Modulus Look-up Table ............................................................................. 177 ix A.3 Thermal Analysis Convergence Table........................................................ 178 A.4 Sample Experimental Data Set, All Runs ................................................... 182 A.5 Sample Experimental Part Dimensions (2 Runs Shown) ............................ 184 A.6 Experimental Data and Calculated Coefficient of Friction (Menges), Run Average............................................................................................................. 185 A.7 Analysis of Variance Tables by Set ............................................................ 187 Appendix B Mold and Canister Drawings.................................................. 190 B.1 Part Drawing............................................................................................... 191 B.2 Mold Insert Drawings ................................................................................. 192 B.3 Mold Assembly Drawings .......................................................................... 205 x LIST OF FIGURES Figure 1.1: Importance characteristics for various tool types……………………….. . 4 Figure 2.1: A schematic of the laser engineered net shaping (LENS™) process…... 36 Figure 2.2: A schematic of the selective laser sintering process……………………. 38 Figure 2.3: A schematic of the stereolithography process…………………………...42 Figure 3.1: Polystyrene monomer……………………………………………………52 Figure 3.2: High impact polystyrene………...…………………………… ............... 53 Figure 3.3: Polyethylene monomer. .......................................................................... 55 Figure 3.4: High density polyethylene linear molecule. ............................................ 55 Figure 3.5: Thin-walled cylindrical pressure vessel. ................................................. 61 Figure 3.6: Section of the part and the core with associated stresses. ........................ 62 Figure 4.1: Schematic of friction apparatus............................................................... 68 Figure 4.2: Friction test apparatus: sled on plate specimen inside furnace and tester… .................................................................................................................................. 69 Figure 4.3: Tensile testing apparatus with tube furnace............................................ 73 Figure 4.4: HIPS specimens after tensile tests.......................................................... 74 Figure 4.5: Elastic modulus at various temperatures for HDPE and HIPS. .............. 74 Figure 4.6: Sprue side of the MUD base mounted in the injection molding machine with SL 5170 cavity insert. ....................................................................................... 76 xi Figure 4.7: Core and cavity inserts, before final machining, made of SL 5170, P-20 steel, and LaserForm ST-100. ................................................................................... 76 Figure 4.8: Canister part with vent holes and no taper.............................................. 78 Figure 4.9: Sumitomo SH50M injection molding machine. ..................................... 79 Figure 4.10: Signal conditioner and computer with front panel for data acquisition, and core side of mold with thermocouple and load cell sensor wires. ....................... 89 Figure 4.11: Thermocouple placement within core insert.......................................... 90 Figure 4.12: Representative thermal traces of the injection molding cycle............... 91 Figure 4.13: Graphs of the thermal analysis results for each material combination... 94 Figure 4.14: Representative ejection force traces ...................................................... 99 Figure 4.15: Digital pictures of HDPE canisters for measuring inside and outside diameter. ................................................................................................................. 101 Figure 5.1: Experimental ejection force results for HDPE, all runs........................ 106 Figure 5.2: Experimental ejection force values for HIPS, all runs.......................... 108 Figure 5.3: Experimental ejection force results from the P-20 and ST-100 inserts. 110 Figure 5.4: Experimental ejection force results from the SL 5170 insert and the combination SL 5170/P-20 insert. ........................................................................... 111 Figure 5.5: Main effects and interactions for HDPE with the P-20 insert............... 114 Figure 5.6: Main effects and interactions for HIPS with the P-20 insert................. 115 Figure 5.7: Main effects and interactions for HDPE with the ST-100 insert. ......... 116 Figure 5.8: Main effects and interactions for HIPS with the ST-100 insert. ........... 117 Figure 5.9: Main effects and interactions for HDPE with the SL 5170/P-20 insert.118 xii Figure 5.10: Main effects and interactions for HIPS with the SL 5170/P-20 insert.119 Figure 5.11: Standard friction test results for HDPE; means and ranges shown in the table. ....................................................................................................................... 122 Figure 5.12: Standard friction test results for HIPS; means and ranges shown in the table. ....................................................................................................................... 123 Figure 5.13: Sample plot of load vs. time for HIPS on SL 5170 from elevated temperature tests. .................................................................................................... 126 Figure 5.14: Sample plot of load vs. time for HIPS on P-20 from elevated temperature tests. .................................................................................................... 127 Figure 5.15: Calculated values for ejection force for HPDE compared with experimental values, averaged across all runs. ........................................................ 134 Figure 5.16: Calculated values for ejection force for HIPS parts from the P-20 and ST-100 cores compared with experimental values, averaged across all runs........... 137 Figure 5.17: Calculated values for ejection force for HIPS parts from the SL 5170 core compared with experimental values, averaged across all runs. ........................ 137 Figure 5.18: Calculated values of the apparent coefficient of static friction for HDPE, all runs. ................................................................................................................... 140 Figure 5.19: Apparent coefficients of friction calculated from experimental results for HIPS, P-20 and ST-100 results only, and results from all runs. ......................... 142 Figure 5.20: Apparent coefficient of static friction for parts from the P-20 insert. . 143 Figure 5.21: Apparent coefficient of static friction for parts from the ST-100 insert. ................................................................................................................................ 144 xiii Figure 5.22: Apparent coefficient of static friction for parts from the SL 5170 insert. ................................................................................................................................ 145 Figure 5.23: Apparent coefficient of static friction for parts from the SL 5170 core with the P-20 cavity. ............................................................................................... 146 Figure 5.24: Average apparent coefficient of static friction for HDPE compared to standard test results. ................................................................................................ 148 Figure 5.25: Average apparent coefficient of static friction for HIPS compared to standard test results. ................................................................................................ 149 Figure 5.26: Defects in the SL 5170 core. .............................................................. 152 Figure 5.27: Simulation results of HDPE injection into SL 5170 insert, no packing. ................................................................................................................................ 153 Figure 5.28: Simulation results of HIPS injection into SL 5170 insert, no packing.154 Figure 5.29: Simulation results of HIPS injection into SL 5170 insert, with packing. ................................................................................................................................ 155 Figure A.1: Sample plots from tensile test data....................................................... 175 xiv LIST OF TABLES Table 1.1: Properties of tooling materials.................................................................... 9 Table 2.1: Vanguard System Specifications from 3D Systems.................................. 40 Table 2.2: LaserForm ST-100 (Sintered and Infiltrated) Material Properties from 3D Systems. .................................................................................................................... 40 Table 2.3: SLA 250 System Specifications from 3D Systems. .................................. 43 Table 2.4: Vantico SL5170 Typical Properties (90-minute UV post cure). ............... 43 Table 4.1: Friction data for polymers on steel. .......................................................... 67 Table 4.2: Friction test matrix. ................................................................................. 71 Table 4.3: Injection molding machine specifications................................................ 80 Table 4.4: Injection Molding Parameters .................................................................. 82 Table 4.5: Typical data for Lutene-H ME9180. ........................................................ 83 Table 4.6: Typical data for BASF PS 495F. .............................................................. 83 Table 4.7: Experimental design for six sets, including process parameters. ............. 86 Table 4.8: Resulting convergence times from the thermal simulation. ..................... 93 Table 4.9: Input conditions for the thermal analysis.................................................. 93 Table 5.1: Experimental ejection force results for HDPE and HIPS according to packing time, cooling time, and packing pressure parameters................................. 104 xv Table 5.2: Results from the designed experiment indicating which factors had a significant effect on ejection force. ......................................................................... 113 Table 5.3: Surface roughnesses of all plates (friction tests) and cores (injection molding experiments).............................................................................................. 129 Table 5.4: Calculated values of ejection force for HDPE from the Menges equation and experimental data. ............................................................................................ 131 Table 5.5: Calculated values of ejection force for HIPS from the Menges equation and experimental data. ............................................................................................ 132 Table 5.6: Calculated apparent coefficient of friction results according to packing time, cooling time, and packing pressure parameters. ............................................. 139 xvi CHAPTER 1 INTRODUCTION 1.1 Background Manufacturing is predominantly a high volume industry and is constantly striving toward greater efficiencies at lower cost. A growing sector of manufacturers, both in the aerospace and consumer markets, however, is targeting small quantity production to meet customer needs. The demand for low volume production is strong in the aerospace industry, where customer organizations such as the military services and NASA need relatively small numbers of end products to accomplish their missions. Product variety is generally higher, and lot sizes smaller, in the defense industry compared to other industries. The U.S. Defense Department and other government organizations have been focused on finding ways to build low volume products more cost effectively (e.g., Kinsella 2000). These organizations and their contractors must implement new methods of producing extremely robust equipment in reduced time at reduced cost (Kaminsky 1996). 1 Small production quantities in the consumer market have historically applied to prototypes and market testing. In the past decade, however, market forces have altered the way industry looks at low volume requirements. Increasing product variety and shorter product lifetimes have led to mass customization, in which products are designed and made to order for individual customers, but produced by methods that still allow for economies of scale. Mass customization is already evident in a number of industries, including pagers, fiber optics, and blue jeans (Victor & Boynton 1998). The concept continues to spread as customers become more particular and manufacturers become more flexible. Thermoplastic injection molding is inherently well suited to high volume production requirements. A quality mold, running with material and process variables under tightly controlled conditions, is capable of producing very large quantities of parts with little or no manual intervention. When coupled with automated material feeding and robotic part removal systems, injection molding operations can be extremely cost effective over large production runs, turning out millions of components per year at a cost of a few pennies per part. At these production scales, the cost of the tooling essentially disappears, and the part cost results almost entirely from material, handling, and overhead. Minimum economic production quantities for injection molded parts are typically large due to tooling costs, which are incurred at the beginning of the product life. Molds are expensive, regardless of part size, and typically require production volumes of tens of thousands of parts in order to amortize their costs. For this reason, injection molding is 2 generally feasible only when the total production run is large enough to recoup the cost of the tool. Many design situations exist in which the complexity and versatility of injection molded thermoplastics would be an ideal solution were it not for the high initial cost of the tooling. For many low volume applications, the ability to use the best engineering solution is inhibited by the inability to cost effectively produce the necessary tooling. Injection molds expressly designed for low volumes have successfully been fabricated for prototyping use and as bridge tools. Small numbers of prototypes are typically built to test out a design for fit and function, and to allow changes to be made before tooling designs are finalized. Since so many consumer products use injection moldings, it is often necessary to build prototype tools to produce the design prototypes, especially in cases where the prototype must be fully functional to answer questions of strength, rigidity, etc. Bridge molds, on the other hand, are built and put into production very quickly. That is, they produce a small volume of parts prior to the completion of the final tooling, thus “bridging” the gap between prototype tools and final production. While prototype tooling can reduce time to market by accelerating the product development cycle, bridge tooling is designed to get a new product to market quickly while the high volume production tooling is still under construction. In both cases, reduced time to market is the prime consideration, and the cost of the tooling will eventually be folded into the total tooling cost and amortized over the lifetime of the product. As opposed to prototype and bridge tooling, molds intended specifically for low volume production have many fewer products over which to amortize costs. In low 3 volume production environments, tooling must be low cost, unless the product is very expensive. Time to market is a secondary consideration in this scenario. For these reasons, the strategies for determining tooling methods will differ between prototype or bridge tooling and low volume production tooling. To distinguish low volume production tools from prototype, bridge, and high volume tools, four characteristics typically dominate: cost, durability, cycle time, and part quality. Part quality means, for example, surface roughness, residual stress, and dimensional accuracy. The importance of these characteristics varies, depending on the tool type, as shown in Figure 1.1. For example, cycle time is very important for bridge and high volume production tools, but not for prototype or low volume production tools. Wear resistance is more important for high volume production tools than for any of the others. The most important characteristics for low volume production tools are cost and part quality. IMPORTANCE CHARACTERISTIC Cost Durability Cycle Time Part Quality High Volume Production TOOL TYPE Prototype Bridge Low Volume Production High Importance Low Importance Figure 1.1: Importance characteristics for various tool types. 4 There are a few ideas that have been implemented for reducing tooling costs for low volume production. For example, a less expensive tooling material can be used that is easier to machine, such as aluminum instead of steel. Universal mold bases with interchangeable inserts are another possibility, though this can be problematic if the piece parts differ widely in design. “Family molds,” in which several components of an assembly are molded together in the same tool, have been used with limited success due to constraints imposed by filling and cooling. Any or all of these approaches can be implemented to minimize tooling costs. The application of rapid prototyping processes for the purpose of making tools, known as rapid tooling, has been the object of much interest for prototype and low volume production. Rapid tooling encompasses many processes based on the rapid prototyping concepts of additive, layer-by-layer manufacturing. While these processes are still finding their way into the injection mold market, they hold significant potential for tools intended to build small quantities of parts. Depending on the process used, rapid tooled molds are made from various materials, which typically have much lower strength and thermal conductivity than the tool steel used in conventionally machined molds. For these reasons, it is generally believed that rapid tooled molds are inadequate for quality production injection molding. If, however, the mold is required only to make a small quantity of products, and if molding conditions are allowed to vary from those used with machined steel molds, rapid tooling may be an economical alternative. These variations may occur at the expense of 5 cycle time, but cycle time is generally not considered to be as critical for low volume production. 1.2 Problem Statement Manufacturers who currently build products in small quantities, such as aerospace systems, can benefit from injection molding tools that will cost effectively produce low volumes of production parts. A growing need for such tools is evidenced by the increasing applications for mass customization. Furthermore, if injection molds for low volume production become economically feasible, then manufacturers will most likely discover their overwhelming potential. Aerospace applications that require small quantities of molded parts, especially for the military, include composites and electronics packaging. In the composites area, injection molding and related processes can be used to mold filled thermoplastics for structural components or for resin transfer, such as for aircraft skins. Future aerospace applications will also include micro molding for microelecromechanical systems. Mass customization refers to the mass production of customized products (Anderson & Pine 1997). The goal of mass customization is to develop, produce, market, and deliver affordable goods and services with enough variety and customization that nearly everyone finds exactly what they want (Pine 1993, p. 44). A modular approach to injection molding using rapid tooled inserts facilitates mass customization at the fabrication level, allowing smaller quantities to be customized economically. 6 Thermoplastic injection molds, in general, must perform several functions, including distribution of the melt, formation of the melt into its final shape, cooling of the melt, and ejection of the part. To meet these requirements, high volume molds are traditionally machined of steel, are very strong, and have good thermal properties. Rapid tooling processes, on the other hand, may be very well suited to building injection molds for small quantity production. These processes have certain advantages for tooling applications. For example, because rapid tooling processes can generate complex geometries as easily as simple ones, they can build mold shapes and cooling lines that are impossible to machine. Also, the capability for local composition control further enhances the appeal of rapid tooled injection molds. The material properties of rapid tools, however, vary from conventional molds, i.e., strength and thermal conductivity can be much lower than for machined steel (Table 1.1). But for small quantity production, the robust material properties exhibited by conventional machined steel molds may not be necessary. With less rigorous molding parameters, such as injection pressures and temperatures, rapid tools might be used successfully for injection molding. The properties of rapid tools may be adequate to meet many small quantity injection molded part requirements, such as those for the aerospace and mass customization industries. This research studied aspects of rapid injection mold tooling in an attempt to find out if this is true. The issue addressed in this research was whether or not rapid tooled injection mold inserts are suitable for small quantity injection molding. Many aspects must be researched in order to confirm any suitability, too many to address in a single project. 7 Therefore, this work has focused on aspects related to ejection force. Rapid tooling materials must be able to withstand the forces inherent in the injection molding process, including forces resulting from ejection of the molded part. Ejection force requirements and the effects of process parameters on ejection force were investigated in this work. Also included was the determination and analysis of friction coefficients from standard test results, injection molding experiments, and an ejection force model. With respect to these areas, this research provides a comparison of rapid tooled inserts to conventional steel inserts, and further provides an assessment of the benefits and limitations of rapid tooled inserts for injection molding small quantities of parts. 8 Process Mold Material Density Tensile Strength 3 kg/m MPa Hardness Conductivity o W/m C Baseline Machining o P-20 Mold Steel [1] 7870 1080 30-35 HRC 47.6 @ 204 C H-13 Tool Steel [1] 7800 1550 52 HRC (R) Moldmax XL [2] (copper-nickel-tin) 8900 760 28-32 HRC 25.1 @ 199oC 63-70 3D Printing - Prometal Bronze/infiltrant 8100 406 60 HRB 7.35 Laser Sintering - 3D Systems Copper polyamide 3450 33.6 75 ShoreD 1.28 @ 40oC S Steel w/bronze 7700 510 79 HRB as machined Rapid Tooling Materials [2] Laser Generating - LENS S Steel 316 Plastic Casting -CIBA Ceramic-filled Epoxy Stereolithography - 3D Systems SL 5170 cured resin 8000 [3] 1220 [1] ed. Rubin 1990 [2] From company literature [3] From www.matweb.com, various ss 316 properties Table 1.1: Properties of tooling materials. 9 800 80 HRB [3] 64 (UFS) 91 ShoreD 59 85 ShoreD 0.92 @ 150oC 49 @ 100oC 56 @ 200oC 15 [3] 0.200 1.3 Research Objective The purpose of this research was to determine the feasibility of using rapid tooled inserts for injection molding small quantities of products. The objective was to quantitatively determine the benefits and limitations of laser sintered and stereolithography tools by: 1) comparing ejection force requirements among materials, 2) learning which process parameters affect them, and 3) determining the friction coefficients between the injection mold insert core and the thermoplastic part. The data generated help to answer the following questions for two thermoplastic molding materials (one amorphous and one crystalline) and two types of rapid tooled mold inserts: • How do ejection forces compare among conventional and rapid tooled injection mold inserts? • How do model-based values for ejection force compare to experimentally measured values? • Do cooling time, packing pressure and packing time affect ejection forces for conventional and rapid tooled injection mold inserts in a similar manner? • What are the coefficients of friction between the thermoplastic materials and the core materials during ejection? • How do standard friction coefficient test results compare to model-based calculations? • Based on these data, what are the potential benefits and limitations for using rapid tooled inserts for small quantity injection molding? 10 1.4 Research Description The present research investigated the ejection portion of the thermoplastic injection molding process. First, ejection forces were measured experimentally for parts produced from steel, laser sintered steel (infiltrated with bronze), and stereolithography resin mold inserts. A full factorial statistical experiment was designed to determine the effects of three process parameters on the ejection force. Second, the experimental ejection force values were compared to calculated values from an ejection force model. Standard friction testing was conducted to determine static friction coefficients to use in the model. Model-based values for the static friction coefficients were also determined and compared with the standard test results. These results, along with observations of tool performance, provide some indication of how successfully the rapid tooled inserts might apply to injection molding. The chosen part for the experiments is a vented, closed-end cylinder, similar to the plastic canisters used to store photographic film. The thermoplastic materials were chosen according to their moldability for the given application and their range of applications for manufacturing and consumer products. High density polyethylene (HDPE), a semicrystalline thermoplastic, and high impact polystyrene (HIPS), an amorphous thermoplastic, are widely used consumer resins, and are known to be well suited to injection molding. The experimental core and cavity pairs were built as inserts that were fitted to a standard mold base. The rapid tooling processes and materials have been chosen 11 according to their potential application for injection molding and to their availability for experimentation. These processes are laser sintering and stereolithography. While the experiments with the laser sintered insert were similar to those with the baseline steel insert, the stereolithography insert posed more of a challenge due to the softness of the material and its insulating qualities. The number of experimental runs was determined by the designed experiment, which varied three input parameters: packing pressure, packing time, and cooling time. These input parameters are key in defining an optimal injection molding process and producing a quality part. For each experimental part, ejection force, temperature at ejection, and part diameter data were collected. The ejection force data were compared with values calculated using a model for estimating ejection force developed by Menges. Apparent coefficients of friction for all material pairs were calculated using the Menges model and data from the experiments. These values were compared with results from standard friction tests. Statistical analysis was performed to determine the effects of the three input parameters on ejection force. 1.5 Organization The remainder of this document is organized as follows. Chapter 2 is the result of a literature search of relevant previous work. It first presents topics related to ejection forces in injection molding, such as ejection force models, shrinkage, friction and 12 adhesion. A section on rapid tooling is also presented that includes a background of rapid prototyping processes and details of the stereolithography and laser sintering processes. Chapter 3 presents the theoretical basis for this work. This includes the materials aspects of the two thermoplastics used in the experiments, HDPE and HIPS, and further definition of the coefficient of friction. The last section in Chapter 3 derives the equation for ejection force. Chapter 4 provides details on how the experimental work was accomplished. It describes the standard friction test, the injection molding experimental design, the part and tool designs and data acquisition. Chapter 5 presents all test and experimental results and statistical analysis, and Chapter 6 presents conclusions, implications and future work. References are listed following Chapter 6. There are two appendixes: Appendix A includes data tables, and Appendix B includes part and tool drawings. 13 CHAPTER 2 LITERATURE SEARCH This chapter presents comprehensive results of a literature search on the topics of ejection force and rapid tooling. Extensive work has been published on topics related to ejection forces, including shrinkage, friction, adhesion, and modeling. Topics cited in rapid tooling include various rapid prototyping processes and works specifically pertaining to stereolithography and laser sintering. 