Jtam.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 47, 1, pp. 3-16, Warsaw 2009 EXPERIMENTAL ANALYSIS OF VELOCITY FIELD STRUCTURE IN ISOTHERMAL COUNTERCURRENT JETS Barbara Wojciechowska Piotr Domagała Stanisław Drobniak Czestochowa University of Technology, Institute of Thermal Machinery, Poland e-mail: bwoj@imc.pcz.czest.pl The paper presents results of experimental analysis of the flowfield in isothermal countercurrent round jets. The velocity measurements were carried out by means of a hot-wire anemometry. The instantaneous si- gnals collected during the experiment were digitally processed and used to determine the statistics of velocity fields and distributions of tur- bulence scales. The results revealed that the fluid aspiration at the jet periphery significantly influences themixing and entrainment in the free flow. Key words: countercurrent jets, turbulence intensity, turbulence scales Notations (·)1 – related to main jet (inner) (·)2 – related to aspirated reverse flow (outer) α – extension collar divergence half-angle Λ – linear Taylor macroscale λ – linear Taylor microscale ν – kinematic viscosity coefficient U – local mean velocity U0 – mean reference velocity at the exit of the inner nozzle τt – correlation time distance D1 – diameter of inner jet D2 – diameter of outer jet I – ratio of the inner to outer velocity 4 B. Wojciechowska et al. L – extension collar length R(τ) – longitudinal velocity autocorrelation function Re – Reynolds number based on D1 T – timemacroscale of turbulence Tu – turbulence intensity u′/U U – local instantaneous velocity u′ – turbulence component of U1 (RMS) x – axial coordinate (attached to the exit plane of internal nozzle) 1. Introduction The countercurrent jet is the subject of a number of publications which appe- ared mostly during the recent decade. The reason for this interest is twofold, i.e. the applicability of this type of flow in many practical technologies and the appearance of interesting flow phenomena like the absolute instability. Thefirstmotivation for the research on counter-current jets, i.e. the enhan- cement of mixing was initiated by the experiment of Strykowski and Niccum (1991), who revealed the great potential of that way of flow stimulation in terms of intensification of transport processes. The phenomenon had a local character, i.e. intensivemixing took place in a limited space of the stream, and additionally was compensated by flow ”laminarisation” (i.e. damping of tur- bulent fluctuations) in the remaining jet regions. However, the enhancement ofmixing in the entire flow area could be obtained by control of the exit shear layer performed by application of extension tubes, as it was shown by Stry- kowski andWilcoxon (1993). The research on counter-current jet control with extension tubes performed by Asendrych (2007) revealed that a substantial modification of large scale vortical structures was obtained in presence of a counterflowwhich, in turn, affected themixing intensity. Summing up, careful control of this flowmay allow for effective use of intensification of themixing, and possible applications of this type of flow concern chemical processes and gas burnerswhere the enhancement ofmixing is of primary importance.An in- teresting perspective for application of counter-current jetswas the substantial increase of diffusion flame blow-off limit shown by Lourenco et al. (1996) and confirmed byAsendrych and Frania (2004). Another special case for practical use of this flow configuration is the thrust vectoring in jet engines (Roy, 2001) where a countercurrent jet may be a new, more effective and more efficient way of the flow control. Experimental analysis of velocity field structure... 5 The secondmotivation for investigations of counter-current jets is the ap- pearance of absolute instability experimentally found in heated jets by Mon- kewitz et al. (1990) and in helium jets by Kyle and Sreenivasan (1993). The theoretical justification for absolute instability triggered by an external coun- terflowing jetwas shown theoretically by Jendoubi andStrykowski (1994), but so far no convincing experimental evidence has been found for the existence of this phenomenon in counter-current jets. The extensive investigations of isothermal and heated counter-current jets performed by Asendrych (2000, 2007), Asendrych andDrobniak (2002), Asendrych andFavre-Marinet (2004), Bogusławski et al. (2002) could only reveal the existence of side jets and the substantial increase of jet spreading rate, which might suggest the appearan- ce of absolute