What if the O-rings were modified before the launch of STS-51L?

[Vahe231991]

We know that the Challenger tragedy was caused by weak O-rings, but if NASA engineers had made modifications to the O-rings to withstand cold temperatures in 1985, would the STS-51L mission have gone smoothly, in which case all Space Shuttle missions planned for the rest of 1986 and 1987 would materialize?

[AS_501]

The O-rings failed to seal because they became brittle in the unusually cold air temp. (~35 F). An elastomers engineer can explain if an alterative O-ring material was available that remained elastic at such temperatures.  The joint design was modified to prevent hot gas from impinging on the O-ring in the first place.  Heaters were also installed in the new joints to keep the O-ring pliable.  But NASA would never risk a cold-air launch regardless of the heaters.  Some other points to ponder about the accident:

-  Scorching was seen on the O-rings as early as STS-2.  The whole program should have come to a screaming stop at that point.

-  What if the the breach occurred on the opposite side of the SRB, away from the ET?  Would it have broken apart before SRB sep? We'll never know.

As always, I defer to our NSF experts on this topic.

[Jim]

We know that the Challenger tragedy was caused by weak O-rings, but if NASA engineers had made modifications to the O-rings to withstand cold temperatures in 1985, would the STS-51L mission have gone smoothly, in which case all Space Shuttle missions planned for the rest of 1986 and 1987 would materialize?

It wasn't just the o-rings.  It was the SRM segment mating interface that needed to be changed. Look up SRM capture feature.

[MDMoery]

Another question one can ask is, what if the closeout crew would have used the proper GSE equipment on the orbiter access hatch during the Monday, January 27, 1986 launch attempt?  If they would have used the correct tool, STS-51L lifts off in 50 degree weather.

What would have happened?  Sometime in the next 18 months, probably a different LOCV Shuttle disaster.

The Challenger investigation did not uncover just o-rings.  It uncovered a large number of problems and a system that was being pushed to breaking point to meet a flight rate that was completely unrealistic for the STS that they actually designed and built (vs the one they sold to Congress in 1972).  Even if the o-ring problem never existed, it is my view that there would have been just a matter of a short amount of time before another type of accident would have occurred.  They did 9 flights in 1985 and were pushing for something like 12-14 in 1986, including as I recall no less than 3 interplanetary probes (Galileo, Magellan, and Ulysses) and the Hubble Space Telescope.

[DaveS]

The big ticket items for NASA in 1986 was Galileo, Ulysses, HST and ASTRO-1 on Columbia to observe Halley's Comet as it passed perihelion. By 1986 Magellan had slipped into June 1988 and was due to launch on a smaller Centaur G upper stage and HST was likely to slip into 1987 due to ground segment readiness.

[AS_501]

The big ticket items for NASA in 1986 was Galileo, Ulysses, HST and ASTRO-1 on Columbia to observe Halley's Comet as it passed perihelion. By 1986 Magellan had slipped into June 1988 and was due to launch on a smaller Centaur G upper stage and HST was likely to slip into 1987 due to ground segment readiness.

Didn't the Challenger accident also put an end to shuttle polar launches from Vandenberg?

[DaveS]

It did. The plans were to ferry Columbia out to Vandenberg for more extensive tests including an FRF at the pad. This was something Enterprise couldn't do as she lacked any semblance of an MPS. This was due to happen sometime after the ASTRO-1 mission, so sometime around April/May 1986. After Columbia's return to KSC, Discovery was to be ferried out to Vandenberg to begin processing for the first flight from there (STS-62A).

[Hog]

The big ticket items for NASA in 1986 was Galileo, Ulysses, HST and ASTRO-1 on Columbia to observe Halley's Comet as it passed perihelion. By 1986 Magellan had slipped into June 1988 and was due to launch on a smaller Centaur G upper stage and HST was likely to slip into 1987 due to ground segment readiness.

Didn't the Challenger accident also put an end to shuttle polar launches from Vandenberg?

Pretty sure the Filament Wound Case (FWC) that bought back performance lost due to the Southerly launch had the improved Capture Device, but only 2 o-rings not the 3 rings the STS SRBs had post 51L.

[libra]

Another question one can ask is, what if the closeout crew would have used the proper GSE equipment on the orbiter access hatch during the Monday, January 27, 1986 launch attempt?  If they would have used the correct tool, STS-51L lifts off in 50 degree weather.

