HomeInfrastructureGetting hot and bothered - Why the railway doesn’t like high temperatures

Getting hot and bothered – Why the railway doesn’t like high temperatures

Listen to this article

Remember the old “clackety-clack, clackety clack” as trains moved around the network?  If you are one of our younger readers, you might not.

The noise came from the train’s wheels passing over the bolted joints between rails.  Every 60 feet. And with two axles/wheelsets on each bogie, and two bogies on each carriage, that’s the “clackety-clack, clackety clack”.

It’s not heard often now.  Sometimes around points, or on old branch lines, but all of the main lines are beautifully silent.

That’s because the railway no longer bolts rails together, it welds them. This forms one long rail that may be several miles long.  No joints, no gaps between rails, no noise.

Lengthy problem

However (isn’t there always a however?), if you think back to your physics O-level or GCSE, you’ll remember that metal grows when it gets hot and shrinks when it gets cold. Now, if you have a piece of steel several miles long, and you constrain it by clipping it down to a load of heavy concrete sleepers, when it gets hot there is only one way it can grow – lengthways.

Using the coefficient of thermal expansion for steel (0.000012 m/mºC if you insist), the result is that, for every degree that a rail heats up, a 5km length will grow by 6cm.  That’s about two and a half inches.

If it heats up by 10 degrees, that’s 60cm or two feet.

Now don’t forget, the rail is constrained by all of those clips so it can only grow lengthways.  Except it can’t – there are other equally-constrained rails at each end – all the way to Scotland (or wherever).

For a while, the rail can stand this.  The track engineers have been clever and have already prestressed the rail to the length it would naturally be at 27ºC. This means that if the rail is actually under 27ºC it is shorter than it should be so it is stretched, if it is over 27ºC it is compressed.  So problems only start over the magic figure of 27ºC.

Steel is a good material, and it can take quite a bit of tension and compression.  However, like all things, it has a limit.

Yesterday, the temperature of rails that had been baking in the sun all day reached over 50ºC.  That’s 23º over the ‘standard’ temperature or 1.38 metres (four and a half feet) of expansion on our 5km rail. And on the rail next to it. And on the rail next to that.

Eventually, something has to give.  The rail buckles, forcing its way sideways or upwards, tearing itself out of its clips, and completely destroying that short length of track where the buckle occurs.

And due to the temperature, when the white Network Rail van arrives, filled with men and women sweltering in orange suits, there is little they can do. If they cut out the damaged piece of track and repair it, the track is now prestressed to 50º not 27º. So when it cools down they will have problems the other way with cracks and breaks.  All they can really do is divert trains around the site unless they make a temporary repair knowing they will have to redo it all again properly later.

Typical failures

Network Rail does what it can.  Some track repairs such as tamping are cancelled in hot weather in case they make the situation worse. In areas of known problems, track and S&C (switches and crossings – or points as they are colloquially known) are painted white – to reflect the heat and keep temperatures down.  But if the sun shines all day, temperatures will inexorably rise.

In the main, prestressing works, otherwise there would be buckled rails everywhere rather than just one or two.  But it was hot yesterday and that brought good examples of three types of failure.

A rail buckled near Stowmarket. Trains were diverted onto the ‘wrong’ track around the failure at low speed, causing disruption to timetables.

At Billingshurst, a conductor rail buckled.  These tend to be aluminium so they actually grow faster than steel in hot weather. The failure was cut out and replaced.

Near Henley-on-Thames, a set of points which connects the main line with a branch is known to stick in hot weather. Rather than risk switching to the branch and then not being able to switch back, closing the main line, the decision was taken to lock the points out of use.  Trains could therefore not access the branch directly from the main line.  However, a local shuttle operated from the other side of the same platform so, by simply changing trains, passengers could still reach their destination.

Blowing hot and cold

So that’s the story of hot rails, and why they buckle. It’s a problem every summer, but only in the hottest weather.

High temperatures bring other problems.  Overhead wires are made of copper and that expands tremendously. Tensioning systems take up most of that growth but, in areas with older installations, it can lead to saggy wire syndrome.

Earthworks dry out and shrink, causing uneven track surfaces, and bright sun can make some signals, both on the railway and for traffic on level crossings, hard to see.

Summer is a real problem.  Roll on winter, when the rails shrink as they get cold and eventually break, earthworks get soggy causing uneven track surfaces, and S&C gets flooded and won’t work.

Ho hum!

13 COMMENTS

  1. What do they do in countries with high temperature change in different seasons? i.e. USA with full 4 season where temperature between summer and
    winter can change upto more than 50 degrees?

    • Here in Manitoba there have been ‘go slow’ orders resulting in the Churchill – Winnipeg trains being several hours late, this despite the relatively small amount of continuously-welded rail on the line.

    • Most of the US’s rail network is rather old, and is still made of short, bolted rails. I’ve actually never heard of buckling rails being an issue before this article, and I’ve never seen them anywhere, so I can only assume that even new tracks here are bolted rather than welded.

      • In Florida, think of Heat Index..of 120 degrees in the summer… for hours… One year one of Amtrak’s “Silver Service” trains derailed South of Palatka, FL killing several people. The cause was Heat Buckling of the Rail.

  2. I’ve heard this story over and over when I was working on the British railways. But I’ve now moved to Spain 7 years ago and nobody talks about this… nor they experience delays during the summer. And, please be aware, Spain gets quite cold during the winter as it has a lot of mountains so I guess the difference in temperature is much greater than in the UK

  3. In my 40 years experience of permanent way engineering covering the entire globe I have never before encountered the device of painting rails in switch and crossing layouts white to help control rail expansion. (The only instance of expansion problems in S&C that I know of is detection of the switch position in switch diamonds.) Does anyone know of what event triggered the introduction of this bizarre (and grossly unsightly) practice?

  4. A mention with respect to CRT (critical rail temperature) management would be worth expanding on in accordance with NR standards and position of stillages? Good piece though.

  5. Outside the box thinking here…. What if you covered the railway sleepers in a UV protective coating in effort reducing potential expansion by reflecting heat? Maintenance on railways potentially less instead of cutting and trimming steel to allow for expansion just cover in a UV protective coating during warmer months. Would this be something worth considering?

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.