I think the extreme cold temperatures may be the key to why the bolts seize up. This is based on some very distant past experience rather than any engineering insight.
Years ago I was a powerplant mechanic in Michigan, part of a traveling repair crew that would go from one company power plant to another rebuilding boilers and steam turbines. The turbine covers were huge, the largest I recall being at the Palisades nuclear powerplant, 80 tons. To keep steam from leaking, the covers needed to be refitted very tightly. After honing the seating surfaces for a smooth fit (no gaskets), they were attached by lots of large, hollow center bolts. The hollow centers were important because that’s where we’d insert big high temperature heating rods.
Every high school science student knows that metal expands when heated, but the physics of cylinder expansion is that this expansion occurs lengthwise, which has the effect of shrinking the cylinder’s diameter slightly. Since bolts are cylinders, they lengthen and get slightly thinner as they heat up, and this can be used to advantage when tightening and loosening them.
So we’d hone the turbine cover’s seating surface for a good fit, lower the cover in place, install the turbine cover bolts, insert the heating rods, heat the bolts to near glowing, and then (Tom & Ray will love this part) tighten the bolts with slug wrenches and sledge hammers.
Slug wrenches are big box end wrenches with giant squared off metal “handles” way too big to wrap your hand around but perfect for receiving the blows of a sledge hammer. That’s right, you tighten the bolts by smacking the tar out of the wrench. When you get the bolt as tight as you can, you re-insert the heating rod, heat it up some more, and the bolt loosens slightly due to it’s lengthwise expansion and diameter-wise contraction, allowing for more tightening by more beating on the slug wrench.
And here’s the part possibly applicable to the Hubble sticky bolt issue… When those turbine cover bolts cool to their normal operating temperatures, they are tight! Really tight! We never had steam leakage problems because as the bolts cooled they would draw the two turbine cover seating surfaces together very, very tightly.
So… Could it be that bolts installed on the Hubble are warmer when installed than they are when removed? I would think in the very cold temps of space, those bolts would shrink lengthwise and fatten up diameter-wise after installation–a lot–thereby increasing the torque necessary to remove them.
Possible solution? I guess I didn’t mention how we’d remove the turbine covers, but the process was the same in reverse. We’d insert the heating rods, heat the bolts to glowing, and slug them loose–righty tighty, lefty loosey. So maybe if NASA can figure a way to heat the bolts before trying to loosen them, the added elbow grease and temptation to beat on the bolts wouldn’t be necessary. And maybe if before installing any replacement bolts they could be brought down to what their normal operating temperature will be, any change in length/diameter would have already taken place so it wouldn’t later have the effect of tightening the bolts.
Dave Caylor
Kauai, Hawaii