O Rings - A Second Look

Isn’t it true there are O-rings in your engine? If so, how come they can withstand all kinds of weather, but not the O-rings of the Challenger Space Shuttle?

  • Jinx :confused:

First off let’s rephrase the question somewhat.

  1. Does the engine in my car have o-rings in it? If you are just talking about the engine, the answer is maybe. If you add in other items like power steering and A/C the answer in almost for sure.
  2. Do these o-rings fail? All the time. However, the failure of an automotive o-ring does not make the car blow up.
    Getting abck to the space shuttle, the o-rings were designed to withstand the max temp inside the SRB, little or no thought was given to to what lower temp the o-ring had to remain plyable to. I sure that when the guys designed those o-rings sat down to design them and specify the material, none of them were concerned about the o-ring remaining plyabe at 32F. After all we are launching in freeking Flordia where it is hot. So they went about there business and designed an o-ring that could stand the multi-thousand degree temp of the SRB ignition, but became brittle when chilled to 32F. Further proof that nature always sides with the hidden flaw, and that Murphy guy can be a real asshole. Also the temps that the o-rings in your engine are exposed to are way lower than the temps that the shuttle o-rings are exposed to when the SRB are burning

The next time yuor engine has to withstand re-enty from orbit, let me know.

Aw, come on Rick. Let’s not speak ill of the dead. (Although from what I read, he might have been a bit of a jerk.)

Pretty much chiming in to agree with Rick here. O-rings are made out of a number of different materials, and no material is perfect. Different applications will have different temperature ranges, different pressures, different fluids, different dynamics. If you use the wrong O-ring, it’ll fail, simple as that. The O-rings in, say, your power steering system are designed for low to moderate temperatures, moderate pressures, and exposure to power steering fluid. They work well for taht; it’s likely that they wouldn’t work so well in a Shuttle application.

If you wanna know even more, try the Parker O-ring handbook

The o-rings being referred to failed shortly after takeoff, and never achieved orbit let alone re-entry.

That kind of thinking would have killed Edison’s “bright” idea. You’re all missing the point. Such items as O-rings become brittle in cold weather, hence they break. Personally, I thought each piston has an o-ring seal on it. If that were to fail due to cold weather alone, you’d be buying a new car every spring. Forget about NASA’s higher stakes here…weather is weather!

If pistons do have such o-rings that withstand temperature change, how come my car’s rubberized materials can outperform NASA’s o-rings at the same temp? Or, do you mean to say the engineers specified “all rubber seals shall be of bargain material meant for a toasty 70 F and up?” Of course not. Shouldn’t their rubber have worked across just as broad a temperature range? Afterall, that part is NOT rocket science!

  • Jinx :confused:

The SRBs don’t even make it that high when they work flawlessly. They are jetisoned off at 150,000 ft. and make it as high as 220,000 before the harsh mistress plays her hand.


Rick’s the man to ask here. He’s a car guru. But I’ll give it a shot: Car pistons do not have rubber O-rings. They have metal rings that go round the pistons. Pistons expand and contract under varying temperatures, so they are smaller than the cylinder. The metal piston rings maintain the seal. The cylinder head has a gasket that is made of either metal (copper?) or some other matierial.

Some thoughts
the disaster.

It is humbling, IMHO. In the 1960’s, using newly invented technologies, human beings left orbit and stepped onto the moon. Yes there were deaths and failures of gear.

The Shuttle Program is no different. Millions of parts, all of which must work perfectly together ( or, if failing, rely on redundancy and backups ).


“… all built by the lowest bidder.”

If pistons do have such o-rings that withstand temperature change, how come my car’s rubberized materials can outperform NASA’s o-rings at the same temp? Or, do you mean to say the engineers specified “all rubber seals shall be of bargain material meant for a toasty 70 F and up?” Of course not. Shouldn’t their rubber have worked across just as broad a temperature range?
Refer back to zut’s post. Even if o-rings are used in cars, it’s not an apples to apples comparision. The same rubber which can withstand colder temperatures may not have the right properties at the greater pressures required in the SRB, for example.

The point is that there are always engineering trade offs, and temperature is one of the factors considered. And yes, since the space shuttle takes off from Florida, the specifications were created based on the assumption that the temperature be above freezing. At the time, a friend from school worked as a tech at the company which manufacturered the o-rings, and verified that this was the spec.

