Thanks to Iron Flight for reminding me of this obscure situation. And for Holly who wants to understand.
This mostly true story is about the Space Shuttle payload bay doors and what to do if they won’t shut properly. And how we forgot about that emergency in our rush to achieve other mission objectives.
Some months ago, I visited Atlantis at the Kennedy Space Center visitor center. Superbly displayed, it is the only shuttle vehicle with the payload bay doors open. In the picture you can see some of the rube-goldberg mechanism that operates the doors critical latches to ensure the doors don’t spring open during aerodynamic flight. A set of push rods and hinges at the sill open and close the doors. A long set of push rod/bell crank mechanisms located in the front and rear edges of the doors operate the latch mechanisms. On the bulkhead just aft of the crew cabin you can see the black cylindrical knobs which the latch mechanisms grab onto to secure the door in place. More latches and their mechanisms ran down the centerline of the doors. If this sounds complicated, it is.
Early in the shuttle days there was a huge concern about what to do if the doors did not close and latch properly at the end of the mission. The doors had to be open during most of the orbit stay time for cooling and to allow satellite deployments or other objectives. But during re-entry and the atmospheric part of the flight, the doors had to be firmly shut and latched down. If they were to spring open and rip off, the vehicle would become uncontrollable and catastrophe would ensue. Flight rules prohibited deorbit with any more than one set of latches not closed; and in that case the aerodynamic maneuvers were to be severely restricted.
Of course, the power to the motors that drove all this mechanism were redundant but that did nothing for a physical jam. During STS-3 a thermal engineering test caused the doors to become banana shaped and prevented – for a short time – closing the doors. This was done on purpose to see what the limits of the spacecraft were. We found out. And never did it again.
Long before STS-1 flew, a set of procedures for a spacewalking crewmember to deal with potential problems was developed and practiced. I got my opportunity to take the class; use the tube cutters to cut a pushrod that had jammed, put clamps around unclosed latches to hold doors tight, and more. There was a whole set of tools flown on every shuttle flight to deal with this contingency and every crew got at least one practice session on how to deal with it.
But the trick is that a space suited crewmember must get to the doors from the inside. Normally this is not a problem; with an empty or near empty payload bay the EV crewmember just translates to the worksite; latches everything down, returns to the airlock door and ingresses. Oh, and quickly gets out of the EVA suit and into the Launch/Reentry Suit and straps down because deorbit must occur shortly after the doors are closed and the radiators can no longer cool the ship.
If a payload blocked the path from the EVA worksite to the airlock, early in the program they were always jettisonable. Not a problem.
Despite the complicated design, the payload bay door and latch mechanisms worked perfectly on every mission. Over time, the concerns about having to deal with a failure faded away, even though the procedure and tools were on every flight.
During the evolution of the shuttle design, the European Space Agency built a laboratory to fit in the shuttle payload bay. You can see the SpaceLab module on display right behind Discovery at the Smithsonian Udvar-Hazy center in Virginia. Before the ISS, this was a facility to do micro-gravity research for up to three weeks in space.
But there is a trick that the designers missed.
If the mechanism on one of the aft payload bay door latches failed, and an EVA crewmember had to go back there to secure the doors down, there was no way for that crew member to squeeze past the SpaceLab module to get to the airlock with the payload bay doors closed. Jettisoning the SpaceLab was not a task that an EVA crewmember could do. The way home was going to be blocked.
Oops.
It may have been the first SpaceLab mission when we discovered this, but my recollection is that it was uncovered later, during a simulated mission. Those darn Sim Sups! They always made us work problems that were unrealistic! Except that most of the time they weren’t.
What to do? No more SpaceLab flights? Unacceptable.
Review the risk? Great history of reliability, low likelihood of needing to do the EVA. Case closed? Not exactly.
What if?
Here is the crazy resolution. If the aft latches had to be secured on a SpaceLab flight, then the crewmember would just stay back there.
There was a good place to strap down at the bottom of the payload bay aft of the SpaceLab module. It would be a short wait from finishing up with the doors until the deorbit burn, probably no more than an hour and a half. From deorbit burn to the ground was about an hour and in normal flight the g loading was light, nothing more than 1.5g. The EMU had plenty of consumables to work for the necessary time, but there are a couple of sticky things to resolve:
- The EMU uses a water sublimator to keep the crewmember cool. As the payload bay repressurizes during entry, the sublimator will quit. If the crewmember had selected full cool and chilled down as much as he/she could stand, it would probably be OK.
- Getting out of the spacesuit: after landing, getting out of the helmet and gloves is not a problem. Getting out of the hard upper torso by oneself is a chore but probably doable in 1 g if one is not in a hurry. The lower pressure garment (pants) would not be a problem.
- Getting the crewmember out of the payload bay, well that is a problem. Remember the doors are latched shut and clamps have been applied to keep them shut. Surely the ground crew could figure something out . . . given several hours . . . .
So that is the story. Accept the risk because we think it is low; have a screwy contingency procedure ready if we’re wrong.
But that is not the way you really want to fly in space.
Wayne Hale's Blog 
Wasn’t this particular problem (the bulky shape of SpaceLab) the reason why the later SpaceHab modules were flat at the top, like a three-quarter SpaceLab of sorts? To provide an astronaut the opportunity to actually get back to the airlock from the aft payload bay in such an event? Or am I misremembering here?
I believe you are correct
Oh good. Having to explain to Congress why someone cut a multi-million dollar door into the side of a shuttle with a sawzall would have been a seriously miserable day for whoever got hauled up to DC.
For 3) there was an easy way to access the midbody/PLB with the PLBDs closed. It was through an access panel called “door 44” located forward of the Fuel Cell Servicing System (FCSS) umbilical panel on the left side of the orbiter. Through the door 44 cavity, techs could easily make their way past Spacelab and to the EV crew member(s). The only thing that would need to be removed is the MLI blanket that separate Bay 10 of the PLB from the midbody.
See, I knew the crack ground crew could figure out rapid access!
I happened to have the TV on in 1996 when the program I was watching was interrupted by news that the Shuttle was having a problem opening its cargo bay doors. This was STS-76 and the problem occurred just after a wave off of that day’s landing attempt at KSC. Even though it was the network science reporter who covered Shuttle missions he didn’t seem too concerned, saying he was confident they would figure it out, and then it went back to their regular programming. Leaving me sitting there in a panic as I knew they would have to make an emergency landing if they couldn’t get the doors open.
They did of course figure it out. Astronauts went into the Spacehab for a closer look at the latches to confirm that they were unlocked, even though faulty sensors were indicating otherwise. They then manually overrode the indicators and got the doors open.
From then on for the rest of the Shuttle program I was always nervous whenever the cargo bay doors needed to be opened or closed.
I remember it well. We had just waved off the last Edwards opportunity of the day for weather and when the auto sequence quit and it looked like the doors might not open, we were on the verge of notifying Honolulu airport that the shuttle would be making an emergency landing there in a couple of hours! But, as you correctly recall, the trouble was quickly diagnosed to a faulty indicator, the doors were opened in manual mode and calm was restored.
I’m guessing that emergency landing in Honolulu would have been at Hickam AFB at 21ºN with their 2 12,000 foot runways? Since this was a 51.6º Shuttle-Mir mission to deliver Dr Lucid and supplies to Mir, any idea what kind of cross-range was required for that emergency landing.
Well within the Shuttle cross range limit on the next orbit
If an emergency landing was required at Honolulu, with the Shuttle Carrier Aircraft having a 1000nm range when loaded with an orbiter and with aerial refueling not an option, do you recall any plans to get Atlantis home? Thank you Wayne, Happy New Year to you and yours.
Wayne, a small correction – I believe the long-duration bottom-sun thermal test and resulting temporary door hang-up was on STS-4. Thanks for this article. I worked on the payload bay door EVA contingency tools a long, long time ago. My first interaction with a Flight Director.s was for a DTO on STS-5, to follow-up on the STS-4 situation and attempt to measure payload bay door to aft bulkhead clearance, using the aft payload bay cameras. I was a youngster in JSC Engineering and didn’t know exactly what a Flight Director was, and couldn’t understand why a fellow named Tommy Holloway was asking me so many questions. Quite a learning experience! 🙂
Memory being what it is, mine says it was STS-3. But I won’t argue about it.
