Topology Optimization for Internal Cooling of Gas Turbine Guide Vanes—A Conjugate Heat Transfer Study

This study explores the feasibility and validity of using topology optimization (TO) to design the internal cooling of an airfoil-like geometry approximating a turbine guide vane. A conjugate heat transfer approach where the fluid flow physics are coupled with a convection–diffusion model for heat transfer is used in the TO. The objective is to minimize the maximum temperature on the outer surface of the vane with a constraint on the mass flow of the internal coolant. Two different flow models are investigated for the TO process: the Stokes model and the Reynolds-Averaged Navier–Stokes (RANS) equations with a simple zero-equation turbulence model. Velocity and temperature fields in topology-optimized designs are then compared to conventional conjugate heat transfer analyses performed on post-processed designs with body-fitted meshes and those using the shear stress transport (SST) RANS turbulence model. Designs obtained with the Stokes model exhibit different flow trajectories and mixing, while the use of RANS equations improves predictions but introduces uncertainties due to turbulence modeling limitations, particularly in the presence of flow separation. Thus, considering these limitations, the findings suggest that a simple flow model, such as Stokes in TO, with a comparatively low computational cost, can yield useful design concepts. However, the simplifications in the governing equations and their impact on physics should be considered carefully, and further aerothermal validation is required. Thus, the study findings, along with advances in robust meshing, enhance the practicality of topology optimization for industrial applications.