Internatonal Journal of Mechancal & Mechatroncs Engneerng IJMME-IJENS Vol: No: 03 24 Computatonal Analyss of Actve Flow Control to Reduce Aerodynamcs Drag on a Van Model Harnald a, Budarso b, Rustan Taraa c and Sabar P. Smanungalt d Department of Mechancal Engneerng Faculty of Engneerng Unversty of Indonesa, Kampus UI-Depo,, Jawa Barat, 6424, Indonesa Abstract Method of actve flow control can be appled to reduce aerodynamc drag of the vehcle. It provdes the possblty to modfy locally the flow, to remove or delay the separaton poston or to reduce the development of the recrculaton zone at the bac as well as the separated swrlng structures around the vehcle. In ths study, a passenger van s modeled wth a modfed form of Ahmed's body by changng the orentaton of the flow from ts orgnal form (modfed/reversed Ahmed Body). Ths model s equpped wth sucton and blowng on the rear sde to comprehensvely examne the pressure feld modfcatons that occur n order to modfy the near wall flow toward reducng the aerodynamcs drag. The computatonal smulaton used s -epslon flow turbulence model. In ths confguraton, the front part of body was nclned at an angle of 3 wth respect to the horzontal. The geometry s placed n a 3D-rectangular numercal doman wth length, wdth and heght equal to 8l, 2l and 2l, respectvely. The sucton and blowng veloctes are set to m/s, m/s, m/s and m/s, respectvely. The results show that aerodynamc drag reductons close to.83 % for sucton and 4.38 % for blowng have been obtaned. Index Terms drag reducton, actve flow control, sucton, blowng, reversed Ahmed body. A I. INTRODUCTION ccordng to the conclusons of Internatonal Energy Agency n World Energy Outloo 2007, the gas emssons wth greenhouse effect wll ncrease close to 7% n 2030 wth strong effects on the envronment and the clmate []. The human actvtes became man cause of the ncrease of the greenhouse gases effect and average global temperature. The actvtes ncluded the transportaton sector Manuscrpt receved May, 20. Ths wor was supported by Incentve of Fundamental Research grant, The Mnstry of Research and Technology Republc of Indonesa under contract no. RD-20-0863 a Dr. Harnald s wth the Department of Mechancal Engneerng Faculty of Engneerng Unversty of Indonesa, Depo, Jawa Barat, 6424, Indonesa (Phone:+62-2-7270032;Fax:+62-2-7270033;e-mal:harnald@eng.u.ac.d). b Prof. Budarso s wth the Department of Mechancal Engneerng Faculty of Engneerng Unversty of Indonesa, Depo, Jawa Barat, 6424, Indonesa (Phone:+62-2-7270032;Fax:+62-2-7270033;e-mal:mftbd@eng.u.ac.d). c Mr. Rustan Taraa s a PhD student at the Department of Mechancal Engneerng Faculty of Engneerng Unversty of Indonesa; e-mal: rustan_taraa@yahoo.com d Mr. Sabar P. Smanungalt s a Master degree student at the Department of Mechancal Engneerng Faculty of Engneerng Unversty of Indonesa; e- mal: sp.smanungalt@gmal.com where the growth number of automoble s rapdly ncreasng and mae the fuel consumpton ncreases as well. It tends to create harmful effects on the envronment because t ncreases ar polluton n the world. Based on these problems t has become a must for automoble ndustry n the world to mmedately create an envronmentally frendly automobles and effcent n fuel consumpton. Fuel consumpton of automoble s related to ts aerodynamcs drag, and the magntude of aerodynamcs drag s hghly nfluenced by separaton flows around ts shape. Meanwhle, the flow around a travelng automoble s complex and presents nonlnear nteractons between dfferent parts of the automoble so that many research nsttutons and ndustral laboratores have been focusng ther nvestgatons automotve aerodynamcs wth numercal studes [2]. It s necessary to modfy locally the flow, to remove or delay the separaton poston or to reduce the development of the recrculaton zone at the bac and of the separated swrlng structures. Ths can be manly obtaned by controllng the flow near