Continuously Welded Rail

 
  TraineeWGH Beginner

I am still confused and skeptical about the given physics of CWR.
In my knowledge of physics there is no way of preventing any material from expanding some way due to temperature variation. An American rail magazine reports that 40 degree (F) temperature change will result in 17 inches of expansion per mile - nearly one and a half foot. Inspite of concrete sleepers and the new clever fixation of steel rail, the only way I can see that his can be avoided is if the steel is especially designed to expand in width and depth and not in lenght. Can someone please help explain the exact physics? I have read so many explanations that do not make sense to me.

Sponsored advertisement

  PeeJay2 Beginner

I am still confused and skeptical about the given physics of CWR.
In my knowledge of physics there is no way of preventing any material from expanding some way due to temperature variation. An American rail magazine reports that 40 degree (F) temperature change will result in 17 inches of expansion per mile - nearly one and a half foot. Inspite of concrete sleepers and the new clever fixation of steel rail, the only way I can see that his can be avoided is if the steel is especially designed to expand in width and depth and not in lenght. Can someone please help explain the exact physics? I have read so many explanations that do not make sense to me.
TraineeWGH

My guess is this: Heat causes an expansion force in the rail, and the sleepers/fasteners/ballast are required to contain this force. An uncontained force = buckle. So basically the compressive force in the rail will just keep rising until something gives. (Try compressing an eraser on a desk and you will see what I mean - believe it or not the rail is slightly elastic). Same thing with the cold - it causes a tensile force as the rail shrinks, and this time it is usually the tensile strength of the rail that is inadequate = broken rail
  a6et Minister for Railways

I'm not doubting him, but in the Australian context, and more importantly the NSW context (as that's where I live), am I right in saying that there are two generations of CWR?

1. Old generation Stephensonian technology. Frequent expansion gaps. Clickety clack.
2. 1st Gen CWR. Rail of a few hundred metres length. Fewer expansion gaps. Increased risk of WOLO. Used with either wooden or concrete sleepers. Old folks whingeing about how "buckled rails never happened in my day". Mid twentieth century.


There are two things worth noting here. One is that jointed - non-welded - rail does not have "expansion gaps", except in very rare instances. The other is that it frequently buckled. Anyone who claims otherwise is talking nonsense.

I cannot recall the date when the flash-butt rail welding plant opened at Chullora, but I can say that the proprietary "Thermite" welding process was used in NSW before WW2, according to an associate who was a Way & Works engineer.
dullsteamer
While I forget the actual lengths that the Chullora facility welded the short rails into lengths that were then placed on unit type flatwagons that were fitted with bolsters and end slides for dropping the rails off at the applicable locations.

In the 60's and later times when full rerailing was carried out those welded lengths were then thermit welded at the time of replacing the old track.  What was noticeable on the track at the time being 106lb was that at set spacings there was holes in the track sides for the purpose to allow for expansion through to tack up some of the slack.  

In those days in all regions when temperatures reached above 35 degrees wollo reduced speeds were introduced until the temperatures went back below 35. It was rare to have buckled rail on the above 100lb rails but would happen on the 94lb rail and below, 94ib was often welded as well but all done with thermit process on the permanent way, once the 106 became the common weight.

Conditions of sleepers even wooden ones could affect things as those locations with long sections of new wooden sleepers with good ballast were much less prone to buckles then one old sleepers.  When the first longish section that was put in place on the up line near between Exeter and Werai around the reverse curves it was very noticeable when hitting the new cement section at speed there was a smooth transition riding on any of the diesels, however when coming off the cement to the wooden sleepers it was a totally different situation with the engines really kicking to the sides, IIRC the speed came off a 70k onto an 80k bend at the bottom of the grade so was very noticeable, made sure you checked the train as you came around the curves.

The heavy rail replacement program during the 80's on the main North from Waratah through to MBK with ballast cleaning and replacement of all sleepers with new cement and 64kg track made a huge difference and never saw or experienced any issues with buckling no matter the heat.
  allan Chief Commissioner

Prestressing is a part of this story: I don't know how it works, but it dates from the 1960s.
  Lockspike Assistant Commissioner

I am still confused and skeptical about the given physics of CWR.
In my knowledge of physics there is no way of preventing any material from expanding some way due to temperature variation. An American rail magazine reports that 40 degree (F) temperature change will result in 17 inches of expansion per mile - nearly one and a half foot. Inspite of concrete sleepers and the new clever fixation of steel rail, the only way I can see that his can be avoided is if the steel is especially designed to expand in width and depth and not in lenght. Can someone please help explain the exact physics? I have read so many explanations that do not make sense to me.
TraineeWGH
Hi Trainee,
Be a sceptical as you like, CWR works. If it doesn't work it hasn't been done correctly. The basic precept for the success of CWR is the fact that after welding the rails are 'adjusted'. This means that the rails are stretched and then a final weld performed, so the rail is under tension most of the time. Generally in Australia rails are adjusted so that at a rail temperature of 35 degrees the rail is neither in compression nor tension. It takes a warmish sort of day before the rails will go into compression. Once the rails go into compression the structure of the track holds the rails in their correct position, so it is important that the ballast, sleepers and fastenings are in good condition. Ballast should be at the correct profile including a full ballast shoulder i.e. 300mm wide for Class 1 track. It is also the strength of the track structure that holds the track in alignment when the rails are in high tension such as on a cold frosty morning; the rails would much prefer to take a shortcut across the inside of a curve!
I've been told, though never tried it myself, that rails after being tensioned are measurably smaller in section.

