Atlas Propulsion

Even though the Viking rocket used alcohol and the Vanguard first stage adopted kerosene as its fuel, the next major advance in alcohol and kerosene propulsion technology came with the Atlas missile. As with the Redstone, North American Aviation designed and built the Atlas engines, which also owed a great deal to NAA’s work for the Navaho. Unlike the Redstone, the Atlas engines burned kero­sene rather than alcohol. (Both used liquid oxygen as the oxidizer.) Kerosene that would work in rocket engines was another legacy of the protean Navaho program. In January 1953, Lt. Col. Edward Hall and others from Wright-Patterson AFB insisted to Sam Hoffman that he convert from alcohol to a hydrocarbon fuel for a 120,000-pound – 122 thrust Navaho engine. Hoffman protested because the standard Chapter 3 kerosene the air force used was JP-4, whose specifications allowed a range of densities. JP-4 clogged a rocket engine’s slim cooling lines with residues. The compounds in the fuel that caused these prob­lems did not affect jet engines but would not work easily in rocket powerplants. To resolve these problems, Hoffman initiated the Rocket Engine Advancement Program, resulting in development of the RP-1 kerosene rocket fuel, without JP-4’s contaminants and variations in density. This fuel went on to power the Atlas, Thor, and Jupiter engines. The specifications for RP-1 were available in January 1957, before the delivery date of the Atlas engines.38

On October 28, 1954, the Western Development Division and Special Aircraft Projects (procurement) Office that Air Force Ma­teriel Command had located next to it issued a letter contract to NAA to continue research and development of liquid-oxygen and

kerosene (RP-1) engines for Atlas. The cooperating air force organi­zations followed this with a contract to NAA for 12 pairs of rocket engines for the series-A flights of Atlas, which tested only two outside booster engines and not the centrally located sustainer en­gine for the Atlas. The Rocketdyne Division, formed to handle the requirements of Navaho, Atlas, and Redstone, also developed the sustainer engine, which differed from the two boosters in having a nozzle with a higher expansion ratio for optimum performance at higher altitudes once the boosters were discarded.39

Using knowledge gained from the Navaho and Redstone engines, the NAA engineers began developing the MA-1 Atlas engine system for Atlases A, B, and C in 1954. (Atlas B added the sustainer engine to the two boosters; Atlas C had the same engines but included improvements to the guidance system and thinner skin on the pro­pellant tanks. Both were test vehicles only.) The MA-1, like its suc­cessors the MA-2 and MA-3, was gimballed and used the brazed "spaghetti" tubes forming the inner and outer walls of the regen­eratively cooled combustion chamber. NAA had developed the ar­rangement used in the MA-1 in 1951, perhaps in ignorance of the originator of the concept, Edward Neu at Reaction Motors. NAA/ Rocketdyne began static "hot-fire" tests of the booster engines in 1955 and of all three MA-1 engines in 1956 at Santa Susana. The two booster engines, designated XLR43-NA-3, had a specific im­pulse of 245 lbf-sec/lbm and a total thrust of 300,000 pounds, much more than the Redstone engine. The sustainer engine, designated XLR43-NA-5, had a lower specific impulse (210 lbf-sec/lbm) and a total thrust of 54,000 pounds.40

Подпись: 123 Propulsion with Alcohol and Kerosene Fuels, 1932-72 Produced in 1957 and 1958, these engines ran into failures of systems and components in flight testing that also plagued the Thor and Jupiter engines, which were under simultaneous develop­ment and shared many component designs with the Atlas. They used high-pressure turbopumps that transmitted power from the turbines to the propellant pumps via a high-speed gear train. Both Atlas and Thor used the MK-3 turbopump, which failed at high al­titude on several flights of both missiles, causing the propulsion system to cease functioning. Investigations showed that lubrication was marginal. Rocketdyne engineers redesigned the lubrication sys­tem and a roller bearing, strengthening the gear case and related parts. Turbine blades experienced cracking, attributed to fatigue from vibration and flutter. To solve this problem, the engineers ta­pered each blade’s profile to change the natural frequency and added shroud tips to the blades. These devices extended from one blade to the next, restricting the amount of flutter. There was also an explo-

Atlas Propulsion

FIG. 3.4 Technical drawing of a baffled injector similar to the one used on the Atlas MA-1 engine to prevent combustion instability by containing lateral oscillations in the combustion chamber. (Taken from Dieter K. Huzel and David H. Huang, Design of Liquid Propellant Rocket Engines [Washington, D. C.: NASA SP-125, 1967], p. 122)

sion of a sustainer engine caused by rubbing in the oxygen side of the turbopump, solved by increasing clearances in the pump and installing a liner.

Another problem encountered on the MA-1 entailed a high – frequency acoustic form of combustion instability resulting in vi­bration and increased transfer of heat that could destroy the engine 124 in less than a hundredth of a second. The solution proved to be rect – Chapter 3 angular pieces of metal called baffles, attached to a circular ring near the center of the injector face and extending from the ring to the chamber walls. Fuel flowed through the baffles and ring for cooling. The baffles and ring served to contain the transverse oscillations in much the way that the 18 pots on the V-2 had done but without the cumbersome plumbing. Together with changing the injection pat­tern, this innovation made the instability manageable. These im­provements came between the flight testing of the MA-1 system and the completion of the MA-3 engine system (1958—63).41 They showed the need to modify initial designs to resolve problems that appeared in the process of testing and the number of innovations that resulted, although we do not always know who conceived them or precisely how they came about. (But see the account below of Rocketdyne’s Experimental Engines Group for some of the explanations.)

FIG. 3.5

Technical drawing showing components of an MA-5 sustainer engine, used on the Atlas space-launch vehicle, 1983. (Photo courtesy of NASA)


С 1983-781












Atlas Propulsion

Atlas Propulsion

National Aeronautics and Space Administration Lewis Research Center

Подпись: 125 Propulsion with Alcohol and Kerosene Fuels, 1932-72 The MA-2 “was an uprated and simplified version of the MA-1," used on the Atlas D, which was the first operational Atlas ICBM and later became a launch vehicle under Project Mercury. Both MA-1 and MA-2 systems used a common turbopump feed system in which the turbopumps for fuel and oxidizer operated from a single gas gen­erator and provided propellants to booster and sustainer engines. For the MA-2, the boosters provided a slightly higher specific im­pulse, with that of the sustainer also increasing slightly. The overall thrust of the boosters rose to 309,000 pounds; that of the sustainer climbed to 57,000 pounds. An MA-5 engine was initially identical to the MA-2 but used on space-launch vehicles rather than missiles. In development during 1961-73, the booster engines went through several upratings, leading to an ultimate total thrust of 378,000 pounds (compared to 363,000 for the MA-2).

The overall MA-3 engine system contained separate subsystems for each of the booster and sustainer engines. Each engine had its own turbopump and gas generator, with the booster engines being identical to one another. The MA-3 exhibited a number of other changes from the MA-2, including greater simplification and bet­ter starting reliability resulting from hypergolic thrust-chamber ig­nition. A single electrical signal caused solid-propellant initiators and gas-generator igniters to begin the start sequence. Fuel flow

FIG. 3.6

Подпись: C-1983-780Подпись: INLETSПодпись:Подпись: PASSAGESПодпись: OXIDIZERПодпись:Подпись: IGNITIONПодпись: FUELПодпись:Подпись: BOOSTED ENGINE THRUST CHAMBER INJECTORAtlas PropulsionTechnical drawing of an injector for an MA-5 booster engine, used on the Atlas space-launch vehicle, 1983. (Photo courtesy of NASA)

National Aeronautics and Space Administration Lewis Research Center

through an igniter fuel valve burst a diaphragm holding a hypergolic cartridge and pushed it into the thrust chambers. Oxygen flow oc­curred slightly ahead of the fuel, and the cartridge with its triethyl aluminum and triethyl boron reacted with the oxygen in the thrust chamber and began combustion. Hot gases from combustion oper­ated the turbopump, a much more efficient arrangement than previ­ous turbopumps operated by hydrogen peroxide in rockets like the 126 V-2 and Redstone.

Chapter 3 The MA-3 sustainer engine had a slightly higher specific impulse of almost 215 lbf-sec/lbm but the same thrust (57,000 pounds) as the MA-2 sustainer. The total thrust of the boosters, however, went up to 330,000 pounds with a climb in specific impulse to about 250 lbf-sec/lbm. Both specific impulses were at sea level. At altitude the specific impulse of the sustainer rose to almost 310 and that of the boosters to nearly 290 lbf-sec/lbm. The higher value for the sustainer engine at altitude resulted from the nozzles that were de­signed for the lower pressure outside Earth’s atmosphere. The MA-3 appeared on the Atlas E and F missiles, with production running from 1961 to 1964.42

Most of the changes from the MA-1 to the MA-3 resulted from a decision in 1957 by Rocketdyne management to create an Experi­mental Engines Group under the leadership of Paul Castenholz, a

design and development engineer who had worked on combustion devices, injectors, and thrust chambers. He “enjoyed a reputation at Rocketdyne as a very innovative thinker, a guy who had a lot of en­ergy, a good leader." The group consisted of about 25 mostly young people, including Dick Schwarz, fresh out of college and later presi­dent of Rocketdyne. Bill Ezell, who was the development supervi­sor, had come to NAA in 1953 and was by 1957 considered an “old- timer" in the company at age 27. Castenholz was about 30. Before starting the experimental program, Ezell had just come back from Cape Canaveral, where there had been constant electrical problems on attempted Thor launches. The Atlas and Thor contracts with the air force each had a clause calling for product improvement, which was undefined, but one such improvement the group sought was to reduce the number of valves, electrical wires, and connections that all had to function in a precise sequence for the missile to operate.

Подпись: 127 Propulsion with Alcohol and Kerosene Fuels, 1932-72 The experimental engineers wanted a system with one wire to start the engine and one to stop it. Buildup of pressure from the turbopump would cause all of the valves to “open automatically by using the. . . propellant as the actuating fluid." This one-wire start arrangement became the solid-propellant mechanism for the MA-3, but the engineers under Castenholz first used it on an X-1 ex­perimental engine on which Cliff Hauenstein, Jim Bates, and Dick Schwarz took out a patent. They used the Thor engine as the start­ing point and redesigned it to become the X-1. Their approach was mostly empirical, which was different from the way rocket devel­opment had evolved by the 1980s, when the emphasis had shifted to more analysis on paper and with a computer, having simulation precede actual hardware development. In the period 1957 to the early 1960s, Castenholz’s group started with ideas, built the hard­ware, and tried it out, learning from their mistakes.

Stan Bell, another engineer in Castenholz’s group, noted a further difference from the 1980s: “We were allowed to take risks and to fail and to stumble and to recover from it and go on. Now, everything has got to be constantly successful." Jim Bates added that there were not any “mathematical models of rocket engine combustion processes" in the late 1950s and early 1960s. “There weren’t even any computers that could handle them," but, he said, “we had our experience and hindsight."

The reason the engineers in the group moved to a hypergolic ig­niter was that existing pyrotechnic devices required a delicate bal­ance. It proved difficult to get a system that had sufficient power for a good, assured ignition without going to the point of a hard start that could damage hardware. This led them to the hypergolic cartridge

(or slug) used on the MA-3. In the process of developing it, however, the group discovered that a little water in the propellant line ahead of the slug produced combustion in the line but not in the chamber; there the propellants built up and caused a detonation, “blow[ing] hell out of an engine," as Bill Ezell put it. They learned from that experience to be more careful, but Ezell said, “there’s probably no degree of analysis that could have prevented that from happening." There were simply a lot of instances in rocket-engine development where the experimenters had to “make the right guess or assump­tion"; otherwise, there was “no way to analyze it. So you’ve got to get out and get the hard experience." Ezell also opined that “with­out the Experimental Engine program going, in my opinion there never would have been a Saturn I," suggesting a line of evolution from their work to later engines.43

The experiences and comments of the members of Castenholz’s group illuminate the often dimly viewed nature of early rocket en­gineering. Without the product-improvement clauses in the Atlas and Thor contracts, a common practice of the Non-Rotating Engine Branch of the Power Plant Laboratory at Wright-Patterson AFB, the innovations made by this group probably would not have occurred. They thus would not have benefited Thor and Atlas as well as later projects like Saturn I. Even with the clauses, not every company would have put some 25 bright, young engineers to work on pure experiment or continued their efforts after the first engine explo­sion. That Rocketdyne did both probably goes a long way toward explaining why it became the preeminent rocket-engine producer in the country.

The changes in the Atlas engines to the MA-3 configuration as 128 a result of the experimental group’s work did not resolve all of the Chapter 3 problems with the Atlas E and F configurations. The Atlas lifted off with all three engines plus its two verniers (supplementary en­gines) firing. Once the missile (or later, launch vehicle) reached a predetermined velocity and altitude, it jettisoned the booster en­gines and structure, with the sustainer engine and verniers then continuing to propel the remaining part of the rocket to its destina­tion. The separation of the booster sections occurred at disconnect valves that closed to prevent the loss of propellant from the feed lines. This system worked through the Atlas D but became a major problem on the E and F models, with their independent pumps for each engine (rather than the previous common turbopump for all of them). Also, the E and F had discarded the use of water in the regen­erative cooling tubes because it reacted with the hypergolic slug. The water had ensured a gentle start with previous igniters. With

the hypergolic device, testing of the engines by General Dynamics had produced some structural damage in the rear of the missile. Design fixes included no thought of a large pressure pulse when the new models ignited.

On June 7, 1961, the first Atlas E launched from Vandenberg AFB on the California coast at an operational launch site that used a dry flame bucket rather than water to absorb the missile’s thrust. The missile lifted off and flew for about 40 seconds before a failure of the propulsion system resulted in destruction of the missile, with its parts landing on the ground and recovered. Rocketdyne specialists analyzed the hardware and data, concluding that a pressure pulse had caused the problem. The pulse had resulted in a sudden up­ward pressure from the dry flame bucket back onto fire-resistant blankets called boots that stretched from the engines’ throat to the missile’s firewall to form a protective seal around the gimballing engines. The pressure caused one boot to catch on a drain valve at the bottom of a pressurized oil tank that provided lubrication for the turbopump gearbox. The tank drained, and the gearbox ceased to operate without lubrication. To solve this problem, engineers re­sorted to a new liquid in the cooling tubes ahead of the propellants to soften ignition and preclude pressure pulses.

Подпись: 129 Propulsion with Alcohol and Kerosene Fuels, 1932-72 Repeated failures of different kinds also occurred during the flight-test program of the E and F models at Cape Canaveral. Control instrumentation showed a small and short-lived pitch upward of the vehicle during launch. Edward J. Hujsak, assistant chief engineer for mechanical and propulsion systems for the Atlas airframe and as­sembly contractor, General Dynamics, reflected about the evidence and spoke with the firm’s director of engineering. Hujsak believed that the problem lay with a change in the geometry of the propellant lines for the E and F models that allowed RP-1 and liquid oxygen (ex­pelled from the booster engines when they were discarded) to mix. Engineers “did not really know what could happen behind the mis­sile’s traveling shock front" as it ascended, but possibly the mixed propellants were contained in such a way as to produce an explosion. That could have caused the various failures that were occurring.

The solution entailed additional shutoff valves in the feed lines on the booster side of the feed system, preventing expulsion of the propellants. Engineers and technicians had to retrofit these valves in the operational missiles. However, the air force decided that since there could be no explosion if only one of the propellants were cut off, the shutoff valves would be installed only in the oxygen lines. A subsequent failure on a test flight convinced the service to approve installation in the fuel lines as well, solving the problem.44 Here


Подпись:Atlas PropulsionChapter 3

was a further example of engineers not always fully understanding how changes in a design could affect the operation of a rocket. Only failures in flight testing and subsequent analysis pinpointed prob­lem areas and provided solutions.


Meanwhile, a much smaller American effort at rocket develop­ment began at the California Institute of Technology (Caltech) in 1936. A graduate student of aerodynamicist Theodore von Karman, Frank J. Malina, together with Edward S. Forman and John W. Par­sons—described respectively by Malina as “a skilled mechanic" and “a self-trained chemist" without formal schooling but with “an un­inhibited fruitful imagination"—began to do research for Malina’s doctoral dissertation on rocket propulsion and flight.22 Gradually, 16 the research of these three men expanded into a multifaceted, pro­Chapter 1 fessional rocket development effort. As with the work under von Braun in Germany, there were many problems to be overcome. The difficulty of both endeavors lay partly in the lack of previous, de­tailed research-and-development reports. It also resulted from the

many disciplines involved. In May 1945, Homer E. Newell, then a theoretical physicist and mathematician at the U. S. Naval Re­search Laboratory, wrote that the “design, construction, and opera­tional use of guided missiles requires intimate knowledge of a vast number of subjects. Among these. . . are aerodynamics, kinemat­ics, mechanics, elasticity, radio, electronics, jet propulsion, and the chemistry of fuels."23 He could easily have added other topics such as thermodynamics, combustion processes, and materials science.

Malina and his associates consulted existing literature. Malina paid a visit to Goddard in 1936 in a fruitless attempt to gather un­published information and cooperation from the secretive New Englander.24 Initially as part of the Guggenheim Aeronautical Lab­oratory at Caltech (GALCIT), directed by von Karman (and after 1943-44 as the Jet Propulsion Laboratory [JPL]), Malina and his staff used available data, mathematics, experimentation, innovations by other U. S. rocketeers, and imagination to develop solid – and liquid-propellant JATOs (jet-assisted takeoff devices), a Private A solid-propellant test rocket, and a WAC Corporal liquid-propellant sounding rocket before Malina left JPL in 1946 and went to Europe. He ultimately became an artist and a promoter of international cooperation in astronautics.25 (Incidentally, in 1942 several of the people at GALCIT founded the Aerojet Engineering Corporation, later known as Aerojet General Corporation, to produce the rocket engines they developed. It became one of the major rocket firms in the country.)26

Подпись: 17 German and U.S. Missiles and Rockets, 1926-66 Under the successive leadership of Louis Dunn and William Pick­ering, JPL proceeded to oversee and participate in the development of the liquid-propellant Corporal and the solid-propellant Sergeant missiles for the U. S. Army. Their development encountered many problems, and they borrowed some engine-cooling technology from the V-2 to solve one problem with the Corporal, illustrating one case where the V-2 influenced U. S. missile development. The Cor­poral became operational in 1954 and deployed to Europe beginning in 1955. Although never as accurate as the army had hoped, it was far superior in this respect to the V-2. At 45.4 feet long, the Corporal was less than a foot shorter than the V-2, but its diameter (2.5 feet) was slightly less than half that of the German missile. However, even with a slightly higher performance than the V-2, its range (99 statute miles) was only about half that of the earlier missile, mak­ing it a short-lived and not very effective weapon.27

In 1953, JPL began working on a solid-propellant replacement for the Corporal, known as the Sergeant. In February 1956, a Sergeant contractor-selection committee unanimously chose the Sperry Gyro-

scope Company (a division of Sperry Rand Corp.) as a co-contractor for the development and ultimate manufacture of the missile. Mean­while, on April 1, 1954, the Redstone Arsenal, which controlled de­velopment of the missile for the army, had entered into a supple­mental agreement with the Redstone Division of Thiokol Chemical Corporation to work on the Sergeant’s solid-propellant motor. The program to develop Sergeant began officially in January 1955.28

The Sergeant missile took longer to develop than originally planned and did not become operational until 1962, by which time the U. S. Navy had completed the much more capable and important Polaris A1 and the U. S. Air Force was close to fielding the significant and successful Minuteman I. The technology of the Sergeant paled by comparison. JPL director Louis Dunn had warned in 1954 that if the army did not provide for an orderly research and development program for the Sergeant, “ill-chosen designs. . . [would] plague the system for many years." In the event, the army did fail to pro­vide consistent funding and then insisted on a compressed sched­ule. This problem was complicated by differences between JPL and Sperry and by JPL’s becoming a NASA instead of an army contractor in December 1958. The result was a missile that failed to meet its in-flight reliability of 95 percent. It met a slipped ordnance sup­port readiness date of June 1962 but remained a limited-production weapons system until June 1968. However, it was equal to its pre­decessor, Corporal, in range and firepower while being only half as large and requiring less than a third as much ground support equip­ment. Its solid-propellant motor could be ready for firing in a matter of minutes instead of the hours required for the liquid-propellant Corporal. An all-inertial guidance system on Sergeant made it vir­tually immune to enemy countermeasures, whereas Corporal de­pended on a vulnerable electronic link to guidance equipment on the ground.29

Thus, Sergeant was far from a total failure. In fact, although not at the forefront of solid-propellant technology by the time of its com­pletion, the army missile made some contributions to the develop­ment of launch-vehicle technology—primarily through a smaller, test version of the rocket. JPL had scaled down Sergeant motors from 31 to 6 inches in diameter for performing tests on various solid propellants and their designs. By 1958, the Lab had performed static 18 tests on more than 300 of the scaled-down motors and had flight – Chapter 1 tested 50 of them—all without failures. Performance had accorded well with predictions. These reliable motors became the basis for upper stages in reentry test vehicles for the Jupiter missile (called Jupiter C) and in the launch vehicles for Explorer and Pioneer satel-

lites, which used modified Redstone and Jupiter C missiles as first stages.30

Because von Braun’s group of engineers developed the Redstone and Jupiter C, this was an instance where purely American and German-American technology blended. It is instructive to compare management at JPL with that in von Braun’s operation in Germany. At JPL, the dynamic von Karman served as director of the project until the end of 1944, when he left to establish the Scientific Ad­visory Board for the U. S. Army Air Forces. Malina held the title of chief engineer of the project until he succeeded von Karman as (acting) director. But according to Martin Summerfield, head of a liquid-propellant section, there was no counterpart at GALCIT/JPL to von Braun at Peenemunde. Instead, Summerfield said, the way the professionals in the project integrated the various components of the rockets and the various developments in fields as disparate as aerodynamics and metallurgy was simply by discussing them as colleagues. He seemed to suggest that much of this was done infor­mally, but like Peenemunde, JPL also had many formal meetings where such issues were discussed. In addition, a research analysis section did a good deal of what later was called systems engineering for JPL.31

Dunn succeeded Malina as acting director of JPL on May 20, 1946, becoming the director (no longer acting) on January 1, 1947. Whereas Malina operated in an informal and relaxed way, Dunn brought more structure and discipline to JPL than had prevailed pre­viously. He was also cautious, hence concerned about the growth of the Lab during his tenure. From 385 employees in June 1946, the number grew to 785 in 1950 and 1,061 in 1953, causing Dunn to create division heads above the section heads who had reported to him directly. There were four such division heads by Septem­ber 1950, with William Pickering heading one on guided-missile electronics.

