Surveyor 4: The Impact of a Low Probability Event

While the tragic fire which killed the crew of Apollo 1 on January 27, 1967 effectively put the brakes on the Apollo program (see “The Future That Never Came: The Unflown Mission of Apollo 1”), work continued at an accelerating pace on NASA’s remaining automated precursor missions to the Moon. With a total of nine of these approved missions remaining to fly at the opening of 1967, NASA pushed hard to get them off the ground so that they could provide information vital for the success of the Apollo program after it resumed following the outcome of the Apollo 1 investigation.

The launch of Lunar Orbiter 3 from LC-13 on an Atlas-Agena D on February 5, 1967 – NASA’s first lunar mission of the year. (NASA/KSC)

First up was Lunar Orbiter 3 launched on February 5, 1967 which completed NASA’s photographic reconnaissance of future Apollo and Surveyor lunar landing sites (see “Lunar Orbiter 3: Preparing for Apollo”). On April 17, NASA launched Surveyor 3 which landed on the lunar surface three days later for its two-week long on-site investigation of another potential Apollo landing zone (see “Surveyor 3: Touching the Face of the Moon”). Just as Surveyor 3 was wrapping up its mission, NASA launched Lunar Orbiter 4 on May 4, 1967 which systematically mapped the entire lunar nearside for scientists at a resolution an order of magnitude or more better than was possible Earth-based telescopes (see “Mapping the Moon: The Mission of Lunar Orbiter 4”). As the summer of 1967 began, NASA’s next lunar mission, to be called Surveyor 4, was already being prepared to explore yet another possible lunar landing site for Apollo.

The Surveyor Spacecraft

Work began on NASA’s Surveyor program in May 1960 under the responsibility of Caltech’s Jet Propulsion Laboratory (JPL) in Pasadena, California (for details on the early history and development of Surveyor, see “Surveyor 1: America’s First Lunar Landing”). Built by Hughes Aircraft Company (whose space division is now part of Boeing), Surveyor was arguably the most advanced lunar spacecraft of its day. The basic 2.4-meter tall structure consisted of a simple 27-kilogram tetrahedral frame made of tubular aluminum alloy members. In each of the three lower corners was a landing leg equipped with an aircraft-style shock absorber and a footpad of crushable honeycomb aluminum. The total span of the legs, once deployed, was 4.3 meters. Rising from the apex of the frame was a mast upon which was mounted a gimballed planar high-gain antenna and a solar panel supplying up to 85 watts of electrical power to the lander’s rechargeable silver-zinc batteries. From the footpads to the top of its mast, Surveyor stood three meters tall.

Diagram showing the major components of the third and fourth Block I Surveyor engineering models to fly to the Moon. Click on image to enlarge. (JPL/NASA)

Buried inside the spacecraft’s frame was a Morton Thiokol-built 91-centimeter diameter TE-M-364 solid propellant rocket motor that would provide between 35.5 to 44.5 kilonewtons of thrust, depending on the motor’s temperature at ignition. This 656-kilogram motor, which would later be used as the third stage in various Delta launch vehicles models flown in the 1970s and as the final “kick stage” for the Pioneer and Voyager missions to the outer planets, would be used to negate most of Surveyor’s motion towards the Moon as the lander approached the lunar surface.

Surveyor also carried a second propulsion system for midcourse corrections and attitude control during the main retrorocket burn as well as for the final descent. This system consisted of three vernier engines fueled by monomethylhydrazine hydrate with MON-10 (a mixture of 90% nitrogen tetroxide and 10% nitric acid) serving as the oxidizer. These engines could be throttled by command of the spacecraft’s flight control subsystem producing between 130 and 460 newtons of thrust each. Yaw, pitch, and descent rate were controlled by selective throttling of the engines while roll was controlled by swiveling a single gimballed vernier. During the trans-lunar coast, Surveyor’s attitude was controlled by a set of six nitrogen gas jets, each providing 270 millinewtons of thrust.

