- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
DEVELOPMENT OF AIRBORNE ARMAMENT
1910 - 1961
Prepared under the provisions of Air Force Regulation 210-3 and Air Force Systems Command Supplement No. 1 thereto as part of the United States Air Force Historical Program.
This document contains information affecting the national defense of the United States within the meaning of the Espionage Laws, Title 18, U.S.C., Sections 793 and 794. Its transmission or the revelation of its contents in any manner to an unauthorized person is prohibited by law.
THIS DOCUMENT MAY NOT BE RELEASED TO ANY INDIVIDUAL OR AGENCY OUTSIDE THE DEPARTMENT OF THE AIR FORCE WITHOUT THE PRIOR APPROVAL OF THE ORIGINATING ORGANIZATION OR A HIGHEST AUTHORITY IN THE DIRECT LINE OF COMMAND
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
DEVELOPMENT OF AIRBORNE ARMAMENT
1910 – 1961
Office of Information
Aeronautical Systems Division
Air Force Systems Command
RALPH R. SPRINGER
Lieutenant Colonel, USAF
Director of Information
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
TABLE OF CONTENTS
I Early Optical Bombsight Development
Chapter II Bombing with Radar
Chapter III Bombing-Navigation Systems: The K-Series
Chapter IV The "Bomb Director for High Speed Aircraft"
NOTES FOR CHAPTERS I - IV
V Supersonic Bombing in the B-58.
Chapter VI The B-70 Bombing and Navigation System
NOTES FOR CHAPTERS V - VI
VII Early Bomber Defense
Chapter VIII Complete Defensive Systems
Chapter IX New Concepts for Tail Defense
Chapter X Tail Defense for the B-52
NOTES FOR CHAPTERS VII - X
XI Tail Defense for the B-58
Chapter XII Defensive Subsystem for the B-70
Chapter XIII The Quest for New Defensive Techniques
NOTES FOR CHAPTERS XI - XIII
XIV The A-Series Computing Optical Sights
Chapter XV The K-19 Computing Optical Sight
Chapter XVI Computing Sights for Atomic Weapons Delivery
Chapter XVII The MA-7 Fire Control System for the F-101
Chapter XVIII The MA-8 Fire Control System for the F-105
Chapter XIX The MA-10 Fire Control System for the F-104
Chapter XX Advanced Systems
NOTES FOR CHAPTERS XIV - XX
XXI The Early E-Series Fire Control Systems
Chapter XXII "Universal Computer" Interceptor Systems
NOTES FOR CHAPTERS XXI - XXII
XXIII The F-106 Fire Control System
Chapter XXIV The Controversial Long Range Interceptor
NOTES FOR CHAPTERS XXIII - XXIV
Index for Volumes I, II, III, IV (omitted due to HTML and Google allowing rapid search)
LIST OF ILLUSTRATIONS
Mark 1-A Bombsight
Evolution of Bombing Stabilization Systems
The Norden Bombsight
Initial Radar Bombing Systems.
Experimental Housing for Eagle Antenna on a B-24
The NOSMO System
AN/APQ-24 Bombing-Navigational System
The K-1 Modification Program
The K-1 System in the B-36
An MA-2 Radar Scope Photo of Long Island
B-58 Bombing and Navigation Equipment
The Advantages of Sidelooking Radar
Applications for Sidelooking Radar
Location of B-70 Offensive Subsystems
In January 1957 the historical office at Wright Air Development Center turned to the ambitious task of recording the evolution of airborne armament. The undertaking was not quite as awesome as one might suppose, since by common agreement, lethal weapons such as guns, bombs, and rockets were largely excluded, as were most of the details of the development of electronic countermeasures. Nonetheless, well within the subject's boundaries were all airborne fire control systems—defensive and offensive—and all bombing systems, which possessed in common the need for directive intelligence. In unsophisticated equipment it might be no more than a mechanical or optical coupling dependent upon human control. The other extreme was a completely automatic fire control device which, at least in theory, could operate with perfect efficiency even if the crew abandoned the aircraft.
Early assessment indicated that the assignment was beyond the capacity of one man unless he were willing to accept the prospect of two years of specialized research and writing. Consequently, in order to complete the work within a reasonable period, two historians prepared semi-autonomous portions. Dr. Julius King, Jr., did the research and writing for a section on bombing systems and another on fighter fire control; Mr. Gary P. Baden worked in the areas of interceptor fire control and bomber defense. In their first incarnations these sections saw light as "capsule monographs" published with semiannual histories of "Wright Air Development Center for July-December 1956 and January-June 1957.
[Published respectively, in May 1957 and November 1957.]
Subsequently, Mr. Baden synthesized the individual parts, added sections not homogeneous with the earlier topics, bought the account up to date, and rewrote great portions to correspond to the plan for the total manuscript. He created the text that follows daring the span of 3 years, interspersed with other historical assignments and nearly 2 years in another Air Force agency. In the process, the manuscript twice outgrew practical size limitations and had to be substantially abbreviated. Even with this effort, it finally filled four volumes. Although other members of the historical staff read and commented on the manuscript, and the final draft received full coordination in the Weapons Guidance Laboratory and individual aircraft project offices, errors of fact or interpretation are Mr. Baden's.
The narrative is in many respects unique, combining narrative style with ingredients having a strongly technical flavor. Because of its classified content, few scholars will even see it, and it must satisfy only the needs of military and civilian members of the Air Force. It is an attempt to take a highly complex technical subject, with tendrils searching in all directions, and compact it into a coherent narrative, accurate and carefully documented, which has utility for the present as well as the future.
As noted elsewhere, during the period when this history was in preparation the parent organization experienced major changes in organization and structure. Initial research was conducted while the historical office was part of Wright Air Development Center, with finishing touches added under the aegis of Wright Air Development Division. As the printing process began, both Wright Air Development Division and the Air Research and Development Command were dissolved, as were the Air Materiel Command and its Wright-Patterson production and procurement element, the Aeronautical Systems Center. Thus this history appears as a product of a successor organization that is not mentioned in the pages of the history: the Aeronautical Systems Division of the Air Force Systems Command. No attempt has been made to revise the narrative to reflect organizational changes of this nature. References to organization, where they appear, relate to structures that existed at the time of the event being discussed. Th- organizational assignments of individuals, both military and civilian, are handled in the same fashion.
Chief, Historical Division
Office of Information
As was true of all transportation devices, the airplane had scarcely passed through its embryonic stage before some ingenious human cast it in a military role. Machine guns and radios were mounted in aircraft, on several occasions in the years before world War I, and a carefree bombing accuracy contest was a fixture of "aeroplane meets" during the early teens of the twentieth century. Bombs were dropped from aircraft during the Balkan Wars of 1912-1913, and observation flights were common to the pilots employed by all military services after 1909- When both opponents possessed aircraft in fair quantities—a situation that, came about for the first time with the start of World War I—it was only natural that both air and ground forces should attempt to destroy enemy airpower. (Anti-air artillery was devised and used during the American Civil War, the Franco-Prussian War, and the Spanish-American War, although the targets in all instances were balloons.) The comradely wave of the hand that marked contact between opposing aviators in September 1914 was supplanted, within three months, by pistols, shotguns, rifles, and finally machine guns. Gun-aiming and bombing-aiming devices were not far behind; by mid-1915 relatively sophisticated techniques of bombing and fire control had been incorporated into combat aircraft on both sides.
It was apparent early that the effectiveness of airborne armament was more directly related to the precision of aiming devices than to the weight of lethal projectiles. The Browning machine gun of the first World War was still in widespread use when World War II came to a close, and the average bomb of 1945 weighed somewhat less than the larger bombs developed for use against Imperial Germany. Apart from ground-to-air fire, which occupied a separate category, aerial warfare rapidly became a matter of solving three aiming problems: attack on another aircraft, defense against another aircraft, and attack on ground installations of various types. To the degree that weapon effectiveness was always a factor, fire control and bombing accuracy were not the only determinants of success in these assignments— but they dominated the scene.
In the twenty years between World War I and World War II, the air services of the world devoted considerable energy to developing strategic and tactical doctrine and to devising mechanisms that would enhance the accuracy of airborne munitions. Without significant exception, the resulting devices were mechanical in nature—withall somewhat complicated at times. All were concerned with ballistic—non variable—projectiles or bombs. Again without exception, each was based on optical sighting. In each instance the basic problem was motion compensation and the calculation of collision points. In the case of a forward-firing fighter aircraft attacking another aircraft, the problem was to hit a rapidly moving target from a vehicle approaching at an angle, at a different rate of speed, and firing projectiles moving at yet another velocity in accordance with the laws of ballistics. The bomber-fighter problem was even more complicated because the defensive fire was affected by the forward speed of the bomber, the closing rate of the attacking fighter, the flight path of the attacker, and the angle at which the defending guns were elevated or depressed. The bombing problem was a complex of air speed, air density, ballistic characteristics, direction of flight, actual ground track and speed,, and the effect of intervening air masses moving at various rates and directions.
To a considerable extent, mechanical devices had been developed to solve all of these problems by the tune World War II began. From the standpoint of armament development, the only mechanical innovation of any consequence during that period was the employment of remote sighting stations for bomber defense fire control. Toward the end of the war, rockets were used for both intercept and ground attack purposes, introducing a new ballistic element, but existent fire control mechanisms adequately coped with the problem. Both allied and Axis air forces used relatively primitive guided bombs of one sort or another, but the guidance was generally limited to correction of terminal errors and invariably was managed by optical sighting and aerodynamic control techniques.
[Obviously, both statements apply only to air-launched munitions.]
The real revolution in airborne armament began with the initial use of electronics. Airborne radar was originally adapted to bombing through overcast and for night-interception of bombers. It arrived momentarily before the reaction engine—the turbojet—expanded the performance envelope of aircraft and nuclear weapons changed the character of bombing. Coupled with computer mechanisms, radar could readily solve problems of three-dimensional warfare, could eliminate concern for the limited utility of optical sighting devices, and could compensate for the vastly increased speed "and altitude of aircraft. Nuclear weapons significantly relaxed accuracy requirements in bombing, and promised to have the same effect on interceptor and bomber fire control problems if they could be adapted to air-to-air warfare. A further innovation was the use of air-launched missiles in three categories: bombardment, air defense, and intercept. Such weapons were no longer pure ballistic projectiles but could change course after separating from the parent aircraft.
By 1960, yet another set of circumstances had affected the environment of air warfare. Rudimentary surface-launched missiles introduced in the waning months of World War II progressed to the status of operationally useful strategic weapons in the next 15 years. With allowances for several limitations—they were non-recallable, they could not intelligently select targets, and they had to be dispatched to known targets—missiles had some advantages over aircraft. (The uncertain value of a single human life complicated any calculations of relative costs in the missile versus bomber controversy.) Whether the manned aircraft would remain a significant element in strategic warfare was a question that still had not been answered in 1961. Equally uncertain were the answers to questions involving the ability of any manned bomber to survive the onslaught of ground-based defensive missiles and the ability of a manned interceptor to operate effectively against any Mach 2 or Mach 3 bomber
The answer to most problems of employing airborne armament was automation. By 1961, all of the manned combat aircraft in development were equipped with electronic and computer elements monitored by men. Highly sophisticated semiautomatic equipment was in widespread use, and the electronic links between ground stations and combat-bound aircraft were vital to the functioning of the total weapon system.
The increasing complexity, and the vastly increased cost, of 1960-1970 weapon systems over their 1945-1955 predecessors introduced complications undreamed of by the immediate postwar planners. Thus the raw cost of developing an effective defensive system for the B-70 weighed heavily against the prospect of developing the aircraft at all. Without significantly compromising the peacetime economy, the national government could not afford to fund massively expensive projects and—to elaborate on the case of the B-70—development of a major subsystem might well cost 100 or 1,000 times as much as development and prototype production of an entire 1945 bomber! By 1961 it was obvious that the era of bombers produced by the thousands had definitely passed. Nor was there any reasonable prospect of re-adopting the futile philosophy of the 1920's, based on the idea that in the event of war the nation would have time to manufacture vast numbers of weapons earlier entered into small scale production. After 1955, any major war promised to be a combat fought entirely with the weapons on hand at the moment of its beginning.
In the final analysis, then, airborne armament was but one element in an equation that defined an air battle which in its own right was one facet of national strategy. The situation was marvelously like that defined by Winston Churchill during World War II-- the "Wizard War" that England waged against the Luftwaffe. Churchill's account of Britain's struggle with technology, of an inner conflict "going on step by step, month by month," of a "secret war, whose battles were lost or won unknown to the public," aptly described the 1961 situation. His conclusions then were all too applicable to the 1960's:
[Winston S. Churchill, The Second World War, Boston, 1949, II, pp. 381-382.]
Yet if we had not mastered its [science's] profound meaning and used its mysteries even while we saw them only in the glimpse, all the prowess of the fighting airmen, all the bravery and sacrifices of the people, would have been in vain. Unless British science had proven superior to German, and unless its strange sinister resources had been effectively brought to bear on the struggle for survival, we might well have been defeated, and, being defeated, destroyed.
Chief, Historical Division
Office of Information
CHAPTER I EARLY OPTICAL BOMBSIGHT DEVELOPMENT
Aerial bombing as such began, according to the records, in 1849 whan Austrian armies bent on crashing the Italian revolt against Hapsburg rule loosed bomb-laden hot-air balloons against the city of Venice. The Austrians won, but it was doubtful that aerial bombardment played a decisive role in the outcome.
Bombing from balloons remained an accepted practice for the next 50 years, until a conference at The Hague in 1899 branded it "inhuman" and banned the whole business. In 1907, however, with military possibilities of aircraft attracting considerable attention, all restrictions against bombing from the air were lifted by international agreement. Some four years later, the Italian aircraft were bombing the Turks in North Africa, thus ushering in a new era in warfare.
In 1911, an American Coast Artillery officer, Lieutenant Riley E. Scott, built a bombsight of his own design and tested it at College Park, Maryland. The device took account of speed and altitude in determining bomb release point. Lieutenant Scott lay prone on the wing of the aircraft to aim the sight and, using hastily prepared bombing tables to make his bombsight settings, was able to place two 18-pound bombs within 10 feet of a four-by-five-foot target from a height of 400 feet. In January 1912, Scott and his bombsight won first place and $5,000 in the Michelin bombing competition (at Villacoublay Airdrome in France) by scoring 12 hits on a 125-by-375-foot target with 15 bombs dropped from 800 meters.
Later in the year, at Hammondsport, New York (in the heart of the wine country) 19-year-old Lawrence Sperry built and successfully flight tested an aircraft gyroscopic stabilizer designed especially for one of Glenn Curtiss' early experimental flying boats. In June 1914, just before the outbreak of World War I, Sperry won 50,000 francs from the French War Department by demonstrating a "stable aircraft" which he flew without touching the controls and with a mechanic walking on one wing.
[The Sperry stabilizer was actually a simple automatic pilot.]
Sperry stabilisers were installed after 1917 in all Allied heavy bombers to keep the aircraft steady on bombing runs, but the idea of gyroscopically stabilizing the bombsight itself did not gain currency until some ten years later. Once this idea caught hold, amazing progress was made in increasing bomb dropping precision.
World War I saw many "firsts" in the history of air warfare, and particularly in aerial bombardment. In 1914, German Zeppelin dirigibles hit England on the first known concentrated bombing
attack and the first long distance raid.1 Bombing from aircraft also started, initially with pilots or bombardiers carrying bombs in a kind of basket and tossing them out with only guess work aiming, or with the bombs suspended by their noses from various parts of the aircraft and released tail first.
However, as the intensity of the air war increased and the role of aircraft was expanded, more precise bombing was required. For a time, this meant nothing more than bomb aiming by sighting off parts of the aircraft (such as struts and engine cylinders) and representing a few dropping angles by lines painted on the fuselage; the lines were usually drawn by individual pilots after trial-and-error experiments. But it was soon apparent that sighting instruments were necessary for greater accuracy.
Sights for low altitude bombing appeared first, and in the German camp. Here, the Goerz bombsight was developed and installed in Gotha heavy bombers. The sight consisted of a telescopic range finder with a rotatable prism at its lower end. Using the range finder, the bombardier could find the altitude and speed of the aircraft and from a prepared table determine the bombing angle at which to set the prism. After adjusting the prism, the bombardier made his bomb run and released the bombs when the target came into view. But the sight (and all other wartime bombsights on both sides) was only as good as the ballistic data for different bomb shapes, altitudes, and speeds and these figures were strikingly unreliable. The lack of bombsight stabilization and of an accurate vertical reference also hampered accuracy.
For the allies, two English and one French bombsight appeared; the English Central Flying School Number Seven bombsight (which took its settings from a direct-reading stopwatch); the French Mark II and an improved low altitude "Course Setting Sight" developed late in 1917 by Lieutenant Commander H. E. Wimperis of the English Air Ministry and Imperial College of Science in London.
The essential feature of the Wimperis sight (called the British Mark III) was drift compensation that permitted the aircraft to bomb from any direction. However, the Mark III was designed for use at altitudes under 1,500 feet and by mid-1918 highly accurate low altitude anti-aircraft gunfire and the appearance of larger bombers created the need for high altitude sights. The solution was another Wimperis design known as the Mark I.
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
The Mark I-A Bombsight
On 7 December 1917, drawings and models of the Mark I were sent to the United States and the joint Army-Navy Aircraft Board at McCook Field, Dayton, Ohio, was authorized by the Signal Corps' Air Service to procure both Mark I's and Mark I-A's (a modification of the Mark I for use in aircraft not built essentially for bombing). The Mark I saw extensive American service during the war. In September 1921 and again in 1923 General Mitchell's Martin bombers used the Mark I to sink three capital ships with 2,000 and 4,000 pound bombs from about 10,000 feet altitude.2
The three most important bombsight developments of the 1920's and 1930's—and those most significant for the future—were the Sparry Gyroscope Company "C-Series," the Georges Estoppey "D-Series," and the C. L. Norden "M-Series." Each group was developed more or less simultaneously, with Air Corps emphasis shifting back and forth between the Sparry and Estoppey designs for the most part. The Norden instruments were under the aegis of the Navy and did not receive Army attention until the early 1930's, when their design had almost fully matured. The "C-Series" sights had gyroscopic stabilization while the "D-Series" sights were stabilized by pendulums. Although the ultimate choice of the Engineering Division was to fall on gyroscopes, the Sperry sights generally took a back seat to the Estoppey series for at least ten years following World War I.
The "C-Series" bombsights were designed by Alexander P. Seversky (or de Seversky). In 1921 the Sperry Gyroscope Company received a development contract for the equipment and Seversky agreed to assist the Sperry organization in developing the C-1. One model was built and tested at McCook Field beginning in 1922. This was the first precision type bombsight to incorporate both the synchronous principle and gyro stabilization.4
[In the synchronous system, the bombardier merely kept the crosshair on the target or aiming point until compensating motors within the system held them. When the crosshairs were thus "synchronized" with the apparent motion of the target, the computer could determine the correct release point, assuming, of course, that the correct allowances for bomb ballistics had been made prior to synchronization.]
The C-2, a subsequent redesign of the C-1, retained the Seversky computer and the synchronous principle, but incorporated a new pilot directing system and Zeiss optics for greater bombing precision. In 1924 Sperry contracted to develop the sight, but nothing was accomplished because of a deadlocked controversy over design.
In 1926 Seversky joined the Sperry organization, and McCook Field entered into an agreement with the company to further the development of the "C-Series." The result, after five years of effort, vas an improved version of the original sight, dubbed the C-4. However, the new unit was cumbersome, its gyro was too small to give adequate stabilization, and it was never adopted by the Army.
[During this interval a variety of sight designs tentatively designated C-3 types were drawn up to determine the best configuration for the C-4.]
Five years later Sperry received a contract to redesign and improve the old C-4, and came up with the N-1. The company also designed another precision synchronous sight, the O-1. But both were bulky, heavy, and complex, and the Air Corps had already concluded that no benefits could be gained from further modification of the "C-Series," so the entire effort was terminated.
While the Air Service had thus been devoting effort to the Sperry (gyroscopically stabilized sight) developments, the bombers of the period were generally equipped with the "D-Series" (pendulum stabilised) sights designed by Georges Estoppey. Estoppey came to McCook Field in July 1921 and immediately began work on what soon became the D-1 sight. At the time, only the improved Wimperis Mark I sight was available, and it was rapidly approaching obsolescence; procurement of an improved bombsight was imperative and the Estoppey D-1 was chosen; it was more accurate than the Mark I, was very light, simple to operate, and quite inexpensive: average cost of the early models was about $300 each.
The "D-Series" sights were time-of-fall devices using a separate synchronous means for determining ground speed, and providing impact location based on the groundspeed observation. The bombsight obtained a vertical reference by means of a pendulum. While the D-1 was designed for high altitude bombing (between 4,000 and 22,000 feet), the entire series came to see most service at relatively low altitudes.
In 1926, Estoppey completely redesigned the D-1. The new configuration, called the D-4, was accepted by the Army and produced in several versions and in considerable quantities by the Gaertner Scientific Corporation between 1926 and 1929.
[Previously, the D-2 and D-3 experimental sights incorporating such functions as automatic timing and bomb release, and prism optics, had been created. The D-3 was built by Sperry from the Estoppey design.]
Further improvements to the "D-Series" resulted in the D-5, in which the open wire optics of the D-4 were replaced by an optical collimator having an illuminated reticle. Although no D-6 sight ever reached the hardware stage, an improved D-5 with a synchronous optical lens system appeared with a D-7 label. The computing mechanism and principles of the D-7 were also used in the C-5, a sight containing a Sperry gyroscope for stabilization rather than a pendulum. Both displayed insufficient accuracy and were built only in limited quantities during 1931 and 1932. The D-4 continued as the Army's standard bombsight until 1939, when the outbreak of World War II stimulated the development of the D-8, a D-4 modified for low altitude use and simplified for production line fabrication. The D-8, however, was only to serve until the more elaborate precision bombsights already under development at Sperry and the Norden Company were available. In any case, it was already quite evident that pendulum stabilization was incapable of compensating for more than the most insignificant aircraft maneuvers, and that gyroscopic stabilization of some kind was the ultimate answer to the development of true precision bombsights. The most likely prospect seemed to the the Norden bombsight.5
(PICTURE NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
In October 1931, Army observers witnessed a Navy demonstration of the Norden Mark XV sight, and were so impressed that the chief of the Air Corps asked the Navy to furnish 25 of the sights for testing.
The first was delivered in April 1933.6 Tests revealed that the Mark XV performed far better than any bombsight known to that time, and was far superior to anything the Army had either in production or development. The Air Corps then purchased 78 more Norden sights. All negotiations with the Norden firm were through the Navy Department.
The Norden sights (called the "M-Series" by the Air Corps) operated on the synchronous principle, had effective gyro stabilization, and were relatively easy to use; some models could be connected mechanically to a special automatic pilot—dubbed the Automatic Flight Control Equipment (AFCE)—also developed by the Norden firm. By January 1936, a total of 100 bombsights had been delivered to the Army, but the Air Corps had been unsuccessful in getting Navy Department approval for desired design improvements by the Norden firm.
The Army wanted the sights to be calibrated for higher speeds and altitudes and the autopilot equipment simplified to reduce maintenance difficulties. Eventually, the Air Corps undertook to have the improvements separately incorporated in the bombsight, principally by the Victor Adding Machine Company. In addition, the Minneapolis-Honeywell Regulator Company received a contract for the fabrication of the C-1 electrical autopilot containing the desired (simplified) maintenance features. By 1942 the Norden bombsight had been released to all Allied nations using United States Army aircraft—except Russia, It was the first true precision aiming device worthy of the name.7
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
The Norden Bombsight
If properly set and operated, the Norden equipment was capable of generating the correct range and course for bombing irrespective of wind, target, motion, altitude, or airspeed. Before range and course could be determined, however, the amount of trail for a given bomb shape, (*) aircraft altitude and airspeed, the actual time of fall of the bomb, and crosswind (**) had to be equated in the bombing solution.
[* - Air resistance acting on a bomb after release caused it to lag behind the drop point and hit somewhere behind the bomber. The distance from a point beneath the aircraft at the instant of bomb impact to the point of bomb impact was called "trail." Trail increased as the bomber's airspeed increased or as its altitude increased. Furthermore, since different bombs encountered different resistance in the air, trail was also a factor of bomb shape.]
[** - Crosswinds brought into the bombing problem a new factor, "drift" In order to fly a given ground track in a crosswind, an aircraft had to "crab" into the wind; the angle formed between the aircraft's true heading and its ground track was called the "drift angle" In a crosswind, the bomb would impact directly behind the aircraft and along its longitudinal axis at the moment of release. But this meant that the bomb would strike the ground at some point downwind of the aircraft's ground track. Thus, in order to score a hit, the bomber had to fly a ground track that ran upwind of the target.
The new variable added to the bombing problem by drift was called "cross trail." This was the least distance from the point of bomb impact to the ground projection of the aircraft's true heading line. Cross trail depended on the amount of trail and the magnitude of the drift angle.8]
Since the bomb left the aircraft in the direction of flight, the course problem solution produced the aircraft heading required to score a target hit, while the range solution determined a release point far enough back from the target to allow correct bomb impact.
The Norden series reached its peak in the M-9, a project which began In February 1942 when Wright Field planned to improve an earlier Norden model (the M-7), When the Air Force was unable to obtain satisfactory production type drawings from the C. L. Norden Company, Victor was given an actual M-7 from which to prepare suitable drawings and specifications for the M-9. The first production sight was received in April 1943, and by July M-9's were flowing off the lines in considerable quantities.9
The M-9 bombsight had two main units, the sight head and the stabilizer. The sight head consisted of an optical telescope and a "rate end" which was actually an analogue computer controlling movement of the optics. The stabilizer, attached to the sight head, contained a vertical gyroscope and a directional gyroscope. The vertical gyro was used to stabilize the optics of the sight head in pitch and roll, while the directional gyro stabilized the entire sight body in azimuth, in addition to providing the heading-reference for the autopilot.
In flying a mission, the bombardier adjusted the sight for bomb ballistic data obtained from bombing tables, which allowed for true airspeed, altitude, and the type of bomb to be dropped. The only remaining unknown factor in the problem was groundspeed, and this was obtained by centering the target in the optical crosshairs and adjusting a range knob and a drift knob until all apparent motion of the crosshairs was "killed" and the sight tracked the target without further need of attention.
From this moment on, the M-9 computer was able to set up a "dropping angle" identified by a pointer in an index window on the sight head; another pointer in the same window moved toward the drop-angle indicator to show the ever-decreasing angle of the line of sight. When the two pointers were in coincidence, either the bombs were released automatically or the bombardier released them manually.
The Norden M-9 was notable for its relative simplicity and light weight. The simplicity of the sight, however, demanded considerable dexterity on the part of the bombardier, particularly during the actual bombing run. He not only had to accurately adjust the two knobs on the sight head (which was no easy task), but he also had to keep an eye on the attitude of the vertical gyro, erecting it manually after un programmed aircraft maneuvers. The addition of radar equipment to bombardment aircraft during World War II made the bombardier's operations even more complex and posed an immediate and urgent requirement for a bombsight with more automatic features than the M-9.10
The "M-Series" bombsights reached their peak of development shortly after the end of the war. The automatic erection feature of previous "M-Series" was added to the M-9; the telescope was replaced by a collimated optical system; and the computer was rebuilt to handle trail values up to 300 mils, automatic cross trail computations up to 150 mils, and to allow manual setting of trail and cross trail values independently of one another. In addition, the equipment was modified to permit the use of an offset aiming point during bomb runs. The Norden sights used in B-29's were subsequently retrofitted to the improved configuration. Nevertheless, the great day of the Norden bombsight was over; other bombing systems incorporating both radar and optics plus more automatic features were already in development.11
The legend of pickle barrel bombing accuracy by the Army Air Forces was admittedly exaggerated. While some practice bomb drops under ideal weather conditions by trained and highly skilled crews gave fantastically good results, and while the combat record of American bombardiers, though naturally not as good as training school scores, was highly creditable, the ideal of hitting the pickle barrel or the mailbox on the corner remained merely an ideal. There were of course a number of reasons why this should be so.