instability, which, however, could not be treated as a direct evidence (this problem still needs futher investigations). The motivation for the present study resulted from the research perfor- med at ITM CzUT within the framework of TIMECOP EU project devoted to modelling of liquid fuel atomization and combustion in jet engines. The counter-current jet was selected as the test flow for investigations of fuel dro- plets evaporation (http://timecop-ae.com/) due to possibility of changing the structure of turbulence at the given point by changing the velocity ratio be- tween the inner and outer jets. This unique feature of the counter-current jet is very convenient for experimental studies of droplet evaporation being performed by means of optical methods, which need careful alignment and, therefore, do not allow for easy traversing of measuring location. However, the analysis of experimental results published so far revealed the lack of data concerning the characteristic scales of turbulent motion, which are the basic parameters for studies of droplet evaporation (Birouk andGokalp, 2006). This finding determined the goal of the present research, which was devoted to the experimental determination of turbulence scales in counter-current jets. 2. Experimental facility and measurements The experiment was performed in an open-circuit wind tunnel equipped with a set of two concentric nozzles shown in Figure 1 and described inmore detail in Asendrych (2007), Bogusławski et al. (2002). The main component of the test rig is an inner nozzle, which ismade of brass. This nozzle generates an air stream with a very low turbulence level and ”flat” profile of mean velocity at the exit of the inner nozzle. The roughness of the inner surface of the nozzle was carefully polished in order to avoid any possible disturbances generated 6 B. Wojciechowska et al. at the nozzle surface. Additionally, to enable control of the flow, an extension collar was mounted at the outer jet exit. Its additional aim was to reduce the influence of the back flowon the velocity field in the channel between the inner and outer nozzle. Fig. 1. The geometry of countercurrent nozzles (Bogusławski et al., 2002) The key parameter for the outer nozzle was the shape of the inner surface, because a very high contraction ratio (∼ 128) had to be applied in order to maintain as low turbulence level as possible. The cubic profile of the inner nozzle enabled one to obtain an extremely low turbulence level at the inner jet exit, which was below the noise level of Constant Temperature Anemometer CTA. Additionally, the design of the outer nozzle allowed for replacement of orifices which determined the inlet width of the suction channel. The outer nozzle and the extension collar were made of aluminium. The geometry of these two nozzles is shown in Figure 1, the dimensions of nozzles applied in the experiment were as follows: • inner diameter D1 =15mm • outer diameter D2 =30mm • extension tube height L=D1 • divergence half-angle α=7◦. Themeasurements of the velocity fieldwere performedalong the jet radius in several control planes covering the area x/D1 =0−11D1 as well as along the inner jet axis and along the line being the extension of the edge of the inner nozzle. The measurements were performed with CTA which had to be precisely adjusted and aligned, which was possible in a given point only. The spreading rate ofmain streamcouldbechangedby theparameter I,whichwas defined as the ratio of the inner-to-outer velocity. The main flow parameters of the reported experiment were as follows: Experimental analysis of velocity field structure... 7 • Reynolds number Re=U1D1/ν ≈ 10000 and 20000 • aspiration intensity expressed as the ratio of mean velocities of the re- verse flow andmain stream I =U2/U1 =0÷0.4. The time scales of turbulence, i.e. Taylormicro andmacroscales have been determinedusing the standardproceduresdescribed inHinze (1975). The time microscale of turbulence τt can be expressed in terms of an autocorrelation function of longitudinal velocity fluctuations R(τ)with the following equation 1 τ2 t =− 1 2 [∂2R(τ) ∂τ2 ] τ=0 (2.1) where R(τ) is a symmetrical function of τ with the maximum value equal to unity for τ = 0, and R(τ) decreases with increasing τ. The value of time turbulencemicroscale τt may be computed from the intersection of osculation parabola with τ axis as shown in Fig.2. Fig. 2. Osculation parabola of the lateral correlation coefficient R(τ) and turbulent length scales τt and T The timemacroscale of turbulence T is given by the following equation T = ∞ ∫ 0 R(τ) dτ (2.2) The time micro- and macroscales of turbulence determined from the auto- correlation functions were transformed to linear scales with the use of Taylor hypothesis (Elsner, 1987) λ=U1τt Λ=U1T (2.3) The measurements were carried out with the use of DANTEC single- channel hot-wire anemometer. The setup is shown schematically in Fig.3. 