Copyright ©1997, American Institute of Aeronautics and Astronautics, Inc.

AIAA Meeting Papers on Disc, January 1997A9715692, AIAA Paper 97-0663

Improvements to the UTA high Reynolds number transonic wind tunnel

W. S. StuessyTexas Univ., Arlington

Jennifer L. PeeplesTexas Univ., Arlington

Baxter R. Mullins, Jr.Texas Univ., Arlington

AIAA, Aerospace Sciences Meeting & Exhibit, 35th, Reno, NV, Jan. 6-9, 1997

The high Reynolds number, transonic Ludwieg concept tunnel at the University of Texas at Arlington has been returned toservice with improved facility pressure monitoring systems. An improved response static pressure measuring system has beenincorporated. The calibrated range of operation has also been extended to allow for testing at higher Reynolds numbers.Comparison of the improved pressure measurement system is compared with the previous system. The effect of increasedReynolds number upon the response of the new system is also discussed. (Author)

Page 1

AIAA-97-0663-

IMPROVEMENTS TO THE UTA HIGH REYNOLDS NUMBER TRANSONIC WIND TUNNEL

W. Scott Stuessy, Jennifer L. Peeples,1^ and Baxter R. Mullins, Jr.*The University of Texas at Arlington

Aerodynamics Research CenterArlington, Texas 76019

ABSTRACT

The high Reynolds number, transonic Ludwiegconcept tunnel at The University of Texas at Arling-ton has been returned to service with improved facil-ity pressure monitoring systems. An improved re-sponse static pressure measuring system has beenincorporated. The calibrated range of operation hasalso been extended to allow for testing at higher Rey-nolds numbers. Comparison of the improved pressuremeasurement system is compared with the previoussystem. The effect of increased Reynolds numberupon the response of the new system is also dis-cussed.

INTRODUCTION

The Ludwieg tube concept was first proposed asa high-Reynolds number, transonic flow simulationby K. Ludwieg of Germany in 1957.' NumerousLudwieg tube tunnels are in existence, both in theU.S. and in Europe. The facility at the University ofTexas at Arlington (UTA) was initially developed atAEDC in the early 1970's as a pilot tunnel to evaluatethe proposed Air Force concept for the NationalTransonic Facility. Extensive evaluation studies ofthe operational characteristics and flow quality wereconducted at AEDC between 1971 and 1975.2

The facility was donated to UTA in 1978 andplaced into service in 1984 after development of thenecessary mechanical, pneumatic, electronic control,and data acquisition systems.3 The tunnel was movedto a new complex in 1986.4 The facility was reacti-vated in early 1996 after several years of inactivityand improvements were made to the pressure meas-urement system.

'Faculty Research Associate, Department of Mechanical andAerospace Engineering, Senior Member, AIAA

Graduate Research Assistant, Department of Mechanical andAerospace Engineering, Student Member, AIAA

* Adjunct Associate Professor, Department of Mechanical andAerospace Engineering, Senior Member AIAA

Copyright © 1996 by W. S. Stuessy. Published by theAmerican Institute of Aeronautics and Astronautics, Inc.with permission.

The three tunnel pressure measurement systems wereall modified by reducing the internal volume of each onethereby improving the response. The new systems werechecked and calibrated to a centerline probe. The facilitywas next checked and calibrated at higher operating pres-sures than had been previously demonstrated.

THEORY OF OPERATION

Ludwieg tube tunnels are based on an unsteady ex-pansion wave concept to acceleration high-pressure airstored within a charge tube to transonic Mach numbers.The expansion wave produces a flow process that is simi-lar to the flow within the driver tube of a conventionalshock tube. A schematic of the UTA Ludwieg tube isshown in Figure 1 and the idealized wave diagram inFigure 2. The components of the tunnel consist of thecharge tube, convergent nozzle, test section, ejector flapsection, diffuser, and starting valve.

