I realize I'm probably an idiot, but I do try to learn from my idiocy. Thanks for helping me learn. But I'm having trouble understanding how your reply takes everything into account, such as the fact that the hyperloop moves pods, not liquid. Pipelines for liquid are certainly a solved problem, but a pipeline of pods is new.
Btw, in railroads (that other long distance transportation method) the expansion is dealt with by putting a small gap between two adjacent rails. That's obviously not possible with the hyper loop as it is currently envisioned but I think that the answer is in there: don't allow the differences to accumulate over long stretches.
If that's the case, then why does the plan assume each station on the endpoints is going to absorb most of the thermal expansion? Or am I misreading? They specifically say "the stations are the ideal points to absorb thermal expansion," and the stations are on the end, so it sounds like they're aiming for a long sleeveless steel tube.
Also, I was basing my criticism on Hyperloop Alpha. Obviously, their design can change in the future. But, as presented, what gives you the impression that sleeves could be incorporated into their design without causing problems for the pods?
I can visualize sleeves which expand into each other, and if they overlap into the direction of travel, so that the edges are pointing away from the pods rather than toward the pods, maybe that wouldn't cause any problems.
Still, the interior margin from pod to wall is very small: http://i.imgur.com/YsLDXBe.png
Also, consider this paragraph:
The Hyperloop travel journey will feel very smooth since the capsule will be
guided directly on the inner surface of the tube via the use of air bearings and
suspension; this also prevents the need for costly tracks. The capsule will bank
off the walls and include a control system for smooth returns to nominal
capsule location from banking as well. Some specific sections of the tube will
incorporate the stationary motor element (stator) which will locally guide and
accelerate (or decelerate) the capsule. More details are available for the
propulsion system in section 4.3. Between linear motor stations, the capsule
will glide with little drag via air bearings.
Sleeves would cause an edge to sort of "protrude outwards" into the path of the pod, wouldn't they? So it'd basically be a small speedbump. Again, the edge is pointed in the direction of travel, so that would basically be like the second half of a speedbump, basically a small "drop". Except since the pod is moving at 760mph, even a small drop can result in quite a lot of force, and can cause wear on the pod itself, couldn't it? And it might feel a little jarring to the passenger.
If the interior needs to remain smooth, what sort of design could fulfill that requirement? (Maybe that's a big "if", but if the pod is going to bank off the walls, then it sounds like that's the case.)
EDIT: Also, as someone else pointed out, pipelines that need to be straight might be a new concept. Many pipelines seem to use "U" bends, which can't be incorporated into hyperloop for obvious reasons: http://demonstrations.wolfram.com/PipelineExpansionLoop/
