Aerodynamic performances of rounded fastback vehicle Giacomo Rossitto, Christophe Sicot, Valérie Ferrand, Jacques Borée, Fabien Harambat To cite this version: Giacomo Rossitto, Christophe Sicot, Valérie Ferrand, Jacques Borée, Fabien Harambat. Aero- dynamic performances of rounded fastback vehicle. Proceedings of the Institution of Mechani- cal Engineers, Part A: Journal of Power and Energy, SAGE Publications, 2017, pp.1211-1221. 10.1177/0954407016681684. hal-01449542 HAL Id: hal-01449542 https://hal.archives-ouvertes.fr/hal-01449542 Submitted on 30 Jan 2017 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Open Archive TOULOUSE Archive Ouverte ( OATAO ) OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible. This is an author-deposited version published in: http://oatao.univ-toulous e.fr/ Eprints ID: 17498 To cite this version : Rossitto, Giacomo and Sicot, Christophe and Ferrand, Valérie and Borée, Jacques and Harambat, Fabien Aerodynamic performances of rounded fastback vehicle. (2017) Proceedings of the Institution of Mechanical Engineers, Part D: journal of automobile engineering. ISSN 0957-6509 Official URL: https://doi.org/10.1177/0954407016681684 Any correspondence concerning this service should be sent to the repository administrator: [email protected] Aerodynamic performances of rounded fastback vehicle Giacomo Rossitto1,2,ChristopheSicot2, Vale´rie Ferrand3, Jacques Bore´e2 and Fabien Harambat1 Abstract Experimental and numerical analyzes were performed to investigate the aerodynamic performances of a realistic vehicle with a different afterbody rounding. This afterbody rounding resulted in a reduction to drag and lift at a yaw angle of zero, while the crosswind performances were degraded. Rounding the side pillars generated moderate changes to the drag and also caused important lift reductions. A minor effect on the drag force was found to result from the opposite drag effects on the slanted and vertical surfaces. The vorticity distribution in the near wake was also analyzed to under- stand the flow field modifications due to the afterbody rounding. Crosswind sensitivity was investigated to complete the analysis of the aerodynamic performances of the rounded edges models. Additional tests were conducted with geometry modifications as spoilers and underbody diffusers. Keywords Rounded edges, drag, vorticity, wake analysis, crosswind, spoilers, underbody diffuser 1Introduction the backlight and the downstream shift of the rotating structures developing in the near wake. Fuller et al.4 For the last decade car manufacturers have been facing analyzed the benefits of rounding the rear side pillars the challenging task of reducing fuel consumption and on the Davis model. They observed that rounded edges CO2 emissions. In response to this, optimization algo- generate a different wake structure dominated by the rithms have been applied to generate new vehicle shapes interaction between the longitudinal vortices and the to minimize the aerodynamic drag. Since the obtained separated region. The rounded edges model resulted in optimum shapes have no brand differentiating details, a drag and lift reduction of 11% and 25% respectively. nowadays stylists are trying to give back a brand signa- The impact of afterbody rounding was mentioned also ture by proposing ‘‘non-conventional shapes. In that during the development of the Tesla Model S by Palin framework, the important rounding of the rear pillars et al.5 Important curved side pillars were avoided to becomes a differentiation strategy. The current study reduce the highly dynamic wake, which caused large suggests to quantify the influence of such afterbody variation in the base pressure. rounding on the flow field and on the drag development From the literature review, it appears that rounding over a fastback vehicle. afterbodies considerably affects the aerodynamic loads Very few papers have addressed the question of the and flow development, but a systematic investigation curvature of the rear edges in aerodynamic vehicle per- into the effects of changing the radii of the side formances. One of the first works was presented by 1 2 Gilhaus et al. and Howell. Thanks to balance mea- 1 surements, it was found that rounded pillars reduced PSA Groupe, Velizy–Villacoublay, France 2Institut Pprime, UPR-3346 CNRS, ENSMA, Universite´ de Poitiers, both drag and rear lift but yawing moment had a pro- France nounced increase. Only recently, were advanced tech- 3Institut Supe´rieur de l’Ae´ronautique et de l’Espace (ISAE-SUPAERO), niques used to achieve better understanding of the Universite´ de Toulouse, France afterbody rounding. Thacker