What would have happened?  Sometime in the next 18 months, probably a different LOCV Shuttle disaster.

The Challenger investigation did not uncover just o-rings.  It uncovered a large number of problems and a system that was being pushed to breaking point to meet a flight rate that was completely unrealistic for the STS that they actually designed and built (vs the one they sold to Congress in 1972).  Even if the o-ring problem never existed, it is my view that there would have been just a matter of a short amount of time before another type of accident would have occurred.  They did 9 flights in 1985 and were pushing for something like 12-14 in 1986, including as I recall no less than 3 interplanetary probes (Galileo, Magellan, and Ulysses) and the Hubble Space Telescope.

Mike Mullane is all too clear about this: the 1981-86 Shuttle system had many different dangers lurking everywhere.

Including the brakes and undercarriage at landing (from memory). 

The flight before the Challenger disaster already had dodged major bullets (the broken temperature probe and the insufficient amount of LOX... two deadly disaster that somewhat cancelled each others) 

[DaveS]

The big ticket items for NASA in 1986 was Galileo, Ulysses, HST and ASTRO-1 on Columbia to observe Halley's Comet as it passed perihelion. By 1986 Magellan had slipped into June 1988 and was due to launch on a smaller Centaur G upper stage and HST was likely to slip into 1987 due to ground segment readiness.

Didn't the Challenger accident also put an end to shuttle polar launches from Vandenberg?

Pretty sure the Filament Wound Case (FWC) that bought back performance lost due to the Southerly launch had the improved Capture Device, but only 2 o-rings not the 3 rings the STS SRBs had post 51L.

Yes and another link in the chain that led to the loss of Challenger and the crew was that the fact that every time the SRBs impacted the ocean it slightly warped the shape of the segments which lead to the "out-of-round" problem that allowed the joints to flex enough to allow for the hot gas blow-bys.

At the time they were unsure if the FWC SRBs segments could be reused. This from Jesse More's testimony to the Roger's Commission:

We have a major question that the program is looking at right now, and we probably won't get any good data on that until later downstream, and our question among others that is currently on the table about the graphite/epoxy cases today, is can we reuse them? You know, currently we reuse the steel cases. The Shuttle returns, it has engines the back, the SRBs are returned. They have parachutes on them. We go back and retrieve the SRBs and go through and refurbishment cycle on to reuse them. For the graphite/epoxy cases , we are doing some final testing at this point in time, and we are not sure whether or not we can reuse those filament wound cases after we fly them and they come back and impact the ocean. We have not made an determination like that, so we are not planning to get out of the steel SRB business at this point in time. We have a lot of additional work on the filament wound cases.

This his answer to question asked by Dr. Arthur Walker, a member of the Commission if NASA had any plans to phase out the steel SRBs once the composite SRBs had phased in. This exchange can be found on page 18 of this document: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19860019504.pdf

[libra]

the fact that every time the SRBs impacted the ocean it slightly warped the shape of the segments which lead to the "out-of-round" problem that allowed the joints to flex enough to allow for the hot gas blow-bys.

I did not knew about that. Very interesting (and damning, too - for reusability triggering more problems down the line).

[Surfdaddy]

Given the NASA culture and the push for flight frequency, it seems inevitable that sooner or later we were going to lose another vehicle if it hadn't been Challenger.

In 1988, the classified military shuttle mission STS-27 was nearly lost due to hundreds of damaged and at least one lost tile from a debris strike during launch. Near burn-through was only thwarted through total luck that under that spot contained an antenna's backing steel plate rather than fuselage aluminum, or it would have melted through. Of course this wasn't publicized at the time due to being a classified mission.

[catdlr]

Given the NASA culture and the push for flight frequency, it seems inevitable that sooner or later we were going to lose another vehicle if it hadn't been Challenger.

In 1988, the classified military shuttle mission STS-27 was nearly lost due to hundreds of damaged and at least one lost tile from a debris strike during launch. Near burn-through was only thwarted through total luck that under that spot contained an antenna's backing steel plate rather than fuselage aluminum, or it would have melted through. Of course this wasn't publicized at the time due to being a classified mission.

https://forum.nasaspaceflight.com/index.php?topic=29360.msg926834#msg926834

https://forum.nasaspaceflight.com/index.php?topic=29360.msg926842#msg926842

[libra]

Given the NASA culture and the push for flight frequency, it seems inevitable that sooner or later we were going to lose another vehicle if it hadn't been Challenger.

In 1988, the classified military shuttle mission STS-27 was nearly lost due to hundreds of damaged and at least one lost tile from a debris strike during launch. Near burn-through was only thwarted through total luck that under that spot contained an antenna's backing steel plate rather than fuselage aluminum, or it would have melted through. Of course this wasn't publicized at the time due to being a classified mission.

The most bizarre (and saddening) aspect of STS-27 was that the clear pictures they got in orbit from the damage, were mis-interpreted on the ground - for a pretty damning reason. The mission was a military one and the necessary encryption of coms and datalinks had the ground seeing only the pictures "low res" and thus missing the damage. Commander Hoot Gibson was rather baffled and furious at such absurdity.

Imagine: you KNOW the damage is there, you have the pictures showing, yet it is not possible to convince the ground of the severity of the damage because the pictures transmission degraded them to low-res.

Then again, what could they have done is probably little if nothing (thinking about the CAIB studies of Columbia rescue resulting in very slim chances of saving the crew via another Shuttle launch).

A very scary and absurd near-miss, really - and coming only the next flight after RTF and two flights obviously after STS-51L. Not only the crew but the entire Shuttle program dodged a big bullet that day, and only thanks to that steel plate / antenna located at the perfect place at the perfect moment.

Sweet Jesus...

[EGREG51]

We know that the Challenger tragedy was caused by weak O-rings, but if NASA engineers had made modifications to the O-rings to withstand cold temperatures in 1985, would the STS-51L mission have gone smoothly, in which case all Space Shuttle missions planned for the rest of 1986 and 1987 would materialize?

It wasn't just the O-rings.  It was the SRM segment mating interface that needed to be changed. Look up SRM capture feature.

Actually  three primary factors  came together to cause the leak at the 304 degree location on the RH Aft field joint:

1.  The segments used to create the field joint  had flown previously (L60 on STS-41B and L06 on QM-4 and STS-51C).   Both segments suffered high splashdown velocities (<89ft/sec) as a result of the being flown with the 115 foot parachutes. NASA was in the process of fixing this flaw by replacing them with 136 foot chutes.   These high splashdown rates resulted in flown segments having an increasing degree of ovality which lead to a larger variation of mating tolerances (loose and tight fit locations around the joint).   In the case of 51L, an extremely tight fit was measured at the 120/300 degree clocking location at mating.  (Rogers Commission Report Appendix C)   This tight fit was a key factor in the subsequent leak

2.  The tight fit had to occur in a location on the joint where it would persist from stacking to launch.   After stacking the joint tended to change tolerances as a result of transportation to the pad and the heating and cooling while on the pad.   This allowed a tight or loose stacking location to move to a more nominal state.   In only one location on the SRB could a tight stack persist to launch.   (STS-61E unstacking report)

3.  Temperature needed to be below 55F.    The temperature had a significant impact on the ability of the joint to seal, but not because of the O-ring.   Testing by Dr. Carleton Moore found that a "cold" O-ring took up to 8 milliseconds to seal while a "warm" O-ring took 3 milliseconds to seal.     This difference was well within the 250 millisecond period in which an O-ring must translate across the gap and seal.    As was shown in later testing presented to the Congressional investigation, the temperature impact on the putty was the more significant environmental variable. At temperatures below 55F the putty could prevent the pressure from reaching the joint for up to 700 milliseconds.     A significant delay in the pressure reaching the joint could, combined with the other two factors,  lead to blow-by of both the primary and secondary O-rings

My calculation of the probabilities of these three factors, (mating of two previously flown out of round segments , tight fit at the critical location and temperature below 55F) indicated that after 24 flights and a late January launch date there was a 18% chance of this type of failure occurring on the next flight.      Had the 51L failure not occurred, the capture feature was in the pipeline at Morton Thiokol for the next order of steel segments and as these new segments came into the pipeline the older segments would be remanufactured to include the capture feature as well.    The capture feature would have removed two of the critical pieces necessary for a Challenger type leak  (1 and 2).

