An experimental validation of residual stresses in weld clad pipelines

The nature and distribution of residual stresses are invariably critical for fatigue life with dissimilar material joints often inducing high tensile residual stresses. A fatigue-resistant concept of weld cladding process pipelines, producing compressive residual stresses, is under investigation to examine how these stresses may be influenced. Simplified weld cladding simulations have successfully illustrated the development and distribution of residual stresses through the joint. The study has highlighted the importance of accurate material data for clad and substrate materials with current analysis assumptions in a simple thick-walled pipe discussed. Experimental validation, using ICHD, measured residual stresses with depth on weld clad specimens, resulting in good correlation between simulation and experiment for a nickel-chromium-based superalloy clad on low alloy carbon steel as discussed. Future work, including a full 3D representation of the cladding process and a comparison of residual stress measurement methods, are also discussed. Figure 1. Axisymmetric FE Model Figure 2. Axisymmetric Model Boundary Conditions Although this model is a simplification of a process with spatially and temporally varying deposition, it is representative of methods such as evaporative deposition or spray coating followed by sintering. In the following sections it will be shown however that it also provides good correlation to experimental measurements for the spatially and temporally varying weld deposition process used. An elastic-perfectly plastic material model was assumed with temperature dependent thermal and mechanical material properties. Focus was placed on coefficient of thermal expansion, specific heat, thermal conductivity, Young’s modulus and yield stress. These properties primarily govern the modelling of the thermal event, degree of constraint and therefore the development of residual stresses (Ahmed et al. 2007), with further contributing factors discussed in the reference given. Poisson’s ratio and emissivity were assumed to remain constant. The nature of residual stresses differs depending on the state of a component, as discussed by Withers and Bhadeshia, with misfit resulting from different parts, phases and regions highlighted as the key characteristic influencing residual stresses (Withers & Bhadeshia 2001). Phase transformations in steels have been shown to influence residual stress development in X70 pipeline steel (Onsoien et al. 2010) for example. Figure 3 shows a large discontinuity in the stress distribution in the vicinity of the low alloy carbon steel substrate and clad interface for both pre-heat and material combinations, with a change from compressive to tensile stresses. At the idealised interface between the substrate and clad there is a step change in the stresses with the nickel-chromiumbased superalloy clad showing a reduction in residual stress, which remains significantly tensile throughout the clad. The form of residual stress graph for the nickel-chromium-based superalloy clad is indicative of yielding and reflects the elasticperfectly plastic model used. The stainless steel clad simulation results on the other hand show a further increase in residual stress in the clad, which is unexpected and is contrary to the experimental results obtained, which show a similar value of residual stress throughout the clad but compressive in nature. 3 EXPERIMENTAL VALIDATION