M. Valentino, C. W. Kauffman and M. Sichel- Experiments on Shock Induced Combustion of Isolated Regions of Hydrogen-Oxygen Mixtures

8/3/2019 M. Valentino, C. W. Kauffman and M. Sichel- Experiments on Shock Induced Combustion of Isolated Regions of Hydrogen-Oxygen Mixtures

1/16

(c)l999 American Institute of Aeronautics & Astronautics

A994 6669 F

AIAA-99-0821EXPERIMENTS ON SHOCK INDUCEDCOMBUSTION OF ISOLATED REGIONS OF

. HYDROGEN-OXYGEN MIXTURESM. ValentinoAir Force Research LaboratoryMunitions DirectorateEglin AFB, FL 32542

C. W. Kauffman, M. SichelDepartment of Aerospace EngineeringThe University of MichiganAnn Arbor, MI 48109


2/16


AIAA-99-0821

EXPERIMENTS ON SHOCK INDUCED COMBUSTION OF ISOLATEDREGIONS OF HYDROGEN-OXYGEN MIXTURESM. Valentino-Air Force Research Labora toryMunitions DirectorateEglin AFB, FL 32542

C. W. Kauffman, M. Siche!Department of Aerospace EngineeringThe University of MichiganAnn Arbor, MI 48 109

ABSTRACTThe interaction of a strong plane shockwave with isolated regions of gaseous mixtureswas examined through a series of shock tubeexperiments. Specifically, the non-uniformmixtures examined consisted of a spherical

bubble of pure hydrogen or hydrogen-oxygenmixtures surrounded by either an oxygen,nitrogen or air atmosphere. Shocks in therange of Mach 1.7 to 3.7 were studied. Theinteraction events where recorded with highspeed shadowgraphs and pressure tracerecordings. No chemical reactions wereobserved in the interactions of shocks withstrengths up to Mach 3.7 with a pure hydrogenbubble due to inadequate mixing o f thereactants. In addition no reaction wasobserved for shocks up to Mach 3.0 interactingwith a premixed stoichiometric bubble.However, shock induced combustion wasobserved for incident shock strengths above

INTRODUCTIONThe interactions of shock waves with fluid non-uniformities have long been studied. Typicallythese studies have looked at the interface of twofluids of different densities. Cases of both a shockmoving from a light gas into the heavy gas and vice

versa have been examined. Markstein [1,2],Richtmyer [3], Meshkov [4], Catherasoo &Sturtevant [5] and Brouillette & Sturtevant [6]examined the instabilities generated by thepassage of a shock wave a t planar interfacesbetween two different density gases. Abd-el-Fattah, et al. [7,8,9] also looked at shock wavesinteracting with interfaces between fluids ofdifferent densities but with an emphasis onexamining the process of shock wave refractionand reflection. Haas & Sturtevant [IO] extendedthis work by looking at shock interactions withcylindrical and spherical interfaces. Theirexperimental images clearly show how the passageof a shock wave over a light gas sphere deforms


3/16


4/16


one atmosphere. The test gas bubble wascreated in the following manner. The test gas(hydrogen or hydrogen-oxygen mixtures) wasconstrained within a soap bubble that wassupported on its bottom by a post located in themiddle of the test section. The soap bubbleswere made with the same Plateaus soapsolution ( 78% distilled water, 20% glycerin bymass 2% sodium oleate) as used by Haas andSturtevant [lo] in their experiments. Thecylindrical post is stainless steel tubing with a3.2 mm outer diameter and a 0.7 mm innerdiameter. A small piece of paper with a pinprickin the center is placed on top of the post tokeep the soap drop from running down theinterior before the gas bubble is made. Thebubbles were then made using a gas tightsyringe to inject the test gas from the bottom ofthe post. The test section was designed withsmall ports directly above and below the post toallow the depositing of the soap drop andinjection of the test gas. Each port had aremovable cap to provide a gas tight seal. Thetop ports opening size was 3.2 mm. Thebottom port was an extension of the supportpost that protruded from the bottom of the testsection.

To minimize the contamination of room airinto the shock tube, the driven section wasinitially filled with a slight overpressure of oneor two psi. This excess pressure was relievedwhen the top and bottom ports were openedjust prior to the bubble being created. All testgas bubbles created were on the order of 6 ccin volume and 19 mm in diameter. The portswere recapped as soon as the bubble was

A/AA-99-0821

proper partial pressure and finally filling the cylinderwith oxygen until the proper total pressure wasachieved. Just prior to opening the access ports onthe shock tube, the contents of the sample cylinderwere released into a balloon. The gas tight syringewas then inserted into the balloon and filled with thetest gas hydrogen/oxygen mixture. Releasing thesample cylinder gas into the balloon shortly beforecreating the bubble, ensured the hydrogen andoxygen were well mixed.

