Development of a high-pressure detonation-driven shock tube … · 2017-01-19 · Description of the Existing Shock Tunnel A schematic of the existing shock tunnel facility is shown

Copyright ©1996, American Institute of Aeronautics and Astronautics, Inc.

AIAA Meeting Papers on Disc, January 1996A9618806, AIAA Paper 96-0853

Development of a high-pressure detonation-driven shock tube facility

Donald R. WilsonTexas Univ., Arlington

Frank K. LuTexas Univ., Arlington

W. S. StuessyTexas Univ., Arlington

Karl R. BurgeTexas Univ., Arlington

AIAA 34th Aerospace Sciences Meeting and Exhibit, Reno, NV Jan 15-18, 1996

The Hypersonic Shock Tunnel at The University of Texas at Arlington is being converted from a pressure-driven to adetonation-driven shock tunnel. This modification was necessitated by a requirement to provide a highpressure/temperature test environment for research to support the development of MHD-augmented hypersonic testfacility concepts. A description of the existing shock tunnel, review of the detonation-driven shock tunnel concept,predicted performance of the detonation-driven shock tunnel, and results from an experimental simulation of the tunneloperation, are provided. (Author)

Page 1

DEVELOPMENT OF A HIGH-PRESSUREDETONATION-DRIVEN SHOCK TUBE FACILITY

Donald R. Wilson*, Frank K. Lu1", W. Scott Stuessy* and Karl R. Surge5

NASA/UTA Center for Hypersonic ResearchThe University of Texas at Arlington

Arlington, TX 76019

Abstract

The Hypersonic Shock Tunnel at The Universityof Texas at Arlington (UTA) is being convertedfrom a pressure-driven to a detonation-drivenshock tunnel. This modification wasnecessitated by a requirement to provide a high-pressure, high-temperature test environment forresearch to support the development of MHD-augmented, hypersonic test facility concepts. Adescription of the existing shock tunnel, reviewof the detonation-driven shock tunnel concept,predicted performance of the detonation-drivenshock tunnel, and results from an experimentalsimulation of the tunnel operation are providedin this paper.

Introduction

The NASA/UTA Center for HypersonicResearch is supporting MSE, Inc. in a NASA-sponsored program to develop MHD-augmentedhypersonic test facility concepts. A principalgoal of this effort is to investigate concepts thatcould provide the basis for development of acontinuous-flow hypersonic wind tunneloptimized for testing advanced air-breathinghypersonic propulsion systems.1 In particular, aneed exists for a facility capable of providingpost bow shock conditions for testing advancedconcepts such as the Pre-Mixed, Shock-InducedCombustor (PM/SIC) Engine.2 In order tosimulate this test environment in an MHD-augmented test facility, preliminary designstudies indicate that accelerator channel staticpressures on the order of 100 atm may berequired.3

'Professor of Aerospace Engineering, AssociateFellowAssociate Professor of Aerospace Engineering,Associate Fellow

^Faculty Research Associate; Member'Graduate Student, Student Member

Copyright ® 1996 by the American Institute of Aeronautics andAstronautics, Inc. All rights reserved.

Unfortunately, the previous experience base forMHD accelerator channel operation was atpressure levels on the order of 1 to 10 atm.4'5Development of MHD accelerators capable ofoperating at high pressure levels will requireimproved understanding of a variety of technicalissues, such as the effect of high pressure onthe electrical conductivity and Hall parameter ofboth equilibrium and non-equilibrium plasmas,the structure and stability of the currentdischarge, and the plasma electrical breakdowncharacteristics. Furthermore, the developmentof novel ionization concepts such as the use ofhigh-energy microwave or electron beams mayprovide an attractive alternative to the use ofalkali metal seeding for attaining the requisiteelectrical conductivity.6'7 Prior experience withthese schemes is also at relatively low pressure,and their feasibility at high pressure will requireexperimental validation.

In order to provide an experimental capability toaddress some of the issues related to high-pressure operation of MHD accelerators, UTAhas proposed to convert its existing pressure-driven hypersonic shock tunner9 into adetonation-driven shock tunnel. Other conceptsfor enhancing the performance of the existingfacility were briefly considered, including the useof an electrical10' or combustion- heated12 lightgas driver, and a free piston driver.13 Althoughthe free piston driver probably has the highestperformance capability, Bakos and Erdos14 haveconcluded that the detonation driver may offercomparable performance with reduced capitolinvestment. Furthermore, an experience basehas been generated at UTA to support thisapproach via an ongoing research program todevelop Pulse Detonation Engine (PDE)concepts.15'16 Much of the technologydeveloped as part of this program would bedirectly applicable to the detonation-drivenshock tunnel.

