Experimental Design and Structural Optimization

Structural optimization problems are mostly solved iteratively by combining finite element and mathematical progamming methods.An alternative approach can be used in which the FEM-analyses are planned a priori, both with regard to their number and the values of the design variables, applying techniques used for planning of experiments. From the outcome of the computations an approximate mathematical model is derived by means of regression analysis. This model can in turn be used as a fast analysis module in optimization software. In this paper the use of experimental design in structural optimization is discussed.The methods have been tested and used extensively for shape optimization of carillon bells, resulting in a new bell which is musically very interesting. They have also been applied successfully to several mechanical engineering problems.