Source: NASA Historical Reference Collection, NASA History Office, NASA Headquarters, Washington, DC.

9 JANUARY 1948

U-55038- 82

Photo of the XS-1 aircraft


14 OCTOBER 1947




As a result of the accelerated transonic flight test program conducted on the XS-1 airplane by the Air Materiel Command at the Muroc Air Field with the cooperation of the National Advisory Committee for Aeronautics, the Secretary of the Air Force has directed that all reliable data available from this program be prepared in a suitable form for presentation to and use by the Aircraft Industry of the United States and selected government agencies. This brochure contains the initial compilation of data available and other pertinent details regarding the project.


Introduction--Purpose of Research Projects

Airplane Parameters

Specification Performance

Resume of Project

Development of the XS-1 Airplane

The Operation of the XS-1 Airplane

Instrumentation, Airspeed Calibration, Tests, Results and Conclusions

Future Program



The limited knowledge of aerodynamics and flight performance in the transonic and supersonic speed ranges possessed in 1944 dictated the initiation of research projects which would increase our meager fund of fundamental and factual information in this field of learning. A comprehensive program for the development of purely research aircraft was laid down and a maximum effort was directed toward its immediate implementation.

The AAF initiated development of the XS-1 and the Bureau of Aeronautics soon followed with the D-558. Both of these projects were initiated to provide flying articles which could explore the high subsonic speed range, and possibly the transonic speed range if subsequent flight testing indicated the feasibility of so extending the range of exploration. The XS-l was to be air launched so that enough endurance could be obtained to get aerodynamic data at very high altitude. The D-558 was to be ground launched under its own power and was to provide aerodynamic data at low altitudes in the high subsonic speed range. To carry the exploration to the supersonic range of speeds, the XS-2, XS-3 and a modified D-558 (Phase II) were put under development. The XS-2 and the modified D-558 were to have sweptback wings so that factual information could be obtained on the performance of such a wing configuration both in the transonic and supersonic speed range, even though it was questionable whether any particular advantage would be realized from sweptback wings once the transonic range was pierced. To explore the optimum wing configuration for higher supersonic speeds, the XS-3 was conceived and at present is in the study phase of development, with the USAF and the Bureau of Aeronautics participating in the project. The study has progressed to the point now where an optimum configuration has been evolved and it is expected that the actual fabrication of a flying article will be commenced in the latter part of 1948. One other research project, the XS-4, was initiated to explore stability and control of a sweptback wing and semitailless configuration in the high subsonic speed range. The XS4 has no horizontal tail surfaces, the empennage consisting only of a vertical fin and rudder. Lateral and longitudinal control will be accomplished by elevons located on the trailing edge of the sweptback wings.

In all of these research projects the National Advisory Committee for Aeronautics contributed a vast amount of technical assistance, both in their design stages, and later in the reduction of flight test data of those which have been flown. The cooperation on the part of the NACA has contributed immeasurably to the success we have gained so far.

When these projects are completed and flight testing accomplished, we hope to have covered the field of supersonic flight research to the extent that we can design, build and fly supersonic tactical planes in the future with the surety we build subsonic planes today. Although the XS-1 has materially increased our fund of knowledge of aerodynamic phenomena in the lower supersonic speed range, there is still much to be learned before we can predict the true nature of ultra high speed flight at any given altitude and under varying flight conditions.

How far we can penetrate the supersonic range and still preserve the tactical utility of our combat aircraft can be determined only after we have been able to actually operate tactical aircraft in this unfamiliar region.

Three view drawing of XS-1 aircraft

Drawing of the inboard profile of the pressurized fuel system of the XS-1 aircraft



LENGTH--Maximum 31'00"

HEIGHT--Maximum 10'10"

SPAN 28'00"



WING AREA--Net 102.5 sq. ft.


WING-NACA 65l-108(a=1.0)

TAIL-NACA 651-006

TAPER RATIO (Root Chord/Tip

Chord) 2:1



SWEEPBACK (L. E.) 52'52"




Design Criteria

LOAD FACTOR--Ultimate 18g



ANALYSIS 8,410.0 lbs.


GROSS WEIGHT--(launched) 13,034 lbs.

EMPTY 6,784.9 lbs.



LIQUID OXYGEN (311 gals.) 2,920.0 lbs.

ALCOHOL WATER (293 gals.)

2,100.0 lbs.

NITROGEN (17.5 cu. ft. at 4500 lb.

Pressure) 301.0 lbs.



50,000 ft.
916 mph.
70,000 ft.
898 mph.


652 mph.

Initial conference between NACA and Air Materiel Command on an Airplane of Transonic Design 15 March 1944
Contract issued to Bell for Design and Construction of Three Airplanes 16 March 1945
Mock-Up Inspection 10 October 1945
Demonstration Flights
1) First Contractor Guide Flights, Orlando, Fla.
3 January 1946
2) First Contractor Guide Flights, Muroc, Calif.
10 October 1946
3) First Contractor Powered Flights, Muroc, Calif.
9 December 1946
Air Force Accelerated Transonic Program Initiated
1) First Air Force Glide Flight
6 August 1947
2) First Air Force Powered Flight
29 August 1947
3) First Transonic Flight by Air Force
14 October 1947
NACA Research Program
1) Acceptance of Airplane No. 46-63 for stability, control and loads investigations
25 September 1947
2) First NACA Glide Flight
21 October 1947
3) First NACA Powered Flight
16 December 1947



R. M. Stanley, Vice President, Engineering


R. J. Sandstrom, Chief, Technical Services

Bell Aircraft Corporation

Early in December, 1944 while the aircraft industry was still concentrating on the production of airplanes with which to end the war, the Army Air Forces discussed with the Bell Aircraft Corporation the possibility of designing a man-carrying supersonic research airplane whose tentative specifications included a minimum speed of 800 m.p.h. for 2 to 5 minutes at 35,000 feet or above and capable of carrying 500 pounds of recording instrumentation. A decision was jointly reached that this airplane should adhere insofar as possible to a conventional arrangement in order to prove whether or not such an arrangement is practical at transonic speeds. Similarly, it was decided that relatively conventional methods of construction to conventional tolerances which could be duplicated in production aircraft would be employed. The experimental supersonic XS-1 airplane was the result.

It was immediately recognized that the many unknowns in this problem would necessitate probing every available source of information. A series of conferences were held with governmental agencies such as NACA, AAF, Bureau of Standards, Army Proving Grounds at Aberdeen, power plant manufacturers and institutions such as GALCIT. Many studies were made investigating various types of power plants including turbo jets, ram jets, and rockets, including unconventional configurations such as the Canard. (Information on the effects of sweepback were at that time quite fragmentary and its use was ruled out on those grounds). Consideration was given to air launching and the use of skids for landing. Wartime considerations of possible tactical development of the aircraft, however, dictated its being able to take off under its own power, and hence a tricycle landing gear was used. The ability to take off under its own power likewise ruled out any serious consideration of ram jets (about which little was known at that time) as a possible power plant.

The XS-1 is of conventional configuration having a wing area of 130 square feet, an aspect ratio of 6, and a wing loading at the beginning of its flight of 100 lbs./sq.ft. It is 31 feet long and has a span of 28 feet. The pilot is contained in a sealed cabin of ogive shape. Behind him within the fuselage are tanks carrying liquid oxygen and alcohol feeding for 4.1 minutes a 6,000 pound rocket motor mounted in the tail. At launching it weighs 13,034 pounds, and lands weighing 6,784 pounds at about 135 m.p.h.

It was anticipated that wing wake interference over the horizontal tail would present one of the greatest difficulties of transonic flight, and for this reason the horizontal stabilizer was placed as high as was considered feasible above the wing wake and was made rapidly adjustable at 1/second through an angle of 15 to accommodate large possible changes of trim. The control surfaces were made conventional and unboosted since the small size of the airplane precluded the probability of unmanageable stick forces.


After reviewing the power plant field, only the turbo jet and rocket were seriously considered for this application. Preliminary studies indicated that an airplane designed around existing turbo jets, using the thrusts given in the engine specifications, would not attain the desired speeds at the desired altitude. It was found that speeds in the region of only M = .90 could be obtained near sea level and altitude performance was considerably less. Turbo jet manufacturers were then contacted on the possibility of increasing their engine outputs at altitude by 100 percent or more for short periods of time by any means they could devise. The general reaction was one of great interest. However, very little theoretical work and practically no testing had been done on thrust augmentation at that time, and the pressure on production models, brought on by the war, prevented any development on such a specialized project.

Although the fuel consumption of an all-rocket powered airplane was high, the rate of climb was also high, averaging better than 20,000 ft/min. between sea level and 35,000 feet with a climbing speed of 500 m.p.h. Thus, the fuel required for climb was relatively low and that required to accelerate from climbing speed to the desired test speed was less than with a combination turbo jet and rocket.


Using the best available information on drag at transonic and supersonic speeds, it was possible to achieve the minimum speed requirements of 800 m.p.h. as specified by the Air Force and simultaneously to achieve altitudes and rates of climb far beyond those of any aircraft hitherto designed. The performance curves shown in Figures 4,5, and 6 are based on expendable fuel load of 8160 pounds and on a launching gross weight of 13,034 pounds.

