Micromechanics-based models provide powerful tools to predict initiation of ductile fracture in steels. A new criterion is presented herein to study the process of ductile fracture when the effects of both stress tria...Micromechanics-based models provide powerful tools to predict initiation of ductile fracture in steels. A new criterion is presented herein to study the process of ductile fracture when the effects of both stress triaxiality and shear stress on void growth and coalescence are considered. Finite-element analyses of two different kinds of steel, viz. ASTM A992 and AISI 1045, were carried out to monitor the history of stress and strain states and study the methodology for determining fracture initiation. Both the new model and void growth model (VGM) were calibrated for both kinds of steel and their accuracy for predicting fracture initiation evaluated. The results indicated that both models offer good accuracy for predicting fracture of A992 steel. However, use of the VGM leads to a significant deviation for 1045 steel, while the new model presents good performance for predicting fracture over a wide range of stress triaxiality while capturing the effect of shear stress on fracture initiation.展开更多
Accurate prediction of ductile fracture requires determining the material properties,including the parameters of the constitutive and ductile fracture model,which represent the true material response.Conventional cali...Accurate prediction of ductile fracture requires determining the material properties,including the parameters of the constitutive and ductile fracture model,which represent the true material response.Conventional calibration of material parameters often relies on a trial-and-error approach,in which the parameters are manually adjusted until the corresponding finite element model results in a response matching the experimental global response.The parameter estimates are often subjective.To address this issue,in this paper we treat the identification of material parameters as an optimization problem and introduce the particle swarm optimization(PSO)algorithm as the optimization approach.We provide material parameters of two uncoupled ductile fracture models—the Rice and Tracey void growth model(RT-VGM)and the micro-mechanical void growth model(MM-VGM),and a coupled model—the gurson-Tvergaard-Needleman(GTN)model for ASTM A36,A572 Gr.50,and A992 structural steels using an automated PSO method.By minimizing the difference between the experimental results and finite element simulations of the load-displacement curves for a set of tests of circumferentially notched tensile(CNT)bars,the calibration procedure automatically determines the parameters of the strain hardening law as well as the uncoupled models and the coupled GTN constitutive model.Validation studies show accurate prediction of the load-displacement response and ductile fracture initiation in V-notch specimens,and confirm the PSO algorithm as an effective and robust algorithm for seeking ductile fracture model parameters.PSO has excellent potential for identifying other fracture models(e.g.,shear modified GTN)with many parameters that can give rise to more accurate predictions of ductile fracture.Limitations of the PSO algorithm and the current calibrated ductile fracture models are also discussed in this paper.展开更多
基金the National Science Foundation (Grant 1344592)the National Natural Science Foundation of China (Grant 51778462)the National Key Research and Development Plan (Grants 2017YFC1500700 and 2016YFC0701400).
文摘Micromechanics-based models provide powerful tools to predict initiation of ductile fracture in steels. A new criterion is presented herein to study the process of ductile fracture when the effects of both stress triaxiality and shear stress on void growth and coalescence are considered. Finite-element analyses of two different kinds of steel, viz. ASTM A992 and AISI 1045, were carried out to monitor the history of stress and strain states and study the methodology for determining fracture initiation. Both the new model and void growth model (VGM) were calibrated for both kinds of steel and their accuracy for predicting fracture initiation evaluated. The results indicated that both models offer good accuracy for predicting fracture of A992 steel. However, use of the VGM leads to a significant deviation for 1045 steel, while the new model presents good performance for predicting fracture over a wide range of stress triaxiality while capturing the effect of shear stress on fracture initiation.
基金the National Natural Science Foundation of China(No.51908416)the Shanghai Pujiang Program(No.19PJ1409500)the Fundamental Research Funds for the Central Universities,China。
文摘Accurate prediction of ductile fracture requires determining the material properties,including the parameters of the constitutive and ductile fracture model,which represent the true material response.Conventional calibration of material parameters often relies on a trial-and-error approach,in which the parameters are manually adjusted until the corresponding finite element model results in a response matching the experimental global response.The parameter estimates are often subjective.To address this issue,in this paper we treat the identification of material parameters as an optimization problem and introduce the particle swarm optimization(PSO)algorithm as the optimization approach.We provide material parameters of two uncoupled ductile fracture models—the Rice and Tracey void growth model(RT-VGM)and the micro-mechanical void growth model(MM-VGM),and a coupled model—the gurson-Tvergaard-Needleman(GTN)model for ASTM A36,A572 Gr.50,and A992 structural steels using an automated PSO method.By minimizing the difference between the experimental results and finite element simulations of the load-displacement curves for a set of tests of circumferentially notched tensile(CNT)bars,the calibration procedure automatically determines the parameters of the strain hardening law as well as the uncoupled models and the coupled GTN constitutive model.Validation studies show accurate prediction of the load-displacement response and ductile fracture initiation in V-notch specimens,and confirm the PSO algorithm as an effective and robust algorithm for seeking ductile fracture model parameters.PSO has excellent potential for identifying other fracture models(e.g.,shear modified GTN)with many parameters that can give rise to more accurate predictions of ductile fracture.Limitations of the PSO algorithm and the current calibrated ductile fracture models are also discussed in this paper.
