Why didn't NASA replaced the SRB of challenger?

[PakledHostage]

The other solution would be to give all the power to the engineers:they can do whatever they want as long as it fit the budget.

Sadly it often isn't that simple. Even though engineers practice under a strict code of ethics, they are human too. Some have higher risk tolerance because they aspire to be managers, and higher risk tolerance often means lower costs and a better performance review. Some have higher risk tolerance because they are too technically incompetent to realize that they're treading on dangerous territory. And others who have a lower risk tolerance are sometimes dismissed as being too conservative, even though they may well turn out to be right (as was the case for Roger Boisjoly at Morton Thiokol). Human factors will always necessitate a system of checks and balances, especially in safety critical industries and when big money or politics are involved.

[Justy]

For instance, it took an airline employee taking a gun through security and shooting his boss and three pilots, causing a crash, for us to change the rules requiring flight crew to be screened

You have a good point with Pacific Southwest Airlines 1771, but you need to take it further to find the real lesson, which I really wish the global security community would learn.

Summary: A man recently fired for theft from the employer is allowed back on that employer's aircraft the same day, without confiscating the company ID that allows him to bypass screening. He brings a gun on board, shoots the manager who fired him who happened to be on the same plane. Flight attendant opens the cockpit door to tell the pilots there is a problem onboard. Man shoots the attendant, pushes past the open door and shoots both pilots; may also have shot another pilot who was in the cabin as a passenger trying to get to the cockpit to save the plane. Plane crashes killing 43 in all.

Possible Solutions: Take away the company ID that gives him those special privileges. Leave the cockpit door locked. Tell the disgruntled fired airline employee he's not allowed on the airline's planes anymore and he can take the bus home.

What We Did: Air crew must be screened for dangerous baggage.

...which doesn't solve the problem. Crew don't NEED dangerous baggage if they want the plane to crash. They already have what they need for that: the ramp crew can bring down an airliner with duct tape (and has, accidentally: Aeroperu 603) or a poorly latched baggage door or stuffing a ham sandwich in the fuel port, the flight attendant has a heavy bucket of scalding hot coffee, the pilot has control of the plane for goodness' sake. All the pilot has to do to cause a crash is STOP DOING HIS JOB, and we're worried about him having a pistol or a metal dinner knife.

The problem was that this guy held on to his trusted status as air crew even after being fired because he couldn't be trusted.

So I say the problem isn't just that we only address problems as reactions... PSA1771 shows that when we do react, we often address the wrong problem.

Edited October 11, 2013 by Justy

[PakledHostage]

Possible Solutions: Take away the company ID that gives him those special privileges. Leave the cockpit door locked. Tell the disgruntled fired airline employee he's not allowed on the airline's planes anymore and he can take the bus home.

What We Did: Air crew must be screened for dangerous baggage.

You left out a detail from the Wikipedia article you cited:

Several federal laws were passed after the crash, including a law that required "immediate seizure of all airline employee credentials" after an employee's termination from an airline position.

[Justy]

Heh. So I did; I gave a link to the article but I was working from memory. Whoops.

I'm glad to hear that actually. I then wonder why, in addition to the very appropriate step of confiscating ID, we took the additional misguided step of physically screening the air crew.

[WestAir]

Heh. So I did; I gave a link to the article but I was working from memory. Whoops.

I'm glad to hear that actually. I then wonder why, in addition to the very appropriate step of confiscating ID, we took the additional misguided step of physically screening the air crew.

We often don't. A lot of crew use a system called "Known Crew Member" to avoid losing their pocket knives or removing their shoes just to pass security. It's not always available, but it usually is.

[Sandworm]

The choice of solid fuel was political, but it was also a military decision. Solid fuel makes for good ICBMs because it is stable. But there weren't going to be enough missile tests to maintain a large solid fuel production facility. So by giving shuttle solid boosters NASA would create a demand that could keep such a facility operational, ready to also supply fuel for missiles.

And the real deadly part of shuttle was the part that mandated the wings, boosters and nearly everything else: the cargo bay. Shuttle was the only craft in history able to bring something back from orbit. Google "Project Azorian" to see the lengths that US intelligence agencies would go to acquire enemy hardware in the 1970s. They were drooling at the concept of plucking a russian sat from orbit.

