Rail101 - The impact of cold snaps on rail failures

In the northern hemisphere the temperature is falling and in this #Rail101 article I explore why rails break more often at this time of year.

There is a seasonal pattern to rail failures, with increased levels of broken rails in the colder months of the year. Look closer still, you typically see a spike in failures immediately after the first cold spell in the season.

Why? Lets explore!

3 key reasons rail fails more often in cold weather

There are a few key factors:

• Rail Stress
• Pre-existing defects
• Material properties

First, I feel the need to point out to non rail experts that rail failures are thankfully rare. In the UK where I'm based there are around 10,000 miles of track (20,000 miles of rail), yet under 100 rail failures recorded per year. So, there is certainly no need for passenger panic!.

Rail Stress

Much new rail is continuously welded into long lengths many hundreds of metres long to provide smooth running and low maintenance track. When temperatures change objects expand or contract very slightly (known as thermal expansion). For objects in our everyday life this change in size is imperceptible. However, when you have rail 1km long, the change in length becomes significant. A 20 deg C temperature change, alters rail length by around 250mm. Rail temperatures in direct sunlight can often be double the ambient air temperature, (think how hot a car roof gets on a sunny day). So, you can have significant changes in rail temperature and therefore changes in rail length to manage. In order to deal with the thermal expansion (and contraction as it cools), rails are usually stretched into position when fitting, leaving the rail under tension (stretched like a guitar string) for most of the time. Compression in rail can result in track buckles/heat kinks as shown in the video below which pose a real risk to safety. (More on this point can be found in my recent article on the challenge hot weather poses for the railway here)

The temperature where the rail stress changes from tension to compression is known as the "stress neutral temperature". Other names for the same thing may be the Stress-Free temperature, Rail Neutral temperature, and I am sure there are others around the world too (Let me know in the comments)

Each network uses a temperature depending on the usual climate in that area. Climate change is already forcing networks to re-evaluate their stress neutral temperatures as temperatures alter over time.

What has this got to do with broken rails?

It is easier to break something under tension than something that isn't. (Try breaking paper by squashing it together rather than tearing it apart). The more tension applied the smaller the force or crack size needed to break something. Colder weather means the rails want to shrink but the track stays put. This means in cold weather the rail is under greater tensile force and so are more likely to break. So, we see more rail failures in winter than in summer.

Still sceptical? Look at the rail failure below and the one at the top of this article taken from opposite sides of the world (Australia and UK). Spot the common feature - the gap that opens up between the rails when the break has occurred. This happens as the tension in the rail is released and it goes back to its natural length. If you were to leave the rail until a hot day, then the gap would close up as the rail temperature increases.

Image Source: Sydney Trains twitter feed

Rail stressing is always balancing act - Too much stress and you increase failures in winter - too little and you are more likely to get track buckles/heat kinks in summer.

Pre-existing defects

Rails get defects in them - it is a fact of life. Over time and the passage of repeated traffic with enough load these defects may slowly grow as fatigue cracks. For example, a fatigue crack may start from the surface and slowly grow over thousands if not millions of loading cycles as each wheel passes over it. As the crack slowly inches through the rail there is less and less material left to support the load so the stress on the remaining material increases. As this happens crack growth may accelerate. Eventually there comes a point where there is not enough material left to support the load and "BANG" the rail breaks. You can see the crack growth accelerating in the image below leaving behind tell-tale "beach marks"

So, what have fatigue cracks got to do with seasonal rail failures?

Remember in the previous section that in winter the rail is under greater tension? That increased tension makes cracks more likely to grow and grow faster giving them more chance to grow big enough to fail.

Imagine a crack has the whole spring, summer and autumn to grow very slowly (as a fatigue crack). In the first cold snap the rail tension increases, and that increase in tensile stress can be enough to accelerate the crack growth ultimately causing the rail to break. This is the main reason you tend to get a spike in rail failures around the first cold snap of the winter. The cracks that have been growing all year (or indeed those that survived previous winters) are now placed under bigger tensile forces increasing both crack growth rate and the likelihood of failure.

Material properties

This is a small effect, but as a materials engineer, I'll include it for completeness. As steel (and indeed most materials) get colder the energy required to grow a crack through it decreases. The colder it is, the more likely an impact on the rail, e.g., from a wheel flat is to cause a rail to break. This is usually through growing a pre-existing defect unless the wheel flat is massive. Of course as winter follows autumn/fall, the period where wheel flats are more likely due to slippery rails, then it can be difficult to untangle these effects. The difference in impact energy between normal and cold temperatures is quite small for rail steels. Even at warm temperatures the impact energy is quite low; my old university professors' words still haunt me "Twice bugger all, is still bugger all!".

The end stuff

I hope this explains why winter and indeed the first cold spell of winter is usually the worst time of year for broken rails, and the knock on effects of train delays.

If you think I've missed other reasons, let me know.

Tag and share to those in your network who you think might benefit or be able to add more insight and follow the #rail101 tag and me for more content like this

Last but certainly not least if you have a rail infrastructure monitoring challenge then take a look at what Sensonic can offer through AI enabled infrastructure monitoring or contact me for more info.