All the temperature sensitive electronics were carried in two thermal boxes mounted to the frame. These compartments were covered with 75 layers of aluminized Mylar insulation and the tops were covered by mirrored glass thermal regulators. Compartment A, which maintained it internal temperature between +4° and +52° C, carried a redundant set of receivers and ten-watt radio transmitters, the batteries, their charge regulators, and some auxiliary equipment. The second box, Compartment B, was designed to maintain the temperature between -15° and +52° C. This compartment carried the computer “brains” of the spacecraft which controlled all aspects of the lander’s operation using a total of just 256 commands. Mounted elsewhere on the frame were star sensors, a pair of radar systems for landing, low-gain antennas, propellant, and helium pressurization tanks.

Diagram detailing the components of the first Block I Surveyor landers as viewed from above. Click on image to enlarge. (NASA)

A total of 30 kilograms of instrumentation were carried by the first “Block I” Surveyors which were considered “engineering models” for more advanced versions of the lander which had been proposed. Most of this instrumentation were engineering sensors such as strain gauges, accelerometers, rate gyros, temperature sensors, and so on to be used to make more than two hundred measurements of the spacecraft’s performance and condition. While not specifically designed for investigating the lunar environment, many of these measurements could be used to determine some of its basic properties including surface mechanical properties and its temperature.

The only true scientific instrument carried by the initial batch of what would be seven “Block I” Surveyors was a slow-scan television camera. The camera was mounted in a 1.65-meter tall mast attached to the spacecraft’s framework. The camera pointed up into a movable mirror that allowed the camera to view 360° of azimuth and from 60° below to 50° above the normal plane of the camera. The 7.6-kilogram camera package was canted at a 16° angle to offer a clear view of the surface between two of the footpads out to the lunar horizon 2½ kilometers away. The camera was fitted with a 25 to 100 mm zoom lens that offered a field of view of between 25.3° and 6.4°.  The aperture could be set between f/4 and f/22 and the lens could be focused from 1.2 meters to infinity.  A shutter was also included so that various integration times could be used to obtain the ideal exposure.  While the nominal exposure time was 150 milliseconds, exposures as long as about thirty minutes could be accommodated.  The typical resolution of the camera was one millimeter at a distance of four meters. By combining a series of images taken in a stepwise fashion at various azimuth and elevation angles, panoramic mosaics of the spacecraft and the surrounding terrain could be created.

Diagram showing the major components of the camera Surveyor used to image its surroundings after landing. Click on image to enlarge. (NASA)

The camera was also fitted with a filter wheel containing clear and three color filters. With the aid of calibration targets mounted at various points of the spacecraft, pictures taken through red, green, and blue spectral filters could be reconstructed back on Earth to yield full-color views of the lunar surface. The camera could only operate in real time via remote control from Earth using a total of 25 commands. The primary means of transmitting images was through the high-gain antenna. Using this powerful antenna, an image would be broken up into 600 scan lines and transmitted back to Earth in 3.6 seconds. The use of the less powerful low-gain antennas, which served as a backup, would permit an image to be broken up into 200 lines and would require 61.8 seconds to transmit.

A second television camera was included in the original Block I Surveyor design which was pointed downwards to provide a view of the lunar surface and a footpad. These images were to be transmitted during Surveyor’s final approach starting at an altitude of 1,600 kilometer to allow the landing site to be pinpointed, along with providing information on the surrounding terrain. As it turned out, however, this camera was never used on the first two flights in order to simplify the already complex landing sequence. Later, the requirement was deleted and the camera was removed altogether since NASA’s Lunar Orbiter missions were providing the needed detailed images to help interpret the Surveyor findings and place them into a regional context.

New Instrument

With the cancellation of the more advanced Block II Surveyor missions on December 13, 1966 which would have explored the lunar surface with a wider array of instruments, program officials decided to incorporate some of the Block II science payloads already completing development on the remaining five approved engineering flights of the more limited Block I spacecraft. The third and fourth Block I Surveyor flights were selected to carry a remote controlled mechanical arm. Formally known as the Soil Mechanics Surface Sampler (SMSS), this arm consisted of a simple tubular aluminum pantograph with a 13-centimeter long, five-centimeter wide scoop attached to the end. One electric motor on the SMSS allowed the pantograph to extend outwards from 58 to 150 centimeters while another opened and closed the door on the scoop. A third motor allowed movement through 112° of azimuth while a fourth provided 42° of motion in elevation. Used in conjunction with Surveyor’s slow-scan television camera, the SMSS would be operated remotely in near-real time by an operator on the Earth to provide information on the mechanical properties of the lunar soil up to a depth of half a meter. The SMSS would give scientists their first chance to touch the surface of the Moon.

