Selecting the appropriate rapid prototyping system for an engineering technology program

Selecting the appropriate rapid prototyping system for an engineering technology program

Sriraman, Vedaraman

Abstract

Modern rapid prototyping (RP) systems can be a valuable tool in an engineering technology curriculum. These systems can range in price from basic configurations that cost approximately $8,000 to more extensive platforms priced over $400,000. These systems also differ in suitability for specific curricular applications, ease of use, safety considerations, and the cost of supplies, maintenance, and installation. To effectively teach the various RP applications used in industry, it is essential that educators select the appropriate technology. This paper will outline a methodology for selecting RP systems for educational use, provide a list of RP machines, specifications and manufacturers, and describe the application of the proposed methodology at

Southwest Texas State University. A summary of RP applications at other institutions is also provided for comparison.

Introduction

Competitive pressures faced by manufacturing companies have led to development of a host of tools and procedures designed to drastically reduce time-to-market cycles, including computer-aided design (CAD), computer-aided manufacturing (CAM), computer-aided engineering (CAE), computer numerical control (CNC), investment casting, virtual prototyping, and rapid tooling.1 These tools, commonly known as time-compression technologies, are based on the concept of rapid prototyping, a broad term used to describe several processes that create physical models directly from a CAD database. RP systems use a variety of techniques to form models, including stereolithography and fused deposition modeling (FDM).2 Detailed descriptions of various RP systems, polymer chemistry, CAD-based requirements, and RP applications are available in the literature.3,4 Charles Hull coined the term stereolithography, or three-dimensional (313) printing, which uses an ultraviolet laser to selectively cure layers of photopolymer. Hull patented the process in 1986 and founded 3D Systems, Inc., which in 1987 developed the first commercially available RP system in the world, the SLA-1.5 Approximately nineteen manufacturers worldwide have since developed RP systems, and products from eight of these manufacturers are commonly available in the U.S.6

The growing importance and acceptance of RP technology in the U.S. is evident in the following statistics. The Rapid Prototyping and Tooling: State of the Industry, 1998 Worldwide Progress Report stated that close to 50% of RP installations worldwide are in North America.7 Several of the most widely sold RP machines are from U.S. manufacturers, and of the 5449 machines sold from 1988 to 1999, 4412 or 81% were from U.S. companies.6

Engineering technology educators play a critical role in providing students with meaningful learning experiences in the tecnologies currently used in industry. In the past, technologies and tools such as solid modeling, CNC, CAE, programmable logic controllers, robotics, and coordinate measuring machines have been successfully incorporated into the engineering technology curriculum. The same must be done for RP. However, selecting the appropriate RP system can be quite a challenge, as educators must choose from among nineteen different RP system manufacturers, various processes and materials, and units priced from $8,000 to over $400,000. This paper establishes a process for selecting a RP system that meets educational goals while considering cost, maintenance, ease of use, and safety.

Curriculum Needs

The curriculum needs of a program should be the single most important factor when considering the purchase of a RP machine. In 1998 it was reported that 28% of all RP models were being used for fit and function applications, 36% served as visual aids for engineering, tooling, quotes and presentations, and 25% were used to make patterns for prototype tooling and metal casting. These technological applications can serve as inputs for determining how RP should be integrated into the engineering technology curriculum, but they should also be reconciled with the special characteristics of an academic environment, such as curriculum need, ease of use, safety, purchase price, and operating and maintenance costs. Thus, several criteria must be adequately addressed before incorporating RP into a curriculum.

It is particularly important to consider the purpose of the RP model, which depends upon the curriculum. A two-year program in engineering design graphics may require only a concept RP machine for form and fit verification, while a four-year program in manufacturing engineering technology may need rapid tooling applications. RP selection also may be influenced by the specific discipline within a curriculum. Thus, the curriculum provides the first essential input to the selection process.

