Origins of inertial navigation.
Draper, who was known as "Doc" to his close friends, fellow professors, and students, was an avid flyer and an expert in aircraft instrumentation. In later years he liked to claim that he got the idea for an inertial navigation system "out of a bottle of whiskey." (1) It was a story that he told so often that the actual facts became muddled over the course of time.
Inertial navigation had been something that Draper had been thinking about for a long time when he broached the subject to Lee Davis. When they learned about the Japanese surrendered, as Doc was found of telling his son James, Davis reached under the pilot's seat and pulled out a bottle of scotch. (2) As they were celebrating the end of the war with Doc's "soothing syrup," Davis talked about the rearrangements in spending that would came from the cancellation of weapons that were no longer needed. (3) Doc saw an unusual opportunity to gain support for a project that had been in the back of his mind since the early thirties: the development of a self-contained inertial navigation system could provide the pilot with an airplane's location in bad visibility without the assistance of external instruments. (4)
Doc knew that Davis understood his suggestion because he had devoted a fair amount of his time studying gyro principles while he was a graduate student at MIT. Both men realized that such a system--if it could be developed--would overcome the huge navigation problem experienced by the crews of the Air Forces's [Ed. note: this refers to the U.S. Army Air Forces, not United States Air Force] long range bombers during the war. The two men discussed the details of just such a system during the remainder of their trip back to Wright Field.
When Doc returned to Cambridge he discussed the idea with key members of his staff. Although the present state of the gyroscopic art precluded their use for inertial navigation, Doc and his colleagues "felt very strongly that self-contained [inertial navigation] systems were possible and that with existing motivation, useful results could be brought to realization in a few years." (5) After reviewing the theoretical and technological issues surrounding the project, Doc returned to Wright Field to discuss it with Davis and a small group of engineers from the Armament Laboratory led by John Clemens. As Doc would write in 1969, "With the war just finished, the problems of accurate bomb and rocket deliveries after long flights over unfriendly ground environments were large in the minds of Colonel Davis and his scientists, Dr. John E. Clemens and Dr. Ben Johnson." (6) Because gyros drifted over time, a stellar sighting system was added to provide the accuracy needed for the long duration flights the system was designed to be used for. Once the aircraft achieved cruising altitude, it would lock onto a celestial body and correct any errors in the flight path caused by drifting of the gyros or other factors. Draper regarded the star tracker as a temporary necessity based was on the limited accuracy of the gyros then available for aircraft use, but he felt that it was a "messy and inelegant," approach to the problem. (7)
On August 23, 1945, the Instrumentation Laboratory submitted a proposal to the authorities at Wright Field for a Stellar Bombing System designed primarily for operation in jet propelled aircraft as a bombsight, noting "the possibility of eventually robotizing the system for use with guided missiles ...." (8) Draper as Donald MacKenzie noted in Inventing Accuracy, was well aware that stellar observations were "subject to interference by weather, aurorae, meteors and countermeasures." (9) Although Doc would have preferred a closed "black box" solution, he was a pragmatic engineer who understood the severe limitations of the gyroscopes then available.
Less than a month later, the Laboratory received a letter contract to study the possibilities of the inertial navigation system Doc had proposed. To help conduct the study Draper recruited Walter Wrigley, a former doctoral student who had spent the war years working as an R&D project engineer for the Sperry Gyroscope Company. (10) One of the key problems that had to be solved in order to construct a workable inertial guidance system was how to accurately indicate the direction of the vertical (the line running from an aircraft's center of gravity to the center of the earth) from a rapidly moving vehicle. To many in the scientific community this seemed an impossible task based on Einstein's general theory that an observer inside a closed box could not distinguish the effects of linear acceleration from the effects of a gravitational field. (11) One physics textbook published in 1942 went so far as stating that it was impossible to construct a device "to indicate the true vertical unaffected by accelerations of the airplane when in curved flight." The authors of this work were unaware, no doubt, of Walter Wrigley's dissertation. (12) In his doctoral thesis, supervised by Draper and submitted on March 9, 1940, Wrigley provided a comprehensive mathematical analysis of the methods available for indicating the direction of the vertical from moving bases. (13) Wrigley's conclusion that a "damped gyroscopic servo-controlled by a pendulum," offered the most practical solution lad the ground work for constructing an inertial navigator adding to Doc's conviction that such systems were now a possibility. (14)
After the Instrumentation Laboratory submitted its initial study of a Stellar Bombing System, the U.S. Air Force, which was established on September 18, 1947, give the green light to proceed with an experimental program designed to the test the possibilities of actually constructing an inertial navigation system. (15) The project was begun on November 21, 1947, under the name of the Stellar Inertial Bombing System (SIBS). (16) It was later changed to FEBE, a variation of the Sun Good, Phoebus, in reference to the use of the sun for stellar tracking purposes.
