AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN 9MR-33 PCRT mR - 33 PROGRESS REPOP NO0. 6 AAF COTRACT W33-058 ae a100 PIRIOD 6 1 May to 1 July, 1949 --- - -. — - --

I I i i i I 1 AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN TUMR.-33 IIST OF FIGUt S Fig. Page 1 Fastex Shadowgraph of Burning Behind a Spherical Flameholder 5 2 Sketch of Surface Section Flame Holder 6 3 Photograph of Flane - no vacuum 7 4. Photograph of Flame - with Vacuum 8 5. Blowoff Velocities of Flaamholder 8 6. Sketch of Separition Experinnt 10 7. Shadowgraph# of Explosi.ns in Cmbustion Chaber 13 8. Streamlines in Flwa with V-flame 17 9. Shadowgraph of Confined Flame 18 10. Photograph of Confined Flam 18 11. Shadovgraph of Unconfined Flame 19 12. Photograph of Unconfined Flame 19 13. Sketch - Flame Front types 15 14. Shadowgraph of Combustion Chamber - No Flow 20 15. Photograph of Downstream End of Shock Tube 22 16. Drawing of Conplete Assembly 23 --

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN Wi-33 ii TABLE OF CONTENTS Page No. I. List of Figures iii II. Summary of Work Conducted During the Period 1 1 May to 1 July - 1949 III. Progress Blowoff Velocities of Flameholders 3 Combustion Chamber Design 11 Pressure and Temperature Effects on' Combustion Processes 14 Flow Associated with the V-flame 15 Detonation 21 Blowdon Equipment 24 Experimental Techniques and Instrumentation 25 IV. Program Planned for Period 7 26 1 July to 1 September - 1949 V. Activities Visited 28 VI. Distribution 29

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN I I I TIMR-33 Page 1 III. Sunmary of Work Conducted During the Period Blowoff Velocities of Flameholders A series of Fastex shadowgraphs of burning behind a spherical flameholder were taken at various fuel-air ration and jet velocities. A cylindrical flameholder, axially mounted in the jet and incorporating surface suction, was constructed and tested. The application of surface suction to this flameholder altered the stability limits and appearance of the flame considerably. It was found possible to greatly alter burning characteristics by various combinations of surface suction and artificially provoked separation. Combustion Chamber Deign Spark shadowgraph pictures were taken of the various stages of an explosion of a propane-air mixture in a acmbustion chamber in an effort to gain an understanding of the flow pattern and propagation phenomena associated with rough burning. High speed movies of the rough burning were taken but failed to show satisfactory detail. Pressure and Temperature Effects on Combustion The new combustion box has been fabricated and pressure tested to 500 psi. The burner heat exchanger was tested to 7000~ exit gas temperature. The installation and instrumentation of the complete pressure temperature system is near completion. Photographs were taken of the V-flame and Bunsen-flame in the new burner assembly. The flame speeds computed for the Bunsen flame agreed quite well with the previous work, however, the V-flame speeds showed a significant discrepancy Flow Associated with the V-flame The experimental flow field upstream of a V-flame in a rectangular cambustion chamber was compared with the flow field described by 4) = C, e o2 ^ i Vt l * The comparison is not good, indicating that this type of theoretical analysis is to no avail. A confined and an unconfined two-dimensional flamel are compared by means of shadowgraphs and photographs. The unoonfined flame does not have the bulbous type flame front that the confined flame does. Detonation Final assembly and alignment of the shock tube, valve housing, and reservoir havi beo ecpl ted. 1. A confined flame rqters to a flame burning in a parallel wall combustion chamber, while the uncrfined fame has only two of the four walls, making it unconfined in two 4imensifs.

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN I I II~~~ I1R-33 Page 2 Blowdown quipmnt The two 3,000 psi air compressor units have been permanently installed. The heat exchanger support stand is complete, with partial completion of the heat exchanger proper. High pressure piping has been selected and a satisfactory bid received. Materials for construction of the compressor building have been selected and ordered. Experimental Techniques A new high voltage power supply for spark shadowgraph photography has been completed. The investigation of the Kerr 'cell shutter was continued. The shock-tube and the blowdwon tunnel instrumentation problems are being studied. The procurement of the Interferometer and schlieren systems was processed. - -

