Stories about spacecraft? I thought you would never ask
When I worked at Spar Aerospace, we had consultants in on a long-term basis from Lockheed Missiles and Space Company who were the experts in the field of three-axis stabilized satellites to help us with the Hermes satellite that was launched successfully on 27 Jan 76. After all, LMSC had actually put some up that worked and that put them ahead of anyone else. On the last day one of the consultants was there, we took him out to the local pub (the other one was a holy roller who kept a Bible prominently displayed on the upper left corner of his desk and would never be caught dead in a pub).
1. The first satellite was a spin-stabilized one. There are solar cells all around so as it spins, there is no need to point an array – but only the equivalent diameter is active, rather than the circumference. But it offered a simple design that was used throughout the 1960’s and beyond. Imagine the left front wheel of a car: if you turn to the right, gyroscopic force will tend to add camber, meaning the top of the tire will move away from the car and the bottom will tuck under. Spin-stabilized satellites make use of gyroscopic force to maintain a stable pointing angle as determined by a star sensor that finds Cassiopeia and the position of the antenna that is pointed to the earth. In this era, most satellites were for communications and pointing accuracy was essential with the low-power transmitters in use.
In order to get the required spin, small thrusters using hydrazine as a propellant were used. They launched one satellite and proceeded to spin it. They got no signal. Not even telemetry. They checked the star sensor output. Lots of light but no edges. Finally it dawned on someone. The thrusters were pointing the wrong way. The satellite had stabilized with the star sensor pointing at the earth and the antenna pointing out into space. There was no way to correct the problem. They lost the satellite.
2. The second satellite was a three-axis stabilized one that always pointed a flat sheet of solar arrays to the sun, one on each side of the satellite, increasing the power available. These satellites were launched into synchronous orbit at a 22,000 mile altitude and appeared to stand still in the sky. The pointing angle of the satellite is controlled by an earth sensor that aligns the satellite with the edges of the earth to maintain a stable antenna position.
In those days, a program that went smoothly without glitches was something of a fantasy. Launch vehicles blew up on the pad. Stuff that was tested exhaustively failed once it was installed. There was plenty of trouble before launch as well. Electronics failed or did something different from what was expected. Parts that should connect together seamlessly didn’t. Integrating the satellite to the launch vehicle was something that was part mystery, part Rube Goldberg and usually took a lot of unscheduled schedule.
But Lockheed seemed to have a program that the gods smiled on. The electronics and optics all worked. The electronics did exactly what it was supposed to do. Integration with the Thor Delta (the same launch vehicle we were using) went off without a hitch. Certainly this program was blessed! All LMSC had to do was deliver it into orbit, despin it, acquire earth (using the on-board earth sensor) and show that it was stabilized and they could collect the $30,000,000 launch cost. At the time, a house cost $20,000, so this would be about ten times that amount today and would buy a community of 1500 houses – a good chunk of change in those days.
The launch went perfectly. The Thor stage with its nine solid-fuel rockets surrounding it lifted off and delivered the Delta stage with the satellite on it into a low orbit 120 miles up. The Delta stage delivered the satellite into an elliptical orbit with the perigee at 120 miles and the apogee at 22,000 miles. The satellite had an additional rocket that was fired at the apogee to circularize the orbit. It all went smoothly. They spun up the internal momentum wheel and fired the small pitch-axis thrusters to eliminate any residual spin, and waited for the attitude control system to stabilize. Nothing happened. The earth sensor did not appear to see the earth. They checked the telemetry. The earth sensor power supply voltages were in spec. The clocks for the digital logic were running. But it was like they were out in deep space. They were starting to get worried now. $30,000,000 and the reputation of the company were at stake. Finally, one program manager decided to look back over the preflight checklist to see if there was anything that would give them a clue. One box was not checked off:
Remove Earth Sensor Lens Cap
They couldn’t stabilize. They lost the spacecraft and their $30,000,000. That would be about $500,000,000 these days.
The CTC Spacecraft
Even if it has been a bad day, whenever I walk past the automotive paint section at Canadian Tire, that large rack of rattle-can aerosol paints, it always makes me smile. Many years ago, an organization I was with bought a case of twelve cans of Canadian Tire Corporation exhaust pipe enamel and sent it 22,000 miles away. If you are wide awake, you will realize something:
It is no longer on this planet.
