Thilo Pirling

Welding process impact on residual stress and distortion

Abstract Residual stress and distortion continue to be important issues in shipbuilding and are still subject to large amounts of research. This paper demonstrates how the type of welding process influences the amount of distortion. Many shipyards currently use submerged arc welding (SAW) as their welding process of choice. In this manuscript, the authors compare welds made by SAW with DC gas metal arc welding, pulsed gas metal arc welding, Fronius cold metal transfer (CMT), autogenous laser and laser hybrid welding on butt welds in 4 mm thick DH36 ship plate. Laser and laser hybrid welding were found to produce the lowest distortion. Nevertheless, a considerable improvement can be achieved with the pulsed gas metal arc welding and CMT processes. The paper seeks to understand the relationship between heat input, fusion area, measured distortion and the residual stress predicted from a simple numerical model, and the residual stresses validated with experimental data.

Effect of high pressure rolling on weld-induced residual stresses

Abstract The formation of large residual stresses continues to be a problematic side effect of all common welding processes. In this work, localised high pressure rolling of gas metal arc welds to relieve these residual stresses has been investigated using strain gauging and neutron diffraction. Rolling was found to remove undesirable tensile stresses and even induce large compressive ones, though only when applied after rather than during welding. Strain measurements taken during combined welding and rolling operations show that this is because material at the weld line continues to yield as it cools. This erases any beneficial effect on the stress distribution of rolling at high temperature. A method of rolling using an oscillating force is also presented and found to be just as effective as the equivalent static force process.

Residual strain measurement for arc welding and localised high-pressure rolling using resistance strain gauges and neutron diffraction

Neutron diffraction and foil resistance strain gauges have been used to study the state of residual stress introduced by localised high-pressure rolling of structural steel plates, and compare it to that caused by gas metal arc welding. Rolling creates a region in which the residual stress state is highly compressive in the rolling direction. Furthermore, this region is sharply defined, making it potentially very suitable for cancelling out the tensile residual stresses caused by welding. It is also demonstrated that non-destructive strain measurements made during the welding and rolling processes can be used to indicate residual elastic strain and stress, and that this method shows good agreement with conventional neutron diffraction measurements. Determination of residual stresses in this way requires consideration of the effect of curvature on the values of strain measured at the surface of the object.

Assessment of the effect of residual stresses in elastic-plastic fracture of dissimilar welded components

Abstract Residual stresses in welds pose a significant threat to the structural integrity of a component, especially in the presence of defects and are required to be accounted for in assessing component safety. Although the R6 assessment procedure suggests various approximate methods for incorporating these effects in defect assessment, most of them are overly conservative and not very cost-effective. A more reliable approach is to characterise the weld residual stresses around a defect and study how they interact with primary load. The current paper analyses the effects of weld residual stresses on the fracture of a dissimilar weld in the presence of defect. The weld is made between modified 9Cr–1Mo steel and 316LN stainless steel using autogenous electron beam welding. A C(T) specimen was extracted from the centre of the weld and a crack introduced in the fusion zone using electro-discharge machining. The residual stresses around the crack were measured on a grid of measurement points at mid-thickness of the C(T) specimen using neutron diffraction on the strain diffractometer SALSA at ILL, Grenoble. The measured residual stresses around the crack-tip were incorporated into a finite element model and the interaction of these with applied load was predicted under fracture.