Operations research fell at the applied end of the research spectrum. It concerned when to certify an airplane for take-off and landing. It was an opportunity to save lives. Even in an area so close to applications or development, the NACA shaped its research to illuminate general problems to benefit the entire industry. During World War II icing research was unquestionably "development," as distinct from fundamental research. The anti-icing hardware developed for military aircraft during the war was only a short-term solution to the icing problem. A fundamental understanding of what constitutes the icing cloud was necessary. In 1944 the NACA began a program to compile statistical data to define icing conditions, culminating in the 1950s in guidelines for the design of ice-protection systems, which became the basis for federal regulations in certifying these systems. As it evolved, icing research became a discipline that combined both theoretical and experimental approaches.
Icing on aircraft is caused by flight through an icing cloud consisting of small supercooled droplets that strike the aircraft surfaces and freeze into a porous white mass called rime ice. If larger supercooled droplets strike the aircraft, not all of the water freezes on impact. As water accumulates and enough heat is dissipated, the water freezes into a clear, hard mass called glaze ice, a much more serious problem because it is difficult to remove. Ice adds weight and impairs the aerodynamic efficiency of an airplane, often leading, in severe conditions, to a crash.1 After the bombing of Pearl Harbor, military planners suggested that the most promising route for an invasion of Japan was across Alaska and the Aleutian Islands. Anticipating severe aircraft icing problems, the Army asked for a broad attack on many fronts. General Electric, the Massachusetts Institute of Technology (MIT), and all three NACA laboratories joined the effort to find ways to protect aircraft from this silent enemy.
In aircraft powered by piston engines during World War II, ice accumulated on the propellers and the plane's leading edges, blocked the air intakes of the engines, and choked the carburetor. Icing of windshields cut down pilots' visibility. Before the war aircraft manufacturers used a simple, inexpensive, mechanical method to remove ice. An inflatable rubber boot was fitted over the leading edges of the wings and tail. When ice began to build up, the boots were inflated and deflated to knock off the ice. However, the boots impaired the aerodynamic effectiveness of the aircraft's surfaces and increased drag. These clumsy and increasingly ineffective de-icing methods had to be replaced with a more aerodynamically efficient system of ice prevention. Heat to prevent ice build-up seemed to promise a more satisfactory method to conquer the icing problem.2
 Much of the wartime research in icing by the NACA focused on the pioneering development of a thermal anti-icing system for military aircraft. Although the idea of diverting the heat from the engine's exhaust to the leading edges of the wings and tail was not new, prior to the war, aircraft manufacturers were unwilling to incur the additional production costs required by a thermal anti-icing system. Under Army auspices, engineers at Langley, led by Lewis A. Rodert, designed and installed a thermal system in the wings of a Lockheed 12 airplane, the NACA's first so-called "flying laboratory." Rodert continued this work at Ames. After flight testing the new system, in 1941 Rodert's group issued its first report demonstrating the feasibility of using exhaust-heated wings to protect against icing.3
During construction of the Altitude Wind Tunnel at Lewis Laboratory, an Icing Research Tunnel was added to the original plan to take advantage of its extensive and sophisticated refrigeration system. It was hoped that the new icing tunnel in Cleveland would be useful in assisting in the development of Rodert's thermal system, Proposed in 1942 and completed in spring 1944, the icing tunnel was not part of the original plan for the laboratory. However, it is one of the few facilities built during the war still used for its original function. Because the designers of the icing tunnel, Alfred Young and Charles Zelanko, thought that the icing problem would be solved within a short time, they proposed a tunnel suitable also for aerodynamic testing of engine components.
Young and Zelanko were aware that the design of an icing tunnel was far more complex than that of a simple wind tunnel. The need to simulate the atmospheric conditions of an icing cloud presented them with an extremely demanding engineering problem. In hindsight, it is apparent that neither the science of meteorology nor the state of engineering knowledge was adequate for the job at hand. As Zelanko later recalled, they knew how to design ordinary wind tunnels, but the problems of creating icing conditions were almost without precedent. "Logic, theory, and speculation were the only design tools that were available."4 Only when the tunnel's spray system was replaced in the early 1950s could tunnel tests yield accurate icing data. Until then, flight research provided more reliable information on the phenomenon of icing.
Icing tunnels were not entirely unknown before the proposed Cleveland tunnel, but they were smaller and did not have the advantage of the extensive refrigeration system Carrier had created. However, Young and Zelanko studied the tunnel of the B. E Goodrich Company (Akron, Ohio), which manufactured rubber pneumatic de-icing boots, and MIT's special icing wind tunnel. They proposed a tunnel with a maximum speed of about 400 miles per hour with a 7-foot x 10 foot test section, later scaled down to a 6-foot x 9-foot test section. To simulate icing conditions in the tunnel, air was cooled to -30°F by passing it over the fins of the heat exchanger. A spray system introduced an atomized stream of water into this refrigerated airstream. The unnaturally large size of the water droplets remained the problem that defied solution.5
As the tunnel began to take shape, Wilson H. Hunter took charge of the Icing Research Section. A 1930 graduate of Yale in mechanical engineering, Hunter brought firsthand experience in the design of rubber de-icers from the B. F. Goodrich Company. However, by June 1944, when Hunter supervised the tunnel's first test, the era of the mechanical de-icing had passed. The wind tunnel testing program focused on testing components of Rodert's thermal protection system. This work was unabashedly hardware development. More basic research would have to wait until the feverish activity of meeting wartime demands could yield to a more measured step.
