SOME confusion and a number of conferences occurred at the Naval Research Laboratory during the fortnight before and the three weeks after Deputy Secretary of Defense Reuben Robertson issued his September 1955 memorandum outlining the military departments' respective obligations under the joint satellite program.1 With technical responsibility assigned to the Navy, it was a foregone conclusion at the Pentagon that the Laboratory, under the administrative aegis of the Office of Naval Research, would direct the project. As it happened, the Secretary of the Navy waited until 27 September formally to designate ONR as administrator and not until 6 October did the Chief of Naval Research officially notify the Laboratory that it was to take charge.2 ONR, a Navy officer once remarked, served the Laboratory chiefly as a post office-transmitting funds, inquiries, and, occasionally, directives from higher authority. But an organization chart drawn up some- what later shows how many echelons of authority stood above NRL. Although a long and distinguished record in research and engineering had won a measure of independence for the Laboratory, its chiefs were still bound by orders from ONR and higher ranking officialdom. Small wonder that uncertainty reigned during late August and September 1955.
As time was precious, the military and civilian directors of the Laboratory felt obliged to make a number of tentative decisions well before they received notice of their authority to proceed. They had to decide how large a proportion of their staff to assign to the satellite program, who was to head it, how much freedom of action he and his principal assistants should have, in consultation with them how to refine the budget estimates and initiate procedures to ensure a prompt and orderly flow of money, how to handle publicity in keeping with whatever security classification the Navy imposed upon the project, and, most urgent of all, what ground rules to lay down in negotiating a contract with an industrial company for design and production of the launch vehicle.
The Laboratory's key men realized that the primary contract would have to deviate in some respects from the routine kind whereby the company selected through competitive bidding undertook to meet requirements for a prototype by building, test-firing, and making successive changes in a series of models. In this case, time forbade that orderly procedure. NRL, acting for the government on an intricate task involving a number of federal agencies, would have to exercise constant supervision over the producer and be free to rewrite specifications if necessary as the work progressed. Performance of mutually interdependent systems, designed in part by a process of extrapolation and analogy, rather than solely by proven, predetermined specifications, would spell success or failure in this novel experiment. Hence, before making any other decisions, the day after NRL learned of its victory, Captain Samuel Tucker, the Laboratory's director, invited two Martin Company executives to his office to hear the good news and talk over plans. For from the inception of the NRL satellite plan its originators had taken for granted that, were their proposal accepted, they would again be working with the engineers who had built the Viking. At Rosen's suggestion, during early summer the Martin Company had prepared several analyses of aerodynamic problems affecting the project and had made a formal presentation to the Stewart Committee. The committee indeed had indicated that its selection of the NRL scheme was predicated partly on the assumption that Martin experience would expedite design and production of the launch vehicle.
Two additional Vikings were already
on order for continuation of NRL's work on upper-atmosphere and
guided missile research, including a study for the Air Force on
electromagnetic wave propagation in rocket flames; cancellation
of that commitment and diversion of Vikings 13 and 14 to use as
satellite launching test vehicles should get the program off to
a good start. The Martin Company consequently was sure of its
When the two Martin vice presidents appeared in Captain Tucker's office on August 25, their confidence in their impregnable position became evident. Tucker, Wayne Hall, NRL's deputy director of research, and Rosen all felt the chill in the atmosphere when the Martin men disclaimed interest in a contract unless the company had full control. As the job was unlikely to be a big money-maker, prestige was the sole inducement. Where- as the public always spoke of the Navy Viking rocket, this time the label must be the Martin satellite project Equable but firm responses from the NRL men, however, led the manufacturers to agree to a more formal conference on August 31 and to transformation of the existing contract for Vikings 13 and 14 into one releasing the vehicles for the satellite program.3
To expedite matters at the meeting held on the 31st, William G. Purdy, chief spokesman there for the Martin Company, came armed with a draft "change-order" of the Viking contract, immediate acceptance of which, completed by filling in a few blanks, would constitute, he opined, a suitable contract for the satellite project and permit work at the Baltimore plant to begin at once. After stating that the government had designated ONR as the "procuring agency" for a program aimed at creating small scientific earth satellites, "the first of which is intended to be operational within approximately eighteen months," the draft change-order declared that the Martin Company as the prime contractor would "carry responsibility for the development, design, manufacture, and test of the complete system." ONR should:
assign to this program a Project
Officer, who shall be the Government representative authorized
to commit the Government on contractual and technical decisions.
. . The Contractor shall rely on the Project Officer for the proper
coordination of Government furnished services, facilities and
materials and for approvals, where required, of Contractor activities.
Subsequent to acceptance by the Government of the system specification,
changes in the specifications will require joint approval of the
Project Officer and the Contractor.
The contractor should then proceed
immediately "to furnish the necessary services, materials
and facilities" in accordance with the following requirements:
a) The mission of the system shall be to place a satellite of____pounds and____cubic inches in an orbital plane within____degrees of the plane of the equator and having a minimum altitude in any point in the orbit of____miles.
b) The Government shall furnish the complete satellite package, including instrumentation, antennae and all ground equipment associated with the transmission and reception of scientific information. The configuration of the satellite shall be established by mutual agreement between the Contractor and the Project Officer within 60 days of the date of this letter.
c) The vehicle shall be launched from Patrick Air Force Base, with the Government furnishing propellants and such handling and test facilities, together with the necessary personnel to operate the same, as are not included in the Contractor's responsibilities under (e) below.
d) The Contractor shall develop, design, manufacture and test the complete vehicle, including all subsystems, except those defined by (b) and (c) above. Certain components may he sub-contracted at the discretion of the Contractor.
e) Handling and testing equipment
facilities presently available to the Government shall be utilized
Additional handling and testing equipment
required shall be furnished by the Contractor.
