跨音速涡轮叶栅尾流中速度场PIV检测方案(CCD相机)

智能文字提取功能测试中

Proceedings of Montreal 2018 Global Power and Porpulsion Forum7th -9th May 2018 www.gpps.global GPPS-NA-2018-0013 ON THE CHALLENGE OF FIVE-HOLE-PROBE MEASUREMENTS ATHIGH SUBSONIC MACH NUMBERS IN THE WAKE OF TRANSONICTURBINE CASCADES Marcel Boerner， Martin Bitter， Reinhard Niehuis Institute of Jet Propulsion Bundeswehr University Muenchen marcel.boerner@unibw.de Neubiberg， Germany Five-hole-probes are common use in turbomachinery flowinvestigations， even though， inserting a probe into a flow fieldinevitably induces perturbations to the flow which can falsifythe measurement results， especially when exposed to transonicflows. The objective of the investigations presented here is toevaluate the Mach number measurements of a five-hole-probe(5HP) in the wake flow of a transonic turbine cascade at enginerelevant Reynolds numbers by comparing them to the resultsof particle image velocimetry (PIV). Furthermore， PIV mea-surements were performed with inserted probe to investigatethe influence of the probe on the wake flow field. Togetherwith a sensitivity study of 5HP measurements in flow regimesclose to Ma=1， the results demonstrate how the measurementuncertainty can be improved in high subsonic flow regimes. INTRODUCTION Experimental investigations on turbine and compressorcascades are essential for the validation of numerical setupsand to develop improved numerical models to enhance the pre-diction accuracy of flow phenomena. Measurements on cas-cades in particular are necessary to give information about realprofile losses and performance of new blade designs. Recently，numerous research projects are focusing on high-speed turbo-machinery profiles because of the benefit arising from increas-ing the spool speed especially in low pressure turbines of air-craft engines. Therefore， five-hole-probes (5HP) are in generalwell-suited to determine the desired flow quantities. But a ma-jor challenge of using such intrusive measurement techniquesis the sensitivity of the flow to a change in static pressure，especially in transonic flow regimes (Fransson et al.， 1988).Hancock (1988) showed theoretically that the sensitivity mea-suring the true static pressure has to be zero at Ma=1. Thisconstraint was affirmed in several experimental investigations， e.g. (Fransson et al.， 1988； Kost， 2009； Langford et al.， 1982).On the contrary， non-intrusive measurement methods， e.g. par-ticle image velocimetry (PIV)， feature the benefit of causingless flow perturbations. In fact， the use of PIV in high-speedflows has been successfully demonstrated in many cases. Forexample， Sartor et al. (2012) investigated a shock-induced sep-aration over a bump and achieved good agreement with otherinvestigations from literature. From 300 um above the wall upto the free stream region the measuring field covers the wholesubsonic and low supersonic flow regime up to 420ms- inone camera frame. Another example is given by Klinner andWillert (2013) for the use of PIV in a cascade wind tunnelwith a transonic compressor profile featuring a detached pas-sage entry shock wave at the leading edge. Therein， the relax-ation time of the tracer particle in the flow was small enough todetermine the shock wave position in accordance to Schlierenimages. However， one drawback of most optical measuringtechniques is the lack of high accuracy information about thethermodynamic state of the flow field， which is why measure-ments with pneumatic multi-hole-probes are common practiceand still beneficial for turbomachinery applications (Raffel etal.，2007). The experimental investigations presented here focus ona two-dimensional midspan wake flow of a transonic turbineprofile with high subsonic exit flow Mach numbers close tounity. In order to evaluate the comparability between intrusiveprobe and non-intrusive optical measurement methods in thishighly sensitive flow regime， pneumatic wake traverses witha five-hole-probe were compared to planar (2D2C-)PIV mea-surements focusing on the potential interactions between themeasurement setup (i.e. the probe and its support) and the exitflow field of the turbine cascade. Most investigations lack thisinformation about the influence of the probe on the flow fieldand comparisons of intrusive and non-intrusive measurement techniques are rare. Veerman et al. (2004)， for example， in-vestigated the wake vortex system downstream of an aircraftwing with an high lift device by the use of stereoscopic PIVand a traversable rake consisting of ten five-hole-probes. Thechange of the flow field by the introduction of the probe rakeincluding its support was well recognizable in the PIV velocityfield at a free stream velocity of about 60 ms-1. However， thequantitative velocity results for this low-speed case were stillin good accordance to each other. It has to be investigated， ifthe same agreement can be demonstrated in high-speed flowsclose to Ma =1， which is the objective of the work presentedhere. EXPERIMENTAL SETUPHigh-Speed Cascade Wind Tunnel The experimental investigations presented in this paperwere carried out in the High-Speed Cascade Wind Tunnel(HGK) at the Institute of Jet Propulsion (ISA) of the Bun-deswehr University Muenchen. The test facility shown inFig. 