|
Post by PhantomWolf on Aug 22, 2005 7:19:41 GMT -4
Would it be possible to put some sort of mesh over the tank that would hold the foam in place, or at least mean that any that fell off would only be big enough to fit through the holes in the mesh?
|
|
|
Post by Captain Kidd on Aug 22, 2005 13:13:03 GMT -4
But that adds weight which reduces payload capacity. And what if the mesh detached and "netted" the shuttle?
|
|
|
Post by snakeriverrufus on Aug 22, 2005 20:38:20 GMT -4
Or we could just go back to the old way of applying the foam. IIRC, OSHA did not require NASA to change the method of application. Does anyone remember the details?
|
|
|
Post by JayUtah on Aug 23, 2005 18:29:35 GMT -4
The bipod ramp used the older CFC (Freon) foam. While the new foam (sort of like New Coke) is drastically more prone to shedding than the Freon-applied foam, it was the Freon foam that broke away and hit Columbia. NASA was granted an exemption by the EPA to continue to use the Freon-applied foam, which is still used where the foam must be applied by hand. I'm not sure why NASA still uses the weaker foam for the rest of the tank.
Retention layers are not always the best policy. Surface retention layers must undergo extreme thermal stress as well as aerodynamic buffetting. Embedded retention layers would probably make the problem worse: instead of shedding bits and pieces of foam, entire sheets of the material might peel away.
|
|
|
Post by gwiz on Aug 24, 2005 2:55:01 GMT -4
The bipod ramp used the older CFC (Freon) foam. While the new foam (sort of like New Coke) is drastically more prone to shedding than the Freon-applied foam, it was the Freon foam that broke away and hit Columbia. NASA was granted an exemption by the EPA to continue to use the Freon-applied foam, which is still used where the foam must be applied by hand. The large piece that broke away on the latest flight was from the ramp around some pipework that is hand-applied.
|
|
|
Post by JayUtah on Aug 24, 2005 11:08:30 GMT -4
Both Columbia and Discovery impacts were from hand-applied areas, the hand-applied foam being that which is still applied using Freon. There is some confusion since ETs built later than the one used on Columbia's final flight used non-CFC blown foam for the bipod ramp. And there is also some confusion because of early political attempts to blame the EPA for Columbia -- the rumor still persists that Columbia was destroyed "because of some stupid environmental regulation."
Hence NASA has been concentrating not on the blowing agent but on the manual foam layup procedures, which seems a prudent course of action given that they can't even get the "good" foam to stick. But there is oodles of evidence that the non-CFC blown foam is several times more susceptible to all shedding modes than the "good" foam. Just because we have two high-profile incidents of CFC foam having been shed doesn't mean there aren't hundreds more foam strikes attributed to the "bad" foam. NASA could immediately reduce the danger of foam strikes by returning to the CFC-11 application procedure even for machine-applied foam.
|
|
|
Post by Jason Thompson on Aug 25, 2005 16:48:19 GMT -4
Just out of curiosity, why was the decision made to use external rather than internal insulation for the EFT? Since the orbiter is bolted to the outside, I would have thought that the last thing you'd want from a design perspective is external insulation that could shear off under thermal and aerodynamic stress.
|
|
|
Post by JayUtah on Aug 25, 2005 17:17:01 GMT -4
Because you want the aluminum to get cold. It performs better at cryogenic temperatures, and that means you can use less of it.
|
|
Al Johnston
"Cheer up!" they said, "It could be worse!" So I did, and it was.
Posts: 1,453
|
Post by Al Johnston on Aug 25, 2005 17:20:30 GMT -4
Isn't the ET mostly made of Kevlar composite?
From what I recall of my time as a metallurgist, cryogenic temperature metals tend to be horribly brittle
|
|
|
Post by JayUtah on Aug 25, 2005 18:34:59 GMT -4
Isn't the ET mostly made of Kevlar composite?
No. It's made of aluminum. Standard construction methods: rings, skins, and stringers, with some special milling for weight savings.
From what I recall of my time as a metallurgist, cryogenic temperature metals tend to be horribly brittle
Except aluminum, especially in the 5xxx series of alloys. Aluminum alloys have no ductile-to-brittle transition at low temperature. Their performance at low temperatures is so good that they are categorically exempt from ASTM and ASME low-temperature testing protocols. Tensile and yield strengths in aluminum actually increase at cryogenic temperatures, from 10% to 40%. (Yes, really.)
Aluminum alloys are the magical material of aerospace, where cryogenic fuels are concerned.
In fact, you have to be careful when you hydrotest booster tanks, not only because water is heavier than liquid hydrogen but because the tank material will be weaker during the hydrotest than in operation.
|
|
|
Post by Obviousman on Aug 26, 2005 1:08:31 GMT -4
Except aluminum, especially in the 5xxx series of alloys. Aluminum alloys have no ductile-to-brittle transition at low temperature. Their performance at low temperatures is so good that they are categorically exempt from ASTM and ASME low-temperature testing protocols. Tensile and yield strengths in aluminum actually increase at cryogenic temperatures, from 10% to 40%. Ooooh, I love it when you talk dirty! (Sorry, but I couldn't resist. As a non-engineer, tech-speak sometimes goes right over my head...)
|
|
Al Johnston
"Cheer up!" they said, "It could be worse!" So I did, and it was.
