N Ab cou of sta opp hav Nu un for con mi ent An coe An she sta nu exp dif int of alt pro req sou cur urb sus the suf rea far tur ide be [2] wh Engineerin www.etasr Numer a Co National R bstract—This p unter-rotating a two stage ve age. The mode position to eac ve been perfo umerical analy steady simulat r the two stage nformal interf nimize numer tropy changes nsys Fluent so efficient for e nother, more edding, impin ages at differ merical resu perimental dev fferent free stre Keywords-CF Concerns re terest of resea renewable en ernative to c oportion of quirements in urces. Both v rrently employ ban areas. V stainable energ eir robustness ffer from self ach a self-sus rms make use rbines, the VA eally suited fo often unpredi ] are limited t hen compared ng, Technology r.com rical a ounter Ion M CFD De Research and D Gas Turbin Bucharest ion.malael@ paper investig vertical axis w ertical H-type el has an inne ch other. Both ormed in ord ysis is based tions using a D s to self-adjust faces are defin rical artifacts s. Fully turbul oftware. One k each stage at qualitative, ou ngement and rent positions ults have be vice which wa eam speeds. FD; counter rota I. IN egarding glob archers for inn ergy. Wind en conventional p wind power January of vertical and h yed, with a he VAWTs repre gy landscape f s in terms of f-starting lack stained operat e of the large AWT indepen or urban enviro ictable [1]. Co to a TSR of o d against the D y & Applied Sci Malael & D and Ex r-Rota Malael epartment Development I nes COMOTI t, Romania @comoti.ro ates the conce wind turbine turbine with t er and an oute numerical an er to validate on the use DOF type of an t their rotation ned between t such as art lent URANS w key outcome w different tip utcome is the overall intera and scenari een validated as analyzed in ating; wind tun TRODUCTION bal warming novation and r nergy is emerg power plants peaked at 2017, surpass orizontal axis eightened inter esent a maj for several dec wind orienta , requiring sta tion point. W er, but less ve dence of the onments wher onstructively, one unit, maki Darrieus types ience Research Dragan: Numer xperim ating V Institute for ept of a conc (VAWT), cons three blades on er stage, rotat d experimenta e this new co of 2.5-dimens nalysis which n speed. Sliding these subdoma tificial relatio were carried o was the mome wind speed v e analysis of v action betwee ios. Ultimately d using a the wind tun nnel; wind turbi have favore research in the ging as a rene . In Romania 20% of na sing nuclear p s wind turbine rest for VAWT or part of cades, mostly d ation although art-up assistan While offshore ersatile, axial inflow makes re flow pattern the Savonius ing them ineff s [3]. On the h V rical and Exper mental Vertica centric sisting n each ting in al tests oncept. sional, allows g mesh ains to ons or out in entum values. vortex en the y, the scaled nnel at ine d the e field ewable a, the ational power es are Ts for urban due to h they nce to wind wind s them ns can types ficient other han star hav from or con of a two cou on t con gen rota of t spe sec that and can con bein app win pos effe pap sup Vol. 8, No. 4, 20 rimental Efficie Effici al Axi National nd, Darrieus rting, making h ve seen a signi m the classica rigid [7] and nfigurations. T an original var o stages are unter-rotating top of one ano nstructive adva nerator. Howe ating and in p the machine, eed ratio and ondary stage i t would otherw d hence to ele n approach Bet In profiling th nsidered the N ng well under plications. Tab nd turbine. By sition of the tu ect. The CAD per is presente pporting system 018, 3282-3286 ency Evaluation iency E is Win Valeri CFD D Research and Gas Turbi Buchare valeriu.drag VAWTs hav high torque ap ificant number al Savonius and flexible [8] The focus of th riation of the D assembled c configurations other, have bee antages in term ever, having parallel does li leaving the c power coeffi in series, a sig wise be inacc ectric power. T tz's limit [11]. II. CFD he blades of t NACA 0021 ai rstood by its u ble I presents y inverting the urbine blades D assembly of ed in Figure 1 m can be obser Fig. 1. 