The NACA made extensive use of a simple instrument to measure ice accumulation. It consisted of rotating metal cylinders of graduated diameters. By measuring the thickness of the ice  buildup, the researcher could determine the mean size of the water droplets, distribution, and liquid water content in the icing cloud. The group at Ames grappled with the problem of defining the characteristics of the icing cloud using a statistical approach. What range of liquid water content, drop size, and temperature defined an icing encounter? The Ames project under Alun Jones was strengthened by the additional talent of William Lewis, a meteorologist on loan from the U.S. Weather Bureau. J. K. Hardy, on loan during the war from the British Royal Aircraft Establishment at Farnsborough, tackled the problems of heat transfer in a heated wing, a study that culminated in an important paper published in 1945, "An Analysis of the Dissipation of Heat in Conditions of Icing from a Section of the Wing of the C-46 Airplane."6
In 1946 Lewis Rodert took charge of the icing program at the Cleveland laboratory. In a March 1947 memo, Rodert stressed the importance of gathering statistical data on icing clouds. These data would allow the Civil Aeronautics Administration to craft regulations to certify new aircraft if they met guidelines in providing ice-protection equipment. Reliable data were needed because of the "general disagreement over what constitutes a safe design basis for the heated wing."7 The Cleveland and Ames laboratories collaborated on this program.
By 1948 the program had logged a total of 249 icing encounters and had taken about 1000 measurements of liquid water content and droplet diameter. This was a collaborative flight research program carried out by the Air Force in the upper Mississippi Valley, by Ames in the West, and by Lewis Laboratory in the Great Lakes area. John H. Enders, an intrepid pilot from Cleveland's Flight Research program, flew the dangerous missions over Lake Erie, bringing back his Consolidated bomber encrusted with ice. In October the NACA group presented a tentative table of design conditions to the NACA Subcommittee on Icing Problems. This report defined the parameters for which icing protection was necessary for safe operation.8 The statistical data in this NACA report became the basis for the design criteria for federal requirements for aircraft icing protection adopted by the Civil Aeronautics Administration in the mid-1950s. In reflecting on the success of this effort, William Lewis wrote in 1969, "Considering the fact that we had data from 167 encounters in layer clouds and 73 in cumulus, statistical extrapolation to a probability of 1/1000 was more than daring, it was down right foolhardy." Nevertheless, more recent research using a much larger data base confirmed the conclusions of the 1949 report.9
In 1949 Abe Silverstein asked Irving Pinkel to take over as the new associate chief of the Physics Division. Supervision of icing research fell among Pinkel's duties. It appeared that NACA work in icing had reached a natural point of termination. Ames closed down its program and  transferred equipment and personnel, including William Lewis and D. B. Kline (also on loan from the Weather Bureau), to Cleveland. Silverstein thought that icing research was too close to development and should be phased out, but in Pinkel's view the icing group was "very able and much maligned." Previously, Pinkel had worked with the icing group as a member of the Wind Tunnels Analysis Panel to encourage a more analytical approach to the problems of icing. Born in Gloversville, N.Y., he graduated with honors from the University of Pennsylvania in 1935. Pinkel began his government service in Pittsburgh, Penn., as a physicist with the Bureau of Mines, where he worked on synthesizing liquid fuels from coal. In 1940 he joined his older brother Ben at Langley, where he focused on the nonstationary aerodynamic forces of airplane flutter. He transferred to Cleveland in 1942 to work on the hydraulics problems of engine lubricating systems. A member of Silverstein's Special Projects Panel after World War II, he contributed pioneering papers on lift and thrust developed by heat addition to the supersonic flow beneath an airplane wing.10
Pinkel listened attentively to his staff's arguments for the continuation of icing research. His engineer's sixth sense told him that terminating it was premature. "I saw that the people who were involved in it were, in fact, trying to create an engineering discipline out of it, which seemed the correct thing to do."11 Instead of closing down the division, he broadened it.
But what did it mean to make an engineering discipline out of the work? The cut-and-try testing of components had characterized most of the NACA's prewar and wartime icing research. Postwar engineering demanded analytic skills. Testing without analysis as a guide was expensive and time consuming. The icing group at Lewis was composed of men with mechanical engineering degrees, the majority of whom took their first jobs with the NACA immediately after graduation. Women with a mathematical bent, but usually only high school graduates, assisted the men as computers and data takers. As a whole, the staff at Lewis lacked the analytic skills that basic engineering research demanded. After the war, a process of self-selection took place. As the manpower of the icing division was reduced, the engineers who remained began to form a corps,....
 ....which developed an attachment to icing as a discipline. They began to assimilate and integrate new knowledge from the fields of heat transfer and meteorology.
Like many of their colleagues in other areas, such as supersonic aerodynamics, they taught themselves what they needed to know. With ties to the National Research Council of Canada and the Mt. Washington Research Station and access to Soviet data, they were in the mainstream of world icing research.