DISCUSSIONPure Hydrogen Bubble

The first series of tests looked at a shock waveinteracting with a hydrogen bubble surrounded byan oxygen atmosphere. The range of shock Machnumbers investigated ran from 1.7 to 3.7. Table 1gives a summary of these tests. This table includesthe measured shock wave velocity, post shockpressure, post shock temperature and an indicationof chemical reaction. It should be noted that theMach numbers have been adjusted slightlyupwards from those reported in a previous paper,Valentino et al. [16], discussing the preliminaryresults of these tests due to a refinement of theshock wave velocity data.

The Mach 1.7 and 2.2 tests were conducted toprovide a bridge between the high Mach tests ofthe present work and the low Mach tests (Mach1.05 to 1.25) of Haas and Sturtevant [lo]. Also.because of the relatively benign conditions behtndthe Mach 1.7 shock many of the fluid dynamrcfeatures of the interaction process are more easrlyseen than in the images for the stronger shock


5/16


stoichiometric hydrogen-oxygen mixtureindicates that the Mach 2.9 case would be onthe border of the third limit while the Mach 3.7case would be well into the mild ignition area.So the temperature and pressures are suchthat ignition would occur in a stoichiometrichydrogen oxygen mixture. But, in order to seea chemical reaction for the un-premixed case athand, the pressure and temperature mustremain at an elevated level long enough forsufficient mixing to take place along with therequired induction time of the HJO, reaction.

However there was no evidence of anychemical reaction when the shock interactedwith a hydrogen bubble surrounded by anoxygen atmosphere from either theshadowgraph images or the photo detectorsignals. Figure 4 shows the shadowgraphimages for the Mach 3.1 case (previouslylabeled Mach 2.96 in Valentino et al. [16]). Thebubble does seem to remain closely coupled tothe incident shock and thus should be subjectto the post shock temperatures and pressuresthroughout the 200 ps these tests wereobserved. Therefore one reason that nochemical reaction occurs it that there has notbeen sufficient mixing of the hydrogen andoxygen despite the roll up of the bubbles intovortex rings. Unfortunately, the shadowgraphimages of the present tests do not reveal anydetails to the interior structure of the vortices.

Premixed Hydrogen-Oxygen BubbleAlthough no chemical reactions were seenfor any clean runs with a pure hydrogen

AIAA-99-0821

bubble. For the cases of small bubbles formingunder the main bubble it is likely the small bubblesrupture and mix quickly after the shock passesallowing ignition to occur.

Encouraged by the results of these bad runs,a series of controlled runs of a shock waveinteracting with a premixed hydrogen-oxygenbubble were examined. The premixed bubble wassurrounded by an oxygen atmosphere for themajority o f the runs. However, several cases wererun to look at the effects of a nitrogen or an airatmosphere.Oxygen Atmosphere

Figure 5 shows the series of shadowgraphimages taken of the Mach 3.5 shock interactingwith a premixed bubble of hydrogen-oxygen at anequivalence ratio of 1.0. An oxygen atmospheresurrounds the premixed bubble. The images aretaken 4 p.s apart for a total of 68 ps. The interactionprocess appears the same as non-reacting runs upthrough 28 PLS. However, at 32 f~s a strong blastwave is seen just emerging from the top of thebubble. This wave continues to spread out fromthe bubble through the remaining frames and isclearly seen at 40 ps. The blast wave appears tosupport and accelerate the incident shock, whichshows signs of curvature by the image at 52 f~s.The curvature of the incident shock is more evidentat 64 and 68 f.~s as the blast wave has grown to fillthe test section from top to bottom. Smallerpressure waves can be seen between the bubbleand the blast wave in the images from 48 j.ts to 68f.~.s. At 64 f~s a reflected wave can be seen underthe bubble from the blast wave encountering the


6/16


Table 2 gives a summary of the resultsfrom all the cases with a premixed hydrogen-oxygen bubble surrounded by an oxygenatmosphere. Chemical reactions wereobserved for all the cases with incident shockwave strength of Mach 3.1 and above. Thepressure and temperature were 12 atm and815 K respectively for the lowest strengthshock, Mach 3.1, where chemical reaction wasobserved. Figure 6 shows these data pointsplotted on the explosion limit map of Lewis andvon Elbe [17]. This figure shows that theseconditions are in the mild ignition region to theright of the third limit. The Mach 3.0 casewhere no reaction was observed had a postshock pressure and temperature of 11 atm and780 K respectively. This condition is also in themild ignition area.