The results of a design study to explore thefeasibility of converting the UTA shock tunnel to

1American Institute of Aeronautics and Astronautics

a detonation-driven facility are presented in thispaper. The basic simulation requirementsneeded to support the MHD acceleratordevelopment program were:

PressureTemperatureMach number

0.01 -100atm2200 - 4000 K«2.0

These values were felt to be representative ofaccelerator channel entrance flow conditions foran MHD-augmented facility capable of providingthe test section environment needed for testingof advanced hypersonic engine concepts.Experimental simulations of several detonation-driven shock tunnel modes of operation werealso conducted by configuring the existing PDEresearch combustor as a shock tube. Theresults of these tests, together with thenumerical simulations conducted as part of thedesign study led to the selection of the designconfiguration for the UTA facility. Predictedfacility performance maps for simulation ofaccelerator entrance flow conditions arepresented.

Description of the Existing Shock Tunnel

A schematic of the existing shock tunnel facilityis shown in Fig. 1. The shock tube consists of a15.24 cm (6 in) diameter, 3 m long driver tube,separated from a 15.24 cm diameter, 8.23 m(27 ft) long driven tube by a double-diaphragmsection. Both driver and driven tubes are ratedfor a pressure of 41.3 MPa (6,000 psi).Diaphragms are normally 10 gauge (3.42 mm)hot-rolled ASTM A36 steel plates, scored tovarious depths in a cross potent pattern. Thenozzle, test section and diffuser were donated toUTA by LTV Aerospace and Defense Co. (nowLoral Vought) of Dallas, Texas. A 7.5° half-angle conical nozzle with interchangeable throatinserts enables Mach numbers of 5 through 16to be obtained. Presently, the tunnel isconfigured for Mach 8. The exit diameter of thenozzle is 33.6 cm (13.25 in). A secondarydiaphragm made of 0.0127 mm (0.005 in) thickaluminum sheet is located in the nozzle throatregion, and is used to separate the driven-tubegas from that in the test section. The testsection is of semi-free jet design, 53.6 cm (21.1in) long and 44.0 cm (17.5 in) in diameter. Two23.0 cm (9 in) diameter optical windows arelocated on opposite sides of the test section.

Instrumentation and model mounting ports arelocated in the floor of the test section, and in thefloor and roof of the diffuser. A 30.5 cm (12 in)diameter, 213 cm (84 in) long diffuser connectsthe test section to a 4.25 m3 (150 ft3) vacuumtank.

The pneumatic system consists of a Haskellmodel 55696 two-stage gas-driven boosterpump capable of charging the driver tube to41.3 MPa (6000 psi). The Haskell pump isnormally connected to the facility aircompressor system, consisting of a Clark CMB-6 5-stage air compressor, twin-tower desiccantdrier, and 16.6 MPa (2400 psi) storage bottles.Alternatively, the Haskell pump can be fed froma manifold of 15.2 MPa (2200 psi) heliumstorage bottles. The vacuum system consists ofa Sargent-Welch model 1376 (300 l/min) pumpused to evacuate the driven tube, a Sargent-Welch model 1396 (2800 l/min) pump used toevacuate the test section/diffuser/vacuum tank,and a vacuum pressure measurement systemconsisting of two Baratron type 127A pressuretransducers and the associated valve system toenable full range coverage from 1000 to 0.001Torr.

Standard shock tube instrumentation, consistingof a variety of Kulite and PCS pressuretransducers, Medtherm thin film heat fluxsensors, and Medtherm fast-responsethermocouples is available. The data iscollected by a LeCroy 40 channel dataacquisition system. Eight high-speed channelsare provided by LeCroy model 6810 waveformrecorders (5 MHz), and the remaining low-speed32 channels by a LeCroy model 8212A datalogger (40 kHz) The data is recorded into on-board memory within the CAMAC (IEEE-583)crate, and transferred via a GPIB-488 bus to a486 PC computer for post-run processing.

