Large-Eddy Simulation and Conjugate Heat Transfer Around a Low-Mach Turbine Blade

Abstract

International audienceDetermination of heat loads is a key issue in the design of gas turbines. In order to optimize the cooling, an exact knowledge of the heat flux and temperature distributions on the airfoils surface is necessary. Heat transfer is influenced by various factors, like pressure distribution, wakes, surface curvature, secondary flow effects, surface roughness, free stream turbulence, and separation. Each of these phenomenons is a challenge for numerical simulations. Among numerical methods, large eddy simulations (LES) offers new design paths to diminish development costs of turbines through important reductions of the number of experimental tests. In this study, LES is coupled with a thermal solver in order to investigate the flow field and heat transfer around a highly loaded low pressure water-cooled turbine vane at moderate Reynolds number (150,000). The meshing strategy (hybrid grid with layers of prisms at the wall and tetrahedra elsewhere) combined with a high fidelity LES solver gives accurate predictions of the wall heat transfer coefficient for isothermal computations. Mesh convergence underlines the known result that wall-resolved LES requires discretizations for which y+ is of the order of one. The analysis of the flow field gives a comprehensive view of the main flow features responsible for heat transfer, mainly the separation bubble on the suction side that triggers transition to a turbulent boundary layer and the massive separation region on the pressure side. Conjugate heat transfer computation gives access to the temperature distribution in the blade, which is in good agreement with experimental measurements. Finally, given the uncertainty on the coolant water temperature provided by experimentalists, uncertainty quantification allows apprehension of the effect of this parameter on the temperature distribution

Transient heat transfer between a turbulent lean partially premixed flame in limit cycle oscillation and the walls of a can type combustor

“…It can be concluded that the prediction of the amplitude of pressure oscillations is highly dependent on the grid size, and can be over predicted by a factor of 2 or even 3 ( in the case that the first layer of grid in the solid is as far as 1 mm away from the fluid-solid interface). 13 Considering that the spatial scales of the solid temperature profiles are different in case of steady state or transient oscillatory heat transfer, care has to be taken into account the different requirements for the meshing in these two situations. Here, the knowledge of analytical solutions for transient solid thermal behavior can be used for interpretation, as available from the semi-infinite solid approach.…”

“…Unfortunately but no attempt has been made to evaluate the influence of transient heat transfer on characterization of Limit Cycle Oscillations (LCO) and creep life of the combustion system. Duchaine et al [11][12][13] used separate solvers and codes to solve the transport equations in the fluid and solid domains, developing a strategy to couple the solvers in an accurate and stable fashion. In this method, the accuracy of results relies heavily on the coupling and interface data exchange between the flow solver and the solid heat conduction code.…”

Large eddy simulation of flows in industrial compressors: a path from 2015 to 2035

“…Wall temperatures are unknowns of the problem that are resolved by the coupled simulations. This implies coupling a flow solver with a conduction solver [86] as well as the radiative transfer equation when compression is important, leading to high static temperatures (according to the Stefan-Boltzmann law, radiative heat transfer scales as T 4 ). …”

LES of knocking in engines using dual heat transfer and two-step reduced schemes

“…The energy equation inside the solid domain is solved by the AVTP solver [41]. Spatial discretization is handled with a second-order Galerkin scheme [42] and temporal integration uses a first-order forward implicit scheme.…”