I was called in to do two hours’ work on a proposal one winter Saturday at 9:00 AM – that miserable unpaid work that engineers have to endure in certain companies. But by the time I left at 11:00 AM, the story I had heard made the day worthwhile.
Several years before, we had been working on the Communications Technology Satellite, CTS, later renamed Hermes when it was launched on 27 Jan 76. While I was working on that proposal, I heard the real truth about how things were put together.
CTS was the first synchronous communications satellite with 200 watts output, so it could be received with an 8-foot diameter antenna. Modern satellite TV users have a 20” dish, but before CTS, you needed the equivalent of a radio telescope antenna, 30 foot diameter – not the thing for consumers. CTS was designed to broadcast into communities in the far north where the 8-foot diameter receiving antenna could be hauled up on the back of a pickup truck.
It was parked at 119 degrees west at synchronous altitude. To get there, it was launched on a Thor-Delta rocket with the Thor stage taking it to a low orbit of 120 miles and the Delta stage sending it into an elliptical orbit between 120 miles and 22,000 miles. At the apogee, a solid-fuel rocket on the satellite itself burned for 28 seconds to circularize the orbit.
This was getting pretty close to the limit for the Thor-Delta. It can lift 1550 pounds into synchronous orbit and the satellite was slightly over 1500 pounds. The rocket on board the satellite was the most beautiful piece of spun stainless steel I have ever seen. It had a large combustion chamber about 4’ diameter narrowing down to a convergent-divergent nozzle and the whole piece of metal stood about 4 feet high. It was 0.100” thick and this had raised a concern: toward the end of the burn that the nozzle would distort due to the heat and the satellite would not be able to reach synchronous speed. The rocket engine in the satellite was 723 pounds of which 692 pounds was propellant.
Someone suggested the simple idea that painting the outside surface of the nozzle with a flat black paint would enable it to radiate enough heat to retain its shape, so we contacted a supplier of space-qualified paint. Now it should be noted that satellites are trailing-edge technology. Nothing new ever goes into a spacecraft without an arduous testing process. Everything has to have an established reliability and preferably years of use before it can be put on a spacecraft. Every part, chemical or process has to be qualified by being subjected to rigourous qualification, or “qual” testing and until the entire suite of tests was over, it was never a slam dunk; parts could pass or fail qual. A few companies specialize in space-qualified parts but normal military testing (as done back in the 1970’s) was usually sufficient if the highest grades of components were used. (Nowadays, most satellites are built with commercial parts – but it was unthinkable then to risk a $62 million launch on parts with no history.) Anytime you wanted to add anything to the list of qualified items, you had to fill out a NONSPAR – a non-standard parts approval request and NASA would pass judgment on it. That was an arcane process that kept us on tenterhooks until it was complete.
We called up one supplier of space-qualified paint and his reply was, “Sure, we can give you immediate delivery. How many tank cars do you want and where is your railway siding?”
So we found a polite way to tell him no and called another supplier, who said, “Pints and quarts, sure, no problem. Fifteen months delivery.” This meant they had not qualified or even formulated the paint yet and there was a definite possibility that the paint would fail qual and leave us with nothing.
So one of our engineers came up with an obvious idea – what about automotive exhaust pipe enamel? We went to the nearby Canadian Tire store at Dufferin and Castlefield in Toronto and bought a case of 12 cans (which were $2.67 a can at the time). We did not use the spray nozzle on the can – that would be too ghetto for any spacecraft – we had a very good De Vilbiss spray system, similar to what would be used for refinishing a car. We established that the paint had to be really roasted onto the metal at high temperature or it would not adhere. But once it was baked on, the adhesion was great. It would not flake or chip when tested for scratch resistance or discolouration under salt spray, the bane of anything stored near the ocean at Cape Kennedy.
