Xingsi Han e-mail: xingsi@chalmers.se Sinis ˇa Krajnovic ´ 1 e-mail: sinisa@chalmers.se Division of Fluid Dynamics, Department of Applied Mechanics, Chalmers University of Technology, SE-412 96 Gothenburg, Sweden Very Large Eddy Simulation of Passive Drag Control for a D-Shaped Cylinder The numerical study reported here deals with the passive ﬂow control around a two-dimensional D-shaped bluff body at a Reynolds number of Re ¼ 3:6  10 4 . A small circular control cylinder located in the near wake behind the main bluff body is employed as a local disturbance of the shear layer and the wake. 3D simulations are carried out using a newly developed very large eddy simulation (VLES) method, based on the stand- ard k  e turbulence model. The aim of this study is to validate the performance of this method for the complex ﬂow control problem. Numerical results are compared with avail- able experimental data, including global ﬂow parameters and velocity proﬁles. Good agreements are observed. Numerical results suggest that the bubble recirculation length is increased by about 36% by the local disturbance of the small cylinder, which compares well to the experimental observations in which the length is increased by about 38%. A drag reduction of about 18% is observed in the VLES simulation, which is quite close to the experimental value of 17.5%. It is found that the VLES method is able to predict the ﬂow control problem quite well. [DOI: 10.1115/1.4024654] 1 Introduction The topic of ﬂow control for aerodynamic drag has attracted increasing interest in recent years, not only from industry but also from academic research [1]. The main contribution of the aerody- namic drag is the pressure drag that arises from separated ﬂow in the rear and the vortex shedding in the wake. This makes it neces- sary to control the ﬂow locally in the wake region in order to elon- gate the near-wake region and suppress or delay the shear layer interactions. Various methods for bluff body ﬂow control have been proposed and investigated. The major achievements of bluff body ﬂow controls were recently reviewed [2]. There are several ideas that have been pursued for the ﬂow con- trol purposes, which can be generally classiﬁed as passive control or active control method. The present study concerns the passive ﬂow control, which is of great interest for many industrial applica- tions. Passive control devices, such as riblets [3] and vortex gener- ators [4], have shown to be quite effective in delaying ﬂow separation. There are also many other methods proposed [5,6]. A simple method is discussed here that uses a small secondary body placed behind the main bluff body in which the drag has to be reduced [7–10]. It was found that the wake behind the main body is disturbed by the small body, tending toward complete shedding suppression for some certain locations of the small body, which results in an obvious drag reduction. The early work of Sreenivasan and Strykowski [7] concerns the ﬂow around a main circular cylinder disturbed by a much smaller cylinder. Since then, several studies have investigated this ﬂow conﬁguration at various Reynolds number both experimentally and numerically [7,8,11–13]. In order to avoid the effect of ﬂow attachment on the main bluff body, such as a circular cylinder, a D-shaped bluff body was recently introduced and the ﬂow conﬁg- uration was experimentally studied [9,10,14]. In the present study, numerical method will be used to investigate the same ﬂow con- ﬁguration at a Reynolds number of Re ¼ 3:6  10 4 . Although the efﬁciency of this ﬂow control method has been extensively conﬁrmed in experimental investigations, it is still not easy to recover the effects in a numerical study. As pointed out by Spalart and Mclean [1], the situation of computational ﬂuid dy- namics (CFD) studies is difﬁcult for ﬂow control and there are major issues with computing cost and turbulence modeling. This is because numerical study of ﬂow control requires accurate pre- dictions of ﬂow structures, even small ones. To recover the ﬂow control effects in experiment by numerical method, the natural ﬂow should be accurately predicted ﬁrst, and then on the basis of the results of that prediction, the interactions of the control device and the ﬂow structures in the natural ﬂow should be accurately predicted. For the bluff body ﬂow, the ﬂow control effects are highly related to the shear layer instabilities and vortex move- ments. A numerical method should be able to reproduce the pro- cess accurately so as to be able to reproduce the ﬂow control effects. Reynolds-averaged Navier–Stokes (RANS) turbulence models cannot accurately represent small-scale ﬂow dynamics associated with ﬂow control, and therefore, the RANS calculation is generally untrustworthy in ﬂow control problems [1]. Large eddy simulation (LES) is a candidate for studying ﬂow control problems, as it can resolve the large ﬂow structures well, and was found that LES can be used to simulate ﬂow control problems of vehicle-like bluff bodies at moderate Reynolds numbers [15–17]. However, the LES technique still needs a large computational effort for high Re number ﬂows encountered in engineering aero- dynamics. With the rapid development of the hybrid RANS–LES methodology, attempts have been made to study aerodynamic problems using hybrid methodology. Guilmineau et al. [18] inves- tigated the main ﬂow features of the Ahmed body, applying detached eddy simulation (DES) approach. Several versions of the DES models based on the k  x shear stress transport turbulence model are used. Hemida and Krajnovic ´[19] used DES to study ﬂow around a double-deck bus inﬂuenced by the crosswind. The ﬂow around a rudimentary landing gear is numerically studied using partially averaged Navier–Stokes (PANS) and LES meth- ods, and PANS shows clear advantages compared with LES [20]. In the present study, a newly developed VLES method [21,22], one of such kind of hybrid methodology, is further extended to the ﬂow control problem. The new VLES method is a uniﬁed simula- tion approach that can change seamlessly from RANS to direct numerical simulation (DNS), depending on the numerical resolu- tion. It has been evaluated in several canonical ﬂow applications and was found that good predictions can be obtained using a quite coarse mesh resolution. As discussed before, the ﬂow control problem is challenging and especially for turbulence modeling. In 1 Corresponding author. Contributed by the Fluids Engineering Division of ASME for publication in the JOURNAL OF FLUIDS ENGINEERING. Manuscript received October 8, 2012; ﬁnal manuscript received May 24, 2013; published online July 23, 2013. Assoc. Editor: Michael G. Olsen. Journal of Fluids Engineering OCTOBER 2013, Vol. 135 / 101102-1 Copyright V C 2013 by ASME Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 11/11/2013 Terms of Use: http://asme.org/terms