The close of the ware made available to us the B-29 originally desired, permitting air launching with its resultant increase of operational safety. With a wing loading of 100 pounds/sq. ft. and a large quantity of high energy fuel aboard, the first few seconds after ground take-off would have been quite hazardous for the test pilot. This hazard is eliminated when his flight begins at a safe altitude thousands of feet above the earth, permitting him to jettison fuel prior to landing in case of engine failure, or bail out in case of more serious mishap. Almost equally attractive is the tremendous increase in potential performance resulting from launching the airplane with the high potential and kinetic energy which the B-29 can supply by virtue of its altitude and speed characteristics. The flight can now begin at a point which it previously would have reached only after the expenditure of better than half of its fuel and hence the potential speed and altitude of the XS-1 are nearly doubled by this unique but highly practical method of launching.

Except at very low altitudes, speed and rate of climb are not limited by drag but are limited by fuel capacity. Hence, more endurance automatically means more performance. Likewise, a premium is placed upon reduction of structural weight to the barest minimum in order to reduce the important factor of inertial in an aircraft of inherently short endurance.

In the course of these design studies, it was likewise found that almost without exception the maximum in rate of climb and high speed can be achieved by the use of motors of large capacity which, although they consume fuel rapidly, impart to the airplane higher performance than is available from a more modest motor which consumes fuel at a lesser rate. The XS-1 is a compromise since it is felt that giving it shorter endurance (irrespective of improvements in high speed) would have defeated the purpose of scientific research.


After a comparitive analysis of the design studies made around the various power plants, it was decided to proceed with the all-rocket powered airplane. Many propellants were considered before liquid oxygen and ethyl alcohol were selected. Hydrogen peroxide was discarded because the motors in operation at that time exhibited low specific impulses. It was available in this country only in limited quantities, and it must be handled with a great deal of care. Acid and Aniline was not considered suitable for a man-carrying aircraft. These propellants are spontaneously combustible, which necessitates special precautions being taken to keep them separated, since any dual leak into the air might initiate a disastrous fire. Acid is also injurious to personnel, which complicates the handling problem. Nitro-Methane, being a monopropellant, would simplify the fuel system considerably, but it is subject to detonation under certain conditions that were not to well understood, and it was not felt that it had been developed to the point of reliability desired. Gasoline and liquid oxygen were studied, but, with a regeneratively cooled motor, it would have been necessary to carry along a third tank containing water for this purpose, as gasoline alone is not a suitable cooling medium. The propellants selected, liquid oxygen and alcohol, are readily available, have a good specific impulse, are relatively safe and easy to work with, are not spontaneously combustible, and are non-injurious to personnel (small spatters of liquid oxygen are not harmful). The motor is regeneratively cooled by circulatingthe fuel through a cooling jacket before it is injected into the combustion chamber. To aid in cooling, one part of water is mixed with three parts of ethyl alcohol. It was found that the addition of this amount of water had a small effect on the overall impluse but aided appreciably in the cooling problem.


The motor assembly which was designed and built by Reaction Motors, Inc., consists of an assembly of four 1500 pound thrust rockets, operating at a chamber pressure of approximately 230 p.s.i.

Because of the high combustion temperature, approximately 5000R, and the relatively long time the motors must operate, some form of cooling was required. Injecting an excess of fuel to reduce the combustion temperature would have been very inefficient; "sweat" cooling, in which the cooling fluid is forced through a porous combustion chamber and nozzle wall and vaporized, was just a gleam in someone's eye and hasn't progressed much farther up to the present time. Regenerative cooling with some film cooling was employed on this motor. That is, the fuel was circulated around the combustion chamber in a cooling jacket before being burned. In the region of the converging section of the nozzle where the hot gases impinge directly against the nozzle wall, the cooling problem is more difficult.

The rockets may be run separately or in combination as selected by the pilot. The motors are started by a small igniter located in the forward end of the cylinder, which ignites a stream of fuel and gaseous oxygen by means of a spark plug. When the chamber pressure reaches a value of approximately 50 p.s.i. the propellant valves open admitting the fuel and liquid oxygen. No provision has been made to throttle the individual rockets and hence the pilot has at his command only 25%, 50%, 75% and 100% of full thrust.

To supply the propellants to the rocket motor, a turbine driven pump was specified. The development of this item was recognized quite early as being one of the items most likely to interfere with the early flight of the airplane since it involved the development of a highly reliable piece of apparatus of a type hitherto unknown. To forestall probable flight delays a supposedly temporary installation of pressurized tanks was employed. This has reduced the airplane's endurance by about 1.5 minutes, has increased its landing weight by approximately 2000 pounds, and has run the landing wing loading up to approximately 55 pounds/sq. ft.

Had the B-29 not become available, the reduction in performance resulting from a reduced fuel weight from 8160 pounds to the 4680 pounds of fuel which could be carried in the pressurized version would have been quite serious, and would have limited the XS-1 to subsonic performance. The B-29 became available, however, in sufficient time to permit even the pressurized version to fulfill nearly all the requirements which had originally been set up. As will be seen from Figure 9, the high speed is in excess of 800 m.p.h., and it is felt that the endurance is adequate for research in this speed range.


Because both existing XS-1 aircraft have been completed in the pressurized version, the remainder of this paper will deal only with design features of this configuration.

Three hundred and eleven gallons of liquid oxygen and 293 gallons of fuel are carried in spherical steel tanks located ahead of and behind the wing respectively. They were designed to operate at an internal pressure of approximately 350 p.s.i. and to a design factor of one and one-half above the yield point. Because of the danger from embrittlement at temperatures of 300F, the loxygen tank was made from annealed stainless steel of 3/16" wall thickness. Not thus handicapped, the fuel tank was made from normalized 4130 steel with a wall thickness of 1/8". Individual segments of the spherical tanks were arc welded together with suitable bosses and provisions for transmitting 18g load into the fuselage shell. Transverse baffles were installed to prevent fore and aft fuel sloshing and the resultant change of center of gravity of the airplane. It has been found that a mildly objectionable sideways sloshing still results and hence it appears that fore and aft baffling will in the future be desirable as well.

Inasmuch as it was considered that the pressurized tanks would be temporary, the basic fuselage structure was unaltered and hence the tanks of spherical shape nested within the fuselage structure were of necessity reasonably small in capacity. Had the tanks gone out to the fuselage external contour as would be possible with integral tanks, a considerable increase in the capacity would have resulted.

No insulation around the loxygen tank was provided. As a result, the external surface of the fuselage in this area frosts up as soon as the tank is filled but the frost is very light and powdery and blows away in flight without altering the fuselage contour in this region. If the tank is allowed to remain filled overnight in a humid climate, solid ice will condense in this area and can build up to objectionable thickness. This is, of course, of no practical consequence since it does not occur during flight operations.

To pressurize the tanks, gaseous nitrogen was employed. Because of the strong oxidizing power of the oxygen and the low temperatures involved, no contamination either in the form of water vapor or oil vapor could be tolerated. It was early discovered that gaseous nitrogen was readily absorbed by the liquid oxygen. This plus the shrinkage resulting from the low temperature of the latter greatly increased our anticipated requirements of gas and it was found necessary to go to 4500 lb. pressure in order to carry sufficient gas on board the airplane successfully to expel the liquid propellants. Because of the high pressures involved, spherical steel tanks varying in wall thickness from " to 5/8" and of diameters varying with the space which could be found for them were located throughout the airplane. These tanks were all manifolded together and connected to a pressure regulator reducing the tank pressure to 1500 p.s.i. Because of leakage troubles, most of the high pressure joints are of welded construction using as few detachable fittings as possible.

Nitrogen gas at 1500 pounds pressure was used for operation of landing gear and flaps to obviate the necessity of batteries or hydraulic pumps toperform this function in more conventional fashion. This method has worked out quite satisfactorily and is considered very reliable and fast in operation.

Three pressure regulators are employed, one as aforementioned to reduce 4500 pound source to 1500 p.s.i., following which it is reduced once more through each of two separate regulators delivering 340 p.s.i. to each of two propellant tanks. The development of these pressure regulators consumed a great amount of design effort due to the unorthodox pressures dealt with and the low temperatures encountered both as a result of gas expansion and chilling from the loxygen tank. Each regulator is controllable from the cockpit and can be adjusted within rather wide limits to deliver any desirable pressure should variations in engine thrust be desired during flight. They may be shut off in the event of emergency, permitting propellant tanks to be depressurized at will.

A source of uncontaminated nitrogen gas at 4,500 p.s.i. was not obtainable, no pumps capable of delivering gas at this pressure were available and hence gas was obtained through the boiling of liquid nitrogen in a nitrogen evaporator.

Referring to Figure 10, the sphere located on top of the evaporator is 36 inches in diameter with a wall of stainless steel 3 inches thick cast in two halves and welded together. This sphere has been hydrostatically tested to 9000 p.s.i.

During the testing of the evaporator we had a rather interesting experience. While operating the evaporator, it was noticed that droplets of liquid dripped from the frosted lines in which flowed liquid nitrogen. This could not be condensed moisturein view of the low temperatures of the lines, and was quite baffling in view of the many pressure tests that had been made. It was finally discovered that liquid oxygen was being condensed out of the air on coming in contact with the cold nitrogen lines.