Detonation-Driven Shock Tunnel Concept

The detonation-driven shock tube was firstproposed by Bird17 in 1957, and has since beenstudied by several investigators.14'18"23 Adetonation process is typically established in adriver tube filled with a near-stoichiometricmixture of hydrogen and oxygen, although othergas combinations are possible. The initialpressure level prior to detonation can be quitelow, thus eliminating the need for thick metal

American Institute of Aeronautics and Astronautics

diaphragms. The detonation process producesa relatively low molecular weight driver gas athigh temperature and pressure levels. Thesudden pressure rise produced by thedetonation wave causes the primary diaphragmto rupture, thus establishing a shock wave in thedriven tube filled with air. Two modes ofoperation are possible. In the "upstreampropagation" mode (Fig. 2), the ignition sourceis placed just upstream of the primarydiaphragm, producing a detonation wave thatpropagates from right to left through the drivertube. The pressure rise following the detonationwave ruptures the primary diaphragm toestablish the flow in the driven tube. Theeffective driver tube conditions for this modeare the pressure and temperature at state 4' onthe wave diagram in Fig. 2. In the "downstreampropagation" mode (Fig. 3), the ignition sourceis located at the upstream end of the drivertube, producing a detonation wave that travelsfrom left to right through the driver tube,rupturing the primary diaphragm on impact.The effective driver tube conditions for thismode are those for state 4" on the wavediagram shown in Fig. 3. For either mode,further performance enhancement is possible byhelium dilution of the hydrogen/oxygen drivertube mixture. The helium dilution raises thesonic speed in the driver gas, and alsosomewhat reduces the danger associated withpremature detonation of the hydrogen/oxygenmixture. Performance calculations by Yu, etal.21 indicate that the performance degradationcaused by the slight lowering of the detonationtemperature is more than adequately offset bythe increased sonic speed of the driver-tubegas.

Facility Design/Performance Analysis

Design and performance analysis calculationsfor the detonation-driven shock tunnel weremade with the aid of the TEP24 computer code,a Windows™ version of the NASA CEC762*code This code provides real-gas calculationsof a variety of basic gasdynamic processes,such as Chapman-Jouguet detonation waves,shock tube and isentropic nozzle performance.A quasi-one-dimensional flow model isassumed, and real gas calculations based onboth equilibrium and frozen flow models areavailable. All of the calculations presented inthis paper are based on the equilibrium flow

assumption. The basic geometric configurationshown in Fig. 1 was retained, along with the6000 psi pressure limitation for both the driverand driven tube.

The TEP™ code was first used to calculatedetonation tube performance for stoichiometricmixtures of hydrogen and oxygen, for a range ofinitial pressures and varying levels of heliumdilution. The TEP™ code was also used forshock tube performance calculations, howeverthe code does not include the unsteadyexpansion between driver and driven tube. Acode developed at UTA that calculated thedriven-tube pressure ratio, p2/pi, and shockspeed as a function of the effective shock tubepressure ratio, pVpi, was used for this purpose.This is a perfect gas code, but priorcomparisons with real gas codes indicate thatthe driven tube incident wave pressure ratio andwave speed is generally within 5 percent. TheTEP™ code was then used to calculate thetemperature ratio across the expansion wave,the properties following the incident andreflected waves in the driven tube, and thesubsequent isentropic nozzle expansion.

A series of preliminary calculations quicklyindicated that operation of the detonation driverin the "upstream" mode (Fig. 2) precluded theattainment of the desired 100 atm staticpressure level in the subsequent nozzleexpansion, without exceeding the 6000 psipressure limit of the driver tube. For this mode,the maximum pressure occurs when thedetonation wave reflects from the end wall ofthe driver, and pressure levels of the order of9000 psi were predicted for operating conditionsthat would produce driven tube p5 pressurelevels high enough to obtain a static pressurelevel of 100 atm when expanded to Mach 2.Thus, the operating mode employingdownstream propagation of the detonation wave(Fig. 3) was selected as the basis for the facilityperformance calculations.

The predicted performance map for the"downstream propagation" mode correspondingto Mach 2 nozzle flow is shown in Fig. 4. Astoichiometric mixture of hydrogen and oxygenin the driver (no helium dilution) was assumed.The corresponding performance envelopesusing air and helium as driver tube gases arealso shown for comparison. The performanceenhancement provided by the detonation driver


is considerable, and although the specifiedstatic temperature levels were easily achieved,a maximum pressure of only 50 atm waspredicted. However, by either reducing thenozzle expansion to a Mach number of 1.6, orby adding helium dilution on the order of 50percent, the target pressure level of 100 atmcould be reached.

Simulation Experiments

UTA has been actively involved in an on-goingPulse Detonation Engine (PDE) researchprogram.15'16 The PDE research combustordeveloped as part of this program wasreconfigured as a detonation-driven shock tubein order to provide experimental validation ofthe proposed concept.

PDE Test Facility

A schematic diagram of the PDE researchfacility is shown in Fig. 5. Principal componentsinclude the research combustion chamber (Fig.6), fuel/oxidizer system, fuel/air system, ignitionsystem, purge air system, vacuum system anddata acquisition/control system. A completedescription of the facility as provided in Refs.15,16.

The combustion chamber is constructed from7.62 cm (3 in) diameter steel tube sections ofvarying lengths. Each section has provisions formounting pressure transducers, heat fluxsensors, and thermocouples at 7.62 cmintervals along the principal axis of the tube.One section of 7.62 cm length contains the arcigniter plug. This section may be placed ateither end of the tube, or at intermediatelocations. The fuel and oxidizer, as well as thepurge air, are injected into the chamber throughan end flange that also contains a pressuretransducer. Mylar diaphragms of 0.254 - 0.381mm (0.01 - 0.015 in) thickness are used to sealthe open end of the tube. These diaphragmsrupture upon impact of the detonation wave,and the combustion products are exhaustedinto the facility exhaust system. The vacuumsystem is used to pump the sealed chamber to apressure on the order of 690 Pa (0.1 psia) andthen the fuel and oxidizer are injected into thechamber from standard high pressure storagebottles through a pressure regulation system.Matheson series 6103 flash arresters areinstalled in the lines to prevent flashback into

the fuel and oxidizer tanks in the event of apremature ignition during the filling operation.The pressure in the chamber is monitoredduring the filling operation by a precise Baratronmodel 127A vacuum pressure transducer, andthe method of partial pressures is used to setthe mixture ratio.

The fuel/oxidizer mixture is ignited by a highenergy electric arc. A Miller snap-start highfrequency arc welder power supply capable ofproviding open circuit voltages up to 30 kV isused to preionize the gap between two flush-mounted electrodes. Once breakdown occurs inthe gap between the two electrodes, a capacitorbank containing two 11,000 u.F capacitorsconnected in series to provide a maximum loadvoltage of 150 volt discharges a high current arcacross the gap. The estimated energy deliveredin the arc is approximately 3-5 Joule.

A DSP Technology data acquisition/controlsystem is used to collect data during a testfiring. This system provides 48 data channels,each containing an independent amplifier, 12bit, 100 kHz analog-to-digital converter, and 20kB memory unit. All channels are sampledsimultaneously. Transient chamber pressureswere measured with PCB model 111a24dynamic pressure transducers, rated at 6.89MPa (1000 psi), and having a response time of1 usec. The data acquisition system wasconnected to a 486-DX 33 MHz IBM-compatiblePC via a GPIB 488 bus.

Experimental Results - PDE Combustor

Prior to presenting the results of the shock tubesimulation experiments, a brief review ofselected results from the basic PDEexperimental program15'16 will be given. Inparticular, phenomena relevant to theestablishment of detonation waves in the driverwill be reviewed. In the PDE program,detonations with near stoichiometric mixtures ofhydrogen, propane, or methane and oxygenhave been achieved. A typical pressure tracefrom a hydrogen-oxygen run at an initialpressure of 2 atm is shown in Fig, 7. Thepressure traces are from pressure transducerslocated at 7.62 cm (3 in) intervals, with the firstsensor located 7.62 cm from the igniter. Theaverage wave propagation speed betweenadjacent pressure transducers was calculatedfrom


AxAr (D

and average wave speeds for different initialpressures are compared in Fig. 8 for thehydrogen-oxygen runs. In addition, Fig. 8includes calculations of wave speed based onthe measured pressure ratio of the propagatingwave, however this approach does not provideconsistent results. Also shown in Fig. 8 are thesonic speed and the theoretical Chapman-Jouguet detonation wave speed. These valueswere calculated from the TEP code. Theobserved wave pattern is clearly pressuredependent. At 0.5 atm, the wave appears to bea deflagration (subsonic) wave, whereas at 1atm a weak (low supersonic) detonation isobserved. At 2 atm, a weak detonation wave isformed that appears to transition to a Chapman-Jouguet detonation wave between the third andfourth pressure transducer. The transitionlocation is also dependent upon fuel/oxidizercombination, mixture ratio, and turbulenceenhancement. Surprisingly, the ignitionenergy level was not found to be a significantfactor in either the location or occurrence of thetransition. This may be due to the fact thatsufficient energy was available from the ignitionsystem to produce the weak detonation wave forall of the levels tested. Further tests areplanned to explore the effect of ignition energylevel, and in particular, to determine if anenergy threshold for direct initiation of a CJdetonation wave exists that is within thecapability of the current system. The ability todirectly initiate a CJ detonation will clearly havea significant influence on the "upstreampropagation" mode of operation.