Подпись: 19 German and U.S. Missiles and Rockets, 1926-66 In August 1954, Dunn resigned from JPL to take a leading role in developing the Atlas missile for the recently established Ramo – Wooldridge Corporation. At Dunn’s suggestion, Caltech appointed Pickering as his successor. A New Zealander by birth, Pickering continued the tradition of having foreign-born directors at JPL (von Karman coming from Hungary and Malina, Czechoslovakia). Easier to know than the formal Dunn, Pickering was also less stringent as a manager. Whereas Dunn had favored a project form of organization, Pickering returned to one organized by disciplines. He remained as director until 1976. Howard Seifert, who had come to GALCIT in 1942 and worked with Summerfield on liquid-propellant develop-

ments, characterized the three JPL directors in terms of an incident when some mechanics cut off the relaxed Malina’s necktie because he was too formal. Seifert said they would never have cut Dunn’s necktie off without losing their jobs, and they would not have cut Pickering’s necktie either, but he would not have fired them for that offense alone. He added that Dunn had a rigid quality but undoubt­edly was extremely capable.32

Despite all the changes in personnel and management from Ma – lina and von Karman, through Dunn to Pickering, and despite the differences in personalities and values, one constant seems to have been a not-very-structured organization, not well suited for dealing with outside industry and the design and fielding of a weapon sys­tem, as distinguished from a research vehicle. Even Dunn’s project organization seems not to have been compatible with the kind of systems engineering soon common in missile development.33

It may be, however, that JPL’s rather loose organization in this period was conducive to innovations that it achieved in both liquid – and solid-propellant rocketry (to be discussed in ensuing chapters). In addition to the direct influence they had upon rocketry, many people from JPL besides Louis Dunn later served in positions of im­portance on other missile and rocket projects, carrying with them, no doubt, much that they had learned in their work at JPL, as well as their talents. Thus, in a variety of ways—some of them incal­culable—the early work at JPL contributed to U. S. rocketry, even though the Lab itself got out of the rocket propulsion business in the late 1950s.34

The Space Shuttle Main Engines

Despite the experience with Centaur and the Saturn upper-stage en­gines, the main engines for the Space Shuttle presented a formidable challenge, mainly because of the extreme demands placed upon the engines in a system that also used solid-propellant rocket boosters

but still required a great deal of thrust from the main engines. In a partly reusable system, NASA’s requirements for staged combus­tion and extremely high chamber pressure made development of the space shuttle main engines (SSMEs) extraordinarily difficult.

The story of this development began in one sense on June 10, 1971, when—with the general configuration of the Space Shuttle still in flux—Dale D. Myers, NASA’s associate administrator for Manned Space Flight, communicated to the directors of the Manned Spacecraft Center (MSC), the Marshall Space Flight Center (MSFC), and the Kennedy Space Center (KSC) the management plan for the Space Shuttle. This gave lead-center responsibilities to MSC but retained general direction of the program at NASA Headquarters in Washington, D. C. MSC would have responsibility for system engineering and integrating the components, with selected person­nel from MSFC and KSC collocated in Houston to support this ef­fort. Marshall would have responsibility for the main propulsion elements, while Kennedy would manage the design of launch and recovery infrastructure and launch operations.77

Подпись:Myers had managed the Navaho missile effort for North Ameri­can and had become vice president of the Space Division, where he had been the general manager for the Apollo spacecraft. He had also overseen North American Rockwell’s studies for the Space Shuttle. In addition, he had experience with aircraft projects. Thus, he came to his new job with a strong background in all aspects of the shuttle (as launch vehicle, spacecraft, and airplane). At Marshall, von Braun had moved on in 1970 to become deputy associate administrator for planning at NASA Headquarters.

His deputy director for scientific and technical matters, Eberhard Rees, had succeeded him as Marshall center director until Rees re­tired in 1973, to be succeeded by Rocco A. Petrone, who had earned a doctorate in mechanical engineering from MIT. Petrone had come from NASA Headquarters and returned there in 1974. He was suc­ceeded by William R. Lucas, a chemist and metallurgist with a doctorate from Vanderbilt University who had worked at Redstone Arsenal and then Marshall since 1952 and become deputy direc­tor in 1971. Petrone reorganized Marshall, deemphasizing in-house capabilities to oversee and test large project components and giving more authority to project officers, less to lab directors, a change Myers approved. As Rees put it, Myers was “somewhat allergic to ‘too much’ government interference" with contractors, preferring less stringent oversight than Marshall had provided in the past.78

In February 1970, Marshall had released a request for proposals for the Phase B (project definition) study of the space shuttle main

engine. Contracts went to Rocketdyne, Pratt & Whitney, and Aero­jet General. The engine was to burn liquid hydrogen and liquid oxy­gen in a 6:1 ratio at a combustion-chamber pressure of 3,000 pounds per square inch, well above that of any production engine, including the Saturn J-2, which had featured a pressure of about 787 pounds at the injector end of the 230,000-pound-thrust version. The shut­tle engine was to produce a thrust of 415,000 pounds of force at sea level or 477,000 pounds at altitude. Although Rocketdyne had built the J-2 and a development version, the J-2S, with a thrust of 265,000 pounds and chamber pressure of 1,246 pounds per square inch, Pratt & Whitney had been developing an XLR129 engine for the Air Force Rocket Propulsion Laboratory. The engine actually delivered 350,000 pounds of thrust and operated at a chamber pres­sure of 3,000 pounds per square inch during 1970.79 Pratt & Whitney thus seemed to have an advantage in the competition.

At Rocketdyne, seasoned rocket engineer Paul Castenholz, who had helped troubleshoot the F-1 combustion-instability and injector problems and had been project manager for the J-2, headed the SSME 206 effort as its first project manager, even though he was a corporate Chapter 5 vice president. He saw that there was not time to build sophisti­cated turbopumps, so he decided to build a complete combustion chamber fed by high-pressure tanks. The NASA study contract did not provide funds for such an effort, so Castenholz convinced North American Rockwell to approve up to $3 million in company funds for the effort. By 1971, testing the engine at Nevada Field Laboratory near Reno, Rocketdyne had a cooled thrust chamber that achieved full thrust for 0.45 second. The thrust was 505,700 pounds at a cham­ber pressure of 3,172 pounds per square inch, exceeding the perfor­mance of Pratt & Whitney’s XLR129 by a considerable margin.80

Funding constraints led to combining Phase C and D contracts (to include actual vehicle design, production, and operations), so on March 1, 1971, Marshall released to the three contractors a request for proposals to design, develop, and deliver 36 engines. In July NASA selected Rocketdyne as winner of the competition, but Pratt & Whitney protested the choice to the General Accounting Office (GAO) as “manifestly illegal, arbitrary and capricious, and based upon unsound, imprudent procurement decisions." On March 31, 1972, the GAO finally decided the case in favor of Rocketdyne, with the contract signed August 14, 1972. This protest delayed de­velopment, although Rocketdyne worked under interim and letter contracts until the final contract signature.81

It was not until May 1972 that Rocketdyne could begin signifi­cant work on the space shuttle main engine in something close to its

final configuration, although some design parameters would change even after that. By then, however, NASA had decided on a “parallel burn" concept in which the main engines and the solid-rocket boosters would both ignite at ground level. The space agency had already determined in 1969 that the engine would employ staged combustion, in which the hydrogen-rich turbine exhaust contrib­uted to combustion in the thrust chamber. It was the combination of high chamber pressure and staged combustion that made the SSMEs a huge step forward in combustion technology. In the mean­time, they created great problems for the shuttle, but one of them was not combustion instability, the usual plague for engine devel­opment. Castenholz and his engineers had started development of the engine with an injector based on the J-2, which had shown good stability. For the shuttle, according to Robert E. Biggs, a member of the SSME management team at Rocketdyne since 1970, the firm had added “two big preventors [of instability] on an injector that was basically stable to begin with." He evidently referred to coaxial baffles, and they seem to have worked.82

Подпись:The XLR129 had been a staged-combustion engine, and its success had given NASA and industry the confidence to use the same concept on the shuttle. But timing for such an engine’s igni­tion was both intricate and sensitive, as Rocketdyne and Marshall would learn. Rocketdyne’s design used two preburners with low – and high-pressure turbopumps to feed each of the propellants to the combustion chamber and provide the required high pressure. The XLR129 had used only a single preburner, but two of them provided finer control for the shuttle in conjunction with an engine-mounted computer, subcontracted to Honeywell for development. This computer monitored and regulated the propulsion system during start, automatically shut it down if it sensed a problem, throttled the thrust during operation, and turned off the engine at mission completion.83

By the winter and spring of 1974, development of the Honeywell controller had experienced difficulties relating to its power supply and interconnect circuits. These problems attracted the attention of NASA administrator James C. Fletcher and his deputy, George M. Low. The latter commented that Rocketdyne had done a “poor job" of controlling Honeywell, which itself had done a “lousy job" and was in “major cost, schedule, and weight difficulty." Rocketdyne had fallen behind in converting test stands at Santa Susana for test­ing components of the engine, including turbopumps. A cost over­run of about $4 million required congressional reprogramming. In a program that was underfunded to begin with, this was intolerable,

so pressured by Fletcher and Low, Rockwell International, as the firm became in 1973, shifted Castenholz to another position, replac­ing him ultimately with Dominick Sanchini, a tough veteran who had led development of the main-engine proposal in 1971. Despite 27 successful years devoted to the rocket business, with important achievements to his credit, Castenholz would no longer contribute directly to launch-vehicle development.

Meanwhile, about the same time, Marshall made J. R. Thompson its project manager for the space shuttle main engine. Trained as an aeronautical engineer at Georgia Institute of Technology, where he graduated in 1958, Thompson had worked for Pratt & Whitney before becoming a liquid-propulsion engineer at Marshall on the Saturn project in 1963, the year he earned his master’s degree in mechanical engineering at the University of Florida. He became the space engine section chief in 1966, chief of the man/systems inte­gration branch in 1969, and main-engine project manager in 1974.84

In May 1975, both component testing (at Santa Susana) and pro­totype engine testing began, the latter at NASA’s National Space 208 Technology Laboratories (the former Mississippi Test Facility). Typ- Chapter 5 ically, there was about a month between testing of a component at Santa Susana and a whole engine in Mississippi. But test personnel soon learned that the highly complicated test hardware at Santa Susana was inadequate. As Robert Frosch, NASA administrator, said in 1978, “We have found that the best and truest test bed for all major components, and especially turbopumps, is the engine it­self." Consequently, because of insufficient equipment to test com­ponents as well as engines, the program gradually ceased testing at Santa Susana between November 1976 and September 1977.85

There were many problems during testing, especially with turbo­pumps and timing. The timing problems involved “how to safely start and shut down the engine." After five years of analysis, as Biggs explained, Rocketdyne engineers had “sophisticated com­puter models that attempted to predict the transient behavior of the propellants and engine hardware during start and shutdown." Test personnel expected that the engine would be highly sensitive to minute shifts in propellant amounts, with the opening of valves be­ing time-critical. Proceeding very cautiously, testers took 23 weeks and 19 tests, with replacement of eight turbopumps, to reach two seconds into a five-second start process. It took another 12 weeks, 18 tests, and eight more turbopump changes to momentarily reach the minimum power level, which at that time was 50 percent of rated thrust. Eventually Biggs’s people developed a “safe and repeat­able start sequence" by using the engine-mounted computer, also

called the main-engine controller. “Without the precise timing and positioning" it afforded, probably they could not have developed even a satisfactory start process for the engine, so sensitive was it.

Подпись:Following purging of the propulsion system with dry nitrogen and helium to eliminate moisture (which the propellants could freeze if left in the system), then a slow cooldown using the cryo­genic propellants, full opening of the main fuel valve started the fuel flow that initially occurred from the latent heating and expan­sion that the hardware (still warmer than the liquid hydrogen) im­parted to the cryogenic propellant. However, the flow was pulsating with a pressure oscillation of about two cycles per second (hertz) until chamber pressure in the main thrust chamber stabilized after 1.5 seconds. Then oxidizer flowed to the fuel and oxidizer preburn­ers and the main combustion chamber in carefully timed sequence such that liquid oxygen arrived at the fuel preburner 1.4 seconds af­ter the full opening of the main fuel valve, at the main combustion chamber at 1.5 seconds, and at the oxidizer preburner at 1.6 seconds. Test experience revealed that a key time was 1.25 seconds into the priming sequence. If the speed of turbine revolution in the high – pressure fuel turbopump at that precise moment was not at least 4,600 revolutions per minute, the engine could not start safely. So, 1.25 seconds became a safety checkpoint.

If any “combustor prime" coincided with a downward oscilla­tion (dip) in the fuel flow, excessively high temperatures could re­sult. Other effects of inaccurate timing could be destruction of the high-pressure oxidizer turbopump. Also, a 1 to 2 percent error in valve position or a timing error of as little as a tenth of a second could seriously damage the engine. Because of these problems, the first test to achieve 50 percent of rated thrust occurred at the end of January 1976. The first test to reach the rated power level was in January 1977. Not until the end of 1978 did the engineers achieve a final version of the start sequence that precluded the problems they encountered over more than three years of testing. There were also issues with shutdown sequencing, but they were less severe than those with safe engine start, especially critical because astronauts would be aboard the shuttle when it started.86

One major instance of problems with the high-pressure tur­bopumps occurred on March 12, 1976. Earlier tests of the high – pressure liquid-hydrogen pump, both at Santa Susana and in Mis­sissippi, had revealed significant vibration levels, but not until the March 12 test had engineers recognized this as a major problem. The prototype-engine test on that day was supposed to last 65 sec­onds to demonstrate a 50 percent power level, rising to 65 percent

for a single second. The test did demonstrate 65 percent power for the first time, but engineers had to halt the test at 45.2 seconds because the high-pressure fuel turbopump was losing thrust. After the test, the pump could not be rotated with a tool used to test its torque. Investigation showed that there had been a failure of the turbine-end bearings supporting the shaft. Test data showed a ma­jor loss in the efficiency of the turbines plus a large vibration with a frequency about half the speed of the pump’s rotation. Experts immediately recognized this as characteristic of subsynchronous whirl, an instability in the dynamics of the rotors.

Although recognizing the problem, test personnel evidently did not know what to do about it in a system whose turbine-blade stresses and tip speeds were still close to the limits of technology in 1991 and must have been at the outskirts of the state of the engi­neering art 15 years earlier. In any event, to speed up a solution, the program assembled a team that ultimately included the premier ro – tordynamics experts in government, industry, and academia, from the United States and Great Britain. The pump was centrifugal, 210 driven by a two-stage turbine 11 inches in diameter that was de­Chapter 5 signed to deliver 75,000 horsepower at a ratio of 100 horsepower per pound, an order-of-magnitude improvement over previous tur­bopumps. The team studied previous liquid-hydrogen turbopumps like that on the J-2, which had exhibited subsynchronous whirl. Following a test program involving engine and laboratory tests, as well as those on components and subsystems, the investigators found 22 possible causes; the most likely appeared to be hydro­dynamic problems involving seals that had a coupling effect with the natural frequency of the rotating turbines. Efforts to decrease the coupling effect included damping of the seals and stiffening the shaft. The fixes did not totally end the whirl but did delay its in­ception from 18,000 revolutions per minute, which was below the minimum power level, to 36,000 revolutions per minute, above the rated power level.

As these design improvements increased operating speeds, in­vestigators learned that a mechanism unrelated to subsynchronous whirl was still overheating the turbine bearings, which had no lu­brication but were cooled by liquid hydrogen. The team’s extensive analysis of the cooling revealed that a free vortex was forming at the bottom of the pump’s shaft where coolant flowed. This vortex reduced the pressure, hence the flow of coolant. In a piece of cut- and-try engineering, designers introduced a quarter-sized baffle that changed the nature of the vortex and allowed more coolant to flow. This fix and the elevation of the whirl problem to above the rated

power level permitted long-duration tests of the engine for the first time by early 1977.87

This problem with the fuel pump had delayed the program, but it was not as diabolical as explosions in the high-pressure liquid – oxygen turbopump. If a fire started in the presence of liquid oxygen under high pressure, it incinerated the metal parts, usually remov­ing all evidence that could lead to a solution. After solution of the fuel-pump-whirl problem, there were four fires in the high-pressure oxygen turbopump between March 1977 and the end of July 1980. This turbopump was on the same shaft as the low-pressure oxygen turbopump that supplied liquid oxygen to the preburners. The com­mon shaft rotated at a speed of nearly 30,000 revolutions per min­ute. The high-pressure pump was centrifugal and provided as much as 7,500 gallons of liquid oxygen at a pressure higher than 4,500 pounds per square inch. An essential feature of the pump’s design was to keep the liquid oxygen fully separated from the hydrogen – rich gas that drove its turbines. To ensure separation, engineers and technicians had used various seals, drains, and purges.

