Robustness of Residual Stress Computation: From SATOH Test to Welding Plate

In order to evaluate the robustness of numerical welding simulation, sensitivity analyses on mechanical models and parameters have been conducted for given thermal fields. The study is based on two simple TIG welding experiments. The material is a 316L stainless steel where no phase transformation is present. Special attention was paid to the mechanical behavior of the Heat Affected Zone (HAZ). Both experiments are performed on a simple geometry of reduced dimensions. The first one is the uniaxial SATOH test on a beam clamped between two rigid bodies. The specimen is homogeneous heated and cooled under restrained condition in order to reproduce thermo-mechanical phenomena that take place in the HAZ of a structure during welding. The second one is a two-pass TIG weld deposit on a plate. On this mock-up (identified as 24–25, [8]), quite a few experimental measurements were performed in order to compare with numerical results of different models. At the first step, we research a reference solution which gives the best agreement with the thermal and mechanical experimental results. The second step consists in determining through numerical simulations which mechanical parameters have the largest influence on residual stresses induced by the welding process. This paper presents the numerical analysis of the two experiments weld reference. Experimental verifications were made at each stage of the calculation, in order to validate the numerical model. In a further step, sensitivity computations have been conducted to examine the most influential parameters on the computed results.