[Mr. Scruffy]

Uh, yeah, never thought of that unique property of the shuttles. Makes a lot of sense. Just shooting down Sputnik XXVII would be sort of lame and not much better than pouting - but stealing the thing, picking it right from the skies and bringing it home for investigation and as trophy, now that´s sort of cunning and ´in your face´. Then put an US-flag on it and write in front of its name tag on the chassis: ´The probe formelly known as...´ and relaunch it.

[Vat]

The old 80/20 rule applies to aerospace as well - in this case, 80% of incidents (hate the word "accident") are human related, 20% are hardware related.

There's an old engineering theory about safety being like a block of Swiss cheese. Ostensibly the cheese looks solid, but there are holes throughout the structure. If you think of the holes as problems with the various safety systems - well, sometimes the holes line up.

Each of the three major disasters involving loss of life for NASA were avoidable - Apollo 1 and Challenger as mentioned above, and Columbia was brought down by a known issue (indeed, it was identified as a risk in the design phase, and Columbia's captain was advised of the issue a week after launch).

[PakledHostage]

The old 80/20 rule applies to aerospace as well - in this case, 80% of incidents (hate the word "accident") are human related, 20% are hardware related.

For what it is worth, (at least in the aviation industry) there is a non-trivial difference between an accident and an incident:

An aviation incident is defined an occurrence, other than an accident, associated with the operation of an aircraft that affects or could affect the safety of operations.

An aviation accident is defined an occurrence associated with the operation of an aircraft, which takes place between the time any person boards the aircraft with the intention of flight until such time as all such persons have disembarked, where a person is fatally or seriously injured, the aircraft sustains damage or structural failure or the aircraft is missing or is completely inaccessible.

Edited October 13, 2013 by PakledHostage

[WestAir]

If by human related you mean pilot related, that could be slightly unfair. A vast majority of pilot related incidents (or accidents) are the result of other failures, somewhere, that took the pilot out of the norm. For instance, if a crew encounters a failed glide slope system on the approach to landing that crew is now in a situation not normal to them. If they lose situational awareness and crash into a mountain, sure it's pilot error, but the mechanical failure of a glideslope system must also be taken into account.

I've rarely read of an accident, aerospace or otherwise, where a professional crew of pilots, astronauts, what have-you, ruined a perfectly good craft by sheer 100% human error. There's always something unusual going on preceding it.

In this case, a very cold night before launch was the unusual, non-human event that contributed.

[Sandworm]

I've rarely read of an accident, aerospace or otherwise, where a professional crew of pilots, astronauts, what have-you, ruined a perfectly good craft by sheer 100% human error. There's always something unusual going on preceding it.

"The dead pilot of a twin-engine plane that crashed into a suburban Vancouver apartment building was an 82-year-old with decades of flying experience whose aircraft was recertified for flight after an accident last year, officials said."

http://usatoday30.usatoday.com/news/world/2007-10-21-715558556_x.htm

Total pilot error. Aircraft was fine, a 100% recoverable situation well within the norms. The human failed the aircraft (heart attack/stroke). It does happen.

[PakledHostage]

In this case, a very cold night before launch was the unusual, non-human event that contributed.

To be fair, there were established operating limits that should/could have justified delaying the launch until conditions were within limits... Likewise there are established procedures for flying a missed approach. It isn't unreasonable, in a well designed system, to expect a successful outcome when those types of situations are encountered by the people responsible. If they don't succeed, then either the system or the people failed. In the case of the Challenger, it is pretty clear that the people failed.

Edited October 13, 2013 by PakledHostage

[Nibb31]

Uh, yeah, never thought of that unique property of the shuttles. Makes a lot of sense. Just shooting down Sputnik XXVII would be sort of lame and not much better than pouting - but stealing the thing, picking it right from the skies and bringing it home for investigation and as trophy, now that´s sort of cunning and ´in your face´. Then put an US-flag on it and write in front of its name tag on the chassis: ´The probe formelly known as...´ and relaunch it.

Actually, it doesn't make sense at all. Stealing enemy satellites from orbit was just another of the totally unrealistic promises of the Space Shuttle program.

