Robustness in reliability-based design

This paper examines the effect of variabilities in the design variables on the reliability and robustness of system performance. A design of experiments and a response table method can be implemented to create a robust design. A nonlinear optimization algorithm can be implemented for reliability-based design. This paper proposes a method to achieve both reliability and robustness in design. The trade-off between robustness and reliability creates an optimization problem. An algorithm is formulated to enable the designer to determine the optimum design under different weights for reliability and robustness objectives. The procedure is demonstrated with a numerical example.