Under Pinkel's direction, the icing group focused on extending the basic understanding of the icing cloud. They launched a new research program with the commercial airlines and the Air Force to gather statistical icing data from areas throughout the world. They developed a simplified pressure-type icing rate meter that could record the frequency and intensity of icing conditions automatically. The meter was installed in about 50 commercial and military aircraft that flew over the North Atlantic, the Continental United States, Alaska, the Pacific and parts of Asia. Over a period of five years, data were gathered for 1800 icing encounters. The icing meter was superior to the rotating multicylinder method, not only because it did not require the constant attention of research personnel, but, more important, because of the far greater number of encounters and range of conditions that it was now possible to record. The rotating multicylinder provided data for a small number of flights into deliberately sought, abnormally severe, conditions. Now routine flights into all degrees of severity could be monitored. In addition, weather reconnaissance information gathered over a two-year period by the Air Weather Service and the Strategic Air Command was made available to the icing group at Lewis. All these data were entered on IBM punch cards for analysis using one of the first IBM computers at Lewis."12 In general, the data supported the findings of the laboratory's original study, adding, in particular, new knowledge about the vertical and horizontal extent of icing encounters. Because of difficulties in the calibration of the meter, the results were not entirely consistent for the icing rate and liquid water content.13
Another problem the icing research group tackled was the complex analysis of the path of a water droplet. As an aircraft moves through an icing cloud, only a small percentage of the supercooled droplets actually strike an aircraft and freeze. The majority of the droplets follow streamlines that move around the leading surfaces of the airplane. With the knowledge of the water content and droplet size (determined by the rotating multicylinder technique), it was possible to calculate the droplet trajectory. The rate of icing on a given surface or shape of a component could then be predicted.14 The Lewis group extended the available data over a range of airfoil shapes and thicknesses through the use of a differential analyzer. Harry Mergler of the Instrument and Computing Division adapted this early computer designed at MIT by Vannevar Bush to meet the needs of icing research. Although calculating the trajectory of a droplet was still a laborious process, the differential analyzer reduced the time it took for a single calculation from weeks to about four days. Women who had formerly worked the problems on simple mechanical calculators made the transition to the gargantuan differential analyzer with comparative ease. Often they were given the distinction of having their names appear as co-authors of the reports of these studies, no doubt because their contribution went beyond the sheer drudgery of "number crunching. In addition to reducing the time for arduous calculations, the differential analyzer was a potential teacher of higher mathematics. The calculus unfolded mechanically through the gearing and shafts; the chart recorders taught the operators "to grasp the innate meaning of the differential equation." The machine trained them to think in the same logical steps that the machine ground through in making a calculation. Through the back door, women assigned to toil  on this giant calculator gained access to the abstruse realm of partial differential equations - a realm in which not all engineers hired during World War II were comfortable.15
With new information now available on the phenomenon of the natural icing cloud, the tunnel engineering staff set about the redesign of the spray system for the icing tunnel, long recognized both within the NACA and by outside authorities as inadequate.16 The droplet size was ten times larger than the statistical definition derived from actual icing encounters, and instrumentation for measuring the properties of the artificial icing cloud was lacking. Vernon Gray, known as "Mr. Icingologist," took this problem under his wing. He directed the efforts of Halbert Whitaker, a fluid systems engineer, to develop a system that would produce an atomized spray. Day after day, through trial and error, the team perfected the design, which consisted of a battery of approximately 80 spray nozzles mounted on six horizontal bars. The system produced an icing cloud of approximately 4 feet by 4 feet. To verify whether tests in the tunnel accurately reproduced actual flight conditions, the heat transfer data obtained by J. K. Hardy at Ames in his famous study of the heated wing were compared to tests on the same wing mounted in the icing tunnel. The data were in agreement.17
The spray system was a complex engineering achievement. It was now possible to run tests that were much closer to natural icing conditions. At this point the tunnel became a valuable research tool, and a group that included Dean Bowden, Thomas Gelder, Uwe von Glahn, Vernon Gray, and James P Lewis inaugurated a new program to test cyclic de-icing systems. As the era of high-speed, high-altitude turbojet fighters and transports dawned in the early 1950s, it became clear that thermal anti-icing systems required an excessive amount of heat. Bleeding this heat from the engine severely impaired performance. In a cyclic system, ice is permitted to form on aircraft surfaces. This ice is melted and removed by aerodynamic forces at intervals by short, intense heating periods. The new program at Lewis Laboratory may have been stimulated by new research on intermittent heating by L. M. K. Boelter's heat transfer group at the University of California at Los Angeles."18 The study of thermal cyclic de-icing systems required a level of sophistication in the field of heat transfer and aerodynamics unknown in the wartime era.19
Between 1949 and 1955 flight research also continued, In addition to checking the results of tests in the tunnel with the performance of the new systems during actual icing encounters, the flight research group evaluated the rotating multicylinder method for determining droplet size, distribution, and liquid water content with a view to understanding both the strengths of the technique and, more important, its limitations. This....