Подпись:Despite such precautions, on March 24, 1977, an engine caught fire and burned so severely it removed most physical evidence of its cause. Fortunately, investigators used data from instrumentation to determine that the fire started near a complex liquid-oxygen seal. Since it was not evident what a redesign should involve, testing on other engines resumed, indicating that one of the purges did not prevent the mixing of liquid oxygen and fluids draining from hot gas. On July 25, 1977, engineers tried out a new seal intended as an interim fix. But it worked so well it became the permanent solu­tion, together with increasing the flow rate of the helium purge and other measures.88

On September 8, 1977, there was another disastrous fire originat­ing in the high-pressure oxygen turbopump. Data made it clear that the problem involved gradual breakdown of bearings on each end of the turbopump’s shaft, but there was no clear indication of the cause. Fixes included enhanced coolant flow, better balance in the rotors, heavier-duty bearings, and new bearing supports. The other two fires did not involve design flaws but did entail delays. In 1972, the shuttle program had expected to launch a flight to orbit by early March 1978. The engine and turbopump problems and many others involving the propulsion system were but some of the causes for not making that deadline, but engines would have kept the shuttle from flying that early if everything else had gone as planned.

By March 1978, the expected first-flight date had slipped to March 1979, but an engine fire and other problems caused even a Septem-

FIG. 5.3

Space Shuttle Columbia launching from Pad 39A at Cape Canaveral on the first shuttle mission, April 12, 1981. (Photo courtesy of NASA)


The Space Shuttle Main Engines

ber 1979 launch to be postponed. By early 1979, turbopumps were demonstrating longer periods between failures. By 1980, engines were expected to reach 10,000 seconds of testing apiece, a figure it had taken the entire program until 1977 to reach for all engines com­bined. But there continued to be failures in July and November 1980. Thus, not until early 1981 was the space shuttle main engine fully qualified for flight. Problems had included turbine-blade failures in the high-pressure fuel turbopump, a fire involving the main oxidizer valve, failures of nozzle feed lines, a burnthrough of the fuel pre­burner, and a rupture in the main-fuel-valve housing. But finally on April 12, 1981, the first Space Shuttle lifted off. After much trouble­shooting and empirical redesign, the main engines finally worked.89

The large number of problems encountered in the development of the space shuttle main engines resulted from its advanced design. The high chamber and pump pressures as well as an operating life of 7.5 hours greatly exceeded those of any previous engine. Each shut­tle had three main engines, which could be gimballed 10.5 degrees in each direction in pitch and 8.5 degrees in yaw. The engines could be throttled over a range from 65 to 109 percent of their rated power level (although there had been so many problems trying to demon­strate the 109 percent level in testing that it was not available on a routine basis until 2001). Moreover, the 65 percent minimum power
level (changed from the original 50 percent level) was unavailable at sea level because of flow separation. During launch, the three main engines ignited before the solid-rocket boosters. When computers and sensors verified that they were providing the proper thrust level, the SRBs ignited. To reduce vehicle loads during the period of maxi­mum dynamic pressure (reached at about 33,600 feet some 60 sec­onds after liftoff) and to keep vehicle acceleration at a maximum of 3 Gs, the flight-control system throttled back the engines during this phase of the flight. Throttling also made it feasible to abort the mis­sion either with all engines functioning or with one of them out.90

At 100 percent of the rated power level, each main engine pro­vided 375,000 pounds of thrust at sea level and 470,000 pounds at altitude. The minimum specific impulse was more than 360 lbf-

The Space Shuttle Main EnginesFIG. 5.4

The space shuttle main engine firing during a test at the National Space Technologies Laboratories (later the Stennis Space Center), Janu­ary 1, 1981, showing the regenerative cooling tubes around the circumference of the combustion chamber.

(Photo courtesy of NASA)

sec/lbm at sea level and 450 lbf-sec/lbm at altitude. This was sub­stantially higher than the J-2 Saturn engine, which had a sea-level specific impulse of more than 290 lbf-sec/lbm and one at altitude of more than 420 lbf-sec/lbm. The J-2’s thrust levels were also substan­tially lower at 230,000 pounds at altitude. Not only were the SSMEs much more powerful than the earlier engines using liquid-hydrogen technology but they were also vastly more sophisticated.91

The Atlas Space-Launch Vehicle and Its Upper Stages, 1958-90

Even before it began service as a missile, the Atlas had started to function as a launch vehicle. In December 1958, an entire Atlas (less two jettisoned booster engines) went into orbit carrying a re­peater satellite in Project Score. Then, simultaneously with their role in Project Mercury, modified Atlas missiles served as space – launch vehicles for both the air force and NASA in a variety of missions. The basic Atlas was standardized, uprated, lengthened, and otherwise modified in a variety of configurations, often indi­vidually tailored for specific missions. Engineers mated the vehicle with a number of different upper stages, of which the Agena and Centaur were the best known and most important. In these various configurations, Atlas space boosters launched satellites and space­craft for such programs as Samos, Midas, Ranger, Mariner, Pioneer, International Telecommunications Satellite Consortium (Intelsat), the Fleet Satellite Communications System (FLTSATCOM), the Defense Meteorological Satellite Program, and the Navstar Global Positioning System. Following the end of the period covered in this book, some Atlases even used strap-on solid motors to supplement their thrust at liftoff.33

After initial failures of three Atlas-Ables in 1959-60, Atlas-Agena 62 had a number of problems but became a successful launch combi – Chapter 2 nation. From February 26, 1960, until June 27, 1978, Atlas-Agenas flew approximately 110 missions, many of them classified. Mean­while, in 1962 NASA urged the air force to upgrade the Atlas D basic launch vehicle to a standardized launch configuration known as Space Launch Vehicle 3 (SLV-3), which was much more reliable than the Atlas D (96 versus 81 percent successful). A further up­grade after 1965 known as SLV-3A featured longer tanks, allowing heavier payloads in conjunction with other modifications. Because of the classified nature of many Agena missions, precise and reliable

statistics are not available, but by May 1979, on Thor, Atlas, and Titan boosters, Agena had proved itself to be a workhorse of space, achieving a reported success rate of higher than 93 percent.34

With the exception of Agena, most of the upper stages used with Atlas were derivatives from other programs. The Centaur, however, was a derivative, in a sense, of Atlas in that it used the steel-balloon tank structure envisioned by Charlie Bossart and developed for the Atlas missile. Adapting that structure to the liquid-hydrogen fuel used on the Centaur proved to be a major challenge, however. It required a lot of engineering changes when problems occurred, a major reorganization of the way Centaur was managed, and a great deal of testing. But after initial delays, it worked well.35

If Agena was the workhorse of space, Centaur was the Clydes­dale. Its powerful engines enabled it to carry heavier payloads into orbit or farther into space than Agena could manage. The Centaur could do this because it burned liquid hydrogen as well as liquid oxygen. Hydrogen offered more thrust per pound of fuel burned per second than any other chemical propellant then available—about 35 to 40 percent more than RP-1 (kerosene) when burned with liq­uid oxygen.36

This added performance allowed various versions of Atlas – Centaur to support such NASA missions as landing on the lunar surface in the Surveyor project and orbiting High-Energy Astron­omy Observatories, as well as placing 35 communications satel­lites into orbit through 1989. As with other upper stages flying on Atlas vehicles, not all of the Centaur missions were successful, but most were.37

Подпись: 63 U.S. Space-Launch Vehicles, 1958-91 The intellectual push for Centaur came from Convair Division of General Dynamics engineer Krafft Ehricke, who had worked for von Braun at Peenemunde and Huntsville and for Bell Aircraft be­fore moving to Convair. When General Dynamics managers asked him to design an upper stage for Atlas, he and some other engineers, including Bossart, decided that liquid hydrogen and liquid oxygen were the propellants they needed. Aware to some degree that liquid hydrogen’s very low density, extremely cold boiling point (-423°F), low surface tension, and wide range of flammability made it unusu­ally difficult to work with, Ehricke faced funding limitations under an air force contract that precluded performing as many tests as the propellant required—an important restriction on normal rocket­engineering practice.38

This, among other issues, prevented Convair engineers from dis­covering problems occasioned by liquid hydrogen’s unique properties as early as they otherwise might have done, necessitating redesign.

Other problems arose with Centaur engines, designed by Pratt & Whitney Division of United Aircraft Corporation. The extreme cold of liquid hydrogen required completely new design features, includ­ing the use of aluminum coated with Teflon in place of rubber gas­kets to seal pipe joints. Despite such problems plus burnthroughs of the combustion chamber that necessitated redesigns, Pratt & Whit­ney engineers conducted a successful engine run in September 1959, less than a year from the date of the initial contracts with their com­pany and Convair.39

However, explosions in engines in late 1960-early 1961 revealed other problems. One of these required an adjustment to the method of feeding the hydrogen to the combustion chamber. Because of such difficulties and resultant delays, an Atlas-Centaur did not launch on a test flight until May 8, 1962, 15 months later than planned. At the point of maximum dynamic pressure, 54.7 seconds into the launch, an explosion occurred as the liquid-hydrogen tank split open. Engineers did not discover the real cause of the problem until five years later, but meanwhile the delays and problems resulted in a complete reorganization of the Centaur program to provide better control and coordination. Funding also improved.40

Solutions to further problems and programmatic changes fol­lowed, but finally, on May 30, 1966, an Atlas-Centaur successfully launched Surveyor 1 to the Moon on the first operational Atlas – Centaur flight. Atlas-Centaur performed satisfactorily on all of the Surveyor launches, although two of the spacecraft had problems. But five of the seven missions were successful, providing more than 87,000 photographs and much scientific information valuable both for Apollo landings and for lunar studies. On Surveyors 5-7 the At­lases used longer tanks with greater propellant volumes and pay­load capacity than the earlier versions. With the longer tanks, the weight of payload that the Atlas-Centaur combination could place in 300-nautical-mile orbit rose from 8,500 pounds on the shorter version to 9,100 pounds.41

64 The longer-tank Atlas (SLV-3C) and the original Centaur (known Chapter 2 as Centaur D) launched on March 2, 1972, with a Delta third-stage solid-propellant motor, the Thiokol TE-M-364-4 (Star 37E), on the spectacular Pioneer 10 mission that was NASA’s first to the outer planets and the first to reach escape velocity from the solar system. Well before this launch, NASA, which had taken over the program from the air force, had decided to upgrade the Centaur with an im­proved guidance/control computer. The new computer allowed General Dynamics to simplify the Atlas to the SLV-3D configura­tion by removing the autopilot, programming, and telemetry units

FIG. 2.7

An Atlas – Centaur launch vehicle with the Mariner 9 space probe undergoing radio­frequency interference tests at

Kennedy Space Center in 1971. (Photo courtesy of NASA)


The Atlas Space-Launch Vehicle and Its Upper Stages, 1958-90

from the earlier, long-tank SLV-3C and having the Centaur perform those functions. The new Centaur had two configurations, the D-1A for use with Atlas and the D-1T for use with Titan space-launch vehicles. The differences between the two configurations involved details of external insulation, payload-fairing diameter, battery ca­pacity, and the like.42

Подпись: 65 U.S. Space-Launch Vehicles, 1958-91 The first use of the Centaur D-1A and SLV-3D was on the launch of Pioneer 11, which had the same mission as Pioneer 10 plus making detailed observations of Saturn and its rings. As on Pioneer 10, the mission also employed the Star 37 motor in a third stage. Launched on April 5, 1973, Pioneer 11 returned much data about Saturn, in­cluding discoveries of Saturn’s 11th moon and two new rings. Be­tween 1973 and May 19, 1983, 32 SLV-3Ds launched with Centaur

FIG. 2.8

Launch of a Titan-Centaur vehicle from Cape Canaveral Air Force Station, Febru­ary 11, 1974. The two solid – rocket motors and the core stages of the Titan appear below the Centaur upper stage. (Photo courtesy of NASA)

D-1A upper stages. With the first launch of the Intelsat У with more relay capacity (and weight) on December 6, 1980, the Centaur be­gan to use engines that were adjusted to increase their thrust. Of the 66 total 32 SLV-3D/D-1A (and the slightly modified D-1AR) launches, Chapter 2 only 2 failed. This marked a 93.75 percent success rate, with no fail­ures caused by the Centaur stage.43

During the early 1980s, General Dynamics and Pratt & Whit­ney converted to new versions of Atlas and Centaur. The Atlas G was 81 inches longer than the SLV-3D because of additions to the lengths of the propellant tanks. It developed 438,000 pounds of thrust. Pratt & Whitney made several changes to the Centaur. The first Atlas G-Centaur launched on June 9, 1984, attempting to place an Intelsat У into orbit. It did so, but the orbit was not the intended

one and was unusable for communications purposes. After a modi­fication to fix the problem, there were four successes and one failure (caused by a lightning strike). Then, on September 25, 1989, an At­las G-Centaur launched the 5,100-pound FLTSATCOM F-8 satellite into geosynchronous transfer orbit. This was the last in a series of such navy ultra-high-frequency satellites, part of a worldwide com­munications system for the DoD.44

Meanwhile, forces had been building for commercializing launch- vehicle services. The air force had become unhappy with the idea, promoted by NASA, that all DoD payloads should be transported on the Space Shuttles instead of expendable launch vehicles. There was already competition from the Ariane launch vehicle in Europe, with prospects that other countries would sell launch-vehicle ser­vices to communications-satellite purveyors and other users. On January 28, 1986, the explosion of the shuttle Challenger grounded the remaining shuttles for more than two years. Early in 1987, General Dynamics announced that it would sell Atlas-Centaur as a commercial launch vehicle. NASA then signed a commercial con­tract with the company. General Dynamics decided to designate the commercial vehicles with Roman numerals, the first being Atlas I. All would have Centaur upper stages. On July 25, 1990, the first Atlas I successfully launched the joint NASA/Air Force Combined Release and Radiation Effects Satellite into a highly elliptical geo­synchronous transfer orbit.45

Through this launch, the Centaur had had a 95 percent success rate on 76 flights. This included 42 successes in a row for Centaur D-1 and D-1A between 1971 and 1984. The Centaur had led to the use of liquid-hydrogen technology on both upper stages of the Saturn launch vehicle and in the space shuttle main engines (SSMEs). It had thus made major contributions to U. S. launch-vehicle technology.46

Подпись: 67 U.S. Space-Launch Vehicles, 1958-91 Partway through the history of Centaur and the various Atlas models used to launch it, the air force contracted with General Dy­namics, beginning on February 14, 1966, to modify Atlas Es and Fs that had been in storage since their decommissioning as missiles in 1965. The process began with the newer F models. Rocketdyne in­spected each of the MA-3 engines and fixed or replaced any part that failed to meet specifications. In 1969, the rocket division started a more extensive program of refurbishment to ensure that the engines in storage would work when called upon. After two launch failures in 1980-81, Rocketdyne rebuilt the engines at its plant, performing static tests before installing them on a launch vehicle.47

Six Atlas Ds and four Fs joined forces in launching Orbiting Vehicle One (OV-1) spacecraft, beginning with a failed launch on

January 21, 1965, by a D model and ending on August 6, 1971, with the successful launch of OV-1 20 and OV-1 21 by an F model. A number of the Atlas launch vehicles carried multiple OV-1 satel­lites, each of which included an FW-4S solid-propellant rocket motor built by the Chemical Systems Division (CSD) of United Technologies Corporation, the organization that also provided the solid-rocket motors (SRMs) for the Titan III. Although the satellite failed to orbit for a variety of causes on four of the OV-1 launches, the air force’s Aerospace Research Support Program placed 117 space experiments in orbit to study a variety of phenomena.48

An Atlas F successfully launched a radar calibration target and a radiation research payload for the air force’s Space Test Program on October 2, 1972, using the Burner II solid-propellant upper stage that usually paired with the Thor booster. Another solid-propellant upper stage that operated only once with an Atlas E or F was the Pay­load Transfer System (PTS), which used the same basic TE-M-364-4 Thiokol motor as the Stage Vehicle System (SVS), employed mul­tiple times (as a different upper-stage system from PTS) with Atlas Fs and Es. On July 13, 1974, an Atlas F and the PTS successfully launched Navigation Technology Satellite (NTS) 1 to test the first atomic clocks placed in space to confirm their design and opera­tion and provide information about signal propagation to confirm predictions for the Navstar Global Positioning System (GPS). GPS was then in development and destined to become a vital naviga­tional aid, far more sophisticated and accurate than anything that preceded it.49

SVS, built by Fairchild Space and Electronics Company of Ger­mantown, Maryland, used two TE-M-364-4 motors in two upper stages to place NTS-2 and six Navigation Development System (NDS) spacecraft into orbit between June 23, 1977, and April 26, 1980. The NDS-7 launch failed on December 18, 1981, when the Atlas E launch vehicle went out of control. The other seven satel­lites all supported the development of GPS.50

68 The air force used a different upper stage, known as SGS-II, to – Chapter 2 gether with the Atlas E to launch NDS-8 through NDS-11 between July 14, 1983, and October 8, 1985, all four launches being success­ful. McDonnell Douglas Astronautics Company made the upper stage, using two Thiokol TE-M-711-8 (Star 48) motors, also featured on the Payload Assist Module (PAM), which the Space Shuttle and Delta launch vehicle had employed since 1980. Thiokol began de­veloping the motor in 1976. It used the same hydroxy-terminated polybutadiene (HTPB)-based propellant as Thiokol’s Antares III

rocket motor, a third-generation, third-stage propulsion unit for the Scout launch vehicle.51

The Atlas Es and Fs used other upper stages to launch satellites, including one Agena D. On June 26, 1978, an Atlas F—modified to mate with the Agena and to carry the Seasat-A oceanographic satellite—placed its payload into orbit. The other major upper stage used by the Atlas Es and Fs was the Integrated Spacecraft System (ISS), with a Thiokol TE-M-364-15 motor (Star 37S). In 1977-78, this was the latest in the Star 37 series of motors, also used as an upper stage on the Thor for launching weather satellites. Beginning with a launch of Tiros N from an Atlas F on October 13, 1978, the ISS served as an upper stage for launching the NOAA-6 through NOAA-11 polar orbiting meteorological satellites plus a number of DMSP satellites. The only failure in the series was NOAA-B on May 29, 1980.52

In February 1983, the air force began operating a derivative of the SLV-3D known as the Atlas H. It used most of the basic systems on the SLV but employed GE radio-inertial guidance. The particular solid-propellant upper stage used with the Atlas H and previous At­las Es and Fs to launch the White Cloud Naval Ocean Surveillance System (NOSS) satellites was classified. The White Cloud NOSS satellites provided the DoD (primarily the navy) with ocean surveil­lance. Overall, the Atlas E and F launch vehicles had only 4 failures in 41 launches by the end of 1990, yielding a success rate of more than 90 percent. All 5 launches with the Atlas H were successful.53

Подпись: 69 U.S. Space-Launch Vehicles, 1958-91 Conceived as a missile, the Atlas became a successful and ver­satile launch vehicle, mated with a great variety of upper stages. Featuring a controversial but “brilliant, innovative, and yet simple" concept (the steel-balloon tank design), both the Atlas and the Cen­taur proved to be flexible and effective. With commercialization, the Atlas and the Centaur continued to provide launch-vehicle ser­vices beyond the period of this book and into the 21st century. The Centaur proved to be especially difficult to develop because of the peculiar properties of liquid hydrogen. But it was also hampered by initial funding arrangements and other avoidable problems. As with many rocket programs, engineers found that the existing fund of knowledge was inadequate to predict all of the problems that would occur in developing and launching an extraordinarily com­plex machine. Unforeseen problems continued into the 1990s, and engineers had to relearn the lesson that continual and sophisticated testing was the price of success, even if it did not always preclude unanticipated failures.54

Shuttle Solid-Rocket Boosters

The solid-rocket motors for the Titans III and IV carried the evo­lution of solid-propellant technology from the significant achieve­ments of the Polaris and Minuteman to a new level. The next step yielded the still larger solid-rocket boosters (SRBs) on the Space Shuttles. After UTC had developed the 7-segment solid-rocket mo­tors for the Titan, NASA decided in March 1972 to use SRBs on the shuttle. Even before this decision, the Marshall Space Flight Center had provided contracts of $150,000 each to the Lockheed Propulsion Company, Thiokol, UTC, and Aerojet General to study configura­tions of such motors. Using information from these studies, NASA issued a request for proposals (RFPs) on July 16, 1973, to which all four companies responded with initial technical and cost proposals in late August 1973, followed by final versions on October 15.

