Document
NASA
CONTRACTOR REPORT LOAN COPY: RETURN TO AFWL (WLIL-2) KIRTLAND AFB, N MEX
EFFECTS OF PERIODIC BLOWING
THROUGH FLUSH TRANSVERSE SLOTS
ON TURBULENT BOUNDARY LAYER
SKIN FRICTION
by Jack G. S$zngler
Prepured by a,’ LING-TEMCO-VOUGHT, INC. :I I 1.
Dallas, Texas Or
f
NATIONAL AERONAUTICSAND SPACEADMINISTRATION . WASHINGTON, D. C. . OCTOBER1966 TECH LIBRARY KAFB, NM 0060257 NASA CR-634 EFFECTS OF PERIODIC BLOWING THROUGH FLUSH TRANSVERSE SLOTS ON TURBULENT BOUNDARY LAYER SKIN FRICTION By Jack G. Spangler Distribution of this report is provided in the interest of information exchange. Responsibility for the contents resides in the author or organization that prepared it.
Prepared under Contract No. NASw-956 by LING-TEMCO-VOUGHT, INC.
Dallas, Texas for NATIONAL AERONAUTICS AND SPACE ADMINISTRATION For sale by the Clearinghouse for Federal Scientific and Technical Information Springfield, Virginia 22151 - Price 62.50 EFFECTS OF PERIODIC BLOWINGTHROUGH FLUSH TRANSVERSE SLOTS ON TURBULENT BOUNDARY LAYER SKIN FRICTION by Jack G. Spangler SUMMARY As part of an experimental investigation of techniques for reducing turbulent skin friction drag, small transverse disturbances were intro- duced into a turbulent boundary layer. The boundary layer was formed on the test section wall of a facility designed especially for low-speed boundary layer studies. Periodic jets of air were injected into the boundary layer from a continuous circumferential slot in the test sec- tion wall resulting in reduced values of skin friction drag and boundary layer turbulence. Skin friction measurements, velocity profiles, and turbulence intensity measurements are presented as a function of blowing rate, pulse frequency, and location in the flow field. The experiments show an interesting relation between turbulence intensity and skin fric- tion drag, but no promise is shown for realizing a net drag reduction due to the power requirements of the particular system.
INTROlXJCTION The turbulent boundary layer has long posed a problem to aero- dynsmicists and others concerned with the performance of vehicles op- erating in a viscous medium. The high skin friction drag and heat transfer rates in turbulent flow, relative to those in laminar flow, are sufficiently detrimental to vehicle performance to have justified the interest of numerous investigators in searching for ways to main- tain a fully laminar boundary layer. Unfortunately, in spite of these efforts, most practical situations involve the eventual occurrence of turbulent flow, and consequently the engineer or scientist is con- stantly trying to find methods for minimizing the undesirable char- acteristics of the turbulent boundary layer.
The Aerophysics Group of the LTV Research Center has been con- ducting a study of the effects of introducing discrete vortices into a fully developed turbulent boundary layer.
The initial effort in this study was an investigation of the effects of continuous longi- tudinal vortices , particularly on the skin friction drag.
The results of these experiments have been reported in reference 1.
This report deals with the results to date of a study on the effects of periodically occurring transverse vortices in a fully developed turbulent boundary layer.
Interest in this study arose from the results of an investigation by Eggers and Hermach (reference 2) in which they concluded that reductions in turbulent heat transfer rate by as much as 22 percent on a supersonic body were due to the periodic passage through the boundary layer of large scale vertical motions. Assuming the validity of the analogy between heat transfer and momentum transfer to hold for this case, the results of reference 2 imply that similar reductions in skin friction drag were effected although they were not measured or reported.
Based on these findings, an experiment was designed wherein trans- verse vortices would be introduced into a turbulent boundary layer formed in the boundary layer channel facility of the LTV Research Center.
If skin friction reductions were accomplished, the results would be analyzed to determine the mechanism involved.
Analyses would also be made to determine if the energy consumed in generating the vortices would be small enough to allow a net reduction in the system drag to be realized.
SYMBOLS A test section surface area lx average skin friction coefficient, 2 To/q dX cf $ change in Cf due to blowing ACf equivalent increase in Cf due to power requirements ncf P P power consumed in blowing pressure differential across blowing slot AP volume flow rate of blowing Q stream dynamic pressure, 3oUm2 cl t time u local mean velocity rms value of longitudinal velocity fluctuation IA' Rm local free stream velocity UC3 friction velocity, .T& UT I- P hot-wire signal V x longitudinal distance downstream of blowing slot distance normal to wall Y T wall shear stress with blowing OB 7 wall shear stress with no blowing ONI percentage change in 'r. due to blowing, 100 (TV - 'co )/T~ A7O B NB NB fn pulse frequency frequency of turbulent velocity fluctuations f )c bandwidth V kinematic viscosity density P MPERIMEXVTAL FACILITY AND INSTRUMEXCATION A. BOUNDARYLAYERCRANNEL The basic experimental facility used for the vortex-boundary layer experiments is an open-circuit wind tunnel designed for study of the boundary layer produced on the test section wall. A complete description of this facility (shown by sketch in Figure 1) may be found in references 1 and 3 and only a brief review will be given here.
The test section of the boundary layer channel is 25 feet long with an inside diameter of 8.030 inches. The test section is an as- sembly of six transparent Plexiglas tubes bored, honed, and polished to an overall tolerance in diameter of f .OOl-inch and a wall waviness of .OOl-inches per inch of length. Air is drawn through the channel by a Roots-Connersville vacuum blower isolated from the test section by a choked sonic throat. Dsmping screens and a x):1 area ratio entrance nozzle reduce the turbulence level of the incoming air to a very low value. For turbulent flow studies one section of the channel is re- placed by an assembly of interlocking rings which can be interchanged to allow various instrumentation to be placed at any azimuthal or longitudinal location in the channel.
B. SKIN FRICTION BALANCE The primary objective of this study was to determine the ef- fects of periodic transverse vortices on the turbulent,skin friction.
Thus it was desirable to be able to measure skin friction values directly and accurately. At the beginning of this program a skin friction balance was designed and built for this purpose. This instru- ment is described in detail in reference 4. Briefly, the skin friction balance is of the floating element type with an interacting electrical current and magnetic field providing a restoring force which may be controlled to null out the skin friction drag force.
The drag force is thus a linear function of the electrical current required to null the balance. A schematic of the balance is shown in Figure 2.
