Jtam.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 45, 3, pp. 569-586, Warsaw 2007 EXPERIMENTAL INVESTIGATIONS OF JET FLOWS Elżbieta Fornalik Janusz S. Szmyd AGH, University of Science and Technology, Faculty of Non-Ferrous Metals, Poland e-mail: elaf@agh.edu.pl The paper describes two kinds of jet flows: the confined and impinging jet flow. A triaxial thermoanemometer probe with a temperature sensor wasused to study the confined jet flow.The time-averageanalysis ofdata resulted in a turbulence characteristic. It allowed the estimation of the budget of turbulent heat flux. The impinging jet flow was investigated by the Particle Image Velocimetry (PIV) method together with heat transfer analysis done with a liquid crystal film. Key words: confined jet, impinging jet, turbulent heat flux, PIV 1. Introduction In the industry, there aremany kinds of jet flows which aremainly turbulent. Investigations of such flows in the industrial scale are very difficult. Therefore, experiments are done in laboratories with applications of various methods. The confined jet flows can be found in ejectors, combustors, jet pumps and other devices, in which mixing layers of different mean velocities appear (Razinsky and Brighton, 1971). That kind of jet flows was studied by many researchers. It would be impossible to list all of them, however some of them have to be mentioned. Craya and Curtet (1955) described theory of the mi- xing zone of streams. Curtet (1958) introduced a similarity criterion to the appearance of recirculation. Becker et al. (1962) named this parameter as the Craya-Curtet number and investigated its influence on the flow structure. An experimental analysis of the confined jet was done by Barchilion and Curtet (1964), Brighton and Jones (1964). Champagne and Wygnanski (1971) inve- stigated the turbulence characteristic for various conditions. Their work was continued by Razinsky and Brighton (1971). Non-isothermal flows were stu- died byFrota andMoffat (1982), Heist andCastro (1998), Keffer et al. (1977), Chevery and Tutu (1978), Suzuki et al. (1987), Kang and Suzuki (1982) and 570 E. Fornalik, J.S. Szmyd Kang et al. (1979). The confined jestwas investigated under various conditions with various experimental methods. The impinging jet has a wide range of industrial applications in processes involving heating, cooling anddrying.Advantages of impinging jets come from the effective removal of heat flux and easy adjustment to locations where it is necessary. The turbine cooling system consists of the jet array. The streams coming fromtheupstream-positioned jetsmerge, producing the crossflow.Ma- ny researchers, includingMetzger andKorstad (1972), Bouchez andGoldstein (1975), Sparrow et al. (1975), Saad et al. (1980), Goldstein and Franchett (1988) andHuang et al. (1996) studied various aspects of the jet impingement heat transfer under the influence of the crossflow. A triaxial thermoanemometer probe with a temperature wire was used to study the confined jet, whereas in the case of the impinging jet the Particle Image Velocimetry (PIV) accompanied by liquid crystal temperature measu- rements. 2. Confined jet The name ”confined jet” describes two streams of a fluid: primary circular stream entraining andmixingwith the secondary annular stream in a confined chamber. If the ratio of velocities is large enough, the recirculation zone may appear. In the presented study, special attention is paid to the budget of turbulent heat transfer. 