The ability to bring back a satellite from orbit was demonstrated only a handful of times, with satellites that were specifically designed for being retrieved by the Shuttle. The mass and configuration of those satellites were well known, and their orbits were lowered to facilite the recovery. The shuttle carried custom built cradles in its payload bay with all the properly located attachment points, designed to properly maintain center-of-mass requirements for the re-entry.

To capture an uncooperative hostile military satellite would be totally different. You would need to have extremely detailed information about its center of mass and any fixation points, good enough to custom-build a cradle to fit that specific satellite inside the Shuttle. A small mistake in the mass or balance of your payload, and you risk re-entering with the wrong angle of attack and you burn up. A small mistake in the size and shape and it might not fit in the payload bay. You have to assume that the satellite will be cooperative and won't perform active evasion or collision manoeuvers, because that would make rendez-vous impossible or even dangerous. You would also have to assume that there are no anti-tampering devices or countermeasures on the sat, which would be an unrealistic expectation on a military satellite. You don't want your captured satellite to blow shrapnel through the wings of your Shuttle or to puncture your astronaut's EVA suit.

In fact it's so trivial to prevent a satellite from being captured that trying to capture a hostile satellite with a valuable manned spacecraft wouldn't be worth the risk.

Edited October 14, 2013 by Nibb31

[drakesdoom]

any fixation points

For that matter why would a military grade once up and burn up satellite have any mounting points that are not blown off during the initial launch?

[Mr. Scruffy]

Actually, it doesn't make sense at all. Stealing enemy satellites from orbit was just another of the totally unrealistic promises of the Space Shuttle program.

The ability to bring back a satellite from orbit was demonstrated only a handful of times, with satellites that were specifically designed for being retrieved by the Shuttle. The mass and configuration of those satellites were well known, and their orbits were lowered to facilite the recovery. The shuttle carried custom built cradles in its payload bay with all the properly located attachment points, designed to properly maintain center-of-mass requirements for the re-entry.

To capture an uncooperative hostile military satellite would be totally different. You would need to have extremely detailed information about its center of mass and any fixation points, good enough to custom-build a cradle to fit that specific satellite inside the Shuttle. A small mistake in the mass or balance of your payload, and you risk re-entering with the wrong angle of attack and you burn up. A small mistake in the size and shape and it might not fit in the payload bay. You have to assume that the satellite will be cooperative and won't perform active evasion or collision manoeuvers, because that would make rendez-vous impossible or even dangerous. You would also have to assume that there are no anti-tampering devices or countermeasures on the sat, which would be an unrealistic expectation on a military satellite. You don't want your captured satellite to blow shrapnel through the wings of your Shuttle or to puncture your astronaut's EVA suit.

In fact it's so trivial to prevent a satellite from being captured that trying to capture a hostile satellite with a valuable manned spacecraft wouldn't be worth the risk.

All these points may be valid - but still wouldnt stop a good salesman from selling the idea...

[Vat]

If by human related you mean pilot related, that could be slightly unfair.

No, I didn't mean just pilot related - so yeah, that would be unfair. I'm talking more along the lines of things done incorrectly by people that were avoidable and lead to the accident.

Many, many, many examples - actually, if you include maintenance, engineering and adminstration/policy and procedures as well as air crew, 80/20's probably conservative.

[magnemoe]

All these points may be valid - but still wouldnt stop a good salesman from selling the idea...

Yes, it would be easier to capture an dead satellite. Also far more political safe as it would be salvage not an act of war.

You still need to know attachment points and center of mass, even trough this should be possible to patch around if the satellite was small.

[Mr Shifty]

The Shuttle could also do on-orbit repair of satellites; it had to have special equipment to attach and de-spin the satellites. (Most satellites are spin stabilized.) It only used its on-orbit repair capability on 10 missions (half of them to the HST), but no other spacecraft has ever had the capability.

[Nibb31]

Yes, but again those capabilities made no sense.

For a repair or retrieval, you spend months training for EVA operations, developing procedures and flight rules, fabricating special tools and fixtures. You need to pay lots of highly qualified engineers for several months to do that sort of thing, and then there is the risk that something goes wrong during the mission and you lose the satellite or even the crew if you're unlucky. There is also the extra cost and mass when building the satellite in making parts repairable or adding provisions for retrieving it, which is wasted if everything actually goes well.