This -6 instrument is capable of measuring wall shear stress values from 10 to lO-4 lb/in* with an accuracy of +l percent for turbulent boundary layer flows.
C. HOT-WIRE ANEMCMETER Measurements of time-averaged mean velocities and fluctuating velocity components were made with constant-temperature hot-wire anemo- These instruments are transistorized circuits of the Kovasznay meters.
type. Output linearizers for the hot-wire signals were available but were not used because of unsatisfactory stability. The remainder of the circuit includingthe bridge and amplifier were quite stable and gave very repeatable results. All of the hot-wire probes were made of .0002- inch diameter tungsten wire spot welded to sewing needle supports.
D. PRESsuREMEAS~ All pressure measurements were made with a Transonics Type 121 Equibar Pressure Meter which employs a fast response differential capacitance transducer. This instrument was used along with appropriate probes for determining channel flow rates, blowing chamber pressures, and calibrating all hot-wire probes.
PRELIMINARY EXPERIMENTS A. FORMATIONOF TRANSVERSE VORTICES In the quest for a means of producing transverse vortices, the best thought that came to mind was to simulate the action that would be created by a transverse fence mounted perpendi.cularly on the test section In this way wall that could be alternately extended and retracted.
transverse vortices would be formed behind the fence when it was ex- tended and would shed and be swept downstream when the fence was re- tracted. Because of the axi-symmetrical configuration of the boundary layer channel, a mechanical fence would have to be quite complex and any mechanical system would undoubtedly cause intolerable vibrations.
Consequently a system was devised that would simulate the action of a fence by the periodic injection into the boundary layer of a thin jet of air through a continuous circumferential slot in the test section wall. It was hoped that the momentary pulses of air would interact with the vorticity in the boundary layer to cause the periodic formation of It should be pointed out that in the axi-symmetric transverse vortices.
boundary layer channel transverse vortices are actually continuous rings which, as will be shown later, are unstable and frequently add compli- cations to the experiments.
As a first step in the development of an air pulsing system, a pilot model chamber was built. Compressed air was fed into the chamber through a throttling valve and a rotating "on-off" valve which gave a ratio of flow on to flow off of 1:g. The rotary valve was belt driven by a variable speed A-C motor. The chamber was a rectangular box with inside dimensions which could be altered by the arrangement of duxmnywalls. A slot was provided in one exterior wall.
The length and width of the slot could also be varied.
A hot wire probe was placed in the chamber slot to monitor the pulse form of the emerging air jet. After a series of tests, a suitable combination of chamber volume and slot size was obtained which gave a single sharp air pulse that was uniform along the length of the slot for each half cycle of the rotary valve.
The proportions determined in the rectangular box were in- corporated into an annular chamber which was co-axially mounted in the boundary layer channel. Different sets of lips for the circumferential slot allowed the pulsed air jet to be directed either normal to the or tangentially downstream in the boundary wall, tangentially upstream, The incoming air to the system was regulated to 80 psig which layer.
kept the metering needle valve choked. The system was calibrated so that the total mass injection into the boundary layer was known at all Originally times. A schematic of the system is shown in Figure 3.
the pulses from the rotary valve were introduced into the annular but a check showed uniformity of the plenum chamber at one location, jet around the circumference of the test section to be unsatisfactory.
Subsequently a manifold was incorporated to distribute the air to four equally spaced inlets to the plenum chamber which gave acceptable results. Once the uniformity of the pulse shape was established two fixed hot-wire probes were permanently mounted inside the chamber slot to allow continuous monitoring of the periodic jet flow.
INITIAL TESTS B.
The first measurements of the effects of the periodic blowing into the turbulent boundary layer revealed a multitude of resonance problems in the system. Since the boundary layer channel is essential- ly a long straight tube, it has a fundamental "organ pipe" frequency and harmonics. The fundamental frequency is about 10 cycles per second. When the injected air was pulsed at frequencies less than 100 cycles per second the tube resonated at the expected frequencies.
This was a potential problem since the experiments were planned for this frequency range. The reasoning here was that to simulate the vortex spacing encountered in reference 2 the injected air should be pulsed at about 20 cycles per second.
Also, considering the spectral distribution of the energy of fluctuations in a turbulent boundary layer, almost all of the energy, at the test conditions used, occurred at frequencies below 100 cycles per second.
It was somewhat surprising to find that initial measurements with the skin friction balance showed pulsing frequencies less than 200 cycles per second to be ineffective on the skin friction and frequencies in the neighborhood of 300 cycles per second were necessary to get significant changes. Thus the organ pipe resonance of the channel was not too important but two other problems came to light.
The enclosure of the skin friction balance was found to respond as a Helmholtz resonator when air was pulsed into the boundary layer channel. Consequently the mass of air in the clearance space around the floating element would oscillate. The configuration was such that partial rectification of this oscillatory flow occurred causing a re- sultant force to act normal to the floating element and thus introduced an error into the balance reading. Several modifications were made to the balance to correct this situation. The most important modification was to mount the balance at the bottom of the boundary layer channel and eliminate the internal air volume by filling the entire enclosure with oil. This essentially eliminated resonance problems with the balance but it restricted the placement of the balance to the bottom of the channel and made handling and operation of the instrument con- siderably more tedious.
The other resonance problem concerned the annular plenum chamber. It was found that the chamberand slot configuration also behaved as a Helmholtz resonator with certain resonance frequencies that strongly affected the pulsed air jet. More will be said about this in the discussion of the results.
C. INITIAL RESULTS Skin friction measurements were made at several locations downstream from the slot for numerous settings of blowing rate and pulse frequency. The first measurements were made for the case of blowing normalto the wall which was expected to be the best method for producing transverse vortices and reducing friction drag. To the contrary, no reductions in skin friction were found and in some in- stances slight increases were noticed. Tangential blowing was then tried for situations with the jet pulses being directed along the wall either upstream or downstream. Again the upstream blowing caused no favorable effects, but the downstream blowing created local skin fric- tion reductions that were dependent on both blowing rate and pulse frequency and were as large as 20 percent in some instances.
As mentioned previously, the plenum chamber was designed for operation at frequencies under 100 cycles per second but significant drag reductions required pulse rates on the order of 300 pulses per second. The hot-wire probes mounted in the slot showed that the pulse form was quite different at these frequencies from what had been de- It seemed logical that refining the chamber to give single sired.
sharp pulses in the high frequency range rather than ill defined pulses with considerable harmonic content would improve the effect on skin friction. This was accomplished by altering the internal volume of the chamber and adding packing material for damping. A reasonable pulse form was obtained but the maximum jet velocities were reduced and the best possible local skin friction reductions were then only 7 percent. A compromise setup was finally reached by removing the packing material. This left a somewhat irregular pulse form but skin friction reductions of 20 percent were once more possible.
