Conjugate heat transfer for LES of gas turbine engines

Current design tools for gas turbine engines invoke a variety of simplifying assumptions to estimate heat transfer to solid/metal engine components (e.g., isothermal boundary conditions). These approximations are often not valid, result in inaccurate predictions of heat transfer, and ultimately compromise the thermal integrity of propulsion and power systems. Wall-modeled large eddy simulation (WMLES) has been recently used to predict heat transfer rates in a variety of relevant flow environments, including separated flows and reactive effusion cooling. Despite these advances, WMLES often does not include conjugate heat conduction in the solid materials due to added computational cost and temporal stiffness from the vast timescale separation between the flow, combustion, and solid-side conduction. In this proposal, Cascade Technologies and The Pennsylvania State University present a plan to build efficient computational tools to predict conjugate heat transfer using WMLES on massively parallel computers. The proposed algorithms will be validated against a battery of cases relevant to turbine blades and effusion cooling, with comparison to detailed experimental data and state-of-the-art RANS simulations. The team will additionally supplement existing experimental databases to advance understanding of conjugate heat transfer in the broader gas turbine community.