Source: NASA Historical Reference Collection, NASA History Office, NASA Headquarters, Washington, DC.

Air Force - NACA Conference

on XS-1 Flight Research

A compilation of the papers presented

TABLE OF CONTENTS

Introduction

List Of Conferees

TECHNICAL PAPERS PRESENTED

Resume of XS-1 Airplane Project by Walter C. Williams

Airspeed Installations and Calibration by Milton D. McLaughlin

Stability and Control Characteristics by Hubert M. Drake

Drag Measurements in Flight on 10-Percent- and 8-Percent Thick Wing XS-1 Airplanes by John J. Gardner

Over-All Loads and Buffeting Measurements by Harold R. Goodman

Pressure-Distribution Measurements by De E. Beeler

INTRODUCTION

LIST OF CONFEREES

Ridley, Capt. J. L. Flight Test Division, Air Material Command

Robinson, R. G. NACA

Rumph, L. B., Jr. Curtiss-Wright Corporation

Schaefer, E. B. Boeing Aircraft Company

Schuldenfrie, Marvin N. Bureau of Aeronautics, Navy Department

Scoles, A. B. Fairchild Aircraft

Sears, Dr. W. Cornell University

Sibila, A. I. Chance-Vought Aircraft

Smith, A. M. Douglas Aircraft, El Segundo

Smith, Capt. F. G. Muroc Air Force Base

Smith, R. B. Douglas Aircraft, El Segundo

Smith Capt. R. M. Engineering Plans, Air Material Command

Snyder, G. Boeing Aircraft Company

Soper, Maj. R. E. Equipment Laboratory, Air Material Command

Soulé, Hartley A. NACA

Stack, John NACA

Storms, H. A. North American Aviation

Truszynski, Gerald M. NACA

Tucker, C. Northrop Aircraft, Incorporated

Vensel, Joseph R. NACA

Voglewede, T. J. NACA

Wade, E. Grumman Aircraft Engineering Corporation

Walker, J. A. NACA

Wassell, J. B. Lockheed Aircraft Corporation

Wattendorf, F. L. Engineering Plans, Air Materiel Command

Weed, D. W. Republic Aviation Corporation

Whitten, J. B. NACA

Wiener, B. A. Bureau of Aeronautics, Navy Department

Williams, Walter C. NACA

Woollen, Lt. Comdr. W. S. Bureau of Aeronautics, Muroc Air Force Base

Worthington, Capt. Roy H., Jr. Muroc Air Force Base

Yeager, Capt. Charles Flight Test Division, Air Materiel Command

Zalovcik, J. A. NACA

FOREWORD

by

Dr. J. D. Hunley

As is well known, almost 50 years ago on October 14, 1947, U.S. Air Force Capt. Charles E. Yeager, flying an XS-1, became the first pilot to exceed the speed of sound in any aircraft. What is not so widely known is the extent to which NASAÃŒs predecessor agency, the National Advisory Committee for Aeronautics (NACA), contributed to this achievement and to subsequent flight research using a sister ship to YeagerÃŒs airplane and later versions of the X-1, as it came to be called.

All of these specific design features contributed to the achievement of safe supersonic flight.

Air Force Major General Albert Boyd recognized the importance of the NACAÃŒs contributions to the joint effort when he commented that the results of the X-1 flight research "stand as a monumental tribute to both the USAF and the NACA, since the sonic barrier monster was not only completely licked, but a blow-by-blow account of its defeat was recorded for future use." Recognition of the partnership occurred at an even higher level when President Harry S. Truman presented the coveted Collier Trophy for achieving supersonic flight not only to the Air ForceÃŒs Chuck Yeager but to Larry Bell of Bell Aircraft and John Stack of the NACA.

The series of papers reproduced here reflects the early research results from the XS-1 program. Taken together, the papers illustrate the role the NACA played in the partnership with Bell and the Air Force and point forward to the continuing flight research the NACA performed with the various versions of the X-1 until 1958. Included are a summary report by Walter C. Williams, who headed the NACA contingent at the Muroc Flight Test Unit (later the Dryden Flight Research Center), and reports by other engineers, including such well-known researchers as Hubert M. Drake on stability and control and De E. Beeler on pressure distribution. Mr. Drake, for instance, provides some very interesting comments on p. 21 about features of the XS-1 that complicated the process of flight research--including the short duration of powered flight using rocket engines and the constant movement of the center of gravity resulting from the high rate of propellant consumption. Despite such complications, information from these and other research reports contributed significantly to knowledge about aeronautics in the transonic and supersonic regions and to the design of future supersonic aircraft.

Please note that the document is reproduced with the original pagination, which did not include the use of even-numbered pages. Thus, there are no pages numbered 2, 4, and so forth, and to reduce unnecessary duplication costs, pages like 1 and 7 that contained only the title of the paper, reproduced again on the next page, are also excluded. Copies are faint in some cases because the only available original was also faint, but all words are legible.

Reprints in Dryden History, of which this collection is the first, are designated to make the results of flight research available to interested persons in an inexpensive format and thereby to further knowledge about the CenterÃŒs history. Suggestions for further documents for reproduction are welcome.

Dr. Hunley is the historian for the Dryden Flight Research Center and the author or editor of numerous articles and books, including The Birth of NASA: The Diary of T. Keith Glennan (NASA SP-4105).

December 1996

Resume OF XS-1 AIRPLANE PROJECT

By Walter C. Williams

At the time of the Wright Field Conference on the XS-1 project, January 9, 1948, the first phase of the joint Air Force-NACA accelerated transonic test program had been completed; that is, the XS-1-1 with the 8-percent-thick wing and 6-percent-thick tail had been flown at speeds in excess of the speed of sound. Since that time, additional flights have been made at sonic speeds with this airplane including an all-out flight with full thrust in a dive from 50,000 feet. Recorded data were at 46,000 feet. This value was, of course, obtained from the pilotÃŒs indicator which is connected to the airspeed boom on the nose of the airplane and is believed to be nearly correct. Also, the XS-1-2, which is assigned to the NACA for transonic research and has a 10-percent-thick wing and 8-percent-thick tail, has been flown beyond a Mach number of 1.0. At the time of the Wright Conference, one man, Captain Charles Yeager, had flown at supersonic speeds--leading the way. Since then four other pilots have flown beyond a Mach number of 1.0. Mr. Herbert Hoover and the late Mr. Howard Lilly (NACA engineer test pilots) and Captain James Fitzgerald and Major G. E. Lundquist (Wright Field Flight Test Division test pilots) accomplished supersonic flight in the order named in the XS-1 airplanes assigned to their respective organizations. In the period between conferences 15 flights were made on each XS-1 airplane.

The progress of the tests of the two XS-1 airplanes in detail has been about as follows: In the case of the thin-winged XS-1, an NACA multiple recording manometer was added to the instrumentation which was initially very brief. Measurements have been obtained of the chordwise pressure distribution over one span station on the wing and one on the horizontal tail through a range of Mach number and lift coefficient. The important results of these measurements are presented in the following papers. Also, additional instrumentation was added to determine the difference in pilot-static pressures as measured by the wing airspeed boom and the nose airspeed boom. The XS-1-2 has been thoroughly instrumented from the start for measurement of stability and control characteristics and over-all aerodynamic loads. Data have been obtained primarily on the longitudinal stability characteristics and data from the strain gages have yielded pertinent loads data, including some measurement of buffeting loads. A drag study has been made using the instrumentation primarily installed for the stability research investigation. Most of the tests of both airplanes have been in the range of altitudes from 30,000 to 40,000 feet although some Air Force flights were made at 50,000 feet.

It might be expected that from the number of flights made a considerable amount of data would have been obtained. It must be remembered, however, that engine duration with full thrust is about 2 minutes and the usable portion of a high-speed flight may be considerably shorter considering time for climb and acceleration to high speed. In order to obtain data in a specific maneuver at any Mach number above about 0.90, it is necessary to plan a flight specifically for that maneuver. The pilot will only have time in powered flight to climb from launching altitude to test altitude, accelerate to test speed, and perform the maneuver. Additional data may be obtained, but it will be only on a "catch as catch can" basis. In the short flight time available, the weight changes from slightly more than 12,000 pounds to about 7000 pounds; and with these weight changes, there are changes in center-of-gravity position. It is easy to see that the collecting of a large quantity of data under similar conditions necessary to the determination of the effects of Mach number on the aerodynamic parameters with this machine is a difficult and time-consuming task. It is felt that progress has not been too slow when it is considered that all the data presented at the present conference as well as the data presented at the Wright Field conference were obtained from a total flight time above a Mach number of 1.0 of about 13 minutes in the case of the thin-winged airplane and about 3 minutes in the case of the thick-winged airplane, for a total of 16 minutes.

