Robert Andrews in 1970.
Hello, my name is Robert E Andrews and I had the good fortune of working in the aerospace business from its early stage in the ’50s through the end of the century.
I attended Drexel University in Philadelphia after high school, matriculating into the mechanical engineering major. I didn’t like this field, lost interest, and dropped out after two years. However, during these years, I did learn a skill that gave me employment. I got a job with the helicopter pioneer Frank Piasecki as a draftsman. I also applied for work with General Electric’s new Missiles and Space Department in Philadelphia, which was located just across the street from Drexel.
In June of 1958, I got a letter from GE asking that I come in for an interview. My supervisor at Piasecki encouraged me to go. He realized working for a big company like GE would offer me many career opportunities. I got the GE job working in the Navigation and Control Engineering drafting room.
I found out that all employed in engineering here required a Secret clearance. So, I was placed in a small room by myself with a drafting table and remained isolated for about six weeks. After my clearance was granted, I moved into the main drafting room with about sixty other draftsmen, designers, and lofts men. (A loftsman develops full scale patterns, or drawings, of complex shapes.)
One of my early assignments was to lay out a printed circuit board. My electrical engineer gave me a schematic and identified each part type, be it a TO5 transistor, carbon resistor, tantalum foil capacitor, et cetera. My task was to fit all the components on a fixed size board and connect them according to the schematic with copper circuit runs on the back of the board. These were the days of single sided boards and jumper wires were not allowed. I took to this like a duck to water. I was given more boards to layout and did well.
My supervisor observed that I was anxious to learn new things. He assigned me to the Discoverer project to develop system schematics, interconnection diagrams, and wire harness definition. I was working for the Electrical Systems engineer. His was a very responsible and highly visible position. And like most engineers working in this new field, he was in his early 20’s, just three years older than me. There were senior engineers in management positions, but there were no senior spacecraft engineers. Everyone was learning on the job.
Discoverer was started in early 1958. Officially, it was a US Air Force program to launch payloads into space and recover them. The Discoverer spacecraft included the Agena upper stage, manufactured by Lockheed, atop a Thor rocket. General Electric was responsible for the satellite recovery vehicle, or SRV, mounted on the nose of the spacecraft. Discoverer was scheduled for its first launch in early 1959 from Vandenberg Air Force Base on the central California coast.
As part of my introduction to the Discoverer program, I was taken to the monkey lab. The Discoverer cover story was that it was a biological experiment to see how monkeys would perform under the conditions of spaceflight. So there were technicians and medical employees in this room full of screaming caged monkeys trying to train them to do simple tasks like hitting a switch when a light came on so they would get a piece of apple. What a terrible place that stunk to high heaven. I hoped I never had to interface with this group again.
CORONA mission being prepared for launch at California’s Vandenberg AIr Force Base in 1963. The success of the mission depended on many systems and components, including the reentry vehicle at the nose, which was built by General Electric. (credit: Peter Hunter)
One day, a fellow draftsman in my group told me I was to come with him to a meeting. We went to a remote area in the building where there was very little personnel traffic. I didn’t know it even existed. We came to a locked door and he had a key to open the door. Inside was a GE employee who introduced himself as a security agent. I thought I was going to lose my Secret clearance. Instead, he told me I was going to be briefed on a highly classified program called CORONA. He told me that what I was about to learn, I could not discuss with anyone who was not also cleared for this information. I could not talk about this with family, friends, or other GE employees forever. He asked if I accepted those conditions and after I did, he gave me a brief overview of the real purpose of Discoverer, which was to carry a camera into space to photograph the Soviet Union and then return the film inside the GE-built reentry vehicles.
The security officer also gave me instructions about meeting the others cleared on the CORONA program. My friend who brought me to this briefing was of course a member of this closed society. There was a formal procedure to follow. To meet someone new, you must have a third party that knows you and the other person are both cleared. This must be done in a closed “black” area. The third party gives the introduction: “Robert Andrews, this is a formal introduction to John Jones on Program C”. A handshake follows. Note that the word CORONA is shortened to C. I don’t believe I ever heard CORONA spoken again.
Now I learned about film takeup reels, film transfer chutes, film cutters, those extra wires in the harness bundles to drive the reel motors. Now a complete picture came into focus.