Structural design criteria were arbitrary and based principally upon ignorance. Since it was not possible to predict with any degree of certainty the loading distributions and possible buffeting in transonic flight, the airplane was designed to withstand 18g ultimate.

Since the wing thickness is only 8% of its chord, the thickness of its covering is quite unusual, being slightly more than one-half inch at the inboard end and machine tapered to conventional thickness at the tip. This tapering process has not proven to be difficult for our shop to perform. Due to its heavy gauge, the problem of local stiffening and preservation of reasonably accurate contour was simple. Some 240 pressure orifices and 12 strain gauges are installed in the left wing with a similar quantity in the empennage.

The stabilizer-construction is quite similar to the wing and has a thickness ratio of only 6%. Again provision has been made for the installation of alternate stabilizer shapes. As previously mentioned, the stabilizer itself is rapidly adjustable in pitch.

The pilot's windshield does not protrude but conforms to the fuselage's ogive nose shape. As a result, visibility during landing is marginal. It is of double thickness plexiglass to prevent fogging; the intervening airspace between the two panels is dehydrated. The pilot's exhalation is likewise dehydrated through a cannister built onto his oxygen mask. It was desired to avoid the use of plate glass in the windshield due to the difficulties in producing double curvature safety glass. The increased potential flight speeds which have been gained by air launching, however, has made it mandatory that a new windshield be installed capable of withstanding elevated temperatures. As a result, plate glass is now being substituted for plexiglass in the windshield.

The pilot enters through a rectangular door on the right side of the fuselage located in such fashion that in case of bail-out he is forced to go downward, thereby increasing his chances of missing the tail. His cabin pressure is sealed with him just prior to launching and no replenishment is necessary for a flight of this airplane's short duration. Internal cabin pressure is maintained at 3 p.s.i. above atmosphere which is sufficient to maintain pilot safety irrespective of altitude. The cabin is tested to leak at a rate no greater than 1 p.s.i./hour at 3 p.s.i. internal pressure. We find this is actually quite conservative, since replenishment from the pilot's exhaled breath through his oxygen mask is actually quite appreciable.

A wheel control was chosen rather than a stick to permit the pilot's use of two hands effectively in case the need arose. A thrust selector control, instrumentation switches, stabilizer control switch, and emerency power shut off are all located on the control wheel so that the pilot need not remove his hands during flight.


To fulfill the role of a mother airplane, a standard B-29 was supplied by the Army Air Forces. Its bomb doors were removed and the XS-1 was suspended from a standard Army Air Forces D-4 bomb shackle. An enclosed ladder was installed between the B-29 and XS-1 in such a manner that the XS-1 pilot or flight personnel could go back and forth between the two airplanes while in the air.

For loading purposes the XS-1 is lowered into a pit, the B-29 rolled over the pit and the XS-1 pulled up and partially into the B-29 by means of a hoist located in the "Mother" airplane.

It could not be predicted with certainty that simply dropping the XS-1 from the B-29 bomb bay would be successful. Accordingly, pressure surveys were made with the B-29 and XS-1 in flight, employing those pressure sampling tubes already built into the XS-1 wing. These data indicated that a positive separating force existed. As a further precautionary measure, one of the B-29 bomb bay actuators was rigged up as an ejector to expel the XS-1 vertically, giving at the same time a slight nose-down pitching moment to insure a clean break-away. In addition a steel tube was pierced through the XS-1 fuselage ahead of the tank area extending about six inches out on each side of the fuselage. Approximately three inches to the rear, slanting wooden guide posts were bolted to the B-29 to guide the XS-1 against backward drift, and to prevent any tendency to yaw during the drop. Fresh red paint was smeared on the tube and the guide rails just prior to flight so that in case the latter functioned, their usefulness would be proven. The flight plan called for the initial drop to be made with inboard B-29 propellers feathered and B-29 flaps neutral, inboard engines at low power, cruising power, then all engines at high power, with each succeeding flight going to higher launching speeds watching the behavior of the XS-1 in each case. The latter was accompanied on all its flights by a "chase airplane," and all early flights were photographed both from cameras mounted on the wing and stabilizer tips of the B-29 and from accompanying aircraft.

During February of 1946, ten glide flights under the foregoing plan were made at Pinecastle, Florida. This site was selected because Muroc Dry Lake is normally flooded at this season and 10,000 ft. runways were available at Pinecastle (although actually only 3000 ft. was ever required). The XS-1 was launched at approximately 27,000 feet and glided to the ground in about 12 minutes, making a normal landing at a speed of about 110 m.p.h. Post flight inspection revealed no fresh red paint and hence no interference. On subsequent flights it was found that the airplane could be safely launched from the B-29 entirely irrespective of the speed, power, or flap setting of the latter and without the aid of the ejectors or any form of guide rails and with the XS-1 pilot in perfect control of his machine at all times during launching.

The foregoing glide flights were made with the airplane in its lightest possible configuration and were completed prior to the receipt of the rocket motor. Following the acceptance of the latter and after extensive ground tests of the complete airplane with its power plant, final preparations were then made for power tests to be conducted at Muroc, California, where facilities were set up to handle this operation. A 15,000 gallon storage tank for liquid oxygen and a 3,300 gallon tank for liquid nitrogen were erected. A mixing tank for the alcohol and water was provided by reworking a small Army fuel trailer and a loading pit was dug.

After four additional glide flights, during one of which a complete fuel load was jettisoned, the first powered flight was made. On December 9, 1946 in perfect weather, the XS-1 was launched at 27,000 feet over Muroc Dry Lake. Approximately 10 seconds later, No. 1 rocket chamber was ignited, followed shortly thereafter by No. 2. The speed of the aircraft picked up so rapidly, however, that thrust was cut back to 25% for reasons of safety and a slow climb was made to 35,000 feet. The cabin noise level was reported very low. At 35,000 feet, 50% thrust was again applied and a Mach No. of .79 quickly attained. All power was then shut down and the airplane glided down to 15,000 feet and with all four rocket motors running 100% thrust was applied during a brief climb. The pilot experienced very high accelerations which he compared to that of a fighter during take-off employing water injection. On its first powered flight, the XS-1 reached a Mach No. of .795 during brief application of 50% thrust, demonstrated satisfactory motor control and general handling characteristics and adhered closely to the flight plan.

During subsequent powered flights, the airplane has gone to successively higher altitudes and higher arispeeds until on October 14, 1947 man's first supersonic flight was achieved by Captain Charles Yeager at an altitude of 42,000 feet. The handling characteristics of the airplane in the high subsonic range were relatively conventional and certainly devoid of those characteristics normally associated with compressibility effects. Stability and controllability remained positive throughout with very little buffeting and without any abrupt changes in trim.


The behavior of conventional airplanes at Mach numbers ahove 0.8 is still largely unknown. The information heretofore received from actual flight test has been generally based upon the reports of badly frightened pilots who have encountered during dives from high altitudes alarming loss of control of their machines. In a large portion of such cases, the airplane has been completely out of control. It has derived its power principally from gravity which likewise cannot be controlled by the pilot. Deprived of any means of effecting the reduction in speed which would give him safety and confronted with the ground rapidly rising up to meet him, the pilot has not been in a position to do objective reporting of the phenomena attendant with high speed flight. He has likewise been reluctant to repeat his experience and hence investigate the various manifestations of such phenomena as are associated with flight at critical Mach numbers.

There has been adequate evidence from flight test to show that the behavior of an airplane at a given Mach number does depend somewhat upon "q" and that its behavior at high altitude is not the same as at low altitude for a given Mach number. The aerodynamic forces which can act upon the airplane are greatly reduced at high altitude, presenting reduced structural problems, and tending to reduce the magnitude of buffet stress. The reduced drag encountered at high altitudes likewise means greater accelerating force to minimize the time during which the pilot need be subjected to a given critical Mach number while increasing his speed.

It was with many of the foregoing factors in mind that the decision was made to power the XS-1 with rocket motors and to provide a sturdy pressurized cabin whereby the pilot could operate with equal safety and equal engine power irrespective of atmospheric density. He has been given a machine that is so sturdy that it is probable that the pilot would break before the machine. He has been given an aircraft so powerful that it is capable literally of climbing and accelerating vertically. It is capable of surpassing during steep climb those Mach numbers hitherto attainable at high altitudes only by diving. With the ability to shut off all power instantaneously and with the ability to conduct at least the early stages of supersonic flight during a climbing attitude, the pilot's ability to retreat from danger at any instant should go a long ways toward promoting the objectivity of his tests. The XS-1 has proven conclusively the feasibility of supersonic flight with aircraft of conventional arrangement.



The purpose of this portion of the presentation is to present the operational procedure and flight characteristics of the XS-1 airplane during the accelerated test program which has recently been conducted at Muroc Lake. The activities of the Flight Test Division personnel in connection with this program will be the primary part of this report. For obvious reasons I shall make it as brief as possible and shall depend upon your questions during the ensuing question period to clarify any points which I do not cover in detail.