The future program on the Flight Test Division XS-1 involves exploratory flight in two directions to establish an upper and lower limit in altitude for supersonic flight with this airplane. The low-altitude portion of the program will include determination of the control characteristics of the airplane at high transonic speeds below 30,000 feet altitude. The instrumentation in this airplane is being increased so that the instrumentation will be comparable to that in the thick winged Machine. The manometer has been removed but strain-gage equipment has been installed since it is felt that, at the lower altitudes, buffeting loads will be of primary interest. Incidental to the research on the thick-winged airplane, it was found that at 30,000 feet altitude full thrust would take the airplane just to a Mach number of 1.0. It would be expected that the thin-winged airplane would be able to reach a Mach number of 1.0 at a somewhat lower altitude. The instrumentation in the XS-1-2 will remain essentially the same for the present, and tests will be conducted to determine more completely the longitudinal stability and control in accelerated flight, the variation of maximum lift coefficient over the Mach number range, and the directional and lateral stability and control characteristics. More measurements will be made of over-all aerodynamic loads and buffeting loads. Some additional instrumentation will be installed when available for better and more complete drag measurements. At a later date, multiple manometers will be installed, and complete spanwise and chordwise measurements of the pressure distribution over the wing and horizontal tail will be made. There are other dynamic research investigations that are of considerable interest such as aero-dynamic heating, boundary-layer measurements, wake surveys, and the like. Some of this information may be obtained with the XS-1 airplane concurrently with the research that has been outlined but will probably be delayed until completion of this work and possibly will be done with future research airplanes with more duration.

AIRSPEED INSTALLATIONS AND CALIBRATION

By Milton D. McLaughlin

In the course of the flights on the XS-1 airplane, we have accumulated a certain amount of information on the problem of measuring airspeed in supersonic airplanes. The story isnÃŒt complete as yet but there are enough data at hand to give a fairly good general picture of the problem and the factors to be considered in the choice of an airspeed installation. The purpose of this paper is to summarize what has been learned so far in this program.

Airspeed is determined by measurements of total and static pressure. The total-pressure measurement gives little trouble in the speed range in which we are working. There is a correction for the total-pressure loss through shock waves associated with the head itself and also for the loss through shock waves associated with the airplane if the head is in their wake. These errors are quite small and can be calculated. The main difficulty in measuring airspeed is in measuring the ambient static pressure.

Figure 1 shows the calibration of the wing airspeed installation. The calibration is presented as M", the uncorrected Mach number, against MÌ or M, where MÌ is the Mach number corrected for static-pressure error and M is the Mach number corrected for static-pressure error and the total-pressure loss behind a normal shock wave.

The magnitude of the theoretical correction for total-pressure loss is shown as the difference between the two calibration curves at Mach numbers above one. As you will see a little later the full value of this correction should not be applied to this particular wing-boom installation. For the present, we are ignoring this small correction and are presenting the test data as a function of MÌ. Note that the airspeed calibration diverges from the 45___ line at Mach numbers above and below one. The difference between measured and true static pressure which causes this divergence will be shown as another calibration curve in figure 2.

At the top of the figure is a diagram of the XS-1 showing the wing-boom airspeed installation. The airspeed head is the Kollsman high-speed pitot-static head. The static-pressure orifices are located 96 percent of a wing chord ahead of the leading edge of the wing. The calibration curve is shown at the bottom of the figure. It is presented as _______ against MÌ, _______ being the difference between the true and indicated static pressure and _____Ì being the impact pressure as corrected for static-pressure error. The ordinate as presented in this form is non-dimensional. The error in the static pressure is of the order of 1 percent at Mach numbers of 0.7 and 0.8. Above Mach numbers of 0.87 the error increases positively to about 8 percent at Mach numbers slightly above one. Here a discontinuity occurs and the error drops to a slightly negative value where it increases in the positive direction again.

Our first thought was that the discontinuity was associated with the airspeed head itself. In the meantime an airspeed head was developed at the Langley Laboratory which was known from bomb-drop tests to have a calibration curve free from any large discontinuities. The physical high-speed head in figure 3. The main differences between the two are that the NACA high-speed head has no bulge aft of the static orifices and also the static orifices are located more to the rear. On the Kollsman head the static orifices are located in the top and bottom of the head and in the NACA head they are spaced equally around the circumference of the head. Figure 4 shows the calibration curve obtained from the bomb-drop test. Shown at the top of this figure is the bomb-drop model showing the airspeed-head installation. At the bottom of the figure is the calibration curve obtained. Although the absolute accuracy of the calibration is not sufficient for direct use it does serve to illustrate the absence of any large discontinuities associated with the head itself. The head is rather isolated from the effects of the wing and body, being _______ fuselage diameters ahead of the relatively slender nose of 8 1/2 fuselage diameters ahead of the maximum-diameter station of the fuselage.

The Kollsman head in the XS-1 was replaced with this new head and a calibration flight obtained. There was no significant change in the calibration curve. The new heard gave essentially the same errors as did the Kollsman. In the meantime at Langley, wing-flow tests were run on the Kollsman head, with results as shown in figure 5.

The calibration curve is presented here as __________ against M. The ordinate __________ is the same as __________ since __________ and __________. The blocking error due to the bulge is about 1 percent and this error disappears between Mach numbers 0.98 and 1.0. As this Mach number range did not occur at the head on the XS-1 due to the discontinuity previously noted, no significant difference due to the Kollsman head could have been noticed.

The evidence is pretty conclusive, therefore, that the errors in the static-pressure measurements are not due to the head but are due to the presence of the airplane. An attempt will be made to explain the discontinuities in the calibration curve in figure 6. Shown at the top of the figure is a diagram of the XS-1 showing the wing-boom airspeed installation. The static orifices are located 0.96 wing chord in front of the leading edge of the wing, 2.5 fuselage diameters outboard of the fuselage, and 35 degrees aft of the nose of the fuselage. Shown in the figure are the approximate locations of the fuselage and wing bow waves for Mach numbers of 1.03 and 1.3. At the bottom of the figure is the wing-boom calibration curve as presented in figure 2. At Mach numbers of 0.7 and 0.8 and below, the calibration curve changes with lift coefficient so no calibration curve has been shown. At an approximate Mach number 0.87 the blocking due to the wing begins to show an effect so the static-pressure error increases in the positive direction to an error of about 8 percent at Mach numbers just above one. At these Mach numbers the wing and fuselage bow waves are forming in front of the wing and fuselage, respectively. As the Mach number increases the bow waves grow stronger and start to move toward the airplane, the fuselage bow wave attempting to attach itself on the fuselage nose and the wing bow wave attempting to form just in front of the wing.

Approximate locations of these waves are shown for a Mach number of 1.03. It may be noticed in the figure that at this Mach number the wing bow wave is passing over the airspeed static orifices in the wing boom. This is shown by the dotted line on the calibration discontinuity. The large pressure decrease is due to the passage of the bow wave. This bow wave does not always pass over the static orifices at a Mach 1.03 but in the immediate vicinity. It is believed that the reason for this is the slightly different angles of yaw and attack of the airplane. After passage of the wing bow wave the wing boom head is exposed to the pressure field around the fuselage, which may be visualized as extending out along Mach lines from the fuselage surface. Just after passage of the wing bow wave the principal influence comes from the negative pressure portion of the fuselage near maximum thickness. With increasing Mach number, the Mach lines slant back, and the head is influenced by the positive pressure region over the nose. The positive-pressure error is expected to reach a maximum just before passage of the fuselage bow wave. At the highest MÌ reached so far in the calibration flights the fuselage bow wave apparently has not passed over the static holes. It seems logical to expect (from the flattening of the error curve) that at speeds a little higher than these the head will emerge from the influence of the airplane and take on the characteristics of an isolated head.

To have the head in the wake of the fuselage bow wave and under the influence of the fuselage flow field seems undesirable. Changes in fuselage yaw angle, for example, might be expected to introduce static errors. An example of the change in pressure fields with respect to the wing due to yaw may be seen in figure 7. Shown in the figure is a photograph, taken by means of a Schlieren setup, of a model in the Langley 9-inch supersonic tunnel. The velocity in the test section is at a Mach number 1.9. From an inspection of the figure we can see that the model has a slight degree of yaw which would induce errors in a wing-boom airspeed installation. A better location for an airspeed installation would seem to be well forward of the fuselage nose. There we would expect to get blocking due to the fuselage, but once the fuselage bow wave had passed we would expect to get pressure measurements corresponding to an isolated head.

On the XS-1 the nose airspeed installation was utilized for the pilotÃŒs indicating airspeed and altitude instruments. To get some data on the nose-type installation we connected the static side of the Kollsman head on the nose boom to the static side of the Kollsman head on the wing boom through a sensitive differential-pressure cell. By adding this differential pressure to the wing-installation error we get a calibration curve for the nose installation. Figure 8 shows the results obtained so far.

Presented is a diagram of the XS--1 showing the nose-boom installation and the calibration curve obtained in this installation. The static orifices on the nose boom are located 0.60 fuselage diameter forward of the nose of the fuselage and 2.77 fuselage diameters ahead of the position of maximum diameter of the fuselage. From the calibration curve it may be seen that the low-speed blocking is greater than on the wing installation. The blocking is slightly lower, however, than that predicted by theoretical calculations, increasing from about 6 percent at low Mach numbers to 15 percent at Mach numbers around 1.07. Here the error decreases to approximately zero or within 2 percent of true static pressure which is within the accuracy of our combined calibration curves.

The error remains on the order of 2 percent to our highest Mach number 1.17 for which we have a calibration for the nose installation. There is no reason to believe that it should not stay the same at higher Mach numbers.