Flow in the tunnel is initiated by charging the entiresystem to the desired charge tube pressure level, and thenrapidly opening a starting valve to initiate flow. This ac-tion generates an unsteady expansion wave that propa-gates upstream through the diffuser, test section, nozzle,and into the charge tube to initiate the flow towards theopen valve at the downstream end of the system. Once theexpansion wave clears the convergent nozzle, steady flowwithin the test section is maintained for the time durationit takes the expansion wave to travel the length of thecharge tube, reflect, and return to the test section. The33.8 m (111 ft) length of the charge tube provides a theo-retical steady flow period of 185 msec; however, this timeis reduced to about 120 msec due to the time period actu-ally required to open the starting valve and by the un-steady flow phenomena associated with the flow initiationin the plenum cavity surrounding the porous walled testsection. Subsequent wave reflections occur, but theircomplex nature results in a deterioration of flow quality.

Normal isentropic flow acceleration through a con-verging nozzle would normally be limited to the attain-ment of sonic flow at the nozzle exit; however, the effectof mass removal relieves the choking effect and permitsacceleration of the flow to supersonic speeds in the con-stant area test section. The range of suction mass flow

1American Institute of Aeronautics and Astronautics

33.8m

HIGH PRESSUREAIR CHARGINGSYSTEM

" ~-t

NOZZLE

THRUST COLLARCHARGE TUBE

(33.3 cm dia)

• 2.1 m BUILDINGWALL

£tTEST SECTION LINERPLENUM CHAMBER

V EJECTOR FLAP CYLINDER

-MODEL SUPPORT SECTION-SLIDING SLEEVESTARTING VALVE

EXHAUSTSILENCER

PLENUM EXHAUST

\_ THRUSTSTAND •EXHAUST SPHERE

Fig. 1. Elevation view of The University of Texas at Arlington's HIRT facility.

rates for the HIRT facility allows acceleration to amaximum Mach number of about 1.2.

The starting process is greatly complicated by thepresence of the ventilated walls and the surroundingplenum cavity required for collecting the suction flowand discharging it to the surroundings. Figure 3 illus-trates the starting process for the plenum cavity.When the initial expansion wave generated by open-ing the downstream main starting valve reaches thetest section, the initial effect is to pump flow from theplenum cavity into the test section through the porouswalls. This flow from the cavity into the test sectionis reversed by rapidly pumping the cavity to the at-

STEADYFLOWTESTTIME

INCIDENTWAVES

DISTANCE

Fig. 2. The Ideal Wave Diagram for the HIRT Tunnel.

mosphere through the plenum exhaust system before thetest section flow can approach an equilibrium steady flowcondition.

PACHJTY

The capabilities of the tunnel include a Mach numberrange of 0.5 to 1.2, with a corresponding Reynolds num-ber range of 4xl07 to 43xl07 per meter (IxlO6 to llxlO6

per inch) as shown in Figure 4. A unique capability of theHIRT tunnel is the ability to independently vary Mach andReynolds numbers over the full operating range of thetunnel. The test section employs a conventional AEDCporous wall design with a rectangular cross section meas-uring 18.5 cm (7.28 in) x 23.2 cm (9.15 in), and is 64 cm(25.4 in) long. The porous walls minimize shock wavereflections from the tunnel walls as well as provide forsubstantial alleviation of tunnel wall interference effects.The porous wall design also allows acceleration to lowsupersonic Mach numbers using the porous walls and afixed-area-ratio convergent nozzle. Data from AEDC in-dicate a practical limit for airfoil testing of about 15x106

chord Reynolds number, determined by the practicalmodel size limitation of the tunnel.

Ludwieg TubeThe charge tube is 35.5 cm (14 in) in diameter and

33.8 m (111 ft) long. It can be charged to a maximumpressure of 4.55 MPa (660 psia), which produces a steadyflow stagnation pressure of about 3.44 MPa (500 psia).The nozzle is 47 cm (18.5 in) in length, and has a con-traction ratio of 2.27. The nozzle provides a transitionfrom the circular geometry of the charge tube to the rec-tangular test section geometry. The test section measures

American Institute of Aeronautics and Astronautics

Flentm Teat &j ectorCharge tube Nozzle cavity /section S flaps ,

Diaphragm (ruptured toinitiate secondary flow)

'_ Variable orificeball valveFlex-flo valve

Fig. 3. Plenum cavity starting process.