....was "normal engineering" in the sense that it involved no dramatic scientific or technical breakthroughs, but an extension and refinement of previous knowledge and practice. For example, using dimensionless parameters, Rinaldo J. Brun, William Lewis, Porter Perkins, and John Serafini produced a technical report that calculated the impingement rate of cloud droplets on the rotating multicylinder. The report shows a solid command of boundary layer theory, heat transfer, experimental physics, and instrumentation that sets it strikingly apart from the simple empiricism of the reports of the mid-1940s. The discipline had indeed been placed on a rational basis.20
Ironically, the steady improvement of the jet plane contributed to the phasing out of icing research at Lewis Laboratory in 1957. By the late 1950s, the turbojet engine had come into general use for commercial flights. In contrast to the more recent development of the turbofan engine, early turbojets were overdesigned. They were not affected by the "bleeding off" or diverting of compressor air, which was piped through the wings for icing protection. The aircraft manufacturers could tailor inexpensive icing protection systems to specific designs without compromising the performance of the airplane, Moreover, since turbojet engines provided greater power than  the reciprocating engines they replaced, higher altitudes were possible. Aircraft could fly over the icing clouds rather than through them, leaving take-off and landing as the only potentially dangerous icing situations. However, as Irving Pinkel pointed out, "With new technology old ideas can get a breath of life again." Problems solved for one set of conditions recur as technology advances in other areas. When the more efficient turbofan engine revolutionized engine design in the mid-1960s there was less compressor bleed air available for ice protection. Production of helicopters and small general aviation aircraft increased in the late 1950s and 1960s. Because they fly at lower altitudes, these aircraft are much more sensitive to icing. Even though the Icing Research Tunnel at Lewis Laboratory was officially closed in 1957, a few tests of hardware under development by industry were permitted. In 1978, in response to increased industry demand, NASA officially reinstituted the tunnel testing program after the icing tunnel was renovated and 2 its instrumentation updated. The tunnel now runs more tests than other wind tunnels of NASA.21 However, the nature of the work has changed. Industry now contracts with NASA to run tests on hardware under development.
In 1947, the Committee on Operating Problems, chaired by the charismatic executive of American Airlines, William Littlewood, debated the question of whether the NACA should enter a new area of aircraft safety - the control or prevention of fires after a airplane crash.22 The Committee directed Lewis Laboratory to make a preliminary study of the crash fire problem. At the same time, it took steps to form a new Subcommittee on Aircraft Fire Prevention. The crash fire issue was of direct concern to the aircraft industry - to the aircraft manufacturers and to increasingly successful commercial airline operators like American, United, Trans World (TWA), and Pan American. They pushed for a greater involvement with the crash fire problem on the part of the NACA.
The previous year the problem of air crash fires received wide publicity. Of a surprisingly high total of 121 air carrier accidents during that year, 22, or 18 percent, involved fire. Of that number, 5 percent were attributed to fires in the air; the remaining 13 percent were due to fires following a crash.23 A new generation of commercial passenger planes made by Martin, Douglas, Lockheed, and Convair made air travel more reliable and comfortable. Even more important, the volume of travelers increased; and air travel, once only for the wealthy elite, became affordable. Yet, from the industry point....
 ....of view, the full development of commercial aviation was being held back by the public perception of its dangers.
Aircraft manufacturers tended to be more concerned with fires in the air because they were often caused by design flaws, but fires after a crash resulted in greater loss of life. Their causes, however, could not always be determined, and they were often blamed on pilot error or airline operating practices.24 Thus, the operators of commercial airlines were especially concerned that the crash fire problem be studied systematically. The airplane of the late 1940s had become a complex system whose very design contained inherent fire hazards. The new commercial airliners traveled long distances without refueling. To achieve maximum seating density, large fuel tanks were placed in the wings. For stability and aerodynamic efficiency, designers favored mounting the engines at the center of the wings. This design, however, brought ignition sources and fuel into dangerous proximity. In a crash, if the wings broke apart, a spark from the engine or the hot gases from the engine's exhaust quickly ignited the fuel, In addition to the danger of the wing-engine configuration, there were numerous other potentially dangerous ignition sources: cabin heating and ventilation, pressurization systems, and hot gas de-icing equipment. Electrical systems required miles of wires, often packed together next to other lines carrying extremely flammable hydraulic fluids or lubricating oils.25
If less theoretical than high-speed aerodynamics or heat transfer, nevertheless the study of crash fires bore directly on the NACA's charter to find practical solutions to the problems of aviation. The crash fire problem excited the imaginations of a panel on "Reduction of Hazards Due to Aircraft Fires" set up at Lewis Laboratory. Chaired by Abe Silverstein, with Lewis Rodert serving as the coordinator of the fire research program with outside agencies, the panel consisted of fuels experts R. F Selden, Louis C. Gibbon, and W T. Olson, as well as Henry C. Barnett and Gerald J. Pesman.26