Given a "perfect" bombsight, a perfect bombardier, and a perfectly stable airplane, if the bombardier synchronized his crosshairs from 25,000 feet altitude so that they remained absolutely centered on the target for the last portion of the bomb run, the bomb should hit the target dead center. Why might this not happen? At the 10,000-foot level, a strong wind, blowing at right angles to the direction of the wind at the altitude of the bomber, could exert sufficient force to deflect the bomb anywhere from 50 to 300 feet away from the target at the moment of impact.
[The existence of different atmospheric conditions below the bomber was one problem which defied adequate solution, even in 1960. The best that the bombardier could do was to assume that conditions were the same down to the target. Generally, only small differences which did not seriously affect the bombing solution appeared.]
This was but one probable source of error. There were many others.
Nevertheless, the American aim of precision bombing produced bombsights and bombardiers that chalked up a record never approached by any of the other air forces—enemy or allied—in World War II. There seen to have been two principal reasons for the American success. One arose from the fact that the American philosophy from the beginning was to obtain the utmost accuracy in bombdrops. European and Asian nations tended to follow General Douhet's dictum that the best utilization of bombardnent aircraft was to neutralize an "area," to "saturate" it, to terrorize its inhabitants. This basic difference in philosophy affected the type of equipment designed and built and the results obtained.
A second reason followed from the first. In accord with the Army's— desire for precision bombing, the Ballistics Research Laboratories of the Ordnance Department, through hundreds and thousands of test drops, compiled by far the most voluminous and the most accurate bomb shape ballistic data in the world. The compilation began in the early 1920's and never ceased. The degree of imperfection of foreign bombsights could be attributed, in many cases, to the fact that no country outside the United States, even as late as 1942, had assembled ballistic data comparable to that of the Ordnance Department. Thus errors in ballistic data alone tended to qualify attempts by other countries to do precision bombing.
CHAPTER II BOMBING WITH RADAR
World War II proved the accuracy of the American precision bomb-sights. It also spotlighted shortcomings, not the least of which was inability to bomb except under conditions of good visibility. Darkness, cloud cover, haze, fog, smoke from fires and explosions in the target area: any of these made the optical bombsight worse than useless.
Worst of all, these obstacles appeared more often than not, making "Bombing Through Overcast" a vital technical requirement.12
In the summer of 1940, the Western Electric Company proposed the development of "A Device for Detecting Objects Through Overcast." This was essentially a crude airborne radar set based on a British model known as 'Mickey." It was to be capable of detecting the presence of and pinpointing ship positions at night and during inclement weather. As the company planned to use some Sperry components in the search radar, Wright Field authorized both organizations to work together on the development.
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
Initial Radar Bombing Systems
The radar, known as the ASV-10, was interconnected with a precision ranging unit and a bombing computer. These basic elements, together with a control box, radar scope, and antenna comprised the bombing system tentatively called the H2X. By early 1943, the H2X had evolved into two radar bombsights, the AN/APQ-13 and the AN/APS-15. The APQ-13 was very similar to the British Mickey and was the principal radar bombing set housed in B-17's and B-24's. The APS-15, also used in these aircraft, incorporated the best features of separate radar developments at Western Electric (Bell Telephone Laboratories), Sperry Gyroscope Company, and the National Defense Research Committee Radiation Laboratory at the Massachusetts Institute of Technology.13
The Army Air Forces' philosophy of precision bombing led to further radar developments, particularly at Bell Telephone Laboratories, culminating in the AN/APQ-7 high altitude, high resolution bombing radar. This, the Eagle project, got under way in April 1943 (at about the time APQ-13 and APS-15 radar bombing systems were going into full production). The APQ-7 was specifically designed for use against such point targets as factories, docks, and bridges. Not the least important factor in its relative success was the bulky "Eagle" antenna (designed and developed by the Radiation Laboratory at Massachusetts Institute of Technology) mounted in the leading edge of a wing-shaped fairing (Eagle Wing) under the bomber's fuselage just ahead of the bomb bays. The 1,100-pound APQ-7 and antenna were capable of scanning the ground from 10 to 30 miles ahead of the aircraft, but not through 360 degrees. The system featured a built-in impact-predicting computer adaptable to 11 different bomb types.
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
Experimental Housing For Eagle Antenna on a B-24.
APQ-7 equipment was tested throughout 1943 and 1944 in B-24, B-25, and finally B-29 aircraft. The unit, which required a separate radar operator and bombardier, went into operational use for the first time in May 1945 during B-29 raids against Japan from the Marianas. Some men familiar with radar bombing systems have claimed that no other set ever gave the clarity and resolution of the APQ-7. Unfortunately, the clarity was principally due to the unconventional antenna which could neither be reduced in size nor compressed enough to meet the aerodynamic exigencies of higher speed bombardment aircraft. For this reason, use of the APQ-7 was limited to the B-29.
However, the two-man synchronous procedure was cumbersome, complicated, and time-consuming, and by early 1944 operational units were calling loudly for soma kind of device that would make it possible for one man to operate the Norden sight using information obtained directly from a radar scope. Project Nosmo (Norden Optical Sighting Mechanized Operations) was established at Wright Field to bring about the desired optical-radar conjunction.
By the spring of 1945, successful Nosmo devices had been service tested and installation had begun in operational bombers using APQ-13, APS-15, and APQ-7 radars. The first model, known as Adapter Assembly AN/APA-46, allowed a straight-line approach to the target during which the bombardier could check the radar against optical sighting at four of five predetermined points. The equipment, which was surprisingly small in size and simple in construction, utilized an auxiliary Norden bombsight computer to control movement of the radar scope marker and allowed synchronous operation of the radar with the optical sight. With improved Nosmo, called AN/APA-47, the bombardier could check the synchronization of the radar and optical portions of the bombing system at any time during the bomb run, eliminating the need to perform this operation at only a few predetermined points.
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
The NOSMO System
By July 1945, a Nosmo unit was being made an integral part of the Eagle bombing system, and the entire complex received the designation AN/APQ-7A, Although the Nosmo devices received only limited combat use (all hostilities ceased shortly after they were introduced), the concept of giving one bombardier the discretion of using either radar or optical bomb aiming had been proved. On this foundation the postwar bombing-navigation systems were developed.15
The postwar era also witnessed the arrival of Jet propelled aircraft—aircraft potentially capable of flying supersonically and at stratospheric altitudes—and new and stringent bombing requirements. The materiel command's Engineering Division, had started to think about the problems inherent in supersonic high altitude bombing even before the war ended. But concrete research and development guidelines awaited a meeting of the Army Air Forces Scientific Advisory Group in 1946.
The advisory group was able to make a number of general recommendations for the design of bombing systems to be housed In future high performance bombers. First of all, since optics were of no practical value at 1,000 knots, high resolution search radar capable of rapid scanning would be required; the transmitted wavelength would have to be as short as possible while the antenna rotated rapidly to illuminate the target as much as possible. Second, the diameter of the antenna would have to be increased to the utmost, limited only by the aerodynamic characteristics of the carrier aircraft. Third, the bombing computer should require only a minimum of adjustment after target recognition.
This latter goal could be reached, the group believed, by "pre-synchronization"—designing computers capable of measuring and "remembering" aircraft groundspeed, drift angle and altitude prior to the bomb run. A computer which already "knew" these things need only be set once on the actual target in order to indicate to the pilot his correct course. Two such computers were already under development, the Western Electric APA-44 and the Sperry SRC-1. Nevertheless, the group warned that neither of these devices was calibrated for high enough speeds or altitudes.
Another recommendation was that the Army Air Forces consider using offset aiming points at relatively large distances from the target. Radar, by virtue of its "map-drawing" ability, made offset bombing relatively simple. Thus, the aiming point might be taken something like 15 miles in advance of the release point in the case of a 1,200 mile-an-hour aircraft. At this distance, a bombsight suitable for a B-24 would also display sufficiently fine detail for a faster bomber. However, a compass accurate to about one-tenth of a degree would be required for precision offset bombing, and none was available in 1946.
The advisory group emphasized the need for a precision true-airspeed meter. Airspeed, direction, and prevailing wind data (obtained from the radar) would permit the computer to compile precise course information,
An alternate suggestion was to develop a Doppler-principle radar.
[The Doppler principle was stated by Christian Doppler (1803-1853), a German mathematician. Doppler observed that as the distance between a source of constant vibrations and an observer diminished or increased, the frequencies appeared greater or less.]
Such equipment would render a true airspeed meter unnecessary and would also eliminate the need for determining drift by sighting with the search set. The disadvantage in the Doppler idea was that it would probably require an additional radar set and antenna, thus creating a space and weight problem.
The advisory group urged that the entire radar, altimeter, compass, airspeed, navigational and bombing equipment of a bombardment aircraft should be designed and treated as a single integrated unit. Thus, the concept of a "bombing-navigational system" rather than just a "bombsight" began to receive emphasis.
The first synchronous radar bombing system, the AN/APQ-23 (essentially the APQ-13 search radar combined with the new CP-16 computer) had been under development since March 1944 for possible installation in B-17, B-24, B-29, and other heavy bombardment aircraft. The APQ-23 system supplied range, azimuth, distance, and drift Information to both pilot and bombardier and could perform offset as well as direct bomb aiming3 The drawback of the system was the drawback of the APQ-13— a lack of resolution and the consequent limitation of the radar sight to the high altitude bombing of large and clearly distinguishable targets. By the end of the war, Western Electric had improved the resolution of the radar by developing a special 60-inch rotating-dish antenna, but the APQ-23 was still in the test stage when hostilities ceased.17
The project was completed after World War II, and the unit was installed in B-29's and B-50's. In its final configuration, the APQ-23 operated satisfactorily up to about 30,000 feet altitude at ground speeds as groat as 440 knots. The radar had a tracking range of 15 nautical miles and could use offset aiming points as far as 30,000 feet from the target in range and at any azimuth bearing. The accuracy of the APQ-23 against an "ideal" point target (one which could be precisely located because of land-water contrast) was about 35 mils --this would be a ground distance of 350 feet from a ten thousand foot altitude. The APQ-23 was often tied in with the Norden optical sight by means of a Noamo type interconnection. Improved models saw service in the Korean fighting.18
The development of the Western Electric Company APA-44 electronic analogue computer, however, was the key factor in Air Force attempts to equip modern bombers with a truly integrated bombing-navigation system. The APA-44 was under development in 1945 and ultimately became a prime component of three bombing-navigation systems. An early model was delivered to Dayton late in 1945 and then subjected to an extensive flight test program primarily at Boca Raton, Florida.19
Three APA-44's had been built and delivered to the Air Force by the end of August 1946. The first model had already undergone eight months of flight testing before being returned to Bell Telephone Laboratories to be completely rebuilt. Together with a modified APQ-7 radar (designated the APS-24), it was installed in a B-29 aircraft in November to form the AN/APQ-16 radar bombing system. The third APA-44 model was installed in a B-17 and employed with a new Western Electric search radar, the APS-22, to form the AN/APQ-34 bombing-navigation system. The second model of the computer remained at the contractor's facility for design flight tests,20 completed in February 1947. At this juncture, the APQ-34 at Boca Raton was just beginning flight testing in a B-17, but engine trouble had thus far prevented B-29 flights with the APQ-16. February 1947 also marked the appearance of the development model of a new Western Electric search radar that was to make the other systems obsolete before they ever became operational.
This was the APS-23, a high resolution, high altitude radar using the best features of the APS-22 and incorporating the common 360-degree scan, provisions for "sector scan" (in which the antenna oscillated through any desired portion of a circle), and a "displaced center" scan with vertex of the sweep starting at or even below the bottom of the cathode ray tube. The company had been working on the APS-23 since mid-1945 and also intended to use this radar in a combined optical and electronic bombing system to be operated by one man.
The APS-23, when combined with the APA-44 computer, formed the APQ-24 bombing-navigation system, a fully synchronous radar system without provision for an optical bombsight tie-in. The AN/APQ-24 mock-up inspection took place at Boca Eaton in May 1947, and after the contractor complied with recommended modifications, system installation in a B-29 began early in June.21
[Tests with the APQ-34 and APQ-16 continued until December 1947.]
At 30,000 feet, the APQ-24 had a search range of from 150 to 200 miles against large cities and could efficiently map an area of about 75 miles radius. Design accuracy in the bombing mode was 25 mils against an "ideal" point target and 35 mils against an inland target without land-water contrast to aid in identification.22
Because of the extreme complexity and sensitivity of the system, the computer and radar had to be calibrated more precisely than any previous bombing system. Thus, the success of the APQ-24 in operational use depended on field adherence to strict calibration requirements. Moreover, alignment of the equipment had to be performed under flight conditions, not just on the ground. These factors did not override Air Force optimism however, and the equipment was earmarked for the B-50, B-36, and B-45.23
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
AN/APQ-24 Bombing-Navigational System
As of 1 April 1949, some 300 APQ-24 systems had been shipped by Western Electric; 100 had already been installed in B-50 and B-36 aircraft at Boeing (Seattle) and Consolidated-Vultee respectively. Within four months B-29 flight tests of the APQ-24 had ended and a widened flight program using B-50, B-45 and B-36 aircraft was getting under way. In September 1949, production of an improved and modernized system began at Western Electric and the contractor was planning to modernize the systems already installed. In addition, Western Electric was considering the possible benefits that might be derived from integrating the APQ-24 with the Doppler portion of an automatic navigation system (AN/APN-66) being developed by General Precision Laboratories, Incorporated.
Operational experience with the APQ-24 in training exercises within the Continental United States turned up a serious difficulty— low reliability. Indeed, the radar-abort rate during Strategic Air Command simulated war missions was somewhere between 25 and 30 percent.24
Air Force and industry officials meeting at Strategic Air Command headquarters, near Omaha, on 2 February 1951 thoroughly aired the reliability problem. Here, SAC officials insisted that the APQ-24, did not even approach the scope definition obtained with the World War II APQ-7, and that one of the command's greatest problems was target identification. Furthermore, while the APQ-24 system was relatively simple from the operator's point of view, its reliability was uncertain and its many hundreds of tubes and score or more of inter-connecting black boxes presented a maintenance nightmare. A further warning concerned the more complex "K System" (also using the APS-23) than being developed for both the B-36 and the new B-47: space limitations in the B-47 would make it impossible to carry along a bombing system mechanic, as in B-50's and B-36's; and in the B-36, the mechanic was often stymied because some APQ-24 components were inaccessable to him. The Strategic Air Command also indicated extreme concern about the possibility of the enemy's jamming the bombing system, a factor already under study by Western Electric; Air Force headquarters had recently flashed to its development agencies an urgent demand for non-jammable electronic subsystems.25
After this, General Curtis E. LeMay, strategic command chief, demanded that the Armament Laboratory (of the Engineering Division) and Western Electric (in the person of Dr. D. A. Quarles) vastly improve reliability in the APQ-24. The general said it was his opinion that the Armament Laboratory had "fallen completely flat on its face" in the attempt to provide a reliable and maintainable bombing system. He also turned verbal guns on electronics firms, expressing the belief that some contractors would have to "close up their doors and starve to death if they sold to the public equipment like that they are selling to the government."
He made an impassioned plea on behalf of his command, emphasizing that too much research and development effort was being directed toward what he considered non-essential equipment. He insisted that SAC s ability to accomplish "atomic offensives" was the item of primary importance to the conferees and that if such offensive ability were lacking, the United States would be lost. He scotched an attempted rebuttal by the Armament Laboratory's chief. Colonel G. A. Blake, by-saying that the good bombsights the Air Force had (meaning the Norden M-9) "were developed for us by the Navy." In order to impress those attending the conference with the sense of urgency he felt, General LeMay suggested to Dr. Quarles that the government, in attempting to improve bombing system reliability, might be forced to "buy Western Electric and throw out all .... civilian business in order to get the AN/APQ-24's modified . . . ."
The Strategic Air Command chief requested a concerted effort by the Armament Laboratory, Air Proving Ground Command, and Western Electric, aimed at improving and modifying the APQ-24 installations in all operational aircraft. His plan was to fix one airplane first, then one squadron, and then move on to one wing. He offered full cooperation to those who would be involved in the actual work. "I'll do anything, " he said. "I'll run any tests you want, I'll do anything I can possibly do to help improve the efficiency [of the Strategic Air Command] 22%. That's the big deal .... I look at this big gap in my efficiency and I can't do anything about it, except complain and call people nasty names and try to get them mad enough to do something."
How mad his listeners got was uncertain, but something they certainly did. The material and proving ground commands immediately established Project Reliable, primarily an attempt to relocate boxes in the B-36 (and later in the B-47) so that in-flight calibration, adjustment, and maintenance could be performed. In addition, engineers made provisions for more efficient cooling of tubes and components, and, where possible, more reliable vacuum tubes replaced those with the poorest operating histories. The project quickly came to include the "K-Series" systems in both B-36 and B-47 aircraft, while the "reliable" improvements to the APQ-24 and K-3A systems installed in B-36 and RB-36 aircraft were completed by the end of 1951, similar improvements to the K-2's and K-4A's in B-47's continued on into 1953.26
By the end of 1951, the APQ-24 reached the configuration it was to maintain until eventually phased out of the Air Force inventory. At that tine it was installed in B-50D's, B-36a's, and some B-45A's; while the newer operational aircraft were already being fitted with "K-Series" bombing-navigation systems which added an optical feature to the radar presentation.27
Thereafter, the APQ-24 was used as a "guinea pig" for bombing system improvements considered for the more versatile "K-Series" systems, primarily because the APS-23 was common to both. In February 1953, the Armament Laboratory installed one of the three "modernized" APQ-24 systems in a B-29 to flight tost a number of items, including a 10-inch (rather than a 5-inch) radar scope and a variable frequency radar of both high and low power designed to counteract enemy
[The Bell laboratories also investigated a Doppler radar tie-in, a tunable search radar to replace the APS-23, and several other items to improve APQ-24 performance and operational flexibility, but the results of these studies rapidly came to affect the K-Systems rather than the obsolescing APQ-24.]
Anti-jamming had become essential to all Air Force electronic systems and subsystems. Later improvements to the APS-23 included rapid-scan antennas, high definition radar scopes, data storage tubes (which could hold a radar picture for a considerable time and then reproduce it on the main system scope whenever the operator desired), K-band radar heads with tunable features, and flush-mounted radar bombing antennas for supersonic aircraft, Western Electric also developed a low power tunable modification kit which converted the APS-23 into the AN/APS-64.28
CHAPTER III: BOMBING-NAVIGATION SYSTEMS : THE K-SERIES
The wartime Wright Field analysis of bombing problems during the oncoming age of jet propulsion had one immediate effect, a 1 May 1944 request sent by the Engineering Division to the Armament Laboratory for a program leading to a combined optical and electronic bombing system suitable for jet bombers. Within six weeks the laboratory had queried various companies working in the bombing field and had recommended a system that would use "the best" in both radar and optics. The Sperry Gyroscope Company SRC-1 (Standard Radar Computer) bombing and navigation computer, although still in the design stage, seemed to offer the best possibilities.29
On 24 October, Sperry sent the Armament Laboratory a final SRC-1 development proposal. Six days later, the Engineering Division decided to procure two models (under Project MX-754) as the first step toward a one man bombing system in which the SRC-1 would be coupled to radar and optical elements.30
Events moved rapidly. By mid-1945, Western Electric had contracted to develop the APS-23 for the new system (as well as the APQ-24), and Sperry had completed negotiations with Eastman Kodak for an optical periscopic bombsight to operate in conjunction with the SRC-1. In November 1945 the Engineering Division halted the effort at Eastman and placed contracts with Farrand Optical Company for the design and fabrication of a vertical retractable optical periscope (designated Y-1) and with the Perkin-Elmer Corporation for a T-2 horizontal (non-retractable) periscope. The Massachusetts Institute of Technology also received a contract for a "coordinate converter" to transform the two-axis data from the horizontal Y-2 sight into three-axis data usable by the Sperry computer.31
The vertical periscope was for aircraft like the B-36 which possessed considerable headroom at the bombardier's station. A horizontal periscope would be necessary in slim-nosed jet bombardment aircraft. Periscope sights overcame the need for a flat glass bombsighting "window" in the aircraft's nose, as in the K-9. Such a plate would have a serious degrading effect on the speeds of jet aircraft.
Both the Y-1 and the Y-2 were being designed with sight lines stabilized in roll and pitch by means of remote signals from stable platforms. The first development model of each was scheduled for flight tests in B-29's.
The entire bombing-navigation system was to be capable of continuously computing and displaying ground speed, ground track bearing, wind velocity and wind direction, together with the latitude and longitude of the aircraft's position. The bomber was to be free to engage in evasive maneuvers both prior to and during the bomb run, such maneuvers to include changes in altitude as well as in azimuth. The Armament Laboratory expected that the first Sperry computer, would be installed in a test aircraft by March 1947.32
The SRC-1 computer, which for a time bore the tentative designation AN/APA-59, was based (as was the APA-44) on a "ground position indicating" circuit first developed by the Radiation Laboratory at the Massachusetts Institute of Technology. Indeed, it was this circuit which made the APQ-24 and the "K-Series" systems "bombing-navigation systems" rather than simply "bombsights." The APA-59 itself was designed under the assumption that navigation to and from the target was an integral, vital part of any bombing mission and that bombing and navigation problems had many common features.
In both operations, the system would have to determine the relative position of a recognizable landmark and track this landmark either optically or by means of radar.
[In the bombing mode, the terrain feature was referred to as a "target," an "initial point," or an "offset point." In the navigation mode, it might be called a "way point," a "check point," or the "destination." The tracking operation, if done in the bombing mode, would be called a "wind run" in the navigation mode, a "drift run."]
Any line of demarcation between navigation and bombing functions was purely arbitrary because an aircraft using equipment designed properly and built to sufficiently close tolerances might very well begin the bomb run the moment it left the ground. Moreover, if system components could be made completely automatic, eliminating the human operator, the equipment complex became a "missile guidance" system. This, then, was the direction in which bombing-navigation system developments were beginning to move.33
The APA-59 was a synchronous device. It incorporated a computing circuit that compensated for crosswinds, thus eliminating a major difficulty in earlier systems. The operator centered the crosshairs on the aiming point and then allowed then to drift away under the influence of existing crosswinds. At a particular moment, called the "memory point" he replaced them on the aiming point and the computing circuit determined the wind values as a proportion of the amount of correction necessary. The computer then introduced a "memory" of this wind value into all subsequent crosshair corrections made by the bombardier and evasive maneuvers performed by the pilot.34
By mid-September 1946, the radar portion of the bombing system (including the AFA-59 and the APS-23, plus interconnection equipment) had been designated AN/APQ-31. However, because of the lack of an APS-23, the Armament Laboratory planned to flight test the system with a Y-1 periscope and an APS-22 in a B-29 test aircraft.35
Within a month, Farrand's Y-1 had entered the detail drawing stage. Final outline drawings of the APA-59 computer were nearing completion at Sperry. Perkin-Elmer was still experimenting with various Y-2 configurations. MIT reported that engineering layout of the coordinate converter for the Y-2 was almost complete and that the first unit would be delivered in time for test with the first horizontal periscope.36
Specifications for individual components of the entire bombing-navigation system were completed in April 1947, but the B-29 flight test schedule had slipped badly due to delays at both Parkin-Elmer and Farrand and aircraft rewiring problems. The Y-1 periscope schedule had been disrupted primarily because of an afterthought requirement for an alternate optical (or B-scope) presentation and an integral scope camera in the single eyepiece. All this, however, pushed the program back far enough to allow initial installation of the APS-23 (instead of the APS-22) in the B-29.37
One month later the Army Air Forces planned to install the APA-59 bombing-navigation system in XB-53, XB-48, B-45A and XB-47 aircraft. The contractors for these projected jet bombers were receiving up-to-date information about the various components of the system.38
Sperry had installed and ground checked the radar portion of the bombing-navigation system and the computer (now called the A-1) in the B-29 by 1 February 1948. On 3 February, the equipment was for the first time flight checked as a unit. Flights on 9 and 16 February were generally satisfactory and showed that the maximum radar range obtained with the APS-23 equalled the exact line-of-sight range corresponding to the aircraft's altitude. The first vertical periscope still had not been delivered.39
On 26 March 1948, Sperry contracted to fabricate two "polar navigation attachments" for the A-1 computer. These "attachments" followed the design pattern submitted in response to the laboratory's March 1947 requirement. Three days later, Air Force headquarters published a set of military characteristics for precision bombing equipment which was to be designed:
... to solve bombing problems including the bomb run from a previously selected initial point, followed by location of the target and an accurate computation of all categories of bombs, individual and clustered as well as mines, both fixed and controllable, during all possible conditions of visibility.
By this time, the materiel command was under pressure to get the required equipment into operational units with all dispatch. Its Procurement Division had money on hand, and since there was no assurance that funds would be available six months later, the division was eager to award a production contract. A query to the Armament Laboratory on whether the system was "ready," followed at once. Although Air Force engineering tests were not scheduled to begin for six months, the Armament Laboratory took a "calculated risk" and told procurement officials to buy as many systems as they could. This meant, of course, that major defects uncovered in flight testing could be overcome solely through retrofit, generally a costly procedure. Nevertheless, contracts for 118 bombing-navigation systems were let to the appropriate firms.
This decision was to haunt both organizations for several years thereafter. The immediate result was the production of some 71 bombing-navigation systems which were not satisfactory for first line use. These were later replaced in combat aircraft and in some cases were reclaimed for bombardier training. But no matter which of the many K-systems to emerge were used in what aircraft, all were continually plagued by the need for a seemingly endless series of modifications, improvements, and changes emanating from systems deficiencies, technological advances, and changing system requirements.41
By March 1948, all components of the development system were finally completed, interconnected, and installed in a B-29 (at MacArthur Field). One month later, the bombing-navigation system (with the Y-1 vertical periscope) officially received the designation K-1 (after having been called the F-1 for several months). On 28 May, the first B-29 functional flight test of the K-1 system occurred. It was confined to a number of optical tracking runs at 10,000 feet altitude, producing a report that resolution of objects both on land and water at various degrees of magnification was very good. Tracking was easily accomplished and, all in all, results of the test were gratifying. The same equipment, but with the Y-2 horizontal periscope, later emerged as the K-2 system.42
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
The K-1 Modification Program
At the completion of this program, the Engineering Division was to install an autopilot in the B-29 prior to the start of Air Force engineering flight tests at Eglin Air Force Base later in the year. Sperry received word from Perkin-Elmer in mid-June that the first Y-2 horizontal periscope with coordinate converter would be delivered before the end of the month. By mid-August, the system contractor had built five experimental A-1 computers, and Wright Field was preparing production specifications for the Y-2 horizontal periscope and coordinate converter.
The K-1 had been demonstrated in all four of its bombing modes by 1 September, Skip bombing releases were considered satisfactory. Optical impact-predicting drops were outside specification limits until maintenance specialists located and remedied a faulty contact; thereafter results in this mode were good. However, neither the radar nor optical synchronous releases satisfied required accuracies on two bombing missions late in August, Sperry blamed a marginal wind multiplier unit in the computer which had given trouble during earlier tests and was scheduled to be replaced. Also at fault was an interconnection between the computer and the autopilot which was not as accurate as the computer and thus could not feed sufficiently precise heading data to the autopilot. Sperry was already fabricating an improved link that would keep the aircraft on an exact bomb release heading.44 Flight testing at MacArthur Field ended in October 1948, and the B-29 was flown to Warner-Robins Air Force Base, Georgia, to be readied for Air Force evaluation. The tests had proved that bombing runs could be as short as 15 seconds using the "memory point" tracking feature. The tracking range of the APS-23 radar reached out some 150,000 feet from the aircraft; projections indicated that the computer was capable of solving the bombing problem at ground speeds up 695 knots. The. total complex now weighed about 1,500 pounds.