2.1. Experimental apparatus A schematic diagram of the experimental apparatus is shown in Fig.1. The set-up has been located at the laboratory of the Institute of Thermal Machinery at the Technical University of Częstochowa. The air is drawn from atmosphere through air filter (1) by blower (2). The pure and regulated air stream is divided in two in air divider (4); i.e. into the primary circular stream and the secondary annular stream.The air of the primary stream is heated by electrical heater (6). The temperature difference between the circular and an- nular streams is 20K. The regulated and heated air stream is then discharged from aluminium inner nozzle (11) with diameter d = 0.040m. The nozzle is designed to obtain coaxial streams (the circular primary stream and the an- nular secondary stream) at the inlet of the test section. Finemesh screens (8) and honeycomb (9) are used to condition the secondary stream. The size of the measuring section (plexi-glass tube) is: diameter D = 0.192m, length Experimental investigations of jet flows 571 l=2m.Themean velocity of the primary air stream equals U1 =20m/s and the mean velocity of secondary stream is U2 =1m/s. Fig. 1. A schematic diagram of the experimental set-up in the case of confined jet; 1 – air filter, 2 – blower, 3 – thyristor system, 4 – air divider, 5 – orifice, 6 – electric heater, 7 – autotransformers, 8 – fine mesh screen, 9 – honeycomb, 10 – outer nozzle, 11 – inner nozzle, 12 – test section 2.2. Procedure Geometry and initial conditions were determined by the similarity crite- rion: i.e. the Craya-Curtet number which expresses the influence of the input momentum flux distribution on the confined jet system. The Craya-Curtet number is defined by Ct= Um U∗ U∗ = √ (U 2 1−U 2 2)b 2+ 1 2 (U 2 2−U 2 m) (2.1) Um =(U1−U2)b 2+U2 The Craya-Curtet number was set up to be Ct = 0.5, for which the re- circulation zone was large enough for the investigations. The location of the recirculation zone was preliminary monitored by a hot-wire probe with two parallel tungsten filaments. The test sectionwas divided into 17 cross sections and in each cross-section 23 measurement points were situated. The measurements were done by the DANTEC hot-wire anemometry systemwith four channels, one of which was for temperature and the other three for velocity components. It allowed simul- taneous measurements of velocity and temperature. The sampling rate was 572 E. Fornalik, J.S. Szmyd 20 kHz per channel. The sampled raw data were stored for further analysis to obtain themean velocity components, Reynolds normal and shear stresses, temperature and its fluctuationand turbulentheat fluxcomponents.The tech- nique of estimation of all the abovementioned quantities is precisely described in Fornalik et al. (1998). 2.3. Results The non-dimensional value of the velocity component was set up as the ordinate axis inFig.2.There canbe recognised an internal, circular streamof a high velocity (0¬ r/d¬ 0.6) andan external, annular streamof a lowvelocity (0.6 ¬ r/d ¬ 2.1). In the first cross-section (x/d = 1) the primary-stream potential core can be observed, but it quickly disappears. Themost important information, which could be read from Fig.3, is related to formation of the recirculation zone. The negative values of velocity appear in the vicinity of the wall on the 4th cross-section (x/d = 4), and going downstream successively, the reverse flow is developing.Themeasured recirculation zone is a little larger than 0.68m x/d = 17 (beyond the investigated zone), therefore it is not possible to specify the reattachment point. Fig. 2. Distribution of the longitudinal velocity component Experimental investigations of jet flows 573 Fig. 3. Distribution of the angular velocity component The flow was assumed to be two-dimensional (symmetrical), but the pri- mary stream was slightly twisted in a few first measured profiles. The third angular component of the velocity vector Uϕ appeared and it affected all me- asured quantities. The distribution of its non-dimensional value is presented in Fig.3. In magnitude, it is much smaller than the other two components of the velocity vector. However, in the first four cross-sections, a change of the direction can be observed. In the first cross-section (x/d = 1), the circular stream is twisted, which is suggested by the change in the sign of the value in the range of r/d from 0 to about 0.6. The value of the velocity for the angular stream changes its sign in the range of r/d from 1.2 to about 2.2, which also suggests twist of the stream. Going downstream, themagnitude of the third velocity component becomes smaller, and from the fifth cross-section (x/d=5) only the circular stream is disturbed. The data related to the distribution of temperature, Reynolds normal and shear stresses and also to turbulent heat flux components were presented in Fornalik and Szmyd (2005). The problem of heat transfer budget will be di- scussed in detail in this paper. 