The problem Doc now faced was obtaining precision sensors that could produce the accuracy needed over the five to ten hours that would be required during the long distance flights the system was designed for. Flight tests of an ARMA (17) Stable Element commonly used in U. S. Navy fire control systems to determine the vertical was installed in an Air Force DC-2 in an attempt to satisfy the Air Forces desire to use existing technology. (18) The equipment, which was large and heavy, proved unsuitable for the task. Doc and his staff at the Instrumentation laboratory concluded that new sensors would have to be developed without dependence on anything available from existing technology. A rigorous analysis of the use of inertial space references for navigation purposes completed by the Instrumentation Laboratory in February 1947, pointed to the gyroscope, rather than the accelerometer as the key sensor. (19) Their analysis was based on the aircraft bombing mission, an application distinguished by long flight times and a low-acceleration environment in which the heading errors produced by gyrosopic drift was the primary inaccuracy in system.
As Walter Wrigley had suggested, a stable platform indicating the true vertical could be constructed using three servo-controlled, single-degree of freedom gyroscopes. To understand how this works, image that the gyros are fixed to a flat board mounted on gimbals so that it is free to move in all directions in such a manner that they detect the motion of the board about its roll, pitch and yaw axes. Suppose this assembly is placed in an aircraft with the board aligned parallel to the horizon so that a perpendicular line through its center establishes the direction of gravity and thus the direction of the vertical. Let's also assume that the gyros are constructed so that their output, as they precess, are proportional to velocity of the change in direction experienced by the gyro rotors due to the forces of acceleration acting on the board as it begins to move through space. Integrating these signals provides a measure of how much the board has been displaced along each of the three axes of roll, pitch and yew. These signals can then be sent to small motors attached to the platform's gimbals. When the aircraft's attitude or heading changes, the integrated signals from the gyros will cause the servo motors to move the platform back to its original starting position keeping the board in a horizontal position thereby maintaining a true indication of the vertical.
The second sensor needed was a highly accurate accelerometer. Let's suppose that two of thee accelerometers have also been placed on the board perpendicular to one another so that they can measure the acceleration in the north-south and east-west directions. When the signals from these accelerometers are integrated twice ( a = v v = d ) they provide a measure of the distance traveled over an interval of time.
To start the system, the stabilized platform is aligned to the horizontal and positioned so that the sensitive axis of the north-south accelerometer is pointed to the north. The latitude and longitude of the starting point and destination is then set into the system, and the integrators are trimmed to zero. As soon as the aircraft begins its takeoff run, the accelerometers will sense the resulting accelerations providing a measure of how far the aircraft has moved. These distances are then converted into corresponding changes in latitude and longitude and added to the starting point coordinates to show the aircraft's new position.
Although the concept involved in building an inertial navigation system was now straight forward, several difficult problems had to be overcome before a working unit could be fabricated. The most difficult of these was to develop a set of gyroscopes, accelerometers, and the other components needed for the system that were small enough to fit into an aircraft, yet accurate enough to provide the precision required during the long flight times specified by the Air Force. A separate issue was the need to take into account the affects of gravity as the stable platform moved over the earth' surface so that the platform remained at right angles to the earth's radius.