Figure 9 shows a comparison between the airspeed calibration curves for the wing and nose-boom installations. Notice the greater blocking effect from the fuselage on the nose boom. But at higher Mach numbers the nose-boom-installation error is about zero whereas the wing boom has an increasingly large error. The low-speed blocking which seems to cause the greatest error in the nose can be reduced by lengthening the nose boom.

The main trouble with a fuselage-nose installation is the length of boom required to get very far ahead of the nose in terms of fuselage diameters. The _____ diameters distance used in the bomb tests would be _______ feet in the XS-1 airplane.

A study of the minimum boom length required is in progress now at Langley, where the wing-flow method is being used to establish how the magnitude of the blocking error and the discontinuity in the static calibration varies with fuselage shape and the distance ahead of the fuselage. The first results from these tests are shown in figure 10.

Shown here is the body used in the tests and the results for two different length nose installations. As can be seen the test body has a more slender nose than the XS-1. At 0.6 diameter and 0.75 Mach number, the blocking is slightly lower than that of the XS-1 nose boom, but the increase in blocking near Mach number one is considerable less than that

Figure 1. - XS-1 wing-boom airspeed calibration.

Figure 2. - XS-1 wing boom airspeed installation and calibration.

noted for the airplane, probably because of the more slender nose of the model as compared with the XS-1 airplane. Correspondingly, the final drop through the bow shock wave occurs at Mach number 1.025 as compared with 1.075 fro the XS-1.

At 1.5 diameters, on the other hand, the blocking is relatively small and remains small until it disappears at about Mach number one.

It is of considerable interest that the blocking observed at Mach number 0.75 agrees fairly well in all three cases with the computed low-speed blocking for the particular body shapes and nose-boom lengths. For this reason, a simple rule is suggested for required nose-boom length for other fuselage shapes: If the boom is long enough that the calculated or experimental blocking at low speeds is below 2 percent, then the blocking should remain of the same order until it disappears near Mach number one. For the XS-1, the boom length for 2-percent low-speed blocking is 1.7 diameters, or 7.8 feet. A good lead on the variation of low-speed blocking error with boom length for several nose shapes may be obtained from NACA TN No. 1496 by William Letko entitled "Investigation of the Fuselage Interference on a Pitot-Static Tube Extending Forward from the Nose of the Fuselage."

In connection with total-pressure measurements one point may be worth noting. For the wing installation, the pitot head is in the presence of a multiple-shock system. The total-pressure loss within the Mach number range of the calibration is the sum of the oblique fuselage bow-wave shock loss plus either the wing bow-wave or the pitot-head bow-wave shock loss. The wing or pitot-head shocks will occur from the local Mach number as determined by the fuselage interference and not from the free-stream Mach number. Thus, at very low supersonic Mach numbers the loss will be slightly greater than that through a normal shock at stream Mach number (but still negligible), whereas at M = 1.3 the loss is estimated to be only _______ percent of the free-stream impact pressure instead of the 3 percent full normal-shock loss. The loss will become equal to the normal-shock loss after passage of the fuselage bow wave.

Figure 3. Comparison of physical characteristics on the Kollerman airspeed head and the NACA airspeed head.

Figure 4. Drop model and calibration curve for NACA head.

Figure 5. Calibration curve from wing-flow tests of Kollerman head.

Figure 6. Diagram of XS-1 wing-boom installation and calibration curve showing approximate position of wing and fuselage bow waves for Mach numbers 1.03 and 1.3.

Figure 7. Schlieren photograph of a model in the Langley 9-inch supersonic tunnel. Test-section velocity equals Mach number 1.9.

Figure 8. Diagram showing XS-1 nose airspeed boom installation and the airspeed calibration curve.

Figure 9. A comparison between the XS-1 nose-boom airspeed and wing-boom airspeed installations.

Figure 10. Model used and calibration curves obtained from a wing-flow investigation of fuselage shapes and nose-boom lengths.

STABILITY AND CONTROL CHARACTERISTICS

By Hubert M. Drake

Flight testing with the XS-1 airplane is complicated by a number of factors not encountered in normal flight testing. These originate mostly in the rocket power plant which permits thrust to be varied only in increments of about 1500 pounds making it difficult to obtain steady flight conditions. In addition, the high rate of propellant consumption inherent in rocket engines causes large changes in the weight and center-of-gravity position. In the course of perhaps less than 3 minutes, the wing loading changes from 94 pounds per square foot, drop condition, to 54 pounds per square foot. The center of gravity moves forward during powered flight from 22.5 percent mean aerodynamic chord at drop to 20.3 percent mean aerodynamic chord at the completion of powered flight. Jettisoning the remaining propellants moves the center of gravity back to 24.2 percent mean aerodynamic chord. This makes difficult the obtaining of data for comparable normal-force coefficients and center-of-gravity positions. Since the duration of powered flight is so short, it is difficult to obtain steady flight or any series of maneuvers at high Mach numbers. The data presented are obtained primarily in powered flight with the center of gravity at about 21 percent mean aerodynamic chord.

The instrumentation of the XS-1 with the 10-percent-thick wing consists of telemetering and internal recording instruments which are used to measure airspeed, altitude, control positions and forces, normal acceleration, transverse acceleration, longitudinal acceleration, rate of roll, and angle of yaw or pitch, as well as elevator and stabilizer angles. The instrumentation of the XS-1 with the 8-percent-thick wing is not quite so complete; so, unless otherwise indicated, the results presented on stability and control are from tests of the airplane with the thicker surfaces.

Figure 1 shows the variation of elevator position and force with Mach number in level flight for two stabilizer angles. These flights were made at 41,000 feet pressure altitude and the airplane normal-force coefficient varied from 0.36 to 0.25 as the Mach number was increased. The center of gravity varied from about 21.5 to 20.8 percent mean aerodynamic chord in the course of the two flights.

The flight with _____ stabilizer was made first. As the Mach number increased, the elevator required for trim increased in an upward direction going against the stop at M = 0.93 with a force of about 30 pounds. Powered flight was terminated at a slightly higher Mach number as the nose-down tendency continued. On the next flight the stabilizer angle was reduced to ________. The elevator trim and force curves are essentially similar to those obtained from the previous flight. The elevator moved up as the Mach number increased almost reaching the stop at M = 0.98. The trim then changed in the nose-up direction to M = 1.0 above which it again reversed and there was a gradual nosing-down tendency. These trim changes are caused mainly by the changes in the pitching moment of the wing. This is discussed more completely by Mr. Beeler in the paper entitled "Pressure-Distribution Measurements." Above a Mach number of about 0.93, the pilot felt that he could move the elevator through a fairly wide range of angles without affecting the airplaneÃŒs flight path. Therefore, the positions recorded above this Mach number may not be the only trim conditions possible.

To illustrate this, figure 2 shows a time history of a push-down made at M = 0.99. The elevator was deliberately moved from full up through three quarters of the total travel. The airplane responded somewhat, the normal acceleration decreasing about 0.7g and the up load on the tail increasing by about 200 pounds. The pilotÃŒs impression was that there was no change in the attitude of the airplane as the elevator was moved. Comparison with a push-down made at a Mach number of 0.30 shows that the elevator was 8 times as effective at this Mach number, producing twice the accelerations with one-quarter the elevator travel used in the push-down at M = 0.99.

If the elevator effectiveness nearly vanishes at M = 0.99, the stabilizer effectiveness at this speed is of extreme importance. Figure 3 shows a time history of a portion of a flight at a Mach number of 0.99 at 30,000 feet. The elevator was first moved full up producing a small increment of normal acceleration. Then it was moved back down and the stabilizer was moved from _____ to _____. This stabilizer deflection produced a total change of about ______ in normal acceleration even through the elevator was moved at the same time in such a direction as to reduce the change in normal acceleration. At a Mach number of 0.82, this change of stabilizer angle would have produced a change of about _____ showing that the stabilizer effectiveness has decreased to about 35 percent of its value at M = 0.82.

The variation of relative effectiveness of elevator and stabilizer between a Mach number of 0.78 and 0.99 is shown in figure 4. The values below 0.93 were obtained from figure 1 which showed the variation of elevator trim position with Mach number for two stabilizer incidences. The extension, shown as a dashed line, was obtained from the single maneuver at M = 0.99 shown in figure 3 using the portion where the elevator was held fixed. The relative elevator effectiveness decreased about 50 percent as the Mach number increased from 0.93 to 0.99. The shape of the curve between 0.93 and 0.99 and above 0.99 has not been established as yet but there has been no evidence of any reversal of effectiveness and there are indications that the relative effectiveness increases at Mach numbers above 1.0.

A time history of a turn at a Mach number above 1.0 is shown in figure 5. The push-down shown previously is presented for reference. The Mach number is not constant during the turn, increasing from 1.01 to 1.05. The elevator is more effective in producing acceleration at Mach numbers of about 1.07 than it was in the push-down at M = 0.99. There is no evident lag or dead spot in the response to the elevator control at this Mach number.

Figure 6 shows the values of stick force per g and ________ obtained from these maneuvers and from several turns at lower Mach numbers. These data show that the control required to produce a unit change in normal-force coefficient increases from a minimum of the order of __ at M = 0.68 to a peak value of about _____ at M = 0.99. The value of _______ then decreases as Mach number increases. This variation is caused both by the decrease in the effectiveness of the elevator in producing lift on the tail and by the increase in stability caused by the rearward movement of the wing aerodynamic center. This change in aerodynamic center is more completely covered in the paper on pressure distributions. The stick force per g has a variation with Mach number similar to that of _____, increasing to a peak value of about 73 pounds at M = 0.99 and decreasing at Mach numbers above 1.0.

