Effect of Welding Direction and Bead Pattern in Alloy 52 / SA508 Repair Weld

As nuclear power plants age and their lifetimes are being extended, the possibility and need to perform repairs of safety critical and hard to replace components is ever increasing. For example, defects in the reactor pressure vessel caused by exposure to high temperature, pressure, and corrosive environment together with neutron irradiation are often repaired by different repair welding techniques. Moreover, the need for such repairs may come at short notice requiring that qualified and optimized techniques and solutions are readily available. Developments of repair welding techniques using robotized gas metal arc welding cold metal transfer to repair a linear crack like defect beneath the cladding, which extended into the reactor pressure vessel steel have been presented in previous works [8–9]). In the latest piece of research [10], the repair welding of a thermally embrittled and cladded low-alloy steel plate with two groove excavations filled using Alloy 52 was presented. In the paper, the two welds were characterized with micrographs and microhardness measurements. This work further evaluates in more detail the differences and similarities of the repair welds welded using two different welding directions, 0-degree and 45-degree, and corresponding bead patterns. Residual stresses were measured from the two repair-weld cases using the contour method. Despite significant differences in the weld bead order and consequent welding procedure, the resulting residual stresses were very similar. It was expected that the crisscross weld bead pattern would cause the subsequent weld layers to induce stresses counteracting the previous layer and thus reduce the overall residual stress field. However, this does not appear to be the case. Both weld areas showed tensile stresses around 300 MPa, which is close to the yield stress of the weld material. Balancing compressive stress is induced to the base material with somewhat lower magnitude, peaking around 200 MPa. This indicates that the main determinant of the residual stress field is the weld material yield behavior. The microstructural characterization of the two weld orientations included microhardness and nanohardness measurements across the low-alloy steel and Alloy 52 weld fusion boundary, where the hardness peak was at the coarsegrained heat-affected zone adjacent to the fusion boundary. The 0-degree weld gives a higher microhardness peak than the 45-degree weld, indicating a slightly higher mismatch in properties, but the nanohardness measurements could not confirm this. Also, in the microstructural analysis, no great differences are seen other than few weld defects, especially voids. The elemental analysis using energy dispersive X-ray spectrometry across the fusion boundary shows expected minor dilution of alloy elements, e.g. chromium, which affects the materials corrosion properties. Electron backscatter diffraction mapping and nanoindentation measurements were performed across the weld interface for both welding directions. In addition, elevated temperature measurements were carried out on the weld to understand the evolution in mechanical properties in service conditions. Finite element modeling was used to simulate the welding using the actual welding conditions and approximate material constants for both bead patterns as input parameters. The resulting deformation, strains and stresses were predicted. The computed weld residual stress state was rather similar for both cases, although the 45-degree welding simulation produced a stronger tempering effect on the subsurface bead layers. The general conclusion of these studies is that no significant differences caused by the welding direction and bead pattern can be observed. The more complex bead pattern may render the weld more susceptible to more welding defects, such as lack-of-fusion and porosity, of which some evidence was found.