The two Viking rockets on order were to be used to advance the development of the satellite vehicle. Within ninety days of signing this contract the Martin Company would submit a firm proposal which would include a system specification, a test plan, and an estimate of the time and cost required for completion of the entire program. Here, from the manufacturer's standpoint, was a straightforward plan of procedure.
The Laboratory and ONR representatives objected: the scheme relegated NRL scientists to the sidelines and demanded of them a delegation of responsibility which, in light of the Defense Department's mandate to the Navy, was out of the question. A single project officer, corresponding to the company's operations manager, could not make final decisions for the government: any significant departure from the proposal approved by the Stewart Committee would have to have both committee and NRL concurrence, lest the end product fall short of the scientific performance required. Captain Tucker pointed out that the contractor would be dealing with NRL, not with ONR, despite the latter's acting as the Laboratory's agent in drawing up the contract. Everyone present acknowledged the importance of definition of the mission, prompt formulation of operational and logistic plans, definition of mutual responsibility, and centralized management; but the Navy representatives could not establish then and there a clear-cut division of responsibility between NRL and the contractor. Neither could they map out explicit operational and logistic plans as yet, since the Army and Air Force were also involved. For the present, Martin should put into writing its recommendations about "system responsibility," vehicle specifications accompanied by delivery schedules, test firing schedules, firing responsibilities and procedures, launching sites, range requirements, the company's requirements for vehicle instrumentation, and a list of items the government should furnish. NRL itself would provide telemetering and tracking equipment and take charge of its operation. While the company was drafting its recommendations, the Laboratory staff would prepare specifications, undertake to integrate them with Martin's, and define the Scientific Officer's duties. A second meeting should take place on 8 September, by which time both parties ought to be able to arrive at a mutually satisfactory modus operandi. When the conference broke up, the only matters on which both sides had reached agreement were the necessity of further meetings, the "Confidential" classification for most components of the vehicle, and NRL's providing the satellite package.4
During the week following this exchange, Laboratory specialists by twos and threes carried on a series of exploratory negotiations. Because a reliable engine for the third stage of the satellite launcher was a pressing need and the Thiokol Chemical Company, under contract with Army Ordnance, was working on a solid-fuel T~5 engine at Redstone Arsenal, on l September James M. Bridger and Alton E. Jones of NRL and two Martin engineers met with a Thiokol Company official and a man from the Arsenal to discuss arrangements whereby Thiokol could use the Arsenal facilities in designing and developing a third-stage power plant for the NRL project. The group concluded that before proceeding further NRL and Martin must agree on specifications and the Thiokol Company must map out a research and development program on a schedule acceptable to the Arsenal. A few days later, while Milton Rosen, Bridger, and General Electric Company engineers talked over preliminary plans for a new GE engine for the first-stage rocket, NRL telemetry experts, headed by Daniel G. Mazur, flew to Patrick Air Force Base in Florida to examine the instrumentation available there, to inquire about safety rules and test schedules, and to explain what facilities the Navy hoped to have the Air Force supply. Mazur brought back some discouraging news: owing to high priority military projects scheduled for tests during the next eighteen months, space for satellite vehicle tests would be hard to arrange; safety rules to forestall danger to ground installations and to Atlantic shipping lanes would require a power-cutoff device in each stage of the rocket; and, as the base lacked the type of telemetry equipment wanted for the satellite and its launcher, NRL would have to either supply and man its own or else modify its original plans.5
Meanwhile in Washington the Navy, accepting the name suggested by Mrs. Milton Rosen, christened the satellite venture Project Vanguard. At the same time consensus at NRL and ONR ran that the Laboratory should establish a separate Vanguard unit which would include the entire Rocket Development Branch and some men from other branches and divisions. Captain Tucker and Edward O. Hulburt, the Laboratory's research director, then selected John P. Hagen to head the project. Although he had had no part in drafting the satellite proposal, he was the obvious choice, Hulburt averred. Superintendent of the Laboratory's Astronomy and Astrophysics Division, Hagen possessed unquestioned stature as a scientist. If his slight build, his bespectacled brown eyes, his mild-mannered professorial mien and utter lack of facade led outsiders at first into thinking him a misfit for a difficult administrative job, his scientific knowledge and his imperturbability in dealing with temperamental prima donnas quickly impressed high-powered industrialists as well as his own staff. In spite of later complaints that he was incapable of "dynamic leadership," he commanded the respect and affection of his colleagues. Indeed one of his team attested that a more forceful chief would probably have cracked under the strains the job imposed; Hagen's patience and tolerance were essential to seeing it through.
Under Hagen the intense, hard-driving
rocket expert Milton Rosen was to be technical director. The youthful.looking
Homer E. Newell, named "Science Programs Coordinator,"
would work closely with the National Academy's IGY panels. Hagen
early asked to have Thomas Jenkins, the Laboratory's Deputy Comptroller,
as his budget officer; Jenkins, a soft spoken young man, thenceforward
wore two hats, one for the Laboratory as a whole, one for Vanguard.