3D W 6 n of a Counter- Evalu nd Tur iu Dragan Department d Developmen ines COMOTI est, Romania gan@comoti.r ve difficultie pplications diff r of VAWT v d Darrieus typ combinations his paper is to Darrieus [9] co coaxially and s, such as sta en proposed ea ms of their link the two hal ittle to further lassical relatio cient unaltere gnificant part cessible is con This way, the D SIMULATION this new wind irfoil which ha use in other s s the geometr e leading edge we achieve th f the embodim , where the th rved. Wind turbine mod 3282 -Rotating Vertic uation rbine nt Institute for I ro es during rot fficult. Recent ariations, devi pes to hybrids of the two p o provide a con onfiguration, w d in series. O cking two H- arlier [10], off kage to the ele lf-turbines co r the aerodyna on between th ed. By placing of the wind p nverted into m e proposed VA d turbine desig as the advanta single stage si ric features of e and trailing he counter rot ment studied in hree blades an del cal … of tation years iating [4-6] prime ncept where Other -types fering ectric ounter amics he tip g the power motion AWT gn we age of imilar f this edge tating n this nd the T bee sub tur con bet and as res has ele Th aro nu equ sim vor dep cha spe spe coe the CF rel sel ord wa ana Da of the typ reg sus opt Engineerin www.etasr TABLE I. G Parameter Blade chor Turbine diam Airfoil Turbine hei Blade numb Velocity For the num en defined div bdomain, roto rbine and stat ncept. Three tween the subd d vice-versa. well as a deta solve the prob s been set to ements was set Fig For this num he coarse me ound 850k an merical model uations. The mulations is ve rtex interactio pendent on th aracteristics th eed ratio) CF eed of the turb efficient is no e constant rota FD simulation ly on the degre lecting mome der to determi ay, the self-st alyzed. This arrieus configu attack [15]. M e post stall re pically low [ gime, traditio stained operat timization top ng, Technology r.com GEOMETRY OF THE rs Inner Tu rd 45 meter 22 ight ber y merical analys vided in four or for the inn tor for subdo fully conform domains, whic Figure 2 show ail of the inter bable No.s in t no more than t to 1.1. g. 2. Mesh of merical study th esh has aroun nd the fine ls work by sol state of the ery complex. A ons showing he flow time hat are being FD studies [1 bine is depende on-uniform thr ational speed s of the count ee of freedom entum inertia ine the operati tarting charac regime is urations-since Meaning that, egion will be [16]. Since th nal H-types tion, hence m pic. In the DO y & Applied Sci Malael & D E COUNTER-ROTA Value urbine Outer 5 7 5 4 NACA 0021 500 3 10÷16 sis, a computa subdomains: ner turbine, r omain for aro mal interfaces ch connect the ws the structur rface between the boundary n 1, and the g f the CR-VAWT c hree structured nd 500k elem one around lving the Navi art in the w Authors in [12 that the flow history not ju considered in 3]. The insta ent on the torq roughout a fu analysis less ter rotating wi m (DOF) metho for each win ional point of cteristics of t very import the turbine w from airfoil p used where th his is an inh have difficul making this reg OF method, th ience