As part of a Department of Energy (DOE) funded program assessing advanced manufacturing techniques for Small Modular Reactor (SMR) applications, the Nuclear Advanced Manufacturing Research Centre (AMRC) and the Electric Power Research Institute (EPRI) have been developing Electron Beam Welding (EBW) parameters and procedures based upon SA508 Grade 3 Class 1 base material. The transition shell, a complex component connecting the lower assembly to the upper assembly is a shell that flares up with varying thicknesses across its section. The component due to its geometry could be built by near net shape powder metallurgy hot isostatic pressing instead of conventional forging techniques. The demonstrator transition shell here is built from several sub-forging as a training exercise. The complex geometry and joint configuration were selected to assess the EBW as a suitable technique. This paper presents results from the steady state welding in the 60-110 mm material thickness range, showing that weld properties meet specification requirements. Weld quality was assured by Time-of-Flight Diffraction (ToFD). The transition shell was completed by welding a flange to the assembly. The presented transition shell assembly represents 6 welded sections all fabricated in below 100 min total welding time.

Creep rupture strength is the principal material property prioritized in designing power generation plants against the steady-state stress due to internal pressure. Increasingly plants must cycle so there is a possibility of life reduction due to creep-fatigue interaction. Grade 92 steel is one of the creep strength enhanced ferritic (CSEF) steels which has superior creep strength compared to other CSEFs. It is expected to be widely used in coal-fired ultra-super critical plants as well as in LNG-fired combined cycle plants. However, at present there is insufficient information regarding the creep-fatigue behavior of this material. A joint study has been conducted to understand the behavior of this steel under creep-fatigue condition and see how accurate the failure life can be estimated. Three kinds of base materials as well as two kinds of welded joints have been tested under strain-controlled cyclic loading with or without hold times as well as under constant load creep condition. Continued decrease in the number of cycles to failure was observed with the extension of hold time in all the base metals and cross-weld specimens. It was found that the modified ductility exhaustion approach based on inelastic strain, as well as its extension employing the inelastic strain energy density, made reasonably accurate predictions of failure lives under a wide range of test conditions. Temperature- and rate-dependencies of fracture limits in terms of inelastic strain and energy density were able to be uniquely expressed using simple thermal activation energy parameters.