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN TMR-33 Page 3 IV. Progress Bl!owoff Velocities of Flameholders A series of Fastex shadowgraphs of burning behind a spherical flameholder were taken for various fuel-air ratios and jet velocities. A representative sequence of three frames for each run is shown in Figure 1. These shadowgraphs were taken at approximately 5,000 frames per second using a carbon are light source. Projection of the films indicated that a camera speed of 5,000 frames per secend is not sufficient to stop the motion in the wake of the sphere during burning. It further indicated that the irregularities in the wake contour are a random phenomena having no particular characteristic frequency. The presence of density striations in the wake can be seen clearly in the fourth sequence of Figure 1. Additional equipment for improving the quality and size of the image has been obtained and provisions were made to increase the speed of the camera to 8,000 frames per second. A surface suction flanmholder was constructed and tested. It consisted essentially of a streamlined body with a spherical tail and a series of No. 80 drill holes just upstream of the separation point to bleed off the boundary layer. The faemeholder was mounted on a hypodermic tubing, which was connected to a water aspirator which in turn evacuated the flameholder during burning. Because of the large pressure drop through the hypodermic tubing and the low capacity of the water aspirator, no conclusions were drawn from this experiment. A second type of flameholder was then devised incorporating surface suction and is shown schematically in Figure 2. The vacunu was applied with a Nash Hytor MD673 Vaeuum Pup. A flow tube in the vacuum line was calibrated so that the amount of mixture removed could be measured. At full vacuum the flow through the bleed holes was approximately 2.5 cubic feet per minute of 0.073 pounds per cubic feet air. This value was substantiated by both the calibrated flow tube and the pump specification curve. Figure 3 is a photograph of a flae burning off the cylindrical holder with no vacuum applied at a fuel-air ratio of 0.0865 and a jet velocity of 137 feet per second. Figure 4 is a photograph of the same flame at the same conditions after full vacuum was applied. The flow through the ten No. 60 (0.020) drillholes is approximately 2.5 cubic feet per minute. The total flow through the jet is 42 cubic feet per minute. Calculations based on the Blasius solution for the arder of magnitude of the thickness of the boundary layer indicate a boundary layer flow between 0.5 and 2.5 cubic feet per minute. The use of smoke tracers showed little or no disturbance of the flow pattern with vacuum on. Therefore, it is believed that most of the boundary layer is being removed with no appreciable alteration of the flow pat4jrn about the holder..

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN 7 IIII L II _ I III III II~~~~~~~~~~~~~ — TMRw33 Page 4 Figure 5 is a oomparative plot of the stability limits on the rich side for this type of flameholder and for spherical flameholders. As was shown in WUR 29,* a single curve was obtained for all spherical flameholders on the rich side* The curve for the flameholder shown in Figure 2 without vacuum concurs exactly with the curve obtained for the four spheres tested previously. The rich limit stability curve for this flameholder with vacuum is also shown in Figure 5. With vacuum applied there is an appreciable decrease in the stability limits. A further extension of this experiment is shown schematically in Figure 6. A ring of 0.026" diameter wire was used to artificially create separation at various points along the flameholder. Three distinct regions were found, as shown in Figure 6. With the ring located at any point in region 1, ignition was secured easily and the flame burned stably. The pilot zone was somewhat larger due to the fast that the flow separated at the ring. Region I extended a distance of 0.125 inches measured from the end of the flameholder. At the extreme point of region I it became impossible to ignite the mixture using the initial ocmbination of jet velocity T? s, 37 feet per second and fuel-air ratio F/A - 0.0826 or any other combination of Vi and F/A. Region II represents the length in which it is impossible to obtain burning off the flameholder. At the end of region II, a distance of 11/16 inch from the downstream end of the holder, it again became possible to obtain ignition and steady burning. The combustion, however, appeared to be unsteady and very incomplete. When surface suction is applied with the ring located anywhere in region I, the flame is altered n much the same manner as in Figure 4. The downstream combustion ceases and only a pilot flame remains. With the ring located anywhere in region II, ignition and steady burning is impossible without the application of vacuum. In other words, surface suction allows the stabilization of a flame behind the holder where it was not previously possible. When the ring is located anywhere in region III, the effect of vaeuum is again the same as shown in Figures 3 and 4. It is to be noted that the action of surface suction becomes more effective in restabilizing the flame as the location of the ring approaches the point dividing regions I and II. The design of a surface suction flameholder was prompted by previous work with smoke tracers and tests on the stability limits of spherical flameholders. It was felt that perhaps the boundary layer served as a means of getting fuel into the pilot zone and consequently was intimately connected with the stability limits of a particular holder. The foregoing experiments are hardly conclusive, but a possible explanation lies in the consideration of boundary layer growth and separation. Future work is planned along these lines with the intention of investigating this aspect of the problem in the case of spherical and circular baffle flameholders. * Progress Report No. 4 - tMR-29 -- University of.Ulchigan AAF Contract W33-038 ac-21100 L