In their preliminary report, members of the panel recommended a research program that would tackle the crash fire problem from many points of view. The main focus of the program was the investigation of the origin and rate of propagation of fires in actual crashes. The systematic study of full-scale airplane crashes had previously been carried out by the U.S. Army Air Corps in a study of single-engine fighter aircraft between 1924 and 1928 and in England by W G. Glendinning. However, there was no precedent for the scale and complexity of the research program envisioned by the NACA panel. Members of the panel argued that full-scale crash tests were justified because unknown scale effects made data from simulated crashes of models unreliable. They proposed basic research on the ignition characteristics of various inflammable liquids-fuels, hydraulic fluids, and lubricants - along with a study of fire-extinguishing agents. Factors involved in general layout would also be considered. Was the placement of engines on the tips of the wings, as far from the fuel tanks as possible, feasible? Could better fuel tanks be designed? What about developing a safety fuel that was less volatile than the gasoline-type fuels currently in use? Like the point of icing research, the ultimate aim of the crash fire program was to provide the federal government with reliable data to establish codes for the design of safer aircraft.27
This ambitious plan raised the hackles of the Civil Aeronautics Administration. In addition to gathering information on accidents, the CAA also conducted a limited research program that involved crash testing components at its Experimental Station in Indianapolis, Ind. With this program now regarded as inadequate, Harvey L. Hansberry of the Civil Aeronautics Administration objected that a similar full-scale program now under consideration had been proposed by the CAA in 1946 and turned down because of its excessive expense.28 However, the NACA program had  the strong backing of the Aircraft Industries Association, representing the manufacturers, and the Air Transport Association, representing the operators. In their view, the Civil Aeronautics Administration was partly responsible for the public perception of the dangers of air transportation and the financial losses due to recent crashes. Lewis Rodert pointed out after a visit to the West Coast aircraft industries, "The CAA is presently being blamed for the present situation because, in the opinion of many aircraft engineers, the CAA research has not been adequately broad nor penetrating and the airworthiness requirements relating to the fire hazard have not been realistic."29
In 1949, when Irving Pinkel inherited the icing program, he also assumed leadership of Lewis's Crash Fire Program. His background in fuels made him a natural choice for leadership of the program. However, at first the work did not particularly appeal to him. He was told that the request to the NACA had come directly from President Truman.30 The main players on the crash fire team were Dugald O. Black, Arthur Busch, Gerard J. Pesman, G. Merritt Preston, and Sol Weiss. They were backed by an intrepid group of pilots and technicians in the hangar who worked closely with William Wynne, Ernest Walker, and other members of the photographic branch.
The NACA obtained about 50 twin-engine cargo planes from the Air Force. These planes had been used in the Berlin airlift and were so service weary that they were flown to Cleveland with their doors open so that the pilot could jump, if necessary. The Army granted the NACA permission to use the grounds of one of its World War II arsenals in Ravenna, Ohio, as the site for the full-scale crash tests. John Everett designed and supervised the construction of a 2000-foot runway.
The crashes were carefully choreographed, They were to be survivable, assuming that a fire could be prevented. A plane was sent down the runway at take-off and landing speeds by remote control. It ran into a barrier that tore off the landing gear and damaged the propellers. The engine, however, had to remain attached to the main body of the aircraft. The plane then slid through a set of poles to rip open the wing tanks before sliding into an open field. The airplane was painted white and the fuel dyed red for ease of photography.
Each airplane carried various instruments to record temperatures, fire location, distribution of combustible mixtures, and times at which various failures occurred. These instruments converted the data into electric signals recorded on meters located in a fireproof, insulated box on the airplane. Seven additional motion-picture camera stands were located at various points near the runway, To correlate the exterior photographs with the data simultaneously photographed inside the box, a timing light on the top of the fuselage flashed at intervals of one second or less.31
The program's goal was to uncover the mechanism of the crash fire and the exact nature of the structural breakup of the airplane. What was the rate, pattern, and area over which the liquid fuel spread? Did it form into a spray? What, if anything, could be done to prevent it? One by one the old myths tumbled before the facts: the mistaken idea of pilots that turning off the ignition before a crash prevented fire; the belief that fuels with low volatility were safer than conventional gasoline.
By 1957 the group under Pinkel, now chief of the Fluid Systems Division, had extended their careful engineering analyses from piston engine aircraft to planes powered by turbojets. They received strong support from the airlines, particularly United, for work on a design for an....
....inerting system, despite a projected weight penalty of 1200 pounds. Pratt & Whitney was less enthusiastic. Representatives objected that over 50 holes drilled into the engine case of the J-57 for a spray system might affect the engine's reliability.32 The team designed an inerting system that could cool the hot exhaust system and cut off the fuel and electrical systems within this short interval. With the inerting system to prevent the ensuing crash fire, all but the most severe crashes were survivable.
However, the NACA had no political clout to force the airframe and engine manufacturers to install the new system. Despite the desire of some of the airline operators to purchase aircraft with crash fire protection, the manufacturers were under no obligation to provide this added equipment. The Civil Aeronautics Administration, possibly alienated by the NACA's strong showing in a field that it had previously dominated, did not take up the cause to force compliance through regulation. Unlike the icing criteria, which became the basis for federal design standards, for fire safety the NACA had to rely on persuasion, and the manufacturers were not ready to make the investment. Although the Crash Fire Program had provided convincing evidence that the inerting system could prevent crash fires, the, added weight of the system forestalled acceptance.
Even with a motion-picture film, convincingly narrated by David Brinkley, the NACA failed to entice the manufacturers to incur the added expense of the system.33
Although the aircraft manufacturers did not adopt the NACA inerting system in the mid-1950s, gradually designers incorporated safety features similar to those recommended by the NACA. 34 Fires following crashes became rare.
