DESTINATION MOON: A History of the Lunar Orbiter Program
Early Design, Fabrication, and Testing Problems
[121] One of the first hardware items to cause Langley and Boeing concern was the velocity control engine. The Boeing Company had proposed using the same Marquardt 100-pound-thrust rocket motor that the Apollo Program was using in the attitude control system of the Command Module. Lunar Orbiter would use this rocket for velocity control. During preliminary testing for Apollo requirements, the Marquardt rocket developed problems which caused Lunar Orbiter Program officials to have second thoughts about it. On April 21, 1964, Captain Scherer, with members of his staff and representatives of the Project Office at Langley, visited Marquardt to determine the seriousness of the problems and their implications for Lunar Orbiter.
His group learned that the Apollo mission requirements called for the rocket to be used in a pulse mode. It would have to fire reliably in short pulses thousands of times during an Apollo mission in order to change the Command Module's attitude as desired. Testing showed [122] that the rocket was not firing correctly in the pulse mode. This, however, did not affect its use in Lunar Orbiter, because as the spacecraft's velocity control engine it would be fired only at specific times in a single-burn mode.14 Despite this difference in use Scherer recommended that until the Marquardt rocket proved reliable for Apollo such alternatives as the JPL Surveyor vernier engine should be studied.15
The Marquardt rocket was not so critical to the program's mission as another piece of hardware: the photographic subsystem's velocity-over-height sensor (V/H sensor). It could not be replaced easily by another component of a different kind, and its function was critical to the performance of the photographic subsystem. An image tracker which scanned a portion of the image formed by the 610 mm lens, it compared outputs derived from successive circular scans to measure the rate and direction of image motion before taking a photograph.16
The limitations of the V/H sensor determined in part the parameters of any photographic mission. It had to determine precisely the image-motion compensation values [123] for photography below 950-kilometer altitude, where the spacecraft's velocity relative to the Moon's surface would affect the ground resolution of all photography. Above 950 kilometers the image-motion compensation could be deleted without significantly affecting ground resolution. At that high or higher altitudes the ground resolution of the high-resolution pictures might be reduced from 20 to 3 meters, but the case would be altogether different in an elliptical orbit which brought Lunar Orbiter as low as 46 kilometers above the Moon's surface. At this low altitude the camera would have to compensate for image motion to avoid "smearing" in a photographic exposure.17
Kosofsky and Broome have detailed why the V/H sensor is vital to low-altitude photography:
The performance required of the image motion compensation apparatus is particularly exacting in the case of the Lunar Orbiter's high-resolution camera, as can be seen from the following figures. The design exposure speed is 1/25 see, because of the very low exposure index of the film used (Kodak SO-243 film, with exposure index about 3). The spacecraft's orbital velocity at the low point of the orbit is around 1.6 km/sec, so that it moves 64 m across the target area during an exposure. In order to achieve 1-m ground resolution, the uncompensated image motion must be no more than the scale equivalent of 0.6 m. The allowable error in image motion compensation is thus 1%, which must be allocated between the mechanical limitations of the [124] platen servomechanism and the errors in the information supplied to it by the velocity/height (V/H) sensor.18

Eastman Kodak held total responsibility for producing the photographic subsystem for Boeing. However, it subcontracted work for certain components of the subsystem to Bolsey Associates. One of these components was the V/H sensor. Although both Eastman Kodak and Bolsey had very qualified men to design and build the components, management of their operations did not always run smoothly and adhere to schedules, as will be discussed later.

