Wednesday, February 9, 2011
The Politics of Design
The entries are in for the second round of NASA's Commercial Crew Development (CCDev) program. Years from now, if this program is adequately funded and pursued, it will no doubt go on to produce some of the most cost-effective investments in NASA's history. CCDev's aim is to nurture a private-sector human spaceflight industry capable of delivering safe and inexpensive (well, relatively inexpensive, anyway) travel to Earth orbit to anyone interested. If we're serious about developing an enduring human presence in space (and I am), this is the road that will take us there, and Barack Obama has been the best president for NASA since Kennedy for kick-starting this development.
Like any good private-sector competition, bidding is open to just about anyone. There were some predictable entries like SpaceX's proposal to accelerate crewed Dragon development, and some less orthodox ones as well. By far the biggest oddball was ATK and Astrium's proposal for a mash-up of the Ares I and Ariane V first stages. I allow the following video to explain the concept as generously as possible:
There are a number of odd things, inaccuracies, and bogus claims in that video that should be addressed:
-Why is the International Space Station shown thousands of miles away from the Earth? Look at any picture of the ISS ever taken. Seriously. Any picture ever taken. See how big Earth is? That's where the ISS lives. ATK is an aerospace company that works with NASA every day, and goofing a visual like this is unprofessional in a bizarre way.
-ATK and Astrium have a long track record of successful missions. Fair enough. But the stages they're proposing for Liberty are radically different beasts than the Space Shuttle SRB and the Ariane V core stage. The four-segment Space Shuttle SRB has virtually nothing to do with the five-segment SRB developed for Ares I, since the geometry of the motor case is completely changed and the load paths are so different. Liberty's upper stage would have to be air-started and survive the punishing vibration and bending loads applied by the first stage, requiring enough analysis and redesign from the stock Ariane V first stage that it would be practically a new vehicle. To imply that the flight heritage of the shuttle and Ariane have a one-to-one mapping with Liberty reliability (as the animation showing shuttle and Ariane components morphing into Liberty does) is at best overoptimistic.
-We are told that the Liberty configuration is the safest and most reliable orbital launch vehicle design available, since it has two engines with one engine per stage. So do Atlas V and Delta IV. Falcon 9 has nine engines on the first stage, but I'm not convinced this automatically makes it less reliable. A nine-engined vehicle can lose one or two and still make it to orbit, and engine-out capability has mattered in the past. Regardless, the specifics of the design and its flight history are probably more important to reliability and crew safety than the high-level configuration, so this is a bit of a stretch to label it "safest and most reliable" on such loose grounds.
-Including the chart of launch abort effectiveness was a terrible decision. This is just wrong. There has been much debate about this at the NASA Spaceflight forum, and credible sources there have said in no uncertain terms that Ares I (and by extension Liberty) would have a less forgiving abort environment than a current launch vehicle like Atlas V or Delta IV. The blackout zones shown on the chart for "other vehicles" can be closed with a minimal loss of payload, and a liquid-fueled rocket engine can be shut down benignly before a problem becomes catastrophic. Since solid-fueled engines can't be shut down once ignited, an abort requires detonating the first stage explosively, and the launch abort system has to outrun the many heavy flaming pieces of metal capable of burning through crew capsule parachutes that result from such an event. The notion that solids are more inherently crew-safe than liquids is a myth that just needs to die. All else equal, you're safer flying on a liquid-fueled vehicle, full stop.
-I thought it was interesting that SpaceX's Dragon wasn't included in the compatible spacecraft. Clearly Liberty would have enough lift capacity to launch it, but it seems like ATK just doesn't want to talk about smaller space companies like SpaceX, instead focusing on Boeing and Orbital Sciences. Also, I don't know what they're getting at with Orion. Ares I was designed to launch Orion, but continually failed to meet its design targets for launch weight, requiring weight-loss programs to make Orion launchable by Ares. Liberty has less lift capacity than Ares. Are they talking about a trimmed-down, orbit-only version of Orion? It would be nice to have some clarity...
-The video makes the case that development costs can be reduced by using NASA facilities at the Kennedy Space Center. Very well, but who will pay for the facilities? Will ATK pay to rent out the parts of KSC necessary to launch Liberty? That would be prohibitively expensive. Will these operations be paid for out of NASA's operating budget? That would be an unfair subsidy to ATK and would obfuscate the real cost of operating Liberty. This is not a convincing selling point, to say the least.
-The go-fast stripes on the first stage look very pretty. I have no rebuttal to this fact.
Outside of the video, ATK claims that Liberty could be ready for its first launch by 2013 and will cost about $180 million per launch. This is incredible. Ares I has already spent four years in development the earliest it could realistically fly according to the Augustine Committee was 2015. It was unlikely, according to the committee, to ever be more cost-effective than the Atlas V. Liberty is, for all intents and purposes, a brand new vehicle, requiring many man-hours of redesign, development, and testing. If ATK and Astrium want to sell this vehicle for $180 million per launch and take a loss, that's their business choice, but I can't imagine any world in which Liberty would be profitable flying at that price.
The reason this proposal exists, then, is not because of the technical or business merits of the concept, but because of the politics surrounding ATK, NASA, and the entire aerospace industry right now. ATK runs a very profitable business assembling solid rocket segments into SRBs for the Space Shuttle program. This is good for the company, good for the state of Utah (where ATK is based), and good for NASA since it spreads the political footprint of its contractor base a little more. While Ares I was in development, it looked like this business could continue by transitioning to the assembly of solid rocket segments for the Ares I first stage. When Ares I was effectively given a vote of no confidence by the Obama administration last February, much rage ensued. Liberty is ATK's last punch in the fight to keep Ares I and its orbital solid rocket business alive.
One of the great things about going to school in Texas is that you're never far from the resources and experience of the Johnson Space Center in Houston. I've had the chance to talk to a number of people at JSC about Ares I and the design choices behind it, and I have yet to meet anyone who's actually worked on a project related to the rocket and come away thinking it was the best technical solution to problem of sending people to the Moon safely and reliably. It's first and foremost a politics machine, not a a physics machine or an economics machine.
Ultimately Ares I, and its Liberty echo, are solutions looking for a problem. CCDev is about spacecraft, capsules and spaceplanes to get people to their destinations in orbit and down to the ground in an orderly fashion. We have launch vehicles already, wonderfully reliable ones with proven track records in the form of Atlas V and Delta IV. I'm sure Falcon 9 will join their ranks soon once it has a dozen or so cargo transport flights under its belt. On a level playing field, Liberty will lose this competition because it doesn't address what NASA and private human spaceflight need. I hope the playing field really is level.
It brings me no joy to attack a new idea about crew transport to low Earth orbit. Rockets are just really cool and a robust human spaceflight industry is something I desperately want to see become reality. But I don't see this ATK proposal fitting into that reality. There have been many bad ideas in the history of human flight, and some of them sank billions of dollars and many years of precious engineering time before fizzling out. Liberty is a bad idea and the CCDev money belongs to ideas with more promise.