574 E. Fornalik, J.S. Szmyd The governing equations are presented below (Eqs. (2.2)1,2,3) in the cylin- drical co-ordinate system (suitable for examined geometry). They are expres- sed in four groups representing (1) convection, (2) production due to mean velocity and mean temperature fields, (3) diffusion and (4) transfer due to pressure correlation, molecular transfer due to viscosity and thermal conduc- tance of the fluid. The last (fourth) term is taken as a closure term, because the pressure-temperature correlations cannot bemeasured and the molecular transport of turbulent heat flux is impossible to obtain. The closure term is identified with the pressure-temperature correlation (the molecular transport due to viscosity and thermal conductance can be neglected, because of low influence on the budget), Kasagi (1991) Ux ∂uxθ ∂x +Ur ∂uxθ ∂r ︸ ︷︷ ︸ (1) =−uxθ ∂Ux ∂x −urθ ∂Ur ∂r −uxux ∂T ∂x −uxur ∂T ∂r ︸ ︷︷ ︸ (2) + − ∂ ∂x (uxuxθ)− 1 r ∂ ∂r (ruxurθ) ︸ ︷︷ ︸ (3) + − 1 ρ θ ∂p ∂x +ν ( θ ∂2ux ∂x2 +θ ∂2ux ∂r2 + 1 r θ ∂ux ∂r ) +a ( ux ∂2θ ∂x2 +ux ∂2θ ∂r2 + 1 r ux ∂θ ∂r ) ︸ ︷︷ ︸ (4) Ux ∂urθ ∂x +Ur ∂urθ ∂r ︸ ︷︷ ︸ (1) =−uxθ ∂Ur ∂x −urθ ∂Ur ∂r −uxur ∂T ∂x −urur ∂T ∂r ︸ ︷︷ ︸ (2) + − ∂ ∂x (uxurθ)− 1 r ∂ ∂r (rururθ)+ 1 r (uϕuϕθ)+ 2 r Uϕuϕθ ︸ ︷︷ ︸ (3) + (2.2) − 1 ρ θ ∂p ∂r +ν ( θ ∂2ur ∂x2 +θ ∂2ur ∂r2 + 1 r θ ∂ur ∂r − 1 r2 urθ ) +a ( ur ∂2θ ∂x2 +ur ∂2θ ∂r2 + 1 r ur ∂θ ∂r ) ︸ ︷︷ ︸ (4) Ux ∂uϕθ ∂x +Ur ∂uϕθ ∂r ︸ ︷︷ ︸ (1) =−uxθ ∂Uϕ ∂x −urθ ∂Uϕ ∂r −uxuϕ ∂T ∂x −uruϕ ∂T ∂r ︸ ︷︷ ︸ (2) + − ∂ ∂x (uxuϕθ)− ∂ ∂r (uruϕθ)− 2 r (uruϕθ)− 1 r Uruϕθ− 1 r Uϕurθ ︸ ︷︷ ︸ (3) + +ν ( θ ∂2uϕ ∂x2 +θ ∂2uϕ ∂r2 + 1 r θ ∂uϕ ∂r − 1 r2 uϕθ ) +a ( uϕ ∂2θ ∂x2 +uϕ ∂2θ ∂r2 + 1 r uϕ ∂θ ∂r ) ︸ ︷︷ ︸ (4) Experimental investigations of jet flows 575 In Figs. 4-6, the budgets of particular components of turbulent heat flux are presented. In Fig.4, the distribution of the axial component is shown. A distance from the centre of the channel is set up as the abscissa, and as the ordinate – the loss or the gain for the budget. A positive value of the parti- cular term (loss) means that the larger quantity of turbulent heat flux flows up from the control volume than comes in. A negative value (gain) describes the opposite situation. Every term is marked as a line of different type (look- up the legend). Two terms are visibly dominant: the production term and the pressure-temperature term,which almost compensate each other. Theproduc- tion term is a gain for the budget in contrast to the pressure-temperature term (loss). They are dominant in the case of each turbulent heat flux component, but they take the highest values for the radial component presented in Fig.5. The most intensive heat transfer processes are observed in the shear region between streams (r/d≈ 0.5). In the first graph (x/d=2, r/d= 0-0.1), it is found that all terms are equal to zero, which is characteristic for a potential core. This is the potential core of primary stream. The potential core of se- condary stream disappears before the place at which the first cross-section is measured. In the second graph (x/d = 3), it is not observed – the potential core of primary stream vanishes due to mixing of the streams. Fig. 4. Distribution of the budget for the axial component of turbulent heat flux in various cross-sections 576 E. Fornalik, J.S. Szmyd Fig. 5. Distribution of the budget for the radial component of turbulent heat flux in various cross-sections It is worth to say that the distributions of diffusion and convection terms are very interesting. They are nearly complementary like the production and closure terms (but in a smaller scale). Near the centre of the channel (r/d= 0.1), the diffusion term makes a gain for the budget in contrast