Wayne Hale has an interesting excerpt from a recent speech by Admiral Gehman on the Columbia loss. I may incorporate this into my space safety book. Hard to believe that it will be ten years in February.
17 thoughts on “NASA’s Safety Culture”
Comments are closed.
This is really just the public vs. private debate. The reason NASA made a decision to kill those people (make no mistake about that) was that it would have no real repercussions. If a private company kills people it likely will put them out of business. This helps to focus.
That can happen with private companies, too. Just look at the Titanic, the Ford Pinto, or the Concorde. It’s just that the NASA culture was worse, since the Shuttle had been designed and built under the same pressures to shave the margins to meet time and budget constraints. I suspect they learned not to raise issues unless they had a workable solution in hand, and for some issues there just isn’t a workable solution, like major tile damage, no abort system, etc.
They developed an ablative tile replacement that could be spread almost like Bondo to replace a mssing tile, a project they abandoned in 1979 but which was eventually successfully tested in space. I don’t know if that became essential equipment, since it’s much easier to just not look at the tiles and keep your fingers crossed.
When a private company goes out of business, they can’t screw up anymore. The government never goes out of business and the screw ups just end up in other departments.
Screw-up companies don’t always go out of business, though. Sometimes they get bailed out, sometimes there are customers who through stupidity or lack of choice keep them going.
Banks are an obvious example. Airlines are another: despite decades of subsidies, there are usually some on the edge bankruptcy.
Of course, but if bailed out that’s usually other people’s money. Business people buy off the assets of failed companies which is what’s supposed to happen.
Sold your stock in British Petroleum, eh?
In all fairness, few people thought that a foam strike would be a serious problem for the leading edge RCC. Even after the accident, it took some time and some hard tests to convince everyone (including me) that such a thing had happened. As a matter of fact, my personal hypothesis (based on Iraqi celebration of the event) was that someone in the OPF had substituted plastique for the RTV normally used as a gap filler. Doped with something like lead stypnate or sodium azide, it would survive boost aeroheating without a problem, but then detonate (with tile removal) on entry.
It took a lot of testing to confirm that the foam problem was, indeed, a problem…
You can’t think of everything.
Oh, yes. That was before I became a Left-wing Democrat. Now I believe that we deserved Columbia, no matter how it was caused. Only tax increases on the rich will begin to compensate…
I think you mean that it is a mistake to have wasted that money on NASA when there are so many poor and starving old/young here on earth that we must feed before we can even begin to think about polluting the sky with rocket exhaust.
Oh, and we ruined Earth let’s not ruin the rest of the solar system.
I and most other folks in the JSC organization I worked in at the time of the Columbia accident (won’t name the org for personal reasons, but it is one of the most respected orgs on-site and many high-ranking NASA managers have come from it’s ranks) knew nothing or very little about RCC impact damage resistance in late January and early February 2003. But as soon as I heard the estimate of the impact velocity and mass of the piece of bipod ramp foam, I calculated the kinetic energy and my immediate gut feel was: That’s a lot of kinetic energy and I doubt any non-metallic material could survive such an impact. When I heard others in my org say they doubted a piece of ET foam could damage a large, convex-curved piece of RCC, I told them the kinetic energy and each of them immediately said, #&!!!#%%!!!!, that’s NOT GOOD, and became believers that the foam could damage the RCC, just on gut-feel. This was within a week or so after the accident and long before the impact tests at SWRI that finally cracked an RCC panel.
How the MMT members and the supporting engineers (all very smart, experienced, and common-sense-filled persons), who must have known estimated impactor mass and velocity earlier than I did, convinced themselves that RCC could withstand such an impact without critical damage, is something I’ve always wondered about. Perhaps another example of the NASA organizational behavior and thinking Admiral Gehman described.
