In this paper the overall Atlas launch-vehicle program in support of Project Mercury is discussed. The paper includes the areas of both management and operations. Implications to be drawn from the presentation are that sound planning by experienced Air Force personnel early in the program; strong' top-level management support; great attention to engineering, manufacturing, and operational detail; and strong individual motivation have been responsible for the success of this portion of Project Mercury. The procedures used in the launch-vehicle program were not conceived or promulgated by any one individual overnight. Rather, they grew from the experience of many and were further shaped by the program itself as it progressed.
This paper presents the management, aspects of the launch-vehicle system in redirecting a ballistic-missile weapon system into a launch- vehicle system for manned space research. Early agreements between the U.S. Air Force and the National Aeronautics and Space Administration (NASA) established the program responsibilities and identified the management interfaces. Specific guidelines were laid down by the Air Force Chief of Staff to provide effective support to NASA within the military framework of what was then known as the Air Force Ballistic Missile Division. Definitive policies were established to insure maximum launch-vehicle safety for the pilots. The initial overall Mercury systems engineering as it affected the launch-vehicle was performed by U.S. Air Force/NASA technical panels and then gradually shifted to the Air Force and its technical contractor, Space Technology Laboratories, and more recently the Aerospace Corporation, for more specific systems engineering.
The basic Atlas "D'' system as it existed at the beginning of the program is described to provide a basis for the explanation of the launch-vehicle modifications that, were required to support, the mission. A brief description is given of the problems that were associated with the individual launch-vehicle flights and the results of the postflight evaluations. A more detailed postflight evaluation is given of the MA-9 flight.
During the mid-1950's, the U.S. Air Force conducted a number of studies dealing with manned space flight. Many plans had been formulated and several of the programs had reached a detailed development plan state when, in August 1958, the President directed the assignment of the man-in-space effort to the National Aeronautics and Space Administration. On October 7, 1958, the Space Task Group was organized at Langley Field, Virginia, to manage the shell established and later named, Project Mercury.
During the period from October 1958 until April 1959, a series of meetings took place between NASA and the Air Force Ballistic Missile Division to define the AFBMD support required by the NASA-Space Task Group. The problems considered included: definition of the scope of NASA's effort, definition of launch-vehicle requirements, definition of procurement procedures, launch schedules, and launch facilities. It is interesting to note that at the time of the first NASA visit to AFBMD on October 21, 1958, the proposed program envisioned over 25 flights using the Redstone, Thor or Jupiter, and Atlas launch-vehicles. Spacecraft  orbital weight was to be approximately 2,100 pounds for a 120 nautical-mile orbit. Additional meetings culminated in the issuance to AFBMD of NASA Order HS-24 on November 23, 1958, which specifically requested that the Air Force supply one "C" series Atlas to support Project Mercury. The order specified that this was the initial request of a proposed program which would require approximately 13 boosters of the Atlas and Thor class. On December 8,1958, AFBMD received NASA Order HS- 36 which requested nine "D" series Atlas boosters. Subsequent amendments to HS-36 deleted HS-24, changing the total requirements to 10 Atlas "D" vehicles, later to 14 "D's," eliminating the Thors. Further discussions between the two agencies resulted in the agreement that the Air Force would have full responsibility for the development, procurement, production and launch of the Atlas vehicles for Project Mercury (see fig. 5-1). The final meeting of this series was held between General Schriever, then Commander AFBMD, and Dr. Glennan, Administrator of NASA, on April 7, 1959, in Washington. The basic memorandum of understanding between NASA and the USAF grew from this conference.
A program office was established within the AFBMD to manage the launch vehicle effort, and the services of the Space Technology Laboratories (STL) were requested within the framework of the Atlas weapons system program to support Mercury. Specific guidelines were laid down by the Commander of AFBMD in order that maximum responsiveness to NASA requirements could be assured.
The early systems engineering was accomplished within the framework of technical panels established by NASA. Participants in the panel work were drawn from various NASA organizations, McDonnell, AFBMD, STL and the Atlas manufacturer, General Dynamics/ Astronautics. Once the initial problem areas had been defined, technical panels were sub. divided into working groups with specific technical areas assigned to assure that thorough treatment was given to all engineering problems. Through the medium of the technical panels, basic trajectory conditions were developed. The launch-escape system concept was born and specific requirements were developed Reliability goals were established, and systems restraints were imposed. In order to implement, in detail, the general systems approach ' developed through the technical panels, the Air Force called upon STL to perform these tasks . It was necessary to institute a special systems engineering and technical direction effort for the Mercury/Atlas program, and the STL Mercury Project Office was established in the Fall of 1959 under the direction of Mr. B. A.Hohmann. In the summer of 1960, when the Aerospace Corporation was organized, the task was transferred to this new organization. The majority', of the STL Mercury office personnel transferred to Aerospace continued to perform their original jobs. The basic responsibilities of the systems engineering and technical direction group were to develop the technical requirements, monitor the systems and launch-vehicle development, provide trajectory calculations and guidance equations, analyze both ground and flight-test results, assure production acceptability of the launch-vehicle, assist in administering the pilot safety program, and provide systems integration of the Atlas associate contractor's systems.
 The Space Systems Division and the Aerospace Corporation program offices together were the focal point for detailed management of the launch vehicle program. Program requirements reached this level along a formal path (see fig. 5-2) established from Headquarters NASA to Headquarters USAF, to the Air Force Systems Command (AFSC), to Space Systems Division (SSD), to the Deputy for Launch Vehicles (SSV) to the program offices. A shorter and less formal but equally binding path existed from Manned Spacecraft Center directly to the program offices. Direction received along either path was translated by the program office personnel into action items and routed to the proper agency for accomplishment. Contractual direction and configuration management were controlled by the SSD Program Office originally through the Atlas Weapons System Program Office and later through the SSD Standard Launch Vehicle III (SLV III) Office. Subsystem offices within SSD were responsive to the Mercury launch vehicle program office in the areas of guidance and propulsion systems. Technical direction was handled informally by direct contact between the Aerospace program office and the contractors and formally through the SSD program office.
The Atlas associate-contractor team consisted of General Dynamics/Astronautics (GD/A) who furnished the Atlas airframe and basic vehicle, Rocketdyne Division of North American Aviation (R/D) who furnished the propulsion system, General Electric (GE) who provided both the airborne and ground portions of the guidance system, and Burroughs Corporation who provided the A-1 Computer for inflight guidance in conjunction with the GE system. GD/A performed the launches at the Atlantic Missile Range (AMR) under the supervision of the 6555th Aerospace Test Wing and the other contractors provided appropriate launch services. Other valuable members of the Atlas team were the Air Force's Western and Eastern Contract Management Regions whose personnel insured the contractors; compliance with contract provisions and performed quality control and technical inspection functions.
Early in the Mercury program, Major General O. J. Ritland, as Commander of BMD recognized that a safety program should be instituted to protect the Mercury pilot. Accordingly, he directed that studies be conducted to determine what efforts were required to insure
 safe powered flight and to assure the program management that the launch vehicle was indeed ready for manned flight. This study resulted in the Pilot Safety Program for Mercury-Atlas launch vehicles (see fig. 5-3), a program which has dominated the management of the launch-vehicle portion of Project Mercury.
