April

A potential problem area exists with respect to three hydraulic systems. If it should be declared by Propulsion and Vehicle Engineering, Astrionics and Quality Laboratories that the three systems must be checked, the launch date [25 April] cannot be met.1

Marshall engineers had made one significant change in the SA-2 booster design, placing additional baffles in the propellant tanks to prevent a recurrence of the sloshing experienced in the latter part of the SA-1 flight. The countdown on 25 April went smoothly; the only hold came when a ship strayed into the flight safety zone, 96 kilometers downrange. The successful flight was terminated with a dramatic experiment. When SA-2 reached an altitude of 105 kilometers, launch officials triggered the command destruct button. Project "High Water" released 86,000 kilograms of water from the dummy upper stages, giving scientists a view of a large disturbance in the upper regions of the atmosphere. A massive ice cloud rose 56 kilometers higher in a spectacular climax. 2

SA-3 (16 November 1962)

A tropical storm greeted the SA-3 vehicle's arrival at the Launch Operations Center on 19 September 1962. Three days of rain and high winds delayed erection of the booster, and conditions were still unfavorable when the launch team resumed work on the 21st. Aeronautical Radio Incorporated engineers, hired by NASA to review Saturn operations, reported: "The erection operation was safely performed but is rather hazardous, with technical personnel climbing around on top of the horizontal booster to install hoisting equipment. This operation was performed on the slick plastic covering of the S-1 stage in a wind of up to [37 kilometers per hour]." The Aeronautical Radio team considered the preparation prior to stage erection (removing the end ring segments) "a relatively slow, inefficient, and dangerous operation, with a considerable amount of trial and error," and recommended more familiarity with the instruction handbooks. During the eight-week checkout, the Washington, D.C., firm found other shortcomings such as "the use of metallic hammers to urge recalcitrant components into place." The observers noted that proper tools were not always handy, "and expediency sometimes prevailed." They concluded, however, that the "efficiency and dedication" of Hans Gruene's Launch Vehicle Operations Division** was instrumental in the success of the Saturn test.3

SA-3 lifted from Cape Canaveral on 16 November 1962. Debus asked von Braun not to invite outside visitors, as the United States armed services were still on alert for the Cuban missile crisis. The rocket incorporated a number of important new features. The first two Saturns had used 281,000 kilograms of propellant, about 83% of the booster's capacity. Marshall, wanting information for the new Saturn IB program, flew SA-3 with a full propellant load to test the effects of a lower acceleration and a longer firststage flight. The flight also tested the retrorockets that would separate the two live stages on SA-5, the first launch of the upcoming block II series. SA-3 flew three other important prototypes: the ST-124 stabilized platform, a pulse code modulated data link, and an ultrahigh-frequency link. The stabilized platform was a vital part of the Saturn guidance and control system, containing gyroscopes and accelerometers that fed error information to the control computers, which provided steering signals to the gimballed engines. The data link's importance lay in its ability to transmit digital data, a vital ingredient in plans for automation of checkout and launch procedures. The ultrahigh-frequency link would be used to transmit measurements, such as vibration data, that could not be handled effectively on lower frequencies.4

SA-4 (28 March 1963)

SA-4 ready for launch from LC-34, March 1963.

SA-4 set records for the shortest launch checkout (54 days) and the longest countdown holds (120 minutes) of the block I series. At T-100 minutes on launch day, test conductor Robert Moser called a 20-minute hold while the launch team adjusted the yaw alignment of the ST-90 gyro guidance platform. Readings from a ground theodolite showed that the platform was not properly aligned on the launch azimuth. An operator oriented the Watts theodolite on a geodetic survey line and then turned the head of the instrument to the launch vehicle. The alignment prism in the ST-90 platform reflected a light directed from the theodolite. If the platform was aligned properly, the reflection from the prism appeared in the center of the theodolite's scope. In this case, the problem was with the theodolite and not the gyro platform.

