Kidneys dying in space.

[darthgently]

1 hour ago, K^2 said:

That is wildly exaggerated, TBH. Your typical interplanetary transfer is in tens of km/s. A 2RPM rotation involves a rotation around a point 250m from the center of mass (this can be more or less than 500m of cable, depending on your counterweight, naturally) and has velocity w.r.t. combined CoM of only 50m/s. This is well within the mid-course correction burns you'll be doing anyways. Likewise, loss of gravity, while disorienting, is about as dangerous as riding on a drop tower at an amusement park. You do still have the spin of the rocket itself, but it's likely to be a fairly stable rotation and the resulting centrifugal effects will be minor enough for this to be at worst comparable to a slip-and fall on flat ground. You can never exclude a risk of injury in any sort of a fall, but you can also slam your hand closing a hatch, so you know.

What is a real risk is the cable snap-back. Back of an envelope estimate, a lot of materials you'd consider for a cable would stretch by about ~10% before failing. You want a good safety margin, so you're probably going to be looking at a little less than 5%. Call it a 10m stretch from CoM, where the cable's most likely to snap. A fully-fueled Starship is 5kT, which means you're looking at 1/2 * 50MN * 10m = 250MJ of energy stored in the cables. It's not a LOT, but it's enough to get the tip of the cable flying at you at a decent speed. (Yes, I know the energy is split between multiple cables, but so is the mass of the cable you have to get moving...) With a very rough estimate of 20T of cables in there, you'd get something like 50m/s average speed, but the tip is likely to be traveling closer to the speed of a bullet. The odds of that tip hitting the ship are not zero, and it will absolutely slice clean through.

There are safety measures you can take. The tension wave will propagate at the speed of sound through steel, which is on the order of a few km/s. That technically gives you enough delay for emergency severing of the cable. If you put accelerometers everywhere along the length of the cable, you can cut all the cables from the ship, causing both ends to snap together instead, resulting in very low chance of any part of the cable hitting the ship with significant velocity. You obviously want this system to be rock solid, but what emergency system isn't? And we are looking after a catastrophic failure, which is very unlikely to begin with. We build suspension bridges using essentially the same tech.

 

All of that said, when people talk about testing a centrifuge on ISS, people do talk about small radius centrifuge. As mentioned earlier in this thread, we do now have strong evidence that even a sub-5m radius centrifuge is viable for artificial gravity with a trained crew. Especially if you don't mind ramping up the speed over a few days when the trip starts, and winding it down on arrival. And that is an entirely different structure. Something with a revolution period of about 4 seconds and compact enough to fit in the inner hull of the Starship. If we're going to see artificial gravity on a trip to Mars, that is far more viable than any tethered design. And it's also something we can comfortably test on the ISS with very reasonable expense. I mean, it's still a full sized module requiring some amount of orbital assembly with current launchers, but again, if we're looking at testing this for interior of a Starship, the best way to test it would be to assemble one within the Starship, dock it to ISS, and just keep it docked for a few months. What better proof of Starship's capability to take a crew to Mars could you possibly ask for?

I disagree about an unannounced loss of gravity not being a big problem.  I think it would run afoul of about any NASA or industrial standard of basic safety concerns and could easily result in serious injury or death depending on what was happening at the time.  Also, consider that it wouldn't be a vanilla free-fall as each craft would be tumbling at the same rate of previous rotation.  Anything not stowed would be adrift in a slowly tumbling craft.  

I also disagree about the unplanned course change of two halves of your craft not being a big problem, considering the crew would simultaneously be dealing with the complications of an unannounced loss of gravity.   Consider that one craft has become two, so now two craft require corrections and they both have a rotational component also.  I'm getting a relative separation rate of ~102m/s, did I not divide by 2 where I should have?

But, yes, I definitely should have listed cable snap-back as a huge problem.  People get killed and maimed in gruesome ways every year on ships and in industry from snapped cables and chains.

1 hour ago, K^2 said:

There are safety measures you can take. The tension wave will propagate at the speed of sound through steel, which is on the order of a few km/s. That technically gives you enough delay for emergency severing of the cable.

So whatever system severs the cable would have to assess, decide, and sever in milliseconds?  Possible I suppose with very fast processing and a very, very fast severing mechanism

 

[kerbiloid]

Spoiler

 

[RCgothic]

Sometimes it's acceptable to just accept the remote chance of a fault because covering it off completely would be grossly disproportionate. 

The chances of a well-designed cable breaking in normal operation are remote. So we're talking fault cases. What sort of faults could be encountered in interplanetary space? Micrometeoroid impact? And the chances of that severing all the strands of a cable are? If the strands are cross-coupled every so often the broken end wouldn't even rebound especially fast.

Any strike large enough to take out multiple strands would probably be a LOCA in the far higher probability of striking the main spacecraft anyway.

Any other plausible fault cases? I'm struggling to think of any 

[kerbiloid]

4 hours ago, RCgothic said:

Sometimes it's acceptable to just accept the remote chance of a fault because covering it off completely would be grossly disproportionate. 

The chances of a well-designed cable breaking in normal operation are remote. So we're talking fault cases. What sort of faults could be encountered in interplanetary space? Micrometeoroid impact? And the chances of that severing all the strands of a cable are? If the strands are cross-coupled every so often the broken end wouldn't even rebound especially fast.

Any strike large enough to take out multiple strands would probably be a LOCA in the far higher probability of striking the main spacecraft anyway.

Any other plausible fault cases? I'm struggling to think of any 

Expectations

Spoiler

 

Reality

Spoiler

 

***

At least, you can't keep it balanced due to mass changes.

[darthgently]

9 hours ago, RCgothic said:

Sometimes it's acceptable to just accept the remote chance of a fault because covering it off completely would be grossly disproportionate. 

The chances of a well-designed cable breaking in normal operation are remote. So we're talking fault cases. What sort of faults could be encountered in interplanetary space? Micrometeoroid impact? And the chances of that severing all the strands of a cable are? If the strands are cross-coupled every so often the broken end wouldn't even rebound especially fast.

Any strike large enough to take out multiple strands would probably be a LOCA in the far higher probability of striking the main spacecraft anyway.

Any other plausible fault cases? I'm struggling to think of any 

The safety severing circuit or sensors gets hit by a cosmic ray and triggers what is basically a RUD?  Do we put FTS on manned rockets?  Why not?  Because in deep, deep space, a safety severing system isn't far off from FTS in level of effect on crew and mission.  There are better ways to get artificial gravity with an internal centrifuge as K^2 described.

With torch drives we could do brachistochrone trajectories at somewhere between Earth and Mars gravity the entire way with a flip over midjourney