Diagram showing the major components of the Soil Mechanics Surface Sampler (SMSS) carried by Surveyor D. Click on image to enlarge. (NASA)

The SMSS was attached using the mount which originally held Surveyor’s now-deleted descent camera. The SMSS had to be modified to use the existing wiring harness to supply power and commands from the ground to operate the device. Since only a single telemetry channel was available, the original mix of sensors on the SMSS to measure strain, acceleration and position at various points were removed. Instead, only a measurement of the current being drawn by the individual SMSS electric motor in use at any time would be sent back to the ground to allow engineers to estimate roughly the force being applied by the arm. The SMSS would be capable of reaching anywhere from 0.6 to 1.6 meters from the lander in a three meter arc starting from Footpad #2 and from one meter above the surface to as much as 0.45 meters below. All together, the SMSS hardware had a mass of 3.8 kilograms while its separate electronics compartment came in at 2.9 kilograms. Although some minor issues were encountered, the SMSS was successfully used for the first time on the Surveyor 3 mission launched on April 17, 1967 giving scientists their first chance to touch the surface of the Moon by remote control (see “Surveyor 3: Touching the Face of the Moon”).

This diagram illustrates the range of motion of Surveyor’s Soil Mechanics Surface Sampler (SMSS). Click on image to enlarge. (NASA)

Surveyor’s launch vehicle was one of NASA’s most advanced rockets called the Atlas-Centaur. The Centaur upper stage used liquid hydrogen and liquid oxygen (LOX) as propellants – the first rocket stage to do so. This combination provided up to half again as much thrust than a like mass of conventional propellants then in use. The Atlas booster used with the Centaur was a modified version of the Atlas D ICBM. The forward propellant tank was altered to accept the wider and heavier upper stage and a new MA-5 engine assembly providing ten percent more liftoff thrust than when the baseline MA-2 system was used.

A cutaway diagram of the Atlas-Centaur with the Surveyor lunar lander. Click on image to enlarge. (NASA)

In order to maximize its payload and launch widow flexibility, the Atlas-Centaur was designed to first place its payload into a low parking orbit. The pair of RL-10 engines powering the Centaur would then reignite at the proper injection point to send the Surveyor lander on its way to the Moon. Due to problems with the development of the Centaur and its in-orbit restart capability, the initial two Surveyor missions were forced to use direct ascent trajectories to the Moon which required only a single burn of the RL-10 engines. With the successful test flight of Atlas-Centaur 9 (AC-9) on October 26, 1966 where the Centaur finally demonstrated its in-orbit restart capability by sending a dynamic model of a Surveyor lander, designated SD-4, into a simulated lunar trajectory, the way was finally clear for the technique to be used in future Surveyor missions. Although the “parking orbit” mode was successfully employed by AC-12 to launch Surveyor 3, for the fourth Surveyor mission, designated “Surveyor D” before launch, the direct ascent profile was used one last time. The available launch vehicle, AC-11, was the last Centaur configured for the direct ascent mode and adequate launch windows existed for the Surveyor D mission with a July 1967 departure.

Diagram illustrating Surveyor’s lunar trajectory options. Click on image to enlarge. (NASA)

Unlike the later Apollo lunar landing missions, Surveyor was designed to make a direct descent to the lunar surface from its translunar trajectory about 65 hours after launch with no intermediate stop in lunar orbit. Since Surveyor was designed with the capability of landing on the Moon with an approach trajectory substantially off of the local vertical, most of the lunar hemisphere facing Earth was accessible to Surveyor. Early flights, however, were limited to the equatorial mare regions of what was called the “Apollo landing zone” which appeared to be the safest landing sites based on orbital photography. It was intended that the initial Surveyor missions would provide ground truth data on these proposed sites to support the upcoming Apollo lunar landing missions.

