Simulation Analysis of Ceramic Shell Representation and Process Parameters in IN718 Investment Casting

Abstract

This thesis investigates how ceramic shell representation and selected process parameters influence the simulated solidification behaviour and predicted defect tendency of an investment-cast IN718 aerospace component. Casting simulations are used to compare different thermal conditions around the ceramic shell, with focus on temperature history, heat-flux response, filling and solidification behaviour, shell representation and porosity-related defect indicators. The study treats the ceramic shell as part of the thermal system, rather than only as a geometric boundary, since shell thickness and surrounding boundary conditions affect heat extraction, feeding conditions and last-to-freeze regions. The cast material is not varied as an independent parameter. IN718 is treated as the selected alloy system through which the thermal effects of shell representation and process parameters are interpreted. The simulations show that insulation wrapping has a strong influence on the thermal response of the mould and casting system. Among the investigated process parameters, shell preheat temperature has the strongest influence on the evaluated outputs. Higher shell preheat gives preferable porosity indications, while lower shell preheat gives less favourable indications, especially when combined with lower mass flow. The mass-flow-rate variation has a smaller influence, while the enclosure emissivity has negligible influence for the investigated setup. The comparison between constant and non-uniform shell representations shows that shell representation can affect the local thermal response and predicted defect tendency. However, this effect is smaller than the influence of shell preheat temperature. A limited qualitative comparison with available process observations shows some agreement between simulated defect-indicator regions and observed porosity indications, but the comparison is not sufficient for quantitative validation. The results should therefore be interpreted as simulation-based sensitivity trends that support process understanding, rather than as verified predictions of exact casting quality or as a final production specification.