In a conference between personnel of the Bell Aircraft Corporation, the NACA, and AMC on 30 June 1947, it was decided that an expedited test program on the XS-1 would be conducted at Muroc Lake. The primary function of the Flight Test Division was to be the operation of one of the XS-1 airplanes and the B-29 mother airplane. In addition such maintenance and technical assistance would be made available as required. The primary aim of this program was to be the investigation of transonic flight phenomena as experienced in the XS-1 up to a Mach number of 1.1. It was decided to proceed through this range of Mach numbers on consecutive flights as rapidly as common sense and technical considerations would permit. The flight program was initiated on 6 August with a series of glide flights, and some indication of the success of the program may be had from considering the fact that the original goal was reached on the ninth powered flight on 14 October. Immediately following this flight a recapitulation of the progress that had been made to that date was made in a conference at Wright Field and an additional flight test program was determined. To this date a total of 16 powered flights have been completed with results that have been detailed in other portions of this presentation.

Since the primary purpose of this paper is to give you some of the details of the operation of the airplane involved, I shall make an attempt to cover at some length one of the powered flights that have been conducted and then briefly sketch the pertinent points of other flights. Let us consider a normal flight and the details of operation. On the day before the flight the XS-1 pressure system is loaded with nitrogen gas and a pressure check is made up to normal operating pressure to check for any leaks that might have developed since the system was last pressurized. The XS-1 is then loaded into the mother airplane the evening before the flight to provide as little delay as possible in getting an early start the following morning. Early the next morning the B-29 is towed to the liquid oxygen and nitrogen storage tanks where immediate action is taken to bring the nitrogen pressure up to operating conditions. While this is being accomplished the lox (liquid oxygen) and fuel tanks are serviced so that a minimum of delay is incurred after the nitrogen pressure has been raised to the needed 4800 pounds per square inch. Just prior to disconnecting the pressure lines the B-29 propellers are pulled through and pre-flight inspection is completed. A last minute briefing with the chase pilot is accomplished to insure that all personnel concerned understand throughly the nature of the day's flight and the responsibility of each person. As the B-29 is towed into a position where the engines may be started, the nitrogen bleed for the XS-1 rocket motor is connected to a bottled source in the B-29 to conserve source pressure in the main nitrogen tanks in the XS-1. This is made even more critical due to the unavoidable loss in source pressure from cooling as altitude is gained, and from absorption of nitrogen by the liquid oxygen. A minimum of delay is necessary in all operations after this point to insure sufficient pressure for dropping at as high an altitude as is practicable. The take-off of the mother ship is normal and immediate climb to altitude is started After an altitude of 7000 feet is attained, I proceed to the cockpit of the XS-1 by way of the ladder from the B-29 bomb bay. In this I am assisted by the project engineer who hands down the door of the XS-I which I fasten from the inside of the airplane. As the B-29 continues to climb, the cockpit of the XS-1 is pressurized and I proceed to connect up the radio, oxygen mask, and othernecessary harness to secure myself in the cockpit. Three minutes before drop time I am in contact with the crew of the B-29 and I notified of the proximity of the drop. It is now time to load the first stage regulator, and in turn the fuel tank and lox tank are pressurized. These tanks maintain a pressure of approximately 330 pounds to propel the fuel and lox to the motor. However, due to lag in the pressure lines, this pressure drops to about 300 pounds when all four chambers are burning and the rate of fuel flow is high. Negative accelerations on the airplane also produce some variation in this pressure. Immediately before the drop a final cockpit check is made, instrumentation switches are turned on and final preparations for the drop are completed. The crew in the B-29 disconnects the nitrogen source from the B-29 and the XS-1 is ready for the drop. An indicated air speed of 250 mph, or slightly more, is desirable to maintain proper control of the XS-1 and to allow for a minimum loss of altitude in the glide before ignition of the rocket engine. Upon a counted signal transmitted from the B-29, the XS-1 is dropped from the mother airplane and an acceleration of approximately .6 g is experienced as the glide is stabilized. After dropping approximately 800 feet all chambers are ignited in rapid sequence and a climb is started which stabilizes at an angle of from 30 to 40. As the climb is started, the horizontal stabilizer is set at 2 leading edge up, and a Mach number of approximately .85 is usually maintained throughout the climb.

Approximately 10,000 feet below the test altitude two of the chambers are turned off to reduce the angle of climb and to establish level flight at approximately .85 Mach number before continuing the tests. The procedure as outlined up to this point is approximately the same for all flights with only those variations necessitated by operational difficulties. For instance, on one flight the release solenoid for the bomb shackle which holds the XS-1 in the mother airplane failed to operate. By the time the airplane could be released mechanically the air speed had fallen below the desired launching speed and the XS-1 was released in a partial stall. Although this was not considered too slow for safety of release, the XS-1 lost approximately 5000 feet before a powered climb could be initiated. On another occasion the reduction of source pressure in the XS-1 was so rapid that the minimum safe release pressure of 3700 pounds was encountered and a release at 20,000 feet was necessary. It has been pointed out elsewhere that the lower altitude of release penalizes the performance of the airplane under its own power since more fuel must be used to climb from the lower altitudes.

I shall now attempt to give you my impression of the flying characteristics of the XS-1. All comments hereafter will be related to Mach numbers as that is the primary speed indication for this test program. The Mach numbers mentioned will be true Mach numbers as corrected and reported by the NACA to eliminate the possibility of confusion which might be induced by a consideration of the correction due to static source error through the range of Mach numbers encountered in this flight test program. During the first portion of the flight test program the progress was quite slow and a number of flights were used to become familiar with the airplane and its operation. The Mach number was increased in fairly small increments to insure maximum safety in the test program, and as more familiarity with the airplane was attained the program was accelerated.

Rapid acceleration is realized as each of the four power units are ignited, and as climb is initiated an effective shift in CG to the rear seems to require nose down trim. The first powered flight runs were made at the 30,000 to 35,000 foot level, and an effort was made to obtain buffet boundaries and stall information. This was necessary to correlate model test data, and the informationobtained on each flight was used in planning succeeding flights. From a Mach number of .8 to approximately .87 very little change in trim was noticeable from the cockpit. An analysis of the data showed that this was partially because of the light stick forces involved. An elevator displacement of only 1 degree to 2 degrees was necessary to compensate for pitch change in the nose down direction. At a Mach number of .87 a light buffeting was encountered which in turn seemed to induce a very moderate lateral instability. The right wing became noticeably heavy and an aileron displacement of approximately 3 degrees was necessary to maintain level flight.

This right wing heaviness appears to be noteworthy since it continues to exist through the transonic range to a Mach number of approximately 1.3 with an increase in aileron forces as the Mach number increases. It is believed that this is possibly caused by the scuffed surface of the right wing.

As the Mach number was increased from .87 the buffeting became more severe and a nose down trim change was noted. The forces were quite light and the movement of the control column remained the best means of indicating the trim change. At approximately .90 Mach number trim change previously mentioned reversed and the tendency was for the nose to rise and in the range of approximately .92 Mach number the buffeting became quite severe.

At this point in the program it was decided from a correlation of model test data that the one degree per second actuator for the stabilizer might prove to be too slow for proper control during subsequent flights and an interruption in the program was made to install a faster motor. In the first flight after the new stabilizer actuator was installed the Mach number was increased to .94. At this point the trim change again reversed to a nose down tendency but it was still easily controllable and approximately 3 of up elevator provided level flight. From .94 to .96 the elevators and rudder became increasingly ineffective until at the latter figure they could be moved throughout their range of displacement with very slight response from the aircraft. At approximately .95 the buffeting decreased rapidly and became nonexistent at .96.

Up to this time a stabilizer setting of 2 leading edge up was used in all of the high speed test runs. The next flight was therefore initiated to investigate the effectiveness of control by the stabilizer at the higher speeds above .96 since the setting had only been varied in climbs up to this time. As the speed was increased on this flight the stabilizer was changed to 1_ leading edge up and returned to 2 leading edge up successively at .84, .88 and .95 Mach numbers. The acceleration experienced in the cockpit was approximately the same for all speeds and it was decided that the stabilizer was still effective even though the elevator and rudder had lost their effectiveness. The ailerons remained effective throughout the range. With the stabilizer setting of 2 the speed was allowed to increase to approximately .98 to .99 Mach number where elevator and rudder effectiveness were regained and the airplane seemed to smooth out to normal flying characteristics. This development lent added confidence and the airplane was allowed to continue to accelerate until an indication of 1.02 on the cockpit Mach meter was obtained. At this indication the meter momentarily stopped and then jumped to 1.06 and this hesitation was assumed to be caused by the effect of shock waves on the static source. At this time the power units were cut and the airphne allowed to decelerate back to the subsonic flight condition. When decelerating through approximately .98 Mach number a single sharp impluse was experienced which can best be described by comparing it to a sharp turbulence bump. This point is mentioned since it has occurred on all subsequent flights and an explanation of its cause does not seem to be completely obvious.

As was mentioned previously, the program was interrupted at this point and a decision was made in conference with Bell Aircraft and AMC personnel to continue the program to attempt to attain a full-power stabilized high speed run. Perhaps it should be mentioned at this time that the high speed attained thus far was accomplished on 75% power in an unstabilized flight condition.