Because the booster cases would be recoverable, unlike those for the Titan III, and because they had to be rated to carry astronauts, they needed to be sturdier than their predecessors. Lockheed, UTC, and Thiokol all proposed segmented cases without welding. Al­though Aerojet had been an early developer of such cases, it ignored a requirement in the RFP and proposed a welded case without seg­mentation, arguing that such a case would be lighter, less costly, and safer, with transportation by barge to launch sites from Aero-

Space Shuttle solid-rocket booster in a test stand at a Thiokol test site in 1979. (Photo courtesy of NASA)


Shuttle Solid-Rocket Boosters

jet’s production site. Had Aerojet won the contract, it is possible that the Challenger disaster never would have occurred. However, the source evaluation board with representatives from five NASA centers and the three military services ranked Aerojet last, with a score of 655 for mission suitability. By contrast, respective scores for Lockheed, Thiokol, and UTC were 714, 710, and 710. The board selected Thiokol as winner of the competition, based on its cost, the lowest of the three, and also its perceived managerial strengths. NASA announced the selection on November 20, 1973.19

Since Thiokol had plants in Utah, NASA administrator James C. Fletcher’s home state, the decision was controversial. Lockheed pro­tested, but the General Accounting Office decided on June 24, 1974, that “no reasonable basis" existed to question the validity of NASA’s decision. Thiokol, meanwhile, proceeded with design and develop­ment based on interim contracts, the final one for design awarded on June 26, 1974, followed by one for development, testing, and produc­tion on May 15, 1975.20

Part of the legacy from which Thiokol developed the technol­ogy for its SRBs came from the air force’s Large Segmented Solid Rocket Motor Program (designated 623A), of which Aerojet’s test­ing of 100-inch-diameter solids in the early 1960s had been an early part. In late 1962 the Air Force Rocket Propulsion Laboratory at Edwards AFB inaugurated a successor program. Its purpose was

to develop large solid motors that the DoD and NASA could use for space-launch vehicles. The air force provided funding for 120- and 156-inch-diameter segmented motors and for continuation of work on thrust-vector-control systems. NASA then paid for part of the 156-inch and all of a 260-inch program. In the course of test­ing thrust-vector-control systems, Lockheed had developed a Lock – seal mounting structure that allowed the nozzle to gimbal, and Thiokol later scaled it up to the size required for large motors, call – 270 ing it Flexseal.

Chapter 7 Lockheed tested both 120- and 156-inch motors in the program, and Thiokol tested 156-inch motors with both gimballed (Flexseal) and fixed nozzles. These tests concluded in 1967, as did those for 260-inch-diameter motors by Aerojet and Thiokol. There were no direct applications of the 260-inch technologies, but participation in the 120- and 156-inch portions of the Large Segmented Solid Rocket Motor Program gave Thiokol experience and access to designs, ma­terials, fabrication methods, and test results that contributed to de­velopment of the solid-rocket boosters for the Space Shuttle. The firm also drew upon its experience with Minuteman.21

The design for the solid-rocket booster was intentionally conser­vative, using a steel case of the same type (D6AC) used on Minute – man and the Titan IIIC. The Ladish Company of Cudahy, Wiscon­sin, made the cases for each segment without welding, using the rolled-ring forging process that it had helped develop for the Titan IIIC. In this process, technicians punched a hole in a hot piece of metal and then rolled it to the correct diameter. For the shuttle, the diameter turned out to be 12.17 feet (146 inches), with the overall length of the booster being 149 feet. Each booster consisted of four segments plus fore and aft sections. The propellant consisted of the same three principal ingredients used in the first stage of the Min – uteman missile, ammonium perchlorate, aluminum, and PBAN polymer. Its grain configuration was an 11-point star in the forward end converging into a large, smooth, tapered cylindrical shape. This combination yielded a theoretical specific impulse of more than 260 lbf-sec/lbm.22

Marshall Space Flight Center sought “to avoid inventing any­thing new" in the booster’s design, according to George Hardy, proj­ect manager for the solid-rocket booster at Marshall from 1974 to 1982. The best example of this approach was the PBAN propellant. Other propellants offered higher performance, but with cost and hu­man-rating being prime considerations, Thiokol employed a tried – and-true propellant used on the first stage of Minuteman and in the navy’s Poseidon missile. As Thiokol deputy director for the booster,

Shuttle Solid-Rocket Boosters

FIG. 7.4 Technical drawing of the Space Shuttle solid-rocket booster showing its segments and internal-burning core with other components, including its nozzle with gimbal actuators for directional (vector) control of the thrust. (Photo courtesy of NASA)

John Thirkill, said in 1973, “Over the last fifteen years, we’ve loaded more than 2,500 first stage Minuteman motors and around 500 Poseidon motors with this propellant."23

The configuration of the propellant grain caused the thrust to vary, providing the boost required for the planned trajectory but keeping the acceleration to 3 Gs for the astronauts. For the first six shuttle missions, the initial thrust was 3.15 million pounds per booster. The 11-point star in the forward section of the SRB had long, narrow points, providing an extensive burning surface. As the points burned away, the surface declined, reducing the thrust as the point of maximum dynamic pressure approached at about 60 sec­onds into the launch. At 52 seconds after liftoff, the star points had burned away to provide a cylindrical perforation in both the forward and rear segments of the booster. As this burned, expanding its di­ameter, the thrust increased slightly from the 52nd to about the 80th second. Thereafter, it tapered to zero as the burning consumed the propellant at about the 120th second, when the SRBs separated from the rest of the shuttle. The separated boosters, slowed by para­chutes, soon fell into the ocean.24

A major drawback of the PBAN propellant was that about 20 per­cent of its exhaust’s weight consisted of hydrogen chloride, which not

only was toxic and corrosive but could damage the ozone layer that protected Earth from excessive ultraviolet radiation. NASA studies of the possible ozone depletion showed, however, that it would be slight, so there was no need to shift to a less powerful propellant.25

Once the Ladish Company had forged the motor cases in Wiscon­sin, the segments traveled by railroad to a firm named Cal Doran near Los Angeles. There, heat treatment imparted greater strength and toughness to the D6AC steel. Then the segments went further 272 south to Rohr Industries in Chula Vista, near San Diego, for the Chapter 7 addition of tang-and-clevis joints to the ends of the segments. On these joints, shuttle designers had departed from the Marshall ad­vice “to avoid inventing anything new." Although the shuttle field joints resembled those for Titan IIIC, in many respects they dif­fered. One key change lay in orientation. For the Titan solid-rocket motor, the single tang pointed upward from a lower segment of the case and fit into the two-pronged clevis, which encased it. This pro­tected the joint from rain or dew dripping down the case and enter­ing the joint. In the shuttle, the direction was reversed.

A second major difference lay in the Titan joint’s having used only one O-ring, whereas the shuttle employed two. Insulation on the inside of the Titan motor case protected the case, and with it, the O-ring, from excessive heating. To keep the protective mecha­nisms from shrinking in cold temperatures and then possibly al­lowing a gas blow-by when the motor was firing, there were heating strips on the Titan. Both the Titan and the shuttle used putty to improve the seal provided by the O-ring(s), but the shuttle added the second O-ring for supposed further insurance. It did not include heating strips, however. One further difference in the joints was in the number of pins holding the tang and clevis together. Whereas the Titan motor had used 240 such pins fitting into holes in the tang and clevis and linking them, the shuttle had only 177, despite its larger diameter.26 There is no certainty in counterfactual history, but perhaps if the shuttle designers had simply accepted the basic design of the Titan tang-and-clevis joints, the Challenger accident would not have occurred because of leaking hot gases through a field joint that ignited the external tank.

Unlike the field joint, the nozzle for the solid-rocket boosters did follow the precedents of the Titan solid-rocket motors and the Large Segmented Solid Rocket Motor Program. The shuttle employed car­bon-phenolic throats to ablate under the extreme heating from the flow and expansion of the hot gases from the burning propellant in the motor itself. In the case of the shuttle, the propellants burned at a temperature of 5,700°F, so ablation was needed to vaporize and

thereby prevent thermal-stress cracking followed by probable ejec­tion of portions of the nozzle. As of June 1979, the expansion ratio of the nozzle was 7.16:1, used for the first seven missions. Start­ing with the eighth mission, modifications of the nozzle increased the initial thrust of each motor from 3.15 million to 3.3 million pounds. These changes extended the length of the nozzle exit cone by 10 inches and decreased the diameter of the nozzle throat by 4 inches. The latter change increased the expansion ratio to 7.72:1, thereby adding to the booster’s thrust.27

Подпись:The nozzle was partially submerged, and for gimballing, it used the Flexseal design Thiokol had scaled up in the 156-inch motor testing from the Lockheed’s Lockseal design. It was capable of eight degrees of deflection, necessitated among other reasons by the shuttle’s now-familiar roll soon after liftoff to achieve its proper trajectory. Having less thrust, the space shuttle main engines were incapable of achieving the necessary amount of roll, and the liq­uid-injection thrust-vector-control system used on the Titan solid – rocket motors would not have met the more demanding require­ments of the shuttle. Hence the importance of the Lockseal-Flexseal development during the Large Segmented Solid Rocket Motor Pro­gram supported by both NASA and the air force.28

Although there were only four segments of the solid-rocket boosters that were joined by field joints, there were actually 11 sec­tions joined by tang-and-clevis joints. Once they had been through machining and fitting processes, they were assembled at the factory into four segments. The joints put together at the factory were called factory joints as distinguished from the field joints, which techni­cians assembled at Kennedy Space Center. Thiokol poured and cast the propellant into the four segments at its factory in Brigham City, Utah, usually doing so in matched pairs from the same batches of propellant to reduce thrust imbalances. At various times, the solid – rocket motors used four different D6AC-steel cases, with slight variations in thickness.29

In part because of its simplicity compared with the space shut­tle main engine, the solid-rocket booster required far less testing than the liquid-propellant engine. Certification for the SSMEs had required 726 hot-fire tests and 110,000 seconds of operation, but the solid-rocket boosters needed only four developmental and three qualification tests with operation of less than 1,000 seconds total— 0.9 percent of that for the SSMEs. There were, however, other tests. One was a hydroburst test on September 30, 1977, at Thiokol’s Wa­satch Division in Utah. This demonstrated that, without cracking, a case could withstand the pressures to which it would be subjected

during launch. A second hydroburst test on Sept ember 19, 1980 (with only the aft dome, two segments, and the forward dome), was also successful. There were other tests of the tang-and-clevis joints that put them under pressure until they burst. They withstood pres­sures between 1.72 and 2.27 times the maximum expected from liftoff through separation.30

The first developmental static test, DM-1 on July 18, 1977, at Thiokol’s Wasatch Division was successful, but the motor deliv – 274 ered only 2.9 million pounds of maximum thrust compared with Chapter 7 an expected 3.1 million. There were other anomalies, including ex­cessive erosion in parts of the nozzle. Modification included addi­tional ammonium perchlorate in the propellant and changed nozzle coatings. DM-2 on January 18, 1978, was another success but led to further adjustments in the design. It turned out that the rubber insulation and polymer liner protecting the case were thicker than necessary, leading to reduction in their thickness. This lowered their weight from 23,900 to 19,000 pounds. There were also modi­fications in the igniter, grain design, and nozzle coating to reduce the flame intensity of the igniter, the rate of thrust increase for the motor, and erosion of portions of the nozzle. As the motor for DM-3 was being assembled, a study of the DM-2 casing revealed that there had been an area with propellant burning between segments. This required disassembling the motor and increasing the thickness of a noncombustible inhibitor on the end of each segment. Designers also extended the rubber insulation to protect the case at the joints. This delayed the DM-3 test from July to October 19, 1978.

Again, the test was satisfactory; but although the thermal protec­tion on the nozzle had been effective, the igniter once more caused the thrust to rise too quickly. Designers could see no evident solu­tion to the rapid rate of thrust increase, an apparent tacit admis­sion that engineers did not fully understand the complex combus­tion process. It did seem evident, though, that the rate had to rise quickly to preclude thrust imbalances between the two motors, so the engineers went back to an igniter design closer to that used in the DM-1 test and simply accepted the rapid thrust rise (for the moment, at least). On February 17, 1979, DM-4 ended the four de­velopmental tests with a successful firing. The qualification tests, QM-1 through -3 from June 13, 1979, to February 13, 1980, were all successful. These seven tests furnished the data needed to qualify the solid-rocket motor for launch—excluding the electronics, hy­draulics, and other components not Thiokol’s responsibility. Other tests on booster recovery mechanisms, complete booster assem­blies, loads on the launch pad and in flight, and internal pressure

FIG. 7.5

Testing of a



following the



(Photo courtesy of NASA)


Shuttle Solid-Rocket Boosters

took place at Marshall and at the National Parachute Test Range, El Centro, California. The program completed all of these tests by late May 1980, well before the first shuttle flight.31 Of course, this was after the first planned flight, so if the main-engine development had not delayed the flights, presumably the booster development would have done so to some degree.

Smaller Solid-Propellant Stages and Boosters

Подпись: 276 Chapter 7 Even early in launch-vehicle history, some missile programs had al­ready begun to influence solid-propellant developments. In 1956, a creative group of engineers at Langley’s Pilotless Aircraft Research Division (PARD) began formulating ideas that led to the Scout launch vehicles. This group included Maxime A. Faget, later famous for designing spacecraft; Joseph G. Thibodaux Jr., who promoted the spherical design of some rocket and spacecraft motors beginning in 1955; Robert O. Piland, who put together the first multistage rocket to reach the speed of Mach 10; and William E. Stoney Jr., who be­came the first head of the group responsible for developing the Scout, which he also christened. Wallops, established as a test base for the National Advisory Committee for Aeronautics’ (NACA) Langley Memorial Aeronautical Laboratory in 1945, had a history of using rockets, individually or in stages, to gather data at high speeds on both aircraft models and rocket nose cones. These data made it pos­sible to design supersonic aircraft and hypersonic missiles at a time when ground facilities were not yet capable of providing comparable information. It was a natural step for engineers working in such a pro­gram to conceive a multistage, hypersonic, solid-propellant rocket that could reach orbital speeds of Mach 18.32

Propulsion for the Saturn Launch Vehicles

There were other developments relating to kerosene-based engines for Thor and Delta, among other vehicles, but the huge Saturn en­gines marked the most important step forward in the use of RP-1 for launch-vehicle propulsion. The H-1 engine for the Saturn I’s first stage resulted from the work of the Rocketdyne Experimental Engines Group on the X-1. Under a contract to the Army Ballistic

Propulsion for the Saturn Launch VehiclesПодпись: 131 Propulsion with Alcohol and Kerosene Fuels, 1932-72 Подпись: Missile Agency (ABMA), let on September 1, 1958, Rocketdyne suc-ceeded in building on its X-1 development to deliver the first production version of the H-1 in a little over half a year. This version had only 165,000 pounds of thrust, however, less than the Thor MB-3, Block II, but the H-1 went through versions of 188,000, 200,000, and 205,000 pounds as the Saturn project evolved, with Saturn I using the first two, and Saturn IB, the final pair.45 One problem with the 165,000-pound-thrust H-1 was that it still used a 20-gallon pressurized oil tank to lubricate the turbopump gearbox, as had the Thor-Jupiter engine. Later, 188,000-pound versions of the engine eliminated this problem by using RP-1 with an additive to lubricate the gearbox. This modification required a blender that mixed fuel from the turbopump with the additive and supplied it to the gearbox. The H-1 also featured a simplified starting sequence. Instead of auxiliary start tanks under pressure to supply oxygen and RP-1 to begin operation of the turbopump, a solid- propellant device started the turbines spinning. The engine kept the hypergolic ignition procedure used in the Atlas MA-3 and the later Thor-Jupiter engines.46 Rocketdyne delivered the first 165,000-pound H-1 engine to ABMA on April 28, 1959. Von Braun and his engineers conducted the first static test on this engine 28 days later, with an 8-second, eight-engine test following on April 29, 1960. On May 17, 1960, a second static test of eight clustered engines lasted 24 seconds and generated a thrust of 1.3 million pounds. That fall, the engine
H-1 engines, which were used in a cluster of eight to power the first stage of both the Saturn I and the Saturn IB. (Photo courtesy of NASA)

passed its preliminary flight-rating tests, leading to the first flight test on October 27, 1961.47

Meanwhile, Rocketdyne had begun uprating the H-1 to 188,000 pounds of thrust, apparently by adjusting the injectors and increas­ing the fuel and oxidizer flow rates. Although the uprated engine was ready for its preliminary flight-rating tests on September 28, 1962, its uprating created problems with combustion instability that en­gineers had not solved by that time but did fix without detriment to the schedule. The first launch of a Saturn I with the 188,000-pound engine took place successfully on January 29, 1964.48

Development had not been unproblematic. Testing for combus­tion instability (induced by setting off small bombs in the combus­tion chamber beginning in 1963) showed that the injectors inher­ited from the Thor and Atlas could not recover and restore stable combustion once an instability occurred. So Rocketdyne engineers rearranged the injector orifices and added baffles to the injector face. These modifications solved the problem. Cracks in liquid-oxygen domes and splits in regenerative-cooling tubes also required rede­sign. Embrittlement by sulfur from the RP-1 in the hotter environ­ment of the 188,000-pound engine required a change of materials in the tubular walls of the combustion chamber from nickel alloy to stainless steel. There were other problems, but the Saturn person­nel resolved them in the course of the launches of Saturn I and IB from late 1961 to early 1968.49

Because the H-1s would be clustered in two groups of four each for the Saturn I first stage, there were two types of engines. H-1Cs used for the four inboard engines were incapable of gimballing to steer the first stage. The four outboard H-1D engines did the gim – 132 balling. Both versions used bell-shaped nozzles, but the outboard Chapter 3 H-1Ds used a collector or aspirator to channel the turbopump exhaust gases, which were rich in unburned RP-1 fuel, and deposit them in the exhaust plume from the engines to prevent the still-combustible materials from collecting in the first stage’s boat tail.50

The first successful launch of Saturn I did not mean that devel­opers had solved all problems with the H-1 powerplant. On May 28, 1964, Saturn I flight SA-6 unexpectedly confirmed that the first stage of the launch vehicle could perform its function with an en­gine out, a capability already demonstrated intentionally on flight SA-4 exactly 14 months earlier. An H-1 engine on SA-6 ceased to function 117.3 seconds into the 149-second stage-one burn. Telem­etry showed that the turbopump had ceased to supply propellants. Analysis of the data suggested that the problem was stripped gears in the turbopump gearbox. Previous ground testing had revealed to

Подпись: FIG. 3.9 Launch of a Saturn IB vehicle on the Skylab 4 mission from Launch Complex 39B at Kennedy Space Center on November 16, 1973. (Photo courtesy of NASA)
Propulsion for the Saturn Launch Vehicles

Rocketdyne and Marshall technicians that there was need for rede­sign of the gear’s teeth to increase their width. Already programmed to fly on SA-7, the redesigned gearbox did not delay flight testing, and there were no further problems with H-1 engines in flight.51

Подпись: 133 Propulsion with Alcohol and Kerosene Fuels, 1932-72 None of the sources for this history explain exactly how Rock­etdyne increased the thrust of each of the eight H-1 engines from 188,000 to 200,000 pounds for the first five Saturn IBs (SA-201 through SA-205) and then to 205,000 pounds for the remaining ve­hicles. It would appear, as with the uprated Saturn I engines, that the key lay in the flow rates of the propellants into the combustion chambers, resulting in increased chamber pressure. After increasing with the shift from the 165,000- to the 188,000-pound H-1s, these flow rates increased again for the 200,000-pound and once more for the 205,000-pound H-1s.52








FIG. 3.10 Diagram of the engines used on the Saturn IB and Saturn V. The Saturn IB used eight H-1s on its first stage and a single J-2 on its second stage; the Saturn V relied on five F-1 engines for thrust in its first stage, with five J-2s in the second and one J-2 in the third stage. (Photo courtesy of NASA)

For the Saturn V that launched the astronauts and their spacecraft into their trajectory toward the Moon for the six Apollo Moon land­ings, the first-stage engines had to provide much more thrust than the eight clustered H-1s could supply. Development of the larger F-1 engine by Rocketdyne originated with an air force request in 1955. NASA inherited the reports and other data from the early de – 134 velopment, and when Rocketdyne won the NASA contract to build Chapter 3 the engine in 1959, it was, “in effect, a follow-on" effort. Since this agreement preceded a clear conception of the vehicle into which the F-1 would fit and the precise mission it would perform, designers had to operate in a bit of a cognitive vacuum. They had to make early assumptions, followed by reengineering to fit the engines into the actual first stage of the Saturn V, which itself still lacked a firm configuration in December 1961 when NASA selected Boeing to build the S-IC (the Saturn V first stage). Another factor in the design of the F-1 resulted from a decision “made early in the program. . . [to make] the fullest possible use of components and techniques proven in the Saturn I program."