To compensate for the earth's rotation Wrigley applied Schuler's Principle of an "earth-radius pendulum." Maximilian Schuler was working to improve his cousin's gyrocompass in 1923 when he hypothesized that a solution to the vertical could be achieved if the vehicle traveling over the earth was attached to a pendulum whose center of gravity was a the center of the earth. (20) As the vehicle moved, the pendulum would continue to indicate the direction of the vertical.
Of course a pendulum of this size could never be built, but Wrigley realized that the disturbing effects of gravity on the stable platform could be removed by designing into it a simple feedback loop that continuously caused the platform to remain horizontal. "Such a system could be seen as working as if it kept horizontal by an earth-radius pendulum." (21) In order to work properly, such a feedback system would have the same 84-minute (22) natural frequency period of Schuler's pendulum. Thus concept, apparently coined by Wrigley, is named Schuler Tuning. (23)
Although several other firms were actively engaged in developing inertial navigation systems for various U. S. Air Force programs, some in the scientific community remained skeptical of this unproven technology. (24) George Gamow, a prominent physicist and a member of the Air Force's Scientific Advisory Board (SAB) was highly skeptical of this approach to the guidance problem facing the U. S. military services as they attempted to develop air-breathing intercontinental guided missiles that were capable of attacking the Soviet Union. In February 1948, Gamow, noted for his brilliant mind an ebullient sense of humor, was working on the long-range navigation problem as a consultant to the Johns Hopkins Applied Physics Laboratory (APL). (25) On the 13th of that month, a day Gamow referred to as "Black Friday," he issued a scathing memorandum addressed to Ralph E. Gilbert, the newly appointed the director of APL, attacking the concept of an inertial navigation system. (26) In his memorandum, titled "Vertical, Vertical, Who's got the Vertical," Gamow argued that the inertial navigation systems then under development by the Instrumentation Laboratory and other contractors working for the Air Force, were impractical because they wold have to work flawlessly. These instruments would have to be capable of indicating an aircraft's initial position and velocity with perfect accuracy if such a system was to function as intended. However, an aircraft's position, as he stated in his memorandum, "is completely undetermined, unless its initial position and velocity are known exactly and the integration [of velocity] is carried on faultlessly all the way through." At the time it was written no instrument existed that could satisfy this criteria, but Gamow underestimated the capabilities of Doc's laboratory, which, as we shall see in the pages to come, was able to develop such a device.
But this was not the only criticism leveled by Gamow. Of greater concern was the issue of correcting the errors that would inevitably be introduced by inaccuracies and outside inputs. Doc would address this issue later on, but for the time being Gamow's memo, which was laced with vivid caricatures mocking the current attempts at developing an inertial navigation system, caused quite a row within the inertial navigation community, according to those interviewed by Mackenzie for Inventing Accuracy. (27) Because Gamow was a member of the Advisory Board's panel on guidance and control, his ideas carried considerable weight and had the potential of catastrophically derailing Doc's program as well as those of his competitors. As Doet and Soderqvist noted in their book on the Historiography of Contemporary Science, "The multiplicity of groups working on the [inertial navigation] problem aggravated the task of responding to Camow's criticism," which became essential if Draper and the other contractors expected the Air Force to continue funding the development of this yet to be proven technology.