Research Dragan: Numer ATING IN-SERIES V Unit Turbine 70 mm 450 mm - mm - m/s ational domai stator for the rotor for the und of the tu have been de e rotors to the s red mesh gen n the two stage layer, the y+ growth ratio o concept d meshes were ments, the me 1200K. The ier-Stokes syst wind turbine 2] studied the b w pattern is h ust the current n constant TSR antaneous rota que, but the mo ull rotation, m accurate [14] ind turbine co od, which wor nd turbine ro f the turbine. I the turbine ca tant-particular works at high a polar diagram he lift-drag ra herently ineff lties attaining gime an inter e parameters h V rical and Exper VAWT ts m m m in has inner outer urbine efined stators erated es. To value of the e used. edium CFD tem of CFD blade- highly t flow R (tip ational oment making ]. The ncept, rks by tor in In this an be rly to angles m, only atio is ficient g self- resting of the ana allo turb opt pre T M the SST reco turb k-ε area bou insu pur of intr tran (ω) turb carb imp cur and thu 120 foll tech mil mo clos Dev the Vol. 8, No. 4, 20 rimental Efficie alysis were th owing the sof bine at every t timal point. sented in Tabl TABLE II. GE Parameters Mass Momentum inert Due to the tim unsteady Re T, was used i ord of accur bines [17-18]. model and the a, which is co undary layer [ ured by the u rpose of activa interest. Thes roducing a diff nsport equation 1 Γ . k k k k k u v t x ρ x G Y                The transport ): 1 Γω ω ω ω ω u t x ω ρ x x G Y D                In order to va bine a scale w bon fiber for portant, a mold ring cycle was d previous exp s consisting in 0°C (~2°C/min lowed by co hnology two lled Al 4.0 Cu ld is presente sed loop win velopment Ins wind turbin 018, 3282-3286 ency Evaluation he wind veloc ftware to com time step, until For CR-VAW le II for each tu EOMETRY OF THE Inner Tu 192.2 tia 0.0037 me-dependent eynolds averag in this study, rately predict This model i e k-ω model, u onsidered fully 19]. The trans use of connec ating each of th se connecting ffusive term in n for the kinet Γk k k v w y z k x y               t equation of Γω ω ω ω v w y z ω ω x y y D              III. EXPER alidate the cou wind turbine all six blades ding method w s established perimental wor n a heating sta n), temperatur ooling at roo molds with u 1.2 Pb 1.1 M ed in Figure 3 nd tunnel of stitute for Gas ne concept to 6 n of a Counter- city and the m mpute the cha l the simulatio WT the mom urbine stage. E COUNTER-ROTAT Value urbine Outer 21 26 767 0.0 characteristics ged Navier-St because of i ting the perf is a combinati using k-ε mod y turbulent, a sition between cting function he models, dep functions hav the transport tic turbulent en Γk k k y z z        f the specific Γω ω ω y z z        RIMENTAL SETU unter rotating model was m . Since dimen was used inside based on the rk made with age from room re was held co om temperatu NACA 0021 Mg 0.8 Mn alu 3. Tests were f the Roman Turbine - COM o various lev 3283 -Rotating Vertic moment of in aracteristics o on converged t mentum inert TING IN-SERIES VA Un r Turbine 64.41 g 013386 kg s of the simula tokes based m its successful formance of ion of the stan del for the free and the k-ω fo n the two mod ns, which hav pending on the ve another ro equation of ω nergy (k): k z    (1) rate of dissip ω z    (2) UP vertical axis manufactured u nsional precisi e an autoclave material data the same mat m temperature onstant for 90 ure. To use were from uminum alloy e conducted in ian Research MOTI by exp els of the st cal … nertia, of the to the tia is AWT nits g gm2 ation, model track wind ndard e flow or the dels is ve the e area ole of ω. The pation wind using