Concern with safety continued to shape operations research programs at Lewis Laboratory. From the Crash Fire Program, Pinkel and his group moved into the investigation of the crashworthiness of airline seats, restraining harnesses for passengers, design of seats to reduce impact forces, maximum seating density, and lightning hazards. In the late 1950s, when NASA's Mercury Program required engineers with experience in protecting human beings from the buffeting of crash landings, Lewis was ready with an experienced cadre of individuals.
As the Mercury Program got under way at Langley, some of the members of Pinkel's division became the nucleus of the Flight Operations Division at Cape Canaveral, Fla. G. Merritt Preston, who had authored many of the reports on the Crash Fire Program, became the Division's first manager.35 Other early members of the Space Task Force who transferred to Langley to work under Robert Gilruth were Elmer Buller, A. M. Busch, W. R. Dennis, M. J. Krasnican, Glynn S. Lunney, Andre J. Meyer, W, R. Meyer, W J. Nesbitt, Gerard J. Pesman, Leonard Rabb, and Scott Simkinson. Simkinson, whose background at Lewis was in testing full-scale engines, perhaps exemplified what Lewis-trained individuals could contribute to the space program. He knew hard ware and could "smell a problem a mile away."36
Other members of Pinkel's division contributed their expertise to the group at Lewis in charge of the Mercury capsule's instrumentation, automatic separation from the Atlas rocket, its stabilization and control systems, and its retrorockets for reentry into Earth's atmosphere. Their first assignment, the altitude control system for Big Joe (the test vehicle to precede manned flight), presented the first opportunity to apply their knowledge of aviation to flight beyond the atmosphere. Substituting a safer fuel, cold-gas nitrogen, for the hydrogen peroxide rocket thrusters used in the X-15, Harold Gold, Robert R. Miller, and H. Warren Plohr worked with Minneapolis-Honeywell to design a system that automatically moved the capsule into the desired alignment after separation from the Atlas booster. 37
Testing this new control system required the engineering ingenuity of a score of Lewis engineers. David S. Gabriel conceived a gigantic tinker-toy-like simulator called MASTIF (Multiple Axis Space Test Inertia Facility), which could turn and tumble a 3000-pound space capsule on  three axes inside three sets of gimbals. Louis L. Corpas executed the details of the design, and Frank Stenger developed the system of air jets to push it at 60 revolutions per minute. This "gimbal rig" was placed in Lewis's vacuum chamber, actually the old Altitude Wind Tunnel, converted by a team of engineers working under William Fleming to the new task of simulating atmospheric conditions up to 80,000 feet.38
In mid-1959, with engineering imaginations conditioned by issues involving human safety, James W Useller, a mechanical engineer, and Joseph S. Algranti, a test pilot, saw the potential of the gimbal rig for astronaut training. They enlisted several local physiologists and ten test pilots (including several women) to test the effect of roll, pitch, and yaw on human physiology. In 1960, beginning with astronauts Gus Grissom and Alan Shepard, each Mercury astronaut submitted to a ride of a tolerable limit of 30 revolutions per minute. They learned how to activate the nitrogen jets that acted as brakes to bring them out of their dizzying spin while the external cages continued to whirl about them.39
 Because of the expertise developed over years of research on the crash fire problem, NASA looked to Lewis Laboratory for assistance in determining the cause of the Apollo fire in January 1967. Irving Pinkel joined the investigation panel and spent tension-filled days and nights at Cape Canaveral. The panel concluded that the fire did not originate in the pure oxygen atmosphere of the astronauts' suits, as people had first surmised, but was caused by a short circuit that began in the cabinet that housed the environmental control unit. Wires of the electrical system ran under the door to the cabinet. Every time the door was opened, it chafed on the wires placed next to a combustible nylon netting. In short, the capsule had been poorly designed by NASA's contractor, North American Aviation. The fire had been entirely preventable. However, there was also blame to be assigned to NASA. One author called the disaster "murder on pad 34," citing sloppy management practices and a hurried atmosphere that precluded high engineering standards and careful supervision of its contractors by NASA.40
In response to the tragedy, James Webb asked Irving Pinkel to create and head an internal agency called the Aerospace Safety Research and Data Institute to serve as a clearinghouse for safety information, primarily for NASA and its contractors. The role of the government as a clearinghouse for information was not a new idea to Pinkel. That was the NACA's forte in the days when NACA reports put the latest innovations in engine technology within the reach of industry's engine designers. Located at Lewis Laboratory, the institute's functions were to organize existing safety information, to find gaps, and to fill them with appropriate research. The information was computerized to facilitate access and quick responses to requests. Within four years the institute had fallen prey to the 1972 cuts in appropriations.
1. G. M. B. Dobson, Exploring the Atmosphere (Oxford: Clarendon Press, 1968), p. 84.
2. George Gray, Frontiers of Flight (New York: Alfred A. Knopf, 1948), p. 307-329.
3. Lewis A. Rodert, William H. McAvoy, Lawrence A. Clousing, ''Preliminary Report on Flight Tests of an Airplane Having Exhaust-Heated Wings. NACA Wartime Report A-53, 1941. In 1946 President Truman presented the prestigious Collier Trophy to Rodert for his wartime thermal de-icing work.