the wall wth or wthout addtonal energy usng actve or passve devces [3]. Sgnfcant results can be obtaned usng smple technques [4,]. Many actve control technques whch have been developed by focusng on local nterventon n wall turbulence deal wth steady blowng or sucton [6,7,8]. A blowng devces nstalled n an ONERA D profle can shft or even prevent the flow separaton to occur [9]. A local sucton system located on the upper part of the rear wndow s capable of elmnatng the rear wndow separaton on smplfed fastbac car geometry. Aerodynamc drag reductons close to 7% have been obtaned []. Other numercal wors [] usng Lattce Boltzmann method to an Ahmed body model ndcated some mportant parameters of actve control to mprove the aerodynamcs performance of a vehcle. Most of prevous numercal study n flow around automoble s body used the geometry suggested by Ahmed [2] as shown n Fg.. Ahmed body allows reproducng flows phenomena around the vehcles and mang a possblty to apply actve control. However, to be practcally mplemented n controllng the flow separaton n the automotve applcaton the actve control methods stll need further comprehensve nvestgatons to obtan some fundamental nsghts of the governng mechansm of separaton control. Hence, the current nvestgaton was a part of a long-term fundamental
Internatonal Journal of Mechancal & Mechatroncs Engneerng IJMME-IJENS Vol: No: 03 2 nvestgaton to develop an actve control to the turbulent flow separaton whch s a fundamental phenomenon governng the aerodynamcs performance of vehcle body. In ths study, a passenger van s modeled wth a modfed form of Ahmed's body by changng the orentaton of the flow from ts orgnal form (modfed/reversed Ahmed Body). Ths model s equpped wth sucton and blowng on the rear sde to comprehensvely examne the pressure feld modfcatons that occur n order to modfy the near wall flow toward reducng of aerodynamcs drag. The wor was conduted by usng a Gambt 2.4 software to generated the grd. Meshng type was tetra/hybrd element wth hex core type and the grd number was more than.7 mllon n order to ensure detal dscretzaton and more accurate calculaton results. The boundary condton were nlet velocty of 6.7 m/s. Mean free stream at far upstream regon was assumed n a steady state condton and unform. The Reynolds number assocated wth length L of the geometry s Re = 2.98. The blowng and sucton velocty are set m/s, m/s, m/s and m/s, respectvely. Detals of computaton condton are gven n Table. Table. Computaton condton Fg. Orgnal Ahmed model (dmensons n mm) [2] II. MODELLING AND NUMERICAL SIMULATION The numercal smulatons presented n ths paper were conducted on a modfed/reversed Ahmed body whch has geometrcal rato 0.2 to the orgnal sze. The model geometry was defned by ts length (l = 0.26m), wdth (w = 0.0972 m) and ts heght (h = 0.072 m). In ths confguraton, the front part of body was nclned at an angle of 3 o wth respect to the horzontal. The body s placed n a 3D rectangular numercal doman of length L = 8l, wdth W = 2l and heght H = 2l as shown n Fg. 2. (a) Computaton Condton Model settng 3D, Steady state Flud Ar Flud propertes Densty.22 g/m 3 Vscosty 0.00007894 g/ms Boundary Van model Wall condton Pressure outlet Pressure outlet wthout flow Velocty nlet Velocty nlet control Wall Wall Boundary Van model Wall condton wth Pressure outlet Pressure outlet sucton/blowng Velocty nlet Velocty nlet Wall Wall Sucton/blowng Velocty nlet Sucton2/blowng2 Velocty nlet Sucton/blowng m/s, m/s, m/s and m/s velocty The governng equatons were solved numercally by fnte volume approach usng a commercal solver Fluent 6.3 [3]. The turbulence model used n the computaton was a standard -epslon model showed n Eq. () and (2) t u Y t C M x S x j t G Gb x j t u C G C G 2 2 x S x j x j 3 b () (2) where : C ε =.44, C 2ε =.92, C μ =0.09, S =.0, S ε =.3. (b) Fg 2. (a).computatonal flow doman. (b).geometrcal dmensons of van model A dmensonless coeffcent, called drag coeffcent and related to the drag force actng on the bluff body, s defned as follows: Fd Cd (3) 2 V S 2