However, a one-dimensional analysis of aMach 3.0 shock interacting with the interfacebetween pure oxygen and a stoichiometricmixture gives the temperature and pressureinside the mixture to be quite low due to thehigh speed of sound of the mixture whencompared to the surrounding oxygen. Evenafter the shock transmits through the mixtureand reflects off the downstream interface thetemperature and pressure of the mixture wouldonly be 675 K and 9.5 atm respectively. Theseconditions are well into the non-explosionregion. The same one-dimens ional analysis forthe Mach 3.1 case yields a mixture temperatureand pressure of 700 K and 10.1 atmrespectively. This is closer to the third limit yetalso in the non-explosion region. But,

AIAA-99-0821

surrounded by either a nitrogen or air atmosphere.No chemical reactions were observed in theshadowgraphs for any of these tests. All of thesetests were at Mach 3.2 to 3.5, and thus were atconditions where chemica l reaction were observedfor stoichiometric bubbles surrounded by oxygen.The photo detector did pick up a faint amount oflight with a long de lay time for one Mach 3.4 shockfor a stoichiometric bubble surrounded by anitrogen atmosphere. However a second test atthe same condition revealed no light emission,indicating that this condition must be close to anexplosion limit. Norma l shock theory predicts thetemperature and pressure behind a Mach 3.4 shockmoving through nitrogen or air to be 935 K and 13.3atm respectively. This condition is well into the mildignition region. A one-dimensional analysis of ashock interacting with the interface between astoichiometric bubble and a nitrogen or airatmosphere gives the temperature and pressureinside the bubble after the internally reflected shockto be 795 K and 12.5 atm. This condition is alsowithin the mild ignition region. Both of theseconditions are plotted on Figure 6. The pressuresare the same and the temperatures are about 15 Kcolder for the same Mach 3.4 condition when anoxygen atmosphere surrounds the bubble, yet noreaction is observed for the nitrogen or air case andreaction is observed for the oxygen case.

So the question remains as to why no reactionis observed when the oxygen atmospheresurrounding the stoichiometric bubble is replacedwith nitrogen or air. The answer lies in the kineticmechanism that regulates the third limit. Thegoverning mechanism at this limit is:


7/16


AIAA-99-0821

system boundary. If that boundary is oxygenas opposed to air or nitrogen, the HO, radicalhas an increased probability of hitting an 0atom in a reaction such as

HO,+0 t) mOH+O,giving a slight edge to the chain branchingpaths. Conversely the nitrogen or airatmospheres increase the probability the HO,radical is removed from the system giving aslight edge to the chain breaking mechanism.So the nitrogen or air atmospheres act in a wayto move the third limit towards the right on theexplosion limit plot. This is analogous to thefindings of Lewis and von Elbe [17] that thethird limit is shifted to the right when thediameter of the spherical vessel used in theirexperiments was reduced allowing the HO,radical to reach the walls sooner and beremoved from the reactions. Thus with thetemperature and pressure of the hydrogen-oxygen mixture of the present experimentsomewhere between the normal shock solutionand the one-dimensional analysis results, thenitrogen and air atmospheres must shift thethird limit enough to the right to put the actualcondition in the non-explosive region.

CONCLUSIONSThe interaction of a strong plane shockwave with isolated regions of gaseous mixtureswas examined experimentally. Specifically, thenon-uniform mixtures examined consisted of aspherical bubble of pure hydrogen or hydrogen-

above for stoichiometric mixtures o f hydrogen andoxygen surrounded by an oxygen atmosphere. Thechemical reaction was identified by the presence ofstrong blast waves appearing in the shadowgraphimages. No reaction occurred when the incidentshock strength dropped below Mach 3.1. Thisdemarcation between reaction and non-reactioncorresponds to the classical third explosion limit fora hydrogen-oxygen mixture.

When a nitrogen or an air atmosphere wassubstituted for the oxygen atmosphere surroundingthe stoichiometric mixtures of hydrogen andoxygen, no chemical reactions were observed for.incident shock strengths up to Mach 3.5. It isargued that the nitrogen and air act to shift the thirdlimit to the right on the explosion limit diagram.

ACKNOWLEDGMENTSThis work was supported by the Air ForceResearch Laboratory, Munitions Directorate, Eglin

AFB and the PALACE KNIGHT program of theUnited States Air Force. Additional financialsupport was provided by the University of Michigan,Department of Aerospace Engineering.

1.

2.

REFERENCESMarkstein, G. H. 1957 A shock tube study offlame front-pressure wave interaction. SixthSymp. (Into) on Combustion, pp 387-390.ReinholdMarkstein, G. H. 1964 Nonsteady FlamePropagation, MacMillian, NY, Chapter D.


