Continuous Welded Rail

Deepgreen said:

Of course, more expansion joints might provide relief if the SFT remains 27C.

Click to expand...

As posted previously, there are no expansion joints in CWR. Having any expansion joints makes it not continuously welded, and there would be no point in stressing it.

Bald Rick said:

As posted previously, there are no expansion joints in CWR. Having any expansion joints makes it not continuously welded, and there would be no point in stressing it.

Click to expand...

Good point! Too much focus on buckled rails recently - I was simply thinking of ways to combat rail heat stress generally.

I'm now very confused with the situation regarding expansion joints - I'm sure I've seen them fitted on the UK rail network, but if they aren't appropriate for CWR where are they appropriate for?

aleggatta said:

I'm now very confused with the situation regarding expansion joints - I'm sure I've seen them fitted on the UK rail network, but if they aren't appropriate for CWR where are they appropriate for?

Click to expand...

Two areas, usually:

Jointed track - usually fishplated, with rails cut straight vertically and perpendicular to the rail. Expansion in the rails (usually of 60ft length) is accommodated in the gap between the rails, which closes up in warm weather. When it gets really hot the joint closes totally, and then you run the risk of buckles. A key maintenance task in spring is to oil fishplates to give them the necessary movement to allow the expansion.

Areas where CWR join up to jointed track or to certain types of switches and crossings (that are known as ‘unstrenghtened’). This is typically an adjustment ‘switch’ (colloquially known as a ‘breather’), where the joint is diagonal over about 300mm if viewed from above. Where you see these, you will usually find short pieces of rail bolted to the centre of the sleepers across the joint; this prevents the sleepers moving relative to each other and thus allows the ‘free’ rail to expand up to the anchored rail.

This is the simple version, it gets a bit complex of course.

Bald Rick said:

Areas where CWR join up to jointed track or to certain types of switches and crossings (that are known as ‘unstrenghtened’). This is typically an adjustment ‘switch’ (colloquially known as a ‘breather’), where the joint is diagonal over about 300mm if viewed from above. Where you see these, you will usually find short pieces of rail bolted to the centre of the sleepers across the joint; this prevents the sleepers moving relative to each other and thus allows the ‘free’ rail to expand up to the anchored rail.

Click to expand...

Thanks for the reply! This is the example I have seen and it makes sense, I'm puzzled as to why they couldn't fit this type of expansion joint at regular intervals on long stretches of CWR, the main benefits of CWR being noise and comfort these seem to be much quieter than your standard fishplate joint, and would appear to help with the issue of expansion on extreme heat.

as the wagons will be perhaps 64ft or what ever that is in metres times 10 so 600ft so 40 foot to spare so a 2-3% expansion will be easily contained

aleggatta said:

Thanks for the reply! This is the example I have seen and it makes sense, I'm puzzled as to why they couldn't fit this type of expansion joint at regular intervals on long stretches of CWR, the main benefits of CWR being noise and comfort these seem to be much quieter than your standard fishplate joint, and would appear to help with the issue of expansion on extreme heat.

Click to expand...

The main reason for CWR is the to reduce track damage and wear. Jointed track is vulnerable to rail end breakage. The fisplate bolt holes are a weak point and when jointed track was the norm, breakage at that weak point was a major failure mode. In addition, the rails would bend as heavy axle loads pass over them causing the ballast to loosen beneath the adjacent sleepers. Thus, jointed track needed so much more maintenance that today's high speeds might not be economically viable.

aleggatta said:

Thanks for the reply! This is the example I have seen and it makes sense, I'm puzzled as to why they couldn't fit this type of expansion joint at regular intervals on long stretches of CWR, the main benefits of CWR being noise and comfort these seem to be much quieter than your standard fishplate joint, and would appear to help with the issue of expansion on extreme heat.

Click to expand...

Any sort of expansion joint automatically makes it not CWR, and you lose the stress in it. Besides, a 60C increase in rail temperature needs, roughly, a 10mm gap per 20 metres of rail, hence why jointed track has gaps more than this every 20 yards!

AM9 said:

The main reason for CWR is the to reduce track damage and wear. Jointed track is vulnerable to rail end breakage. The fisplate bolt holes are a weak point and when jointed track was the norm, breakage at that weak point was a major failure mode. In addition, the rails would bend as heavy axle loads pass over them causing the ballast to loosen beneath the adjacent sleepers. Thus, jointed track needed so much more maintenance that today's high speeds might not be economically viable.

Click to expand...

Jointed track needs inspecting more frequently* and the fishplates all need checking, tightening, oiling etc. Another reason, is that CWR is much less likely to buckle than jointed track in high heat. In jointed, any slight rail creep (rail moving longitudinally) will lead to tight joints - and once the joints are tight in hot weather the risk of a buckle increases very quickly.