Propulsion for the Saturn Launch VehiclesIn 1955, the goal had been an engine with a million pounds of thrust, and by 1957 Rocketdyne was well along in developing it. That year and the next, the division of NAA had even test-fired

Propulsion for the Saturn Launch Vehicles







Propulsion for the Saturn Launch Vehicles






Propulsion for the Saturn Launch Vehicles

Propulsion for the Saturn Launch Vehicles


such an engine, with much of the testing done at Edwards AFB’s rocket site, where full-scale testing continued while Rocketdyne did the basic research, development, and production at its plant in Canoga Park. It conducted tests of components at nearby Santa Su – sana Field Laboratory. At Edwards the future air force Rocket Pro­pulsion Laboratory (so named in 1963) had three test stands (1-A, 1-B, and 2-A) set aside for the huge engine. The 1959 contract with NASA called for 1.5 million pounds of thrust, and by April 6, 1961, Rocketdyne was able to static-fire a prototype engine at Edwards whose thrust peaked at 1.64 million pounds.53

Подпись: 135 Propulsion with Alcohol and Kerosene Fuels, 1932-72 Burning RP-1 as its fuel with liquid oxygen as the oxidizer, the F-1 did not break new ground in its basic technology. But its huge thrust level required so much scaling up that, as an MSFC publica­tion said, “An enlargement of this magnitude is in itself an inno­vation." For instance, the very size of the combustion chamber— 40 inches in diameter (20.56 inches for the H-1) with a chamber area almost 4 times that of the H-1 (1,257 to 332 square inches)—required new techniques to braze together the regenerative cooling tubes. Also because of the engine’s size, Rocketdyne adopted a gas-cooled, removable nozzle extension to make the F-1 easier to transport.54

The engine was bell shaped and had an expansion ratio of 16:1 with the nozzle extension attached. Its turbopump consisted of a single, axial-flow turbine mounted directly on the thrust chamber with separate centrifugal pumps for the oxidizer and fuel that were driven at the same speed by the turbine shaft. This eliminated the

The huge first stage of the Saturn V launch vehicle being hoisted by crane from a barge onto the B-2 test stand at the Mississippi Test Facility (later the Stennis Space Center) on January 1, 1967. Nozzles for the F-1 engines show at the bottom of the stage. (Photo courtesy of NASA)

Propulsion for the Saturn Launch Vehicles136

Chapter 3

need for a gearbox, which had been a problematic feature of many earlier engines. A fuel-rich gas generator burning the engine pro­pellants powered the turbine. The initial F-1 had the prescribed 1.5 million pounds of thrust, but starting with vehicle 504, Rock – etdyne uprated the engine to 1.522 million pounds. It did so by in­creasing the chamber pressure through greater output from the tur­bine, which in turn required strengthening components (at some expense in engine weight). There were five F-1s clustered in the S-IC stage, four outboard and one in the center. All but the center engine gimballed to provide steering. As with the H-1, there was a hypergolic ignition system.55

Perhaps the most intricate design feature of the F-1 was the in­jection system. As two Rocketdyne engineers wrote in 1989, the

injector “might well be considered the heart of a rocket engine, since virtually no other single component has such a major impact on overall engine performance." The injector not only inserted the propellants into the combustion chamber but mixed them in a pro­portion designed to produce optimal thrust and performance. “As is the case with the design of nearly all complex, high technology hardware," the two engineers added, “the design of a liquid rocket injector is not an exact science, although it is becoming more so as analytical tools are continuously improved. This is because the basic physics associated with all of the complex, combustion pro­cesses that are affected by the design of the injector are only partly known." A portion of the problem lay with the atomization of the propellants and the distribution of the fine droplets to ensure proper mixing. Even as late as 1989, “the atomization process [wa]s one of the most complex and least understood phenomena, and reli­able information [wa]s difficult to obtain." One result of less-than – optimal injector design was combustion instability, whose causes and mechanisms still in 1989 were “at best, only poorly known and understood." Even in 2006, “a clear set of generalized validated design rules for preventing combustion instabilities in new TCs [thrust chambers] ha[d] not yet been identified. Also a good uni­versal mathematical three-dimensional simulation of the complex nonlinear combustion process ha[d] not yet been developed."56

Подпись: 137 Propulsion with Alcohol and Kerosene Fuels, 1932-72 This was still more the case in the early 1960s, and it caused huge problems for development of the F-1 injector. Designers at Rocket – dyne knew from experience with the H-1 and earlier engines that injector design and combustion instability would be problems. They began with three injector designs, all based essentially on that for the H-1. Water-flow tests provided information on the spacing and shape of orifices in the injector, followed by hot-fire tests in 1960 and early 1961. But as Leonard Bostwick, the F-1 engine manager at Marshall, reported, “None of the F-1 injectors exhibited dynamic stability." Designers tried a variety of flat-faced and baffled injec­tors without success, leading to the conclusion that it would not be possible simply to scale up the H-1 injector to the size needed for the F-1. Engineers working on the program did borrow from the H-1 effort the use of bombs to initiate combustion instability, saving a lot of testing to await its spontaneous appearance. But on June 28, 1962, during an F-1 hot-engine test in one of the test stands built for the purpose at the rocket site on Edwards AFB, combustion instabil­ity caused the engine to melt.57

Marshall appointed Jerry Thomson, chief of the MSFC Liquid Fuel Engines Systems Branch, to chair an ad hoc committee to ana-

lyze the problem. Thomson had earned a degree in mechanical en­gineering at Auburn University following service in World War II. Turning over the running of his branch to his deputy, he moved to Canoga Park where respected propulsion engineer Paul Castenholz and a mechanical engineer named Dan Klute, who also “had a spe­cial talent for the half-science, half-art of combustion chamber de­sign," joined him on the committee from positions as Rocketdyne managers. Although Marshall’s committee was not that large, at Rocketdyne there were some 50 engineers and technicians assigned to a combustion devices team, supplemented by people from uni­versities, NASA, and the air force. Using essentially cut-and-try methods, they initially had little success. The instability showed no consistency and set in “for reasons we never quite understood," as Thomson confessed.58

Подпись: 138 Chapter 3 Using high-speed instrumentation and trying perhaps 40-50 de­sign modifications, eventually the engineers found a combination of baffles, enlarged fuel-injection orifices, and changed impinge­ment angles that worked. By late 1964, even following explosions of the bombs, the combustion chamber regained its stability. The en­gineers (always wondering if the problem would recur) rated the F-1 injector as flight-ready in January 1965. However, there were other problems with the injector. Testing revealed difficulties with fuel and oxidizer rings containing multiple orifices for the propellants. Steel rings called lands held copper rings through which the propel­lants flowed. Brazed joints held the copper rings in the lands, and these joints were failing. Engineers gold-plated the lands to create a better bonding surface. Developed and tested only in mid-1965, the new injector rings required retrofitting in engines Rocketdyne had already delivered.59

Overall, from October 1962 to September 1966, there were 1,332 full-scale tests with 108 injectors during the preliminary, the flight­rating, and flight-qualification testing of the F-1 to qualify the en­gine for use. According to one expert, this was “probably the most intensive (and expensive) program ever devoted primarily to solving a problem of combustion instabilities."60

The resultant injector contained 6,300 holes—3,700 for RP-1 and 2,600 for liquid oxygen. Radial and circumferential baffles divided the flat-faced portion of the injector face into 13 compartments, with the holes or orifices arranged so that most of them were in groups of five. Two of the five injected the RP-1 so that the two streams impinged to produce atomization, while the other three in­serted liquid oxygen, which formed a fan-shaped spray that mixed with the RP-1 to combust evenly and smoothly. Driven by the

FIG. 3.13

Fuel tank assembly for the Saturn V S-IC (first) stage being prepared for transportation. (Photo courtesy of NASA)


Propulsion for the Saturn Launch Vehicles

52,900-horsepower turbine, the propellant pumps delivered 15,471 gallons of RP-1 and 24,811 gallons of liquid oxygen per minute to the combustion chamber via the injector.61

Подпись: 139 Propulsion with Alcohol and Kerosene Fuels, 1932-72 Despite all the effort that went into the injector design, the turbo­pump required even more design effort and time. Engineers experi­enced 11 failures of the system during development. Two of these involved the liquid-oxygen pump’s impeller, which required use of stronger components. The other 9 failures involved explosions. Causes varied. The high acceleration of the shaft on the turbopump constituted one problem. Others included friction between mov­ing parts and metal fatigue. All 11 failures necessitated redesign or change in procedures. For instance, Rocketdyne made the turbine manifold out of a nickel-based alloy manufactured by GE, Rene 41, which had only recently joined the materials used for rocket en­gines. Unfamiliarity with its welding techniques led to cracking near the welds. It required time-consuming research and training to teach welders proper procedures for using the alloy, which could withstand not only high temperatures but the large temperature dif­ferential resulting from burning the cryogenic liquid oxygen. The final version of this turbopump provided the speed and high vol­umes needed for a 1.5-million-pound-thrust engine and did so with minimal parts and high ultimate reliability.62

Once designed and delivered, the F-1 engines required further testing at Marshall and NASA’s Mississippi Test Facility. At the lat­ter, contractors had built an S-IC stand after 1961 on the mud of a swamp along the Pearl River near the Louisiana border and the Gulf of Mexico. Mosquito ridden and snake infested, this area served as home to wild pigs, alligators, and panthers. Construction workers faced 110 bites a minute from salt marsh mosquitoes, against which nets, gloves, repellent, and long-sleeved shirts afforded little protec­tion. Spraying special chemicals from two C-123 aircraft did reduce the number of bites to 10 per minute, but working conditions re­mained challenging. Nevertheless, the stand was ready for use in March 1967, more than a year after the first static test at Marshall. But thereafter, the 410-foot S-IC stand, the tallest structure in Mis­sissippi, became the focus of testing for the first-stage engines.63

Despite static testing, the real proof of successful design came only in actual flight. For AS-501, the first Saturn V vehicle, the flight on November 9, 1967, largely succeeded. The giant launch vehicle lifted the instrument unit, command and service modules, and a boilerplate lunar module to a peak altitude of 11,240 miles. The third stage then separated and the service module’s propulsion sys­tem accelerated the command module to a speed of 36,537 feet per second (about 24,628 miles per hour), comparable to lunar reentry speed. It landed in the Pacific Ocean 9 miles from its aiming point, where the USS Bennington recovered it. The first-stage engines did experience longitudinal oscillations (known as the pogo effect), but these were comparatively minor.64

Euphoria from this success dissipated, however, on April 4, 1968, when AS-502 (Apollo 6) launched. As with AS-501, this vehicle did 140 not carry astronauts onboard, but it was considered “an all-important Chapter 3 dress rehearsal for the first manned flight" planned for AS-503. The initial launch went well, but toward the end of the first-stage burn the pogo effect became much more severe than on AS-501, reaching five cycles per second, which exceeded the spacecraft’s design speci­fications. Despite the oscillations, the vehicle continued its upward course. Stage-two separation occurred, and all five of the engines ignited. Two of them subsequently shut down, but the instrument unit compensated with longer-than-planned burns for the remain­ing three engines and the third-stage propulsion unit, only to have the latter fail to restart in orbit, constituting a technical failure of the mission, although some sources count it a success.65

As Apollo Program Director Samuel Phillips told the Senate Aero­nautical and Space Sciences Committee on April 22, 1968, 18 days

after the flight, pogo was not a new phenomenon, having occurred in the Titan II and come “into general attention in the early days of the Gemini program." Aware of pogo, von Braun’s engineers had tested and analyzed the Saturn V before the AS-501 flight and found “an acceptable margin of stability to indicate" it would not de­velop. The AS-501 flight “tended to confirm these analyses." Each of the five F-1 engines had “small pulsations," but each engine ex­perienced them “at slightly different points in time." Thus, they did not create a problem. But on AS-502, the five 1.5-million-pound engines “came into a phase relationship" so that “the engine pulsa­tion was additive."66

All engines developed a simultaneous vibration of 5.5 hertz (cy­cles per second). The entire vehicle itself developed a bending fre­quency that increased (as it consumed propellants) to 5.25 hertz about 125 seconds into the flight. The engine vibrations traveled longitudinally up the vehicle structure with their peak occurring at the top where the spacecraft was (and the astronauts would be on a flight carrying them). Alone, the vibrations would not have been a problem, but they coupled with the vehicle’s bending frequency, which moved in a lateral direction. When they intersected (with both at about the same frequency), their effects combined and mul­tiplied. In the draft of an article he wrote for the New York Times, Phillips characterized the “complicated coupling" as “analogous to the annoying feedback squeal you encounter when the microphone and loud speaker of a public address system. . . coupled." This cou­pling was significant enough that it might interfere with astronauts’ performance of their duties.67

Подпись: 141 Propulsion with Alcohol and Kerosene Fuels, 1932-72 NASA formed a pogo task force including people from Marshall, other NASA organizations, contractors, and universities. The task force recommended detuning the five engines, changing the fre­quencies of at least two so that they would no longer produce vi­brations at the same time. Engineers did this by inserting liquid helium into a cavity formed in a liquid-oxygen prevalve with a cast­ing that bulged out and encased an oxidizer feed pipe. The bulging portion was only half filled with the liquid oxygen during engine operation. The helium absorbed pressure surges in oxidizer flow and reduced the frequency of the oscillations to 2 hertz, lower than the frequency of the structural oscillations. Engineers eventually applied the solution to all four outboard engines. Technical people contributing to this solution came from Marshall, Boeing, Martin, TRW, Aerospace Corporation, and North American’s Rocketdyne Division.68 This incidence of pogo showed how difficult it was for

FIG. 3.14

Propulsion for the Saturn Launch VehiclesLaunch of the giant Saturn V on the Apollo 11 mission (July 16, 1969) that carried Neil Armstrong, Edwin Aldrin, and Michael Collins on a trajectory to lunar orbit from which Armstrong and Aldrin descended to walk on the Moon’s surface. (Photo courtesy of NASA)


Chapter 3

rocket designers to predict when and how such a phenomenon might occur, even while aware of and actively testing for it. The episode also illustrated the cooperation of large numbers of people from a variety of organizations needed to solve such problems.

The redesign worked. On the next Saturn V mission, Apollo 8 (AS-503, December 21, 1968), the five F-1s performed their mission without pogo (or other) problems. On Christmas Eve the three astro­nauts aboard the spacecraft went into lunar orbit. They completed 10 circuits around the Moon, followed by a burn on Christmas Day to return to Earth, splashing into the Pacific on December 27. For the first time, humans had escaped the confines of Earth’s imme­diate environs and returned from orbiting the Moon. There were no further significant problems with the F-1s on Apollo missions.69

The development of the huge engines had been difficult and unpre­dictable but ultimately successful.

The American Rocket Society, Reaction Motors, and the U. S. Navy

While JPL’s rocket development proceeded, there were several other efforts in the field of rocketry that contributed to the development of U. S. missile and launch-vehicle technology. Some of them started earlier than Malina’s project, notably those associated with what became (in 1934) the American Rocket Society. This organization, first called the American Interplanetary Society, had its birth on April 4, 1930. Characteristically, although Goddard became a mem­ber of the society, founding member Edward Pendray wrote, “Mem – 20 bers of the Society could learn almost nothing about the techni – Chapter 1 cal details of his work." Soon, society members were testing their own rockets with the usual share of failures and partial successes. But their work “finally culminated in. . . a practical liquid-cooled regenerative motor designed by James H. Wyld." This became the

first American engine to apply regenerative cooling (described by Oberth) to the entire combustion chamber. Built in 1938, it was among three engines tested at New Rochelle, New York, Decem­ber 10, 1938. It burned steadily for 13.5 seconds and achieved an exhaust velocity of 6,870 feet per second. This engine led directly to the founding of America’s first rocket company, Reaction Mo­tors, Inc., by Wyld and three other men who had been active in the society’s experiments. Also, it was from Wyld that Frank Ma – lina learned about regenerative cooling for the engines developed at what became JPL, one example of shared information contributing to rocket development.35

Reaction Motors incorporated as a company on December 16, 1941. It had some successes, including engines for tactical missiles; the X-1 and D-558-2 rocket research aircraft; and an early throttle­able engine for the X-15 rocket research airplane that flew to the edge of space and achieved a record speed of 6.7 times the speed of sound (Mach 6.7). Reaction Motors had never been able to develop many rockets with large production runs nor engines beyond the size of the X-15 powerplant. On April 30, 1958, Thiokol, which had become a major producer of solid-propellant rocket motors, merged with Reaction Motors, which then became the Reaction Motors Division of the Thiokol Chemical Corporation. In 1970, Thiokol decided to discontinue working in the liquid-propellant field; and in June 1972, Reaction Motors ceased to exist.36

Подпись: 21 German and U.S. Missiles and Rockets, 1926-66 Despite its ultimate failure as a business, the organization had shown considerable innovation and made lasting contributions to U. S. rocketry besides Wyld’s regenerative cooling. A second impor­tant legacy was the so-called spaghetti construction for combus­tion chambers, invented by Edward A. Neu Jr. Neu applied for a patent on the concept in 1950 (granted in 1965) but had developed the design earlier. It involved preforming cooling tubes so that they became the shells for the combustion chamber when joined to­gether, creating a strong yet light chamber. The materials used for the tubes and the methods of connecting them varied, but the firm used the basic technique on many of its engines on up through the XLR99 for the X-15. By the mid-1950s, other firms picked up on the technique or developed it independently. Rocketdyne used it on the Jupiter and Atlas engines, Aerojet on the Titan engines. Later, Rocketdyne used it on all of the combustion chambers for the Sat­urn series, and today’s space shuttle main engines still use the concept.37

Another important early contribution to later missile and launch- vehicle technology came from a group formed by naval officer Rob-

ert C. Truax. He had already begun developing rockets as an ensign at the Naval Academy. After service aboard ship, he reported to the navy’s Bureau of Aeronautics from April to August 1941 at “the first jet propulsion desk in the Ship Installation Division." There, he was responsible for looking into jet-assisted takeoff for seaplanes. He then returned to Annapolis, where he headed a jet propulsion project at the Naval Engineering Experiment Station (where Robert God­dard was working separately on JATO units nearby). Truax’s group worked closely with Reaction Motors and Aerojet on projects rang­ing from JATOs to tactical missiles. Among the officers who worked under Truax was Ensign Ray C. Stiff, who discovered that aniline and other chemicals ignited spontaneously with nitric acid. This information, shared with Frank Malina, became critical to JPL’s ef­forts to develop a liquid-propellant JATO unit. In another example of the ways technology transferred from one organization or firm to another in rocket development, once Stiff completed his five years of service with the navy, he joined Aerojet as a civilian engineer. He rose to become vice president and general manager of Aerojet’s liq­uid rocket division (1963) and then (1972) president of the Aerojet Energy Conversion Company. In 1969 he became a Fellow of the American Institute of Aeronautics and Astronautics (into which the American Rocket Society had merged) for “his notable contri­butions in the design, development and production of liquid rocket propulsion systems, including the engines for Titan I, II, and III."38