Other appointments came from the Rocket Development Branch, the
progenitor of the Vanguard plan: Leopold Winkler as Rosen's engineering
consultant on problems of mechanical design, James M. Bridger
on rocket propulsion and control, John T. Mengel on satellite
tracking, and Daniel G. Mazur on telemetry and data reduction.6
The sad-eyed, big-eared Winkler, with his air of resignation to
human vagaries, was as even-tempered as the short, slightly rotund,
humorous Mazur was explosive in the face of crisis; tall, lanky,
handsome Bridger tended to the laconic but at times was as emphatic
as Mazur; Mengel, heavily built, sandy-haired and square-jowled,
displayed the quiet self- possession of a man sure of his knowledge,
yet ready to learn from others. All four shared a passionate interest
in the satellite program and each brought a special expertise
Cryptic, unsigned notes on an informal
conference held shortly thereafter among representatives from
Quarles' office, ONR, and the Laboratory revealed the extent of
uncertainties. Asked about what the Laboratory would expect to
receive from the National Science Foundation, Admiral Fourth
said, "NSF is out," whereupon Charles Weaver of Quarles'
staff observed that the Foundation could supply some funds. When
the Admiral suggested that the Navy simply do what it could for
$20 million, Weaver reminded him that that sum, the figure which
the National Security Council had accepted, was not binding and
in any case did not include the cost of satellite instrumentation.
Apparently everyone assumed the availability, free, of a good
deal of expensive equipment already in existence and a number
of services-test and launching facilities, for example, powerful
radar, and telemetry instruments, as well as Army Engineer Corps
construction, maintenance, and operating teams for the tracking
A decade later perusal of the early
discussions of costs might awaken in the reader the suspicion
that a lack of candor prevailed among participants fearful lest
realistic figures cancel the entire program. But the Vanguard
comptroller, the person most familiar with the financial problems,
saw "guesstimates"-all too often too low-as inescapable:
there were no cost data to draw upon applicable to a satellite
program. In 1955 no one had ever built a multistage launcher except
as a vertical sounding rocket; no one had tried out an electronic
tracking system for an artificial earth-circling body; and no
one could predict accurately the expenses of developing reliable
instrumentation capable of functioning for days in space. The
vagueness that shrouded early financial planning was a handicap
to the men at NRL responsible for getting work started. For a
first allotment of money they had to wait until December 1955
and in the interim had to draw on the Laboratory's small Naval
How to map out financial procedures was more difficult than setting up an informal organization fur Vanguard. On 8 September Captain Tucker issued a new budget estimate raising to $28.8 million the $20-million figure submitted to the Stewart Committee in mid-August. Deputy Secretary of Defense Robertson's directive of 9 September reiterated the National Security Council's decree that the satellite venture must not interfere with military programs of higher priority, instructed each service to "provide for the immediate implementation" of action on its share of the joint undertaking, and specified that all information intended for public release must first be submitted to the Office of Security Review for clearance. But the paragraph covering budgetary arrangements merely noted that "the Navy Department will manage the technical program with policy guidance from the Assistant Secretary of Defense (R&D) and will provide the funds required . . . with the understanding that reimbursement will be made as soon as funds can be made available from other sources." The decision as to how much money, when, and from what sources was left in abeyance.
Cryptic, unsigned notes on an informal conference held shortly thereafter among representatives from Quarles' office, ONR, and the Laboratory revealed the extent of uncertainties. Asked about what the Laboratory would expect to receive from the National Science Foundation, Admiral Fourth said, "NSF is out," whereupon Charles Weaver of Quarles' staff observed that the Foundation could supply some funds. When the Admiral suggested that the Navy simply do what it could for $20 million, Weaver reminded him that that sum, the figure which the National Security Council had accepted, was not binding and in any case did not include the cost of satellite instrumentation. Apparently everyone assumed the availability, free, of a good deal of expensive equipment already in existence and a number of services-test and launching facilities, for example, powerful radar, and telemetry instruments, as well as Army Engineer Corps construction, maintenance, and operating teams for the tracking station network.
A decade later perusal of the early discussions of costs might awaken in the reader the suspicion that a lack of candor prevailed among participants fearful lest realistic figures cancel the entire program. But the Vanguard comptroller, the person most familiar with the financial problems, saw "guesstimates"-all too often too low-as inescapable: there were no cost data to draw upon applicable to a satellite program. In 1955 no one had ever built a multistage launcher except as a vertical sounding rocket; no one had tried out an electronic tracking system for an artificial earth-circling body; and no one could predict accurately the expenses of developing reliable instrumentation capable of functioning for days in space. The vagueness that shrouded early financial planning was a handicap to the men at NRL responsible for getting work started. For a first allotment of money they had to wait until December 1955 and in the interim had to draw on the Laboratory's small Naval Industrial Fund.7In mid-September they were still engrossed in contract negotiations with the Glenn L. Martin Company.
For both NRL and the Martin Company resumption of the contract discussions proved an uncomfortable experience. So far from achieving a meeting of the minds, the exchanges on the 8th of September and again on the 12th resulted in a sharpening of differences. As foreshadowed in the earlier session, the principal items of contention were two: who was to have overall "systems responsibility," and whether unilateral changes in specifications were to be permissible. NRL insisted that the Laboratory itself must hold the reins, the Martin men that the company must, once a firm contract had spelled out requirements. William Purdy declared that Martin's authority must embrace not only fabrication and testing of the launching vehicle but also control of supporting facilities and operations and of ground instrumentation for tracking and telemetering. In addition, Martin demanded freedom to "make design decisions within a system specification," an equal voice with NRL in defining the system specification, and, once that was fixed, no changes without company acquiescence. NRL replied that it would not accept "a closed-door policy on government-directed changes." The impasse on the major issues was so complete that NRL representatives wondered briefly whether they dared risk looking for some other prime contractor. Obviously that step was no solution. The Martin Company had the Viking experience and had prepared satellite launching studies during the spring and summer. A switch to another company would require the express approval of Assistant Secretary Quarles and the Stewart Committee and would certainly be costly in time. One young NRL engineer later said bitterly of the Martin negotiators, "They had us over a barrel, and they knew it."8
The irritation was mutual. The Martin representatives were apparently following the line laid down in the company's front office, but they also felt that concurrence with NRL's stated policy would subject them to needless harassment-a succession of Laboratory scientists invading the Baltimore plant and breathing down the necks of company executives and engineers, or worse, a series of peremptory directives from Washington making impossible demands and causing endless delays in what would be at best a very tight time schedule. Purdy and Robert Schlechter, who had both worked harmoniously with Milton Rosen and other members of NRL's Rocket Sonde Branch in developing the Viking rockets, were particularly irked at the reluctance on the part of the Laboratory to delegate authority to a contractor who had proved trustworthy and efficient in the past. But because the contract was important to the company as well as to the government agency, the men confronting each other across the conference table agreed to put to one side the question of specification changes and division of responsibility until Rosen and Schlechter had reviewed and revised the outlines of specifications and test schedules so as to permit drafting and signing of a "letter of intent." The letter of intent should serve as a preliminary to a final contract that would pin down reciprocal obligations and financial arrangements.
