Transterrestrial Musings

Comments

Time to shut down NASA for good.

Two very damning pieces of evidence out today.

From the WaPo, we learn that the damage analysis of the insulation strike was piss-poor. The analysis did NOT include the leading edge of the wing as a potential contact location, and the analysis assumed that the debris was foam insulation - and NOT ice. One further suspects that the analysis done on the impact of the foam insulation on the orbiter did not consider how much harder the foam becomes at very low temperatures.

From the AP, we learn that two years ago, an Atlantis reentry was adjusted in order to lower the thermal stress on the orbiter. Why? Because a debris strike at launch was thought to have damaged the TPS. And after landing, damage was found.

If these two news reports are accurate, then Dittemore's briefings have been little more than a charade. The debris strike COULDN'T have caused the catastrophic failure - if you ASSUME that the debris wasn't ice and that it didn't hit a vulnerable spot on the wing. There was NOTHING that could have been done EVEN IF NASA had known the left wing was damaged - except that at least one prior orbiter mission had been changed to reduce the thermal stress of reentry.

IF this was happening for the first time, it would be cause for a major restructuring of NASA, but of course it's not the first time, it's the second time. Time to close the doors and start over somewhere else.

Posted by G.Haubold at February 8, 2003 12:15 PM

The details continue to surface. At this point I have but a single question: Engineers are trained to think in terms of worst-case scenarios when performing an analysis -- why wasn't that done in this instance?

Posted by gojou at February 8, 2003 12:42 PM

I explained that here. It doesn't absolve NASA, because it's still their fault for perpetuating a program philosophy that relies entirely on the Shuttle, but it does explain it.

Posted by Rand Simberg at February 8, 2003 02:29 PM

Actually, no, Rand, you didn't.

Two years ago with the Atlantis, NASA decided to change the reentry trajectory to reduce thermal stress.

A week ago with Columbia, after taking a larger hit from launch debris, NASA decided to do NOTHING.

This dialogue from Apollo 13 just isn't applicable here:

"Gene, we think they may be entering a little hot."

"Anything we can do about it?"

"No."

"Then they don't need to know, do they?"

There was something they could have tried, and they did NOT. Yes, there were too few options, NASA's bad about that but still - with an honest analysis of potential danger, NASA could have decided to try something - at the very least adjust the reentry trajectory to reduce thermal load on the orbiter - in the extreme, maybe try to launch additional supplies up to the Columbia to buy time to come up with a more creative fix.

Posted by G.Haubold at February 8, 2003 07:15 PM

I'm not saying they had no options (though in reality they probably didn't, due to lousy programmatic decisions made years or decades ago). Just that they didn't think they did. That Atlantis incident was probably just a psychic bandaid. In reality, there's no trajectory change that's going to significantly reduce the risk of a broken TPS.

Posted by Rand Simberg at February 8, 2003 08:37 PM

I would argue that it's one thing to perform an analysis resulting in a questionable outcome and try to interpret the results in a more favorable light, even if circumstances don't support that interpretation; it's quite another to completely fail to consider what is clearly the worst case scenario.

However, my question was of a rhetoric nature. I happen to believe the worst case scenario was analyzed and the logical conclusion reached. I also believe we're not being told so.

Posted by gojou at February 9, 2003 01:56 AM

How was the worst-case damage scenario not analyzed? They took the steepest impact angle, doubled the highest velocity (which quadruples the energy), and used the largest estimate of foam size and mass; then they impacted it onto the most-fragile area of the TPS, using code which over-predicts damage.

The RCC leading-edge composite is very tough -- that's a large part of why they used it there, it's by no means the fragile beast that the other tiles are. And the foam will not absorb water, so assuming it was saturated with ice is a bit strange; the debris cloud was indicative of foam, not ice, in any event.

I hear a lot of armchair quarterbacking here, and not much realism. And this is coming from an engineer who has worked with NASA on far too many occasions, and whose opinions of their engineering abilities (in a very practical and applied sense) is low... from my experience, they err to the side of unworkable conservatism far more than they tend to cover up real dangers, ever since Challenger.

Posted by Troy at February 9, 2003 09:42 AM

Well, the next question we need the answer to is: how did the sensor readings on the right wing of the Atlantis compare to the sensor readings on the left wing during its adjusted reentry? Since NASA admits that adjustments were made to the Atlantis reentry path (to protect its RIGHT wing), the sensor data would tell how much thermal benefit was created, if any.