What transpired during the next twelve months, as far as I can determine, does not show up in the historical record. Although Draper was not a member of the SAB Guidance and Control Panel and was not mentioned by name in Gamow's memorandum, its content was undoubtedly of great concern. The Armament Laboratory was not a traditional source of research monies for guidance work. The situation facing Draper and Leighton Davis, his patron, was clearly put forth my Michael Dennis: "Few organizations were capable of supporting Doc's research; of others on the funding food chain perceived Draper's research as a technological 'dead end,' then Davis and Draper were in jeopardy." (28) Where, when, or if Gamow's memo was circulated or discussed is not known. But Draper, who was member of the SAB's Guided Missile Panel had the connections and political clout to do something about it. Using his contacts within the SAB, he arranged to conduct a classified conference on guidance at MIT in February 1949. The meeting, which was held under the auspices of the SAB, was titled a "Seminar On Automatic Celestial and Inertial Long Range Guidance Systems. (29) Although the stated purpose of the meeting was "a means of promoting a wider dissemination of information on the basic theory involved" in the guidance problem, Doc used it as a clever means of refuting Gamow's contemptuous opposition to inertial navigation. (30)
Doc invited every major firm and component manufacturer working in the field to demonstrate the progress that had been made in the past few years. Although Gamow was also invited, he probably recognized that the "meeting was 'stacked' against him" and decided not to attend. (31) The instrument errors that Gamow claimed would make inertial navigation unusable could be corrected--according to Draper--by a process he termed "smoothing." As put forth in Doc's opening statement to the scientists who had come to MIT for the seminar on guidance, "the amount of smoothing that can be used is limited by the fact that any increase in smoothing always brings with it an increase in the time required for a system to solve its guidance problem." (32) Doc went on to explain the importance of solving the conflict between smoothing and solution time, which would have a prominent place in the papers to be presented.
This was Doc's hidden agenda for presenting the details of FEBE, the experimental inertial navigation system being assembled by the MIT Instrumentation Laboratory under Doc's supervision. Although it had yet to be flown, it was nearing completion and was soon to be tested. Doc staff was responsible for presenting 8 of the 25 sessions conducted during the course of the three-day meeting. This was twice as many as the Lab's nearest competitor: the North American Aviation Company that was working to develop an inertial navigation system for the Navaho intercontinental missile on a another Air Force contract.
FEBE was a demonstration system engineered to validate the design assumptions needed to create a true inertial navigation system--a so called "black box" that would function without any external inputs. It was designed to investigate the dynamics of a closed loop automatic navigation system, study the various instruments and their organization, and to establish a correlation between the results of flight tests and theory. (33) Although FEBE could operate at night using navigational stars, the sun was selected as the celestial reference so that records of the actual ground track could be more easily made to ascertain the system's accuracy. The sensors used in FEBE were based on Marine gyrocompasses and the gyroscopic elements of Doc's World War II anti-aircraft fire control systems. (34)
The system, which weighted 4,000 pounds when fully assembled, was installed in a B-29 so that it could be systematically tested in the in 1949. It was flown for the first time on May 5, 1949. (35) This "shakedown" flight was followed by nine more experimental flights designed to test the system's accuracy and see how it behaved over long distances. Because of equipment malfunctions on two flights and an abnormally erratic reading on another, only six of the flights produced acceptable results. (36) When averaged together they yielded mean error of five nautical miles. Although this made FEBE unsuitable for the bombing mission, the results were encouraging enough that the Air Force issued a follow on project to the Instrumentation Laboratory to design, build and test a navigation and guidance system that would depend only upon the inertial and gravitational inputs. This project was named Space Inertial Reference Equipment (SPIRE).
To construct SPIRE, Doc's team at the MIT Instrument Laboratory designed an inertial platform using three single-degree-of-ffeedom gyros tha.t the lab had developed for improved accuracy. The system was loaded into a B-29 on loan from the U.S. Air Force on January 23, 1953, and given a one-hour shakedown flight on Friday, February 6. (37) Draper was so confident in its success that he secretly planned to demonstrate the system enroute to a top-secret government sponsored symposium on inertial navigation that was scheduled to begin in Los Angles, California, on Monday, February 9.
The flight was uneventful and the navigation system seemed to be working fine until they reached the Rocky Mountains. Just south of Denver, north of Colorado Springs they climbed to twenty thousand feet to clear the mountains. The weather had been clear until they reached the Rocky Mountains when they ran into dense cloud cover. An hour or two from Denver, Chip suddenly noticed that that the B-29 was turning to the right about ten to twelve degrees as the encountered some unexpected air turbulence.