18.5 cm (7.3 in) by 23.2 cm (9.15 in), and is 64.5 cm(25.4 in) long. The porous walls are of conventionaldesign, consisting of two stacked plates with 60-deginclined holes and having a tapered porosity patternin the upstream one thkd of the test section. The po-rosity can be varied manually from 3.5 to 10 percentby moving one plate relative to the other. This ad-justment capability is provided on all four walls. Thevariable porosity walls provide for wave cancellationfor supersonic flow in the tunnel, and provide an ef-fective means for partially alleviating tunnel wallinterference effects. They also allow the attainment oflow supersonic Mach numbers with a fixed area ratioconvergent nozzle. The test section is surrounded bya plenum cavity that has a volume approximately 1.75times the test section volume. Details of the nozzle,test section, diffuser, and starting valve are shown inFigureS.

The ejector flaps at the downstream end of the testsection are 7.62 cm (3.0 in) long and can be openedto provide a maximum gap height of 2.29 cm (0.9 in).Flow in the plenum cavity is exhausted to atmosphere

1.2

0.2 1.00.4 0.6 0.8MACH NUMBER

Fig. 4. Operational capability of the Ludwieg tubetransonic wind tunnel.

1.2

through the system as shown in Figure 3. Eight 5.08 cm (2in) I.D. flexible hoses are attached around the test sectionand connected to a common manifold. Downstream of themanifold is a diaphragm holder consisting of two platesheld together by a Grayloc clamp. Mylar diaphragm mate-rial of the required thickness is ruptured at a specifiedtime interval from the start of the run by means of a ple-num exhaust cutter (PEC) that is actuated by a pneumaticcylinder. Downstream of the diaphragm assembly, the15.2 cm (6 in) diameter plenum exhaust line vents to at-mosphere through a 6-inch variable orifice ball valve forwhich the open area is preset at a prescribed value prior to

The first section of the diffuser contains another tran-sition section with contours designed to change smoothlyfrom the rectangular cross-section to a circular cross-section. Downstream of the transition is a conical diffuserwith a 3-deg wall angle and an exit diameter of 40.6 cm(16 in). Provisions for mounting a model support sting areprovided at the diffuser entrance.

A 30.5 cm (12 in) and a 40.6 cm (16 in) diametersliding sleeve valves (SSV) are available as a main start-ing valves. The sliding sleeve valve is driven by a pneu-matic cylinder and requires approximately 60-80 msec toopen (depending on the charge tube pressure and pneu-matic cylinder pressure). The larger SSV takes longer toopen than the smaller one.

The 16-in diameter sliding sleeve valve has twenty-seven, 3-in diameter ports in 3 rings around the valvebody. The valve is opened and closed by a pneumaticcylinder and the valve can be opened in approximately 80msec. Each of the 27 ports is a short pipe nipple whichcan be capped to adjust the mass flow rate out of thevalve. The main tunnel flow, as well as that from bothbranches of the plenum system, empties into a large ex-haust sphere. From there the flow exits the buildingthrough a 1.22 m (4 ft) diameter duct and into a silencerwhich turns the flow upward before being exhausted toatmosphere.

SubsystemsA control board is used to regulate and control the

pressurized air from its supply system. A low pressurecompressor provides control air for various pneumaticallycontrolled remote valves. A high pressure compressor isused for pressurization of the charge tube as well as pres-surizing the FFV, SSV-OPEN, SSV-CLOSE and PECactuators. Each subsystem has one accumulator that ischarged to a required pressure level. In the event that arun needs to be aborted, dump valves and vent lines areprovided to vent pressurized air to the atmosphere.

