Evolution, incumbency and risk


Monument to the conquerors of space Moscow (Copyright)

Engineers as the agents of evolution

Recently I had the chance to visit the Monument to the conquerors of space when I was passing through Moscow. The monument itself is impressive, in all its snow clad titanium majesty as was the long and patient queue in about -14C. If you’re in the neighbourhood I recommend it as a great way to spend half a day, the museum not the queue.

But what was for me the best part was to actually look at the technology and reflect a little bit on how different cultures and organisations can come up with quite different approaches and what this says about risk and how systems evolve. Looking at the crew couches in the Soyuz module you notice immediately that they’re individually mounted on a set of rugged shock absorbers. That’s interesting given that the Soyuz is designed for a hard landing because it reduces the amount of mass that needs to be shock protected as compared to using external legs for the whole vehicle. A not so small side advantages is that you need to worry less about landing orientation, leg deployment failure is not an issue, you can land at a higher speed, and so on.

At the vehicle level shock reduction is achieved by dropping the impact speed with retro rockets, introduced with the Voskhod capsule, rather than vehicle energy attenuation systems. Of course the Soviet designers could incorporate a retro-rocket system because their boosters could haul a large payload into orbit, around 5 tons with the original Voskhod as an example. As to the original decision to use retro-rockets? That was prompted by the decision to go with a three man crew for the Voskhod spacecraft, as the crew could no longer eject from the spacecraft the designers needed a way to slow the vehicle sufficient to protect the occupants, and a retro-rocket pack fitted the bill.

Now NASA’s approach was quite different, they performed a number of studies looking at the g loads that crew would be exposed to during hard landings and decided that as the crew could be potentially disoriented by the shock loads, water recovery would be the way to go. As they were playing catch-up with the Soviets program, and mass limited by their early ICBM based boosters (Redstone, Atlas and Titan), chute recovery into water seemed the simplest, least risk, quickest and most lightweight recovery option.

Given their mass limitations the Americans also tended to incorporate shock attenuation into the vehicle structure, rather than paying a mass penalty for a retro-rocket system, with the Mercury capsule having a deployable airbag while the Apollo capsule had collapsible ribs to attenuate shock impact, although on Apollo the Americans also included a one shot shock absorbing strut arrangement into a palletised couch design. From the start it seems American spacecraft were designed to be lightweight, single use devices.

Of course the difference between the design teams recovery concepts also reflected the difference between the launch sites of the two countries, the American launches were over water with the likelihood of an abort landing into the ocean, while the Russians launched over land with a similar likelihood of an abort landing to land.

The Russian Soyuz capsule design also jettisons the heat shield once it’s done it’s job which reduces the work that the combined chutes and retro landing system has to do. Of the American capsule designs only the Apollo capsule jettisoned a heat shield, in that case the forward heat shield, needed to protect the crew capsule from the higher reentry temperatures of a lunar transfer orbit, the rear part of the heat shield being retained. Of course if you intend to re-use the reentry vehicle, as the Russians do, then jettisoning that burned out shield makes sense, and you can tuck the retrorockets away underneath. While if you don’t, as the Americans didn’t, then not so much.

What’s interesting is how a set of slightly different initiating conditions led to quite different design forms of solutions, and also that both design teams stuck with their basic forms once decided, it seems that in engineering historical contingency plays as strong a part as it does in evolution. Complex systems like spacecraft are built on the hard lessons of engineering failures, so once a design strategy and architecture is successfully established it’s very hard to change it because to do so would introduces new uncertainties, the prospect of which tends to keep engineers up at night. Thus it’s highly unlikely that either Russians or Americans will give up their current reentry strategies for capsules, see for example Space X’s election of water landings for its Dragon spacecraft even though dropping a reusable space vehicle into salt water every flight makes much less sense.

In an evolutionary sense this parallels what we see in the biological sciences, where evolution seems to pass through an initial stage of ‘wild’ experimentation until a dominant and successful form, or forms, emerges which thereafter evolve more slowly and in a perfective and incremental sense.