MULTIOBJECTIVE DESIGN OPTIMIZATION OF CRUCIFORM SPECIMENS FOR BIAXIAL FATIGUE LOADING AS A FUNCTION OF MATERIAL AND MANUFACTERING LIMITATIONS

Cruciform specimen design, for biaxial fatigue loading applications, requires a large number of variables in order to be fully defined. Two of the most important variables are the base material thickness and the minimum center thickness that can be machined in order to reduce the specimen thickness on the gauge area. While the first one is constrained by the commercially available material sheet thickness, the second one is limited by the ability to machine the material to a minimum thickness, which is a function of the manufacturing process and the material itself. Combining this fact with the use of a new generation of biaxial fatigue testing machines, that uses quite efficient electrical motors but with limited load capacity, the cruciform specimen geometry needs to be optimized, in order to achieve the higher stress levels possible on the specimen center, while maki ng sure the geometry is possible to manufacture. Using a cruciform geometry, with an elliptical fillet between the specimen arms and a revolved spline to reduce the specimen center thickness, a multi-objective optimization was used to achieve the optimal values for the design variables. The first objective function was defined to achieve the maximum stress level on the specimen center, which leads to fatigue crack initiation without the need for a notch. The second objective function was defined to achieve the center thickness itself, in order to avoid manufacturing problems, being necessary to have the higher value possible. Using the Direct Multi-Search algorithm, which is a derivate-free optimization method, a Pareto Front was obtained for each one of the base material thickness, ranging from 1 to 10 mm, defined in the Renard series of preferred numbers. In this Pareto front different optimal configurations exist and the end user can choose the applicable one for the used material and manufacturing process. Finally a full map of optimal design configurations was produced, in order to serve as a standard design procedure for cruciform specimens.