Well, depending on how you think about the problem, you can make some assumptions and visualize it in ways that don’t make it sound like such a severe impact, focusing on the way the foam won’t hit as a solid object and will compress and deform to lower the peak forces, etc. It migh’ve gone something like this:
The energy is about the same as a 750 grain .50 caliber bullet at the muzzle, but it’s not a very dense impactor and it’s hitting a curved surface so the energy isn’t delivered all at once, and it is distributed over a fairly large area. So if the stopping distance for the foam was six inches (compressing and wrapping around the LE) and you assume constant acceleration, you’re only looking at about 31,000 pounds of force, If that was distrubted over a 5″x5″ area, it would only be 1250 psi. So for a while you can think that it’s not too bad compared to material strengths, and after all, it shouldn’t be worse than a bird strike.
As long as you don’t think of it in terms of someone taking a 6″ diameter pneumatic ram at 1250 psi to the leading edge, or imagine a small armored vehicle parked on the leading edge, or note that the pressure is a couple orders of magnitude greater than the stagnation pressure at max Q, you can imagine that the impact shouldn’t be too bad. I assume that’s what they did for a while.
I’m not sure that reflects an organization problem, rather than the steps we naturally go through to gain insight into a situation we haven’t looked into previously, before someone comes in with some hard design calculations or the results of a test. I think the organization problem was that nobody ran through the calculations on a LE impact or did a test when the foam issues first surfaced. Tiles were being damaged, and those were from grazing impacts at shallow angles, so what would happen if the impact was square on? What are the critical areas that could be hit? Maybe someone did ask those question and didn’t like the answer.
George,
The scenario you describe about considering a foam impact being much different than a pneumatic ram impact, at the same kinetic energy, is pretty much the same thought process the STS-107 MMT went through, from the record of the MMT meetings and discussions I had with lots of people after the accident. After the accident I never had an opportunity to talk with the STS-107 MMT chairman or the Program Manager at that time, so I do not know their true reasons for going with the “rosy scenario” that a supposed RCC expert presented to them, and for turning-off requests for photos from ground assets.
The tipping point was STS-112, when a similar piece of foam came off the same location on the ET, and hit and DENTED a metal cover on an externally-mounted left SRB avionics box. I cannot remember the thickness of that metal or the size and depth of the dent, but why the MMT in January 2003 decided to not pay attention to the STS-112 damage to a metal component by a piece of foam, I do not know. But before that, as Gehman and the CAIB Report said, the management decision to not open an in-flight anomaly for the STS-112 incident was the missed opportunity to pause and determine if the Program really understood the effects of high-speed impacts of large pieces of foam on all critical components of the stack, especially the Orbiter TPS. All sadly reported in the CAIB Report, and re-iterated in the Gehman speech that Hale’s blog references.
Wide-spread organizational head-in-the-sand behavior or not? I’ll have to leave that call to psychiatrists and other people who know more than me about behaviors of individuals and organizations under pressure, when careers and lives are possibly at risk.
One factor that I’ve always wondered about was the nature of the foam at impact. The foam was likely very cold since it was used to insulate a cyrogenic tank. Some materials become harder when extremely cold and others become very brittle. If, as I suspect, the foam became harder, then it could’ve done a lot of damage.
That’s a good point, Larry. I hadn’t thought about the temperature. The only thing I can think of that they could’ve done to solve the foam problem, other than more tests and inspections or going to a different adhesive, is put a coating on the outside of the foam, perhaps a lightweight nylon, kevlar, or nomex “stocking”. But any weight it would add would reduce the payload by an equivalent amount.
If I remember right, they changed the formulation of the foam for some reason. The old formula would not have caused the problem because it produced smaller chunks?
Rand, you mentioned on Wayne Hale’s blog that a Mars mission can’t happen if you require a rescue capability, and suggest an armada approach.
But what if instead of a rescue capability you added a self-destruct, and instead of using astronauts, crew the mission with lawyers, even Congressmen? Then there wouldn’t be any ethical dilemmas or the need for expensive safety measures other than keeping an armed guard posted next to the destruct button to keep people from pushing it prematurely. 🙂
George, yes your idea has merit. But the last thing you want is politicians to make it to mars. Natural law and contract law. Let’s do it right this time.