It may be pointed out that this one phenomenon, the almost complete loss of elevator control in the Mach number range around 1.0, shows the correctness of the original premise on which these research airplanes were obtained. This premise was that the first transonic flights should be made in level or climbing flight where the pilot would have positive speed control through power. if an attempt had been made to dive the airplane through this speed range, the pilot would have found himself in a position of being unable to reduce power, in this case gravity, and without control effectiveness to recover from the dive. Use of the adjustable stabilizer would be required for recovery with the possibility of large unintentional accelerations being imposed. An account of a dive recovery using the stabilizer is given later by Mr. Goodman in the paper entitled "Over-All Loads and Buffeting Measurements."

This region of elevator ineffectiveness produces disturbances on the airplane even in level and climbing flight. At the Wright Field Conference on January 9. 1948, Captain Charles Yeager mentioned that in decelerating through the region of about M = 0.98, there was experienced a sharp impulse. Figure 7 shows a time history of a deceleration through this region at 41,000 feet. The deceleration is at a rate of about 0.02 Mach number per second. As the Mach number decreased through the ineffective range the elevator was held relatively fixed; therefore, as the elevator became effective at M = 0.97 to 0.96, it was not at the correct angle for trim and as a result there was some acceleration developed. The magnitude of the acceleration developed was about 0.7g. Larger accelerations would be developed if the elevator were held in a neutral or down position when the speed was reduced. This disturbance has occurred every time the airplane is decelerated from speeds above M = 0.97. It is minimized by a slow rate of Mach number decrease. It does not occur in accelerating through this range, apparently because the rate of change of Mach number is smaller.

Figure 8 shows a comparison between the curves of the elevator trim position of the XS-1 with the 8-percent-thick wing and with the 10-percent-thick wing. The short-dash and dot-dash curves were presented in figure 1 and are for the 10-percent-thick wing and the solid-line and long-dash curves are for the 8-percent-thick wing. In general, the variation in trim is similar for the two airplanes. At low altitudes there was a distinct change in trim in a nose-up directions at about M = 0.9 but at the higher altitudes of these flights this change of trim smoothed out into a flat spot. As mentioned previously, the trim changes are more thoroughly discussed in the paper on pressure-distribution measurements. Although the trim changes near M = 1.0 appear more pronounced for the 10-percent-thick wing, the ineffectiveness of the elevator about M = 1.0 leaves some doubt as to the validity of a comparison between the airplanes in this Mach number region.

As yet there has not been very much done on the lateral stability and control of the XS-1 airplanes. Figure 9 shows the variation with M of the amount of rudder required to produce one degree sideslip. The value of _____ remains nearly constant at about 1.5 to the highest Mach number for which data are available below 0.99. This quantity increases very rapidly to a value of about 16.5 at a Mach number of 0.99. This trend is similar to the elevator-effectiveness variation. The sideslip angle was measured by a standard yawmeter on the right wing tip and, therefore, there may be some doubt concerning the exact yaw angle of the point at M = 0.99 since the characteristics of this installation at high Mach numbers have not been determined. The pilot could not see any noticeable sideslip at M = 0.99 even with full rudder deflection so the effectiveness is certainly very low or the value of _____ is very high. There has been no evidence of any reversal of effectiveness and effectiveness increases again at Mach numbers above 1.0. As yet, no measurements have been made of the directional stability parameter _____ by use of the period of the directional oscillation.

No measurements of aileron effectiveness have been made above M = 0.8 as yet. PilotsÌ reports, however, state that the ailerons remain effective at all speeds. It may be that the ailerons lose some of their effectiveness in the same region as the elevator and rudder but the rate of roll is so high the pilot does not notice any loss of effectiveness.

The lateral oscillation mentioned in the Wright Field Conference is still present and is annoying. The oscillation is undamped and of high frequency having a period of about 1.3 seconds, but of small amplitude. The motion usually shows up on the rate of roll, sideslip, and transverse acceleration records, sometimes predominantly as rate of roll, other times predominantly as sideslip. There is no evidence in the records or to the pilot of control motion which might be exciting the oscillation. When the controls are used in an attempt to control the oscillation, the records only show a change from a regular to an irregular motion.

The oscillation is intermittent in occurrence, usually starting after perhaps 10 to 20 percent of the powered flight. Thereafter it can occur at any time. It has been noticed even at low altitudes in the landing approach.

The effect of the oscillation on the tests of the XS-1 has been considerable. Precision flying is very severely interfered with since the oscillation may occur in straight flight, turning flight, sideslips or other maneuvers. In early flights it appeared that the oscillation might limit the maximum Mach number to about 0.95 because in these flights the oscillation would cause the lox tank to lose pressure and reduce the thrust even to the point of stopping the engine. Thinking the spoilers on the wing might be sucking up and providing the exciting force, they were screwed down. The oscillation did not occur as often thereafter and it was possible to achieve higher Mach numbers. When the oscillation occurs, however, it is of the same character as before.

Although as yet the primary cause of the oscillation has not been determined, apparently the spoilers were a contributing cause. Another possible cause of the oscillation is at present under investigation. This is the possibility that the snaking may be caused by an angle-of-attack effect. The occurrence of the oscillation appears to be related to lift coefficient since, in many cases, the oscillation has damped out as the lift coefficient was increased at low Mach numbers.

The Mach number also has an effect on the oscillation at transonic Mach numbers which would also indicate an angle-of-attack effect as the cause. This is indicated in figures 10 and 11. Figure 10 shows a time history of the oscillation as the Mach number is increased from 1.06 to 1.16 and the normal-force coefficient decreases from 0.33 to 0.16. The controls, except for the elevator, are essentially fixed. Previous to the start of this record, the oscillation was not noticeable. The amplitude increases as Mach number increases and normal force decreases but becomes constant in amplitude at a Mach number of about 1.13. Figure 11(a) shows that with the Mach number and normal-force coefficient, the oscillation damps. The record was ended at this point because the airplane was disturbed as it went through the range of Mach numbers where the elevator was ineffective. A few angle-of-attack measurements have been made in this speed range and indicate that the constant normal-force coefficient and decreasing Mach number. It appears, therefore, that the angle of attack, as it affects the inclination of the principal axis of the airplane, may be the primary cause of the snaking.

The airplane with the thin surfaces has not encountered snaking to the same extent as has the airplane with the thick surfaces but snaking has been noted at supersonic speeds at high altitude and at subsonic speeds at 30,000 feet.

CONCLUSIONS

From preliminary flights at transonic speeds with the XS-1 having a 10-percent-thick wing and an 8-percent-thick horizontal tail the following conclusions may be drawn:

1. There is a gradual nose-down change in trim between M = 0.78 and 0.99, a change in the nose-up direction to M = 1.0, followed by a nose-down change.

2. The elevator effectiveness decreases about 75 percent as the Mach number increases from 0.78 to 0.99 and is so low at Mach numbers above 0.93 that it has been difficult to obtain stabilized trim conditions in the powered-flight time available.

3. The elevator forces required to fly to a Mach number of 1.05 in level flight at 40,000 feet are low, on the order of 30 pounds.

4. The elevator effectiveness is at a minimum at M = 0.99 increasing at Mach numbers above and below this value. This change in effectiveness causes a disturbance in decelerating from Mach numbers above 1.0.

5. The trim curves are similar for the XS-1 with the thick and with the thin surfaces.

6. The rudder effectiveness is approximately constant as Mach number increases from 0.35 to 0.9 but the effectiveness almost vanishes at M = 0.99.

7. Snaking has been encountered at supersonic speeds and appears to be affected by both lift coefficient and Mach number.

Figure 1.- Variation of elevator position and force with Mach number.

Figure 2.- Time history of push-down at M = 0.99 compared with push-down at M = 0.30.

Figure 3.- Time history of elevator and stabilizer movement in pull-up at M = 0.99.

Figure 4.- Variation of relative elevator and stabilizer effectiveness with Mach number.

Figure 5.- Time history turn at a Mach number above one compared with push-down at M = 0.99.

Figure 6.- Variation of _______ and ______ with Mach number.

Figure 7.- Time history of deceleration from supersonic flight.

Figure 8.- Comparison of trim curves for XS-1 with 8-percent thick and 10-percent-thick wings.

Figure 9.- Variation of rudder effectiveness with Mach number.

Figure 10.- Time history of lateral oscillation. M = 1.06 to 1.16.

Figure 11.- Time histories of lateral oscillation.

DRAG MEASUREMENTS IN FLIGHT ON 10-PERCENT- AND 8-PERCENT-

THICK WING XS-1 AIRPLANES

By John J. Gardner

INTRODUCTION

In the course of loads and stability and control tests on the XS-1 airplanes, it has been possible to obtain drag measurements by using the records of these other tests. Drag measurements have been obtained on both the 8 percent wing and 10 percent wing airplane.

Before presenting the results of the drag measurements the method used in making the measurements and the accuracies of this method should be explained.