The basic objectives of the program have been to assure design reliability and adequate pilot safety. Recognizing that the Atlas had been designed as a weapons system and had not been required to meet the reliability expected of a manned system, program personnel established these objectives. The first was to be met through quality of production and end-product excellence. The quality of production would be assured through education and motivation of all personnel associated with manufacture of the hardware, through special component selection and marking procedures, and through special handling techniques. End-product excellence could be assured by requiring that no shortages would be tolerated at the time of launch-vehicle acceptance, and that the vehicle must be complete and up to date with no provisions for field modifications. This assurance would be gained by means of a detailed and highly critical factory roll-out inspection. The inspection would be conducted by experienced and well qualified personnel from both the Aerospace and SSD program offices. NASA observation was invited.
The second objective of assuring adequate safety would be met by providing reliability: augmentation and by special test-site operations. The abort sensing and implementation system (ASIS) was designed to bridge the gap between the existing reliability of the launch vehicle and the near perfection required of a manned system. The ASIS was an automatic system designed to sense an impending catastrophic failure and initiate spacecraft escape prior to the failure. The ASIS itself had to be an extremely reliable system. This reliability was obtained first through a design based upon redundant sensors and circuitry. Then rigid design reviews, stringent ground testing, and finally flight testing were conducted for the system.
The special test-site operations started with unique Mercury handling procedures for the launch vehicle and a requirement that complete documentation be maintained on all prelaunch operations. The documentation, in turn, led to assurance that the vehicle was indeed flight ready upon completion of the required pre-launch testing. The flight readiness was certified by the Mercury-Atlas Flight Safety Review Board. This board was established as a high-level Air Force and Aerospace board chaired for all manned flights by the Commander, SSD.
At the time of the original NASA order for Mercury-Atlas launch vehicles in the fall of 1958, the U.S. Air Force development flight test program was principally concerned with. the Atlas "C" model. The "D" model (see fig. 5-4) which was scheduled to begin testing in 1959, was considered the operational system and was therefore selected as the most suitable for use as the Mercury launch vehicle. The following paragraphs give a general description of the basic Atlas "D" vehicle from which the launch vehicle for the Mercury spacecraft was developed.
The Atlas launch vehicle comprises of two main sections, the body or sustainer section and the aft or booster-engine section. The booster- engine section is connected to the sustainer thrust ring by a mechanical system which permits separation. The Atlas is considered a 1 1/2-stage missile in that only the boost engines and  associated hardware are jettisoned at the completion of the first stage of firing.
The sustainer section is made up of a thin wall, fully monocoque structure pressure vessel and derives its rigidity from internal pressurization. The sustainer body is a welded structure of corrosion-resistant stainless-steel sheets varying in thickness from 0.048 inch to 0.015 inch. The tank is approximately 50 feet in length The forward end consists of a thin dome on which the liquid oxygen boil-off valve is mounted. The base of the dome is joined to the first skin of a conical section whose upper diameter is approximately 70 inches. The conical section joins a cylindrical section 10 feet in diameter. The lower end of the tank is conical, tapering to a point. A hemispherical diaphragm called the intermediate bulkhead divides the tank into a forward section for liquid oxygen and an aft section for RP-1 fuel. A thrust ring joins the conical aft section to the cylindrical portion of the tank. Annular baffles in the tanks serve to dampen propellant sloshing. The sustainer engine with its associated
equipment and subsystems is gimbal-mounted to the sustainer thrust cone which is the aft end of the fuel tank. Vernier engine thrust chambers are gimbal-mounted on opposite sides of the structure at the extreme aft end of the cylindrical portion of the tank. Equipment pods containing electronic and electrical units are attached to the tank skin 90° around the tank from the verniers.
The aft section or booster-engine section consists of two booster engines, structure, and associated equipment. It is attached to the thrust ring at the aft end of the tank section by a mechanism which releases it for separation. The motion of this section is controlled during separation by jettison tracks. A radiation shield protects the aft section from the heat radiated from the engine exhaust.
The propulsion system consists of a Rocketdyne MA-2 rocket-engine group made up of two main assemblies: the booster section (see fig. 5-5) consisting of two booster engines having 154,000 pounds of thrust each and the sustainer-vernier group (see fig. 5-6) consisting of
 one sustainer engine having 57,000 pounds of thrust and two vernier engines having 1,000 pounds of thrust each. All are single-start, fixed- thrust rocket engines utilizing liquid oxygen and a liquid hydrocarbon fuel (RP-1) as propellants.
The booster engine is composed of two identical thrust chambers and a power package. Two dual turbopumps in the power package deliver the propellants under high pressure to the thrust chamber. The turbopumps are driven by high-speed turbines, energized by highvelocity gas supplied by a single gas generator. The power package also includes the hydraulic pump used for lubrication of the turbopump gears. The booster gas generator consists of a spherical combustion chamber and an exhaust manifold. After start, liquid oxygen and fuel are supplied to the combustion chamber under pressures developed by the turbopump. The combustion gases are routed to the turbopump turbine wheels by the exhaust manifold after which the gases pass through the heat exchanger to heat and expand helium for vehicle-system pressurization and then are vented overboard. High-pressure propellants exiting from the turbopumps are routed through valves which control the flow of propellants to the fuel manifold and oxidizer dome. The two thrust chambers are bell shaped and made up of tubes running lengthwise from the top of the chamber to the bottom of the skirt. Fuel is routed through these tubes to cool the chamber walls. A pyrotechnic igniter initiates combustion of the fuel-oxidizer mixture. Thrust loads are transmitted to ,the missile through gimbal mounts on each chamber allowing the chambers to be swiveled a maximum of 5° in pitch and yaw about the vehicle centerline.
The sustainer engine is gimbal mounted to the thrust cone of the fuel tank. The assembly is similar to that of the booster engines. The sustainer engine dual turbopump supplies propellants to the vernier engine in addition to the sustainer engine. The sustainer engine fuel-lox mixture is continuously controlled during flight by the Propellant Utilization Subsystem (PIN) in order to maintain optimum mass ratio of the propellants and thus reduce unusable residuals to a minimum. The sustainer engine gimballing is controlled in pitch and yaw within an arc of ±3°. The sustainer engine is used for steering only after the booster engines have been shut down. The sustainer is operated throughout the flight and is at full thrust at lift- off.
The vernier engines are installed on the aft airframe in two separate units. Propellants for starting the vernier engines are provided by pressurized start tanks and are supplied by the sustainer turbopump for the remainder of the flight. The thrust chamber is double walled and also contains fuel for cooling of the thrust chamber walls. The vernier engines provide roll control throughout flight; pitch and yaw control during staging; and pitch, roll, and yaw during the vernier solo phase during flights in which this phase of operation is utilized. Mercury-Atlas vehicles do not have a vernier solo period. The chamber can be moved through an arc of approximately 140° in pitch and 50° in yaw.
The automatic start. sequence of the rocket engines is accomplished by initiating propellant flows into the thrust chambers, the firing of igniters, and the burning through of igniter detector links. These must be accomplished in the proper sequence and total time, or automatic shutdown of the engine will occur.
