Surely the only thing holding back hot high pressure gas isn’t a rubber packing or O-ring is it? The reason I ask is there is a certain amount of tin-foil nuttery available on various websites that posit that the cause was NASAs switch to a non-asbestos type putty which provides protection around the packings/o-rings.
Hey, I was just reading about that on another thread about asbestos – the one where somebody fell through their roof and was worried about the potential asbestos inhalation. What a coincidence.
The nuttery is not true, of course.
The composition of the segments is solid propellant inside, then a layer of insulation against the case wall, the steel case, and the cork TPS outside. The joints between the segments are a tang & clevis groove joint (see here for original joint design). You can see that there’s an aluminized potting compound (referred to as “putty”) which is basically smeared in where the insulation joins and into the channel where the O-rings lie, then the sections are tied together with pins that go around the circumference. Nominally the pressure of the case pushes the joint closed. The primary purpose of the O-rings is to make a seal such that when the case is under pressure and sees some bending load that tends to pry the joint open, the putty can’t escape and flow out. What they found in practice is that you’d get some small voids in the putty where it gets squeezed by handling or because the O-rings were insufficiently elastic at cold temperatures, and the hot gases inside the booster could push past the insulation and actually erode the O-ring. Also not intended was the fact that the joints flexed more than allowed for, permitting gases to slip past the potting, and the fact that the design of the groove was not optimum for the O-ring seal. This was a problem that was well-known long before the STS-51-L launch–see above–but because it hadn’t caused anything else to fail, or otherwise compromised performance it was not considered a failure. In fact, one analyst elected that since it only eroded the O-ring by 1/3 of it’s thickness that the part actually had a factor of safety of 3, despite the fact that O-ring erosion was neither part of the design allowable nor anticipated in function by the manufacturer.
By the time of 51-L (the 31st launch) program-level managers had forgotten about or dismissed early evaluations of flight failures, and were eating their own press about what a milk run flying the Shuttle was. (The same was decidedly not true for the engineers responsible for risk evaluation and quality assurance; when Feynman did his “loose cannon” thing at KSC he found that while anonymous estimates of flight failures from a group of engineers were in the 50-200 range, a program manager estimated a failure rate of 1:100,000, a completely absurd number even if the STS were without significant design flaws.) Despite warnings by both NASA and Thiokol engineers about launching the STS at cold temperatures (51-L was the coldest temperature at launch, and also had some of the highest wind shear, exacerbating both aero bending loads and O-ring compliance) the program people were unanimous about the launch being a go. In this case, a lack of resiliance of the cold O-rings, possibly compounded by bending loads on the SRB resulted in potting compound being squeezed or burned out, O-rings burning through very shortly after ignition, and the aformentioned jet of hot propellant gases escaping and doing damage to truss structure and ET integrity.
Post-Challenger the joint was redesigned (see diagram here) to be self-reinforcing so that a bending load would tend to close it regardless of how the bending loads were applied, and a third O-ring added (all of which were properly sized per O-ring design specs). The internal insulation was also modified to protect the joint from seeing internal gases, and a heating strip was added to heat the O-rings in the case of low ambient temperatures. I believe they also changed the potting compound, but I don’t know the details on why or what was improved, and whether asbestos was used before or not. The design still isn’t ideal–you’d prefer to have no field joints in the case–but it’s about the best practical way to use the SRB as intended by the Concept of Operations, and despite careful examination and analysis they haven’t found any repeat of the type of problems that resulted in the destruction of Challenger. The Shuttle SRBs are, in fact, the most reliable large solid booster flown, and the only solid rocket motor man-rated for orbital flight. There were plans to build an Advanced Solid Rocket Motor with a fiber-wound composite case and monolithic propellant pour like the large Trident and Peacekeeper motors (so, no segments) which would have been built at the integration facility and would have increased payload mass substantially, but the program was delayed by budget adjustments and eventually ran over budget without ever building a single live test article.
Stranger
I swear I read somewhere, Von Braun said something to the effect that “solid fuel boosters have no place in manned space-flight” or something like that, which isn’t the last word on the subject but probably should be considered pretty carefully.
