NASA Test Shows Foam Could Be Culprit

Ben Hutchings writes "The BBC has a report on an impact simulation that aimed to recreate the impact of insulating foam on Columbia's wing. The result was a large hole that probably could not be repaired in orbit even if it was known about."

i have often wondered

[JeanBaptiste]

how much supplies do they have on the ship? as in: so they discover a problem that wont allow them to re-enter... do they have enough food and stuff to allow them to stay up there a few more days, until possibly another shuttle could be launched with repair materiels, or at least to ferry the astronauts safely back to earth?
what about the ISS? could they have docked there for a while?

So What Now?

[CrankyFool]

It'll be interesting to see what the reaction to this failure will be.

Challenger didn't really rock the way we did Shuttle missions because the problems that led to its explosion were not core to how the Shuttles are built -- someone / some process screwed up and there was a relatively reliable way to make sure it wouldn't happen again.

Columbia, on the other hand, was destroyed because the design of the Shuttle is so fragile that once you develop an external problem, you're dead -- since they're using tiles that are individualized, there are no spares they could carry that would help them fix this sort of problem.

Hopefully, this will be a step in the right direction -- either a radical redesign of the Shuttle, or its abandonment in favor of a more robust solution.

A pinch of salt ...

[BillsPetMonkey]

The impact was so violent that it popped a lens off one of the cameras recording the experiment and prompted gasps from about 100-strong astonished crowd.

When I hear of "entertaining" demonstrations to prove a point, I'm reminded of magicians before an audience and furrow my brow.

Is the real "secret" here a less visually spectacular flaw, not in a bodypart but in the design process and it's assumptions?

another story

[pyros]

It's amazing to think that prior shuttle launches have had foam break off and strike the wing without this happening (according to Discovery Channel [discovery.com]). Makes me wonder what was different, perhaps just the size of the foam chunk. It's good to know they finally tested it out to measure the impact. Tragic that people died first. Here's a link to another article [voanews.com] on VOANews.com [voanews.com]

Accouting for angle and all?

[tevenson]

Does this also account for the the angle at which the foam in the wing? They don't mention it so I thought it was a question worth asking.

My understanding was that the foam glanced off the wing at high speeds and wasn't simply "shot" into it from a right angle. I may be completely wrong (and would love to be corrected) on my misunderstanding.

This obviously wasn't the same kind of foam we use to sleep on when we go camping.

Re:Minor curiosity...

[wass]

but what exactly would they do if the damage was too severe to be repaired?

in this case, where heatup during reentry would be a huge problem with a damaged wing, I was wondering if they could bring the shuttle in at a very oblique trajectory consisting of many orbits of slightly-decreasing radii to aerobrake it orders-of-magnitude more gradually than they currently do now.

850 km/h in 2 seconds?

[Psychic Burrito]

Can the impact speed be really that fast? Before the piece of foam fell away from the shuttle, it was moving at the same speed. To impact at 850 km (530 miles) hour, the piece of foam would have to slow down 850 km/h during the short distance between falling off and hitting the wing... during 2 seconds or so. Are the numbers really feasible?

Excuse the raw humor

[Stephen Samuel]

This pretty much literally blows a big hole in any argument that the nasa probe people were over=estimating the kind of damage that a 'little' chunk of foam could do to the shuttle's wing.

I think that this final test is a smoking bun because it shows that pieces of foam can do much more than just cause minor holes in the wing. that might allow a fatal stream of air into the shuttle wing. If Columbia had had a hole in it's wing like this test created, it probably wouldn't have made it anywhere near as close to the landing point as it did.

I'm guessing that this was something of a worst-case scenario, and it pretty much blew the socks off the testers.

(having gotten in my weekly quota of pun, I'm now gonna go do some real work).

Challenger's O-ring led to new O-ring design...

[aksansai]

The Challenger (as well as Columbia, and the newer vehicle that was being built - Discovery) had a flaw in the design of its O-ring that NASA itself knew could cause problems in flight. The design itself worked (proven by earlier flights of the shuttles). However, the design was not resilient to, as you said, external problems that were not properly thought up before-hand, such as massive fluctuations in temperatures (which led to the failure of the seal on the booster rocket).