While trying to determine the variation of skin friction with frequency it was found that in each of several narrow bsndwidths the balance would show large increases in the measured drag force. With the earlier resonance problems still in mind it was suspected that these anomalies were again due to balance resonance. However, a hot- wire probe placed in the boundary layer just above the floating balance element showed that the local velocity also increased sharply when the measured drag increased and likewise the velocity decreased as the drag decreased. The two instruments checked qualitatively and quantitatively verifying that the anomalies were actually in the flow.
Inspection of the signal from the hot-wires in the slot showed that the jet velocities became much larger at these anomalous points in the frequency range even though the total mass flow through the blowing system was held constant. Also the apparent frequency of the It was concluded that these large pulses was doubled at these points.
were resonance points in the plenum chamber and the result was a strong oscillatory flow in and out of the slot superimposed on the regular pulse flow. The hot-wires, being equally sensitive to flow in or out of the slot acted as rectifiers indicating twice the true pulse fre- quency. The result was a very strong jet pulse outward followed by a The difference in the two mass flows was slightly weaker pulse inward.
the amount being let into the chamber through the calibrated flow system.
This was verified by integrating the pulse profile over a period of one cycle, assuming the slightly weaker pulses to be in the inward direction.
Agreement with the measured mass input was good. Further inspection of the hot-wire signals showed that this phenomenon occurred to a lesser extent over most of the frequency range above 200 pulses per second.
The integrated difference between the inward and outward pulses always agreed reasonably well with the measured net flow rate and the individual pulses became very large only at the aforementioned resonance frequencies.
Up to this time no information had been obtained concerning the action in the boundary layer caused by the jet pulses, i.e., were vortices being formed or not? Signals from hot-wire probes did not show any regu- lar disturbance pattern in the boundary layer except within a region ex- tending about two inches downstream from the slot.
In this region very small oscillations appeared in the signal when the probe was near the The signal generally had the form of a rectified sine wave, but wall.
at the resonance frequencies of the plenum chamber the signal broke up into strong irregular fluctuations typical of turbulent bursts.
At this point an attempt was made to render the jet pulses visible. It will be recalled that the test section of the boundary layer channel is completely transparent except for the plenum chamber and the mounting rings for the skin friction balance and hot-wire tra- verse. Also a collimated, high-intensity light beam can be directed down the tube axis through a window in the transition section near the sonic throat. This allows flow patterns to be observed visually by - injecting smoke into the air stream. Due to the rapid diffusive action of the turbulent boundary layer nothing could be learned about the action of the jet pulses by flooding the boundary layer with smoke. Instead smoke was injected into the plenum chamber so that each pulse that came out was actually a puff of smoke. A mechanical chopper was used to interrupt the light beam to achieve a strobe effect. With this set up, the emerging jet pulses could be seen but they were very faint.
Several interesting discoveries came out of the flow visualiza- tion studies. First and most surprising was the size of the transverse disturbances. Vortices of the same order in size as the boundary layer thickness were desired, but with downstream tangential blowing the only distinct transverse disturbances that could be seen were very small, ap- pearing to be approximately l/10 to l/8-inch in thickness, and occurring only for particular values of blowing rate and pulse frequency. These disturbances could be seen originating at the blowing slot and continu- ing downstream a distance of two to three inches. At this distance they faded out and it could not be determined conclusively whether the disturbances died out or the smoke just became too diffused to see any more. The disturbances were very uniform around the entire channel circumference and appeared to be stable. The distance of the distur- bances from the wall seemed to be approximately l/lo-inch. It was also noted that the occurrence of these small disturbances coincided with reductions in skin friction as indicated by the balance.
Similar tests with other blowing rates and frequencies gave very different results. Instead of transverse disturbances being shown by the smoke, various longitudinal patterns appeared. In most cases these patterns seemed to consist of longitudinal vortices of various strengths depending on the amount of blowing and the pulse frequency. In some cases the vortex streamers lay right on the wall originating at the slot.
In other cases the longitudinal vortices could be seen to arise out of the break-up of an initially transverse disturbance. In general the flow pattern indicated an instability in the transverse disturbances which led to the formation of longitudinal vortices.
These patterns were somewhat difficult to identify at the normal channel flow rates, but by slowing the channel down to very low flow rates the action was quite clear.
When the pulsing frequency was set at one of the previously mentioned resonance points of the plenum chamber, the smoke showed very large disturbances coming out of the slot which broke up into large longitudinal vortices and regions of strong turbulence. It was this action that was responsible for the anomalous high skin friction drag readings for these frequencies.
Further attempts were made to identify the nature of the small, well defined transverse disturbances. The smoke had shown them to be uniform and stable but no definite vortex rotation could be seen due to the small size. A hot-wire probe was placed in the flow such that the disturbance centers passed over it. The probe signal showed small double pips each time a disturbance passed.
This would be indicative of a vortex since the hot-wire should show a signal increase as the leading and trailing edges of the vortex passed it. Also the mean velocities measured by the hot-wire showed a slightly higher value in the upper region of the disturbances and a slightly lower value in the lower region as compared to the case of no disturbances. This result agrees with the concept of a vortex in a steady flow rotating in a direction such as to These observations all suggest that effectively be rolling downstream.
the small transverse disturbances were vortices but this has still not been proved conclusively.
Although the cases of blowing normal to the wall and blowing tangentially upstream were found to give no benefits and were eliminated from the final investigations, one additional comment should be made about them. While the apparatus was set up for flow visualization studies, these two cases were examined with smoke and hot-wires for periodic disturbances in the same manner as described for the down- No discrete periodic disturbances were found at any stream blowing.
conditions for the upstream blowing.. For the normal blowing a few combinations of blowing rate and frequency were found which caused the appearance of periodic disturbances very similar in all respects to those that were formed by the downstream blowing. However, the dis- turbances for normal blowing were always about one half of an inch off of the wall which was well outside of the viscous sublayer region. The fact that these disturbances had no effect on skin friction, despite their similarity to the ones near the wall, suggests that it is es- sential that the interaction mechanism of the disturbances with the mean flow occur near the wall.
FINALEXPERIMENTS TEST ENVIRONMENT A.
