Determination of residual stresses in HSLA-100 steel weldments as a function of welding parameters using x-ray diffraction

This project was initiated by the Cardcrock Division of the Naval Surface Warfare Center (CDNSWC) and the Office of Naval Research to study the effects of various processing parameters on the residual stress state of HSLA-100 bead-on-plate weldments. Three groups of samples were provided to Virginia Tech by CDNSWC. The first (GPX) was a sample of unwelded, as-received base material; the second group (RS-) consisted of an experimental matrix of differently processed bead-on-plate weldments which were all ground prior to welding in order to prepare the surface for welding; the third group (SR-) was a pair of weldments vacuum annealed at 1200°F for one and two hours, and then welded (without grinding) using welding parameters identical to weldments from the second group. X-ray diffraction was used to measure the surface residual stress state of all samples. It was found that the surface residual stress states of the GPX plate and the unwelded SR-group plates showed no statistically significant difference in magnitude, though the variation of the stress state over the surface of the plates seemed to decrease with increasing annealing time.

The severe, non-uniform grinding was determined to play a very large role in the residual stresses generated in the welds, sometimes changing both the magnitude and the shape of the stress patterns. Residual stresses in plates that were ground before welding were always more tensile than those that were not ground. Grinding also caused a large compressive-to-tensile stress gradient in the transverse direction. The grinding made it difficult to determine the effects of different welding parameters on the residual stress state.

Assuming that the stresses closest to the weld bead are exclusively residual stresses due to welding, preheat temperature reduced the tensile nature, or increased the compressive nature, of the residual stresses. This is due to the preheat reducing the effect of shrinkage stresses induced after the austenite transformation upon cooling of the weldment. Because of the effects of grinding and the small sample sizes, no definitive conclusions could be drawn about the effects of heat input and plate thickness. It was shown that grinding was the dominant parameter on the residual stress state in these HSLA-100 bead-on-plate weldments. Because the angle and force of grinding are purely at the discretion of the operator, it is very difficult to determine the effects of different welding parameters on the residual stresses generated in bead-on-plate weldments ground prior to welding.