2.1 Ejection Force 2.1.1. Ejection Force Models Several researchers have developed force equations for the ejection of parts from injection mold cores based on mechanical or thermo-mechanical models. Most of these equations derive from the friction-based concept FR = f × p A × A (see derivation in Chapter 3), where FR is the ejection (or release) force, f is the coefficient of friction between the mold and the part, pA is the contact pressure of the part against the mold 14 core, and A is the area of contact. While area is a straightforward measure, friction coefficient and contact pressure have various interpretations or methods of estimation. A number of models and variations are summarized below. The version of the ejection force equation developed by Menges et al for a vented cylinder defines contact pressure as p A = E (T ) × ∆d r × s m 2.1 and therefore ejection force is: FR = f × E (T ) × ∆d r × s m × 2πL 2.2 where E(T) is the elastic modulus of the thermoplastic part material at ejection temperature, ∆dr is the relative change in diameter of the part immediately after ejection, sm is the thickness of the part, and L is the length of the part in contact with the mold core (Menges, Michaeli, & Mohren 2001). The rationale for this formulation is that shrinkage of the part is constrained by the core, thus causing stresses to build up in the cross sections of the part and resulting in forces normal to the surfaces restrained from shrinking. When the part is ejected from the mold, the stored energy-elastic forces can recover spontaneously. The relative change in circumference, measured immediately after ejection, is used as a measure of tensile strain in the cross section of the part while it is still on the core. The strain multiplied by the elastic modulus, the surface area in 15 contact, and an assumed friction coefficient then gives an estimate of the force required to remove the part from the core. Malloy and Majeski (1989) referenced the ejection force equation as used by Menges et al and a more detailed version by Glanvill, as shown below. Their paper examined ejection variables with respect to designing ejector pins. Burke and Malloy (1991) further discuss aspects of contact pressure and coefficient of friction. They showed that ejection force is affected by cooling time, surface finish, direction of polish, and draft angle. Their version of the ejection force equation for a box-shaped part (not vented) is as follows: FR = f × E (T ) × α × (TS − TE ) × 8s m L + (W1W2 PA ) 1 −ν 2.3 where α is the coefficient of thermal expansion (contraction), TS is the temperature at the onset of shrinkage (determined using a secondary empirical calculation), TE is the temperature at ejection, ν is Poisson’s ratio, W1 and W2 are the widths of two sides of a rectangular core, and PA is atmospheric pressure. The authors applied this equation to determine the apparent coefficient of friction at various surface finishes. Michalski (2000) used a version of the equation for closed cylindrical sleeves to measure ejection force for film canisters. This version of the ejection force equation took into account vacuum forces and an adjusted value for f due to the taper of the part (Menges, Michaeli, & Mohren 2001). 16 Glanvill (1971) is another oft cited reference for ejection force. His equation defines contact pressure as pA = α (Tm − Te ) × E 1 ν − 2t 4t 2.4 where Tm is the softening point of the thermoplastic and t is thickness. Thus, FR = α (Tm − Te ) × E × πL × f 1 ν − 2t 4t 2.5 Hopkinson and Dickens (1999, 2000a, 2000b) used Glanvill’s equation to predict ejection force for parts molded with stereolithography tools. The authors have done extensive work with stereolithography tools as described later in this chapter. A model by Pham and Colton (2002) was developed from Glanvill for stereolithography molds, taking into account friction and shrinkage, as well as the stairstep (roughness) aspect of stereolithography molds with draft angles θ. They defined two components of force, one due to friction and another due to the stair-step surface as follows: FR = F fric.therm + Fdef . stair 2.6 17 The ejection force is FR = A × ( f eq cosθ − sin θ ) × Ptherm 2.7 where contact pressure is Ptherm = α p ∆T p rp − α m ∆Tm rm 1 rm E p rp2 + rm2 1 −ν m + + ν p r2 − r2 E m p m 2.8 where r is the hydraulic radius, the p subscript refers to the part, and the m subscript refers to the mold. This model derives from the ejection force equation for a general mold with a core feature and uses an approximation for thick-walled cylinders. The model was applied, along with finite element analysis and experimentation. Results showed the Pham and Colton model to be more accurate than Glanvill in this case. Colton, Crawford, Pham, and Rodet (2001) showed that the ejection force model for stereolithography molds gives reasonable results when compared to experimental results. In this work, the build orientation of the stereolithography tool had no effect on mechanical properties. Mechanical properties of the mold were shown to degrade with higher temperatures. Brittle fracture of the molds occurred below the glass transition temperature, while yielding occurred above the glass transition temperature. Palmer and Colton (2000) used this model to predict ejection failures of stereolithography mold features based on height ratio, aspect ratio, and draft angle. 18 Height ratio was the most critical factor in determining feature life, while aspect ratio had no conclusive effect. As expected, larger draft angles increased feature life. Fatiguebased chipping failures also occurred. Cedorge and Colton (2000) studied the stair step effect of stereolithography tools. Surface roughness resulting from the stereolithography build process depended on layer thickness and draft angle. The authors showed a trade-off between these two parameters in terms of ejection force. For tools built with thin layers, ejection force decreased with draft angle, while for thick layers, ejection force increased with draft angle. Colton and LeBaut (2000) showed that ejection force decreases with number of shots in a stereolithography mold. This was because the mold gradually heated up, and shrinkage was less because the mold and part temperatures were closer together. The authors also found that the stereolithography material continued to cure and become harder. Another version of the ejection force equation was presented by Shen et al (1999) for hollow, thin-walled cones, taking into account draft angle θ and vacuum forces: FR = 2πEεs m L cosθ ( f − tan θ ) × + 10 B 1 −ν 1 + f sin θ cos θ 2.9 where ε is elastic strain in the thermoplastic and B is the projected area of the core surface in the core axis direction. The first term in this equation refers to contact pressure, which was determined from a force and stress analysis for a hollow, thin-walled cone. The second term refers to the friction force, and the third term refers to the vacuum 19 force. Experiments by Shen, et al showed agreement with model results, though molding parameters were not discussed. Pontes, et al derived a thermo-mechanical model for amorphous materials based on average internal stress (Pontes et al 2001, Pontes, Brito and Pouzada 2002, Pontes, et al 2002; see also Jansen and Titomanlio 1996). The model assumed that stresses in each layer of the part start to develop when the layer solidifies, and relaxation in the solid polymer is negligible because of the high cooling rate. For a cylindrical part σ θθ = E (Te ) E (Te ) t × (− βPs + α (Ts − Te )) − × ×δ r 1 −ν 1 − ν 2 Dm te t r* 2.10 where σ θθ is the average circumferential stress before ejection, β is compressibility, PS is the pressure as each layer of polymer solidifies, Dm is the center thickness coordinate, δr is thickness shrinkage, te is the time of ejection and t r* is the time of solidification. The first term of this equation represents pressure induced effects, the second term represents thermal contraction, and the third term represents thickness shrinkage, which reduces average internal stresses. The authors found that ejection force decreased with increasing surface temp at ejection (for polystyrene), increased slightly, then decreased with increasing holding pressure (for polystyrene and polypropylene), and decreased with increasing holding pressure (for polycarbonate). Experimental results agreed with the model. 20 Kabanemi, et al (1998) derived a numerical model for prediction of residual stresses, shrinkage and warpage for thin, complex injection molded products. Wang, Kabanemi, and Salloum (1997, 2000) presented the numerical approach to predict ejection force from mold-part constraining forces and friction forces. It included finite element thermoviscoelastic solidification analysis to account for stress and volume relaxation of polymers under cavity-constrained conditions, and predicted distribution of ejection force among ejector pins. The model worked well for a rigid polymer (polycarbonate), but HDPE had significant post-molding shrinkage and warpage that was not taken into account. Several examples of research using models for injection force have been described in this section. Many researchers have used the Menges or Glanvill models, while others have derived their own models. Much of this work has shown the effects of various parameters on ejection force and has illustrated the many different variables that need to be taken into account. The present work follows up on these ideas by determining the effects of three parameters on ejection forces for three mold insert materials and two thermoplastic materials, and by applying an existing ejection force models (from Menges) and comparing the results to experimental values. The present work is unique in that it includes three different injection mold inserts in the same experiment, and two of these are made by rapid prototyping processes. It is also unique in that it includes values for modulus at temperature, standard measurements of coefficients of friction at elevated temperatures, and near real time measurements of part diameters (to determine shrinkage and part thickness). 21 2.1.2 Shrinkage An important aspect of the above ejection force models is shrinkage of the thermoplastic part. Shrinkage influences the contact pressure of the part against the core and can affect strain and friction. The extent of shrinkage that occurs depends on material properties and process conditions. The following works, most by researchers previously mentioned, address shrinkage in the context of thermoplastic injection molding and ejection forces. Malloy and Majeski (1989) explained aspects of shrinkage that relate to the injection molding process. They stated that shrinkage values for thermoplastics are often given in ranges because they vary both parallel and perpendicular to flow and with process conditions. Standard shrinkage values, however, have limited value in determining ejection forces since ejection is normally at elevated temperatures. Deep gates, long holding times, and high holding pressures in the injection molding process can compensate for shrinkage of the part. In calculating the ejection force, accurate values of the coefficient of thermal expansion may not be available since it is a function of temperature and pressure in the process. Therefore, shrinkage (strain) values can be used instead. Burke and Malloy (1991) stated that shrinkage results from thermal contraction and directional distortion. Thermal contraction is due to atomic vibration in which atoms move closer together at lower energy levels, and directional distortion results from orientation of polymer molecules during flow, and their subsequent relaxation back to a 22 coiled state after flow ends. Shrinkage is greatly influenced by ejection temperature, is material dependent, and varies for amorphous and semicrystalline polymers. Semicrystalline materials exhibit greater shrinkage due to phase transformation of the crystalline portion: random amorphous coils and high free volume in the melt reduce to orderly packed chains in the crystal lattice. Amorphous polymers, on the other hand, contract much more gradually. Michaeli et al (1999) modeled the development of material properties and crystallization due to processing. They found that the temperature at which the crystallization peak occurs decreases, and the crystallization interval widens, with increasing cooling rate. That is, crystallization starts earlier at lower cooling rates. Menges, et al (2001) stated that, for both amorphous and crystalline thermoplastics, holding pressure exerts the greatest effect on shrinkage (degressive effect) in the injection molding process. The temperature of the material is the second major factor influencing shrinkage. Higher temperature results in higher thermal contraction potential, but also lowers viscosity for better pressure transfer. With a longer holding time, the effect of improved cavity pressure predominates for crystalline materials. Menges et al provided shrinkage values for some thermoplastics, but stated that the best data are found through experience. Pantani and Titomanlio (1999) found that higher pressure histories inside the injection mold cavity – obtained by increasing either holding time or holding pressure – result in a lower final shrinkage and in a delayed start of shrinkage inside the mold for a polystyrene plate. 23 In mold experiments with polycarbonate, Pontes et al (2001) found that increasing holding pressure reduces contact pressure by decreasing diametrical shrinkage. Holding time, however, had no effect on the shrinkage due to fast solidification of this material. In their ejection force modeling work, the authors described average circumferential stress to include volumetric shrinkage due to thermal contraction (and crystallization) and thickness shrinkage, which reduces stress. Ejection force, then, depended on elastic modulus, friction coefficient, part thickness, and variation of the volumetric shrinkage. In their model, initially, ejection force increased (or plateaued) with increasing holding pressure because of thickness shrinkage, while at higher holding pressures a reduction in volumetric shrinkage reduced ejection force. As indicated by the research described in this section, shrinkage varies with parallel and perpendicular flow, injection and ejection temperature, holding pressure and time, material structure and properties, and pressure histories. In the present work, while shrinkage is not analyzed directly, it is measure and used in the Menges model to calculate ejection force. Shrinkage influences the contact pressure of the part on the core (see Chapter 3). Furthermore the effects of cooling time, packing pressure, and packing time on ejection force are determined in part by the shrinkage characteristics of the thermoplastic material. 24 2.1.3 Friction and Adhesion Friction is another important aspect in determining ejection forces. Friction between the thermoplastic part and the injection mold core not only depends on the mechanical relationship between the two surfaces, but also on an adhesive component inherent in the properties of the two materials at processing conditions. The following works address friction and adhesion, some in general terms, others specifically as they apply to polymers and injection molding. Contact between two solids occurs only at asperities (ed. Eley 1961, Ch. V). Extremely high pressures are produced at these contact points and, in metals, plastic flow occurs. Under plastic conditions, the area of real contact is directly proportional to the load and is independent of the apparent area of contact. During sliding at slow speeds, with no temp increase, fragments of one metal can strongly adhere to the other (cold welding). The frictional force is then the force required to shear the junctions formed in this way. With softer metals junctions are more ductile and easily deformed, and appreciable adhesion may occur. Relatively smaller adhesions occur in plastics due to higher elastic recovery. Adhesion may thus occur by reducing elastic stress or by increasing ductility. A number of concepts relating polymer friction and adhesion to thermoplastic injection molding (with steel molds) were presented by Burke and Malloy (1991) and are summarized below. More on friction and adhesion theory is presented in Chapter 3. 25 • Plastics have relatively low modulus values, which lead to frictional values that are not always directly proportional to load. This is attributable to adhesion and deformation. • Theoretical calculations show that van der Waals forces, which attract molecules with permanent dipoles, and London dispersion forces, which cause dipoles created by motion of electrons in the molecule, are great enough to produce bonds exceeding the cohesive strength of most adhesives. • In the surface energy theory, a liquid may wet and spread over a solid surface if the critical surface tension of the solid is greater than that of the liquid. Heat decreases viscosity and improves wettability; heat and pressure promote wetting and spreading. Molten polymers on steel under injection molding conditions are a good environment for wetting and spreading. • Wetting and spreading does not necessarily imply adhesion. Apparently both physical adsorption and surface energy criteria must be met for adhesion to occur. An increase in adhesion will increase the apparent coefficient of friction, which depends on the specific polymer-steel combination. • Surface roughness causes mechanical coupling and increases surface area over which van der Waals forces can act. Imperfect surfaces lead to inherent voids or trapped gas bubbles and imperfect molecular fit, limiting the bond strength. • The coefficient of static friction increases with increasing surface roughness, depending on viscosity and pressure applied. A highly viscous material under low pressure may not wet the steel. The direction of polishing affects part 26 ejection. The coefficient of friction decreases with increasing cooling time because shrinkage decreases the mechanical anchorage of the polymer, i.e., it no longer completely penetrates irregularities in the mold. Looking at polymer adhesion from the standpoint of wear, Briscoe (1981) summarized several fundamental aspects, including cohesive wear and interfacial wear. Cohesive wear mechanisms occur adjacent to the interface, e.g., abrasion and fatigue wear induced by tractive stresses. Interfacial wear processes dissipate frictional work in much thinner regions and at greater energy densities, e.g., transfer wear and chemical or corrosive wear. In interfacial wear, frictional work originates from adhesive forces emanating from the contacting solids. These forces generate localized plastic surface deformation and transfer of relatively undegraded polymers to the counterface in certain systems. Also discussed in this paper is natural adhesive or transfer wear, specifically, initial adhesion. Briscoe stated that initial junction strength is a function of the interaction of surface forces and mechanical properties of the contact. For polymers, the surface forces consist of van der Waals, coulombic and possibly hydrogen bonding forces. The higher the surface free energy of the polymer, the greater the adhesive force. Polymers above the glass transition temperature will adhere more strongly because they conform to surface imperfections and have a relatively low level of stored elastic strain. Very clean metal surfaces may promote chemical bonding. Essentially brittle and highly 27 elastic crosslinked polymers tend to fail at the interface. This includes polymers below the glass transition temperature and crosslinked systems. Czichos (1983, 1985) investigated contact deformation, static friction, and tribological behavior of polymers. In his work contact deformation was measured for four crystalline thermoplastics in loading and unloading conditions. A model was proposed that takes into account elastic, viscoelastic, and viscoplastic components. Polymer to polymer coefficients of friction were measured, using a pin-on-disc configuration, and plotted against sliding distance. Experimental frictional work was plotted against the work of adhesion using the Dupre equation (see Chapter 3), showing that a reasonable correlation exists. Also, coefficient of friction and wear rate were plotted against surface roughness for four polymers against steel. The author found that adhesion was the primary influence for very low surface roughness, while abrasion was the primary influence for higher roughness. Benabdallah and Fisa (1989) measured the static friction coefficient between a steel surface and three thermoplastics with surface roughnesses varying between 0.4 µm RMS to 40.5 µm RMS. Parameters measured were normal load, relative displacement and tangential force. In these tests, static coefficient of friction decreased with increasing normal loads ranging up to 160 N. This is explained by an increasing influence of the adhesion component of friction. Also, the friction coefficients decreased with increasing surface roughness since, with smoother surfaces, there is more adhesion. The authors present a model for friction coefficient µs based on this work: 28 µs = αFNn 2.11 where FN is normal load, α is a proportionality constant that depends on polymer surface roughness and n is experimentally determined for each polymer. Benabdallah (1993) investigated static shear strength during contact between a bulk plastic and a metallic plate, both with smooth surfaces. The experimental equipment included one apparatus to measure static friction force and another to measure the real area of contact. The adhesion component of friction was approximated by the measured static friction force. The author determined bulk shear strength of plastics experimentally following ASTM D732-78 and found surface energies from the Young equation (see Chapter 3). In this work the friction force was assumed to consist only of the adhesion component due to the smoothness of the contacting surfaces. That is, the deformation (or ploughing) component was not considered. Friction force equaled the maximum tangential load, which corresponded to the minimum force required to initiate motion. The paper includes plots of the adhesion component of friction against the calculated work of adhesion according to the Dupre equation, Wa = 2φ (γ 1γ 2 ) 2 , where the 1 interaction parameter, φ, equals 1, and the static shear strength against the real contact pressure (the ratio of applied load and real area of contact). Benabdallah then combined the Young equation with the geometric mean equation to obtain: 29 γ L (1 + cosθ ) ( ) 2γ d L 1 2 ( ) = γ p S 1 γ p 2 L γ d L 1 2 ( ) + γ Sd 1 2 2.12 This equation is in the form y = mx+b. Given known surface energies of six liquids, contact angles were measured and x and y plotted. Then the square of the intercept determined the dispersion component, the square of the slope determined the polar component, and the addition of the two gave the total surface energy of the solid γs. The author concluded that the adhesion component of friction increases with the real area of contact and is large when the surface energy of the plastic material is high. It was also found that a correlation may exist between the adhesion component of friction and the work of adhesion when evaluated as a function of the real area of contact. Menges and Bangert (1981) measured static coefficients of friction for determining opening and ejection forces in injection molding. This work looks at the effects of various parameters, including surface (contact) pressure, cooling time, mold temperature, holding pressure, and surface roughness. Various thermoplastics were studied, but results were reported only for polypropylene. In all cases, the friction coefficient decreased with surface roughness in the range 1 to 35 microns. In general the friction coefficient also decreased with increasing cooling time. The effects of other parameters were varied. The friction coefficient results varied from standard measurements for polypropylene. Balsamo, Hayward and Malloy (1993) conducted studies on ejection forces and coefficients of friction. The authors found, first, that external lubricants have a large 30 effect on ejection force. They also measured static friction coefficients for polystyrene, polypropylene, a polycarbonate/polyethylene alloy, and filled polycarbonate parts on nickel, steel, and polytetrafluoroethylene (PTFE)/nickel plated mold cores. While noting that the friction test does not exactly duplicate the injection molding environment, they found that friction coefficients are generally lowest on PTFE/nickel surfaces and highest on steel surfaces. Some coefficients changed significantly with temperature. In Dearnley’s work (1999) to study low friction surfaces for injection molds, steel rings were coated with TiN (polished), CrN (polished and spark eroded), and MoS2 (polished and spark eroded) and used as a core around which an acetal ring was molded. Coating thickness, surface hardness, and surface roughness were measured, and friction force was determined experimentally. Spark eroded surfaces were found to have higher roughness values and higher friction forces compared to polished surfaces. The author attributed this to mechanical interlocking. Polished CrN had the lowest friction forces even though polished TiN and MoS2 had lower roughness values. This was attributed to possible differences in chemical behavior at the interface, e.g., lower surface energy (or wettability) of CrN coatings. Pontes et al (1997) studied the effects of processing conditions on ejection forces for tubular moldings. Parameters included surface roughness, injection temperature and holding pressure. Two thermoplastics were used, one amorphous and one crystalline. For polyphenylene ether (PPE), ejection force increased with injection temperature and decreased with holding pressure, as would be expected. For polypropylene, ejection force decreased initially with surface roughness (less than 0.75 microns), then increased. 31 For polypropylene, ejection force decreased as injection temperature increased, indicating that the core surface temperature was different than ejection temperature, and that deformation ability increased with higher temperature. For polypropylene, ejection force decreased with increasing holding pressure, as was expected. The diametrical shrinkage was determined from shrinkage at room temperature and the coefficient of thermal expansion. Calculated values for the equivalent coefficient of friction of PPE (amorphous) were near the lower range of published values. The authors concluded that adhesion appears to be an important factor for semicrystalline materials molded on surfaces with low surface roughness. Sasaki et al (2000) molded cylindrical parts with polypropylene, polymethyl methacrylate (PMMA), and polyethylene terephthalate (PET) at a range of surface roughnesses from 0.016 to 0.689 microns Ra. In all cases, ejection force increased significantly when surface roughness approached zero. Optimum surface roughness (in terms of ejection force) for polypropylene and PET was approx. 0.2 microns and for PMMA was 0.009 microns. For lower values of roughness, the meniscus force or van der Waals force was thought to be the greatest factor, whereas for higher values of roughness, the “engraving or scratching” of the surface came into play. Ejection force was also measured on polypropylene and PET parts from cores with various coatings. Here, tungsten carbide/carbon coating was found to be most effective for reducing ejection force. TiN (HCD), TiN (Arc), DLC, and CrN coatings also showed ejection force reduction effects. 32 Ferreira et al (2001) friction tested polycarbonate and polypropylene using a special prototype apparatus. The testing procedure included heating the specimens to processing temperatures, applying a normal load (so that the specimen replicated the mold surface), cooling to ejection temperature, then pulling the specimen. At room temperature, the coefficient of friction of polycarbonate at 0.32 was similar to published values at 0.31. At high temperature, the coefficient was much higher at 0.47. For polypropylene at high temperature, the coefficient of friction was 0.19, much lower than published values, 0.36. A similar approach to imprinting the mold surface onto the specimens was used in the standard friction tests in the present work (see Chapter 4). In related work, design of experiments was used to determine the effect of polish direction, surface roughness, and temperature on the coefficient of friction (Ferreira et al 2002). Results showed that testing temperature and surface roughness had a significant effect on the coefficient of friction for polycarbonate. For polypropylene, none of the parameters had a significant effect on the coefficient, except possibly the interaction of polish direction and roughness. Friction values for both polymers were higher than published values. Muschalle (2001) measured the coefficient of friction for polycarbonate (amorphous) and polypropylene (semi-crystalline) materials against steel with two different surface roughnesses, machining directions, and temperatures. Results showed that the friction coefficient for polycarbonate was higher at higher temperature, while that for polypropylene was lower at higher temp. Also, the coefficient for polypropylene was 33 higher when temperature and pressure caused surface reproduction of the metal on the plastic. The works described in this section indicate the many aspects of friction, which includes both a mechanical and an adhesive component. No unifying theory seems to exist for friction, but, rather, it can be explained by one or another or a combination of concepts. This can be seen in the friction testing in the present work, where coefficients of static friction are influenced by adhesion and/or mechanical components of friction to varying degrees (reference Chapters 5 and 6). 2.2 Rapid Tooling 2.2.1 Background One of the most promising techniques for low volume, net shape manufacturing tools is rapid tooling, i.e., the application of rapid prototyping processes for the purpose of making tools. Rapid tooling processes are additive and produce a tool or pattern from a CAD model. Direct rapid tooling processes generate the tool itself from the CAD file, while indirect rapid tooling processes require intermediate steps and usually generate a pattern from which a tool is made. A wide range of materials can be used in rapid tooling processes, from waxes and resins to ceramics and metals. Those processes that use metal materials and build tools directly tend to be more suitable for production tools. These 34 include laser sintering, 3D printing, and laser generating (Karapatis, van Griethuysen & Glardon 1998). To date most rapid tooling technology for production parts has been aimed at meeting the same process requirements as conventional tooling. Those direct and indirect rapid tooling processes that use metal materials have been most successful with this approach. Tools from non-metal processes, however, can possibly be used under non-conventional molding conditions. These include, for example, stereolithography and cast epoxies. Several rapid tooling processes are described in the following paragraphs. Laser generating processes deposit highly dense metal materials and come very close to meeting the material properties of conventional molds. The laser engineered net shaping (LENSTM) process, for example, focuses a high power Nd:YAG laser and creates a molten puddle on a substrate, into which metal powder is injected (Keicher, Gorman & Taute 2001). A schematic of the process is shown in Figure 2.1. An injection mold insert with conformal cooling channels was successfully built for a high volume automotive part using the LENSTM process (Optomec 2001). 