TURBULENT HEAT TRANSFER IN LATTICE-STRUCTURED MATRIX COOLING CHANNELS

Abstract

Matrix cooling is an internal cooling concept for gas turbine blades that combines structural reinforcement with enhanced heat transfer through a rib-crossing lattice geometry. The repeated flow redirection generates strong shear layers, impingement, and vortical structures that significantly influence turbulent transport and heat transfer. This study presents a numerical investigation of turbulent flow and heat transfer in a lattice-structured matrix cooling channel using Reynolds-averaged Navier–Stokes (RANS) and Large-Eddy Simulation (LES) approaches, validated against experimental data at Re = 24100. Steady RANS simulations are performed using k–ω, k–ω SST, Realizable k–ε, and Launder–Sharma k–ε models. The RANS results reproduce global trends of heat transfer and pressure drop, with area-averaged heat transfer coefficients predicted within approximately ±3% of the experimental value, while showing limited capability in resolving local heat transfer variations. LES computations are conducted using several subgrid-scale (SGS) models, including WALE, One-Equation Eddy Viscosity (OEEVM), Smagorinsky, and the Localized Dynamic k-Equation Model (LDKM). A mesh-refinement study demonstrates that increasing grid resolution improves LES accuracy, reducing the discrepancy in area-averaged heat transfer from more than 25% on coarse meshes to approximately 5-8% on fine and very-fine meshes. The choice of SGS model significantly affects heat transfer predictions, whereas pressure drop remains comparatively insensitive. The results quantify the trade-off between predictive accuracy and computational cost and provide guidance for selecting turbulence modeling strategies for engineering analysis of matrix cooling configurations.