F I ENGINEERING RESEARCH INSTITUTE:i| Te UNIVERSITY OF MICHIGAN Page PROGRESS REPORT NO. 1 AAF CONTRACT W-S33-38-ao-21100 PERIOD 1 JULY - 1 SEPTEMBER, 1948 REPORT UMR- 1. This document contains information affecting the national defense of the United States within the meanin of te Espionage Act, 50 U.S. G., 31 and 32. Its transmission or the revelation of its contents in any manner to an unauthorized person is prohibited by law

TM-21'I ENGINEERING RESEARCH INSTITUTE' Page R. 0MR~-21! UNIVERSITY OF MICHIGAN PROGIRSPS EPOT NO. 1 AAF CONTRACT W 33-o58 ac-21100 PERIOD 1 JULY - 1 S3EPT1,1948 REPORT UMR-21 AEONAUTICAL RESEARSB C' E R Willow Run Airport Tpysiuanti XMioMoan

T1M~21 'ENGINEERING RESEARCH INSTITUTE I Page UNIVERSITY OF MICHIGAN UNIVERSITY OF MICHIGAN I - - I~~~~~~~~~ lFZINTION OF SYM4BOIS AND TWEiS Normal Flie Speed or Normal Propagationi Rate is defined as the rate at which a flame front moves in a stagnant combustible mixture relative to the unburned gas in a direction noal to the front. a s peed of sound F/A = Fuel-air ratio by weight V 1M - = Mach Number a P1 = Ambient air preasure P2 ~ Static pressure after shook T = Ambient air teamturare To T3otal temperature of air stream V vJ - velocity Jet velocity y p/c V = Ratio of specific heats

iVD.,. 01 ] ENGINEERING RESEARCH INSTITUTE PaIM-21 'UNIVERSITY OF MICHIGAN ii TABLE OF CONTENTS Page Definition of Terms i Purpose of Project I Procedure Sumiary 6 Progress 7 Blov-off Velocity 7 Large Scale Combustion Chamber 8 Temperature and Pressure Effects on Combustion Processes 8 Detonation 8 Program for Next Period U Activities Visited 12 References 13 Distribution 14

ENGINEERING RESEARCH INSTiTUTE Paeg.URT-21 I UNIVERSITY OF MICHIGAN iii LIST OF FlGUB BFgure No. Pag 1 Flame Holder and Nozzle Assembly 7-A 2 Schematic Diagram of Blow-off Equipmnt 7-B 5 ]Blow-off Velocity vT-0t.Air Ratio 7-C 4 Schmlti Diagram a of a Stll Scale 7-D Pressure Apparaft for StUaOH Oem Effects of Pressure on the Coatotion Processes 5 Modified Bullet Nose 9

L [ tvr0I ENGINEERING RESEARCH INSTITUTE pag IJI-1 rUNIVERSITY OF MICHIGAN I, Purpose This contract is a basic study of the fundamntals of combustion as related to the specific variables stated in xhibit A. In order to carry out such a study economically, the problem will first be investigated on smnll scale versatile equipment. Success on this equipment will dictate the procedure to be followed on the larger and more cumbersme equipment. he purpose of Contract W 33-038 a-21100 as stated in Exhibit "A" i3reproduoed here in full. A; Determine the effects of the following conditions upon the ocmbustion process, maTinnxr flame temperature, and physical and chnical.characteristics of the flame at various points along the combustion chamber. If possible, each condition is to be imposed upon the oombustor independently while the other variables are held constant. After determining the effect of each condition separately, the cumulative effect of two or ore of the conditions are to be determined. A single base fuel of petroleum origin should be used for these tests and the F/A ratio varied as required. (1) Inlet air temperatures up to 1000~F. (2) Inlet air pressures up to 200 p.s.i.a. (3) Inlet air velocities up to and possibly exceeding a Mach number of 1. B. Determine the effect of turbulence upon combustion and flame characteristics. Controlled turbulence shall be introduced in the combustion chamber and each type of turbulenice evaluated while other conditions are being held constant. Turbulence may be induced by the following methods: (1) Various designs of combustion chasbers (2) Air jets (3) Flme holders.of various shapes in the combustion chamber. C. Deterxine the effect of various types of turbulence upon combustion superimposed upon the conditions listed in A above. D. Determxii the effects of fuel characteristics upon combustion and flame characteristics. Repeat tests under A and B above using various liquid petroleum fuels. 'Cpare the results with those obtained with the base fuel to evaluate the effect I I - I -- ----