For this and about a thousand other little details expressed clearly and elegantly, thanks Stranger, I feel I could always depend on your kindness.
von Braun was an admirer of Robert Goddard (from whom his early designs bear much credit) and was definitely a proponent of liquid boosters. The higher performance and controllability of liquid-fueled rockets gives them significant advantages over solids, which ignite and thrust at whatever rate available surface and chamber pressures will allow for; the only effectve way of moderating thrust is either to perform deliberate energy-wasting maneuvers, or inject additional cold propellant into the plume to affect exhaust momentum and expansion. Early solids were quite variable in performance from unit to unit and uneven mixtures had a tendency to crack from thermal stress or due to outgassing of volatiles, expel whole chunks of propellant, or spontaneously detonate from internal shock resonance.
However, by the early 'Seventies, all the major competitors in the STS proposal had a large amount of experience with large solids either as a primary rocket or as assist boosters such as used on Titan-IIIC. The Air Force had also done a lot of research on large (156 and 260 inch) solids for heavy ICBM and quick launch satellite boosters where reliability was paramount. The major thrust (no pun intended) for using solid motors for the Shuttle boosters, however, was cost and schedule. Several original proposals involved using various winged or semi-winged “fly-back” or recoverable and partially reusable Saturn-IC variants, but they couldn’t get the IC within the cost guidelines. Other proposals involved using LOX-kerosene pressure-fed fully reusable fly-back boosters (as was eventually proposed for the second generation Russian Buran Shuttle system), but development costs and schedule risk were major concerns; for all of their risks and lower performance, the solid fuel SRBs represented a much less complex development, and would ultimately be a lower risk flight item than complicated liquid boosters with plumbing and and pressurization system with cryogenic oxidizer. The estimated savings were to the tune of US$700M in FY1971 dollars. Additionally, the loss of a solid booster case and avionics would be less costly to the program than the loss of a pressure-fed reusable by a factor of about 5, despite the cost of recovery operations, refurbishment, and remanufacture of solid boosters.
The concern about failure of field joints was known from the very beginning; unfortunately, the understanding of behavior of field joints on such a large diameter booster was apparently unknown to designers of the SRB, and by the time problems became apparent upper management at Morton-Thiokol and NASA refused to acknowlege it lest it be used as an example of failed design. (Never mind that the STS was essentially an experimental design that, with refinement, could be vastly improved in capability and reliability, and simplified in operation and maintenance.) This short-sighted attitude and an undue rationale of “it hasn’t failed yet, so safety margins must be getting better and better” was the underlying cause of failure of the SRB from a known problem (and a similiar cultural attitude caused the institutional blindness that resulted in Columbia’s spectacular re-entry failure). This problem could have been resolved before with a redesign at a price significantly less than what it cost to replace the destroyed orbiter and having the Shuttle fleet grounded for two and a half years with no alternative heavy lift capability readily available. A major portion of NASA’s problem is that they’re not allowed politically to admit to error, ignorance, or even unforseen problems or cost impacts.
Stranger
Bolding mine
Could you elaborate on this please? Sorry if it seem as if I’m beating a dead horse here but are you saying that the astronauts would have died regardless? To my watching the “explosion” precluded them being able to use emergency measures.
There were no “emergency measures” to be taken. There is no plausible abort mode at the altitude in which the failure of the ET structure failed in which the Orbiter could be piloted to an RTLS abort or the crew could safely egress, and there were no means at the time of the Challenger failure for the crew to bail out. Columbia originally had ejection seats but only for the pilot and copilot and these were removed once Columbia was altered from test configuration to operational configuration, and these were only useful for a small portion of the ascent (and not at all for landing). Challenger and later Orbiters never had an ejection escape system. There is a system now for a manual bail-out procedure, called the Inflight Crew Escape System (ICES) but it was only designed for a scenerio in which the Orbiter was in controlled flight and simply couldn’t make a designated abort landing site, resulting in a Loss Of Crew and Vehicle (LOCV). Nobody really has any confidence that the complex, manual ICES system will actually permit safe egress in any real emergency; it’s mostly a face-saving measure to assure the public that NASA has done everything possible to ensure the safety of the crew, without admitting that in any practical sense it can do nothing more than maximize the likleyhood of a successful ascent.