By 15 September the air had cleared slightly. The associate director of the Laboratory informed the Martin representatives that an "initiating contract" was ready for their scrutiny, that John Hagen was to be the administrator of the project with Rosen as technical director, that a redefinition of the mission had relaxed the eighteen-month requirement for launching but called for a "program of quality" at reasonable cost; the company was to submit cost figures the next day. Discussion then produced a specification outline that both groups considered a sound basis for "initiating system definition." Either group could request further meetings on specifications, and the same rules were to apply to subcontracted items as to Martin's. Disposition of the earlier contract for Vikings 13 and 14 was to be a separate negotiation. The flight test program presented by the Martin engineers, however, met with criticisms from James Bridger and Daniel Mazur because it was a minimum one with inadequate backup in some areas, and because more exact information on trajectories was needed, since Martin's computations assumed at each stage mass ratios which were far from optimum. Mazur pointed out, furthermore, that tests without the right telemetering instrumentation would be useless and that the Air Force Missile Test Center's safety doctrine was likely to impose severe "instrumentation weight problems"; design of the controls in the second stage in particular would require close attention to keep them within weight limits. John Mengel added that instrumentation for the third stage of vehicles spending substantial time above 200 miles of altitude was another problem to consider carefully.9 Those objections notwithstanding, the way now looked clear to signing a workable letter of intent within a few days.
The agreement as it took form labeled the Glenn L. Martin Company "the supplier of the launching vehicle" which was to orbit a satellite that "will enhance the prestige of the United States." The satellite was to be trackable by radio, optical instruments, or both. The contractor was to prepare the vehicle for launching, to suggest ways of helping the scientific phases of the undertaking, and by mid-January 1956 to have a complete time schedule worked out. The NRL Scientific Officer, the director of the Laboratory, or whomever he designated, namely Milton Rosen, was to have access to all data and, if he wished, to attend all conferences. Within forty-five days of signing the letter of intent the company was to submit to the Scientific Officer its recommended specifications and within sixty days have them in finished form. Any changes the government deemed essential were to go to the contractor in writing and any proposed by the latter must have Rosen's written approval. The company must notify him of the effects of changes on deliveries and costs. The detailed design, if up to specifications, need not have express government endorsement. The payload was to weigh 21.5 pounds. The minimal altitude of projection at the final stage must be 300 miles to ensure a perigee of more than 200 and an apogee of less than 800 miles. Eastward launchings from Cape Canaveral would give an inclination of the orbit to the equator of about 30°, but all vehicles were to be capable of achieving a 45° angle. Until a regular reporting system was evolved, the contractor was to submit semimonthly progress letters to NRL. A final contract containing the definitive specifications for the vehicle and a careful enumeration of government and contractor obligations should replace the letter of intent within 120 days.10
By 23 September, when the Navy and Glenn L. Martin officials signed the initial contract for Vanguard, Martin had set $13 million as its probable costs, although, in keeping with standard governmental procedures of "incremental financing," the sum written into the contract was only $2,035,033.11 Despite the unresolved questions of specifications, both parties thought the agreement reasonable.
Gratification at NRL evaporated a few days later: the Martin Company had just won the Air Force contract to build the airframe of the mighty Titan missile and was in the process of assigning a majority of its best men to start that job in Denver. Vanguard would have to get along without Purdy's hand at the helm in Baltimore and without other experienced engineers who had helped develop the series of Vikings step by step. While realizing that the Titan job would be bigger and more remunerative for the contractor, people at NRL were dismayed. So also were members of the Stewart Committee who had counted on Martin's putting, not some but all, its top-flight designers and engineers to work on Vanguard. The IGY staff at the National Academy too felt uneasy. Indeed all the principal sponsors of Vanguard believed that Martin was relegating the satellite project to the role of poor relation; denied a high priority within the Department of Defense, it had now become a second-string project in the manufacturer's books also. Martin executives denied the allegation when they learned of it; they were fully aware, they declared, that the satellite project was important and had assigned to it a fair proportion of their engineering talent. They had chosen Elliott Felt, Jr., to head the Vanguard team at the Baltimore plant. Though new to the responsibilities of top level management, he was an able, hard-headed engineer who had worked on controls for Vikings 9 and 10. Still, as Titan would require 1,000 designers and engineers and Vanguard some 300, the company was obviously going to have to spread its manpower resources thin. Whether or not the Titan contract did in fact seriously impede progress on the satellite vehicle perhaps mattered less than NRL's conviction that it was being shortchanged, for that belief shook the Laboratory's confidence in the contractor's good faith.