But what about those people in the monkey lab? They believed this spacecraft was for their payloads. And they will never find out they are just a cover for the CORONA program.
There were many failures in the first flights of CORONA. Due to the lack of diagnostic instrumentation, the engineers had to sometimes infer the cause of failure from observations. Finally, they had success in August 1960. I will always remember Chief Engineer Charlie Robinson coming into the drafting room to announce that Discoverer 13 had been recovered from space. The response was euphoric. This was our equivalent of landing on the Moon. This was the first object in history to be recovered from space. All our hard work paid off.
A CORONA film return capsule. (credit: National Air and Space Museum)
Months later, my Electrical System engineer came to my drawing board and said “Bob, we have a major change to the thrust cone”. The thrust cone is the part of the Satellite Recovery Vehicle that contained the hardware to deorbit it from space. After separation from the Agena/camera vehicle, two solid rocket motors spun the SRV to about 60 rpm to stabilize the vehicle. Then a large solid rocket motor fired to slow the SRV and cause reentry. After SRM fire, a second set of solid rocket motors fired to despin the SRV to near zero rpm. This was required so that after reentry when the parachute was deployed, the lanyards wouldn’t get twisted by a rotating SRV body.
There were five rocket motors and each represented a single fault failure. A spin rocket failure would result in rotation about an axis that is not aligned with the axis of the retro rocket. So when ignited, its thrust would send the SRV into an unpredictable vector with no chance of recovery. A despin rocket failure would result in the SRV tumbling and could prevent proper reentry attitude and subsequent parachute deployment. The retrorocket did have dual bridgewire igniters to help improve its reliability.
During an earlier mission, an SRV did not reenter after separation from the Agena. It was postulated that one of the four small rocket motors could have been the cause. So, our GE engineers were going to replace them with a cold gas system.
Two identical systems were designed, one for spin up, the other for despin. A very simple design of a small spherical tank connected to parallel, normally closed explosive valves, and then piped to two small nozzles on opposite sides of the thrust cone. After firing the one set of explosive valves, the spin nozzles caused the SRV to rotate in one direction. Later, the despin valves were blown to stop the SRV’s rotation. There were no single point failures (excluding loss of tank pressure) in this design.
My task was to change the electrical drawings to reflect this new design. And my Electrical System Engineer Al Gross threw me a curve. He said, “Bob, we need to connect two new circuits with a couple of diodes. I don’t know where you are going to put them, but I don’t want to add them to the Recovery Programmer as this will trigger a requal program,” That would require a lot of work and paperwork to requalify the programmer. I told him that we were not allowed to put electrical parts into the wiring harnesses. My solution was to solder the diodes into a small bulkhead connector and create a new harness interface to it. Al put me in for an award for my creative design. It was during this period that I decided I wanted to be an electrical engineer.
Drawing of an early 1960s version CORONA reconnaissance satellite showing how the film was collected in a reentry vehicle at the front of the spacecraft. The reentry vehicles were built by General Electric and used on CORONA satellites until 1972. A similar reentry vehicle was used for the GAMBIT-1 and then GAMBIT-3 satellites, which operated until the 1980s. (credit: NRO)
GAMBIT 1: 1961–1967
I received a significant promotion around this time. It was from draftsman, who made detail drawings, to designer, who is responsible for high-level layout drawings. Management had recognized my potential and I was very thankful to be moving forward in my career. It was then I got that familiar tap on the shoulder with the instructions to “follow me.” I went again to yet another secluded office with a security officer. Here I was told about a new program where we at GE were to design the entire spacecraft, not just the SRV. The program was called GAMBIT but, for reasons I never understood, was referred to as Program Z. Again, all the same restrictions and protocol as CORONA, but with new people that may or may not have C clearance.
GAMBIT was a more powerful reconnaissance satellite with a bigger camera. GE was building the spacecraft that would hold the camera stable and point it while in orbit. GE was also building a modified version of the CORONA SRV that would return the film to Earth. GAMBIT would also launch from California, atop a more powerful Atlas-Agena rocket.
My drawing board was moved into an office with the Navigation and Control Engineers to enhance communications. Other support engineers (thermal, power, communications) also had their small offices spread around the six-story building in Philadelphia. Management saw this was very bad for interfacing between disciplines and decided to move us into the Valley Forge facility that was still under construction 20 miles away. When we moved in, they were still pouring concrete floors in parts of the building.
