Aerodynamic flow around a sport utility vehicle—Computational and experimental investigation Bahram Khalighi a,n , Shailesh Jindal b , Gianluca Iaccarino c a General Motors Global R&D, Warren, MI 48090, USA b ANSYS Inc. Ann Arbor, MI, USA c Mechanical Engineering Department, Stanford University, USA article info Article history: Received 29 November 2011 Received in revised form 31 March 2012 Accepted 13 April 2012 Available online 15 May 2012 Keywords: Immersed boundary CFD Aerodynamics of ground vehicles Steady RANS CFD Aerodynamics of SUVs abstract Standard CFD methods require a mesh that fits the boundaries of the computational domain. For a complex geometry the generation of such a grid is time-consuming, and often requires modifications to the model geometries particularly when one begins with the ‘‘raw’’ CAD data. This paper evaluates the newly developed Immersed Boundary (IB) approach which does not require mesh to be conformal to body and thus would speed up the process of the grid generation. The IB approach starts directly from CAD (STL) files, and has the potential in mitigating the process of CAD cleanup and surface meshing for the CFD simulations. The Reynolds-Averaged Navier–Stokes (RANS) solver based on the Immersed Boundary technique is used to investigate the aerodynamic flow field around a generic Sport Utility Vehicle (SUV). The simulations are compared with the experimental data for the same vehicle geometries. The experimental data include particle image velocimetry (PIV) velocity, surface pressure and drag coefficient measurements. The results show that the CFD simulations were able to track the flow very well for the generic SUV (both qualitatively and quantitatively). The predicted drag coefficients for the generic SUV model was within 5% of the measured values. Finally, the IB approach was used to predict the flow structures and the drag coefficient for a production SUV (Chevy Tahoe). In addition, the same geometry was simulated using the standard body-fitted approach for direct comparison with the IB simulations. The flow results of both simulations were similar. The simulated drag coefficients for the IB and the body-fitted approaches were within 3% and 3–7% of the measured value, respectively. It should be noted that, the paper is about a comparison of the IB to body-fitted approaches for automotive aerodynamics, and these approaches are simulations of wind tunnel measurements. Since the wind tunnel testing is a cold process using an ‘‘engine-off’’ condition, there was no need to perform the simulations with thermal boundary conditions activated for engines, exhaust, heat exchangers, etc. & 2012 Elsevier Ltd. All rights reserved. 1. Introduction The use of Computational Fluid Dynamics (CFD) codes by the engineering community to predict aerodynamic flow around ground vehicles has increased dramatically in the last few years (Gaylard, 2009; Holloway et al., 2009; Kitoh et al., 2009; Bagal, and Mulemane, 2010; Keating et al., 2008; Ahmed, 1998; Malviya et al., 2009). This rise in interest and use has resulted from improvements in the predictive capabilities of codes, reductions in the cost of computing technology, and inflation of the costs to perform experiments and to maintain experimental facilities. Most industrially relevant geometries are usually defined in the CAD environment and must be translated and cleaned up to generate water-tight surfaces for simulations using the standard body-fitted approach. This process is very tedious and time consuming. In addition, during this process small details are usually eliminated and overlapping surface patches are trimmed. A smooth water-tight surface mesh is then made which serves as a boundary condition for the volume mesh. There are mainly two types of approaches in volume meshing, structured and unstructured meshing (Owen, 1998; Thompson et al., 1985). In structured meshing, the governing equations are transformed into the curvilinear coordinate system aligned with the surface (Owen, 1998). It is trivial for simple shapes, however, it becomes extremely inefficient and time consuming for complex geometries. In the unstructured approach, there is no transformation involved for governing equations (Thompson et al., 1985). The integral form of governing equations is discretized and either a finite-volume or finite-element scheme is used. The information regarding the grid is directly incorporated into the discretization. Unstructured grids are in general successful for complex geometries. Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/jweia Journal of Wind Engineering and Industrial Aerodynamics 0167-6105/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jweia.2012.04.009 n Corresponding author. E-mail addresses: bahram.khalighi@gm.com (B. Khalighi), shailesh.jindal@ansys.com (S. Jindal), jops@stanford.edu (G. Iaccarino). J. Wind Eng. Ind. Aerodyn. 107–108 (2012) 140–148