A6et,
The trial section of concrete sleepers on the Up between Exeter and Werai gave valuable experience in how to install and maintain CWR and concrete sleepers. At the Werai end the concretes finished on a curve, bad mistake. Anywhere where a change of track stiffness occurs tends to create a 'hole' in the road immediately on the more flexible side. A classic location for this phenomenon is on bridge ends, also on the end of turnouts, and on the end of a section of concrete sleepers. Compounding this problem at Werai were the springs of water under the track, creating a permanently wet formation. Further exacerbating the situation was the fact that as trains braked for the curve they incrementally pushed the rails along creating an excess of steel at the bottom of the hill (this phenomenon tends to be worse in jointed track). All these factors made for a location difficult to maintain in good condition, hence the bumpy ride around that bottom curve. Note that the Dn track at that location shared the problem of the springs, but was a lot easier to maintain a good top and line.
  dthead Site Admin

Location: Melbourne, Australia
A old thread might be good background reading, but there again it is 14 years old:


https://www.railpage.com.au/f-p2098328.htm#2098328
  a6et Minister for Railways

I am still confused and skeptical about the given physics of CWR.
In my knowledge of physics there is no way of preventing any material from expanding some way due to temperature variation. An American rail magazine reports that 40 degree (F) temperature change will result in 17 inches of expansion per mile - nearly one and a half foot. Inspite of concrete sleepers and the new clever fixation of steel rail, the only way I can see that his can be avoided is if the steel is especially designed to expand in width and depth and not in lenght. Can someone please help explain the exact physics? I have read so many explanations that do not make sense to me.
Hi Trainee,
Be a sceptical as you like, CWR works. If it doesn't work it hasn't been done correctly. The basic precept for the success of CWR is the fact that after welding the rails are 'adjusted'. This means that the rails are stretched and then a final weld performed, so the rail is under tension most of the time. Generally in Australia rails are adjusted so that at a rail temperature of 35 degrees the rail is neither in compression nor tension. It takes a warmish sort of day before the rails will go into compression. Once the rails go into compression the structure of the track holds the rails in their correct position, so it is important that the ballast, sleepers and fastenings are in good condition. Ballast should be at the correct profile including a full ballast shoulder i.e. 300mm wide for Class 1 track. It is also the strength of the track structure that holds the track in alignment when the rails are in high tension such as on a cold frosty morning; the rails would much prefer to take a shortcut across the inside of a curve!
I've been told, though never tried it myself, that rails after being tensioned are measurably smaller in section.

A6et,
The trial section of concrete sleepers on the Up between Exeter and Werai gave valuable experience in how to install and maintain CWR and concrete sleepers. At the Werai end the concretes finished on a curve, bad mistake. Anywhere where a change of track stiffness occurs tends to create a 'hole' in the road immediately on the more flexible side. A classic location for this phenomenon is on bridge ends, also on the end of turnouts, and on the end of a section of concrete sleepers. Compounding this problem at Werai were the springs of water under the track, creating a permanently wet formation. Further exacerbating the situation was the fact that as trains braked for the curve they incrementally pushed the rails along creating an excess of steel at the bottom of the hill (this phenomenon tends to be worse in jointed track). All these factors made for a location difficult to maintain in good condition, hence the bumpy ride around that bottom curve. Note that the Dn track at that location shared the problem of the springs, but was a lot easier to maintain a good top and line.
Lockspike
Better drivers had the brakes releasing prior to the restart of the timber sleepers in order to gain momentum around the curve and for the grade prior to MV. if the distant signal was showing clear the aspect of train management like that was a big part of many of the drivers I was with in the 60's, and when I was driving in the 70's.  We rarely used Dyno even if they were working which was more often than not not working.

I agree that the test section should have been extended at least as far as being on the main straight on the MV side of Werai station. The forces of the engines and vehicles on a curve coming from the more secure bedding aspect of the concrete onto the much less bedding of the timber sleepers showed up by the ballast movement on the sleepers edges, the hairy legs had to go out each day to check and repack the ballast where needed.
  Lad_Porter Chief Commissioner

Location: Yarra Glen
A old thread might be good background reading, but there again it is 14 years old:


https://www.railpage.com.au/f-p2098328.htm#2098328
dthead
That link apparently does not work.  Try this one:

https://www.railpage.com.au/f-p1881755.htm

There are a couple of "Rails that grow"videos mentioned in there.  I think they were training videos for the NSWGR, very interesting at the time, but now unavailable.
  gordon_s1942 Chief Commissioner