Internatonal Journal of Mechancal & Mechatroncs Engneerng IJMME-IJENS Vol: No: 03 26 In ths expresson, ρ represents the ar densty, V s the free stream velocty, S s the cross secton area and F d s the total drag force actng on the car projected on the longtudnal drecton. Note that the drag force F d can be decomposed nto a sum of a vscous drag force and a pressure drag force as: F d w sn ds p cos ds (4) Combnng (3) and (4), the drag coeffcent can be expressed as: C d w sn ds 2 V S 2 C p cos ds S () (a). Sucton velocty, U sc = m/s Where w = (du/dy) w s the wall shear stress evaluated from the wall velocty gradent and C p = (p-p )/(V 2 /2) s the pressure coeffcent evaluated from the wall pressure dstrbuton. III. RESULTS AND DISCUSSION A. Pressure Dstrbuton Some selected results from computatonal are descrbed n the followng fgures. Fg. 3 shows computatonal result of the pressure coeffcent dstrbuton wthout flow control n the rear part of the reversed Ahmed body wth the upstream velocty at 6.7 m/s evaluated at several span wse locatons. From Fg. 3, t can be seen that the mnmum value of pressure coeffcent s -.389 at y/h = and z/w = -/2 n the rear sde of the body. (b). Sucton velocty, U sc = m/s (c). Sucton velocty, U sc = m/s Fg 3. Pressure coeffcent dstrbuton wthout flow control n the rear part of the reversed Ahmed body Meanwhle, Fg. 4 (a-d) descrbes the effect of varous sucton veloctes n the rear part of the reversed Ahmed body to the pressure coeffcent dstrbuton wth the upstream velocty at 6.7 m/s. The sucton veloctes are m/s, m/s, m/s and m/s. Closer nsght to Fg. 4 (a-d) ndcates that by ntroducng sucton, the locaton of mnmum value pressure coeffcent shfted sgnfcantly to y/h = ; z/w = -/4 and y/h = ; z/w = /4 n the rear sde of the body. The mnmum value of pressure coeffcent s summarzed n Table 2. (d). Sucton velocty, U sc = m/s Fg 4. Pressure coeffcent dstrbuton wth varous sucton velocty n the rear part of the reversed Ahmed body
Internatonal Journal of Mechancal & Mechatroncs Engneerng IJMME-IJENS Vol: No: 03 27 Table 2. The mnmum value of pressure coeffcent wth sucton Sucton velocty, U sc (m/s) Pressure Coeffcen, C p -.0097 -.488 -.2339 -.3372 Fgures (a-d) shows the effect of varous blowng velocty n the rear part of the reversed Ahmed body to the pressure coeffcent dstrbuton wth the upstream velocty at 6.7 m/s. The blowng veloctes are m/s, m/s, m/s and m/s, respectvely. By the ntroducton of a blowng n the rear part of the body t can be seen from Fg. (a-d) that the locaton mnmum of value of pressure coeffcent shfted sgnfcantly to y/h= ; z/w= -/4 and y/h = ; z/h = /4 n the rear sde of the body for blowng velocty m/s, m/s and m/s. For blowng velocty m/s, the locaton mnmum value of pressure coeffcent at y/h = /2; z/w = -/4 and y/h = /2; z/w = /4 n the rear part of the body. Table 3 summarzes the alteraton of the mnmum value of pressure coeffcent. (c). Blowng velocty, U bl = m/s Table 3. The mnmum value of pressure coeffcent wth blowng Blowng velocty, U bl (m/s) Pressure Coeffcent, C p -0.846-0.74-0.7680-0.9477 (d). Blowng velocty, U bl = m/s Fg. Pressure coeffcent dstrbuton wth varous blowng velocty n the rear part of the reversed Ahmed body B. Turbulence Intensty The characterstcs of turbulence ntensty n the rear part of the van model wthout and wth flow control are presented n Fgs. (6) and (7). Fg. 6 shows the turbulence ntensty wthout flow control n the rear part of the reversed Ahmed body wth the upstream velocty at 6.7 m/s. It can be seen that the maxmum value of turbulence ntensty s about.99% wthout flow control. (a). Blowng velocty, U bl = m/s (b). Blowng velocty, U bl = m/s Fg 6. Turbulence ntensty wthout flow control n the rear part of the reversed Ahmed body.