.-_~l21 ^ ~ ENGINEERING RESEARCH INSTITI TE Page 2 UNIVERSITY OF MICHIGAN of fuel characteristics upon combustion. Z. Using the infontion obtained in A and B above, attempt to correlate the data so as to obtain a ccmbustion parameter involving all the above variables and suitable for ccabuttion chamber design calculations. F. Using the resulting data from investigations under D above, attempt to detenaine and express practical reaction rate equations for the fuels or fuel types used in the work. G. Investigate the feasibility of maintaining a standing detonation wave in a high velocity air fuel mixture. Such an investigation may be carried on in a small bore burning tube to avoid the use of excessive amounts of air and fuel. Initial tests will be made using hydrogen and oxygen or hydrogen and air. Depending upon the success of these tests, gaseous and liquid hydrocarbons will be also used as the fuel. To assist in holding the detonation wave, the use of stratified mixtures, detonation traps or other mhanical devices should be investigated. This investigation is to be of a basic nature and does not include a study of the application of a standing detonation wave. H. A seiple combustion chamber of the Contractor's own design shall be employed with auxiliary equipment capable of deliver-, ing air to the combustion chamber at pressures up to 200'ps.l.a.i, velocities at Mach number equaling, or possibly exceeding, and teaperatures at 1000~F. The combustion chamber shall be constructed in suh a manner that various methods of introducing turbulence may be eployed. I. The combustion chamber may be a simple burning tube of circular cross s#otionb and of a reduced scale to avoid the necessity of excessively large auxiliary equipment. The burner should be large enough however so that results will not be adversely masked by surface-volum ratio. Present estimates indicate that burners with a two to five inch diameter may be used to obtain reliable data, J. Investigate the effect upon flame speed of the follcor variables: (1) F/A ratio (2) Fuel characteristics ad types (-) Turbulenoe

1ENGINEERING RESEARCH INSTITTE! Pae MJRU-21I UNIVE:RSITY OF MICHIGAN - |,a-e35 (4) gnition intensity (5) Flame holder geometry K. Inveatigate the divergent-convergent flow phenomena associated with a V-flaem ooifined in a e amltion chamber. Various shapes of ccmbustion chambers shall be used and an attempt made to arrive at the combustion chamber fom most suitable for V-flame front stability. An attept shall be made to arrive at a mathematioal analsis of the potential flow through a V-flame front in order that combustion chamber design may be placed on a more sound mthematical basis as far as fundamentals are concerned. L. Conduct preliminary investigations of detonation phesnosna in stationary hages for the purposes of finding the effect of the following variables: (1) F/A ratio (2) Ignition intensity (5) Fuel characteristics and types (4) Shock waves 1. The research program called for hereunder shall not be considered an exact guide or limitatin on the work to be perfonued. Depending upon the results obtaind during the course of these investigations, deviaticas and/or additions to the outline may be applied at the discretion of t.kvt vernm nt or upon the suggestion of the Contractor with the approal of the Goverment.

!11~-2l ENGINEERING-RESEARCH INSTITLT TF Page.....u.-.21 ____ ~ _______._ UNIVERSITY OF MICHIGAN... 4 Il. Procedure to be followed on the initial hase of the Contract Generally, cambustion is accepted as a phenomenon associated with multiple diffusion processes from the burn gases to the unbured gas. The classical analysis of Mallardo-Ie bhatelier *ilh is based upon the heat transfer from the burnd gas to the uburned gas has been used by such investigators as Lewis anl yon Elbe to explain e rate of propagation of flames in gaseous mixtures. More recently, other investigators have considered the diffusion of active species such as atomic hydrogen to explain flame propagation rates, A satisfactory theory and analysis to explain flame propagation rates not jet been proposed, Past combustion work at the University of Michigan (Ref.1 and 2) has shown some unexpected variations in the V flames when the temperature and geometry of the holders were varied. In such oases, higher flame speeds were observed for hot holders. At high Jet velocities, ca. two to three hundred feet a second, it was possible to produce a small pilot flae in the wake of the flame holder which failed to ignite the bulk of the combustible mixture. The existence of such a pilot flam as well as the foregoing mentioned phenomenon indicates that the aerodynamics associated with combustion could be as important as the chemical considerations. An understanding of the aerodynamic effects would give much insight into the phenomenon of blow off and combustion instabilities that exist at high free stream velocities. It is not unlikely that the aerodynamics could influence flame propagation in gaseous mixtures. Accordingly, to have an understanding for the variables involved, various sizes of flame holders are being tested for blow-off speeds at various fuel air rJties. There appears to be evidence that the velocity at blow-off divided by the normal propagation speed will be a constant. (Bef. 3 and Fig. 3) Further, a correlation between blowoff velocities and the velocity gradients in the vicinity of the holder is to be investigated. If velocity gradients prove to have an effect on blow-off speeds, it is quite possible that the flame speeds will also be a function of these gradients. In order to have a knowledge of the magnitude of the velocity gradients involved in the case of a "V" flame, a potential flow analysis of the "V" flame is in progress. Experimental verification of the latter analysis will be made on both the open "V" flame and the confined "V" flame. I.A