The flight tests further revealed the need for periscopic improvements. The Armament Laboratory soon composed a set of requirements for a redesigned optical device incorporating the best features of the Y-1 and Y-2, but lighter in weight, having a variable barrel length to facilitate installation in aircraft of differing fuselage depths, and having magnification powers of one, four and eight diameters. Retractability was not considered absolutely necessary in the redesigned scope. As of mid-September 1948, the Air Force contemplated installation of the K-1 bombing-navigation system in B-45C, B-49A, B-36B, B-52, B-54A and B-55 aircraft, and the K-2 in the B-47A and B-47B.45
Warner-Robins K-1 tests began in December 1948. Early in January test crews dropped 15 bombs from 10,030 feet with an accuracy of 11 mils, well within required limits. By this time, the second computer, incorporating all changes found necessary during flight tests of the number one system, was ready for Armament Laboratory bench testing; bench test models of both the K-1 and K-2 systems were completed in March 1949. Boeing Airplane Company personnel at Wichita now planned to install a prototype K-2 system in the second XB-47 model within two months. In March, Air Force headquarters also raised the priority of the "K-Series" and APQ-21, test programs to the highest within the Air Materiel Command. One month later, Perkin-Elmer delivered the second Y-2 periscope, a coordinate converter, and power supply elements to Sperry for tests with the first production A-1 computer.
As 1949 progressed, "K-Series" systems found their way into more test aircraft. A B-50 was delivered to Warner-Robins in September and readied to accept the equipment for bombing to 40,000 feet and above. The prototype K-1 was completely installed in B-36 number 69 in mid-November 1949. The K-2 installation in the second XB-47 at Wichita was completed at about the same time; part of the delay resulting from the need to have the Massachusetts Institute of Technology replace the faulty coordinate converter originally sent to Boeing. By the end of the month, both the B-36 and XB-47 had made their initial flights; but functional tests were not to begin until after aircraft "debugging."47
Throughout 1949, new bombing-navigation system requirements were emerging from Air Force headquarters in response to recommendations from research and development organizations as well as from using commands. In May, the Pentagon directed the Engineering Division at Wright Field to study modifying the "K-Series" systems to provide the automatic drift and ground speed features of the Doppler navigation system (under development by General Precision Laboratories, Incorporated). The Doppler navigation system was called the AN/APN-66, and its radar portion (the part to be used with the K-system) the AN/APN-81. In December 1949, the Pentagon asked the Engineering Division to begin other studies aimed at improving optical-radar bombing-navigation systems for precision operation at all latitudes, very high altitudes, (at least 60,000 feet) and high speeds (900 knots and more). Air Force headquarters further asked that "automatic data inputs" be provided to the greatest extent possible.48
[The Armament Laboratory established a program for development of an automatic ballistic computer at Raytheon Manufacturing Company to meet the "automatic data inputs" requirement issued by the Pentagon in December 1949. The automatic ballistic computer would insert proper trail and time-of-fall values into the bombing system until the instant of bomb release, and thus free the bombardier from the task of determining these values and setting them manually into the computer. In June 1950, the project was shifted to Perkin-Elmer, where the feasibility of this concept was established, but requirement for use in the K-systems never appeared.]
In February 1950 the Armament Laboratory asked Sperry to study a union of the APN-81 Doppler radar with the K-1 and K-2 systems, limiting extra weight to 200 pounds or less if possible, Sperry was also to examine the feasibility of incorporating a number of additional features into the K-systems, such as fully automatic navigation, "automatic crosshair laying," automatic tracking of targets until after bomb burst (for bomb damage assessment) and automatic navigation in polar regions.49
Concurrently, a new element of urgency was added to the "K-Series" project when a small group of general officers in Washington saw photographs of a jammed K-system radar scope. The devastating effect of jamning on system operation highlighted by these pictures so alarmed the group that they hurriedly tacked a Top Secret classification on the pictures and immediately issued a requirement for an "inertial extrapolator"— a device to render the system immune from electronic countermeasures. The inertial extrapolator was essentially a guidance device which permitted inertial control of an aircraft for one hour before bomb release—or over a distance of 300 nautical miles. It was to supply the whole system with the rate and position information normally obtained by visual or radar tracking so that the bombardier had only to perform "final positioning of the bombing crosshairs for completion of the bombing run." This was the "K-Secure" program.
On 24 April 1950, the Director of Research and Development in Washington formally initiated the project, announcing the need for a "simple" inertial component for "K-Series" bombing-navigation systems to he installed in operational B-36 and B-47 aircraft by February 1952. The Pentagon was ready to authorize a crash program, "if necessary," but actually it was doubtful that such a program could be carried out on any other basis. On 4 May 1950 the Air Materiel Command published the procurement directive necessary to insure funding of the K-Secure program.
The inertial extrapolator never became part of the "K-Series" systems. Sperry began work on an expedited basis in May 1950, but quickly realized that the deadline could not possibly be met. The Armament Laboratory informed the Pentagon that early 1953 seemed the most optimistic delivery date. Thereafter the program ran into continuing difficulties. In May 1951 a laboratory study of the Sperry design disclosed that the size of the extrapolator was too great to allow installation in B-47 aircraft.
Finally, a short haul solution was approved. A "tunable" radar-one in which the operator could switch from a jammed to an unjammed frequency—was developed and incorporated in the "K-Series" systems. The inertial program continued, but the equipment was not developed in time for use with the "K-Series"; eventually it was incorporated in other advanced bombing-navigation systems.50
During 1950, changes in the K-system periscopes and computer led to a rash of new designations. Farrand designed a non-retractable vertical periscope similar to the Y-1 and called it the Y-3. Sperry redesigned the A-1 computer, adding an improved amplifier, tracking computer and navigation control, and the new device was called the A-1A. The bombing-navigation system incorporating the Y-3 periscope and the A-1A computer, dubbed the K-3A, came to be the standard system for B-36 aircraft and early production model B-52's.
At the same time, Perkin-Elmer developed a Y-4, horizontal periscopic sight which had binocular optics, was slimmer than the Y-2, and incorporated mounts for a camera which could photograph either the radar or the optical display. Together, the A-1A computer, the Y-1, sight and the APS-23 radar made up the K-4A bombing-navigation system installed in operational B-47 aircraft. The K-3A and K-4A systems were designated "alternate standard" on 6 November 1950.51
During January and February 1950, Warner-Robins was the scene of B-29 tests with the experimental K-1 system, Consolidated-Vultee engineering flight tests of the prototype K-1 in B-36 number 69, and B-50 tests of the bombing and navigation features of the first production K-1.52 In the spring and summer of the year an extensive and intensive program, known as Project Wibac, was established at the Boeing facility in Wichita to solve a host of B-47 problems, one of which was the K-system. Boeing had already complained to both the Air Force and Sperry that the K-2 could not be induced to work properly and that it was next to impossible to fit all the black boxes in the airframe. Moreover, reliability had suddenly become a major headache. Air Force and contractor personnel continued to hammer at B-47 reliability under Project Wibac until that undertaking was abandoned in favor of Project Reliable, which began in February 1952.
Boeing completed flight tests of the K-2 in the XB-47 at Wichita in April 1950, and Air Force flight testing began at Edwards Air Force Base in May. Comparative accuracies obtained with the APQ-24 system and the K-1 during recorded bomb drops indicated that the K-1 (in the radar mode) was not measuring up to its design accuracy while the APQ-24 was doing better than required. The flight tests of the K-2 system in the XB-47 were abruptly terminated on 7 June 1950 when the aircraft was assigned to atomic weapons effects testing scheduled for the Pacific Proving Ground the following spring (Operation Greenhouse). K-2 tests were to continue in a B-47A as soon as one could be made available at Edwards.53
Air Force accelerated service tests of the prototype K-1 system in the B-36 began on 24 June 1950 and were completed on 14 November, During about 100 hours of flying time covering 58 bomb drops and four long range navigation flights, each of which exceeded 2,000 nautical miles in range, the average error was less than one percent of the distance traveled, thus meeting specification requirements. The system operated for 86 flight hours without malfunction, leading the Armament Laboratory to report optimistically that the K-1 system was extremely reliable under controlled conditions and when maintained properly.
In September 1950, the Armament Laboratory officially called for incorporation of the APN-81 Doppler radar with the K-1 in order to provide more accurate ground speed and track data (the output of the APN-81 was also to be used in performing automatic navigation and automatic crosshair laying).
[Necessity for the Doppler integration had been established by the results of overwater flights and flights in the polar regions. The K-systems were incapable of obtaining accurate wind or drift values when flown over either water or snow. The Doppler addition would make tracking runs unnecessary.]
Bids from both Sperry and International Business Machines Corporation were evaluated on 7 February 1951 and the latter was selected. The Air Force gave IBM an A-1A computer to be modified for the APN-81 tie-in on 26 April.
By this time, the Armament Laboratory had under consideration a special computer to determine aircraft vertical velocity, altitude variation, and air speed (or ground speed) variation. This would enhance the existent limited capacity for evasive action during the bomb run and would also provide precise and immediate inputs to the automatic ballistic computer then under development at Perkin-Elmer. An "automatic astro-compass," utilizing a star-tracking telescope stabilized remotely by the bombing system gyros was also being considered for the "K-Series;" theoretically, it would provide heading information sufficiently precise to allow extremely accurate navigation over long distances, and bombing from offset points many miles from the target, Sperry promptly began studying the possibility of linking an astro-compass to the automatic ballistic computer.54
The vertical periscopes did not remain immune to the continuing drive for improvement. First, Farrand altered the Y-3 vertical periscope to utilize a shallower optical dome; the revision was designated Y-5, and scheduled for installation in the B-52, Next, the Armament Laboratory published a design exhibit for a 50-inch-long vertical periscope, designated the Y-7, which incorporated the shallow dome of the Y-5. Both Farrand and Eastman Kodak were to manufacture the Y-7. This was the "ultimate" vertical periscope for the "K-Series" system, and was intended for installations in the B-52, the B-36, and the "large nose" B-45, In January 1952 Farrand contracted to design and construct two prototype Y-7's. One month earlier, Farrand had delivered the first model of the Y-5 vertical scope to Boeing's Seattle plant for installation in the XB-52 aircraft.52
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
The K-1 System in the B-36
As of 15 May 1951, the Armament Laboratory felt that bombing-navigation system development had reached a "plateau" where system accuracy was being limited by factors other than mere instrumental precision. Of these factors, the inaccuracy of heading and speed inputs and uncertainties in target tracking because of poor radar resolution were surpassed only by poor reliability primarily arising from premature production. In Strategic Air Command missions, the B-36 abort rate due to K-system malfunctions was from 25 to 30 percent and the B-47 rate was later to approach 54 percent!
Within a month, however, definite improvement in radar reliability was evident. Early that month, 20 B-36 aircraft on a maximum effort mission over Birmingham, Alabama, registered 700 to 900-foot average errors; none aborted. Nine days later, on 16 June, the Armament Laboratory concluded that the B-36 abort rate due to K-system malfunctions was steadily decreasing, and that radar bombing accuracy was steadily increasing as aircrews and maintenance men gained familiarity with the equipment. The average errors were less than 1,500 feet from 40,000-foot bombing altitudes, the larger portion of the bombs falling within 1,000 feet (25 mils) of the target. And these results were recorded by production K-systems—not by the improved "Reliable" configuration.56
K-systems were now installed in 43 B-36 aircraft, eight B-47's, and 11 B-50's for training purposes. At this point, the K-3A system was scheduled to go into B-36's, and the K-4A into B-47's. These configurations would represent some 76 modifications to 26 different components since the initial delivery of production units. Project Reliable was responsible for 18 of the 76 changes, and future changes to increase altitude, air speed, and tracking ranges would probably be necessary. The "Reliable" K-4A system (when its final configuration was established) was scheduled to be installed in B-47B number 31.57
In August 1951, the Armament Laboratory eased the Doppler requirement by deciding that the APN-81 would not have to provide displays showing the direction and magnitude of wind forces and that semi-automatic navigation would be satisfactory. The semi-automatic feature meant that the bombardier could, by taking radar position fixes whenever possible, reset the "present position" indicator of the Doppler set and thus correct any navigational errors before they reached serious magnitudes.
A radome for the Doppler radar was delivered to Wright Field on 31 August for installation in a flight test B-29. About a month later, International Business Machines delivered electrical schematic drawings covering the AN/APN-81 tie-in to the K-system, On 10 January 1952, the first "debugging" flight of the B-29 Doppler test aircraft took place without major difficulties. By the end of March, International Business Machines was modifying A-1 computers to accept APN-81 information.58
A new complication appeared early in 1953. The K-systems had originally been modified to accept inputs from the initial development model (SA-1) of the APN-81. The production version of the APN-81 (the XA-2) was incompatible with the production K-system. In March 1953, however, IBM successfully modified a K-system computer to be compatible with the XA-2 model of the APN-81. This happened several months prior to the delivery by General Precision Laboratories of the XA-2 Doppler set itself.59
During a bombing system phasing group meeting on 1 May 1953, Wright Air Development Center recommended that the APN-81 and tie-in components be integrated with the K-3A system which was slated to be installed in the early production model B-52's. By this time, the bulk of APN-81 components had become great enough to warrant a completely new mock-up of the K-3A system in the YB-52 aircraft. At the conference, the Strategic Air Command tried to keep the Doppler radar out of the B-47 project, but was overruled by Air Force headquarters.
In May, the director of requirements in the Pentagon also ruled that the modified B-47's emerging from Project Reliable would be equipped with a K-system representing the standard installation scheduled for future production model B-47E aircraft. However, the prototype aircraft was to use an APS-23 fixed frequency radar set rather than its more recent successor, the low power, tunable APS-64, not yet available. The Reliable aircraft's bombing-navigation system would be capable of tracking at 50 nautical miles range, have a 10-inch radar scope, contain equipment for indirect bomb damage assessment, and have an attachment for presetting of data on two different offset aiming points near a target. (These features would also be necessary in the B-52).60
A mock-up inspection of the Western Electric 10-inch indicator for the APS-23 radar was held on 23 July 1953 at the Boeing Plant in Wichita but revealed that the scope was unsatisfactory for either the B-47 or the B-52, Principal difficulty was that the equipment was unpressurized and thus limited to operation below 25,000 feet altitude. Air Force headquarters decided that unpressurized 10-inch scopes without integral camera provisions would not be used in B-47 or B-52 aircraft and in the B-36 would be used for training only.
The Pentagon directed development of a Motorola pressurized 10-inch indicator with an integral camera. Until it became available, the B-47 and B-52 would continue to use the 5-inch scope.61
As 1953 ended, production K-systems (K-4A) installed in B-47 aircraft were beginning to exhibit gratifyingly increased reliability and accuracy due to relocation of components within the airframe, modification and improvement of critical parts of the system, better shock mounting, and the combination of simplified and flexible maintenance procedures with better test equipment. Not the least important factor in increasing K-system accuracy was the growing experience of Strategic Air Command combat crews with the operation of the equipment. Within a year, most K-system deficiencies had been eliminated in new production aircraft and rapid retrofitting of B-47's in the field was well under way; a satisfactory pressurized 10-inch Motorola radar scope, the most irksome deficiency remaining, was scheduled for production by January 1955.62
Wright Field received two experimental models of a Kollsman KS-18 Automatic Astro-Compass in October 1953. The first was installed in a B-29 aircraft and flight tested briefly and successfully in Dayton, then flown to the Boeing plant at Wichita where the equipment was transferred to a test B-47 for evaluation in conjunction with a K-4a bombing-navigation system. The second model was also installed in a B-29 at Wright Field, where it was tested in conjunction with an N-1 magnetic compass and aircraft autopilot.63
In November 1953, Wright Air Development Center recommended installing Y-7 periscopes in the B-52's K-3A system in lieu of the less advanced Y-5. Farrand finally completed development of the Y-7 in February 1955; the periscope met all specifications, was standardized, and went into production. By January 1956, Eastman Kodak had completed 125 Y-7's. 64
During the first six months of 1954, flight tests of the D-3A bombing-navigation system in the YB-52 were underway. The Strategic Air Command insisted that the suitability of the K-3A for the Stratofortress be established since the ultimate B-52 system, the IBM developed ASB-4 (also MA-2), would not be in production in time for the first operational B-52 squadrons. In the fall of 1954, the improved K-3 (MA-6A) incorporating tunable radar (APS-64) began flight tests in a B-47 test bed.
From the beginning, MA-6A tests demonstrated the incompatibility of the APS-64 radar with the experimental Motorola 10-inch radar scope. The radar presentation proved unsuitable for bombing and the reliability of the combined equipment was very low. This situation was especially critical since the APS-23 was going out of production in April 1955 and by that time only enough sets to support the B-47 or the B-52 program would be in existence—but not both. Accordingly, the Air Research and Development Command and the Air Materiel Command jointly decided late in 1954 to install the APS-23 in production B-52's until the APS-64 passed suitability tests. The B-47 remained committed to the tunable AFS-64. In this configuration, the K-4A emerged as the MA-7A bombing-navigation system.
The MA-7A was in production by November 1955 and development of the B-47 bombing and navigation system was essentially complete. Production continued through 1956, and in 1957 the B-47B and B-47E aircraft in the field were being retrofitted with the MA-7A. Bombing and navigation system development for the B-52, however, was far from complete. The MA-6A was still being considered for the Doppler radar and the astro-compass and beyond that the ASB-4 was on the books.65
Wright Air Development Center had delivered a B-47 with Doppler-augmented K-system to the Air Force Armament Center, in July 1955, for tests aimed principally at the MA-6A. The test installation also contained provisions for automatic crosshair-laying, semi-automatic fix taking, and dead reckoning navigation. The first flight at Eglin on 22 July 1955 revealed that the Doppler set caused "Wander" in the wind values calculated by the D-system, a condition which made the equipment completely unsatisfactory for bombing. General Precision Laboratories went to work on the APN-81, and Sperry on the computers already modified by International Business Machines, in an attempt to resolve the difficulties. By July 1956, the Doppler tie-in was at last working satisfactorily during test flights.
Early in 1957, the Air Force received the first two production prototype computers (ME-5) modified to accept Doppler information and to operate with the tunable search radar. Flight tests were underway by the turn of the year, and continued into the fall of 1958. By that time both the astro-compass and the Doppler radar had proven they could operate satisfactorily with the new computer, but a scarcity of funds forced suspension of the program.
With the possibility of adding the new computer, the astro-compass, and the Doppler radar to the MA-6A already demonstrated, the next phase of development was charted. However, this phase had not begun by the end of 1959 and all B-52B through B-52D aircraft in SAC were appointed with a standard MA-6A. The B-52E and B-52F models delivered through the middle of 1960 were employing the ASB-4;, while the B-52G's delivered by this time had the ASB-9 bombing and navigation system, a version of the ASB-4 without optical features. Finally, the B-52H's to be produced were to contain the ASB-9A, essentially the ASB-9 with the terrain clearance features required for low altitude operations. This project was under way by mid-1960. The Air Force had revived the computer, astro-compass, and Doppler program for the MA-6A at this juncture, and was in the midst of preparing these units for incorporation with extant models of the B-52C and D, along with terrain clearance devices.66
The real shortcoming of the K-Series equipment and its siblings, the MA-6A and the MA-7A, lay in their effect on the responsiveness of the Strategic Air Command. Irksome maintenance difficulties, the need for elongated periods of pre-flight calibration, and the frequency with which airborne adjustments were required imposed delays on the readiness of the strategic fleet. In an age of limited air defense warning and the undoubtedly high target rating of Strategic Air Command bases, the mobility and responsiveness of the bomber fleet, still the prime national weapon, was critical.
Yet it seemed probable that an emergency would find the Strategic Air Command bombers ready for combat. Early reliability problems had been considerably lessened in their effect by 1957, maintenance was no longer the overwhelming obstacle it once had been, and familiarity with the equipment had improved the performance of flight crews.
From oven a thoroughly pessimistic outlook, it would have to be conceded that the work begun in 1944 had, by 1957, created the ability to bomb under all weather conditions from altitudes above 50,000 feet and at speeds in excess of 650 knots. This was no small achievement. But bombing at supersonic speeds and above 60,000 feet posed new, and still more difficult, problems. Furthermore, the increasing effectiveness of countermeasures to radar bombing techniques, and the inherent disadvantages of radar navigation, were rapidly relegating radar-oriented bombing-navigation systems to obsolescence. Since optical bombing techniques and visual-reference navigation had been proven inadequate before the end of World War II, the problem posed was to find yet a third basic medium for the control of strategic air warfare.
CHAPTER IV THE "BOMB DIRECTOR FOR HIGH SPEED AIRCRAFT"
The answer to problems of complexity and requirements for high performance was initially conceived under the "Bomb Director for High Speed Aircraft" project. Here was to be a system capable of directing a bomber to a prescribed location with relatively little reliance on radar navigation, and then bombing from over 60,000 feet during supersonic flight. More than that, the high speed bomb director was to be simple in design and easily and quickly prepared for use. This was indeed a handsome goal.
At a July 1947 conference called by the Deputy Chief of Air Staff for Research and Development, Army Air Forces guidelines for the development of the "High Speed Bomb Director" (HSBD) or the "Bomb Director for High Speed Aircraft" (BDHSA) were created. The Pentagon wanted an electronic bombsight for long range strategic bombers and also desired the concurrent development of "auxiliary visual methods." The primary requisite for the radar portion would be excellence of target definition during the bombing phase. Navigation features were not to jeopardize this in any manner. The final system was required to have a "consistent bombing accuracy under combat conditions" of 25 mils or better.67
Even earlier, on 22 May 1947, the Engineering Division had approved a project aimed at a reliable, lightweight bombing-navigation system containing optical and radar features (with possible inertial or Doppler tie-ins) for use in very high speed bombardment aircraft. The Armament Laboratory advocated using hermetically sealed computer units and "modular" units in the radar display section of the system. These "beer can" plug-in components could be quickly and easily interchanged in flight or on the ground without the necessity of calibrating the system every time a unit was replaced. The computer was to be an electronic analogue type, rather than a mechanical unit.68
Perkin-Elmer contracted in 1948 to develop all but the radar portion of a high speed bomb director system having mechanical simplicity without operational complexity. The Armament Laboratory decided that an entirely new radar development would be a necessary key to the project since the APS-23 was a product of already obsolescent technology.69
In June 1950, Raytheon Manufacturing Company undertook to develop the new radar. The Armament Laboratory specified that the set should be designed for high reliability and constructed with a maximum number of hermetically sealed, plug-in units similar to those used in the computer and interconnection components. This was to be a high-power, tunable unit employing advanced techniques to obtain the highest possible resolution.70
More specific requirements for the high speed bomb director resulted from a conference held in the Pentagon on 24 and 25 August 1950. Here, the priority of development was raised to 1-A (from 2-C) and the completion of operational suitability tests set for January 1952, a considerable acceleration of the original program. First production items were due in January 1953.
The military characteristics proposed during the conference were issued on 25 September 1950 as the requirements for a "Precision Bombing System" and included an operational altitude of 60,000 feet at a speed of 650 knots for the first units. The desired ultimate goal was 125,000 feet and 1,250 knots. The incorporation of a star-tracking device for precise heading information was desired as soon as possible. The high speed bomb director was to have an "inertial extrapolator"—like that designed for the K-system—that would allow offset bombing from points as far as 100,000 feet from the target and navigation to the target or offset point from 300 nautical miles out. Eventually, inertial navigation from 1,000 nautical miles out was to be possible.71
In late 1950, the laboratory decided that Perkin-Elmer was too small for large scale production and asked International Business Machines to take over the entire system, including the radar. The Perkin-Elmer contract was continued, but only for fabrication of one experimental model of the bomb director (Model XY-1) to be delivered to Wright Field on or about 1 October 1951. Raytheon had been carrying on radar work more or leas independently of Perkin-Elmer, but was now to come into the revamped project as a subcontractor to International Business Machines. The prime contractor was to be responsible for integrating the radar, optics, and bombing-navigation computer with data sources such as Doppler radar and heading reference systems, and to add such attachments as an automatic ballistics computer, a "topographical comparator" (for matching a map with the radar display) and advanced navigational components.72
This decision was followed by an Air Materiel Command production contract with International Business Machines in January 1951; the Perkin-Elmer contract was amended to call for certain additional design features in the experimental model. The XY-1 model was now to be delivered to the Air Force in February 1952.73
However, on 29 August 1951, the Pentagon published a revised set of military characteristics for a "Bombing System for High Altitude Supersonic Bomber," which, while labeled "proposed," provided the guidelines for the project. The system was now to be developed in two steps: first, equipment that would solve the bombing problem at altitudes up to 60,000 feet; second, a device able to operate above 120,000-foot altitudes. True airspeed limits for the former were to be Mach 1.0, and for the latter Mach 2.2. The requirement for near automatic navigation, evasive maneuvers, and simplified presentations remained the same, but the new characteristics called for automatic ballistic inputs up to the instant of bomb release (instead of 10 seconds prior to it), a provision for adjusting the aircraft position counters at any time, and the ability to guide air-to-surface missiles at ranges up to 150 nautical miles.74
On 8 December 1951, the Air Force announced that the high speed bomb director was to be ready for operational use in early 1957. (The bomber that became the B-58 was scheduled to be ready at that time.) Almost simultaneously, the Armament Laboratory directed International Business Machines to allow manual settings of trail and time of fall in the bombing computer and to leave development of the automatic ballistics interpolator to Perkin-Elmer.75
Perkin-Elmer delivered to Wright Field on 6 March 1952 a bomb director with attachments for navigation, offset bombing, and automatic ballistics computations. It was installed in a B-29 by the first week in June 1952 and satisfactorily passed through a series of shake down flights (tied in with an APS-23 search radar) in the vicinity of Dayton. By the late summer of 1952, the bomb director had been designated MA-2 and the first prototype had been scheduled for delivery during February 1953. This system was to be installed in a B-47B for flight testing at the Air Force Armament Center. Before the year was over, however, the delivery date of the second International Business Machines' prototype, the XY-2, had slipped to April 1953.76
In October 1952, ARDC programmed the MA-2 for the B-52. At the same time, Baltimore directed a feasibility study to see whether the MA-2 might not serve as a satisfactory "interim" system for the B-58 Hustler, in case the Sperry system planned for that aircraft did not meet production schedules. International Business Machines agreed to perform the study in conjunction with the airplane manufacturer (Consolidated-Vultee).77
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
By the end of 1952, IBM had expanded the "modular" construction of the MA-2 far beyond the original design philosophy. Not only were there to be as many interchangeable plug-in units as possible within the various system components, but the components themselves were to be "building blocks" which night or might not be installed, depending on aircraft configuration or on the mission the system would be called upon to perform. The versatility of the system could be enhanced by adding one or more "attachments," which included an offset bombing computer, a navigation computer, an automatic ballistics computer, a Doppler radar set and antenna (APN-81), an "inertial extrapolator," autopilot tie-in equipment, and a radar map-matcher. The astro-compass was added to the lot later.78
At this juncture, the B-29 housing Perkin-Elmer's MA-2 (XY-1) system was undergoing flight tests at the Air Force Armament Center. When these missions, including bomb drops, were completed, the aircraft was to be returned to Wright Field for installation of the navigation attachment and a 10-inch radar scope. Modification of the B-47 to accept the IBM MA-2 (XY-2) system was delayed by lack of installation information, forcing retardation of flight tests.79
In April 1953, the first Raytheon experimental tunable radar was accepted and shipped to Wright Field, where it was installed in a T-29 for airborne evaluation. The tests continued through January 1954 when a second Raytheon set replaced the first model. Flight tests of the second model continued through July 1954.
