Optimization of Composite Structures Using Response Surface Approximations

Response surface techniques have been originally developed for improving designs based on testing in the laboratory or the field. These techniques fit simple functions, typically low order polynomials, to the experimental data, using least square fit, in a process known as regression. The least square fit filters out some of the variability and experimental errors, especially if the number of experiments is substantially larger than the number of unknown coefficients in the fitted function. The design optimization process then can employ the simple fitted function, and requires minimal effort.