반응표면 모델에 의한 맥퍼슨 스트럿 공차 설계

Tolerance design of McPherson strut is performed according to the minimization of four responses (toe, camber, caster and kingpin inclination) analyzed by using the each response surface model made from stochastic design of experiments. After the each response surface model is constructed, the variation of response can be estimated by making approximation model as described in this paper and we can get the robust optimal solutions by using nonlinear programming method such as sequential quadratic programming method and generalized reduced gradient one. The design of experiments and the optimization algorithm are used together to obtain the optimal tolerance design of McPherson strut in this paper. The first order response surface models of the six lower arm pivot points (design factors) are constructed by using Taguchi's orthogonal array and the tolerance sensitivity analysis is also performed by applying variation approximation instead of Monte Carlo method.