International Business Machines finally shipped to Wright Field the XY-2 model of its MA-2 system on 6 November 1953 and the unit was installed in the B-47 which had been ready since mid-September. Installation was completed (with an APS-23 radar) and flights began in April 1954, continuing until October of the same year. Raytheon delivered a third high speed bombing radar to Wright Field in August 1954, and it promptly replaced the APS-23 in the B-47. At a mock-up inspection of the MA-2 system in the B-52, held in April 1954, the Strategic Air Command agreed to the details of the installation. The Air Force still planned to install a modified K-3A bombing-navigation system in the B-52 until the MA-2 was ready.80
In July 1954, WADC completed a study of the MA-2 as an insurance bombing-navigation system for the B-58 aircraft. The B-58 required a combination of Doppler radar, inertial and stellar observation components for navigation; as well as a highly accurate vertical reference; high altitude radar bombing and bomb-pod-launching equipment; automatic flight control and stabilization equipment; and a guidance and control system for the bomb pod. Wright Field's contention that the MA-2 could be modified to perform these functions found support at Air Research and Development Command headquarters and in Washington. The Sperry system, however, progressed satisfactorily after early difficulties, and the adoption of a modified a MA-2 for the Hustler never became necessary.81
[The decision to employ the MA-2 solely in the B-52 negated the requirement for a system operating at 125,000 feet and 1,250 knots; the B-52 was a "50,000-foot plus" and 650 knot aircraft.]
B-47 tests of the MA-2 (XY-2) ended successfully in April 1955. The first MA-2 (XY-3) production prototype was delivered to Boeing's Seattle facility the following month where it was installed in B-52 aircraft number 008 for accelerated flight testing designed to qualify the system quickly as a replacement for the K-3A and MA-6A equipment already in production B-52's. At this point, the XA-2 was scheduled to go into the 118th production B-52. However, by January 1956, B-52 number 177 at Seattle, and number 52 at Wichita appeared to be the earliest aircraft in which production installation could take place and that not until October 1957.82
The MA-2 (XY-3) operated very much as did the "K-Series" systems. When departing on a mission, the operator switched the navigation computer to the automatic-fix mode and positioned the crosshairs on a nearby landmark whose coordinates had been temporarily set into the computer as a "destination." The navigation computer then established the starting point of the mission and set the coordinates of the first check point (about 600 nautical miles away) into the device. Based on the geometric relationship of the "present position" and destination, the computer determined the ground track to be flown and steered the aircraft through the autopilot. As many 600-mile legs as necessary could be flown by repeating the same procedure. When the aircraft was 50 nautical miles from its final destination, the operator switched to the bombing mode.
During the bomb run, he could use the optical system, if visibility conditions permitted; the bad-weather alternate, the high speed bombing radar, gave either a plan position indicator presentation (360 degrees), a sector scan, or an off-center sweep on a 10-inch scope. By making a "memory point run," the operator could find wind direction and velocity and determine the correct bomb run course for the autopilot to fly. In the XY-3 system, the operator set in ballistic data manually. If the crosshairs drifted off the aimpoint during the final run in, the operator performed a final memory point operation to adjust them. The bombing computer then proceeded to vary the components of the bombing problem with changes in wind, altitude, course, and airspeed and automatically released the bombs at the correct instant to score hits on the target.
By 1 September 1955, research and development on the basic MA-2 bombing-navigation system was essentially complete, although the B-52 with the XY-3 model was still in flight test. On 2 September, Wright Air Development Center requested four service-test MA-2 systems, the last of which was to be delivered by November 1957. The four systems were to be integrated with inertial extrapolators and to have the ability to launch and guide air-to-surface missiles from the B-52. One of the four systems was to be used for operational suitability testing after completion of service tests in July 1958.
The XY-3 system installed in the B-52 had completed 200 hours of flight testing by the end of October 1955. The reliability of the bombing navigation equipment compared favorably with that of the APQ-24 and "K-Series" systems then in operational service; however. the tests resulted in recommendations for some 52 design changes. Once the 19 most important of these had been incorporated, the MA-2 (XY-3) was to be considered satisfactory for installation in operational B-52 aircraft.85
The B-52 at Eglin resumed flight testing on 15 December 1955 and was to continue until the end of February 1956, when Phase II instrumentation (for Doppler radar and astro-compass tie-in) would be installed for functional testing scheduled to begin in August 1956.
In November 1955, Wright Air Development Center began installing an XY-3 system in a B-50 aircraft in order to evaluate such advanced features as an automatic astro-compass and an inertial extrapolator designed to provide 200 nautical miles of completely passive navigation.86
In January 1956, the Weapons Guidance Laboratory (successor to the Armament Laboratory, 1 July 1955) announced a slight slippage in the MA-2 installation program; the system was now to be installed in B-52 number 192 at Seattle and number 62 at Wichita. Aircraft delivered prior to these would receive MA-6A systems (K-3A's modified to have extended tracking ranges and low power tunable search radar equipment). One month later the Air Force decided that the 49th MA-2 system installed in the B-52 (incorporating some 21 component changes over the basic MA-2) would be designated AN/ASB-4. Previous systems were to be MA-2 (XY-3) production prototype models for use as air and ground test systems and in aircrew training.
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
An MA-2 Radar Scope Photo of Long Island
The MA-2 test program was now moving along rapidly. The Phase I functional evaluation of the XY-3 in the B-52 ended, on 9 February, on schedule, whereupon the system was removed and shipped to International Business Machines for modification to the ASB-4 configuration and returned to Seattle by the end of June 1956. On 1 March, aircraft 008 was in the shops at at Seattle; here the APN-89 Doppler radar (APN-81 modified for use with the MA-2) and a Kollsman MD-1 automatic astro-compass were also to be installed.87
The armament center successfully completed operational suitability testing of the MA-2 (XY-3) installed in a B-47E and 15 MA-2 (XY-3) systems had been installed in B-52E's for squadron flight testing by October 1956,
International Business Machines had delivered two prototype ASB-4's to Seattle on 30 June 1956. One was immediately installed in the B-52 (already equipped with Doppler radar and the astro-compass, plus tie-in devices). The AN/ASB-4 plus Doppler, astro-compass, the AN/AJA-1 True Heading Computer, and the O-32 radar recording camera now comprised the AN/ASQ-38 Offensive Weapon Control System. Engineering flight tests by the contractor began in August and were to continue at Eglin and Seattle under armament center auspices until December 1957. However, as December ended, the project was still to be completed. The delay was primarily the product of a late decision to change the test aircraft from a B-52B to a B-52G; the new configuration was to include modifications to the MD-1 automatic astro-compass tie-in equipment, a series of minor alterations to give the operator the ability to lock the aircraft's airspeed and altitude just prior to bomb release, and the deletion of the entire optical portion of the system. The available data showed gratifying performance for the ASB-4, with reliability improving on each flight. The APN-89 and the MD-1 were also showing considerable promise.88
During 1956, International Business Machines was fabricating four experimental inertial extrapolators (designated MA-1) for delivery In mid-1957. Two of these were to be flight tested in B-52 aircraft and two would go to Wright Field, one for installation in the test B-50 and one to back up the flight modele All had been delivered by December 1957.
[In the midst of this effort, MA-1 installation was changed to include one in a B-47, one in a T-29, and another for bench test at the contractor's plant. A fourth set had to be used in the program already established to develop a bombing-navigation system for Weapon System 110A, a chemically powered, long range, supersonic bomber.]
Raytheon and IBM were, by this time, also studying ways to increase the anti-jamming ability of the high speed bombing radar, and hoped to have a flyable test model sometime in 1958. Production delivery of APN-89 Doppler radar sets by General Precision Laboratories began in March 1957. On 13 December, the first production B-52E, with an ASB-4; system, was delivered to the Strategic Air Command; 249 aircraft were already in the field with MA-6a equipment.89
Thus, by the end of 1957 the B-52 was receiving bombing-navigation equipment which could out perform the MA-6A in many respects and had promise of more significant advantages in the future. The new system was designed for higher reliability and easier maintenance through use of modular construction. Good reliability had been demonstrated in the flight test program, but extensive operational use would be the true reliability test. Furthermore, the solution to the problem of providing the strategic command with a system which required no elaborate preflight preparation had not been found.
By the end of 1957, the MB-1 had emerged as the major problem in the ASQ-38 endeavor. Flight test failures had caused elimination of the set from production installation for the first two complete wings of ASB-4, equipped B-52's. Plans still called, however, for adding the MD-1 to these aircraft by retrofit and installation in production aircraft beginning in February 1958.90 Although the ASB-4 systems being delivered to SAC were generally performing well, some "objectionable characteristics" had appeared. While it was true these did not prevent "effective utilization of the system:" the Air Force was concerned and foresaw requirements for greater reliability and a number of "engineering refinements. . . ."
A strike at the Kollsman plant, lasting from 20 January 1958 through 17 February 1958, halted all work on the MD-1. Work resumed immediately after settlement and the unit was scheduled for installation in production versions of the last 14 B-52F and all the B-52G aircraft. Initial models began arriving at the Boeing Wichita plant in April, but Boeing soon complained of a very high rejection rate which threatened to "seriously affect" the production program for the remaining B-52F's. On 21 April a group representing the project office, WADC, IBM, and Kollsman met at the Wichita plant to "isolate the causes" of rejection and soon agreed that rough handling might be the culprit. Personnel from WADC and the Air Force Plant Representative's office at the factory combined to follow a subsequent group of six sets from Kollsman to Boeing, making sure that each item was handled with great care; factory bench tests then showed the problem had been solved.91
But flight tests in July 1958 also displayed erratic MD-1 behavior. Some of the fault could be traced to the Time Standard Amplifier, the device which drove the astro clock and served as a "primary reference" for "running all time sensitive circuits. ..." Temperature change effects and the incompatibility of some components were also involved.
Kollsman made some fast adjustments, but was not able to demonstrate complete resolution of MD-1 failures. The Air Force decided that if subsequent tests showed the clock mechanism did not lose more than six seconds per hour the set could be used even though it meant more work for the operator. Test results showed the MD-1 to be within this stricture and the unit began to be installed in B-52F's at the Boeing plant in Wichita in September 1958.92
The first complete model of the ASQ-38 was delivered to SAC before the end of 1958 and included an improved radar capable of longer search ranges. However, the first six months of 1959 once again revealed low reliability and a concomitant reduction in acceptance rate at Boeing. The project office, declaring the need for "immediate corrective action," appealed to the contractor group to do everything possible to improve quality control and instituted a "component rejection program" at the Boeing plant in Wichita.
Designed to "evaluate, analyze, and take action to reduce the number of rejects," the program was supported by special committees covering the main components of the systems. Before the end of May, committees on the radar and the MD-1 had been established, and soon every major element was included. Results were gratifying. By the end of 1959 the ascending rejection rate had been arrested in many cases, and in some instances a noteworthy downward trend had ensued.93
The need to provide the B-52 with the ability to operate at low altitudes was the dominant element in the bombing and navigation program throughout most of 1959. The Strategic Air Command announced the requirement for flight at 500 feet or less in June 1958, but the ASQ-38 was not designed for employment at these altitudes. It needed, above all, terrain clearance equipment, modifications to the radar "to reduce picture blooming at low altitudes," and higher precision in the Doppler radar.
Planning for the changes required in the weapons control system consumed the greater part of the following year, until, in July 1959, SAC stressed the urgency of the requirement and a presentation to Air Force headquarters was scheduled. SAC wanted to buy an integrated terrain clearance unit before extensive flight testing, but the project office and WADG hastened to add that "some degree of technical risk" was involved in such a procedure.
The Weapons Board in Washington reviewed the program on 4 August 1959 and agreed to the SAC plan. The B-52G was now in production with the ASB-9 system, essentially the ASB-4 without optics.
The B-52H was to be the initial recipient of the equipment for low level operation while in production. The B-52G was to be appointed by retrofit. These bombing and navigation systems were entitled ASB-9A, a part of the AN/aSQ-38(V) Offensive Weapons Control System, B-52E and F aircraft containing the ASB-4 would have the AN/ASQ-38(U) designation. Proposals for retrofitting the B-52C and the B-52D were also under consideration.
A number of points remained to be settled at the end of 1959. For one, ARDC wanted permission to conduct flight tests before laying definite plans for incorporation of the terrain clearance unit and attendant devices by retrofit in the B-52C and the B-52D. Of greater consequence was the Strategic Air Command's stand against terrain clearance installation in any B-52 before it was established that the structure of the aircraft could support such a load and that the equipment did not degrade the high altitude flight capacity of the aircraft. SAC was also adamant in insisting that the B-52 fleet would actually be able to operate at low altitudes when the prescribed units wore installed.95
Thus, by 1960 the B-52 series had been appointed with a variety of bombing and navigation systems. Development of each category had been completed, except for the terrain clearance features (and associated items) required for low level flight. All aircraft were eventually to gain this capacity and the older systems—MA-6A—were to be improved so that their performance was not too far from that of the newer units.
The B-52 had cone a long way since the first K-system was installed. The road included a modification of the K-system, (MA-6A) and entirely new system (MA-2) and modifications and improvements (ASB-4, ASB-9, ASB-9A) to that device. The key problem, aside from the difficulty encountered in developing new and advanced bombing and navigation equipment, had been low reliability. Massive corrective efforts had been instituted and by 1960 they had generally succeeded in reducing the problem to manageable proportions.
1. ConAC Manual 50-11. "Air Force Armament" (hereafter cited as ConAC Manual 50-11), May 1949. III, 3-4, in Hist. Br., WADD files (hereafter cited as HBF); rpt., Historical Development of Armament Equipment, 1943 (hereafter cited as Hist. Dev. of Arm. Equip,, 1943), prep. by Arm. Lab., Eng. Div., AMC; rpt. Brief History of Bombsight Development by the Army (hereafter cited as Bombsight Hist.), 22 Oct. 1943, prep, by Arm. Lab., Eng. Div., AMC, in AF Records Center, St. Louis, Mo., files; J. E. Clemens and B. B. Johnstone, ed., Introduction to the Theory and Practice of Bombing (hereafter cited as Theory and Practice of Bombing), July 1955, published by Raytheon Manufacturing Co. (hereafter cited as Raytheon) for Aero. Res. Lab., WADC XI-3, XI-4, XI-5; L. A. Dubridge, E. M. Purcell, G. A. Morton and G. E. Valley, Radar and Communications. May 1946, prep, for the AAF Scientific Advisory Bd.. p.76; Keith Ayling, Bombardment Aviation, (Harrisburg, 1944), in HBF.
2. ConAC Manual 50-11, I, 1-3; Clemens and Johnstone, Theory and Practice of Bombing. VI-47, XI-4, XI-6, XI-9, XI-10, XI-13, XI-16; Hist. Dev. of Arm. Equip., 1943; Bombsight Hist., 22 Oct. 1943.
3. Clemens and Johnstone, Theory and Practice of Bombing, II-5.
4. Ibid., XI-15, XI-17; Bombsight Hist., 22 Oct. 1943; rpt., The Sperry Bombsight, A History of Its Development, about July 1931, prep, by Sperry Gyroscope Co., (hereafter cited as Sperry Hist.), in HBF.
5. Bombsight Hist., 22 Oct. 1943; Sperry Hist; Clemens and Johnstone, Theory and Practice of Bombing, XI-11, XI-16-XI-18; Hist, Dev. of Arm. Equip., 1943; Bombsight Data Sheet, 1943, D-8 Bombsight, in HBF.
6. Bombsight Hist., 22 Oct. 1943; W. F. Craven and J. L. Cate, Ed., The Army Air Forces in World War II. I, 598, in HBF; Clemens and Johnstone, Theory and Practice of Bombing. XI-19.
7. Bombsight Hist., 22 Oct. 1943; Claraens and Johnstone, Theory and Practice of Bombing, XI-19, XI-20; Craven and Cate, The Army Air Forces in World War II, I, 598; Hist. Dev. of Arm. Equip., 1943.
8. Rpt., The Bombing Problem (hereafter cited as the Bombing Problem), 28 June 1947, prep, by Bombing Br., Arm. Lab., Eng. Div., AMC, in WGL Library files.
9. Bombsight Data Sheet, K-1 to M-9 Bombsights; Eng. Div. R&D Projects Rpt. (hereafter cited as R&D Rpt.) Jan. and July 1943, in HBF.
10. Clements and Johnstone, Theory and Practice of Bombing, IV-3, IV-5, IV-7, IV-9, IV-11; ConAC Manual 50-11, 111, 3-16.
11. R&D Rpt. Jan. 1945 and July 1946; ltr., Col. R. E. Jarmon, Ch,, Arm. Lab., Propulsion and Accessories Subdiv., Eng. Div., ATSC, to Hq. ATSC in Europe, 28 Apr. 1945, subj.: Current Status of Bombing Projects, in HBF.
12. Bombsight Data Sheet, 1943, T-1 Bombsight (British Mark XIV); Clemens and Johnstone, Theory and Practice of Bombing, XI-21 to XI-23; Hist. Dev. of Arm. Equip.. 1943; R&D Rpt. July 1942, Jan. and July 1943, Jan. and July 1944, and July 1945.
13. Clemens and Johnstone, Theory and Practice of Bombing, XI-22, XI-23; ltr., Col. F. O. Carroll, Ch., Exp. Eng. Sect., Mat. Center, to Asst.. C/S. Mat. Cmd., 9 Oct. 1942, subj: Precision Bombing at Night - Development Projects, in HBF.
14. R&D Rpt. July 1945; Bombsight Hist., 22 Oct. 1943; ATSC TI-2028, Add. No. 23, 5 Jan. 1945, B-29 Installation of Eagle for the 315th Wing; ConAC Manual 50-11, 111, 6-9 and 6-10, in HBF.
15. R&D Rpt. July 1945; Clemens and Johnstone, Theory and Practice of Bombing, XI-24 and XT-25; ltr., Jarmon to Hq. ATSC in Europe, 23 Apr. 1945.
16. Inter-office memo, Brig. Gen. F. 0. Carroll, Ch., Eng. Div., to Ch., Arm. Lab., Eng. Div., Mat. Cmd., 1 May 1944, sub: Bombsight Development) ltr. (1st Ind.) Col. F. C. Wolfe, Ch., Arm. Lab., to Ch., Eng. Div,, Mat. Cmd., 21 June 1944, subj: Bombsight Development; Dubridge, et al., Radar and Communications, 16-18, 76, 81-83, 85-88; ltr., Gen. M.S. Fairchild, Vice C/S, USAF to CG, AMC, 12 July 1948, subj.: Precision Bombing Standards, in HBF.
17. R&D Rpt. July 1945 and July 1946.
18. Rpt., AN/APQ-23A Radar Bombing System, Sep. 1948, prep, by Arm. Lab., Eng. Div., AMC, in WGL Library files; ConAC Manual 50-11, 111, 6-25.
19. R&D Rpt. July 1945 and July 1946.
20. Eng. Div. Weekly Information Rpt. (hereafter cited as Eng. Div. WIR), 23 Aug., 27 Sep., and 4 Oct. 1946, in HBF.
21. Eng. Div. WIR 12 Feb., 7 Mar., 28 Mar., 7 Apr., 11 Apr., 2 May, 23 May, 6 June, 13 June, 29 Aug., 10 Oct., and 5 Dec. 1947, and 6 Feb. and 1 July 1948.
22. R&D Rpt. July 1946; Eng. Div. WIR 27 Sep. 1946, 5 Feb., 2 May, and 29 May 1947; Clemens and Johnstone, Theory and Practice of Bombing, IV-47; Quarterly Rpt., AAF Tech. Comm., Elect. Subcomm,. meeting of 29 Apr. 1947, E.O. 103-17, Radar Set AN/APQ-24; Periodical Rpt., AAF Tech. Comm., Arm. Subcomm., meeting of 30 July 1947, E.O. 103-17, Radar Bombing and Navigation System AN/APQ-24, in HBF; rpt., AN/APQ-24 System, Sep. 1948, prep, by Arm. Lab., Eng. Div., AMC, in WGL Library files
23. Eng. Div. WIR 26 Dec. 1947 and 17 June 1948.
24. Eng. Div. Monthly Proj. Rpt. (hereafter cited as Eng. Div. MPR), Mar., Apr., May, Aug., Oct., Nov. 19Ù9, Apr. 1950, and Jan. 1951, in HBF; Transcript (edited), subj,: Conference in Electronic Bombing Systems at Hq. SAC, 2 Feb. 1951, in WGL Library files.
25. SAC conference transcript, 2 Feb. 1951.
26. Ibid.. ADF Monthly Proj. Rpt, (hereafter cited as ADF MPR) May 1951; WADC R&D Proj. Information Rpt. (hereafter cited as WADC R&D PIR), Aug. and Dec. 1951; interview, Maj. W. F. Kroemmelbein, K-Sys. Sect., Strat. Bombing Br., WGL, by J. King, Hist. Br., WADC, 9 Aug. 1957.
27. WADC R&D FIR Dec. 1951; presn., Strategic Bombing Operations, by R. J. Nordlund, Ara. Lab., ADF, .to Fire Control Panel, Ord. Comm., RDB, 15 May 195l (hereafter cited as Nordlund presn.); History of Wright Air Development Center. 1 July-3I December 1952. II. 383, In HBF.
28. History of Wright Air Development Center, 1 July - 31 December 1952. II, 384; AF Supplementary Progress Rpt. Card (hereafter cited as ARDC 82), Proj. S-556-360, 20 Jan. 1953; RDB Proj. Card (hereafter cited as ID 613), Proj. 5041, 9 Mar. 1954; ARDC R&D Proj. Plan (hereafter cited as ARDC 100), Proj. 5l59, 2 Oct. 1954; and ARDC R&D Mgmt. Rpt. (hereafter cited as ARDC 111), Proj. 5041, 25 Mar. 1955, in HBF.
29. Presn., The K-1 Optical-Radar Navigational-Bombing System, by Arm. Lab., 1 Mar. 1951 (hereafter cited as K-1 presn.); inter-office memo., Carroll to Ch., Arm. Lab., 1 May 1944; ltr., (1st Ind.), Wolfe to Ch., Eng. Div., 21 June 1944; ltr., Sperry, to Dir., ATSC, 24 Oct. 1944, subj.: Formal Proposal Covering 2 Standard Radar Computers (SRC-1), in WGL Library files.
30. Memo rpt. TSEPS 556-1060, 19 Nov. 1946, subj.: Optical-Radar Bombing-Navigational Computer for Jet Propelled Aircraft, prep, by Capt. M. B. Esch, Arm. Lab., Eng. Div., AMC, in WGL Library files; R&D Proj. Card (hereafter cited as RDB 1A), Proj. R-556-340, 29 Sep. 1951, in WGL Library files.
31. K-1 presn., 1 Mar. 1951; memo. rpt. TSEPS 556-1060, 19 Nov. 1946.
32. R&D Rpt., Jan. 1946, July 1946.
33. Memo, for record (unsigned and undated—between 1 June and 18 Nov. 1954), subj.: The Extent of the Navigation Field, in HBF; Clemens and Johnstone, Theory and Practice of Bombing, IV-11.
34. Clemens and Johnstone, Theory and Practice of Bombing, IV-11, IV-13, IV-14; conf. rpt., Ad Hoc Working Committee of Ground Position Indicators, 6 Aug. 1946, in HBF.
35. Eng. Div. WIR, 20 Sep. 1946; K-1 Presn., 1 Mar. 1951.
36. Eng. Div. WIR, 23 Aug., 4 Oct., 31 Oct., 18 Oct., and 25 Oct. 1946.
37. Eng. Biv. WIR, 14 Mar., 21 Mar., 26 Mar., 4 Apr., 11 Apr., 18 Apr., and 25 Apr. 1947.
38. AAP Tech. Comm., Arm. Subcomm., Quarterly Rpt., Meeting of
29 Apr. 1947, E.O. 556-340, Optical Radar Bombsight AN/APA-59, in HBF; Eng. Div. WIR, 2 May 1947.
39. Eng. Div. WIR, 23 Jan., 20 Feb., 5 Mar., and 19 Mar. 1948.
40. Supp. Agmt. No. 3, Contr. W33(038)ac-6256, 26 Mar. 1948; Statement of Military Characteristics, 29 Mar. 1948, Precision Bombsight, in WGL Library files; interview, R. L. Perry, Ch., Hist. Br., by J. King, Hist. Br., WADC, 22 Aug, 1957.
41. K-1 Presn., 1 Mar. 1951; WADC R&D PIR, Dec. 1951; RDB 1A, Proj. R-556-340, 10 Dec. 1952, in HBF; RDB 1 A, Proj, R-556-340, 29 Sep. 1951, in WGL Library files.
42. Eng. Div. WIR, 20 Feb., 2 Apr., 9 Apr., 6 Hay, and 3 June 1948.
43. E. 0. 556-340, Add. No. 4, 17 June 1948; Supp. Agmt. No, 4, Contr. W33(038)ac-6256, 23 June 1948, in WGL Library files; Eng. Div. WIR, 24 June 1948, 29 July 1948, 19 Aug. 1948.
44. Eng. Div. WIR, 3 Sep. 1948.
45. E.O. 556-340, 21 Sep. 1948; chart Bombers vs. Equipment, Sep. 1948, prep, by Arm. Lab.; rpt., K-1 Vertical Optical-Radar Bombing-Navigational System, Sep. 1948, prep, by Arm. Lab.; Eng. Div., AMC, AMC TI 2249-11A, 21 Feb. 1949, subj.: "O" Program FY 1949, Procurement of Aircraft, in WGL Library files.
46. Eng. Div. MPR, Jan., Feb., Mar., Apr,, and June, 1949; TT, MCM-3-3. CO, AMC, to CG, WRAMA, 14 Mar. 1949; RDE 1A, Proj. R-556-340, 29 Sep. 1951 and 10 Dec. 1952, in WGL Library files; Eng. Div, WIR, 24 Sep 1946; R&R, Col. B.I. Funk, Ch., Ac. and Missiles Sect., Proc. Div., to Mr, L. L. Haas, Equip Lab., Eng. Div., AMC, 14 Mar. 1949, subj.: B-47 Inverter Requirements for K-2 System, in WGL Library files.
47. Eng. Div. MPR, Aug., Sep., Oct., and Nov. 1949; RDB 1A, Proj. 556-340, 31 Dec. 1949, in WGL Library files.
48. R&R, Lt. Col. W. C. Williams, Ch., Strat. Bombing Br. to Ch., Arm, Lab., Eng. Div., AMC, 26 Oct. 1950, subj.: Request for Allocation of Production K-Systems; rpt., Bombing Systems Phasing Group Mtg., 28-29 May 1953, in WGL Library files; AMC TI 2073-30, 28 Dec. 1949, subj.: Modernization of Optical-Radar Bombing-Navigational Systems; conf. memo, 6 Aug. 1946, in HBF.
49. ARDC 111, Proj. 5147, 25 Mar. 1955, in HBF; AR DC 111, Proj. 5147, 14 Mar. 1956; DD Form 613, Proj. 5147, 26 June 1956, all in K-Sys. Sect., WGL (hereafter cited as K-WGL) Library files; DD 613, Proj. 5147, 24 Jan. 1956, in K-WGL files; E.O. 556-340, 10 Feb. 1950; ltr., Mr. C. W. Whall, Strat. Bombing Br., Arm, Lab., Eng. Div., AMC, to Sperry, 15 Feb. 1950, subj.: SAC, AMC, Sperry, BTL Conference at AMC, in WGL Library files.