A university student did an excellent case study [utexas.edu] on the Challenger incident, including the O-ring design "flaw," and what NASA did to improve upon the design.

If it were in NASA's tome of simulated problems, there would have been a way to make sure a rescue would have been possible. Even if we had to park the shuttle in orbit (or on the international space station) until a rescue could have been performed. It tires me to listen to the people that say "well, they would have run out of oxygen if they were not able to return immediately."

Fact: humans will never be able to calculate for every single variable in a system. It's just impossible. I completely agree with you. We will continue to develop better designs that will hopefully prevent further destruction and loss of life.

Re:i have often wondered

[jmichaelg]

The ISS was out - it was on a different orbit and the Shuttle didn't have enough fuel to make the transition.

On the other hand, I have also wondered why the hell they couldn't send up an empty shuttle and bring everyone back on it. Moreover, once the Columbia had been emptied, they could have tried to bring it back with out bleeding off speed using S turns. The Columbia broke apart as it was slaloming and had just loaded up the damaged wing. Had they known the wing was busted, they may have been able to slide slip the whole way in and kept the damaged wing trailing on the backside the whole way down.

All those ideas go out the door when the shuttle manager said "Even had we known, there was nothing we could have done." For that sentiment alone, he deserved to go - it was a far cry from Gene Kranz'es "failure is not an option" attitude when Apollo 13 blew an oxygen tank.

Re:850 km/h in 2 seconds?

[a_timid_mouse]

OR the shuttle accelerated that much during those two seconds while the foam decelerated due to friction. Have you ever watched a shuttle launch on NASA TV? Take a look at the speeds and rate of acceleration that those things hit during launch. You might be surprised.

Velocity of Foam at Impact

[DonJefe68]

I have a question - and maybe the question will demonstrate my pig ignorance of the physics of deceleration, but I've not been able to figure this out. In the test they shot this piece of foam at approx 500 mph into the wing at approximately the same distance as the foam flew from the tank to Columbia's wing. Here's my question: At the moment of separation from the external tank, the piece of foam should have had nearly the same velocity as the shuttle/external tank, relative to a stationary object. Immediately upon separation, the foam would have started decelerating and the shuttle was still accelerating, but it seems hard to imagine that at the moment of impact, the differential velocity of the shuttle/tank versus the foam piece would have resulted in an impact at 500 mph.

I mean, it's not like the shuttle flew into a stationary object while it (the shuttle) was going 500 mph (similar to a jet hitting a bird or whatever). Was the shuttle really accelerating that quickly so that in the one or two seconds between foam separation and impact on the wing the shuttle had gained 500 mph in velocity relative to the foam piece? My faith in scientists is such that I imagine this must be the case (since the alternative is that they missed this question) but I would love to have someone with enough knowledge of the science to clue me in.

Re:Minor curiosity...

[Xzzy]

I thought the popular strategy was to make sure the shuttle could dock with the ISS, and allow the astronauts to get back to earth in the soyuz module the station has.

Granted I got this info from the media so it could be a pointless thing to say, but it sure sounds good.. especially since I don't think they can just lob another shuttle into space on a whim.

It would also still leave a broken shuttle up in space, which I imagine makes for an interesting engineering problem once the business of keeping people alive is done.

Astronauts knew of wing damage

[Muttonhead]

The astronauts knew their wing was damaged because one of them wrote to his brother on earth via email:

"Sen. George Allen, R-Va., said in a televised speech on Tuesday that the brother of Columbia astronaut David Brown disclosed receiving an e-mail from orbit that conveyed the crew's "concern" about the left wing, the Richmond (Va.) Times-Dispatch reported in Wednesday's paper. According to the report, the senator said Doug Brown, who lives in Virginia, told him his brother's e-mail said the crew had taken a photo of the left wing.

Story [upi.com]

Columbia could not have reached the ISS

[AzrealAO]

Columbia was in the wrong sort of orbit to be able to rendevous with the ISS, nor was it capable of generating enough delta-v to enter a rendevous orbit.

This is one of the reasons the board recommended that all future shuttle flights (apart from the already scheduled Hubble Servicing Mission), fly to the ISS, or in Orbits that are capable of rendevousing with the ISS.