Wing the preliminary experiments, the effects of the periodic blowing on skin friction drag were investigated for several different The largest percentage re- flow rates in the boundary layer channel.
ductions in skin friction occurred over the middle portion of the Due to the large number of variables which channel operating range.
seemed to affect the results it was decided to fix as many parameters as possible in order to keep the scope of the experiments within reason.
Consequently a single channel flow rate was arbitrarily picked and maiy- tained during all further testing. !Ihis particular flow rate (4.0 ft. / set) provided a fully developed turbulent boundary layer with a locsl thickness and free stream velocity at the slot station of 2.5 inches and 13.8 feet per second respectively.
SKIN FRICTION DISTRIBUTIONS B.
The first detailed investigation was the determination of the skin friction drag as a function of position, blowing rate, and pulse frequency. hta were taken with the skin friction balance at stations and 26.0 inches downstream from the blowing slot.
3.3, 5-3, 6-9, 15.7, Two sets of data were taken at each station; one set for a fixed pulse frequency with various blowing rates and the other set for a fixed blowing rate and variable pulse frequency.
In each case the wall shear stress was measured with the skin friction balance, the pressure drop across the blowing slot was measured, and oscilloscope records were made of the hot-wire signal in the slot. The shear stress data for the first four stations are shown in Figures 4 and 5 plotted as percentage change in shear stress based on the local shear stress for the case of no blow- ing. Negative values indicate shear stress reductions and positive values indicate shear stress increases. !l?ime averaged values of the pressure differential across the slot are also shown in Figure 5. The pressure differential is the difference between the average values of the plenum chamber pressure and the free stream static pressure. More will be said about the pressure differential in the discussion of power requirements.
The data in Figure 4 were obtained by tuning the pulse frequency to give a good skin friction reduction at a nominal blowing rate and then holding this frequency (340 pps) and varying the blowing rate over its The effect is similar at each of the measuring stations with full range.
the maximum reduction shifting to higher blowing rates as the distance from the slot increases. It may be seen that the maximum reduction in skin friction occurs in the middle portion of the test region of the boundary layer channel. The maximum reduction increases from 16% at the 3.3-inch station to 20% at the middle stations, dropping off to 6.54 at the 15.7-inch station. No data are shown for the 26.0-inch sta- tion as there was no measurable effect persisting this far downstream.
The data in Figure 5 were taken at a fixed blowing rate of 26.4 x 10m3 ft3/sec. The frequency range from 100 to 370 pulses per second was The data show that the skin friction varies considerably with covered.
frequency. The primary reason for this is because of the change in the pulse with frequency. One exception is a sharp drag reduction indicated near 210 pulses per second which was caused by a resonant condition in the balance due to a low oil level. It was confirmed that this particular effect did not exist in the flow and should be neglected.
The shape of the pulse does not seem to be as critical as the magnitude of the jet velocity and, as mentioned previously, the velocity is very sensitive to resonance conditions in the plenum chamber.
For instance a large skin friction increase occurs at the forward stations between 220 and 280 pulses per second. The pressure differential across the slot also shows a local peak. Several oscilloscope pictures of the signal from the hot- wire in the slot are shown in Figure 6 which help to explain this situa- tion. It should be kept in mind that the scope shows the hot-wire voltage which is approximately proportional to the one-fourth power of velocity.
The center of the grid is the base line for zero velocity and velocity increases downward in the pictures. The scope does not show the signal returning to the base line when the flow is reversed because of the frequency cut-off characteristics of the hot-wire amplifier.
Picture (a) in Figure 6 shows the pulse form at a frequency of 190 pulses per second. Two pulses are shown for each cycle indicating that the flow in the slot was oscillating. The stronger pulse is the outward flow with a velocity of 29.2 feet per second.
Picture (b) was taken at a frequency of JO0 pulses per second and in general is similar to picture (a). The outward flow velocity in (b) is 2‘7.2 feet per second. Pictures (c) and (d) are at 240 and 270 pulses per second respectively. The outward velocities are 53.7 and 45.2 feet per second even though the blowing rate is the same as in (a) and (b). Referring to Figure 5 the skin friction is seen to be increased by approximately 30s in this region. By reducing the blowing rate to 9.6 x 10-3 ft3/sec and repeating the measurements at 240 and 2'70 pulses per second, the skin friction was found to decrease. The shear stress and pressure dif- ferential data for this blowing rate are also shown in Figure 5 and identified by *. Figure 6, (e) and (f) show the hot-wire signal for the reduced blowing rate. The velocities are now 32.9 and 29.2 feet per second respectively, the shear stress is reduced by lo%, and the .038 to .014 millimeters of pressure differential is decreased from mercury. This indicates that driving the plenum chamber at a resonant point gives a more efficient operation but more will be said about this in the discussion of power requirements.
Some of the data from Figures 4 and 5 are combined and shown in Figure 7 to illustrate the shear stress distribution downstream from the slot at various blowing rates. Unfortunately the skin friction balance could not be placed closer to the slot than the 3.3-inch sta- The lower blowing rates show the skin friction increasing from tion.
A minimum may occur in these curves at a this station on downstream.
station closer to the slot but this is only speculation. The curves for the higher blowing rates show a progressively higher skin friction at the forward station with a minimum occurring near the 6.0-inch sta- tion. The skin friction increases from this point downstream until the effect of the blowing vanishes for all conditions at the 26.0-inch sta- tion.
The total reduction in skin friction drag at any blowing rate may be found by integrating the shear stress curves in Figure 7 over the surface area of the boundary layer channel in which the reductions exist. To do this, some assumption must be made about the shear stress in the region between the slot and the 3.3-inch station. The data imply that the shear stress is actually increased in this region but for evaluation purposes the curves for the lower blowing rates are as- sumed to drop off to zero effect at the slot. If anything, this gives a liberal estimate of the total skin friction reduction for these blow- ing rates.
A blowing rate of 26.4 x low3 cubic feet per second at 3h.Q pulses per second was picked for detailed studies of the mechanism in- volved in the skin friction reduction. The data for this blowing rate give an average skin friction reduction of 6.1% when integrated in the manner described above.
C. POWER RJQUIREMENTS One of the questions that comes to mind when a technique is found for reducing drag concerns the penalty that must be paid. This penalty can be thought of in terms of an effective drag increase by converting the power consumed in the system to an equivalent drag as follows. Since the type of drag in question is skin friction drag, it is convenient to work with the skin friction coefficient given as 0 avg cf = P u2 co The net skin friction coefficient may be expressed as - ACf + ACf 'fnet = 'f P The percentage change in net drag may be indicated by re-writing equation (2) C AC, Acf fnetl-w + 1, (3).
cf cf cf The power consumed is given by P=QAP (4).