External Tank

Another major part of the shuttle propulsion system was the exter­nal tank (ET), the only major nonreusable portion of the launch ve­hicle. It was the largest (and, when loaded, the heaviest part of the Space Shuttle), at about 154 feet in length and 27.5 feet in diameter. NASA issued a request for proposals for design and construction to Chrysler, McDonnell Douglas, Boeing, and Martin Marietta on April 2, 1973. All four bidders submitted their proposals on May 17. The source selection board gave the highest technical ratings to 214 Martin Marietta and McDonnell Douglas. Martin argued that it Chapter 5 alone among the bidders had relevant experience, with the Titan III core vehicle being situated between two large solid-rocket motors. Martin’s costs were by far the lowest of the four, although the board recognized that it was bidding below true expected costs—“buying in" as it was called. But as NASA deputy administrator George Low said, “We nevertheless strongly felt that in the end Martin Mari­etta costs would, indeed, be lower than those of any of the other contenders." Consequently, on August 16, 1973, NASA selected Martin Marietta (Denver Division) to negotiate a contract for the design, development, and testing of the external tank, a selection that, this time, the other competitors did not protest. NASA re­quired assembly of the structure at the Marshall-managed Michoud facility near New Orleans.92

The external tank seemed to some to pose few technological de­mands. James Kingsbury, head of Marshall’s Science and Engineer­ing Directorate, stated, “There was nothing really challenging tech­nologically in the Tank. . . . The challenge was to drive down the cost." Similarly, Larry Mulloy, who was Marshall’s project manager for the solid-rocket booster but also worked on the tank, stated, “There was no technological challenge in the building of the Ex­ternal Tank. The only challenge was building it to sustain the very large loads that it has to carry, and the thermal environment that it is exposed to during ascent" and do so within a weight limit of about 75,000 pounds. As it turned out, however, there was in fact

a major challenge, only fully appreciated after loss of Space Shuttle Columbia on February 1, 2003, to a “breach in the Thermal Protec­tion System on the leading edge of the left wing" resulting from its being struck by “a piece of insulating foam" from an area of the external tank known as the bipod ramp. During reentry into the atmosphere, this breach allowed aerodynamic superheating of the wing’s aluminum structure, melting, and the subsequent breakup of the orbiter under increasing aerodynamic forces.93

Подпись:The external tank had to carry the cryogenic liquid-hydrogen and liquid-oxygen propellants for the three shuttle main engines. It also served as the “structural backbone" for the shuttle stack and had to withstand substantial heating as the shuttle accelerated to supersonic speeds through the lower atmosphere, where dynamic pressures were high. This heating was much more complex than on a launch vehicle like the Saturn V. At the top, the tank needed only to withstand the effects of high-speed airflow. But further down, the tank’s insulation had to encounter complex shock waves as it passed through the transonic speed range (roughly Mach 0.8 to 1.2). As the airflow became supersonic, shock waves came from the nose of the orbiter, the boosters, and the structural attachments connect­ing the tank, boosters, and orbiter. As the waves impinged on the sides of the external tank, they created heating rates up to 40 British thermal units per square foot per second. This was much smaller than the heating of a nose cone reentering the atmosphere, but it was substantial for the thin aluminum sheeting of which the exter­nal tank was formed to reduce weight.94

As designers examined the requirements for the external tank, they found that not even the arrangement of the hydrogen and oxy­gen tanks involved a simple application of lessons from the Centaur and Saturn. In both, the liquid-hydrogen tank was above the liquid – oxygen tank. Since liquid oxygen was 6 times as heavy as liquid hydrogen, this arrangement made it unnecessary to strengthen the hydrogen tank to support the heavier oxygen during liftoff. Also, with the lighter hydrogen on top, the inertial forces necessary to change the attitude of the vehicle were lower than would have been the case had the reverse arrangement prevailed. For the shuttle, however, the engines were not directly under the tanks, as was the case for the Saturn upper stages and Centaur. Instead, they were off to one side. With the heavy oxygen tank on the bottom of the ex­ternal tank, its weight would have created an inertial force difficult to overcome by gimballing of the SSMEs and the SRB nozzles. Es­pecially after the separation of the solid boosters, the weight of the oxygen tank would have tended to cause the orbiter to spin around

FIG. 5.5

Technical drawing of the Space Shuttle vehicle showing its component parts, including the external tank. (Photo courtesy of NASA)




External Tank



External Tank
External Tank

External Tank

External Tank

MS F C 7F> SA 4106—2C

the tank’s center of gravity. Placing the oxygen tank on top moved the shuttle stack’s center of gravity well forward, making steering much more feasible. But it also forced designers to make the liquid – hydrogen tank (and also an intertank structure between it and the oxygen tank) much sturdier than had been necessary on the Saturn upper stages.95

This, in turn, compounded a problem with the ET’s weight. The initial empty weight allowance had been 78,000 pounds, but in 1974, the Johnson Space Center in Houston (renamed from the Manned Spacecraft Center in 1973) reduced the goal to 75,000 pounds. Moreover, NASA asked Martin Marietta if it could not only reduce the weight but do so at no additional cost. In fact, the space agency suggested that it would be helpful actually to reduce the cost. Even though Marshall lowered the safety factor for the ET, the initial tank used on shuttle flights 1-5 and 7 weighed some 77,100 pounds. But through concerted efforts, Martin Marietta was able to achieve a 10,300-pound weight reduction for the lightweight tanks first used on flight 6 of the shuttle. The firm attained the weight re­duction through a variety of design changes, including eliminating some portions of longitudinal structural stiffeners in the hydrogen tank, using fewer circumferential stiffeners, milling some portions of the tank to a lower thickness, using a different type of aluminum that was stronger and thus allowed thinner sections, and redesign­ing anti-slosh baffling.96

The resultant external tank included a liquid-hydrogen tank that constituted 96.66 feet of the ET’s roughly 154 feet in length. It had semi-monocoque design with fusion-welded barrel sections, forward and aft domes, and five ring frames. It operated at a pres­sure range of 32 to 34 pounds per square inch and contained an anti-vortex baffle but no elaborate anti-slosh baffles, because the lightness of the liquid hydrogen made its sloshing less significant than that of liquid oxygen. The feed line from the tank allowed a maximum flow rate of 48,724 gallons per minute from its 385,265- gallon (237,641-pound) capacity. The intertank structure was much shorter at 22.5 feet. Made of both steel and aluminum, it, too, was semi-monocoque in structure with a thrust beam, skin, stringers, and panels. It contained instrumentation and a device called an umbilical plate for supply of purge gas, detection of hazardous gas escaping from the tanks, and boil-off of hydrogen gas while on the ground. The intertank also had a purge system that removed the highly combustible propellants if they escaped from their tanks or plumbing fixtures.

Подпись:Above the intertank was the liquid-oxygen tank. Its 49.33 feet of length, combined with those of the intertank and the liquid – hydrogen tank, exceeded the total length of the ET because it and its liquid-hydrogen counterpart extended into the intertank. The liquid-oxygen tank was an aluminum monocoque structure oper­ating with a pressure range of 20-22 pounds per square inch. It al­lowed a maximum of 2,787 pounds (19,017 gallons) of liquid oxy­gen to flow to the main engines when they were operating at 104 percent of their rated thrust. Containing both anti-slosh and anti­vortex mechanisms, the tank had a capacity of 143,351 gallons, or 1,391,936 pounds, of oxidizer.97

The thermal-protection system for the external tank had to withstand the complex aerodynamic heating generated by the shut­tle structure and keep the cryogenic propellants from boiling. The tank was coated with an inch of foam similar to that used on the Saturn S-II. Unlike the S-II insulation, however, which had to pro­tect only against boil-off and not against the formation of ice on the foam from the liquid hydrogen and the liquid oxygen, that on the ET could not permit ice formation, because if ice came off the tank during launch, it could easily damage the critical and delicate thermal-protection system on the orbiter. Thus, the external tank’s insulation had to be thicker than that on the S-II. It was in fact so effective that despite the extreme temperatures inside the tanks, the surface of the insulation felt “only slightly cool to the touch." For the first two shuttle flights, there was a white, fire-retardant

latex coating on top of the foam, but thereafter, following testing to determine that the foam alone provided sufficient protection during ascent, the shuttle team dispensed with this coating, saving 595 pounds and leaving the orange foam to add its distinctive color to the white of the orbiter and solid-rocket boosters at launch.98

Like the main engines, the external tank underwent extensive testing before the first shuttle launch. The entire propulsion sys­tem was, of course, designed under Marshall oversight, with Cen­ter Director Lucas continuing von Braun’s practice of using weekly notes for overall communication and systems engineering. In view of this, the Columbia Accident Investigation Board was perhaps un­fairly critical in 2003 when it wrote:

In the 1970s, engineers often developed particular facets of a design (structural, thermal, and so on) one after another and in relative isolation from other engineers working on different facets. Today, engineers usually work together on all aspects of a design as an inte­grated team. The bipod fitting [in the area where foam separated on 218 Columbia’s last flight] was designed from a structural standpoint,

Chapter 5 and the application process for foam (to prevent ice formation) and

Super Lightweight Ablator (to protect from high heating) were devel­oped separately.

However, the board went on to note in all fairness:

It was—and still is—impossible to conduct a ground-based, simul­taneous, full-scale simulation of the combination of loads, airflows, temperatures, pressures, vibration, and acoustics the External Tank experiences during launch and ascent. Therefore, the qualification testing did not truly reflect the combination of factors the bipod would experience during flight. Engineers and designers used the best methods available at the time: test the bipod and foam under as many severe combinations as could be simulated and then in­terpolate the results. Various analyses determined stresses, thermal gradients, air loads, and other conditions that could not be obtained through testing.99

Design requirements specified that the Space Shuttle system not shed any debris, but on the first shuttle flight, the external tank produced a shower of particles, causing engineers to say they would have been hard-pressed to clear Columbia for flight if they had known this would happen. When the bipod ramp lost foam on shuttle flight 7, wind-tunnel testing showed that the ramp area was

designed with an aerodynamically too steep angle, and designers changed the ramp angle from 45 degrees to a shallower 22 to 30 degrees. However, this and a later “slight modification to the ramp impingement profile" failed to prevent the destruction of Space Shuttle Columbia on February 1, 2003. It is beyond the scope of this history to discuss the Columbia accident further, but despite advances in analytical capabilities until 2003, the board was unable to pinpoint the “precise reasons why the left bipod foam ramp was lost."100

Подпись:This was so even though the board included a staff of more than 120 people aided by about 400 NASA engineers in a lengthy and extensive investigation lasting months. The reasons a definitive ex­planation was impossible included the fact that foam did not “have the same properties in all directions" or the “same composition at every point." It was “extremely difficult to model analytically or to characterize physically. . . in even relatively static conditions, much less during the launch and ascent of the Shuttle." Factors that may have caused the foam to separate and damage the wing in­cluded “aerodynamic loads, thermal and vacuum effects, vibrations, stress in the External Tank structure, and myriad other conditions" including “wind shear, associated Solid Rocket Booster and Space Shuttle Main Engine responses, and liquid oxygen sloshing in the External Tank." Even in 2003, “Non-destructive evaluation tech­niques for determining External Tank foam strength have not been perfected or qualified." 101

With statements such as, “In our view, the NASA organizational culture had as much to do with this accident as the foam," the ac­cident investigation board clearly implicated more than technol­ogy in the causes of the Columbia accident. But a major cause was NASA and contractor engineers’ failure to understand the reasons for and full implications of foam shedding from the external tank. As well-known space commentator John Pike said, “The more they study the foam, the less they understand it." And as a newspaper article stated, “Getting every ounce of the foam to stick to the ex­ternal tank has bedeviled NASA engineers for 22 years. . . . Why foam falls off any area of the tank remains a scientific mystery." In the more sober language of the CAIB report, “Although engineers have made numerous changes in foam design and application in the 25 years the External Tank has been in production, the problem of foam-shedding has not been solved." 102

Whatever the larger causes of the accident, from the perspective of this book, this was but one more instance in which engineers did not have the design, development, and operation of rockets “down

to a science." Despite countless billions of dollars spent on research­ing, developing, and operating a large number of missiles and rock­ets; despite a great deal of effort on NASA’s and contractors’ parts to understand and correct this particular problem, there were aspects of rocketry (including this one) that eluded the understanding of engineers and even scientists such as investigation board member Douglas D. Osheroff, a Nobel Prize-winning physicist from Stan­ford University. Osheroff had conducted some simple experiments with foam that helped him understand the “basic physical proper­ties of the foam itself" but also demonstrated “the difficulty of un­derstanding why foam falls off the external tank." As he said, “At­tempts to understand [the] complex behavior and failure modes" of the components of the shuttle stack were “hampered by their strong interactions with other systems in the stack."103

The Scout Family of Space-Launch Vehicles, 1958-91

With its development overlapping that of Atlas and other launch ve­hicles, the Scout series of boosters was unique in being the first mul­tistage booster to operate exclusively with solid-propellant motors. It remained the smallest multistage vehicle in long-term use for or­bital launches. And it was the only launch vehicle developed under the auspices of Langley Research Center, which made many con­tributions to space efforts but, as the oldest of NASA’s component organizations, had a long heritage of aeronautical rather than space – related research. Like the Delta, with which it shared many stages, Scout proved to be long-lasting and reliable. But in contrast with Delta, it suffered through a difficult gestation and childhood.55

Because, like Delta, it used much technology that had already been developed elsewhere, Scout’s problems lay less in the design – and-development area than was true with many other rockets, al­though there were several developmental difficulties. But Scout’s problems were primarily matters of systems engineering and quality control. Following a series of early failures, the program underwent a reliability improvement and recertification process, after which one Scout engineer stated that he and his colleagues had “all un­derestimated the magnitude of the job" when they had undertaken its development. “The biggest problem we had was denying the ex­istence of problems that we did not understand." Once the project accepted that it had such problems and examined them, it learned from the process and went on to produce a long-lived, reliable, small launcher used by NASA, the DoD, and foreign countries.

Scout’s payload capability increased almost fourfold by its fi­nal flight in 1994. At that time, it had launched a great variety of scientific and applications payloads, Transit navigation satellites, and experiments to help understand the aerodynamics of reentry, among other types of missions. Counting two partial successes as failures, Scout had 103 successful missions out of 118 for an overall 70 8 7 percent success rate, according to one source. The 15 failures

Chapter 2 were mostly in the early years, with 12 of them occurring by June 1964. In the 91 missions since that time, only 6 failures or partial failures occurred for a 94 percent success rate.56

During 1956, the idea for Scout arose at Langley’s Pilotless Air­craft Research Division (PARD) on remote Wallops Island in the At­lantic Ocean off Virginia’s Eastern Shore. There, several engineers conceived of a four-stage solid-propellant launch vehicle. Between July 1, 1960, and March 29, 1962, Scout had nine developmental flights from Wallops, with six of them counted as successes. Several

FIG. 2.9

The Scout Family of Space-Launch Vehicles, 1958-91Подпись: 71 U.S. Space-Launch Vehicles, 1958-91

Подпись: developmental problems led to upgrades of the third- and fourth- stage motors.57 While NASA was in the process of developing Scout, the air force worked with the civilian agency on a military version called the Blue Scout. Meetings with the air force had begun before the creation of NASA, as early as June 4, 1958. By the end of February 1959, the air force had assigned primary responsibility for the development of the Blue Scout to its Ballistic Missile Division, with a project office at BMD being set up under Maj. (soon-to-be Lt. Col.) Donald A. Stine. Because of the payloads the air force expected to launch, its vehicle required thicker walls and more mounting studs for the third and fourth stages. By September 1960, the air force had evolved its designs to include a Blue Scout 1, Blue Scout 2, and

Launch of Scout ST-5 (Scout Test 5) on June 30, 1961, from Wallops Island, Virginia—a failure because the third stage of the Scout did not ignite, which prevented the satellite from going into orbit. (Photo courtesy of NASA)

Blue Scout Junior. All of them used the Castor I (the regular Scout second-stage motor) in place of the usual Algol I in the first stage, and the Antares I (the regular Scout third-stage motor) in the sec­ond stage. The third stage used the Aerojet 30-KS-8000 motor, also known as Alcor. The motor for the fourth stage of most Blue Scout Juniors was a unit designed by the Naval Ordnance Test Station, known as NOTS model 100A. The B version of the same motor later became the fifth-stage propulsion unit for NASA Scouts using that many stages.58

The first Blue Scout Junior launched on September 21, 1960, be­fore the fourth-stage motor’s development was complete, but ap­parently the vehicle did use a NOTS 100A. In all, there were 25 known Blue Scout Junior launches from Cape Canaveral, Vanden – berg (or the navy’s nearby Point Arguello), and Wallops Island, with the last one on November 24, 1970. All were suborbital, 22 of them successful, for an 88 percent success rate, although the telemetry and payloads sometimes failed. The configurations of the vehicles varied, depending on the mission, with some launches using only three stages and a supersonic combustion ramjet test employing only one.59

The Blue Scout 1 was a three-stage version of the Scout. Its first (successful) launch occurred at Cape Canaveral on January 7, 1961. Blue Scout 2 was a four-stage vehicle. Most sources list only three flights in 1961. But other sources continue to list launch vehicles as Blue Scouts, so the precise history of the vehicle is quite nebu­lous. The navy procured some of them at least as late as fiscal year 1967, and the air force, until fiscal year 1976. Of the first 92 Scouts, NASA paid for 54; the navy, 19; and the air force, 14, with the other 5 being funded by the Atomic Energy Commission or European us­ers. Whereas earlier Blue Scout vehicles had been launched by uni­formed (“blue suit") air force personnel, on January 10, 1970, an agreement between NASA and the DoD stated that NASA would contract for Scout launches from Vandenberg AFB for both itself 72 and the DoD. Thus, it appears that there was a gradual blurring of Chapter 2 the lines between Blue and NASA Scouts. But whichever they were called, they continued to perform launch services for the armed forces as well as the civilian space agency.60

Meanwhile, a number of Scout failures in the early years led in 1963 to a major review of the program. This revealed that no two Scout failures had been caused by the same problem. But the large number of failures, including many recent ones, suggested a need for greater procedural consistency and for requalifying all Scout vehicles then in storage awaiting launch. As Eugene Schult, head

of the Langley Scout Project Office in 1990, remembered, “We did things differently at Wallops than at the Western Test Range. The Air Force had its own way of doing things; the contractor had his ways; and we had our ways. It was a problem trying to coordinate them."61

To address these problems, a team from NASA, the LTV Missile Group of the Chance Vought Corporation (the airframe and prime contractor for Scout), and the air force initiated procedures that brought the manufacturing and launch teams in closer contact to improve coordination and quality control. (Obviously, this entailed exchange of information as well.) In addition, all 27 existing Scout vehicles went back to the LTV plant for disassembly and X-ray or microscope inspection. Standardization became the order of the day. The first recertified Scout, S-122R, with the R indicating that it had been refurbished and recertified, launched from Vandenberg December 19, 1963. It was the beginning of a series of 26 launches through October 1966 with only 1 failure, for a 96 percent success rate.62 Standardization and quality control had greatly improved re­liability, showing the value of improved management and better systems engineering.

In this period and after, Scout continued to develop, with new stages replacing those already in use. These changes increased the payload and other capabilities of the Scout system. Beginning on April 26, 1967, Scout also began launching (under agreement with Italy) from the San Marcos platform off the coast of Kenya, Africa, on the equator. From there, Scout could place satellites into orbits not achievable by launches from Cape Canaveral, let alone Wallops and Vandenberg, the three U. S. launch sites for the vehicle. As a re­sult of a long series of improvements, the payload capacity of Scout increased from only 131 pounds into a 300-mile circular orbit for the original Scout in 1960 to 454 pounds by October 30, 1979. The Scout continued in operation through August 5, 1994, with all of the remaining launches using this last (G-1) configuration.63

Подпись: 73 U.S. Space-Launch Vehicles, 1958-91 Operating for nearly three and a half decades, Scout obviously was successful. Neither its payload capacity nor reliability matched those of the Delta. But it filled a niche in the launch-vehicle spec­trum, or it would not have lasted so long. In the process, it had to overcome some initial growing pains. Many of its motors and other components experienced developmental problems, including the Castor I, Antares I, and Altair II, as well as heat shields, a fourth – stage frangible diaphragm, and the nozzles on the Algol IIA and IIB. Thus, like other missiles and launch vehicles, Scout also suffered from the frequent inability of designers to foresee problems their

handiwork might face. But as in so many other cases, engineers were able to correct the problems once they understood them and/ or brought their experience and knowledge to bear on available data.