A propellant utilization system shown in figure 5-7 is used to effect emptying of the propellant tanks as nearly simultaneously as possible. This subsystem continuously senses the mass of the propellants remaining in the tanks and computes the error resulting from a comparison of the mass ratio of the remaining propellants with a nominal mixture ratio. This error signal then adjusts the rate of fuel flow by repositioning the sustainer-engine fuel-control valve to allow the burning of more or less fuel in order that the required mass ratio can be maintained. This assembly is made up of two manometers, each enclosing a mandrel coated with a dielectric material, and a computer-comparator. The unit senses the propellant messes by functioning as a variable capacitor by area contact with a column of mercury balanced against the liquid-propellant head in each tank. The mandrels are shaped in such a way that the capacitance is analogous to the mass of propellant remaining in each tank.
The airborne pneumatic system provides the structural rigidity for the main propellant tanks and also provides the necessary head to
 prevent the turbopumps from cavitating at low acceleration levels. This pneumatic system, presented schematically in figure 5-8, is used throughout the missile for control, reservoirs, lubricant tanks and the pressurization of the vernier engine propellant tanks. The pneumatic system also provides the actuation force for the first stage separation latches. The pressurization medium is helium, and liquid nitrogen is used to refrigerate the vehicle borne helium supply during the prelaunch phase of the countdown. Five spherical titanium storage vessels are used for the primary supply and are jettisioned with the booster section at staging. The control helium bottle is retained with the sustainer section and provides control pressure for the sustainer section. Tank pressurization is maintained by helium throughout booster-engine operation only. After first stage separation, no helium is required since oxidizer vaporization will keep the pressure in the oxidizer tank above the allowable minimum limits, and main fuel-tank pressure decay will not reduce this pressure beyond the minimum of allowable limits throughout the remainder of the flight. A liquid- oxygen tank boil-off valve is used to maintain proper cryogenic conditions of lox during tanking and holds.
The electrical subsystem (see fig. 5-9) is composed of a 28 v d-c main missile battery and a 115 v d-c three-phase 400 cps inverter. Battery power is provided to the inverter, propulsion subsystem, flight control subsystem, propellant utilization system and abort sensing and implementation system (ASIS). A power changeover switch is used to transfer both a-c and d-c
power from external to internal. The position of this switch is manually selected in the launch control blockhouse. The main battery is a remotely activated unit consisting of 20 silver zinc cells connected in series and housed in a sealed canister. The inverter is a rotary-type inverter using a magnetic amplifier voltage and frequency regulator and associated noise filters. The inverter is three phase-WYE connected.
The flight control subsystem consists of a flight programer, an autopilot, and 10 gimbaled thrust-chamber actuator assemblies. The subsystem stabilizes and steers the vehicle along the desired flight path by controlling the direction of the engine thrust vectors. Steering commands are generated on the onboard flight programer during the boost phase. Shortly after first-stage separation, the airborne portion of the guidance subsystem is enabled to provide steering commands to the autopilot for the remainder of the sustainer phase. The autopilot (see fig. 5-10) consists of a gyro package, a servo amplifier package, a programer, an excitation transformer, and engine-position feed-back transducers. On the standard Atlas "D", the main gyro package is located at station 991 and contains three rate gyros, three displacement gyros, and associated electronic equipment. The programer is a transistorized electrical timing device which controls the various flight sequential functions such as roll and pitch programs, staging filter changes, guidance enable, and so forth throughout the entire flight. The programer has two major sequences, the first of which is initiated at 2-inch motion of the missile and the second at receipt of the staging command from the ground-based portion of the guidance subsystem. The servo-amplifier package provides the integrating circuits and includes the necessary filters to insure proper flight attitudes and rates.
The guidance subsystem (see fig. 5-11) consists of the ground-based General Electric Mod III-A X-band radar system, the Burroughs A-1 computer system, and the airborne General Electric Mod III-A guidance group. The Mod III system consists of a position-tracking radar subsystem which determines the position vector of the missile with respect to the guidance station, plus a rate subsystem, which by Doppler techniques measures the missile velocity. In addition, the tracking radar serves as a data
link to provide operational commands to the missile-borne equipment. Position and rate data from the radar are transmitted to the Burroughs A- 1 computer for processing in accordance with the guidance equations. The computer generates corrective commands which are then fed back into the radar to be transmitted as steering signals to the launch vehicle.
Although weapon-system Atlas vehicles do not require telemetery transmission, research and development vehicles have such a requirement. Two telemetry subsystems were used on Mercury flights. The standard subsystem was used on flights through MA-4 (Atlas 88D). Subsequent flights utilized a lightweight telemetry subsystem (see fig. 5-12) which will be described in the next section.
Two additional systems are installed for the use of range safety personnel. The first is the range safety command system which receives, decodes, and activates the arming, engine shutdown, and destruct functions. The other system is the Azusa radio tracking system which monitors launch vehicle space position and velocity. The Azusa system data are sent to the Atlantic Missile Range IBM 7090 computer which continuously predicts the instantaneous impact point (IIP) of the launch vehicle.
The Atlas "D" vehicle had been chosen for the task of launching Mercury on the basis of its being the most reliable launch vehicle available with the requisite performance during the time period of the program. It was not possible to start at that point to design a "man
rated" vehicle to perform the Mercury flights without several years' delay to the program. Therefore, to capitalize on the reliability inherent in the basic design of the vehicle which had been demonstrated in Atlas development flight tests, a ground rule of the booster program was to make a minimum number of changes to the launch vehicle. Only those changes necessary to adapt the vehicle to the requirements of the Mercury mission or those required to improve the safety of the vehicle for manned flight would be authorized. As with any development program, flight-test experience established the need for incorporation of additional modifications with the major purpose being the enhancement of reliability and pilot safety. It should be recognized, however, that an extremely conservative approach was taken with regard to such changes. Modifications required extensive ground testing, and no critical modification to be used in a manned flight was incorporated until it had been successfully flown on at least one other Atlas. The following paragraphs describe the major system modifications incorporated in Mercury-Atlas launch vehicles. These changes are shown schematically in figure 5-13.
In the first category of changes required by the Mercury mission, one of the most important of the changes was the addition of a new autopilot
rate gyro package in a position considerably ahead of that used on the standard Atlas "D". This addition was dictated by the longer Mercury payload and its effect on the flexible Atlas tank during flight. The modification provided optimum attitude rate sensing with resulting minimum engine deflections for more efficient performance of the launch vehicle. The standard rate gyro installation was retained for abort system sensing.
Additional changes in this category include the deletion of the vernier solo phase of operation and relocation of the retrorockets from the launch vehicle to the spacecraft for use as posigrade rocket motors. In the vernier solo mode of operation the vernier engines remain in operation after sustainer engine cut-off, which allows very delicate adjustments to vehicle velocity. Deletion of this mode permitted n reduction in weight and mission complexity with a resultant improvement in performance and reliability. Relocation of the retrorockets was feasible since the Mercury spacecraft was lighter and the posigrade rockets would thus be more efficient in separating the spacecraft from the launch vehicle. The standard Atlas used these retrorockets to "back off" the launch vehicle from the payload. This relocation of the Atlas retrorockets to the spacecraft retropack required that the shill skin of the lox dome be protected from the rocket exhaust. This was accomplished by developing a fiberglass shield that attached to the mating ring and covered the entire dome. A wet-start technique was also incorporated in the engine starting sequence to minimize starting transients. Another change required for the Mercury mission affected the guidance system. Because the trajectory of the Mercury-Atlas flight differed greatly from that  of the weapon system vehicles, new guidance antennas were required to insure maximum signal strength throughout powered flight. Extensive theoretical and model work was required to develop antennas which would have suitable radiation patterns.