Dividing the power by the free stream velocity gives the effective drag and normalizing this by the product of free stream dynamic pres- sure and surface area gives the equivalent drag coefficient for the power. Thus P AC, = 2 e - c (5).
qA WJ m P There is one problem in using this analysis and that is knowing the correct value for the pressure differential.
In these tests, it was fairly easy to measure the time averaged pressure but nothing was known about the actual pressure level during the 10s of the cycle that the blowing takes place. A conservative estimate of the power requirement may be made by using the average pressure differential.
If this tech- nique does not show a net drag decrease then there is no need to look further because the actual pressure differential will cause an even higher net drag.
Application of equation (3) to all of the data shows that no net drag reduction was realized at any condition except the previously mentioned plenum chamber resonant point with reduced blowing.
The re- sult for this case was a net drag decrease of 4% based on the average pressure differential. The fact that the net drag was decreased justi- fied a closer look at the actual blowing power being used. An attempt was made to determine the damping characteristics of the plenum chamber so that the true power required to drive it at its resonant frequency could be calculated. The results showed the damping rate to be so large as to require a lengthy investigation.
The feeling at this point was that it would be of more value to study the mechanism involved in the skin friction reduction than to try to optimize the system.
It should be emphasized that any effort toward developing a practical system for drag reduction would necessitate a much more thorough study of the technique for producing the disturbances.
I2 D. MEAN VFILOCITY MlWSIJREMENTs The skin friction data have shown conclusively that periodic blowing in the manner described reduces the wall shear stress in a The problem now becomes one of determining turbulent boundary layer.
A change in wall shear stress requires a how this is accomplished.
corresponding change in the shape of the velocity profile. To find what changes were effected by the periodic blowing mean velocity pro- files were measured at the 3.3, 6.9, and 15.7-inch stations with and A single element hot-wire probe was used for all without blowing.
measurements. These data are shown plotted in the universal law of the wall form in Figure 8. The solid curve in the figure shows the empirical profile expression derived by Coles (reference 5) for in- compressible flow and widely used as a standard for turbulent boundary The data for no blowing are in good agreement with the em- layers.
pirical curve.
The velocity profiles with blowing (26.4 x 10m3 cubic feet per second, 340 pulses per second) show two interesting features. The sublayer region appears to be extended by an amount related to the local shear stress reduction, and the turbulent region appears to have the same shape and slope as the standard turbulent profile. This im- plies that the sublayer is effectively thickened without the outer turbulent layer being affected. This result does not agree with the effects that would be expected if the skin friction reduction were due to large scale vortex-boundary layer interactions as hypothesized in the beginning of the study.
E. TURBUIBNCEEFFECTS The fact that the mean velocity profiles were affected only in the sublayer stimulated interest in the behavior of the turbulent of the boundary layer.
structure Since the sublayer is actually a region in which turbulent velocity fluctuations are damped, any change in its thickness indicates a corresponding change in the turbulent energy structure of the boundary layer.
The boundary layer turbulence was examined by measuring spectral distributions and integrated totals of the turbulent, u', velocity fluctuations. The data were measured with the same hot- wire probe used for the mean velocity measurements. Spectral distri- butions were obtained using a Hewlett-Packard 302A wave analyzer as a notch filter and recording its output on a Flow Corporation 12Al random signal voltmeter. The 12Al was used directly to measure average totals of the turbulent, u', fluctuations. These data are presented in Figures 9 through 14.
Figures 9, 11, and 13 compare the root-mean-square values of the longitudinal turbulence intensity across the flow field with and without the periodic blowing. The blowing data show a slight trend toward a reduced turbulence level in the inner portion of the boundary layer but in general no strong effect shows up.
Figures 10, 12, and 14 compare the spectral distribution of energy at various positions in the boundary layer the turbulent, u', The data are plotted as the square of the with and without blowing.
divided by the bandwidth over which they were rms value of, u', This bandwidth was 6.5 cycles per second for the 302A wave measured.
The data were presented in this form so that the spectral analyzer.
curves could be integrated to get the total energy for comparison with the values measured with the 12Al random signal voltmeter.
The turbulence spectra with blowing are consistently below the spectra without blowing in the inner region of the boundary layer, and the difference in the two diminishes with increasing distance down the channel. At the 3.3-inch station the turbulent energy reduction varies from approximately 25% at 20 cycles per second to 50% at 100 cycles per second. Beyond 150 cycles per second there is no signifi- cant effect. The turbulence reduction decreases with increasing Figure 10 shows some effect at 1.000 inch distance from the wall.
from the wall but the stronger effects are within .25O inches. Figure 12 shows the effect to be within .050 inches from the wall, and Figure 14 shows an appreciable effect only at .OlO inches. These data indi- cate that the periodic blowing has reduced the turbulence level through- out a significant portion of the boundary layer near the station where blowing is applied, and the turbulence intensifies from this reduced " level as the flow progresses down the channel until the intensity is back up to the values for no blowing.
A curious feature of these data is that the spectral distri- butions show a much larger reduction than the total, u', component of Also the integrated values of each spectra are Figures 9, 11, and 13.
consistently lower than the corresponding values measured directly with the random signal voltmeter. The probable reason for this discrepancy lies in the low frequency range of the energy spectra. The 12Al volt- meter responds to fluctuations down to a frequency of 2 cycles per second but the 3O2A wave analyzer is operable only down to 15 cycles per second. The fact that all integrated spectra are lower than the totals measured with the 12Al implies that large contributions to the total turbulent energy lie in the frequency range below 15 cycles per second. Also, the discrepancies in the turbulence reduction as indi- cated by the spectral curves and the total measurements suggest that a possible cross-over of the spectral curves with and without blowing occurs at some low frequency. In other words, although the turbulence level is consistently reduced above 20 cycles per second with periodic blowing it possibly is increased at lower frequencies. Further investi- gations are necessary to resolve this question.
F. EFFECTS OF CONTINUOUS BLOWING The data obtained for the case of periodic blowing into the turbulent boundary layer raise many questions and do not provide suf- ficient answers to determine conclusively the mechanism causing the skin friction reduction. The idea of large scale vortices causing a reduction in the slope of the velocity profile at the wall due to a simple super-position of flow fields is discounted because of the - A more extremely small observed size of the actual disturbances.
reasonable hypothesis is that the turbulence reduction is due to an altered mixing process in the boundary layer caused by the small disturbances.