Posts: 1,453
|
Post by Al Johnston on Aug 26, 2005 10:33:36 GMT -4
;D It was quite a long time ago And I had very little to do with alumin ium
|
|
|
Post by JayUtah on Aug 26, 2005 13:52:58 GMT -4
The ferrous metals, for example, have the brittleness properties you describe, and absent any specific knowledge of aluminum's special properties (and correct spelling :-)) it would be prudent to assume they were the same as for nearly all other metals.
...especially in the 5xxx series of alloys.
Metallic alloys are generally coded in series that share progressively more detailed properties. Aluminum alloys have four-digit numbers where the first digit specifies roughly the qualitative constituents. The 2000-series, for example, is alloyed primarily with copper and so all the 2000-series alloys have all the basic properties that come from mixing aluminum with copper. Individual alloys sport various helpful impurities or changes in the proportions, or have rolling or heat-treating processed applied to them to alter their materials properties. Rolling, for example, can make a metal anisotropic, or stronger in one dimension than in the other.
There are negative properties. Some aluminum alloys, for example, cannot be welded without destroying their properties. And so an engineer has to manage the give-and-take between desired production methods and appropriate materials. If he designs with one of the unweldable alloys, he must devise a method for manufacturing his design that does not involve welding.
Aluminum alloys have no ductile-to-brittle transition at low temperature.
Ductility is the property of a material to shrink in one dimension as it is placed under tensile load in the other dimension. Highly ductile materials are those that can be drawn into wires by pulling them through progressively smaller holes.
If you get a bar of steel and put it in the standard static tensile strength test (pull hard at both ends until it breaks) the steel will actually become narrower at a point along the bar, stretching and becoming thinner like taffy until it eventually fails. The degree to which a material stretches before breaking, and the forces required to accomplish each step of the process, give insight into the material's properties. And indeed, much can be inferred reliably from static tension tests.
Materials that do not stretch at all before breaking are non-ductile. There is a very qualitative difference between this "brittle" failure and ductile-tensile failure. A metallurgist can look at the separation faces and tell whether the material has broken under ductile tension (i.e., it has torn) or under brittle failure (i.e., it has fractured).
Steel is often used because of its ductility. That is, steel will actually stretch and deform under load and still retain some of its strength where other materials would have failed. This was brought to public attention in the investigation of the Titanic failure. The steel in the hull was expected to bend and deform under impact, but still to retain its watertight integrity. This did not happen.
As temperature decreases, many materials -- ferrous metals in particular -- undergo what what is called the ductile-to-brittle transition. That is, the very qualitative change occurs between ductile tensile failure and brittle tensile failure. Steel, at cold temperatures, fractures under tension like glass; it does not stretch and then fail. The transition is abrupt.
Aluminum is fairly unique among metals in that it retains its ductile properties even at cryogenic temperatures. So while steel would shatter like glass at the temperature of liquid hydrogen, aluminum continues to give and bend in response to loads. This makes it the material of choice for cryogenic facilities.
Their performance at low temperatures is so good that they are categorically exempt from ASTM and ASME low-temperature testing protocols.
ASTM is the standards body for measuring and categorizing the behavior of materials, especially under adverse conditions. You recall from our prior discussions that ASTM methods tested the steel used in the WTC towers and qualified them for fire resistance under certain circumstances.
ASME is the professional association of mechanical engineers. They publish, among other things, tables of the common properties of various materials used in engineering design.
Normally these standards and testing bodies will test at a variety of temperatures, specifically looking for brittleness. Aluminum is known to retain its properties at very low temperatures, and so these committees consider it a waste of time to test many of the material properties at low temperatures; they are assumed to be comparable to more medial temperatures.
Tensile and yield strengths in aluminum actually increase at cryogenic temperatures, from 10% to 40%.
The tensile strength of a material is the pulling-apart force per unit area of cross section that can be applied to pull a material before it separates. The tensile strength of most materials increases as temperature decreases, but the ductile-to-brittle transition of most other materials makes the behavior at or near that limit undesirable.
The yield strength of a material is that force which must be applied in order to induce plastic deformation. Deformation is generally divided into plastic and elastic. Elastic deformation is the change in shape of a material under stress such that when the stress is removed, the material returns to its previous shape. Plastic deformation is a change in the shape of a material that remains after the stress is removed. Most materials exhibit a combination of both kinds of deformation. Steel, for example, can be stressed in elastic deformation, but does not always return precisely to its former shape. The deformation in that case is mostly elastic, but partially plastic. The yield strength of a material is important when considering lifetime loads. If a load plasticly deforms a structural member, even if it does not fail, the new shape may be insufficient to sustain further loads.