It will practically always be more cheaper to just build and launch a new satellite. Building a couple of replacement Hubbles would have been much cheaper than flying 5 servicing missions.

[magnemoe]

Yes, but again those capabilities made no sense.

For a repair or retrieval, you spend months training for EVA operations, developing procedures and flight rules, fabricating special tools and fixtures. You need to pay lots of highly qualified engineers for several months to do that sort of thing, and then there is the risk that something goes wrong during the mission and you lose the satellite or even the crew if you're unlucky. There is also the extra cost and mass when building the satellite in making parts repairable or adding provisions for retrieving it, which is wasted if everything actually goes well.

It will practically always be more cheaper to just build and launch a new satellite. Building a couple of replacement Hubbles would have been much cheaper than flying 5 servicing missions.

Downside of using the shuttle for repairs was that it was so expensive to launch.

The training and special tools is the cheap part here.

Now if you had an small shuttle for repair work it makes more sense.

Even smarter to have satellites who don't need special orbits close to an space station so you can take them in for service or use an service craft operating out of the space station.

This has the benefit that if you find you need another part the satellite will just stay offline until the next service run.

[Sandworm]

Repairs only make sense if the repair vehicle (shuttle) weighs less than the object being repaired. Otherwise, it's almost certainly cheaper to launch a total replacement.

But that assumes no downside to de-orbiting or parking the broken object. So every time shuttle repaired something I had to think that there was a compelling reason why the object couldn't instead be thrown away. Perhaps they were dangerous. Or perhaps they contained material or information that they didn't want falling down on potentially enemy territory. Given hubble's commonalities with then-current spy satellites, perhaps they needed to keep it up there at all costs.

Rather than stealing soviet sats, those advocating shuttle's hardware return capabilities were probably looking at future orbital weapons programs. Such devices would have needed a secure method to bring down hardware for refit as weapons systems could never be de-orbited or left to rot (the 'Space Cowboys' scenario). Nor could the public ever be allowed to see them. This use of shuttle to maintain weapons systems surely is what triggered the Buran program.

Edited October 14, 2013 by Sandworm

[OdinYggd]

It's not so much it was leaky, as that cold weather caused the O-rings to not seal.

The failure was situated at one of the SRB field joints- where they are assembled at in the VAB while full of fuel.

At the time the SRBs used two O-rings, and it was known that on ignition it would leak a little initially before the rising pressure forced it out of its groove and into position to form a seal.

Because of the low temperature the O-ring didn't deform fast enough, the leak became too hot and too severe for the O-ring to survive it. Casing began to melt, as did orange tank. The rest is visible in the footage of the flight- you can see the jet of fire coming off the faulty field joint and striking the tank shortly before detonation.

Following that disaster they redesigned the SRB field joints to use 3 o-rings instead of 2 in order to ensure a more reliable seal, and enforced limits on allowed launch temperatures so that the O-rings would not be unacceptably cold at SRB ignition.

But the day of the challenger launch, they didn't expect it to be a problem. Up until that point the O-rings had always formed a seal moments after SRB ignition without issue.

[magnemoe]

Repairs only make sense if the repair vehicle (shuttle) weighs less than the object being repaired. Otherwise, it's almost certainly cheaper to launch a total replacement.

If you had an pretty much reusable system, it would make sense to repair an expensive satellite. Otherwise you are right.

But that assumes no downside to de-orbiting or parking the broken object. So every time shuttle repaired something I had to think that there was a compelling reason why the object couldn't instead be thrown away. Perhaps they were dangerous. Or perhaps they contained material or information that they didn't want falling down on potentially enemy territory. Given hubble's commonalities with then-current spy satellites, perhaps they needed to keep it up there at all costs.

Rather than stealing soviet sats, those advocating shuttle's hardware return capabilities were probably looking at future orbital weapons programs. Such devices would have needed a secure method to bring down hardware for refit as weapons systems could never be de-orbited or left to rot (the 'Space Cowboys' scenario). Nor could the public ever be allowed to see them. This use of shuttle to maintain weapons systems surely is what triggered the Buran program.

For space stations you are right you would need to build and service them. Same with other large structures.

However placing nuclear weapons in space has no benefit over using ballistic missiles from submarines.