How about a *real* test?

[whitroth]

I read the news story...including the part where they said, "we could do this again, and get a different result".

So, first, how about doing this at *least* three times?

THEN, take the average, and put the damn thing in front of a horizontally-mounted rocket engine, to simulate actual re-entry, and see if it happens...or if, as has happened in the past, the shockwave keeps the heat from penetrating.

Gee, if that happened, then they'd have to go back to looking for another cause...like (google for it) the diehard's analysis that it was stress corrosion cracking in the hydraulic lines that control the elevons. Loosing control of them would rip the wing *right* off.

But then, stress corrosion cracking shold have been caught...*if* they hadn't cut safety inspectors by 75%, and if the managers, in their own meetings, cared more for safety than for "being a team player, and meeting the schedule".

NASA's management strucure needs flattening, anyway - there's maybe 1 chief for 2 indians. Is that sane, to y'all?

mark

Re:computer modeling

[Psion]

Because prior to this accident, the concensus opinion at NASA was that the foamed insulation was low-mass and crumbled easily enough that it didn't pose a threat to the vehicle. In fact, the mixture of foam being used had been in place for five years. In STS-87, the first time it was applied to Orbiter Columbia, foam debris caused 308 hits on the orbiter, some resulting in deep gashes. After changes in the method of application, the foam was rendered more secure, but chunks continued to break off in future flights.

Still, none of those flights exhibited the kind of damage that would lead to the Columbia tragedy until now. It seems perfectly obvious to "monday-morning quarterbacks" that the foam was a problem, but five years of experience suggested otherwise.

Re:Minor curiosity...

[Longbow]

Aerobraking only works in the upper atmosphere and requires you to remain in orbit. Basically you incrementally lower your closest approach by dipping into the atmosphere, as you stated, however, at some point you cease to be "in orbit" and just reenter. That point is still about 100 km up, so the reentry stresses and heat would not be much different from a nominal shuttle entry and would have likely destroyed the vehicle anyway.

whatever happened to starlite?

[ed.han]

back in 1993, british amateur inventor maurice ward created a plastic he called "starlite", which would withstand temperatures of up to 2700 celsius (that of a nuclear explosion). does anybody know what temperatures are reached during reentry, or for that matter, what happened to ward & starlite? i tried googling but didn't find anything interesting.

ed

Bad Documentation Kills.

[Tackhead]

> It seems perfectly obvious to "monday-morning quarterbacks" that the foam was a problem, but five years of experience suggested otherwise.

Five years of experience, or One Fucked Up Powerpoint Slide [edwardtufte.com]?

Just like poor presentation of temperature data killed Challenger, poor presentation of the foam data killed Columbia.

Stupid goddamn PHBs and their fucking PowerPoint slides.

Rescue missions do NOT require immediate shuttle

[vasqzr]

They could have used (basically an ICBM) a satellite launch rocket, put a supply shipment up there, and let them sit up there for a few weeks.

not unexpected

[mantera]

According to the legendary aeronautical designer, Burt Rutan, the shuttle is a very expensive and very dangerous vehicle. http://www.wired.com/wired/archive/11.07/space.htm l

Stop attempting to walk before you Crawl

[shthd]

Perhaps NASA should start looking at new designs with potentially fatal flaws. Have they not been using this design for something like 15-20 years now? I agree, but maybe they should wait until they have a plan with somehting really revolutionary on the table. Perhaps in ten years they could build a scramjet thing with much greater capacity.

In order to develop scramjets, NASA needs to ressurect the X-15 program. Hypersonic flight [nasa.gov]. With newer materials and newer rockets, they could go higher and faster than ever before. The X-15 reached 62.5 miles and the pilots even got their Astronaut wings. If that 62.5 mi altitude sounds familiar, it should. That's what you need to win the X-prize. It did close to 200 missions in 9 years and nowhere near the cost of the shuttle program. The X-15 would make a perfect platform to test designs. As a matter of fact it flew one mission with a mock scramjet aboard.