1 is an open loop， continuously operating wind tunnel withan open test section optimized for linear cascades. The windtunnel is placed inside a pressure tank of 11.8 m length and adiameter of 4m. The wind tunnel flow is supplied by a sixstage axial compressor with an adjustable bypass for stall pre-vention enabling a compression ratio up to 2.14. The drivingunit for the compressor is placed outside the tank and consist-ing of a 1.3 MW a.c. electric motor， a hydraulic coupling and agear box. Finned coolers downstream of the compressor outletcompensate the heat of compression and regulate the desiredambient temperature. Flow straighteners and a settling cham-ber follow the cooler to obtain uniform flow conditions beforethe flow is re-accelerated within a nozzle towards the flow pathupstream of the outlet where the cascade is placed. The rectan-gular flow path downstream the nozzle is 300 mm wide and theheight can be adjusted from 250 mm to 500 mm. Furthermore，different turbulence generators can be inserted into the nozzleto vary the turbulence level. Since the ambient pressure in thevessel can be reduced down to approximately 3.5 kPa by twovacuum pumps， the Reynolds and Mach number can be variedindependently to achieve realistic turbomachinery conditions， which makes the facility unique among very few comparablefacilities world wide. Important parameters of the operatingrange of the wind tunnel facility are given in Tab. 1. For moredetailed information the reader is referred to Sturm and Fottner(1985). The operating point of the turbine cascade is defined bythe theoretical exit flow Mach and Reynolds number (Ma2，th，Re2，th) assuming an isentropic flow through the cascade. Thetheoretical exit flow Mach number is derived by with the total pressure pti upstream and the static pres-sure pe downstream of the cascade. Combined with the totaltemperature T at the cascade inlet and the chord length I thetheoretical exit flow Reynolds number can be calculated by with the Sutherland constants C and C2. In this investiga-tion， a Reynolds number based on the dimensions of five-hole-probe Re2，th，p is more suitable. This can be calculated analo-gous to Eqn. 2 by using the probe head diameter dp instead ofthe profile chord length l. TABLE 1 Operating range of the High-Speed Cascade WindTunnel test facility Reynolds number 2·10 m-1≤Re/l≤ 1.6·10'm-1 Mach number 0.20.8. In order to achieve better performance in the transonicflow regime other head shapes of pneumatic probes should beconsidered. Initial investigations with a wedge type probe al-ready demonstrated improvements by reducing the uncertaintyin transonic flows and the measurement results are less affectedby the high gradients. Continuing investigations are sched-uled on that. Furthermore， reducing the size of the probe headwill be on focus， because concentrating the pressure taps in asmaller region is favorable in gradients flows， too. An opti-mized probe design seems to be highly promising to enhancethe Mach number measurements in the wake flow of transonicturbine cascades. NOMENCLATURE Symbols C1， C2 Sutherland constants d diameter kMa，ko，ky coefficients of the five-hole-probe 1 chord length Ma Mach number pressure R dynamic pressure specific gas constant of air Reynolds number temperature t blade pitch turbulence intensity velocity pitch-wise coordinate X stream-wise coordinate Greek heat capacity ratic Subscripts 0...4 numbering of five-hole-probe pressure tapsconditions upstream of cascade2 conditions downstream of cascade 5HP five-hole-probe aX axial direction (normal to cascade outlet plane) C pressure tank ambient valuehigh highest turbine exit flow Mach number caseprobereference at u/t=0re J t total condition th theoretical (isentropic conditions) Abbreviations 2C2D two components in two dimensional plane 5HP five-hole-probe CCD charge-coupled device DEHS Di-Ethyl-Hexyl-Sebacat HGK High-Speed Cascade Wind Tunnel ISA Institute of Jet Propulsion PIV particle image velocimetry Nd：YAG neodymium-doped yttrium aluminum garnet ACKNOWLEDGMENTS The research leading to this paper was based on the AGTurbo project in cooperation with the MTU Aero EnginesAG. The transonic turbine profile used here was developedby MTU. The authors would like to thank MTU for the per-mission to use the profile and the thought-provoking impulsesinitiating this investigation. Furthermore， the authors wouldlike to thank the technical staff at ISA for their great supportin gaining the experimental results. REFERENCES Bohn， D. (1977).