American Institute of Aeronautics and Astronautics

12-IN SLID ING SLEEVESONIC PLENUM PLENUM EXHAUST MODELSUPPORT STARTVALVE

CHARGE NOZZLE SHELL (10 PLACES TYP) STRUT (SHOWN CLOSED)TUBE \ -—— ————-f——————\———130.4"13.94

EJECTORTEST SECTION WALL' FLAPSSUPPORT STRUCTURE

INSTRUMENTATION LEADSAND PRESSURE TUBE BUNDLE

Fig. 5. Sectional view of the nozzle, test section, diffuser and both large and small sliding-sleeve starting valves ofthe HIRT facility.

Solenoid valves are used to release the accumu-lated pressure into a pneumatic actuator or cylinder atthe prescribed time to start the wind tunnel. Two so-lenoid valves are used to control the charge tubepressure. One is used to control the flow into thecharge tube. The other is used to bleed the chargetube in case pressurization over the required levelshould occur, or in case the run needs to be aborted.

Data Acquisition / Electronic Control SystemsLocated adjacent to the test section, the tunnel

control system5 operates solenoid valves that open thesliding sleeve valve to initiate tunnel flow, cuts a dia-phragm to start the plenum exhaust system flow, andclosing a flex-flow valve in the plenum exhaust flowline after culmination of the starting transient. Thecontrol system also sends a signal to the data acquisi-tion system to initiate data acquisition. The slidingsleeve valve is then closed after the completion of thetest run. Once the run is completed, the data is trans-ferred from the data acquisition system to the maincomputer located inside the control room. The data isthen processed while the tunnel is being prepared forthe next run.

The instrumentation includes a single KuliteXTS-1-190-200 pressure transducer with a pressure

rating of 1.38 MPa (200 psi) and can be operated up to2.75 MPa (400 psi), several Kulite ITQS-500F-500SGpressure transducers with a pressure rating of 3.44 MPa(500 psi), and an Optoelectronics precision ThermometerT-100 temperature sensor. The pressure transducers areused to measure the charge tube total, charge tube static,test section plenum cavity, and test specific pressuresduring the test. The temperature sensor is used to obtainthe temperature prior to a test run. The standard procedureis to use the charge tube stagnation pressure transducer asa reference signal for calibration of the static pressuretransducers during charging of the tunnel. The Kulitetransducers used for normal data acquisition are not tem-perature compensated, and do not maintain their calibra-tion over long periods of time. The normal calibrationprocedure that was developed at AEDC included cali-brating the transducers against an accurate temperaturecompensated transducer during the charging cycle. Thesame procedure is used at UTA. The ITQS-500F-500SGtransducers are calibrated against the single XTS-1-190-200 transducers during the charging operation at threepoints near the pressure expected during the test. The sin-gle XTS-1-190-200 pressure transducer is temperaturecompensated and has very little drift over time while theother transducers can drift with time. This method sub-stantially reduces the errors associated with the drift.

American Institute of Aeronautics and Astronautics

The instrumentation is connected to a DSP tech-nology data acquisition system which has the capa-bility of 100 kHz sampling rate, 12 bits of accuracy,and 48 channels each with its own amplifier andanalog to digital converter to allow for simultaneoussampling of all channels. The system has 512 Kilo-bytes of memory available for distribution to thechannels being utilized. The data acquisition systemis connected to a 486-DX, 33 MHz IBM-compatiblePC via a GPIB 488 bus for data retrieval, storage, andmanipulation.

Once the tunnel flow has been started and theplenum diaphragm has been cut, the steady state con-ditions are defined by the interactions of the secon-dary flow through the ball valve, the ejector flapopening, the wall porosity, the tunnel blockage, andthe main valve characteristics. The wall porosity andthe test section blockage (ratio of model cross sec-tional/area test section area) are of secondary impor-tance in determining the mean test section flow.