In 1957, after a five-stage rocket vehicle at Wallops had reached speeds of Mach 15, PARD engineers began to study in earnest how to increase the speed of solid-propellant combinations even further. The group learned that Aerojet had developed the largest solid – propellant motor then in existence as part of its effort to convert the Jupiter to a solid-propellant missile for use aboard ship. Called the Jupiter Senior, the motor was 30 feet long and 40 inches in di­ameter, and it weighed 22,650 pounds, more than 3 times as much as the contemporary Sergeant missile’s motor. Using a propellant of polyurethane, ammonium perchlorate, and aluminum, the Jupiter Senior motor provided a thrust of up to about 100,000 pounds for 40 seconds in two successful static firings in March-April 1957. It eventually amassed a record of 13 static tests and 32 flights without a failure, and it prepared the way for the Aerojet motors used in Polaris and Minuteman.

About the time that the PARD engineers learned about Jupiter Senior, they found out Thiokol had discovered a way to improve the Sergeant motor by shifting from the polysulfide binder used on the missile to a polybutadiene-acrylic acid binder with metallic addi­
tives. This offered a possible 20 percent increase in specific impulse. These developments led Stoney to analyze a four-stage vehicle with the Jupiter Senior as stage one, the improved Sergeant as stage two, and two Vanguard X248 motors as the third and fourth stages. Even after Sputnik, in early 1958 NACA Headquarters told the PARD team that it would not be receptive to developing a fourth launch vehicle when Vanguard, Jupiter C, and Thor-Able were well along in development or already available.33

Подпись:However, with plans moving forward for what became NASA, in March 1958 NACA Headquarters asked Langley Aeronautical Laboratory to prepare a program of space technology. As a neces­sary part of this program, Langley included Scout—only later an acronym meaning Solid Controlled Orbital Utility Test system—to investigate human space flight and problems of reentry. The pro­gram called for $4 million to fund five vehicles for these purposes. By May 6, 1958, when Scout had become part of the space program, further analysis suggested that the third stage needed to be larger than the X248. But by then, plans for America’s space efforts were becoming so extensive that the extra costs for such development were hardly significant. By that time Langley had also arranged a contract with Thiokol for four improved Sergeant motors.34

Langley assigned Stoney as project officer but gave Thibodaux’s Rocket Section at PARD the responsibility for the initial five con­tracts needed to develop the Scout. A contract with Aerojet for the first stage became effective on December 1, 1958. The name of the Aerojet first stage changed from Jupiter Senior to Aerojet Senior, also called Algol I. As developed by December 1959, the motor was 29.8 feet long and 40 inches in diameter. With a steel case and poly­urethane-aluminum-ammonium perchlorate propellant configured in an eight-point gear (a cylinder with eight gear-shaped, squared – off “points" radiating from it), it yielded a specific impulse of only about 215 lbf-sec/lbm and a low mass fraction of 0.838 but an aver­age thrust of upward of 100,000 pounds depending upon the ambi­ent temperature.35

This was a far cry from the huge boosters for the Titans and Space Shuttle, but it became the first stage of the successful Scout pro­gram. Thiokol’s second stage presented some problems. The firm was finding it difficult to adapt a new propellant for what came to be called the Castor I (or TX-33-35) second-stage motor. Initial static firings had been successful, but then Thiokol encountered unspecified difficulties that had to be overcome. Although the grain design was the same as for the Sergeant, the Castor used a poly­butadiene acrylic acid-aluminum-ammonium perchlorate propel-

lant (also employed as an interim propellant on Minuteman I) and was 20.5 feet long to Sergeant’s 16.3 feet, with an identical 31-inch diameter. Once Thiokol engineers overcame developmental prob­lems with the propellant, the Castor yielded a specific impulse of almost 275 lbf-sec/lbm to only 186 for Sergeant (although the two figures were not comparable because for Scout, the Castor was used at altitude with a larger expansion angle than for Sergeant, which launched on the ground). According to figures supplied by Thiokol, 278 Castor’s average thrust was 64,340 pounds compared to Sergeant’s Chapter 7 41,200.36

The initial four-stage Scout with all stages live flew on July 1, 1960, in the first of nine developmental flights labeled ST-1 through ST-9 (for Scout Tests 1-9), all launched from Wallops. NASA treated them as operational missions, having them carry Explorer space­craft, ionospheric probes, and one reentry payload. The December 4, 1960, launch of ST-3 was the first attempted orbital mission with the Scout. The Algol IA first stage performed properly, but the Cas­tor IA second stage did not ignite because of human failure to detect a defect in the ignition system. ST-4 on February 16, 1961, thus became the first entirely solid-propellant launch vehicle to achieve orbit.37

From these (by later standards) somewhat primitive beginnings, the Scout went on to incorporate more sophisticated technologies. For instance, the Antares IIA, designed by Allegany Ballistics Labo­ratory and produced by Hercules at its Bacchus Works in Magna, Utah, marked a considerable improvement over the first stage – three motor for the Scout, also a Hercules product. The Antares IIA featured a composite, modified, double-base propellant includ­ing ammonium perchlorate, HMX, nitrocellulose, nitroglycerin, and aluminum. Even with a smaller nozzle expansion ratio, this yielded an increase in average thrust from about 13,000 pounds for the older motor to about 21,000 for X259 Antares IIA. This stage first launched on March 29, 1962, in an early (A1) version. The A2 completed its development in June of that year, almost simultane­ously with Hercules’ second stage for Polaris A2, which also used a composite-modified double-base propellant, although one without HMX and with a lower level of performance (probably reflecting the fact that the Polaris motor had an earlier date of development completion by about six months). For Polaris, HMX usage awaited the A3 version, with the Antares IIA actually using it before the missile, reversing the usual practice for launch vehicles to borrow technology from missiles.38

About this time, some earlier Scout technologies found use in other programs. For example, the thrust-augmented Thor (TAT), which entered the launch-vehicle inventory in 1963, incorporated three Thiokol TX-33-52 (Castor I) solid-propellant rocket boosters to supplement the power of the liquid-propellant first stage. The TAT consisted of a Thor with about 170,000 pounds of thrust and three Castor I solid-propellant rocket boosters, which increased liftoff thrust to 331,550 pounds.39

Подпись:The TAT soon gave way to a further improved vehicle with a re­placement for the Castor I. The air force’s Space Systems Division had announced contracts for a long-tank Thor (called Thorad) to replace the thrust-augmented Thor in January 1966. Douglas would provide the new Thor, with thrust augmentation continuing to be provided by Thiokol; only for the Thorad, the three solid motors would be Castor IIs. The Thorad was more than 70 feet long, as compared with 56 feet for the TAT. The added length came mainly in the form of the extended tanks that increased the burning time of the first stage. For the Castor II (TX-354-5), basically developed (as the TX-354-3) in 1964 for the Scout second stage, among other appli­cations, Thiokol kept the steel case used on Castor I but substituted carboxy-terminated polybutadiene for the polybutadiene-acrylic acid used as the binder for the earlier version, keeping aluminum and ammonium perchlorate as fuel and oxidizer. This increased the specific impulse from under 225 for the Castor I to more than 235 lbf-sec/lbm for Castor II and the total impulse from 1.63 million to 1.95 million pounds, improving the payload capacity for the Thorad by 20 percent over the TAT.40

A further major advance in propulsion technology for the Scout came in 1977-79 when, under contract to the Vought Corporation, Thiokol produced a new third-stage motor at its Elkton Division in Maryland. This was the Antares IIIA (TE-M-762, Star 31), employ­ing a HTPB-based binder combined with ammonium perchlorate and aluminum. This propellant increased the specific impulse from about 285 for the composite-modified double-base propellant used by Hercules in the Antares IIB to more than 295 lbf-sec/lbm. In ad­dition to the higher-performance propellant, Thiokol used a com­posite case made of Kevlar 49 and epoxy. Introduced commercially in 1972, Kevlar 49 was DuPont’s registered trademark for an aramid (essentially nylon) fiber that combined light weight, high strength, and toughness. Lighter than fiberglass, it yielded a mass fraction of

FIG. 7.6

Scout S-131R on August 10, 1965, with a new Castor II second stage. (Photo courtesy of NASA)



® 1965-L-06I38



0.923 compared with Antares IIB’s already high 0.916. Burning much longer than the propellant in the Antares IIB, the one in the Antares IIIA produced a lower average thrust, but its total impulse at the high altitudes in which it operated was 840,000 pounds compared with 731,000 pounds for the Antares IIB. No doubt because of the higher erosive propensities of the Antares IIIA motor, which had a higher chamber pressure than the Antares IIB, the newer motor used 4-D carbon-carbon (pyrolitic graphite) for the nozzle-throat insert.41

Launched initially at the end of October 1979, the Scout version G1 with the Antares IIIA as its third stage appears to have been the first launch vehicle to use an HTPB propellant, but the first use of the substance may have been on an improved Maverick tactical (air-
to-surface) missile. Thiokol provided the Maverick’s initial motor, with development starting in 1968, but under contract to the Air Force Rocket Propulsion Laboratory at Edwards AFB, Aerojet had begun in August 1975 to develop the improved motor with HTPB propellant. By October 1976, Aerojet had produced 12 demonstra­tion verification motors. Aerojet did get production orders for some version of an improved Maverick motor. But interestingly, an orga­nization called the ATK Tactical Systems Company, producer of the Maverick heavy warheads, later claimed to be under contract to pro­vide a rocket motor very similar to the Aerojet design with an HTPB propellant, an HTPB liner, an aluminum case, an 11-inch diameter, and a glass phenolic exit cone, all features of Aerojet’s motor.42

Подпись:Another motor that used HTPB was Thiokol’s Star 48, used on the Payload Assist Module (PAM)—a third stage on the Delta and an upper-stage motor used from the Space Shuttle. Thiokol began developing the motor in 1976. The firm made the Star 48 motor case with titanium and used the recently developed advanced com­posite, carbon-carbon, for the nozzle’s exit cone. The PAM was an offshoot of Minuteman, stage three, which Thiokol began produc­ing in 1970 essentially using the original Aerojet design. The pur­pose of the PAM on the shuttle was to propel satellites from a low parking orbit (about 160 miles above Earth) to a higher final orbit, including a geosynchronous transfer orbit. It used the same basic HTPB-aluminum-ammonium perchlorate propellant as Thiokol’s Antares IIIA rocket motor.43

The Inertial Upper Stage (IUS) featured a much more problematic set of motors using HTPB propellant. Designed under management of the Space and Missile Systems Organization (SAMSO—a recom­bination of the air force’s Ballistic and Space Systems Divisions) primarily for use with the Space Shuttle (for when orbiter payloads needed to be placed in geosynchronous orbit), IUS became a difficult stage to develop for a variety of complicated reasons. Many of them were technical, but the major ones involved management. Some, but far from all, of the management problems resulted from the fact that the IUS, which initially stood for Interim Upper Stage, was conceived as a temporary expedient until a more capable Space Tug could fly with the shuttle. When the Space Tug was not terminated but “just slid out year-by-year under budget pressure," as one air force general expressed it, the IUS shifted from being a minimal modification of an existing upper stage such as Transtage or Agena to become, starting about 1978, a projected “first line vehicle in the Space Transportation System." Yet “considerable cost reduction pressure [remained ] as an outgrowth of the interim stage thinking."

Moreover, the air force was developing the vehicle under “a con­tract structure which strongly incentivized performance, but only provided limited cost incentives."44

The IUS ultimately overcame its birth pangs to become “an inte­gral part of America’s access to space for both military and civilian sectors." It had its beginnings in 1969 when presidential direction gave impetus to studies leading to the Space Shuttle. Since the shut­tle would be incapable of reaching geosynchronous and other high 282 orbits, the IUS ultimately became the solution. The DoD agreed to Chapter 7 develop it, proposing in 1975 that it use solid propellants to hold down costs.

In August 1976, the air force selected Boeing Aerospace Com­pany as the prime IUS contractor. The contract provided incentives for meeting performance and cost targets, but Boeing was liable for only 10 percent of cost overruns. Moreover, cost projections for IUS had been based on assumptions, according to Maj. Gen. William Yost of the air force, that “the M-X and Trident missile programs would develop most of the solid rocket motor technology. . . needed by the IUS. Unfortunately, the schedules for those programs slipped far enough that the IUS program became the leader in developing the solid rocket motor technology necessary to meet our perfor­mance requirements." This led the contractor to increase “spend­ing to insure that he will achieve his performance goals and earn the performance fees." As a condition of revising the contract with Boeing in 1979, the air force insisted that the firm’s apparent “man­agement deficiencies be resolved," and Boeing appointed a new pro­gram manager, assigned senior managers to oversee major subcon­tractors, and instituted formal review to correct the problems.45

Boeing had begun a planned 18-month preliminary design phase in August 1976 when it won the contract, to be followed by a 28- month development phase. This would have made the IUS avail­able by June 1980. Soon after winning the basic contract, Boeing subcontracted with CSD to design and test the solid motors to be used in the IUS. CSD chose to use a hydroxy-terminated polybu­tadiene propellant, as had Thiokol in the Antares IIIA motor for Scout. CSD selected a carbon-carbon material for the nozzle, which would be manufactured using a new process that held costs to a low level. It was making the case out of Kevlar. Thiokol was also using the same or similar materials on the contemporary Antares motor, raising questions about the extent to which CSD was taking the lead “in developing the solid rocket motor technology" needed for the IUS (as Yost claimed), but it appears that CSD and Thiokol

were, in effect, competing for that lead from 1977 to 1979, with Thiokol winning the contest.46

Подпись:At first things seemed to be going well with motor development. CSD conducted a series of tests in 1977 to prove the adequacy of the nozzle and motor. It subjected the nozzles to successful 85-second tests at the Air Force Rocket Propulsion Laboratory on June 10 and July 15. A follow-on 145-second test of the nozzle at the laboratory on October 7 was again successful. Moving to the Arnold Engineer­ing Development Center (AEDC), CSD subjected a full-scale mo­tor with 21,000 pounds of propellant to a 154-second test, which it completed successfully. On May 26, 1978, a further test of the nozzle material using the carbon-carbon made with the new, cost­saving technique again occurred without problems in a 140-second test firing at the Rocket Propulsion Laboratory.

But on October 19, 1978, a test of the Kevlar case at AEDC re­sulted in its bursting at only 750 pounds per square inch of water pressure instead of CSD’s prediction of 1,050 pounds. The firm de­cided that defective manufacturing equipment caused the failure. After redesigning the equipment and strengthening the structure of the case, AEDC conducted six more tests between January and September 1979. Five of them were successful, with the cases with­standing higher pressures than specified. By this time, the design of the IUS had evolved into two stages with similar larger and smaller motors. A test firing of the large motor was scheduled on Octo­ber 17, 1978. Inspection of the motor revealed some propellant that was improperly cured, resulting in softness and blistering, delaying the test. With the propellant recast, the test occurred on March 16, 1979, with a 145-second firing that generated more than 50,000 pounds of thrust. Engineers vectored the nozzle several times, dem­onstrating its ability to direct the thrust for course corrections. A follow-on test of the small motor occurred on June 25, again with the nozzle moving. Both tests were successful.47

Most other tests in 1979 went well in most respects, but cracks had appeared in the nozzle of the small motor. Moreover, a special feature of the nozzle for the smaller motor was an extendable exit cone, which was added to the design in 1978. This was a series of conical pieces that in the final design (as of 1983) telescoped out (ex­tended) and fit over one another to increase the nozzle expansion ratio from 49.3:1 without the extension to 181.1:1 with the pieces extended. Although the design, which would be used only on some missions, increased the weight of the motor, it added about 15 lbf- sec/lbm to the specific impulse. Unfortunately, about half the exit

cones for the small motors were defective. Finally, five motors proved to have more “bad" propellant. Boeing and CSD said they could still be tested, but Aerospace Corporation, advising SAMSO, disagreed.

On October 1, 1979, SAMSO formed a tiger team of experts from several organizations (including NASA, the Rocket Propulsion Lab­oratory, and Aerospace Corporation) to investigate technical con­cerns and management. This resulted in the management changes at Boeing already mentioned and a change in one supplier. CSD had 284 been making the large Kevlar motor cases, and Brunswick Corpora – Chapter 7 tion made the small one, which the team found to be superior. As a result, Brunswick became the supplier of both sets of cases.

During 1980, the production team solved the other problems. For example, the cracks in the nozzle of the smaller motor proved to result from unequal expansion of two materials. A silica-phenolic insulation material expanded faster than the carbon-carbon next to it. The solution was to wrap the silica phenolic with graphite to limit expansion. The problem with the exit cones resulted from the methods of the supplier, Kaiser, still learning about the prop­erties of carbon-carbon. A change in tooling and ply patterns plus improved quality-control procedures provided the solution. The degraded propellant had all come from a single batch and was us­able in tests. As a result, three rocket motor tests of each motor (small and large) during 1980 at the AEDC were successful. (All of these development tests at AEDC simulated conditions the motors would actually face in flight at altitude.) There were further prob­lems with propellant cracks, delamination of the carbon material in the extendable exit cones, and the mechanism for extending the exit cones, but engineers solved them, too.48

The various technical and managerial problems had led to more than two years of delay and to cost overruns that basically dou­bled the originally projected cost of the IUS. Although many of the problems resulted from contractual arrangements and the initial, interim character of the upper stage, many of them involved fabri­cation methods and quality control. They showed that despite more than two and a half decades of continuous rocket development, rocket engineering in the United States still required constant at­tention to small details and, where new technology was involved, a certain amount of trial and error, although Thiokol’s success with Antares IIIA showed that sometimes the process of innovation could go more smoothly. (But not always, as Thiokol’s later prob­lems with the shuttle solid-rocket boosters showed.) Because the IUS was designed principally for use on the Space Shuttle, NASA’s

delays with that program made the stretch-out of the IUS schedule less problematic than it could have been.49

Подпись:On October 30, 1982, the first Titan 34D and the first IUS to­gether successfully launched a Defense Satellite Communications Satellite II and the first DSCS III into geosynchronous orbit from Cape Canaveral. As planned, the second-stage burn achieved low – Earth orbit, with the first IUS motor carrying the third stage and the satellites into transfer orbit. The second IUS motor placed the payloads in geostationary orbit, with hydrazine thrusters making final adjustments in the placement of each satellite. During launch the telemetry failed, attributed to a leak in the seal of a switch. But the guidance/control system, flying “blind" (without telemetry) or external control from Earth autonomously carried out the provi­sions of the flight plan, as designed.50

As completely designed, the IUS was roughly 17 feet long and had a maximum diameter of 9.25 feet. Fully loaded, the large, first – stage motor (SRM-1) carried 21,400 pounds of ammonium perchlo – rate-HTPB-aluminum propellant, but the propellant load could be reduced as required for specific missions (as was done with the first launch). The smaller, second-stage motor (SRM-2) could carry up to 6,000 pounds of the same propellant. The propellant-delivered specific impulse of SRM-1 was upward of 295, that for SRM-2 about 290, increased to more than 300 lbf-sec/lbm with the extendable exit cone.51

To mention just one other use of an HTPB propellant, this tech­nology came to the Delta with the Castor IVA strap-ons. A Cas­tor IV (TX-526) had actually replaced the Castor II strap-ons in De­cember 1975 for the Delta model 3914, but it was a reversion from the carboxy-terminated polybutadiene used in the older strap-on to polybutadiene-acrylic acid (PBAA) as the binder. The reason for the shift may have been cost, since the Castor IV at 29.8 feet long and 40 inches in diameter contained much more propellant than the 19.8-foot by 31-inch Castor II, and CTPB was more expensive than PBAA. But in the early 1980s, Goddard Space Flight Center (man­ager of the Delta program) shifted to an uprating with the Castor IVA. Tested and qualified in 1983, the new motors were not intro­duced then because of the impending phaseout of the Delta in favor of the Space Shuttle. With the post-Challenger resurrection of ex­pendable launch vehicles, McDonnell Douglas proposed incorporat­ing the Castor IVAs on Delta II as a low-risk improvement. The new strap-ons kept the steel case and graphite nozzle throat material. But they used the HTPB-aluminum binder with a higher loading of

FIG. 7.7

The Inertial Upper Stage attached to the Magellan spacecraft in the payload bay of Space Shuttle

Atlantis. (Photo courtesy of NASA)



solids. This increased the average thrust for the same-sized motor from 85,105 to 98,187 pounds.52

Analysis and Conclusions

Although the solid-propellant breakthrough achieved by the Polaris and Minuteman programs provided many technologies to launch ve­hicles, others followed. These included the carbon-phenolic throat, segmenting, and the tang-and-clevis joints for the Titan SRM; Flex-
seal nozzles used on the Space Shuttle’s huge solid-rocket boosters; and the use of HMX in the propellant for the Antares IIA stage of the Scout launch vehicle. Although it apparently found its first use on an improved motor for the Maverick tactical missile, HTPB propellant seems to have first appeared on a launch vehicle in the Scout G1.