It was known that the Reynolds number of the.per1odi.c flow in the slot was always below the critical value such that the jet pulses were lsminar in all cases. Thus each pulse was actually a small quantity of very low turbulence air which mixed into the turbulent flow of the boundary layer. !ihe net effect should have been a reduced turbulence level in the boundary layer as observed. The reduced tur- bulence level would then allow the sublayer to thicken and hence re- duce the shear stress at the wall. If this hypothesis is correct then any introduction of low turbulence air into the boundary layer should cause a similar effect. This is to say the periodicity of the blowing is not essential to the mechanism.
To test this idea some additional experiments were performed with continuous instead of periodic blowing. Skin friction reductions were again accomplished but only at higher blowing flow rates than were The 26.4 x 10-j cubic feet per necessary with the periodic blowing.
As the flow rate was increased the second flow rate had no effect.
skin friction decreased until local reductions as large as 24% were ob- tained. This occurred at a blowing rate of 643 x 10-3 cubic feet per second. As the blowing rate was further increased the skin friction increased until values higher than those for no blowing were finally reached.
Some of the data for steady blowing at 60 x 10m3 cubic feet per second are shown in Figures 15 and 16. Figure 15 again compares the velocity profiles in the law of the wall form. The data for con- tinuous blowing are seen to be quite similar to the data for periodic blowing shown in Figure 8. Figure 16 shows an example of the turbulence reduction at two continuous blowing rates. Again the results for the higher blowing rate are similar to the results in Figure 10 for periodic blowing. Notice that the steady blowing at 26.4 x 10-3 ftj/sec has no effect on the turbulence as compared with pulsing at the same blowing rate in Figure 10 (b). The power requirement for continuous blowing was found to be four to five times as high as for periodic blowing in order to achieve comparable skin friction reductions.
An investigation of similar effects has been conducted at NASA, Langley Research Center (references 6 and 7) for continuous blowing through slots in supersonic flow. The effectiveness of single and multiple slots, both flush and rearward facing steps, as compared to distributed blowing through a porous surface was studied. An exact comparison of the LTV and NASA data is difficult because of insufficient local skin friction measurements in the NASA tests. However, average skin friction reductions of the same order as reported here were found for some of the flush slot configurations. It is interesting to note that the NASA investigators also place emphasis on the exact mixing process of the injection air with the boundary layer in determining the skin friction.
CONCLUSIONS These experiments are by no means sufficient to completely describe the mechanism responsible for reducing skin friction drag. A thorough understanding of the phenomenon will require much more work. However, several conclusions may be drawn from the results presented herein.
A skin friction reduction has been accomplished by the intro- 1.
duction of low turbulence air into a turbulent boundary layer.
The fact that this is possible with either periodic or continuous blowing discredits the importance of any vertical motion for the size disturbances used in this study.
The turbulence intensity near the wall is reduced coincident 2.
with the skin friction. This effect, along with the effects on the law of the wall, indicate that the skin friction re- duction is due to a thickening of the viscous sublayer rather than a gross modification of the velocity profile as might be expected with large scale vortices such as in reference 2.
The power required to accomplish the skin friction reduction 3.
is prohibitively large to allow a net drag decrease except possibly for certain resonant conditions in the system. This suggests that further work toward developing a more efficient mechanical apparatus might be worthwhile.
4. The turbulence reduction is probably due to the mixing of low turbulence air with the turbulent flow in the boundary layer.
The fact that stronger effects occur with periodic rather than continuous blowing at a given blowing rate suggests that a more efficient mixing process is achieved with the former due to the higher jet velocities.
ACKNOWLEDCEMENTS The work reported herein is part of a project supported jointly by independent research and development funds of Ling-Temco-Vought, Inc., and the National Aeronautics and Space Administration under Contract No.
NASw-956 (Fluid Physics Branch of the Research Division, OART).
REFERENCES 1. Spangler, J. G., and Wells, C. S., %ffects of Spiral longitudinal Vortices on Turbulent Boundary Layer Skin Friction", NASA CR-145, (1964); also LTV Research Center Report No. 0-71000/4R-19 (1964).
2. Eggers, A. J., and Hermach, C. A., "Initial Experiments on the Aerodynamic Cooling Associated with Iarge-scale Vertical Motions in Supersonic Flow", NASA mf A54L13 (1955).
"A Facility for Basic Boundary 3. Wells, C. S., and Spangler, J. G., Layer Experiments", LTV Research Center Report No. 0-71000/2R-32.
(1962) l 4. Spangler, J. G., "A Sensitive Magnetic Balance for the Direct Measurement of Shin Ekiction Drag", IZV Research Center Report No.
0-7lW/3R-% (1963).
Cbles, D., "Measurements in the Boundary Iayer on a Smooth Flat Plate 5.
in Supersonic Flow", Ph.D. Thesis, California Institute of Technology, 1953.
6. McRee, D. I., Peterson, J. B., Jr., and Braslow, A. L., "Effect of Air Injection Through a Porous Surface and Through Slots on Turbulent Shin Friction at Mach 3”, NASA !CN Ik2427, August 1964.
Peterson, J. B., Jr., McRee, D. I., Adcock, J. B., and Braslow, A. L., 7.
"Further Investigation of Rffect of Air Injection Through Slots and Porous Surfaces on Flat-plate Turbulent Shin Friction at Mach 3”, NASA TN IL33ll, March 1966.
\ TRANSITION- SECTION TEST SECTION ONTRACTION SECTION ’ WWING SECTION -
NULL INDICATOR
(DIFFERENTIAL TRANSFORMER WITH FLOATING CORE)
r 2D.C. CONDUCTOR
MAGNETIC FIELD
ii ii/-
mea* FLUX LINES
FLEXURE PIVOT
ROTATING BEAM
I-
-r
1 FLOATING
ELEMENT
Schematic of skin friction balance.
FYgure2 - I -
u/////////‘II.-III---------
BOUNDARY LAYER CHANNEL ASSEMBLY VARIABLE SPEED / AC MOTOR ROTARY VALVE ASSEMBLY - tzE CALIBRATED / STILLING CHAMBER Periodic blowing apparatus.
P&u= 3
-8
-12
-16
16 24 32 40 48 56
Q x 103- ft3bc
Station 3.3 inches, co =
-lc 340 pps, T = 6.72 x 10 lb/ft2
(a) ONEl Variation of wall shear stress with Ngure 4 blowing rate.