“Untersuchung zweier verschiedeneraxialer Uberschallverdichterstufen unter besondererBeriicksichtigung der Wechselwirkungen zwischen Lauf-und Leitrad". PhD thesis.Aachen： RWTH Aachen. Bohn， D. and H. Simon (1975).“Mehrparametrige Ap-proximation der Eichraume und Eichflachen von Unterschall-bzw.Uberschall-5-Loch-Sonden”. tm-Technisches MessenLieferung 470. DOI： 10.1524/teme.1975.468479.jg.81. Carscallen， W. E.， T. C. Currie， S. I. Hogg， and J. P.Gostelow (1999).“Measurement and Computation of EnergySeparation in the Vortical Wake Flow of a Turbine Nozzle Cas-cade". ASME J. Turbomach. 121.4， p. 703.DOI： 10.1115/1.2836723. Eckert， E. and W. Weise (1942).“Messungen der Temper-aturverteilung auf der Oberflache schnell angestromter unbe-heizter Korper”. Forsch Ing-Wes 13.6， pp. 246-254. DOI： 10.1007/BF02585343. Fransson，T.， F. Schaller， and K. Schulz (1988).“Changeof Flow Conditions due to the Introduction of an AerodynamicProbe during Calibration”. In： 9th Symposium on MeasuringTechniques in Transonic and Supersonic Flow in Cascades andTurbomachines.Oxford， UK. Gostelow， J. P. and W. E. Carscallen (2010).“Investigationof Eckert-Weise Energy Separation in the Wakes of TurbineBlades and Circular Cylinders”. In： 17th Australasien FluidMechanics Conference. Auckland， New Zealand. Grimshaw， S. D. and J. V. Taylor (2016).“Fast settlingmillimetre-scale Five-Hole-probe”. In： ASME Turbo Expo2016： Turbomachinery Technical Conference and Exposition.Seoul， South Korea. Hancock， P. E. (1988).“A Theoretical Constraint at M=1for Intrusive Probes and Some Transonic Calibrations of Sim-ple Static-Pressure and Flow-Direction Probes". In： 9th Sym-posium on Measuring Techniques in Transonic and SupersonicFlow in Cascades and Turbomachines.Oxford， UK. Hoenen， H. T.， R. Kunte， P. Waniczek， and P. Jeschke(2012).“Measuring Failures and Correction Methods forPneumatic Multi-Hole Probes". In： ASME Turbo Expo 2012：Turbine Technical Conference and Exposition. Copenhagen，Denmark， p. 721. DOI： 10.1115/GT2012-68113. Klinner， J. and C. Willert (2013).“Application of Tomo-graphic PIV in a Transonic Cascade". In： 10th InternationalSymposium on Particle Image Velocimetry. Delft， Netherlands. Kost， F. (2009).“The behaviour of probes in transonicflow fields of turbomachinery”. In： 8th European Conferenceon Turbomachinery (ETC). Graz， Austria. Langford， R. W.， K. R. Keeley， and N. B. Wood (1982).“Investigation of the Transonic Calibration Characteristics ofTurbine Static Pressure Probes”. In： ASME 1982 InternationalGas Turbine Conference and Exhibit. London， England. DOI：10.1115/82-GT-280. Raffel， M.， C. E. Willert， Steven T. Wereley， and J.Kompenhans (2007). Particle Image Velocimetry： A PracticalGuide. Experimental Fluid Mechanics. Springer. Sartor， F.， G. Losfeld， and R. Bur (2012). “PIV study on ashock-induced separation in a transonic flow”.Experiments inFluids 53.3， pp.815-827. DOI： 10.1007/s00348-012-1330-4. Sturm， W. and L. Fottner (1985).“The High-Speed Cas-cade Wind-Tunnel of the German Armed Forces UniversityMunich". In： 8th Symposium on Measuring Techniques forTransonic and Supersonic Flows in Cascades and Turboma-chines. Genua， Italy. Truckenmiiller， F. and H. Stetter (1996).“Measurementerrors with pneumatic probes behind guide vanes in transonicflow-fields”. In： 13th Symposium on Measuring Techniques forTransonic and Supersonic Flow in Cascades and Turboma-chines. Ziirich， Switzerland. Veerman， H. P. J.， F. L. A. Ganzevles， and K. Pengel(2004).“Wake Vortex Investigations by Means of StereoscopicPIV and 5-Hole Probe". In： Particle Image Velocimetry： Re-cent Improvements： Proceedings ofthe EUROPIV 2 Workshopheld in Zaragoza， Spain， March 31 -April 1， 2003. Ed. byStanislas， M.， Westerweel，J.， and Kompenhans， J. Berlin，Hei-delberg： Springer， pp. 67-82. Vinnemeier， F.， L. Simon， and W. Koschel(1990).“Ko-rrektur des Kopfgeometrieeinflusses einer Fiinfloch-Druck-sonde auf die MeBergebnisse”. tm- Technisches Messen 57.DOI： 10.1524/teme.1990.57.jg.296. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives . International License CC-BY-NC-ND .      Five-hole-probes are common use in turbomachinery flow investigations, even though, inserting a probe into a flow field inevitably induces perturbations to the flow which can falsifythe measurement results, especially when exposed to transonic flows. The objective of the investigations presented here is to evaluate the Mach number measurements of a five-hole-probe (5HP) in the wake flow of a transonic turbine cascade at engine relevant Reynolds numbers by comparing them to the results of particle image velocimetry (PIV). Furthermore, PIV measurements were performed with inserted probe to investigate the influence of the probe on the wake flow field. Together with a sensitivity study of 5HP measurements in flow regimes close to Ma=1, the results demonstrate how the measurement uncertainty can be improved in high subsonic flow regimes.