Two other problem areas became evident by September 1964 when Boeing commenced tests on the thermal model of Lunar Orbiter. The first was an overload on the power system because of increased need for electricity during periods when the spacecraft could not use its solar panels. The Inertial Reference Unit placed the greatest demand on the power system, and tests revealed that a battery with a greater capacity was probably needed to meet the demand. Boeing and Langley engineers also examined the possibility of changing the orbit design to give the spacecraft a longer period of sunlight instead of having to go to a heavier battery.
Review of the power system difficulties and subsequent [125] findings showed that under the planned night flying conditions the Orbiter's 12-ampere-hour battery would require an excessive charging rate, approximately 4.5 amperes, to meet the power needs of the other spacecraft subsystems. This high rate could cause battery failure, and Boeing engineers had worked out three possible solutions: 1) Install a heavier, higher capacity battery, 2) turn off some equipment during the night periods, and 3) increase the time of the spacecraft's exposure to the Sun by altering the orbital parameters to be approximately 1,850 kilometers at apolune and 46 kilometers at perilune. The third solution would affect the spacecraft's photographic capabilities because the increased period of orbit would necessitate a decrease in the spacecraft's orbital inclination to the Moon's equator.19
During the Lunar Orbiter Program's First Quarterly Review at the Langley Research Center Scherer pointed out that, "if the initial orbit [of Lunar Orbiter] is made elliptical with a higher apolune, the day to night ratio would be improved and could be used to solve the problem."20 Langley and Boeing adopted the third solution after Thomas Yamauchi, head of Boeing LOPO's System Engineering Section, [126] had worked out the rationale for the orbit change. The change did not greatly affect photography and eliminated the need for a heavier battery.
The second problem concerned the spacecraft's fuel and oxidizer tanks which Boeing was purchasing from the Bell Aero Systems Company. Off -the-shelf hardware developed for the Apollo Program, the tanks had failed to pass qualification tests because of repeated rupturing of their teflon bladders. These bladders held nitrogen gas under pressure, and it was apparently seeping through the thin-walled bladders and saturating the fuel for the velocity control engine.21 The Lunar Orbiter Program required extra qualification tests of the tanks, but this threatened to triple their cost. Langley requested the Office of Advanced Research and Technology to review the problem of the tanks while it looked into possible alternative solutions.22
On August 26, 1964, the Langley Research Center held the First Quarterly Review of the program to discuss all known problems which had come to light since the Boeing contract had been signed. Boeing representatives summarized their operations for Langley and Headquarters officials on [127] the first day of the review and then devoted the second day to detailed presentations on specific areas of the program to NASA personnel working directly in each area.
The Lunar Orbiter Program Office rated Boeing's total performance as very good, but noted that Boeing had treated its relationship with the Eastman Kodak and RCA subcontractors superficially. No representatives from EK of RCA were present at the Langley review, and officials of the Lunar Orbiter Program felt that a Boeing-Eastman Kodak-RCA team presentation at subsequent reviews would be very desirable.23 Boeing, of course, was still in the process of signing contracts with these two firms.
During the review NASA and Boeing people treated the technical problem areas very thoroughly and discussed other difficulties related to spacecraft design and engineering. Boeing showed three more areas where work was required to attain the maximum functional efficiency in the spacecraft's configuration. The first was the spacecraft weight, a factor limited by the lifting capability of the launch vehicle. Boeing was aiming for a 370-kilogram spacecraft after separation from the Agena and before any midcourse maneuver. The preliminary Lunar Orbiter design had indicated a 390-kilogram spacecraft, but two major steps had [128] successfully reduced this figure. First, Boeing had decided to use integrated logic circuits in the control assembly electronics, since this would save some 6 kilograms over the use of discrete parts and perform just as well. Second, the need to use one-pound thrusters in the attitude control subsystem to compensate for thrust vector misalignment was eliminated when Boeing engineers redesigned the system.
Originally the attitude control thrusters had been located on the solar panels to take advantage of the greatest moment. However, a close reexamination of this design convinced Boeing and Langley engineers that controlling the thrust vector through the spacecraft's center of mass would be substantially more difficult with one-pound thrusters located far out on the solar panels. Attitude changes could be executed easily, but they would cause perturbations in the spacecraft's thrust vector which would have to be counteracted if the spacecraft were not to assume a slightly altered trajectory each time the thrusters were fired. The process of counteracting changes in attitude would require considerable fuel consumption on a thirty-day mission.
Boeing solved this design problem by eliminating the four thrusters on the solar panels together with all of the plumbing necessary to get gas out to them. This reduced weight and the quantity of attitude control gas. Next the [129] velocity control rocket was gimbaled. The change required addition of two gimbals, their actuators, and bearings, but now the rocket's nozzle could be moved to compensate for any perturbations caused by the attitude thrusters. This resulted in a weight saving of about 3 kilograms. The attitude control thrusters were half-pound thrusters located at the perimeter of the heat shield. They were coupled so that when one of the four fired in one direction, its opposite number would fire in the opposite direction with the same amount of thrust for the same duration, changing the spacecraft's attitude without affecting the thrust vector.24 This design change brought Lunar Orbiter's overall weight at the time of the Langley review 25 * to approximately 382 kilograms.
The participants of the review also tackled the problem of the Marquardt rocket motor, specifically the weight of the rocket's propellant versus the transit time from the Earth to the Moon and the specific impulse required to make the injection into lunar orbit. If the spacecraft was to achieve an initial elliptical orbit of 925 by 46 kilometers, it would require a total velocity change of slightly less than 1,100 meters per second. This meant that an Orbiter [130] weighing about 370 kilograms at separation from the Agena would require a specific impulse of 290 seconds. The Marquardt rocket, which had yet to pass qualifying tests for the Apollo Program, might not be able to achieve this high a specific impulse. (Although specific impulse is expressed in seconds, it is not a measure of duration. It is a measure of efficiency and indicates the thrust a rocket can provide at a certain rate of fuel consumption per second.) One possible solution to the problem, if the specific impulse of the rocket proved indeed too low, was to reduce the total impulse and alter the spacecraft's trajectory in order to place it in a more convenient initial elliptical orbit before transfer to final orbit.26
After reviewing the Marquardt rocket, the participants of the First Quarterly Review took up the examination of the last major problem to be considered at that time: Could the photographic system withstand the intense vibrations of the launch? The Eastman Kodak Company claimed that the vibration test levels were too high and that flight data on the launch vehicle did not warrant the high levels which Boeing had stipulated in its Environmental Criteria document. Boeing and Langley Lunar Orbiter Project Office people decided to reexamine the flight data of the [131] Atlas-Agena launch vehicle before making a decision on Eastman Kodak's complaint.27
This action ended the intensive two-day review of the program's major problem areas, and work proceeded. Two months later another review convened, and still more technical and engineering problems surfaced. They did not, however, threaten the comprehensive progress of the program toward its goals.

* Please note that footnote 25 was missing from the document I worked with so this is my best guess as to where it belongs. Chris Gamble, html editor.