Parameters Sensitivity Analysis And Steel Joints Research’s Application Of GTN Model

The GTN (Gurson-Tvergaard-Needleman) model is a powerful tool to analyse the damage and fracture of metal materials. While at present, there are still two major issues which need to be solved in using the GTN model to analyse the damage and fracture of the steel joints. The workload of parameters identification is large, and some parameters may be identified with large deviations; The convergence of the existing subroutines with GTN model is not good in the finite element analysis, the engineering application of GTN model is hindered. This article will focus on solving these two major issues, the details are as follows:(1) The sensitivity analysis of single GTN model parameter change was carried out. A finite element model of an uniaxial tensile notched bar was established by using the ABAQUS software.29 analysis tests were designed. GTN model parameters’values were changed one by one near a given reference value. Ultimate load and fracture displacement were obtained and selected as the result indicators. The curves were used to fit the relationship between the result indicators and the different levels of GTN parameters, and the sensitivities of the indicators to the parameters were evaluated. The results indicated that, the influences of the parameters to the fracture displacement were much more significant than to the ultimate load. Sensitivities of the ultimate load to the parameters were ranked from large to small as the initial void volume fraction f0, the volume fraction of void nucleating particles fn, the plastic strain standard deviation of nucleation sn, the mean plastic strain for void nucleation εn, the model correction coefficient q2, the model correction coefficient q1, the void volume fraction at final failure fF. Sensitivities of the fracture displacement to the parameters were ranked from large to small as fo,fn, q2, εn,fF, q1, sn. All of these conclusions were based on GTN model parameters changing near the given reference values.(2) The sensitivity analysis of multi GTN model parameters change was conducted. The influence of GTN model parameters on the analysis results was studied, when these parameters changed in their common range. The parameters were divided into q1,q2and f0,fn,fF, εn, sn two groups. The finite element model of the uniaxial tensile notched bar was still used in this study. Orthogonal tests were applied to arrange values of parameters. The influence of the parameters to the predicted ultimate load and fracture displacement were analyzed. The results indicated that, the influence of every parameter to the ultimate load was not more than 3%, while the influences of parameters to the fracture displacement were significant. The fracture displacement was most sensitive to fo, fn, the second sensitive to q2,fF, εn, sn, q1. q1 and q2 determined by different methods led to different fracture displacements and fracture displacements’sensitivity to fF. fo, fn, εn, sn influenced the analysis results similarly when q1 and q2 were determined by those two methods.(3) In order to solve the issue of poor convergence of analyzing the practical engineering problems with the GTN model, the VUMAT subroutine used in the ABAQUS software was written. The new subroutine was used to analyse the failure of cross baffle welded pipe joints. GTN parameters values were determined by the uniaxial tensile tests. The finite element analysis models of joints were established. Joints’finite element analysis results and experimental results were compared. The results indicated that the new subroutine could be used to analyse joints under complex working conditions. GTN model analysis could predict the failure of the steel joints.