Defects that form during the investment casting of turbine blades for aircraft engines pose a serious hazard to flight safety and greatly increase the costs associated with manufacturing. In this study, we conducted numerical simulations based on a retained melt modulus model to analyze the flow of metals through molds in order to determine the probability of shrinkage defect (PSD) in heat-resistant steel (SCH12) turbine blades for aircraft engines. The results of preliminary simulations and experiments were used to guide the establishment of simulations based on four casting schemes. Our objective in this stage was to identify the casting scheme with the lowest PSD. We then implemented the best of the four schemes with the bottom of the casting system immersed in cold water. We also incorporated virtual thermo-dynamical sensors in the simulations to characterize the effects of water depth on the rate and direction of solidification in the mold cavity. We then applied the optimal investment casting conditions to the fabrication of turbine blades in an established foundry. X-ray analysis detected almost none of the detrimental defects commonly associated with this type of casting, thereby demonstrating the efficacy of the proposed scheme.