35 Figure 2.1: A schematic of the laser engineered net shaping (LENS™) process (Castle Island 2003). Three other rapid tooling processes that use metal materials are laser sintering, direct metal laser sintering, and 3D printing. In the laser sintering process, a laser is scanned over powdered material with a binder coating, and the part is built layer by layer. Laser sintering is described in detail later in this section. Direct metal laser sintering (DMLS) is a similar process in which the metal itself is sintered without any polymer binder. 3D printing processes spray a binder material in an ink-jet-printing fashion onto successive layers of metal powder. All of these processes include steps for debinding, sintering and infiltration. 36 Stereolithography, like laser sintering, employs a scanning laser, but uses liquid resin build materials. The stereolithography process is described in detail later in this section. Indirect processes generate a pattern using rapid prototyping techniques, then build a tool from that pattern. Although indirect processes have more steps, they benefit from a wider range of material choices. At the Pennsylvania State University, a powder metallurgy process was used to make a mold insert (Weaver et al 2000). From a three dimensional model, a pattern was generated using a three dimensional plotting process, a negative of the tool was cast in silicone rubber, and a slurry of steel and ceramic was cast into the negative to make the final tool insert. Such a tool has excellent mechanical properties and is capable of high volume production. Another option for indirect tooling is cast epoxy, in which a blend of resin and aluminum filler is cast over a rapid prototyped pattern. The resulting tool insert is machineable, capable of withstanding typical molding pressures and temperatures, and can produce low volumes of prototype or production parts. The two rapid tooled injection mold inserts for this research were built using laser sintering and stereolithography processes. These represent two very different processes among the spectrum of rapid tooling techniques. While the laser sintering material is more like conventional tool steel, the stereolithography material is unlike what you would expect in a production tool. The two processes were chosen first for their availability, and second based on their economic potential for producing small quantities of parts. In the remaining paragraphs of this section, these processes will be described in detail. 37 Laser sintering is an additive layer process in which a laser melts powdered material by cross sections to build a part. The process is versatile in that it can use any of several powdered materials including polymers, ceramics, and metals. For the Selective Laser Sintering (SLS®) process, developed by the University of Texas at Austin, early processing materials included wax, polycarbonate, unreinforced nylon, and glass reinforced nylon (McAlea et al 1995). Today metal materials can be sintered and infiltrated for higher densities. The laser scans across each layer of metal powder coated with a polymer binder, and fuses the binder to create the tool (Figure 2.2). Later the “green” part is sintered and infiltrated with copper or bronze (Beaman et al 1997, Kai & Fai 1997). Figure 2.2: A schematic of the selective laser sintering process (Castle Island 2003). 38 The selective laser sintering process was used to make one of the injection mold inserts for this research. This process involves a polymer-coated 420 stainless steelbased powder, known as LaserForm ST-100, and a 3D Systems Vanguard machine. The sintered ST-100 material was subsequently infiltrated with bronze. Specifications of the Vanguard are shown in Table 2.1 and material properties of ST-100 are shown in Table 2.2. 39 Model Number Laser Wavelength Power Beam Diameter Max. Scan Speed Min. Layer Thickness Build Chamber LC-100 DEOS CO2 Laser 10.6 microns 100 W max at part bed 450 microns 10,000 mm/s (394 in/s) 0.10 mm (0.004 in) 381w x 330d x 457h mm (15w x 13d x 18h in) Table 2.1: Vanguard System Specifications from 3D Systems. 3 Density Thermal Conductivity 7.7 g/cm o o 49 W/m K @100 C CTE Tensile Yield Str. (0.2%) Tensile Strength Young's Modulus Elongation Compression Yld Str (0.2%) Hardness, HRB 56 W/m K @200 C o 12.4 ppm/ C 305 MPa 510 MPa 137 GPa 10% 317 MPa 87 As infiltrated 79 As machined o ASTM D792 ASTM E457 o ASTM E831 ASTM E8 ASTM E8 ASTM E8 ASTM E8 ASTM E9 ASTM E18 Table 2.2: LaserForm ST-100 (Sintered and Infiltrated) Material Properties from 3D Systems. 40 When the 3-dimensional part is initially built on the Vanguard System, the laser heats the metallic particles above the glass transition temperature of the polymer coating. The polymer softens and deforms, then fuses with other particles at each contact surface. The temperature is such that melting of the metal does not occur, only viscous flow of the polymer coating. The metal powder is then bound together by the polymer to form the “green” part. After the build is complete, the green part is removed from the machine and excess powder is brushed away. A furnace cycle follows in a reducing atmosphere to burn off the polymer, sinter the steel powder, and infiltrate the part with bronze. Infiltration eliminates any voids within the steel, resulting in a fully dense part (Bourell et al 1994, McAlea et al 1995). Stereolithography, a non-metal process, uses a laser to scan a vat of liquid resin and build a part layer by layer. Stereolithography resin, in direct comparison to conventional mold steel, has vastly different mechanical and thermal properties. With enhancement, such as a metal backing, a metal coating, or water cooling channels, a stereolithography resin mold still underperforms an aluminum one under traditional molding conditions (Li, Gargiulo & Keefe 2000). Nevertheless, there are several examples of research in stereolithography tooling, as described in the next section. Stereolithography was one of the first rapid prototyping processes to emerge, and the 3D Systems stereolithography apparatus (SLA) was a pioneer rapid prototyping system in the late 1980s (Kai & Fai 1997). The SLA system consists of a control computer, a control panel, a laser, an optical system, and a process chamber. The SLA 250, appropriate for many applications, has been widely used across the globe and, in 41 fact, was used to make mold inserts for this research. Specifications for the SLA 250 machine are shown in Table 2.3. The SLA uses a photo-curable liquid resin as a build material. Many resins are available depending on the type of laser in the machine and the requirements of the part to be built. For this research, one of the rapid tooled injection mold inserts was built using the stereolithography process and SL 5170 resin from Vantico. Properties of this resin are shown in Table 2.4. The SLA set-up includes a vat of the photo-curable liquid resin, inside which an elevator table is set just below the resin surface (Figure 2.3). A solid model CAD file in .STL format is loaded into the machine. The model is sliced by the control unit into cross sections, which are solidified by the SLA laser one at a time. After each layer is solidified, the elevator drops just enough to cover the solid layer with a new coat of liquid resin. The part is built in this manner from the bottom up. When completed, the elevator raises the part out of the vat, and the excess liquid resin is removed. Figure 2.3: A schematic of the stereolithography process (Castle Island 2003). 42 Laser Wavelength Power Beam Diameter Max. Drawing Speed HeCd 325 nm 24 mW 0.20-0.28 mm 762 mm/s Min. Layer Thickness Elevator Resolution Max. Part Weight Vat Capacity Max. Build Envelope 0.1 mm 0.0025 mm 9.1 kg 32.2 L 250 x 250 x 250 mm Table 2.3: SLA 250 System Specifications from 3D Systems. Tensile Strength Tensile Modulus Elongation at Break Glass Transition Temp CTE Thermal Conductivity Hardness, Shore D Density 59-60 MPa 3737-4158 MPa 8% o 65-90 C o 90 ppm/ C ASTM D638 ASTM D638 ASTM D638 DMA TMA (T<T g) o 0.200 W/m K 85 3 1.22 g/cm DIN 53505 Table 2.4: Vantico SL5170 Typical Properties (90-minute UV post cure). 43 The mechanisms that are the basis for the stereolithography process are free radical and cationic photopolymerization. Polymerization is the process by which monomers are linked into larger, chain-like molecules called polymers. Further linking leads to the crosslinking of these chains. In free radical polymerization, heat or light energy decomposes an initiator to generate free radicals that catalyze the polymerization process. In cationic photopolymerization, cationic photoinitiators cause reactions that open molecular ring structures to catalyze the polymerization process. Free radical photopolymerization is associated with acrylate resins, and cationic photopolymerization is associated with epoxy resins. The resins used in the 3D Systems’ SLA machines are UV-curable photopolymers made up of photoinitiators and reactive liquid monomers. In the SLA polymerization process, sufficient crosslinking is required to prevent the polymer molecules from dissolving back into monomers. Furthermore, since the cured resin must withstand forces during recoating, the polymer molecules must be sufficiently strong. Increasing the laser power results in a higher polymerization rate and thus a faster build rate, but brittleness also results, due to lower molecular weight. Cure depth must be deep enough to prevent delamination, but not so deep as to cause distortion and, therefore, inaccurate parts (Kai & Fai 1997, Beaman et al 1997). 2.2.2 Stereolithography and Laser Sintering for Injection Molding Tools Rapid tooling processes lend themselves well to injection molds because of their ability to generate complex shapes as easily as simple ones. Complex shapes that are 44 difficult or impossible to machine, detailed internal structures, and thin walls can be readily generated. This allows the integration of conformal cooling channels within the mold, which lower residual thermal stresses and can reduce cycle times (Sachs et al 2000). Some rapid prototyping processes have the ability to vary material composition during fabrication. This local composition control benefits rapid tooling because it allows tailoring of various material properties, such as conductivity, corrosion resistance, and hardness (Cho, Sachs, & Patrikalakis 2001). It is theoretically possible, for example, to build an injection mold with a core of highly conductive material, such as copper, and surround it with a wear resistant material, such as stainless steel. Rapid tooling processes lend themselves well to low volume production because they reduce the requirements for labor intensive machining, minimize material waste, and, in some cases, use less expensive materials. Thus they have the potential to reduce tooling costs enough to make low volume injection molding economically feasible. Research in the area of rapid tooling for injection molding is varied. The work described in section 2.1.1 includes some research with rapid prototyped tools. Additional work specifically pertaining to stereolithography and laser sintering, is summarized below. Laser sintering uses powdered metals for tooling and is not as challenging as stereolithography in terms of strength and thermal conductivity, so there are more examples of its use for injection molding prototyping in industry (e.g., Campbell 2000). Laser sintering with copper polyamide has been used to build small, low volume mold inserts, such as those for a brake reservoir and a glass guide (Nelson et al 1998). In these 45 cases, the advantages of laser sintering included durability for up to hundreds of parts, low cost and lead time, and cycle times that are comparable to those with conventional tools. Pham, Dimov and Lacan (2000) studied characteristics related to laser sintered tool accuracy, including shrinkage of the tool material and finishing requirements. Mold insert accuracy requires fine tuning of scaling and offset factors due to shrinkage and careful planning of tool finishing processes. Two case studies indicate successful use of laser sintering for injection molding and gravity die casting. In cases for two injection molded parts, Dalgarno and Stewart (2001) studied cycle time effects of conformal cooling and molding costs based on tool durability, and compared laser sintered tool results with conventional tooling. Due to tool finishing requirements, they found no lead time advantage for the laser sintered tooling process. The laser sintered tools, however, did exhibit cycle time savings with conformal cooling channels and economic benefits at low demand rates. Other work includes optimization of shapes for heating and cooling lines in mold inserts made with the direct metal laser sintering process (commercialized by EOS), which sinters bronze particles and infiltrates with epoxy resin (Hopkinson & Dickens 2000a). Hopkinson and Dickens have investigated stereolithography tools for injection molding, including tool failure, tool strength and ejection force. In a comparison of stereolithography with aluminum injection mold tooling, they found that the low thermal conductivity can be advantageous since the tool surface stays above its glass transition temperature for easier ejection. Also, tool degradation due to thermal cycling is reduced, 46 and the ability to mold long thin slots is enhanced. Ejection forces in this work were calculated using the equations developed by Glanvill and Menges et al (Hopkinson & Dickens 2000b). In other work, models were developed to predict tool strength and ejection force (Hopkinson & Dickens 2000c, 2000d). Heat transfer through the tool was measured and modeled. Then the results of the heat transfer analysis were used in a finite element analysis model to predict tool strength. The model showed a decrease in tool strength with increased cooling time before ejection. Ejection forces were predicted based on a modified equation by Glanville and Denton. Longer cooling times were found to lead to higher ejection forces, as expected, due to part shrinkage onto the core. The predicted values, however, were approximately 30 percent lower than actual values, and the measured values contained some inherent variation. Harris and Dickens (2001) explored two design variables for stereolithography injection mold inserts, namely, layer thickness and draft angle. They found that ejection forces increase with increasing stereolithography layer thickness and decreasing draft angles, thus increasing the risk of mold breakage. Interestingly, the linear changes in these two variables cause nonlinear changes in the ejection force, suggesting that optimum values must be found that balance ejection force requirements with desired part design and economy of stereolithography process. A later paper describes their study of the morphology of thermoplastic materials injection molded from stereolithography and aluminum tools (Harris & Dickens 2003). Parts from the stereolithography tool had longer cooling times and higher crystallinity. Experimental work demonstrated that 47 crystallinity can be controlled by using a nucleating agent or by adjusting melt temperature. The work of Dickens and Rudgley (2001) demonstrates the successful use of stereolithography resin inserts to mold an engineering polymer. With much lower injection pressure, speed, and clamping force, poly ether ether Ketone (PEEK), an engineering polymer, was injected into a room temperature stereolithography mold insert. The low thermal conductivity of the insert allowed the mold to fill completely at the lower pressure level. The part was ejected at a higher temperature so that the insert flexed during ejection. The part was molded successfully, and the slower cooling resulted in higher crystallinity as compared to parts molded in a conventional tool. 2.2.3 Summary There are many rapid prototyping processes in use or under development today, some of which have been described in this section. A subset of the rapid prototyping processes can be applied to make tools, including injection mold inserts. The present work investigates tools from two of these processes, laser sintering and stereolithography. Laser sintering with powdered metal has been successfully used to build injection molds for limited quantities of parts. Stereolithography with epoxy resin has been the subject of research for injection mold inserts, but has not been used for production to any significant extent. This work follows up much of the work described previously in this section by taking a systematic look at inserts from these two processes for molding two different thermoplastic materials. Ejection forces and friction coefficients are measured, compared 48 with model-based calculations, and baselined against a machined steel insert. The data collected help to determine the applicability of these rapid tools to injection molding, at least in terms of their ability to withstand the forces of ejection. 49 CHAPTER 3 THEORY This chapter includes necessary theoretical background on polymeric materials, the coefficient of friction, and an ejection force model. First amorphous and crystalline aspects of the thermoplastic materials used in this work are presented. Some friction theory follows, including discussions on the deformation and adhesion components of friction. The final section derives the primary ejection force equation used in this research. 3.1 Thermoplastic Materials Chemical structures and some properties of the two thermoplastics used in this work are presented in this section because they relate to the shrinkage, friction, and strength characteristics of the materials. These characteristics explain much of the behavior of these materials in the present work during testing and experimentation and described in Chapters 5 and 6. 50 A thermoplastic material is a polymer that has a linear macromolecular structure and will repeatedly soften when heated and harden when cooled. Examples of thermoplastics include styrenes, acrylics, polyethylenes, vinyls, and nylons. A crystalline thermoplastic has sections of crystallinity, i.e., periodic ordering of molecules, whereas an amorphous thermoplastic lacks any long range molecular order. The characteristic differences between amorphous and crystalline polymers determine processing parameters and influence the properties of an injection molded part. Amorphous polymers have a second order transition, or glass transition temperature, above which the material flows, and below which the material is glassy (Trantina & Nimmer 1994). In general, they have lower and more uniform shrinkage, greater post-mold stability, and high melt viscosities. Amorphous polymers also tend to be more susceptible to chemical attack. Crystalline polymers have a well-defined melting point below which crystals are formed, and above which the crystals dissolve and the material flows. In general they shrink more, and shrink more anisotropically, have low melt viscosities (long flow lengths), and have more temperature dependent mechanical properties. Crystalline polymers also tend to be more resistant to solvents. 3.1.1 High Impact Polystyrene Polystyrene is a vinyl polymer (i.e., formed from hydrocarbon monomers with double carbon bonds) having a phenyl group attached to every other carbon atom in its hydrocarbon chain (Figure 3.1) (University of Southern Mississippi 2002). In atactic 51 polystyrene, the phenyl groups are distributed on either side of the carbon atoms in a random fashion. Thus, it is amorphous because its unwieldy and asymmetric structure is not conducive to regular crystal formation. Figure 3.1: Polystyrene monomer. Polystyrene is formed using the free radical vinyl polymerization process. This process depends on the use of initiators that, upon splitting, produce free radicals. The unpaired electrons in the free radicals attack the double carbon bonds, pair with one electron from that bond, and cause the other electron to become a free radical. The chain reaction continues in this way to propagate the polymer. High impact polystyrene (HIPS) is formed by adding polybutadiene rubber monomers during the polymerization process. HIPS is a graft copolymer that has a polystyrene backbone chain with polybutadiene grafted onto it (Figure 3.2). The polystyrene provides strength to the material, while the polybutadiene renders it less brittle. 52 Figure 3.2: High impact polystyrene. 53 The HIPS material used in this work is BASF PS 495F. Its glass transition temperature is 100 oC (212 oF). This material is more brittle at room temperature compared to high density polyethylene. The properties of PS 495F are given in Chapter 4. 3.1.2 High Density Polyethylene Polyethylene is also a vinyl polymer with a very simple hydrocarbon chain (Figure 3.3). High density polyethylene (HDPE) has linear molecules (Figure 3.4) that can pack more tightly together, as opposed to low density polyethylene that has branched molecules. Because of its regular symmetric structure, HDPE is conducive to crystal formation and is considered a crystalline polymer. HDPE cannot be produced using free radical vinyl polymerization because some termination reactions result in branching of the molecules. Instead, the Ziegler-Natta vinyl polymerization process is used. The Ziegler-Natta process involves transition metal catalysts and co-catalysts based on the Group III metals, and it can produce polymers of a specific tacticity (University of Southern Mississippi 2002). 54 Figure 3.3: Polyethylene monomer. Figure 3.4: High density polyethylene linear molecule. The HDPE used in this work is Lutene-H ME9180 from LG Chem. Its crystalline melting point is 133 oC (271 oF). More on material selection is presented in Chapter 4. 3.2 The Adhesion Component of Friction A few introductory concepts of adhesion are presented in this section because adhesion plays an important part in the friction between the part and the injection mold 55 core. This can be seen in the present work, especially in the case of a HIPS part molded in an epoxy insert, as described in Chapters 5 and 6. In the basic friction equation, the friction force F between two sliding bodies is equal to the normal force N pressing the bodies together, multiplied by a constant, i.e., the coefficient of friction µ. The force required to initiate motion between the two bodies is typically higher than the force required to maintain motion. Thus the coefficient of static friction, µstatic, is defined as the ratio of the force necessary to initiate motion to the normal force: µstatic = Fbreakaway N 3.1 Friction is comprised of a deformation component and an adhesion component, the latter of which is typically more prominent for polymer materials. While the deformation (or mechanical) component of friction tends to be more easily defined, the adhesion component is rather more complex. The following paragraphs include theoretical background on adhesive bonding and adhesion theory. In adhesive bonding, the surface tension of the adhesive should be less than the free surface energy or critical surface tension of the adherend (ed. Cagle 1973). This allows the adhesive to wet and spread. Wettability or tendency to adsorb can be measured by the contact angle (between the adhesive and the surface to be bonded) or the work of adhesion. 56 Forces in the wetting and spreading phenomena include chemical bonds, mechanical entanglement, physical and chemical adsorption, electrostatic forces of attraction, and combinations thereof. Physical adsorption involves secondary attractive forces, i.e., van der Waals forces: molecules with permanent dipoles, dipoles induced by permanent dipoles in neighboring molecules (Debye forces), and London dispersion forces. Dispersion forces are dipoles produced by the motion of electrons and are independent of molecular polarity. Dispersion forces are considered to be the major attractive force even when polar groups and hydrogen bonding groups are present. Hydrogen bonding is demonstrated by molecules with hydroxyl groups. If the critical surface tension of the solid is greater than the surface tension of the liquid, a good bond can occur. Surface free energies of metals range from 100 to 3000 ergs/cm2, while organic liquids (including molten thermoplastics) have surface free energies of less than 100 ergs/cm2. Heat serves to increase the ability of the adhesive to adsorb, dissolve, and disperse. Heat also decreases viscosity, thus increasing wetting and adsorption. Pressure and heat together improve wetting and spreading of more viscous materials. Some of the equations of adhesion theory that derive from surface energy are introduced below. Surface energy or surface tension is represented by γ, where subscripts S and L represent solid and liquid, respectively (Wu 1982). Interfacial energy is represented by γLS. The Young equation relates contact angle θ, formed between a drop of liquid and a solid surface, to interfacial tensions as follows: 57 γ LV cos θ = γ SV − γ SL 3.2 where θ is the contact angle of a liquid on the plane surface of a solid, γLV is the surface tension of the liquid in equilibrium with its saturated vapor, and γSV is the surface tension of the solid in equilibrium with the saturated vapor of the liquid. The Dupre equation for the work of adhesion Wa defines the work required to reversibly separate the interface between two bulk phases and can be written as Wa = γ L + γ S − γ LS 3.3 The Young-Dupre equation relates the work of adhesion, a thermodynamic parameter, to two easily determined parameters, the contact angle and the liquid-vapor surface tension: Wa = γ LV (1 − cosθ ) 3.4 When the surface energy of the liquid is smaller than that of the solid, θ will be small, and adsorption will occur. Various molecular forces are linearly additive, and the work of adhesion can be separated into two terms, a dispersion component and a polar component: Wa = Wad + Wap 3.5 58 The geometric mean relation is used when the interface is made up of a low-energy and a high-energy material. Then the dispersion component of the work of adhesion is: ( Wad = 2 γ Ld γ Sd ) 1 2 3.6 If dipole-dipole interaction is predominant, then the polar component of the work of adhesion is: ( Wap = 2 γ Lp γ Sp ) 1 2 3.7 It can be seen that if one of the materials is non-polar, then the polar component of adhesion is zero. If the surface energies of the two materials due to polarity are similar, the polar component of adhesion will be maximized (Wu 1982). In the present work, adhesion is found to be high between HIPS and SL 5170 resin (see Chapter 5). The work of adhesion, or the work required to separate these two material surfaces, is high compared to the other material pairs studied. The high work of adhesion may be due to dispersive interactions, polar interactions, or both. 59 3.3 Ejection Force Model Derivation The ejection force model derived in this section is a key component of the present work. It is used for theoretical comparison to experimental measurement, both for the ejection force and the coefficient of static friction, for all the injection mold insert and thermoplastic material combinations used. 3.3.1 Model derivation Ejection force equations are derived from the empirical law of the friction phenomenon, presented above, in which the friction force between two surfaces is proportional to the normal force pressing the two surfaces together: F = µN 3.8 where N is the normal force and µ is the coefficient of friction, a characteristic constant of the materials involved. For deep parts with cores and cavities, the friction force is equal to the release force FR, and the normal force results from the product of the contact pressure P and the area of contact A (see Burke 1991): 60 FR = µPA 3.9 The stresses in an injection molded cylindrical part before ejection can be modeled as stresses in a thin-walled cylindrical pressure vessel as shown in Figure 3.5 (Beer & Johnston 1981, p. 326). The radius of the core is r, and t is the wall thickness. The stresses exerted on a small element of wall will be determined. The sides of the element are respectively parallel and perpendicular to the axis of the cylinder. The vessel and its contents are axisymmetric, so there are no shear stresses on the element, and σ1 and σ2 are principal stresses. The hoop stress is represented by σ1, and the longitudinal stress is represented by σ2. Figure 3.5: Thin-walled cylindrical pressure vessel. For thin-walled pressure vessels, the term t/2r is considered sufficiently small such that the stresses do not vary across the wall, and thus the core radius may be used in the calculation in lieu of the mean radius of the wall section. Also, in this case the 61 longitudinal stress is assumed to be insignificant relative to the hoop stress, so only σ1 will be calculated here. A detached portion of the part and the core, bounded by the xy plane and by two planes parallel to the yz plane and separated by a distance ∆x, is used to determine the hoop stress σ1 (Figure 3.6). The forces in the z direction acting on this free body are the elementary internal forces on the wall sections σ1 dA and the elementary pressure forces acting on the projected area of the core p dA. Figure 3.6: Section of the part and the core with associated stresses. The resultant of the internal forces σ1 dA equals the product of σ1 and the crosssectional area of the wall 2t ∆x. The resultant of the pressure forces p dA equals the product of p and the area 2r ∆x. The sum of the forces in the z direction are: 62 ΣFz = 0 : σ 1 (2t∆x ) − p (2r∆x ) = 0 3.10 Solving for hoop stress σ1: σ1 = pr t 3.11 Rearranging equation 3.11 to solve for the pressure force, i.e., the contact pressure P in this case: P= σt rc 3.12 Next, Hooke’s Law is applied, assuming elasticity in the solidified part. According to equation 3.12, contact pressure P is proportional to tensile (circumferential) stress σ. Stress σ is directly proportional to the elastic modulus and strain Eε: σ = E (T )ε 3.13 where E(T) is the elastic modulus at the ejection temperature, and ε represents engineering strain. The injection molding case involves changing temperatures, thus strain can be represented by thermal strain as follows: 63 ε = α (TM − TE ) 3.14 where α is the coefficient of thermal expansion, TM is the melt temperature, and TE is the temperature at ejection. Combining equations 3.13 and 3.14 gives: σ = E (T )α (TM − TE ) 3.15 As previously mentioned, the term t/2r must be sufficiently small in order to apply the equation for thin-walled pressure vessels. For example, ensuring that t < 2rc is 10 a good rule of thumb (Popov 1976, p 290). Combining equations 3.12 and 3.15 gives: P= E (T )α (TM − TE )t rc 3.16 With the area of the cylinder A = πDc L and equations 3.9 and 3.16: FR = µE (T )α (TM − TE )tπDc L rc 3.17 Menges et al (2001) approximate strain by the relative change in diameter ∆dr of the cylinder immediately after ejection. With this change, then, the ejection force is: 64 FR = µE (T )∆d r tπDc L rc 3.18 3.3.2 Additional Consideration for Strain The description of strain in equation 3.14 may be a simplification considering other transformations within the material. For example, in addition to thermal strain, there may also be crystallization strain for crystalline materials, reaction strain for thermosets, and hydrostatic strain due to the compressibility of the material (Jansen & Titomanlio 1996). Total strain would be the sum of thermal strain, hydrostatic strain, and crystallization and reaction strain as applicable. In this work, reaction strain does not apply, and crystallization strain would only apply to HDPE, since HIPS is amorphous. By using a measure of relative change in diameter in place of thermal strain, as shown in equation 3.18, all aspects of strain are taken into consideration. Thus ∆dr represents total strain, and improves the accuracy of the ejection force model. 