Equally troublesome from NRL's standpoint was Martin's new "get. tough" policy, based apparently on the company's ambition to enlarge its position in the industry and taking the form of a determination to yield as little as possible to extravagant demands of impractical scientists and government bureaucrats. Proud of having wider experience than any other firm in the country in designing and building rockets and ballistic missiles, the company considered itself entitled to speak with authority. At the first informal meeting in Captain Tucker's office on 25 August, Martin executives had outlined their position, but the NRL group had believed they would modify it when they understood more fully the purposes of the program. Unhappily, the contretemps between company officials and government scientists in Washington assumed new proportions with every passing week of the autumn.12
In essence the clash arose from conflicting philosophies, although not at first recognized as such. The "definition of mission" as stated in the initial contract fixed the Martin Company's commitment: to produce a workable launcher for an earth-circling satellite before the end of 1958. A hard enough job in itself, in Elliott Felt's view it should not be complicated by extraneous inquiries from the company's customer into why something worked or did not; if it failed, investigation should be confined to finding something that would function properly. The scientist, Felt observed wryly, always wanted to know the whys of success or failure; the practical engineer was intent on getting performance. It was for that reason that Martin had hoped to have a commissioned Navy officer, instead of a scientist, put in charge for the government. Felt was obviously oversimplifying, using the term "scientist" loosely, but the difference in attitude of mind he was pointing to was unmistakable before the end of 1955. NRL, as a scientific research body, contended that the study of failures could and should be as constructive as chance success; learning the whys was part of the job, an essential process in building up knowledge to draw upon for future work. Where the manufacturer was ready to try shortcuts and rely upon empirical data, the Laboratory held out for a more carefully analyzed approach. Hence NRL's insistence that it must not only supervise plans but monitor their execution. The ensuing controversies, intensified at times by the greater freedom the Air Force allowed the contractor on Titan, would wax and wane over the next two years but never entirely subside.13
In beginning their common task, both NRL and Martin recognized the importance of determining promptly the dimensions and estimated weight of the vehicle, the aerodynamic loads, and the thrust and efficiency of the engines. The first- and second-stage rockets had to have more powerful propulsion systems than either the Viking or the existing Aerobee-Hi carried. For the solid-fuel third stage even some of the preliminary requirements were still uncertain. And the weight and configuration of each stage would affect those allowable in the other two as well as overall performance. Had perfect harmony obtained between the Vanguard team in Washington and the men at the Martin plant, the problems confronting them in their race against time must still have been gargantuan. Until the major specifications were fixed for each part of the whole launcher, work at the drawing boards could not progress far.
A first necessity was to decide upon the optimum distribution of weight within the vehicle. Taking as a point of departure the calculation of Joseph Siry of NRL that a launch vehicle of about twenty thousand pounds gross weight would produce the required orbital velocity, at the end of September Bridger and Richard L. Snodgrass met with Robert Schlechter and other Martin engineers to discuss how much weight to allot to each of the three stages. The give-and-take at that session illustrates the difficulties of handling technical questions involving a number of unknowns. Although a large part of the vehicle was of standard design, recourse to extrapolation would be necessary at many points. When one of the Martin representatives argued that since nobody as yet had ascertained the minimum weights possible for various parts, a proposed "optimization parameter study" would be a waste of time, Snodgrass suggested that a rough guide based on preliminary estimated weights would save trouble, for if scrutiny of the three-stage system indicated that the second stage must not weigh over 2,000 pounds, the designer would not consider a control system suited to a second stage weighing 4,000 pounds; a second study based on detailed design weights could follow and a final study based upon hardware weights could modify that. While the GLM men doubted the feasibility of keeping to Siry's 20,000-pound figure, they agreed, since they had as yet no surer calculation to work from, to start with it in the endeavor to achieve a vehicle that could reach a peak altitude of 299.94 miles with a tangential velocity of 26,420 feet per second, an excess over orbital velocity of 2,890 feet per second. Martin engineers later pointed to their skillful employment of the analog technique as one of the significant contributions Vanguard designers made to the nascent science of space exploration.14
Concurrent with the weight optimization studies, work on the desired flight trajectory had to proceed, even though the computations had to be tentative until the weights of the vehicle stages were fixed and although a new plotting would have to be prepared for every flight. By means of special "calculation programs" employing both digital computers and Reeve Electronic Analog Computers (REAC), Navy scientists and Martin aerodynamic engineers arrived at attainable flight paths with an allowance of a six-degree deviation which would satisfy range safety requirements and still permit the transmission of scientific data to ground stations throughout the life of the satellite's batteries. In plotting the optimum path from the initial vertical lift at takeoff to the horizontal altitude of the orbiting satellite, analytical work undertaken during the early winter at the Naval Ordnance Research Center at the proving ground in Dahlgren, Virginia, provided valuable data to start with. "The Vanguard 3-D [three.dimensional] trajectory program," wrote a Martin engineer in 1960, "was probably the most important single tool of the project in defining and solving the design and flight problems." In the later design specifications covering the trajectory, the only change from those of the preliminary version was, first, the extension of the apogee to 1,400 miles and eventually the removal of any limit. In actuality, the first Vanguard satellite to attain orbit would have a 2,460-mile apogee but still proved able to transmit signals receivable on earth.15