Internatonal Journal of Mechancal & Mechatroncs Engneerng IJMME-IJENS Vol: No: 03 28 (a). Sucton velocty, Usc = m/s (a). Blowng velocty, U bl = m/s (b). Sucton velocty, Usc = m/s (b). Blowng velocty, U bl = m/s (c). Sucton velocty, Usc = m/s (c). Blowng velocty, U bl = m/s (d). Sucton velocty, Usc = m/s Fg 7. Turbulence ntensty wth varous sucton velocty n the rear part of the reversed Ahmed body (d). Blowng velocty, U bl = m/s Fg 8. Turbulence ntensty wth varous blowng velocty n the rear part of the reversed Ahmed body
Internatonal Journal of Mechancal & Mechatroncs Engneerng IJMME-IJENS Vol: No: 03 29 Furthermore, Fg. 7 (a-d) shows the effect of varous sucton veloctes n the rear part of the reversed Ahmed body to the turbulence ntensty dstrbuton wth the upstream velocty at 6.7 m/s. Meanwhle n Table 4 the maxmum turbulence ntenstes under sucton veloctes of m/s, m/s, m/s and m/s are summarzed. These results ndcate that there s a decrease n the turbulence ntensty by the use of sucton mechansm n the rear part of the body. The largest decrease of turbulence ntensty about 0.20 % s obtaned at sucton velocty of m/s. Table 4. The maxmum value of turbulence ntensty wth sucton Sucton velocty, U sc (m/s) Turbulence ntensty, (%).70.7.84 2.23 n the flow topology for U sc < 0.30 U o. Conversely, durng the second phase n fg. 9, the drag shows tendency to ncrease and the reducton obtaned are.83% for U sc = 0.30U o and 4.3% for U sc = 0.90U o. Hence, from Fg. 9 t can be ndcated that there s an optmum sucton velocty to obtan drag reducton.e. U sc = 0.30U o. Closer nsght to Fg. n the frst phase, the control performance ncrease very rapdly wth the blowng velocty. The reducton obtaned are 4.38% for U bl = 0.06U o. These result suggest sgnfcant modfcatons n the flow topology for U bl < 0.06U o. Conversely, durng the second phase n Fg., the drag show a tendency to ncrease and the reducton obtaned are 4.38% for U bl = 0.06U o and.32% for U bl = 0.90U o. Hence, Fg. ndcates that there s an optmum blowng velocty to acheve maxmum drag reducton.e. U bl = 0.06U o Fgures 8 (a-d) descrbes the effect of varous blowng veloctes n the rear part of the reversed Ahmed body to the turbulence ntensty the upstream velocty at 6.7 m/s. The blowng velocty are m/s, m/s, m/s and m/s. Meanwhle n Table the maxmum turbulence ntenstes under blowng veloctes of m/s, m/s, m/s and m/s are summarzed. These results ndcate that there s a decrease n the turbulence ntensty by the use blowng mechansm n the rear part of the body. The largest decrease of turbulence ntensty about 0.36% was obtaned at blowng velocty m/s. Table. The maxmum value of turbulence ntensty wth blowng Blowng velocty, U bl (m/s) Turbulence ntensty, (%).63.67.77.94. Generally, the turbulence ntenstes obtaned by the use of sucton and blowng are lower than those wthout flow control n the rear part of the reversed Ahmed body. Ths alteraton seems to occur due to addtonal energy from sucton and blowng whch modfy the flow structure so that the separaton can be decreased or elmnated. Fg 9. Drag coeffcent as a functon of sucton velocty. C. Aerodynamcs Drag Reducton The nfluences of sucton and blowng are analyzed accordng to the aerodynamc forces appled to the geometry, wth and wthout control. The results are presented through aerodynamc drag reducton wth respect to the reference confguraton (wthout control). The mean drag coeffcents, obtaned at dfferent veloctes of sucton and blowng are ndcated n Fgures 9 and, respectvely. In the fgures, the drag reductons obtaned wth respect to the reference confguraton (wthout control) are also ndcated. From Fg. 9 n the frst phase, the control performance ncrease very rapdly wth the sucton velocty. The reducton obtaned are 4.70 % for U sc = 0.06U o and.83 % for U sc = 0.30U o. These results suggest sgnfcant modfcatons Fg. Drag coeffcent as a functon of blowng velocty. Generally, by comparng the results obtaned from the use of sucton wth those from the use of blowng n the rear part of the reversed Ahmed body, t can be fgured out that the drag reductons obtaned by sucton s greater than by blowng.
