The Effect of Prior Austenite Grain Size on Transformation Behaviour as a Function of Cooling Rate in Nuclear Pressure Vessel Steel SA508 G4N

Abstract

A key element of many nations’ future baseline energy supply is earmarked to be based on nuclear fission. A critical component of a civil nuclear reactor is the Reactor Pressure Vessel (RPV). The RPV is exposed to a harsh environment of high pressure and high temperature concurrently with neutron irradiation. The majority of RPVs in-service use low alloy steel grade SA508 Gr.3 (Mn–Mo–Ni) or its international equivalent. However, there is a drive to run at higher pressures and temperatures for future nuclear reactors, which requires higher RPV wall thicknesses than currently operating. This has led to the development of SA508 Gr.4N (Ni-Mo-Cr), which has higher hardenability, toughness, and strength, allowing for the design thickness of the reactor wall to be more than 200 mm, increasing the efficiency, power, and lifespan of the nuclear reactor to potentially well over 50 years. A potential major problem going to larger size RPVs will be a possible variation in Prior Austenite Grain Size (PAGS) generated because of the complex thermomechanical processing route the RPV material will experience and its interaction with varying cooling rates between the surface and centre during quenching. There has been much debate about the effect of PAGS on microstructure evolution in RPV materials, in particular the bainite transformation. In this research, the impact of PAGS on microstructure evolution during heat treatment has been investigated in order to understand the complex relations between processing, microstructure and physical properties to help understand if there are going to be complications during the operational stages of the RPV. Austenitisation was undertaken at different times and temperatures to produce 3 different PAGS sizes, fine (23 µm), medium (84 µm), and coarse (412 µm). They were then subjected to three different cooling rates: 0.07 °C/s and 0.3 °C/s and 3 °C/s. A lath microstructure dominated at fine PAGS at the slowest cooling rate of 0.07 °C/s, while at the largest PAGS the microstructure was mostly granular bainite. The average bainitic lath width is inversely proportional to the size of the PAGS. A larger prior austenite grain size resulted in a greater volume fraction of granular bainite and blocky MA islands. A fully martensitic microstructure was present at the highest cooling rate of 3 °C/s for all PAGSs, while a combination of lower bainite and martensite were present at the intermediate cooling rate and increasing PAGS also lead to a slightly greater level of tempered martensite volume fraction. The effect of tempering times (6h and 24h) at a tempering temperature of 550 °C was also studied. In case of the slowest cooling rate at 0.07 °C/s, bainitic lath width increases during theiii tempering stage, also some laths coarsen and consume neighbouring laths and there is also some indication of laths bulging out. In addition, the carbides inside the coarse laths appear to be coarsening at a much faster rate than the fine carbides within a fine lath structure. Some of the martensite/austenite islands (MA) within granular bainite decomposed with tempering time to form new bainite and fresh carbides that increased the bainite volume fraction, while other untransformed MA increased its size and then disappeared with longer tempering times. Moreover, a larger PAGS led to a greater amount of carbide clusters forming at austenite triple junctions (for the longer tempering time), furthermore some of the blocky-like shape MA stabilised more than needle-like shape. For certain martensitic regions of the microstructure, laths were observed coarsening, which consumed adjacent laths. At a cooling rate of 3 °C/s with tempering time, carbides were not seen in the medium PAGS. Carbides were observed at triple junction boundary for coarse PAGS with 6h tempering, and after 24h tempering