In what can only be described as having extremely high aspirations by competing in the Big Boys league, Omnipless tenders on and is awarded a contract perhaps under false pretenses. This was SHUCS or SpaceHab Universal Communication System with the ultimate customer being NASA.
Much of what they committed to would require extraordinary effort. With little capital or knowledge about engineering equipment for conditions in space, any snags, glitches or deviations from plan could financially cripple Omnipless.
This is the story of a bunch of intrepid South African engineers who beat the odds and met their target albeit by a narrow margin. It had been a close-run thing.
Main picture: SHUCS finally in space: The antenna is in a roughly horizontal attitude on top of a positioner, on top of a Spacehab module at the back of the Shuttle cargo bay.
SpaceHab Approaches Omnipless
(SHUCS – SpaceHab Universal Communication System)
A company called SpaceHab (Space Habitat) had a brilliant idea. There was a problem with turning around the Space Shuttle. After return from orbit the Shuttle first had to be refurbished before any integration of the next payload could be done. SpaceHab’s concept was to make a module in which the next mission’s experiments could be installed and integrated separately from the Shuttle and when required, the module could be quickly fitted in the Shuttle cargo bay with only a few connections to be made and a minimal number of integration tests done. Their problem was that radio communication was abysmal – completely analogue and there was a satellite missing so that there was a gap in the coverage during orbit. In fact NASA could barely transmit system updates let alone provide capacity for anything else. Continuous, real time communication with experiments was not possible. SpaceHab decided that they would use the existing Inmarsat satellite network and stuff NASA. On surveying the field they realised that Omnipless’s field portable antenna offered everything that they needed in terms of mass and ruggedness. No one even came close. The only problem was that we had to qualify it for space.
They approached Johan who was agreeable. He wished to break into the aerospace market in the future and thought that this would give us street cred when the time came. With his customary chutzpah, (this boertjie from Belleville knew no limits) he said easy peasy, we have space experienced guys. That was me. I had designed stuff for SA’s recce satellite in the bad old days and made some prototypes but never tested them before the wall came down on our cosy military industrial complex. So we did a dog and pony show in Tellumat’s impressive boardroom, showed them the production and test facilities at Tellumat. Meanwhile we actually worked out of a house on a plot outside Hout Bay. We assured them that the specialist space tests could be done at the Houwteq facility which had actually been mothballed and only had one security guard and his dog and 2 or 3 chaps who kept the EMC (electromagnetic compatibility) facility running. I had 6 months and kR200 to change the design, qualify it to NASA’s satisfaction and deliver 3 antennas to the States.
The Pad: New improved Development Facility. This was on another smallholding down the road from the first facility. Space at last. The house was reputed to have been bawdy house hence Johan’s name of The Pad. It came with a huge horizon pool (right)and nice views across the valley – great.
Stoep of The Pad overlooking the Houtbaairivier Valley. No I didn’t wear my Bart Simpson tie to meet with SpaceHab
The SHUCS Progenitor
The basis for the SHUCS antenna was the flat panel TES (Transportable Earth Station) antenna developed for the Inmarsat-B system. It measured 774mmx774mmx15.5mm thick and weighed 2.4kg. This was by far the lightest on the market. It comprised 4 separate panels that clipped together enabling it to be broken down into an extremely compact package that could be carried as hand luggage. Each panel was a sealed unit and electromagnetic signals were coupled capacitively between them making it extremely rugged and reliable.
The internal construction was a patch antenna array on top of a stripline feed network trapped between 2 foil ground planes.
The Inmarsat-B TES antenna showing its modular construction
Redesign for SHUCS
The first problem with space is the outgassing of materials in a vacuum. Most materials give off a chemical mist that makes a cloud around the Shuttle and ends up condensing on optical lenses etc. None of the polystyrenes and glues that we used were suitable. The first problem was then to replace our materials with low outgassing versions that weren’t lossy to RF (radio frequency.) The polystyrene problem was solved by importing a sheet (at kR10 a pop) of specialised Rohacell foam from France that had to come by sea and chewed up 3 months of our timeline. The glues were replaced with specialised epoxy which were used incredibly sparingly as it is very lossy. Special space qualified spray paint had to be procured from USA.