"Chipper," Doc exclaimed over the intercom, "what the hell's going on up there?" (38)
"Doc," Chip replied, "the system is commanding a turn to the right."
Doc and the crew monitoring the system in the back of the plane knew that something was awry because they could see the gimbal turning with respect to the aircraft. But they couldn't see what was happening to the rudder. There was a note of panic over the intercom, but Doc remained calm.
"Let's not do anything," he said. "Let's leave it alone. Let's see what it's going to do."
Unbeknownst to those on board the B-29, they had encountered a weather front and were being blown southward. SPIRE, sensing the wind drift adjusted the rudder so that the aircraft would stay on track. When they broke out of the cloud cover over the San Joaquin Valley they were right on course.
The aiming point for the flight was the intersection of the runways at their planned destination. An indicator light had been installed in the left side of the cockpit to show when they were over the aiming point. When it came on, Chip looked down to see that they were passing over the apron area near the airports building about eighteen hundred feet from the aiming point. Although the accuracy was classified at the time, the system error over the 12hour, 2,600-mile flight (according to Collins) was one one-hundredth of a percent (0.00013). Over the years, Chip's enthusiasm for this accomplishment clouded his memory for the actual error according to the data recording during the flight was nine nautical miles. (39) Nevertheless Doc was ecstatic with SPIRE's results, for they demonstrated beyond a shadow of a doubt that a completely self-contained system using sensors that relied solely on inertial principles could be successfully used to navigate long distances.
During their flight across the country Doc and Roger Woodbury plotted the B-29's progress on a long role of paper, showing the intended course and the actual course using photographs of prominent landmarks taken through the nose of the B-29 to verify their results. (40) After landing at 9:28 in the evening Doc and the rest of the SPIRE team stayed up all night putting Lambert conformal maps on the big board behind the seminar podium adding a brightly colored tape showing the exact track the B-29 had followed across the continent.
When the symposium began the next day, Doc was introduced as the first speaker. "Gentlemen," he began, "we have a system that works. We did it." (41) Then, to the astonishment of the other attendees, he went on to describe the historic flight he had just made, "giving credibility to the enormous potential of inertial guidance." (42)
SPIRE was the forerunner of the modern inertial navigation systems that the aviation community depended upon before the advent of GPS. It also established MIT's Instrumentation Laboratory as the leader in inertial navigation and guidance, forming the foundation for the Laboratory's future development of the guidance systems for the Thor, Polaris, Titan, Poseidon, and Trident ballistic missiles.
Thomas Wildenberg is an independent aviation historian and a frequent contributor to Air Power History. He is a former Smithsonian Fellow having served successive terms as a Ramsey Fellow at the National Air and Space Museum in 1998 and 1999. Mr. Wildenberg is the recipient of a number of awards recognizing his scholarship. These include the Surface Navy Association Literary Award, an award from the Air Force Historical Foundation for the best article in Air Power History in 2009, and the John Laymen Award. He is currently working on a biography of Charles Stark Draper.
(1.) Draper Oral History, p. 79.
(2.) James Draper, interviews with author November 15, 2012.
(3.) Davis suggested that some funds assigned to purchase bomb shackles might not have to be used. See Draper, "On the Evolution of Accurate Inertial Guidance Instruments," p. 24.
(4.) Draper, "Remarks on Instrumentation Laboratory," p. 19.
(5.) Draper, "Origins of Inertial Navigation," p. 455.
(6.) Draper, "Remarks on Instrumentation Laboratory," p. 19.
(7.) Mackenzie, "Stellar-Inertial Guidance," p. 196.
(8.) Ferguson, FEBE entry, CSDL-HC Finding Aid. Note: because this endeavor involved navigation, it did was not assigned to the Armament Laboratory to which Davis was assigned, but to another organization at Wright Field. See Draper, "On the Evolution of Accurate Inertial Guidance Instruments," p. 24.