Preparing for the Fourth Mission

The primary objectives of the Surveyor D mission were to perform a soft landing on the Moon at Sinus Medii near the center of the Moon’s visible face and then obtain television images of the lunar surface. Secondary objectives included using the SMSS to manipulate lunar surface material while being observed using the television camera as well as obtaining information on the bearing strength, radar reflectivity and thermal properties of the lunar surface using data from Surveyor’s collection of engineering sensors.

A map of Sinus Medii including the intended landing site for the Surveyor D mission. (NASA)

The planned landing site in Sinus Medii was a 60-kilometer wide circle centered at 0.42° N, 1.33° W about 25 kilometers northwest of the intended landing site of the unsuccessful Surveyor 2 mission launched in September 1966 (see “Surveyor 2: Things Don’t Always Go As Planned”). This site was chosen because it was considerably rougher than the Surveyor 1 and 3 landing sites and would provide verification of yet another landing site in support of Apollo project objectives. In order to reach this target point, Surveyor would need to perform a much larger gravity turn during descent due to the initial approach which was about 36° to the local vertical – much larger than the 6° and 25° angles of the Surveyor 1 and 3 missions, respectively, with their more westerly equatorial landing sites.

Based on the experience with the previous Surveyor flights, a number of improvements were made to spacecraft number SC-4 to be used on the Surveyor D mission. Most notable among these were modifications made to the logic circuitry of the landing radar system to prevent a repeat of the three-bounce landing of Surveyor 3 caused by a loss of radar lock with the lunar surface during the last seconds of the descent preventing an automatic vernier engine shutdown. The spacecraft also included a pair of metal bars (one magnetic and the other nonmagnetic) attached to one of the footpads. Observations of these 5.1 by 1.3 by 0.3 centimeter thick bars by the television camera would allow scientists to determine the magnetic properties of the lunar soil. The launch mass of SC-4 for the Surveyor D mission was 1,039 kilograms – the heaviest of the series so far. After it had jettisoned its spent retrorocket and consumed all of its propellants during descent, the spacecraft would have a mass of just 283 kilograms after landing.

This diagram shows the placement of two metal bars on the landing pad of Surveyor D to be used to assess the magnetic properties of the lunar soil. Click on image to enlarge. (NASA)

In order to reach the intended landing site within the limits imposed by mission requirements, the spacecraft, launch vehicle and tracking network, launch of Surveyor D had to take place during one of the launch windows available daily between July 13 and 17, 1967. The first day’s launch window of just 6½ minutes on July 13 was the shortest opening at 7:02 AM EDT. The restrictions on the following day’s launch window opened up allowing a launch anytime between 7:53 and 8:30 AM EDT. Launch windows on subsequent days were later in the morning and of shorter duration.

This schematic illustrates the effects of the various constraints on the launch windows of the Surveyor D mission for launch between July 13 and 17, 1967. Times are in GMT. Click on image to enlarge. (NASA)

The first major piece of mission hardware to arrive at Cape Kennedy (which reverted to its original name of Cape Canaveral in 1973) was the Surveyor SC-4 spacecraft on April 24, 1967. This was followed by the arrival of the Atlas 291D booster for AC-11 on April 29, the nose fairing and interstage adapter on May 1 and finally the Centaur upper stage four days later. The Atlas was erected on Pad A of Launch Complex 36 (LC-36) on May 3 with the Centaur added to the stack three days later.

Here we see Atlas 291D, which would serve as the booster for Atlas-Centaur 11, being worked on prior to shipment to Cape Canaveral for the Surveyor D mission. (General Dynamics)

Following initial verification testing of SC-4 which was completed on May 24, the spacecraft was prepared for acceptance testing with the launch vehicle. Temporarily fitted with a dummy retrorocket and test instrumentation, SC-4 was encapsulated in its nose fairing and added to the top of AC-11 on May 31. With the completion of this first round of joint tests, SC-4 was demated on June 3 to begin its final preparations for launch. Following the final readiness tests for AC-11, the encapsulated Surveyor spacecraft was added to the stack for the final time on July 9. After additional tests and a practice countdown, AC-11 and SC-4 were now ready for their first launch attempt on July 13.