After the resumption of the flight test program the speed was gradually increased in relatively larger increments to a corrected Mach number of 1.35. The buffeting and instability as mentioned heretofore was encountered at the same Mach number each time and as might be expected appeared much less serious as altitude was increased. The maximum Mach number of 1.35 was attained in a flight on 6 November 1947 at an altitude of 48,000 feet. This corresponds to a true air speed of 910 mph. Since I was unable to ignite the No. 3 chamber on this flight due to malfunction of the igniter, this speed was attained in a shallow dive from approximately 52,000 feet.

It is probably desirable at this time to mention the extent of the maneuvering which has been accomplished at higher Mach numbers. At Mach numbers in excess of 1, gentle turns and climbs have been made without difficulty, and except for the expected higher control forces the flight characteristics seemed to be quite normal. Stalls have been performed at various altitudes at subsonic Mach numbers to investigate the variation of maximum lift coefficient with Mach number.

At the conclusion of each test the small amount of fuel remaining in the airplane is jettisoned and a relatively low speed glide to lower altiudes is started. On some flights additional accelerated stall information has been obtained in the glide since a definite weight fix is attainable when all the fuel has been jettisoned. The practice of jettisoning the last portions ot the fuel has been used to prevent possible malfunction of the rocket motor by permitting the fuel to become completely exhausted while the motor is running.

An unaccelerated stall with gear and flaps down is usually performed at a lower altitude for a landing reference and to insure that all fuel and lox has been jettisoned. The stalling speed in this condition is approximately 135 mph and touch down is usually made at approximately 140 mph. Ten to twelve thousand feet of roll is necessary to stop the airplane without brakes.

With reference to the planning of succeeding test flights, this is usually accomplished in conference with NACA and Flight Test Division personnel on the day preceding the proposed flight. The NACA personnel express their desires as to what data would best augment their technical report, and offer advice as to what might be expected in the range to be explored next. The final decision made during the course as to how much advancement in the program is to be made on any particular flight is left with the Flight Test Division crew. As has been indicated indicated previously some changes are necessarily made during the course of the flight.

It is expected that two or three more flights will be required to complete the current phase of the test program. Since the last flight on 6 November, some delay has been experienced due to maintenance difficulties on the mother aircraft and the installation of a new windshield on the XS-1. In view of speeds already obtained, it is considered quite likely that a Mach number of 1.6 can be attained during the completion of this phase with 100% power and possibly a slight dive.




Walter C. Williams

NACA Aeronautical Engineer

Since the acceptance tests of the XS-1 airplanes were completed, two airplanes have been put into operation for research purposes. The No. 2 airplane, which has a 10 percent thick wing and 8 percent thick horizontal tail, has been instrumented in accordance with the original plans for a systematic detailed investigation of stability and control and aerodynamic loads in the transonic speed range to be conducted by NACA personnel using NACA instrumentation. The Air Forces are aiding in this program by flying and maintaining the B-29 mother ship. A detailed description of the instrumentation and test program with the XS-1-2 airplane was given in a paper presented at the Institute of Aeronautical Sciences annual meeting in New York in January of 1947. The NACA program is just getting underway, having been delayed by numerous mechanical difficulties. Two powered flights have been made and some data have been obtained up to a Mach number of 0.85.

The No. 1 airplane which has an 8 percent thick wing and 6 percent thick tail is being flown by the Air Forces Air Materiel Command Flight Test Division for an accelerated transonic flight research program. The purpose in this series of flight tests is not to make detailed investigations but rather to fly at speeds in excess of the speed of sound in as short a test program as possible, making in each flight as large an increase in Mach number as compatible with safety. The flying of the XS-1 is being done by Capt. Charles E. Yeager of the Flight Test Division and the data reduction and analysis are made from NACA instrumentation by personnel of the NACA Muroc Flight Test Unit. The Air Forces program has been underway since August and 16 flights have been made. Flight up to and beyond a Mach number of 1.0 has been obtained on 6 flights at altitudes from 37,000 feet to 49,000 feet. These speeds were obtained either in level flight, shallow dives, or shallow climbs. The maximum Mach number reached was 1.35 at 49,000 feet using three of the four rocket cylinders.

The present paper presents for the most part results or the XS-l-1 airplane obtained during the Air Forces accelerated transonic research program, although some information for XS-1-2 airplane from the NACA research and Bell acceptance tests has been included. Some of the data, up to a Mach number of 0.92, were presented previously at the Ames Laboratory transonic airplane design conference in November of 1947.


No detailed description of the airplanes will be given as a preceding paper has covered this subject thoroughly. It may be well, however, to recall a few of the characteristics of the XS-1 airplanes that influence the interpretation of the results. The XS-1 airplanes are very small. The wing area is only 130 square feet. The airplanes are very dense; the wing loading varies from 92.5 to 52 pounds per square foot during a flight. The surfaces are thinner than those of airplanes for which comparable data have been obtained in the past. The wing sections of the No. 1 and No. 2 airplanes are 65,-108 and 65,-110, respectively. The horizontal tail sections are symmetrical 65 series sections of 6 percent and 8 percent thickness, respectively. There is essentially no aerodynamic balance on any of the control surfaces. The exposed mass balances used on control surfaces of the airplanes give an aerodynamic balancing action which represents only a small part of the unbalanced aerodynamic hinge moment. The friction in the control system, particularly the ailerons and rudder, is quite high as illustrated in Fig. 1 which shows the force required to move the controls of the No. 2 airplane on the ground as a function of control surface deflection.


The instrumentation in the two XS-1 airplanes is considerably different. In the XS-1 assigned to NACA, complete instrumentation to describe the motion of the airplane and controls is installed. The measurements include airspeed, normal, longitudinal and transverse acceleration, angular velocity in roll or pitch, sideslip angle, aileron, elevator, rudder and stabilizer position and aileron, elevator, and rudder forces. These data are all internally recorded on standard NACA instruments. In addition, the measurement of some of the more essential quantities, airspeed, altitude, normal acceleration, elevator, aileron and rudder position, is duplicated using a 6-channel telemeter.

The instrumentation in the Air Forces XS-1 is not as complete with regard to measurements of airplane characteristics but somewhat more attention has been paid to the measurement of airspeed since it was realized that there would be considerable importance in accurately determining the Mach numbers obtained. Airspeed is measured by means of telemetering and with the NACA standard recorders. One of the airspeed recorders is of large range (150 inches of water) to cover the entire speed range and altitude range, whereas the other recorder has a low range (50 inches of water) for more precise measurement of the low values of q corresponding to the Mach numbers which were expected to be reached at extreme altitude (60,000 feet). Altitude is also measured by telemetering and standard NACA recorders. Other quantities that are telemetered include elevator, stabilizer and aileron position, elevator stick force, and normal acceleration. As the tests have progressed, internal recording instruments have been added to measure elevator position and normal, longitudinal and transverse acceleration.

In addition to the instrumentation described, an SCR584 radar set is used during tests of both airplanes to obtain data for use in airspeed calibration. All measurements, recorded, telemetered and radar are synchronized through use of common timing circuits.


In the calibration of the airspeed system, the assumption is made that the sole error at Mach numbers below 1 is in the measurement of the static pressure. This error results from the influence of the airplane pressure field on the measured pressure. At Mach numbers above 1 there is an additional error resulting from the energy loss in the bow shock wave.

The error in static pressure resulting from the airplane pressure field was determined using the radar pressure-altitude method. This method is described fully in NACA Wartime Report No. L-43 by Zalovcik and Wood entitled "Static-Pressure Error of an Airspeed Installation on an Airplane in High Speed Dives and Pull-Outs". The method consists essentially of measuring simultaneously the geometric altitude of the airplane with radar and the pressure altitude as measured by the pilot-static tube mounted on the XS-1 one chord length ahead of the left wing tip and recorded in the airplane. The geometric altitude measured by radar is converted to pressure altitude by means of a pressure-altitude survey made just prior or subsequent to the test run. The pressure altitudes as measured by radar and the pilot-static head are then compared and the difference between these quantities is the error in static pressure.

A typical pressure survey is shown in the lower part of Fig. 2, where the pressure in inches of mercury is plotted as a function of geometric height. These data are obtained by measuring simultaneously the geometric altitude using radar and the pressure altitude using the recording instruments in the test while the test airplane is flying at a speed for which the static pressure error has been previously determined. The curve without tests points has been determined from calculations using the standard lapse rate and is shown for comparison only. The survey brackets the expected altitude range for the test run.

The basic determination of the static-pressure correction was made by flying the test airplane, at relatively low speeds, by the B-29 mother ship from which a trailing static bomb was suspended. A large enough range of lift coefficients was covered by making pull-ups so that the effect of lift coefficient could be determined. The upper part of Fig. 2 gives the static-pressure error obtained by this procedure expressed as_(symbol) as a function of lift coefficient. The data of the figure show that the lift coefficient has no effect on the error of static pressure as measured on the XS-1.

Fig. 3 shows a time history of the static pressures as measured by the pilot-static head on the test airplane and as derived from the radar altitude using the pressure survey. The data are for a typical high speed run through a Mach number of 1. The difference between the curves represents the error in static pressure.

