| Summary: | The structural integrity and operational safety of nuclear pressure vessels is based upon mechanical properties measured from test material representative of their individual component forgings. Conservatism within the nuclear industry has meant that a limited number of alloys have been used for nuclear forgings and as a result of the operational experience generated, the industry relies upon process optimisation to achieve the levels and balance of mechanical properties required. Of the production processes used, heat treatments offer a significant contribution to the final mechanical properties formed. Owing to the bespoke nature and cost of nuclear forgings experimental heat treatment optimisation cannot be regularly used. In such scenarios computational modelling presents itself as a powerful tool for heat treatment optimisation which can be exploited in the design of components. A semi-empirical methodology was used to develop an industrially relevant quenching model that captures realistic physical phenomena, using a non-expensive computational simulation, with sufficient detail to quantify the key metallurgical and heat transfer events. The model development is also used to improve the current understanding of heat transfer in large components. Differential scanning calorimetry was used to investigate the phase transformations of the low alloy pressure vessel steel grades SA508-3 and SA508-4N. As part of the investigation, the critical transformation temperatures for each alloy is established and latent heat release, in the form of effective specific heat capacity, is measured. The measured transformation data is incorporated in a material file and used in finite element modelling to replicate the effects of latent heat on cooling curves. The simulated results of the developed model demonstrated good agreement with thermocouple data taken from the ¼ and ½ thickness positions of a production forging. The results also highlighted the importance of accurately representing values of thermal conductivity, particularly the transition in magnitude at the Curie point The model was then used in inverse analysis to establish boundary conditions representative of the Sheffield Forgemasters' quench tank. The heat transfer coefficients derived showed that classical boiling events occurred for a short period relative to the total quench time. An investigation into result sensitivity showed that, for large components, a constant value of thermal conductivity is sufficient to derive cooling results that are reasonably accurate when compared with thermocouple data from representative forgings at positions of significant distance from the surface. Finally, a full scale heat treatment trial was conducted to further validate the modelling methodology and to establish a second set of heat transfer coefficient values for comparative purposes. The trial was also used to investigate the use of welded thermal buffers in quenching and concluded that buffers are an acceptable means to remove end affects so long as the attachment of the buffer to the forging eliminated quench water ingress to the buffer forging interface.
|
|---|