* in the accident report that @ChiefPlanner posted, the track was patrolled 2-3 times a week, presumably because it was a CWR island in a jointed area. With all CWR that would be weekly at best.

This thread has got me curious, if I remember correctly, one of they key buckling metrics was the unsupported length, but I can't figure out what the unsupported buckling length of a rail is? Is it between each sleeper or something else? The pictures of buckled rail look like it's buckled out of the load over a longer distance than that.

For ballasted track it would be very difficult (probably impossible) to determine the unsupported length as there are many factors to consider which would vary from location to location.

All the following provide lateral resistance

• Sleeper weight & friction on underside of sleeper,
• Heaped ballast shoulders
• Well-consolidated ballasted in the sleeper beds
• Lateral resistance plates
• The fastenings between rail and sleeper to prevent "twisting"
• Clip toe loading to minimise longitudinal rail stresses

But as all of the above aren't rigidly fixed in place and their effective lateral resistance can vary a lot along the length of the line, track is more susceptible to buckle at the weak points. (E.G, timber sleepers, wet beds, collapsed ballast shoulders, slender rail sections)

Rail on concrete slab track, on the other hand, would have an unsupported length equal to that of sleeper spacing (A little less if you consider the
width of the base plate or fixing type)

A question to the more knowledgeable folk.

A little scenario. A long length of CWR in a high temperature environment. The track is in “perfect” condition, so good sleepers, ballast etc.

So the rail is effectively supported and “clamped” into position. As the temperature rises, the rail expands. Where does that expansion go? Is it a case of no matter how good the track, at some point it’s going to buckle if it gets hot enough?

carriageline said:

A question to the more knowledgeable folk.

A little scenario. A long length of CWR in a high temperature environment. The track is in “perfect” condition, so good sleepers, ballast etc.

So the rail is effectively supported and “clamped” into position. As the temperature rises, the rail expands. Where does that expansion go? Is it a case of no matter how good the track, at some point it’s going to buckle if it gets hot enough?

Click to expand...

It doesn’t expand, at least not if it has been correctly stressed, and is correctly fastened. I’ll try and explain.

A standard length of CWR is around 200 metres long. If it is installed with the rail temperature at, say, 10C, it will be pulled (stretched if you like) by around 30mm, using hydraulic jacks. In simple terms, one end is anchored to the sleepers by being clipped up, the other end is pulled by the jacks, and all the rail inbetween is placed on rollers which help it ‘stretch’ evenly. The technical officer on site has to do the calculation regarding temperature and rail length to determine the amount of pull required.

When the rail is pulled to the correct length (to the millimetre), it is then welded and clipped up to fasten it in place. It gets a bit tricky on long pulls with sudden changes of temperature - I’ve been on sites where the rail has been in position, but a sudden break in cloud has cause the rail to warm up quickly and someone needs to lop another 10mm off quickly!

With the rail stressed in this way, when it warms up, the stress is simply relieved, ie the tension in the rail caused by the pull reduces. When the rail temperature goes over the stress free temperature (usually 27c), the tension in the rail starts to build the other way, ie in compression. The rail can’t expand as it is fixed in place by the strength and integrity of the whole track system - the clips, the sleepers and particularly the ballast (which needs to be well consolidated, and heaped on the sleeper ends - the ballast ‘shoulders’).

As long as the force of the compression is less than the restraining capability of the track, then all is well. A buckle occurs when the compressive force exceeeds that of the restraining capability of the track. This usually occurs at a weak spot, typically an area where the ballast is insufficient on the shoulders or not sufficiently consolidated in the whole bed, and more often than not on a curve. The buckle is usually to the outside of the curve as the compressive force pushes it out.

Having said all that, whilst the rail can’t expand in length, it does of course does expand width and depth, but in fractions of a millimetre!

Bald Rick said:

Having said all that, whilst the rail can’t expand in length, it does of course does expand width and depth, but in fractions of a millimetre!

Click to expand...

However, rail will move longitudinally in stress transition zones - The more obvious one where the forces in the rail tend towards zero at the adjustment/breather switch.

Stress transition zones can occur in the middle of CWR sections. Tunnels are a good example as there is a varying difference in rail temperature compared with that outside in the sunshine.

civ-eng-jim said:

However, rail will move longitudinally in stress transition zones - The more obvious one where the forces in the rail tend towards zero at the adjustment/breather switch.

Stress transition zones can occur in the middle of CWR sections. Tunnels are a good example as there is a varying difference in rail temperature compared with that outside in the sunshine.

Click to expand...

Yes, of course - I was trying to do the GCSE version; A level is for another time (and I’d have to get out my books )

Yeah, no worries. They're complex beasts are transition zones....bit of black magic thrown in.

civ-eng-jim said:

However, rail will move longitudinally in stress transition zones - The more obvious one where the forces in the rail tend towards zero at the adjustment/breather switch.