The classical Zeldovich, von Neumann andDoring (ZND) model28 for a detonation wave isillustrated in Fig 9. Their model assumes apropagating normal shock wave thatcompresses the gas from its initial state to ahigh pressure state referred to as the vonNeumann spike. The pressure ratio across thiswave is given (for a perfect gas) from classicalshock wave theory,27

A(2)

in the combustible mixture after a short ignitiondelay, and the state gradually approaches anequilibrium state predicted by Chapman-Jouguet theory

Pi + 72(3)

The corresponding temperature rise associatedwith the shock wave initiates chemical reactions

The Chapman-Jouguet wave propagation speedfor a stoichiometric hydrogen/oxygen mixture at2 atm is 2879 m/sec, corresponding to a shockMach number of 5.35. Substituting this valueinto Eqs. (2) and (3) gives Ps/pt = 33.23 andp2/pi = 19.8. The experimentally observedpressure ratios are on the order of 12.6, which iscloser to but lower than the CJ pressure ratio of19.8. The duration of the von Neumann spike isonly expected to be on the order of severalmicroseconds,28 thus the 100 kHz samplingfrequency of the DSP data acquisition systemmay miss the peak pressure level. In fact,subsequent tests in which the pressure traceswere recorded on a 300 MHz digital oscilloscopeverified that this often occurs. The reason forthe discrepancy between the observed pressureratio and the CJ pressure ratio is not obvious,but is consistently observed in all of our results.

Shock Tube Configuration

The PDE research chamber was reconfiguredas a detonation-driven shock tube by inserting aMylar diaphragm at a flange interface toseparate the driver tube section containing theappropriate fuel/oxidizer/diluent mixture fromthe driven tube containing air. The length of thedriver tube was 53.3 cm (18 in), and the driventube was 30.5 cm (12 in). The driver tubecontained two transducers, located 15.24 cm (6in) apart, whereas the driven tube contained 4transducers located at 7.62 cm (3 in) intervals.The first transducer in the driven tube was 3.81cm (1.5 in) downstream of the diaphragm.Tests were run with the igniter located at bothupstream and downstream ends of the drivertube.

Experimental Results - Simulated ShockTube

For the simulated detonation-driven shock tubeexperiments, the driver tube was filled withnear-stoichiometric mixtures of hydrogen andoxygen. The initial pressure in the driver tube


was limited to 2 atm to prevent over-pressurization of the pressure transducersresulting from reflection of the detonation wavefrom the end wall or primary diaphragm. A0.381 cm (.015 in) thick Mylar diaphragmseparated the driver and driven tubes. Thedriven tube was open to atmosphere for theshock tube simulation experiments. Both theupstream and downstream propagation modeswere simulated. Also, the effect of turbulenceenhancement in the driver tube and heliumdilution of the hydrogen/oxygen mixture wereinvestigated.

Transient pressure traces at selected axialstations for the "upstream propagation" modeare shown in Fig. 10, and the plot of wavepropagation speed vs. distance is shown in Fig.11. The two pressure traces to the left of Fig.10 are from transducers located in the drivertube, whereas the other four traces are fromtransducers in the driven tube. Although a largepressure is generated by the upstreampropagating detonation wave, the diaphragmwas not ruptured until the return of the reflecteddetonation wave to the diaphragm. The velocityplot shows that transition to a CJ detonationwave occurred in the vicinity of 27 cm from theignition source, which is comparable tomeasurements of detonation wave formationdistance from the basic PDE experiments. TheCJ detonation wave was clearly not establishedprior to passage of the wave past the firsttransducer upstream of the diaphragm, and wespeculate that the gradual rise in pressureassociated with the detonation wave formationwas insufficient to cause diaphragm rupture. Athinner diaphragm is clearly needed in this case.Diaphragm rupture did occur upon impact of thereflected detonation wave, generating a strongshock wave (Ms » 3.5) in the driven tube. Notethe abrupt rise in pressure upon passage of theshock wave, compared to the gradual rise inpressure associated with passage of thedetonation wave. This "precompression"phenomena appears to be a characteristic ofweak detonation waves, and generallydisappears when a full CJ detonation wave isformed.15 Relatively good agreement isobserved between the measured pressure ratioand the value predicted from Eq. (2) using themeasured wave speed in the forward part of thedriven tube. The measured pressure ratiosindicate an attenuation of wave speed that is notobserved from the time-of-flight measurements.

The tests with turbulence enhancement via aShchelkin spiral29 produced rather surprisingresults. Turbulence enhancement actuallydelayed the formation of a CJ detonation wavein the driver tube, with a maximum wavepropagation speed of only about 2000 m/sec,compared to the nearly 3000 m/sec achieved forthe CJ detonation wave. The maximum drivertube pressure levels were actually higherthough, and little difference was seen in thestrength of the shock wave generated in thedriven tube.

