Cold-formed steel member cross-section shape optimization by Knowledge-Based Global Optimization method

One of the great advantages of cold-formed steel is the immense flexibility that the material affords in forming cross-sections. This flexibility would seem to readily lend itself to optimization of member cross-section shapes. However, little research has been conducted to explore this general problem; in part, because most existing optimization attempts have used prescriptive design standards that apply to a very limited set of cross-sections for their objective function evaluation. A newly introduced design procedure, the Direct Strength Method, or DSM, provides a method for evaluating the strength of cold-formed steel sections of arbitrary geometry. However, such an arbitrary problem, with a highly nonlinear objective function, is still difficult to solve by traditional optimization methods. A new technique, known as Knowledge Based Global Optimization, or KBGO, has recently been introduced which appears to hold promise for general optimization problems of this nature. In this paper, KBGO is applied to the member shape optimization of a cold-formed steel column to demonstrate the potential of this new optimization technique when combined with new advanced analysis methods (DSM) for member strength prediction. The results demonstrate that KBGO is indeed effective for such a problem, and can also shed some light on key features that are important for a successful member cross-section.