Anxious to get subcontractors started on the job, in mid-September 1955 Martin with NRL concurrence had entered into negotiations with the General Electric Company to build the power plant for the first-stage booster. Reaction Motors Inc., the maker of the Viking engine, was working on a motor with a 75,000-pound thrust, but Rosen and Bridger thought that much power excessive for Vanguard and, furthermore, they dared not wait to see how the only partly developed RMI engine would perform. What GE had to offer looked like a better choice." Martin itself planned to supply the tankage and vector control actuators, but requested GE to furnish a self-contained unit which was to include the thrust structure, gimbal ring, engine components, and engine starting equipment. To use tooling methods originally employed in building Vikings, the diameter of the cylindrical tanks constituting the rocket casing was to be forty-five inches; the configuration of the GE engine must mate with that dimension, just as engine inlet pressures and temperatures must integrate with the Martin tankage and pressurization systems. The new GE X-405 engine was expected to achieve 27,000 pounds thrust with a specific impulse of 254 seconds at sea level. The term specific impulse is a measure of a rocket engine's efficiency-the higher the specific impulse, the better the engine. It is equal numerically to the pounds of thrust an engine produces by burning propellant at a rate of one pound a second. Delivery of the first of ten X-405 engines was to be on l October 1956. General Electric engineers accepted the terms, but Rosen asked for more detailed specifications on propellant utilization and insisted that GE supply a complement of spare parts and an itemized price list. With those changes incorporated, the purchase order from Martin to GE was signed on 1 October.17
A contract for the second-stage engine was harder to arrange. The preliminary specifications based on modifications of the small Aerobee-Hi rocket called for hypergolic propellants-fuels that ignite spontaneously when mixed-and regenerative cooling of the thrust chamber. Those features together with pressure feeding of the fuel into the thrust chamber promised to provide reliable means of starting the engine at an altitude of about thirty-six miles, even though igniting a pump-fed engine at a high altitude was a still untried system.18 The Aerojet General Corporation, maker of the Aerobee-Hi, had expressed interest in a contract in mid-August, but the Bell Aircraft Corporation also had claims to consider because of its extensive experience in developing liquid-fuel engines, notably for the Nike antimissile missile. While Martin approached both contractors, Bridger took counsel with the Air Force and Redstone Arsenal. The Air Force put the competitors on a par; the Army gave Bell's performance a slightly higher rating. When the two companies presented their proposals for second-stage propulsion systems complete with structure and tankage, Martin engineers judged Bell's 8,000-pound-thrust, pressure-fed, integral-tank design technically superior to Aerojet's 5,000- to 7,500-pound-thrust, turbopump system. But Aerojet's request for a chance to prepare a pressurized integral-tank design postponed a decision.19
While Martin was drawing up a purchase
order to the Minneapolis-Honeywell Company for the development
of the rocket's guidance and control system, Aerojet won the engine
contract, partly, at least, by putting in a bid of $1.03 million,
less than half Bell's more carefully figured price of $2.675 million.
As things turned out, the costs of the second-stage engine package
exclusive of the thrust vector control actuator and its hydraulic
tanks would exceed $4 million. The development of the entire second
stage, with the complex of equipment it had to accommodate, came
to be one of the principal bêtes noires of the project
during the next two years. Yet eventually it would constitute
one of Vanguard's notable contributions to the design of reliable
Bridger's optimism proved ill-founded.
By early November, although Martin increased the allowable weight
to 365 pounds, Thiokol declared it could not build a rocket of
the type needed unless hydrostatic proof pressure were reduced,
the case diameter were increased from 11 to l4 1/2 inches, and
25 pounds were added to the gross weight, or else the allowable
burning time of the fuel were shortened. Confronted with Thiokol's
ultimatum, Martin and NRL felt compelled to restudy the original
specifications and in the interim to seek proposals from four
other companies for the third-stage engine. There the matter stood
foreground, Aerojet General's second-stage engine.
In the autumn of 1955, however, it was the third-stage rocket that appeared to be the most formidable part of the undertaking. In August the Thiokol Chemical Company had seemed the most likely subcontractor and, with Redstone Arsenal collaboration, had prepared an exploratory study of how to meet the Martin-NRL tentative specifications. But at the end of September a Thiokol spokesman had declared that Martin's demands went beyond the present state of rocketry art. Thiokol could meet the weight specifications for metal parts, but the required impulse would need twenty more pounds of propellant. He saw no way of producing the slow burning time in solid fuel of the diameter GLM called for, and, he contended, under those circumstances it was impossible to reach the thrust level specified; the safety requirement, moreover, was excessive, higher than that set for aircraft rockets. In reporting upon this gloomy analysis, James Bridger noted half-humorously: "All is Not Lost, however." By increasing the diameter of the case, allowing "more optimum expansion," a less exaggerated safety requirement, and use of the new TRX-217 propellant which, tests indicated, could give a burning rate of 0.092 inch per second instead of the 0.28 inch per second of the earlier design, the solid-fuel rocket wanted for Vanguard should be realizable. Slow burning was important to minimize the jolt caused by a rapid buildup of thrust which might damage the delicate instruments in the satellite. As soon as a compromise was reached on those technical requirements, he concluded, progress should be rapid, provided funds were forthcoming promptly.
