Abstract: The constructed Back-Propagation Neural Network (BPNN) can be used to learn the stress field around the crack tip. Then the approximation of the stress field around the crack tip was achieved. The trained neural network has the ability of association and memory. At the same time, it can calculate the stress field of the component with a arack in reality quickly and correctly, confirming to the plastic region. With the trained network, the component can to be analyzed easily and conveniently.
Keywords: neural network, crack, stress.