[AS_501]

Given the NASA culture and the push for flight frequency, it seems inevitable that sooner or later we were going to lose another vehicle if it hadn't been Challenger.

In 1988, the classified military shuttle mission STS-27 was nearly lost due to hundreds of damaged and at least one lost tile from a debris strike during launch. Near burn-through was only thwarted through total luck that under that spot contained an antenna's backing steel plate rather than fuselage aluminum, or it would have melted through. Of course this wasn't publicized at the time due to being a classified mission.

The most bizarre (and saddening) aspect of STS-27 was that the clear pictures they got in orbit from the damage, were mis-interpreted on the ground - for a pretty damning reason. The mission was a military one and the necessary encryption of coms and datalinks had the ground seeing only the pictures "low res" and thus missing the damage. Commander Hoot Gibson was rather baffled and furious at such absurdity.

Imagine: you KNOW the damage is there, you have the pictures showing, yet it is not possible to convince the ground of the severity of the damage because the pictures transmission degraded them to low-res.

Then again, what could they have done is probably little if nothing (thinking about the CAIB studies of Columbia rescue resulting in very slim chances of saving the crew via another Shuttle launch).

A very scary and absurd near-miss, really - and coming only the next flight after RTF and two flights obviously after STS-51L. Not only the crew but the entire Shuttle program dodged a big bullet that day, and only thanks to that steel plate / antenna located at the perfect place at the perfect moment.

Sweet Jesus...

Didn't they also discover a small crack in one of the SSME turbopump bearings during the post-flight tear-down?
Many thanks to EGREG51 and Libra for their detailed insights....I never knew the depth of these issues.

[EGREG51]

In the summer of 1986, I first read the Rogers Commission report at the University of Iowa library.   I came away from my reading of Volume I with a single question:
 
    "If it was the cold O-rings that  caused the leak, why did the leak occur in only one single spot the size of my fingernail?"

  It seemed to me, as a layman, that if cold O-rings were really the cause then the joint should have leaked like a shaken two-liter soda bottle with a loose cap, spraying from many locations.  In fact, Roger Boisjoly and Allan McDonald of MTI predicted that would happen and the vehicle would be destroyed on the pad.  Some really unique set of conditions must have occurred at the 304 degree location that hadn't happened anywhere else on that joint, on the other joints of 51L or on any of the other 66,048 inches of O-ring from the 24 prior flights.
    It took me many years and many interviews to solve that mystery.    The key person who was closest to the answer was James Kingsbury of MSFC.   Unfortunately Jim and Dr. Moore acquired the data and testing results in March 1986 and formulated the root cause, two months too late.  Their 3 point theory was confirmed by the recovery of the RH aft field joint in April 1986.

Some of Jim's thoughts on the matter are here:

https://www.nasa.gov/sites/default/files/atoms/files/19930303_james_e._kingsbury_oral_history_interview.pdf

[Jim]

Yes and another link in the chain that led to the loss of Challenger and the crew was that the fact that every time the SRBs impacted the ocean it slightly warped the shape of the segments which lead to the "out-of-round" problem that allowed the joints to flex enough to allow for the hot gas blow-bys.

 the segments just sitting on their side did this.

[Jim]

    It took me many years and many interviews to solve that mystery.   

There was no mystery and you didn't solve anything.

[EGREG51]

    It took me many years and many interviews to solve that mystery.   

There was no mystery and you didn't solve anything.

It was a mystery to me since I  bought into the "official" line of an O-ring which was unable to expand across the gap opening because of a loss of resiliency.    If that caused the leak at 304 why not at 324 or 104 degrees?   It made no sense that it was just cold enough to impact the O-ring sealing ability at one location but it sealed  at the same temperature conditions 8 inches away.  It seemed like a "goldilocks temperature" theory.    There had to be a more complete explanation.

You are correct, I solved nothing, the NASA accident analysis team solved the issue in 1986, but that report and it's conclusions were never made public.    My curiosity lead me to speak to those in the know, like Jim Kingsbury, Jim Beggs, Joe Kilminister, Bob Lund, Joe Sutter and Larry Mulloy and they provided me with the complete root cause.

It was the same mystery Dr. Feynman pondered after the Rogers Commission report was released.   Unfortunately he didn't live long enough to solve it himself.  I attach a transcript of an interview with his thoughts on the mystery.