An Interactive Virtual System for Simulation and Optimization of Rapid Prototyping

Abstract

The paper describes the development of a computer system for simulation and optimization of rapid prototyping (RP) processes. The system provides a test-bed for virtual prototyping by integrating product design and RP with simulation and realistic visualization techniques. It enhances the dimensional accuracy and reduces the build-time of product prototypes. The virtually fabricated parts may be exported in VRML format over the Internet for effective communication between the manufacturer and the customer. The designer may use the system to design-build-break as many parts as required at a relatively low cost andin a short period of time. Therefore, virtual simulation of RP processes facilitates tuning of the control parameters according to the requirements, and hence reduces the number of physical prototypes needed to produce a part. 1.0 Introduction Innovative products and shorter time-ta-market cycles are essential for the success of every company. A typical product development cycle (concept->design->evalutaion>redesign) consists of several sub-processes, such as prototyping, si~ulation and optimization. An important way to reduce the product development cycle time i80to .accelerate the prototyping processes. This is often achieved by using rapid. prototyping (Rf) techniques. A typical RP process, as shown in. Fig. includes .pre-processing,physical p~pcessing and post processing stages. The pre-processing stage involves creation of 3D CAli;; models and process planning. The process planning stage· generates appropriate laser path signals based on the control parameters specified by the designer. The physical processing stage drives the laser to build the part in layers. The method of achieving this differs amongst various processes. The post-processing comprises a few operations, which include removal of support structure, and curing and cleaning and of the part, as needed for the RP process. The process planning stage is almost identical to all the RP processes. However, it significantly influences the quality of the part, which can be measured by the accuracy, buildtime, efficiency and strength. It involves part orientation, slicil1gofthe CAD model and laser path generation. For liquid-based processes,support structures are required. All these steps affect the part .quality. The orientation of an RP part affects its accuracy, build-time and strength. Forexalnple, orienting the part with minimum z-height possible will result in fewer slices, and hence a reduction in build-time. Layer thickness. affects the accuracy and buildtime. The accuracy may be improved when the part is built with a smaller thick.ness,but the build-time will increase inversely. Hatch spacing refers to the distance between the laser paths. It affects the build-time and strength. Typically, the user has to determine theseproc~ss planning parameters. Thus the quality of the part depends on the users'experiengys,which are often inconsistent. To overcome this problem, several semior fully automatic algorithms have been proposedto optimise the process planningtoehhancepart quality[1-4]. However, most of these methods choose only a single criterion as the objective function, while other