When conducting life assessment analysis of welded structural components in nuclear power plants, residual stress needs to be taken into account. In addition, when considering structural size components, to account for size-effect dependence, fatigue crack growth criteria are necessarily conservative to ensure plan safety. To reduce the over-conservatism of currently used structural integrity criteria, exhaustive and representative integrity tests must be done on full-size specimens. However, when considering welding residual stress fields in structure-size components such as electron beam-welded pipes, the stress profile is uncontrolled and tends to be mainly geometry-dependent. A very steep stress gradient must also be studied, linked to the redistribution of the residual stress field with crack growth and the particular cyclic plasticity behaviour of 316L steel. To define an experimental campaign, a numerical model has been developed to predict the introduction of a specific residual stress profile using a thermal shock to create inhomogeneous plasticity. A parametric numerical analysis on Cast3M has been conducted to design an experimental rig and evaluate the possibility of controlling the residual stress profile introduced by the temperature gradient during the thermal shock. Especially, an exhaustive parametric analysis has been used to validate the stability of the numerical model results: studying, in particular, eventual anisotropy effects and small deviations from modelling assumptions. The conclusion drawn from these calculations is the consistency of the numerical model with very stable results. Finally, these numerical results are compared with experimental data from a thermal shock experiment and neutron diffraction measurements to validate the numerical model and the methodology.