The final hold came at T-19 minutes as a result of a LOX bubbling test. Andrew Pickett's propulsion group performed the test late in the countdown to verify the flow of helium to the LOX suction ducts of the eight engines. The decreasing temperature of the LOX indicated a proper flow of helium, but the propulsion panel did not register a signal that the LOX bubbling valve was open. Without the signal the terminal sequencer would shut down. Pickett's team, along with Isom Rigell's electrical engineers, improvised a bypass for the valve signal on the sequencer. The propulsion team assured a proper LOX temperature for the Saturn and then initiated the bypass manually as the sequencer brought the vehicle to liftoff.5

In SA-4's most important test, officials deliberately shut down the number 5 engine 100 seconds after liftoff. Booster systems rerouted propellants to the seven other engines. Contrary to some predictions, the shutdown engine remained intact and the imbalance of hot gases on the engine compartment heat shield had no ill effect. The SA-4 vehicle simulated all block II protuberances on the dummy second stage, e.g., fairings and vent ducts, to determine the aerodynamic effects of a live second stage. Block II antenna designs were also flown. The SA-4 vehicle employed a new radar altimeter and two experimental accelerometers for pitch and yaw measurements. After the successful flight, the von Braun team in Huntsville looked confidently toward two-stage missions.6

Pad damage from the first four launches did not surpass expectations. Restoration cost an average $200,000 and took one month. LVOD officials were particularly interested in assaying pad damage after the launch of SA-3. One of the mission's goals was to determine the effect on the pad of an increased propellant load with the consequent slow acceleration and longer exposure to rocket exhaust. The damage was comparable to the first two launches. The only effect readily attributable to the slower acceleration was increased damage to the pedestal water deluge system (the torus ring) and a warping of the flame deflector.7

The LOX fill mast at the base of the rocket had to be replaced after each launch. The 21-meter cable mast assembly extending up alongside the rocket also crumpled during each of the first two launches. After watching the long aluminum fixture collapse the second time, officials replaced it with an umbilical swing arm. The Huntsville engineers converted a swing arm intended for the SA-5 launch and shipped it to the Cape in early August. At LC-34, Consolidated Steel and Ets-Hokin-Galvin began work on the new umbilical tower two weeks after the SA-2 shot.# The swing arm, mounted in August, suffered very little damage in the SA-3 launch.8

* Boilerplate means a full-scale model of a flight vehicle flown on research and development missions, without some or all of the internal systems.

** See chap. 7. From 1 July 1962 to 24 April 1963, LVOD was a division of MSFC. Since Debus and Gruene served as Director and Deputy Director of both the Launch Operations Center and LVOD, this was an administrative distinction with little or no bearing on launch activities.

# Saturn construction became rather complicated at times. LOD personnel observed that the column splices connecting the new construction to the existing 8-meter base were not consistent with Maurice Connell & Associates design drawings. In a letter to the Corps of Engineers, Debus stated, "Upon investigation, it appears as though the Jacksonville District Office had instituted changes in the original design without the concurrence of LOD, who has the design responsibility." The fabricator of the first phase steel had apparently erred in the column's angle of slope. The Corps solution, using one-inch diameter interference body bolts, was satisfactory; but the construction teams were using one-inch hightension bolts, which had only two-thirds the necessary strength. Debus requested that the Corps get LOD's approval in future modifications.

Previous Page

Table of Contents

Next Page

A Second Saturn Launch Complex-LC-37

Block I - the first four Saturn launches had gone up from launch complex 34. With the block II launches (SA-5 through SA-10), the program would move to new facilities at launch complex 37. The second complex had originated with the Hall Committee study of 1959, which found that an explosion would render LC-34 useless for a year [chapter 2-6]. On 29 January 1960 Debus asked Dr. Eberhard Rees to approve a second Saturn complex. Since LC-37 would serve primarily as insurance for LC-34, no major design changes were anticipated. Taking into account the rising costs on complex 34, Debus estimated the price of LC-37 at $20 million (roughly one-third more than LC-34's costs as of January 1960). In his report, Debus warned Rees that LC-37 would likely be sited at the undeveloped north end of Cape Canaveral, 1,220 meters north of LC-34 and 425 meters from the Atlantic Ocean. A complex at that location "would require utility capacities of unusually large magnitudes and the cost to Saturn, as the initial [user] could be excessive."9 In February 1960, representatives from the Missile Firing Laboratory, Army Ballistic Missile Agency, and the Air Force Missile Test Center estimated demands for water, power, roads, communications, and instrumentation at LC-37 and discussed the cost of extending these to the proposed site. Eventually, development for LC-37 included a new electrical power substation and transmission lines, a 3,785,000-liter water reservoir, and a pumping network, at a price of $2.5 million. 10

Proposed launch complex 37.