with the convection term. The opposite situation can be observed near the place of themost intensive streamsmixing (r/d=0.5), where the diffusion term takes positive values (loss) and the convection term takes negative values (gain).The diffusion transfers the heat from themixing region of the jets (positive values – loss) toward the centre of the channel and walls, where it takes negative values (gain). The heat is transferred by convection from the circular stream (centre of the channel, r/d=0.1), where it takes positive values (loss) to the secondary stream (r/d> 0.5), where its sign is negative (gain). Generally, it can be said that going downstream (x/d=5, 6, 7), the heat transfer becomes less intensive, which does not depend on the way of transfer. The distribution of the budget for the radial component of turbulent heat flux presented in Fig.5 is very similar to the budget for the axial component. The only difference is in the magnitude of particular terms. The difference in the distribution of the budget can be observed in Fig.6. It presents the angular component of turbulent heat flux. The magnitudes of all terms are much smaller than in the case of the axial and radial components. Experimental investigations of jet flows 577 Fig. 6. Distribution of the budget for the angular component of turbulent heat flux in various cross-sections The heat transfer from the circular stream toward the annular one is well visible. The region of intensive mixing between the streams (r/d = 0.5) can be easily identified, since it is the region where the heat transfer changes the direction. The production term is dominant, especially in the first three cross- sections (x/d = 2-4). Turbulent heat flux is transferred form the circular primary stream to the annular secondary stream. The diffusion and closure terms compensate the production term, what can be found for x/d = 3 and x/d=4. The convection term has small influence on the heat flux transfer. It has some effect only for the circular stream causing heat flux transfer from its outer region to its central part. Going downstream, the heat transfer becomes less intensive. 2.4. Comments Themost important area for the heat transfer in the case of the confined jet with the recirculation zone appeared in the mixing region of circular and annular streams (the shear region). Two terms were dominant in the budget of turbulent heat flux: production term and pressure-temperature term. The heat transfer processes were the most intensive in the radial direction. The 578 E. Fornalik, J.S. Szmyd angular component of the heat flux budget showed different tendency, but its magnitude was the smallest. 3. Impinging jets 3.1. Experimental apparatus Figure 7 is a schematic illustration of the experimental apparatus used in the present study. Water used as the working fluid was driven by the head difference between constant-head upper tank (1) and an outlet gate of hori- zontally installed test section (4). The gate was made of an acrylic plate to enable optical access for flow visualizations. Two round impinging jet noz- zles (9), with the diameter d of 6mm, were mounted flush with the bottom wall of the test section. The main flow supplied into the test section through contraction (3) served as a crossflow for the impinging jets. Fig. 7. A schematic diagram of the experimental set-up in the case of impinging jets; 1 – upper tank, 2 – wire gauge, 3 – contraction, 4 – test section, 5 – lower tank, 6 – pump, 7 – flowmeters, 8 – valves, 9 – jet nozzles Dimensions of the test section are shown inFig.8 togetherwithgeometrical parameters, jet nozzles arrangements and coordinate system set up in the present study. The test section was rectangular, 432mm wide. The height of the test section, HD, was 30mm in the case of the jets normal to the target wall, and 21mm to keep the same impinging distance in the case of oblique jets. The two jets were situated in an in-line arrangement with a streamwise distance of 10d. The skew angels between the directions of the impinging jets