A summary of measured driver tube pressurelevels, driver and driven tube wave propagationspeeds is shown in Fig. 12 for the "upstreampropagation" mode. The shock speed in thedriven tube appears to correlate better with themaximum pressure level attained in the drivertube rather than the detonation wave speed.Helium dilution did not exhibit the anticipatedperformance enhancement. In fact, shockvelocities were actually lower. Furthermore,examination of the pressure traces for thehelium dilution runs indicated a rather irregularwave pattern in both the driver and driventubes. It appears that either inadequate mixingof the gases prior to ignition of the mixture or alowering of the partial pressure of thehydrogen/oxygen mixture due to helium dilutionmay be a principal factor. Several investigatorshave emphasized the importance of achievinggood mixing to obtain strong detonation wavefronts.23'30

Similar results were achieved for "downstreampropagation" of the detonation wave, althoughlower detonation wave speeds and somewhatlower driven-tube shock speeds were observedfor this mode. In general, CJ detonation wavesdid not form in the driver tube, probably due tothe shorter distance traveled by the wave. Thisis not expected to be a problem for the full-scaledriver tube, since the total length isapproximately a factor of 10 greater than thedistances required for transition to a CJdetonation wave in the basic PDE experiments.The pressure rise resulting from the impact ofthe downstream-propagating detonation waveshould produce immediate rupture of thediaphragm, thus establishing the necessaryconditions for strong shock formation in thedriven tube.


Concluding Remarks

Results of a design study to upgrade theperformance of the UTA Hypersonic ShockTunnel by converting from a pressure-driven todetonation-driven mode of operation have beenpresented. Experimental simulations of thedetonation-driven shock tube were conducted byreconfiguring an existing pulse detonationengine research chamber. Both upstream anddownstream propagation of the detonationwaves were investigated. The "downstreampropagation" mode was selected as the basis ofoperation of the full scale facility. This was duein part to the necessity of achieving a highpressure level for proposed MHD acceleratorchannel experiments without exceeding the6000 psi pressure limit of the existing tunnel.Furthermore, the gradual rise in pressure behindthe upstream-propagating detonation waveduring its transition from a weak to a CJ wavemay cause unsteady wave formation in thedriven tube due to changing pg pressure levels.Hardware requirements for the facilitymodification have been identified, and costestimates for necessary modifications havebeen prepared. We anticipate that the tunnelmodifications will require approximately threemonths to implement. The resulting tunnelmodifications will provide the requiredsimulation capability needed to support researchon MHD-augmented hypersonic test facilityconcepts. Furthermore, implementation of thedriver modifications should greatly enhance thehypersonic test capability at UTA.

Acknowledgments

The authors would like to acknowledge thecontributions and support provided by Dr. Y.M.Lee and Mr. Gloyd Simmons of MSE, Inc.during the course of this study, and to expresstheir thanks to Mr. Dennis Bushnell of NASA-Langley for the original suggestion to investigatethe detonation-driven shock tunnel concept.Also, valuable discussions with Dr. John Erdosof GASL are gratefully acknowledged. Thestudy was funded in part by MSE, Inc. throughContract No. 95-C294-F (Dr. Y.M. Lee, technicalcontract monitor). Additional support wasprovided by the State of Texas AdvancedTechnology Program through Grant No. 003656-056 (Dr. D. Johnston, program monitor).

References

1. "Proceedings of the MARIAH Workshop,"MSE, Inc., Fairmont Hot Springs, Montana,November 1995.

2. Chinitz, W., Bakos, R.J. and Erdos, J.I.,"Experimental Requirements for the Studyof Shock-Induced Premixed Combustion,"AIAA Paper No. 94-3099, 1994.

3. "Proceedings of the Magnetohydrodynamics(MHD) Technical Review and PlanningWorkshop," NASA, Washington, D.C.,February 1995.

4. Rittenhouse, L.E., Pigott, J.C., Whoric, J.M.and Wilson, D.R., "Theoretical andExperimental Results With a LinearMagnetohydrodynamic (MHD) AcceleratorOperated in the Hall Current NeutralizedMode,"AEDC-TR-67-150, 1967.

5. Pate, S.R., Siler, L.G., Stallings, D.W. andWagner, D.A., "Development of an MHD-Augmented, High Enthalpy, Shock TunnelFacility," AIAA Journal, Vol. 12, No. 3,March, 1974, pp. 289-297.

6. Simmons, G.A., Nelson, G.L., Hiers, R.S.and Western, A.B., "An Unseeded Air MHDAccelerator Concept for High Mach NumberHypersonic Propulsion Testing," AIAA Paper89-2535, 1989.

7. Simmons, G.A., Nelson, G.L and Lee,Y.M., "Analysis of Unseeded NonequilibriumMHD Accelerator Concept for HypersonicPropulsion Ground Testing Applications,"AIAA Paper 92-3994, 1992.