L ENGINEERING RESEARCH INSTITUTE P UM. -21.. UNIVERSITY OF MICGAN 5.I The effect of inlet air teP'erature up to 1000 F. and inlet air pressures up to 200 p.s.i.a. will first be studied as related to ~t nor.al f lame spede. Equipment for this investigation will consist of a small Buneen burner and 't" flame apparatus which can be subjected to air pressures from 3 to 4 p.s.i.a. to 200 p.a.i.a. With such a system the air flows required are very small and can be maintained with a small capacity air compressor. Such small scale equipment also makes it possible. to easily obtain these temperatures with a small heat exchanger. The University of Michigan has at its disposal a hot wire anemoaeter capable of measuring both the intensity and scale of turbulence,- The effect of b**ulene upon combustion will first be investigated by controlled turbulence obtained by the inser.tion of grids upstrea of the. V" flame, Such turbulence could easily be masured with a hot wire anmmter. After the completion of this initial investigation, turbulene of a more ccmplicated nature will be introduced by Variations in combustion chamber design, introduction of air ets, and flame holders of various shapes. The feasibility of maintaniing a standing detonation wave in a duct will first be investigated by use of a shock tube capable of Mach numbers up to 5 or 6. The teaperatures behind such shock waPv6 will be well above the igaition temperatures of the fuel-air mixtures used, and therefore should produce combustion. The combustio in.turn should further increase the pressure behind the shock wave which would strengthen th&e shock wave traversing the tube. Such a shook wave followed by combustion would then be a detonation wave. The shock tube will give sufficient data to determine the temperature an pressures required to iaittate combustion as well as final velocities obtained after the combustion has developed. In this manner, it will be possible to investigate the temperatures and Mach numbers required to initiate and maintain a standing detonation wave. Since a detonation wave must confom to (1) the law of conservation of mass, (2) the momentum equation, and (3) the energy equation, an analysis is being made on this basis. It is possible that suh an analysis ill show some relation between the speed of the detonation wave and heat release. I - ~

I ENGINEERING RESEARCH INSTITUTE Page UtMR-21 __ UNIVERSITY OF MICHIGAN 6 3:1, SuBmary A, 4aneral A large portion of Period I was used in the design and setting up of small scale test equipment to be used in the initial investigations. The equipment is described under individual headings. B, Blaw-off Velocities Blow-off velocities of spherical flameholders were easured for two diameters of spheres. The results, based on this limited data, indicate that a correlation between blow-off velocity and nomal flam speed exists. C. Large Scale Combustion Chamber The major components of the 1hrge scale cmBbustion chamber have been assembled d tested. The burner has been successfully operated with the aid of a hydrogen pilot flame. The approximate Xlimts of the flow have been established both for the air system and fuel system. D. T erature and Pressure Effects on Combustion Processes The small scale equipment required for the study of pressure effects on combustion has been assembled, and tests will proceed as soon as the flow instruments have been calibrated. 3. Detonation Initially the phenomenon of detonation is to be investigated in a shock tube capable of Mach number equal to five or six. The equipment fr auch a shock tube has been partially designed. In addition theoretical work on the phenomenon of detonation has been carried out.! - - -