Canadian Tire exhaust pipe enamel was qualified and flown on the CTS spacecraft. But before that, we had the problem of writing up the NONSPAR. We used the stock number which at the time was three numbers followed by a dash and three more numbers and listed it as “Paint, CTC Corporation” in the description and we buried it among other chemicals we were getting approved. NASA didn’t ask any questions. We certainly didn’t volunteer any information. According to the rules, only the paint from that particular case was qualified – approval is granted only for a certain manufacturer and run and there was no way to determine if any other paint was the same formulation or even came from the same company, since it is a private-label brand.
But in a world where product quality is a sore point for buyers of many items, isn’t it nice to know that you can purchase space quality products without even knowing it?
Hermes Post-Launch
Apogee motor firing went off perfectly and the satellite remained in its position at 119° west synchronous orbit as planned. CTS remained in service from the launch date of 27 Jan 76 until its demise due to earth sensor failure on 24 Nov 79. Since it had a planned two-year lifespan, the project was a success. That doesn't mean everything went smoothly. We had some component failures which were mitigated by redundant systems.
You have to understand the psychology pre-launch. To a man, all of us were saying, "I hope it blows up on the (launch) pad." None of us had any confidence that his own system was ready for prime time.
The earth sensors were linear photodiode arrays which were part of an optical train that included a mirror mounted on flex pivots (crossed leaf springs that formed an "X" shape) with the mirror pivoted and vibrating at 4.4 Hz. The length of the earth was known in pixels so the prime earth sensor built up a measurement in the direction of vibration (along the equator) had a less accurate measurement was in the latitude direction based on the total length of the image. The start of the bright area was one edge of the earth and the end was the other, so the prime earth sensor had very good accuracy and resolution in the longitude dimension and not so much in the latitude dimension. Someone had the bright idea that the flex pivots would possibly be a problem, so the redundant earth sensor was mounted at right angles, so the flex dimension corresponded to latitude and the image size dimension corresponded to longitude. It was an inspired decision – the prime earth sensor failed due to flex pivot breakage immediately after takeoff and we ran on the redundant sensor until it failed due to flex pivot breakage on 24 Nov 79. Within five minutes on its final day, the main antenna had drifted out of reception range and we were left to communicate via the omnidirectional TT&C (telemetry, track and command) system. But with the 1200-watt solar arrays moving out of alignment with the sun, the main 200-watt transmitter was shut down for the last time.
This wasn’t the only problem. In synchronous orbit, you run into at least partial eclipse 84 days of the year and full eclipse 72 days of the year. Since we launched in winter, we didn’t see any effect until a couple of months later when we got the word: the prime power converter had failed. The solar arrays were two fan-folded structures about 21 feet long on each side of the spacecraft. They were carefully designed so that all magnetic torques were balanced and they were gathered in series-parallel into 78-volt buses to feed the converter. When a solar cell gets cold, the voltage goes up and it appears that just before the failure, the array output reached 130 volts. The arrays were being cooled to -50°C in eclipse. From that time on, we went with the redundant power converter and shut the array output off for about eight minutes after coming out of eclipse. The electronics could run off the battery supply for another few minutes per day and the array had warmed to an acceptable temperature (and voltage) by then.
On day 100, I was met at the door as soon as I walked into the building. A colleague said, "We went into AFP last night." This was chilling news for me. Backing up a bit, I was REE (Responsible Equipment Engineer) for the Attitude Control Electronics Assembly. For political reasons, we had to break a cardinal rule of spacecraft: "You never buy a guidance system – ALWAYS build it". Being REE is the same as being the goalie in hockey – you are the last line of defense against a bad design. And RCA had done bad designs, so bad that we forced them into two intermediate design reviews between PDR (preliminary design review) which is approval of the conceptual design and CDR (critical design review) which is the authority to begin build of the qual units. I was blissfully unaware of the political and career carnage my design reviews had caused, but they had saved a lot of schedule by making sure everything sailed through qual without a hitch. One of the things added to the guidance system was an AFP (automatic failure protection) mode where the pitch momentum wheel was held at constant speed and reaction control thrusters were shut off if anything drifted too far off course or telemetry found a problem. If you went into AFP, something was very wrong. We were on tenterhooks for hours as the news slowly trickled in: an experimental solar array section had blown out and the force was enough to throw the spacecraft off alignment. BIG SIGH OF RELIEF! We powered the attitude control electronics back up and it acquired earth and remained stabilized for the rest of its life.
