Numerical computation of convective heat transfer in complex turbulent flows - Time to abandon wall functions?

In applying finite-difference and other numerical approaches to the computation of elliptic or three-dimensional turbulent flows it is usual to adopt wall functions to bridge with a single cell the very thin region adjacent to the wall where viscosity modifies the turbulence structure. The paper suggests that such approaches are now unnecessary from the standpoint of computational economy and, in many cases, may be replaced to advantage by a fine-grid, near-wall treatment that extends the numerical computations through the buffer zone to the wall itself. Examples are given of three flows currently being explored at UMIST where the abandonment of the wall-function approach has led to a marked improvement in the prediction of convective heat transfer rates.