DISCUSSION

The accelerometer method was used in obtaining the drag measurements. (See fig. 1) In this method the simple relation D = T - _____ is used. The thrust term in this equation is determined by the relation

_______________

where

T thrust

___ chamber pressure

___ throat area of rocket nozzle

___ thrust coefficient of rocket system

___ nozzle exit area

___ design discharge pressure at nozzle exit, 14.7 psi

___ Ambient pressure of surrounding medium

In obtaining ___ it is necessary to tap into a line of the engine system at the point indicated by the pressure gage. This line carries a small flow of _____ gas to the spark-plug system discharging through the chamber to atmosphere. A small orifice between the high-pressure _____ source and the spark-plug system restricts the flow to an approximately constant value irrespective of a variation in cylinder back pressure. It is therefore necessary to correct ____ for the pressure drop in this line due to the ____ flow. The value ______ is also corrected to absolute by use of a value of _____. The maximum error in determination of ____ is estimated as 3 percent. The values ____, ____, and ____ are obtained from the manufacturerÃŒs engine test reports. The exit section of nozzle is designed to give proper expansion of the stream at sea-level conditions. The value of ____ is therefore 14.7 psi. The term ___ (_____) is a correction to the thrust for altitude operation. At sea level it is zero. The manufacturerÃŒs tests of the engine were made at sea level and determined the quantities in this relation T = _______. The estimated maximum probable error in ____ has been stated as 3 percent. It is estimated as 1 percent in _____ and _____. The possible error in thrust determination because of an error in the altitude correction is less than 1 percent as this pressure thrust is small in comparison with the velocity-thrust term. The maximum probable error then in the over-all thrust determination is approximately 5 percent.

The second part of the drag relation is W (weight of airplane) times _____ (acceleration of the airplane, g units, along the flight path). (See See fig. 2) The value of W is determined by using gross weight at drop and time history of propellant and nitrogen consumption during flight. The W term is accurate to 1 percent. The value of _____ (flight-path acceleration) is determined in the following manner. In the airplane there is a recording accelerometer that records the accelerations along the airplane longitudinal axis. (___) and normal to this axis (____). Between the longitudinal axis and the flight path there is an angle ___. The angle ____ is determined from values of ______ and _____ given in transonic-tunnel and wing-flow data on XS-1. Then,

____________________

The maximum probable error in the determination of _____ is approximately 0.03g. At gross weight this is a maximum error of approximately 300 pounds; at empty weight, 200 pounds. Expressed as an error in drag at high altitudes and approximately 0.8 Mach number of these tests, it could be a maximum of 25 percent of the drag. At the high speed it could be a maximum of 3 percent. By considering all the possible errors in the drag determination at the low speeds, _____, and high altitudes of these tests, the error could be 30 percent; at the high-speed condition, it could be a maximum of 10 percent. It might be pointed out herein that the accuracy at low speeds could be improved considerably by conducting the tests of this low range of M at lower altitudes and therefore at higher qÃŒs. Having described the method of measuring the drag and the accuracies of the method, the presentation of the drag results can now be given.

In See figure 3 is shown the variation of _____ (total drag coefficient) with M (airplane Mach number) for the 8 percent wing and 6 percent tail airplane. The results of two flights are shown (flights 15 and 16). It might be of interest to mention the flight conditions of these tests. In both flights, climbs were started at 25,000 to 30,000 feet on three rockets in the M range of 0.8. The climbs were continued to approximately 45,000 feet (_______). The climbs were not steady, and this accounts for some of the scatter of the points. The ____ in these tests varied from an average value of 0.4 at low speed to 0.2 at high speed. Sufficient data were not available to present drag curves for constant ___ conditions. The disagreement between the two flights at the high M range is not fully understood. The difference between the conditions of _____ for the two flights in this Mach region does not explain the variation between the two curves. It might be mentioned that the group of points for flight 15, Mach number approximately 1.1, were obtained in a 20-second leveled stabilized run after climbing; whereas the points in the Mach number range of 1.1 to 1.32 for flight 16 were obtained in a shallow dive while the airplane was accelerating. It is possible that the difference in these two flight conditions might account for a part or all of this disagreement. In both of these flights, poweroff points were obtained. Several points in the group for flight 15 at a Mach number of approximately 1.1 represent power-off operation. In flight 16, power was cut at a Mach number of approximately 1.32. Drag points were obtained in deceleration to a Mach number of 1.1 for this flight. The scatter of the power-off data points from both flights about the faired curves was random and typical of the rest of the data. No significant variation between power-on and power-off points was noted.

In See figure 4 is shown the variation of _____ with M for the 10 percent wing and 8 percent tail configuration. The points with the solid line through them are the flight test results. This flight was made somewhat similar to the thin-wing tests. A climb was begun at approximately 30,000 feet in _____ range on four rockets. It was continued to an altitude of approximately 45,000 feet reached at _____. A level stabilized run was then made for approximately 20 seconds represented by the group of points at ______. Again this climb was not steady and this accounts for part of the scatter of points. The _____ in this flight varied from an average low-speed value of 0.4 to a high-speed value of 0.2. Sufficient data were not available to show the effect of ____ on drag coefficient _____.

Also shown on this plot is a comparison of the flight results with wind-tunnel and drop-model tests. The long-dash line represents drag results from tests of a ____ -scale drop model tested by the Langley Flight Research Division. The short-dash line represents results of a _______-scale model tested in the Langley 8-foot high-speed tunnel. All three curves are compared on the basis of same _____ at a given Mach number. The agreement between the three tests is good at __________.

In See figure 5, a comparison of the drag of the 10 percent wing and 8 percent wing airplanes against Mach number is given. At ______ to 1.1 the L/D of the 10 percent wing airplane is approximately 1.5 and for the 8 percent wing airplane is approximately 2.0.

At M = 1.1, the drag of the 10-percent-thick wing airplane is 60 percent greater than that of the 8-percent-thick wing configuration.

This drag difference of 60 percent is more than would be expected if the drag of the 8-percent-thick wing airplane is assumed to be divided evenly between the wing and the fuselage-tail combination, and the drag of the wing is assumed to vary as the square of the thickness ratio. This drag increase is approximately double what might be expected if this assumption is made. However, it might be expected that the blunt fuselage of the airplane would cause considerable interference with the wing, making the separation of wing and fuselage drag difficult.

In See figure 6 is shown the variation of ____, the chordwise force coefficient, with airplane Mach number. This data was obtained from a work-up of pressure-distribution measurements made in flight at midsemispan station of the 8 percent wing. In the Mach number range M = 0.92 to 1.30, the chordwise force can be taken as the section drag at zero lift. Then it is seen that the peak section drag occurs at an airplane Mach number of 0.92. In tests of free-fall models with similar wings attached to cylindrical bodies having low interference with the wing, this section zero-lift drag reached a maximum at a Mach number of 0.97 to 0.98. It is then evident that at the midsemispan station of these tests, the fuselage has increased the local-stream Mach number by 0.05. In the discussion on the XS-1 airspeed calibration, it was pointed out that at the airspeed boom at the wing tip, the fuselage had raised the local-stream Mach number by 0.02. It can be expected that at the wing root the effect will be more than the 0.05 effect at the midsemispan. From these results, it can be seen that the fuselage is affecting the velocity field of the wing and this influence extends even beyond the wing tips. It is evident then that with such an effect existing, it would be difficult to divide the over-all airplane drag into wing drag and fuselage-tail drag.

In See figure 7 is shown the variation of the ratio of section normal-force coefficient and section pressure drag coefficient (_____) with section normal-force coefficient (____). These data were worked up from a midsemispan pressure-distribution survey on the 8 percent wing. The Mach number of these tests was approximately M = 1.16. The ratio of _____ can be assumed to be approximately the same as L/D of the section and the _____ value can be assumed to be section _____. It is noted then that a maximum section L/D of approximately 5 is reached at a section ____ of 0.4 and is maintained over the _____ = 0.4 to 1.0 range. It can be expected to drop off at _____ values above 1.0 range.

In the _____ range of 0.2 to 0.3, the section L/D is 3 to 4. The L/D for the 8 percent wing airplane at this wing condition was approximately 2. Adding the fuselage and tail combination to the wing then reduced the wing L/D from 3 to 4 to an over-all airplane L/D of 2. From these data, it is evident that at the high altitudes and highest Mach number of these drag tests, the airplane was operating at a _____ below that for maximum L/D. If the wing is operated in the region of ____ giving a section L/D of 5, the airplane L/D might be expected to rise to slightly more than 3. It is evident then that while the 8 percent wing airplane is flying fast, it is not going to go very far until considerable improvement is made on its L/D for the supersonic speeds.

See Figure 1.- Determination of rocket thrust T.

See Figure 2.- Determination of flight-path acceleration N f.p..

See Figure 3.- Drag of 8 percent wing XS-1 airplane.

See Figure 4.- Drag of 10 percent wing XS-1 airplane.

See Figure 5.- Comparison of drag of 8 percent wing XS-1 airplanes.

See Figure 6.- Variation of section chordwise force coefficient with airplane Mach number obtained from midsemispan pressure distributions on 8 percent wing airplane.

See Figure 7.- Variation of ratio of section normal-force coefficient and section pressing drag coefficient and section pressure drag coefficient with section normal-force coefficient.

M = 1.16.