I}MN 3d 5000 [:: RAIMES PER S:ECON INCZREASING T IME DIREICtION OF FL oV IA FU!LLAIR RATI O 'AH 0.0l0o1 JE[ VELOCITY 92 F~/SFC.. A F:Fl. " ALL RATIO 0',0O665 jFr V:OCITrY - 219 Fl/S.Ct I B 1 C C. JA F [F.A. -"AIR RATIO Oi 0,0579 JET VELOCITrY.295 FI/SEC, 4 A F['EtL -AIR R.ATIO o 0,0536.7F VEI.OCI~TY 193 fTSEC, 5 A FU.EtL -AR RAYiO:, 0.071 JET VEiOC ITYc 34JO f'//SEC,.B B 3C 4.8 4I c BU" Li5C f.:. s.: i:j..............!:::AST AX SIAO.\V.R,.APIIlS,Of U!NNtNINGi Bw:HIND A $t'HERICAIL1AM)IGII isd,) '" H,!:',A. H [R (B; ('H] ~~~~~~~~~~~~ 7~~~~~C$H:.tjA\t-ii~ w ~ ~ &V~,'?:~ t"::.;" (;::~;; '~"i;~~~i;:;.- [..~?;:.~a; ~'~~;:i;;:;.~::ii:~4::~::i':~'~' ~:::::-:L.:,~:;. ', ~:''-:.:e~w;,:'~:;~;c~h~::,':. f;'~':'t~: l ~:-:':..: ~.1: ~,

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN TIMR-33 6e.o42 J De/6L Page 6 6 A/ c'L ~ AP s e,~'c2: E Q c/4 Z 1 - q c e L )~~~~~~ 4/L~k~d~ir /D,/14 DAbC 6 75,/;C 0/J4 _ _ _ _ _ _ _ _ _ '7 7/a A Cl P,=.. HQoL s I 0l,.,, 1.__.1 -. 11-..,_ 5 _ -.

UMR 33...E Figtite 3

iJhMR 33 PAG.E 8 F-gufre 4

MI-S33 t n )f a * 12. a a v i rr k d.

I\ i I I I i @ I N d P Q.N,"I rW 1iN ^(s I" - (tIA3 - q itS t~C~f (N (Z I F |l f \ ns-^ 5 Iz Qff N. I -jI 1f tt t 13, S( Ch~ Ef Q` C &&' lo fJ 10 -1 - %I,-" I If > m x >c mi *4 n >> t-;> m:~ 5 M ov n n m fK m X Po FI m 2 rZ C co *q IP II * 0 $ 3 C S Z