65 CHAPTER 4 EXPERIMENTATION This chapter describes the details about how data were collected for this research. For friction testing, this includes a description of the test apparatus and the test matrix and procedure. The process for measurement of thermoplastic modulus at temperature is then presented. For the injection molding experiments, the mold and part design, the injection molding machine and process parameters, and the experimental design and procedure are all explained. The last section on data acquisition summarizes core temperature, ejection force, and part diameter measurements. 4.1 Friction Testing Most published data on coefficients of friction for thermoplastics result from room temperature tests against steel or against like materials. For example, the ASM Handbook lists friction data for polymers on steel as shown in Table 4.1 (ASM International 1992). Actual friction coefficients during the injection molding process are difficult to determine because of the rapidly changing temperature and pressure 66 environment that exists. In order to have reasonable values against which to compare friction values determined from the experimental injection molding data, standard friction testing was conducted using the same mold insert materials and thermoplastics used in the experiments while more closely simulating processing conditions. Fixed Specimen Steel, 52100 Steel, carbon Steel, mild Moving Specimen HDPE HDPE Polystyrene polystyrene Test Static Geometry CoF pin-on-disc --pin-on-flat 0.36 pin-on-flat 0.43 thrust washer 0.28 Kinetic CoF 0.25 0.23 0.37 0.32 Table 4.1: Friction data for polymers on steel. The coefficients of static friction of HDPE and HIPS were measured against P-20 mold steel, LaserForm ST-100, and SL 5170 stereolithography resin following a modified ASTM D 1894, “Standard Test Method for Static and Kinetic Coefficients of Friction of Plastic Film and Sheeting.” A schematic of the friction apparatus is shown in Figure 4.1. Measurements were made first at room temperature; second, at ejection temperature; and third, after the specimen was heated to a higher temperature, pressed against the plate, and cooled to ejection temperature. The purpose of these tests was to compare coefficients of friction among injection mold insert materials, and at elevated 67 temperatures that more closely resemble processing conditions. The polymers tested were identical to the polymers used in the subsequent molding experiments. A. Sled B. Plane C. Supporting Base D. Gage E. Tensile Tester Crosshead F. Braided Wire G. Pulley Ref. ASTM D 1894, figure 1. Figure 4.1: Schematic of friction apparatus. 4.1.1 Friction Test Apparatus The friction tests were conducted in accordance with the modified ASTM D 1894 procedure. The equipment used consists of a Coefficient of Friction Sled Fixture (Material Testing Technology Co.) installed in an Instron model 4507 tester equipped with a Sensotec 25-lb load cell and a Bemco furnace (Figure 4.2). Data were collected at a rate of 100 samples per second. Thermocouples were mounted at two locations on each plate specimen to measure temperature. 68 Figure 4.2: Friction test apparatus: sled on plate specimen (left) inside furnace and tester (right). Modifications to the ASTM D 1894 standard test include elevated temperatures for some tests, a slower pull speed of 25 mm (1 inch) per minute instead of the recommended 150 mm (6 inches) per minute, and a shorter pull distance, i.e., approximately 25 mm (1 inch) vs. the recommended 125 mm (5 inches). The slower pull speed was more appropriate for measuring static coefficient of friction since the higher speed caused a sudden jerk of the sled, and an unreliable measurement of force. The 69 shorter pull distance was used because, for determining static friction, only the force necessary to set the sled in motion is required. 4.1.2 Test Matrix and Procedure The friction test matrix is shown in Table 4.2. Two thermoplastics and three mold insert materials were tested under three temperature conditions. The thermoplastic specimens were 63.5 mm (2.5 inches) square and attached to the sled using double-sided, high temperature fiberglass tape. The mold insert material specimens were approximately 125 mm (5 inches) wide by 250 mm (10 inches) long and were positioned on the base plate of the apparatus. The surface roughnesses of the plate specimens were 0.7 microns (28 microinches) for P-20, 0.2 microns (8 microinches) for ST-100, and 3.6 microns (142 microinches) for SL 5170. The first temperature condition was room temperature, 22 oC (71 oF). The second condition was ejection temperature, i.e., the temperature at which molded parts are ejected from the injection molding machine. The second temperature condition simulated the environment in which friction is encountered in the injection molding process, i.e., 50 o C (120 oF) for P-20 and ST-100 plate materials and 55 oC (130 oF) for SL 5170 plate material. The third condition consisted of first heating to an elevated temperature (120 o C (250 oF) for HDPE specimens and 150 oC (300 oF) for HIPS specimens), then cooling down to the ejection temperatures mentioned above before testing. While at the elevated temperature, a 0.9 kg (2-lb) mass was placed on the sled. When the plate cooled to ejection temperature, the weight was removed and the specimen tested. The purpose of 70 the third temperature condition was to imprint the surface of the plate specimen onto the sled specimen. This simulated the environment as well as the surface condition that occurs during ejection of a molded part. Elevated Temp Ejection Temp Ambient Plate Specimen P-20 Steel P-20 Steel ST-100 (Sintered) ST-100 (Sintered) SL 5170 (Resin) SL 5170 (Resin) P-20 Steel P-20 Steel ST-100 (Sintered) ST-100 (Sintered) SL 5170 (Resin) SL 5170 (Resin) P-20 Steel P-20 Steel ST-100 (Sintered) ST-100 (Sintered) SL 5170 (Resin) SL 5170 (Resin) Sled Specimen Temporary Load HDPE HIPS HDPE HIPS HDPE HIPS HDPE HIPS HDPE HIPS HDPE HIPS HDPE HIPS HDPE HIPS HDPE HIPS None None None None None None None None None None None None 0.9 kg (2 lbs) 0.9 kg (2 lbs) 0.9 kg (2 lbs) 0.9 kg (2 lbs) 0.9 kg (2 lbs) 0.9 kg (2 lbs) Initial Plate Temp o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o 50 C (120 F) o o 50 C (120 F) o o 50 C (120 F) o o 50 C (120 F) o o 55 C (130 F) o o 55 C (130 F) o o 120 C (250 F) o o 150 C (300 F) o o 120 C (250 F) o o 150 C (300 F) o o 120 C (250 F) o o 150 C (300 F) Soak Time N/A N/A N/A N/A N/A N/A 1 min 1 min 1 min 1 min 2 min 2 min 1 min 1 min 1 min 1 min 2 min 2 min Plate Temp at Pull o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o RT 22 C (71 F) o o 50 C (120 F) o o 50 C (120 F) o o 50 C (120 F) o o 50 C (120 F) o o 55 C (130 F) o o 55 C (130 F) o o 50 C (120 F) o o 50 C (120 F) o o 50 C (120 F) o o 50 C (120 F) o o 55 C (130 F) o o 55 C (130 F) Table 4.2: Friction test matrix. The test procedure began with insertion of the proper materials. The plate specimen was cleaned with acetone prior to each set of tests. For room temperature tests, the sled was pulled until it moved along the plate. In the ejection temperature tests, the furnace was ramped until the plate reached the specified initial temperature. After the apparatus was soaked for the specified amount of time, the sled was pulled until it moved 71 along the plate. In the elevated temperature tests, a weight was placed evenly on top of the sled, and the furnace was ramped until the plate reached the specified initial temperature. After the specified soak time, the plate was cooled to the specified temperature at pull (ejection temperature). The weight was then removed, and the sled was pulled until it moved along the plate. In all cases the pulling force and time were recorded. Each set of test conditions was repeated five times, using a new sled specimen for each test. 4.2 Measurement of Elastic Modulus The equation for ejection force requires values for the elastic modulus of the molding material at the ejection temperature. Elastic moduli for the HDPE and HIPS materials used in this research were measured at various temperatures using ASTM D 638 “Standard Test Method for Tensile Properties of Plastics” as a guide. The testing apparatus was an Instron model 1322 tensile tester with a tube furnace. An extensometer with a 2-inch gauge and 50 percent strain was used to measure elongation (Figure 4.3). 72 Figure 4.3: Tensile testing apparatus with tube furnace. ASTM Type I (dogbone) specimens were molded from each thermoplastic material and were tested at room temperature, 30oC, and at ten degree increments until no elastic region was detected (Figure 4.4). HDPE was tested through 70oC, and HIPS was tested through 50oC. At least three samples of HIPS were tested at each temperature, while at least 5 samples of HDPE were tested at each temperature. Data from the tensile tests are included in Appendix A, and a graph of the results is shown in Figure 4.5. Modulus values for HIPS at higher ejection temperatures than 50oC were determined by extrapolating the graph. The lookup table generated from this graph and used in the calculations of friction coefficient can also be found in Appendix A. 73 Figure 4.4: HIPS specimens after tensile tests. Modulus (MPa) 4000 3500 HIPS PS 495F 3000 HDPE ME 9180 2500 2000 1500 1000 500 0 20 30 40 50 60 70 80 Temperature (°C) Figure 4.5: Elastic modulus at various temperatures for HDPE and HIPS. 74 4.3 Injection Molding 4.3.1 Mold Design and Materials Since this research involved the study of different rapid tooling materials, a modular mold design was employed. The mold base was a steel Master Unit Die (MUD) having a core and cavity that could be removed and replaced with those of other materials (Figure 4.6). The baseline core and cavity were made of P-20 steel, a typical mold steel, and the two rapid tooled core and cavity sets were made of SL 5170 epoxy resin and LaserForm ST-100 material (Figure 4.7). The SL 5170 insert was built at NASA Marshall Space Flight Center using the stereolithography process, and the ST-100 insert was built at General Pattern Company (Blaine, MN) using the laser sintering process, as described in Chapter 2. Machining allowances were included in the design of each rapid tooled insert so that it could be machined to fit properly into the mold base. The baseline steel insert core had a surface finish of Ra = 0.7 microns (28 microinches). The stereolithography insert also had a surface finish of Ra = 0.7 microns (28 microinches), and the laser sintered insert had a surface finish of Ra = 0.3 microns (12 microinches). 75 Figure 4.6: Sprue side of the MUD base mounted in the injection molding machine with SL 5170 cavity insert. Figure 4.7: Core and cavity inserts, before final machining, made of SL 5170 (left), P-20 steel (center), and LaserForm ST-100 (right). 76 The experimental part was a closed-end, straight cylinder with a 32 mm (1.26 in) outside diameter, 49.6 mm (1.95 in) height, and 1.2 mm (0.05 in) wall thickness (Figure 4.8). Design drawings for the canister and its injection mold are included in Appendix B. The canister was designed with four vent holes in the base to prevent vacuum forces that would result during ejection from the core. The part was similar in size and shape to the canisters used to store 35 mm photographic film. This particular part was selected because it required a simple core and cavity that eliminated the effect of corners, and its non-tapered design allowed for a significant and measurable ejection force. The dimensions of the SL 5170 inserts were modified to alleviate problems with core swelling and parts sticking in the cavity. The core diameter was reduced by 0.1 mm, (0.005 in) to prevent interference between the stripper plate and the core when the core swelled with temperature. All parts from the SL 5170 core had this slight increase in diameter. The cavity wall was tapered 0.42o, leaving the canister base dimension intact, in an attempt to prevent parts from sticking in the cavity. Only the parts from the experiments using both the SL 5170 core and cavity inserts were thicker at the rim, i.e., had an outside diameter of 32.8 mm (1.29 in), and were slightly tapered on the outside of the wall. 77 Figure 4.8: Canister part with vent holes and no taper. The mold was a single-cavity design with a heated sprue connecting directly to the base of the canister. The hot sprue allowed more control over packing pressure, i.e., the packing material did not prematurely freeze at the gate. The core and cavity each had a housing that fixed it to the MUD base. This ensured that there were no bolt holes through the core or cavity inserts. The cavity insert had a square profile large enough to provide for the possibility of adding cooling channels in the future. The core insert, on the other hand, had a round profile so that material requirements could be reduced and the machining process simplified. The core insert also had one flat surface for orientation. If necessary, cooling channels could be added inside the core. Three thermocouples were positioned at different depths inside each core insert. The ejection system employed a stripper plate with a circular hole that fit around the base of the core. The stripper plate was supported by four ejector pins that connected to the mold ejector plate. Subminiature load cells for measuring ejection force were 78 positioned between each ejector pin and the mold ejector plate. Drawings of the mold insert design are included in Appendix B. 4.3.2 The Injection Molding Process A Sumitomo Injection Molding Machine, model SH50M, was used for the experimental portion of this work (Figure 4.9). It was a horizontal press with a fully hydraulic, 50-ton clamping system. Machine specifications are given in Table 4.3. Figure 4.9: Sumitomo SH50M injection molding machine. 79 Model Clamping System Clamp Force Distance Between Tie Bars Overall Size of Platen Opening Stroke Ejector Type Ejector Stroke Ejector Force Screw Diameter Injection Capacity Injection Rate Nozzle Contact Force Machine Weight SH50M Fully Hydraulic 50 metric tons (55.1 short tons) 325 x 325 mm (12.8 x 12.8 in) 470 x 467 mm (18.5 x 18.4 in) 440 mm (17.3 in) Hydraulic, cross multipoint ejection (5pts) 70 mm (2.8 in) 2.2 metric tons (2.42 short tons) 28 mm (1.1 in) 3 3 70 cm (4.3 in ) 3 3 99 cm /s (6.0 in /s) 4670 kgf (10297 lbf) 2.2 metric tons (2.42 short tons) Table 4.3: Injection molding machine specifications. The procedure for defining injection molding process parameters was intended first to establish the volume of material required, and then to determine the velocity required to completely fill the mold with no flashing. Barrel zone temperatures were set based on commonly used temperatures for injection molding the given thermoplastic material. For the steel and sintered inserts, with velocity at 50 percent of maximum and packing pressure at zero, the screw position was initially set for a short shot, and then gradually extended until the part filled. Velocity was increased if the part froze before the entire shot could be injected, and decreased if flashing occurred. All experiments, except those using the SL 5170 insert, were run with one stage packing pressure, 25% maximum screw rpm, 5% maximum back pressure, and 15% maximum ejection velocity. 80 The velocity and temperature parameters for each set of experiments are shown in Table 4.4. Two levels of packing time were defined at 2 and 6 seconds. Three levels of cooling time were defined at 5, 10, and 15 seconds. Packing pressure levels were defined at 0, 5, and 10 percent of maximum (0, 10.93, and 21.87 MPa). Clamping force was 20 metric tons. Since the SL 5170 insert was expected to be less durable, temperature and velocity settings were reduced as far as possible such that the mold would still fill. As much as possible, the number of test runs on this insert were minimized. Screw rpm, back pressure, clamp force, and ejection velocity were the same as above. Packing time levels remained the same (2 and 6 seconds), while the number of cooling time and packing pressure levels were reduced from three to two. Cooling times were greatly increased to 120 and 150 seconds to allow for the low thermal conductivity of the stereolithography resin. Packing pressure levels were 0 and 5 percent (0 and 10.93 MPa). 81 HDPE with P-20 Steel and LaserForm ST-100 Inserts Velocity 35% or 56 mm/s (2.2 in/s) Temperature Profile: Sprue o 210 C Nozzle Front o o 210 C 199 C Middle o 193 C Rear o 177 C HIPS with P-20 Steel and LaserForm ST-100 Inserts Velocity 40% or 64 mm/s (2.5 in/s) for P-20, 35% or 56 mm/s (2.2 in/s) for ST-100 Temperature Profile: Sprue o 221 C Nozzle Front o o 221 C Middle o Rear o 213 C 204 C 191 C Front Middle Rear HDPE with SL 5170 Insert, SL or P-20 Cavity Velocity 25% or 40 mm/s (1.6 in/s) Temperature Profile: Sprue o 177 C Nozzle o o 177 C o o 171 C 166 C 160 C Front Middle Rear HIPS with SL 5170 Insert, SL or P-20 Cavity Velocity 40% or 64 mm/s (2.5 in/s) Temperature Profile: Sprue o 210 C Nozzle o o 216 C 202 C o 193 C o 182 C Table 4.4: Injection Molding Parameters The thermoplastic materials used in the experiments, HDPE and HIPS, are described in Chapter 3. Prior to molding, both materials were dried for two hours in a desiccant dryer, with dew point at -40 oC (-40 oF) and air temperature at 71 oC (160 oF). Materials data for HDPE (Lutene-H ME9180) are shown in Table 4.5, and for HIPS (BASF PS 495F) are shown in Table 4.6. 82 Melt Flow Index Density Tensile Strength @ Yield Tensile Strength @ Break Flexural Modulus Vicat Softening Temperature 18.0 g/10 min 3 0.958 g/cm 2 290 kg/cm (4125 psi) <1000% 2 10,000 kg/cm (142 kpsi) o o 123 C (253 F) ASTM D 1238 ASTM D 1505 ASTM D 638 ASTM D 638 ASTM D 790 ASTM D 1525 Table 4.5: Typical data for Lutene-H ME9180. Melt Flow Index Impact Strength, Izod Tensile Strength @ Yield Tensile Elongation @ Break Flexural Modulus Vicat Softening Temperature 7 g/10 min 112 J/m (2.1 ft-lb/in) 20 MPa (2900 psi) 55% 1655 MPa (240 kpsi) o o 101 C (214 F) Table 4.6: Typical data for BASF PS 495F. 83 ASTM D 1238 ASTM D 256 ASTM D 638 ASTM D 638 ASTM D 790 ASTM D 1525 4.3.3 Design of Experiments Statistical design of experiments has been used to design the six sets of experiments run in this work. Each set was blocked by insert material and thermoplastic material, and then randomized by packing time, cooling time and packing pressure. In the first four experimental sets, there were two levels of packing time, three levels of cooling time, and three levels of packing pressure as described above. Each combination of factors was repeated eight times. The experimental design with process parameters is shown in Table 4.7. The last two experimental sets, using the SL 5170 insert, were designed to be smaller than the other sets due to the expected low durability of the insert material. In this case there were two levels each of packing time, cooling time, and packing pressure as described above. Each combination of factors was repeated five times. A much longer cooling time was required for this insert to accommodate its low thermal conductivity. It was initially intended that cooling time levels be set at 150 and 180 seconds. However these levels were reduced to 120 and 150 seconds to reduce the shrinkage on the core, and thus ejection force requirements. Also note that only a limited number of parts could be processed using the SL 5170 cavities due to deformation that caused sticking of parts (see Chapter 6). The designed experiment, therefore, was carried out using the SL 5170 core with the P-20 cavity. While the use of steel cavity material greatly changed the thermal performance of this insert and reduced the temperature at ejection, it also allowed a complete experiment to be performed in which ejection force from the SL 5170 core could be measured. 84 Design of experiments (DOE) analyses in Minitab® have been performed using packing time, cooling time and packing pressure as factors and ejection force as the response. Using analysis of variance (ANOVA), the effects of variables and their interactions on each response were determined. ANOVA tables for each data set are included in Appendix A. DOE results and graphs identifying main effects and interactions are included in Chapter 5. 85 SET 1 P-20 HDPE 8 Reps Run 1 Tp = 2 s Tc = 15 s Pp = 0% Run 10 Tp = 6 s Tc = 15 s Pp = 10% Run 2 Tp = 2 s Tc = 5 s Pp = 10% Run 11 Tp = 6 s Tc = 5 s Pp = 0% Run 3 Tp = 2 s Tc = 15 s Pp = 10% Run 12 Tp = 6 s Tc = 5 s Pp = 5% Run 4 Tp = 2 s Tc = 15 s Pp = 5% Run 13 Tp = 6 s Tc = 15 s Pp = 5% Run 5 Tp = 2 s Tc = 10 s Pp = 10% Run 14 Tp = 6 s Tc = 10 s Pp = 0% Run 6 Tp = 2 s Tc = 5 s Pp = 5% Run 15 Tp = 6 s Tc = 10 s Pp = 5% Run 7 Tp = 2 s Tc = 5 s Pp = 0% Run 16 Tp = 6 s Tc = 15 s Pp = 0% Run 8 Tp = 2 s Tc = 10 s Pp = 0% Run 17 Tp = 6 s Tc = 5 s Pp = 10% Run 9 Tp = 2 s Tc = 10 s Pp = 5% Run 18 Tp = 6 s Tc = 10 s Pp = 10% SET 2 P-20 HIPS 8 Reps Run 1 Tp = 2 s Tc = 15 s Pp = 0% Run 10 Tp = 6 s Tc = 15 s Pp = 10% Run 2 Tp = 2 s Tc = 5 s Pp = 10% Run 11 Tp = 6 s Tc = 5 s Pp = 0% Run 3 Tp = 2 s Tc = 15 s Pp = 10% Run 12 Tp = 6 s Tc = 5 s Pp = 5% Run 4 Tp = 2 s Tc = 15 s Pp = 5% Run 13 Tp = 6 s Tc = 15 s Pp = 5% Run 5 Tp = 2 s Tc = 10 s Pp = 10% Run 14 Tp = 6 s Tc = 10 s Pp = 0% Run 6 Tp = 2 s Tc = 5 s Pp = 5% Run 15 Tp = 6 s Tc = 10 s Pp = 5% Run 7 Tp = 2 s Tc = 5 s Pp = 0% Run 16 Tp = 6 s Tc = 15 s Pp = 0% Run 8 Tp = 2 s Tc = 10 s Pp = 0% Run 17 Tp = 6 s Tc = 5 s Pp = 10% Run 9 Tp = 2 s Tc = 10 s Pp = 5% Run 18 Tp = 6 s Tc = 10 s Pp = 10% SET 3 ST-100 HDPE 8 Reps Run 1 Tp = 2 s Tc = 15 s Pp = 0% Run 10 Tp = 6 s Tc = 15 s Pp = 10% Run 2 Tp = 2 s Tc = 5 s Pp = 10% Run 11 Tp = 6 s Tc = 5 s Pp = 0% Run 3 Tp = 2 s Tc = 15 s Pp = 10% Run 12 Tp = 6 s Tc = 5 s Pp = 5% Run 4 Tp = 2 s Tc = 15 s Pp = 5% Run 13 Tp = 6 s Tc = 15 s Pp = 5% Run 5 Tp = 2 s Tc = 10 s Pp = 10% Run 14 Tp = 6 s Tc = 10 s Pp = 0% Run 6 Tp = 2 s Tc = 5 s Pp = 5% Run 15 Tp = 6 s Tc = 10 s Pp = 5% Run 7 Tp = 2 s Tc = 5 s Pp = 0% Run 16 Tp = 6 s Tc = 15 s Pp = 0% Run 8 Tp = 2 s Tc = 10 s Pp = 0% Run 17 Tp = 6 s Tc = 5 s Pp = 10% Run 9 Tp = 2 s Tc = 10 s Pp = 5% Run 18 Tp = 6 s Tc = 10 s Pp = 10% SET 4 ST-100 HIPS 8 Reps Run 1 Tp = 2 s Tc = 15 s Pp = 0% Run 10 Tp = 6 s Tc = 15 s Pp = 10% Run 2 Tp = 2 s Tc = 5 s Pp = 10% Run 11 Tp = 6 s Tc = 5 s Pp = 0% Run 3 Tp = 2 s Tc = 15 s Pp = 10% Run 12 Tp = 6 s Tc = 5 s Pp = 5% Run 4 Tp = 2 s Tc = 15 s Pp = 5% Run 13 Tp = 6 s Tc = 15 s Pp = 5% Run 5 Tp = 2 s Tc = 10 s Pp = 10% Run 14 Tp = 6 s Tc = 10 s Pp = 0% Run 6 Tp = 2 s Tc = 5 s Pp = 5% Run 15 Tp = 6 s Tc = 10 s Pp = 5% Run 7 Tp = 2 s Tc = 5 s Pp = 0% Run 16 Tp = 6 s Tc = 15 s Pp = 0% Run 8 Tp = 2 s Tc = 10 s Pp = 0% Run 17 Tp = 6 s Tc = 5 s Pp = 10% Run 9 Tp = 2 s Tc = 10 s Pp = 5% Run 18 Tp = 6 s Tc = 10 s Pp = 10% SET 5 Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 SL 5170 Tp = 2 s Tp = 2s Tp = 6 s Tp = 6 s Tp = 2s Tp = 2s Tp = 6 s Tp = 6 s HDPE Tc = 150s Tc = 120s Tc = 150s Tc = 120s Tc = 150s Tc = 120s Tc = 150s Tc = 120s 5 Reps Pp = 0% Pp = 0% Pp = 0% Pp = 0% Pp = 5% Pp = 5% Pp = 5% Pp = 5% SET 6 Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Run 8 SL 5170 Tp = 2 s Tp = 2s Tp = 6 s Tp = 6 s Tp = 2s Tp = 2s Tp = 6 s Tp = 6 s HIPS Tc = 150s Tc = 120s Tc = 150s Tc = 120s Tc = 150s Tc = 120s Tc = 150s Tc = 120s 5 Reps Pp = 0% Pp = 0% Pp = 0% Pp = 0% Pp = 5% Pp = 5% Pp = 5% Pp = 5% Table 4.7: Experimental design for six sets, including process parameters. 86 4.3.4 Experimental Procedure An experimental set was defined by the mold insert and thermoplastic material. The procedure followed for each experimental set is as follows: 1) Begin Set 2) Setup hardware and load thermoplastic material 3) Load parameter settings 4) Begin Run i) Adjust DOE settings for packing time, cooling time and packing pressure ii) Process test parts to bring insert to temperature iii) Begin Iteration (a) Inject part (b) Capture ejection force and temperature during ejection (c) Digitally photograph part immediately after ejection (d) Wait for core to return to desired temperature iv) Repeat iteration 8 times (5 times with SL 5170 insert) 5) Repeat run for all DOE combinations 6) Go to next set Ejection force data were used to determine the initial force required to release each part from the core. Thermal data were used to determine the elastic modulus of each thermoplastic at ejection. Inside and outside diameters of the canister parts, 87 measured from digital photographs, were used to determine part thickness. All of these data were used with the Menges ejection force model to calculate the ejection force and the apparent coefficient of static friction. Statistical analysis of variance, as described above, was conducted using packing time, cooling time, and packing pressure as input factors, and ejection force as the response. Experimental results and conclusions are included in Chapters 5 and 6. Experimental data are included in Appendix A. 4.4 Set-up and Data Acquisition The data required from the injection molding experiments included thermal (core temperature at ejection), load (ejection force), and dimensional (part diameter and thickness at ejection) measurements. Type J thermocouples were used to measure temperature, subminiature load cells were used to measure force, and digital imaging was used to measure diameter. The three thermal and four load sensors collected data through a National Instruments SC-2311 signal conditioner with associated thermocouple and strain gage input modules. A LabVIEW™ program was used to read the ejection force and temperature data and write them to a Microsoft Excel® spreadsheet. Pictures of each part were taken with an Olympus Camedia C-740 digital camera and processed using Adobe PhotoShop software. The inside and outside diameters of each part were measured in pixels and converted to inches and millimeters by reference to a scale, a picture of which was taken during each set of experiments. The data acquisition equipment is shown in Figure 4.10. 88 Figure 4.10: Signal conditioner and computer with front panel for data acquisition (left), and core side of mold with thermocouple and load cell sensor wires (right). 4.4.1 Temperature Measurement and Thermal Model Three NANMAC Type J thermocouples were used to measure temperature in the core at ejection. The thermocouples were installed by friction-fitting them into holes drilled at three depths into the core (Figure 4.11). Representative thermal traces of the injection molding cycle are shown in Figure 4.12. 89 Figure 4.11: Thermocouple placement within core insert. 90 Core Temperature (C) vs. Time End of Core Mid Core Base of Core 55 50 45 Cooling Time Injection Mold Open 40 35 Core Temperature (C) vs. Time End of Core Mid Core Base of Core 55 50 45 Cooling Time Mold Open 40 Injection 35 Figure 4.12: Representative thermal traces of the injection molding cycle. 91 Since the distance between each thermocouple and the surface of the core was more than a millimeter, there was a time lag from heating of the surface to heating of the thermocouple. A thermal analysis was run to determine the convergence time between the thermocouple reading and the actual temperature at the surface of the core (Carpenter 2004). The analysis simulated injection of the thermoplastic into the mold insert for each material combination. The simulation was run using ABAQUS, and the results show the time required for temperature at the thermocouple to match the temperature at the surface of the core (Table 4.8). Initial and boundary conditions for the thermal analysis are shown in Table 4.9, and graphs of the results are shown in Figure 4.13. The table of temperature values is included in Appendix A. The graphs show that the P-20 and ST100 thermocouple readings and surface temperatures converge within 5 seconds for HDPE, and within 10 seconds for HIPS. Convergence in the SL 5170 insert requires 110 seconds for HDPE and 120 seconds for HIPS. All processing times allow for these convergence times so that thermocouple readings are accurate. It can be seen from the simulation of the SL 5170 core with the P-20 cavity that convergence times are approximately 10 seconds for HDPE and 20 seconds for HIPS. The longer cooling times, therefore, are not necessary. In the experiments with this combination insert, however, the longer, more conservative cooling times were used. 92 Insert Material P-20 Steel P-20 Steel ST-100 ST-100 SL 5170 SL 5170 SL 5170/P-20 SL 5170/P-20 Thermoplastic Material HDPE HIPS HDPE HIPS HDPE HIPS HDPE HIPS Convergence Time (s) 4.3 9.3 4.3 9.3 120.4 121.1 10.4 14.9 Table 4.8: Resulting convergence times from the thermal simulation. P-20 Density (kg/m3) Specific Heat (J/kg.C) Thermal Conductivity (W/m.C) Mold Temperature Polymer Injection Temperature Material Properties ST-100 SL 5170 HDPE HIPS 7,870 7,700 1,220 958 1,050 486 475.2 1,674 2,200 2,000 47.6 49 0.2 0.39 0.16 P-20 and HDPE P-20 and HIPS 50 °C 50 °C 50 °C 50 °C 30 °C 33 °C 30 °C 33 °C 210 °C 221 °C 210 °C 221 °C 177 °C 210 °C 177 °C 210 °C Initial Temperature Condition SL 5170 SL/P-20 SL/P-20 ST-100 and ST-100 and SL 5170 HIPS and HDPE and HIPS and HDPE and HIPS HDPE Table 4.9: Input conditions for the thermal analysis. 