Experimental investigations of jet flows 579 and the crossflow in thehorizontal planewere set tobe +90◦ for theupstream- side jet, Jet 1, and −90◦ for the downstream-side one, Jet 2, while the pitch angle measured up from the nozzle-installed wall was fixed at 45◦ for both jets. The pitch angles of both jets were changed to be 90◦ for the vertical impingement case. Fig. 8. Test section and twin jet nozzles; (a) vertical nozzles, (b) inclined nozzles, in-line arrangement, (c) inclined nozzles, staggered arrangement 3.2. Procedure TheReynolds number of the crossflow, Re, based on the hydraulic diame- ter of the test section was set to be equal to 5000. The velocity ratio of the jet to the crossflow, VR, was set up to be 5. Local Nusselt number, Nu, based on the hydraulic diameter is defined as follows Nu= α(2HD) λ (3.1) where α is the heat transfer coefficient calculated from Eq. (3.2) α= qW TW −T0 (3.2) 580 E. Fornalik, J.S. Szmyd where λ is the thermal conductivity, TW and T0 are the local temperature of the target wall and the inlet flow temperature, respectively, qW is the target wall heat flux calculated from the electric power input. The heat conduction loss towards thebacksideof the targetwallwasneglected since itwas relatively small, approximately 3% of the total heat flux. The heat radiation loss from the surface of the target wall was also neglected. The cross-sectional images of the flow patterns were taken through the acrylic plate with a CCD camera set up downstream. Velocity vectors were estimated with the PIV system from the scattering images of fine TiO2 par- ticles injected into the jet flows and illuminated by a planar Nd:YAG pulse laser light. The temperature distribution of the heat transfer target surface was me- asured with thermochromic liquid crystal sheets attached to the back of the target surface. The images of the color liquid crystal surface were taken with another CCD camera installed above the test section. The colors of the ima- ges were transformed into the temperature distribution by the neural network technique described in Nakabe et al. (1998). Afterwards, the Nusselt number could be estimated. 3.3. Results Thevelocity vectormapsof twinvertical jets togetherwith thedistribution of the Nusselt number are shown in Fig.9. The measured cross-section were at streamwise locations: (a) x/d = 10, (b) x/d = 20, (c) x/d = 30 and (d) x/d = 40. Two large longitudinal counter-rotating vortices are visible. Their positions correspond very well to the position of area of enhanced heat transfer presented in Fig.9e. Values of the Nusselt number in the range from z/d=−5 to z/d=5 are larger than outside this area. The velocity vector maps of two inclined jets in the in-line arrangement together with the distribution of the Nusselt number are shown in Fig.10. Themeasured cross-sections were at four streamwise locations: (a) x/d=10, (b) x/d=20, (c) x/d=30 and (d) x/d=40. The three large-scale longitudi- nal vortices plus one small-scale vortex can be recognized. These vortices can be observed even far downstream from the locations of the nozzles. It can be seen that the vortices forming each of the pairs are rotating in opposite direc- tions, and that the areas of high spanwise velocity, close to the upper target wall, are related to the region of enhanced heat transfer, which is presented in Fig.10e. In Fig.10e, the distribution of the Nusselt number at the same four loca- tions as for the velocity vectormaps are presented. The area of enhanced heat transfer iswell visible.Two regions of highNusselt numbers can be recognized. Experimental investigations of jet flows 581 Fig. 9. Results for the twin vertical jets (Re=5000, VR =5); (a)-(d) cross-sectional velocity vectors: (a) x/d=10, (b) x/d=20, (c) x/d=30 and (d) x/d=40, (e) distribution of the Nusselt number The highest value of the Nusselt number is close to the location at x/d=10 and z/d = 4, Nu = 250 value almost equal to the maximum value in the case of two vertical jets in Fig.9e. The area of enhanced heat transfer beco- meswider, however