9308492 Barton Computer simulation models are the laboratory for many today's design engineers. Circuits are simulated before they are prototyped, exhaust systems are represented as finite element simulations before they are built, and factories are modeled with discrete event simulation before the ground is broken for a new plant. "Laboratory" tools for: (1) modeling random manufacturing variations, (2) for identifying key design parameters, (3) for developing and displaying simulation metamodels and (4) for aiding in robust design optimization exist in engineering research. Yet they have not been transferred to engineering practice. A new graduate course, "Using Simulation Models for Engineering Design," presents recent research results in simulation sensitivity analysis, metamodeling, and optimization for use in engineering design case studies. The case studies are based on computer aided design models that have been developed and used in industry. The department advisory group assessed the initial offering as a significant contribution to a revised engineering curriculum, but cited these key shortcomings: (1) the students spend unnecessary time programming because software tools to exercise the case study software are lacking, and (2) refinements and additions to the lecture notes would make it easier to understand and implement these tools. This research to overcome these shortcomings.
