Design and analysis of computer experiments have been widely investigated. This study presents numerical procedure to optimize the sheet metal forming process. Metamodels based on responses from numerical experiments may form efficient approximations to functions in engineering analysis. They can improve the efficiency of engineering optimization substantially by uncoupling computationally expensive analysis models and (iterative) optimization procedures. This paper investigated the kriging metamodel approach. At the same time, the practical nonlinear engineering problems such as square drawing are also optimized successfully by proposed method. The results prove Kriging model is an effective method for nonlinear engineering problem in practice.