By far the most important change made to the Atlas in support of Project Mercury was the development and installation of an entirely new system, the Abort Sensing and Implementation System (ASIS). This system was designed to bridge the gap between the admittedly less than perfect reliability of the basic Atlas weapon system design and that near- perfect reliability desirable for a manned flight system. From a very searching and thorough analysis of Atlas flight data, it was seen that certain missile parameters deviated from a norm sufficiently ahead of catastrophic failure to be used as warnings. It was decided to develop an extremely reliable automatic system to monitor these parameters and to signal the spacecraft escape system when a catastrophe was imminent.
The parameters that were considered the most significant for abort indications (see fig. 5 -14)
were the liquid oxygen tank pressure, the differential pressure across the intermediate bulkhead, the missile attitude rates about all three axes, rocket-engine injector manifold pressures, sustainer hydraulic pressure, and the launch vehicle a-c power. Dual sensors for each of these parameters were incorporated into the Atlas system and operation outside a predetermined tolerance band then caused the ASIS to drop out the 28 volt power being supplied to the catastrophic failure detection relays. This drop-out of voltage provided an additional measure of safety in that if the abort sensing system failed in itself, the loss of power to the spacecraft would also cause an abort. This system was developed at GD/A under the direction of the Air Force and its systems engineering contractor and with the coordination of the NASA Manned Spacecraft Center group. This subsystem with its sensors was flown "piggy back" on Atlas research and development vehicles prior to the first Mercury-Atlas flight at which time it was flown in the open- loop configuration. The first closed-loop flight of this system was the MA-3 mission. The flight very successfully demonstrated the capability of the ASIS when the launch vehicle was destroyed by the range safety officer. The ASIS satisfactorily signaled an abort to the spacecraft in sufficient time to permit adequate separation of the spacecraft from the Atlas explosion.
To provide additional safety measures with the automatic abort, commanded abort, and range safety command destruct, a 3-second delay was incorporated between the signal that commanded engine shutdown and the signal that ignited the destruct package on the launch; vehicle. With this change, the launch vehicle could not be destroyed by command for a period of 3 seconds after the engines were shut down. This delay was incorporated to provide adequate separation of the spacecraft from the launch vehicle prior to a command destruct. To provide protection to the launch area, a lockout was incorporated from lift-off to 30 seconds that prevented an abort command from signaling engine shutdown. The spacecraft launch-escape motor had sufficient thrust to provide adequate separation from the Atlas during this period. Immediately after the failure of the MA-1 (Atlas 50D) mission, a special board was convened to investigate the cause of the failure. A number of separate phases of investigation were performed under the direction of the board. These included extensive analyses by Aerospace and GD/A of the thermal environment, discontinuity stresses, and aerodynamic loads. Wind-tunnel tests were performed to gain more knowledge of the aerodynamic conditions imposed on the total flight vehicle in the transonic and maximum dynamic-pressure regions. Analyses conducted by NASA Space Task Group personnel indicated the possibility of concentrated loads being introduced into the  Atlas through the forward structural ring which mated with the spacecraft adapter. None of the investigations or analyses were able to pinpoint the exact cause of the initial failure of the vehicle, but there was no question of the fact that the failure had occurred in the area of the forward lox tank and the spacecraft adapter.
Because of the failure of MA-1 in July 1960 and the successful flight from a structural standpoint of Big Joe I (1OD) in September 1959, a coordinated decision was made by BMD and NASA to increase the thickness of the four forward skins of the Atlas lox tank on future Mercury-Atlas launch vehicles to approximately the same dimensions as those on 10D. At the same time it was agreed that the spacecraft adapter would be stiffened. In order to fly the MA-2 mission with Atlas 67D, a thin-skinned vehicle, without undue delay a temporary modification was made. A stainless steel reinforcing band was installed about the lower flange of the mating structure (Station 502 ring) and the first skin aft.
Early in the Mercury program, it was decided to incorporate the electronic "square" autopilot in place of the electromechanical "round" autopilot. The reason for selecting the relatively new electronic system over the proven round autopilot was to obtain improved reliability, improved maintainability due to modular plug in packaging, much increased flexibility to allow for most types of mission changes, and ease of manufacturing by eliminating much of the hidden, point-to-point wiring, and the mechanical setup of- the programer. The improved reliability was a result of including such design features as electronic switching in place of mechanical switching, electronic integration in place of electromechanical integration, and improved circuit board design.
Initial flight testing in the Atlas program was accomplished by using an early type of telemetry system. The weight and power requirements to operate the early system were high, and oscillator stability degraded over a short operating time span. A transistorized, lightweight system was developed by GD/A to support the Centaur flight test programs and appeared to be well suited to the Mercury program (fig. 5-12). NASA requested the Air Force to incorporate the new lightweight system as soon as practicable. This system was first flown on launch vehicle 100D.
Normal cut-off of the sustainer and vernier engines is initiated by a discrete signal from the Burroughs computer to the ground guidance station. The ground guidance station then retransmits this signal to the airborne decoder which in turn signals engine shutdown. A partially redundant path for the sustainer-engine cut-off (SECO) discrete transmission was developed early in the program. This path enabled the Burroughs computer to forward the signal to the launch vehicle through the range safety command transmitter, to the airborne receiver and then to the engine relay control. This path was not wholly redundant because no duplication existed in the computer function for generating the SECO time; therefore, a single failure mode still remained. As a result, discussions with the AMR range personnel brought out the capability of the Azusa system to provide a completely redundant SECO discrete signal. The Azusa system in conjunction with the IP 7090 computer continuously computed the instantaneous launch-vehicle impact point (IIP) for Range Safety purposes. With certain modifications to the IP 7090 program it w as possible to obtain the time at which orbital velocity was attained. This time was provided electrically by land line to the NASA Flight Director. The Flight Director used this signal as a backup in the event of a failure or malfunction of the Mod III guidance system. This backup SECO system was susceptible to guidance noise; therefore, it was discontinued after the MA-E, mission.
The SECO discrete transmitted to the launch vehicle through the range safety command system as described above, was originally tied to the output of the guidance decoder which obtained a SECO discrete through the guidance system. Both SECO signals used the same path from the guidance decoder and the range safety command receiver to the engine shutdown relays. Additional engineering was required to reroute the signal to provide a completely redundant path.
It is pointed out later in the paper that a problem was discovered with the guidance system at low antenna elevation angles. After a thorough study of the hardware involved, it  was concluded that the excessive noise in received signals was cyclic in nature and was caused by an as yet undetermined atmospheric phenomenon. To reduce the effect of the noise in the over-all guidance loop, first the guidance equations were modified to provide additional smoothing, and second, the rate station base legs were increased from 2,000 to 6,000 feet. Although the latter modification did not reduce the actual noise being received. the deleterious effect of the noise on the received signals was reduced by approximately 3 to 1. The third and more complex phase of the study was the development of a mathematical model of the noise to permit a more detailed analysis of the trajectory equation changes that were necessary to minimize this effect. These changes were made to the guidance equations and used on the MA-9 mission.
