Stretch-zone analysis by image processing for the evaluation of initiation fracture toughness of a HSLA steel

In this work, stretch zone width and stretch zone depth of ductile fracture surfaces of a high-strength low-alloy steel at various prestrained conditions are computed using digital image processing techniques, where the images are taken by scanning electron microscopy. Histogram equalisation is applied on the input image for enhancement of the image quality. Various edge detection filters, such as Laplacian, Sobel, and Kirsch are separately applied to the digital image to detect the edges of the stretch-zone. The images obtained after processing are binarized through thresholding for identifying the stretch zone boundaries and to compute the stretch-zone width. It is observed that the Laplacian filter produced better results compared to the other filters. The measured values of stretch-zone width and stretch-zone depth by image analysis agreed well with those obtained manually; however they showed discrepancy with the same obtained indirectly from the fracture-resistance curve. The reasons for such a discrepancy are highlighted in this paper. From the estimated stretch-zone width, initiation fracture toughness of the HSLA steel has been obtained.