8 B. Wojciechowska et al. The constant temperature anemometer was used to recover the instantaneous velocity of flow. Fig. 3. Scheme of the measuring equipment Velocity measurements were performed with DANTEC gold-plated wire probe 55P05. Instantaneous voltage signals of CTA were sent to the signal conditioners and data analysis system. In order to enable continous monito- ring of the experiment, all signals were visualized by oscilloscopes, and their average values were controlled by mean voltage meters. The signal proces- sing, i.e. recovering the instantaneous velocity value U as well as evaluation of the statistical moments and spectral density functions, was performed fully digitally by an HP signal analyzer. 3. Characteristics of the flow field in isothermal countercurrent jets Fig. 4. Comparison of present experimental data for mean velocity with literature (Drobniak andKlajny, 2002) (measurements along the jet axis) Experimental analysis of velocity field structure... 9 Figures 4 and 5 present the comparison of present experimental data for themean velocity and turbulence intensity with the literature data (Drobniak and Klajny, 2002) for a similar Reynolds number along the symmetry axis of the jet (Drobniak andKlajny, 2002). The results for themean velocity (Fig.4) are in a very good agreement with the literature data. Also the data for the turbulence intensity (Fig.5) are very similar except for the first region where some differences can be noticed. These differences are most probably due to much lower turbulence intensity at the jet exit obtained during the present experiment, whichmay be attributed to the corrected shape of the nozzle and better selection of gauzes in the plenum chamber. Fig. 5. Comparison of present experimental data for turbulence intensity with literature (Drobniak and Klajny, 2002) (measurements along the jet axis) Fig. 6. Comparison of mean velocity radial distribution for different suction ratios for Re=10000 InFig.6, onemay observe the influence of suction ratio I onmean velocity profiles for Re=10000. Velocity plots on the left hand side are for the suction ratio I = 0, and the plots on the right-hand side correspond to the case 10 B. Wojciechowska et al. of maximum suction I = 0.4. The case with no suction presents classical spreading of the jet with gradual decay of potential kernel. For the case with maximum suction the decay of potential kernel of the main jet is the most distinct influence. Onemay also notice the reversed flow in the first two cross- -sections. The spreading rate for this case is smaller in the first cross-sections than in the reference case due to large suction intensity. Fig. 7. Decay of mean velocity for Re=20000 for various ratios of suction intensity I (measurements along the jet axis) The change of flow parameters at the symmetry axis of the jet for Re = 20000 is shown in Figs.7 and 8, both the decay of the potential core (Fig.7) and the increase of turbulence intensity (Fig.8) caused by suction are visible. These pictures present measurements obtained in consecutive cross- sections from the jet exit up to normalized distance x/D1 =11D1. Both the mean velocity andRMSof velocity fluctuations have beennormalizedwith the inner jet velocity U0 at the exit. The influence of suction is not monotonous, the smallest value of suction parameter I = 0.1 causes faster decay of the potential core and bigger increase of turbulence intensity than for I = 0.2. For larger values of I, the faster decay of mean velocity and increase of Tu is restored, for all values of I themaximumof Tumoves upstream.Onemay see that due to the change of variable I, the flow parameters change considerably at every point located at the jet axis. The decay of mean velocity at the jet axis for Re = 20000 is shown in Fig.7 and for Re = 10000 in Fig.9. The distance x and velocity U are again normalized by the jet diameter D1 and the inner jet velocity at the exit U0. The same tendency is visible for both Reynolds numbers applied in present investigations for the mean velocity at the axis, but for lower Re the decay of the potential core is more conspicu- ous, andmoremonotonous decay of velocity caused by the increasing suction intensity is visible. Experimental analysis of velocity field structure... 