A fuel tanking test that was being accomplished between the first and second launch attempts of the MA-6 mission brought out a problem that necessitated a major airframe change. The plastic foam material that is used for insulating the base of the liquid oxygen tank from the fuel tank is contained between two hemispherical bulkheads which separate the lox and fuel. A more detailed description of this problem is contained in a description of the MA-6 mission. The limited need for the insulation material coupled with the undesirable feature of removing the bulkhead in the field indicated the need for eliminating the insulation bulkhead from all future Mercury vehicles. A change in the production line stopped further installations of this material.
A major modification in the propulsion system was required to eliminate the possibility of combustion instability. Early in the Atlas program, it was found through flight test experience that combustion instability in the booster engines could cause catastrophic failure of the entire missile. The probability of the occurrence was low; however, the need for maximum safety in the manned space program dictated the need for corrective action. Initially, rough combustion monitors were incorporated and the Atlas was held down for an additional period of time, to allow sensing of the engine vibration characteristics. A rough combustion cut-off (RCC) system then would automatically shut down the engine if combustion instabilities occurred. Again, a thorough ground and flight test program was required before installation on Mercury-Atlas launch vehicles. Another modification provided redundancy in the electrical portion of the propulsion system to insure engine shutdown at SECO. Action was taken also to reroute electrical circuitry to insure proper valve sequencing during start in the high-pressure liquid oxygen plumbing.
Another major modification was made to the booster engine turbopumps. Flight and component testing experience show that incidents had occurred where the lox pump impeller had rubbed against the inlet adapter of the pump. This rubbing caused sufficient heat to ignite the lox and in some cases cause an explosion in the turbopump. Extensive analyses and tests could not pinpoint the exact cause for rubbing; however, the effect of the rubbing could be eliminated by lining the inlet adapter with a plastic material. Months of component and system testing and engineering review were required to provide positive assurance of the suitability of this modification.
Limited changes were made to the pneumatic system specifically for Mercury. Considerable effort was expended however on analyzing tank pressure oscillation that occurs during lift-off under certain payload conditions. The necessary precautions were taken-until this problem was resolved. To resolve the entire problem a complex computer model was developed to represent the dynamic conditions existing in the pneumatic system and structure of the Mercury-Atlas vehicle. It was found at the conclusion of the study that earlier characteristics of the helium regulator which controls pressurization gas to the oxidizer tank tended to drive the system into a resonant condition. The new regulator that was used with Mercury did not have the unstable characteristics; therefore, flight restrictions were removed.
The propellant utilization (PU) system was modified to insure an outage of lox rather than fuel in the event abnormal flight characteristics caused the vehicle to expend the total propellant. Early studies had indicated that a safer engine shutdown would be possible in this propellant depletion shutdown case if the lox supply was the first to be consumed. The PU system normally monitors the propellant levels to maintain the proper ratio of onboard  propellants For the Mercury-Atlas the system was modified to drive the mixture ratio to the lox-rich condition at 10 seconds prior to SECO to reduce the ratio of lox to fuel. More recently, a revised method of calibration and a slightly modified mandrel have been developed to provide a more accurate method of maintaining proper propellant ratios.
A normal phenomenon associated with the Atlas vehicle is a roll oscillation that occurs with the missile as the vehicle becomes free of the launcher mechanism. Ordinarily this roll is of small magnitude, and quickly corrected as the autopilot is enabled. A review of flight test history showed that certain vehicles were displaced at roll rates which approached the abort threshold established for the ASIS in roll. Two parallel studies were accomplished to review this problem area. One study reevaluated the abort thresholds to determine if the roll rate limit could be increased. The other study attempted to determine the cause for the roll oscillation in order that a proper modification could be made. It was determined that limited opening of the threshold in roll could be accomplished. The study into the cause for the roll included developing a mathematical model of the launcher mechanism, analysis of control forces required to rotate the missile similar to that demonstrated in flight, base recirculation, engine alinement, and a review of engine acceptance data at Rocketdyne. It was readily apparent that the canted turbine exhaust duct contributed to the clockwise roll moment. This force could cause only half of the roll moment experienced by the missile. Acceptance data from the engine supplier showed that a group of 81 engines had an average roll moment in the same direction of approximately the same magnitude as that experienced in flight. Although the acceptance test-stand and flight-experience data on individual engines did not correlate, it was determined that offsetting the alinement of the booster engines could counteract this roll moment and minimize the roll tendency at liftoff. This change was flight tested and found to correct the roll moment satisfactorily; therefore, the change was incorporated for MA-9 in Atlas 130D.
The first Mercury-Atlas launch was that of Big Joe 1 Atlas, number 10D, on September 9 1959. Atlas 10D was built originally as an R and D vehicle but had received the initial Mercury modifications. The payload was a boilerplate spacecraft. The purposes of the flight were to test the spacecraft's ablative heat shield, afterbody heating, reentry dynamics attitude control and recovery capability.
Two flight readiness firings (FRF) were performed on Big Joe 1. The first, on September 1, 1959, ended immediately after T-0 because the ignition stage delay timer commanded shutdown of the rocket engines when neither sustainer nor main engine ignition followed normal vernier ignition. There was no booster or stand damage. The second FRF was successfully completed on September 3,1959, with normal ignition, transition to main stage and shutdown by the engine timer after approximately 19 seconds of running time.
During the launch on September 9, 1959, engine ignition, thrust buildup and lift-off were normal, and launch vehicle performance was completely satisfactory throughout the booster phase. However, after booster engine cut-off (BECO) the booster section failed to jettison and remained attached to the vehicle for the duration of the flight. The sustainer continued to power the vehicle until propellant depletion some 14 seconds prior to normal cut-off. The malfunction resulted in the vehicle failing to achieve planned maximum velocity and in exceeding planned maximum altitude.
Although the injection conditions were considerably different from the preplanned values, the spacecraft reentry satisfied the NASA test objectives. By extrapolating the acquired data, NASA Space Task Group was able to derive the information which was required for spacecraft design. The spacecraft was recovered and returned to Cape Canaveral. Since the data from Big Joe 1 satisfied NASA requirements, a second Mercury launch, Big Joe 2 (Atlas 20D), which had been scheduled for the fall of 1959, was cancelled and the launch vehicle was transferred to another program.
 The first of the Mercury-Atlas series, MA-1, was launched at 8: 13 a.m. e.s.t. on July 29, 1960, from AMR Launch Complex 14. The vehicle consisted of Atlas 50D and Mercury Spacecraft number 4, the first production spacecraft, and adapter. The spacecraft primary test objectives concerned structural integrity, afterbody heating and reentry dynamics from a temperature critical abort. Launch vehicle objectives concerned the capability to release the spacecraft at the desired insertion conditions and the evaluation of the open-loop operation of the Abort Sensing and Implementation System (ASIS). A single successful FRF was accomplished on July 21,1960.
Lift-off and flight of the vehicle were nominal until 57.6 seconds after lift-off when a shock was registered by both the launch vehicle and spacecraft axial accelerometers. The vehicle at that time was at approximately an altitude of 30,000 feet and 11,000 feet down range. The sequence of sensing of the shock indicated that the disturbances occurred in the area of the adapter and the forward portion of the lox tank. All Atlas telemetry was lost at 59 seconds, which is believed to be the time of final missile destruction. Spacecraft telemetry however, continued until 202 seconds, which was the time of landing on the sea, approximately B miles downrange. The only launch vehicle primary test objective accomplished was successful evaluation of the open-loop performance of the ASIS which generated an abort signal at 57.6 seconds due to loss of normal a-c voltage.