8. Stuessy, W.S., Murtugudde, R.G., Lu, F.K.and Wilson, D.R., "Development of the UTAHypersonic Shock Tunnel," AIAA Paper 90-0080, 1990.

9. Lu, F.K., "Initial Operation of the UTA ShockTunnel," AIAA Paper 92-0331,1992.

10. "Summary of Capabilities of HypersonicShock Tunnels at Calspan ATC," CalspanAdvanced Technology Center, Buffalo, NY.


11. "Description of the Aachen Shock TunnelTH2," Shock Wave Laboratory, TechnicalUniversity of Aachen, Aachen, Germany,January 1991.

12. Nagamatsu, H., Sheer, R., Osberg, L andGary, K., "Design Features of the GeneralElectric Research Laboratory HypersonicShock Tunnel," GE Research Lab ReportNo. 61-RL-2711C, May 1961.

13. Eitelberg, G., Mclntyre, T.J., Beck, W.H.and Lacey, J., "The High Enthalpy ShockTunnel at Gottingen," AlAA Paper 92-3942,1992.

14. Bakos, R.J. and Erdos, J.I., "Options forEnhancement of the Performance of Shock-Expansion Tubes and Tunnels," AlAA Paper95-0799, 1995.

15. Stanley, S.B, Stuessy, W.S. and Wilson,D.R., "Experimental Investigation of PulseDetonation Wave Phenomena," AlAA Paper95-2197, 1995.

16. Stanley, S., Surge, K. and Wilson, D.,"Experimental Investigation of PulseDetonation Wave Phenomena as Related toPropulsion Application," AlAA Paper 95-2580,1995.

17. Bird, G.A., "A Note on Combustion DrivenShock Tubes," In "Hypersonic Facilities inthe Aerodynamics Department, RoyalAircraft Establishment," P.A. Hufton (Ed.),AGARD Report 146, 1957.

18. Coates, P.B. and Gaydon, A.G., "A SimpleShock Tube With Detonating Driver," Proc.Royal Soc. (London), A283, 1965, pp.18-32.

19. Balcarzak, M.J. and Johnson, M.R., "TheGaseous-Detonation Driver and ItsApplication to Shock Tube SimulationTechniques," In Proc. 5th Intl. Shock TubeSymp., Z.I. Slawsky, J.F. Moulton and W.S.Filler (Eds.), U.S. Naval Ordinance Lab,White Oak, MD, 1966, pp. 1111-1128.

20. Lee, B.H.K., "Detonation Driven Shocks in aShock Tube," AlAA Journal, Vol. 5, No. 4,April 1967, pp. 791-792.

21. Yu, H.R ., Esser, B., Lenartz, M. andGronig, H., "Gaseous Detonation Driver fora Shock Tunnel," Shock Waves, Vol. 2,1992, pp. 245-254.

22. Tamagno, J., Calleja, J. and Erdos, J.,"Exploratory Tests of Detonation-DrivenExpansion Tube Performance," NASA CR-191580, General Applied ScienceLaboratories, Ronkonkoma, NY, March1994.

23. Engers, R.J., Calleja, J.F. and Bakos, R.J.,"A Detonation Driven Injectant Heater forPulse Facility Testing Applications," AlAAPaper 95-3153, 1995.

24. "Thermal Equilibrium Program for WindowsUser's Manual," Software and EngineeringAssociates, Inc., 1992.

25. Gordon, S. and McBride, B.J., "ComputerProgram for Calculation of ComplexChemical Equilibrium Compositions, RocketPerformance, Incident and ReflectedShocks, and Chapman-JouguetDetonations," NASA SP 273, 1971. .

26. Angelone, J.P., "Shock Tube - HighTemperature Gasdynamic Studies," M.S.Thesis, The University of Texas atArlington, 1978.

27. Anderson, J.D., Jr., Modem CompressibleFlow, 2nd ed., McGraw-Hill Publishing Co.,New York, 1990.

28. Bussing, T. and Pappas, G., "AnIntroduction to Pulse Detonation Engines,"AlAA Paper 94-0263,1994.

29. Shchelkin, K.I., Soviet Journal of TechnicalPhysics, Vol. 10, 1940, pp. 823-827

30. Ting, J.M., Bussing, T.R.A. and Hinkey,J.B., "Experimental Characterization of theDetonation Properties of Hydrocarbon Fuelsfor the Development of a Pulse DetonationEngine," AlAA Paper 95-3154, 1995.


t ,1 '

DRIVER TUBE(15 cm dia)

3m 8.2mr- DRIVEN TUBE (15 cm dia)

