The hypothesis is that designers utilize a hierarchical sequence of mathematical models from simple, highly idealized, cheap models towards the most rigorous, complex, and expensive models. They stop when the idealization error is sufficiently small or they run out of time or money. Currently, the error estimate is usually based on the judgement of the design engineer. The objective is to develop algorithms to automatically estimate the error and control the analysis. The concept should be generally applicable to evaluating designs, but to illustrate the method this research project will be focused on plate and shell theory.