The failure investigation and results are discussed in the section Atlas Modification for Mercury in this paper.
The MA-2 mission was flown by using the Atlas 67D and a production Mercury spacecraft. Test objectives for this flight were concerned with the ability of the spacecraft to withstand reentry under the temperature- critical abort conditions and with the capability of the Atlas to meet the proper injection conditions. This Atlas "D" modified for the Mercury mission, was unique in the program in that it incorporated a stainless steel reinforcing band installed around the vehicle between stations 502 and 510. A thin sheet of asbestos was in stalled between the reinforcing band and the tank skin. This modification was installed as a precaution against the type of failure which had occurred on the previous MA-1 flight. Atlas 67D had accomplished a successful Flight Readiness Firing on November 19, 1960.
Launch countdown was satisfactory. Although 70 minutes of hold and recycle time were required, none of this time was required for the launch vehicle systems. Lift-off occurred at 9:10 a.m. e.s.t. on February 21, 1961. Ignition and transition to main stage were normal, and lift-off was clean. The launch-vehicle flight was uneventful. All test objectives were fully met, and the spacecraft was successfully recovered. This launch was the first one which was preceded by a full Flight Safety Review Board in accordance with the Mercury-Atlas Booster Pilot Safety Program.
Atlas 100D, the launch vehicle for the MA-3 mission, was launched from Complex 14 at AMR at 11:15 a.m. e.s.t. on April 25, 1961. The mission was terminated by the range safety officer after approximately 43.3 seconds due to failure of the launch vehicle to follow its roll and pitch programs. Although the launch-vehicle was destroyed as a result of a malfunction, considerable benefit was derived from the flight test. First, the satisfactory closed-loop performance of the ASIS was demonstrated when the booster engines were shutdown and escape rocket ignition was initiated automatically by the ASIS. The escape was so successful that the spacecraft was recovered some 20 minutes after launch and reused on the next flight.
Second, because of the nature of the failure an intensive reexamination of the complete electrical circuitry and its design, manufacture and installation for both the launch complex and the Atlas was conducted. The maIfunction which caused flight termination was isolated to the flight programer or associated circuitry. The programer either failed to start or started and then subsequently stopped without initiating the roll and pitch program. The programer was subsequently recovered, examined, and tested. The most probable cause of the flight failure was traced to contamination of one of the programer pins which under vibration could have caused the failure. The extensive review  that was conducted to analyze the flight failure also revealed other deficiencies in the flight control systems. Changes were made to the system to eliminate these possible failure modes and to improve the overall system reliability.
On August 24, 1961, the Flight Safety Review Board for the MA-4 mission (Atlas 88D) performed a thorough review of all pertinent problem areas and all recent Atlas flight test problems. At the completion of the meeting, the Flight Safety Review Board approved the use of Launch Vehicle 88D for the MA-4 mission. The launch was delayed for a 1-week period, and during this period of time a transistor malfunction in one of the flight control canisters aroused considerable concern. An investigation into the factors associated with this failure necessitated an Air Force Program Office decision to delay the flight in order that flight control equipment could be reworked to eliminate this failure mode. The contractor responded to this decision with a concentrated effort to rework and test the equipment in time to support a mid-September launch. On September 12, 1961, the Flight Safety Review Board reconvened. The flight control canister rework was reviewed in detail and the Board concluded that 88D was suitable for launch. The 88D was scheduled for a 250-minute countdown starting at 2:50 a.m. e.s.t. on September 13, 1961. There were four holds and a recycle which resulted in a total count of 374 minutes. Propulsion system performance was normal throughout the start sequence, additional hold-down period and flight. Thrust chamber vibration levels were normal during the hold-down period and chamber pressures were nominal. Lift-off occurred at 9 :04 a.m. e.s.t. The flight control systems satisfactorily generated the missile roll and pitchover programs and responded correctly to guidance discrete and steering commands. An oscillation in the pitch plane was evident from T+15 seconds to T+21 seconds. Missile bending was evidenced by an accelerometer located on the lox-dome, launch-vehicle flight control rate gyros, and by spacecraft rate gyros. A change to a launch-vehicle automatic hydraulic actuator had been incorporated on the MA-4 launch vehicle and the flight control gains had been modified. A postflight modal analysis of the MA-4 data showed that marginal stability characteristics existed with these changes; therefore, additional filtering was deemed to be necessary for future Mercury flights. Propellant slosh amplitudes during the booster phase were low and considerably less than that observed on launch vehicle 67D. The spacecraft injection conditions on the flight of 88D were of the poorest quality of all Mercury-Atlas flights. Tolerance limits were not exceeded; however, a thorough study was required to determine the cause. An analysis of the flight data brought to light tracking phenomena associated with low incident angles. Under certain conditions the guidance system could be affected by varying atmospheric refraction towards the end of flight when the vehicle was approaching the horizon. Limited experience had been obtained at these low elevation angles with the Mod III guidance system. A continuing study was conducted by SSD, GE, Aerospace Corporation, and Space Technology Laboratories in conjunction with the AF Electronic System Division and its technical stag to determine the source and limitations of this phenomenon knowledge gained from this study was later used to rewrite the trajectory equations to reduce the effects of refraction anomalies. The postflight evaluation of the launch vehicle 88D mission indicated that all flight objectives were successfully achieved.
On November 28, 1961, the Flight Safety Review Board met to consider all aspects of the MA-5 (93-D) mission. Included in the Board review were the autopilot changes that resulted from the previous flight and a thorough discussion of the activities and studies conducted in the evaluation of the guidance phenomena. Additional problems associated with other Atlas space and weapons flight test were reviewed. The Board committed the vehicle to launch.
A number of holds were required during the countdown on November 29, 1963. The data link between the GE ground guidance station and the Mercury Control Center dropped out temporarily, requiring a 4-minute hold, and a 3-minute hold was called at T-7 minutes to resolve a pulse beacon anomaly. Ignition and transition into mainstage were accomplished satisfactorily and within expected limits. There was no indication of the pitch oscillation observed on the launch of 88D. Following  lift-off a slight oscillation was noted in the pitch channel during the roll program which is common to all launches. The usual flight oscillation clue to slosh was observed from T+86 seconds to T + 100 seconds. Staging transients were normal. Approximately 30 seconds before sustainer engine cut-off, a slight oscillation appeared in the pitch channel. This condition persisted for 15 seconds, but the magnitude of the oscillation was of no significance. All flight test objectives were met and the performance of the launch vehicle was within expected tolerance limits.