Fig. 4 gives the results of the high speed calibration. On this figure the static pressure error is given in terms of _p/q' and is plotted as a function of Mach number M'. q' M' are the dynamic pressure and Mach number uncorrected for loss in total head. It is believed that the increase in positive static pressure error above a Mach number of 0.8 is caused by the increase in positive pressure behind a local shock wave which formed on the airspeed head itself. This shock moves back as the Mach number is increased at a Mach number just above 1.0 the shock passes over the static holes, and there is a jump in the static pressure curve. As the Mach number is increased further, the static pressure is affected by the bow wave at the nose of the head. As this shock attaches and bends back, the static pressure as measured by the pilot-static head increases in the positive direction. These data are in general agreement with results obtained in tests made of a static pressure tube mounted in the high speed flow over the wing of an airplane and reported in Wartime Report No. L-90 by Zalovcik and Daum entitled "Flight Investigation at High Mach Numbers of Several Methods of Measuring Static Pressure on an Airplane Wing."

Fig. 5 gives the results of the airspeed calibration in terms of the error in Mach number. In this figure M' ' (Mach number uncorrected for static pressure error and loss in total head) is plotted as a function of M' and M. This figure shows the effect of total head loss. No calibration was made for loss in total head but the theoretical correction for loss in total head was applied. Measurements which are reported in the previously mentioned paper by Zalovcik and Daum showed that actual loss in total head is in agreement with that given by theory. As can be seen in this figure, the maximum Mach number reached to date is approximately 1.35. It is believed that calibration for Mach numbers above 1.0 is correct within .04 in Mach number.


The Air Forces test program has for the most part consisted of straight speed runs. The data obtained, therefore, are mostly those which pertain to trim in straight flight, and are presented separately for each flight since there were changes in altitude and small changes in stabilizer position and lift coefficient between flights which would affect those data which apply to longitudinal trim.

Fig. 6 shows the variation with Mach number of elevator position, elevator force for two stabilizer setting (+2.2 and +1) as measured in flight at 30,000 feet for a Mach number range of 0.70 to 0.94. The data given in this figure and subsequent figures are for essentially constant lift coefficient. With the stabilizer set at an incidence angle of 1.0, the pilot did not fly beyond a Mach number of 0.876 because it was difficult to hold steady flight due to the wheel forces required for trim, the relatively far forward position of the stick, and the buffeting expected at higher Mach numbers. Data were obtained for a stabilizer incidence of 2.2 up to a Mach number of 0.94 and this stabilizer setting was used for most of the subsequent flights to higher speed. This flight was made with two stabilizer settings to obtain some measure of the elevator effectiveness and to aid in determining a stabilizer position that would not lead to running out of elevator control in either the upward or downward direction. It should be noted that the variation of elevator effectiveness with Mach number, which is indicated by the increasing distance between the elevator position curves for the two stabilizer settings with increasing Mach number, is such that the elevator effectiveness is reduced by more than 50 percent between a Mach number of 0.70 and 0.87. This reduction in elevator effectiveness will affect the magnitude of the trim changes but should not affect the trends of the trim changes.

In approaching the flight up to and beyond Mach number of 1, previous knowledge obtainedfrom tests made in the NACA 8-foot high speed tunnel and from NACA wing-flow tests of a model of the XS-1 airplane was used as a guide in formulating the flight test program. The 8-foot tunnel model was an exact model of the XS-1-2 airplane and had the 10 percent thick wing and 8 percent thick horizontal tail. The wing-flow model was a model of the XS-1 airplane also, but the fuselage was smaller and shorter than that for the actual airplane. These data are discussed at this time because the Mach number range covered by the model tests was approximately the same as that covered in the last figure for the airplane tests. The next figure (Fig. 7) compares the variation of elevator position required for trim with Mach number as measured in the model tests and in the flight tests. This figure shows that the same general trends were predicted by the model tests as were experienced in flight. It may be that some of the differences can be accounted for by the differences in wing and horizontal tail thicknesses of the models and airplane.

In Fig. 8 the variation of elevator position and force with Mach number is shown for a test run made at 37,000 feet pressure altitude. The maximum Mach number reached was approximately 1.00. It should be noted that additional trim changes not previously expected occurred after the trim changes predicted in the Mach number range from 0.8 to 0.94. Also the nose up trim change which occurred around Mach number 1.0 is much larger than those experienced at lower Mach numbers.

The variation of elevator position with Mach number at 43,000 feet up to a Mach number of 1.055 is shown in Fig. 9. The curves on this figure are discontinuous because data were arbitrarily selected for two different constant values of lift coefficient. These data show the same general trend as the previous figure with a large trim change occurring just past Mach number of 1.0.

Fig. 10 gives the variation of elevator position and force with Mach number as obtained in tests made at a pressure altitude of 49,000 feet up to a Mach number of approximately 1.25. It should be noted that above a Mach number of 1.0 there is a continuing trim change in the nose down direction. The maximum elevator force experienced in flying the XS-1 in the transonic speed zone is shown in the figure and occurs just past a Mach number of 1.0. The force measured was 25.5 pounds which is well within the capabilities of the pilot. It should be remembered, however, that these data were obtained at 49,000 feet altitude. The value of q for a Mach number of 1.03 at 49,000 feet is approximately 180 pounds per square foot. If this speed were reached at sea level, the value of q would be 1570 pounds per square foot and the elevator force would be of the order of 223 pounds or almost 9 times the force at altitude. It should also be noted that the XS-1 has a very small elevator. The elevator chord is 20 percent of the horizontal tail chord and the root mean square chord of the elevator is only 5.6 inches. The average fighter probably has an elevator chord double that of the XS-1 which would make the forces four times heavier.

The next figure (Fig. 11) shows the effects of altitude and stabilizer position on the longitudinal trim characteristics. In this figure the variation of elevator position with Mach number from the previous figures is shown, with the altitudes and stabilizer positions noted. Although the changes in stabilizer position are small, it should be appreciated that the relative effectiveness of the elevator is low above a Mach number of 0.8 and that small changes in stabilizer position may make appreciable difference in the elevator angles for trim. The data in this figure show that although the variation of elevator angle with Mach number is somewhat different for each condition shown, the general trends are the same.

The next figure (Fig. 12) gives data obtained in flight at 49,000 feet and shows the variation of elevator position and force with Mach number up to 1.35. The pilot changed the stabilizer setting several times during the run. It can be seen in this figure that the changes in elevator angle for trim as well as the forces are reduced over those required when a constant stabilizer setting was used.

There has been no detailed investigation of the longitudinal stability and control characteristics of the XS-1 at high Mach numbers but there are indications that the apparent airplane stability, as measured by stick force per G and rate of change of elevator angle with change in lift coefficient, is greatly increased as the Mach number is increased. Large amounts of elevator are required at Mach numbers beyond 1.0 to produce relatively small changes in acceleration or lift coeffficient. A time history of a gradual pull-up made at a Mach number of 1.34 is shown in Fig. 13. In order to show the effects of Mach number on the stability parameters, Fig. 14 was prepared which gives data obtained in tests made in the Bell and NACA tests as well as for the Air Force program. Data are shown here for both XS-1 airplanes (thick and thin wings) and for accelerated turns and gradual pull-ups. Although the data are not directly comparable, they are all that are available at present and do show the trend of the parameter. The longitudinal stability of the XS-1 is low but positive at subsonic Mach numbers, and is lowest at about a Mach number of 0.675, as predicted from tests made in the the NACA 7- by 10-foot wind tunnel and from calculations made by personnel of the Bell Company. The stability apparently increases rapidly for the Mach numbers above 1.0, although the data are insuffficient to establish the actual quantitative variation.

The buffet boundary and limit lift as defined for the XS-1 with the 8 percent thick wing are shown in Fig. 15. These data were obtained in level flight or in gradual turns with the stabilizer set on incidence angle of 2.2. Limit lift has been determined from measurements where lift ceased to increase although increasing up-elevator deflection was being applied. The limit lift has not been determined for Mach numbers above 0.90 but the highest lift coeffficient reached above Mach number of 1.0 was 0.48 at a Mach number of 1.06. The highest lift coeffficient reached at the maximum Mach number of 1.35 was 0.2. Although buffeting has been experienced in level flight for all flight beyond the buffet boundary, it has not been disconcerting to the pilot because the buffeting is not severe.

The directional stability of the XS-1 appears to be very high at all speeds although there are no recorded data available at present to define the directional stability at the higher flight speeds. Fig. 16 gives data obtained in a steadily increasing sideslip made at a Mach number of 0.78 during the NACA tests of the XS-1-2 airplane. These and similar data at lower speeds obtained during the Bell tests indicated the high directional stability and the positive dihedral effect up to a Mach number of.78. Values of the directional stability parameter Cn__computed from sideslip data obtained in flight compare very closely with the value of 0.0035 obtained from the tests conducted in the NACA 7- by 10-foot wind tunnel. The degree of dihedral effect also corresponds closely to that obtained in the tunnel tests.