ion is e. The asheet terial, up to min, this CNC y. The n the h and osing tream vel kin con me Fig qu con Fig pre and sev ben Fig ma tur Engineerin www.etasr locity. The ex nd permanent nfiguration. Fi This PMG [2 eaning it has gure 4 shows antify the pow nnected to an gure 5 the mai esented, showi d the bearings In order to d veral experim nch. The conf gure 6. With th ap was defined To find the rbine rotors t ng, Technology r.com xperimental m t magnet ge ig. 3. Blades 20] was design two rotors w s a cross sec wer generated n electric circu in parts of this ing the armatu . Fig. 4. PM Fig. 5. P determine the mental tests w figuration of t he results from d. Figure 7 sho Fig. 6. PM e mechanical the relation b y & Applied Sci Malael & D odel is connec enerator with molds for inner tu ned for this wi hich spin in o ction of this by the turbine uit which acte s permanent m ures of the coil MG cross section PMG main parts internal effic were carried this testing be m PMG tests a ows this PMG MG testing bench power gener between the m ience Research Dragan: Numer cted to a one h a purpose urbine ind turbine co opposite direc PMG. In ord e, the generato ed as a resist magnet generat ls, the shaft, ca n ciency of this out on the t ench in presen an electric effic map. h rated by the measured elec h V rical and Exper of the built ncept, ctions. der to or was tor. In tor are asings PMG testing nted in ciency wind ctrical qua ord mea sch pre whe carr stat the win inte sys min con ball man Vol. 8, No. 4, 20 rimental Efficie antities, torqu der to determ asured with hematic diagr sented. F Fig. 8. Typically, wi en a certain w ried out at 4 w tic Pitot tube i velocity meas nd turbine up ernal settings. tem was a nimized in ord ncept. In this e l bearings. nufactured. In 018, 3282-3286 ency Evaluation e and rotatio mine the PM a digital torq ram of the Fig. 7. Efficien Schematic diag ind turbines c wind speed is r wind speeds, is fixed on the surements to d pstream and In the experim challenge be der to have littl embodiment, th To support n Figure 9 the e Fig. 9. Exp 6 n of a Counter- nal speed we MG efficiency que detector. experimental ncy map of the PM gram of experime can generate e reached, there with the max e 3-D traversin determine the double check mental model ecause the fr le impact on th he bearing sys the turbine experimental m perimental model 3284 -Rotating Vertic ere determine , the torque In Figure 8 l system [21 MG ental system electric power efore the tests imum of 20m ng system alon flow velocity k the wind tu design the be friction should he efficiency o stem consists o e a casing model is show l cal … ed. In was 8 the 1] is only were m/s. A ne for from unnel earing d be of the of six was wn. rot sim est po eas im Fig wh Engineerin www.etasr In order to tating vertical mulations were timate the evo sitions of the sier to observ mpact on the gures 10-13 t here the stream Fig. 10. V Fig. 11. Vo Fig. 12. Vo ng, Technology r.com IV. determine the l axis wind H e carried out, u olution of the t turbine blades ve the recircul overall perfor the vorticity m m velocity was orticity magnitud orticity magnitude orticity magnitude y & Applied Sci Malael & D RESULTS e nominal po H type Darrie using a URAN turbine, the vo s, has been plo lation areas th rmance of th magnitude var s 12m/s is pres de for theta 0 degr e for theta 30 degr e for theta 60 degr ience Research Dragan: Numer oint of this co eus turbine, s NS DOF metho orticity for dif otted. This ma hat have a ne he wind turbin riation for the sented. rees, for V=12m/s rees, for V=12m/ rees, for V=12m/ h V rical and Exper ounter everal od. To fferent akes it gative ne. In e case s /s /s of max con hig whe valu the exp Vol. 8, No. 4, 20 rimental Efficie Figures 14-15 outer and inn ximum value ntribution of th hest. The num ere the exper ues between 1 power coef perimental test Fig. 13. Vor Fig. 14. Mo Fig. 15. Mo Fig. 16. 