50. Ltr. (1st Ind.), Col. R. L. Johnston, Ops. Office, Eng. Div., AMC, to Dir/R&D, DCS/D, USAF, 26 June 1950, subj.: Secure Bombing System, in HBF; DF, Col. M. F. McNickle, Ch., Arm, Lab., Weap. Comp. Div,, WADC, to Ac. and Missiles Sect., Proc. Div., AMC, 10 July 1951, subj.: Optical-Radar Bombing Systems-B-47 Type Airplanes, in WGL Library files; ltr., Col. M. L. Hawkins, Ch,, Ac. Weap. Br., Dir/R&D, DCS/D, USAF, Military Characteristics, Secure Bombing System, 24 Feb. 1950, prep, by Dir/Req., USAF; R&R, Lt. Col. W. C. Williams, Ch., Strat. Bombing Br., Arm. Lab., to Ch., Arm. Lab., Eng. Div,, AMC, 26 Oct. 1950, subj,: Request for Allocation of Production K-Systems, in WGL library files; memo, for record, (unsigned and undated—between 1 June and 18 Nov. 1954), subj.: The Non-Technical Parameters of the Self-Contained Navigation Field, in HBF; Kroemmelbein interview, 9 Aug. 1957.
51. Status of USAF Equip. - Request for Type Classification (hereafter cited as Form 8lA), 16 Jan. 1950, Sight, Bomb, Vertical Periscopic, Type Y-3; Form 81A, 20 Jan. 1950, Computer, Bombing Navigational, Type A-1A; Eng. Div. Memo. Rpt., MCREXG-556-1143, 9 Feb. 1950, K-1 Bombing System Conference; Form 81A, 20 Jan. 1950, Sight, Bomb, Horizontal Periscopic, Type Y-4; rpt., B-47B Mock-Up Inspection, Seattle, K-4A Bombing and Navigational Radar System, 3 Oct. 1950; Form 81A, 20 June 1950, Bombing Navigational System, Vertical Optical and Radar, Type K-3A; rpt., Weapons Guidance Laboratory Equipment, March 1957, prep, by Anal. and Design Br., Dir/Dev,, WADC, Mar. 1957; Form 81A, 20 June 1950, Bombing Navigational System, Horizontal Optical and Radar, Type K-4A, in WGL Library files.
52. Eng. Div., MPR, Jan., Feb., Mar,, and Apr. 1950.
53. Memo, for file, J. W. Schaefer, Arm. Lab., 29 May 1950, subj.: Comparison of APQ-24 and K-1 Bombing Systems - Case 26660-1; R&R, Lt. Col, D. W. Graham, Ch., Ac. and Missiles Sect., Proc. Div. to Arm. Lab., Eng. Div., AMC, 8 May 1950, subj.: Edwards Air Force Base Operation for XB-47 #46-66, K-2 Tests, both in WGL Library files; Kroemmelbein interview, 10 Oct. 1957; Eng. Div. MPR, May and June 1950.
54. K-1 Presn., 1 Mar. 1951; Eng. Div. MPR, Sept., Oct., and Nov. 1950; RDB 1A, Proj. 556-340, 30 June 1950; R&R, Col. G. A. Blake Ch., Arm. Lab., to Ch., Equip. Lab., 18 Dec. 1950, subj.: Precise Astro-Compass for use with K-Type Bombing Navigational System, in WGL Library files; R&R, Col. G. A. Blake, Ch., Arm. Lab., Eng. Div,, to Aero, Equip, Sect.. Proc, Div,, AMC, Feb. 1951, subj.: Purchase request No. 85444, in WGL Library files; WADC MPR, May 1951.
55. Exhibit MCREXG-335, 27 Jan. 1951, Sight, Bomb, Horizontal Periscopic, Type Y-6; Form 81A, 10 Nov. 1950, Sight, Bomb, Vertical Periscopic, Type Y-5; RDB 1A, Proj. R-556-( ), Development of Type Y-6 Horizontal Periscopic bombsight, 19 Sep, 1951; Form 81A, 23 Aug. 1951, Sight, Bomb, Vertical Periscopic, Type Y-7; R&D Proj. Amend. No. 1, Proj. R-556-340, 31 Oct. 1951, all in WGL Library files; ARDC R&D Task Plan (hereafter cited as ARDC 98), Proj. 5l47-50494, 24 Sep. 1953, in K-WGL files; R&D PIR, Dec. 1951.
56. Nordlund presn., 15 May 1951; undated memo., Non-Technical Parameters of the Self-Contained Navigation Field, in HBF; rpt., K-System, Write-up and Justification, 1951, prep, by Arm. Lab.; K-System, 16 June 1951, in WGL Library files.
57. Nordlund presn., 15 May 1951; R&D PIR, June 1951; minutes, K-Series Conf. held at Arm. Lab., 6-8 June 1951, in WGL Library files.
58. R&D PIR, Aug., Sep., and Oct. 1951; RDB 1A, Proj. R-556-340,
29 Sep. 1951; Exhibit MCREXG-234B, 3 Nov. 1952, Modification of K-Series Optical Radar Bombing Navigational System, in HBF; WADC R&D PIR, Jan. 1952; R&D Proj. Amend. No. 2, Proj. R-556-340, 31 Jan. 1952; Weap. Comp. Div. Proj. Info. Rpt., 31 Mar. 1952, in WGL Library files; RDB 1A, Proj. R-556-340. 10 Dec. 1952, in HBF.
59. ARDC 82, Proj. R-556-391, 6 Jan. 1953, In HBF.
60. Rpt., Bombing Systems Phasing Group Meeting at WADC, 30 Apr.-1 May 1953; ltr.; Col. W, B. Putnam, D/Ops., APGC, to CO, 3,200th Proof Test Gp. and 3,206th Test Support Gp., Eglin AFB, 3 June 1953, subj.: Test Program for Operational Suitability Test of RELIABLE Configuration, B-47 Aircraft, Project No. APG/ SAS/126-A; ltr. (1st Ind.), Brig. Gen. J. O. Guthrie, Dep. Dir/Req., ISAF, to CG, APGC, 1 May 1953, subj.: Operational Suitability Test of RELIABLE Configuration in B-47B No. 50-054; TT, AFMPE-EN-1 59284, Cmdr., USAF, to Cmdr., AMC, and Cmdr., ARDC, 24 June 1953, no subj., in WGL Library files,
61. ARDC 82, Proj. R-556-340, .10 June 1953; TT, AFDRD-SA-52781, C/S, USAF, to Cmdr., SAC, Cmdr., TAC, & Cmdr,, ADC, 13 Aug. 1953, in WGL Library files.
62. History of Wright Air Development Center, 1 July-31 December 1953, II, 315-316; History of Wright Air Development Center, 1 January-30 June 1954, III, 319; ARDC 82, Proj. S-426-278, 12 Feb. 1954.
63. WADC WIR, 20 Nov. 1953; ARDC 111, Proj. 5147, 25 Mar. 1955; DD 613, Proj. 5147, 24 Jan. 1956; DD 613, Proj. 5147, 26 June 1956, in K-WGL files; WADC WIR, 6 Nov. 1953; Kroemmelbein interview, 9 Aug. 1957; History of Wright Air Development Center. 1 January-30 June 1954, III, 355; History of Wright Air Development Center, 1 July-31 December 1954 II, 311, 312; ARDC 111, Proj. 5147, 25 Mar. 1955, in HBF; DD 613, Proj. 5147, 26 June 1956, in K-WGL files.
64. WADC WIR, 6 Nov. 1953; ARDC 111, Proj. 5147, 25 Mar. 1955, in HBF ; DD 613, Proj. 5147, 24 Jan. 1956, in K-WGL files.
65. History of Wright Air Development Center, 1 January-30 June 1954,. III, 355; History of Wright Air Development Center,
1 July-31 December 1954, II, 311, 312; History of Wright Air Development Center, 1 January-30 June 1955. II, 278, DD 613, Proj. 5021, 9 Jan. 1956, in HBF.
66. DD 613, Proj. 5147, 24 Jan. 1956; DD 613, Proj. 5147, 26 June 1956, in K-WGL files; interview, P. G. West, Strat. Bombing Br., WGL, by Gary P. Baden, Hist. Br., WADC, 3 Jan. 1958; interview, Maj. G. Dubraan, B-52 WSPO, by Gary F. Baden, Hist. Br., WADD, 22 Aug. 1960, USAF Standard Aircraft Characteristics (hereafter cited as USAF/SAC), B-52B, 1 Oct. 1959; B-52C & D, 1 Oct. 1958; B-52E, 1 Oct. 1958; B-52F, 16 Nov. 1959; B-52G, 16 Nov. 1959; and semi-annual history, B-52 weapon system, 1 January - 30 June 1960, prep, by B-52 WSPO, DSM, WADD, in HBF.
67. Ltr., Brig. Gen. L. P. Whitten, AC/AS-4, to CG, AMC, 18 July 1947, subj.: Bombsight- Equipment - Policy for Development, in MA-2 Sect., WGL (hereafter cited as MA-2 WGL) files.
68. ARDC 98, Proj. 5021-5011, 24 July 1953, In MA-2 WGL files; DD 613, Proj. 5021, 21 June 1951; interview, M/Sgt. H. D. Ford, Strat. Bombing Br., WGL, by J. King, Hist. Br., WADC, 15 Aug. 1957 (hereafter cited as Ford interview).
69. Nordlund presn., 15 May 1951.
70. Eng. Div. MPR, July 1950; ARDC 98, Proj. 5021-50113, 24 Julv 1953; RDB 1A, Proj. R-556-370, undated, (proj. approved 21 Aug. 1950), in MA-2 WGL files; BD 613, Proj. 5021, 21 June 1954, in HBF.
71. Ltr., Col. J, F. Babcock, Ch., Arm. Div., Dir/R&D, DCS/D, USAF, to CG, AMC, and CG, ARDC, 21 Mar. 1951, subj.: Precision Bombing Systems; ltr., Col. J. F. Babcock, Ch., Arm. Div., Dir/R&D, DCS/D, USAF, to CG, AMC, 20 Oct. 1950, subj.: Airborne All-Weather Tactical Bombing System, in MA-2 WGL files.
72. Eng. Div. MPR, Jan. 1951; AIF MPR, Apr. 1951; Nordlund presn., 15 May 1951; Ford interview, 15 Aug. 1957.
73. ADF MPR, Apr. 1951; WADC R&D PIR, June 1951; DD 613, Proj. 5021, 21 June 1954.
74. Ltr., Col. J. F. Babcock, Ch., Arm. Div., Dlr/R&D, DCS/D, USAF, to CG, ARDC, 29 Aug. 1951, subj.: Precision Bombing System; Revised Statement of Proposed Military Characteristics 14 Aug. 1951, Bombing System for High Altitude Supersonic Bomber; memo., Capt. G. T. Fouse, Ch., HSBD Unit, to Lt. Col. W. M. Hilt, Ch., Strat. Bombing Br., Arm. Lab., WADC, 25 June 1952, subj,: Military Characteristics, in MA-2 WGL files,
75. WADC R&D PIR, Sept. and Dec. 1951; USAF GOR SAB-51, 8 Dec. 195l, Strategic Bombardment System, in MA-2 WGL files.
76. WADC R&D PIR, Apr. 1952; ARDC 82, Proj. 556-352, 20 Sept. 1952, in HBF) History of Wright Air Development Center. 1 January-30 June 1952. II. 173-1751 History of Wright Air Development Center, 1 July-31 December 1952. II. 377-378; ARDC 98, Proj. 5021-50113, 24 July 1953, in MA-2 WGL files; DD 613, Proj. 5021, 21 June 1954, in HBF; Ford interview, 15 Aug. 1957.
77. DD 613, Proj. 5021, 21 June 1954, in HHP; History of Wright Air Development Center, 1 July-31 December 1952, II, 330; ARDC 100, Proj. 5021, 29 Mar. 1954, in MA-2 WGL files.
78. Form 81A, 26 Feb. 1952, Bombing Navigational System, Optical and Radar, Type MA-2, in MA-2 WGL files.
79. ARDC 82, Proj. R-556-352, 22 Nov. 1952; ARDC 82, Proj. R-556-352, 22 Dec. 1952; ARDC 82, Proj. R-556-352, 22 Jan. 1953, in HBF.
80. ARDC 98, Task 5021-50113, 24 July 1953, in MA-2 WGL files; DD 613, Proj. 5021, 21 June 1954; ARDC 111, Proj. 5021, 21 Mar. 1955 in HBF; History of Wright Air Development Center, 1 July-31 December 1953, II, 371; 386; History of Wright Air Development Center, 1 January-30 June 1954, III, 358-359; WADC WIR, 30 Oct., 27 Nov. 1953.
81. History of Wright Air Development Center, 1 January-31 June 1954, III, 371, 372; History of Wright Air Development Center, 1 July-31 December 1954, II, 311;, 331; History of Wright Air Development Center, 1 January-31 June 1955, II, 276.
82. History of Wright Air Development Center, 1 July-31 December 1954,. II, 314; DD 613, Proj. 5021, 9 Jan. 1956, in HBF.
83. Preliminary Handbook, Operation Instructions, Bombing Navigational System, Optical and Radar, Type MA-2 (XY-3), prep, by International Business Machines Corp. (hereafter cited as IBM), 1 Apr. 1955; Preliminary Handbook Service Instructions, Bombing Navigational System, Optical and Radar, Type MA-2 (XY-3), prep, by IBM, 1 July 1955, in MA-2 WGL files.
81,. ARDC 171, 2 Sep. 1955, Proj. 5021, in MA-2 WGL files.
85. ARDC 111, Proj. 5021, 28 Dec. 1955, In MA-2 WGL files.
86. DD 613, Proj. 5021, 9 Jan. 1956, in HBF; ARDC 111, Proj. 5021-50113, 31 Oct. 1955; ARDC 111, Proj 5021, 23 Feb. 1956, in MA-2 WGL files.
87. DD 613, Proj. 5021, 9 Jan, 1956, in HBF; ARDC 111, Proj. 5021, 30 Max, 1956; DD 613, Proj. 5021, 14 Feb, 1957; ARDC 81, Proj. 5021, 6 Apr, 1956, in MA-2 WGL files.
88. Interview, Maj. J. W. Sullivan, Strat, Bombing Br., WGL, by Gary P, Baden, Hist. Br., WADC, 3 Jan 1958; ARDC 111, Proj. 5021, 29 May 1956; ARDC 111, Proj. 5021, 13 Aug. 1956; DD 613, Proj. 5021, 14 Feb. 1957, in MA-2 WGL files; ARDC Dir/Sys. Mgmt. Weekly Activity Rpt. (hereafter cited as DSM WAR), 28 June 1957, in HBF.
89. ARDC 171, Proj. 5021, 5 Oct, 1956; ARDC Form 171, Proj. 5021, 18 Jan. 1957; ARDC 111, Proj. 5021, 27 Mar. 1957; DD 613, Proj. 5021, 14 Feb. 1957, in MA-2 WGL files; interview, W. D. Winters, Nav. Br., VCL, by Gary P. Baden, Hist. Br., WADC, 3 Jan. 1958.
90. DSM WAR, 3 May, 17 May, 14 June, 28 June, 19 July 1957.
91. DSM WAR, 15 July and 13 Dec. 1957, 24 Jan., 14 Feb., 21 Feb., 14 Mar., 27 June 1958.
92. DSM WAR, 1 Aug., 8 Aug., 22 Aug., 29 Aug., 3 Sep., and 19 Sep. 195
93. Semi-Annual Hist., B-52 Weapon System, 1 January - 30 June 1959, prep. by B-52 WSPO, DSM, ARDC; Semi-Annual History, B-52 Weapon System, 1 July - 31 December 1959, prep, by B-52 WSFO, DSM, ARDC, in HBF.
94. Ibid., USAF/SAC, B-52G, 19 Nov. 1959; interview, Maj. G. Dubman, B-52 WSPO, DSM, by Gary P. Baden, Hist. Br., WADD, 22 Aug. 1960.
95. DSM WAR, 11 Dec. 1959.
CHAPTER V SUPERSONIC BOMBING IN THE B-58
Apart from influencing the development of the "Bomb Director for High Speed Aircraft" which became the MA-2, the strategic bomber requirements of 8 December 1951 were directly responsible for a program intended to result in a bombing-navigation system specifically engineered for the B-58. The weapon objective, as more fully defined by Wright Air Development Center on 7 May 1952, was a strategic bomber with an unrefueled combat radius of 2,300 nautical miles carrying a 10,000-pound air-to-surface missile. The development contractor was to study the use of unconventional techniques in designing the weapon system, such techniques to include external stores and "expendable portions" of the aircraft itself. The bomber defense system, bombing-navigation-missile-guidance system, and the flight control system were to be "integrated" insofar as possible and the entire aircraft with its associated equipment was to utilize components already developed or nearly ready for production. The bomber was to be able to fly at supersonic speeds for as much of its mission as possible.1
Air Force headquarters on 1 September 1952 redefined the original directive to require that the system be capable of carrying the TX-13 special weapon plus "other atomic weapons" of up to 8,500 pounds total weight, an air-to-surface missile of comparable weight, or 8,500 pounds of biological, chemical, or high explosive bombs and clusters. The bombing-navigation-missile-guidance system was required to be "highly automatic , , , , accurate, reliable" and immune to electronic counter-measures. This meant that the system would have to be based on technology other than radar. The entire weapon system was to be operational in wing strength by early 1957.2
The weapon system itself grew out of "Generalized Bomber Studies" (Gebo Project) begun in 1946 by the Consolidated-Vultee Aircraft Corporation. In the ensuing five years, the contractor studied some 110,000 possible configurations of long range subsonic and supersonic strategic bombers utilizing both turboprop and turbojet engines. The original concept for extending the range of the aircraft was to use the "parasite" method; that is, the small bomber would be carried under a B-36 or B-60 until as close to the target as possible, and then dropped to perform its mission. However, in December 1951 the parasite concept was dropped and range extension was to be provided by in-flight refueling.3
Shortly thereafter, Boeing was chosen to design a competing bomber and both contractors proceeded with generalized Phase I development. In July 1952, the results were received at Wright Air Development Center, where the Consolidated configuration was chosen. This choice received official approval from Air Research and Development Command headquarters in November.
The proposed Consolidated-Vultee aircraft was to have a gross weight of some 140,000 pounds and four afterburning engines. The expendable pod was to be either a free-fall bomb or a rocket-powered air-to-surface missile. Perhaps by premonition, Wright Air Development Center immediately accepted the premise that the aircraft probably would not be operational until at least 1959, two years later than the original target date.4
On 31 December 1952, Consolidated-Vultee (now called Convair Division of General Dynamics Corporation) issued a report containing the results of a "Navigation-Bombing and Missile Guidance System" study. After investigating the potentialities of the "K-Series" systems, the "Bomb Director for High Speed Aircraft," various map-matching systems, and combinations of inertial and radiating devices, the contractor proposed basically a Doppler-inertial complex using an astro-compass for directional information. Recommended components were almost all well along in the development cycle; only a few specialized computers and some tie-in equipment would be new.
Navigation by the Doppler-inertial unit was to be accurate enough to obviate the need for a long-range, high resolution search radar. Search radar would be needed, therefore, only to correct the position information displayed by the autonavigator and to sight on the target or offset point during the last few moments of the bombing run.
Convair expected to employ a radar with a 40-inch antenna. The design provided for a non-fading-type scope display to give a videolike picture of the target area. The display could be either line-of-sight or North-oriented at the operator's choice. The scheme of bombing envisioned large offset distances with extremely precise navigation playing a major role in the mission.5
The Sperry Gyroscope Company, the General Electric Company, and the Minneapolis-Honeywell Regulator Company ware included as potential subcontractors for the bombing system. In April 1953 Sperry was chosen for the job. At the time, Convair had under consideration five different design configurations of the disposable pod to be carried by the bomber: a 50-nautical mile air-to-surface missile with rocket propulsion and inertial guidance; a guided but unpowered missile; a free fall bomb; a photo reconnaissance pod; and a ferret reconnaissance pod.
New military characteristics for the aircraft, now designated the XB-58, were issued in September 1953. The aircraft was to cruise at speeds between Mach 0.9 and 0.95 and have a "dash" top speed of Mach 2.0 in the combat zone; the powered missile was to cruise at speeds from Mach 2.75 to 3.0. The bombing-navigation system was to be all but invulnerable to electronic countermeasures, capable of bombing accurately without direct reference either to the target or an offset point, and capable of operating at the maximum speed of the aircraft (Mach 2.0). The system was also intended to be fully compatible with the automatic flight control and bomber defense system, and ability to provide acceptably accurate navigation in the event of electrical failure was a prime requirement. A single operative system capable of performing all these functions would never be installed in any aircraft, since only those portions associated with the specific pod carried on a given mission would be "installed" for that mission.
By mid-June 1954, the air-to-surface missile (for some reason designated a "rocket-powered bomb pod") was being designed for 175 nautical miles range, using a completely inertial three-axis guidance system. The missile guidance portion of the B-58 electronic and control system had only to launch the "pod" on the correct computed heading, after which the self-contained "inertial system would direct the missile to the target and a pressure sensing device would explode the warhead at the desired altitude above the terrain.
The bombardier-navigator operating the navigation, bombing, and missile guidance system had to perform multiple functions. For example, he had to make vernier corrections to the system-computed position based on intermittent radar sightings or star sightings, he inserted star-change and coordinate-change data into the system when necessary, he conducted air data runs, he armed the pod and monitored its release, he monitored the entire system for evidence of malfunctioning, he maneuvered the aircraft to a limited extent in emergencies, and he tracked the target or offset point by radar when possible at any time up to release. Not only wore these duties multifarious and time consuming in the extreme, but the operator had to rely on the trouble-free functioning of a myriad of sensors, computers, interconnection boxes, electronic tubes, and miles of multi-wire cables, many of which were located in portions of the aircraft remote from his station.6
[He had some 75 indicators and control knobs either to view or operate. The B-58 bombing system section of the Weapons Guidance Laboratory was pressing (without much success) for an "integrated capsule" containing the operator and all components of the bombing system.]
By early February 1955, Sperry had essentially completed the study and component development phases of the bombing-navigation and missile guidance system project. Sperry's reports of progress identified two major areas of difficulty. First of all, limitations on the weight of the system implied the extensive development of miniaturized components. The problem here, of course, was that such newly designed items would be highly unreliable until "debugged."
Secondly, tremendous intricacy of system assignments together with the precise computation accuracy required of the system forced the contractor into a fantastically complex design. Sperry also noted that heat generated by high speed flight undoubtedly would pose serious cooling problems.7
At this point, the bombing system being designed for the B-58 differed from that envisaged in Convair's December 1952 report in only two major items. Sperry had modified the original cathode ray "malfunction system" display by providing several switches which, when operated, would indicate the functional status of the part of the circuit being checked, and the unpowered guided pod had been deleted in favor of a free-fall bomb pod incorporating roll-stabilization only. This meant that there would be only two bomb configurations for the B-58: a rocket-powered air-to-surface missile configuration, and a roll-stabilized free-fall bomb.
Raytheon took on the B-58 Doppler and search radar development and Kollsman the astro-tracker components. Both devices were, by May 1955, in a laboratory test phase that was showing "encouraging results." Sperry, concentrating on achieving vastly improved reliability, was also beginning an investigation of "redundant" circuitry, "extremely trouble free" components, and ways of furnishing a favorable environment for the equipment.8 As of l4 October 1955, the Air Force Armament Center was planning to employ the ninth B-58 in support of Convair flight tests of the bombing-navigation and missile guidance system, starting in November 1957. The armament center also scheduled final engineering evaluation of the system to begin in February 1958.9
By the beginning of December 1955, Convair had completed limited flight testing of some of the Sperry bombing systems components in a B-36 aircraft. Sperry now planned to deliver the first bombing system (designated the YMH-1) in segments; the missile-pod control system in March 1957, the bombing-navigation system in July 1957, and the missile pod guidance system in September 1957.10
The bombing system contractor had assembled all major computing circuits of the bombing-navigation and missile guidance system into a laboratory test unit by 1 November 1956. Drawings for the complete search radar portion were released to production at Raytheon at about the sane time. The first Kollsnan astro-tracker was expected in January 1957, and Sperry hoped to have the first complete B-58 bombing-navigation system at the Convair plant in July 1957, for installation in the ninth aircraft. Before the end of 1956 selected components, such as a second Doppler radar model, a remote compass transmitter, a search radar, and a breadboard computer were being installed in a KC-97F.11
On 15 March 1957, Convair proposed an "improved B-58" to the Air Force -- essentially an aircraft with a transistorized bombing-navigation system to reduce weight and a digital rather than an analogue computer to improve speed and accuracy. An additional feature of the proposal was a 350-nautical mile air-to-surface missile with a circular error of only 3,000 feet. As late as the end of June the Air Force had not acted on the proposal and it seemed unlikely that it would be accepted, since Convair had learned, on 10 May, of a Pentagon decision to cancel the entire missile-pod program.
[This eliminated the requirement for missile-pod control and guidance equipments.]
An official contract change reflecting the deletion received approval seven days later.
Sperry was some five months late in delivering the first YMH-1 and installation began in December 1957, with the aircraft scheduled to fly in April 1956. By the end of 1957, the weapons Guidance Laboratory felt that the major navigator problems had been surmounted.
At this point only 13 YB-58A "state-of-the-art" aircraft and 30 production models were on order.12
Installation of the system in a B-58 (aircraft number 9) was delayed by a lack of test equipment (under development concurrently with the bombing and navigation system itself), some component failures, and the results of poor quality control at Sperry. The Air Force also planned to have Convair install another bombing and navigation system in a B-58 (number 11) for climatic hangar tests at Eglin, and later, arctic tests in Alaska.13
The system for this aircraft arrived at Convair's Fort Worth plant in April, but was soon in the same predicament as the initial unit. Aircraft number 9 finally flew on 13 May, but only with a partially operative bombing and navigation system (the radio altimeter, astro-tracker, and a number of other components had been "wired off" because of failure to produce satisfactory ground operation), and the equipment that performed at all was unsatisfactory. By this time the difficulties at Sperry and Convair had combined to threaten a 60 day delay in outfitting B-58's 12 through 25 with bombing and navigation systems. A Convair "team of top management, production, engineering and test personnel" had been at Sperry since early in May, lending a hand in overcoming existing problems.
Later flights with aircraft number 9 were more successful, although some elements of the system still had to be proven. On 29 May 1958, a group representing the project office and Convair met to discuss the situation. While recognizing that Sperry was taking every step possible to eradicate its difficulties, the group found that system installation for the next group of 14 aircraft would be delayed by two months, and that installation in the first five aircraft might be as much as three months behind schedule. In an attempt to make up the lost time the conferees agreed to compress the flight test program.15
Airborne tests in aircraft number 9 showed a marked improvement in system performance during June 1958. The second B-58 (number 11) flew for the first time on 26 June, entered the climatic hangar at Eglin twelve days later, and passed the rigors of cold and warmth on 16 July 1958. Before the end of July, system deliveries to Convair had reached the point where eight complete sets wore on hand. Seven of these were in B-58's, and the initial flight test aircraft had been appointed with a fully operative unit. Furthermore, Sperry was now conducting a limited flight evaluation of selected bombing and navigation system deficiencies in a KC-97 and Raytheon was using a C-131 for Doppler tests; results in the latter case were "very good."16
Thus, mid-1958 was essentially the point at which the bombing and navigation system for the B-58 passed the stage of design and development since only a few small adjustments had yet to be made. However, the system still had to be proved reliable and capable of being maintained in the field by SAC. The Air Force affirmed this situation in August. After reviewing the programs, a joint ARDC and AMC survey team proclaimed that no "apparent major design deficiencies existed," but that reliability and maintenance difficulties remained and might not be eliminated by the end of system testing in October 1959. Still, the survey team was able to isolate the main contributors to the present situation (insufficient test equipment at the time of initial system installation in aircraft, poor quality control, minor design deficiencies, and Sperry's excessive use of "out-of-specification tolerance conditions") and Sperry was soon exerting effort in each area.17
In September 1953, to place the final stamp of recognition on the system, the Air Force dubbed the Sperry unit the AN/ASQ-42V Offensive Weapon Control System. Testing continued into October, but was again delayed by the dearth of check out equipment and some component malfunctions. By the end of the month the first "fully checked" system had been installed in a B-58, number 14, and within three weeks enough units had been received to outfit aircraft 12 through 18.
