Most safety related valves in EDF’s nuclear plant must prove their ability to sustain thermal shocks of approximately 240K amplitude. This paper evaluates the simulation of a globe valve tested for thermal shocks. Since the physical test campaign showed inadequate internal sealing, the simulation focuses on the residual deformation of the hard alloy, planar seat, welded on successive body designs. This deformation is the result of the thermal loadings first induced by the welding process, then by fluid flow inside the valve. A chain of 3D simulations successively computes: a welding temperature transient in the body, the resulting strain hardening — especially in the seat vicinity —; temperature transients in the flow and the valve parts, and the resulting strains in the body causing a bump deformation of the seat surface. This end result agrees with measurements on the tested valve specimen. We show that inaccurate results are obtained on simpler assumptions, such as no welding, and we give insights on the dominant effect of the first hot, cold, hot transient over other profiles. Finally, the agreement we obtain on deformation predictions is toned down by an unsatisfactory sealing prediction, as well as the complexity and duration of the simulation chain compared with physical testing.
Influence of heat flux loading patterns on the surface cracking features of tungsten armor under ELM-like thermal shocks