The historic flight of Astronaut John Glenn was conducted on board Atlas launch vehicle 109D and Mercury Spacecraft number 13. This was the flight for which the Atlas Pilot Safety Program had been conceived and for which the launch vehicle team had been preparing so long. Major General O. J. Ritland, then Commander, SSD, convened the Fight Safety Review Board on January 26,1962, to determine the suitability of Atlas 109D for support of the MA-6 mission. In addition to reviewing the readiness of 109D, the Board reassessed the critical problem areas in the development of the Atlas in support of the Mercury program. This reassessment included all major developments, flight-test incidents and corrective action, the results of additional reliability tests and analyses conducted specifically for Mercury, the performance and test status of the abort system, performance margins experienced on past flights and the prediction for MA-6, the configuration differences between the previous Mercury vehicle and 109D, and the production and test history of 109D prior to its arrival at AMR. One minor, last-minute problem with a faulty pin connection in the staging umbilical necessitated a second session of the board on January 26, 1962. The condition was repaired, and a complete series of tests to validate all the pin connections in the connector was satisfactorily accomplished. After the second session the Board committed 109D for the launch of MA-6. Adverse weather in the launch area forced the cancellation of the first launch attempt on January 27, 1962. After a tanking test was conducted on January 30, fuel was detected in the insulation bulkhead located between the fuel and liquid oxygen tanks. The insulation bulkhead is located beneath the intermediate bulkhead that structurally separates the two tanks and is composed of a plastic foam material vented to the fuel tank and supported by a thin steel membrane. Test of the plastic material indicated that sufficient fuel could be retained in the insulation material to overload the membrane sup porting the insulation bulkhead under flight accelerations. Inasmuch as it was not possible to assess the amount of saturation accurately, a decision was made to remove the insulation material and the supporting structure. The extent of the repair on Atlas 109D at AMR constituted a major but necessary rework of the vehicle in the field. Because of the extent of the repair a highly qualified group of personnel from Aerospace, 6555 ATW and GD/A were selected as a special committee to review all procedures associated with the task. This group also was responsible for validating of the complete task. The primary reason the task was authorized as a field modification was because it had been successfully performed in the field only weeks before on Atlas 121D, the Ranger 3 launch vehicle which flew successfully.
The combined Atlas-Mercury countdown was begun at 11:30 p.m. e.s.t. on February 19, 1962. A built-in hold of 90 minutes was scheduled to begin at T-120 minutes. At T- 280 minutes, a telemetry check indicated the Azusa impact predictor was "no-go." The ground station was checked and found to be operating satisfactorily. The tower decks around the transponder were raised, but still the Azusa system could not achieve a satisfactory lock. A decision was made to change the transponder which was accomplished by T- 273 minutes. The test was resumed and Azusa was declared "go" at T-213 minutes. No hold time was involved. At T-149 minutes, during the flight control system test, there was a sudden drop in the rate beacon automatic gain control (AGC). The first backup beacon was substituted for the original unit during the built-in hold. This hold was extended for 30 minutes and then extended another 15 minutes to complete installation and retesting. Ten additional minutes of hold were required for the spacecraft. At T-60 minutes a 30-minute hold was requested by Mercury Control Center which was then extended an additional 5 minutes. At T-45 minutes a 15-minute delay was  instituted to catch up with the countdown procedures. Lox tanking began at 8:30 a.m. e.s.t. Lox pump problems caused a 25-minute delay in the count. A 2-minute hold at T-6.5 minutes was requested by Mercury Control. The count then proceeded normally to T-0. Lift-off of 109D and Astronaut Glenn occurred at 9:47 a.m. e.s.t. Propulsion system operation during ignition was satisfactory. The longitudinal oscillation normally expected at lift-off were nominal and damped out by approximately 25 seconds after lift-off. Performance of the guidance system was satisfactory. The missile was acquired by radar at the normal time, and tracking was maintained continuously throughout SECO. Steering began at 155 seconds with 60-percent pitchup and 23-percent yaw right commands of 10 and 5 seconds duration, respectively. These initial commands were acceptable for the planned trajectory. Thereafter pitch steering did not exceed 10 percent and yaw steering 5 percent until the end of the flight. Flight control system performance was satisfactory. All monitored programer pitch functions occurred at the proper time. Staging sequence was normal and no evidence of pitch oscillation buildup occurred during the flight. Insertion accuracies were good and well within the tolerance requirements established by NASA. Postflight evaluation of the mission indicated that all systems functioned satisfactorily, and no significant anomalies were apparent.
Atlas 107D was shipped to AMR on March 7, 1962, to support the MA-7 flight of Astronaut Carpenter in spacecraft number 18. The vehicle was erected on March 14, 1962, and no serious problems were found during the prelaunch activity. A joint spacecraft and launch vehicle flight- acceptance composite test (FACT) was conducted on May 4, 1962. The Flight Safety Review Board met on May 23 under the chairmanship of Lt. General Estes, then Commander of SSD, for the purpose of determining the readiness of 107D to support the second Mercury manned orbital launch. The combined Atlas-Mercury countdown began at T-390 minutes at 11 :00 p.m. e.s.t. May 23, 1962. The count proceeded very smoothly and without delay until T-11 minutes when the NASA flight director called a 15-minute hold because of unfavorable ground visual conditions. An additional 15-minute hold for the same reason was requested. At 7:17 a.m. e.s.t. an additional 10-minute hold was requested to analyze airborne refractometer test data to determine its effect on the ground guidance system. At 6:28 a.m. e.s.t. an additional 5-minute hold was called to complete the analysis of the refractometer data. Countdown was resumed at 7:34 a.m. e.s.t. and proceeded normally to T-0.
The Atlas vernier-sustainer and booster ignition and transition to mainstage were normal. Lift-off transients were very small and the normal pitch oscillation seen during the roll program was of minimum magnitude. Guidance lock-on was normal. No yaw command was necessary at the time of guidance enable. A slight pitchup was commanded, after which no steering commands were required until just before SECO. Staging transients were very small. An anomaly occurred in the sustainer hydraulic system when at T+192 seconds telemetry data showed that the sustainer engine control hydraulic pressure had begun to drop. The number two ASIS pressure switch activated at T+265.1 seconds when system pressure dropped below the abort level. The number one ASIS switch, which is on a separate sensing line, did not activate and therefore no abort signal was generated. Other telemetry measurements did not show corresponding hydraulic pressure drop. Test simulations conducted after the flight duplicated flight test indications when the sense line was cold soaked at liquid oxygen temperatures. Action was taken to modify future Mercury vehicles by insulating the sense lines. Guidance accuracies for the flight were improved as a result of the extension of the ground based rate system base legs. This was the first Mercury flight to incorporate this modification.
Atlas launch vehicle 113D scheduled to support the MA-8 mission on October 3, 1962, incorporated the baffled injector modification in the two booster engines. Sufficient ground and flight test experience had been conducted to provide adequate assurance of the additional flight safety possible with this modification. However, recent ground and flight test failures of  the sustainer turbopump created a new atmosphere of concern in the engine area. Investigation of these failures did not reveal any specific cause. Therefore, additional testing was required to determine the susceptibility of 113D to a similar malfunction. An extensive analysis of these past failures did point out that two conditions were common to the failures. The first condition was that the failure occurred during the period of time the fuel control valve was moving into the control position during start. Secondly, the malfunction had always occurred during the initial test of the system in that configuration. For these reasons it was determined that conducting an FRF on 113D in its launch configuration should expose the turbopump to this failure mechanism. Accordingly, an FRF was conducted on September 8. Post FRF evaluation indicated that the propulsion system was flight ready.
