1. This section addresses the crossflow instability that causes the breakdown to turbulence in three-dimensional boundary layers that are characteristic of swept-wing flows. The review of Saric12provides an extensive list of References for the recent experiments, including the DLR experiments in Germany on a swept flat plate, a Russian swept-flat-plate experiment, the Centre d'Etudes et de Recherches de Toulouse (CERT)/ONERA experiments on swept wings, the Institute of Fluid Science (IPS) work in Sendai on cones and spheres, and the ASU swept-wing experiments. The archival databases appropriate for CFD validation include the following. The DLR efforts in Germany are reported in Refs. 129-133. The CERT/ONERA efforts in France are reported in Refs. 134-138. The Russian efforts are summarized in Ref. 139. The ASU work is summarized in Refs. 25, 32, and 140-144. The IPS work in Sendai is described in Refs. 145-149. These papers established the existence of both traveling and stationary crossflow vortices, saturation of the stationary crossflow vortex, the nonlinear secondary instability leading to transition, and the sensitivity to freestream disturbances and surface roughness. Here are some great challenges to the computationalist.
2. One of the important results to come out of the DLR group is the set of data that show early saturation of the disturbance amplitude and the failure of linear theory to predict the growth of the instability. They also report distorted mean profiles, similar to those of Michel et al.153Dagenhart et al.142and Kohama et al.32due to the presence of the stationary corotating vortices. A similarity between the DLR and ASU experiments is the high N factors and the high amplitude of the mean-flow distortion (10-20%). It is not surprising that linear theory fails.
3. The NPSE approach has recently been validated for three-dimensional flows subjected to crossflow disturbances in Refs. 26-29. Here a detailed comparison of NPSE results with the experimental measurements of Reibert et al.25show remarkably good agreement. The configuration is an NLF(2)-0415 45-deg-swept airfoil at -4-deg angle of attack. A spanwise array of roughness elements is used near the airfoil leading edge to introduce 12-mm spanwise-periodic crossflow disturbances into the boundary layer. The initial conditions for the NPSE calculation were obtained by solving the local LST equations at a 5% chord location for the fundamental [mode (0, 1)] and adjusting its rms amplitude such that the total disturbance amplitude matched that of the experiment at 10% chord.
4. Figure 1 shows the comparison of experimental N-factor curves with LPSE, NPSE, and LST. The NPSE results include curvature effects. It is clear that the linear theories fail to accurately describe the transitional flow for this situation. After a region of linear growth, the disturbance modes achieve large amplitudes and interact nonlinearly, saturating at about 30% chord location. There is a large region of nonlinear interaction from 30-50% chord before transition occurs. 0 ""0.1 0.2 . 0.3 O4 0.5