Identifying plastic properties of friction stir-welded material by small punch beam test

Average plastic properties of friction stir-welded AA2024-T3 are obtained by coupling novel small punch beam testing with a neural network algorithm. The small punch beam test utilizes a cylindrical punch head and miniature rectangular beam specimens. The specimens may be manufactured by material removed from in-service components with minimal effect on mechanical performances. Specimen preparation, material model, and identifying procedure are systematically presented. Predicted load–displacement results agree well with the experimental results and the identified strain–stress relationship demonstrates useful agreement with tensile test. Since the load–displacement curve is insensitive to base material properties, knowledge of these properties is not required in the proposed method.