The use of eigenstrain theory and fuzzy techniques for intelligent modeling of residual stress and creep relaxation in welded superalloys

Abstract

Ni-base superalloys are used in a wide range of applications where components made from these alloys are exposed to extreme conditions of high temperature and high pressure, and dependable performance is critical for mission success, the safety of human lives, and multi-million commercial investment. To ensure robustness and reliability of highly demanding engineering solutions, it is crucial to advance the development of cutting-edge computational design tools based on artificial intelligence and fuzzy techniques, combined with the application of materials characterization and materials design. In the present study, the use of the contour method in combination with eigenstrain theory provided new insights into 3D residual stress states in Ni-base superalloy samples. As-welded and heat-treated specimens were made using bead-on-plate design to investigate the effect of complex fabrication conditions on welds process in large components. The widely used relief of residual stresses during post-weld heat treatment was simulated using eigenstrain-creep model. Furthermore, artificial intelligence (AI) based eigenstrain-contour and eigenstrain-creep models, that use fuzzy techniques, were developed by the present authors for materials used in advanced ultra-supercritical coal-powered plants, showing good match with experiments. The present study reports the combination of eigenstrain theory with artificial intelligence for the modelling of welding residual stresses and simulation of post-weld heat treatment process and highlights the benefits of AI-based eigenstrain-contour and eigenstrain-creep methods on the development of robust and reliable aeroengine components.