Hacker News Reader

Comment by Gustomaximus | Hacker News Reader

I dont know if this an elephant in the room. I'm guessing engineers have thought about this and just not discussed it publicly. They have some pretty smart guys.

Also to nitpick, from "the ends of a viable hyperloop track will have to move hundreds of meters" - this doesn't have to be true. There can be absorption points along the track. Overlaps. Or maybe they tunnel the entire way where its cooler. Or wrap the tunnel in something cooling and reflective. We dont have train tracks moving hundreds of meters do we? I dont profess to know the solution, but do know there are solutions somewhere.

View on HN

Wouldn't they just put accordion joints on the junction at every pylon? It's not like the tube is going to be built as one gigantic piece. It has to be assembled from parts.

The vacuum is the problem. With HSR you also have small gaps in the system to allow for expansion but nothing there requires a perfect surface.

AFAIK, high speed rails do not have gaps. Expansion is controlled by brute force : lots of clamping and massive supports.

They can have expansion joints around bridges, but generally speaking you're correct, they're thermite-welded and adjusted to correctly handle thermal stresses without kinking (heat expansion) or fracturing (cold contraction).

> There can be absorption points along the track

Sure, but these have to hold a vacuum. No one has figured out a way to do that.

I am assured by those in a position to know that thermal expansion joints which hold a near-vacuum are an entirely solved and essentially trivial problem.

I do have some concerns about whether they're a trivial problem within this context. An engineering solution which works in small numbers for PhDs in a lab isn't necessarily scalable to mass low-cost manufacturing, deployed in the field and irregularly serviced by workers of uncertain provenance. But evidently there is a way to do this, at least.

And, more importantly, can be certified for real world use. Considering the speeds used, scrutiny will be extremely high. Even if you use it for freight at first, the track will be close to settlements and any accident could end up in a disaster. The behaviour of trains on tracks is very well understood (learning from accidents), the behaviour of a vacuum train tube is not.

This is a very fair point. There might be components that are routinely used in non-life-safety-critical applications, but this is no guarantee that they will pass certification scrutiny when they are in the critical path for thousands of lives every day.

I worked on an autonomous vehicle project here in the UK about 10 years ago (the Heathrow Pod). The certification process was absolutely intense. And this was for a system that was capped at 25mph; had we gone any faster than the regulatory burden would have become very significantly more onerous (below 25mph it wasn't necessary to test vehicles and infrastructure to destruction; above 25mph, it is).

In this domain, there's a huge difference between a cool engineering testbed, a private pilot project, and running actual service for the public. If Hyperloop is going to be doomed by anything, failure to appreciate this fact is probably at the top of the list.

Except for the people who invented the accordion and bellows.

I work in a lab with vacuum systems that get heated and cooled and we just have segments of flexible bellows like this:

http://www.pchemlabs.com/subcatagoryb.asp?pid=flexible-bello...

Not sure if the same thing will work at larger scales, but I'm not sure its a fundamental problem.

Wowza, 5" bellow is $450, how cheap could they make 6' one, even in bulk?

Not exactly like that, but yeah, the same sort of idea I'm sure would work.

Gas pipelines hold hundreds of bar of pressure; I don't see how one atmosphere is a problem.

Gas has no problem flowing around corners. Gas pipelines are usually not straight lines. Hyperloop will have to be a straight line, unless you want to kill passengers.

Gas pipelines use Expansion Loops [1] to pick up expansion slack. The problem is that, in order to push people through the tube at 700mph at acceptable G's, you can't put bends like that in the HyperLoop. If you try to send people down a pipeline at 700mph, you would end up with paste at the other side. Gas pipelines are surprisingly un-straight, bendy, and bumpy

[1] https://static.interestingengineering.com/images/import/2017...

Isn't this what a syringe does?

Reply on news.ycombinator.com