et al.3 showed that round- Corresponding author: ing the edge between the roof and the rear slant of the Giacomo Rossitto, PSA Groupe, Centre Technique Velizy A, Velizy - Ahmed body results in a 10% drag reduction. Authors Villacoublay, France. attributed this reduction to the fully attached flow over Email: [email protected] backlight edges was not reported. To complement the meters long and has a test section which is 2 meters recent study proposed by the same authors6 over a sim- high, 5.2 meters wide and 6 meters long. It has a maxi- plified car model (Ahmed body), the present study mum free stream velocity of 53 m/s. The wind tunnel addresses this question on a realistic car model blockage ratio was 1.4%. The wind tunnel experiments equipped with various rear pillars curvatures. Special were conducted with a fixed ground and the model is care is taken to understand how the modified flow on placed over a false floor. All of the data were obtained the backlight interacts with the near wake and how this at 40 m/s which gives a Reynolds number based on the promotes drag and lift changes. length of the model of 2.6e6. Starred velocities are nor- Four rear ends were analyzed by combining PIV malized by the free stream velocity. A six components data to balance surface pressure measurements to fully balance was used to measure the aerodynamic forces characterize the flow structures and the associated acting on the model. The drag and lift coefficients were aerodynamic forces. The zero yaw case was first calculated as follows explored before focusing on the crosswind effects. F Complementary numerical simulations were systemati- d, l Cd, l = 1 2 cally applied to complete the experimental data and to 2 rU0S help the physical analysis. Geometry modifications, by where F is the force measured by the balance, r the den- means of spoilers and underbody diffusers, were inves- sity of the air, and S the frontal surface area of the tigated to understand their sensitivity to the side edges model. The precision of the balance was 0.001 for the rounding. drag coefficient Cd and 0.002 for the lift coefficient Cl. Static surface pressure was recorded by 40 pressure 2 Experimental and numerical set-up probes over the vertical surface of the rear end. Thirty- five probes were located on the driver side and the rest The model and its relevant dimensions are reported in were located on the passenger side to check the symme- Figure 1. Four rear ends are tested; they differ by their try of the static surface pressure over the vertical base. side pillars curvature. The curvature radius is given as a Even thought the non-symmetric underbody should percentage of the model span, i.e. 300 mm. The model have induced asymmetry in the flow the comparison equipped with sharp pillars having 0% radius is referred between the pressure probes, not reported here for brev- to as S0 and it is considered as the reference case. The ity, resulted in negligible differences. Those probes were others models are S8, S20, and S40. All of the rear ends connected to a SCANdaq 8000 acquisition system. The have the same curvature at the end of the roof to avoid acquisition rate was 40 Hz for 3000 samples giving 75 s flow separation. The corresponding radius is chosen to of time recording. The static surface pressure coefficient maximize the room for the rear passengers for a fixed at one point i was computed from the expression backlight angle of 23 degrees. The model features a rea- listic non symmetric underbody with an exhaust line. It P(i) À P0 CP(i)= 1 2 does not have open front air intake. The horizontal pro- 2 rU0 jection of the slanted surface, j = 440 mm, will be used as a reference length. Starred spatial coordinates are where P(i) is the static pressure of point (i), P0 the sta- normalized by the reference length. tic pressure measured upstream of the model. The static The experimental results reported in this work were accuracy of the system is 613 Pa, i.e. 0.015 Cp. obtained from tests conducted in the PSA Groupe wind Furthermore, PIV measurements were performed in tunnel of La Ferte´Vidame. The Eiffel wind tunnel is 52 the wake of the model. The laser sheet was set by a 2*120 mJ Nd:Yag Quantum Big Sky Laser. The Dantec Flowsense 4M mkII camera (2024 pixels by 2024 pixels) was equipped with 105 mm lenses which generated a 462 mm by 462 mm field of view. 2D PIV was per- formed on the symmetry (x,z) plane Y*=0. For all the PIV measurements, post-processing was performed with a final interrogation window of 16*16 pixels, after an initial window of 32*32 pixels, with an overlap of 50% in horizontal and vertical directions.