An Inverse Analysis-Based Optimal Selection of Cohesive Zone Model for Metallic Materials

Abstract

Five different cohesive zone models (CZMs), including bilinear, polynomial, trapezoidal, exponential, and PPR (Park–Paulino–Roesler) models, which are commonly used in simulating fracture failure of metallic materials, are evaluated in this paper. The cohesive parameters of these CZMs are determined by an inverse analysis based on the modified Levenberg–Marquardt method. A finite element (FE) model is developed by employing these CZMs and used to predict fracture behaviors of steel grade 120, which is frequently used for the tool joints of drill pipes. Tensile and fracture tests are conducted to determine material properties and fracture behaviors of the steel grade 120, and the fracture behavior obtained from the experiment is used to determine the CZM parameters and validate the FE model. It is found that the five CZMs, with the cohesive parameters determined by the inverse analysis, can be used to simulate the ductile fracture process of the steel, and that among the five CZMs, the exponential CZM provides the closest results to the experimental data.