ROBUST DESIGN: THE FUTURE OF ENGINEERING ANALYSIS IN DESIGN by

Computational capability today is a commodity making engineering analysis affordable at all stages of design. The ability to do more engineering analysis provides the promise of moving engineering analysis into the design concept phase of product development from the design validation phase where most analyses are now conducted. Some call this simulation based design, with a number of new tools focused on providing this capability (1.). Intuitively, better engineering analysis can produce better and more affordable products. However, more engineering analysis does necessarily not mean better engineering, or better products. How can you tell if: • sufficient engineering analysis has been conducted to ensure expected performance over the product lifetime. • the product has been over-engineered, resulting in unnecessarily high cost. • the product has been under-engineered resulting in safety risks and increased warranty claims. • the engineering analysis has considered the most important customer factors. • there are aspects of the design that will not be learned until test and operation. A way to ensure better engineering analysis is to follow one or more design exploration methods to systematically explore questions about a product design. Design exploration methods have been well developed over the past century. These methods - Taguchi analysis, design of experiments, Monte Carlo simulations, sensitivity analysis, design scans - provide the means to coherently and comprehensively explore a design, and understand the trade-offs between cost and performance. The integration of these design exploration methods with traditional engineering analysis is the next generation of engineering simulation in design. This paper and presentation provides a perspective on how this can be accomplished.