Bridger's optimism proved ill-founded. By early November, although Martin increased the allowable weight to 365 pounds, Thiokol declared it could not build a rocket of the type needed unless hydrostatic proof pressure were reduced, the case diameter were increased from 11 to l4 1/2 inches, and 25 pounds were added to the gross weight, or else the allowable burning time of the fuel were shortened. Confronted with Thiokol's ultimatum, Martin and NRL felt compelled to restudy the original specifications and in the interim to seek proposals from four other companies for the third-stage engine. There the matter stood till mid-December.21
Organization of the Vanguard teams meanwhile had been taking form at the Martin plant and at the Laboratory. In Baltimore Raymond B. Miller, Jr., was made contracts manager; energetic, thirty-year-old Donald Markarian became project engineer, with Robert Schlechter, Leonard Arnowitz, Joseph E. Burghardt, Russell Walters, and Sears Williams as his assistants. By the end of October, under this supervisory staff, about fifty designers and aerodynamic, electromechanical, and propulsion engineers were working on Vanguard plans. After the readying of production lines and test facilities began, the number rose to some three hundred. At NRL matters moved more slowly, largely because John Hagen wanted to work out a clearcut division of responsibility between one unit and another and to be sure that the head of each, irrespective of his individual attainments, could work effectively not only with his fellows at the Laboratory but also with his counterparts in other government agencies and industry.22
As Hagen himself had to devote much of his time to arranging for collaborative services from other branches of the government, to obtaining money from the DoD Emergency Fund or the National Science Foundation, and to the preparation of memos, speeches, and formal reports, additional administrative staff was necessary. So J. Paul Walsh was transferred from another division to become Vanguard deputy director and planning manager; a young man, Walsh brought to his new assignment the vigor and technical leadership which had marked him early in his career in testing nuclear devices in the South Pacific. Charles De Vore, public relations officer for the Laboratory, took charge of that task for Vanguard during its first two years. At the same time, Rosen engaged a few new men, but most of his then thirty-five man staff was already part of NRL and familiar with particular aspects of its work; he estimated that within six months he would need fifty-three men and within a year sixty-six. In late November he set up a technical board consisting of himself, his three branch chiefs (Bridger, Mengel, and Mazur), his engineering consultant Winkler, Walsh, and James Fleming of the Laboratory's Applications Research Division. Homer Newell, in charge of satellite instrumentation, had no separate staff as such; instead he drew on the special talents of some thirty men working in various divisions of the Laboratory but not assigned directly to Vanguard; later nearly a hundred experts worked with him in seeking solutions to such problems as excessive heating of the instruments.
To compute the orbits of close-in
satellites Hagen appointed a committee composed of Joseph
Siry of NRL, Gerald M. Clemence and R. L. Duncome of the Naval
Observatory, and Paul Herget of the University of Cincinnati.
For military liaison officers assigned full time to the Laboratory,
the Navy first named Captain B. F. Herold and later Commander
Winfred E. Berg; the Army appointed Major John T. O'Hea, the Air
Force, Lieutenant Colonel Asa Gibbs. All told, what John Hagen
called the "in-house management staff" consisted of
fifteen men.23 In view of the official DoD announcement
that the satellite program was to be a three-service undertaking,
the question may arise as to why the Army and the Air Force had
only one representative each on the Vanguard staff, and why Hagen
did not invite Wernher von Braun, for example, to serve as a consultant.
Inasmuch as ABMA in 1956 was making overtures to Hagen to enlist
his support of an Army plea to use the Redstone-Orbiter launchers
as a backup for Vanguard, there is some reason to suppose that
von Braun would have accepted the invitation to sit in on Vanguard
councils. But, as Hagen later explained, Army collaboration had
been wanting when he needed it and he had concluded that the DoD
had never "envisioned a greater degree of interservice participation
than we [NRL] were able to generate."24 As it
was, the Army Corps of Engineers and the Signal Corps gave the
project unstinting cooperation from the beginning. Officers of
both Corps went out of their way to cut through entangling departmental
red tape and to open doors that might otherwise have long remained
closed to civilians and officers at NRL. And after Vanguard field
crews arrived at the Cape, the Air Force in turn tendered all
The entire Vanguard team at its peak numbered 180 persons, including clerical help and shop hands. During l955 some jockeying for positions of authority went on and a few complaints sounded in other divisions of the Laboratory that Vanguard appointees were disrupting the rest of NRL, but although friction among members of the team recurred from time to time, the Vanguard group early came to be a tight-knit organization bound together by an exceptional esprit de corps. The excitement and passionate interest in the job extended also to the men not formally assigned to the project but who, whether working full time or only part of every month on constructing and testing the instruments for the satellites, considered themselves and were considered part of the Vanguard team. "Team" was the key word. Set up in a fashion that defies exact charting, the organization functioned as a unit. No one within it or associated with it acted on any matter of importance without consulting with Hagen and his management staff, for every part of the operation affected every other. If that procedure sometimes caused delays, it prevented confusion and trouble at a later time, and it heightened the sense of mutual responsibility that all members of the team shared.25
Enthusiasm was put to a severe test during the late autumn of 1955. In the midst of skirmishes with Martin officials, NRL and ONR were struggling to secure from the Air Research and Development Command assurances of accommodations for Vanguard tests at the Air Force Missile Test Center at Cape Canaveral. After Mazur's exploratory trip to the base in early September when the Air Force spokesman there outlined the objections to such a plan, and after Vanguard administrators had seen the full text of AFMTC regulations, Vanguard heads had looked for an alternative site. White Sands Proving Ground, where the Navy had equipment for vertical launchings of sounding rockets, had to be ruled out for the satellite program because of the danger to populated areas from falling burned-out first- and second-stage rockets. Use of the Navy installation at Roosevelt Roads in the Caribbean would not involve that peril but would be prohibitively expensive, and inconvenient as well. Hence a formal request from the Navy for space and equipment at the Florida base went to the Air Force on 2 November.
A month later an official Air Force endorsement approved "in principle the support of Project Vanguard at AFMTC" but warned that new construction would probably be necessary or else joint use of existing facilities. The Air Force could not pay for new installations and, if joint use of existing facilities were attempted, "it is possible that other missile programs will be delayed."26 Unfortunately investigation revealed "that the Redstone facility is the only facility at AFMTC that is suitable with proper modifications for Vanguard operation." And, as Captain Tucker informed the Chief of Naval Research and Assistant Secretary Quarles on 15 December-
It is now stated by the Army representative at AFMTC that these facilities will not he available......Since the first [Vanguard] launchings are scheduled for October 1956, it is apparent that even if money were available and an adequate priority obtained, scheduled commitments for the initial phases of the program cannot he met by totally new construction. Therefore, in order to meet the Vanguard schedule, funds for construction, suitable priorities, and sharing of existing launching facilities are mandatory.
