13th International Conference on Fracture June 16–21, 2013, Beijing, China -1- A Mechanism-Based Approach for Predicting Ductile Fracture of Metallic Alloys Xiaosheng Gao Department of Mechanical Engineering, The University of Akron, Akron, OH 44325, USA xgao@uakron.edu Abstract Ductile fracture in metallic alloys often follows a multi-step failure process involving void nucleation, growth and coalescence. Because of the difference in orders of magnitude between the size of the finite element needed to resolve the microscopic details and the size of the engineering structures, homogenized material models, which exhibits strain softening, are often used to simulate the crack propagation process. Various forms of porous plasticity models have been developed for this purpose. Calibration of these models requires the predicted macroscopic stress-strain response and void growth behavior of the representative material volume to match the results obtained from detailed finite element models with explicit void representation. A series of carefully designed experiments combined with finite element analyses of these specimens can also be used to calibrate the model parameters. As an example, a numerical procedure is proposed to predict ductile crack growth in thin panels of a 2024-T3 aluminum alloy. The calibrated computational model is applied to simulate crack extension in specimens having various initial crack configurations and the numerical predictions agree very well with experimental measurements. Keywords Ductile fracture, Unit cell analysis, Porous plasticity model, Stress triaxiality, Lode angle 1. Introduction It is well-known that ductile fracture in metallic alloys is a process of void nucleation, growth and coalescence and this process is strongly affected by the stress state imposed on the material. Based on the fracture mechanism, a straight-forward approach to simulate ductile failure process is to model individual voids explicitly using refined finite elements [1-3]. A distinct advantage of this approach is the exact implementation of void growth behavior. It provides an effective method to study the mechanisms of ductile fracture and to analyze the trend of fracture toughness. However, due to sizeable difference between the characteristic length scales involved in the material failure process and the dimensions of the actual structural component, it is impractical to model every void in detail in structure failure analysis, especially for situations involving extensive crack propagation. For this reason, various forms of porous material models have been developed to describe void growth and the associated macroscopic softening during the fracture process. The Gurson-Tvergaard-Needleman porous plasticity model [4-6], which assumes voids are spherical in materials and remain spherical in the growth process, has been widely used in modeling ductile failure process and ductile crack extension. Gologanu, Leblond and Devaux [7, 8] extended the GTN model and derived a yield function for materials containing non-spherical voids. The GLD model can be applied to predict crack propagation in many processed materials, such as rolled plates. In literature, the stress triaxiality ratio, defined as the ratio of the mean stress to the equivalent stress, is often used as the sole parameter to characterize the effect of the triaxial stress state on ductile fracture. However, recent studies show that the Lode parameter must be introduced to distinguish the stress states having the same triaxiality ratio [3, 9-11]. In this study, we describe a procedure to calibrate the material specific porous plasticity model so that it can accurately capture the material behavior in the fracture process zone with the influence of the stress state. A numerical approach is proposed to predict ductile crack growth in thin panels of a 2024-T3 aluminum alloy, where the GLD model is used to describe the void growth process and the material failure criterion is calibrated using experimental data. Model predictions are compared with experimental data for fracture specimens having various initial crack configurations.
RkJQdWJsaXNoZXIy MjM0NDE=