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN r TIAR-33 Page 11 Combustion Chamber Desig This period's work was directed toward trying to learn more about the instability that apparently is present under certain conditions of combustion in a duct. In order to gain an initial insight into the problem of rough combistion, single explosions in a small scale combustion chamber were studied. It has not yet been definitely established that an isolated explosion can be regarded as being representative of the rough burning phenomena; however, it may give se understanding of the problem. Thus a small 1 inch square combustion chamber 15 inches long with transparent walls was made so that spark shadowgraphs could be taken. The walls were of ordinary safety glass, and did not rupture due to the heat as long as combustion runs were not over 5 seconds. Since most pictures were taken of single explosions, the glass was not affected by the heat. A Ford coil furnished the ignition spark for the propane-wair mixture. Flameholders were of the simple flat plate or two-dimensional orifice type as shown in the accompanying photographs (Figure 7). The photographs are spark shadowgraphs taken With a synchronized shutter and triggering circuit. Considerable time was spent in improving the quality of the shadowgraphs and devising a technique for operating the combustion chamber and shadowgraph equipment simultaneously. This is described under Experiental Techniques and Instrumentation, page 25. The spark triggering circuit contacts and the contacts for the ignition spark in the combustion chamber are connected to the shutter and the frame which holds the film pack. By suitable spacing of these contacts, one stroke of the shutter will ignite the pre-set fuel-air mixture flow in the combustion chamber, then expose the film and trigger the spark at a predetermined time interval (At) after the ignition spark was discharged. Thus by setting the time interval (At), pictures of various stages of the explosive burning could be obtained. One difficulty experienced was the unpredictable delay in the ignition of the fuel-air mixture, especially at the higher air velocities, so that photographs taken with the same.t were seldom similar. The shadowgraphs shown in Figure 7 are some that were taken in an effort to obtain sm idea of the flow patterns and propagation phenomena associated with the initial stages of combustion. The air-fuel flow is from left to right and the ignition spark arcs to the flameholder from the wire visible in the photographs. Each one of these pictures except Numbers 2 and 8 represents a different explosion. Number 2 shows ignition of the mixture behind the holder, but it apparently blew cut as there was no explosion. There was no ignition in Number 8. Number 3 shows propagation of the canbustion upstream as well as downstream. It appears from Number 3 and Number 9 that the flow is stopped momentarily as the explosion propagates in both directions from the holders. Number 4 and Number 5 show the initial flow of the burning gases past the flameholders. Number 6 and Number 7 show the propagation of the flame toward the exit end of the 15 inch long caobustion chamber. Number 7 presents, apparently, some evidence of initial propagation in the boundary layer region. Number 8 shows the second flameholder (3/8 inch by 1 inch) with no turning. Number 9 shows the propagation of the combustion upstimek with little evidence of any I w

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN 1. r IR-33 Page 12 propagation downstream. The propagation of the combustion upstream seems to be more of a "frontal" movement as compared to the "striated" propagation downstream (Number 5 and Number 6). Number 10 shows normal smooth burning off the flameholder. The cause of the mottled appearance of the combustion chamber is not understood unless it is a diffraction effect from the impurities in the glass. It was also noticed in other pictures that even though the ignition spark arced to the top flameholder (Number 1 through Number 7), the combustion was also symnetrio, i.e., the mixture was ignited practically simultaneously behind the two flameholders. This would suggest propagation down the side walls of the chamber to the bottom holder. Several high speed movies of rough burning were taken using a 300-watt Concentrated Arc Lamp. They failed to show satisfactory detail, due probably to the coarse grain of the ground glass. Also, the camera could photograph only a small area of the chamber because of the insufficient brightness from the lamp. w I

V::

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN I I I I [ I I I J ___ --- 1R-33 Page 14 Pressure and Temperature Effects on Combustion Processes The combustion box for the new burner has been fabricated and tested hydraulically to 500 psi. Windows for the box are being made from 1 inch bullet-proof plate glass of four laminations, the type used in fighter aircraft windshields. The heat exchanger for the new burner was checked to 700~F gas exit temperature. In these preliminary tests the burner was not lagged. With sufficient lagging it is believed that the heat exchanger will heat the gas temperature to 10000~ as specified by the contract. The series of burner nozzles was checked for rectilinear flow and they are believed satisfactory. Installation of the new pressure-temperature system is near completion. Photographs were taken of "V" and Bunsen flames for each of the four nozzle sizes. The type flameholder used for the "V-flae runs was a rod holder giving the flame an approximately two-dimensional appearance for the preliminary runs. However, it is felt that with the three-dimensional nozzles a three-dimensional flameholder, i.e., a sphere, should be used to preserve flow symmetry. Such a flameholder assembly is now being constructed. The Bunsen flame speeds observed in the above test, computed by the Bunsen angle method, checked quite well with previous tests using the Gouy method for computing flame speeds frm laminar flo in Bunsen tubes. However, the flame speeds computed for the V-flames using the method outlined by Morrison and Dunlap in BM-21*were considerably lower than the data computed by the Bunsen angle and Gouy methods. One possible reason for this discrepancy was the use of two-dimensional flameholders with a three-dmensional nozzle. * UM-21 - "Measurement of Flame Speeds with the V-Ulame" By R. B. Morrison and R. A. Dunlap, May, 1948

AERONAUTXCAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN I i!m I II I I II ~ IIIII I ~I _II__ TMR-33 Page 15 Flow Associated with the V-flame In a previous progress report (Ref. 1) it was suggested that a particular solution to the Laplacian,, t,;-.0= such as t=c, e'cos - * j, v fwhich satisfies all the boundary conditions at the walls for flow in a channel, may describe the flow pattern preceding a flame in a csnbustion chamber if the constant C1 could be determined. Since the boundary conditions at the flame froni are not known, except as related to the flow field immediately after the flame, it was necessary to determine the constant l1 experimentally. To do this the streamline flow pattern was traced with Ti C014 moke and values of x and y determined. Evaluating the constant with two points on ax experimental streamline, and plotting the remainder of the streamline (Figure 8) shows that the particular solution found above does not satisfy the actual flow conditions, indicating that this type of analysis is to no avail. As illustrated in Figures 9 and 10, one of the characteristics of a stable V-flame In a lamaiar flow combustion chamber is what appears to the eye as a wide flarum front, but which apparently consists of a series of "bulbs", moviar up the front. This type of flame front does not appear in the unconfined two-dimensional flame. Figure 9 is a shadowgraph of the confined V-flame and Figure 10 is a photograph of the same. Figure 11 is a shadowgraph of the unconfined two-dimensional V-flame and Figure 12 is a photograph of same. Figures 10 and 12 would indicate that the flame front of both the unoonflned and confined flames are of approximately the same width, but as the sadowgraphs show (Figures 9 and 11), the apparent width of the uletIted flame is due only to a folding of the flame surface into the laas due to cooling of the glass (see Figure 13) and is not du t t the bulbous type front of the eoflined flame. It should be notid&Bre that the striations an. the speckled appearance of the flowe hown in the shadowgraphs is due to the glass used for the chamber walls* This becomes evident if Figure 14 is compared with Figure 11. Flame Front lame Front ill i Walls Confined Flame Unconfined Two-dimensional Flame Figure 13 A cross section of the confined and unconfined flame showing the different type flame fents. Ref. 1 - Progress Report No. 4 Tfr- M29 tItv a ty of Michigan AAF Contract 1W3-038 ac 21100 I I I I

AERONAUTICAL RESEARCH CENTER -UNIVERSITY OF MICHIGAN U1MR-33 Page 16 A possible explanation for the wide type flame front of the confined flame is that it is the envelope of a series of "cylinders of combustion", formed by some means of discontinuous ignition. It was thought possible that the discontinuous ignition was due to vortices shedding off the flameholder with the Karmen vortex frequency. This does not seem to be the case, however, since three different size flameholders were tried and the frequency of the bulbs remained the sme.

VIEW OF EOTTO N ALL PIPE.. F -1 INCS5 TO w rTHI^: PT ARE FOR 30001* SERSlICE VtERTICA.L SE C-TION OOWt4 -TO S U POR TTINc I-taUCTUR C TE AT BOTTOM 4 i= BLOWOFF FLAIGCE. Gm 4 -4,_JL_.T.,11. I DRAIN Zi rOSTCR REGUU^TOR 0-300 PSI GAGRE W5y5 CM -E Z,-, HEAT EXCH AN GER VERTICAL SECTION -ro FLOOR SUPPORTi Nr4 STR )CTURE ENGINE - UN IT \ FEIGURE 18 5CHEMAT1C DIAGRAM OF BLOWDOWN SYSTEM m Ern

3flOTB )008 B'HO3RJU.H3OM R.HOABA KIA ttI-et G..Y.",.00 Ra31~* l JAllUXln.baiajrn9a ae"nl (ia *donsi A sriU ofr x 30f.oit o x 7 wnivaBTn'i A,U.0U I 30AM bid I~e

t::!::?:.'F ~,) ~., -C 0 0 C 4 M ~~,j

i~r ~~ 9H ak.~d ~ ~6 ~ I: 9 ' X 8 ~I I g~~~ X:gt:..cU, o:!,,~i.:i;(?i~i 9 SB ds.D tX~ 9f i>AS:X