93 P-20 and HDPE Thermocouple Reading Core Surface Temperature 250 Temperature (C) 200 150 100 50 0 0 5 10 15 20 25 30 Time (s) P-20 and HIPS Thermocouple Reading Core Surface Temperature 250 Temperature (C) 200 150 100 50 0 0 5 10 15 20 25 30 Time (s) Figure 4.13: Graphs of the thermal analysis results for each material combination. (continued) 94 Figure 4.13 (continued.) ST-100 and HDPE Thermocouple Reading Core Surface Temperature 250 Temperature (C) 200 150 100 50 0 0 5 10 15 20 25 30 Time (s) ST-100 and HIPS Thermocouple Reading Core Surface Temperature 250 Temperature (C) 200 150 100 50 0 0 5 10 15 20 25 30 Time (s) (continued) 95 Figure 4.13 (continued.) SL 5170 and HDPE Thermocouple Reading Core Surface Temperature 200 180 Temperature (C) 160 140 120 100 80 60 40 20 0 0 20 40 60 80 100 120 140 160 180 Time (s) SL 5170 and HIPS Thermocouple Reading Core Surface Temperature 250 Temperature (C) 200 150 100 50 0 0 50 100 150 200 Time (s) (continued) 96 Figure 4.13 (continued.) SL Core P-20 Cavity and HDPE Thermocouple Reading Core Surface Temperature 200 180 Temperature (C) 160 140 120 100 80 60 40 20 0 0 10 20 30 40 50 60 Time (s) SL Core P-20 Cavity and HIPS Thermocouple Reading Core Surface Temperature 250 Temperature (C) 200 150 100 50 0 0 10 20 30 Time (s) 97 40 50 60 4.4.2 Ejection Force Measurement Four Sensotec subminiature load cells, rated at 100 lbs each, were used to measure ejection force. Each load cell was installed between one of four ejector pins and the ejector plate. The four pins, mounted on bearings to reduce friction, were attached to the stripper plate, which removed the canister from the core. Total load was measured during ejection of each part, and the ejection force was determined from the initial peak load required to release the part from the insert core. Sample ejection force traces are shown in Figure 4.14. After each experimental run, the ejection force was measured without a part on the core. This no-load ejection force, i.e., the force required to simply move the ejection mechanism, was subtracted from the peak load measurement of every canister to determine the actual force required to release the part. 98 Total Load (lbs) vs. Time (0.1s) 60 Part Release 50 40 Part Sliding off Core 30 20 10 0 Total Load (lbs) vs. Time (0.1 sec) 100 Part Release 90 80 70 Part Sliding off Core 60 50 40 30 20 10 0 Figure 4.14: Representative ejection force traces 99 4.4.3 Diameter and Thickness Measurement The inside and outside diameters of each canister were measured immediately after ejection to determine the relative change in diameter due to shrinkage and the canister wall thickness. The relative change in diameter and the thickness measurements are required in order to use the Menges equation (see Chapter 3). A digital picture of each canister was taken immediately after ejection (Figure 4.15). A picture including a scale was taken for each set of experiments. Using Adobe Photoshop® software, the pictures were magnified and the inside and outside diameters of each canister were measured in four places: vertically, horizontally, and at 45-degree angles. The four measurements were averaged for each diameter value. The scale reference picture was also magnified and measured to determine the number of pixels per inch. The resolution for each set of measurements varied between 0.0009 and 0.0010 inches per pixel. 100 Figure 4.15: Digital pictures of HDPE canisters for measuring inside and outside diameter. 4.4.4 Calculation of Static Friction Coefficient Ejection forces were calculated using equations 3.18 and 3.25, as derived in Chapter 3, and the data described above. The apparent coefficients of static friction were calculated using equation 3.18. Spreadsheets for these data are included in Appendix A, and results are given in Chapter 5. 101 CHAPTER 5 RESULTS AND ANALYSIS This chapter summarizes results from the injection molding experiments and standard friction tests. First, the data from each are discussed individually. Next, experimental ejection forces are compared with those calculated from the ejection force model, using friction coefficients from the standard tests. Then, calculations of the coefficient of static friction, using data from the injection molding experiments, are presented and compared to standard test results. Analysis of variance results from the designed experiment are also included. The last section of the chapter presents a qualitative analysis of the rapid tooled injection mold inserts, i.e., some observations of how these tools performed. 102 5.1 Injection Molding Experiments 5.1.1 Experimental Results and Discussion Ejection forces from the injection molding experiments for HDPE and HIPS are shown in Table 5.1. Measured ejection force results are listed by levels of packing time, cooling time, and packing pressure. Other experimental data, including diameter and temperature measurements, are included in Appendix A. The following discussion comments on the ejection force results according to thermoplastic material and mold insert material. 103 Packing Time s 2 2 2 2 2 2 2 2 2 6 6 6 6 6 6 6 6 6 Cooling Time s 5 5 5 10 10 10 15 15 15 5 5 5 10 10 10 15 15 15 Packing Pressure % 0 5 10 0 5 10 0 5 10 0 5 10 0 5 10 0 5 10 Packing Time s 2 2 2 2 Cooling Time s 120 150 180 150 Packing Pressure % 0 0 0 5 2 2 2 2 6 6 6 6 120 120 150 150 120 120 150 150 0 5 0 5 0 5 0 5 Experimental Ejection Force HDPE HIPS P-20 LaserForm P-20 LaserForm Steel ST-100 Steel ST-100 N N N N 177.15 182.30 343.97 366.29 183.29 190.02 376.37 389.51 186.52 209.79 401.48 375.28 176.51 177.44 346.11 375.71 185.89 196.61 385.56 393.54 172.87 194.33 408.26 394.47 191.90 196.15 384.60 366.31 174.68 201.68 381.94 393.56 173.08 208.75 403.70 398.83 184.69 173.95 376.97 363.86 193.58 185.81 395.23 378.14 173.88 184.41 393.38 388.07 171.28 172.12 369.19 369.54 174.63 178.83 390.81 360.58 175.19 180.87 391.73 374.38 170.00 170.97 351.62 340.67 180.05 186.82 394.77 370.57 185.53 184.13 424.46 399.88 HDPE SL 5170 SL 5170 w/P-20 N N 239.06 193.21 HIPS SL 5170 SL 5170 w/P-20 N N 1334.27 1136.12 1512.25 274.21 299.65 258.76 278.18 313.33 297.38 321.09 317.93 695.76 826.28 610.12 845.02 770.24 939.13 702.28 892.15 Table 5.1: Experimental ejection force results for HDPE and HIPS according to packing time, cooling time, and packing pressure parameters. 104 5.1.2 HDPE Experimental Ejection Force Results Figure 5.1 shows experimental ejection force results for HDPE. Ejection forces for HDPE from the P-20 core, averaged per run, are generally lower than from the ST100 core, which are lower than from the SL 5170 core. Ejection forces from the ST-100 core at low level packing time are higher than at high level packing time. The parts with higher ejection force also had lower shrinkage values, as measured by the relative change in diameter immediately after ejection. Ejection forces for HDPE from the SL 5170 core with the P-20 cavity are higher than from the SL 5170 insert because the P-20 cavity draws away much of the heat, resulting in more shrinkage of the HDPE against the SL 5170 core. 105 Ejection Force HDPE 330 310 290 Ejection Force (N) 270 250 230 210 190 170 150 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Run P-20 ST-100 SL 5170 SL/P-20 Figure 5.1: Experimental ejection force results for HDPE, all runs. 106 16 17 18 5.1.3 HIPS Experimental Ejection Force Results Figure 5.2 shows experimental ejection force results for HIPS. In contrast to HDPE, ejection forces from the P-20 core, averaged per run, are greater than from the ST-100 core for 12 out of 18 runs. Ejection forces from the SL 5170 core are much higher than from the other two. Also in contrast to HDPE, ejection forces for HIPS from the SL 5170 insert are larger than those from the combination SL 5170/P-20 insert. This is because, as will be seen in the standard friction tests, HIPS and SL 5170 react more strongly with each other at higher temperatures. 107 Ejection Force HIPS 1600 1400 Ejection Force (N) 1200 1000 800 600 400 200 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Run P-20 ST-100 SL 5170 SL/P-20 Figure 5.2: Experimental ejection force values for HIPS, all runs. 108 15 16 17 18 5.1.4 Experimental Ejection Force Results from the P-20 and ST-100 Inserts From both the P-20 and ST-100 inserts, ejection forces for HDPE were lower than for HIPS for all experimental runs (Figure 5.3). 5.1.5 Experimental Ejection Force Results from the SL 5170 and SL 5170/P-20 Inserts Due to the problems with cavity deformation and parts sticking in the core, only two HDPE runs and three HIPS runs were completed with the SL 5170 core and cavity. With this insert, ejection forces for HIPS were much higher than for HDPE (Figure 5.4). The same result can be seen from the SL 5170 insert with the P-20 cavity. 109 Ejection Force P-20 450 Ejection Force (N) 400 350 300 250 200 150 100 50 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run HDPE HIPS Ejection Force (N) Ejection Force ST-100 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run HDPE HIPS Figure 5.3: Experimental ejection force results from the P-20 and ST-100 inserts. 110 Ejection Force SL 5170 1600 1500 1400 1300 1200 Ejection Force (N) 1100 1000 900 800 700 600 500 400 300 200 100 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run HDPE HIPS HDPE with SL/P-20 HIPS with SL/P-20 Figure 5.4: Experimental ejection force results from the SL 5170 insert and the combination SL 5170/P-20 insert (all completed runs are shown). 111 5.2 Statistical Analysis 5.2.1 DOE Results Analysis of variance results for the designed experiment described in Chapter 4 are shown in Table 5.2. Numbers are given for the slope of the line that defines the effect (in Newtons per unit) and the correlation coefficient. The slope of the line indicates the magnitude of the effect. The correlation coefficient applies only for three-level parameters and indicates how well the line (defined by the slope) fits the data. If the correlation coefficient is low, then the effect is not linear. A check mark indicates that there was an interaction between the parameters shown. 5.2.2 Main Effects and Interactions As shown in Table 5.2, ejection force increased with packing time for the SL 5170 insert, and decreased with packing time for the ST-100 insert. Packing time also had an effect from the P-20 insert on the HIPS ejection force, but not on the HDPE ejection force. Ejection force decreased with cooling time for the P-20 insert. Note that, for HDPE, the correlation coefficient was low, indicating that the effect was not linear. Packing pressure had an effect on ejection force for HIPS with the SL 5170 and P-20 inserts, but did not for HDPE. There were no interaction effects for the SL 5170 insert. The effects on ejection force from these parameters differed between the baseline steel insert and the rapid tooled inserts. Furthermore, there were differences in effects 112 between the two thermoplastics, especially with the P-20 insert. Main effects and interaction plots for each set of experiments are shown in Figures 5.5 through 5.10. HDPE Insert Material Main Effects on Ejection Force Packing Packing Cooling Pressure Time Tp Time Tc Pp P-20 Steel Sintered ST-100 -3.86 SL 5170/ P-20 8.68 HIPS Insert Material P-20 Steel -.40 .31 -1.10 1.00 Sintered ST-100 -2.99 SL 5170/ P-20 20.46 -2.90 .96 Tp-Tc Tp-Pp Tc-Pp Tp-Tc-Pp a a Main Effects on Ejection Force Packing Packing Cooling Pressure Time Tp Time Tc Pp 1.82 Interactions .88 .82 a Interactions Tp-Tc Tp-Pp Tc-Pp Tp-Tc-Pp a a a a a a a 36.21 Table 5.2: Results from the designed experiment indicating which factors had a significant effect on ejection force. 113 Main Effects Plot - Data Means for EF Set 1 Packing Time Cooling Time Packing Pres EF Set 1 41.5 41.0 40.5 40.0 39.5 2 6 5 10 0 15 5 10 Interaction Plot - Data Means for EF Set 1 2 6 5 10 15 0 5 10 42.0 Packing Time 6 40.5 39.0 2 Cooling Time 42.0 15 40.5 10 39.0 5 Packing Pres 42.0 10 40.5 5 0 39.0 Figure 5.5: Main effects and interactions for HDPE with the P-20 insert. (Ejection force is shown in pounds.) 114 Main Effects Plot - Data Means for EF Set 2 Packing Time Cooling Time Packing Pres 89.0 EF Set 2 87.5 86.0 84.5 83.0 2 6 5 10 0 15 5 10 Interaction Plot - Data Means for EF Set 2 2 6 5 10 15 0 5 10 90 Packing Time 6 85 80 2 90 Cooling Time 15 85 10 80 5 Packing Pres 90 10 85 5 80 0 Figure 5.6: Main effects and interactions for HIPS with the P-20 insert. (Ejection force is shown in pounds.) 115 Main Effects Plot - Data Means for EF Set 3 Packing Time Cooling Time Packing Pres 43.7 EF Set 3 42.9 42.1 41.3 40.5 2 6 5 10 0 15 5 10 Interaction Plot - Data Means for EF Set 3 2 6 5 10 15 0 5 Packing Time 10 45.0 6 42.5 40.0 2 Cooling Time 45.0 15 42.5 10 40.0 5 Packing Pres 45.0 10 5 0 42.5 40.0 Figure 5.7: Main effects and interactions for HDPE with the ST-100 insert. (Ejection force is shown in pounds.) 116 Main Effects Plot - Data Means for EF Set 4 Packing Time 86.4 Cooling Time Packing Pres EF Set 4 85.8 85.2 84.6 84.0 2 6 5 10 15 0 5 10 Interaction Plot - Data Means for EF Set 4 2 6 5 10 15 0 5 10 88 Packing Time 6 84 80 2 88 Cooling Time 15 84 10 80 5 Packing Pres 88 10 84 5 0 Figure 5.8: Main effects and interactions for HIPS with the ST-100 insert. (Ejection force is shown in pounds.) 117 80 Main Effects Plot - Data Means for EF Set 5b Packing Time 70.6315 Cooling Time Packing Pres EF Set 5b 68.4793 66.3271 64.1748 62.0226 2 6 0 12 0 15 0 5 Interaction Plot - Data Means for EF Set 5b 2 6 0 12 0 15 0 5 Packing Time 70 6 65 60 2 Cooling Time 70 150 65 60 120 Packing Pres 70 5 65 0 60 Figure 5.9: Main effects and interactions for HDPE with the SL 5170/P-20 insert. (Ejection force is shown in pounds.) 118 Main Effects Plot - Data Means for EF Set 6a Packing Time Cooling Time Packing Pres 195 EF Set 6a 185 175 165 155 2 6 0 12 0 15 0 5 Interaction Plot - Data Means for EF Set 6a 2 6 0 12 0 15 0 5 200 Packing Time 6 175 150 2 200 Cooling Time 150 175 150 120 Packing Pres 200 5 175 0 150 Figure 5.10: Main effects and interactions for HIPS with the SL 5170/P-20 insert. (Ejection force is shown in pounds.) 119 5.3 Standard Friction Testing Results Coefficient of friction results from the standard tests are shown in Figures 5.11 and 5.12. In general the data show some expected trends. For example, at room temperature, the friction coefficient of HIPS is larger than that of HDPE on all three plate materials. Also, the friction coefficients of both thermoplastics on the SL 5170 plate are higher than those on the metal plates. 5.3.1 HDPE Standard Friction Results Temperature did not make a dramatic difference in friction coefficient for HDPE on P-20, ST-100, or SL 5170. For the P-20 plate, the difference in coefficient between temperatures was not statistically significant. With the ST-100 and SL 5170 plates, there was an increase in the friction coefficient in the ejection temperature tests as compared to the room temperature tests. This is probably because the adhesion component of friction became more apparent in the heated tests. From ejection temperature to elevated temperature for the ST-100 plate, the difference in the coefficient of friction was not statistically significant. For the SL 5170 plate, there was actually a slight decrease in friction coefficient from ejection temperature to elevated temperature, an unexpected result. In all cases, elevating the initial temperature to imprint the surface of the specimen was presumed to increase the friction coefficient. This was not the case for any of the HDPE tests. No change in coefficient (in the case of P-20 and ST-100) and the slight decrease in coefficient (in the case of SL 5170) may be due to shrinkage of the sled 120 specimen from the plate and a reduction in the area of contact due to the imprinted pattern. HDPE is a crystalline polymer and will shrink more than HIPS, which is an amorphous polymer. As discussed in Chapter 3, compared to amorphous materials, a crystalline structure can arrange itself into a tighter, more orderly fashion as the polymer cools. It is noted, however, that the imprinted pattern was not very pronounced on the HDPE specimens. Therefore, the temperature to which the specimens were heated was probably not high enough to sufficiently soften the polymer. The relationship of HDPE friction coefficients among the three plate materials was as expected. The coefficients for HDPE on SL 5170 were highest because of the nature of polymer on polymer materials and because the surface roughness of this plate was higher than that of the other two. The coefficients of HDPE on ST-100 were lowest because the surface roughness of this plate was lower than that of the other two. The conductivity of the ST-100 plate is also highest, which may have contributed to a reduction in adhesion by dissipating heat at a faster rate in the heated tests. 121 Plate Material Surface Roughness P-20 Steel 0.7 microns LaserForm ST-100 SL5170 Resin 0.2 microns 3.6 microns HDPE Static Friction Coefficient Room Temp Ejection Temp Elevated Temp +0.08 +0.04 +0.06 0.26 -0.07 0.31 -0.02 0.28 -0.02 +0.02 +0.04 +0.01 0.21 -0.03 0.26 0.25 -0.05 -0.01 +0.04 +0.02 +0.07 0.37 -0.07 0.45 -0.05 0.38 -0.02 Static CoF of HDPE 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Room Temp On P-20 Steel Ej Temp On ST-100 Elev Temp On SL5170 Static CoF of HDPE 0.6 0.5 0.4 Room Temp 0.3 Ej Temp 0.2 Elev Temp 0.1 0 On P-20 Steel On ST-100 On SL5170 Figure 5.11: Standard friction test results for HDPE; means and ranges shown in the table. 122 Plate Material Surface Roughness P-20 Steel 0.7 microns LaserForm ST-100 SL5170 Resin 0.2 microns 3.6 microns HIPS Static Friction Coefficient Room Temp Ejection Temp Elevated Temp +0.04 +0.04 +0.11 0.36 -0.04 0.32 -0.04 0.35 -0.21 +0.08 +0.03 +0.17 0.32 0.13 0.54 -0.29 -0.05 -0.02 +0.05 +0.11 +2.65 0.43 -0.03 0.56 -0.13 5.47 -2.08 Static CoF of HIPS 5.47 SL 5170 elev temp 0.6 0.5 0.4 0.3 0.2 0.1 0 Room Temp On P-20 Steel Ej Temp On ST-100 Elev Temp On SL5170 Static CoF of HIPS 5.47 0.6 0.5 0.4 Room Temp 0.3 Ej Temp 0.2 Elev Temp 0.1 0 On P-20 Steel On ST-100 On SL5170 Figure 5.12: Standard friction test results for HIPS; means and ranges shown in the table. 123 5.3.2 HIPS Standard Friction Results The HIPS friction results were much more diverse than the HDPE results and will be discussed by individual plate material. HIPS on P-20 Steel – Temperature had only a slight effect in this case. While one would expect the friction coefficient to increase with temperature, the coefficient actually decreased in the ejection temperature test compared to the room temperature test. This may be explained by a softening of the polymer to allow asperities in the two surfaces to slide over each other more easily. Adhesion was probably not dominant in this case. The elevated temperature test resulted in a small increase in average friction, but this difference is not statistically significant. HIPS on ST-100 – The results from these tests show the same trend by temperature as the P-20 results, except that it is much more pronounced. As in the case of the P-20 Steel plate, the decrease in coefficient from room temperature to ejection temperature may be explained by a softening of the polymer to allow asperities in the two surfaces to slide over each other more easily. The marked increase in coefficient at the elevated temperature may be caused by the increase in area of contact due to the imprinted pattern, from an increase in adhesion, or both. Note also that the surface roughness of the P-20 is larger than that of the ST-100. At room temperature and ejection temperature, this may have contributed to higher friction on the P-20 plate, while at elevated temperature it may have caused more adhesion on the ST-100 plate. HIPS on SL 5170 – Adhesion had a very prominent effect in the heated tests on the SL 5170 plate. At ejection temperature, the friction coefficient is significantly higher 124 than at room temperature. Whereas the softened polymer slid over the metal plates with less resistance, it adhered to the resin plate, causing a larger peak frictional force. The elevated temperature test shows an extreme case of this adhesion, due to a strong interaction between the two materials. Secondary forces such as dispersion forces, polarity, or hydrogen bonding, contribute to this interaction. Adhesion between two materials, for example, will be maximized when their polarities are similar. While the imprinted pattern of the SL 5170 plate into the softened HIPS material may also have contributed to the spike in frictional force, the mechanical interaction was not a dominant influence. Figure 5.13 shows a sample plot from the HIPS on SL 5170 test. The graph shows an initial large peak followed by a lower peak as the specimen begins to slide. The initial peak may be explained by the strong force required to overcome adhesion, followed by a lesser force required to overcome roughness once the specimen is “unstuck.” The HIPS on SL 5170 at elevated temperature was the only test to show this phenomenon. The secondary molecular forces that caused HIPS and SL 5170 to adhere to each other were not a prominent factor with other material pairs. Polarity has no effect between HIPS and the two metal plates, because the metal plates become oxidized, and there is an insulating layer between the polymer specimen and the plate. Polarity has no effect between HDPE and any of the plates because HDPE is a nonpolar material. Other test results looked more like the sample HIPS on P-20 graph shown in Figure 5.14. 125 4.000 3.500 Friction Force (lbf) 3.000 2.500 2.000 1.500 1.000 0.500 0.000 Time 12 24 36 48 60 72 84 Time (sec) Figure 5.13: Sample plot of load vs. time for HIPS on SL 5170 from elevated temperature tests. 126 0.250 Friction Force (lbf) 0.200 0.150 0.100 0.050 0.000 Time 12 24 36 48 60 72 84 Time (sec) Figure 5.14: Sample plot of load vs. time for HIPS on P-20 from elevated temperature tests. The relationship of HIPS friction coefficients among the three plate materials was as expected, except for the elevated temperature test on ST-100 compared to P-20 Steel. One would anticipate that the coefficient of friction on ST-100 would be lower because its surface roughness is lower than that of the P-20. Instead, however, the smoother finish may have allowed the adhesion component of friction to dominate in this case. In the room temperature and ejection temperature cases, the friction coefficient of HIPS on ST-100 was lowest, and in all cases the friction coefficient of HIPS on SL 5170 was highest. 127 5.4 Reliability of the Data In conjunction with the experimental and test discussions above, a few parameter discrepancies and data variations must be noted. First, there are differences in injection molding process parameters among experimental sets. These are necessary because of the nature of the materials used. For example, HDPE and HIPS have different processing temperatures and injection velocities. Also, the temperatures and pressures used with the SL 5170 resin mold insert are as low as the processing window will allow so that minimal deformation or degradation occurs. A complete list of process parameters is included in Chapter 4. Second, there may be some variation in the data due to the following: • HDPE parts were flared slightly at the rim in ejections from the P-20 and ST-100 cores. This flaring occurred due to the magnitude of the ejection force applied by the stripper plate and the softness of the warm thermoplastic material. This flaring may have increased canister diameter measurements slightly, and can affect ejection force calculations. • The SL 5170 core was susceptible to swelling. Experiments with the SL 5170 core and the P-20 cavity were run with significantly lower ejection temperature because, if the heat was increased to raise ejection temperature, then core swelling was excessive. Core swelling may have affected diameter measurements as well. 128 • The shape of the SL 5170 cavity was modified to include a taper to facilitate ejection. Those few experimental parts from the SL 5170 cavity required adjustments to the ejection force equation based on this geometry. Shrinkage and ejection forces may have been affected as well. • Surface roughnesses vary among the three insert cores. The plates used in the friction tests were intended to have the same surface roughness as their corresponding injection molding core, but this was not accomplished in all cases (Table 5.3). Comparisons of friction coefficients among materials and between friction test data and experimental injection molding data must take this into account. • The length of the part was not measured at the time of ejection. This introduces a small amount of error, especially for HDPE, in the calculation of ejection force (see next section) because the lateral shrinkage is not taken into account. Surface Roughness, Ra (microns) P-20 Plate 0.7 P-20 Core 0.7 ST-100 Plate 0.2 ST-100 Core 0.3 SL 5170 Plate SL 5170 Core 3.6 0.7 Table 5.3: Surface roughnesses of all plates (friction tests) and cores (injection molding experiments). 129 5.5 Calculation of Ejection Force Using the Model Using the values for coefficient of static friction from the standard tests at elevated temperature, ejection forces have been calculated using the Menges model derived in Chapter 3. These values, along with the experimental values for ejection force, are shown in Tables 5.4 and 5.5. The difference between calculated values and actual experimental values for ejection force is significant, and excessive in some of the HIPS cases. 130 HDPE P-20 ST-100 Packing Cooling Packing Experiment Calculation Experiment Calculation Time Time Pressure s s % N N N N 2 5 0 177.15 84 182.30 33 2 5 5 183.29 78 190.02 39 2 5 10 186.52 82 209.79 46 2 10 0 176.51 92 177.44 36 2 10 5 185.89 84 196.61 43 2 10 10 172.87 77 194.33 41 2 15 0 191.90 114 196.15 82 2 15 5 174.68 80 201.68 46 2 15 10 173.08 86 208.75 45 6 5 0 184.69 89 173.95 59 6 5 5 193.58 82 185.81 71 6 5 10 173.88 67 184.41 64 6 10 0 171.28 79 172.12 59 6 10 5 174.63 72 178.83 68 6 10 10 175.19 63 180.87 64 6 15 0 170.00 82 170.97 67 6 15 5 180.05 86 186.82 74 6 15 10 185.53 76 184.13 78 SL 5170 SL Core with P-20 Cavity Packing Cooling Packing Experiment Calculation Experiment Calculation Time Time Pressure s s % N N N N 2 150 0 239.06 73 2 180 0 193.21 83 2 2 2 2 6 6 6 6 120 120 150 150 120 120 150 150 0 5 0 5 0 5 0 5 274.21 299.65 258.76 278.18 313.33 297.38 321.09 317.93 286 251 355 248 261 155 255 163 Table 5.4: Calculated values of ejection force for HDPE from the Menges equation and experimental data. 131 HIPS P-20 ST-100 Packing Cooling Packing Experiment Calculation Experiment Calculation Time Time Pressure s s % N N N N 2 5 0 343.97 218 366.29 700 2 5 5 376.37 235 389.51 639 2 5 10 401.48 205 375.28 603 2 10 0 346.11 237 375.71 753 2 10 5 385.56 258 393.54 698 2 10 10 408.26 210 394.47 612 2 15 0 384.60 284 366.31 767 2 15 5 381.94 220 393.56 592 2 15 10 403.70 273 398.83 655 6 5 0 376.97 234 363.86 353 6 5 5 395.23 240 378.14 330 6 5 10 393.38 73 388.07 330 6 10 0 369.19 205 369.54 541 6 10 5 390.81 168 360.58 438 6 10 10 391.73 131 374.38 343 6 15 0 351.62 182 340.67 541 6 15 5 394.77 228 370.57 555 6 15 10 424.46 264 399.88 560 SL Core with P-20 Cavity SL 5170 Packing Cooling Packing Experiment Calculation Experiment Calculation Time Time Pressure s s % N N N N 2 120 0 1334.27 2268 2 150 0 1136.12 4094 2 150 5 1512.25 1495 2 2 2 2 6 6 6 6 120 120 150 150 120 120 150 150 0 5 0 5 0 5 0 5 695.76 826.28 610.12 845.02 770.24 939.13 702.28 892.15 13059 6277 19679 6825 5405 2812 8383 1905 Table 5.5: Calculated values of ejection force for HIPS from the Menges equation and experimental data. 132 5.5.1 Calculated Ejection Force for HDPE Figure 5.15 shows that the calculated values for ejection force for HDPE from the Menges model are lower than the measured values by 50 to 70 percent on average, except for those from the SL 5170 core with the P-20 cavity. Based on results from the standard friction test at elevated temperature, the calculated values for ejection force were expected to be lower than the actual values because, as previously mentioned, the elevated temperature was probably not high enough to measure an accurate value of the coefficient of static friction for HDPE. Furthermore, the standard test environment was not identical to the injection molding environment in terms of temperatures and pressures and their respective histories. Calculated ejection force values for HDPE from the SL 5170 core with the P-20 cavity were closer to actuals, i.e., within 16 percent on average. This is because the ejection temperatures during these experiments were lower than the others, so the friction coefficient from the standard test is more comparable. 