-8
-12
-16
I -
-20
L
0 8 16 24 32 40 48
Q Xt03- ft%ec
= 6.68 x lO-4 lb/ft2 (b) Station 5.3 inches, w = 340 PPS, T ONB (Continued) Variation of wall shear Figure 4 stress with blowing rate.
-4
-8
-16
-20
-24
8 16 24 32 40 48 56
Q X 103-ft3/sec
(c) Station 6.9 inches, a, = 340 pps , ‘corn = 6.68 x 10N4 lb/ft2 Figure 4 (Continued) Variation of walJ. shear stress with blowing rate.
-I
-2
-5
-6
32 40 48
Q X IO3 - ft3/sec
\ (d) Station 15.7 inches, a, = 3b-pps, 7om = 6.59 x lo-' lb/ft' Figure4 (Concluded) Variation of wall shear stress with blowing rate.
.05
.I” .04
I6
E .03
L .02
t 8
Q .Ol t
Q
-8
O-Ar, III-AP
-16
I I 1 I I I -- I I I I I I I
-24
140 180 220
260 300
340 380
w -PPS
-4
(a) Station 3.3 inches, Q = 26.4 x 10q3 ft’/sec 9 b, = 6*72x lo
lblPt2 - Variation of wall shear stress and pressure derential Ntw-e 5 across blowing slot with pulse frequency.
.06
g .05
I6
; .04
L .03
Q .02 8
.OI
-
0 0
I
-8
I-O
a
-16
BALANCE RESONANCE
- ATo
-24
El -AP
I-
-32
220 260 300
140 180
u - PPS
(b) Station 5.3 inches, Q =I 26.4 x 10B3 ft3/sec, b, z 6.68 x 10B4 lb,ft2 (Continued) Variation of wall shear stress and pressure Ww= 5 differential across blowing slot with pulse frequency.
.06 8
i .02 -8
.OI
- -16
\o
9 -24
t-
BALANCE RESONANCE
a
t-
-32
0 -AT
340 380
too I40 I80 220 260 300
CA) - PPS
(c) Station 6.9 inches, Q = 26.4 x 10-3 ft3/sec, 7 -4 = 6.68 x 10 lb/n" ONB Continued) Variation of wall shear stress and pressure
Wwe 5
61 fferential across blowing slot with pulse frequency.
.06
.05
.04
.03
.02
.Ol
-4
-8
ru a3
-12
;p -16
BALANCE RESONANCE
a
O-AT0 o-AP
k
-20
loo
140 180
340 380
W - PPS
(d) Station 15.7 inches, Q = 26.4 x 10~~ ft3/sec, ho_- t 6.59 x lo-' lb/ft2 (Concluded) Variation of wall shear stress and-NJ3 Figure5 pressure differential across blowing slot with pulse frequency.
(b)
(a)
w =3OOPPS
w =19OPPS
Q = 26.4 X 10m3 f t3/sec Q = 26.4 X 10m3 fts/sec
tad = 240 PPS W =270 PPS
Q = 26.4 X 10s3 ft3/sec Q = 26.4 X 10m3 f?/sec
(e) (f)
W = 240 PPS w =270 PPS
Q = 9.6 X 10s3 f t3/sec
Q= 9.6 X low3 ft’/sec
Pulse forms as measured by a hot-wire probe in Figure 6 the blowing slot. Gain = 0.5 volts/cm, sweep = .002 set/an.
- DISTANCE FROM SLOT - INCHES
2 I 4 6 8 IO I2 I4 I6 I8 2
‘PS
%3u=7 longitudinal wall shear stress distributions.
A WITHOUT BLOWING
o WITH BLOWING
31;
5)
IO
Y" lb
(al F;ti;-:;3$ches, LD= 340 ppsL4Q = 26.4 x 10s3 ft3/sec, .
> T = 6.72 x 10 lb/ft?
OHB Figure8 bm of the Wdll velocity profile.
A WITHOUT BLOWING
0 WITH BLOWING
YU*
-T
(b) Station 6.9 inches, a~ = 340 pps, Q = 26.4 x 10m3 ft3/sec, Arom-8.ti, 'c = 6.68 x 1O-4 lb/ft: ONB Figure 8 (Continued) Law of the wall velocity profile.
A WITHOUT BLOWING
o WITH BLOWING
5)
Y" +
1)
Q = 26.4 x 10m3 ft3/sec, (c) Stati,; 2.7 inches, ol= 340 p--, ATE= . o,~ = 6.59 x lo lb/f-t: ONE (Concluded) Law of the wall velocity profile.
Figure 8 I
3.5 4.0
2.0
1.0 1.5
0.5
)/- INCHES
Iongitudinal turbulence intensity, station 3.3 inches.
Figure 9
4.8
3.2
o-WITH BLOWING
WITHOUT BLOWING
2.4
1.6
0.8
-
*
n
v v v
0 3
160 200
0 40 80 120
f -CPS
(a) Station 3.3 inches, y = 0.010 inches Longitudinal IxrlxiLence spectrum.
Figure 10
O-WITH BLOWING
WITHOUT BLOWING
r-
40 80
f - CPS
Station 3.3 inches, y = 0.025 inches (b) spectrum.
(Continued) Longitudinal turbulence Figure 10
40 80 120 160 200
f -CPS
(c) Station 3.3 inches, y = 0.050 inches Longitudinal turbulence spectrum.
Figure 10 (Continued)
O-WITH BLOWING
WITHOUT BLOWING
120 160 200
0 40 80
f -CPS
(d) Station 3.3 inches, y = 0.100 inches Figure 10 (Continued) Longitudinal turbulence spectrum.
m
0-W I TH BLOWING
-
WITHOUT BLOWING
0” I6
W
160 200
f -CPS
(e) Station 3.3 inches, y = 0.250 inches Figure 10 (Continued) Longitudinal turbulence spectrum.
O-WITH BLOWING
WITHOUT BLOWING
m
h
u
v
a
0 40 200
80 120
- CPS
f
(f) Station 3.3 inches, y = 0.500 inches Figure 10 (Continued) Iongitudinal turbulence spectrum.
I6
CJ~ WITH BLOWING
I2
WITHOUT BLOWING
8.
n
”
160 200
0 40 80 120
f -CPS
(g) Station 3.3 inches, y = 1.000 inches Figure 10 (Conti nued) Longitudinal turbulence spectrum.
O-WITH BLOWING
WITHOUT BLOWING
I
I I I I I I I I 1 I
0 40 80 120 160
f -CPS
(h) Station 3.3 inches, y = 2.000 inches Longitudinal turbulence spectrum.