Once the curriculum-based needs have been established, the RP system that best suits those needs can be selected. For example, programs in engineering design graphics or mechanical engineering technology will, in general, use RP models for visualization of form and fit and for limited functional testing. If this is the sole or predominant use for RP, there are several basic systems available for less than $50,000, as shown in table 1. The JP-5 systems from Schroff Development, including 3D modeling software, are priced as low as $7,900. These particular systems offer manual stacking and adhesion of the paper layers, so they may not be as fast or as accurate as other processes. If the program under consideration is manufacturing engineering technology, RP models can be applied over the entire product cycle, from design through manufacturing. In this case, RP will also be used to generate rapid tooling for processes such as metal casting, injection molding, and prototyping tooling. However, system accuracy and material selection must be carefully considered, since prototypes are cast from wax or other materials that are burned out of a mold.

The following section describes several popular RP systems. This information can be useful when deciding which system best meets the curriculum needs of a specific program.

Comm”al RP Systems

There are a number of RP systems that are suitable for use in an academic environment. Selected models from the major U.S. manufacturers are listed in two categories: systems priced below $100,000 as shown in table 1, and systems priced more than $100,00 as shown in table 2. Each manufacturer may offer several models. For example, Stratasys, Inc. offers the FDM 2000, FDM 3000, FDM 8000, and the Quantum. System characteristics may vary for each model, such as prototype material, process type, build envelope dimension, accuracy limits, and initial cost. Each of these characteristics warrants closer examination.

BUILD MATERIALS

The materials used to build prototypes can limit a machine’s suitability for certain applications. For example, a foundry program that is implementing rapid tooling in the investment casting process may require a system that can generate prototypes in wax or other materials that will bum out of a ceramic mold easily. An engineering design graphics program that primarily will use prototypes for visualization and form verification may find any prototype material suitable, whether it is paper, wax, acrylonitrile– butadiene-styrene (ABS), nylon, or epoxy.

BUILD ENVELOPE

Another important consideration is the build envelope, which roughly corresponds to the L x W x H dimensions of the largest prototype that may be built on the system. However, because most prototypes are either built on or inside a shell of supporting material, actual prototype size will often be less than the space suggested by the build volume proportions. Most additive processes require a base to support the prototype and structures for overhangs. A subtractive process requires a supporting shell with a wall thickness of at least 1/4″ around the entire model. The build envelope has another consequence as well. For design verification purposes, build times and costs can be greatly reduced by constructing scaled models. However, tooling and manufacturing applications often call for full-scale models.

RP PROCESS AND ACCURACY

RP model accuracy is affected by the unique characteristics of a particular process, such as 3D Systems’ stereolithography (SLA), DTM Corporation’s selective laser sintering (SLS), Stratasys’ fused deposition modeling (FDM), or Cubic Technology’s laminated object manufacturing (LOM). Process characteristics plus the structure and dynamics of the RP machine and its optical and mechanical subsystems determine the accuracy of the model, which in turn determines its appropriateness for an application. For example, models built on Stratasys’ Genisys system and other concept modelers may have an accuracy of +/-.013 inches, which is suitable for form verification purposes. However, for tooling applications, a Stratasys FDM 3000 model with accuracy of +/-.005 inches is more appropriate.

Since no national standard exists for determining the accuracy of RP machines, it is important to discuss this issue with the manufacturer before purchasing a system. Also, accuracies will vary depending on the size and shape of the part being built. In general, the z-axis is often the most inaccurate on RP models, a factor that should be carefully considered during system selection.

MAINTENANCE AND SUPPLIES

All RP machines will require routine maintenance and supplies, and various companies offer convenient, although costly, maintenance contracts. At Southwest Texas State University, departmental technicians were able to perform minor RP system repairs, but major repairs had to be performed by factory technicians at a higher cost. Also, all RP machines use expendable materials, which are costly and normally only available from RP companies, but supply prices are being reduced as secondary vendors enter the market. Southwest Texas State University offset most of their supply and maintenance costs by performing limited prototyping for local industry during noninstructional time periods. However, educational instruction has been and will remain the primary purpose of these systems.