Major General Ben I. Funk, Commander, SSD, conducted the Flight Safety Review for the MA-8 mission at 9:30 a.m. e.s.t. on October 2, 1962, to determine the flight readiness of Atlas 113D. NASA concurred with the board's recommendation that the vehicle was in suitable condition to support the MA-8 mission.
MA-8 (Atlas 113D) was launched at AMR Complex 14, 7:15 a.m. e.s.t. on October 3, 1962. The performance of the propulsion system was satisfactory. Telemetered values of all measurements were indicative of normal system operation. Because of the incorporation of the production baffled thrust chamber injectors on the booster engines the missile hold- down time was not extended; and the rough combustion cut-off system was installed open loop on the booster and the sustainer engine for instrumentation purposes only. Flight control data indicated the usual clockwise roll transient at lift-off; however, in this case the transient condition approached 80 percent of the abort threshold. Longitudinal oscillations and pitch oscillations during the initial portion of the flight were nominal and slosh amplitudes were within expected values. All monitored programer switch functions occurred at the proper times and staging sequence was normal. A low amplitude roll limit cycle was apparent from approximately 252 seconds to SECO. Performance of the guidance system was satisfactory with negligible steering commands required after responding to the initial inputs. Insertion conditions were very close to nominal.
Atlas 130D was the sixth consecutive launch vehicle to place a Mercury spacecraft into earth orbit. It was the tenth and final launch vehicle used in the Mercury-Atlas program. 130D was accepted at the General Dynamics/Astronautics plant at San Diego, California, on March 15, 1962. Acceptance of this vehicle marked the attainment of a long standing goal of the SSD-Aerospace launch vehicle program offices: acceptance of a Mercury-Atlas launch vehicle without discrepancies or contractual deviations.
The Flight Safety Review Board convened on May 13, 1963, with Major General Ben I. Funk, Commander, SSD, as chairman, to review the status of Atlas 130D to support the MA-9 mission. The MA-8 launch-vehicle performance and the MA-9 launch-vehicle predicted performance were reviewed. All differences between the MA-8 and MA-9 vehicles were discussed, as well as the flight qualification of these changes. The history of manufacturing and testing of 130D at the manufacturer's plant and the prelaunch history at AMR were reviewed. Atlas flight-test experiences were updated to insure that no related problems existed and the board agreed that 130D was ready for flight. An initial launch attempt was made on May 14, 1963; however, the diesel engine used for retracting and stowing the gantry caused a delay in the count when it malfunctioned. Subsequently, the launch was postponed until the following day because of a malfunction in the radar at Bermuda.
The Atlas prelaunch operation, which began on time at midnight of May 14, 1963, was scheduled for a 390-minute countdown plus one planned hold of 90 minutes duration at T-140. There was one unscheduled hold of 4 minutes duration at T-11 minutes 30 seconds, to investigate a signal fluctuation in the Mod III ground guidance system. The anomaly was attributed to an outside source of radiation, and the countdown was resumed. The whole launch vehicle countdown had been exceptionally smooth, and no further delays were encountered. Ignition, transition to mainstage and lift-off were normal with no additional  hold-down beyond the normal approximately 2 seconds between flight lock-in and release. Lift-off occurred at 8 :04:13 a.m. e.s.t., on May 15, 1963. As the vehicle came off the launcher arms it rolled counterclockwise approximately 0.3° before this minor transient was corrected by autopilot control initiation at 40" motion. The expected slight longitudinal oscillation associated with lift-off occurred during the first few seconds of missile motion and damped normally. At two seconds after lift-off the roll program was enabled and 130D rolled toward its climbout heading of 72°. The roll program was completed at 15 seconds, and the booster pitch program w as enabled. Slight lofting took place during the early portion of the booster powered flight; however, the vehicle intercepted the planned trajectory at 125 seconds. Propellant sloshing became noticeable at 55 seconds, reaching a maximum amplitude at 98 seconds and decaying to a negligible value by 120 seconds. Propellant slosh during this period of time is normal, but the amplitudes on this flight were higher than on most previous Mercury launches. Postflight review of the 130D flight control gains indicated there were within tolerance but below nominal. Higher than normal propellant slosh amplitudes could be expected under these conditions.
Booster engine cut-off (BECO) was accomplished at 132.5 seconds with booster section staging at 135.4 seconds. Space position at BECO was very close to planned. At BECO the sustainer engine was nulled in pitch and yaw to assure proper clearance of the booster section during the jettison phase. After booster jettison the sustainer was reactivated in pitch and yaw. The sustainer-stage pitch program was initiated at staging plus 5 seconds and was completed at 159 seconds after lift-off. Entrance into the guidance steering mode was relatively smooth with the initial steering response being slightly up and to the right. After the initial correction, only extremely small steering commands were transmitted. SECO occurred at 303.03 seconds, approximately 1 second earlier than planned. Burnout conditions of the launch vehicle were very close to those planned and were within a few feet per second high in velocity, 500 feet low in altitude, and 0.005° low in flight path angle.
A detailed analysis of flight test data has shown that the launch vehicle performance was very close to nominal. An over-all vehicle postflight trajectory simulation did indicate that the effective specific impulse of the total launch vehicle system was within, but on the high side, of the tolerance band.
The pneumatic system operated satisfactorily, and no anomalies were noted. The tank pressure oscillation which normally occurs at lift-off was of very low magnitude and of no significance to the flight. Adequate pressures were maintained in both lox and fuel tanks and well above the abort limits at all times.
The propellant utilization system exhibited very smooth characteristics throughout the flight and was holding at the nominal position during the period prior to sustainer engine cutoff, indicating that the propellant mass ratio was correct. The PU system of this flight utilized a slightly reshaped mandrel and improved calibration techniques compared to previous Mercury flights.
The sustainer and booster engine hydraulic systems behaved in a normal manner with only slight booster position response to auto-pilot system demands occurring during the propellant sloshing period.
The a-c power supply frequency and the main battery voltage were within specified limits through powered flight. The a-c voltage ran 0.4 to 0.7 volt above the nominal but within the tolerance band. Slight vehicle lofting occurred as a result of this minor shift in a-c voltage.
The flight control system functioned satisfactorily and properly stabilized the launch vehicle. All guidance discrete and steering command functions of the flight control system were properly carried out. GE and Azusa data indicated that the total magnitudes of the booster phase roll and pitch programs were extended slightly beyond nominal but were still well within allowable limits. The major contributor to these excesses was the higher than normal inverter voltage output during the launch to BECO phase of powered flight. It should be noted that the effect of higher than nominal engine performance during boost phase tended to counteract the effect of higher than nominal  inverter voltage on the pitch program. As previously pointed out, the propellant slosh was greater than that on most previous flights but its effect on attitude rates was negligible. A low-amplitude roll limit cycle was evident from BECO to SECO. This motion had been noted on previous Mercury-Atlas flights and was not considered detrimental to the mission.
All instrumentation measurements functioned properly throughout the flight, and the telemetry quality was such that a very thorough analysis of all flight parameters was possible.
The range safety command system was not required until the auxiliary sustainer cut-off signal (ASCO) was transmitted 0.04 second after the BECO guidance discrete signal in accordance with the computer program logic.
Performance of the ASIS was satisfactory. Review of launch-vehicle data did not reveal the existence of any undetected abort condition. Switching functions to change abort logic and parameter levels were accomplished i the planned manner from launch throughout powered flight to SECO.