Aluminum's yield strength increases dramatically at low temperature. That is, at low temperature aluminum becomes more "springy" and able to accept greater loads without permanently deforming.
These properties make the shuttle's external tank actually much stronger when filled with fuel than standing empty. This means that the flight dynamics forces can be computed with those more advantageous materials properties. Since you can have more strength with less aluminum, you plan for the cryogenic case.
That's why you want the aluminum in direct contact with the fuel. It's actually a big loss if you insulate the tank structure itself from the effects of the fuel. So given that you want to prevent heat transfer into the aluminum from outside, how do you do it?
|
|
Al Johnston
"Cheer up!" they said, "It could be worse!" So I did, and it was.
Posts: 1,453
|
Post by Al Johnston on Aug 26, 2005 15:15:52 GMT -4
IUPAC (the International Union of Pure and Applied Chemistry) give the official spelling of the element name as "aluminium", but they make allowances for American eccentricity Sailors on Liberty Ships in the convoys to Russia in WWII found out the hard way that the ductile-brittle transition in steel can occur uncomfortably close to 0 oC. In conventional (at that time) rivetted ships, this would not have been so much of a problem, but the welded construction of the new vessels meant that a crack could propagate all around the hull, and the ship could literally split in half with virtually no warning. Ouch. An answer to the ET problem might be some form of sandwich construction, with the insulation between stress-bearing layers. I can see that there would have been cost and performance reasons for not doing that originally, and it's too late in the life-cycle of STS for such a major redesign now. The CEV solves the problem by reverting to putting the payload in the traditional position at the top of the stack.
|
|
|
Post by JayUtah on Aug 26, 2005 19:13:38 GMT -4
...but they make allowances for American eccentricity [smug grin] In conventional (at that time) rivetted ships, this would not have been so much of a problem, but the welded construction of the new vessels meant that a crack could propagate all around the hull, and the ship could literally split in half with virtually no warning.Yes, but that too is a can of worms. Kaiser shipyards used welded construction, but at least one of the other shipyards stuck with rivets. And the fact that the same design could be assembled both ways is actually one of the downfalls of the welded construction. In order to use rivets you have to design lap joints, where one member overlaps the other so that rivets can be driven through both. If you weld a lap joint instead, you have to use fillet welds. A butt weld would be stronger, but it requires a butt joint. And that means having two major variants on the Liberty Ship design. Lap joints are also faster to cut. Remember that the Liberty Ships were not necessarily designed to survive their maiden voyages. And so they had to be built fast and cheap. A butt joint has to be cut and fitted up with greater precision than lap joints. That takes time. And so even if the entire production of Liberty Ships had been welded, they likely would have kept with lap joints. The fact that all the mysterious Liberty Ship failures occurred in the welded models and none in the riveted models suggests that temperature-related embrittlement is not the only cause. Likely the cold revealed one of the many flaws we know were present in the welding practices at Kaiser shipyards. An answer to the ET problem might be some form of sandwich construction, with the insulation between stress-bearing layers.Yes. The structural and heat-transfer requirements obviously required a separate solution for each. But now the outermost layer incurs an additional requirement to stay in place under aerodynamic (and possibly structural) forces. That requires either replacing the foam with an insulation that also has good fluid-flow and yield properties, or providing an additional retention layer. The latter has obvious weight penalties, but also places a new structural burden on the underlying foam. If you stick the retention layer to the foam, the foam has to have sufficient shear strength to hold it there. If you stick fasten it through the foam to the skin, you now have a heat conduction path. Keeping the material thin so that it doesn't require excessive shear strength means it has to have high tensile strength and will likely also be highly flexible. That's undesirable in the case where foam does break loose -- your retention layer then does more harm than good by pulling off more foam. And none of this is necessarily going to help if you have spalling, which is one of the common foam failure modes. Pockets of ice from water than has permeated the foam turn to steam and pop out pieces of the foam. You could argue that the outer layer would reduce water permeation and thus steam-related spalling. The outermost material's thermal properties can't be completely ignored either. The outer layer deals with aerodynamic heating on the ascent, so it would have to retain good structural properties at high temperature, again without too much of a mass penalty. I'm stumped. The CEV solves the problem by reverting to putting the payload in the traditional position at the top of the stack.There is much to be said for that. The Saturn V shed material (ice, etc.) like crazy, but no one cared since there was nothing crucial below. Side-staging or parallel staging isn't inherently bad if the payload can stand it. Ice shedding would have been a problem even for the original shuttle design. The delicacy of the orbiter TPS is the limiting factor here, and likely would have been the limiting factor in the earliest STS design. Its inability to withstand safely even the most minor impact imposes an unrealistic requirement on the rest of the system. You can have a delicate TPS if you just keep it out of harm's way, such as by traditional serial staging.
|
|