8/16


AIAA-99-0821

7. Abd-el-Fattah, A.M., Henderson, L.F. &Lozzi, A. 1976 Precursor shock waves at aslow-fast gas interface. J. Fluid Mech. 76,157-176

8. Abd-el-Fattah, A.M. 81 Henderson, L.F.1978a Shock waves at a fast-slow gasinterface. J. Fluid Mech. 86, 15-32.9. Abd-el-Fattah, A.M. & Henderson, L.F.1978b Shock waves at a slow-fast gasinterface. J. Fluid Mech. 89, 79-9510. Haas, J.F. & Sturtevant, B. 1987 Interactionof weak shock waves with cylindrical andspherical gas inhomogeneities. J. Fluid

Mech.181,41-7611. Marble, F.E., Hendricks, G.J. & Zukoski,E.E. 1987 Progress toward shockenhancement of supersonic combustion

processes. AIAA Paper 87-l 88012. Marble, F.E., Zukoski, E.E. & Jacobs, J.1990 Shock enhancement and control ofhypersonic mixing and combustion. A/AAPaper 90-l 98113. Waitz, I.A., Marble, F.E. & Zukoski, E.E.1991 An investigation of a contoured wallinjector for hypervelocity mixing

augmentation. A/AA Paper 91-226514. Lee, S-H., Jeung, I-S. & Lee, S. 1996aApplication of shock-enhanced mixing tocircular cross-sectional combustor. A/AA


9/16


AIAA-99-0821

HeavyGas -

(a) (b) w


10/16


9 AIAA-99-0821

All dimensions inmm

?X?SSQShocktube Cross 0 0 0 0 0-~ Shocktube Test SectionSectionFigure 2. Shock tube schematic

Test Section T (c)l999 American Institute of Aeronautics & Astronautics


11/16

AIAA-99-0821



12/16

AIAA-99-0821



13/16

AIAA-99-0821

loo- I I I I , I I I I

10

1.0

0.1

0.01

nem, 0 *o

N o Explosion \

\ \P\?--\Q\ c-,-\8.\ +\

/Explosion


14/16

Test Atmosphere Test Gas Shock Shock Post Post IndicationWave Wave Shock Shock of ChemicalMach Velocity Pressure Temperature ReactionNumber Ws) p,

Wm )d5 S12027 02 H, 1.7 550 3.2 425 No72% S11257 0, 4 2.2 720 4.8 543 Noizi-5 S12067 0, 4 2.9 930 9.9 750 No9if Si 2057 02 4 2.9 960 10.6 750 No-IA2, so9177 02 l-4 3.1 1020 12.2 815 NoE SO4238E. 02 4 3.6 1170 15.3 1110 No

SO4268 0, Hz 3.7 1210 15.3 1150 No

Table 1 Summary of Results from the Pure Hydrogen Bubble TestsSpeed of sound used for given atmosphere: 02 - 326 m/sCalculated from normal shock relations ba sed on Mach and given atmosphere ideal gas properties


15/16

Test Atmosphere Test Gas Shock Shock Post Post IndicationEquivalence Wave Wave Shock Shock of ChemicalMach Pressure Reactionatio Velocity Temperature

4 Number Ws) p2Wm )27 SO4208 02 1 o 3.0 970 10.9 780 No2z SO4028 02 1.0 3.1 1010 11.9 815 Yesti so71 68 02 1.0 3.3 1080 12.6 890 Yes0,if SO4286 02 1.0 3.5 1140 13.9 970 Yes7-L9 SO71 38 02 1.0 3.5 1140 13.6 970 YesEz0)2 Table 2 Summary of Results: Premixed Hydrogen-Oxygen Bubble in an Oxygen AtmosphereaD Speed of sound used for given atmosphere: 02 - 326 m/sv)T . Calculated from normal shock relations based on Mach and given atmosphere ideal gas properties2ESz


16/16

Test Atmosphere Test Gas Shock Shock Post Post IndicationEquivalence Wave Wave Shock Shock of ChemicalRatio Mach Velocity Pressure Temperature* Reaction

4 Number ow p*(atm)SO8028 N* 1.0 3.2 1130 12.6 860 No3

1SO721 8SO7228

N2 1.0 3.4 1210 13.9 9354 1.0 3.4 1200 13.9 935

NoYes4

SO8058 Air 1.0 3.4 1180 14.3 935 NoSO8048 Air 1.0 3.5 1190 14.3 975 No

Table 3 Summary of Results: Premixed Hydrogen-Oxygen Bubble in Nitrogen or Air AtmosphereSpeed of sound used for given atm osphere: N2 - 350 m/s, Air - 344 m/sCalculated from normal shock relations based on Mach and given atmosphere ideal gas properties%hadowgraphs not available, but photodetector did not pick up any lightNo blast wave seen on shadowgraph images, but a small amount of light picked up by photodetector

Documents

M. Valentino, C. W. Kauffman and M. Sichel- Experiments on Shock Induced Combustion of Isolated Regions of Hydrogen-Oxygen Mixtures