it)~~r,:0 0 fgios JYa~'b 'ti. fi z Y ':11F~~ ~~~~~~~~~~~%i~i~1~iii~~iiii.~l

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN I LUR-33 Page 21 Detonation Final assembly of the five shock tube sections has been completed. (See Figures 15 and 16.) Figure 15 shows the detail of the downstream end of the shock tube, which includes the blowout flange and the metering system to obtain the desired fuel-air mixtures. figure 16 is an overall drawing showing complete assembly with the important details brought out. The assembly included inserting the sealing 0 rings and the neessary testing of the assembly for leaks. The piping to deliver fuel and air has been installed and tested under vacuum and under pressure. The valve housing and the slide valve which opens the high pressure reservoir to relatively low pressure shock tube test section has been installed and partially tested. The blowout membrane located at the extreme downstream end of the shock tube which allows for the ignited gases in the test section of the shock tube to blow out into the laboratory, has been installed. It has been decided that the fuel-air ratios in the shock tube be obtained by using partial pressures of air and gaseous fuel. As mentioned above, the neeessary piping and instrmentation has been installed and tested. Starting with air at one atmosphere in the shock tube downstream of the ralVe, the pressure will be reduced with a vacuum pump to some valve below atmospheric. After this reduced -pressure has been obtained in the shock tube, gaseous fuel will be injected until the total pressure again has reached one atmosphere. By regulating the partial pressure of the air and of the gaseous fuel, it will be possible to obtain any desired fuel-air ratio in the shock tube section. A

~:z::e ('4; wo0 (9:OF O i: '-:;:;ii

PACE 23 i FIG. 16 - COMPLETE ASSEFMBLY OF SHOCK TUBE

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN m I I,, - UfR-33 2Page 24 Blowdown Eui nt The foundation for the engine-oampressor units has been poured and vibration-insulated from the concrete platform for the campressor building, The compressor building has been designed and building materials received. The engine compressor units have been uncrated and mounted on the foundation. The Waukesha gasoline engines which drive the compressors were prepared for operation and performed satisfactorily. The air compressors, however, were not prepared for operation due to lack of compressor oil, which has been ordered. The final installation of propane lines running from the 1,000 gallon tank located south of the laboratory has been made. Information was received from the Waukesha Motor Company concerning the possibility of using propane for engine fuel. A propane-gasoline combination carburetor appears to be satisfactory for this purpose. The heat exchanger support stand was fabricated in the laboratory from steel pipe and is ready for the heat exchanger shell. The heat exchanger shell was found to have several drilled holes in it which had to be welded up before it could be hydraulically tested to 400 psi approximately; it is planned to operate the present heat exchanger to a maximum of 150 psi air pressure. At this point, the hydraulic tests are almost completed, after which the shell will be lined with fire brick and filled with Alumite grog. A cylindrical form for the refractory liner has been cut from an extra twelve-inch pipe and a sample refractory ring made. Some trouble may be experienced with contraction of the refractory in the twenty-foot length of the heat exchanger. It was suggested that the refractory be overlapped and joined so that any contraction cracks would not expose the shell to radiation from the hot bed. Due to the difficulty of cutting uniform grooves in the thin circular segments, it was decided to join them as originally planned and take a chance on contraction cracks. Considerable time was spent in discussion of the icing problem that may occur at the pressure-regulating valve during blowdown operation. The Naval Ordnance Plant in Chicago, Illinois, suggested that we heat the valve by external means if the problem occurred. It is planned that the condensed moisture-s to be drained from the storage tanks as they are being pumped up. (Since the amount of water vapor which can be contained by a given volume \f air is primarily a function of temperature, the air at 200 atmospheres storage pressure should contain no more water than an equal volume of air at one atmosphere pressure and the same temperature contains, provided the condensed moisture is drawn off. It is expected, however, that the icing problem will not be too severe. The high pressure piping system has been outlined in detail and sent out for cost estimates* To date one satisfactory bid has been received. Final selection of a pressure regulating valve is alo in process, awaiting information from a subsidiary supplier. I

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN I r I I I I I tI5R-33 Page 25 Experimental Techni ques and Instrumentation A new high voltage power supply using a power frequency (60 cycle) transformer has been completed and is being used in conjunction with a Liebessart spark gap to take shadowgraph exposures. Excellent results have been obtained with this equipment. The h-v power supply used previously consisted of a commercial 6,000 volt generator of the R-F type, charging a 2,000 volt, 8 Mfd. condenser with an ignition coil used as a trigger. This supply had a slow charging rate (about one minute) and was also unreliable, so a new one was built using an 8,000 volt 1 Mfd. condenser to supply the spark. This condenser is charged to maximum in about one half second with the new power supply. A few pictures were taken with the spark arcing abross an unconfined air gap behind a.040" diameter hole. Because of the large amount of light loss with this arrangement, the spark had to be fairly close to the subject so that considerable detail was lost (see Figure 7). In order to improve the spark, an electrode arrangement suggested by the principle of the Liebessart gap was tried. The light from the partially confined arc emerges from its longitudinal axis through two.040" diameter holes properly aligned. With this arrangement the spark source could be eight feet from the subject so that the properly aligned.040" diameter holes illuminated approximately a 6" circle at the plane of the subject. This light was more nearly parallel than formerly and still of sufficient brightness. Of the several kinds of film used the Contrast Process Panchrcmatic has been the most satisfactory. Considerations have been given to the Kerr Cell and its possible use as a photographic shutter for use in taking high speed motion and still pictures. The Kerr Cell seems to offer considerable possibilities along this line, and further study will be given to it. Thought is also being given to the instrumentation of the shock tube which is nearing completion. There are several alternative methods that may give results. One method which appears promising is to use photoelectric cells to pick up the shock passage and record the output on a linear time base, such as the sweep of an oscilloscope. Specifications for the interferometer optical system have been received and the requisition for it has been issued. Pending approval of Wright Field an order will be issued to the Optron Laboratories, Dayton, Ohio, for the manufacture of the optical system. Bids for the manufacture of the frame and suspension system have been received and are currently being processed. The photographic equipmenthas also been ordered.. m -- --