Aircraft 15 had a completely operable system installed soon after and on 16 December was accepted by the Air Force. All previous aircraft and attendant bombing and navigation systems had been handled by Convair, as weapon system contractor, and it was now the turn of the military flight test group at Carswell Air Force Base, near the Convair plant, to conduct its own set of airborne trials. However, B-58 number 15 crashed before the year was out and the program was stalled until a new vehicle, suitably appointed, could be obtained. The climatic hangar test aircraft (number 11) left for Alaska on 31 December 1958 and returned, after a successful endeavor, on 5 March 1959. By this time the total flight test time for the ASQ-42 amounted to 86 hours, but in no instance had an entire system operated successfully throughout a flight.
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
B-58 Bombing and Navigation Equipment
The Air Force flight test group received aircraft 17 with an operating ASQ-42, on 21 March 1959 and conducted its initial flight on 1 April, Earlier aircraft containing older versions of the bombing and navigation system were then retrofitted with newer components, and in June the basic character of the test program was altered.
Until this time the airborne evaluations had aimed at demonstrating that the various components of the system could meet Air Force performance specifications. Now, flight tests were to determine the "functional, reliable and maintainable performance of the [entire] B&N system. . . ." In order to expedite procedures, added personnel were assigned to Convair, the test group, and the office of the Air Force Plant Representative at the Convair facility, and the test group was scheduled for more aircraft and additional logistical support.19
By July these moves were completed and the test group had accomplished over 81 hours of flight test, while the Convair program totalled same 245 hours, Reliability, specifically in terms of ground maintenance and check out, remained a problem.
Specific "high failure rate items" and engineering deficiencies of the ASQ-42 had 'been isolated by mid-1959 and before the end of the year attempts to provide adequate solution for each problem were beginning to bear fruit. To overcome low reliability and marginal operation of the search radar, a Model D set was placed in the test aircraft. This unit, with a redesigned antenna, approximately twice the power, and a number of other new features, displayed significant improvement in performance and reliability. The Doppler radar had failed to maintain lock-on at high altitudes and was subjected to excessive electronic interference but the addition of a low frequency "cut-off filter" to overcome interference generated by "engine reflected signals," and a number of other adjustments solved the problem. Imprecise radio altimeter readings remained an unsolved mystery until December 1959, when an investigation revealed that the tail compartment of the B-58 pod acted as a "ringing chamber" for altimeter signals and returned signals of greater strength than those from the ground. The chamber was covered with electrical wave absorbent material and the problem faded into oblivion.
Other difficulties remained as the year 1959 ended, but it seemed certain that the major problems of the ASQ-42 would soon be a thing of the past. By December 1959, flight tests had shown that the dead reckoning navigational error had been reduced from a circular error of probability (CEP) of 10 to 11 nautical miles per hour in January 1959 to 4 to 5 nautical miles per hour as the year ended. Radar bombing accuracy in September 1959 showed a very poor 5,000 foot CEP, in December it was a creditable 1,500 feet. Turn-around time had been 9 to 10 days in the summer of 1959, it was approximately 4 days in December. Overall system operation on an average mission was increased to the point where, in December, systems operated for between 70 and 80 per cent of each flight. No serious maintenance problems remained and the airborne reliability of the first complete production aircraft was rated at 74 per cent. These were noteworthy achievements.
The future of the ASQ-42 held the need for operation at low altitudes. The requirement for flight "on the deck" had been established in May 1959, when the project office received authority to conduct a preliminary design study. Actual development was authorized in January 1960.20
CHAPTER VI THE B-70 BOMBING AND NAVIGATION SYSTEM
As early as 1945 the Armament Laboratory had expressed an interest in navigation based on star tracking. Stellar navigation, unlike navigation relying on radar and optical systems, grow more accurate as altitude increased. It could be accomplished—theoretically— with no earth reference. Stellar reference points had known positions with respect to the earth, so through star tracking, precise terrestrial position could be computed easily and solutions to bombing and navigation problems could be obtained automatically. Moreover, the ability to deliver a bomb load without reliance on the transmission of radar or radio signals was obviously very desirable.
At the same time, the Air Force also looked to the possibilities inherent in the development of inertial systems as well as inertial systems monitored by star tracking. Inertial devices could derive data entirely by sensing acceleration and the direction of motion of the air vehicle. The Massachusetts Institute of Technology (MIT) undertook exploratory work in both areas.
In 1946, MIT took an important step along the stellar-inertial path under what came to be called Project Febe. The project, which was essentially a feasibility study, considered the use of stellar data to determine ground position, ground speed, and drift with respect to earth coordinates within accuracy limits applicable to effective bombing. Preliminary tests of inertial components were so encouraging that MIT abandoned theoretical work on star-tracking feasibility and tackled the larger problem of designing a pure inertial autonavigator, without stellar monitoring, that would give position accuracies of one nautical mile after 10 to 12 hours of flight.22
Under Project Spire (Space Inertial Reference Earth) MIT attempted to design, fabricate, and test an experimental pure inertial system. Progress was excruciatingly slow. By December 1957 the project had been terminated, although astonishingly good flight test results had been recorded in its final few months. Flight-derived acceleration data had supplied nearly all the information required to compute the actual course, position, and velocity of an aircraft with respect to the earth. This data, if fed into a bombing computer, would allow correction for errors due to cross trail (*) and could also be employed to compute a bomb release signal.
[* - The consequence of bombing in a cross wind; a bomb fell along a line determined by the heading rather than the course of the releasing aircraft.]
The computer "knew" when flight-derived position data showed the aircraft to be precisely where pre-flight calculations indicated it should be at the instant of bomb release.
But the Spire system tipped the scales at almost 3,000 pounds, its gimbal system was over four feet in diameter, and its console panels were extremely bulky. It incorporated no provision for allowing the aircraft to fly a "dog-leg" course, and the target could not be changed once the flight had begun. Experimental equipment was neither pressurized nor insulated and thus was not suited for extreme operational environments, although this could have been provided for production systems. Another major handicap was the system's need for approximately 48 hours of pre-flight gyroscope calibration and alignment.
In early 1955, the Massachusetts Institute of Technology was designing a lighter and mere flexible inertial system—called Spire Jr.—which scientists hoped would overcome the shortcomings of the older model. Spiro Jr. was flight tested in a C-97 with aircraft control provided through the autopilot. The junior version was about one-half the size and weight of its forebear, and could be completely aligned and readied for installation while in a hangar. Tests indicated that the unit could be removed from the hangar and placed in the C-97 in four and a half minutes—but calibration in the hangar still took two days. On one six-hour C-97 flight, position errors of four miles in range and one and one-half miles in track were recorded.23
Air Force interest in some form of a "correctable" inertial system, particularly one using star tracking to correct for gyro drift errors, also led into a number of other investigations, ultimately involving six contractors for almost a decade.
[Hughes Aircraft Company, IBM, North American, Pacific Mercury Television Manufacturing Company, AC Spark Plug Division of General Motors, and Sperry]
In May 1946, Hughes contracted to develop an experimental red-sensitive automatic star-tracker for use with a stellar-inertial system. The tracker was to be capable of detecting and following stars in full daylight. In addition, Hughes was to study means of directing and controlling an aircraft over a course between two predetermined points on the earth's surface using an inertial device corrected by continuous star sightings.
Later in that year Hughes reported that an experimental star tracker (the Mark I) had been successfully tested on the ground. The instrument had actually converted daylight star images into controlled electrical signals. Hughes modified the design of the original model during the following year and subsequently built a Mark II star-tracker, which was mounted on a five-gimbal gyro-stabilized platform. Flight tests in November 1947 indicated the desirability of simplifying the system somewhat, so Hughes began work on a three-axis gimballed stellar-inertial navigation device.
It was a long process, but by February 1950 Hughes had built the simplified model and had begun flight tests. Within two months, the device had demonstrated destination errors of approximately five miles after four hours of continuous flight. Although this did not meet existing specifications for navigation, bombing, or guidance systems, the Hughes equipment had done sufficiently well to warrant Air Force continuation of the program.24
By mid-1951, Hughes had completed two engineering models of the star-tracker itself and had performed limited experimental work with a telescope of reduced size using a refrigerated infrared detector cell. This detector allowed a considerable increase in the number of stars that could be tracked and the smaller telescope proved practical. Hughes, however, pleading the press of other work for the Air Force, backed out of the star-tracking development area and Wright Air Development Center transferred the project to the Pacific Mercury Television Manufacturing Corporation. Some project personnel and two experimental star trackers also went to North American Aviation, where they wore employed in stellar-inertial projects intended for the Navaho (SM-64) program.
Pacific Mercury continued development work from the point at which Hughes had withdrawn and by mid-1953 had constructed three star tracker models employing miniaturized telescopes. At that point the Air Force terminated the original project, alleging "excessive cost." The corporation delivered one model of the tracker to Wright Field for further testing, while the others went to AC Spark Plug Division of General Motors Corporation for incorporation into that contractor's stellar-inertial bombing system ("Sibs").25
The development at AC was to result in the fabrication of one experimental and one preproduction system based on the Spire design but incorporating stellar monitoring of the inertial component. The original specifications called for system operation at 60,000-foot altitudes, at ground speeds as great as 1,050 knots, and at true air speeds of 800 knots. AC's system was to be capable of operating at any time during the day or night. The Air Force also anticipated that it would include some type of "precision bomb director" employing either a radar or optical sight for the terminal phase of a bombing mission.26
The contractor began fabricating the first test components in August 1951, planning to use the Pacific Mercury star tracker, and completed the initial (non-flying) experimental system in December 1952. Tests of subassemblies then began.27
The goal was a system operating in five general "phases." During the preflight phase, pilot and crew would plan check points, targets, and navigational stars and select primary and alternate routes. Ground teams would then prepare punched tapes containing the information necessary to enable the system to follow the aircrew's mission plan. During this phase, the system itself would be aligned and calibrated, largely through the use of portable instrumentation.
The navigation phase of the mission, which could cover any distance, was to be flown with completely automatic stellar-inertial guidance For the bomb run, however, the navigator switched on the radar, after which the crosshairs would automatically align themselves on a predetermined check point with a maximum uncertainty of one nautical mile. The navigator had only to center the crosshairs on the checkpoint or target to reduce system error immediately to less than 1,000 feet. Under ordinary circumstances, the bombing phase would be completely automatic and would require no radar ground sighting (although radar could be used). Radar would be employed on the final or "return navigation" phase but only for a short period to perform indirect bomb damage assessment. Automatic stellar-inertial guidance would then take the aircraft back to a pre-selected air base.28
In May 1954, the Armament Laboratory divorced the stellar-inertial developments for guided missiles from the manned aircraft area and coupled bomber efforts under a single project entitled "High Altitude Strategic Bombing Components and Techniques." The two principal goals were equipment to launch initially guided air-to-surface missiles and improved aircraft devices incorporating stellar monitoring for navigation and bombing, utilizing as many automatic features as possible. The air-to-surface missile was to have a maximum circular probable error of one nautical mile with the entire crew of the launching aircraft "incapacitated." With crew attendance, this error was to be reduced to about 1,500 feet.
The project included a number of tasks. One called for International Business Machines to modify the developmental MA-2 bombing system, (which relied on radar) so that it could "align and launch" air-to-surface missiles. Although the promising "Sibs" development, based on a five-axis gimballed platform, became part of the project, the gimballed unit appeared to be too bulky for bomber use.
[A "catch-all" task, bearing the cogently descriptive title of "Crutch," was also included. Under the assumption that pure inertial systems wore "far from perfect," the task objective was to investigate "crutches" of various sorts (star trackers, stabilized astro-compasses, damping devices, ground fixing devices, and Doppler radar devices) to "assist" a basic initial component.]
Sperry undertook another task— carrying the short title "Tempo"—which called for study and component development of a three-axis stellar/inertial device of small volume and weight.29
By mid-February 1955, International Business Machines was studying alignment and launching techniques and was fabricating experimental components to operate the MA-2 in conjunction with air-to-surface missiles. At this juncture, the MA-2 was planned for the B-52 and was considered as a back-up for the B-5B. The goal was to provide a 100-nautical mile "passive navigation ... bombing phase" for dropping free-fall bombs and to allow missile launch at approximately l50 miles from a target. Circular probable error of the free-fall bomb was not to exceed 3,000 feet, and of the missile, 5,000 feet at extreme range.
By this time Sperry's two "Tempo" systems, incorporating infrared star trackers, were in flight test but Sperry reported three major problems: difficulty in obtaining the extreme manufacturing precision required for vital components, such as gyroscopes and acclerometers; inability to demonstrate continuously reliable star detection under flight conditions; and the need to simplify the intricate pre flight procedures.30
Running concurrently with all these developments were the extensive inertial and stellar-inertial programs at North American Aviation. North American began working in the general area of pure inertial guidance in 1946 with most of the early work being directed towards guidance for what ultimately became the Navaho (SM-64) missile.
In 1950, while continuing the basic Navaho program, North American began working on the development of stellar-inertial guidance for bombers. One of its efforts was associated with Brass Ring, a relatively short-lived attempt to employ a B-47 as a drone bomber.
[Brass Ring was an attempt to devise means of delivering a massive thermonuclear weapon to targets beyond the round trip range of existent bombers. A human crew was to perform take-off and course-setting assignments before bailing out; automatic navigators controlled the balance of the mission. Crew bail-out seemed necessary because of the one-way nature of the mission and also because of the general expectation that the then-untested thermonuclear bombs would inevitably destroy any drop aircraft within many miles of the blast.]
North American made use of much of the star tracker work earlier accomplished by Hughes, hired some of the Hughes engineers, and in June 1951 acquired two of the experimental star trackers Hughes had built.
By this time, North American had also developed an experimental pure inertial system, the X-1, and an experimental stellar-inertial system, the X-2. The X-2 was mounted on a truck for tests and evaluations early in 1951; in February of that year the contractor reported that "the first known successful tracking of dim stars in daylight from a moving vehicle has been accomplished and is consistently repeatable. . . ."
The X-2 was later flight tested in a YC-97, the program lasting until May 1953. However, the X-2 and the X-1 were breadboard models built solely to test the firm's original concepts. In 1952, North American began working on a version of the X-2 for operational employment in air vehicles. This was the XN2B. Although the ultimate Navaho missile was to be guided by a pure inertial system, the XN2B was included in that program to permit tests of experimental versions of the missile and for use until the accuracy requirement of a pure inertial system could be met. The XN2B was tested in a T-29 beginning in June 1954 and was experimentally installed in the X-10 (the test vehicle for the Navaho) later in 1954. The N2B, a refined unit, was tested still later. North American also developed an N2C version as a back-up guidance system for Northrop's Snark missile.31
On 15 April 1955, these semi-independent efforts took on a new meaning. This was the publication date of an ARDC System Requirement (SR-22) for a "Piloted Strategic Intercontinental Bombardment weapon System"---Weapon System 110A. The requirements outlined in this document clearly indicated the need for an inertial system with some sort of monitoring. Thus there arose the possibility of a manned bomber application for the work at North American, IBM, and other firms, hitherto largely pursued without reference to employment in a specific weapon system. First, however, the details of design for the 110A system had to be established. The Weapons Guidance Laboratory took definite stops in this direction three days before issuance of the system requirement by publishing an exhibit outlining functions required of the new bombing and navigation system.
System Requirement 22 set a target date of 1963 for an operational wing of Weapon System 110A aircraft, designated this system as the B-52 successor, and assigned a IA priority to the program. To meet the target date, a weapon system mockup was required by October 1957, a first flight by September 1960, and initial SAC inventory aircraft by November 1962. This vehicle was to be "a conventionally powered bombardment system capable of intercontinental nuclear weapon delivery ... thereby permitting maximum freedom from political, logistical, and security problems." Based in the continental United States, the 110A system was to have the capacity to strike at and destroy three types of targets within the "Soviet Bloc": the "military, logistic, and control strengths" associated with strategic, tactical and defensive air weapons; the "military, logistic, and control strengths" associated with land and naval weapons; and
the industrial and economic strengths of the enemy that enable him to sustain or increase the expenditures of warfare, to advance the technology of warfare, to outlast or compete with our corresponding strengths as they become strained or exhausted in warfare. Destroy additionally as required, the control and psychological strengths of the enemy that enable him to direct and organize his strengths in warfare or to adhere to any course of action involving warfare.
The requirement also reflected refined strategic concepts in the statement that "winning the air battle by destroying the enemy's air bases is the most important initial task of a war;" other "highly resistant targets" such as atomic storage sites and missile launching installations, were secondary.
The weapon system designed to perform these tasks was, first of all, to "be capable of reacting quickly." This meant that "in a combat alert status [it] must be capable of taking off on a combat mission within 3 minutes after the crew takes stations." Other aircraft, those not on alert status, were to be airborne within 30 minutes. This was a very stringent requirement, one that was to weigh heavily in the design of the bombing and navigation system for the 110A aircraft.
The weapon system was also to be capable of operating at all times—day and night—and under all types of weather conditions. It has to overcome what the requirement termed the "limitations of present systems" in terms of range, combat zone altitude, speed, and defensive capability.
Equally as important as operational performance was the desired reliability requirement, which called for "85% of the available (in commission) aircraft" to undertake a defined mission "with no component failures." This was an extremely rigid requirement, ranking with the alert strictures as a basic factor in weapon system design.
The 110A system was to have a minimum unrefuelled radius of 4,000 nautical miles, with 5,500 nautical miles desired. It was to cruise at not less than Mach .9 "unless a significant increase in maximum radius of action or in combat dash zone performance may be attained." In the combat zone, "maximum possible supersonic dash was required," and this speed was to commence at least 1,000 nautical miles from the target, with 2,000 nautical miles desired. Target altitude was to be 60,000 feet, with 70,000 feet desired. The "capability of this weapon system for low altitude operations" was also to be investigated.
Target destruction was to accrue from conventional bombing and employment of an air-to-surface missile. The missile was to be included as an "alternate load'' employed to increase "survival probability" against "intense local defenses or weapons effects." It was to range 300 nautical miles, with 700 nautical miles desired, at a speed of Mach 3, with an accuracy of "1,500 feet CEP at 300NM."
[That is, an average miss distance of 1,500 feet.]
The bombing and navigation system itself was to take advantage of the "inherent flexibility" of a manned system and employ automatic features "only to the extent necessary for mission success." In essence, this meant the 110A bombing and navigation system would take "maximum advantage ... of man-machine combinations."
A "careful balance between accuracy, reliability, and serviceability" was required, plus "minimum susceptibility to passive detection," a "high degree of security from jamming," "enroute [position] fixing and terminal sighting," accuracy in "reference point bombing from distances of up to 200 NM" and "indirect bomb damage assessment." Tracking ranges of the system were to extend to 100 nautical miles. System component malfunctions were to be minimized "by the use of such techniques as multiple circuitry and modular construction." Moderate accuracy, a 1,500 foot circular error of probability (CEP) for bombing, was required although this figure could be increased to 5,000 feet "if the loss can be translated into increased reliability and/or improved maintenance characteristics." Obviously, nuclear destruction required less accuracy, and precision could be sacrificed to further insure the weapon system would reach the target.
The WADC conception of a bombing and navigation system for the 110A (and the nuclear powered system 125A), (*) published as an "exhibit" three days before approval of the system requirement, further defined the concepts destined to mold the new program.
[* - A system common to both programs was the first objective, but if this was not possible common components were to be employed wherever possible.]
It called for a study which in time could serve as an outline for actual development and posed a number of hypothetical questions for consideration, but it was also an indication of early Armament Laboratory analyses of the problem at hand.
During the remainder of the year, the laboratory gave continual attention to the 110A bombing and navigation system. The key requirements dominating all considerations were those calling for rapid response, high reliability, high serviceability, air-to-surface missile launch and free-fall bombing, strong resistance to jamming, reference point bombing, and moderate accuracy.
Another factor was the need for an air weapon by 1963. The laboratory felt that this date could only be met by a bombing-navigation system under development since 1953. Thus, either something already under development would have to be considered or components already under development would have to be mated for the new environment.32
Regardless of the eventual equipment choice, the bombing and navigation system (*) for the 1I0A had to be compatible with System Requirement 22.
[* - The system was officially the Bombing-Navigation-Missile Guidance System. Missile guidance, although related to the functioning of the bombing and navigation equipment, was essentially a factor of missile development and is not treated here.]
This meant, first of all, that it would have an inertial platform. The requirement called for reference point bombing. Reference point bombing meant that the bomber had to be able to cover 200 nautical miles from an established reference point to the target without employing radiation devices for guidance. A precise determination of position and course from the reference point would permit the bomber to proceed to its target over the final 200 nautical miles while giving no indication of its presence.
The only known device which could fulfill this requirement was an inertial system. But in 1955, (and in 1960) inertial platforms were not accurate enough to solve the bombing and navigation problem alone. They were particularly fallible in providing azimuth data, the heading of the bomber with respect to north. Azimuth, velocity, time, and altitude were of course the critical factors in determining the position of the air vehicle.
One of the best ways to get accurate azimuth data was to relate the heading of the aircraft to the sub-stellar position of known stars. This meant employing a star tracker. And a star tracker could best be used with an inertial platform when it was physically located atop the latter. Logically, therefore, the bombing and navigation system should be stellar-inertial.
The inertial platform, acting essentially as a memory device, had to be supplied with basic navigation data. This took time, and the System Requirement allocated only three minutes from alert signal to the time the bomber was to be airborne. Under this condition, the platform would have to receive data after the vehicle was airborne. And this data would have to be of three kinds: azimuth, position, and speed (or velocity). Azimuth was already provided for. Speed and position information still had to be provided.
The laboratory felt that the best way to obtain speed data (in relation to the earth) was to employ a Doppler radar to measure slowly varying changes in speed; tied to the platform, it could also help reduce platform oscillations. The platform also contained two acceleroneters to sense rapid velocity changes. This combination of speed data and azimuth data (fron the star tracker) would provide the platform with the basic ingredients for navigation. Position was far more difficult to obtain. It was not absolutely necessary— although desirable—to determine exact position at the beginning of flight, so the laboratory eventually decided upon a graduated sequence of events designed—at each step--to provide increasingly precise position information.
The Weapons Guidance Laboratory now had a stellar-inertial system with a doppler radar. The next problem was to devise a way to mould those sensing devices into an integrated system. This meant a computer, ideally a high capacity, precise digital computer. A digital computer had never previously been used in a bombing and navigation system; its only application to aircraft in 1955 was a planned use in the highly sophisticated MX-1179 system being developed by Hughes for the F-106 interceptor.
The storage capacity associated with a digital computer also played a part in star tracker operation. Star trackers had to have exceedingly accurate data on the location of stars in order to be able to seek and find the stellar bodies. If this data was not available, interminable scanning would he required, and star trackers lost accuracy rapidly when scanning through a large cone; the initial attempt at scanning had to be no more than one-fifteenth of a mile in position and one-tenth of a degree in azimuth from the star to be tracked. Thus, the computer would have to store, before flight, a large number of exact star locations (later set at 100). This data would be used to decide which stars could best be tracked and to direct scanning accordingly. Furthermore, if the 110A system was to be housed in a supersonic vehicle, acceptable navigation at these speeds would require new position information every half second or so, and only a digital computer had sufficient capacity and precision for those computations.
International Business Machines, one of the contractors interested in the new project, was consulted early in the design stage. The first estimate of the nascent system supplied by the contractor indicated that the ultimate device would have some 2,000 transistors, 4,000 diodes, and 600 connectors. After one look at these figures the laboratory quickly concluded that the reliability requirement of the system was out of reach. The obvious solution to the problem thus presented was to take "maximum advantage ... of man-machine combinations" as the original system requirement had suggested.
The entire system was of course designed to take advantage of a capacity for intelligent judgement and decisions on the part of trained operators. However, human operators could also be used for in-flight maintenance to monitor equipment and to decide how best to compensate for equipment failures. Since IBM predicted that the computer would probably fail before any other element of the whole, and the human operator could not completely compensate for such a loss, the laboratory decided to add an emergency computer. It was smaller and less efficient, but supplemented by human intelligence it could do the job if necessary. In addition, the human operator was to have a direct operational function. Since he could observe a radar scope (optics were out of the question at the 110A's cruise and combat altitude), the laboratory included a search radar to be used as a conventional bombing instrument and in navigation. In essence, this completed the theoretical design of the bombing and navigation system for the 110A.
But techniques for operational employment of these devices also had to be devised. The first objective of system operation was star tracking and 95 percent of the time devoted to preparing for star tracking was flight time.
In order to track stars, the inertial platform had to be erected with precision and had to be fed highly accurate aircraft velocity and azimuth data. The Doppler radar provided velocity information. This was no problem but the determination of azimuth, of the relationship of the aircraft to true north, was another matter. After considering and discarding alternate processes, the laboratory decided to obtain precise azimuth data by the graduated process.
First, a rough azimuth could be obtained by a quick determination of aircraft heading at take-off. This could be done before flight, but it meant that every base used by the 110A system would have to obtain precise runway headings before the aircraft could be placed there on alert status.
[This technique was actually established late in the program but is treated here because of its early place in system operation.]
On the basis of this data and the Doppler data, the gyroscopes could be erected, but the platform required still more precise data before it could function with the extreme accuracy required for proper system operation.
The next step was a "gyro-compassing phase." Here the erected inertial platform could be rotated so that it defined a line relative to the axis of the earth's rotation; this reference was to be used to obtain a reasonably precise orientation with respect to true north.
Ultimately, the precision position information required for actual star scanning had to be provided by use of the search radar. With fairly accurate position data available, the search radar could be used to illuminate a known point on the earth. With this point established, the information required for the determination of exact aircraft position—in relation to the known terrestial spot--was available, and the computer could supply the star tracker with instructions on which star to seek and about where the star could be found. After one star was located, another was sought and located. Thereafter, the system could continually establish both azimuth and position with a highly precise degree of accuracy. The whole process would take about 30 minutes.
At the reference point, which was established at 200 nautical miles from the target (but theoretically could be any time after alignment), the search radar could take a final position fix. All radiation then stopped and navigation was thereafter dependent entirely on the inertial system.
Having settled on a technique, the Weapons Guidance Laboratory had to decide which particular pieces of equipment (computers, radars, gyroscopes, etc.) could do all that was necessary for the 110A bombing and navigation system. This meant that a contractor would have to study and then develop items synchronized with the overall weapon system development plan. Money and scheduling thus became factors. In short order they became the dominant factors in the entire 110A weapon system program.