SAFETY

Safety considerations are another essential input to the selection process. Most lower-priced concept modelers, shown in table 1, are designed for use in an office environment. Also known as 3D printers, these basic RP systems are very safe to operate because they use no lasers.

Higher-priced RP systems, shown in table 2, are also safe to operate as long as manufacturer instructions are followed carefully. Properly containing the laser light to avoid eye damage is a key safety concern with laser-based machines. These RP machines have safety interlocks that disable the laser when critical machine panels are open. For example, the Cubic Technology 1015 system has four microswitches that must be in a closed state before the machine can operate: one each on the hood and paper and another two on the loading doors. This feature ensures that the laser beam stays within the machine enclosure during operation.

The temperature of machine parts during prototype construction is another concern. For example, the Cubic Technology system laminates paper layers together with a shielded roller that can exceed 500degF, and the extrusion nozzle on the Stratasys system can reach these same temperatures. Thus, students must receive proper instruction on machine operation and safety, and they must wear appropriate protective gear during system operation.

COST

While cost may not be a significant factor in industry, it is particularly important in academia. Estimated costs listed in tables I and 2 should be used only as a general guideline. Costs can vary considerably depending on system options, service contracts, transportation, installation, and educational discounts.

Case Study

RP initiatives began at Southwest Texas State University in 1992. Since then, with the assistance of six different grants, RP has been successfully integrated into the manufacturing curriculum. The RP systems include a Cubic Technology Model 1015 (operational 1/95), a Stratasys Model FDM 3000 (operational t/00), and a VIPLOM process (operational 12/99), which is used to impregnate LOM models with epoxy, thereby increasing their strength, moisture resistance, and durability.

The first step in RP system selection was to determine curriculum-based needs. Selected courses and their RP requirements are as follows:

TECH 1413 – ENGINEERING DESIGN GRAPHICS

This is a freshman-level design and drafting course in which the primary goal was to introduce basic RP concepts and use models for visualization and form verification.

TECH 2310 – MACHINE DRAFTING

This is a sophomore-level course that involves machine elements, manufacturability, geometric dimensioning and tolerancing, surface finish, and assembly design. While visualization and form verification are related applications, RP models are mostly used to study manufacturability and assembly-related issues. In a final class project, students design and build assembly parts on the RP system and then assemble the product using the prototyped components.

TECH 4330 – FOUNDRY AND HEAT TREATMENT

This is a senior-level course in metal casting with a laboratory portion that involves projects in sand and investment casting. Traditionally, wooden patterns for the sand casting projects have been built in the departmental wood technology laboratory, but this process was time consuming and often inaccurate because students were not always familiar with woodworking operations. Molds for investment wax casting projects are manually prepared out of soft materials such as polyurethane and silicone, but this process is also time consuming and has accuracy and mold durability problems.

Once these course requirements were identified, the strategy for selecting the RP system was developed as follows. In the two drafting courses, the end material was not significant since any prototype could serve for visualization purposes. Also, build envelope size was not a limiting factor because scaled models could be employed. Thus, system selection was primarily dictated by the requirements of the foundry course. For sand casting applications, RP products from 3D Systems, DTM Corporation, and Cubic Technology were considered. The first two were eliminated based on cost, as funding was capped at $200,000. In 1994, the 3D Systems Model SLA 250 cost $250,000 (now $145,000) and the DTM Sinter Station 2000 cost over $400,000 (now the Sinter Station 2500 costs about $250,000). The 3D Systems model also presented issues related to storing and preserving the quality of the photopolymer as well as postcuring requirements, but these problems have largely been eliminated due to the new epoxy resins currently in use.

As a result, the Cubic Technology Model LOM 1015 was selected for RP implementation. This particular system is well suited for applications at Southwest Texas State University because (1) it builds RP models with layers of adhesive-backed paper that eventually have the look, texture and machinability of wooden models, (2) it was very favorably priced in 1994 at $100,000, including a computer, (3) its venting requirements of a 300-500 CFM fan were modest, and (4) storage of the raw material posed no problems. Since 1995, SWT has recorded over 5000 hours use on this RP system, which has supported the sand casting and engineering design graphics activities very well.