The Test Center Commander also advises
that he must have authorization and funds by 1 January 1956 for
any construction necessary to the project.27
Although the Laboratory won permission to draw for this purpose on the DoD Emergency Fund and had worked out with the Martin Company a statement of Vanguard test requirements, the road ahead looked thorny: the badly wanted high priority for the satellite project was not forthcoming. Lack of priority meant that the Vanguard team could hope for little consideration in obtaining a launch pad equipped with the necessary pipelines and wiring or the privilege of having Vanguard share a blockhouse with a guided missile project. To build those facilities from scratch would take more time and money than the IGY program could afford. And the question remained: how, without adding excessive weight to the rocket and consequently sacrificing velocity, to incorporate in the launch vehicle the safety devices upon which the AFMTC insisted? The Center's headquarters declared that the first and second stages of the vehicle must have both a power cutoff and a destruct receiver capable of functioning from the first moment of flight, and demanded for each Vanguard trajectory such data as fuel weights at successive intervals from launching through the burnout of the final stage, the predicted number of fragments created by a destruct explosion, their velocity, and the time delay between activation of the firing circuits and the first motion. Altogether, the permission to launch Vanguard from Cape Canaveral appeared to introduce nearly as many problems as it solved.28
Negotiations for other facilities, on the other hand, proceeded fairly smoothly. Martin had asked in September and the Navy had arranged to have the Malta Test Facility north of Schenectady reactivated for tests of the General Electric engine. Hagen's staff had endorsed the recommendation of Rosen and Robert Schlechter that static testing of the fully assembled first stage not be attempted at Malta because shipment of tankage and instrumentation from the Martin plant to GE would be expensive and would delay production in Baltimore. Doubts later arose about the wisdom of the decision to dispense with full-scale first-stage static tests at Malta. Schlechter for one believed that they would have saved endless troubles at Cape Canaveral and in the long run have proved an economy, but even severe critics of the quality of the rockets sent from Martin to the Cape point out that equal or worse difficulties might have followed from choosing the other course.29 Although static firing tests could not be conducted in the heavily populated Baltimore area, NRL agreed to finance the building of other special test facilities on Martin property at Strawberry Point on Chesapeake Bay and to meet the cost of constructing at the main plant an elaborate steel tower to permit assembly and inspection of the entire first-stage rocket in a vertical position. The immovable gantry on a forty- by twenty- by three-foot concrete pad would be disassembled after the Vanguard program was over.
In seeking approval of having the government foot the bills for these "costly" installations, Admiral Rawson Bennett, Chief of Naval Research, felt impelled to explain to the Secretary of the Navy that construction and testing for Vanguard involved problems for which there was no existing precedent. No contractor had ever undertaken this kind of job and no one had the testing facilities needed.30 Nor did that list include the new devices NRL had to have in its shop. For environmental testing, for example, the Laboratory would need a pressure chamber big enough to hold a thirty-inch spherical satellite in which the pressures could be raised from that of 0.01 millimeter of mercury to that of 1 atmosphere; the chamber should provide means of cooling or heating the satellite by radiation; for vibration tests an electromagnetic shaker capable of giving a 625-pound force, a longitudinal and centrifugal accelerator and a shock simulator for use in testing the spin-stabilized third-stage rocket. Equipment already on hand could be adapted to some of this work. Neither the Navy brass nor the Office of the Secretary of Defense protested over expenditures for testing facilities, even though the cost appeared likely to run to nearly $80,000 for Martin alone.31
Dealings with the Army Corps of Engineers also went well when NRL approached the Corps about constructing satellite tracking stations. Before discussing plans with the Army, Hagen, Walsh, Mengel, and Roger Easton, Mengel's collaborator in devising the Minitrack system, concluded that eight stations spaced about 850 miles apart in a north-south "fence" along the 75th meridian from Maryland to Santiago, Chile, would provide a sixty percent probability of making a measurement on each orbit of the satellite at an average altitude of 200 miles. NRL proposed to place a station for testing the system and training field crews at Blossom Point, south of Annapolis, Maryland. The Navy itself would man a radar station at Antigua in the Bahamas to observe satellite performance at the moment of third-stage separation, but requested the Army Engineers to build, maintain, and operate the tracking stations in Cuba, Panama, and South America. The Engineers were willing to cooperate. Selection of the exact sites in foreign territory would have to await State Department approval and the results of an exploratory trip by a team composed of members of the Inter-American Geodetic Survey, the Army Map Service, and NRL representatives. The survey of possible sites was set for late February 1956.32
As 1955 drew to its close, men responsible
for the satellite venture could see a good deal of progress. At
the National Academy plans for choosing scientific experiments
to attempt in the instrumented bird were moving forward under
the aegis of an IGY satellite panel.33 At NRL a well-organized
team had been formed to ensure that a satisfactory launcher was
available at the earliest possible moment, that a flyable
bird was ready, and that tracking and data reduction were as accurate
as scientific knowledge and technical knowhow permitted. The comptroller
had set up orderly procedures which promised to ease, if not obliterate,
difficulties so getting money released for approved purposes.
NRL bad enlisted the help of Army engineers in establishing tracking
stations. The Martin Company had aligned three major subcontractors
and had several studies in process calculated to ensure sound
design of the vehicle. The chief monkey wrench in the works was
still the failure of the Laboratory and the primary contractor
for the launching vehicle to reach agreement on final specifications.