Engineering analysis methods, such as the finite element method, are employed extensively to optimize complex engineering designs, but their success in conceptual product development is rather limited since numerous designs must be analyzed to cover the design space, and unfortunately, modern analysis methods can be tedious and time consuming in such scenarios. We propose here a novel analysis methodology for conceptual design wherein, given the simulation results and performance of one of the designs, one predicts upper and lower bounds on the performance of geometrically similar designs. The methodology rests on sound mathematical principles such as adjoint theory of boundary value problems, and is partly motivated by recent work on shape similarity exploitation in manufacturing wherein the cost of manufacturing a new part is estimated by retrieving the manufacturing costs of geometrically similar parts.

This content is only available via PDF.
You do not currently have access to this content.