Had there be a need to launch rockets from Alaska, they could very well have found material which would have been worked. The problem was not the engineers, but within NASA.

An interesting read is book by Richard Feynman, the Nobel Prize winner and a member of the committee investigating Challenger disaster.

IIRC, there was a program on PBS (I think) several years ago concerning the Challenger disaster and what caused the o-ring failure. The upshot of it was that these particular o-rings were replacements, as the original design was no longer available due to the phasing out of products containing asbestos. They had used existing stock up to this launch, but had to switch to the new compound for Challenger. The replacement o-rings met the original spec, but were less compliant at lower temperatures. The freakish weather at Canaveral that day was outside the design parameters for the new material, so they were stiff, didn’t seal properly, and seven of our best and brightest entered the history books that morning. All because of a ban on asbestos.

If I have mis-remebered this in some significant fashion, I hope someone will correct me.

There are a lot of different ways of answering your question. Firstly, it seems from your second post you are talking about piston rings. Piston rings are not o rings. They are an almost closed c-ring, I guess you could say. They don’t seal perfectly, they have a gap in them, for a start. They don’t have to operate in all kinds of weather. They don’t have to operate at all. An engine in which they are operating badly will still run, and will certainly not blow up. They are not at all comparable, in either design, construction material or purpose to the o-rings that failed in the space shuttle.

Secondly, there are o-rings in an engine. As others have said, there are few that would have to operate from freezing through to thousands of degrees and which are mission critical.

Thirdly, the underlying point is perhaps this: the o-ring that failed could have been designed so it worked. It wasn’t. The designers effed up. This happens. Designing complex things that work is hard. Try it sometime. The usual way you do it, if truth be known, is that you design it as best you are able, then you test, then you design again, rinse and repeat, then you release a working model then when faults show up under full operating conditions (as they inevitably do) you redesign some more.

Designing something complicated, expensive, and which is only operated extremely seldom (and which explodes when it fails) is a tough call.

The internal combustion engine was unreliable when it was first built. It has been in volume production and continuous development for a hundred years or more and building a reliable example is old hat. Trust me, the first few thousand times one was run, components failed.

Pistons do not have o-rings. I think it’s safe to say that o-rings would not survive inside an engine’s cylinder.

Piston rings are made of iron or steel.

Some engines with wet sleeve design use o-rings between the sleeves and the block, but these o-rings are outside the cylinder, not inside. There’s a world of difference between those two environments in terms of heat, pressure, and chemicals.

O-rings are sometimes used on engines, for example to seal where a coolant pipe joins the block. Again, a world apart from being inside the cylinder where the piston is.

As I recall the O-rings can’t withstand the extremely high temp inside the SRBs during flight. They are protected by a ceramic “putty”. The two had to work in tandem, i.e. rubber can withstand enormous pressures but not very high temperatures. The ceramic insulation can withstand emormous temps (think a pot in a kiln) but cannot hold back even a modest degree of pressure.

Upon ignition the pressure & temperature inside the SRB instantly increases tremendously. The force against the putty is transferred to the O-ring which is right against it (on the otherside, where its protected from the heat). If either one failed so would the other.

The extreme cold that day (coldest launch ever) made the o-ring hard and unable to maintain a seal. So a hole was blown thru the putty, past the O- ring to the outside. It was actually sealed by escaping residue soon after liftoff but unfortunately it became unsealed by aerodynamic vibrations. Without the ceramic putty the hot gases quickly burned right thru the rubber O-rings. It was also not a tragic coincidence that this happened right on a support strut. The strut acted as a heat sink (or “cold sink”) for the SRB and made it the coldest part of the joint (i.e. the most likely to fail).

As far as the OP goes, its apples & oranges. Car engines don’t rely on O-rings for any part of the combustion chamber. Besides, the SRBs generate many, many times higher temperatures & pressures than an internal combustion engine does…

::sigh:: Well, this is about the closest explaination to what happened.