Re:happens often

[Psion]

Let's see, prior to STS-86 and while CFC-11 was still used as propellant, sprayed-foam insulation loss was minimal, sporadic, and concentrated around a few problem areas and was characterized by small debris. After that, the loss became common, resulting in significant damage to STS-87 and other flights and was characterized by much bigger chunks shedding off random areas of the external tank.

So, yes, it most certainly is the new formula that caused the problems.

Re:computer modeling

[Psion]

So prior example shouldn't be considered? While I concur that management erred in considering the threat from foam debris (and it's composition) to be negligible, five years of evidence that the system worked reliably argued that STS-107 was fine. Right or wrong, the evidence was on their side. Based on the media focus, it seems obvious that this is something that should have been investigated more clearly, but how many times had something similar happened on prior missions with no significant damage to the vehicle?

Yep, they were lucky those times, but there was no factual evidence to the contrary, and unfortunately, it took this tragedy to provide a single, compelling data point to the contrary. Had falling foam cause more significant, but non-catestrophic damage prior to STS-107, the warnings probably would have been given more attention, but until now, the only cost was expensive tile repair.

If I sit down in a chair fifteen times without it collapsing under me, it is hardly a "dangerous fallacy" to think it will continue to support me fifteen more times. It is reasonable to assume it will eventually break and fail, but no reasonable person will stop and examine their chairs every time prior to sitting in them. The shuttle is a remarkably complex machine, with extraordinary attention paid already to vehicle safety. There's going to be a lot of hand-wringing over this incident and a lot of finger-pointing until things settle, but in the final analysis, no system is perfect. It is humanly impossible to catch all accidents before they occur. Sometimes, unfortunately, it takes a catastrophic failure to highlight a problem before it is corrected.

And guess what? This won't be the last failure in space exploration/exploitation.

AP reports, it's not the first time!

[jafac]

http://story.news.yahoo.com/news?tmpl=story&cid=53 3&ncid=533&e=3&u=/ap/20030708/ap_on_sc/shuttle_ear lier_breach

Gasses have breached the wing on a previous Atlantis flight. And they didn't even know about it until a postflight inspection, AND, it sounds like the damage almost went unnoticed, and the Atlantis would have launched with the damage from a previous flight, and no replacement of the faulty seal.

This damage was caused by the combination of a faulty seal, and falling ice.

The Columbia is being blamed on just the falling foam. But wouldn't you say that the heat shield was a faulty design?

Did the Soviet shuttle use tiles?
The X-33?

I recall during Columbia's first flight - the tile design was questioned in the press. The aluminum structure underneath, of course, is flexible, and it's covering, the tiles, is not. A few tiles popped off on that first flight, and subsequent flights - and it was mentioned that the wrong tiles falling off would have dire consequences.

Sad, that nobody sees this as an unacceptably risky design.

Re:Here's the equation.

[coult]

> It's a fairly simple matter of plugging the in the right
>values for m, g, S, C, p, A, v0, and x0, and we'll settle this
>right now. If anyone knows what those numbers are I
>invite you to share them.

This makes a nice toy model but it won't cut it for estimating the relative velocity of the foam (which can be done more easily just by watching the video and using fifth-grade math: velocity = distance traveled/travel time).

First of all, the foam was very likely tumbling. So A isn't constant. Second of all, the foam was probably rougher in some places than others; so C isn't constant either. To solve both for the velocity of the foam and its orientation requires a nonlinear system of 7 differential equations. You'd also need to know the exact shape and surface characteristics -- clearly impossible to know at this point.

Re:happens often

[Xerithane]

We've all heard about the joke with Gates saying "If cars had progressed at the same rate as computers..."

The first motorized car (Daimler and Benz) that used a combustion engine was in 1886. Look at cars in 1926. That is a lot more innovation than what was done in the first 40 years of computing.

There was a huge boom for car manufacturers prior to the Great Depression that was very similar to the dot-com boom. There was a tremendous amount of innovation, especially in clutch and brake design.

Cars did progress at first like computers did. Then they reached a limit as to how far the progression could take them. Computers will reach that, too.

I seem to recall

[uberdave]

I seem to recall that on the first shuttle flight, when missing tiles were a *huge* concern, that the astronauts had some sort of chemical foam, or gel that they could use. It would burn off like the Apollo heat shields.