As the support functions (manufacturing inspection, test, program office, etc.) began to staff up, we outgrew out allocated space in the new facility. Space was rented in a commercial complex about two miles away and we moved into seven buildings. Engineering was placed in buildings 6 and 7.
As the spacecraft design was progressing, the drafting room became the focal point. There were over 200 people working on every facet of the design. My manager asked me to take the position of supervisor to help with task assignments and to make sure that schedules were met. My people supported the Navigation and Control engineers. From this association, I learned how horizon crossing detectors worked, what role the rate gyros played in the control of the spacecraft, and how rate and position algorithms controlled thruster firings or reaction wheel speeds. It was a great learning experience.
The Air Force began launching our GAMBIT spacecraft from Vandenberg AFB in July 1963. It was a busy time both at our facility and Vandenberg, with a launch every four to six weeks. Manufacturing problems, system test problems, and orbital problems kept our engineers very busy. Design changes, while frowned upon by our program office and our Air Force customer, were a fact of life.
I remember one problem that I am not sure was ever solved. There was a power supply that passed all qualification and system-level tests, but failed after a few hours or days in orbit. This did not happen on every flight, but enough to know something was wrong. The power engineers were looking for some type of transient that triggered this failure. They developed a test plan on a full-up spacecraft. This was in the days before storage oscilloscopes that can capture the waveform of onetime electrical events. So, the test was instrumented with about eight standard oscilloscopes manned by eight technical people trying to observe an elusive transient. The test was scheduled for 24 hours and therefore required 24 people to man the scopes. I was asked to support. The power engineers were spread across the three shifts and tried to tell us what we might see. To make a long story short, after starring at an oscilloscope for eight hours, no one saw anything. The damn thing didn’t fail. I don’t remember if any design changes were made, but subsequent flights were okay.
A major part of the GAMBIT design included the Satellite Recovery Vehicle build under subcontract by our sister department Missiles and Ordnance Systems Department in Philadelphia (where we started this program). As with all subcontracts, Special Military Space Programs (SMSP) (our new name) had people in oversight positions to interface and monitor progress. Primarily they were program office, finance, and engineering people.
Sometime in 1964, I was called to the Chief Engineer’s office. What did I screw up? After a cordial welcome, he got down to business. His SRV subsystem engineer had accepted a position with our subcontractor. The Chief knew of my history with the CORONA recovery vehicle and my knowledge of our SRV. He asked if I thought I could handle the position recently vacated by our SRV subsystem engineer. I was to be the technical interface and responsible for the design interfaces to our GAMBIT spacecraft. I was scared to death. But as comes with youth, I was confident I could do this. I told the Chief I would do my best. There is one caveat he said. “At GE, you are not titled ‘Engineer’ unless you have a degree from an accredited college. If you have an engineering assignment without a degree you are an ‘Engineer Specialist.’ For me to give you this position, you must agree to complete your formal engineering education.” The next day, I contacted Drexel and enrolled in the Evening College Electrical Engineering program. I was committing myself and my family to the most difficult five years of my life. Working more than full time, traveling for business, going into Philadelphia for classes three nights a week, studying all day Sunday, and finding time for my family over a five-year period was a full plate. I don’t recommend it to anyone.
Back in the day, we didn’t have Interface Control Documents (ICDs) between the prime contractor (GE) and the Air Force. We depended on our customer, Air Force Space and Missile Systems Organization (SAMSO), to define interfaces with the other Air Force branches involved with the program. This was not always a good solution. Here’s why.
The last task of the GAMBIT orbit operations was recovery of the capsule (called a “bucket”) by the aircraft. A squadron of specially equipped C-130s at Hickam AFB in Hawaii had that assignment. They trailed a cable hung between two poles that stuck out of the back of the aircraft. The pilot would locate the bucket floating under its parachute and fly over it, catching the parachute lines on hooks on the suspended cable. The bucket could then be winched into the rear of the C-130 and flown back to Hawaii. In Hawaii, the bucket with its film would be loaded into an Air Force jet and flown to the mainland, and then to a facility where the film would be processed.