Investment casting requires wax patterns. Several RP companies claim their nonwax patterns can be burned out of an investment casting, but often there are unique problems associated with this process, such as higher burnout temperatures, long burnout times, and the need for an oxygen furnace. The investment casting industry routinely uses wax and prefers this material for pattern building. Program administrators at Southwest Texas wanted to improve the efficiency of making investment patterns in two ways: (1) to directly build rapid prototypes in wax, thereby eliminating the need for molds, and (2) to build wax prototypes of molds that could then be cast in aluminum or steel for repeated use. Platforms from 3D Systems as well as the Stratasys Model FDM 3000 were considered for building investment patterns. The 3D Systems platforms have good accuracy, but they did not allow for the variety of build materials available in the Stratasys system, namely ABS, medical-grade ABS (ABSi), elastomer, and casting wax. However, with new epoxy-based resins and cost reductions in their Quick Cast software, 3D Systems machines have become a very competitive alternative for investment casting.

The Cubic Technology and Stratasys machines both serve the needs of the engineering design graphics course. The Cubic Technology machine is generally used for large volume, thick-walled parts, while the Stratasys machine is used for smaller, thinner-walled components or finely detailed parts that require ABS, elastomer, or casting wax as the build material. Lastly, it was found that when LOM patterns were repeatedly used for sand casting, durability problems occasionally would occur because of delamination of the paper layers. Therefore, VIPLOM equipment that impreg-nates and strengthens LOM and other porous models with heat-cured epoxy resin was added to the RP equipment already in place at Southwest Texas State. The VIPLOM process prolongs the life of the tooling, in particular the matchplates used for sand casting.

The following is a description of typical RP models built in these three classes along with the associated learning outcomes. Figure 1 illustrates a slanted block that was generated in the TECH 1413 class by the Stratasys Model FDM 3000 for visualization and form verification. Figure 2 illustrates a prototype of a valve assembly that was generated in the TECH 2310 class by the Cubic Technology Model LOM 1015 and used to study assembly requirements and manufacturability issues. The availability of a prototype during early product development, together with the opportunity to make design changes based on manufacturability concerns, enables students to experience concurrent engineering (CE).

Figure 3 illustrates a matchplate pattern for a brake drum that was generated in the TECH 4330 class by the Cubic Technology Model LOM 1015. This pattern was then used to generate a sand mold in the foundry. Gray cast iron was poured into the mold to produce a cast brake drum, as shown in figure 4. This project was an in-depth exercise in CE because students incorporated manufacturing-related concerns such as draft angles, machining allowances, and shrinkage factors during the solid modeling stage. These concerns typically are overlooked during the design stage and discovered only during manufacture, when changes are costly in terms of time and money.

RP Applications at Other Institutions

At Purdue University-Calumet, RP is primarily used for visualization and design verification in engineering graphics and mechanical design courses.8 Their RP equipment includes 3D Systems’ ThermoJet Printer and Schroff Development’s JP-5 system, both of which do not exceed a cost of $50,000. At the University of Texas-Austin, a key application for RP is in the freshman engineering design graphics course. Their approach to graphics represents a paradigm shift whereby students start the design process by generating 3D CAD models that are then subjected to virtual tests and optimized. Once the virtual testing has concluded, RP models generated by their JP-5 system are used for design verification? The Rapid Prototyping and Manufacturing Institute at Georgia Tech supports the university’s educational and research-related RP initiatives. Since both design and manufacturing fall under the broad purview of this institute, Georgia Tech has a wide assortment of RP systems, including 3D Systems’ Models SLA– 250/50 and Actua 2100 and the Stratasys Model FDM 1650.10 Conclusion

Since beginning its RP initiative nearly seven years ago, Southwest Texas State University has successfully integrated RP into its ET curriculum. The following recommendations are based on experiences from this implementation. First and perhaps most important, curriculum-based prototyping requirements should be determined. These requirements will be the primary factor in equipment selection. Other issues to consider in an academic environment are ease of use, cost, safety, and machine characteristics. Institutions that need RP capabilities on a very limited basis may consider working with service bureaus, eliminating operating costs and other issues related to the purchase, maintenance, and storage of an in-house RP system.