11 Fig. 8. Turbulence intensity evolution for Re=20000 for various ratios of suction intensity I (measurements along the jet axis) Fig. 9. Decay of mean velocity for Re=10000 for various ratios of suction intensity I (measurements along the jet axis) For the same two values of Reynolds numbers the evolution of turbulence intensity is shown in Figs.8 and 10. Figure 8 presents data corresponding to Re = 20000, while Fig.10 the data for Re = 10000. The comparison of Tu evolution along the jet axis reveals that for a lower Re number higher values of turbulence intensity and bigger increments of Tu for the same value of I were generally achieved. An other interesting observation is that the plateau of turbulence intensity, which was only slightly conspicuous for I = 0 and Re = 20000 at the distance x/D1 = 3− 4D1 (Fig.8), becomes much better visible for I =0.1. This phenomenon was already reported by Drobniak and Klajny (2002) as related to vortex pairing, and for present investigations it may be of some importance as it changes the spectral content of turbulence in this region. The increase of outer-to-inner ratio stream parameter leads to even more perspicuous presence of this phenomenon. The maximum value I = 0.4 shifts the maximum of Tu much more upstream and allows one to expect even bigger modification of the spectral content of turbulence, which may be of some use in further investigations. 12 B. Wojciechowska et al. Fig. 10. Turbulence intensity evolution for Re=20000 for various ratios of suction intensity I (measurements along the jet axis) Fig. 11. Evolution of mean velocity for Re=10000 and for various ratios of suction intensity I (measurements along the inner jet edge) A summarised view on the influence of suction upon the velocity field along the jet edge is shown in Figs.11 and 12 for Re = 10000. In Fig.11, the distribution of mean velocity is presented, and in Fig.12 the distribution of turbulence intensity is shown for all values of suction intensity which were applied in present investigations. The influence of suction upon the mean ve- locity field (Fig.11) is more visible in the far region of the flow, while in the region close to the exit, the influence of suction is less visible. The opposite effect is observed in the case of Tu intensity (Fig.12), where suction exerts the biggest influence in the initial region of the flow, while in cross-sections from x/D1 = 5D1 all lines corresponding to various levels of suction almost collapse. Onemay notice that along this line one may achieve a very high value of turbulence intensity which reaches as high as ∼ 25%. The next flow parameters which were of interest for present investigations were the scales of turbulence.Figures 13 and14present thedownstreamevolu- Experimental analysis of velocity field structure... 13 Fig. 12. Turbulence intensity evolution for Re=10000 and for various ratios of suction intensity I (measurements along the inner jet edge) Fig. 13. Downstream evolution of linear Taylor microscale for Re=20000 (measurements along the jet axis) Fig. 14. Downstream evolution of linear Taylor macroscale for Re=20000 (measurements along the jet axis) tion of linear Taylor micro andmacroscales for a sample value of Re=20000 and for all values of the suction parameter. These results were obtained along the jet axis. One may notice that in the initial region, there is almost no dif- ference between the micro- and macroscales, but it is not a surprise bearing in mind that the spectrum of turbulence did not develop here. As one moves downstream, both the micro- andmacroscales increase, but the rate of incre- ase for the macroscale is much faster, which confirms the well known rules of 14 B. Wojciechowska et al. increasing the frequency span with the development of turbulence. One may also notice that suction increases both the micro- and macroscale of turbu- lence, but the rate of increase of the macroscale is much faster, which reflects the widening of the spectrum of turbulence in the downstream direction. In Fig.14, one should notice a substantial difference between the two measure- ments and the trend line. This discrepancy has been explained as resulting from the systematic error in two autocorrelation curves, and these pointswere not taken into account in calculation of the trend line. 4. Summary The paper presents experimental study describing flow field characteristics in isothermal countercurrent round jets. The experiment confirmed that the reverse outer flow may substantially change the flow pattern of