The bucket had a flashing strobe light and a colored parachute for visual acquisition, and a radio frequency beacon provided for electronic tracking and was activated after the SRV deployed its parachute. I had gotten word from our Air Force customer that the C-130 pilots could not lock on to the beacon signal during the last mission and had to make a steep dive to catch the bucket under its parachute. I contacted my interface at the recovery vehicle manufacturer at the GE department in Philadelphia. Their response was that they changed the modulation characteristics of the beacon so that the signal would be easier to capture by the tracking receiver.
Not only we, the GAMBIT customer, didn’t get the word, but the C-130 wing didn’t, either. So, within a week, I had travel arrangements to Honolulu. I arrived in January of 1967 at Hickam AFB with a recovery beacon, antenna, coax cable, recovery battery, and power cable for a test with the aircraft. The night of my arrival, I had to activate the silver zinc battery with electrolyte using a hypodermic needle on the bathroom floor of my hotel room, The Royal Hawaiian. The battery had a fairly short life after activation and could not be charged. We also didn’t want to carry a live battery on the flight to Hawaii. (On my retirement from Lockheed Martin, I donated one of these recovery batteries to the Smithsonian and it is on display at the Udvar-Hazy Museum near Dulles International Airport.)
My point of contact was Colonel Popsicle (phonetic spelling) at the 6594 Air Force Wing. After a security introduction confirming we both were cleared on the GAMBIT (Z) program, we outlined the aircraft test. I would connect my equipment on a cart and pull it in front of the C-130 parked on the tarmac. Holding the antenna in my hand, I would walk back and forth across the front of the direction-finding antennas of the aircraft. Inside the aircraft were technicians from the direction-finding manufacturer (I think they were Philco guys.) They were re-tuning the aircraft receiver to recognize my beacon signal. After a few hours, they believed they were successful. We all regrouped in the colonel’s office. He and his pilots wanted a flight test to confirm that day’s results.
The flight test was outlined by the colonel as follows: Two C-130s would be required. In aircraft #1 the rear door would be open and I would be harnessed in there with my equipment used in today’s test. Aircraft #2 would be the same one used in today’s test with the now re-tuned receiver. Aircraft #2 would follow aircraft #1 to confirm tracking the recovery beacon.
This all sounded like a good plan except for one issue: I didn’s like the part where I was flying in a plane with open doors over the Pacific Ocean. I quickly assured the colonel that I could train one of his people to operate my equipment. No need for me to be on the aircraft. He accepted that suggestion from this cowardly civilian. He gave me a very sharp young airman who learned my equipment operations in little time. I later learned that the test was successful, but I was gone from Hawaii by then.
While in the colonel’s office, I got a call from the home office in Valley Forge. They got word from our Air Force customer that another program using the same recovery vehicle (probably CORONA) had a recovery failure. That program thought maybe the retrorocket didn’t ignite. Perhaps the igniter wasn’t installed properly. My chief engineer wanted me to stop by Vandenberg AFB on my way home from Hawaii to talk to the GE tech that installed retro igniters.
At Vandenberg, I discovered that installing the igniters was no simple task. The tubular igniter was inserted into the retrorocket through the engine nozzle. It was threaded onto a screw inside the top of the engine. The electrical ignition wires embedded in the igniter exited the nozzle and were routed to the outside of the nozzle, where they were taped for support. Finally, the igniter connector was mated to the thrust cone harness. I was given a demonstration of the entire process and was confident that this tech knew his job. Having completed my assignment, I finally got back to Pennsylvania after a few weeks on the road. My wife, daughter, son, and three-month old baby girl were happy to see me. Our only contact during this time had been a few phone calls. My wife knew I was in Hawaii, but didn’t know why.
Later, I was again contacted by the 6594 Wing. They had a problem catching the last capsule. Its rate of descent was much greater that others before and the pilot had to dive after it but did catch it. They sent me a film record of this recovery taken from the open hatch of the C-130 aircraft. In the film, I saw the parachute come into view, the recovery hooks of the aircraft grabbing the chute, and then another event that shouldn’t have happened. The ablative heat shield (fore body) was still attached to the capsule. The snap of the chute capture was enough to pull the capsule out of the fore body and the film record showed the fore body falling into the Pacific Ocean. This was definitely an anomaly. The parachute cover, the recovery capsule (holding the film), and the fore body are sandwiched together by explosive bolts. These bolts are fired at the time of chute deployment. When the chute blossoms, it decelerates both the capsule and fore body pulling them apart. The fore body is long gone by the time the aircraft is near. But for some reason, this ablative heat shield was still stuck to the rest of the vehicle when the C-130 pilot caught it.