The second recommendation is to use a building block approach in which students are gradually exposed to various design and manufacturing applications of RP as they progress from freshman- to senior-level classes. According to Terry Wohlers, a RP applications expert, RP technology is here to stay.7 Michael Schrage, an engineering management consultant from MIT, affirms that strong prototyping cultures produce strong products.7 ET educators must ensure that their students learn the integrated application of time compression technologies, and selecting the appropriate RP system could very well determine the success of this endeavor.

Acknowledgements

The authors acknowledge support from National Science Foundation grants DUE-9950816 and DUE-9551467.

References

1. Graver, T. W., L. F. McGinnis, and D. W. Rosen. Engaging Industry in Lab-Based Manufacturing Education: RPM at Georgia Tech. Atlanta: RPMI, 1997.

2. Bertoline, G., and E. Wicke. Fundamentals of Graphics Communications. New York: McGraw Hill Company, 2001.

3. Winek, G., and V. Sriraman. “Rapid Prototyping: The State of the Technology.” Journal of Engineering Technology 12, no. 2 (Fall 1995): 38-43.

4. Sriraman, V., G. Winek, and R. Habingreither. “Tooling Applications of Rapid Prototyping.” Journal of Engineering Technology 16, no. 1 (Spring 1999): 34-38.

5. Jacobs, P. F. Rapid Pro to typing & Manufacturing– Fundamentals of Stereolithography. Dearborn, Mich.: Society of Manufacturing Engineers, 1992.

6. Wohlers, T. T. Rapid Prototyping & Tooling-State of the Industry: Annual Worldwide Progress Report. Fort Collins, Col.: Wohlers Associates, Inc., 2000.

7. Wohlers, T. T. Rapid Prototyping & Tooling-State of the Industry: 1998 Worldwide Progress Report. Fort Collins, Col.: Wohlers Associates, Inc., 1998.

8. Higley, J. Personal communication, August 27, 2001.

9. Barr, R. Personal communication, August 27, 2001.

10. Georgia Institute of Technology, Rapid Prototyping and Manufacturing Institute. RPM] Equipment. Retrieved October 3, 1998, from http://rpmi.marc.gatech.edu/about/equip/htm.

Vedaraman Sriraman * John DeLeon * Gary Wick

Dr. Sriraman is an associate professor of Technology and the program coordinator of the Manufacturing Engineering program at Southwest Texas State University. He has presented at and published in several American Society for Engineering Educators (ASEE) conferences and journals. He has also secured numerous grants from sources such as the National Science Foundation and the Society of Manufacturing Engineers (SME). Dr. Sriraman is also the faculty advisor to the SME student chapter. His teaching and research interests include the areas of CAD, rapid prototyping, quality assurance, automation, and manufacturing systems.

Dr. DeLeon is an associate professor of Technology and the assistant dean of the University College at Southwest Texas State University. He has presented at and published in several National Association of Industrial Technology (NAIT) and ASEE conferences and publications. He has secured several internal and external grants to help update technology laboratories and curriculum. His teaching and research interests involve CAD, rapid prototyping, quality assurance, and industrial safety.

Dr. Gary Winek is a professor of Technology and head of the Construction program at Southwest Texas State University. He was instrumental in securing the first NSF grant at Southwest Texas State for a Cubic Technology RP machine, which became operational in January of 1995. Since then, he has co-authored a second NSF grant for a Stratasys RP machine as well as published numerous articles and given presentations on the topic. His teaching and research interests include construction management and industrial design and manufacturing.

Copyright American Society for Engineering Education Spring 2002

Provided by ProQuest Information and Learning Company. All rights Reserved