the inner jet proving that it can be utilised for active flow control. As the key parameter, the outer-to-inner flow ratio (the ratio of bulk velocities of the reverse stream and the main jet) was foundwith its critical value around 0.2. This tendency was visible for both Reynolds numbers applied in the investigations, but it proved to be more tangible for lower numbers Re. Another novel element of the paper is the presentation of the distribution of micro- andmacroscales of turbulence, which to the best knowledge of authors has not been published so far. Acknowledgements The research was performed under the TIMECOP EU project (contract No. STREP 030828) and SPBTIMECOP funded byMNiSzW. This work received funding from the European Community through the project TIMECOP-AE(Project#AST5-CT-2006-030828).It reflectsonly theauthor’sviews, and the Community is not liable for any use that may be made of the information contained therein. References 1. Asendrych D., 2000, Intensification of transport processes in free round co- untercurrent flows, Cieplne Maszyny Przepływowe, Turbomachinery, 117, II, 27-32 Experimental analysis of velocity field structure... 15 2. Asendrych D., 2007, Active flow control by countercurrent jets, J. of Theor. and Applied Mech., 45, 3, 463-478 3. Asendrych D., Drobniak S., 2002, Experimental analysis of the flowfield in non-isothermal countercurrent jets, Advances in Turbulence, IX, Proc. 9th Eur. Turbulence Conf., Castro I.P. et al. (Edit.), 839 4. Asendrych D., Favre-Marinet M., 2004, Diffusion of jets with annular counterflow and small diameter ratio,AIAA Journal, 42, 11, 2385-2387 5. AsendrychD.,FraniaT., 2004,Zastosowanieprzepływuzwrotnegodoopty- malizacji spalania w strudze swobodnej, XVI Krajowa Konferencja Mechaniki Płynów, Waplewo 6. Birouk M., Gokalp I., 2006, Current status of droplet evaporation in tur- bulent flows,Progress in Energy and Combustion Science, 32, 408-423 7. Bogusławski A., Asendrych D., Drobniak S., Kubacki S., 2002, CFD simulations of concentric jets with external counterflow,QNET-CFD Network Newsletter, 1, 3, 27-30 8. DrobniakS.,KlajnyR., 2002,CoherentStructures inAxisymetricFreeJets, J. of Turbulence, 3, 001 9. Elsner J.W., 1987, Turbulencja przepływów, PWN Warszawa, ISBN 83-01-0613-4 10. Hinze J.O., 1975,Turbulence, McGraw-Hill Book Company, NewYork, ISBN 0-07-029037-7 11. Jendoubi S., Strykowski P.J., 1994, Absolute and convective instability of axisymmetric jets with external flow,Phys. Fluids, 6, 9, 3000-3009 12. Kyle D.M., Sreenivasan K.R., 1993, The instability and breakdown of ro- und variable density jet, J. Fluid Mech., 249, 619-664 13. Lourenco L., ShenH., Krothapalli A., Strykowski P.J., 1996,Whole- field measurements on an excited premixed flame using on-line PIV, 8th Int. Symp. on Applications of Laser Techniques to Fluid Mech., Lisboa 14. Monkewitz P.A., Bechert D.W., Barsikow B., Lehman B., 1990, Self- excited oscillations and mixing in a heated round jet, J. Fluid Mech., 213, 611-639 15. RoyG.D., 2001,Deflagrative anddetonative combustionflows: reserchaccom- plishments and chellenges in the newdecade,Proc. 5th ISAIF,Gdansk,Poland, 45-65 16. StrykowskiP.J.,NiccumD.L., 1991,The stability of countercurrentmixing layers in circular jets, J. Fluid Mech., 227, 309-343 17. Strykowski P.J.,WilcoxonL.G., 1993,Mixing enhancement due to global oscillations in jets with annular counterflow,AIAA Journal, 31, 3, 564-570 18. http://timecop-ae.com/ 16 B. Wojciechowska et al. Eksperymentalna analiza struktury pola prędkości w izotermicznych strugach przeciwbieżnych Streszczenie Artykuł przedstawia wyniki eksperymentalnej analizy pola prędkości izotermicz- nych strug przeciwbieżnych. Eksperyment przeprowadzony został z wykorzystaniem pionowego tunelu aerodynamicznego wyposażonego w układ dwóch dysz do genera- cji osiowosymetrycznych, koncentrycznych strug przeciwbieżnych. Pomiar prędkości został wykonany przy użyciu termoanemometrii. Uzyskane w trakcie badań wyniki pokazują, że zewnętrzna struga zwrotnama istotnywpływna rozwójwewnątrz strugi osiowosymetrycznej. Najistotniejszy parametr niniejszych badań, tzn. stosunek pręd- kościwewnętrznej i zewnętrznej strugi,wykazałwartość krytyczną I =0.2.Prawidło- wość tę zaobserwowano dla obydwu badanych liczb Reynoldsa, chociaż dla niższych liczbReynoldsa tendencja ta byłabardziejwidoczna.Nowymelementempoznawczym było określenie rozkładu charakterystycznych skal turbulencji, które według wiedzy autorów nie były prezentowane w literaturze dla badanego przepływu. Manuscript received April 16, 2008; accepted for print July 16, 2008