The Surveyor 4 Mission

After some minor issues during fueling of the Atlas booster, erratic behavior was noted in the Centaur’s propellant utilization valve during prelaunch preparations on July 12, 1967. The Surveyor encapsulated in its nose fairing had to be briefly lifted off of the top of the launch vehicle to allow technicians to access and tighten the valve forcing the countdown of Surveyor D to be recycled to start during the evening of July 13.

Here we see Atlas-Centaur 11 on the pad at LC-36A prior to the launch of the Surveyor D mission. (NASA)

The final countdown for the Surveyor D mission proceeded as planned with only the usual mix of minor issues encountered by the end of the last scheduled hold at T-15 minutes. The countdown then resumed as planned until T-39 seconds when an unscheduled 29-second hold was needed to complete topping off the Centaur’s liquid hydrogen tank. Finally, at 7:53:29.2 AM EDT (11:53:29.2 GMT) on July 14, 1967, AC-11 lifted off from LC-36A for the start of the Surveyor D flight only 29.2 seconds into its launch window.

The launch of Surveyor 4 from LC-36A on July 14, 1967. (NASA)

A nearly flawless performance by the Atlas 291D booster was followed by a single, successful 437-second burn of the pair of RL-10-A3-3 engines of the Centaur. Despite the Centaur burning for 5.2 seconds longer than planned, it had successfully placed what was now designated Surveyor 4 into a trajectory towards the Moon which would miss the pre-launch aim point by only 176 kilometers after a transit of about 62 hours. Surveyor 4 separated from its spent upper stage 12 minutes, 36.9 seconds after launch and proceeded to deploy its appendages and lock onto its celestial attitude references for the cruise to the Moon. Five seconds after spacecraft separation, the Centaur turned and started venting its 170 kilograms of residual cryogenic propellants as part of a retromaneuver to move it safely away from Surveyor and the Moon. The Centaur would safely pass 22,300 kilometers from the Moon about 70 hours after launch.

Shown here are the events and nominal times of the direct ascent launch profile of the Surveyor 4 mission. Click on image to enlarge. (NASA)

With Surveyor 4 settling into its cruise to the Moon after its star sensor finally locked onto Canopus 6½ hours after launch, engineers checked all of the spacecraft’s systems and worked to determine its trajectory in order to plan its midcourse maneuver nominally scheduled for 16 hours after launch during the first tracking pass of NASA’s primary Goldstone tracking station in California. With the orbit calculation ten hours after launch showing only a 176-kilometer miss, ground controllers decided to postpone the midcourse maneuver by 22 hours during the second tracking period by Goldstone. While this delay would tend to increase the midcourse velocity change, or delta-v, required to meet a particular trajectory requirement (which was already quite modest compared to Surveyor’s capabilities), the delayed maneuver would help to improve the accuracy of the final targeting.

This diagram illustrates the typical sequence of events for Surveyor to perform a midcourse correction. Click on image to enlarge. (NASA)

At 02:30:02 GMT on July 16, Surveyor 4 ignited its three vernier engines for 10.5 seconds for a delta-v of 10.13 meters per second – only 1.4% shy of the intended value. The midcourse maneuver not only adjusted the trajectory to a slightly altered aim point at 0.43° N, 1.6° W, but it modified the descent profile a touch so that retrorocket burnout would occur at a slower descent speed of 153.7 meters per second instead of 161.5 meters per second to provide a little more margin during the final descent. Touchdown was predicted to occur after a total transit time of 62 hours, 11 minutes and 43 seconds.