Above a Mach number of 0.8 the pilot has reported no decrease in directional stability. The pilot has rather reported that in the Mach number range from 0.9 to 1.0 deflection of the rudder has caused no change in sideslip angle which is indicative of a loss in rudder effectiveness and/or an increase in directional stability. He also reported that the dihedral effect throughout the speed range covered is positive.

From the tunnel tests and calculations made by the 7- by 10-foot tunnel staff and Bell Aircraft personnel, predictions of the stability boundaries indicated that with the high directional stability the XS-1 would be in the region of spiral divergence and well removed from the boundary of oscillatory instability. There have been experienced, during the XS-1 tests, however, poorly damped lateral oscillations. These oscillations have occured with both of the XS-1 airplanes and are most noticeable to the pilot in the range from 0.7 to 0.9 Mach number. They may not be transonic phenomena but are discussed since other airplanes have experienced similar difficulties. Time histories of an oscillation which was measured during the NACA tests of the XS-1-2 airplane are shown in Fig. 17, which gives the variation with time of rolling velocity, sideslip angle, transverse, longitudinal and normal acceleration, and the control positions. It should be noted that these oscillations are control-fixed phenomena. Several possible reasons for the damping of these lateral oscillations have been investigated but so far no real explanation is available. It was thought that these oscillations were possibly caused by fuel sloshing since the fuel weight is quite high in comparison to the size of the airplane. A series of rudder kicks to induce lateral oscillations were made, therefore, with varying amounts of fuel on board. The airplane relative density, lift coefficient and Mach number were held essentially constant by making the tests at increasing altitudes as the fuel was consumed. These tests showed that the fuel had little effect on the damping of the short period oscillations.

Tests were also made in the 8-foot high speed tunnel at Langley in which the XS-1 model was yawed to check for the possibilities of low directional stability near zero sideslip and of separation of flow from rear of the fuselage at high speeds. These tests, however, failed to show either of these characteristics which could cause the poor damping.

Another possible reason for the poor damping, which also appears the most likely, is the effect of the angle of the principal inertia axis on the stability. This matter is discussed fully in NACA Technical Note No. 1193 entitled "Effect of Product of Inertia on Lateral Stability," by Leonard Sternfield where it is shown that the stability may be critically dependent on the angle of attack of the principal inertia axis, a mass characteristic, and is decreased when the angle is decreased.

The pilot has reported no loss in aileron effectiveness up to the highest Mach number tested. It may be, however, that there is a reduction in the aileron effectiveness parameter, pb/2V, but the true speed is so high that the rolling velocities are still high. The pilot therefore would not notice any decrease in aileron effectiveness.

Although diffficulties have been experienced in recent tests of other airplanes at transonic speeds with one dimensional flutter or buzz, one probable contributing factor to the absence of this oscillation is the large amount of friction in the control system. The friction in the ailerons is of the order of 20 foot pounds measured at the aileron. Although the excitation forces may be present, they are of small magnitude because of the small size of the control surface and the low dynamic pressure. The aerodynamic hinge-moment coefficient for the dynamic pressure corresponding to a Mach number of 0.85 at 30,000 feet and neglecting the effects of Mach number on the hinge-moment coefficient, is of the order of 6.9 foot pounds per degree. Hydraulic dampers are installed but have not been used. The airfoil section as mentioned before is a 65,-108 which is a thin surface with low camber. The ailerons themselves are flat-sided.

Fig. 18 shows the variation of the right aileron angle required for trim as a function of Mach number and indicates the effect of Mach number on lateral trim. The abrupt change in trim shown at a Mach number of 0.98 may be faired out when more data are obtained. The pilot has not reported a discontinuity in lateral trim as shown by the figure.


The data obtained in flight up to and beyond the speed of sound with the XS-1 airplane shows that most of the difficulties expected in the transonic range have been experienced. In most cases, however, the difficulties have not been as serious as expected and, although conditions are not normal, the airplane can be flown under control. In detail, the following has been noted:

1. The airplane has experienced longitudinal trim change from a Mach number of 0.8 up to the highest number reached. The largest control force associated with these trim-changes has been 25.5 pounds and the pilot has been able to control the airplane. At lower altitudes or with a larger similar airplane, the control forces may be unreasonably large.

2. The elevator effectiveness has decreased more than 50 percent in going from a Mach number of 0.7 to 0.87. There is evidence of further reduction in elevator effectiveness above a Mach number of 0.87.

3. The apparent longitudinal stability in accelerated flight is greatly increased above a Mach number of 0.75. This increase in apparent stability is probably a combination of increase in the actual stability of the airplane and decrease in elevator effectiveness.

4. Buffeting has been experienced in level flight but has been very mild. The limit lift coefficient is decreased above a Mach number of 0.80. A lift coefficient of 0.48 was reached at a Mach number of 1.06.

5. Static directional stability of the XS-1 as measured in steady sideslips is high. In spite of this high directional stability, a poorly damped lateral oscillation has been experienced which cannot be readily explained.

6. Lateral control as reported by the pilot is apparently good throughout the speed range tested. There has been no evidence of aileron buzz or associated phenomena. The airplane becomes right wing heavy above a Mach number of 0.8 but can be trimmed with aileron.



Charles L. Hall

Before considering what future plans are being made by the Air Force for using the XS-1 airplanes let us first briefly review the purpose of procuring the original three airplanes. The basic idea behind this project was to obtain a piloted type aircraft capable of flight in the transonic region. As you have just learned this goal has been achieved, even though we have not as yet made any transonic flights for any specific reason except to fly transonically. The question now arises as to what can we do with an airplane capable of this performance.

To go back to the original three airplanes for a moment let us recall the specific assignments agreed upon for these articles. The first airplane was used by Bell at Pinecastle for glide flights early in 1946 for pilot familiarization and to determine preliminary flight characteristics. This article was fitted with 10% thick wing panels and 8% thick horizontal tail surfaces. After completion of demonstration flights by the Bell Company as previously mentioned this airplane was assigned to the National Advisory Committee for Aeronautics for continuance of transonic flight research. Since the data to be obtained by the NACA is in the nature of basic research information which could only be gathered by a very lengthy and detailed flight program, it became apparent that it would be desirable to start an accelerated flight program with the sole purpose of reaching transonic speeds in the shortest possible period consistent with safety to the pilot. This program was accomplished on the second airplane, the results of which have been described by Captain Yeager and Mr. Williams. This airplane was fitted with 8% thick wing panels and 6% thick horizontal tail surfaces. Recalling Mr. Stanley's speech you will recall that the first two airplanes were equipped with the nitrogen pressure fuel system instead of the turbine pump system originally intended. The third XS-1 airplane was to be fitted with the turbine pump system but during the past summer we were forced to temporarily stop further work on this article since additional funds for the completion of the turbine pump project were not available. This airplane will probably be finished with the nitrogen pressure fuel system unless current tests of a hydrogen peroxide turbine pump indicate a safe and practicable application to the XS-1. We are somewhat reluctant to design the plane to accommodate the H202 turbine pump until we know a little more about the handling qualities of H202 under field conditions similar to those that exist at Muroc. If we determine that the transportation, storage, handling, and use of H202 can be accomplished with reasonable safety we would very much like to modify the third XS-1 airplane to accomodate a low pressure fuel system. By so doing we could realize about a 60% greater endurance, the weight of the heavy nitrogen tanks being supplanted by an increased load of propellant. A set of 8% thick wings and 6% thick horizontal tail surfaces are being procured for the No. 3 X-1 because we know the actual characteristics of these thin airfoils in the transonic and supersonic speed ranges. We have yet to learn the actual aerodynamic characteristics of the thicker airfoils that were originally constructed for this plane. When completed the No. 3 XS-1 will serve as a spare for the No. 1 and No. 2 airplanes currently undergoing flight testing at Muroc.

With the understanding that we now have three XS-1 airplanes assigned to two separate flight programs both of which have specific limitations and goals, the question of using this type of airplane for applied research and equipment testing remains to be considered. We are greatly encouraged in our belief that supersonic tactical aircraft can be developed in the future, but we must not let our enthusiasm over our recent success with a purely research airplane throw us into a feeling of complacency. We have barely scratched the surface in our determination to solve the mysteries of the supersonic flight region. To fly a man-carrying airplane at supersonic speeds is indeed a notable achievement. But to extend our achievements to the point where a combat weapon can fly and fight effectively at such speeds is really a tremendous undertaking which must be tackled with vigor and determination right now. Not only is it a fact that we have few tactical aircraft components and systems which we know are suitable for supersonic flight, but it is also a fact that we know precious little about designing suitable ones to replace those we have. What actual effect will ultra high speeds have on temperature changes in the airplane? Will such temperature changes be of sufficient magnitude to force the scrapping of our present systems and components and cause their complete redesign to new specifications? What sort of automatic gun laying and sighting equipment will be required to make our tactical aircraft truly effective in combat? How are we going to drop or launch bombs effectively at these speeds when even now we are greatly concerned about this very thing in the tactical employment of our new subsonic bombers? Will our most modern machine guns and automatic cannons function satisfactorily or will the supersonic air blast on their muzzles cause them to recoil with destructive force? How are we to provide for emergency crew ejection and keep within the tolerance of the human body to withstand acceleration forces and air blasts? And once the crew is safely ejected what kind of parachutes and special survival equipment will be required to lower the crewmen to a safe altitude. Answers to all of these questions and a myriad of others we must have before we can get even a fairly accurate idea of just how to design our supersonic combat airplane and its allied equipment. Added to all of these questions which we must have answered soon there are a vast number of ideas and theories which we wish to explore to improve or refine our present equipment and to increase our factual knowledge of scientific phenomena that occur in high speed flight. The guidance and control of missiles, the effects of high speed acceleration and extreme altitude on the human body, the optimum design for turbo jet and ram jet engines for supersonic flight, radar detection and tracking of relatively small targets travelling at supersonic speed, etc., are examples of the things we would like to take under intensive study immediately. Suitable test facilities which would enable us to undertake these studies and get the answers we so sorely need do not exist at present. Realizing the importance of this type of work the Secretary of the Air Force has recently directed the Air Materiel Command to procure four additional XS-1 airplanes for this purpose. These airplanes will be used by the various Sections and Laboratories of the Engineering Division. It should be clearly understood that the information to be obtained from the use of these airplanes in no way duplicates the basic information to be obtained by the NACA in their long term flight research program. In the fields of aerodynamics, structures, and stability and control, it may appear that our proposed program duplicates that of the NACA. lt should be understood, however, that in these fields the Air Materiel Command program is set up to yield practical design information rather than basic theoretical data.