018, 3282-3286 ency Evaluation 5 show the va ner turbine fo e is obtain he inner turbin merical data rimental mode 10m/s and 16m fficient for n ts is presented. rticity magnitude omentum coeffici omentum coeffici Tip speed ratio v 6 n of a Counter- ariation of the r different ve ed at V=12 ne in the conce was validated el was investi m/s. In the ch numerical sim . for theta 90 degr ient variation for t ient variation for t variation within str 3285 -Rotating Vertic e torque coeffi elocity values. 2m/s, where ept efficiency d in wind tu igated for vel hart from Figu mulation and ees, for V=12m/s the outer turbine the inner turbine ream velocity cal … ficient . The the is the urbine locity ure 17 d for s tw adv tes Ae VA fou can thi rot com tes ma exp per opt the con sta con exp pro aer De fun [1] [2] [3] [4] [5] [6] Engineerin www.etasr Fig. 17. P A new conc o stage, in se vantage regar sted both in erodynamic m AWT model und that the na nnot be easily is method, only tors, leaving mputed using sts of the conce anufactured perimental re rformance but timization the e inner and nsideration. A age solidity an ncept towards perimental m ovide less fric rodynamics of This work w evelopment In nds on wind tu P. A. Costa R E.Vieira da Sil effects of blad turbine in urba T. Wenlong, M Savonius win Renewable En R. Gupta, A. bucket Savoni bladed Darrieu 1981, 2008 J. Gavalda, J. M Darrieus—Sav No. 4, pp. 457- A. Ghosh, A. B of flow physic rotor”, Journal K. Sahim, D. Radius Rotor ng, Technology r.com Power coefficient V. C ept for VAW eries H-type D rding the pow CFD and w modeling of c was done us atural rotating y predicted thr y an inertial m the acceler the unsteady ept feasibility specifically esults from w t an aerodynam e blade airfoil outer turbine A better under nd specific sp its full potent model where a ction, allowing f the concept. ACKNOW was carried ou nstitute for G urbine field. REFE Rocha, J. W.Carn lva, D. Albiero, C de pitch angle o an environments”, M. Zhaoyong, B. nd rotor with d ergy, Vol. 117, pp Biswas, K. K. S ius rotor with a us rotor”, Renew Massons, F. Diaz vonius wind mach -461, 1990 Biswas, K. K. Sh cs of a combine of the Energy In Santoso, D. Pusp r in Combined y & Applied Sci Malael & D t variation within CONCLUSION WT was propos Darrieus turbin wer extractio wind tunnel w counter rotati ing DOF me rate of the tw rough classica momentum is s ration to ae SST RANS m were also con for this wind tunnel mic optimizatio l selection and e diameter w rstanding of th peed will lead tial. Future wo a magnetic b g more focus t WLEDGMENT ut within Nati Gas Turbines ERENCES neiro de Araujo, C. F. de Andrade, on the performan , Energy, Vol. 148 Zhanga, Y. Li, “ different convex p. 287-299, 2018 Sharma, “Compar combined three- wable Energy, Vo z, “Experimental hine”, Solar & W harma, R. Gupta, “ ed three bladed D stitute, Vol. 88, N pitasari, “Investig d Darrieus-Savo ience Research Dragan: Numer stream velocity sed, consisting nes. The envis on capabilities with good re ing straight b ethods since i wo stages differ l calculations. specified for th erodynamic e model. Experim nducted with a application. showed accep on is needed. I d the ratio be will be taken he relation be d to furtherin ork will focus o earing system to be placed o ional Researc COMOTI res R. J. Pontes Lim F. O. M. Carneir nce of small-scal 8, pp. 169-178, 20 “Shape optimizati x and concave rative study of a -bucket Savonius ol. 33, No. 9, pp study on a self-a Wind Technology, “Computational a Darrieus Savoniu No. 4, pp. 425-437 gations on the Ef onius Wind Tu h V rical and Exper g of a sioned s was esults. bladed it was rs and . With he two effects mental PMG The ptable In this etween n into etween ng this on the m will on the ch and search ma, M. o, “The le wind 018 ion of a sides”, a three- s–three- . 