Large eddy simulations (LES) were made of flows around a generic ground vehicle with sharp edges at the rear end (an Ahmed body with a 25° angle of the rear slanted surface). Separation of the flow at the rear results in large regions with recirculating flow. As the separation is determined by the geometry, the Reynolds number effects are minimized. Resolution requirements of this recirculating flow are smaller than those in LES of wall attached flows. These two consequences of the geometry of the body are used to predict the experimental flow at relatively high Reynolds number. Recommendations are presented for the preparation and realization of LES for vehicle flows. Comparison of the LES results with the experimental data shows good agreement.

The flow around a finite circular cylinder with a height-to-diameter ratio of 6 was studied using large eddy simulation (LES). The cylinder was mounted on a ground plane where a thin boundary layer with a thickness of only 7% of the cylinder's diameter was formed. The flow resulting from the present LES was used to present a detailed picture of both the instantaneous and the time-averaged flow. Among the results of the present study is the explanation for the formation of two time-averaged tip vortices from two pairs of instantaneous vortices. The main recirculation region on the cylinder's top was found to be one vortex with both legs attached to the top surface, in agreement with some previous investigations. The mean flow in the near wake region was found to contain two horseshoe vortices, one large arch vortex stretching from the cylinder's top and another vortex resulting from the flow moving down below the rear stagnation point on the cylinder. The instantaneous horseshoe vortex near the ground was found to be unsteady and changed shape over time. The time-averaged flow showed three complete horseshoe vortices, in agreement with previous knowledge. The downwash process above the free end into the near wake occurs at a rather constant angle with respect to the plane normal to the streamwise direction. The same angle was preserved in the inclination of the arch vortex in the time-averaged flow.

Results of a large eddy simulation (LES) are used to explore the flow around a generic car model. A new, refined picture of this flow is established. Many parts and aspects of this flow are studied and explained. The development of the instantaneous flow and its resulting time-averaged flow are depicted. Large differences are found between the instantaneous and the time-averaged flows. Special attention is given to the flow above the rear slanted surface. The origin, the development, and the interactions of the instantaneous vortices in this part of the flow are presented for the first time. This instantaneous flow is shown to be very unsteady and to contain a large number of different vortices that range in size from those of the size of the body over the intermediate hairpin-like vortices to very small coherent structures. Besides the variety in the length scales, the flow covers a wide spectrum of the time scales from the relatively steady motion of the cone-like trailing vortices on the slanted edges to highly frequent collisions of the hairpin-like vortices in the region of the attachment on the rear slanted surface.

A new procedure for the optimization of aerodynamic properties of trains is presented, where simple response surface (RS) models are used as a basis for optimization instead of a large number of evaluations of the Navier—Stokes solver. The suggested optimization strategy is demonstrated in two flow optimization cases: optimization of the train's front for the crosswind stability and optimization of vortex generators (VGs) for the purpose of drag reduction. Besides finding the global minimum for each aerodynamic objective, a strategy for finding a set of optimal solutions is demonstrated. This is based on the use of genetic algorithms on RS models. The resulting Pareto-optimal solutions are used to explore the extreme designs and find trade-offs between design objectives. For the optimization of VGs, three different RS models are used: polynomial functions, radial basis neural networks (RBNN), and RBNN-enhanced polynomial RSs. The three approaches produce different results, and the combination of RBNN and polynomial functions in the last approach is found to be the best as it enables the construction of high-order polynomial functions, and the model's fit with the data is the best.

A brief review of large eddy simulation (LES) applications for different bluff-body flows performed by the author and his co-workers is presented. Examples of flows range from simple cube flows characterized by sharp edge separation over a three-dimensional hill where LES relies on good near-wall resolution, to complex flows of a tall, finite cylinder that contains several flow regimes that cause different challenges to LES. The second part of the paper is devoted to flows around ground vehicles at moderate Reynolds numbers. Although the present review proves the applicability of LES for various bluff-body flows, an increase of the Reynolds number towards the operational speeds of ground vehicles requires accurate near-wall modelling for a successful LES.

Large-eddy simulations are used to investigate the origin of the wake asymmetry and symmetry behind notchback Ahmed bodies. Two different effective backlight angles, ${\beta _1} = 17.8\mathrm{^\circ }$ and ${\beta _2} = 21.0\mathrm{^\circ }$ , are simulated resulting in wake asymmetry and symmetry in flows without external perturbations, in agreement with previous experimental observations. In particular, the asymmetric case presents a bi-stable nature showing, in a random fashion, two stable mirrored states characterized by a left or right asymmetry for long periods. A random switch and several attempts to switch between the bi-stability are observed. The asymmetry of the flow is ascribed to the asymmetric separations and reattachments in the wake. The deflection of the near-wall flow structures behind the slant counteracting the asymmetry drives the wake to be temporarily symmetric, triggering the switching process of the bi-stable wake. The consequence of deflection that forces the flow structure to form on the opposite side of the slant is the decisive factor for a successful switch. Modal analysis applying proper orthogonal decomposition is used for the exploration of the wake dynamics of the bi-stable nature observed.