Подпись:Although sometimes innovations occurred without many appar­ent problems, as in Thiokol’s use of HTPB in the Scout’s Antares IIIA, the IUS, employing many similar technologies, faced a whole host of difficulties, many of them technical. The field joints for the shuttle caused the Challenger tragedy, and when Hercules devel­oped the solid-rocket motor upgrade for the Titan IVB, technical problems delayed launch of the first uprated launch vehicle until well beyond the period covered by this history. Rocket engineers continued to advance the state of their art, but often they could do so only by trial and error. There was no such thing as a mature rocket science that could guide them effortlessly through the design of new technologies, but accumulated experience, data, computers, instrumentation, and telemetry allowed practitioners to solve most problems.

Подпись: Conclusions and EpilogueDURING THE LAST 45 YEARS, LAUNCH VEHI­cles have propelled countless spacecraft and sat­ellites into space, in the process revolutionizing life on planet Earth. Americans have become de­pendent upon satellites for everything from what they watch on television to how they wage war. Space telescopes and other spacecraft have greatly expanded our knowledge of the universe. What en­abled the United States to develop the technology for access to space so quickly? One major contribu­tor was the cold war, whose terminus is the end point for this book. Had it not been for the Soviet threat, symbolized by Sputnik, the enormous ex­penditures needed to develop U. S. missiles, launch vehicles, and satellites would have been lacking.1

There seems to be no accurate compilation of total expenditures on missiles and launch vehicles during the cold war. Obviously, however, the out­lays were enormous and constituted a virtual sine qua non for the speedy development of the technol­ogy. An early (1965) estimate of the total costs for ballistic missiles to that point in time suggested a figure of $17 billion, equal to some $106 billion in 2005 dollars. Including missile sites, which were irrelevant to launch vehicles, this figure also cov­ered factories for producing propellants, engines, airframes, and guidance/control systems; test fa­cilities; ranges with their testing, tracking, and control equipment; laboratories; and much else.2 A further indicator of the huge costs of missiles and launch vehicles was the $9.3 billion (nearly $55 billion in 2005 dollars) spent on the Saturn launch – vehicle family.3

Fears of Soviet missile attacks and the spend­ing they stimulated were one factor in the develop­ment of launch-vehicle technology. They also pro­vided the context for a second major contributor, the work of heterogeneous engineers in stimulating Congress, several presidential administrations, and the American people to invest the money needed

Подпись: FIG. 8.1SCOUTMonkey Baker posing with a model of a Jupiter vehicle, one of which launched it into space in an early example of the use of a missile as a launch vehicle, part of the space race inaugurated by the Soviets’ launch of Sputnik. (Photo courtesy of NASA)

for rapid development. Without individuals like Trevor Gardner, John von Neumann, Bernard Schriever, Theodore von Karman, Wernher von Braun, and William F. “Red" Raborn, funding for mis­siles and rockets (with their frequent failures in the early years) would not have been forthcoming.

As missiles and launch vehicles increased in size and complexity, it is not surprising that many of them experienced failure. Ameri­cans recognized the arcane nature of the technology by their use of the term “rocket science" to describe it. Ironically, the rocket “scientists" could not always predict the problems the technology encountered in operation. Methods of testing rockets and missiles, technical reports, computer tools, and other supporting infrastruc­ture continued to grow. But as recently as 2003, when the Columbia disaster occurred, NASA discovered once again that it did not fully understand all aspects of rocket behavior despite extensive experi­ence. Hardly an isolated case, this major accident simply reempha­sized that predictability of rocket behavior had been problematic from Robert Goddard’s inability to reach the altitudes he had fore­cast until very near the present. If we define rocket science as a body of knowledge complete and mature enough to allow accurate predic­tions of problems, then clearly, such a science does not exist. Maybe it will someday, but what we currently have is rocket engineering.

Of course, there are other ways of defining science. And recent science has hardly been immune from uncertainties, such as those

regarding the big bang theory about how the universe arose. More­over, the success of rocket developers in resolving unanticipated problems and getting their creations to work certainly compares 290 favorably with scientists’ accommodations of unexpected data by Chapter 8 adjustment of theories. The difference lies in science’s basic quest to understand the universe as compared with rocket engineers’ ef­fort to make their vehicles meet design goals. These engineers used any available resources to reach that end, including science. Cer­tainly the engineers, especially those engaged in developing engi­neering theory (often called engineering science), wanted to under­stand how rockets worked. But often they had to “fix" problems in the absence of such fundamental understanding. In such cases, they had to resort to trial and error, finding a solution that worked without necessarily understanding why it worked. Accumulated knowledge, engineering theory, and intuition helped in correcting problems, but the solution did not always work when a particular technology had to be scaled up to a larger size. Sometimes, in fact, innovative solutions ran counter to existing theory.

This reality shows that the basic process of developing rockets constituted engineering, not science. Such an argument tallies with the general theses of Edwin Layton, Walter Vincenti, and Eugene S. Ferguson about engineering in general as different from science— especially their points about engineers’ focus on doing as opposed to scientists’ knowing, on the importance of design for engineers, and on the role of art in that design. Vincenti, especially, proposed more historical analysis of the ways engineers sometimes must make de­cisions in the absence of complete or certain knowledge.4

Research for this book did not start with the thought of apply­ing Vincenti, Layton, and Ferguson’s arguments to rocket technol­ogy. Instead, as I gathered information about the process of missile and launch-vehicle development, I became increasingly convinced that it fit the mold of engineering, not science. This is particularly true in the areas of injection, ignition, and propulsion of liquid pro­pellants and of combustion instability in solids. Problems in these areas occurred in the design of the V-2, the H-1, F-1, and space shut­tle main engines as well as many solid-propellant motors. A. O. Tischler, NASA assistant director for propulsion, in 1962 called in­jector design “more a black art than a science."5 With the passage of time, the art became less “black," but art it remained.

Problems with rocket design were not exclusive to propulsion. As early designs had to be scaled up or modified with new and bet­ter materials to improve performance, unanticipated problems con­tinued to occur through the end of the period of this book and be-

yond. Failure to understand the behavior of foam covering the Space Shuttle’s external tank as it rose through the atmosphere continued beyond the Columbia disaster. Problems in developing the solid – rocket motor upgrade for the Titan IVB persisted beyond the end of the cold war. These and other instances demonstrate the continu­ing uncertainties accompanying rocket engineering, especially in an environment where speed and cost control limited basic research.

Подпись: 291 Conclusions and Epilogue In the face of this unpredictability, it is noteworthy that missile and launch-vehicle technology evolved as quickly as it did. Design and development engineers did exceptionally well to find innova­tive solutions to problems and allow the technology to advance as successfully as it did.

Another key to the speedy development of rocket technology was the process of innovation. Sadly, known sources often shine little light on the individuals or processes involved. Interviews and corre­spondence with rocket engineers sometimes yield information.6 But even the principals in a particular development frequently cannot remember who came up with a discovery or how it came about. En­gineers typically worked in large teams to design rocket systems or components. And many innovations involved more than one firm. Otto Glasser at the Western Development Division offered an inter­esting analogy for the difficulty of finding out who contributed sig­nificantly to innovation under such circumstances: “If you were to back into a buzz saw could you tell me which tooth it is that cut you?"7

Many innovations did not arise from initial design but occurred in response to problems during testing. Examples of these that seem to fit Glasser’s “which tooth?" analogy include the process of roll­ring forging developed by UTC, Westinghouse, and the Ladish Com­pany and UTC’s tape-wrapped, carbon-phenolic nozzle throat for the Titan solid-rocket motors. The companies doing the innovating are clear, but we do not know which individuals were the principal innovators. How Aerojet engineers fixed problems with the Trans – tage (ranging from a weakness in a nozzle extension to malfunction­ing bipropellant valves) likewise remains somewhat mysterious. Regardless, all of these technologies appear to exemplify trial-and – error engineering.

What we do know about innovations in rocket development sug­gests that they did not follow a single pattern. Hugh Aitken pro­vided a felicitous description in his book about radio technology, saying that it involved “a process extending over time in which information from several sources came to be combined in new ways."8 In the case of missiles and launch vehicles, large numbers of firms, institutions, and organizations helped provide the requi-

site information. Among those that contributed were firms such as Aerojet, Rocketdyne, Pratt & Whitney, Douglas, the Martin Com­pany, UTC/CSD, and Thiokol; and other organizations like the air 292 force’s Western Development Division and its successors, the ar – Chapter 8 my’s counterparts at Redstone Arsenal, the navy’s special projects office and its successors, NASA Headquarters and various centers (notably JPL, Langley, Lewis, and Marshall for rocketry), the Naval Ordnance Test Station and its successors, the Air Force Rocket Pro­pulsion Laboratory (under various names), the Arnold Engineering Development Center, the Chemical Propulsion Information Agency (CPIA) and its predecessors, and the Armour Institute (later Illinois Institute of Technology).

During developmental planning, representatives from entities like these met to exchange ideas and information. Then, if prob­lems arose for which there were no known explanations and/or no evident solutions, as often happened, engineers and other experts from perhaps different organizations met to brainstorm and trou­bleshoot. With the enduring problem of combustion instability, for instance, numerous university researchers, as well as other engi­neers, have long been seeking both understanding and solutions.9 Surviving sources indicate that the general process was often com­plex, with no record of specific contributors except the authors of technical reports. But the authors themselves often wrote in the passive voice, masking individual participants beyond the authors themselves, who presumably were involved. Engineers sometimes remembered (but how accurately?) some details of solutions but not always the precise process.

Among the factors that conditioned rapid development of mis­sile technology, the existing literature points to interservice com­petition, often as a problem but also as a spur to innovation.10 Vir­tually unnoticed in the literature but probably more important was interservice and interagency cooperation. The CPIA was a key pro­moter of cooperative exchange of information, but not the only one. For instance, the air force saw the navy’s Polaris program as a com­petitor for roles, missions, and funding. The Polaris competition encouraged the air force’s development of its own solid-propellant missile, Minuteman. Yet ironically, Polaris itself might not have been possible without technologies the air force developed. Min – uteman, in turn, borrowed the use of aluminum as a fuel from the navy. Likewise, the air force reluctantly accepted NASA as a devel­oper of rocket technology, and the army was not happy to lose the von Braun team and JPL to the civilian space agency. Both services, however, cooperated with NASA (and vice versa), with the air force

FIG. 8.2

A static test of a Space Shuttle solid-rocket booster at the Morton Thiokol test site in Wasatch, Utah, on January 20, 1989. (Photo courtesy of NASA)



loaning important managers like Samuel Phillips to help NASA with its programs. Moreover, many astronauts came to NASA from the U. S. Marine Corps, the navy, and the air force.11

Technology transfer also contributed to rapid rocket develop­ment, but its details remained almost as elusive as those involving innovation. Federal contracting agencies often precluded contrac­tors from treating innovations developed under government con­tract as trade secrets or company property. Lockheed, for example, was unable to protect its Lockseal technology, permitting Thiokol to increase its size and use it to vector the exhaust from the shuttle solid-rocket boosters under the name Flexseal.

Engineers frequently learned about the rocket technology of one firm or organization, established their credentials, and moved to an-

other organization, carrying their knowledge with them. This helped transfer technology and promote overall rocket development. For instance, after Charles Bartley developed early rubberized, compos – 294 ite solid propellants at JPL, he founded the Grand Central Rocket Chapter 8 Company, which participated in solid-rocket development and then became part of Lockheed Propulsion Company in 1960-61.12 Bart­ley also transferred JPL technologies to Thiokol when it entered the rocket business. Barnet R. Adelman had worked with both liquid and solid propellants at JPL. He then became technical director of the Rocket Fuel Division at Phillips Petroleum and then director of vehicle engineering for the Ramo-Wooldridge Corporation. At Ramo-Wooldridge, he became a major supporter of the Minuteman missile, along with Col. Edward N. Hall. Next, he helped found United Research Corporation (later UTC/CSD). The knowledge he carried with him undoubtedly helped UTC develop the solid-rocket motors for the Titan III and IVA.13 Adelman and other UTC execu­tives’ knowledge of people in the solid-propellant field also enabled them to hire other experienced engineers who furthered the success of the firm.

The previous chapters have covered many other examples of technology transfer through people moving from one organization to another. A well-known example was the von Braun team, in­cluding Krafft Ehricke, who brought German V-2 technology to the United States. Importantly, though, the technology they transferred was only part of the story. Americans, notably Rocketdyne engi­neers, learned much from the V-2 and its German developers, but they went on to create major innovations of their own in develop­ing successor engines, extending from the Redstone to the space shuttle main engine. This synergism contributed in complex ways to American rocket development.

Management systems constituted another factor in the rapid de­velopment of missile and launch-vehicle technology. Without such systems (and the systems engineering they fostered), the various components of rockets would not have worked together with in­creasing reliability to launch their payloads as time passed. Such systems (and the individual management skills that complemented and implemented them) were especially necessary as the numbers of people from industry, government, and universities increased and became interdependent. Growth was rapid. The Atlas missile had only 56 major contractors in 1955. By decade’s end, the number had escalated to roughly 2,000. To keep this many organizations on schedule and to ensure quality control, Gen. Bernard Schriever availed himself of a systems engineering-technical direction con-

tractor (Ramo-Wooldridge). Beyond that problematic arrangement, the Western Development Division instituted a management-con­trol system to keep track of schedules and to deal with problems as they arose. Graphs, charts, and computer tracking permitted Schriever and his successors to keep their projects more or less on schedule.

Подпись: 295 Conclusions and Epilogue The enormously complex Polaris system, likewise, led Adm. “Red" Raborn of the navy to oversee development of the Program Evaluation and Review Technique, analogous to Schriever’s system. Critics complained about both systems, but without some such ar­rangements, early missile development could hardly have been as successful and rapid as it was.

George Mueller and Sam Phillips brought these kinds of sys­tems to NASA, enabling it to land astronauts on the Moon in less than a decade from President Kennedy’s 1961 exhortation to do so. These systems allowed Apollo to stay on schedule and within bud­get while still achieving configuration control. The overall success of missile and launch-vehicle development owed much to such arrangements.

The basic processes of rocket engineering did not change abruptly with the end of the cold war about 1991. But the context in which research and development had to occur suffered a drastic shift. It became less urgent for new technologies to appear, while Congress exercised stringent oversight over costs and schedules. Basing itself on studies in the late 1980s and early 1990s, the air force responded to the new environment with the Evolved Expendable Launch Ve­hicle (EELV) program. This replaced Titan II, Delta II, Atlas II, and Titan IV with a new series of boosters capable of launching 2,500 to 45,000 pounds into low-Earth orbit with a 98 percent reliability rate. This exceeded capabilities of previous launch vehicles at costs 25 to 50 percent below earlier figures.

The air force provided four $30 million contracts in August 1995 to Lockheed Martin, Alliant Techsystems (which acquired Hercu­les Aerospace Company, the former Hercules Powder Company, in March 1995), Boeing, and McDonnell Douglas to develop proposals for the EELV vehicles. Out of this process, McDonnell Douglas (later acquired by Boeing) and Lockheed Martin won contracts for actual development.14 The further development of this program is beyond the scope of this book, but the Lockheed Martin Atlas V’s success­ful launch on August 21, 2002, and that of Boeing’s Delta IV on No­vember 20, 2002, suggested that the EELV launchers would capture most of the military market. The use of a Russian RD-180 engine on the Atlas V symbolized the radical change that had occurred

since the end of the cold war. The air force reportedly believed as of August 2002 “that through 2020 the two new EELV families w[ould] reduce the cost per pound to orbit to $7,000 compared with $20,000 296 for the old booster fleet," saving “$10 billion in launch costs"—a Chapter 8 50 percent reduction “compared with launching the same military

payloads on old Delta, Atlas and Titan boosters."15

Similar concerns about cost affected NASA’s efforts to develop new launch vehicles.16 NASA also had to contend with concerns about safety after the Columbia disaster. In 2005-2006, the agency developed a concept for a safer pair of launch vehicles that would use technologies from the Space Shuttle or other existing hardware without the problems of an orbiter with wings that could be struck by debris from an external tank. A crew launch vehicle (named Ares I) would consist of one enlarged solid-rocket booster derived from the shuttle, a stage powered by a Rocketdyne J-2X engine (evolved from the J-2 used on the Saturn V), and a crew capsule at the top of the stack, equipped with an escape rocket to separate the crew from the launch vehicle in the event of problems. In this Apollo-like cap­sule, the crew would face little danger from debris separating from the shuttle on launch, as happened with tragic consequences dur­ing the Columbia launch in 2003 and again (without significant damage) during the launch of Space Shuttle Discovery in 2005 and subsequent shuttle launches in 2006.

For future space exploration, NASA also planned a heavy-lift launch vehicle named Ares V with two solid-rocket boosters, a cen­tral element derived from the shuttle’s external tank, five RS-68 engines modified from the Delta IV, and an earth-departure stage propelled by a J-2X engine. Like Ares I, this vehicle would be config­ured in stages, but the first-stage solid-rocket boosters would flank the central booster tank, with the liquid engines above the tank. Ares V thus resembled a Titan IV in its general configuration. NASA calculated that these two future vehicles would be 10 times as safe as the shuttle and would not cost as much as a completely new vehicle because of the use of proven technologies.17 However, both vehicles were in a state of development and could easily change as further possibilities came under study.

It would be foolhardy to predict how the struggle for cheaper access to space will play out in the new environment ushered in by the fall of the Soviet Union. However, maybe this history of the uncertainties and difficulties of developing launch vehicles in a different environment will highlight the kinds of problems rocket engineers can expect to encounter in a more cost-constrained atmo­sphere. Rocket reliability improved significantly in the 50 years or

Подпись: 297 Conclusions and Epilogue so following the serious beginnings of U. S. missile and rocket de­velopment. Recently, launch vehicles have experienced only a 2-5 percent failure rate. By comparison, the first 227 Atlas launches failed 30 percent of the time. Nevertheless, in 2003 the Colum­bia Accident Investigation Board stated, “Building and launching rockets is still a very dangerous business and will continue to be so for the foreseeable future while we gain experience at it. It is unlikely that launching a space vehicle will ever be as routine an undertaking as commercial air travel."18 Knowledge of this reality, coupled with the historical experiences recounted previously, may help contemporary rocket engineers design both cheaper and bet­ter launch vehicles. Perhaps Congress and the American people can also benefit from knowing the kinds of challenges rocket engineers are likely to face.