On earlier recovery capsules returned to our factory, I noticed the scorching on the capsule cover caused by the heat of reentry. This cover is inboard from the ablative fore body, but still subjected to high temperature. I had the GE SRV manufacturer do a thermal analysis of the interfaces between the recovery capsule and the fore body to determine if mechanical interface tolerances could be impacted by higher than predicted heat from reentry. As a result, there were dimensional changes implemented on all future SRV’s. This retention of the fore body was never repeated in a GAMBIT flight.
During my time on the GAMBIT 1 program, we were invited to submit a proposal on the next program, GAMBIT 3. It was a two-SRV vehicle design with a new camera system. I believe we put together a good technical design based on our GAMBIT 1 vehicle experience, and I attended the orals at SAMSO. However, GAMBIT 1 program management at GE did not have a good relationship with the Air Force customer. Due to cost overruns, there was a continuing battle as to who was responsible to pay for changes. Because of this and some technical problems, GE wasn’t selected to win this contract.
Fortunately, this didn’t have a major impact on employment at GE. Most of the engineers had completed their tasks by this time and moved on to other projects at GE Space Systems. Manufacturing, quality control, and testing still had a backlog of GAMBIT 1 vehicles to produce. And there was another major program cranking up and needing engineers: MOL.
I ended my assignment as GAMBIT SRV subsystem engineer in early 1967 and was transferred to the MOL Program.
Frame grab from a declassified National Reconnaissance Office video showing a technician pointing to a General Electric reentry vehicle for the Manned Orbiting Laboratory. This reentry vehicle would have been loaded with film by the astronauts on MOL, and returned to Earth before the end of the mission. Robert Andrews was responsible for developing the interface system that allowed the astronauts to eject and activate the reentry vehicle. (credit: NRO)
As the GAMBIT 1 program was winding down with about six spacecraft left to launch, the engineering support work was just about over. In early 1967, I got that familiar greeting from an associate and asked to take a ride with him. About five miles from Valley Forge we pulled into a parking lot of a nondescript office building. He escorted me to the security office where I became a cleared employee of the Manned Orbiting Laboratory (MOL) program. I was then introduced to the engineering manager who had asked for my services. After a general overview, he told me they were going to include a CORONA-type SRV into the MOL design. My task was to define the crew controls necessary to prep the SRV for deorbit from MOL and any necessary changes to the SRV to interface to these controls. I was not involved with the major GE task of developing the crew controls to operate the camera systems.
After cleaning out my GAMBIT office, I began work in the closed area in the basement of the MOL building which contained about 30 other engineers. These were the days where almost everyone smoked. Not pleasant.
After a few months, I had a strange feeling that, except for my boss and a few others, no one was interested in my progress. There were no design reviews that I had been accustomed to on previous assignments. No customer contact. But this effort to have an SRV on MOL made perfect sense, to me, anyway.
The MOL flight plan was to have the two astronauts take photos for about a month, load the film cassettes and themselves into the Gemini, and then return to Earth. But what if, in the first few days in orbit, they photographed something very important that people on the ground should see? Do they cut the mission short, take the film, and fly the Gemini home? Not only would that have wasted an expensive mission, but it would have prevented the astronauts from continuing to take photos of whatever had been so important in the first place. Having an SRV onboard where they could load that film, get it back to Earth, and continue the mission made perfect sense. I was convinced my task was worthwhile, but I always felt that this idea was a derived requirement of GE’s and not an Air Force customer requirement.
I did an electrical design and laid out the crew controls required to ready the SRV for separation and deorbit. I remember that the SRV was mounted with the fore body facing the exit of the air lock launch tube during the entire mission. The thrust cone with its film chute was accessible to the crew. A film leader attached to the take-up reel in the capsule was also accessible. I don’t know how the crew was to attach the film from the cassettes to this leader, but my electrical design provided crew control to drive the take-up reel motor and move the film into the SRV. I don’t recall any discussions about the crew opening up the SRV. This would be a major task.