-jrn ——— \ ——— m ——————— m ——————— r~EUi i

t^—— ———— ——n ——— \U —— t

tr^ ^~"~n 1

THRUST STAI^D TH

—— L

C"T

i rVACUUM

61 i2.1 m

^NOZZLE

iN —QTJPTinM-

n

n

nn

— L i —— i

\TANK1.8 A

t \

\\

LDIFFUSER

Fig. 1 Elevation view - hypersonic shock tunnel

Fig. 2 Wave diagram for a detonation-drivenshock tube with upstream propagation ofthe detonation wave

Fig. 3 Wave diagram for a detonation-drivenshock tube with downstream propagation ofthe detonation wave

A4/AI = I M = 2.0

7000

6000 -

5000 -

4000 -

3000 -

2000 -

1000 -

O Detonation DriverQ Air Drivera Helium Driver

10 20 30 40

Pressure (atm)

50

Fig. 4 Performance map for the detonationdriven shock tunnel, H2/O2=2.0, Mach 2nozzle


ft

Oelonation Chambe

Vacuum Pump

-&r<J- ft *w

Fuel Oxygen

J

-A -A •,

(12 ln|

•A >A

, 2 1 . 7 6 5 cm ( 9 . 7 5 in l Dl«

-19.685 cm (7.75 t n l' Ola -BC

- H . 6 0 5 cm ( 5 . 7 5 Inl

- 7 . 6 2 cm ( 3 in) Di»

' 9-! .-505 cm ( .75 in) DIAEQUALLY SPACED

„ PHESSURE PORt

- prr POTT

Section A-A

Fig. 5 Schematic diagram of PDE test facility Fig. 6 PDE research combustion chamber

3000.0 -

2 5 0 0 . 0 -

TJCu.«

„; 2000.0 -

M3n 1500.0 -<uMcu

1000.0 -

500 .0 -

0.0 -0.

_.— — — ~

00 0.

———————— (Ii1

jI

- ~-—"~

05 0.

l\' \

\\

10 0.

i \

15 0.Time, t,

v/_^''

20 0.millisec

'> l\l\ l\! \ i *

1 \ 'i \ i1 \1 T-\

i — lf~~* ,i . '

j25 0.

/

\ /iv"vVi /w4"

ft30 0.

y^\V \

c\;

35 0.

Distance From the Ignition Source In the Direction of the Open End— — 7 .62 cm-. (3 in) •-•-• 15.24 cm (6 in) ——— 22.86 cm (9 in)— — 30.48 cm (12 in — - 38.1 cm (15 in) ——— 4 5 . 7 2 cm (18 in)

- 500 .0

- 4 0 0 . 0

•Htoa

- 300.0cu

<DM3V)to

- 2 0 0 . 0 JJCL*

- 100.0

- 0 . 040

Fig. 7 PDE pressure plots for H2/O2=2.0 at an initial pressure of 2 atm


• .5 atm" 1 atm (Base)•*• 2 atm

O .5 atmO 1 atm (Base)A 2 atm— Sonic Velocity

— — C-J Vel. at .5 atm— —' C-J Vel. at 1 atm———• C-J Vel. at 2 atm

Solid Markers are Time of Flight VelocitiesOutlined Markers are Pressure Ratio Velocities

3000.0 - 10000.0

0.0 T————————=T"————————r7.62 15.24 22.86 30.48 38.10

Distance from Ignition Source, X (cm)45

0.072

Fig. 8 PDE velocity plot tor H2/O2-2.0 as a function of initial pressure

0)M3COto0)kPu

P*

P2

1P« /

^-/

Ei<pansian

yc0

•H

013

Com

b

Jlid2^4

| S

ho

e

P1

Distance

Fig. 9 ZND detonation wave model

/\

Fig. 10 Shock tube pressure plot for H2/O2=2.0 at an initial pressure of 2 atm

£J?

3000

IOOO

1000

CJ speed

a.n

Sonic speed

0 4 I It 18 20 74 Til 37 30 4O 44 48 fl fl

Distance from ignition source (in)

Fig. 11 Shock tube velocity plot for H2/O2=2.0at an initial pressure of 2 atm

DRIVER TUBE PRESSURE, psia

DRIVER TUBE WAVE SPEED, m/sec

B pmven TUBE WAVE SPEED, m/sec

BASELINE TURBULENCE 25* HELIUM 40% HELIUM SOU HELIUM

CONFIGURATION

Fig. 12 Summary of upstream propagationtest runs


Documents

Development of a high-pressure detonation-driven shock tube … · 2017-01-19 · Description of the Existing Shock Tunnel A schematic of the existing shock tunnel facility is shown