4. In 1987 the Icing Research Tunnel was designated an International Historic Mechanical Engineering Landmark by the American Society of Mechanical Engineers, Extensive documentation from National Archives Record Group 255 and NASA Lewis Records on the design and construction of the Icing Research Tunnel is available as part of the application submitted by William Olsen to the ASME. The quotation is from the booklet prepared for this occasion.
5. Alfred W. Young and Charles N. Zelanko, "Memorandum for the Construction Administrator," National Archives, Record Group 255, Box 131. For a full description of the icing tunnel prior to 1986 refurbishing, see Uwe von Glahn, "The Icing Problem-Current Status of NACA Techniques and Research," NASA TM-81651, August 1981, appendix A. For accounts of NACA wartime icing research, see George Gray, Frontiers of Flight, p. 307-29; Edwin P. Hartman, Adventures in Research, NASA SP-4302 (Washington, D.C.: U.S. Government Printing Office, 19701, p. 69-77; Elizabeth Muenger, Searching the Horizon: A History of Ames Research Center, 1940-1976, NASA SP-4304 (Washington, D.C.: U.S. Government Printing Office, 1985), p. 19-22; James Hansen, Engineer in Charge, NASA SP-4305 (Washington, D.C.: U.S. Government Printing Office, 1987), p. 1100-111. See also comments on current use by A. Richard Tobiason, "Overview of NASA's Programs," in Meteorological and Environmental Inputs to Aviation Systems, NASA Conference Publication 2388, 1983, p. 35.
6. NACA Report 831 (1945). See discussion by Hartmann, Adventures in Research, p. 76. British studies indicated that for bombing missions the increased weight of icing protection systems meant that the bomb load would have to be reduced. See R. V Jones, The Wizard War: British Scientific Intelligence 1939-1945 (New York: Coward, McCann and Geoghegan, 1978), p. 387-388. The first report describing the rotating multicylinder method was Bernard Vonnegut, A.A.F. Tech. Report, No. 5519, 1946. The Mount Washington Monthly Research Bulletin, vol. II, no. 6, June 1946, published an anonymous description. See also Victor F. Clark, "Conditions for Run-Off and Blow-Off of Catch on Multicylinder Icing Meter." Harvard-Mt. Washington Icing Research Report 1946-7, AF Tech. Rep. 5676, p. 190-218. See also Irving Langmuir, "Super-Cooled Water Droplets in Rising Currents of Cold Saturated Air," Report No. RL-223 (1943-1944) in The Collected Works of Irving Langmuir, vol. 10 (New York: Pergamon Press, 1961), p. 199. Biographical information can be found in Dictionary of Scientific Biography (New York: Scribners, 1973), vol. 8, p. 22-25.
7. Lewis Rodert to Chief of Research, 12 March 1947, NASA Lewis Records, 34/lcing. The Civil Aeronautics Authority and Air Safety Board were established by the Civil Aeronautics Act of 1938. In 1940 they were reorganized into the Civil Aeronautics Administration and the Civil Aeronautics Board. The Federal Aviation Administration was set up in 1958.
8. Alun R. Jones and William Lewis, "Recommended Values of Meteorological Factors to Be Considered in the Design of Aircraft Ice-Prevention Equipment." NACA TN 1855, March 1949.
9. William Lewis, "Review of Icing Criteria," p. 3. From Aircraft Ice Protection, Federal Aviation Administration, April 1969 (typescript). I am indebted to the late Dr. William Olsen for providing me with a copy of this paper. These data are now the basis of icing certification in Federal Aviation Regulation (FAR 25), Appendix C. The exact date for the adoption of these regulations by the Civil Aeronautics Administration is unclear. Copies of correspondence (courtesy of William Olsen) indicate that as late as 1956 the regulations, although proposed, had not yet become law. NACA data were used in "Engineering Summary of Airframe Icing Technical Data," Federal Aviation Administration ADS-4, 1963 (known throughout the world as the "Icing Bible"), and "Engineering Summary of Powerplant Icing Technical Data," FAA RD-77-76, 1977.
10. Transcript of interview with Irving Pinkel by Walter T. Bonney, 22 July 1973, NASA History Office Archives, Washington, D.C., p. 13-14. Transcript of interview with author, Lewis Research Center, 30 January 1985, p. 19.
12. The results of the statistical survey are summarized in Porter J. Perkins, "Summary of Statistical Icing Cloud Data Measured Over the United States and North Atlantic, Pacific, and Arctic Oceans During Routine Aircraft Operations," NASA RM 1-19-59E, 1959. For an extremely helpful overview of NACA icing research see Uwe von Glahn, "The Icing Problem-Current Status of NACA Techniques and Research", reprinted in Selected Bibliography of NACA-NASA Aircraft king Publications, Lewis Research, August 1981, NASA TM-81651, I am not certain whether the IBM computer referred to in von Glahn's report is the IBM 601, a system with mechanical storage used between 1954 and 1956, or the IBM 604, the first electronic computer in use between 1955 and 1957.
13. See William Lewis, "Review of Icing Criteria," p. 4. The meter is described in detail in Porter J. Perkins, Stuart McCullough, and Ralph D. Lewis, "A Simplified Instrument for Recording and Indicating Frequency and Intensity of Icing Conditions Encountered in Flight" NACA Research Memorandum, E511316, 1951.