This schematic diagram shows the major milestones of the Surveyor 4 voyage to the Moon. Click on image to enlarge. (NASA)

As the local afternoon of July 16, 1967 wore on at the spacecraft control center at JPL in Pasadena, California, Surveyor 4 was being prepared for its final descent to the lunar surface. At 01:24:44 GMT on July 17 (6:24:44 PM PDT on July 16 at JPL), Surveyor 4 began to change its attitude to align itself for descent. At 02:01:54.784 GMT, the altitude marking radar mounted inside of the nozzle of Surveyor’s retrorocket hit the preprogrammed 96.5-kilometer mark to begin the countdown to engine ignition. At 2:01:57.493 GMT, the vernier engines ignited followed 1.109 seconds later by the ignition of the main retrorocket. Everything was proceeding as planned until 02:02:39.696 GMT when Surveyor 4 suddenly stopped transmitting 41.09 seconds into the retrorocket’s nominal 42.5-second burn. Just 1.222 seconds later, Goldstone’s 26-meter DSS-11 antenna and the new 64-meter DSS-14 simultaneously lost contact with Surveyor 4. At the point of loss of signal, Surveyor 4 was about 14.9 kilometers above the lunar surface and descending at 326 meters per second with all telemetry readings appearing normal.

Diagram showing the major events in Surveyor’s descent to the lunar surface. Click on image to enlarge. (NASA)

Initially it was assumed that Surveyor 4 had safely landed automatically about 2½ minutes after loss of signal at 2:05:11.3 GMT at 0.43° N, 1.62° W – about 8.61 kilometers from its post midcourse aim point. Commands were transmitted to Surveyor 4 starting with the first post-landing sequence at 02:05:31 GMT assuming that the spacecraft could receive and act upon those commands. In order to regain two-way contact, the spacecraft was systematically commanded to use various combinations of transmitters, antennas and transmitter power levels while in various battery modes. Despite all the effort, nothing was heard from Surveyor 4.

NASA’s new 64-meter DSS-14 at Goldstone, California was used to track Surveyor 4 during its descent and during subsequent attempts to recontact the lander. (NASA/JPL)

The first round of attempts to contact Surveyor 4 stopped as the Moon set as viewed from Goldstone at 08:07 GMT on July 17 but were resumed at 00:51 GMT the following day after the Moon rose again. After station transfer at 06:00 GMT, NASA’s 26-meter DSS-42 antenna in Canberra, Australia began its attempts to contact Surveyor 4 with commands sent to reposition its high gain antenna and solar panel. Efforts to contact Surveyor 4 were officially terminated at 12:16 GMT when the spacecraft failed to respond after 4,255 commands had been sent during the course of 34 hours.

With the failure of Surveyor 4 to meet its objectives, a Technical Review Board was formed to determine the cause and recommend corrective actions. Because of the lack of any evidence of the source of failure in the spacecraft telemetry, it was not possible for the Board to determine the definitive cause of the failure aside from the fact that it was caused by spacecraft hardware. Possible causes identified included 1) breakage of a critical power lead in the wiring harness or the failure of a connector or solder joint, 2) a rupture in the case of the Thiokol TE-M-364 retrorocket, 3) a failure in the transmitter which caused its power to drop to zero or 4) sudden failure of a pressure vessel (e.g. propellant, helium or nitrogen tanks). All of these failure modes could produce the observed sudden loss of signal but the lack of required evidence made it impossible to chose.

Although a definitive cause for the failure of Surveyor 4 was not found, one of the potential causes was a catastrophic failure of the Thiokol TE-M-364 solid retrorocket motor shown here. (NASA)

Since the Board was unable to determine the cause of the failure of Surveyor 4 and the most likely culprits involved low-probability events, no changes to the Surveyor spacecraft were deemed warranted. While the Board did make some general recommendations, the decision was made to push forward towards the launch of the fifth Surveyor planned for September 1967 (see “Surveyor 5: Pulling Success from the Jaws of Failure“).

Follow Drew Ex Machina on Facebook.

Related Reading

“Surveyor 1: America’s First Lunar Landing”, Drew Ex Machina, May 30, 2016 [Post]

“Surveyor 2: Things Don’t Always Go As Planned”, Drew Ex Machina, March 13, 2017 [Post]

“Surveyor 3: Touching the Face of the Moon”, Drew Ex Machina, April 17, 2017 [Post]

General References

J. Jason Wentworth, “A Survey of Surveyor”, Quest, Vol. 2, No. 4, pp 4-16, Winter 1993

NASA Prepares to Launch Fourth Surveyor, NASA Press Release 67-172, July 11, 1967

Surveyor IV Mission Report: Mission Description and Performance, JPL Technical Report 32-1210, January 1, 1968