Starting TransientThe leading edge of the expansion waves gener-

ated by opening the sliding sleeve valve and cuttingthe plenum exhaust system diaphragm should reachthe test section at the same time. The differing char-acteristics of the pneumatic actuation systems neces-sitated an experimental evaluation of the proper timedelay settings for operation of the various startingsub-systems. The determination of the time delaysbetween actuation of the various starting mechanismsis largely accomplished by an empirical process. Thestagnation and static pressures in the charge tube ap-pear to stabilize in about 60-70 msec, but the testsection static pressures do not approach equilibriumvalues until much later. AEDC test data showedstarting times on the order of 80 msec are possible.2

EXPERIMENTAL RESULTS

A centerline probe was installed in the test sec-tion to obtain the axial pressure distribution throughthe test section. The front part of the probe mounts inthe downstream end of the charge tube, whereas thedownstream end mounts in the model support sec-tion of the diffuser / transition section. Pressure tapsare available throughout the nozzle and test section.Eleven ports were chosen in the region where themodels are normally mounted.

These measurements are supplemented by staticand total pressure measurements in the charge tubejust upstream of the nozzle entrance. The total pres-

sure measurement allows a Mach number to be calculatedfor the centerline probe static pressure measurements.

Test Section Mach Number VariationThe axial Mach number distribution within the test

section for a range of test section Mach numbers is shownin Figure 6. The data scatter is within about ±0.50 percent,which is comparable to published data from the AEDCcalibration. The Mach number variation for the datashown in Figure 6 was obtained by control of the plenumexhaust flow rate by variation of the plenum exhaust sys-tem ball valve setting.

Modifications or ImprovementsThe tunnel monitoring probes and instrumentation

consists of a charge tube total system, a charge tube staticsystem, a charge tube temperature system, and a plenumcavity system. The charge tube systems are located justupstream of the nozzle in the charge tube. The total pres-sure system is actually two total pressure probes whosepressure lines are mechanically connected. Two transduc-ers are used to measure the total pressure. The XTS-1-190-200 transducer is used for charge pressure monitoringand for the calibration standard against which all otherfacility transducers are calibrated against. The secondtransducer is used for obtaining the run data.

The charge tube static system consists of a static pres-sure probe and one transducer. The plenum chamber pres-sure system consists of four static probes mounded in thechamber surrounding the test section. Originally, all fourprobes were connected together to mechanically averagethe pressure and was measured with a single transducer.The charge tube temperature probe obtains the tempera-ture just before initiation of the expansion wave.

Test Section Mach Number Variation

Test Section Coordinate (cm)

Fig. 6. Test section Mach Number variation along thecenterline.

American Institute of Aeronautics and Astronautics

The charge tube total and static probes were notchanged or altered, but the external tubing from thetunnel was shortened as much as practically possible.The total pressure system was shortened from 37.5 to28 in. of 1/16-in ID tubing. The static pressure tubingwas shortened form 25.75 to 9.75 in. using the sametubing. The tubing internal to the probes was left un-changed.

The plenum pressure system was totally replacedwith new components before satisfactory results wereobtained. Initially, only the probes themselves werereplaced. The original probes were 14 in. long with a0.180-in ID with four 0.063-in diameter static portsaround the probe. The probes were replaced withprobes of the same length and outside diameter, butwith an ID of 0.083-in. The tubing from the probes tothe transducer was not changed and was 1/16-in OD x0.042-in ID. The response was greatly improved, butgave non-repeatable results after a few tunnel runs.Examining the probes, it was found that debris asso-ciated with clay used to fill model attachment holes,thin-film remaining on model due to machining, oil,and occasionally fine particles left in the tunnel wasfilling and plugging the smaller probe ports. Replac-ing the tubing with 1/16-in ID tubing and providingseparate transducers for each of the four probes, pro-vided very a sound and repeatable response.