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN 0 _ i ii II i II iii LR-33 Page 26 VI. Program Planned for Period 7 (1 July to 1 September, 1949) Blowoff Velocities of Flameholders Experiments to determine combustion blowoff' velocities under vacuum will be initiated. Pressure traverses in the flame region will be made. The shadowgraph method will be further extensively used in the vacuum blowoff and other experiments. A study to determine possible relations between flame ionization and other parameters will be undertaken. Combustion Chamber Design Efforts will be made to obtain full combustion chamber length spark shadowgraphs in a sequence rapid enough to completely cover one explosion in 20-30 frames. Pictures will also be taken of the propane-air mixing. Work is also to be completed on the effect of imposing rotation on the air at the entrance to a combustion chamber using annular flameholder. Pressure and Temperature Effects on Combustion Processes Installation of the new pressure temperature setup will be copleted. Pressure effects on flame speed and d, as outlined in IMR-31, will be conducted for a series of fuel-air ratios from.055 to.090. An attempt will be made to establish an agreement between flame speeds measured by the V-flame method and the flame speeds measured by the Bunsen flame method before the V-flame will be used in the proposed pressure and temperature studies. Flow Associated with the V-flame An attempt will be made to determine the cause of tbh bulbous type flame front of the confined flame. Detonation The mechanism to actuate the slide valve will be completed and the reservoir assembly will be tested with the valve and the shock tube test section assembled. The necessary timing circuit will probably be fabricated during the coming period. The design of a timing circuit applicable to the shock tube will be undertake. Blowdown Rqui ment It is expected that the heat exchanger will be lined with refractory, filled with Alumite grog, and made ready for operation. The high pressure piping will have been purchased and partially installed. The pressure regulating valve will have been ordered. The 3,500 psi air storage tanks should also be ready for operation and ready for testing; this includes the final machining of the top and bottom flanges for the tanks. It is further expected that the 3,000 psi air compressors dl have been prepared satisfactorily and ready to deliver air at high lpreM mN. The compressor I building is expected to be ecmpleted witht a w E's time. I

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN UIMR-33 Page 27 Experimental Techniqeues and Instrumentation Further improvements in the shadowgraph technique will be attempted. An initial design of a Kerr cell will be made and construction of it will be initiated. A study of the instrumentation problems for the shock tube and the blowdown tunnel will be continued. Same of the components for the interferometer and schlieren system will be constructed.

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN IaR-33 Page 28 ACTIVITIES VISITED Activities Visited Subject Discussed A. 0. Smith Company, Milwaukee, Wisconsin I. A. S. Meeting, New York, New York University of Washington, Seattle, Wash. Boeing Airplane Co., Seattle, Wash. N.A.C.A, - Ames Laboratory University of Southern California California Institute of Technology Jet Propulsion Laboratory High Pressure System International Meeting Combustion under Vacuum Detonation Ram Jet Burners Electronic Techniques for Detonation Studies Ram Jet Combustion Low Pressure Combustion Turbulent Flames

AERONAUTICAL RESEARCH CENTER - UNIVERSITY OF MICHIGAN UiR-33 Page 29 DISTRIBUTI CN 7 Black and White Copies 1 Tracing (Opaque Ribbon) TO: Commanding General Air Materiel Counamnd Wright-Patterson Air Force Base Dayton, Ohio 1 - - ~ --