The search for devices and a contractor was conducted under the assumption that a totally new system could not be developed if the weapon system schedule was to be met. Some existing equipment would have to be used. Implicit in this approach was the desire to equip refined versions of the 110A with advanced bombing and navigation devices based on new developments. Therefore, initial attention was given to the K-5; the MA-2 (developed at International Business Machines for the B-52); the B-58 bombing and navigation system (a Sperry product); the Bell Telephone Laboratories studios for the 125A; and Spire Junior in combination with a stellar tracker, a radar, and a computer. By August 1955, the Weapons Guidance Laboratory had narrowed the field to an improved MA-2 and the B-58 system. The reliability, advanced state of development, and attractive performance features of the MA-2 (such as the search radar and anti-jamming qualities) caused the laboratory to favor this system in preference to the B-58 system. Nevertheless, in September 1955, Sperry, Motorola, and IBM all undertook studies of possible 110A and 125A bombing and navigation system configurations.33
By the summer of 1955, the entire 110A program was beginning to take on a more definitive character, one that set it apart from planning for the development of the 125A weapon system. On 27 July 1955 six contractors had attended a meeting at Wright Field to discuss possible 110A proposals. Lockheed, Douglas, Convair, Martin, North American, and Boeing comprised the group, but only the latter two desired to participate. The Air Force thereupon decided to sponsor initial studies at both concerns.
The bombing and navigation portion of the 110A program was disassociated from the 125A by 1 September 1955, essentially because the development schedules of the two weapon systems were no longer compatible. The 110A was now scheduled to enter Phase 1 (the study phase) some 13 months before its nuclear-powered cousin. In December 1955 Boeing and North American were awarded letter contracts covering 110A studies (the contracts were pre-dated to 8 November to cover work begun earlier) and on 7 December ARDC headquarters officially authorized the start of work on the bombing and navigation subsystem by International Business Machines. Because time was short, the Weapons Guidance Laboratory favored contracting for the services of that firm on a sole source basis. This procedure also had the effect of making a modified MA-2 system virtually inevitable for the 110A.34
Before the end of 1953, North American and Boeing also had contributed some opinions on bombing and navigation aspects of the 110A project. North American was essentially in agreement with Wright Field armament experts. Boeing dissented. Whereas North American was ready to employ a system based on "components ... drawn from existing programs which were well advanced" for the "initial" 110A and proposed to investigate ultimate systems for later versions of the weapon system, Boeing favored a new development and gave slight consideration to the generally accepted belief that only a device currently in the development stage could
be available for the first weapon system. The Boeing approach, in the opinion of the Weapons Guidance Laboratory, "might endanger the success of the complete weapon system." The laboratory also questioned the wisdom of Boeing's proposal to rely on a side-looking radar and to forego provision for a conventional forward-looking radar. A side-looking radar increased system resolution but was slower in operation, posed video display problems, and was not a proven technique. Moreover, use of a technique based on side-looking radar might prevent any later addition of a forward-looking radar "if such was found necessary."35
On 15 December 1955, laboratory views on the proposed bombing and navigation system were expressed in another bombing and navigation system exhibit. This document established July 1957 as a mockup date, April 1960 as the date for completion of the first prototype, November 1961 as the time for an initial production model, and November 1962 for the first complete bombing and navigation component of a 110A weapon system in SAC. The exhibit also specified that the study to be conducted for the 110A bombing and navigation system had to consider first "existing components, components in the advanced states of development, and proven techniques. ..." Once again, reliability was a critical requirement. Performance equivalent to that of the MA-2 radar—the High Speed Bombing Radar (HSBR)— was to be a minimum objective. (The HSBR employed forward-looking radar techniques.)36
During the early months of 1956 the entire 110A weapon system program took on a new face. For a time the Air Force again mulled over the possibility of having International Business Machines work on bombing-navigation for the 125A program as well as the 110A, but the idea died. In any case, the contractor submitted a technical proposal for the bombing and navigation system on 9 March and six days later presented to Wright Field personnel a complete development plan. The basic technical aspects of the IBM proposal were accorded a friendly reception, but the firm's development schedule was six to nine months behind the weapon system schedule.
[The time difference depended upon whether the first prototype was to be ready in April or June 1960); this was still to be decided.]
To further complicate the timing imbalance, the project office also reviewed, in March, a request to consider a 110A program which called for an initial flight in October 1959, an initial SAC aircraft in December 1960 or July 1961, and the first SAC wing in either December 1961 or 1962. The official program schedule announced shortly thereafter required an initial production aircraft in November 1961, a first SAC aircraft one year after that, and a SAC wing by late 1963.37
The Weapons Guidance Laboratory advised project officials that the new timing sequence could be met only if system performance requirements were eased, and added that the best bet for the new dates was the MA-2 with stellar-inertial features and a Doppler radar. As it happened, however, the acceleration request was a straw in the wind, an indication of the rapidity with which the entire program was to be changed. No program acceleration resulted.
By April, the laboratory had completed evaluation of the contractor proposal. Although IBM on 30 March 1956 formally agreed to undertake the development, (**) and the laboratory judged the proposal to be "comprehensive and complete," it was not accepted for a number of reasons.
[** - A supplemental agreement to the initial study contract generally determined that work would be only for the 110A. Details of the contractor effort were to come later.]
One was the schedule disparity; IBM could have assembled only a prototype system by June 1961—one year too late to meet the vehicle schedule. The laboratory also feared that the contractor, in his effort to provide the "highest possible decree of operational capability," was risking too much by proposing inclusion of a side-looking radar. Therefore the laboratory advised IBM that a new and scaled down proposal was required, one that emphasized free-fall bombing at the expense of missile launch and side-locking radar. Still the laboratory favored a system with the "growth potential to meet fully desired performance later."38
On 16 May 1956, the contractor presented at Wright Field a revised program based on the views already expressed by the Weapons Guidance Laboratory, The result was tentative approval of a system comprising the MA-2 and its High Speed Bombing Radar, and including the North American N2C, the General Precision Laboratories' APN-96 Doppler radar, and the IBM solid state (transistorized) Dinaboc digital computer then under development. A Goodyear side-looking radar was approved for use in conjunction with the forward-looking radar, but the program was not given primary emphasis.39
A formal contractor proposal was received on 15 June 1956, only to meet with new objections from the Weapons Guidance Laboratory.
This time, the laboratory felt that the contractor had "overestimated the scope of the program," "lost sight of objectives," "provided for unjustified duplication of effort," and had shown neither an adequate management concert nor adequate accounting for risk in the development process. In addition, it all seemed too costly; IBM estimated a total expenditure of some $155 million.40 While the proposal was reviewed and discussed at Wright Field, the 110A program began to feel the effects of new fiscal strictures. Air Force headquarters decided in July to have only one weapon system contractor. By september 195[ILL], North American and Boeing were practically out of money and had been authorized to continue only "on a sustaining basis."
On 25 October, the Pentagon decided to end both study contracts and four days later announced a decision to reduce the bombing and navigation efforts as well. At this juncture, the entire weapon system program schedule was delayed 14 months, changing the availability date for the first aircraft from April 1960 to June 1961.
The bombing and navigation system contractor was advised of the new fiscal and timing decisions, submitted a revised proposal on 15 October, and then discussed the matter 16 days later in Dayton.
Now that more time was available, all parties agreed that it would be possible to include advanced developments in the system.
International Business Machines was ready to expand the scope of the original development effort, especially with regard to definite inclusion of a "High Resolution Side Looking Radar."41
Then in January 1957 the Pentagon approved another 11 month schedule retardation, pushing the first flight to May 1962; this meant an initial SAC aircraft in November 1964, and the first wing in March 1965. The IBM effort received the title "Bombing-Navigation-Missile Guidance System AN/ASQ-28(V)," but the exact character of this program, and indeed of the entire 110A system was uncertain. On 25 February, IBM was told to hold the work to the level reached by 1 March 1957.
The contemporary program instability was to persist until the end of May 1957. However, if the earlier 14 months slippage had permitted the inclusion of more advanced devices in the bombing and navigation system, it was also certain that another eleven months could be used for the same purpose. The problem was to define just how far, and with what objectives in mind, the Air Force wanted to go. The door was now open.
The Weapons Guidance Laboratory was engrossed in planning the new and expanded approach in February and hoped to have the contractor contractually bound by 1 July 1957. For a variety of reasons, the laboratory now favored inclusion of a side-looking radar and expressed particular interest in the possibility of obtaining higher resolution. In addition, the laboratory felt that the General Precision Laboratories' doppler program should be reviewed to determine whether it would be better to substitute the contractor's proposed System 14 doppler radar for the APN-96. Advanced computer and anti-jamming concepts were also scheduled for additional investigation.
After preliminary talks with the contractor, WADC, AMC, and project office personnel attended a full scale, five-day IBM presentation in Dayton late in March. This was followed by a review of a contractor-written proposal in April. By the end of May the future of the 110A bombing and navigation development had been settled.
The bombing and navigation system for the 110A weapon system thereafter took on a dual aspect. There was agreement that the contractor should proceed essentially along the lines laid down before the schedule slippages, with whatever equipment was at hand, to prove the feasibility of the system concept. His prime task was to determine just how well a digital computer would function with a stellar-inertial system and doppler and search radars. Concurrently the contractor was to undertake development of an advanced system based on the same concepts but employing as many improved and new components as possible. The feasibility systems wore dubbed experimental models. The advance systems were called the engineering models; there were also to be development models to support aspects of the engineering model effort. The first of the two planned experimental models (one was for flight and the other for bench testing) was to be bench tested by March 1958. This was to be followed in July 1958 by installation of an experimental model for flight tests, which were to begin by December 1958. By this time, the contractor was also to have completed an investigation of the feasibility of including missile guidance in the system. One and a half years later, in July 1960, the first engineering model was to have entered the bench test phase. Flight testing of an engineering model was scheduled to start by February 1961 and the first production system was to be flown by May 1963.
The experimental models were to incorporate the N2C stellar-inertial platform, the Dinaboc computer, the APN-96, and the High Speed Bombing Radar, with attendant parts and displays. The engineering models were to incorporate an improved stellar-inertial platform (probably something from North American), the System 14 doppler radar, high resolution forward-looking and side-looking radars, a faster and higher-capacity computer, some type of map matching or comparison device, and improved anti-jamming features probably based on the addition of a travelling wave tube to the radars.
One aspect of the computer improvement area was already under investigation. The Dinaboc used germanium switching to get higher speeds, but faced an environmental temperature problem. The Weapons Guidance Laboratory asked IBM to investigate silicon switching to meet the temperature requirements and to consider introducing improved logical techniques to compensate for the lower speeds resulting from use of silicon switches.43
In the midst of tho contractor proposal review in April 1957, Wright Field received notice of a policy which firmly established guidelines for all future bombing and navigation system programs in the Air Force. Originating in the Office of the Deputy Chief of Staff for Development in the Pentagon, the message advised ARDC and AMC that "after careful consideration, it has been decided that wo will no longer develop new BNMGs for each new airplane added to the inventory, especially in view of the present and anticipated limited budgets, together with the missile-manned airplane force structure." Progress thereafter was to be based on the design of systems with "black box growth" potential. Air Force headquarters felt the IBM program was the best bet for continual development of a system which could be available for any new bomber, so the message ordered that the program "be oriented toward development of an advanced type BNMG System that will have sufficient growth potential to be used in any strategic bombardment weapon system built for the 1965-1970 time period." IBM was also to work closely with contractors already developing components that might be valuable additions to the 110A system. All the eggs were now going into one basket.44
However, it was also apparent, that the Bombing and navigation devices for the 110A system were not yet fully defined. In April, the Weapons Guidance Laboratory published a description of the redirected program and in May the contractor published a brochure outlining his plans. Both documents indicated that no decision had been made on the computer and the stellar-inertial platform.
The laboratory wanted IBM to overcome the excessive weight and the cooling problems of the Binaboc computer by investigating "other forms of general computer logic" and demonstrating a "flyable digital computer utilizing silicon transistors and diodes and high temperature components," Furthermore, the laboratory wanted an improved console display including some kind of comparison feature to aid in target identification, but was not ready to authorize inclusion of IBM's ASB-4 Topographical Comparator until the efficiency and capacity of that device was proved. A study of map matching features for platform erection was also authorized.
The contractor was planning for addition of the System 14, doppler radar with growth features to permit operation from aircraft flying at 2,500 knots and 100,000 feet and incorporating "very narrow pencil beamed widths which are extremely difficult to jam and detect . . .;" the laboratory desired "pencil beam illumination of the ground" from the high resolution radars.
The IBM report stated that the side-looking radar was to provide high resolution and
a capability of looking perpendicular to either side of the longitudinal axis of the aircraft for a range of 50 miles. Within this 50 mile strip, a ten mile diameter section may be magnified and displayed to attain a 0.05° beamwidth and a 200 feet resolution.
The search radar was to operate at a range of 250 nautical miles with a tracking capacity of 125 nautical miles. A 125-mile tracking range was "compatible with the tine required for an operator to detect a target and synchronise his crosshairs at the higher speed and altitude of this system."
IBM also outlined a typical mission. Assuming a flight path at Mach 3 and 70,000 foot altitude, the contractor felt the aircraft could, while well outside a 30 mile circle surrounding the target, use the side-looking radar (with a range of 50 nautical miles) to map a ten mile square area containing the target. Twenty seconds after mapping had ceased the target would have been identified and the bombing crosshairs properly placed. At once the aircraft would start a programmed turn and the bomb would be released at the proper time, some 95 miles and 3.2 minutes after crosshair placement.45
As the remainder of 1957 passed with no further schedule or program changes, IBM began to achieve some definite results. For the 110A system, much of this time was consumed by preparations for the selection of one weapon system contractor. A source selection board was created, evaluated both contractors' plans, and on 27 November presented its views to the commanders of AMC, ARDC and SAC. A presentation to the Air Staff followed on 11 December; the Air Council was briefed on 12 December, and Secretary of the Air Force Douglas on the following day. Before the month had passed North American had been selected as the system 110A contractor and in January 1958 the project office began preparing plans for an accelerated program. IBM's opinions on this matter were solicited and received. The accelerated program was also presented to AMC, ARDC, SAC, and high Air Force officials in Washington, and by 24 January the project office could report that it had received "unconfirmed concurrence" of an 18 month 110A program acceleration. On 14 February the 110A became the B-70.46
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
On 7 January, ARDC had published a comprehensive statement of the weapon system requirements for the 110A program, This document outlined the current strategic bombardment concepts within which the 110A was to function. It stated that although "surface-to-surface strategic bombardment missiles having widely different characteristics will be operational as a vital portion of the strategic air arm" during the operational period of the 110A:
total dependence cannot be placed on these missiles' due to uncertainties about their reliability, accuracy, flexibility of employment and relative immobility. Because of these uncertainties, the use of missiles in this era will be limited, initially at least, to unhardened accurately located targets. Such targets comprise only part of the strategic target system.
A second part of the strategic target system is composed of smaller targets, some of which may be hardened to the extent they can be destroyed only be accurate bombing with high yield weapons. For such tasks, a manned bomber is the only known system possessing the needed and proven capabilities. In addition, man provides discretionary capabilities for target discrimination, malfunction correction or override, timely evasive maneuvers and judgement in selection and employment of penetration aids. These attributes, coupled with bomber flexibility of employment . . . are important, considerations to the probability of success in a strategic campaign.
A third important consideration requiring continued use of manned weapon systems is related to the roles which the strategic air arm may be expected to play and kinds of wars in which it may be employed. In some instances, physically demonstrable presence may be sufficient. Alternately it may be required to engage in a major conflict or in a limited war which may also mean limited weapons. Such possibilities attach decisive importance to the economic values associated with heavy payloads, high accuracies, recallability and, in particular, recoverability.
The system designed with these elements in mind, the B-70, was to operate from home bases in the United States and be capable of operating anywhere in the world, in all types of weather, under the alert concept. A 90-percent reliability factor became a requirement. Minimum unrefuelled range was to be 6,000 nautical miles, with 11,000 desired. Bombing altitude was to be 60,000 feet, with 75,000 or higher desired and with a low altitude (500 feet or lower) bombing capacity also required. (This was the first time such a capacity had been declared essential.) At high altitudes, a Mach 3 speed was necessary, with at least Mach .9 for low altitudes.
The An/ASB-28(V), as described at this juncture, was to perform direct and offset aiming, bombing from reference points at least 200 nautical miles from the target, crosshair tracking at approximately 100 nautical miles or more, and high resolution radar terminal sighting giving 250 feet resolution or better at a slant range of 50 nautical miles. With "optimum offset aiming." bombs were to he delivered within 1,500 feet of the target, with 3,500 feet as a maximum acceptable figure. For direct aiming and the 200 nautical mile reference point run, 2,500 feet was desirable and 5,000 feet was acceptable,
Eight weapon loads were listed for the bomber. On its "design mission," a 10,000 pound thermonuclear bomb was required. For alternate loads, two 10,000 pound thermonuclear bombs were specified. Other possibilities included multiple units of small bombs, one ten-thousand pounder and small bombs, one twenty-thousand pounder, the air-to-surface missile and a 10,000-pound thermonuclear bomb or assorted smaller bombs, two missiles and one small bomb, or biological and chemical warfare weapons. The missile was to have a range of 300 nautical miles, but 700 nautical miles or more was desired. Its speed was to be Mach 5, or better if possible, and accuracy was proscribed as a strike within 5,000 feet of the target for a launch 300 nautical miles from the target, with a 2,500 foot accuracy figure, or less, desired.47
Exactly two months later, on 7 March 1958, a revised General Operational Requirement for the B-70 was published by Air Force headquarters. A tentative operational concept appeared six days thereafter. The latter document called for the first test aircraft in January 1962, the first inventory aircraft in SAC by October 1963, and the 45th inventory aircraft in August 1964.
[The scheduling fluctuation was as follows:
First Flight Test A/C
First Inventory A/C
SR-22 Amend. I
7 March 1953
It also raised the unrefueled range requirement by 500 nautical miles (to 6,500), raised cruise altitude to 80,000 feet, and authorized bombing between 500 and 80,000 feet. A revision in April 1958 called for an unrefueled range of 6,873 nautical miles. Where the initial concept required the missile "concurrently with introduction of the aircraft into the inventory," the later paper added that this development was not to "delay the progress or procurement of the basic aircraft."48
The appearance of the 7 March GOR was attended by publication of North American's "Description of Technical Development Program for Weapon System 110A Bombing-Navigation Missile Guidance Subsystem," which reflected much of the analyses at IBM and Wright Field of possible configurations for the ASQ-28. As weapon system contractor, North American assumed cognizance over the IBM program and all other component developments for the B-70. Wright Field agreed that the North American description was to serve as the standard for a new IBM work statement, to be completed by 1 April 1950.
North American outlined the growth requirements of the system (operation at Mach 3.5 and 110,000 feet), suggested that IBM delete the pencil beam radar study, informed IBM that the radar sighting equipment (with antenna, receiver, transmitter, and doppler data processor) should be "similar in performance to the SABRE, AN/APS-75, which "employs pulse-to-pulse frequency tuning with X-band, modified for coherent high-resolution, side-looking operation," and decided that the Goodyear Range Gated Filter Processor would be functionally compatible with the doppler data processor.
The SABRE was a very promising development sponsored by WADC at the General Electric Company. Its introduction, together with the Goodyear equipment, promised to provide for the B-70 bombing and navigation system a sophisticated side-looking radar capable of extremely high resolution and with the anti-jamming features offered by a travelling wave tube (pulse-to-pulse tuning) for the forward-looking mode.
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
The Advantages of Sidelooking Radar
North American wanted a platform "equivalent in performance to the N2J" (a modification of the N2C proposed by the company's Autonetics Division), and a doppler radar similar to the System 14 (now termed the APN-115). Work in map matching was authorized, to determine the "feasibility of providing an automatic map-match capability . . .," as were studies (by IBM) aimed at employing improved radar resolution techniques, data recording, display equipment, and target aids.
The computer portion of the system was also to be subjected to analysis. The goal was "increased utilization of the inherent flexibility of a digital computer to increase the reliability, maintainability and effectiveness of the system ..."
North American allocated 650 hours of RC-121 flight time to IBM for experimental and engineering models and tests. The B-70 contractor also wanted a T-29 for flight tests by Autonetics, a C-54 and B-57 (or equivalent) for General Precision Laboratories' doppler tests, a B-47 or B-66 for General Electric SABRE tests, and a JB-29 for Goodyear.
With some reservations, WADC was in general agreement with North American's views. However, the center did not want to restrict the map-matching studies to the console area (the matching unit could turn out to be part of the radar sighting equipment). WADC also felt that North American should impose stricter controls over IBM, remarking that under the existent plan the bombing and navigation system contractor could of his own volition "initiate work without prior approval from NAA.49
On 1 May 1958, the IBM contract with the Air Force was terminated. North American and IBM immediately began preparing a new contract to bind both parties under the weapon system contractor principle. The contract was to be consumated by September 1953. However, before this could be accomplished, the Air Force had to overcome heavy criticism of its development plans from the Office of Electronics in the Office of the Secretary of Defense. The electronics office had received an Air Force description of the proposed B-70 bombing and navigation system and was far from satisfied, Most of the criticism, which became known in March 1953, was directed at radar developments, especially the side-looking radar. The electronics office felt that, even with "optimistic assumptions," this device would not function beyond 35 nautical miles and would be severely limited by "second-time-around echoes," adding that elimination of the echo problem was "problematical." The agency also expressed doubt about the worth of testing the radar portion of the system in the subsonic B-47 and questioned the probability of developing devices for low level operations when failure could mean later need for a "very costly redesign." In the view of the Pentagon office, the estimated cost of seme $200 million was excessive. The electronics office also suggested a "searching look by technically qualified electronic and aeronautic people to make sure that the proper engineering compromises between electronic and aeronautic performance are being made."50
Lieutenant General S. E. Anderson, ARDC commander, answered on 12 September. Conceding the validity of "some of the questions" posed by the Office of electronics, General Anderson emphasised that the primary targets for the B-70 were such "hard targets" as missile launching cites and that the best way to achieve the radar resolution required for such a strike was with the "side-look doppler processing approach." This method had limitations, he said, but "there is no other more promising avenue available immediately and this technique offers growth potential to a "squint" mode of operation which would approach the desired forward look high resolution capability." (The "squint mode" involved sighting 22.5 degrees off the direct forward look.)
General Anderson noted that a range of 38 nautical miles was currently attainable with the side-looking radar, that this was acceptable as a means of establishing check points when employed with the data available from the inertial system, and tliat ARDC was taking every step possible to increase range. One promising method employed "pulse tagging" techniques whore each pulse was identified and the second-time-around signals could be eliminated. ARDC was also investigating modifications to the antenna to obtain similar results. General Anderson added that the target identification capacity resulting from use of the side-looking radar was a valuable side benefit.
In replying to the criticism of flight test plans, General Anderson held that initial tests could very well be accomplished in the B-47 and that a B-58 would be used when necessary. As for the low level functions, he pointed to the fact that this factor had only become important "in a hardware sense" after publication of the revised General Operational Requirement of March 1953 and that studies were underway to determine just how such a requirement could be satisfied.
(PICTURE NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
The quoted cost figure, he said was for the entire system and could not be appreciably reduced by a decision to "up-grade current equipments to meet the requirements of the B-70 program. . . ." Furthermore, he added, the bombing and navigation system had a "growth capability equivalent to that of the air vehicle in both performance and time" and provided "significant increases in performance, flexibility, and reliability over any equipments currently available or under active development."
General Anderson concluded by advising the Pentagon that the program was "under constant evaluation by the WSPO with the assistance of the most highly qualified technical teams available within ARDC." Still, he did not deny that the program was "faced with many technical problems, the immediate optimum solutions to which are not now available," and welcomed "any constructive comments as to how these technical problems can be overcome." Thus the matter seemed settled.51
At this juncture, component development was well under way. Tests of the APN-115 had just begun and the contractor was trying to use the facilities at Cape Canaveral for additional flights. Finding this impossible he began investigating facilities at the Air Force Flight Test Center at Edwards Air Force Base, California, the Navy Point Mugu range in California, and the Air Force Proving Ground Center at Eglin Air Force Base, in Florida. International Business Machines was hard at work on improvements to the Dinaboc and expected to test an experimental model late in October. It was at this point that the immense complexity of the computer and its relatively low reliability induced IBM and the Weapons Guidance Laboratory to consider seriously the installation of a secondary computer for emergency use.
Goodyear Aircraft Corporation was making progress in the development of a high resolution radar data processor and happily announced that the component count had been reduced from 12,000 to 8,000. General Electric now expected to begin flight tests of the SABRE radar in June 1959, and North American had begun tests of the N2C in a T-29.52
But the gains in development were to be subjected once again to critical analysis from Washington, though from a new viewpoint. The Office of Electronics had been concerned with technical factors during the March episode; the criticism which developed later in 1958 was a product of the uneven fiscal policies which had plagued the B-70 program. The process began, or was renewed, when Air Force headquarters on 27 November informed ARDC headquarters of "concern with the manner in which the B-70 program is developing," and of a Pentagon feeling that "there appear to be many areas in which time and money may be saved through judicious use of currently available equipments, batter utilization of the flight test inventory, and reduced over-all complexity of subsystems." On this basis, the Pentagon directed investigation to determine among other things, the feasibility of reducing "the sophistication of the bombing and navigation subsystem" and the expanse of the flight test program. A summary of the findings was to be presented to the Air Council, the Weapons Board, and the Strategic Air Panel.53
In response to this message, IBM and North American posed six hypothetical bombing and navigation systems to be reviewed in terms of cost and performance, with a system comprising only existent components at one end of the spectrum and the B-70 system (as the most sophisticated) at the other. In January 1959, the resultant analysis was discussed thoroughly at North American by representatives of both firms, project officials, command headquarters personnel, experts from WADC, and spokesmen from the Strategic Air Command. All parties agreed that in the ASQ-28 the Air Force was getting the most for its money. These opinions and supporting data were presented at the Pentagon and were approved. However, one month earlier Air Force headquarters had established an aircraft acceptance program which affirmed that an "equippage date of August 1965 will be used for planning purposes" in the B-70 program.
[This was one year later than the 7 March 1958 schedule previously current.]
The immediate result of these events was a slight compression of the flight test program.
As of 1 March 1959, the flight teat program consisted of Goodyear tests with a JB-29 (which were about ready to begin), General Precision Laboratories' tests in a C-54 (already underway) and a JRB-57 (to begin at Eglin in mid-1959), SABRE tests in an RB-66 (**) (to begin soon), and the IBM experimental system test program in the RC-121, which had started late in 1958.
[The B-47 had earlier been eliminated from the General Electric program.]
North American's T-29 flight tests had ended in January 1959. Stellar-inertial, doppler, and radar sighting tests were scheduled for the B-58, beginning with the first two components early in 1960.
With the "time and money" exercise out of the way by the end of January 1959, the bombing and navigation development program for the B-70 had clear sailing for some ten months. However, technical problems arose in the course of the flight test program, particularly in the case of the APN-115 and the N2C (now officially termed, for the ASQ-28, the N2J). On 12 January 1959, North American proposed substitution of the N3B stellar-inertial system for the N2J. The N3B was based on the same principles as its sibling but was half the size, was capable of more precise operation, and included dual two-degree-of-freedom gyroscopes, as against three single-degree-of-freedom gyroscopes for the N2J. Motor design changes in the N3B and the use of beryllium and titanium in the rotor promised a further reduction of the gyro drift rate. The N2J drift rate was great enough to require very careful calibration on the ground if in-flight alignment was to be accomplished; the N3B "could be aligned very accurately [on the ground] on a continual basis during alert status. ..."
North American had scheduled flight tests for November 1959 to prove the feasibility of substituting the N3B. Computer compatibility was a key question, but it was also evident that development of the device would take from 12 to 18 months. The Weapons Guidance Laboratory, not as confident of the outcome as the contractor, insisted on evaluating flight test data for the related N3A before agreeing to consider the N3B. The laboratory took immediate steps to insure the availability of a T-29 for the tests, proposing to complete the evaluation by December 1959.55
On 1 May 1959 the laboratory decided that insufficient time was available for completion of the N3A tests and decided to approve the N3B for the B-70 and cancel all work on the N2J. Three other factors were also involved in this decision. First, key components of the N3A/N3B were planned for the Minuteman (Weapon System 133A) solid propellant intercontinental ballistic missile, and the laboratory believed that with funds from both weapon system programs to support the effort, and the benefit to accrue to the Air Force from the use of an essentially common component, the risk in substituting the newer device was worth taking. Second. North American was subjecting the device to extensive ground tests. Finally, North American was planning to have 95 percent of the design of the B-70 officially approved by December 1960 and it was necessary to establish the components of the bombing and navigation system as early as possible. These factors led to project office approval.