14. Irving Langmuir and Katherine B. Blodgett, "A Mathematical Investigation of Water Droplet Trajectories:' Technical Report No. 5418, Air Materiel Command, AAF, 19 February 1946.
15. See, for example, Rinaldo J. Brun, Helen M. Gallagher, and Dorothea E. Vogt, "Impingement of Water Droplets on NACA 65AO04 Airfoil at 8° Angle of Attack," NACA Technical Note 3155, 1954. For a description of how the differential analyzer worked, see Larry Owens, "Vannevar Bush and the Differential Analyzer: The Text and Context of an Early Computer.'' Technology and Culture 27:63-95. The quoted passage is on p. 85-86.
16. See "Progress Report Submitted by Professor H. G. Houghton [MITI, for Meeting of Subcommittee on De-Icing Problems, April 9, 1947." National Archives, Record Group 255, 115.24.
17. Thomas F. Gelder and James P. Lewis, "Comparison of Heat Transfer from Airfoil in Natural and Simulated Icing Conditions," NACA TN 2480, 1951. Hardy's original paper is cited in note 14. See also Uwe von Glahn, "The Icing Problem," note 5.
18. See Myron Tribus, "Intermittent Heating for Aircraft Ice Protection," Ph.D. Dissertation, University of California at Los Angeles, 1950.
19. See, for example, Vernon H. Gray and Dean T. Bowden, "Comparison of Several Methods of Cyclical De-Icing of a Gas-Heated Airfoil," NACA Research Memorandum E53C27, 1952.
20. R. J. Brun, William Lewis, Porter Perkins, and John Serafini, "Impingement of Cloud Droplets on a Cylinder and Procedure for Measuring Liquid-Water Content and Droplet Sizes in Supercooled Clouds by Rotating Multicylinder Method," NACA Report 1215, 1955, p. 1-43.
21. The first sign of the reawakened interest in icing research by the Federal Aviation Administration was the republication of old NACA data in "Engineering Summary of Airframe Icing Technical Data," FAA ADS-4, 1963; and "Engineering Summary of Powerplant Icing Technical Data," FAA RD-77-76.
22. Minutes of Meeting of Subcommittee on Aircraft Fire Prevention, 3 February 1948. National Archives, Record Group 255, 115.35.
23. "Preliminary Survey of Existing Data on Aircraft Fire Problem and a Suggested Research Program," 10 December 1947. National Archives, Record Group 255, 115.35.
24. Minutes of the Meeting of Subcommittee on Aircraft Fire Prevention, 3 February 1948. National Archives, Record Group 255, 115.35.
25. Allen W, Dallas, "Views of the ATA on the Aircraft Fire Problem," National Archives, Record Group 255, 115.35.
26. Minutes of Meeting of Panel on "Reduction of Hazards due to Aircraft Fires," 31 October 1947. National Archives, Record Group 255, 115.32.
27. Preliminary Survey, 10 December 1947. National Archives, Record Group 255, 115.35.
28. Minutes of the NACA Subcommittee on Aircraft Fire Prevention, 3 February 1948. National Archives, Record Group 255, 115.32.
29. Lewis A. Rodert to Director, "Visit of Mr. Lewis A. Rodert to West Coast Aircraft Industry in Regard to Fire Problem," 12 March 1948. National Archives, Record Group 255, 115.35.
30. Transcript of interview with Irving Pinkel by V. Dawson, 30 January 1985, p. 13.
31. Irving Pinkel, G. Merritt Preston, and Gerard J. Pesman, "Mechanism of Start and Development of Aircraft Crash Fires," NACA Technical Report 1133, 1953. Also typescript, unsigned and undated, by a member of the photographic branch, author's files.
32. Irving Pinkel to Associate Director, "Visit of Airline and Airframe Personnel to NACA-Lewis on January 13, 1958." NASA Lewis Records, Box 297, 116-9.
33. Film TF-26: G.M. Preston and I. I. Pinkel, "NACA Crash Fire Research," 1953. Cited in NACA Film Catalogue, Lewis Flight Propulsion Laboratory. Available at NASA Lewis Photo Laboratory.
34. Correspondence from R. S. Sliff, Federal Aviation Administration, Larry H. Hewin, U.S. Army Air Mobility Research and Development Laboratory, Marc R Dunnam, Air Force Aero Propulsion Laboratory (ASITC), and Franklin W. Kolk, American Airlines, to Irving Pinkel supports this conclusion.
35. In May 1963 G. M. Preston received a medal for outstanding leadership in the Mercury Program. The roles of Scott Simkinson, Joe Bobik, and others at Cape Canaveral are discussed in Charles Murray and Catherine Cox, Apollo: The Race to the Moon (New York: Simon & Schuster, 1989).
36. Murray and Cox, Apollo, p. 48.
37. Loyd S. Swenson, Jr., James M. Grimwood, and Charles C. Alexander, This New Ocean, NASA SP-4201 (Washington, D.C.: U.S. Government Printing Office, 1966) p. 201.
38. Ibid, p. 44.
39. Ibid, p. 244-245.
40. Eric Bergaust, Murder on Pad 34 (New York: Putnam, 1968). See also Murray and Cox, Apollo, p. 189-205. Transcript of interview with Irving Pinkel by V. Dawson, 30 January 1985, p. 18.