Structural Fatigue Life Optimization Based on Metamodels

In order to increase the computing efficiency of the structural fatigue life in optimization cycles, the metamodels of the structural fatigue life are researched systematically. Three kinds of metamodels, the polynomial response surface model, the Kriging model and the radial basis function model constructed by the central composition design and the Latin hypercube design were compared. It is found that the radial basis function model constructed by the Latin hypercube has the highest fitting precision and higher fitting efficiency. As an example, the forward strut of an aircraft´s nose gear is optimized by using the radial basis functions of the fatigue life. The research results offer a guidance for selection and construction of the metamodel of the structural fatigue life.