133 Ejection Force HDPE 400 350 300 N 250 200 150 100 50 0 P-20 SL/P-20 ST-100 Experimental SL 5170 Calculated Figure 5.15: Calculated values for ejection force for HPDE using the Menges model compared with experimental values, averaged across all runs. 5.5.2 Calculated Ejection Force for HIPS Calculated values for ejection force, averaged across all runs, for HIPS parts from the P-20 and ST-100 cores are shown with experimental values in Figure 5.16. Those from the SL 5170 insert, and the SL 5170 core with the P-20 cavity are shown in Figure 5.17. The model estimated ejection force for HIPS on P-20 to be 44 percent lower than 134 actuals, and on ST-100 to be 47 percent higher on average. For the SL 5170 core, calculated values were higher on average than actual values by 97 percent. (Note that only three runs of data were collected for the SL 5170 insert.) For the SL 5170 core with the P-20 cavity, calculated values were extremely high (924 percent higher than actuals). As was shown for HDPE, differences between calculated and actual ejection force values for HIPS are due in part to friction coefficients, which vary between standard measurements and actual ejection. Assuming this to be a primary reason for the differences in calculated and actual ejection forces, the measured friction coefficient for HIPS with P-20 was low, and the measured friction coefficients for HIPS with ST-100 and HIPS with SL 5170 were high. In any case, the surface interactions of each material pair were different during the standard tests as compared to during the injection molding experiments. Therefore the friction coefficient measured in the standard test caused some error in the calculation of ejection force when applied to the injection molding case. 5.5.3 Possible Sources of Error One source of error in the calculation of ejection force using the model developed by Menges is the coefficient of friction measurement. As mentioned above, the environment of the standard friction test is not identical to the environment of the injection molding experiment. Therefore, it is likely that the standard measurement of the coefficient of friction, even at elevated temperature, is not accurate. Another possible source of error in the ejection force calculation is the elastic modulus measurement. Again, the environment of the standard modulus measurement 135 does not exactly simulate the injection molding environment, so there is likely to be some error. Furthermore, the elastic modulus of the HIPS material is more sensitive to temperature as compared with HDPE. A small difference between the measured and actual temperatures in the modulus measurements described in Chapter 4 or in the injection molding experiments would cause a significant change in elastic modulus and a resultant change in the calculation of ejection force. A third source of error in the ejection force calculation is the measurement of the relative change in diameter of the part. The digital imaging approach used to measure the inside and outside diameters of the part was probably the most accurate method short of a laser-based, real time, and much more expensive system. However, digital imaging required manual transfer of the part from the injection molding machine to the camera fixture and most likely introduced some error by expanding the time between ejection and capture of the digital data. 136 Ejection Force HIPS 800 700 600 N 500 Experimental 400 Calculated 300 200 100 0 P-20 ST-100 Figure 5.16: Calculated values for ejection force for HIPS parts from the P-20 and ST-100 cores compared with experimental values, averaged across all runs. Ejection Force HIPS 10000 8000 6000 N Experimental Calculated 4000 2000 0 SL/P-20 SL 5170 Figure 5.17: Calculated values for ejection force for HIPS parts from the SL 5170 core compared with experimental values, averaged across all runs. 137 5.6 Calculation of Apparent Friction Coefficients using the Menges Model By rearranging the Menges model equation, and using experimental ejection force and shrinkage data, the apparent coefficient of static friction was calculated for each material combination. Table 5.7 lists the calculated values by experimental parameters. 138 Packing Time s 2 2 2 2 2 2 2 2 2 6 6 6 6 6 6 6 6 6 Cooling Time s 5 5 5 10 10 10 15 15 15 5 5 5 10 10 10 15 15 15 Packing Pressure % 0 5 10 0 5 10 0 5 10 0 5 10 0 5 10 0 5 10 Packing Time s 2 2 2 2 Cooling Time s 120 150 180 150 Packing Pressure % 0 0 0 5 2 2 2 2 6 6 6 6 120 120 150 150 120 120 150 150 0 5 0 5 0 5 0 5 Calculated Friction Coefficient HDPE HIPS P-20 LaserForm P-20 LaserForm Steel ST-100 Steel ST-100 0.59 0.65 0.64 0.54 0.62 0.63 0.47 0.61 0.56 0.58 0.66 0.73 0.61 0.68 0.78 0.58 0.59 0.68 1.40 1.21 1.15 1.24 1.14 1.20 0.60 1.10 1.16 0.73 0.66 0.73 0.73 0.66 0.71 0.64 0.63 0.59 HDPE SL 5170 SL 5170 w/P-20 N N 1.24 0.88 0.55 0.56 0.68 0.51 0.52 0.68 0.47 0.61 0.52 0.56 0.58 1.88 0.63 0.81 1.05 0.68 0.61 0.56 0.28 0.33 0.34 0.27 0.30 0.35 0.26 0.36 0.33 0.56 0.62 0.63 0.37 0.44 0.59 0.34 0.36 0.39 HIPS SL 5170 SL 5170 w/P-20 N N 3.22 1.52 5.53 0.36 0.45 0.28 0.43 0.46 0.73 0.48 0.74 0.29 0.72 0.17 0.68 0.78 1.83 0.46 2.56 Table 5.6: Calculated apparent coefficient of friction results according to packing time, cooling time, and packing pressure parameters. 139 5.6.1 HDPE Apparent Coefficient of Friction Results Figure 5.18 shows calculated values of the static friction coefficient by run for HDPE. The average calculated value of friction coefficient for HDPE from the P-20 core is lower than that from the ST-100 core, which is lower than that from the SL 5170 core. The calculated value of friction coefficient from the SL 5170 core with the P-20 cavity is lower than the other calculated values due to the lower ejection temperature. Coefficient of Static Friction HDPE 1.6 1.4 1.2 CoF 1 0.8 0.6 0.4 0.2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run P-20 ST-100 SL 5170 SL/P-20 Figure 5.18: Calculated values of the apparent coefficient of static friction for HDPE, all runs. 140 5.6.2 HIPS Apparent Coefficient of Friction Results Figure 5.19 shows calculated values of the static coefficient of friction for HIPS from the experimental data. Results from all four inserts are shown. The HIPS coefficient of friction on P-20 is generally higher than on ST-100. Two of the friction values from the P-20 insert are much higher than the others, i.e., 1.88 and 1.05. These values correspond to high packing time, high packing pressure, and lower cooling time, and therefore imply that higher pressure and temperature cause higher friction between the HIPS material and the steel. This phenomenon is also seen with the SL 5170/P-20 insert, but not with ST-100. Friction values for HIPS with the SL 5170 core and cavity are very high, as expected. In general, the coefficient of friction values on SL 5170 core with P-20 cavity are low because ejection temperatures were much lower. 141 CoF Coefficient of Static Friction HIPS 6.0 5.8 5.6 5.4 5.2 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run P-20 ST-100 SL 5170 SL/P-20 Figure 5.19: Apparent coefficients of friction calculated from experimental results for HIPS. 142 5.6.3 Apparent Friction Coefficient Results from the P-20 and ST-100 Inserts From the P-20 insert, the friction coefficient was lower for HDPE than for HIPS for 10 out of 18 runs, most of which correspond to the higher level of packing time (Figure 5.20). From the ST-100 insert, the friction coefficient for HDPE was higher than for HIPS for all runs (Figure 5.21). The HDPE coefficients corresponding to low packing time were higher than those corresponding to high packing time, whereas the HIPS coefficients showed the opposite relation to a lesser degree. Coefficient of Static Friction P-20 2 CoF 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run HDPE HIPS Figure 5.20: Apparent coefficient of static friction for parts from the P-20 insert. 143 Coefficient of Static Friction ST-100 1.6 1.4 1.2 CoF 1 0.8 0.6 0.4 0.2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run HDPE HIPS Figure 5.21: Apparent coefficient of static friction for parts from the ST-100 insert. 5.6.4 Apparent Friction Coefficient Results from the SL 5170 and SL 5170/P-20 Inserts With SL 5170 core and cavity, calculated values for coefficient of friction for HIPS are much higher than for HDPE (Figure 5.22). From the SL 5170 insert with the P20 cavity, HIPS friction coefficients were higher than HDPE for the runs at higher packing pressure, and lower for the runs with zero packing pressure (Figure 5.23). Note, once again, that the coefficients of friction on SL 5170 core with P-20 cavity were low because ejection temperatures were much lower. 144 Coefficient of Static Friction SL 5170 6.0 5.0 CoF 4.0 3.0 2.0 1.0 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Run HDPE HIPS Figure 5.22: Apparent coefficient of static friction for parts from the SL 5170 insert. 145 18 Coefficient of Static Friction SL/P-20 3 2.5 CoF 2 1.5 1 0.5 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Run HDPE HIPS Figure 5.23: Apparent coefficient of static friction for parts from the SL 5170 core with the P-20 cavity. 5.6.5 Comparing Calculated Friction Results to Standard Friction Test Results Figures 5.24 and 5.25 show the average calculated friction coefficient results for HDPE and HIPS, respectively, along with those measured in the standard tests. The standard test values for HDPE are all lower than the calculated values because the temperatures and pressures in the standard tests did not match those in the injection molding experiments, and because of the sources of error discussed in Section 5.5.3. The values shown for the SL 5170 core with the P-20 cavity are reasonably close because the 146 ejection temperature during this injection molding experiment was much lower compared with experiments using the other inserts. Calculated friction coefficients for HIPS on P20 were higher than standard test results. Those for HIPS on ST-100, however, were not all higher than standard test results at elevated temperature. Recall that only two runs of HDPE parts and three runs of HIPS parts were completed with the SL 5170 insert. These average and range calculations, then, are based on only a small amount of data. Note for HIPS that the calculated friction coefficient value corresponding to higher packing pressure (5.53) is comparable to the friction coefficient obtained in the standard test (5.47). 147 Static Friction Coefficient for HDPE 1.6 1.4 1.2 1.0 P-20 ST-100 0.8 SL 5170 SL/P20 0.6 0.4 0.2 0.0 STD RT STD Elev Calculated Figure 5.24: Average apparent coefficient of static friction for HDPE compared to standard test results at room temperature and elevated temperature. 148 Static Friction Coefficient for HIPS 9.0 8.0 7.0 6.0 5.0 ST-100 SL 5170 P-20 4.0 SL/P20 3.0 2.0 1.0 0.0 STD RT STD Elev Calculated Figure 5.25: Average apparent coefficient of static friction for HIPS compared to standard test results at room temperature and elevated temperature. 149 5.7 Other Observations of Rapid Tooled Inserts In addition to the quantitative analyses, a number of qualitative observations were made during experiments with the ST-100 and SL 5170 tools. These are summarized in the paragraphs below. The ST-100 insert seemed to operate as well as the P-20 insert. There were no problems with core swelling or parts adhering to the sintered material. The insert held up well for the number of parts that were processed, and had every indication that it would last for many more. There were, however, differences in data between the ST-100 and P20 experiments, indicating that there are thermal and adhesion differences that affect ejection force and friction coefficient. The SL 5170 insert, on the other hand, operated quite differently from the other two metal inserts, as one would expect. Surprisingly, the resin core held up for more than 105 full parts, with minimal flashing and no catastrophic failure. The following are some of the anomalies of the SL 5170 insert discovered during processing. It was obvious during molding of both HDPE and HIPS that there was adhesion of the parts to the SL 5170 core (no matter which cavity was used). The adhesion was both visible and audible. After the mold opened and before the part was ejected, the portion of the part that adhered was more translucent than the rest. At ejection, the part would audibly snap upon release, then slide off the core. It is quite possible that the force from the stripper plate, in conjunction with the adhesion, forced the part to bow outward somewhat, thus reducing the amount of shear along the core surface. 150 Core swelling was a problem with the SL 5170 material. Control of the core temperature became especially important when using this insert. During the first experimental set, the core diameter was reduced by 0.13 mm (0.005 in) to avoid interference between the swelled core and the stripper plate. During experiments with the SL 5170 core and the P-20 cavity, the planned temperature at ejection could not be maintained. This was because the steel cavity would conduct much of the heat away from the insert, and allowing the temperature to build would cause excessive swelling of the resin core. The reason for the core swelling is not entirely clear. There may have been some resin material that was not completely cured. Or swelling may be characteristic of this material at certain temperature, humidity, and/or pressure conditions. After the SL 5170 core processed about 40 parts, it began to show some internal defects. The defects were barely visible at first, but gradually increased in number and size as more and more parts were processed. Shining a light through the end of the core highlighted the defects, and one could see that they looked like internal delaminations along the layers of the tool (Figure 5.26). The core, however, lasted through the entire experiment without these defects propagating into failure. 151 Figure 5.26: Defects in the SL 5170 core. The issue with the SL 5170 core that most affected experimental results was deflection of the cavity wall during injection. Only a few experimental runs were completed with the SL 5170 core and cavity when the parts began to stick in the cavity upon mold opening. The wall of the cavity apparently gave under injection pressure and elastically deformed. Then, although the part began to shrink, too much material had been forced into the mold, and was held fast by the cavity wall. In a preliminary study of this phenomenon, the SL 5170 insert was modeled with the two thermoplastic materials using ABAQUS finite element analysis (Carpenter 2004). Using the thermoplastic material characteristics from the software database and the geometry of the SL 5170 core and cavity, the injection was simulated for HDPE and HIPS with no packing, and for HIPS with 6 seconds at 5 percent packing. Results show that the walls of the cavity do indeed elastically deform in a way that would cause parts to stick. In Figure 5.27, deformation of the SL 5170 by HDPE injection, no packing, is 152 concentrated near the base of the cavity, with a maximum magnitude of 0.06 millimeters (0.002 inches). In Figure 5.28, similar deformation results with HIPS, although the maximum magnitude is slightly less at 0.05 millimeters (0.002 inches). When packing is included, Figure 5.29, the magnitude of the deformation is greater at 0.07 millimeters (0.003 inches), and it encompasses a greater area along the cavity wall. Figure 5.27: Simulation results of HDPE injection into SL 5170 insert, no packing. 153 Figure 5.28: Simulation results of HIPS injection into SL 5170 insert, no packing. 154 Figure 5.29: Simulation results of HIPS injection into SL 5170 insert, with packing. 155 CHAPTER 6 CONCLUSIONS This chapter presents the conclusions of this research project. First, conclusions regarding the molding of HDPE and HIPS parts using the rapid tooled inserts are presented, including the benefits and limitations of rapid tooled injection mold inserts. Next, the use of a model for determining ejection force and the coefficient of friction is discussed. Implications and future work are also included, and the chapter concludes with an overall summary. 6.1 Molding HDPE and HIPS with ST-100 and SL 5170 Inserts 6.1.1 Benefits and Limitations of Using Rapid Tooled Injection Mold Inserts The general benefits of rapid tools were discussed in Chapter 1. These include the ability to build complex geometries and incorporate conformal cooling lines, and potential savings in lead times and material and labor costs. As an injection molding insert, the ST-100 tool was very capable. Few process problems were experienced in this 156 work. This particular insert withstood the processing of 288 parts plus many test parts with no signs of wear or damage. For these quantities, the ST-100 performed just as well as the P-20, and even showed some advantages for processing HIPS in terms of friction coefficients. For the scope of this work, no limitations of the ST-100 insert are noted. The SL 5170 tool shares the same general benefits as the ST-100 and other rapid tools as mentioned above. Some unexpected benefits, as seen in this work, include its ability to mold both HDPE and HIPS, and durability of the core for processing more than 105 parts. The use of SL 5170 for injection molds, however, is limited because of the deformation that occurs with high pressure and swelling of the material at high temperature. Although the SL 5170 core did not fail catastrophically during these experiments, it is assumed that the fatigue life of the core is limited because of the defects that developed after approximately 40 parts. The interaction of the surface of SL 5170 with those of some thermoplastic materials, such as HIPS, is also a drawback. Adhesion often occurs between the two surfaces, which can accelerate failure of the core and potentially affect the quality of the part. The statistical analysis shows the effects of processing parameters on ejection forces for all three inserts. Packing time, cooling time, and packing pressure affect ejection force differently between the baseline steel and the rapid tooled inserts. Effects are also different between thermoplastic materials. Conclusions from the statistical analysis that relate directly to the insert material are as follows: • Ejection force increases with packing time for the SL 5170 insert. 157 • Ejection force decreases with packing time for the ST-100 insert. • Ejection force decreases with cooling time for the P-20 insert; but this is a nonlinear effect. 6.1.2 Friction and Ejection Force Considerations In the standard tests, the friction coefficients of HDPE were similar on P-20, ST100, and SL 5170 at all temperature conditions. HIPS showed a different friction response than HDPE, and its friction coefficients varied significantly between plate materials in heated tests. Both polymers showed highest coefficients on SL 5170 at all three temperature conditions. The HIPS test on SL 5170 showed the interplay of the adhesion and deformation components of friction and how this affects the friction coefficient. Although the standard friction tests at elevated temperatures may have given a more accurate estimate of the friction coefficient during ejection than those at room temperature, they still did not exactly simulate the actual process. Additional adjustments might be made to the temperatures and normal forces applied in the standard tests to render the results more similar to actual molding conditions. In the injection molding experiments, ejection forces for parts from the ST-100 core were generally similar to those from the P-20 baseline core (170 – 200 N for HDPE and 340 – 430 N for HIPS). HDPE parts from the SL 5170 core had slightly higher ejection forces (190 to 240 N), and those from the SL 5170 core with the P-20 cavity were higher still (250 – 330 N). Conversely, HIPS parts had higher ejection forces from 158 the SL 5170 core with the P-20 cavity (600 – 950 N), and much higher ejection forces from the SL 5170 core (1100 – 1600 N). This seems to indicate that, when ejecting HPDE parts from the SL 5170 core, a lower ejection temperature will increase shrinkage and increase ejection force, and, when ejecting HIPS parts from the SL 5170 core, a higher ejection temperature will increase adhesion and increase ejection force. Given the discussion above and all the results of this research, the following conclusions are drawn: § ST-100 inserts can be used to mold HDPE parts. This insert material performed similarly to P-20, but was affected differently by process parameters. Calculations of apparent coefficient of static friction indicated that friction can be high when packing time is low, but these values did not cause extremely large ejection forces. § ST-100 inserts can be used to mold HIPS parts. Once again, ST-100 performed similarly to P-20, and in some cases had lower ejection forces. § SL 5170 can be used to mold HDPE parts, but with adjusted process parameters or alternative cavity materials to minimize cavity deformation. Ejection temperatures should be relatively high to minimize the load on the core. Minimizing this load may extend core life prior to defect formation. 159 § SL 5170 is not recommended for molding HIPS due to adhesion and very high ejection forces. The coefficient of friction will increase with higher ejection temperatures and packing times due to adhesion, which is enhanced by the secondary forces between the two materials. Maintaining a lower ejection temperature (e.g., by using a P-20 core) reduces the ejection force somewhat. Core life, however, will probably be minimal. 6.2 Using a Model to Determine Ejection Force and the Coefficient of Friction In this work the Menges model was used to determine ejection force for comparison to experimental measurements. This model requires values for the coefficient of static friction between the part and the core, the elastic modulus of the part material at the time of ejection, and the relative change in diameter of the part immediately after ejection. Each of these values is difficult to obtain and introduces error into the calculation. The differences between the calculated ejection force and the actual ejection force varied from 16 percent to 70 percent for HDPE, and from 44 percent to 924 percent for HIPS. While some of the calculations provided good ballpark estimates, others did not, and only one was within 20 percent. The static friction coefficient measurement may have been the largest contributor to the lack of accuracy in the ejection force calculations. Standard friction tests provided values for static friction coefficients that were an improvement over room temperature 160 values, but were still not an exact simulation of the injection molding experiment. More standard testing at a wider range of temperature and pressure environments would be required to determine more accurate values for the static friction coefficient that occur during ejection of an injection molded part. The Menges model was also used to determine the apparent coefficient of friction for comparison to results from the standard friction tests. The specific description of friction coefficient is still largely a mystery. The term includes deformation and adhesion in unknown proportions and affected by certain conditions to unknown extents. The calculation of apparent coefficient of friction from ejection force models, as was done here, results in a value that encompasses a complete surface interaction under the given processing conditions. How this value compares with standard measures of friction coefficient has not been entirely clear. Furthermore, the calculated value of the apparent coefficient of friction includes error from the measurements of elastic modulus and the relative change in diameter, and so is not an apples-to-apples comparison with the standard friction test values. The apparent friction coefficient calculation, however, can be potentially useful in testing a new material for an injection mold insert application. For example, to estimate the ejection forces that will occur, a simple cylindrical mold insert can be built with the new material, and the apparent coefficient of friction can be calculated using the Menges model and injection molding data from the cylindrical mold. This calculated value can then be applied to the model to estimate ejection force for molds having different geometries but the same materials and similar processing parameters. 161 Results from this work, however, are insufficient to validate this application of the model. Additional experimentation with a different part geometry would be required to provide comparison data and prove this concept. 6.3 Implications and Future Work This work has included friction testing of thermoplastics against rapid prototyped materials, following a standard procedure and including higher temperatures; direct measurement of ejection force from steel and rapid tooled injection mold inserts; calculation of ejection forces using a model developed by Menges; and determination of the apparent coefficient of static friction from experimental data using the same Menges model. A good indication of the processing capability of a mold material are its ejection force requirements. The experimental results provide these data for all material combinations. The results were compared among thermoplastic and mold insert materials, and then compared to calculated ejection force values. These results give an indication of the usefulness of the ejection force model. The static friction coefficient results from the standard tests were also compared among materials, and then compared with calculations of apparent coefficients of friction. The friction test data are a useful reference for understanding the basic friction conditions between the thermoplastics, HDPE and HIPS, and the mold insert materials, 162 P-20, ST-100 and SL 5170. Friction results from the standard tests have also pointed out the adhesion phenomenon that occurs between HIPS and SL 5170. The adhesion results from the molecular forces between the two materials and is enhanced by higher ejection temperatures. The statistical results are useful for determining those process parameters that can be adjusted to optimize ejection force for the material pairs studied. Analysis of variance has shown which parameters affect ejection force and how strong those effects are for the given process window. This information can be used, once a decision has been made to use one of these inserts, to design a process that meets ejection force requirements. The experiments have shown the molding of HDPE and HIPS parts with P-20 and ST-100 mold inserts as rather routine. The more interesting results come from the use of the SL 5170 insert. While HDPE parts can be molded with SL 5170 inserts, higher friction coefficients and ejection forces will result. HIPS parts were molded with the SL 5170 insert as well, but with extremely high friction coefficients and ejection forces. These forces caused high cyclic loads on the SL 5170 core, leading to the formation of internal defects. Additionally, deformation of the cavity occurred during injection. These considerations must be taken into account in the application of this rapid tooled material for injection molds. Overall, the data are useful for choosing mold insert materials, for deciding whether or not to use rapid tooled inserts for small quantity production, for development of rapid prototyping materials and processes, or for injection molding part or machine design. 163 As mentioned in the previous section, additional work in static friction coefficient measurement and ejection force comparison with a different part geometry would enhance the current work. Friction testing under various temperature and pressure conditions may provide more accurate values of the friction that exists during ejection. These values could then be used with the ejection force model. Further injection molding experiments with a different part shape would provide ejection force data to compare to those from the cylindrical part. This comparison would indicate whether or not the Menges model could be used to determine ejection forces for new mold insert materials. Other possible areas for future work include further study on the deformation and swelling of the SL 5170 material, failure testing and analysis of the SL 5170 core, optimization of injection molding process parameters, and materials characterization for adhesion. The deformation and swelling of the SL 5170 resin can be investigated to determine their actual causes. If curing of the material is an issue, improvements in the stereolithography process or its post-cure may have an effect. The selection of alternative materials may also be a solution, including more recently developed materials and resins that contain fillers. The defects that developed in the SL 5170 core can be analyzed to confirm whether or not they are in fact delaminations and to determine why they occurred. Improvements in the building or curing of the stereolithography tool might be required, or it may be characteristic of the process or material. Further testing to failure would provide useful data on the actual life expectancy of the core. Any of the designed experiments could be expanded to encompass a broader processing window, leading to the optimization of process parameters. This might be 164 especially useful for the SL 5170 core and the HDPE thermoplastic. Further study of the temperature and pressure parameters would further delineate the effects of the adhesion component of friction and clarify a feasible processing space. Since the role of adhesion and friction in injection molding is not explicitly understood, further study in this area would also be useful. This would include materials characterization and research into the interfaces between the thermoplastics and rapid prototyped materials. Areas that were not addressed in this work include part quality, as-built rapid tools, and conformal cooling lines. First, the performance of rapid tooled inserts cannot be completely assessed without consideration for part quality; this would include dimensional and surface finish quality. Second, the rapid tools used in this work were finish machined. The advantage to using rapid tools is maximized, however, if they are inserted as built and not post-processed. And third, the addition of conformal cooling lines would allow more control of processing temperatures. The nature of rapid prototyping processes is such that they facilitate the incorporation of conformal cooling lines. Further research in each of these areas would contribute to the potential use of rapid tools for injection molding. 6.4 Summary The application of rapid prototyped tools for injection molding, if technically feasible, may allow for small quantity production by reducing the cost of tooling. This 165 work has investigated one aspect of the technical feasibility through testing and experimentation to determine ejection force requirements and coefficients of friction. Friction coefficients between thermoplastics and rapid tooled materials were measured using a modified standard testing process. Injection molding experiments were conducted using three mold insert materials, P-20 steel, laser sintered ST-100, and stereolithography SL 5170 resin. Ejection forces for cylindrical parts molded with high density polyethylene and high impact polystyrene were measured directly and then compared with values calculated from an ejection force model. Process parameters affected the adhesion and deformation components of friction differently, depending on the materials characteristics. Results show that ST-100 is a good candidate for injection molding tools, and that SL 5170 may be a good candidate for molding some thermoplastics, but only in very small quantities. 