Determination of Convective Heat Transfer Coefficients on Surfaces of Coolant Passages
Since the local values for the convective heat transfer coefficient, on the surfaces of the 3- D coolant flow passages have been specified thus far, it is necessary to compute their actual values. This can be done using a reliable 3-D Navier-Stokes code with the best turbulence model available at the time. With the specified inlet coolant temperature and the already determined temperature distribution on the walls of the coolant flow passages, a single run w ith the Navier – Stokes code should predict a detailed distribution of h^y on the 3-D surfaces of the coolant flow passages.
If the predicted values of h^** are significantly different from those used in the previous tasks, the last two tasks should be repeated until convergence.
It is possible for thermal design and optimization to perform relatively efficiently, even in a fully 3-D case, if BEM codes arc used for thermal field analysis and inverse determination of unknown boundary conditions and optimization is performed using a hybrid genetic cvolution/gradicnt search constrained optimization algorithm. One specific conjugate heat transfer design and optimization scenario has been suggested for which all of the individual tasks have been proven to work by the author’s research team. The main uncertainty of the entire conjugate heat transfer design process still rests w ith the issue of reliability of turbulence models in typical 3-D Navicr – Stokes flow-field analysis codes to predict surface temperatures and heat fluxes.