The next was the temperature problem. We were given various orbital scenarios. The worst ended being one where the Shuttle would be sun synchronous for a number of hours at a time so the front of the antenna would be facing the sun and attain a barmy temperature of around 25degC while the back would start trending to -273degC. This seemed intractable until I vaguely recalled reading about radiation shields from 20 years previously at varsity. Essentially, if you have a hot plate facing a cold plate then, say for example, 100W of heat is transmitted between the 2, the heat can be dramatically reduced by inserting another plate in between. The heat would be disproportionately reduced because radiation heat transfer is proportional to the 4th power of the respective temperature differences. So maybe less than 20W would transmitted. I then realised that the very nature of our antenna construction actually was just this. The core of the antenna was the radiation shield because of the ground plane foils. To give our antenna structural strength, it had to made thicker (30-40mm) so the front and back skins were separated by space grade honeycomb from the antenna core. Problem solved.
The next issue was the structural strength. We were only allocated 4kg for a roughly 1m square antenna. Unfortunately they specified that 4 space qualified GPS antennas had to be fitted to ears on the corners and these weighed 200gr each and were in the worst possible place as the antenna was mounted to a positioner via 6 bolts very near to its centre. So we were only left with 3.2kg and had to stop these incredibly heavy (relatively speaking) GPS antennas from vibrating off under the vibrations to be experienced during launch. A further complication was the acoustic loading. The antenna had to be tested for its ability to withstand the sound pressure levels at launch which are rather high and the antenna is a huge flat plate so it catches a lot of it.
It must be noted that a good friend of mine, Louis van Wyk, was subcontracted to do the structural analysis and manufacture.
The structural design was a matter of computer aided stress analysis and required the use of the thinnest fibreglass sheets, front and back, to meet the mass requirements. The main problem was one of construction. The antenna core comprised Al foil sheets bonded to machined Rohacell cores and then honeycomb was bonded to the outside and skins bonded onto this. As epoxy is very lossy, only the barest minimum of epoxy could be used and the shiny foil degreased to the nth degree and the shine broken. Vim and a lot of distilled water worked very well. The epoxy was minimised by only using dots applied through a silk screen. Unfortunately, we only had sufficient Rohacell for the prototype and the required number of deliverable antennas. Oh shucks, calamity. Murphy is never far away. The first antenna core crumpled up in the vacuum bag during autoclaving. This meant that we didn’t have enough material for all the antennas and the lead time on delivery of more stock was untenable. Johan was virtually apoplectic. But the subcontractor saved the day. He had kept every offcut and one antenna that we delivered was made of a mosaic of foam pieces.
The RF redesign was minor as only small adjustments had to be made.
SHUCS Antenna with 4 GPS antennas mounted on ears
Acoustic load test
Houwteq had an acoustic vibration facility. This was a large room that was separately constructed within their test area. It was about 10m tall and about 4mx4m in plan. In the corner was a huge air horn with about a square metre outlet which was driven by a monster compressor outside. The whole structure was separately founded on the bedrock and separated from the rest of the building with a rubber gasket around its perimeter. Enquires with ex Houwteq personnel revealed that special provisions had to be made with Eskom in order to run the system. We placed the issue with Spacehab. After consultation with NASA, it was decided that we could replace the dynamic test with a quasi-static test whereby a dead load could be placed on the antenna. After increasing the maximum sound pressure load by a certain amount to force an equivalent damage level and adding in some fudge and safety factors, they came up with a load of 310kg! Moreover this had to be done 4 times. To demonstrate dual redundancy, we had to first use only 3 of the bolts and then repeat using the other 3 bolts. We also had to do the test with the load on the front face and then turn it upside down and repeat with it hanging from the same 3 bolts. The only problem was that we had already completed our design and were building antennas by the time this number was mooted. We were quite sanguine about it as the company motto was never say die.
The funny part was the testing. SpaceHab sent one of their guys around to monitor and assist during qualification. Luckily he was down to earth and practical and appreciated our boer maak ‘n plan methods. The load to be applied had to be a dead load and evenly distributed. This can be accomplished via sandbags. When the time came to do it, I thought, wtf, I’m not filling sandbags. I have to get the sand, organise suitably sized small bags and fill them a trowel and close them up somehow. Light bulb moment. I tootled off to the nearest PnP in Tokai and cleared out their stock out sugar and then I started on the flour. It was only a small branch then so I only managed to get 2 trolley loads – about 250kg. Each bag was duly weighed on a calibrated scale and gently added to the antenna. By the time the bags had been loaded, enough layers had been built up and then we scrounged various aluminium jigs and fixtures and added them on top. The results were recorded and the test setups were photographed with an early digital camera (about 1M pixels) and added to the qualification report and signed off by the SpaceHab representative. Job done and the sugar and flour were donated to the production staff. During a conference call to NASA to review our qualification tests, they were tickled pink by my approach.
Adding my 75kg to the test before scrounging for Al jigs and bits and pieces (actually I had taken it far past the test spec by standing on it but what the hell)
Quasi-static acoustic load equivalent test: Poor little SHUCS antenna at the bottom of 310kg of sugar, flour and Al Jigs
NASA was also amused by the presentation of the vibration test results. Tellumat had a great vibrator but the data capture system was from the Ark. The vibration analysis was done by a processor that preceded pc’s and there was only one person in Tellumat who could sort out its issues. Unfortunately, due to time pressure, the test was run after hours and we could not get the display on the spectrum analyser to be transferred to the plotter. The solution was to stick post it notes onto the screen with various annotations highlighting important features. The screen was then photographed and published like that in the report. Again Nasa thought it unusual but perfectly acceptable.
Me assembling SHUCS onto its positioner prior to vibration testing
SHUCS on its positioner mounted to a vertical axis shaker at Tellumat. All other orientations of the positioner and angles of the antenna still had to be tested. A lot of work.
The testers arsing around while burning the midnight oil at Tellumat. This little camera was not only good for selfies (nearly another first), but also saved our bacon when the printer wouldn’t print. The control panel of the antediluvian vibrator can be seen in the background. Popeye is Clark, the SpaceHab observer
Thermal Vacuum Test
The most ridiculous test was the space vacuum test. Houwteq had the only facility in SA. Unfortunately, it had not been used in anger for many years and only a few times when Houwteq was alive. We tracked down a chap who had worked there. He agreed to come through from Napier or Bredasdorp where he had probably retired and try to fire it up. The vacuum chamber was a stainless steel cylinder – about 4m diameter and 5m deep. The door was like a bank vault door that you see in the movies complete with a wheel underneath it to take stress off its hinges. The only way to heat or cool things in a vacuum is via radiation. We were required to qualify from -140degC through to +80degC. For the cold side there are radiators inside through which liquid nitrogen is expanded at temperatures around -190degC. For the hot side conventional electric radiators are used. Space vacuum isn’t for sissies. As I recall there are 3 levels of pump. Conventional vacuum pumps, turbo vacuum pumps and then something akin to molecular blotting paper. We had no idea whether the control system would work or if all the seals were still 100.0000%. The slightest leak would destroy any chance of pulling a space vacuum. Our man from Napier checked it out and said we were good to go – phew.
Overkill: 3 small antennas (face down on the anechoic foam) being loaded into the ridiculously large thermal vacuum chamber at Houwteq (L-R: Clark, Ian George, me)
Next problem was to get the liquid nitrogen. For the amount we needed, Fedgas promised to send me a truck from Gauteng as they didn’t have enough in the Cape.
The day arrived along with 20 tons of N2 in a truck that was connected directly to a valve outside. Wonder of wonders, everything worked. We ran the test through the night. By the time we disconnected the truck the next morning we had used roughly 10 tons of N2 to test 3 antennas!
The gas truck and the deserted multi million Rand Houwteq space integration facility
Job Done: SHUCS goes into Space
SHUCS antenna being fitted on top of the Spacehab module. This whole module is then fitted into the cargo bay of the space shuttle
SHUCS finally in space: The antenna is in a roughly horizontal attitude on top of a positioner, on top of a Spacehab module at the back of the Shuttle cargo bay.
What absolute cowboys, but what fun. It was all done on a wing and a prayer but unheralded in that we were the first South Africans to have a product (apart from Pratley Putty) fly in space without anyone knowing about it. Unfortunately, SpaceHab’s test was a real disappointment. When they tested the system in space, their diplexer must have had a small leak and shorted out. However, Inmarsat did report that they had received signals during their handshaking protocols before it blew. Testing back on the ground showed our antenna to be still operational so it was accepted that we did our bit. It’s a pity, but SpaceHab went out of business and their great idea was stillborn.