Figure 10 (Continued)
0.5
0.4
O-WITH BLOWING
“0
z
WITHOUT BLOWING
0.3
ii
\
“t
0.2
c\I
(4
z
-a
K
0.1
120 160
- CPS
f
(i) Station 3.3 inches, y = 4.015 inches (channel centerline) Figure 10 (Concluded) Imgitudinal turbulence spectrum.
I2
IC
o-WITH BLOWING
WITHOUT BLOWING
0 E
X i (
y - INCHES
Figure11 kmgitudinal turbulence intensity, station 6.9 inches.
I -
2.0
“0
O-WITH BLOWING
X
-
W
1.6
WITHOUT BLOWING
cn
\
“t
1.2
t cu - .%I(
0.E
0.4
A
n
w
w v 3
120 160
f -CPS
(a) Station 6.9 inches, y = 0.010 inches Figure12 ikmgitudinal turbulence spectrum.
O-WITH BLOWING
WITHOUT BLOWING
I6
I2
h
n
w v
160 200
80 120
f -CPS
(b) Station 6.9 inches, y = 0.025 inches Hgure12 (Continued) Longitudinal turbulence spectrum.
o-WITH BLOWING
WITHOUT BLOWING
I6
80 120
f -CPS
(c) Station 6.9 inches, y = 0.050 inches (Continued) Longitudinal turbulence spectrum.
Figure 12 , .., .._ . . .-..-...-- -.-- X t: cn \ “t I I I I I I I I I J
f -CPS
(d) Station 6.9 inches, y = 0.100 inches (Continued) Iongitudinal turbulence spectrum.
Figure 12
O-WITH BLOWING
WITHOUT BLOWING
m X 0 I6 W v) \
“t I2
I ’
‘60
‘20
40 80
- CPS
f
Station 6.9 inches, y = 0.250 inches (e) (Continued) Iongitudinal turbulence spectrum.
Figure12
I6
O-WITH BLOWING
c
WITHOUT BLOWING
120 160
40 80
f ^,CPS
(f) Station 6.9 inches, y = 0.500 inches longitudinal turbulence spectrum.
(Continued) Figure 12
O-WITH BLOWING
WITHOUT BLOWING
cut 6
.s
U
f -CPS
IC 1.000 inches inches, y station 6.9
(f3)
spectrum.
Iongitudinal turbulence (Continued) Figure 12
O- WITH BLOWING
“0
X : v) \ E IL
I
0 40 80 120 160 200
f -CPS
(h) Station 6.9 inches, y = 2.000 inches (Continued) kmgitudinal turbulence spectrum.
Figure 12
O-WITH BLOWING
.24
WITHOUT BLOWING
.20
IO
.04
ACPS
f
(i) Station 6.9 inches, y = 4.015 Inches (channel centerline) Figure12 (Concluded) Longitudinal turbulence spectrum.
IO
ONWITH BLOWING
WITHOUT BLOWING
I
2.0 3.5 4.0
0.5 I.0 1.5
- 0
y-INCHES
FQwe 13 Longitudinal turbulence intensity, station 15.7 inches.
odVITH BLOWING
WITHOUT BLOWING
I
n n n v ” 3 v
120 I60
f -CPS
(a) Stat-ion 15.7 inches, y = 0.010 inches Figure 14 Imgitudinal turbulence spectrum.
-_ --- _-- ~-
o-WITH BLOWING
WITHOUT BLOWING
*g 20
X ZII
* I6
\
a
t
t
(v
I2
-
.w
u
80 120 160 200
f -cm
Station 15.7 inches, y = 0.025 inches
b)
Figure 14 (Continued) Longitudinal turbulence spectrum.
I6
80 120
-CPS
f
(c) Station 15.7 inches, y = 0.050 inches Figure 14 (Continued) kmgitudinal turbulence spectrum.
a-WITH BLOWING
WITHOUT BLOWING
“0
X 24
n
40 80 120 I60
f -CPS
(d) Station 15.7 inches, y = 0.100 inches (Continued) Imgitudinal turbulence spectrum.
Figure 14
I6
f -cps
(e) Station 15.7 inches, y = 0.250 inches limgitudinalturbulence spectrum.
(Continued) Figure 14
I6
I2
f -CPS
(f) Station 15.7 inches, y = 0.500 inches (Continued) Lmgitudinal turbulence spectrum.
Figure 14
Q-WITH BLOWING
I6
WITHOUT BLOWING
I2
160 200
40 80
f -CPS
(g) Station 15.7 inches, y = 1.000 inches (Continued) Im@tudinal turbulence spectrum.
Figure 14
I -
o&WITH BLOWING
WITHOUT BLOWING
n
a
80 120
f -CPS
(h) Station 15.7 inches, y = 2,000 inches Figure 14 (Continued) Longitudinal turbulence spectrum.
c-WITH BLOWING
WITHOUT BLOWING
“0
ki
cn
\
c
IL
cuu(
I6
.%r(
U
f -CPS
(i) Station 15.7 inches, y = 4.015 Inches (channel centerline) figure 14 (Concluded) Imgitudinal turbulence spectrum.
A WITHOUT BLOWING
0 WITH BLOWING
5) 31;
IO
I IO 100 1000 10000
YU +
I?
(a) Station 3.3 inches, Q = 60 x 10s3 ft3/sec (steady), ATE = -24.3% Figure 15 Law of the wall velocity profile, T = 6.72 x 1O-4 lb/f%?
ONB
A WITHOUT BLOWING
0 WITH BLOWING
‘5)
313’
IO
I IO
100 1000 10000
YU c
v
(b) Station 6.9 inches, Q = 60 x 10D3 ft3/sec (steady), AT~ P -15.2$ -4 Figure 15 (Continued) Law of the wall velocity profile, T = 6.68 x 10 lb/ftT Om
WITHOUT BLOWING
WITH BLOWING
,(s
IO
u
YU
1)
(c) Station 15.7 inches, Q = 60 x 10s3 ft3/sec (steady), A?o I O& lb/fty Mgure 15 (Concluded) Lsw of the wall velocity profile, T - 6.59 x 1.0-' ONEI
II -STEADY BLOWING
WITHOUT BLOWING
0 40 80 120 160 200
-CPS
f
Figure 16 Imgitudinal turbulence spectrum for various blowing conditions, station 3.3 inches, y = 0.025 inches.
NASA-Lmqdey, 19% CR-63