NOW, since NASA hasn't come out and said "we tried an adjustment on Atlantis and there was NO thermal benefit" I know what I THINK THE ANSWER IS . . . . . and presumably that information will leak in a weak or two and we'll see.

To me, it sounds like NASA went right back into the "Russian Roulette" safety mode that Feynman discussed in his addendum to the Challenger report - you have regular damage to the TPS and then don't have any reentry problems, once twice three four five times and it's hard not to start thinking that minor TPS damage just isn't an issue and complacency (or magical thinking) sets in. And eventually something goes wrong.

I am so sorry that Dr. Feynman isn't with us anymore - his wisdom and insight would add a lot to the commission report on the Columbia.

Posted by G.Haubold at February 9, 2003 09:51 AM

Each branch of the failure tree being constructed by NASA has its own worst-case scenario. Troy has pointed out that the worst-case scenario of one branch has been considered. What about the other branches?

One need not assume the foam was "saturated" with ice. The object shown in the film clip could have been ice frozen to the tank surface and dislodged during launch.

It's interesting to note that the white coloration of the material in question, as shown in the film clip, is consistent with the appearance of ice when light reflects from it. In fact, I've wondered why, if the object was actually foam, why it doesn't appear orange in the clip.

The manner in which the object disintegrates in the clip is also consistent with how ice looks when it shatters.

Given the available evidence, there is no good reason to exclude the possibility that damage to the wing's leading edge might have contributed to the disaster, or that such damage might have been caused by ice. Yet the media reports that this possibility was not considered in the post-launch analysis. I classify that exclusion as a failure to consider at least one possible worst-case scenario.

Posted by gojou at February 9, 2003 10:03 PM

One other thing: Let me restate that I am not a conspiracy theorist. My comments here are informed by my technical background (mechanical engineering degree, specializing in composites) and my understanding of human nature and the nature of government bureaucracy.

Every engineer knows the meaning of KISS. In contemplating the Columbia tragedy I've tried to hew as closely as possible to the KISS principle. The views I've expressed here are the result of that effort.

Posted by gojou at February 9, 2003 10:22 PM

gojou, I appreciate your looking at the problem as an engineer. Now take one step back, and consider that NASA might also have had engineers seriously look at the problem.

If you do that, you'll find that the reason why the foam is there in the first place is largely to prevent the formation of ice, which was recognized as a potential problem early in the program -- and long before the first Shuttle actually flew. This NASA source (page 34 in the PDF, p. 304 in the document's internal numbering) says:

Without proper insulation, ice could form on the tank that might shear off and damage the orbiter tiles during flight. The tank surface and every line and bracket on the outside of the tank had to be insulated to keep the exterior temperature above 32 degrees.

If you add to that the facts that the outside temperature was well above freezing at Columbia's launch, and that the T-3 inspection revealed no ice on the ET, what are you left with?

Just saying "well, I think ice might have been there" isn't good enough...

Posted by Troy at February 10, 2003 07:54 AM

There's no need for the encryption box to be classified; there are plenty of well-known encryption techniques that would work and not be compromised by capture of the box itself.

(There are also lots of bad schemes that would be compromised by loss of the box, and I hope they had enough sense not to use them...)

Posted by mike earl at February 10, 2003 08:11 AM

I don't agree that there was "no need." That might be your opinion, but there's little harm in classifying it, and you can't know that someone might not get a clue from it. Security by obscurity does have some value, and it's better safe than sorry.

Posted by Rand Simberg at February 10, 2003 08:53 AM

Troy: There's a lot more to the possibility of ice formation than the temperature at launch and pre-launch inspection.

The fuel tank is filled with liquid oxygen and liquid hydrogen at temperatures as low as minus 400 degrees Fahrenheit . . . . and it had been raining in Houston for a couple of days prior to launch day. Some experts have noted that ice can form inside cracks in the tank's foam, causing pieces of the insulation to pop off. That could easily have been the case with Columbia. Now, the inspectors looked at Columbia and the tank before launch and saw nothing unusual . . . . but of course the inspectors might have missed something.

Furthermore, with the heavy rains in the days leading up to launch and possible imperfections in the insulation, might moisture have been present UNDERNEATH the insulation? John Lindner, an associate professor of physics at the College of Wooster in Ohio, said ice could have formed in a gap between the tank and its exterior layer of insulating foam, which would make it hidden from the view of the inspection team.