Location: Central Tablelands of NSW
Many years ago I was talking to the Track Supervisor who was bothered by the current temperature readings he had got of 53C from the Rails outside the Signalbox.
At the time it was a generally nice sunny day but no overly hot.
I remember him saying that it wasnt those long areas of track in full sunlight that concerned him but where it was shadowed in a deep Cutting or were there was a tunnel as you got wide variations of heat readings between the shadowed and exposed areas.
I came into the district just prior to the rerailing programs that began in the early 60's where the 10 meter (30 foot) long (90lb) 40kg? was to be replaced with heavier sectioned welded rail.
Timber sleepers were the norm and many did not have plates between the rail and sleeper.
Over the next few years, this plate changed shape too considerably.
Today the area is all steel sleepers with maybe  a few hundred metres of Concrete sleepers laid in a specific area for some reason.
During the next 20 plus years I can only remember one or 2 kicks occurring due to the heat, one surprisingly was opposite the Goods Shed between the two signalbox's at Wallerawang on what didnt seem an excessively warm day.
  neillfarmer Train Controller

Some physics. Firstly temperature. When steel is heated it expands in all directions. This expansion depends on the increase in temperature and the length of the steel. When it cools it contracts. The expansion and contraction can be calculated exactly. Secondly stresses, When steel is loaded it changes its shape as if it were a piece of rubber. Once again these changes in shape can be calculated precisely. CWR is like a long column. The expansion from heat can be counteracted by compression from loading so that the length does not change. It is however under a heavy compressive load. If not supported laterally it will buckle similar to a drinking straw when the ends are pushed in.
So in hot weather the rail expands in all directions. We don't worry about it getting bigger across, or in height but we have to do something about the lengthwise increase. Every sleeper adds load to the long string of the rail restraining its expansion by applying a compressive load.
Buckles happen when the clamping forces between rail and sleeper are insufficient, the rail slips in its fixings and pushes up against track that is more solidly fixed. If the sleepers are not held by the ballast then they become the weak link and they move sideways. A train can disturb the track enough for the rail to sleeper grip to be weakened.
One advantage of closer sleeper spacings to get heavier axle loads is that the additional sleepers also help in restraining the rail's expansion.
In SE Qld the track is adjusted to be stress free at 38C. The theory is that a broken rail (due to too high a contraction in cold weather) can be quickly detected by signalling track circuits, but a buckle cannot. Also trains can ride over a clean break but buckles derail them.
Hope this helps.
  YM-Mundrabilla Minister for Railways

Location: Mundrabilla but I'd rather be in Narvik
Some physics. Firstly temperature. When steel is heated it expands in all directions. This expansion depends on the increase in temperature and the length of the steel. When it cools it contracts. The expansion and contraction can be calculated exactly. Secondly stresses, When steel is loaded it changes its shape as if it were a piece of rubber. Once again these changes in shape can be calculated precisely. CWR is like a long column. The expansion from heat can be counteracted by compression from loading so that the length does not change. It is however under a heavy compressive load. If not supported laterally it will buckle similar to a drinking straw when the ends are pushed in.
So in hot weather the rail expands in all directions. We don't worry about it getting bigger across, or in height but we have to do something about the lengthwise increase. Every sleeper adds load to the long string of the rail restraining its expansion by applying a compressive load.
Buckles happen when the clamping forces between rail and sleeper are insufficient, the rail slips in its fixings and pushes up against track that is more solidly fixed. If the sleepers are not held by the ballast then they become the weak link and they move sideways. A train can disturb the track enough for the rail to sleeper grip to be weakened.
One advantage of closer sleeper spacings to get heavier axle loads is that the additional sleepers also help in restraining the rail's expansion.
In SE Qld the track is adjusted to be stress free at 38C. The theory is that a broken rail (due to too high a contraction in cold weather) can be quickly detected by signalling track circuits, but a buckle cannot. Also trains can ride over a clean break but buckles derail them.
Hope this helps.
neillfarmer
Sums up why the lightweight/cheapskate concrete sleepers on the completely relaid Ararat - Maryborough section were such a false economy.
  gordon_s1942 Chief Commissioner

Location: Central Tablelands of NSW
Oddly rails dont always break down from the head to the foot but I have seen a large lump come out of the running side of the head and  apparently that is NOT good news.
On that current TV show about the Canadian Railways, a rail snapped about 3~inches from the end but I couldnt understand why they were treating it as real problem when the Fish Plate was keeping it together.
A Fish Plate in this case is one of 2 plates that is used to keep the rail joints together. Some are slotted to allow the joint to open and close.
Certainly the break would have broken the track circuit but would not pose any immediate danger of derailment to a train.
  YM-Mundrabilla Minister for Railways

Location: Mundrabilla but I'd rather be in Narvik
Most of the so called railway documentaries on TV are nothing other than a series of contrived disasters waiting to happen.

The fishplate may well have maintained the track circuit continuity.

Sponsored advertisement

Subscribers: a6et

Display from: