Hacker News Reader

Buckling at high temperature is better than fracture at low temperatures. I used to do research in this area. The "holy grail" technology would be a non destructive way to infer the "neutral stress temperature" of continuously welded rail that has been already installed in a quick and non destructive way. Saw lots of weird hopeful technology prototypes based of measuring the hysteresis curves of induced magnetism in the rail, idea being that the stress would have some effect on the magnetic domains. Never worked in the field AFAIK. Here in Australia they will get hoses out on hot days desperately trying to cool the rail down to stop it buckling.

My weird idea would be if sound waves would travel differently down the rail dependent on temperature. Although, the rail is likely not uniform enough to detect the differences?

I found a device that is in use - Verse[1]. It claims to be able to measure the neutral? Although my read says the rail must be below neutral, and therefore under tension, to determine what the neutral is?

They unclip the rail and then lift it with a known force? I guess that lets them measure the tension and lateral forces on the rail?

[1] https://www.pandrol.com/us/product/verse-rail-stress-managem...

https://www.youtube.com/watch?v=Qg9vn50ZOdo

https://www.youtube.com/watch?v=WbNzhMsUOwA

Railway buckling video: https://www.youtube.com/watch?v=jJeiYaSFCf0

Verse PDF: https://railway-news.com/wp-content/uploads/2020/02/VERSE-Te...

That's interesting. Unclipping the rail is a bit of an exercise though. Makes sense that it must be in tension, otherwise it would move on unclipping. I guess they just measure it in the morning. All these types of devices are less practical as soon as you have to disrupt services to do the measurement. Your idea of the measuring the dynamic response is something that has been investigated. Modal testing is very common (using accelerometers on the rail and a force measuring hammer). The effect would depend on the boundary conditions of the rail (so basically something that varies from site to site) so you would need to track how it changes over time. The engineering problem is how to collect this data easily.

> Here in Australia they will get hoses out on hot days desperately trying to cool the rail down to stop it buckling.

Have you got anything to show that?

Never seen or heard of that in NSW, and not heard of any track buckling issues during summer, either.

I've only found mentions online about spraying post buckling during repairs, but pre-emptive spraying does (did) occur. You will hear about speed restrictions on hot days to avoid buckling.

https://www.researchgate.net/figure/Buckling-of-a-railway-tr...

Yeah I have never seen anything like that in NSW either - anyway on 38°+ days one would have to spray continually anyway, and there’s no way you could do that along a whole rail line in a country that routinely experiences droughts and water restrictions in the summer.

It wouldn't help for rail already installed, but I wonder if you could insert two corrosion-resistant pins in the rail at some standard distance apart. The pins would be installed when the rail is at its "neutral stress temperature". Then you could use some Go/No Go gauges to quickly check if the rail has expanded or contracted. By measuring the distance you could get a more accurate assessment of the stress state. The accuracy required might be too much, though.

It's an interesting problem. It's simple in theory, but may not have a perfect solution.

Part of the problem is that even the state of stress at installation isn't known very precisely. I basically did some pre work on a similar idea. Working through if you could track the displacement of sleepers could you figure it out. Problem is the extension in nominal range is linear and there isn't a simple way to work out where zero is.

Train carriages used to go ca-chunk-ca-chunk, ca-chunk-ca-chunk, ca-chunk-ca-chunk. Each carriage has two bogies and each bogie has two pairs of wheels on axels; the first ca-chunk pair was the forward bogie, the second the rearward bogie.

It was as recent as the 1980s in Ireland, I remember the sound when travelling by train as a kid.

You're correct that the rail needs to continually have its stress managed. This could be cutting the rail and adding a new piece, or removing a bit. This is a fairly labor intensive and costly process so the railroads were very interested in coming up with a way to determine when a buckle or break was most likely.

Buckles and breaks were most common just after a train had left a section of track. The normal approach was to issue slow orders when the temperature got too high (buckle) or too low (break). So there was a lot of interest in finding more accurate ways to determine when slow orders were necessary and to optimize when to add/remove rail.