There was a modification I proposed for crew safety. This SRV was loaded with pyrotechnics. It was like the Fourth of July. With the exception of turning on the recovery RF beacon and the flashing strobe light, all events were pyrotechnically activated. The device that presented the most danger to the crew was the retrorocket. This was a solid propellant device (I recall 40 pounds [18 kilograms], but I might be wrong) with a pyrotechnic igniter. All the SRV pyros were fired using power from the recovery battery. I proposed a switch to electrically disconnect this battery during the MOL mission. The switch would be electrically closed by the crew after the launch tube inner door was closed and the outer door opened for SRV separation from MOL. Verification that the battery power was on was provided to the crew at the SRV control console.
We had a few visits from the MOL astronauts to our plant, but they were mostly for morale building of the GE team. None came to review this SRV integration effort. The first time I met a MOL astronaut was in 1985 (more on this later.) I was instructed to identify all our drawings with engineering sketch numbers and transfer them to print control for retention. I suspected that the whole SRV idea was scrapped, perhaps for cost control, as our Air Force customer was definitely under pressure here. In the last year of the program, we must have been asked for a “cost to complete” every few months.
When the GAMBIT program finally closed shop, we moved into their offices close to the main GE plant in Valley Forge. It was like coming home. No more closed basement work space. GE had built new facilities in the “orchard” near us to support MOL. Thousands of employees were working on this project.
My final task on MOL was to lead in the construction of the electrical system schematics. These drawings showed a high-level internal diagram of every electrical component in the vehicle. It also defined every electrical connection between these components. This is the source of information required by the wire harness engineers. These drawings are used by test engineers to identify test equipment interfaces. They are the main source of information when troubleshooting spacecraft system test problems. This was a major task for the MOL project. We were making good progress.
Then, in June of 1969, word spread like wildfire throughout all the offices. MOL was cancelled! Our work was over. Before the end of the year, two out of every three MOL employees were laid off from GE. How many families devastated by this event? How many lives changed?
I was lucky. Besides having a project manager that oversaw my MOL activities, I had a functional manager in the Electrical Power Engineering group in the main plant at Valley Forge. Joe Hayden had a contract with JPL to design the control electronics for a nuclear-powered spacecraft they were developing. Joe had a need for an electrical engineer on this job, so he moved me to his offices in the main GE plant. This spacecraft was to visit the outer planets as it flew out of our solar system. At the time, JPL called the mission the “Grand Tour.” You know it as Voyager. But that’s another story.
After 10 years, I was finally out of the black world. Now I could talk about my work to family and friends. It had been a privilege to be a part of this special group of dedicated people, but also stressful as well. I chose not to return.
To finish the MOL astronaut meeting story, in 1976, GE won the DSCS lll contract to build communications spacecraft for the US Air Force. This was the biggest and longest duration (about 30 years) spacecraft program in the history of GE Space Systems. I was asked to be the electrical systems engineer on this program. We launched the first few spacecraft on the Titan lll. We were then redirected to change to the Space Shuttle. STS had many safety design requirements for any hardware that went into the shuttle bay. We made many mechanical and electrical design changes and had many reviews at the Johnson Space Center (JSC) in Houston. We were on the flight manifest for the maiden flight of Atlantis scheduled for October 1985. A few months before, I was asked to do a briefing to the Flight Readiness Review team at JSC. This team consisted of NASA management and the full Atlantis flight crew. I had met some of the crew at previous meetings, but this was the first time I met the commander. Commander Karol Bobko, one of the original MOL astronauts, stood up and introduced himself. It took me 16 years to meet a MOL astronaut.
After my morning briefing, I returned from lunch with a NASA baseball cap I purchased at the souvenir shop. I sat down next to one of the crew members to listen to the secondary payloads briefings. Bill Pailes, the payload specialist, asked what was in my bag. I took out the hat; he took out a felt tip pen, signed the hat, and passed it on to the next crew member. All five members signed my hat which now resides in a display case in my home office.
I hope you found this article, written by an old greybeard, interesting. It was fun to recall these events that happened so many years ago. Thank you.
Note: we are temporarily moderating all comments submitted to deal with a surge in spam.