Hot Isostatic Pressing of a Varying Thickness, Thick-Walled Vessel (Reactor Circulating Pump Bowl) for a Pressurised Water Reactor (PWR) Application

Abstract

This paper presents an overview of the work undertaken by Rolls-Royce to justify the use of a Hot Isostatically Pressed (HIP) Reactor Circulating Pump (RCP) bowl manufactured in 316L stainless steel for a Pressurised Water Reactor (PWR) application. It presents the work from a design justification/manufacturing quality assurance perspective, rather than from a pure metallurgical perspective. Although the HIP process is not new, it was new in its application to Rolls-Royce designed nuclear reactor plant. As a consequence, Rolls-Royce has implemented an evolving, staged approach, starting with HIP bonding of solid valve seats into small bore valve pressure boundaries. This was followed by powder HIP consolidation of leak-limited, thin-walled toroids, and has culminated in the powder HIP consolidation of thick-walled components such as RCP bowls. The paper provides an overview of each of these stages, the method of manufacture for the RCP bowl, and the approach taken with respect to justification. In previous Rolls-Royce applications of HIP to PWR plant components, the component section thickness has been fairly consistent. For the RCP bowl application, the section thickness varies quite considerably. To assess any variation in properties as a result of section thickness variation, a stepped wedge technology demonstrator was first manufactured and non-destructively and destructively examined to assess whether material properties remain within specification. The paper presents material property results for ‘Near Surface’ and ‘Buried’ samples taken from section thicknesses covering 50mm to 300mm. Yield, Ultimate Tensile Strength (UTS), Charpy ‘V’ notch impact strength and Strauss test results were found to be within specification for all section thicknesses. An analysis of the data shows no statistical difference in material properties across the different section thicknesses for both near surface and buried properties, except for Charpy ‘V’ notch impact strength, for which, above 50mm, mean ‘Buried’ strength is significantly higher than mean strength for ‘Near Surface’. It is observed that the mean ‘Near Surface’ yield and tensile strength appears to increase with section thickness; however, there are insufficient data to demonstrate whether or not the increase is statistically significant. There appears to be a reduction in mean ‘Buried’ yield strength when the section thickness is at 300mm, compared to thinner sections, but again, the current data are insufficient to determine whether or not the reduction is statistically significant. It could be postulated that there is a point at which the section thickness starts to influence the cooling rate, and as a consequence grain growth may be more prominent at the ‘Buried’ position. This could result in yield and tensile material properties being detrimentally affected. However, this is not supported by an analysis of Charpy ‘V’ notch impact strength results, for which a similar reduction would be expected; the highest ‘Buried’ Charpy results occur in the thickest section. Further work is ongoing to understand this observation. Non-destructive examination results of a prototype RCP bowl are presented. This shows no defects identified from dye penetrant surface examination and ultrasonic testing for a rejection level set at a 3mm flat bottomed hole.