Wayne, I think we’re both right. Today I happened to find my copy of Bob Legler’s Shuttle Missions Summary book. Thermal conditioning to close or latch the payloads bay doors is in the events/problems listing for both STS-3 and STS-4. Thanks again for this article.
Jack Lousma remembers the mission ” to carry out extensive thermal testing of Columbia by exposing its tail, nose and top to the Sun for varying periods of time. The crew found that prolonged exposure to the Sun caused the cargo bay doors to warp slightly, preventing them from closing fully. Rolling the orbiter to balance temperatures around the orbiter resolved the issue.” occured during STS-3. This was from his portion of the NASA Johnson Space Center Oral History Project (Interview). Interviewed by Jennifer Ross-Nazzal.
Loved the explanation regarding the doors, especially the comment about simulations.
Identified risks can usually be mitigated. It’s the unidentified risks (we don’t know what we don’t know) that keeps managers up late at night…..
Those damn sim sips!
Sim Sups….
That would be the coffee they drink staying up late at night, dreaming up their evil sim scenarios. Hmm: Sim sup sip scenarios. I’ll let myself out now….. 😉
I had forgotten all about this, Wayne. Thank you SOOO much for bringing the chills back thinking about it again. Really appreciate it!
Heck, if that qualifies me to fly in a shuttle mission, I will gladly volunteer.
A little too late for that now
Was adding a hatch to the aft end of SpaceLab considered? IIRC SpaceLab flew as either one or two modules so the single module presumably already had an opening on its aft end so it wouldn’t be wildly out of line with the overall architecture, I’d think.
Of course, any use would require depressurising SpaceLab which might damage experiments but that would surely be acceptable in an emergency.
No
I checked the Apace Shuttle Missions Summary document that Bob Legler and Floyd Bennett produced and it says that STS-4 was the flight that required thermal conditioning to close the PLBDs. I think Greg is right.
Also, I think it was STS-49 during which one of the aft port PLBD latches did not latch correctly. During post-flight inspection I believe the latch hook was found to have been fractured during closing. The mechanical design was such that when the mechanism gets close to closed, the torque limiters don’t work anymore and the motor keeps on turning and driving the mechanism until it reaches end of travel and turns off the motor. I think the latch hook had not completely made in over the knob and cracked. But that was not observable by the crew.
Just some more interesting tidbits about the PBD’s “The doors are constructed of graphite epoxy composite material, which reduces the weight by 23 percent over that of aluminum honeycomb sandwich. This is a reduction of approximately 900 pounds, which brings the weight of the doors down to approximately 3,264 pounds. The payload bay doors are the largest aerospace structure to be constructed from composite material.
The composite doors will withstand 163-decibel acoustic noise and a temperature range of minus 170 to plus 135 F.”-NASA
HI Wayne !
The fly in the ointment would be hypergolic leaks from the OAMS Pods or ammonia boiler fumes..
Was there any way to cut open a door to allow and escape? I would have hoped there would have been a way to provide an emergency mask to allow a way to breathe (Scott Air Pack arrangement) after landing…
NH3 boilers could be quickly shut down and there was rarely any hyperbolic vapors present. Postulating an additional problem on top of what had already happened is not fair
Hi Wayne amazing story, the PS comment’s was interesting. is it possible for a civilian to qualify for a shuttle mission?
Many non military personnel have been selected to the astronaut corps and flown
One thing I have always wondered along these lines; what was the procedure if the aft door latches failed on an EDO flight? I know the EDO pallet had a stair shape to the upward facing surface and the winch was mounted on the forward face of the pallet, but was there really room enough for a suited astronaught to reach in there and cut things or put clamps on the latches? Remember the EMU isn’t really designed to allow you to work over your head. Did they ever try that out in the WETF during development? If so, how did it work out?