1974- adapting Vol. 7, analysis us wind 7, 2015 ffect of urbine”, [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] Vol. 8, No. 4, 20 rimental Efficie International Jo 3568542, 2018 Y. Kyozuka, “A for the Tidal C Technology, Vo X. Liang, S. Fu study of the e performance o Renewable Ene I. Paraschivoiu, pitch control”, Article ID 5053 ] D. Horia, A. “TORNADO c NCAS Bulletin ] M. J. Werle, W revisited”, Jour 1150, 2008 ] S. Ivanell, J. N wind turbine wa ] D. Hartwanger CFD”, NAFEM ] M. R. Castelli Proposal for a Energy, Vol. 36 ] F. Balduzzi, A No.s in the CF Energy, Vol. 85 ] W. A. Timmer University win Engineering, Vo ] I. Malael, V. efficiency evalu and Materials, V ] H. Dumitrescu, for vertical axi Turbines, pp. 69 ] G. Naccache, M axis wind turb Industrial Aerod ] M. Dranca, M generator for c International U Greece, August ] R. Howell, N. Q study of a smal No. 2, pp. 412-4 018, 3282-3286 ency Evaluation ournal of Rotati An Experimental Current Power Ge ol. 3, No. 3, pp. 4 u, B. Ou, C. Wu, ffects of the rad of a Darrieus- ergy, Vol. 113, pp , O. Trifu, F. Sae International Jou 343, 2009 Dumitrache, F concept and reali , Vol. 5, No. 3, pp W. M. Presz, “Du rnal of Propulsion N. Sorensen, D. H akes”, in: Wind E r, A. Horvat, “3 MS Conference, Ch , A. Englaro, E a new performan 6, No.8, pp. 4919- A. Bianchini, R. M FD simulation o 5, pp. 419-435, 20 r, R. P. J. O. M nd turbine dedica ol. 125, No. 4, pp Dragan, G. Viz uation by using t Vol. 772, pp. 90-9 , V. Cardos, I. M s wind turbines”, 9-81, Springer, 20 M. Paraschivoiu, bine using CFD dynamics, Vol. 17 M. Chirca, V. Za counter-rotating v Universities Powe t 28-31, 2017 Qin, J. Edwards, N ll vertical axis win 422, 2010 6 n of a Counter- ing Machinery, V Study on the Dar eneration”, Journa 39-449, 2008 C. Y. H. Chao, K dius ratio and att -Savonius comb p. 329-334, 2017 ed. “H-Darrieus w urnal of Rotating F. Frunzulica, isation of a roto p. 69–75, 2013 ucted wind/water n and Power, Vo Henningson, “Num Energy, Springer, D modelling of heltenham, UK, J . Benini, “The D nce prediction m -4934, 2011 Maleci, G. Ferrar of Darrieus wind 016 M. Van Rooij, “ ated airfoils”, Jo p. 488-496, 2003 zitiu, “The verti the CFD method 95, 2015 Malael, “The phy , in: CFD for Wi 015 “Parametric stu D”, Journal of W 72, pp. 244-255, 2 aharia, S. Breban vertical axis micr er Engineering C N. Durrani, “Win nd turbine”, Rene 3286 -Rotating Vertic Vol. 2018, Artic rrieus-Savonius T al of Fluid Scien K. Pi, “A computa tachment angle o bined wind tur wind turbine with Machinery, Vol. A. Pal, V. Tu or for small VA turbines and prop ol. 24, No. 5, pp. merical computati pp. 259-263, 200 f a wind turbine June 10-11, 2008 Darrieus wind tu model based on ra, L. Ferrari, “C d turbines”, Rene Summary of the ournal of Solar E ical axis wind t ds”, Applied Mec sics of starting p ind and Tidal Of dy of the dual v Wind Engineerin 2018 n, “Permanent m ro-wind turbine” Conference, Hera nd tunnel and num ewable energy, V cal … cle ID Turbine ce and ational on the rbine”, h blade 2009, urbatu, AWTs”, pellers 1146- ions of 07 using urbine: CFD”, Critical ewable e Delft Energy turbine chanics process ffshore vertical ng and magnet , 52nd aklion, merical Vol. 35,