Once the ET ruptured and vomited a large mass of fuel and oxidizer, the sudden increase in net thrust tore the STS system apart. The structural margins of many of the components are barely over unity at Max Q (the maximum aerodynamic loading) and a flight mode that is not aerodynamically stable will induce structural loads in the Orbiter and ET/SRB support structure that will cause it to literally tell itself apart. I know from the videos that it looks for all the world like the Shuttle “exploded”, but this isn’t at all what happened. The combustion of fuel and oxidizer did little more than char the exterior; the temperatures of combusting hydrogen and oxygen at atmospheric pressure are significantly less than much of the Orbiter sees during descent, and the pressure of the burning propellant was substantially less than stagnation pressure of the atmosphere against the Shuttle’s forward surfaces. Once the STS started to come apart and the starboard booster seperated, the Orbiter spun around and wind pressure from its forward movement tore the thing apart like wet blotting paper. Even if the astronauts had been wearing parachutes and pressure suits (which they weren’t–at that time, launches were done in a “shirtsleeve environment” with a partial pressure “clamshell” helmet) there’s simply no way they could have made their way to the exit portal and ejected, and despite claims that several crewmembers activated their PEAPs (which were only intended for a launch pad emergency, not inflight use) its unlikely that any crew maintained consciousness for more than a few seconds after breakup, as the remaining structure attached to the cabin was in a flat, uncontrolled spin.
So yes, if the propellants had escaped without combusting, the Shuttle still would have broken apart, and the crew would have died.
Stranger
Geez Stranger, if ya can’t offer some informed information to the discussion you should just opt out. 
As always, I am awed by the insight you have provided. Thanks
ETA: FWIW, I stood in my front yard east of Orlando and watched the Challenger destruction. And yes, I will no longer refer to it as the “explosion”.
Yes, from now on I will say, “I remember when the challenger’s ET ruptured and vomited a large mass of fuel and oxidizer, and the sudden increase in net thrust tore the STS system apart. Good thing it didn’t explode!”
Why is this considered inconclusive? What steps did the crew have to do to activate their packs?
The PEAPs are typically mounted behind the seat (although I find from some informal comments that indicate that PEAPs for some of the mission specialist chairs are mounted aside) and are intended for use in the case of an on-pad emergency where the air becomes contaminated. They aren’t intended to serve for a depressurization failure. Actuating a PEAP (which is basically an oxygen pack attached to the partial pressure suit) is done by turning a dial to the open position; however, crew members would have to get up out of their seats to actuate their own (in the case that it is mounted behind) or someone else’s PEAP. Given the initial impulse of somewhere around 20g and the instantaneous depressurization, it’s almost certain that the crew was initially rendered unconscious. It’s possible that some crew recovered consciousness and attempted to activate the PEAPs, although this would have involved getting up out of chairs and moving around, which seems unlikely given the spinning and tumbling behavior of the Main Cabin, but it’s at least equally likely that the PEAPs were activated by the jolt of impact. There is no definitive conclusion on precisely what killed the crew or whether they were conscious upon impact. The PEAPs were removed some time after STS-51-L as being unnecessary and the crew began wearing the modern full pressure ACES suits for ascent.
Regardless, there wasn’t then and isn’t now any means for recovering from a flight instability that causes structural failure, or indeed, any means whatsovever of egressing during powered flight or from an unstable flight mode. Even if the booster just stopped burning prematurely, or the fuel tank ruptured without igniting, the aerodynamic loads from moving through the air at nearly Mach 2 at altitude would have torn the Orbiter apart as it rolled or spun without any combustion or explosion occuring.
Stranger
[sup]Coloring Mine[/sup]
Under the circumstances, the question would seem to be: could the crew have physically accomplished the task of turning the dial. In other words, why is this considered inconclusive when the crew could not physically have done so much as to get out of their restraints without bouncing around the cabin like so many ping-pong balls in a Bingo shuffler?
I understand that the initial impulse of 20g didn’t last, but I remember taking sailing lessons and being amazed at how easily I was tossed around when I jibbed incorrectly. The forces involved there can’t be anywhere near what would have been involved in the Challenger disaster.
After the initial breakup, the Main Cabin would have been in free-fall (or nearly so) until it reached terminal velocity, and was probably in a slow, relatively flat spin. It’s possible come of the crew could have reached the PEAPs to activate them, but it’s also possible that the impact of hitting the ocean’s surface jarred the dials open. We don’t have good telemetry information on the cabin’s motions, so it’s uncertain what conditions were actually like.
Stranger