This paper introduces the application of Gurson model to simulate the ultimate ductile failure of three types of specimens with different constraints. Fracture tests were conducted using three kinds of notched round bar tensile specimens with different notch radii, a flat plate tensile specimen with a centered semielliptical surface flaw and a 1/2TCT specimen. Using the test results of the notched round bar tensile specimens, the Gurson model parameters were determined from the literatures, fracture observation and experimental design calculations. After fixing the Gurson parameters, they were applied to the flat plate tensile specimen and the 1/2TCT specimen model. As a. result Gurson model could simulate the fracture behavior of the flat plate with good accuracy, on the other hand there was a large difference with the test result of the CT specimen. In order to determine the Gurson model parameters universally and quickly, the blind optimization calculations were performed without arbitrariness. Three of five parameters were in the same order of the parameters by the 1st consideration and the others were in the different order. However the blind optimization parameters could show the similar simulation results of the fracture behaviors of the flat plate and 1TCT specimen as the 1st consideration parameters. The further investigation suggested the reason of the different behavior of the CT specimen model may be the different stress distribution, bending component dominant in the cracked ligament of the CT specimen from those of the other models. Additionally the accuracy of the local strain criteria according to ASME Sec. VIII Division 2 was confirmed.

This content is only available via PDF.
You do not currently have access to this content.