Hoping to have LC-37 ready for backup duty by January 1962, MFL originally set a mid-1960 deadline for criteria on the launcher, umbilical tower, and propellant systems. 11 Debus's decision to put a new service structure on LC-37 dashed these plans. Harvey Pierce, a Connell engineer, had prompted the change. Pierce had played an important role in designing LC-34 and more recently on the Hall Committee. On 26 February Pierce had written Debus about some inherent shortcomings in the inverted U service structure and recommended the formation of a study group. 12

By mid-April 1960 Albert Zeiler was directing a two-pronged investigation into problems encountered with LC-34's service structure and concepts for a larger one. The latter reflected NASA's decision to build LC-37 for both C-1 and C-2 versions of the Saturn. 13 The service structure committee met periodically over three months to review 21 concepts proposed by NASA officials and private industry. No proposal proved fully satisfactory; attractive features from several were combined in the final recommendation. The committee concentrated on a half-dozen aspects of the service structure design, posing these alternatives:

• Mobile or fixed structure?

• Bridge crane or stiff-leg derrick for hoisting?

• Protection for the launch vehicle from wind loading or absorption of the rocket's wind loads into the service structure?

• Open or closed service platforms?

• Launch stand above or below ground?

• Collapsible or fixed umbilical tower?

The fixed service-structure designs were attractive since they offered economy and good utilization. The committee, however, feared the effects of a pad explosion on a fixed structure. The fixed design also posed a difficult engineering problem. Long cantilevered platforms with elaborate retracting mechanisms were needed to keep the main structural frame outside the rocket's drift cone (the safety allowance for effect of surface wind at launch). At an 11 July meeting in von Braun's office, Marshall officials discussed the effects of wind drift, thrust malalignment, and loss of one engine on the clearance requirement for a service structure or umbilical tower. The participants agreed to a 12-meter clearance between the vehicle's center line and the nearest obstruction at the 91-meter level. About the same time the Zeiler committee opted for a mobile service structure. 14

The hoisting matter was settled in favor of a stiff-leg derrick mounted on top of the service structure. Although a bridge crane offered more flexibility, its use in the upper reaches of the service structure would obstruct the vertical escape trajectory of a manned payload. In the final design a 40-ton mobile crane positioned at a lower level assisted the 60-ton main hook on the stiff-leg derrick.

The question of wind loads arose because the Saturn was not self-supporting in high winds. One alternative was to design "hard point" connections between the vehicle and service structure platforms. This would require additional structural members on the rocket, increasing its empty weight. It would also add considerable stress to the service platform. The committee chose a design enclosing the launch vehicle in a 76-meter silo of five sections that eliminated wind loads and protected the rocket from flying debris. The silo design also solved the service platform problem. The committee recommended a minimum of ten adjustable work platforms in the structural steel frame silo. Air conditioning would provide the necessary ventilation during propellant loading.

The committee rejected a plan to put the launch stand below ground with the flame diverted into side trenches. Doing so would reduce the height of the service structure, but the higher costs of subsurface facilities, due to Cape Canaveral's high water table, were unacceptable. 15

In designing the umbilical tower, the major concern was separation of the umbilical connections from the launch vehicle at liftoff. The committee studied jointed, collapsing towers; towers supported by cable catenaries (a curved cable suspended from two poles); and pivotal reclining structures. The size and weight of the Saturn umbilical connections - propellant piping, pneumatic lines, instrumentation circuitry, and electrical power lines - rendered all those concepts impractical. The committee recommended a free-standing umbilical tower, with ties to the service structure for support against high winds. Swing arms, entering the silo enclosure through cutouts in the platform mating edge, would connect the umbilicals to the launch vehicle. 16