166 LIST OF REFERENCES Anderson, D. M., Pine, B. J. II 1997, Agile Product Development for Mass Customization, Irwin Professional Publishing, Chicago. ASM International 1992, “Static and Kinetic Friction Coefficients for Selected Materials,” ASM Handbook, vol. 18, p. 73. Balsamo, R., Hayward, D., Malloy, R. 1993, “An Experimental Evaluation of Ejection Forces: Frictional Effects,” ANTEC, pp. 2515-2521. Beaman, J.J., Barlow, J.W., Bourell, D.L., Crawford, R.H., Marcus, H.L., McAlea, K.P 1997, Solid Freeform Fabrication: A new Direction in Manufacturing, Kluwer Academic Publishers, Dordrecht. Beer, F.P., Johnston, E.R. 1981, Mechanics of Materials, McGraw-Hill, New York, pp 325-6. Benabdallah, H., Fisa, B. 1989, “Static Friction Coefficient of Some Thermoplastics,” ANTEC, pp. 1055-1058. Benabdallah, S.H. 1993, “Static Shear Strength and Adhesion Friction of Some Thermoplastics,” Polymer Engineering and Science, vol. 33, no. 9, pp. 543-548. Bourell, D.L., Crawford, R.H., Marcus, H.L., Beaman, J.J., Barlow, J.W. 1994, “Selective Laser Sintering of Metals,” Proceedings of the 1994 ASME Winter Annual Meeting, November, pp. 519-528. Briscoe, B.J. 1981, “Wear of polymers: an essay on fundamental aspects,” Tribology International, August, pp. 231-243. Brydson, J.A. 1999, Plastics Materials, Butterworth-Heinemann, [On line] Available from www.knovel.com, [June 2004]. Burke, C.T. 1991, Ejection Force Measurement for the Injection Molding Process, Master Thesis, University of Lowell, Appendix E. 167 Burke, C., Malloy, R. 1991, “An Experimental Study of the Ejection Forces Encountered During Injection Molding,” ANTEC 1991, pp. 1781-1787. Cagle, C.V. (ed.) 1973, Handbook of Adhesive Bonding, McGraw-Hill, New York. Callister, W.D., Jr. 1985, Materials Science and Engineering An Introduction, John Wiley & Sons, New York. Campbell, C. 2000, “Rapid Tooling for Real World Applications,” Rapid Tooling Technologies for Plastic Applications, Society of Manufacturing Engineers Symposium, October, Detroit, MI. Carpenter, B. 2004, personal communication. Castle Island, “Worldwide Guide to Rapid Prototyping,” [Online]. Copyright Castle Island Co. Available: home.att.net/~castleisland/home.htm [May 2003]. Cedorge, T., Colton, J. 2000, “Draft Angle and Surface Roughness Effects on Stereolithography Molds,” Polymer Engineering and Science, vol. 40, no. 7, pp. 1581-1588. Cho, W., Sachs, E. M., Patrikalakis, N., M. 2001, “Solid Freeform Fabrication with Local Composition Control,” Rapid Prototyping, Society of Manufacturing Engineers, vol. 7, no. 2, pp 1-5. Colton, J. S., Crawford, J., Pham, G., Rodet, V. 2001, “Failure of Rapid Prototype Molds During Injection Molding,” CIRP Annals – Manufacturing Technology, vol. 50, no. 1, pp. 129-132. Colton, J.S., LeBaut, Y. 2000, “Thermal Effects on Stereolithography Injection Mold Inserts,” Polymer Engineering and Science, vol. 40, no. 6. Czichos, H. 1983, “Influence of Adhesive and Abrasive Mechanisms on the Tribological Behaviour of Thermoplastic Polymers,” Wear, vol. 88, pp. 27-43. Czichos, H. 1985, “Contact Deformation and Static Friction of Polymers, Influences of Viscoelasticity and Adhesion,” Polymer Wear and Its Control, American Chemical Society, pp. 3-26. Dalgarno, K. W., Stewart, T. D. 2001, “Manufacture of production injection mould tooling incorporating conformal cooling channels via indirect selective laser sintering,” Proceedings of the Institution of Mechanical Engineers Part B, Journal of Engineering Manufacture, vol. 215, pp.1323-1332. Dearnley, P.A. 1999, “Low friction surfaces for plastic injection moulding dies – an experimental case study,” Wear, vol. 225-229, pp. 1109-1113. 168 Dickens, P., Rudgley, M. 2001, “DirectAIMTM in a New Direction,” Rapid Prototyping and Manufacturing Conference [CD-ROM], Society of Manufacturing Engineers, Cincinnati. Eley, D.D. (ed.) 1961, Adhesion, Oxford University Press, London. Ferreira, E.C., Neves, N.M., Muschalle, R., Pouzada, A.S. 2001, “Friction Properties of Thermoplastics in Injection Molding,” ANTEC Portugal, [Online]. Available: www.dep.uminho.pt/NP/2001_03.pdf, [July 2003]. Ferreira, E. C., Muschalle, R., Neves, N. M., Pouzada, A. S. 2002, “Analysis of parameters determining the friction properties of thermoplastics in injection molding,” ANTEC 2002, vol. 1, pp. 878-881. Glanvill, A.B. 1971, The Plastics Engineer’s Data Book, Machinery Publishing Co. Ltd., The Whitefriars Press Ltd., London, p. 30. Harris, R. A., Dickens, P. M. 2001, “Mould Variables in Stereolithography Injection Moulds,” ANTEC 2001, vol. 1, pp 107-111. Harris, R. A., Dickens, P. M. 2003, “Determining, Understanding, and Controlling the Morphology of Injection Moulded Parts Produced in Stereolithography Moulds,” ANTEC 2003, pp. 889-893. Hopkinson, N., Dickens, P. 1999, “Study of ejection forces in the AIM™ process,” Materials and Design, vol. 20, pp. 99-105. Hopkinson, N., Dickens, P. 2000a, “Conformal Cooling and Heating Channels using Laser Sintered Tools,” Proceedings of the Solid Freeform Fabrication Symposium, Austin, TX, pp. 490-497. Hopkinson, N., Dickens, P. 2000b, “A comparison between stereolithography and aluminium injection moulding tools,” Rapid Prototyping Journal, vol. 6, no.4, pp. 253-258. Hopkinson, N., Dickens, P. 2000c, “Predicting stereolithography injection mould tool behaviour using models to predict ejection force and tool strength,” International Journal of Production Research, vol. 38, no. 16, pp. 3747-3757. Hopkinson, N., Dickens, P. 2000d, “Using stereolithography tools for injection moulding: research into tensile tool failure and unexpected benefits of the process,” Proceedings of the Institution of Mechanical Engineers, vol. 214, part B, pp. 891899. Jacobs, P. F. 1996, Stereolithography and other RP&M Technologies, Society of Manufacturing Engineers, Dearborn, MI, ASME Press, New York. 169 Jansen, K.M.B, Titomanlio, G. 1996, “Effect of Pressure History on Shrinkage and Residual Stresses – Injection Molding with Constrained Shrinkage,” Polymer Engineering and Science, vol. 36, no. 15, pp 2029-2040. Kabanemi, K.K., Vaillancourt, Wang, H., Salloum, G. 1998, “Residual Stresses, Shrinkage and Warpage of Complex Injection Molded Products: Numerical Simulation and Experimental Validation,” Polymer Engineering and Science, vol. 38, no. 1, pp. 21-37. Kai, C.C., Fai, L.K 1997, Rapid Prototyping, Principles and Applications in Manufacturing, John Wiley & Sons, New York. Kaminsky, P.G. 1996, “Lean Logistics: Better, Faster, Cheaper,” Defense Issues, vol. 11, no. 99, [Online] Available: http://www.defenselink.mil/speeches/index.html. Karapatis, N.P., van Griethuysen, J.-P.S., Glardon, R. 1998, “Direct rapid tooling: a review of current research,” Rapid Prototyping Journal, vol. 4, no. 2, pp. 77-89. Keicher, D., Gorman, P., Taute, L. 2001, “A Novel Method for Fabrication and Repair of Tooling?” National Tooling and Machining Association Record, January. Kinsella, M.E. 2000, “Piloted Concepts for Commercial-Military Integration Ready for Implementation,” Program Manager, July-August, p. 30. Li, Y., Gargiulo, E.P., Keefe, M. 2000, “Studies in Direct Tooling Using Stereolithography,” Journal of Manufacturing Science and Engineering, vol. 122, May, pp. 316-322. Malloy, R., Majeski, P. 1989, “Design of Pin Ejector Systems for Injection Molds,” ANTEC 1989, pp. 1231-1235. MatWeb Material Property Data (March 2004), [Online], Available from: www.matweb.com [6 June 2004]. McAlea, K., Booth, R., Forderhase, P., Lakshminarayan, U. 1995, “Materials for Selective Laser Sintering Processing,” Proceedings of the 27th International SAMPE Technical Conference, vol. 27, pp. 949-961. Menges, G., Bangert, H. 1981, “Measurement of Coefficients of Static Friction as a Means of Determining Opening and Demoulding Forces in Injection Moulds,” Kunststoffe, vol. 71, no. 9, pp. 552-557. Menges, G., Michaeli, W., Mohren, P. 2001, How to Make Injection Molds, Hanser Gardner Publications, Munich, Cincinnati. 170 Michaeli, W., Gutberlet, D., Hoffmann, S., Zachert, J. 1999, “Development steps for integrated material modeling for a semicrystalline polymer,” Computational Materials Science, vol.16, pp. 151-157. Michalski, D.D. 2000, A Parametric Study of the Ejection Forces Encountered During the Injection-Molding Process, Master Thesis, The Ohio State University. Muschalle, R. 2001, “Friction Behaviour of Thermoplastics,” Kunststoffe, vol. 91, no. 12, pp. 48-54. Nelson, C., Kepler, J., Booth, R., Conner, P. 1998, “Direct Injection Molding Tooling Inserts from the SLS Process with Copper Polyamide,” Proceedings of the Solid Freeform Fabrication Symposium, August, Austin TX, pp. 451-459. Optomec, Inc. 2001, company literature, Albuquerque. Osswald, T.A. 1998, Polymer Processing Fundamentals, Hanser Publishers, Munich. Palmer, A.E., Colton, J.S., 2000, “Failure Mechanisms in Stereolithography Injection Molding Tooling,” Polymer Engineering and Science, vol. 40, no. 6, pp. 13951404. Pantani, R., Titomanlio, G. 1999, “Analysis of Shrinkage Development of Injection Moulded PS Samples,” Intern. Polymer Processing, vol. XIV, no.2, pp. 183-190. Petrie, E.M. 2000, Handbook of Adhesives and Sealants, McGraw-Hill, [On line] Available from www.knovel.com, [June 2004]. Pham, G.T., Colton, J.S. 2002, “Ejection Force Modeling for Stereolithography Injection Molding Tools,” Polymer Engineering and Science, vol. 42, no. 4, pp.681-693. Pham, D. T., Dimov, S. S., Lacan, F. 2000, “The RapidTool process: technical capabilities and applications,” Proceedings of the Institution of Mechanical Engineers Part B, Journal of Engineering Manufacture,vol. 214, pp 107-116. Pine, B. J. II 1993, Mass Customization – The new Frontier in Business Competition, Harvard Business School Press, Boston. Pontes, A.J., Pinho, A.M., Miranda, A.S., Pouzada, A.S. 1997, “Effect of Processing Conditions on Ejection Forces in Injection Moulds,” O Molde, OctoberDecember, pp. 25-34. Pontes, A.J., Pantani, R., Titomanlio, G., Pouzada, A.S. 2001, “Prediction of Ejection Forces in Tubular Moldings in Amorphous Polymers,” ANTEC, vol. I, pp. 927931. 171 Pontes, A.J., Brito, A.M., and Pouzada, A.S. 2002, “Assessment of the Ejection Force in Tubular Injection Moldings,” Journal of Injection Molding Technology, vol. 6, no. 4, pp. 343-352. Popov, E.P. 1976, Mechanics of Materials, second edition, Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Rosato, D.V., Rosato, D.V. 1986, Injection Molding Handbook, Van Nostrand Reinhold Company, New York. Rubin, I.I. (ed.) 1990, Handbook of Plastic Materials and Technology, John Wiley & Sons, Inc., New York. Sachs, E., Wylonis, E., Allen, S., Cima, M., Guo H. 2000, “Production of Injection Molding Tooling With Conformal Cooling Channels Using the Three Dimensional Printing Process,” Polymer Engineering and Science, May, vol. 40, no. 5, pp. 12321247. Sasaki, T., Koga, N., Shirai, K., Koayashi, Y., Toyoshima, A. 2000, “An experimental study on ejection forces of injection molding,” Precision Engineering, vol. 24, pp. 270-273. Shen, K., Chen, L.-M., Jiang, L. 1999, “Calculation of ejection force of hollow, thin walled, and injection molded cones,” Plastics, Rubber, and Composites, vol. 28, no. 7, pp. 341-345. Titomanlio, G., Jansen, K.M.B. 1996, “In-Mold Shrinkage and Stress Prediction in Injection Molding,” Polymer Engineering and Science, vol. 36, no. 15, pp. 20412049. Trantina, G.G., Nimmer, R. 1994, Structural Analysis of Thermoplastic Components, McGraw-Hill, NewYork. University of Southern Mississippi, “The Macrogalleria,” [Online]. Copyright 1995-2002 by the Department of Polymer Science, Available: www.psrc.usm.edu/macrog/index.htm [November 2003]. Victor, B., Boynton, A. 1998, Invented Here – Maximizing Your Organization’s Internal Growth and Profitability, Harvard Business School Press, Boston. Wang, H., Kabanemi, K. K., Salloum, G. 1997 Wang, H., Kabanemi, K. K., Salloum, G. 2000, “Numerical and Experimental Studies on the Ejection of Injection-Molded Plastic Products,” Polymer Engineering and Science, vol. 40, no. 3, pp. 826-840. 172 Weaver, T.J., Thomas, J.A., Atre, S.V., German, R.M. 2000, “Time compression-rapid steel tooling for an ever-changing world,” Materials and Design, vol. 21, pp. 409415. Wu, S. 1982, Polymer Interface and Adhesion, Marcel Dekker, New York. Wypych, G. 2004, Handbook of Plasticizers, Noyes Publications, [On line] Available from www.knovel.com, [June 2004]. 173 APPENDIX A DATA TABLES 174 HDPE at Temperature - Sample 250 RT 200 Load lbs 30 C 150 100 40 C 50 C 50 60 C 0 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3 Extension i/i HIPS at Temperature - Sample 400 350 Room Temp Load lbs 300 250 30 C 200 150 40 C 100 50 C 50 60 C 0 0 0.05 0.1 0.15 0.2 Elongation i/i Type I Specimen - HDPE Thickness 3.09 mm Width 12.35 mm 2 Area 38.22 mm 0.12 in 0.49 in 2 0.06 in Type I Specimen - HIPS Thickness 3.12 mm Width 12.76 mm 2 Area 39.85 mm Figure A.1: Sample plots from tensile test data. 175 0.12 in 0.50 in 2 0.06 in A.1 Tensile Test Data Table Material HDPE Lutene-H ME9180 HIPS BASF PS495F Temp degC 20.5 20.5 20.5 20.5 20.5 30 30 30 30 30 40 40 40 40 40 40 50 50 50 50 50 60 60 60 60 60 60 70 70 70 70 70 20.5 20.5 20.5 20.5 Sample PERT1 PERT2 PERT3 PERT4 PERT5 PE301 PE302 PE303 PE304 PE305 PE401 PE402 PE403 PE404 PE405 PE406 PE501 PE502 PE503 PE504 PE505 PE601 PE602 PE603 PE604 PE605 PE606 PE701 PE702 PE703 PE704 PE705 PSRT1 PSRT2 PSRT3 PSRT4 30 30 30 40 40 40 50 50 50 60 60 60 PS30A PS30B PS30C PS40A PS40B PS40C PS50A PS50B PS50C PS60A PS60B PS60C Pull Speed mm/min (in/min) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 50(2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) 5(0.2) Slope Slopes Included In Excluded from Analysis Analysis 4469.30 3956.00 3179.90 3783.30 3386.60 2796.60 2054.50 1783.10 2708.60 2832.00 1499.60 1730.50 2233.00 1342.50 1412.80 1654.40 616.06 401.10 677.94 430.26 617.97 377.76 416.34 375.86 235.52 336.33 463.26 200.00 178.57 30871.0 30481.0 31497.0 8011.4 10513.0 10636.0 11191.0 14663.0 14753.0 5874.0 5627.6 176 Modulus Modulus (MPa) (psi) 520 75439 461 66775 370 53675 440 63860 394 57164 326 47205 0 0 315 45720 330 47803 175 25312 201 29210 0 156 22661 164 23847 193 27925 72 10399 0 79 11443 0 72 10431 44 6376 48 7028 44 6344 27 3975 39 5677 54 7820 23 3376 21 3014 0 3447 3403 3517 894 1174 1187 1249 1637 1647 656 0 628 0 0 0 0 499768 493455 509903 129696 170194 172185 181170 237378 238835 95094 91105 0 0 0 Average (MPa) Average (psi) 437 63383 324 46909 178 25791 74 10758 43 6203 22 3195 3455 501042 1085 157359 1511 219128 642 93099 A.2 Modulus Look-up Table Estimate from Modulus Graph HDPE HIPS temp Mpa Mpa 35 252 36 238 37 220 38 209 39 190 40 178 1511 40.1 176.7 1499.9 40.2 175.4 1488.8 40.3 174.1 1477.7 40.4 172.8 1466.6 40.5 171.5 1455.5 40.6 170.2 1444.4 40.7 168.9 1433.3 40.8 167.6 1422.2 40.9 166.3 1411.1 41 165 1400 41.1 163.9 1391 41.2 162.8 1382 41.3 161.7 1373 41.4 160.6 1364 41.5 159.5 1355 41.6 158.4 1346 41.7 157.3 1337 41.8 156.2 1328 41.9 155.1 1319 42 154 1310 42.1 152.6 1302 42.2 151.2 1294 42.3 149.8 1286 42.4 148.4 1278 42.5 147 1270 42.6 145.6 1262 42.7 144.2 1254 42.8 142.8 1246 42.9 141.4 1238 43 140 1230 44 131 1150 45 119 1050 46 110 970 47 100 890 48 90 800 48.1 89.2 792 48.2 88.4 784 48.3 87.6 776 48.4 86.8 768 48.5 86 760 48.6 85.2 752 48.7 84.4 744 48.8 83.6 736 48.9 82.8 728 Estimate from Modulus Graph HDPE HIPS temp Mpa Mpa 49 82 720 49.1 81.2 712.2 49.2 80.4 704.4 49.3 79.6 696.6 49.4 78.8 688.8 49.5 78 681 49.6 77.2 673.2 49.7 76.4 665.4 49.8 75.6 657.6 49.9 74.8 649.8 50 74 642 50.1 73.5 635.8 50.2 73 629.6 50.3 72.5 623.4 50.4 72 617.2 50.5 71.5 611 50.6 71 604.8 50.7 70.5 598.6 50.8 70 592.4 50.9 69.5 586.2 51 69 580 51.1 68.1 573 51.2 67.2 566 51.3 66.3 559 51.4 65.4 552 51.5 64.5 545 51.6 63.6 538 51.7 62.7 531 51.8 61.8 524 51.9 60.9 517 52 60 510 52.1 59.9 504 52.2 59.8 498 52.3 59.7 492 52.4 59.6 486 52.5 59.5 480 52.6 59.4 474 52.7 59.3 468 52.8 59.2 462 52.9 59.1 456 53 59 450 53.1 58.8 445 53.2 58.6 440 53.3 58.4 435 53.4 58.2 430 53.5 58 425 53.6 57.8 420 53.7 57.6 415 53.8 57.4 410 53.9 57.2 405 177 Estimate from Modulus Graph HDPE HIPS temp Mpa Mpa 54 57 400 54.1 56.7 391 54.2 56.4 382 54.3 56.1 373 54.4 55.8 364 54.5 55.5 355 54.6 55.2 346 54.7 54.9 337 54.8 54.6 328 54.9 54.3 319 55 54 310 55.1 53.8 307 55.2 53.6 304 55.3 53.4 301 55.4 53.2 298 55.5 53 295 55.6 52.8 292 55.7 52.6 289 55.8 52.4 286 55.9 52.2 283 56 52 280 56.1 51.8 275 56.2 51.6 270 56.3 51.4 265 56.4 51.2 260 56.5 51 255 56.6 50.8 250 56.7 50.6 245 56.8 50.4 240 56.9 50.2 235 57 50 230 58 48 170 59 45 120 60 43 100 61 41 70 62 39 60 63 36 30 64 34 65 32 A.3 Thermal Analysis Convergence Table P20 and HDPE Time sec 0 3.39E-03 6.77E-03 1.02E-02 1.35E-02 1.69E-02 2.03E-02 2.37E-02 3.05E-02 3.73E-02 4.40E-02 5.08E-02 5.76E-02 7.11E-02 8.47E-02 9.82E-02 1.12E-01 1.25E-01 1.39E-01 1.66E-01 1.93E-01 2.20E-01 2.47E-01 2.74E-01 3.01E-01 3.29E-01 3.56E-01 3.83E-01 4.37E-01 4.91E-01 5.45E-01 6.00E-01 6.54E-01 7.08E-01 7.62E-01 8.16E-01 8.70E-01 9.25E-01 9.79E-01 1.08727 1.19566 1.30405 1.41244 1.52082 1.7376 1.95438 2.17115 2.60471 3.03826 3.47181 4.33892 5.33892 6.33892 7.33892 8.33892 9.33892 10.3389 11.3389 12.3389 13.3389 14.3389 15.3389 16.3389 17.3389 18.3389 19.3389 20.3389 P20 and HIPS Thermocouple Peak Polymer Reading Temperature C 50 50.0009 50.0039 50.0114 50.0256 50.0484 50.0807 50.1228 50.2409 50.3853 50.5482 50.7229 50.9043 51.2692 51.623 51.9591 52.2751 52.5704 52.8457 53.3268 53.7491 54.1225 54.4549 54.7528 55.0215 55.2652 55.4872 55.6904 56.0362 56.3323 56.5867 56.8053 56.9931 57.154 57.2911 57.4074 57.5051 57.5864 57.6531 57.7399 57.7896 57.8099 57.8069 57.7859 57.6994 57.5874 57.463 57.2144 56.9887 56.7936 56.5142 56.2969 56.1476 56.0401 55.9575 55.8894 55.8297 55.7748 55.7225 55.6718 55.622 55.5727 55.5238 55.4753 55.4269 55.3789 55.331 C 210 209.958 210.002 210.054 210.087 210.096 210.085 210.06 210.001 209.954 209.924 209.906 209.887 209.755 209.465 208.977 208.279 207.371 206.267 203.43 200.123 196.485 192.631 188.652 184.616 180.576 176.567 172.619 165.075 157.939 151.228 144.941 139.062 133.574 128.453 123.678 119.226 115.074 111.203 104.436 98.518 93.338 88.8004 84.8222 78.5953 73.708 69.8605 64.8203 61.4853 59.2533 56.9357 55.5814 54.8542 54.4353 54.1744 53.9982 53.8696 53.7694 53.687 53.6164 53.5541 53.4978 53.4461 53.398 53.3528 53.31 53.2692 Time sec 0 0.007719 0.015438 0.023156 0.030875 0.038594 0.046313 0.054032 0.06175 0.077188 0.092625 0.108063 0.123501 0.138938 0.169813 0.200688 0.231563 0.262439 0.293314 0.324189 0.385939 0.447689 0.50944 0.57119 0.63294 0.69469 0.756441 0.818191 0.879941 0.941692 1.00344 1.12694 1.25044 1.37394 1.49744 1.62094 1.74444 1.86795 1.99145 2.11495 2.23845 2.36195 2.60895 2.85595 3.10295 3.34995 3.59695 3.84395 4.33796 4.83196 5.32596 6.31396 7.30197 8.30197 9.30197 10.302 11.302 12.302 13.302 14.302 15.302 16.302 17.302 18.302 19.302 20.302 21.302 Sintered and HDPE Thermocouple Peak Polymer Reading Temperature C 50 50.0074 50.031 50.0758 50.1419 50.2262 50.3247 50.4331 50.5479 50.7851 51.0197 51.246 51.4614 51.6647 52.0251 52.3443 52.6277 52.8806 53.1078 53.313 53.6575 53.9519 54.2074 54.4321 54.632 54.8115 54.974 55.122 55.2575 55.3821 55.497 55.695 55.8664 56.0149 56.1436 56.2551 56.3514 56.4344 56.5057 56.5665 56.6183 56.6619 56.7238 56.766 56.7928 56.8077 56.8137 56.8129 56.7958 56.7707 56.742 56.6856 56.6351 56.5907 56.5522 56.5182 56.4875 56.4591 56.4322 56.4062 56.3809 56.3559 56.3312 56.3065 56.2819 56.2572 56.2326 C 221 220.954 220.995 221.049 221.088 221.104 221.099 221.078 221.048 220.987 220.942 220.914 220.897 220.88 220.764 220.5 220.051 219.4 218.545 217.495 214.75 211.506 207.897 204.036 200.014 195.905 191.762 187.628 183.534 179.502 175.548 168.021 160.901 154.195 147.895 141.986 136.449 131.265 126.412 121.87 117.62 113.642 106.646 100.491 95.0735 90.304 86.1032 82.4017 76.5642 71.9549 68.3107 63.5261 60.3707 58.2612 56.8584 55.9176 55.28 54.8421 54.5364 54.3188 54.1601 54.0413 53.9497 53.8769 53.8171 53.7667 53.723 Time sec 0 0.003387 0.006774 0.01016 0.013547 0.016934 0.020321 0.023708 0.030481 0.037255 0.044029 0.050802 0.057576 0.071123 0.08467 0.098218 0.111765 0.125312 0.138859 0.165954 0.193048 0.220143 0.247237 0.274332 0.301426 0.328521 0.355615 0.38271 0.436899 0.491088 0.545277 0.599466 0.653654 0.707843 0.762032 0.816221 0.87041 0.924599 0.978788 1.08717 1.19554 1.30392 1.4123 1.52068 1.73743 1.95419 2.17095 2.60446 3.03797 3.47148 4.3385 5.3385 6.3385 7.3385 8.3385 9.3385 10.3385 11.3385 12.3385 13.3385 14.3385 15.3385 16.3385 17.3385 18.3385 19.3385 20.3385 178 Sintered and HIPS Thermocouple Peak Polymer Reading Temperature C 50 50.0011 50.005 50.0146 50.0324 50.0602 50.0989 50.1486 50.2838 50.4457 50.6253 50.8155 51.0109 51.398 51.7692 52.1191 52.4457 52.7493 53.0311 53.5207 53.9488 54.3259 54.6608 54.9605 55.2303 55.4747 55.6973 55.9007 56.2464 56.5421 56.7957 57.0134 57.2001 57.3597 57.4956 57.6104 57.7067 57.7866 57.8519 57.9359 57.9831 58.0011 57.9962 57.9735 57.8848 57.7719 57.6476 57.4023 57.1819 56.9929 56.7253 56.5184 56.3756 56.2713 56.1893 56.1201 56.0579 55.9997 55.9436 55.8888 55.8348 55.7811 55.7279 55.6749 55.6222 55.5698 55.5177 C 210 209.958 210.002 210.054 210.087 210.096 210.085 210.06 210.001 209.954 209.924 209.906 209.887 209.756 209.465 208.978 208.28 207.373 206.27 203.434 200.128 196.492 192.641 188.664 184.63 180.592 176.585 172.639 165.099 157.967 151.26 144.975 139.1 133.615 128.497 123.725 119.275 115.125 111.256 104.493 98.5779 93.4006 88.8653 84.8891 78.6654 73.7805 69.9351 64.8984 61.5669 59.3384 57.0278 55.6808 54.9592 54.5444 54.286 54.1109 53.9824 53.8816 53.7982 53.7262 53.6622 53.6041 53.5505 53.5005 53.4533 53.4085 53.3658 Time sec 0 7.72E-03 1.54E-02 2.32E-02 3.09E-02 3.86E-02 4.63E-02 5.40E-02 6.17E-02 7.72E-02 9.26E-02 1.08E-01 1.23E-01 1.39E-01 1.70E-01 2.01E-01 2.32E-01 2.62E-01 2.93E-01 3.24E-01 3.86E-01 4.48E-01 5.09E-01 5.71E-01 6.33E-01 6.95E-01 7.56E-01 8.18E-01 8.80E-01 9.42E-01 1.00337 1.12686 1.25035 1.37384 1.49734 1.62083 1.74432 1.86781 1.9913 2.11479 2.23829 2.36178 2.60876 2.85574 3.10273 3.34971 3.59669 3.84368 4.33764 4.83161 5.32558 6.31351 7.30144 8.30144 9.30144 10.3014 11.3014 12.3014 13.3014 14.3014 15.3014 16.3014 17.3014 18.3014 19.3014 20.3014 21.3014 Thermocouple Reading C 50 50.0091 50.0373 50.0894 50.1647 50.2591 50.3676 50.4857 50.6094 50.8615 51.108 51.3439 51.567 51.7766 52.1456 52.4706 52.7583 53.0143 53.2436 53.4505 53.7972 54.0932 54.35 54.5758 54.7767 54.9572 55.1207 55.2696 55.4061 55.5317 55.6476 55.8476 56.021 56.1716 56.3024 56.416 56.5144 56.5996 56.6731 56.7363 56.7903 56.8364 56.9033 56.9507 56.9828 57.0031 57.0144 57.0189 57.0119 56.996 56.9755 56.9315 56.8898 56.8514 56.8165 56.7844 56.7543 56.7256 56.6977 56.6704 56.6433 56.6164 56.5896 56.5628 56.5359 56.5091 56.4821 Peak Polymer Temperature C 221 220.954 220.995 221.049 221.088 221.104 221.099 221.078 221.048 220.987 220.942 220.914 220.897 220.88 220.764 220.5 220.052 219.401 218.546 217.497 214.753 211.51 207.902 204.042 200.022 195.914 191.773 187.641 183.548 179.518 175.566 168.041 160.924 154.221 147.924 142.017 136.483 131.3 126.45 121.91 117.662 113.686 106.694 100.543 95.129 90.363 86.1656 82.4674 76.6362 72.0331 68.3946 63.6202 60.4733 58.3708 56.9734 56.0368 55.4022 54.9664 54.662 54.4451 54.2867 54.1678 54.0759 54.0025 53.9421 53.891 53.8464 A.3 Thermal Analysis Convergence Table (continued) P20/SLA and HDPE Time sec 0 0.001915 0.003831 0.005746 0.007661 0.009576 0.011492 0.013407 0.017237 0.021068 0.024898 0.028729 0.03639 0.044051 0.051712 0.059373 0.074695 0.090017 0.10534 0.120662 0.135984 0.166628 0.197272 0.227916 0.258561 0.289205 0.319849 0.381137 0.442426 0.503714 0.565003 0.626291 0.68758 0.748868 0.810156 0.932733 1.05531 1.17789 1.30046 1.42304 1.54562 1.79077 2.03593 2.28108 2.52623 3.01654 3.50685 3.99716 4.48746 5.46808 6.44869 7.44869 8.44869 9.44869 10.4487 11.4487 12.4487 Thermocouple Peak Polymer Reading Temperature C 30 29.9969 29.9948 29.9934 29.9927 29.9926 29.9928 29.9934 29.9954 29.9978 30.0004 30.003 30.0074 30.0106 30.0127 30.0138 30.0129 30.0104 30.0072 30.0041 30.0015 29.9989 29.9986 29.9997 30.0014 30.0032 30.0048 30.0063 30.0068 30.0068 30.007 30.0078 30.0094 30.0118 30.0151 30.0259 30.0422 30.0655 30.0978 30.1407 30.1958 30.3603 30.5795 30.8496 31.1638 31.903 32.6928 33.4856 34.2472 35.5264 36.5324 37.292 37.8225 38.1658 38.361 38.442 38.4366 C 177 176.942 176.946 176.979 177.022 177.061 177.092 177.112 177.112 177.087 177.05 177.011 176.958 176.935 176.929 176.921 176.788 176.471 175.946 175.215 174.294 171.891 169.106 166.084 162.933 159.733 156.537 150.385 144.573 139.137 134.079 129.382 125.024 120.978 117.218 110.656 104.896 99.8088 95.2896 91.2527 87.6279 81.631 76.6035 72.3389 68.6838 63.0562 58.6276 55.0774 52.1857 48.0636 45.0333 42.7113 40.9253 39.5234 38.4031 37.4934 36.7442 Time sec 13.4487 14.4487 15.4487 16.4487 17.4487 18.4487 19.4487 20.4487 21.4487 22.4487 23.4487 24.4487 25.4487 26.4487 27.4487 28.4487 29.4487 30.4487 31.4487 32.4487 33.4487 34.4487 35.4487 36.4487 37.4487 38.4487 39.4487 40.4487 41.4487 42.4487 43.4487 44.4487 45.4487 46.4487 47.4487 48.4487 49.4487 50.4487 51.4487 52.4487 53.4487 54.4487 55.4487 56.4487 57.4487 58.4487 59.4487 60 P20/SLA and HIPS Thermocouple Peak Polymer Reading Temperature C 38.3671 38.2508 38.1014 37.9291 37.7419 37.5457 37.3451 37.1432 36.9426 36.7451 36.5517 36.3635 36.181 36.0045 35.8343 35.6703 35.5126 35.3611 35.2155 35.0758 34.9417 34.813 34.6895 34.5711 34.4573 34.3481 34.2433 34.1425 34.0457 33.9527 33.8632 33.7771 33.6942 33.6144 33.5375 33.4634 33.392 33.3231 33.2565 33.1923 33.1303 33.0704 33.0124 32.9564 32.9022 32.8497 32.7989 32.7714 C 36.1194 35.5922 35.1429 34.7562 34.4207 34.1274 33.869 33.6399 33.4357 33.2526 33.0877 32.9383 32.8025 32.6786 32.5651 32.4607 32.3645 32.2754 32.1928 32.116 32.0444 31.9775 31.9149 31.8561 31.8008 31.7488 31.6996 31.6532 31.6093 31.5676 31.5281 31.4905 31.4548 31.4207 31.3882 31.3571 31.3275 31.2991 31.2719 31.2458 31.2208 31.1968 31.1737 31.1515 31.1301 31.1095 31.0897 31.079 179 Time sec 0 3.73E-03 7.46E-03 1.12E-02 1.49E-02 1.87E-02 2.24E-02 2.61E-02 3.36E-02 4.11E-02 4.85E-02 5.60E-02 6.34E-02 7.84E-02 9.33E-02 1.08E-01 1.23E-01 1.38E-01 1.68E-01 1.98E-01 2.28E-01 2.58E-01 2.87E-01 3.17E-01 3.47E-01 4.07E-01 4.67E-01 5.26E-01 5.86E-01 6.46E-01 7.05E-01 7.65E-01 8.85E-01 1.00396 1.12339 1.24282 1.36225 1.48168 1.60111 1.72054 1.83997 1.9594 2.19827 2.43713 2.67599 2.91485 3.15371 3.39257 3.63143 4.10915 4.58687 5.0646 5.54232 6.02004 6.97548 7.93092 8.93092 Thermocouple Peak Polymer Reading Temperature C 33 32.9975 32.9964 32.9962 32.9967 32.9977 32.999 33.0005 33.0035 33.0061 33.0083 33.0098 33.0107 33.0106 33.0092 33.0073 33.0051 33.0031 33.0008 33.0001 33.0005 33.0016 33.003 33.0043 33.0054 33.0064 33.0067 33.0068 33.0071 33.0079 33.0094 33.0116 33.0194 33.0316 33.0492 33.074 33.1075 33.1513 33.2069 33.2752 33.357 33.4529 33.7007 34.0014 34.3499 34.7404 35.166 35.6202 36.0965 37.0936 38.0999 39.0884 40.0395 40.9398 42.4921 43.7884 44.881 C 210 209.931 209.922 209.946 209.986 210.03 210.071 210.105 210.138 210.14 210.119 210.084 210.044 209.976 209.934 209.916 209.912 209.913 209.841 209.64 209.277 208.74 208.03 207.158 206.137 203.634 200.791 197.713 194.482 191.163 187.804 184.444 177.887 171.589 165.596 159.924 154.575 149.54 144.803 140.348 136.159 132.216 125.19 118.91 113.279 108.214 103.642 99.504 95.7464 89.4384 84.097 79.5367 75.6134 72.214 66.912 62.7018 59.1744 Time sec 9.93092 10.9309 11.9309 12.9309 13.9309 14.9309 15.9309 16.9309 17.9309 18.9309 19.9309 20.9309 21.9309 22.9309 23.9309 24.9309 25.9309 26.9309 27.9309 28.9309 29.9309 30.9309 31.9309 32.9309 33.9309 34.9309 35.9309 36.9309 37.9309 38.9309 39.9309 40.9309 41.9309 42.9309 43.9309 44.9309 45.9309 46.9309 47.9309 48.9309 49.9309 50.9309 51.9309 52.9309 53.9309 54.9309 55.9309 56.9309 57.9309 58.9309 59.9309 60 Thermocouple Peak Polymer Reading Temperature C 45.7366 46.3867 46.8622 47.192 47.4015 47.5125 47.5437 47.5107 47.4265 47.3018 47.1453 46.9643 46.7648 46.5514 46.3281 46.0981 45.864 45.6277 45.391 45.1552 44.9213 44.6902 44.4624 44.2386 44.0191 43.8041 43.5939 43.3886 43.1882 42.9927 42.8023 42.6168 42.4361 42.2603 42.0892 41.9227 41.7607 41.6031 41.4498 41.3007 41.1557 41.0145 40.8772 40.7436 40.6136 40.4871 40.3639 40.244 40.1272 40.0135 39.9028 39.8951 C 56.3059 53.9398 51.963 50.2924 48.8659 47.6362 46.5674 45.6312 44.8054 44.0724 43.4179 42.8306 42.3009 41.8211 41.3847 40.9862 40.621 40.2854 39.9758 39.6897 39.4243 39.1777 38.9479 38.7334 38.5327 38.3446 38.1679 38.0017 37.8451 37.6973 37.5576 37.4254 37.3001 37.1812 37.0682 36.9607 36.8584 36.7608 36.6677 36.5788 36.4938 36.4124 36.3344 36.2597 36.1881 36.1193 36.0531 35.9896 35.9284 35.8695 35.8128 35.8089 A.3 Thermal Analysis Convergence Table (continued) SLA and HDPE Time sec 0 0.001915 0.003831 0.005746 0.007661 0.009576 0.011492 0.013407 0.017237 0.021068 0.024898 0.028729 0.03639 0.044051 0.059373 0.074695 0.090017 0.105339 0.135983 0.166628 0.197272 0.25856 0.319848 0.381137 0.442425 0.503713 0.62629 0.748866 0.871443 0.994019 1.23917 1.48433 1.72948 1.97463 2.46494 2.95524 3.44555 4.42616 5.40678 6.38739 7.38739 8.38739 9.38739 10.3874 11.3874 12.3874 13.3874 14.3874 15.3874 16.3874 17.3874 18.3874 19.3874 20.3874 21.3874 22.3874 23.3874 24.3874 25.3874 26.3874 27.3874 28.3874 29.3874 30.3874 31.3874 32.3874 33.3874 Thermocouple Peak Polymer Reading Temperature C 30 29.9969 29.9948 29.9934 29.9927 29.9926 29.9928 29.9934 29.9954 29.9978 30.0004 30.003 30.0074 30.0106 30.013 30.0125 30.0102 30.0073 30.0023 29.9996 29.9991 30.0014 30.004 30.0057 30.0065 30.0069 30.0091 30.0144 30.0238 30.0384 30.0998 30.2049 30.3615 30.5729 31.1932 31.9665 32.8439 34.7255 36.5589 38.253 39.7962 41.147 42.3143 43.3146 44.1668 44.8896 45.5007 46.0156 46.448 46.8098 47.1112 47.3607 47.5658 47.7326 47.8666 47.9722 48.0534 48.1134 48.1551 48.1809 48.193 48.1931 48.1828 48.1634 48.1363 48.1022 48.0623 C 177 176.939 176.944 176.978 177.022 177.064 177.096 177.117 177.118 177.092 177.054 177.014 176.961 176.94 176.91 176.784 176.499 176.03 174.47 172.471 170.206 165.42 160.784 156.469 152.525 148.947 142.997 137.993 133.722 130.023 124.133 119.284 115.184 111.644 106.025 101.419 97.5524 91.6781 87.0491 83.2962 80.1313 77.4634 75.178 73.1934 71.45 69.9031 68.5185 67.2699 66.1363 65.101 64.1507 63.2742 62.4624 61.7077 61.0037 60.3449 59.7267 59.1451 58.5965 58.0781 57.587 57.1211 56.6781 56.2563 55.8541 55.47 55.1026 Time sec 44.3874 45.3874 46.3874 47.3874 48.3874 49.3874 50.3874 51.3874 52.3874 53.3874 54.3874 55.3874 56.3874 57.3874 58.3874 59.3874 60.3874 61.3874 62.3874 63.3874 64.3874 65.3874 66.3874 67.3874 68.3874 69.3874 70.3874 71.3874 72.3874 73.3874 74.3874 75.3874 76.3874 77.3874 78.3874 79.3874 80.3874 81.3874 82.3874 83.3874 84.3874 85.3874 86.3874 87.3874 88.3874 89.3874 90.3874 91.3874 92.3874 93.3874 94.3874 95.3874 96.3874 97.3874 98.3874 99.3874 100.387 101.387 102.387 103.387 104.387 105.387 106.387 107.387 108.387 109.387 110.387 Thermocouple Peak Polymer Reading Temperature C 47.3999 47.3293 47.2581 47.1864 47.1143 47.042 46.9695 46.897 46.8245 46.7521 46.6799 46.6079 46.5362 46.4649 46.3939 46.3233 46.2532 46.1835 46.1143 46.0456 45.9775 45.9099 45.8428 45.7763 45.7104 45.645 45.5803 45.5161 45.4525 45.3895 45.3271 45.2653 45.204 45.1434 45.0834 45.0239 44.965 44.9067 44.849 44.7919 44.7353 44.6793 44.6238 44.5689 44.5145 44.4607 44.4074 44.3547 44.3024 44.2507 44.1995 44.1488 44.0986 44.0489 43.9997 43.951 