The project office in June 1959, after reviewing a North American development testing schedule for determining stellar-inertial and doppler performance, disapproved both the contractor's plan and the requirement for a B-58. The project office felt that the data collected under the testing procedures could not be applied with validity to the B-70, and that the program was planned for too late a date. Specifically, the project office advised the contractor that "early flight test of a completely integrated system was mandatory in order to provide effective and timely engineering data for production design."
NOT INCLUDED, DUE TO POOR MICROFILM QUALITY)
Location of B-70 Offensive Subsystem
Moreover, the project office did not feel that the contractors' plan to use one B-70 for these tests was acceptable and asked North American "to submit a reoriented BNMGS development schedule" which stressed the need for a production prototype "to provide the earliest possible integrated system flight test." The initial portions of these tests were to use a subsonic vehicle; two B-70's would participate at a later date.57
Another problem was the state of the doppler radar program. Doppler flight tests in the C-54 had demonstrated the new techniques employed in the APN-115, but development was slow. General Precision Laboratories was taking "some drastic steps to prevent further slippages in the program," and in July was planning to test a transistorized version. However, during the next three months little progress was evident and the effort received careful scrutiny by the Air Force, IBM and the weapon system contractor. At this point the Eglin tests had been postponed until December 1959.58
Throughout the summer of 1959, fiscal inhibitions continued to influence the B-70 program. This was the case with the requirement for a low altitude capability. North American submitted a proposal for terrain clearance equipment and related items in July, but on 5 August the project office informed the contractor that "in view of the current austerity budget requirement ... proposals to implement a low altitude capability for the B-70" would have to be deferred.59
Some three months later, the Air Force decided to compress the development program—and cut costs—by freezing system design as early as possible. The time was ripe for such a decision in the case of the bombing-navigation system—the ASQ-28—because it had been "developed to the point that the design concepts are clear and design techniques have been defined." But this was not to mean that changes necessary to meet performance specifications or altered Air Force requirements, or resulting from deficiencies revealed through flight tests, were to be prohibited. In November 1959, the date for a fixed configuration production prototype was set at 1 February 1960.60
Unfortunately for such plans, it was also in November 1959 that the Air Force cancelled all plans for producing an operational B-70 and reduced the program to a one-airplane experimental-flight status. On 27 November the Pentagon informed ARDC headquarters that only one B-70 "for earliest flight test possible under budget limitations" was to remain in the program and that the contracts for the development of defensive, mission and traffic control, and bombing and navigation systems for the B-70 were to be terminated. The B-70 vehicle was reclassified the XB-70.61
Command headquarters officially advised Wright Field of these circumstances on 1 December 1959, adding that it would be well for the project office to identify which component development portions of the weapon system program were "vital to the USAF." By 11 December, termination notices to the component contractors, including IBM, were in the mails.62
But the Weapons Guidance Laboratory and weapon system project office were most reluctant to sacrifice all the valuable work that had gone into the ASQ-28. The project office asked the laboratory to try and salvage as much of the ASQ-28 component developments as possible. Then on 15 December 1959 IBM presented a complete resume of progress to date and proposals for continuation.
International Business Machines, after first outlining the configuration of the ASQ-28 and achievements thus far (*), stated that the ASQ-28 program had proved the feasibility of accurate digital computation in bombing and navigation, the employment of a pulse-to-pulse tunable radar, high resolution sighting, the use of advanced doppler techniques, precise stellar-inertial navigation, the missile launch capability, and the efficacy of the man-machine relationship.
[At this juncture, IBM tests in the RC-121 of the experimental system were providing important data on the compatibility of components and the contractor had fabricated the first model of the secondary computer. General Electric was also in the midst of B-66 tests with the SABRE radar; results were thus far satisfactory. APN-115 testing was not so satisfactory, but the transistorized version of the set was expected soon. Goodyear flight tests of the Range Gated Doppler Data Processor for the side-looking radar had demonstrated azimuth resolution of better than 250 feet at 20 miles. North American had completed, in the T-29, 94 flights totalling 342 hours in tests of the stellar-inertial platform, doppler damping, and missile azimuth erection. The RC-121 flights now numbered 58 and 216 hours, and included demonstration of "full system marriage with [the] digital computer," radar tracking and fixing, doppler and doppler inertial techniques, star acquisition, and long range navigation. In the B-66, General Electric had. completed 20 flights for 45 hours to test travelling wave tubes, pulse-to-pulse tuned radar mapping, and range and ground mapping, Goodyear, in the modified B-29 and a C-97, had completed 40 flights lasting 183 hours to evaluate the doppler processor and high resolution mapping. Finally, General Precision Laboratories' tests in the C-54 and B-57 included 46 flights for 225 hours of doppler radar analysis.]
IBM recommended that the Air Force complete one prototype and continue work on system integration and the reliability program. Specifically, the contractor stated that if approval for continuation were received by 1 February 1960, a functional model of the system could be made available by the end of 1961. A development model for flight test, additional development to add improved techniques, and additional tests of the functional model to "incorporate advanced technology consistent with the schedule" could be provided by the end of 1962 if approval was forthcoming by 1 August 1960. Approval by 1 August 1962 would mean that as of the end of 1964 the scope of the program could be increased "to include providing prototype models for laboratory, flight, and experimental test," and this would complete the "production engineering of the system." Finally, should the Air Force agree by 1 August 1962, an initial production system could be provided within twenty eight months.63
The next stop was to secure approval—and money—from Air Force headquarters. By 16 December r:59, Air Force headquarters had agreed to permit continuation of the development of components for the ASQ-28, but had added that "new" R&D funds must be found for commitments to support ASQ-28 fall-out efforts. ..." This was not enough; the project office urged the authorization of money to fund development. The office felt that "a serious gap in the overall Department of Defense development program now exists" and took every possible measure to see that the program was not permitted to suffocate before Air Force headquarters had another chance to investigate the full consequences of the termination decision.64
The door opened a little 12 days later. On 30 December 1959 Lieutenant General B.A. Schriever, ARDC Commander, had witnessed a B-70 presentation prepared by the project office and had agreed to "a special briefing on the bombing-navigation subsystem." The briefing was to cover
a comparison of the design, cost capabilities, status and development process (if applicable) of the IBM systems for the B-52 and B-70 and the Sperry system for the B-58, Other potential systems derived from these systems or from existing guidance system programs that may have potential application to the B-70 should be included.
The briefing should stress use of common equipment and/or elimination of 'gold plating' to provide a suitable system for the B-70 at a minimum cost.
At Wright field, the B-70 project office chief, Colonel E.L. Bishop, was assigned the task of preparing the briefing, with full support from the WADD laboratory complex.65
[On 15 December 1959, Wright Air Development Center and the ARDC Directorate of Systems Management had formally been merged, the resultant organisation being the Wright Air Development Division—WADD.]
The presentation led ARDC to secure $5 million on 26 January 1960 for an "Experimental Bomb-Nav System for the XB-70," but this did not include the high resolution side-looking radar and certain console display features.
[Including map matching, the map indicator, and the in-flight printer.]
These were deferred pending allocation of additional funds.
Command headquarters officially directed implementation of the funding measure on 2 February 1960, adding another million dollars on 26 February and also announcing that a requirement for $15 million more was being considered. Still, the XB-70 was to have "no offensive or defensive capability," only the potential to accept a bombing and navigation system. First flight was now scheduled for December 1962.66
Thus the IBM program was enabled to continue. It represented the most promising bombing-navigation system development effort extant and—by virtue of the Air Force decision of April 1957--the only bombing-navigation program that could bear fruit in the 1965-1970 time period. Wright Field had good reason to contend that it must not be eliminated.
1. ARDC Dev. Directive 00034, 26 Feb. 1952, High Altitude Strategic Bomber Reconnaissance weapons System, in B-58 Bombing Sys. Sect.., WGL (hereafter cited as B-58 WGL) files; WAX Tech. directive 52-14, 7 May 1952, Strategic Bomber/Reconnaissance Weapons Development Program, in B-58 WGL files.
2. USAF Gen. Opl. Req. SAB-52-1, 1 Sept. 1952, Strategic Bombardment System, in B-58 WGL files.
3. Chart, History of the Hustler Airplane, Oct. 1946 - Aug. 1952, prep, by Consolidated Vultee Aircraft Corp., in HBF; RDB 1A, Proj. R-426-256, 29 Feb. 1952 and 13 Oct.1952; RDB Proj. Card (hereafter cited as DD 613), Proj. R-426-276, 21 Mar. 1953; DD 613, Proj. 5036, 1 Oct. 1953, in HBF.
4. DD 613, Proj. 5036, 1 Oct. 1953; RDB 1A. Proj. R-426-276, 29 Feb. 1952; Proj. Status dpt., MX-1964, (MX-1626), 15 Nov. 1952, in HBF; Chart, History of the Hustler Airplane, Oct. 1946 -Aug. 1952; Proj. Status Rpt., MX-1964, 15 Jan 1953.
5. Rpt., FZP-4-009, 31 Dec. 1952, Navigation-Bombing and Missile Guidance System for the MX-1964, prep, by Convair, in B-58 WGL files.
6. WADC R&D Proj. Information Rpt. (hereafter cited as WADC R&D PIR), RDO 448-110, 2 Feb. 1953; Proj. Status Rpt., MX-1964 14 Mar. 1953; WADC R&D PIR, RDO 448-110, 20 Apr. 1953; DD 613, Proj. 5036, 1 Oct. 1953; Military Characteristics No. 345 for the B-58 High Altitude Strategic Bombardment Weapon System, SAB-53-A1, 11 Sept. 1953, in B-58 WGL files.
7. ARDC R&D Mgmt. Rpt. (hereafter cited as ARDC 111) Proj. 5036, 9 Feb. 1955, in HBF; History of Wright Air Development Center, 1 January - 30 June 1955, II. 295; DD 613 Proj. 5036, in B-58 WGL files.
8. ARDC 111, Proj. 5036, 27 May 1955, in B-58 WGL files.
9. ARDC R&D Proj. Plan (hereafter cited as ARDC 100), Proj. 4086, 25 Oct. 1955; ARDC 111, Proj. 5036, Nov. 1955, in B-58 WGL files.
10. ARDC 111, Sys. 102A and 102L, 5 Dec. 1955; DD 613, Proj. 5036, 9 Jan. 1956, in B-58 WGL files.
11. ARDC 111, Proj. 5036, 7 Nov. 1956; Model Spec, USAF Model YB/RB-58A, Convair Model 4, Strategic Bombardment/Reconnaissance Airplane, 31 Dec. 1956, prep, by Convair; DD 613, Proj. 5036, 31 Dec. 1956, in B-58 WGL files.
12. Rpt., FZA-4-269, 15 Mar. 1957, B-58 Model Improvement Program, Primary Navigation System, in B-58 WGL files; ARDC Dir/Sys. Mgmt,, Weekly Activity Rpt. (hereafter cited as DSM WAR), 23 Aug. 1957; interview, W. A. Dynes, Strat. Bombing Br., WGL by Gary P. Baden, Mist. Br., WADC, 3 Jan. 1953.
13. DSM WAR, 4 Apr. and 11 Apr. 1958.
14. DSM WAR, 2 May and 16 May 1958; Semiannual Hist., B/RB-58 Weapon System, 1 January thru 30 June 1959, prep, by B-58 WSPO, DSM, ARDC, (hereafter cited as B-58 Hist., Jan-Jun 1959), in HBF.
15. DSM WAR, 23 May, 29 May, and 2 June 1953.
16. DSM WAR, 3 July, 11 July, 18 July and 25 July 1958.
17. B-58 Hist., Jan-Jun 1959.
18. DSM WAR, 5 Sep., 19 Sep., 3 Oct., 17 Oct., 31 Oct., and 21 Nov. 1958.
19. B-58 Hist., Jan-Jun 1959.
20. Ibid., Semi-Annual Hist., B-58 Weapon System, 1 July thru 31 December 1959, prop, by B-58 WSPO, DSM, ARDC.
21. Memo rpt., TSEPL-2-556-981, 15 Oct. 1945, subj.: Stellar Bombing,, prep, by Arm. Eng. Div,, ATSC, in HBF.
22. Eng. Div., R&D Progress Rpt,, July 1946; quarterly rpt,, AAF Tech. Com., meeting of 29 April 1947; periodical rpt., AAF Tech. Conn., meeting of 30 July 1947; in HBF.
23. Presn., Strategic Bombing Operations, by R. J. Bordlund, Arm. Lab., ADF, to Fire Control Panel, Ord. Conn. RDM, 15 May 1951 (heroafter cited as Nordlund Presentation): ARDC Form 82, Proj. R-554-311, 2) Sept. 1952; WADC Weekly Ops. Rpt,, 5 Mar. 1953; ARDC 111, Proj. 5102, 21 Feb. 1955 in HBF.
24. ARDC 100, Proj. 4431, 31 May 1955, in HBF.
25. ADF MPR, April 1951; DD 613, Proj. R-556-416, 28 April 1953; WADC Weekly Technical Information Report, 3 July 1953, in HBF.
26. Ltr., Brig. Gen A. A. Kessler, Jr., Dir/Proc & Industrial Plng., DCS/M, USAF, to CG, AMC, 7 Nov. 1949, subj.: Stellar Inertial Bombing Systems; RDB 1A, R-556-372, 3 Dec. 1952; Nordlund Presentation; ADF UFR, April 1951, in HBF.
27. WADC R&D PIR, June 1951, Aug. 1951; RDB 1A, Proj. R-556-372, 8 Dec. 1952; History of Wright Air Development Center, 1 July -1 December 1952, II, 387, in HBF.
28. DD 613, Proj. R-556-416, 28 April 1953, in HBF.
29. DD 613, Proj. 5042, 28 May 1954, in HBF.
30. ARDC 111, Proj. 5042, 18 Feb. 1955, in HBF.
31. ARDC 100, Proj. 4431, 13 May 1955; WAX R&D PIR, June 1951; interview, Maj. H. P. Warner, B-70 WSPO, DSM, by Gary P. Baden, Hist, Mr., MA DD, 7 July 1960; The Development of the Navaho Guided Missile, 1945-1953, by J. A. Neal, Hist. Br., WADC, Jan. 1956, pp. 40-47, in HBF.
32. ARDC System Requirement 22, 15 April 1955? WAX Exhibit WCLG-78 12 April 1955, subj.: Bombing Navigation System, in HBF; interview, T. M, Pienkowski, Bombing-Navigation Br., Elec. Office, B-70 biv., DSE, by Gary P. Baden, Hist. Br,, WADD, 26 May 1960, (hereafter cited as Pienkowski interview).
33. Pienkowski interview; WADC Exhibit WCLG-785, 12 April 1955, subj.: Bombing Navigation System, in HBF; DF, Col. N. P. Hays, Ch., Plans & ups. Office, WGL, WADC, to Ch., B-70 WSPO, Bomb. Ac. Div., DSM, ARDC, 25 Aug. 1955, subj.: Offensive and Defensive Subsystems for Weapon System 110A; ltr. Lt. Col. W. C. McLaughlin, Asst. Ch., 110A WSP0, Bomb. Ac. Div., DSM, ARDC, to NAA, 29 Aug. 1955, subj.: Bombing-Navigation-Missile Guidance Subsystem for System 110A, in B-70 Office, DSE files (hereafter cited as B-70 / DSE files) j DSC WAR, 28 March 1956; Amendment I to ARDC Systen Req. 22, 11 Oct. 1955, in HBF.
34. DSM WAR, 4 and 25 Aug., 1 Sept., 10 Nov., 7 Dec, 14 Dec., and 28 Dec. 1955.
35. Ltr., L.L. Waite, NAA, to DSM, ARDC, 5 Nov. 1955, subj.: Bombing-Navigation-Missile Guidance System for Weapon System 110A, in HBF; memo, Ch., Strat. Bombing Dr., to A. Zmeskal, Bomb. Sys. Sect., Plans 4 Ops. Office, WGL, WADC, 16 Nov 1955, subj.: Suborder 55RDZ-SBL-55-140, in B-70/DSE files.
36. WADC Exhibit, WCLG-881, 15 Dec. 1955, subj.: Bombing-Navigation and Missile Guidance Subsystem for Weapon System 110A/L, in HBF.
37. DSM WAR, 14 and 21 Mar. 1956; WADC Form 49, Suborder, 23 March 1956, subj.: System 110A, in B-70/DSE files.
38. DSM WAR, 4 and 18 April, 2 May 1956; nemo., Col L. J. Israel, Ch., WGL, WAGD, to Ch., 110A WSPO, Bomb. Ac. Div., DSM, ARDC,
29 Mar. 1956, subj.: Suborder Mo. RDZSBL-56-259; memo.,..Col L.J. Israel, Ch., WGL, WADC to Ch., Arm. Dev. Sect., Arm. Dr., Equip. Div., Dir/Proc. & Prod., AMC, 12 April 1956, subj.: Evaluation of IBM Proposal for BN-MG for Weapon System 110A; memo., Col. L.J. Israel, Ch., WGL, WADC to Ch., Bombing Nav. Sys. Sect., Weap. Guid. Br., Aero. Equip. Div., Dir/Proc. & Prod., AMC, 7 May 1957, subj.: Redirection of AN/ASQ-28 (V) Program; contract AF33(600)-31315, in B-70/DSE files.
39. DSM WAR, 23 May 1956.
40. Memo., G. T. Fouse, Res. and Tech. Dev. Sect., Ops. Office, to Ch., WGL, WADC, 20 Aug. 1956, subj.: Evaluation of IBM Proposal for the 110A Berthing Navigation and Missile Guidance Subsystem, in D-70/DSE files.
41. DSM WAR, 11 July, 28 Sept., 25 Oct., 1 and 15 Nov. 1956; memo. Lt. Col. W. C. McLaughton, R&D Admin., 110A WSPO, Bomb. Ac. Div., DSM, ARDC, to Ch., Arm. Dev. Sect., Arm. Br., Aero Equip. Div., Dir/Proc. Prod., AMC, 4 Dec. 1956, subj.: AN/ASQ-28 (V).
42. TWX, RDZ-1, Cmdr., ARDC Dat. 1, to Cmdr., AR DC, 2 Jan. 1957 DBK WAR, 1 Mar. 1957; WGL Charts for Review of the ASQ-28 Program, 28 Feb. 1957; memo, Maj. Gen. H. M. Estes, DSM, to Dep. Cmdr./Weap. Sys., ARDC, 26 March 1957, subj.: Bombing-Navigation-Missile Guidance Subsystem for System 110A, all in HBF; WADC Form 49, Suborder, 5 Dec. 1957, subj.: Phase I Work Statement, Weapon System 110A, in B-70/DSE files.
43. Memo., Lt. Col. C. E. Riddle, Asst. Ch., Nav. Br., to Ch., Strat. Bombing dr., WGL, WADC, 10 April 1957, subj.: ASQ-28 BWMGS for 110A WS; memo., Col. L. J. Israel, Ch., WGL, WADC, to Ch., Bombing Nav. Sys. Sect., Weap. Guid. Br., Aero. Equip. Div., Dir/Proc. & Prod,, AMC, 7 May 1957, subj.: Redirection of AN/ASQ-28 (V) Program, contract AF-33(600)-31315, in B-70/ DSE files.
44. TWX, AFDDC 54338, Hq. USAF, to Cmdr., ARDC, 3 Apr. 1957, in HBF.
45. Rpt., IBM, Military Products Division, Airborne Computer Labs., 14 May 1957, subj.: Bombing Navigation Missile Guidance System AN/ASQ-28(V), in B-70/DSE files; memo, and 6 incls., Col. L. J. Israel, Ch., WGL, WADC, to Ch., 110A WSPO, Bomb. Ac. Div., DSM, ARDC, 25 Apr. 1957, subj.: Redirection of AN/ASQ-28 Program; contract AF33(600)-31315, in HBF.
46. DSH WAR, 26 July and 30 Dec. 1957, 10 and 31 Jan., 14 Feb. 1958; ARDC Weapon System Requirement, Strategic bombardment Weapon System, WS-110A, TAB A-C, 7 Jan. 1958, as revised 14 Mar. 1958, in B-70/DSE files.
47. ARDC Weapon System Requirements, Strategic Bombardment Weapon System, WS-110A, TAB A-C, 7 Jan. 1958, as revised 14 Mar. 1958, in B-70/DSE files.
48. Ltr. and incl., Col. J. Tarter, Ch., Bomb. Er., Ops. Control Div., Dir. Ops., DCS/Ops., USAF to SAC, AMC, ARDC, ATC, 13 Mar. 1958, subj.: Tentative (B-70) Operational Concept; ltr. and incls., Col. J. Tarter to SAC, AMC, ARDC, and ATC, 28 Apr. 1958, subj.: Preliminary (B-70) Operational Concept, in HBF.
49. NAA rpt., Description of Technical Development Program for Weapon System 110A Bombing-Navigation Missile Subsystem, revised 7 Mar. 1958; TWX, MCXA-3-18-E, AMC to AFPR, NAA, 20 Mar. 1958, in HBF; Peinkowski interview.
50. TWX, 05-16-01, Cmdr., ARDC, to Cmdr., ARDC Det. 1, 16 May 1958, in HBF; and Peinkowski interview.
51. Ltr., Lt. Gen. S. E. Anderson, Cmdr., ARDC, to Dir/R&D, DCS/D, USAF, 12 Sept. 1958, subj,: Comments on B-70 Bombing-Navigation System; and unsigned draft letter to Ch., Guid, and Weap. Div,, Dir./R&D, DCS/D, ISAF, undated, subj.: Comments on B-70 Bombing Navigation System, in HBF.
52. Draft Memo., [Aug. 1958] filed in B-70 WSPO; memo. Ch., Strat. Bombing Br., WGL, WADD to [unknown], 8 Aug. 1958, subj.: Trip Report ( for 31 July 1958 visit to IBM); ltr,, R. H. Kemp, NAA, to Cmdr., AMC, 10 Sept, 1958, subj.: Contract AF33(611)-36599, Report on Test Sites for the AN/APN-115 Doppler Radar Flight Test Program, in HBF.
53. TWX, AFDRD-51392, Hq. USAF to Cmdr,, ARDC, 20 Nov. 1958, In HBF.
54. Interview,. Maj. R. V. Walker, B-70 WSP0, DSM, by Gary P. Baden, Hist. Br., WADD, 8 June 1960; DSM WAR, 23 Jan. 1959: TWX, AFDDC-52125, Hq. USAF to Cmdr., AMC, 17 Dec 1958; NAA Chart, AN/ASQ-28 (V) Flight Test Phasing and Schedule Chart (Based on original C-1 Schedule), 1 Mar. 1959, in HBF.
55. Memo., M. W. McCabe, Asst. Ch., Inertial Guid. and Computer Br., Guid. Dev. Div., to Ch., Adv. Dev. Sect., Strat. Bomb.
Br., Offensive Sys. Div., WGL, WADC, subj.: B-70 Stellar Inertial Platform; memo, M. W. McCabe, Asst. Ch., Inertial Guid. and Computer Br., to Ch., Inertial Guid. and Computer Br., Guid. Dev. Div., WGL, WADC, 18 Mar. 1959, subj.: Supplemental Information, Form 56, dated 10 Nov. 1958, ID# WCLG-58-81; and memo, M. W. McCabe, Asst. Ch., Inertial Guid. and Computer Br., Guid, Dev. Div., WGL, WADC to Ch., B-70 WSPO, Bomb Ac. Div., DSM, ARDC, 13 May 1959, subj,: Optical Ground Alignment of Inertial Celestial Autonavigator for B-70 Aircraft, in HBF.
56. Memo, M. W. McCabe, Asst. Ch,, Inertial Guid. and Computer Br., Guid. Dev. Div., WGL, WADC, to Ch., B-70 WSPO, Bomb. Ac. Div., DSN, ARDC, 13 May 1959, subj.: Optical Ground Alignment of Inertial Celestial Autonavigator for B-70 Aircraft; DF, F. D. Banta, Ch., Inertial Guid. and Computer Br., Guid. Dev. Div., to Ch., Adv. Dev. Sect., Strat. Bombing Br., Offensive Sys. Div., WGL, WADD, 1 May 1959, subj.: Sub-Order B-70-59-104, Replacement of N2J with N3B, in HBF.
57. Ltr., Col. R. J. Meyer, Ch., B-70 WSPO, Dir/ Strat. Sys., ASC/AMC, to NAA, 29 June 1959, subj.: Proposal for BNMGS Test Bed Program in a B-58, Contract AF33(600)-36599; ltr., Col. R. J. Meyer, Ch., B-70 WSPO, Dir/Strat. Sys., ASC/AMC, to NAA, 8 June 1959, subj.: Contract AF33(600)-36599, BNMGS Development Schedule, in HBF.
58. Memo., Lt. Col. J. J. Jones, B-70 WSPO, Dep/Strat, Weap., DSM, ARDC, to Ch., Ops. Dev. Dir., Ops. Office, Dir/Labs., WADC, 29 Sept. 1959, subj.: Doppler Radar Development Survey; memo., Lt. Col. J. J. Jones, B-70 WSPO, Dep/Strat. Weap., DSM, ARDC, to Ch., Ops. Office, WGL, WADC, 6 Oct. 1959, subj.: MOPTAR Range Survey: memo., Capt. K. D, Hurley to Maj. Warner, B-70 WSFO, Dep/Strat. Weap., DSM, ARDC, 6 July 1959, subj.: Discussion of the APN-115 Program, in HBF.
59. TWX, LMSA-7-315-E, Cmdr., ASC/AMC, to NAA, 27 July 1959; TWX, MSA 1969, Cmdr., ASC/AMC, to AFPR, NAA, 5 Aug. 1959, in HBF.
60. Ltr., N. Silverston, Dep. Ch., B-70 WSPO, Dir/Strat. Sys., ASC/AKC, to NAA, 12 Oct. 1959, subj.: Contract AF33(600)-38669, Design Point Definitive for the ASQ-28, in HBF; DSM WAR, 22 Nov. 1959.
61. TWX, AFDRD-06-3, Hq. USAF to Cmdr. ARDC and Cmdr. AMC, 27 Nov. 1959, in HBF.
62. DSM WAR, 22 Nov., 11 Dec. 1959.
63. DSM WAR, 18 Dec. 1959; rpt., IBM, 15 Dec. 1959, subj.: AN/ASQ-28(V) Bombing Navigation System, in HBF.
64. DSM WAR, 18 Dec. 1959; memo, Lt. Col. J. J. Jones, B-70 WSPO, Dir/Strat. Weapons, DSM, ARBO, to Ch., Ops, Office, WGL, WADD, 22 Dec. 1959, subj.: Disposition of Assets of Terminated ASQ-28 Bomb-Nav Program, in HBF.
65. DSM WAR, 8 Jan. I960; memo, Col. J. C. Maxwell, C/S, to DCS/Res. and Eng., ARDC, 5 Jan. 1960, subj.: Special Briefing on Bombing-Navigation Subsystems; memo, Col. D. S. Dunlap, DCS/Plans and Ops., to Ch., B-70 WSPO, DSM, WADD, 19 Jan. 1960, subj.: Special riefing to the Commander on Bombing Navigation Subsystems, in HBF.
66. DSN WAR, 5 Feb. 1960; ARDC 111, 26 Feb. 1960, subj.: Chemically Powered Strategic Bomber. B-70, in HBF.