Quantitative study on surface crack of 304 austenitic stainless steel under natural magnetic field

Cracks are common defects in stainless steel which often lead to serious industrial accidents. In this paper, magnetic detection without external excitation was proposed. Quantitative defect identification was performed using multiclass classification support vector machine. The magnetic signals of 304 austenitic stainless steel were collected before and after annealing. The width, amplitude and area of magnetic signals were extracted as the input set of support vector machine model, and the prediction accuracy of cracks were compared and analyzed. The results showed that the prediction accuracy of length, width and depth of cracks are 80.70%, 92.71% and 65.63%, respectively. The width and depth of defects are increased by 5.89% and 33.34% respectively. This research provides a potential possibility for quantitative defect identification for nonferromagnetic materials under the natural magnetic field.