OVER-ALL LOADS AND BUFFETING MEASUREMENTS

By Harold R. Goodman

An important phase of the flight tests of the XS-1 airplanes is the evaluation of the aerodynamic loads on the wing and tail surfaces. Objectives of the aerodynamic loads measurements are to obtain information on balancing tail loads due to aerodynamic center changes, wing fuselage zero-lift pitching moment, the location of wing lateral center of pressure, the division of total airplane load between the wing, tail, and fuselage, and the magnitude and occurrence of buffeting loads. Data on over-all loads and wing section characteristics are being obtained, the former through strain-gage installations and the latter by means of pressure-distribution measurements. The data presented herein are overall loads and buffeting measurements as obtained on the XS-1 airplanes.

See Figure 1 shows a plan view of the thick-wing XS-1 airplane and indicates the approximate location of the strain-gage stations. The gages are wired as four active arm bridges and are located on the front and rear spars of the left wing and on both sides of the horizontal tail. Shear gages are mounted on the spar webs and bending-moment gages are mounted on the skin. The strain-gage circuits operate on direct current and the output is recorded on an oscillograph equipped with galvanometer elements which are damped to give accurate frequency response up to 100 cycles per second.

The strain gages have been calibrated by applying known loads to the structure and by determining equations from which the measured responses of these gages may be interpreted directly as loads. The measured loads represent a combination of aerodynamic and inertia loads. The inertia loads, however, may be determined from a knowledge of the accelerations and mass distributions. The accuracy of the measured wing shears is within _____ pounds, the accuracy of the measured tail shears is within _____ pounds, while the accuracy of wing and tail bending moments is within _____ inch pounds.

Information on wing and tail loads at Mach numbers below 0.92 has been published for the SX-1 airplanes. The data presented here are those obtained at higher Mach numbers correlated with data previously published.

Figure 2 shows the variation of the wing, tail, and fuselage loads in terms of percent of total airplane load for the airplane with the 10-percent-thick wing. The data are given for Mach numbers to 1.07 at an airplane normal-force coefficient of 0.35. For comparison, the division of total load is shown for Mach numbers below 0.8 for the same airplane normal-force coefficient. It may be noted that the mean curves through the experimental data shown indicate only a slight variation in the division of total airplane load from the mean of the data for subsonic Mach numbers. The tail load in the transonic region has become more negative as a result of the aft movement of the center of pressure on the wing. It has not been definitely established, at this time, whether the variations in the experimental data are actual or merely due to scatter. However, the significant fact brought out by the plot is that no radical changes have occurred in the division of load between wing, tail, and fuselage.

Figure 3 shows the variation of wing lateral center of pressure for the thick-wing airplane. The lateral center of pressure is obtained by dividing the root aerodynamic bending moment by the root aerodynamic shear and the resulting center of pressure is expressed in terms of percent of the wing span outboard of the wing strain-gage station. The wing strain-gage station is located 31 inches outboard of the fuselage center line. The data, for this Mach number range, are for airplane normal-force coefficients from 0.2 to 0.4. The center of pressure is seen to remain constant at a point about 50 percent outboard of the gage station. The incompressible lifting line and strip theory are 52 percent and 56 percent wing semispan. Thus, the lateral center of pressure is not only constant in the Mach number range considered but is also very close to incompressible-flow theoretical values.

The determination of buffeting boundaries and magnitude of buffeting loads is an important phase of transonic research. Tail buffeting occurs when the air flow over the wing breaks down and the turbulent wing wake strikes the horizontal tail. The breakdown in air flow over the wing may occur as a result of a stall or formation of a shock wave on the wing. Figure 4 shows maximum lift and buffet boundaries for the thick-wing airplane. Airplane normal-force coefficient is plotted against Mach number. Data to M = 0.7 have been published previously. The solid line labeled "abrupt stall" represents maximum lift obtained in abrupt maneuvers. There is a decrease in the maximum normal-force values with Mach number until a Mach number of 0.57 is reached. The maximum lift then increases until a peak normal-force coefficient is reached at a Mach number of 0.65, after which a general decrease occurs with increasing Mach number. In the region beyond M = 0.7, the limit normal-force coefficient may possibly be exceeded if the stabilizer is used for longitudinal control. The upper limit of the maximum lift boundary was obtained with a stabilizer setting of _____. The peaking of the maximum-normal-force-coefficient curve near a Mach number of 0.65 is characteristic of low-drag airfoils where, for a small range of Mach numbers, the chordwise broadening of the low-pressure region more than offsets any reduction in negative pressure peaks that may occur with increasing Mach number.

The gradual stall line at Mach numbers below 0.7 corresponds to level-flight stalls or turns with slow pitching rates. Higher pitching rates in this region give higher normal-force-coefficient values. Beyond a Mach number of 0.7, buffeting occurs regardless of the pitching rate of the maneuver at points indicated by the dashed line. The effect of Reynolds number on the buffet boundaries has not been established as these data are for Reynolds numbers greater than ten million.

Of interest in any discussion of buffeting are the buffeting frequencies. For the XS-1 airplane with the thick wing, it was found that the predominant measured buffeting frequencies corresponded to the wing and tail structural primary bending frequencies; for example, the wing structural primary bending frequency is 13.3 cycles per second and the lowest frequency recorded in flight was approximately 14 cycles per second, which corresponds to the tail structural primary bending frequency. The buffeting frequencies were found to be independent of airspeed.

With the buffet boundary lines established, it becomes necessary to relate the magnitude of the buffet load to some aerodynamic parameters so that the buffet load intensity may be estimated throughout the transonic and supersonic speed ranges. Figure 5 shows the variation of resultant wing and tail buffet load with airplane normal-force coefficient for a Mach number range of 0.7 to 0.9. These data are for the thick-wing airplane. These data indicate that, within the Mach number range, the magnitude of the tail buffeting load increases with an increase in airplane normal-force coefficient and are thus in agreement with the theory that higher tail buffeting loads are associated with higher stall angles of attack. These data also indicate that, at values of airplane normal-force coefficient of 0.35 and 0.5 for the test altitude range from 30,000 feet to 40,000 feet, the magnitudes of the buffet loads are relatively small. The data for the variation of wing buffeting load with airplane normal-force coefficient, for the same Mach number range, increase with increase in airplane normal-force coefficient. This may be expected since, for a constant Mach number, the increased angle of attack required to obtain higher airplane normal-force coefficients would result in a larger area of wing being affected by a breakdown in the flow over the surface, resulting in increased buffeting.

Sufficient data have not been obtained to extrapolate values of resultant buffeting loads of other altitudes. It is expected, however, that the magnitude of the loads will increase with decreasing altitude and increasing dynamic pressure. Further tests will determine, more concretely, the effects of this parameter.

Limited data from later flights to a Mach number of 1.02 at test altitudes of 30,000 to 40,000 feet indicate that the buffeting loads are of the same magnitude as those in the M = 0.7 to M = 0.9 range. These additional data also show that, at the test altitudes through which flights have been made and to the Mach numbers and airplane normal-force coefficients attained, the buffet loads are light and well within the design load limits of the airplane.

With a buffet boundary established for the 10-percent-thick-wing airplane and a trend in the variation of the magnitude of buffeting load with some aerodynamic parameters ascertained, it is of interest to compare these data with those for the thin-wing airplane. Figure 6 shows a comparison of buffet boundaries and limit lift regions as determined from flight tests of the two airplanes. Airplane normal-force coefficient is again plotted as a function of Mach number. These data show the boundaries for the two airplanes where buffeting will occur regardless of the type of maneuver performed. Several points between the buffet boundary and maximum lift have been attained without the pilot noting buffeting. Buffeting boundaries were established on the basis of recorded data from flights where pilotÃŒs reports have not indicated any such phenomena. Below the lower boundaries of the presented envelopes buffeting was not discernible. The upper limits of the envelopes indicate where peak airplane normal-force coefficients were obtained with the use of elevator for the constant stabilizer settings shown in figure 6. The upper limit is not a line of definite demarcation but simply indicates the extent to which the airplane has been flown through the buffet region in an attempt to attain maximum lift. The uppermost limit is determined from test data where the elevator was full up against the stop. There was no indication from the test records that maximum airplane lift had been reached when full-up elevator was used. These data show that as expected the thick-wing airplane will encounter buffeting at lower Mach numbers for the same airplane normal-force coefficient than the thin-wing airplane. Also indicated is that, with the constant stabilizer incidences as shown, the thick-wing airplane can be flown, for the same Mach number, to higher values of airplane normal-force coefficient. It appears, however, that at the higher Mach numbers the two boundaries tend to converge.

The limited strain-gage instrumentation in the thin-wing airplane did not permit an accurate determination of buffet loads encountered in flight tests of that airplane so that the variation of wing and tail buffet loads for this configuration has not been established. However, an indirect comparison of the magnitudes of the buffet loads on the two airplanes was made. From data on the thick-wing airplane, it was found that the change of airplane center-of-gravity normal acceleration ______ during buffeting is similar to the variation of buffet load with ______ for the given Mach number range. That is, for points where higher buffet loads were encountered, higher values of _____ were obtained. A comparison of the variation of ________ of the two airplanes during buffeting with airplane normal-force coefficient was made and it was found that the trend noticed in the thick-wing airplane, higher values of _____ with higher values of airplane normal-force coefficient in the same Mach number range, was also true of the thin-wing airplane. The values of ______ for the thin-wing airplane were approximately 40 percent lower than values of ______ for the thick-wing airplane for buffeting encountered at the same _______ within the same Mach number range. It is believed, therefore, that the variation of the resultant buffet loads on the thin-wing airplane will follow that on the thick-wing airplane with the buffet load magnitudes being lower than those encountered on the thick-wing airplane.