What really happened was that the o-rings were (and frankly, still are) used in an application that o-rings are not really optimal for, i.e. acting as circumferential seals in a joint that sees both axial moments and torsional movement. This alone can put the o-ring in a compromised position; combined with expansion of the segements at the joint interface (due to reuse and wear) and the cold weather that made the elastomer o-rings less resilliant, allowed hot exhaust gases to seep past the potting compound (intended to protect the o-rings) and degrade the o-rings. It was actually well-known that this was occuring–NASA and Thiokol engineers knew that as much as 1/3 of the o-ring section was being degraded in a way that was not intended for any wear at all–but warnings from engineers fell on nearly deaf ears of upper level administrators, who judged the problem to be non-critical. As flights continued and evidence of degradation mounted, admins took this to be a sign of fewer problems, i.e. if it hasn’t failed yet then it is even less likely too in the future, rather than, “Oh shit, we’re in an out-of-design condition, and the further we push the envelope the worse it’ll get.”

Thiokol has, and continues to use this joint design, not only on the Shuttle SRBs but on virtually every solid rocket motor they build. (Granted, for the most part, this occurs at forward and aft domes where bending moment is insignificant and flexure is less critical.) The joint on the SRBs was reinforced but not actually redesigned to eliminate the problem conceptually, and other inspection and qualification measures were undertaken to avoid this issue, but it is an inherent problem with the design. The bigger problem, though, is with the NASA culture, where identification and announcement of problems is criticized rather than commended, an attitude that led to the Columbia disaster as well. (Again, several engineers were aware of the potential for problems but were poo-pooed by managers. Thank you, Linda Ham, you culpability-avoiding twit.)

There are no o-rings inside your car engine. There are piston seals, which occasionally need to be replaced if your car is driven very hard or has sloppy tolerances, but these do not behave like o-rings and are made of babbit metal (cast iron or ductile steel), not elastomer. O-rings in hydraulic systems (with which I am most familiar) fail regularly but almost never in any catastrophic manner; it is indicated by slight leakage of hydraulic fluid.


The book What Do You Car What Other People Think? by Richard Feynman contains a long and fairly detailed account of the O-ring problem on the Challenger. The accound corresponds pretty much with what Stranger wrote.

As was stated above, the solid rocket motor joint is entirely unsuited for an O-ring seal. Such seals are intended to used in a machined channel that allows the pressure being contained access to the back side of the seal. This expands the O-ring and compresses it into the gap and makes the seal. If the channel isn’t machined correctly so that there is too big a gap the O-ring won’t fill the gap properly and is likely to leak. And if the O-ring is not flexible, like when it is cold, it won’t compress properly and rapidly and this allows gas to leak by enough to erode the seal. And on top of all this the rocket motor cases are several feet in diameter, with a relatively thin skin so that it is hard to keep them dimensionally stable and this also can leave large gaps that the O-ring can’t fill. That is to say, the motor cases get out of round and fit together poorly.

Automobile engines have lots of seals, but not many of them are O-rings. The crankshaft has cupped oil seals front and rear. Each valve stem has a seal to keep oil from leaking into the cylinder and these are sometimes O-rings. There is a water pump seal which is usually a paper gasket. The cylinder head-engine block joint has a seal which is a combined metal-asbestos fiber (or other heat resistant material) gasket. The oil filter has a built-in seal on it that seals the joint between it and the engine block. The oil pan has a rubber-cork seal, or gasket and so does the valve cover on overhead valve engines. In fact the set of seals (gasket set) for an engine is extensive and is one of the significant parts-cost in an engine overhaul.

However only the head-block gasket is exposed to the combustion chamber temperatures and that part of it which is so exposed is protected by metal. The other seals that are exposed to such temperatures are the piston compression rings and they are cast iron. Babbit, by the way, is not used in rings. It is a tin, copper and antimony alloy used for bearing liners.

Er, I knew that. Me–>:wally

The reason that Thiokol continues to use o-rings in this application is because…they always have. Great reasoning. And since there is no one else (in the US) building large solid rocket boosters (now that P&W has gotten out of the business and everyone else has been bought up by ATK/Thiokol) they’ll continue to use the design concept, flaws and all, indefinitely.


Can you explain this in a bit more detail? Is it the direction of movement and/or pressure that makes it unsuitable for an O-ring joint, or just the fact that there is movement at all?