AFAIK this problem is still outstanding. The product I worked on years ago never quite succeeded. Railroads are a harsh environment.

maybe you could have a automated drone "chaser" caboose that follows a train a few miles behind looking for buckles and breaks.

For suburban networks signals will go to danger if the rail breaks and interrupts the track circuit

Breaks are "easy" to find, the problem would be buckles where the track is still "connected" but it is one rail is no longer equidistant from the other rail.

Only if track circuits are used in the first place. Modern installation tend to prefer axle counters, and even with track circuits some designs use only one rail for the track circuit and therefore wouldn't be able to detect breaks in the other rail.

It should be noted that whether rails are continuous or not depends a lot on the country and the region. There are still many places around the world where you'll find tracks made of separate pieces. For example, Russia has around 10,000 kilometers of track like that (out of 85,000 km total).

Much of the US rail network is still piece-rail; it's only CWR when replaced, because most rail-miles in the US are freight-only, and almost all are freight-primary.

Russia sees extremes of both heat and cold. That may play a part in deciding to use separate rail piece.

I don't think so. It was all separate historically up until 1960s, but then gradually updated to continuous as budgets allow.

Thing is, even aside from passenger comfort, you also get about 5% less friction from CWR, which translates to more fuel efficiency. And 5% might not sound like much, but when most freight cargo goes by train, and distances are in thousands of kilometers, it adds up pretty fast. Ditto for reduced maintenance on the tracks themselves, which is especially handy for those long tracks going through largely empty areas.

Yah. But you can have these instead [1] https://de.wikipedia.org/wiki/Feste_Fahrbahn#/media/Datei:Oe...

(right track, above middle of picture) Not only on bridges.

AFAIK this is the way to go with any sort of HSR, "Feste Fahrbahn", https://en.wikipedia.org/wiki/Ballastless_track or whatever. I remember them even from the early 80ies, asking people about what that is, and why?

Edit: I see your [1] has a similar picture for an "expansion joint", but it looks functionally different, because not "seamless". With the stuff I know there is only expansion outwards, while there is almost no gap on the rail itself.

> This confused me because trains go "clack clack clack" - I guess that's foley now.

They still seem to go clack clack on some bits.

> When a track is made up from short rails bolted together, we leave small gaps between each one so that expansion doesn’t cause a problem. [0]

They seem to use both long welded pieces and short gapped pieces, i'm guessing they are just selective where they use each. Anecdotally, trains seem to go slower on the clackedy clack bits...

Speculation: I wonder if gapped sections are more resistant to failure IF they are used at slower speed? i.e long welded sections reduce maintenance at high speed; but gapped sections lack lateral temperature stresses, so IF they are used at a slow enough speed they could have very little stress. If so it's possible they use them on track close to stops, busy junctions, or track which is simply in a difficult area to carry out repairs on... essentially anywhere that track repair work will cause more problems than average and it's worth having slower trains on. Correction welcome.

[0] https://www.networkrail.co.uk/running-the-railway/looking-af...

Yeah trains haven’t clack clack clack for a long time now. Possibly as much as 20 years for the lines I’m on.

CWR is common to reduce noise and bumps. The latter in particular is really important at high speeds, because the feel of a bump to a rider is a lot worse.

The neutral temperature is 27C, we may have to follow countries further south in setting it higher.

Ontario suggests their rails are designed for 30C to -30C: https://www.cbc.ca/news/canada/hamilton/headlines/metrolinx-... (2016)

Addressed in the post:

>Why don’t we replace all our rails with ones stressed for higher temperatures?

> <snip> ... there would be the risk of increased tension on the rails in the winter.

Sydney Australia uses 35°C

Iron ore ops in the Pilbara, (northern) Western Australia use 40°C

From an accident report:

> The area surrounding the 67 km point had been subjected to several maintenance activities, including a few days before the derailment. The records showed that a length of rail was installed at about the 67.150 km mark when the rail temperature was about 65°C

> According to BHPB, the stress-free track temperature was 40 °C. This specification was based on midpoint average minimum and maximum temperatures expected for that time of year (summer).

https://www.atsb.gov.au/media/5775153/ro-2015-023_final.pdf