With the preceding data in mind, it is of interest to discover what maximum ______ could be attained and what buffeting would be experienced in a pull-up at supersonic speeds to high _____ where the stabilizer was used for longitudinal control. Figure 7 shows time histories of pertinent measured quantities for this maneuver. The occasion for this recovery was an attempt to reach high Mach numbers with four-rocket operation at altitudes between 45,000 ad 50,000 feet. Due to faulty rocket operation, the pilot had to dive the airplane with only three rockets on in order to obtain higher speeds. At approximately 36,000 feet, the instruments indicated to the pilot that a terminal Mach number had been reached for the power condition and dive angle existing. The pilot attempted recovery by use of elevator. At a Mach number of 1.2, for an incremental change in elevator position of about _____, an increment in normal acceleration of about 0.8g was experienced. The rate of recovery appeared small to the pilot, and he actuated the horizontal stabilizer to complete the recovery. The stabilizer was moved from _____ incidence to _____ incidence at the rate of _____ per second in the manner shown. The elevator force was slacked off shortly before the stabilizer movement and an airplane normal-force coefficient of 0.8 was attained at a Mach number of 1.17. There is no indication from these data that this is the maximum lift, obtainable at this Mach number. All three rockets were then shut down, and Mach number dropped off rapidly with cessation of powered flight. From the time where a _____ or 0.8 was reached at M = 1.17 buffeting of barely discernible amplitude took place and this type of buffeting continued to a point where at a Mach number of 0.88 and a _____ of 0.77 buffeting or an appreciable amount and of constant amplitude appeared. This continued to the time where, at a Mach number of 0.72 and an airplane normal-force coefficient of 0.74, the buffeting ceased to be constant and became intermittent, associated with each up-elevator movement the pilot made. Or, as the airplane attained a higher airplane normal-force coefficient in the speed range, buffeting started; and as the pilot moved the elevator in such a way as to reduce the angle of attack, the buffeting ceased. The pilot then returned the stabilizer to the original trim position and at this instant the longitudinal oscillation (present from the point where the Mach number dropped off with the airplane maintaining high _____) increased in amplitude since the stabilizer movement was in phase with the oscillation. The oscillation ceased after about 2 seconds following return of the stabilizer to the original trim position. The intermittent buffeting ceased with the first large movement of the stabilizer toward the original position for trim.

Figure 8 shows a history of the mean values of _____ attained with varying Mach number during the recovery and throughout completion of the recover superimposed upon a plot of the buffeting boundary and limit lift envelope of the thin-wing airplane. As was mentioned previously, from the point at which the peak _____ was attained a the highest Mach number (M = 1.17) to the point marked "buffeting starts," there was a negligible amount of buffeting as indicated by the trace of the center-of-gravity normal-acceleration record. At this point, buffeting of an appreciable magnitude appeared and continued constantly to the point marked "buffeting stops." It was at this Mach number also that the intermittent buffeting occurred. Whenever the airplane attained a normal-force coefficient of the magnitude of the _____ marked as "buffeting stops," buffeting occurred. When the _____ was reduced, buffeting ceased.

A comparison of the magnitude of this transonic buffeting with other values of buffet intensity was made. It was found that the magnitude of buffeting realized in this maneuver is the same as that attained in a pull-up into the buffet region at subsonic Mach numbers at the same altitude.

Figure 1.- Plan view of thick-wing XS-1 airplane showing approximate location of strain-gage stations.

Figure 2. Variation of wing, tail, and fuselage load in terms of percent of total airplane load with Mach number for CNA = 0.35 (10 percent-thick-wing airplane.

Figure 3.- Variation of wing lateral center of pressure with Mach number.

(CNA, 0.2 - 0.4; 10-percent-thick-wing airplane.)

Figure 4.- Maximum lift and buffet boundaries for 10-percent-thick-wing airplane.

Figure 5.- variation of wing and tail loads with CNA for a Mach number range 0.7 to 0.9, 10-percent-thick-wing airplane.

Figure 6 - Comparison of buffet boundaries and limit lift regions for thick-wing and thin-wing airplanes.

Figure 7.- Variation of measured quantities during a supersonic recovery utilizing stabilizer for longitudinal control.

Figure 8.- Mean values of CNA plotted against Mach number during a supersonic recovery of thin-wing airplane.

PRESSURE-DISTRIBUTION MEASUREMENTS

By De E. Beeler

INTRODUCTION

As was mentioned by Mr. Goodman in the preceding paper entitled "Over-All Loads and Buffeting Measurements," the present phase of testing the XS-1 airplane with the 10-percent-thick wing includes the measurement of stability and control characteristics of the airplane and the measurement of the total wing and tail loads. On completion of these tests, it is planned to determined the distribution of load over the wing and tail from measurement of surface pressures. In order that some load distribution data might be made available at an early date, additional instrumentation was included in the XS-1 incorporating the 8-percent-thick wing. Measurements have been made of the chordwise pressure distribution over a representative section of the wing and tail.

DESCRIPTION

Pressures over the upper and lower surfaces were measured at the midsemispan station of the wing and tail as shown in figure 1. The wing section included a wing spoiler located on the upper surface and the landing flap. Both the wing and the tail incorporate an NACA 65-series airfoil section. The wing thickness is 8 percent of the chord and the tail thickness is 6 percent of the chord. The surface conditions of the section were smoothed by waxing but the spoiler and the flap presented some irregularities in the wing contour.

The data to be presented were obtained during the flight described in the paper presented by Mr. Goodman where the airplane was dived to high Mach numbers and recovered from the dive at fairly high load factors.

INDIVIDUAL SURFACE PRESSURE DISTRIBUTIONS

In figure 2 distributions are given for the upper and lower surfaces which are identified by the circular and square symbols. The distributions plotted are the pressure coefficient ______ where ______ is the difference between the free-stream static pressure and the static pressure at the surface and q is the free-stream dynamic pressure. The pressure coefficients corresponding to sonic velocity are noted by the short lines (_____). Also shown along the chord axis is the approximate location of the normal shock occurring on the upper surface. The location of shock has been determined from an inspection of the pressure distribution, of the variation of local pressure coefficient with Mach number, and of the time history of the individual surface pressures. In general, the location at which supersonic flow over the forward portion of the airfoil is followed by an abrupt pressure recovery has been taken as indicative of the shock.

The distributions shown in figure 2 are for free-stream Mach numbers of 0.75, 0.76, 0.77, and 0.80 and are for a given airplane normal-force coefficient. The normal-force coefficient _____ is based on the wing area and is obtained from a measurement of dynamic pressure and normal acceleration. It may be noted that the normal shock and the corresponding pressure peak on the upper surface moves from about 35 percent of the chord to 55 percent of the chord at a Mach number of 0.80. The distribution over the rear portion of the upper surface remains essentially the same for the Mach number range presented in figure 2. Pressure at the 10-percent-chord station fluctuates rapidly through the range in pressures shown. The distributions are faired through values obtained at the selected Mach numbers and normal-force coefficients. Data obtained on later flights show this fluctuation to occur at other chord stations for different Mach numbers and normal-force coefficients. The shape of the distribution over the lower surface remains essentially the same at the Mach numbers shown.

In figure 3 similar distributions are shown for Mach numbers of 0.82, 0.85, 0.86, and 0.87 at _____ = 0.33. The compression shock has moved to 60 percent of the chord at a Mach number of 0.82 and remains in the vicinity of the 60-percent-chord area fluctuating between 60 percent and 65 percent chord as the Mach number is increased. At Mach numbers of 0.86 and 0.87 there is an indication of separation to the rear of the shock as shown by an increase in negative pressure due to thickening of the boundary layer. The distribution over the lower surface changes considerably with an increase in Mach number above 0.82 with the formation of shock at about 55 percent of the chord at a Mach number of 0.85 and the shock moves rearward to about 65 percent chord at a Mach number of 0.87. There is no indication of separation on the lower surface at these Mach numbers.

Distributions are presented in figure 4 for Mach numbers of 0.88, 0.89, 0.90, and 0.92 at _____ = 0.33. The location of the compression shock in the upper surface remains in the vicinity of 60 percent of the chord up to a Mach number of 0.90 at which time the shock continues again to move rearward approximately reaching the 80-percent-chord station at a Mach number of 0.92. A region of supersonic flow exists over practically the entire upper surface at Mach numbers greater than 0.90. The compression shock on the lower surface moves rearward reaching the 85-percent-chord station at a Mach number of 0.92. Supersonic speeds occur over approximately 70 percent of the lower surface.

Figure 5 shows pressure distributions for Mach numbers of 0.95, 0.96, 0.98, and 0.99 at _____ = 0.33. The compression shock on both surfaces as indicated from these tests is located near the trailing edge. The distribution shape is essentially the same for both surfaces as the Mach number is increased in this range.

Figures 6 and 7 show that the chordwise pressure distributions for Mach numbers from 1.03 up to 1.09 at a constant ______ are similar for both surfaces. A reduction in up load over the leading edge of the upper surface is indicated as the Mach number is increased.

The pressure distributions for Mach numbers of 1.14 to 1.25 shown in figures 8, 9, and 10 are for a slightly lower value of _____ than the previous figures (0.16). Nevertheless these distributions are of the same general type over both the upper and lower surfaces as those above a Mach number of 0.95.

DIFFERENTIAL PRESSURE DISTRIBUTIONS

Up to this point, individual surface distributions have been considered. The individual loadings have been resolved into total section loadings and are shown in figures 11, 12, 13, and 14 for a section normal-force coefficient ____ of 0.35. From figure 11 it may be noted that the load shifts to the rear as the Mach number is increased from 0.75 to 0.80.

Figure 12 shows, for the same section normal-force coefficient (0.35), section loadings for a range of Mach number from 0.87 to 0.92. The occurrence of a down load in the vicinity of the 80-percent-chord station is a result of the movement of the compression shock to the rear on the lower surface while the shock on the upper surface is essentially stationary.

For a Mach number of 0.95 to 1.19 at a section normal-force coefficient of 0.35 ( figs. 13 and 14) the total loading distribution approaches that of a rectangular shape having peak loading conditions at the leading and trailing edges. The magnitudes of the peak loads are about _____ to 2 times that loading over the midchord section. The general shapes of the distributions for this Mach number range are similar.

SECTION CHARACTERISTICS

A summary of the section aerodynamic characteristics is shown in figures 15 as a function of free-stream Mach number. The location of the center pressure C.P. in percent of local chord is shown. At a Mach number of 0.75, the center of pressure is located at 25 percent chord and moves rearward to about 41 percent chord as the Mach number is increased to 0.85. This is a result of the rearward movement of the upper-surface shock and the corresponding peak load to about 60 percent of the chord. As the Mach number is increased further, the center of pressure moves forward to about 25 percent chord at a Mach number of about 0.88. It is within this Mach number range that the upper-surface shock remains essentially stationary and the formation and rearward movement of the lower-surface shock occur. As the Mach number is increased to 0.95, the center of pressure again moves rearward to about 48 percent of the chord due to the rearward movement of the upper-surface shock. At Mach numbers greater than 0.95 there is a gradual rearward movement of the center of pressure due to the larger rate of load reduction over the leading edge of the upper surface than over the rear portion of the section, the center of pressure being located at about 51 percent chord at a Mach number of 1.25.

The variation of the section pitching-moment coefficient about the quarter-chord point _________ is also shown in figure 15. From the changes in pitching moment of the wing as the Mach number is increased from 0.75 to 0.95 it would be expected that large change in trim of the airplane would occur. Also, large load factors may result from these trim changes for flights at low altitude. The occurrence of large load factors, of course, depends on the rate at which the airplane passes through this Mach number range and on the rate and magnitude of the trimming forces that may be available to the pilot.

From measurements of the surface pressures over the rear 15 percent of the chord, a value of section hinge-moment coefficient _____ has been obtained. The changes in the hinge-moment coefficients occurring within a Mach number range of 0.75 and 0.95 are due to the increased loading of the section trailing edge.

The ratio of the section normal-force coefficient to the airplane normal-force coefficient is shown in the lower curve. At Mach numbers up to 0.9 the ratio is fairly constant. The ratio increased rather suddenly at a Mach number of approximately 0.93 and remained at this new level at higher Mach numbers. No conclusions can be made of this variation until additional section loadings are obtained for more spanwise stations.

VARIATION OF MAXIMUM PRESSURE COEFFICIENT

Figure 16 shows the variation of the maximum pressure coefficient ______ measured over the upper and lower surfaces with free-stream Mach number M for section normal-force coefficients of 0.2 and 0.3. The circles correspond to a value of 0.2. Included also are the local Mach number lines _____ for 1.0 and 1.5. The data presented in this figure were obtained after the wing section had exceeded the critical speed. A decrease in the maximum pressure coefficient for the upper surface occurs with increasing Mach number up to a Mach number of 0.85. In the Mach number range from 0.85 to about 0.95 when the shock movement is varying, the pressure coefficient tends to remain somewhat constant resulting in an increase in the maximum local Mach number with freestream Mach number. As the Mach number is increased above 0.95 where the shock is located near the trailing edge, the pressure coefficient tends to follow a line of constant local Mach number equal to about 1.5. Sonic velocity is first reached on the lower surface at a free-stream Mach number of about 0.82. In the Mach number range from 0.85 to 0.95 remains fairly constant which results in an increase in local Mach number with increasing free-stream Mach numbers up until the shock reaches the trailing edge at M = 0.95. Further increase in free-stream Mach number reduces the maximum pressure coefficient occurring over the lower surface and the pressure coefficient tends to follow a line of constant local Mach number equal to about 1.3.

VARIATION OF DISTRIBUTION WITH

Figures 17, 18 and 19 show some pressure distributions for a range of section normal-force coefficient at a constant Mach number of 1.16. The section normal-force coefficient, obtained from a mechanical integration of the distributions, are noted on each distribution. As the section normal-force coefficient is increased from 0.19 to 0.93, the shape of the distribution of the rear portion of the upper surface remain essentially the same. The loading over the forward portion of the upper surface, however, increases with increasing local Mach number resulting in a distribution over the upper surface which approaches that of a rectangular shape as the section normal-force coefficient is increased. Local Mach numbers of 2.0 were obtained over the rear portion of the upper surface at the highest section normal-force coefficient obtained (________).

TAIL-SECTION PRESSURE DISTRIBUTIONS

The evaluation of the pressure distribution over the tail section has not been completely evaluated. However, some data at high Mach number which may be of interest have been obtained. The distribution over the upper and lower surfaces are shown in figure 20 at a constant stabilizer incidence angle _____ of 2.25____ for elevator angles ____ of _____down, _____ up and _____ up. The elevator hinge line is at 80 percent chord. These data are for a Mach number of 1.16. Pressures measured over the elevator show that deflecting the elevator produces the expected load change, that is, an increase in negative pressure over the upper surface with increasing down elevator but it may be noted that the load changes are small. The upper surface pressures over the stabilizer show some change with changing elevator angle. The main portion of this change is due to a change in airplane normal-force coefficient. The data obtained over the lower surface were not sufficient to establish the distribution over the forward portion.

Similar distributions are shown in figure 21 for a range of stabilizer incidence angle for a constant up elevator angle of _______. These data are for a Mach number _______.

The data presented were all that were available at this time. Additional chordwise pressure-distribution data have been obtained over the wing at two span stations for Mach number greater than 1.0 and are now being evaluated.

Figure 1.- Plan form of airplane and airfoil profiles.

Figure 2.- Wing pressure distributions. M =0.75, 0.76, 0.77, and 0.80; CNA = 0.33.

Figure 3.- Wing pressure distributions. M = 0.82, 0.85, 0.86, and 0.87; CNA = 0.33.

Figure 4.- Wing pressure distributions. M = 0.88, 0.89, 0.90, and 0.92; CNA = 0.33.

Figure 5.- Wing pressure distributions. M = 0.95, 0.96, 0.98, and 0.99; CNA = 0.33.

Figure 6.- Wing pressure distributions. M = 1.03, 1.04, 1.05, and 1.06; CNA = 0.33.

Figure 7.- Wing pressure distributions. M = 1.07, 1.08, and 1.09; CNA = 0.33.

Figure 8.- Wing pressure distributions. M = 1.14, 1.15, 1.16, and 1.17; CNA = 0.16.

Figure 9.- Wing pressure distributions. M = 1.18, 1.19, 1.20, and 1.22; CNA = 0.16.

Figure 10. Wing pressure distributions. M = 1.23 and 1.25; CNA = 0.16.

Figure 11.- Section chordwise loading. M = 0.75, 0.76, 0.77, and 0.80; cn = 0.35.

Figure 12.- Section chordwise loading. M = 0.87, 0.88, 0.90, and 0.92; cn = 0.35.

Figure 13.- Section chordwise loading. M = 0.95, 0.96, 0.98, and 0.99; cn = 0.35.

Figure 14.- Section chordwise loading. M = 1.03, 1.05, and 1.19; cn = 0.35.

Figure 15.- Summary of section aerodynamic characteristics.

Figure 16.- Maximum pressure coefficient plotted against free-stream Mach number for upper and lower surfaces. cn = 0.2 and 0.3.

Figure 17.- Wing section pressure distributions. cn = 0.19, 0.21, 0.23, and 0.27; M = 1.16.

Figure 18. Wing section pressure distributions. cn = 0.28, 0.39, 0.48, and 0.57; M = 1.16.

Figure 19.- Wing section pressure distributions. cn = 0.60, 0.70, 0.85, and 0.93; M = 1.16.

Figure 20.- Tail section pressure distributions. Stabilizer angle, 2.25° ; elevator angles, 7.8° down, 4.1° up, and 9.8° up; M = 1.16.

Figure 21.- Tail section pressure distributions. Elevator angle, 4.5° up; stabilizer angles, 2.25° , 1.45° , and -0.4° ; M = 1.16