Internatonal Journal of Mechancal & Mechatroncs Engneerng IJMME-IJENS Vol: No: 03 30 IV. CONCLUSION An actve flow control soluton by sucton and blowng are tested to reduce the aerodynamc drag on reversed Ahmed body. The maxmum drag reductons assocated wth these modfcatons are close to.83% and the ncrease of sucton velocty ncrease at U sc > 0.30U o does not mprove such a reducton sgnfcantly. Lewse, the drag reductons decrease when the sucton velocty dmnshes below 0.30U o. In the other hand, n case of blowng n rear sde on reversed Ahmed body, the drag reducton are close to 4.38% and the ncrease of blowng velocty at U bl > 0.06U o does not mprove such a reducton sgnfcantly. Moreover, the drag reductons decrease when the blowng velocty dmnshes below 0.06U o. The present study also show that ths s a good test case for further development. In addton, the results presented n ths paper gve nformaton about potental of actve flow control by sucton and blowng n the automoble ndustry. The results should, however, be corroborated by expermental results and tested on real car flow confguraton. ACKNOWLEDGMENT The authors would le to than Mr. Freddy Lay, Mr. Andre Grvanzy and Mr. Ahmad Tr Ageng.S for helpng the preparaton of computaton facltes. REFERENCES [] World Energy Outloo 2007, Executve Summary, Chna and Inda Insghts, Internatonal Energy Agency IEA, 2007. [2] M. Gad-El-Ha, Modern developments n flow control, Appl Mech Rev. vol. 996, no. 9, pp. 36 379, 2006. [3] H.E. Feldler, and H.H. Fernholz, On the management and control of turbulent shear flows, Prog Aero Sc, vol. 27, 990. [4] C.H. Bruneau, and I Mortazav. Numercal modellng and passve flow control usng porous meda, Comput Fluds, vol. 37, no., 2008. [] A.Rosho, and K. Koeng, Interacton effects on the drag of bluff bodes n tandem, In: Sovran G, Morel T, Mason WT, edtors. Symposum on aerodynamc drag mechansms of bluff bodes and road vehcles, Plenum Press, 978. [6] P.A. Krogstad, and A. Kourane, Some effects of localzed njecton on the turbulence structure n a boundary layer, Phys Fluds, vol.2, pp. 2990 2999, 2000. [7] Par J, Cho H, Effects of unform blowng through blowng or sucton from a spanwse slot on a turbulent boundary layer flow, Phys Fluds vol., pp. 309 3, 999. [8] M. Sano, and N. Hrayama, Turbulent boundary layer wth njecton and sucton through a slt, Bull JSME, vol. 28, pp. 807 84, 98. [9] T. Ivanc., and P. Glléron., Reducton of the Aerodynamc Drag Due to Coolng System: An Analytcal and Expermental Approach, SAE Paper, No. 200-0-7, 2004. [] M. Roumeas, P. Glleron, and A. Kourta, Drag Reducton by Flow Separaton Control on a Car after Body, Internatonal Journal for Numercal Method n Fluds, vol. 60, pp. 222-240, 2009. [] M. Roumeas, P.Glleron, and A. Kourta, Separated flow around the rear wndow of a smplfed car geometry, Journal of Fluds Engneerng, vol. 30, 2008. [2] S.R. Ahmed, G. Ramm and G. Faltn, Some salent features of the tmeaveraged ground vehcle wae, SAE techncal paper seres, no. 840300, Detrot, 984. [3] User s Gude Manual of Fluent 6.3.2, September 2006