values of theNusselt number are smaller. The comparison between the results of the velocity and Nusselt numbermeasurements reveals a very good agreement between the vortex location and heat transfer enhan- cement region. The enhanced region is related to the paths of the large-scale longitudinal vortices. The velocity vectormaps of two inclined jets in the staggered arrangement together with the distribution of the Nusselt number are presented in Fig.11. The measured cross-sections were at six locations: (a) x/d=0, (b) x/d=5, (c) x/d = 10, (d) x/d = 20, (e) x/d = 30 and (f) x/d = 40. The fo- ur vortices can be recognized. These longitudinal vortices form two pairs of counter-rotating vortices, which can be observed even far away from the lo- cations of nozzles. The number of vortices is different from the case of the 582 E. Fornalik, J.S. Szmyd Fig. 10. Results for the twin inclined jets, in-line arrangement (Re=5000, VR =5); (a)-(d) cross-sectional velocity vectors: (a) x/d=10, (b) x/d=20, (c) x/d=30 and (d) x/d=40, (e) distribution of the Nusselt number in-line arrangement of the jets. The positions of vortices correspond very well to the region of enhanced heat transfer presented in Fig.11e. The distribution of the Nusselt number presents two regions with maximum values, which are bigger more than twice than in the case of vertical jets and inclined jets in the in-line arrangement. The region of enhanced heat transfer elongates from z/d=−10 to z/d=10 and is also wider than in the case of the in-line arran- gement. 3.4. Comments The large-scale longitudinal vortices are generated in the rectangular duct flow as the result of impinging jets. The number of vortices depends on the arrangement of the jet nozzles. For the vertical jets, two vortices are observed, for the inclined jets in the in-line arrangement three, while for the inclined jets in the staggered arrangement four ones are found.Thepath of the longitudinal vortices corresponds verywell to the heat transfer enhancement regions for all kinds of arrangements of the jet nozzles. The interaction between the jets depending on the nozzles arrangements has serious influence on the vertical structure and the enhanced heat transfer regions as well. Experimental investigations of jet flows 583 Fig. 11. Results for the twin inclined jets, staggered arrangement (Re=5000, VR =5); (a)-(f) cross-sectional velocity vectors: (a) x/d=0, (b) x/d=5, (c) x/d=10, (d) x/d=20, (e) x/d=30 and (f) x/d=40, (e) distribution of Nusselt number 4. Conclusions The experimental studies on two kinds of jet flows were presented. Various methods were applied: in the case of the confined jet a four-wire thermoane- mometer probe was used, while in the case of impinging jets, the PIV and liquid crystal film technique was applied. The choice of the method depen- ded on the character of the flow, fluid and investigated qualities. The results, analysis and comments were reported. 584 E. Fornalik, J.S. Szmyd References 1. BarchilonM., CurtetR., 1964, Some details of the structure of an axisym- metric confined jet with backflow,Trans. ASME: J. Basic Eng., 86, 777-787 2. BeckerH.A.,HottelH.C.,WilliamsG.C., 1962,Mixing andflow induc- ted turbulent streams,Proceedings of the 9th Int. Symposium on Combustion, The Combustion Inst., Pittsburgh, 7-20 3. Bouchez J.P., Goldstein R.J., 1975, Impingement cooling from a circular jet in a cross flow, Int. J. Heat and Mass Trans., 18, 719-730 4. Brighton J.A., Jones J.B., 1964, Fully developed turbulent flow in annuli, Trans. ASME: J. Basic Eng., 86, 835-844 5. Champagne F.H., Wygnanski I.J., 1971, An experimental investigation of coaxial turbulent jets, Int. J. Heat and Mass Trans., 14, 1445-1464 6. Chevray R., Tutu N.K., 1978, Intermittency and preferential transport of heat in a round jet, J. Fluid Mech., 88, 133-160 7. Craya A., Curtet R., 1955, On the spreading of a confined jet, Comptes- Rendus Acad. Sci., 241, 621-622 8. Curtet R., 1958, Confined jets and recirculation phenomena with cold air, Combustion and Flame, 2, 