43.9027 43.855 43.8076 43.7608 43.7144 43.6684 43.6229 43.5778 43.5332 43.489 43.4452 C 51.8966 51.663 51.4368 51.2176 51.0051 50.7989 50.5987 50.4043 50.2154 50.0317 49.853 49.6791 49.5097 49.3447 49.1839 49.0271 48.8741 48.7248 48.5791 48.4368 48.2978 48.1619 48.029 47.8991 47.772 47.6476 47.5259 47.4067 47.2899 47.1755 47.0634 46.9535 46.8457 46.7401 46.6364 46.5347 46.4348 46.3368 46.2406 46.1461 46.0532 45.962 45.8724 45.7843 45.6976 45.6125 45.5287 45.4463 45.3652 45.2855 45.207 45.1297 45.0536 44.9787 44.905 44.8323 44.7607 44.6902 44.6208 44.5523 44.4848 44.4183 44.3527 44.288 44.2242 44.1613 44.0993 180 Time sec 124.387 125.387 126.387 127.387 128.387 129.387 130.387 131.387 132.387 133.387 134.387 135.387 136.387 137.387 138.387 139.387 140.387 141.387 142.387 143.387 144.387 145.387 146.387 147.387 148.387 149.387 150.387 151.387 152.387 153.387 154.387 155.387 156.387 157.387 158.387 159.387 160.387 161.387 162.387 163.387 164.387 165.387 166.387 167.387 168.387 169.387 170.387 171.387 172.387 173.387 174.387 175.387 176.387 177.387 178.387 179.387 180.387 181.387 182.387 183.387 184.387 185.387 186.387 187.387 188.387 189.387 190.387 Thermocouple Reading Peak Polymer Temperature C 42.873 42.8349 42.7971 42.7596 42.7225 42.6857 42.6492 42.613 42.5772 42.5416 42.5064 42.4715 42.4368 42.4025 42.3684 42.3347 42.3012 42.268 42.2351 42.2024 42.17 42.1379 42.106 42.0745 42.0431 42.012 41.9812 41.9506 41.9203 41.8902 41.8603 41.8307 41.8013 41.7721 41.7432 41.7144 41.6859 41.6577 41.6296 41.6018 41.5741 41.5467 41.5195 41.4925 41.4656 41.439 41.4126 41.3864 41.3604 41.3345 41.3089 41.2834 41.2581 41.233 41.2081 41.1834 41.1588 41.1344 41.1102 41.0862 41.0623 41.0386 41.0151 40.9917 40.9685 40.9454 40.9225 C 43.3108 43.2596 43.209 43.159 43.1096 43.0608 43.0125 42.9648 42.9176 42.871 42.8249 42.7793 42.7342 42.6896 42.6455 42.6018 42.5587 42.516 42.4737 42.4319 42.3905 42.3496 42.3091 42.269 42.2293 42.19 42.1511 42.1126 42.0745 42.0368 41.9994 41.9624 41.9258 41.8895 41.8535 41.8179 41.7826 41.7477 41.7131 41.6788 41.6448 41.6112 41.5778 41.5448 41.512 41.4796 41.4474 41.4155 41.3839 41.3526 41.3216 41.2908 41.2603 41.2301 41.2001 41.1703 41.1409 41.1116 41.0826 41.0539 41.0254 40.9971 40.9691 40.9412 40.9136 40.8863 40.8591 A.3 Thermal Analysis Convergence Table (continued) SLA and HIPS Time Thermocouple Reading Peak Polymer Temperature sec C C 33 32.9975 32.9964 32.9962 32.9967 32.9977 32.999 33.0005 33.0035 33.0061 33.0083 33.0098 33.0106 33.0099 33.0082 33.0062 33.0028 33.0008 33.0003 33.0009 33.0032 33.005 33.0061 33.0073 33.0098 33.015 33.0239 33.0374 33.0567 33.0834 33.1189 33.235 33.4027 33.6242 33.8989 34.2235 34.5935 35.4695 36.4382 37.4618 38.5097 39.558 41.5477 43.3751 45.0859 46.5935 47.9085 49.0472 50.0281 50.8695 51.5886 52.2012 52.7212 53.161 53.5313 53.8414 54.0994 54.3121 54.4855 54.6249 54.7345 54.8184 54.8798 54.9216 54.9463 54.956 54.9526 210 209.928 209.918 209.943 209.985 210.031 210.075 210.111 210.146 210.148 210.126 210.09 210.013 209.958 209.929 209.922 209.909 209.819 209.603 209.237 207.925 206.137 203.998 199.048 193.892 188.794 183.889 179.243 174.877 170.788 166.967 160.247 154.31 149.033 144.313 140.066 136.223 129.766 124.261 119.51 115.365 111.714 105.825 100.975 96.7511 93.1789 90.1158 87.4574 85.1256 83.0613 81.2185 79.5614 78.0616 76.6961 75.4464 74.2973 73.236 72.2521 71.3368 70.4824 69.6825 68.9317 68.2251 67.5587 66.9287 66.332 65.7657 0 0.003732 0.007464 0.011197 0.014929 0.018661 0.022393 0.026125 0.03359 0.041054 0.048518 0.055983 0.070911 0.08584 0.100769 0.115698 0.145555 0.175412 0.20527 0.235127 0.294842 0.354557 0.414272 0.533701 0.653131 0.772561 0.89199 1.01142 1.13085 1.25028 1.36971 1.60857 1.84743 2.08629 2.32515 2.56401 2.80287 3.28058 3.7583 4.23602 4.71374 5.19146 6.1469 7.10233 8.10233 9.10233 10.1023 11.1023 12.1023 13.1023 14.1023 15.1023 16.1023 17.1023 18.1023 19.1023 20.1023 21.1023 22.1023 23.1023 24.1023 25.1023 26.1023 27.1023 28.1023 29.1023 30.1023 Time sec 41.1023 42.1023 43.1023 44.1023 45.1023 46.1023 47.1023 48.1023 49.1023 50.1023 51.1023 52.1023 53.1023 54.1023 55.1023 56.1023 57.1023 58.1023 59.1023 60.1023 61.1023 62.1023 63.1023 64.1023 65.1023 66.1023 67.1023 68.1023 69.1023 70.1023 71.1023 72.1023 73.1023 74.1023 75.1023 76.1023 77.1023 78.1023 79.1023 80.1023 81.1023 82.1023 83.1023 84.1023 85.1023 86.1023 87.1023 88.1023 89.1023 90.1023 91.1023 92.1023 93.1023 94.1023 95.1023 96.1023 97.1023 98.1023 99.1023 100.102 101.102 102.102 103.102 104.102 105.102 106.102 107.102 Thermocouple Reading Peak Polymer Temperature C 54.4061 54.3307 54.2533 54.1742 54.0935 54.0117 53.9288 53.8451 53.7608 53.6761 53.591 53.5057 53.4203 53.3349 53.2497 53.1645 53.0797 52.9951 52.9109 52.8271 52.7437 52.6608 52.5785 52.4966 52.4154 52.3347 52.2547 52.1752 52.0964 52.0183 51.9408 51.864 51.7879 51.7124 51.6376 51.5635 51.4901 51.4173 51.3453 51.2739 51.2032 51.1332 51.0638 50.9951 50.9271 50.8598 50.7931 50.727 50.6616 50.5968 50.5327 50.4691 50.4062 50.3439 50.2823 50.2212 50.1607 50.1008 50.0415 49.9827 49.9246 49.8669 49.8099 49.7533 49.6974 49.6419 49.587 181 C 60.9927 60.6557 60.3308 60.0171 59.7141 59.4212 59.1378 58.8634 58.5976 58.34 58.09 57.8474 57.6118 57.3828 57.1603 56.9437 56.733 56.5279 56.328 56.1333 55.9434 55.7582 55.5775 55.4011 55.2288 55.0606 54.8961 54.7354 54.5782 54.4244 54.274 54.1267 53.9825 53.8413 53.703 53.5674 53.4346 53.3043 53.1766 53.0513 52.9283 52.8077 52.6892 52.5729 52.4587 52.3466 52.2364 52.1281 52.0216 51.917 51.8142 51.713 51.6135 51.5156 51.4193 51.3245 51.2312 51.1394 51.049 50.96 50.8723 50.7859 50.7008 50.617 50.5343 50.4529 50.3726 Time sec 121.102 122.102 123.102 124.102 125.102 126.102 127.102 128.102 129.102 130.102 131.102 132.102 133.102 134.102 135.102 136.102 137.102 138.102 139.102 140.102 141.102 142.102 143.102 144.102 145.102 146.102 147.102 148.102 149.102 150.102 151.102 152.102 153.102 154.102 155.102 156.102 157.102 158.102 159.102 160.102 161.102 162.102 163.102 164.102 165.102 166.102 167.102 168.102 169.102 170.102 171.102 172.102 173.102 174.102 175.102 176.102 177.102 178.102 179.102 180.102 181.102 182.102 183.102 184.102 185.102 186.102 187.102 Thermocouple Peak Polymer Reading Temperature C 48.8698 48.822 48.7747 48.7278 48.6813 48.6352 48.5895 48.5442 48.4993 48.4548 48.4107 48.367 48.3236 48.2806 48.238 48.1958 48.1539 48.1124 48.0712 48.0303 47.9898 47.9497 47.9099 47.8703 47.8312 47.7923 47.7538 47.7155 47.6776 47.64 47.6027 47.5657 47.529 47.4925 47.4564 47.4205 47.385 47.3497 47.3146 47.2799 47.2454 47.2112 47.1772 47.1435 47.11 47.0768 47.0439 47.0112 46.9787 46.9465 46.9145 46.8828 46.8513 46.82 46.789 46.7581 46.7275 46.6971 46.667 46.637 46.6073 46.5778 46.5485 46.5193 46.4904 46.4617 46.4332 C 49.3572 49.2916 49.2268 49.1627 49.0995 49.037 48.9753 48.9143 48.854 48.7945 48.7356 48.6774 48.6199 48.5631 48.5069 48.4513 48.3964 48.342 48.2883 48.2351 48.1826 48.1306 48.0791 48.0282 47.9778 47.928 47.8787 47.8299 47.7816 47.7338 47.6865 47.6396 47.5933 47.5474 47.5019 47.4569 47.4124 47.3682 47.3245 47.2812 47.2384 47.1959 47.1538 47.1122 47.0709 47.03 46.9895 46.9493 46.9096 46.8701 46.8311 46.7924 46.754 46.7159 46.6782 46.6409 46.6038 46.5671 46.5307 46.4946 46.4588 46.4234 46.3882 46.3533 46.3187 46.2844 46.2504 A.4 Sample Experimental Data Set, All Runs Date-Time and Part ID 20040304-102314 Set 1 Run 1 Rep 1.xls 20040304-102433 Set 1 Run 1 Rep 2.xls 20040304-102552 Set 1 Run 1 Rep 3.xls 20040304-102716 Set 1 Run 1 Rep 4.xls 20040304-102838 Set 1 Run 1 Rep 5.xls 20040304-103000 Set 1 Run 1 Rep 6.xls 20040304-103124 Set 1 Run 1 Rep 7.xls 20040304-103247 Set 1 Run 1 Rep 8.xls 20040304-103857 Set 1 Run 2 Rep 1.xls 20040304-104021 Set 1 Run 2 Rep 2.xls 20040304-104145 Set 1 Run 2 Rep 3.xls 20040304-104314 Set 1 Run 2 Rep 4.xls 20040304-104449 Set 1 Run 2 Rep 5.xls 20040304-104622 Set 1 Run 2 Rep 6.xls 20040304-104756 Set 1 Run 2 Rep 7.xls 20040304-104929 Set 1 Run 2 Rep 8.xls 20040304-105106 Set 1 Run 3 Rep 1.xls 20040304-105240 Set 1 Run 3 Rep 2.xls 20040304-105416 Set 1 Run 3 Rep 3.xls 20040304-105552 Set 1 Run 3 Rep 4.xls 20040304-105730 Set 1 Run 3 Rep 5.xls 20040304-105908 Set 1 Run 3 Rep 6.xls 20040304-110046 Set 1 Run 3 Rep 7.xls 20040304-110226 Set 1 Run 3 Rep 8.xls 20040304-111421 Set 1 Run 4 Rep 1.xls 20040304-111558 Set 1 Run 4 Rep 2.xls 20040304-111738 Set 1 Run 4 Rep 3.xls 20040304-111917 Set 1 Run 4 Rep 4.xls 20040304-112102 Set 1 Run 4 Rep 5.xls 20040304-112246 Set 1 Run 4 Rep 6.xls 20040304-112433 Set 1 Run 4 Rep 7.xls 20040304-112619 Set 1 Run 4 Rep 8.xls 20040304-113245 Set 1 Run 5 Rep 1.xls 20040304-113431 Set 1 Run 5 Rep 2.xls 20040304-113623 Set 1 Run 5 Rep 3.xls 20040304-113819 Set 1 Run 5 Rep 4.xls 20040304-114011 Set 1 Run 5 Rep 5.xls 20040304-114207 Set 1 Run 5 Rep 6.xls 20040304-114404 Set 1 Run 5 Rep 7.xls 20040304-114603 Set 1 Run 5 Rep 8.xls 20040304-114802 Set 1 Run 6 Rep 1.xls 20040304-115008 Set 1 Run 6 Rep 2.xls 20040304-115216 Set 1 Run 6 Rep 3.xls 20040304-115426 Set 1 Run 6 Rep 4.xls 20040304-115636 Set 1 Run 6 Rep 5.xls 20040304-115851 Set 1 Run 6 Rep 6.xls 20040304-120107 Set 1 Run 6 Rep 7.xls 20040304-120326 Set 1 Run 6 Rep 8.xls 20040304-120544 Set 1 Run 7 Rep 1.xls 20040304-120805 Set 1 Run 7 Rep 2.xls 20040304-121025 Set 1 Run 7 Rep 3.xls 20040304-121247 Set 1 Run 7 Rep 4.xls 20040304-121511 Set 1 Run 7 Rep 5.xls 20040304-121737 Set 1 Run 7 Rep 6.xls 20040304-122003 Set 1 Run 7 Rep 7.xls 20040304-122231 Set 1 Run 7 Rep 8.xls 20040304-123344 Set 1 Run 8 Rep 1.xls 20040304-123610 Set 1 Run 8 Rep 2.xls 20040304-123837 Set 1 Run 8 Rep 3.xls 20040304-124108 Set 1 Run 8 Rep 4.xls 20040304-124339 Set 1 Run 8 Rep 5.xls 20040304-124615 Set 1 Run 8 Rep 6.xls 20040304-124851 Set 1 Run 8 Rep 7.xls 20040304-125129 Set 1 Run 8 Rep 8.xls 20040304-125659 Set 1 Run 9 Rep 1.xls 20040304-125945 Set 1 Run 9 Rep 2.xls 20040304-130231 Set 1 Run 9 Rep 3.xls 20040304-130808 Set 1 Run 9 Rep 4.xls 20040304-131055 Set 1 Run 9 Rep 5.xls 20040304-131343 Set 1 Run 9 Rep 6.xls 20040304-131631 Set 1 Run 9 Rep 7.xls 20040304-131921 Set 1 Run 9 Rep 8.xls Peak Load 46.62883 45.436356 45.279495 50.681374 49.785763 49.105278 49.205067 44.488358 47.062386 48.801445 44.790001 48.562714 44.295425 47.351768 45.579613 44.489647 48.056576 42.174576 44.498764 41.299881 41.452736 43.930874 44.221973 41.120594 42.379463 46.715652 41.865509 41.553627 47.430531 41.915798 44.105892 43.67466 44.537437 44.807816 46.723701 42.515388 40.387726 40.945396 42.459793 43.998917 42.385052 50.552158 44.403934 41.823353 45.487274 45.608482 47.657516 47.201088 44.425724 43.235306 42.173809 44.286232 43.368046 41.627769 48.082321 46.882805 45.507973 41.00259 45.06089 43.363491 44.254284 45.699806 45.69239 42.352139 43.582703 49.186348 45.506218 47.338581 46.109573 44.55669 47.892349 45.624733 Net Max Load 42.19622471 41.00375071 40.84688971 46.24876871 45.35315771 44.67267271 44.77246171 40.05575271 42.62978071 44.36883971 40.35739571 44.13010871 39.86281971 42.91916271 41.14700771 40.05704171 43.62397071 37.74197071 40.06615871 36.86727571 37.02013071 39.49826871 39.78936771 36.68798871 37.94685771 42.28304671 37.43290371 37.12102171 42.99792571 37.48319271 39.67328671 39.24205471 40.10483171 40.37521071 42.29109571 38.08278271 35.95512071 36.51279071 38.02718771 39.56631171 37.95244671 46.11955271 39.97132871 37.39074771 41.05466871 41.17587671 43.22491071 42.76848271 39.99311871 38.80270071 37.74120371 39.85362671 38.93544071 37.19516371 43.64971571 42.45019971 41.07536771 36.56998471 40.62828471 38.93088571 39.82167871 41.26720071 41.25978471 37.91953371 39.15009771 44.75374271 41.07361271 42.90597571 41.67696771 40.12408471 43.45974371 41.19212771 182 Temp at Load 49.884712 49.936985 49.973155 49.772371 49.96387 49.982246 50.054485 50.088226 51.658398 51.695778 51.745364 51.559713 51.123327 51.096231 51.17956 51.206355 50.105169 49.97516 49.960869 49.957606 49.962041 50.061731 50.097313 50.055912 50.518533 50.521381 50.503502 50.593096 50.363425 50.376292 50.316962 50.506229 50.957099 51.223117 51.09591 50.86924 51.022172 51.004598 50.932648 50.915217 51.439566 51.171801 51.05388 50.916729 50.910329 50.807639 50.696154 50.641937 51.049432 51.008214 51.035702 51.00606 50.960645 50.942015 50.965229 50.867827 50.820927 50.47199 50.43103 50.381209 50.405053 50.390062 50.404541 50.394667 50.101435 50.1542 50.120294 50.079656 50.213548 50.190848 50.191024 50.181164 Run Avg Load 43.1437 Run Avg Temp 50.0 41.9340 51.4 38.9119 50.0 39.2725 50.5 38.8644 51.0 41.2073 51.0 39.8276 51.0 39.6841 50.5 41.7920 50.2 A.4 Sample Experimental Data Set, All Runs (continued) Date-Time and Part ID 20040305-091132 Set 1 Run 10 Rep 1.xls 20040305-091217 Set 1 Run 10 Rep 2.xls 20040305-091301 Set 1 Run 10 Rep 3.xls 20040305-091349 Set 1 Run 10 Rep 4.xls 20040305-091433 Set 1 Run 10 Rep 5.xls 20040305-091521 Set 1 Run 10 Rep 6.xls 20040305-091609 Set 1 Run 10 Rep 7.xls 20040305-091658 Set 1 Run 10 Rep 8.xls 20040305-092152 Set 1 Run 11 Rep 1.xls 20040305-092303 Set 1 Run 11 Rep 2.xls 20040305-092421 Set 1 Run 11 Rep 3.xls 20040305-092535 Set 1 Run 11 Rep 4.xls 20040305-092653 Set 1 Run 11 Rep 5.xls 20040305-092816 Set 1 Run 11 Rep 6.xls 20040305-092938 Set 1 Run 11 Rep 7.xls 20040305-093058 Set 1 Run 11 Rep 8.xls 20040305-093332 Set 1 Run 12 Rep 1.xls 20040305-093449 Set 1 Run 12 Rep 2.xls 20040305-093609 Set 1 Run 12 Rep 3.xls 20040305-093730 Set 1 Run 12 Rep 4.xls 20040305-093852 Set 1 Run 12 Rep 5.xls 20040305-094011 Set 1 Run 12 Rep 6.xls 20040305-094129 Set 1 Run 12 Rep 7.xls 20040305-094249 Set 1 Run 12 Rep 8.xls 20040305-094533 Set 1 Run 13 Rep 1.xls 20040305-094652 Set 1 Run 13 Rep 2.xls 20040305-094807 Set 1 Run 13 Rep 3.xls 20040305-094919 Set 1 Run 13 Rep 4.xls 20040305-095033 Set 1 Run 13 Rep 5.xls 20040305-095150 Set 1 Run 13 Rep 6.xls 20040305-095307 Set 1 Run 13 Rep 7.xls 20040305-095422 Set 1 Run 13 Rep 8.xls 20040305-100925 Set 1 Run 14 Rep 1.xls 20040305-101047 Set 1 Run 14 Rep 2.xls 20040305-101219 Set 1 Run 14 Rep 3.xls 20040305-101340 Set 1 Run 14 Rep 4.xls 20040305-101504 Set 1 Run 14 Rep 5.xls 20040305-101632 Set 1 Run 14 Rep 6.xls 20040305-101806 Set 1 Run 14 Rep 7.xls 20040305-101943 Set 1 Run 14 Rep 8.xls 20040305-102447 Set 1 Run 15 Rep 1.xls 20040305-102617 Set 1 Run 15 Rep 2.xls 20040305-102747 Set 1 Run 15 Rep 3.xls 20040305-102919 Set 1 Run 15 Rep 4.xls 20040305-103050 Set 1 Run 15 Rep 5.xls 20040305-103224 Set 1 Run 15 Rep 6.xls 20040305-103358 Set 1 Run 15 Rep 7.xls 20040305-103532 Set 1 Run 15 Rep 8.xls 20040305-103715 Set 1 Run 16 Rep 1.xls 20040305-103855 Set 1 Run 16 Rep 2.xls 20040305-104036 Set 1 Run 16 Rep 3.xls 20040305-104221 Set 1 Run 16 Rep 4.xls 20040305-104407 Set 1 Run 16 Rep 5.xls 20040305-104556 Set 1 Run 16 Rep 6.xls 20040305-104735 Set 1 Run 16 Rep 7.xls 20040305-104916 Set 1 Run 16 Rep 8.xls 20040305-105845 Set 1 Run 17 Rep 1.xls 20040305-110035 Set 1 Run 17 Rep 2.xls 20040305-110236 Set 1 Run 17 Rep 3.xls 20040305-110443 Set 1 Run 17 Rep 4.xls 20040305-110647 Set 1 Run 17 Rep 5.xls 20040305-110854 Set 1 Run 17 Rep 6.xls 20040305-111102 Set 1 Run 17 Rep 7.xls 20040305-111310 Set 1 Run 17 Rep 8.xls 20040305-112848 Set 1 Run 18 Rep 1.xls 20040305-113056 Set 1 Run 18 Rep 2.xls 20040305-113310 Set 1 Run 18 Rep 3.xls 20040305-113522 Set 1 Run 18 Rep 4.xls 20040305-113735 Set 1 Run 18 Rep 5.xls 20040305-113954 Set 1 Run 18 Rep 6.xls 20040305-114208 Set 1 Run 18 Rep 7.xls 20040305-114427 Set 1 Run 18 Rep 8.xls Peak Load 47.874283 46.989723 46.881813 45.736076 45.029144 46.069248 48.469646 42.096375 44.725861 44.587673 42.717182 46.337166 44.815929 45.658592 49.946003 48.852619 51.720848 45.144958 43.349007 51.384926 50.378567 41.728935 48.978333 50.947769 41.8722 43.347034 44.334808 45.608547 49.756042 45.773117 40.27779 48.326248 43.114841 41.929817 40.761803 44.940529 44.202026 45.258545 39.655685 43.660015 43.201645 44.945435 44.624767 41.508282 43.016941 40.242275 48.910042 43.095409 45.12886 45.127762 42.580982 41.723831 44.49881 37.737038 37.382568 47.043259 41.832466 40.071449 43.794182 41.143127 39.029778 46.641609 47.707619 47.969162 39.171555 48.006432 47.755684 39.958481 40.581181 47.601276 44.750404 42.72242 Net Max Load 43.44167771 42.55711771 42.44920771 41.30347071 40.59653871 41.63664271 44.03704071 37.66376971 40.29325571 40.15506771 38.28457671 41.90456071 40.38332371 41.22598671 45.51339771 44.42001371 47.28824271 40.71235271 38.91640171 46.95232071 45.94596171 37.29632971 44.54572771 46.51516371 37.43959471 38.91442871 39.90220271 41.17594171 45.32343671 41.34051171 35.84518471 43.89364271 38.68223571 37.49721171 36.32919771 40.50792371 39.76942071 40.82593971 35.22307971 39.22740971 38.76903971 40.51282971 40.19216171 37.07567671 38.58433571 35.80966971 44.47743671 38.66280371 40.69625471 40.69515671 38.14837671 37.29122571 40.06620471 33.30443271 32.94996271 42.61065371 37.39986071 35.63884371 39.36157671 36.71052171 34.59717271 42.20900371 43.27501371 43.53655671 34.73894971 43.57382671 43.32307871 35.52587571 36.14857571 43.16867071 40.31779871 38.28981471 183 Temp at Load 49.551394 49.373647 49.661218 49.484529 49.625281 49.707837 49.691736 49.545989 50.649783 50.050031 49.423058 49.432578 49.194352 48.901438 48.908986 49.013739 49.580786 49.472202 49.345601 49.332055 49.245178 49.557564 49.720498 49.871513 48.736814 48.674564 48.834702 49.227592 49.39116 49.28816 49.367914 49.558032 50.717812 50.769615 50.150435 50.775567 50.935811 50.822927 50.449383 50.229616 50.501971 50.609038 50.686845 50.747482 50.830825 50.822244 50.772311 50.923402 50.471526 50.413878 50.514117 50.402253 50.351899 50.22589 50.763444 50.892408 51.569878 51.321208 50.669801 50.379514 50.390362 50.404251 50.409625 50.403584 49.835092 50.093459 49.993261 50.178301 50.237394 50.096854 50.242349 50.195285 Run Avg Load 41.7107 Run Avg Temp 49.6 41.5225 49.4 43.5216 49.5 40.4794 49.1 38.5078 50.6 39.2605 50.7 38.2203 50.5 39.0911 50.7 39.3858 50.1 A.5 Sample Experimental Part Dimensions (2 Runs Shown) Part No. 1-1-1 1-1-2 1-1-3 1-1-4 1-1-5 1-1-6 1-1-7 1-1-8 1-2-1 1-2-2 1-2-3 1-2-4 1-2-5 1-2-6 1-2-7 1-2-8 ID pixels 1110 1124 1118.64 1114.4 1116 1114 1120.06 1115.81 1113 1118 1112.99 1115.81 1115 1118 1115.81 1112.99 1115 1117 1115.81 1118.64 1110 1123 1114.4 1117.23 1115 1120 1118.64 1112.99 1115 1119 1114.4 1111.57 1121 1119 1118.64 1120.06 1118 1121 1117.23 1122.89 1123 1121 1117.23 1122.89 1122 1119 1117.23 1120.06 1121 1125 1118.64 1118.64 1120 1121 1122.89 1118.64 1119 1119 1115.81 1122.89 1123 1118 1120.6 1120.06 Average pixels 1116.76 ID Run Avg Run Avg inches inches m 1.151299 1.150522 0.029223 1116.468 1.150997 1114.95 1.149433 1115.45 1.149948 1116.613 1.151147 1116.158 1.150678 1116.658 1.151193 1114.993 1.149477 1119.675 1.154304 1.154781 0.029331 1119.78 1.154412 1121.03 1.155701 1119.573 1.154198 1120.82 1.155485 1120.633 1.155291 1119.175 1.153789 1120.415 1.155067 OD pixels 1197 1204 1199.25 1195.01 1202 1196 1199.25 1197.84 1195 1201 1197.84 1200.67 1198 1202 1200.67 1196.42 1195 1203 1197.84 1199.25 1195 1203 1197.84 1199.25 1198 1204 1200.67 1196.42 1199 1203 1199.25 1195.01 1208 1206 1206.32 1204.91 1204 1206 1203.5 1207.74 1210 1204 1204.91 1206.32 1209 1206 1204.91 1204.91 1205 1209 1206.32 1203.5 1205 1207 1207.74 1203.5 1206 1204 1202.08 1209.15 1209 1205 1206.32 1203.5 184 Average OD Run Avg Run Avg rel ∆ dia thickness pixels inches inches m m/m m 1198.815 1.235892 1.236066 0.031396 0.016648 0.001086 1198.773 1.235848 1198.628 1.235698 1199.273 1.236363 1198.773 1.235848 1198.773 1.235848 1199.773 1.236879 1199.065 1.236149 1206.308 1.243616 1.243190 0.031577 0.013008 0.001123 1205.31 1.242588 1206.308 1.243616 1206.205 1.243510 1205.955 1.243253 1205.81 1.243103 1205.308 1.242585 1205.955 1.243253 A.6 Experimental Data and Calculated Coefficient of Friction (Menges), Run Average -1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1 1 1 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 Modulus at Ejection Ejection Ejection o Force (N) Temp ( C) Temp (Pa) 177.1534 51.0 69000000 183.2899 51.0 69000000 186.5225 51.4 65400000 176.5148 50.5 71500000 185.891 50.2 73000000 172.8689 51.0 69000000 191.9032 50.0 74000000 174.6842 50.5 71500000 173.0801 50.0 74000000 184.6922 49.4 78800000 193.5839 49.5 78000000 173.8771 50.7 70500000 171.2827 50.6 71000000 174.6307 50.7 70500000 175.1881 50.1 73500000 170.0038 50.5 71500000 180.0522 49.1 81200000 185.5291 49.6 77200000 P-20 HIPS -1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1 1 1 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 343.9689 376.3673 401.4804 346.1149 385.5599 408.2617 384.5955 381.943 403.7004 376.9719 395.2322 393.3786 369.1949 390.8133 391.7333 351.6246 394.7653 424.4616 51.0 51.1 51.1 50.8 50.8 51.0 50.4 50.2 50.2 50.5 51.1 51.5 50.6 50.8 50.9 50.4 49.6 48.9 580000000 573000000 573000000 592400000 592400000 580000000 617200000 629600000 629600000 611000000 573000000 545000000 604800000 592400000 586200000 617200000 673200000 728000000 0.003089 0.003377 0.002935 0.003301 0.003565 0.002989 0.003788 0.002896 0.00354 0.003199 0.003444 0.001107 0.002796 0.002337 0.001826 0.002424 0.002791 0.003016 0.001138 0.00114 0.001144 0.001134 0.001146 0.001133 0.001139 0.001131 0.00115 0.001124 0.001138 0.00114 0.001139 0.001137 0.001146 0.001141 0.001137 0.001127 0.553423 0.559850 0.684961 0.511977 0.522718 0.681704 0.473993 0.607589 0.516881 0.563075 0.577328 1.876760 0.628923 0.814746 1.047471 0.675713 0.606239 0.562590 ST-100 HDPE -1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1 1 1 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 182.3003 190.0239 209.7913 177.4434 196.6074 194.3292 196.1544 201.6826 208.7471 173.9534 185.8118 184.4146 172.1205 178.8341 180.8674 170.97 186.8184 184.1281 50.9 50.9 50.3 50.3 50.2 51.4 49.1 50.2 49.6 50.8 50 50.6 50.4 50.2 50.2 50.2 49.3 48.6 69500000 69500000 72500000 72500000 73000000 65400000 81200000 73000000 77200000 70000000 74000000 71000000 72000000 73000000 73000000 73000000 79600000 85200000 0.005901 0.006708 0.007395 0.006204 0.006976 0.007264 0.012154 0.007462 0.006877 0.010785 0.011344 0.010469 0.010469 0.010979 0.010258 0.011443 0.01105 0.010809 0.00104 0.001108 0.001116 0.001045 0.001109 0.001122 0.001093 0.001105 0.001113 0.00103 0.001103 0.001122 0.001031 0.001109 0.001119 0.001055 0.001104 0.001114 1.402205 1.207056 1.150452 1.238469 1.142516 1.196203 0.596783 1.099385 1.159108 0.733839 0.658192 0.725377 0.726403 0.659969 0.707842 0.636309 0.631115 0.589072 Insert/ Thermoplastic P-20 HDPE Tpack Tcool Ppack 185 Relative Change in Dia (m/m) 0.013299 0.012021 0.013008 0.013940 0.012097 0.011691 0.016648 0.011819 0.012125 0.012311 0.011043 0.009859 0.012184 0.010684 0.008880 0.012579 0.011084 0.010279 Thickness (m) 0.001075 0.001109 0.001123 0.00108 0.001114 0.001121 0.001086 0.00111 0.001127 0.00107 0.001118 0.001123 0.001066 0.001114 0.001125 0.001074 0.001115 0.001126 CoF 0.589265 0.653811 0.640622 0.537958 0.619705 0.627208 0.470372 0.610927 0.561318 0.583684 0.659497 0.730678 0.609197 0.682887 0.782797 0.577123 0.588728 0.681465 A.6 Experimental Data and Calculated Coefficient of Friction, Run Average (continued) Insert/ Thermoplastic ST-100 HIPS Tpack Tcool Ppack Modulus at Relative Ejection Ejection Ejection Change in Thickness o Force (N) Temp ( C) Temp (Pa) Dia (m/m) (m) CoF 366.2944 49.9 649800000 0.00575 0.001138 0.282649 389.5081 50 642000000 0.005295 0.001143 0.328955 375.2821 50.4 617200000 0.005225 0.001135 0.33628 375.7127 49.6 673200000 0.005929 0.001146 0.269402 393.5372 49.8 657600000 0.005626 0.001147 0.304293 394.4723 50.5 611000000 0.005295 0.001149 0.348226 366.3106 49.5 681000000 0.006008 0.001139 0.257966 393.5601 49.6 673200000 0.004871 0.001097 0.358898 398.8342 49.6 673200000 0.005179 0.001141 0.328905 363.8594 50.6 604800000 0.00307 0.001153 0.557383 378.138 50.4 617200000 0.002821 0.001151 0.618891 388.0728 49.6 673200000 0.002605 0.001144 0.63481 369.5402 49.7 665400000 0.004337 0.00114 0.368619 360.5764 50.1 635800000 0.003636 0.001152 0.444142 374.3817 49.8 657600000 0.002766 0.001145 0.589803 340.6728 50 642000000 0.004431 0.001155 0.340328 370.5696 49.4 688800000 0.004128 0.001186 0.360363 399.8762 49.8 657600000 0.004506 0.001147 0.385893 -1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1 1 1 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 SL 5170 HDPE -1 -1 -1 1 -1 239.0608 -1 193.2119 53.1 51.2 SL/P-20 HDPE -1 -1 -1 -1 1 1 1 1 -1 -1 1 1 -1 -1 1 1 -1 1 -1 1 -1 1 -1 1 42.2 42 40 40.7 41.3 42 40.5 40.9 SL 5170 HIPS -1 -1 -1 -1 1 1 -1 1334.275 -1 1136.124 1 1512.254 55.2 304000000 0.003412 0.001311 3.217736 52.8 462000000 0.00406 0.001309 1.518087 54.1 391000000 0.001753 0.001308 5.533213 SL/P-20 HIPS -1 -1 -1 -1 1 1 1 1 -1 -1 1 1 -1 -1 1 1 -1 1 -1 1 -1 1 -1 1 40.7 42.6 38.1 41.1 42.2 42.7 40.2 42.1 274.2134 299.6501 258.7609 278.1758 313.3349 297.3824 321.087 317.9256 695.756 826.2838 610.1186 845.0178 770.2405 939.1286 702.2841 892.146 186 58800000 0.009477 0.001138 1.237096 67200000 0.009388 0.001137 0.883788 151200000 154000000 178000000 168900000 161700000 154000000 171500000 166300000 1433300000 1262000000 1675000000 1391000000 1294000000 1254000000 1488800000 1302000000 0.013628 0.01166 0.014343 0.010554 0.011732 0.007237 0.01081 0.006996 0.004444 0.002462 0.00579 0.002444 0.002077 0.001119 0.002788 0.000735 0.001197 0.001205 0.001202 0.0012 0.001189 0.001204 0.001187 0.001206 0.00123 0.001212 0.001217 0.001204 0.001206 0.001202 0.001211 0.001194 0.364731 0.454231 0.276662 0.426762 0.45566 0.727173 0.478619 0.743436 0.291433 0.720073 0.169589 0.677233 0.779572 1.826759 0.458249 2.561299 A.7 Analysis of Variance Tables by Set General Linear Model: P-20 and HDPE EF Set 1 versus Packing Time, Cooling Time, Packing Pressure Factor Packing Cooling Packing Type Levels Values fixed 2 2 6 fixed 3 5 10 15 fixed 3 0 5 10 Analysis of Variance for EF Set 1, using Adjusted SS for Tests Source Packing Cooling Packing Packing*Cooling Packing*Packing Cooling*Packing Packing*Cooling*Packing Error Total DF 1 2 2 2 2 4 4 126 143 Seq SS 3.836 149.135 13.938 36.255 3.204 40.352 106.985 899.771 1253.478 Adj SS 3.836 149.135 13.938 36.255 3.204 40.352 106.985 899.771 Adj MS 3.836 74.568 6.969 18.128 1.602 10.088 26.746 7.141 F 0.54 10.44 0.98 2.54 0.22 1.41 3.75 P 0.465 0.000 0.380 0.083 0.799 0.234 0.006 F 8.84 49.29 5.30 7.32 11.79 16.03 20.01 P 0.004 0.000 0.006 0.001 0.000 0.000 0.000 General Linear Model: P-20 and HIPS EF Set 2 versus Packing Time, Cooling Time, Packing Pressure (One Outlier Removed) Factor Packing Cooling Packing Type Levels Values fixed 2 2 6 fixed 3 5 10 15 fixed 3 0 5 10 Analysis of Variance for EF Set 2, using Adjusted SS for Tests Source Packing Cooling Packing Packing*Cooling Packing*Packing Cooling*Packing Packing*Cooling*Packing Error Total DF 1 2 2 2 2 4 4 125 142 Seq SS 89.91 1059.86 119.56 166.61 245.27 689.01 863.76 1349.12 4583.09 187 Adj SS 95.44 1063.93 114.30 158.03 254.54 691.83 863.76 1349.12 Adj MS 95.44 531.97 57.15 79.01 127.27 172.96 215.94 10.79 A.7 Analysis of Variance Tables by Set (continued) General Linear Model: ST-100 and HDPE EF Set 3 versus Packing Time, Cooling Time, Packing Pressure Factor Packing Cooling Packing Type Levels Values fixed 2 2 6 fixed 3 5 10 15 fixed 3 0 5 10 Analysis of Variance for EF Set 3, using Adjusted SS for Tests Source Packing Cooling Packing Packing*Cooling Packing*Packing Cooling*Packing Packing*Cooling*Packing Error Total DF 1 2 2 2 2 4 4 126 143 Seq SS 435.039 146.848 28.902 86.973 6.791 113.575 122.030 1064.275 2004.434 Adj SS 435.039 146.848 28.902 86.973 6.791 113.575 122.030 1064.275 Adj MS 435.039 73.424 14.451 43.487 3.395 28.394 30.508 8.447 F 51.50 8.69 1.71 5.15 0.40 3.36 3.61 P 0.000 0.000 0.185 0.007 0.670 0.012 0.008 F 17.51 2.62 3.76 6.97 3.75 9.00 6.70 P 0.000 0.076 0.026 0.001 0.026 0.000 0.000 General Linear Model: ST-100 and HIPS EF Set 4 versus Packing Time, Cooling Time, Packing Pressure Factor Packing Cooling Packing Type Levels Values fixed 2 2 6 fixed 3 5 10 15 fixed 3 0 5 10 Analysis of Variance for EF Set 4, using Adjusted SS for Tests Source Packing Cooling Packing Packing*Cooling Packing*Packing Cooling*Packing Packing*Cooling*Packing Error Total DF 1 2 2 2 2 4 4 126 143 Seq SS 261.17 78.28 112.18 207.85 111.73 536.86 399.58 1879.43 3587.07 188 Adj SS 261.17 78.28 112.18 207.85 111.73 536.86 399.58 1879.43 Adj MS 261.17 39.14 56.09 103.92 55.87 134.21 99.89 14.92 A.7 Analysis of Variance Tables by Set (continued) Fractional Factorial Fit: SL 5170/P-20 and HDPE EF Set 5b versus Packing Time, Cooling Time, Packing Pressure Estimated Effects and Coefficients for EF (coded units) Term Constant Packing Cooling Packing Packing*Cooling Packing*Packing Cooling*Packing Packing*Cooling*Packing Effect 7.809 -0.485 1.447 3.666 -3.595 0.380 1.057 Coef 66.337 3.904 -0.243 0.723 1.833 -1.798 0.190 0.529 SE Coef 0.9830 0.9830 0.9830 0.9830 0.9830 0.9830 0.9830 0.9830 T 67.49 3.97 -0.25 0.74 1.86 -1.83 0.19 0.54 P 0.000 0.000 0.807 0.467 0.071 0.077 0.848 0.594 Analysis of Variance for EF (coded units) Source Main Effects 2-Way Interactions 3-Way Interactions Residual Error Pure Error Total DF 3 3 1 32 32 39 Seq SS 633.01 265.08 11.18 1236.78 1236.78 2146.06 Adj SS 633.01 265.08 11.18 1236.78 1236.78 Adj MS 211.00 88.36 11.18 38.65 38.65 F 5.46 2.29 0.29 P 0.004 0.098 0.594 Fractional Factorial Fit: SL 5170/P-20 and HIPS EF Set 6a versus Packing Time, Cooling Time, Packing Pressure Estimated Effects and Coefficients for EF (coded units) Term Constant Packing Cooling Packing Packing*Cooling Packing*Packing Cooling*Packing Packing*Cooling*Packing Effect 18.358 -10.220 40.702 -2.700 -0.375 7.045 -4.687 Coef 176.511 9.179 -5.110 20.351 -1.350 -0.188 3.523 -2.344 SE Coef 2.401 2.401 2.401 2.401 2.401 2.401 2.401 2.401 T 73.50 3.82 -2.13 8.47 -0.56 -0.08 1.47 -0.98 P 0.000 0.001 0.041 0.000 0.578 0.938 0.152 0.336 Analysis of Variance for EF (coded units) Source Main Effects 2-Way Interactions 3-Way Interactions Residual Error Pure Error Total DF 3 3 1 32 32 39 Seq SS 20981.5 570.6 219.7 7381.2 7381.2 29153.1 Adj SS 20981.5 570.6 219.7 7381.2 7381.2 189 Adj MS 6993.8 190.2 219.7 230.7 230.7 F 30.32 0.82 0.95 P 0.000 0.490 0.336 APPENDIX B MOLD AND CANISTER DRAWINGS 190 B.1 Part Drawing 191 B.2 Mold Insert Drawings 192 B.2 Mold Insert Drawings (continued) 193 B.2 Mold Insert Drawings (continued) 194 B.2 Mold Insert Drawings (continued) 195 B.2 Mold Insert Drawings (continued) 196 B.2 Mold Insert Drawings (continued) 197 B.2 Mold Insert Drawings (continued) 198 B.2 Mold Insert Drawings (continued) 199 B.2 Mold Insert Drawings (continued) 200 B.2 Mold Insert Drawings (continued) 201 B.2 Mold Insert Drawings (continued) 202 B.2 Mold Insert Drawings (continued) 203 B.2 Mold Insert Drawings (continued) 204 B.3 Mold Assembly Drawings 205 B.3 Mold Assembly Drawings (continued) 206