The near wall region in turbulent boundary layer flows is important for many reasons and is at the focus of many investigations. It is however very difficult to study this region experimentally, partly because of strong demands on spatial resolution, but also because the presence of the wall influences the performance of the measurements. In this contribution we try to detect and quantify problems with very near wall measurements using laser Doppler anemometry, and, in particular, with the determination of the wall shear stress. Several experiments are considered: a wall jet, and smooth and rough boundary layers, both with zero and favorable pressure gradients. Three methods were used to obtain the wall shear stress. Two of them are based only on velocity measurements and are referred to as "the momentum integral method" and "the wall gradient method". The third method, oil film interferometry was used only in the case of the wall jet. In all experiments, problems with spatial resolution were encountered, with the exception of the smooth flat plate at very low speed. This was the only case in which the wall gradient of the mean velocity could be determined with reasonable accuracy. The momentum integral method worked for the boundary layers, but not for the wall jet, where a small secondary motion was present in spite of purely two-dimensional boundary conditions. The momentum integral method was the only one that could be applied to the rough boundary layers. The oil film method worked well in the wall jet, but could naturally not be applied on the rough surfaces.
展开▼