L- ENGINEERING RESEARCH INSTITUTE Page --- LUMP'<~*_-21 j UNIVERSITY OF MICHIGAN 7 ] tV, Progress B. Blow-off Velocities of Fli olders The blow-off velocities of flameholders te ted in an Open Jet because of the s BiplciOty f the gs t -I Spherical flaieho.lers were selected to elimite err due to orientation and because of -beir relativel1 si e flow patterna Four spher"e of the folcw ing diait.a have ben prepared: 0.25:, 0.125, 0.156% and 0 78iches. The data inAltied in tais rport ia based on test- comdwted vith 0.1625 en 0.125 inch iaietr phor- a Th.e method ubed 'to emunt theo flamholder isn amin Figi'- 1. Figure 2 presents a echeatilo diagram of the equi0iea The fuel vas oa al propane. prope flaf van meaiiredi by a rotemster, and the total flec of the wtixtr. was meaured by a of. a pressure tap in te i dn -W and a calibrated. nozzle. The air flow Waa obtai.ted by s-:b.tracting the propane flow frc the total Bixture flcw. -Ehe mixture was ignited with an electric epr at a Jet veloity of ca. 100 ft. per second' The Jet velocity va' then varied to some given value, at which point th fuel-air- ratio Was changed in saral inrw te until- blov-off oocured. This procedure was used on both the rick and lean sides of the blow-off. The pilot flame which xtas on both the rich an lean sides of the curve assislted in determining the blcw-off region inasmuch as it appears shortly before blow-off oc-Ofre, After blaw-off had occurred, amu electric spark was passe^ through the mixture to determine whether or not it was combustible. Several trials were used to eliminate the possibility of the pilot flame having been blown out by a gust. The experimental results obtained thus. far are shown in Figure 3. These curves were obtained from the 0.0625 inch and the 0.125 inch spherical flame holders, A portion of the latter curve has not yet been substantiated, Flame propagation oan be mintained only within the area bounded by the curve. The ymx blcv-off velocity was obtained at a fuel-air ratio of approximately.075 for the 0,0625 inch diameter sphere. There is soe doubt as to the ]

Page 7-A MR-21 L - FLAME HOLDER AND NOZZLE ASSEMBLY FLAME HOLDER WIRE FIG. I

I L F -~~~..__ SCHEMATIC DIAGRAM OF dLOW-OFF EQUIPMENT \) H PROPANE STATIC PRESSURE PROPANE ROTAMETERA N RESERVOIR PRESSURE -, ROPANE T EMPERTURE, STACK PROPANE METERING VALVE, TANK SHUT-OFF VA LVE // r- FLAME HOLDER X | | 5/ NOZZLE V8o 1 / //7//////// '///' //,~~ VALVE -30 GAL MIXING DRUM AIR METERIN( SHUT-OFF 0% 3 /,L i VALVE.. — 100 P.S.I. AIRLINE?- PROPANE TANK i~~/~~/////~% // / /:/// //I FIG. 2

UMR-21 Page 7-C BLOW-OFF VELOCITY V.S. FUEL AIR RATIO FOR.0625 a.125 DIA. SPHERICAL FLAME HOLDERS THE FUEL USED WAS -COMMERCIAL PROPANE li aL UI >L0 I m.01.02.03.04.05.06.07.08.09.10.11 FUEL AIR RATIO BY WEIGHT.12.13 FIG. 3 -- ~ ~ ~ ~ ~ ~ ~ ~ -

I I - _ _ AIR ROTAMETER EXHAUST CONTROL VALVE ---- PUMP COMBUSTION CHAMBER AIR PRESSURE / VALVE MIXING DRUM II 'ER _ iC -jy~~_ In I.... SCHEMATIC DIAGRAM OF A SMALL SCALE PRESSURE APPARATUS FOR STUDYING THE EFFECTS OF PRESSURE ON THE COMBUSTION PROCESSES. A I %31 -r

I U] -2.1 ENGINEERING RESEARCH INSTITUTE Page 8 UNIVERSITY OF MICHIGAN relative position of thse Wto ourrs o the rioh side. c. Large Soale Combustion ohamber The large scale combustion cmber equipmtnt, vfiob consists of a naturallyaspirated Rolls Royce esnge aend the piping to the 4" buner section, has been put into operation and is ready to proceed with a test program. fhitia teets have been aaae on the bwrer section, and all auiliary equipmint (fuel pi, ignition, and control valves) opertd atisfactorily. Blanks for a seriee of fuel nozmlea have beeAn iaa up, and calibration of thee anozzles at different fuel preuree is proceeding. Various meansw of evalxuating oWbuaatiosi chamber perforne are being tudied. D. Temupera:ture and Presure effects on Coabustioa. Pr ea The first period of the otraot sas oeen spae org=anzze facilities aZd test equipment for all sal s "le eAtdie. A aketoh of the apparatus to study effects' of pre ost O obustion prooegse is shown in ig lgr. 4. Thel*ih'ooarbonc fWe to be used is coaarcially pure propane. -hie S: fo is controlled by a eteing valve t h a flw ter to the mxing dru vhere it is mided with ar. ir. l suBi tlied by a low capacity cpreesorS-.to a large hi prerieare stage tank The air flow is then thrott:ad through a ettriBg -alve to a flow meter ed then passsa o the s ixing drun. Tho fuel-air nmxture flows to the pressurzed omb tion ohbr where omt bustion prooesses can be obrexd watgh windcws. ae. - bhaut gaees are controed r g a tthottlln vaTle, B using a vaouum pump and light modifioatios of the aysten, effects of lower than atmaopheric pre^eure on oombustion phenoaena oan be stxUdie. At the present,AewquiPent is being assembed and oalibrated. E. Detonation Ignition of a combustible mixture with a shook wave involves other conditions than teperature and pressure. Il order to confirm this, four bullets were fired through stoichiometric mixtues of propane and air at atmoepheric conditions. The nuzzle velocity of the bullet as specified by the manufacturer was 4140 feet per second. The nose of the bullet was ground flat (See Fig. 5) to produce a normal shook at the noee. i I - —