383-411 9. Fornalik E., JaszczurM., Szmyd J.S., KlajnyR., NakabeK., Suzuki K., 1998, An application of a triaxial probe with temperature sensor to ana- lysis of turbulent heat transfer in a confined coaxial jet,Proceedings of the 4th KSME-JSME Fluids Engineering Conference, Pusan, 513-516 10. Fornalik E., Szmyd J.S., 2005, Turbulent heat transfer in a confined jet, Progress in Computational Fluid Dynamics, 5, 136-143 11. FrotaM.N.,MoffatR.J., 1982,Triple hot-wire technique formeasurements of turbulence in heated flows,Proceedings of the 7th Int. Heat Transfer Confe- rence, Monachium, 491-496 12. Goldstein R.J., Franchett M.E., 1988, Heat transfer from a flat surface to an oblique impinging jet,Trans. ASME: J. Heat Transfer, 110, 84-90 13. HeistD.K.,Castro I.P., 1998,Combined laser-dopplerandcoldwireanemo- metry for turbulent heat fluxmeasurement,Experimetns in Fluids, 24, 375-381 14. Huang Y., Ekkad S.V., Han J.C., 1996, Detailed heat transfer coefficient distributions under an array of inclined impinging jets using a transient liquid crystal technique,Proceedings of the 9th Int. Symposium on Transport Pheno- mena in Thermal-Fluid Engineering, Singapore, 807-812 Experimental investigations of jet flows 585 15. Kang Y., Suzuki K., 1982, Numerical study of wall heat transfer in the recirculating flow region of a confined jet, Heat Transfer Japanese Research, 11, 1, 44-69 16. Kang Y., Suzuki K., Sato T., 1979, Convective heat transfer in an axisym- metric confined jet, Studies in Heat Transfer, Edit. T.F. Irvine et al., Hemi- sphere, 103-125 17. Kasagi N., 1991, Direct numerical simulation data bases: an effective tool in fundamental studies of turbulent heat transfer, Proceedings of the Japan-U.S. Science Seminar on Computers in Heat Transfer Science, Oiso 18. Keffer J.F., Olsen G.J., Kawall J.G., 1977, Intermittency in a thermal mixing layer, J. Fluid Mech., 79, 595-607 19. Metzger D.E., Korstad R.J., 1972, Effects of crossflow on impingement heat transfer,Trans. ASME: J. Eng. Power, 94, 35-42. 20. NakabeK., Higashio A., ChenW., Suzuki K., Kim J.H., 1988,An expe- rimental study on the flow and heat transfer characteristics of longitudinal vortices induced by an inclined impinging jet into a crossflow, Proceedings of the 11th Int. Heat Transfer Conference, Kyongju, 439-444 21. Razinsky E., Brighton J.A., 1971, Confined jet mixing for nonseparating conditions,Transactions of the ASME J. Basic Eng., 93, 333-349 22. Saad N.R., Mujumdar A.S., Messeh W.A., Douglas W.J.M., 1980, Lo- cal heat transfer characteristics for staggered arrays of circular impinging jets with crossflow of spent air,ASME Paper 80-TH-23, 105-112 23. Sparrow E.M.,GoldsteinR.J., RoufM.A., 1975,Effect of nozzle-surface separationdistance on impingement heat transfer for a jet in a crossflow,Trans. ASME: J. Heat Transfer, 97, 528-533 24. Suzuki K., Suga K., Oshikawa Y., 1987, LDV Measurement of turbulen- ce and test of turbulence models in recirculating flow, Proceedings of the 6th Symposium on Turbulent Shear Flows, Toulouse, 15.4.1-15.4.6 Badania eksperymentalne przepływów strumieniowych Streszczenie Zaprezentowane zostały badania eksperymentalne dotyczące dwóch rodzajów przepływów strumieniowych: strugi pierścieniowej oraz strugi uderzającej. Czterow- łóknowa sonda termoanemometrycznawykorzystana zostaławbadaniach związanych ze strugą pierścieniową.Analiza danychpozwoliła nawyznaczeniewielkości charakte- ryzujących przepływ turbulentny, ze szczególnym uwzględnieniem bilansu turbulent- nego strumienia ciepła. W badaniach strugi uderzającej pod działaniem przepływu 586 E. Fornalik, J.S. Szmyd krzyżowego do wyznaczenia pola prędkości wykorzystanometodę PIV (ang.Particle Image Velocimetry). Badaniomprzepływu towarzyszyła analiza wymiany ciepła przy użyciu termoczułych ciekłych kryształów w postaci cienkiego filmu. Zaobserwowane zostały wiry, które w głównej mierze odpowiadały za intensyfikację wymiany ciepła, co zostało potwierdzone wyznaczonymi polami prędkości. Manuscript received January 17, 2007; accepted for print May 28, 2007