So, the presence of ice isn't proven, but it can't be excluded, yet the NASA analysis during the Columbia flight did exclude the possible presence of ice (as best we can tell). Which is just dumb and shortsighted.

Posted by G.Haubold at February 10, 2003 11:37 AM

Not clear to me how rain in Houston would cause icing on an ET in Florida...

Posted by Rand Simberg at February 10, 2003 12:38 PM

Oops. Brain fart. There was reportedly lots of rain in FLORIDA for a couple of days prior to launch, leading some to wonder if the rain-soaked fuel tank got iced up underneath or in nooks and crannies.

Posted by G.Haubold at February 10, 2003 02:27 PM

OK, I finally unearthed weather and temperature info for the launch date -- from NASA's own site, actually. This launch and landing forecast and this post-launch weather summary indicate that freezing temperatures were not a concern on January 16. I am willing to concede that ice formation due to ambient temperature was unlikely.

Mr. Haubold's speculation regaining rain seeping into cracks in the foam and then freezing is interesting. If this were to occur, the foam would provide excellent thermal insulation for the ice, making Troy's observation regarding ambient temperatures being well above freezing less relevant. However, NASA's weather summary for the week preceding the launch indicates total rainfall of 0.04 inches -- probably insufficient to create an ice mass large enough to cause the type of damage we are discussing.

Total rainfall for the two weeks preceding the launch was 0.92 inches -- substantially greater. Is that enough rainfall to create a sufficiently large mass of ice? I believe so. Could the ice survive two weeks without melting? Perhaps. The extremely low surface temperature of the tank, aided by the thermal insulation provided by the foam, might be enough to keep the ice from melting. Of course, this assumes that Columbia sat on the launch pad for the entire two weeks and that the external fuel tank had already been filled by then. Not being a space geek, I don't know the launch rollout schedule and can't speak to whether these assumptions are valid.

Troy, I still hope you're right. Of course you are correct in stating that you just can't assume ice was present. My previous statements were based on the assumption that ambient temperatures were below freezing in the days leading up to the launch and were very low at the time of launch. That assumption was based on my memory of media reports of extremely cold weather in Florida during early January and proved to be invalid. However, if my newest assumptions are valid then I still think there's a slight possibility ice played a part and still believe that possibility should have been analyzed by the engineers.

I guess what it comes down to is that NASA would sound much more intelligent and inspire much more confidence saying, "Yes, we considered the possibility that ice might have played a part, but the evidence discounts that scenario" than saying, "No, we hadn't considered that scenario."

Posted by gojou at February 10, 2003 05:27 PM

gojou, the external tank isn't loaded with its cryogenic propellants until about T-8 hours; that doesn't leave a large window for the ice scenario, does it? Off the top of my head, I don't know how long the ET can remain fueled, but I know from personal experience that they offload the propellants when a launch is scrubbed for a given day.

And regarding your last paragraph: my memory of the early NASA statements about the "foam impact" analysis is that they explicitly stated that they'd considered whether ice might have been present, but discounted it for all the reasons I've been repeating here. That, in fact, is why I've been making these arguments... y'know?

Posted by Troy at February 11, 2003 11:06 AM

Dittemore's statements are here"; relevant portions:

Contrary to recent speculation, he said, the foam is virtually waterproof. He said ice team inspections the day of launch found no signs of any ice buildups on the tank and that in his opinion, the foam did not end up with ice imbedded in it.

"It is very lightweight," Dittemore said. "So it's designed to be resilient and be an insulating material to keep the tank cold...

I don't think there is an embedded ice question here, I don't think this came off as a chunk of foam and solidified with ice. When it hits the wing, this piece of foam disintegrates. Although we're going to look, we did not have icing conditions that day. We didn't have the ice, it's impervious to water. So it's something else. It's something else."There was more, but this is the first link I found.

Posted by Troy at February 11, 2003 11:13 AM

Troy: Great find.

I'm now convinced it might not have been ice, but I'm also convinced that Dittemore's judgement is badly warped.

Look at it this way: The insulation's not supposed to come off at all, but it does. In fact, one launch a year or two ago got several hundred damaged tiles from falling insulation debris. So stuff happens ALL THE TIME that Dittemore and none of the engineers understand completely, or expect.

Just because Dittemore can't figure out how ice might have been present doesn't mean that it wasn't, or that he should be so determined to argue the possibility away. His mind just doesn't seem very open to the idea of icing. THAT makes me a little suspicious. More because he's feeling guilty than trying to hide anything.