I L UMR-21 | ENGINEERING RESlEARCH INSTIT'IT1'E | 1_'-2__I _UNIVERSITY OF MICHIGAN 9 Theoretically the shock had a pressure ratio a total temperature, To, at the nose of 1965QR following formulas: Pe of 15.4 with P1 based on the P2 PI 2 M2 + 1 1(+ 1.T 2 T + ( le0 T = 540o, - 1.4 j I Iigwre 5 The propane air mixture and a spark gap were seled in a cardboard oarton into whioh the bullet was fired at point blank range. The mixture failed to ignite each time the bullet was fired through it. In order to prov the inflaamability of the mixture, it was ignited by a spark after the bullet traversed it. This initiated combustion, blowing the carton apart. In the bullet experinent the stagnation tenerature behind the shook at the nose of the bullet was 1965~ R and the absolute pressure was approximately 225 psia. The accepted ignition temperature of propane-air mixtures is 1374cR (Ref. 4). Dr. Shepherd ignited an ethylene-ir mixture at 708~R to 726~R with a shock wave (Ref.5). The ignition teoperature for ethyiene-air mixtures is \1464~R (Ref. 4). Therefore, Dr. Shepherd ignited the ethyleneair mixture at temperatures below accepted ta,!u.. quoted in the reference. --- ----- -- --- --- ----- ---- — ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..::

ENGINEERING RESEARCH INSTi'ITUTE. Pag UMR-21 UNIVERSITY OF MICHIGAN 10 The bullet experimait in which teperaturee at the nose of the bullet were above the ignition temperature of the gas mixture anA Dr. Shepherd's Ifrk with ethyleneoair mixtures confirm the belief that conditions in addition to temnprature and presaure control ignition behind a shodk wave,

ENGINEERING RESEARCH INSTIT ITE U VoUNIVERSITY OF MICHIGANlt period V* Pograa for Next Period B. Blow-off Velocity It is planned to install a Bunsen burner in the system to enable measurement of flame speed and blow-off velocity simultaneously. C. Large Scale Combustjon Chamber The initial test program has been outlined for testing the effect of high pressure liquid fAel injection upon combustion. It is proposed that the angle of spray from the fuel nozzle as well as the fuel pressure be varied with photographs of the spray being taken while the data on each nozzle is obtained. D. Te perature and Pressure Effects on Combustion Processes A study of the effect of temperature on combustion processes will be commenced upon the completion of the pressure studies. At that time a study of the cumulative effect on both temperature and pressure will be initiated. E. Detonatin At present a shock tube is being d agnti, agd a method of photography capable of 'WtuyiAy ' the shock wave and the ignition of the mixture is being developed. F. Blow-Down Euipment A design study of a blow-down tunnel, utilizing the high pressure storage, pumping and dta~tg equipment available, will be made.

ENGINEERING RESiEARCH INSTITUTE Page 1 UMR-21 _..... 1 IUNIVERSITY OF MICHIGAN ACxTIVITIES VISf!E i, ~ ~~~~ --- ------ Activities Visited Ducted Power Plant Symposium Sponsored by Project Squid Navy Department Washington, D. C. Flame stability2 ejectors, and ducted pcrer plants.

L ENGINE;ERING RESEIAR( 1,NST!'l p11. UR-21 UNIVERSITY OF MICHIGAN 13 Univgan eC 1. University of Michigan Progress Report No. 9 2. University of Michigan Progress Report No. 10 5. University of Michigan Progress Report No. 6, p. 51 4. Marks Handbook 5. Dr. W. C. F. Shepherd, Ministry of Fuels and Pcwer, Sheffield, England. "The Ignition of Gas Mixtures by Impulsive Pressure Effects". Abstract of Papers Third Symposium an Ccambustion and Flame and Explosion Phenamena. Septenber 7-11, 1948, Page 91

L ENGINEERING RESSARCHI INSTITTEI Pa, UMR-21 _________ lUNIVERSITY OF MICHIGAN 14 DISTRIBTIOIN 7 Black and White Copies 1 Carbon Back Tracing To: Comaanding General Air Material Ccaand Wright-Paterson Air Foroe Baa Daytoa Ohio