In general the four additional XS-l airplanes will be copies of the original article. However, since each airplane will be used for specific fields of investigation it may prove desirable to make certain changes in the individual airplanes to improve their usefulness. This part of the program is now under consideration and the extent to which any changes are made will probably be influenced by the availability of funds. It is expectedthat the first of these additional airplanes will be ready for flight in the fall of 1948. To further implement this new program as well as continue the present X-1 program now under way the Air Force is planning to modify a B-50 Boeing bomber to act as an additional mother ship. The use of this airplane is expected to materially improve the performance of the XS-1 airplanes since the higher rate of climb and higher launching altitude will reduce the loss of nitrogen which so readily dissolves in the liquid oxygen and thus reduces the endurance of the airplane. In connection with this problem consideration is also being given to the use of helium instead of nitrogen. Helium does not combine with liquid oxygen to any appreciable degree but its cost, availability, and ease of loading on the airplane are factors which must be considered before any definite decisions can be reached. Means of increasing the range of the airplane by improving the fuel and liquid oxygen tank designs are also under consideration by the Bell Company.

Although detailed plans for the use of these additional airplanes have not been completed, it is contemplated that they will be assigned to projects undertaking studies of the following general nature:

A detailed list of the exact studies to be undertaken follows. This list will be extended to include other studies as their need materializes.

a. Guided Missles

l. Tests of radar target scanning system to be used on guided missile at low supersomc speeds.

2. Performance tests of liquid rocket power plants for missiles.

3. Tests of complete and partial control systems for missiles at supersonic speeds.

4. Radar tracking tests of the XS-1 from a bomber carrying the tracking and guidance equipment for defensive air to air missiles.

5. Tests of auto celestial navigation system stable platforms to determine the magnitude of errors due to Coriolis acceleration.

6. Tests of star followers at extreme altitudes to study the effects of diminishing sky brightness on auto celestial navigators.

7. Tests of large scale supersonic shock diffusers, provided the XS-1 nose can be modified to accomodate them.

8. Studies of antenna attenuation effects caused by the rocket discharge system.

9. Tests of missile ground tracking equipment using the XS-1 as a target operating at supersonic speed and extreme altitude.

10. Tests of missile guidance systems to determine the effect of speed on accuracy of guidance.

b. Aero Medical

1. Ejection seat tests (if possible using radio controlled aircraft).

2. Tests of protective clothing such as pressure suits, air ventilated suits and crash helmets.

3. Measurements of sound range and intensities.

4. Studies of visibility and recognition at high speeds.

5. Studies of human reactions at high altitudes.

6. Environmental data (temperature, humidity, and pressure effects).

c. Armament

1. Muzzle blast and trunnion reactions at high speeds.

2. Tracking control tests.

d. Electronics

1. Tests of following equipment when available: VHF communication equipment; Radar control beacon AN/APN-11, Radio Control equipment AN/APW-40; High speed zero Drag Antennas of various types; Telemetering equipment AN/AKT-5; Mark V I.F.F. interrogator responser AN/APX-6; D/F Equipment AN/CRA-1; Ground Controlled Approach equipments AN/MPN-I, AN/CPN-4, and AN/CPN-18.

e. Aircraft Equipment

1. Recovery parachute for NIKE.

2. Emergency escape equipment for XS-2 airplane.

3. High speed, high altitude Personnel Parachute Canopy.

4. Indicator System, Compensated Gyro Horizon, High Performance.

5. Study of Pitot-Static Design.

6. Skin Temperature Measurements at High Speed.

7. Heat transfer problems.

8. Reduction of Fire Hazards.

9. Fire Detection and Fire Extinguishing problems.

10. Electrical System Operation data to include temperatures of cooling air to D. C. generators, acceleration effects on reverse current relay and voltage regulator, and general battery behavior.

f. Power Plant

1. Flight testing of existing ram jet engines.

2. Flight tests of ram jet acceleration arrangements.

3. Study of converting XS-1 from rocket power to ram jet power.

g. Aircraft Design

1. Structures: Loads, pressures, deflections, control system and canopy loads. Temperature effects.

2. Dynamics: Vibration studies to include basic structure, controls, and engine mounts.

3. Aerodynamics: Determination of dynamic stability derivitives, wing flow model tests, effect of protruberances and a symmetrical wing tip bodies, boundary layer heat transfer, variation of C1 with wing leading edge radius, gust load records, buffet boundary studies.

I have a motion picture which shows a chronological account of the complete XS-1 test program which will portray on the screen many of the things you have heard discussed this afternoon.

The following persons were cleared for attendance at the XS-1 conference held 9 January 1948.




Strategic Air Command



Air Training Command



Tactical Air Command

LEE, R. M.


Tactical Air Command



Tactical Air Command



Air Defense Command



Air Defense Command



Air University



Air University



Air University



Air Proving Ground Cmd. Cmd.



Air Proving Ground Cmd.



Dir R & D, Hq. USAF



R & D Board



Off R & D Hq. USAF



R & D Board



Off R & D Hq. USAF



ADVS Board-Aerodynamics



ADVS Board - Exec Sec.



ADVS Board - Aerodynamics



ADVS Board - Aerodynamics



ADVS Board - Guided Miss.



ADVS Board - Guided Miss.



ADVS Board - Guided Miss



ADVS Board - Guided Miss



ADVS Board - Guided Miss



ADVS Board - Guided Miss



ADVS Board - Fuels & Prop.



ADVS Board - Fuels & Prop.



ADVS Board - Fuels & Prop.




Commanding General AMC



Dep. C.G. AMC



Director of R & D



Dir. Supply & Maint.



Dep. Dir. R & D



Chief, Engr. Div.



Ind. Mobilization Planning



Chief Maint. Div.



Chief, Engr. Oper.



Dep. Chief Engr. Div.



Chief, Elect Sub-Div.



Dir. of Procurement & Ind. ML



USAF Institute of Tech.



Intelligence Dept.



Chief, Flt. Test Div.



Electronics Operations Sect.



Intelligence Dept.



Aircraft Projects Sect.



Electronics Plans Sect.



All Weather Flying Div.



Equipment Lab.



Chief, Armament Lab.



Chief, Aero-Med Lab.



Aircraft Projects Sect.



Engineering Plans



Comp & Systems Lab.



Chief, Power Plant Lab.



Chief, Aircraft Lab.



C & N Lab.



Procurement Div.



Aircraft Projects Sect.



Aircraft Radiation Lab



Aircraft Section



Aircraft Proj. Sect.



(AAF Tech. Committee)



Guided Missiles Sect.



Aircraft Proj. Sect.



Aircraft Proj. Sect.



Flight Test Div.



Aircraft Proj. Sect.



Flt. Test Div.









Propeller Lab.



Off. Dir. R & D



Aircraft Projects Sect.



Material Lab.



USAF, Institute of Tech.



Photo Lab.







Northrop Aircraft Co.



General Electric Corp.



Chance-Vought Co.



Bell Aircraft Co.



Curtiss-Wright Corp.



Pratt & Whitney Co.



Curtiss Propeller Div.



McDonnell Aircraft Corp.



Consolidated-Vultee A/C Corp.



Douglas Aircraft Co.



Bell Aircraft Corp.



Allison Div., GMC



Lockheed Aircraft Co.



Grumman Aircraft Co.



Ryan Aeronautical Co.



Republic Aviation Corp.



Westinghouse Elec. & Mfg. Co.



Wright Aeronautical Corp.



Douglas Aircraft Co.



Reaction Motors, Inc.



Packard Motor Car Co.



Douglas Aircraft Co.



North American Aviation



Boeing Aircraft Co.



Glenn L. Martin Co.



Bell Aircraft Co.



Bell Aircraft Co.
















Navy Dept.



Navy Dept.



Navy Dept.



Navy Dept.



Navy Dept.



Navy Dept.



Navy Dept.