(9.) MacKenzie, Inventing Accuracy, p. 76. There is no evidence to support McKenzie's suggestion that stellar-inertial navigation was "an added attraction of being, superficially at least ... more familiar to Air Force officers, who were by then well accustomed to the use of star-sightings in navigation." On the contrary both Davis and Clemens were well familiar with Draper's capabilities and would not have felt the need to "guild the Lilly." Besides, a study project with no hardware was not very expensive and would have been a "no brainer," with Davis's approval.
(10.) Wrigley biography, Massachusetts Institute of Technology, Instrumentation Laboratory, Biographies Section, p. 5.
(11.) See MacKenzie, Inventing Accuracy, pp. 66-67.
(12.) F. K. Richtmyer and E. H. Kennard, Introduction to Modern Physics, third edition (New York: McGraw-Hill, 1952), p. 51, as cited by Mackenzie, Inventing Accuracy, p. 67.
(13.) Wrigley, "An Investigation of Methods Available for Indicating the Direction of the Vertical from a Moving Base, Ph.D. Dissertation, MIT, 1940.
(14.) Ibid., p. 86.
(15.) Ferguson, FEBE entry, CSDL-HC Finding Aid.
(16.) Bogsian, "An Experimental Automatic Long Range Guidance System, Project FEBE," Preface, in Seminar on Automatic Celestial and inertial Long Range Guidance Systems, Vol. I, p. 141.
(17.) Anna Engineering Company.
(18.) Draper, "The Evolution of Aerospace Guidance Technology at the M.I.T. 1935-1951," pp. 232-33.
(19.) Mackenzie, "Missile Accuracy," p. 201.
(20.) MacKenzie, Inventing Accuracy, p. 76.
(21.) Ibid., p.70.
(22.) If a satellite were circulating the earth every eighty-four minutes, it would have to be at tree-top height.
(23.) Ibid., see note 121, p. 71.
(24.) These programs included Northrop's Snark and North America Aviation's Navaho air breathing missiles. Hughes Aircraft was also working on an Air Force contract to develop a celestial navigation and guidance system.
(25.) Eric Roston in The Carbon Age, page 17 characterized George Gamow as "a hard drinking Soviet emigre to the United States with a brilliant mind, [and] an ebullient sense of humor...."
(26.) George Gamow to R. E. Gibson, dated Black Friday [February 13?] 1948. Cabinet 5, Drawer 4, File Gamow, Applied Physics Laboratory Archives as cited by Dennis, "A Change of State," pp. 416-18.
(27.) MacKenzie, Inventing Accuracy, footnote 130, p. 75.
(28.) Dennis, "A Change of State,"' pp. 419-20.
(29.) Scientific Advisory Board, Office of the Chief of Staff U.S. Air Force, Seminar on Automatic Celestial and inertial Long Range Guidance Systems, Vol. I (hereafter MIT Guidance Seminar), passim.
(30.) Ibid., p. 2. The politics behind Gamow's memorandum, which is beyond the scope of this monograph, is summarized by Dennis on pages 420-21 of his dissertation.
(31.) MacKenzie, Inventing Accuracy, p. 72.
(32.) MIT Guidance Seminar, Foreword, p. 2.
(33.) Ibid., p. 184.
(34.) Draper, "Origins of Inertial Navigation," p. 457.
(35.) Ferguson, FEBE entry, CSDL-HC Finding Aid.
(36.) Draper, "Origins of Inertial Navigation," p. 456.
(37.) Ferguson, SPIRE entry, "CSDL Historical Collection Projects."
(39.) Ferguson, SPIRE entry, "CSDL Historical Collection Projects."
(40.) "Laboratory Aviator Extraordinaire," D-notes, May 20, 1988, p. 3.
(41.) Chip Collins, interview conducted on October 16, 2012,
(42.) Morgan, et al. Draper at 25, p. 11.
|Printer friendly Cite/link Email Feedback|
|Publication:||Air Power History|
|Date:||Dec 22, 2016|
|Previous Article:||Boyd revisited: a great mind with a touch of madness.|
|Next Article:||The World War II training experiences of the Tuskegee Airmen at Oscoda Army Air Field.|