Typical pressure measurements for the chargetube total, static, and plenum pressure systems areshown in Figures 7, 8, and 9. Figure 7 shows theoriginal system and is used as a benchmark to judgethe improvements. Figure 8 shows the total, static,and plenum pressures for the new system. The re-sponse of the static and total pressures are compara-ble to the previous configuration. These two systems

Pressure Plot

Pressure Plot

200 300

Time (milliseconds)

100 200 300

Time (milliseconds)

Fig. 8. Pressure v. Time contours using a modified pres-sure system consisting of only the four new probes.

Pressure Plot

Fig. 7. Pressure v. Time contours as determined us-ing the original pressure system.

0 100 200 300 400

Time (milliseconds)

Fig. 9. Pressure v. Time contours using the new pressureprobe/measurement system.

were only changed by shortening the tubes. The internalvolume of the new system was not significantly reducedand only a small improvement, if any, was expected. Thesteady flow plateau in the data was about 120 millisecondsduration.

The original plenum pressure measuring systemyielded about 30 msec of steady-state pressure. The newplenum pressure measuring system volume was reducedconsiderably with a corresponding increase in measuredsteady-state pressure of nearly 110 msec as shown in Fig-ure 8. Figure 9 shows pressure traces from the systemafter the four transducers were installed and larger tubingused to connect the transducer to the probe. This graph is

American Institute of Aeronautics and Astronautics

at a higher charge pressure. Due to this higher pres-sure the charge tube total and static pressures havetaken longer to respond and the constant pressureplateau has been reduced from 120 msec to about 90msec.

The plenum pressure is of comparable duration.All four pressure measured nearly identical readingsthroughout the run. This supports the theory of freeflowing the plenum cavity.

CONCLUSIONS

The transonic wind tunnel facility at The Univer-sity of Texas at Arlington has been returned to serv-ice. The facility pressure monitoring systems aremuch more responsive after several improvements.The charge tube total and static pressure systems re-sponse is comparable to their previous response. Theplenum pressure system responds comparable to thecharge tube system.

The modified system allows a measured regionof constant pressure flow of about 120 msec for aninitial charge tube pressure of 150 psia. The constantpressure region is reduced to about 90 msec for aninitial charge tube pressure of 225 psia. The transonicfacility has been checked out and calibrated with acenterline probe for Mach numbers near 0.8 and aReynolds number of 14xl07 per meter (3.5xl06 perinch).

The current systems all measure absolute or gagepressures with respect to the atmosphere and Machnumbers are calculated relative to these pressures.Measuring the total pressure and then measuring thedynamic pressure with a differential pressure trans-

ducer would provide a more accurate means of determin-ing Mach number. Mounting a small static probe or apilot-static probe in the test section would provide themost accurate measurement of flow quantities. Plans areunderway to include a small pitot-static probe with a dif-ferential pressure transducer.

REFERENCES

1 Ludwieg, H. "Tube Wind Tunnel - a Special Type ofSlowdown Tunnel," Paper presented at the Eleventhmeeting of the AGARD Wind Tunnel and ModelTesting Panel, Scheveningen, Holland, July 1957.

2 Starr, R. F. and Schueler, C. J., "Experimental Stud-ies of a Ludwieg Tube High Reynolds Number Tran-sonic Tunnel," AEDC-TR-73-168, December 1973.

3 Wilson, D. R. and Chou, S. Y., "Development of theUTA High Reynolds Number Transonic Wind Tun-nel," AIAA Paper 85-031 5, AIAA 23rd AerospaceSciences Meeting, Reno, Nevada, January 14-17,1985.

4 Wilson, D. R., "Development of the University ofTexas at Arlington Aerodynamics Research Center,:AIAA Paper 88-2002, AIAA 15th AerodynamicTesting Conference, San Diego, CA, May 18-20,1988.

5 Engelhardt, J. P., "A Transonic Wind Tunnel Elec-tronic Control Unit and Total Pressure Survey of aVortex," M. S. Thesis, The University of Texas atArlington, December 1989.

American Institute of Aeronautics and Astronautics