How about hedging your statement to "not CURRENTLY feasible using TODAY'S materials and structural concepts (trusses, etc)". Space flight and industrial production of fertilizer were both considered impossibilities when first proposed.
If you're planning on building a closed loop out of steel, and there's nowhere for the steel to go when it expands in summer, the plan may need some rethinking.
The best I can think of is some kind of inner lining, so that it has a segmented hull, like a caterpillar: https://www.google.com/search?q=caterpillar&tbm=isch
But it would need to be very strong lining in order to prevent any leaks between the segments. And any kind of a leak would need to be detectable and would shut down the hyperloop until it's resolved.
Between the wildly optimistic cost projections and the fundamental engineering difficulties, I'm much more inclined to call it a pipe dream than a perhaps-future-possibility.
> But it would need to be very strong lining in order to prevent any leaks between the segments.
The entire point of not requiring vacuum is to avoid having nearly as strict tolerances for leaks.
Regardless, the core structural engineering is the focus of the criticism, not whether it might leak. Also, since Space X's whole point is to compete on cost savings, it was a little strange to hear cost projections that were so optimistic.
Section 4.5.4 otherwise covers this, and does point out that for large leaks, braking would be required, but smaller leaks are intended to be dealt with as part of regular maintenance, with no indication that it'd need to affect operations.
> Regardless, the core structural engineering is the focus of the criticism, not whether it might leak.
That may be so, but most of the criticism I've seen of it appears to just ignore large sections of the whitepaper.
EDIT: For example, with respect to handling expansion/contraction, consider that the world has gas and oil pipelines that has to deal with far worse conditions without leaks.
Because the hyperloop cars are discrete (as opposed to the continuous flow in a pipeline), I'd expect them to cause a lot of stress on the pipeline walls if they had to run around the bends in an expansion loop.