Click to expand...

Is there a tendency for the rail to "work" back and forth near breather switches as temperature fluctuates, so that the expansion range of the switch is gradually used up? Or are the fasteners guaranteed to hold the rail in a fixed position once a certain distance from the switch?

Another idiot question - which in my role of 'man in the pub' I couldn't answer the other evening:

When heat does cause track to buckle
- how much (length-wise) of the track buckles?
- what happens next? Does the track go back into gauge when everything cools down, or does the buckled section need to be replaced?
- how long does sorting things out take?

I appreciate that these are all 'how long is a piece of string' type questions, but (simple) answers would be appreciated...

Fawkes Cat said:

Another idiot question - which in my role of 'man in the pub' I couldn't answer the other evening:

When heat does cause track to buckle
- how much (length-wise) of the track buckles?
- what happens next? Does the track go back into gauge when everything cools down, or does the buckled section need to be replaced?
- how long does sorting things out take?

I appreciate that these are all 'how long is a piece of string' type questions, but (simple) answers would be appreciated...

Click to expand...

To answer in ‘how long is a piece of string’ style:

1) it varies, it can be a small kink, it can be up to 100metres or more
2) no it doesn’t go back into gauge, or even remotely close to its correct position. It needs to be physically moved back.
3) that depends on the amount that has shifted. A small kink would take a couple of hours, a big buckle may take up to 24 hours or so. This also depends on the rail temperature, as it is difficult to do in the heat. Fortunately, most buckles happen between 1200-1700, and by the time resources are in place it is mid to late evening and things have cooled down. If there is an accompanying derailment, it can be longer of course.

Bald Rick said:

To answer in ‘how long is a piece of string’ style:

1) it varies, it can be a small kink, it can be up to 100metres or more
2) no it doesn’t go back into gauge, or even remotely close to its correct position. It needs to be physically moved back.
3) that depends on the amount that has shifted. A small kink would take a couple of hours, a big buckle may take up to 24 hours or so. This also depends on the rail temperature, as it is difficult to do in the heat. Fortunately, most buckles happen between 1200-1700, and by the time resources are in place it is mid to late evening and things have cooled down. If there is an accompanying derailment, it can be longer of course.

Click to expand...

Thank you.

Are there any restrictions on the use of CWR on Waybeam bridges?
Thinking of short span bridges less than 10m.

edwin_m said:

Is there a tendency for the rail to "work" back and forth near breather switches as temperature fluctuates, so that the expansion range of the switch is gradually used up? Or are the fasteners guaranteed to hold the rail in a fixed position once a certain distance from the switch?

Click to expand...

The adjustment switch blades are set (at Stress Free Temperature) to allow 100mm of rail movement but couldn't tell you how much of that got used up in the latest heatwave.

I might be misinterpreting your post, but are you implying that rail will expand and contract (cyclically from night to day, sunshine to cloud et cetera) but over time the rail won't contract back as much as it's expanded and slowly eat into that 100mm spacing? Some sort of heat-induced expansion creep? Sounds plausible but I suspect if there is such a phenomenon it'll be very small. Activities, namely tamping, will have a greater effect on the 100mm and there are maintenance standards requiring routine inspection of the switches.

This chap, who is an academic and also works for a well-known UK design consultancy, has a very good permanent way blog - He covers stress transition zones in this:
https://pwayblog.com/2017/08/30/rail-stress-transition-zone/

Ploughman said:

Are there any restrictions on the use of CWR on Waybeam bridges?
Thinking of short span bridges less than 10m.

Click to expand...

Not that I'm aware of.

civ-eng-jim said:

The adjustment switch blades are set (at Stress Free Temperature) to allow 100mm of rail movement but couldn't tell you how much of that got used up in the latest heatwave.

I might be misinterpreting your post, but are you implying that rail will expand and contract (cyclically from night to day, sunshine to cloud et cetera) but over time the rail won't contract back as much as it's expanded and slowly eat into that 100mm spacing? Some sort of heat-induced expansion creep? Sounds plausible but I suspect if there is such a phenomenon it'll be very small. Activities, namely tamping, will have a greater effect on the 100mm and there are maintenance standards requiring routine inspection of the switches.

This chap, who is an academic and also works for a well-known UK design consultancy, has a very good permanent way blog - He covers stress transition zones in this:
https://pwayblog.com/2017/08/30/rail-stress-transition-zone/

Click to expand...

I was thinking about whether over a long period the stress transition zone would work its way further and further into the CWR section, and if so whether this would result in more and more movement of the end section and possibly risk to the integrity of the CWR because it's no longer stressed as it should be. That link may answer but looks like something I will need longer than a lunch break to study!