AUTOMATED AISI 317L SS THIN SECTION IMAGE POROSITY RECOGNITION AND SEMI-AUTOMATED PRIORITIZATION FOR PREDICTING THE YIELDING POSITION BY MEANS OF IMAGE PROCESSING
Porosity is a common phenomenon in fabricated parts, which cause stress concentration and leads to yielding, brittle fraction and fatigue of structures which makes porosity recognition and analyzing very important to increase efficiency and decrease the defects of manufactured parts. Many researches have been done in order to detecting and recognizing pores of material by different numerical and experimental methods. In this paper, an algorithm is developed in order to detect the porosity of thin section images automatically and prioritize them for a semi-automatically prediction of yielding position. The algorithm results is the following parameters in a porous material: number of pores, position and size of pores, maximum and minimum distances of pores both from each other and from boundaries. The algorithm shows an accuracy of 83% in the prediction of the results for simulated finite element method (FEM) tensile tests, which makes the algorithm creditable to be used as a non-destructive testing method.
Porosity, Thin-section image, Image analysis, Finite element method, Aisi 317L