If the tank wasn't filled with cryogenic fuel (minus 400 degrees F) for about 8 hours prior to launch, I might agree that ice was a remote possibility - BUT. All you need is a large crack in the foam, or a gap between the foam and the tank and add a little moisture and you've got ice. Right? The foam is reportedly sprayed onto the tank, but the tank must flex a lot in transportation and being raised into position - with the tank flexing I'd think some of the insulation could easily get a crack or gap in it. And it's a huge tank. The inspectors could easily miss something as small as the anomalistic piece of whatever that his the shuttle.

And I STILL WANT TO KNOW more about the sensor data from the Atlantis reentry a couple of years ago. The one where they adjusted the trajectory and aircraft attitude to reduce thermal load. Did it, or did it not, reduce thermal load? That's all.

Posted by G.Haubold at February 11, 2003 02:01 PM

Troy, thanks again for the info. I had not been able to find info regarding how soon before launch the cryogenic propellants are loaded, but instinct told me it surely wasn't two weeks -- unlikely in any event, but especially so given the current political climate. That's a long time to have what is, in effect, one hell of a big bomb just sitting around.

The article you quoted states that NASA had considered and rejected the ice scenario as part of its post-disaster investigation. I'm sure they have, but the comment in the last paragraph of my previous post refer to my belief that the ice scenario should have been included as part of the post-launch analysis and, in particular, to statements Dittemore made during the Feb 3 NASA news conference in which he as much as admitted NASA had not, to date, so much as even considered an ice scenario. This quote from this LA Times article supports that contention:

There was enough concern, however, to schedule at least two teleconferences during Columbia's first week in orbit to discuss the debris impact and possible damage to the shuttle's tiles. The teleconferences included representatives from NASA, shuttle prime contractor United Space Alliance, Boeing and Lockheed Martin. Their job was to study the issue and report to senior officials that make up the shuttle's mission-management team.
During the meetings, there was a presentation by a trajectory-analysis group, which gave estimates of the amount of debris that struck Columbia's tiles and where it hit on the shuttle's underside. Tile experts used that analysis to estimate possible damage.

After the teleconferences, the group reached a conclusion: There was the potential for a large area of damaged tile, but the damage would be limited in depth and not endanger Columbia.

"These thermal analyses indicate possible localized structural damage but no burn-through and no safety of flight issue," stated a daily summary report from the shuttle's Mission Evaluation Room issued Jan. 28.

Some, however, felt the finding was flawed.

Analysts assumed that the tank debris that struck Columbia consisted entirely of foam insulation. The possibility that harder tank materials or ice were involved was not considered. There also was concern the tile team's analysis of the depth of the damage was wrong because it did not fully account for the large amount of debris that hit.

In any event, thank you all for the stimulating discussion and thanks to Rand for providing the forum.

Posted by gojou at February 11, 2003 03:45 PM

My Dad worked as an engineer on the Shuttle program in the early days and he remembers meetings regarding the possibility of using a less thermally but more mechanically stressful "skip" reentry, which gives the vehicle a chance to cool off between skips. Of course, the big problem with this is that you don't know exactly how many skips there will be before you reach your terminal reentry phase, which makes selection of your actual landing field extremely problematical and very much a last-minute thing.

I suppose if they did have guidance software for this option, which they probably don't, they could use it as a last resort and have the crew bail out at some predetermined altitude if they couldn't reach a landing field. This sounds pretty dicey, too, because to successfully bail out of an aircraft with as little risk of injury as possible, you'd want to be going pretty slow -- my guess is less than 300 knots indicated, and slower would be better -- which is getting pretty close to stalling speed for the orbiter, as I recall, though I have heard it mentioned as an option by NASA folks before.

The bottom line before you followed such a risky plan is that you would have to believe that you had a problem that would likely result in loss of the vehicle during a normal reentry, and nobody at NASA or any of the contractors really believed that as far as we currently know.

I just heard the last day or so that the Air Force (I think) detected a large object moving away from Columbia near the time it was making its deorbit burn. I'm still not certain we know exactly when the orbiter began shedding parts, and it's entirely possible something may have come loose very early on.

Posted by Mark Scheen at February 11, 2003 06:25 PM

...Actually, Rand, I didn't know I'd written an FAQ on the Damned. I haven't even written one on Bauhaus, Tones on Tale, or even Peter Murphy.

Ah well...

Posted by OM at February 18, 2003 02:31 AM

Post a comment