In August 1960 the launch team approached von Braun about adding a second pad to the LC-37 complex. The additional pad would provide a backup for Saturn C-2 launches and reduce launch time by one-third, since it would eliminate the month needed for pad repairs. Von Braun directed Debus to add a second pad on LC-37 if funds could be secured. Before the meeting adjourned, General Ostrander, Office of Launch Vehicle Program Director, arrived. After reviewing the proposal, Ostrander agreed to provide $700,000 for the initial modification work. 17

Further revision of LC-37 plans occurred in early 1961. In January Debus heard of an Air Force-sponsored study on blast potentials of the Atlas-Centaur rocket. The Arthur D. Little Company findings, Debus informed von Braun, "indicated a problem of considerable magnitude with Saturn complex siting." 18 Since there was little data on liquid hydrogen's explosive characteristics, the calculations were tentative. The Little report, however, reinforced the Hall Committee's conclusions. On 12 January Debus asked Petrone, as Saturn project coordinator, to investigate the explosive potential of liquid hydrogen and determine the cost of extending pad distances beyond 183 meters. The distance between pads was subsequently increased to 365 meters.19

Two Florida firms won the LC-37 design contract: Connell and Associates prepared the service structure and umbilical tower designs, while Reynolds, Smith, and Hills handled the subsurface facilities. The architects' design work extended from February to July 1961. During the same period, Gahagan Company dredged thousands of cubic meters of sand from the Banana River onto the LC-37 site. Vibroflot machines began their work at the complex in mid-July. Blount Brothers Construction Company of Montgomery, Alabama, won the pad B construction bid in August 1961 and started work the following month. The project was 45% complete on 30 March 1962, when the Corps of Engineers awarded Blount Brothers a contract to build pad A. 20

LC-37 under construction, January 1963.

The new construction soon overshadowed the older Saturn facility. LC-37 was nearly three times larger than LC-34. The two umbilical towers rose 82 meters from a 10-meter-square base. Stability of the towers in high wind presented a challenge to the designers. The large number of electrical, propellant, and pneumatic lines running up through the lofty structures gave the tower surface a wind resistance nearly equivalent to a solid wall. At the base of each tower stood a four-story building (one floor was underground) containing a generator room, high-pressure-gas distribution equipment, and a cable distribution center. The building would later house digital computers for the automated checkout.21 Hydrogen burn ponds were an added feature on LC-37. The gaseous hydrogen boiled off from the LH2 storage tank and the S-IV stage and flowed several hundred meters through pipes to the burn pond. The LC-37 launch control center, or blockhouse, was similar to LC-34's, but half again as large. By far the most imposing of LC-37's facilities was a 4,700-ton, 92-meter-high service structure, containing four elevators, nine fixed platforms, and ten adjustable platforms that allowed access to all sides of the vehicle. The six semicircular enclosures could withstand 200-kilometer-per-hour winds. When completed in 1963, the self-propelled, rail-mounted structure was the largest wheeled vehicle in the world. 22

The LC-37 service structure at pad B.

The LC-37 service structure in the open position, February 1963.

The industrial area on the Cape. Hanger AF is in the upper left. The causeway (under construction) leads to Merritt Island in the distance.

Mating spacecraft modules inside Hanger AF, March 1964

Erection of a special assembly building was a third construction project for Saturn I in 1962. Some novel building designs were rejected before deciding on a conventional hangar configuration. The new hangar AF was in the Cape industrial area, a short distance from the Saturn dock. A bridge crane in the hangar's main bay provided a lift capability for the initial upper stage checkout; lean-tos on both sides provided extra office space. The Launch Vehicle Operations Division performed some preliminary checkout work in hangar AF, but half of its big bay was soon given over to Gemini and Apollo spacecraft operations. 23

Previous Page

Table of Contents

Next Page

Was this article helpful?

0 0

Post a comment