CONTINUOUS RAIL EXPANSION 2

[harrisj]

Here's a railway problem I don't understand - haven't found a good explanation on any website, but it can't be rocket science.

It goes like this: Continuously welded rail can see temperature variations from -30C to + 50C or more in some parts of the world. Coefficient of expansion of steel is around 12 x 10-6 / deg C.

So expansion of the rail between intense cold winters and hot summers is 1 in 1000. Now if you are laying rail in 1/2 km runs, an expansion joint is required that will absorb 500 mm of length variation. I have seen slip joints in the rail using something like a scarf joint in wood, but they're not that long.

So how is this variation accommodated? If it is allowed to slide through the rail fixings, how is the friction controlled? If it is constrained, why doesn't the track push out of line on a curve as it heats / cools?

The forces are considerable; if you constrain the track, you will end up with compressive stresses (buckled track) - or tensile stresses (not good for crack propagation).

You will gather that I'm not a railway engineer - but my office overlooks a main rail line and it's puzzling me.

So chip in, railway engineers - tell me how you do it!

Cheers - John

 

[diamondjim]

Each section of the rail is shortened by
the ratio given. I am certain that you
hear the clack, clack, clack, that the train
makes as it passed over these rail section
breaks. So you probably hear the greatest
noise at the cold extreme temperature.

 

[harrisj]

Yes - but for a 500m section, the gap is 500mm! You'd hear more than a 'clack'.

I understand that continuously welded rail is often longer than 500m.

If the rail is not constrained, how do you construct a 500mm long sliding joint that will take the weight of a train?

In the old days, sections were (I believe) 60 feet long. That's only 3/4 inch between sections - lots of clacks but wheels don't fall down the gap.

Still not convinced.

 

[desertfox]

Hi harrisj

I was under the impression that the continous welded rail was pre tensioned during laying and welding, which as I understand it means that a given length of rail is always under tension and therefore when hot weather expands the rails this would just reduce the tension in them amd because
there is no compressive stress buckling cannot occur.

 

[harrisj]

Desertfox:

You may be right. I found a website which implied that the rails were constrained, and the rail just sits there in cold weather with tensile stress as you suggest. But the levels must be quite high. And how is it anchored at the end of the section?

I can't help feeling that, when the track is laid on a curve, as it gets cold and the stress increases, there is a tendency to pull the track into a tighter radius. The track is only sitting in ballast, not rigidly constrained.

I also found that CWR can be several miles long when welded so there is no realistic possibility of expansion joints.

Also, tension is not desirable in a structure prone to cyclic loads. In UK there have been several fatalities relating to catastrophic failure of rails by crack propagation (eg Hatfield).

How do rail engineers get round these problems? Clarification from anyone?

John

 

[Speedy]

"To ensure that temperature variations in continuously welded rails (CWR) do not cause track buckles in hot weather or equally dangerous rail fractures in cold weather, it is crucial that the Permanent Way Engineer knows the temperature at which the rail is neither in compression nor in tension. This is known as the stress-free temperature (SFT)."

From;

http://www.pandrol.com/cstudies/c15_bdy.htm

Speedy

"Tell a man there are 300 billion stars in the universe and he'll believe you. Tell him a bench has wet paint on it and he'll have to touch to be sure."

 

[sancat]

I was just wondering, a 60kg/m rail (120lbs/yd) will have a section of 7650 mm2, and for a 1/1000 stretch will need a force of 160,000 kgs. If we assume that the zero stretch temp is just in the middle of the temp range, that will mean 80 tons in compression or tension depending on cold or hot weather.
That kind of force will mean 10 kg/mm2, that thinking on a SAE 1060 alloy will mean a 15 to 20% of yield.
Seems that numbers fit, but, 160 MT are a LOT of force. How will this force be anchored on ballast?
On the extremes of the rail, the force and stress will be zero, and it will increase, by means of a distributed axial force, up to 10 kg/mm2 that will remain constant in a middle zone. The zone where the stress is not yet 10 kg/mm2 will ellongate/stretch a certain amount that will have to be taken by an expansion joint. Then, a longer rail will have less expansion joints and behave in a similar way than a shorter one.
On the side push that this stress will cause on the ballast, if we assume that rail minimum radius is 1000 mts, and that rail supports are spaced 0.5mts, the angle between adyacent supports will be 0.057 degrees, and thus the force needed to constraint the rail will be 80,000kgs x sin(0.057) = 80 kgs, not so big.
Well, this are just ideas, that could turn to be actually BS since I really dont know much about trains...
sancat

 

[tstanley]

The rail is constrained from side to side by sitting in a "rail seat" in the plate that supports the rail on the tie. Each plate has a spring clip on it (ie pandrol clip or safelok clip) that holds it down onto the plate and supplies enough friction to take care of the expansion between one tie and the next, about 20 inches. The clamping force is actually greater than 80kg from each clip. In this way the forces are taken up by the thousands of ties. Spikes do not hold the rail down to the tie for CWR.

That is my understanding of why Vancouver and Halifax are not pushed out into the ocean by Toronto.

 

[harrisj]

Thanks tstanley and sancat

Seems it just sits there getting stressed - a bit like me really ......

The way it's done (relying on ballast not to move around too much)is not the most elegant technical fix but seems to work. As you say, the numbers stack up but I'm still concerned that rails have tensile forces around 20% of yield in a safety critical structure which is subject to cyclic loads and in which cracks propagate due to wheel wear.

But thanks for a useful explanation.

"No day is without gain"

John

 

[smltwnboy]

While I am certainly not experienced in this field, just a couple of observations...
When looking at the expansion properties of the rail, you
cannot do that exclusively unless the rail is completely isolated, which it is not! You must then consider the SYSTEM, which then requires that you consider the fasteners, ties, and ballast and all of the heat sinking properties related there to... Sounds almost like a nearly infinite heatsink which could explain a few things...

smltwnboy

 

[harrisj]

You're absolutely right in looking at the system as a whole.

However, rails are suspended in mid-air along most of their length, with a chair every 600 - 800 mm or so - I don't think the chair and clip provide much of a heat sink. At the low end of the range, they will nearly reach the air temp which will get down to minus 30C. At the upper end, air temp will be nearly 30C plus the solar input - rails aren't always bright and shiny and metal objects sitting in the summer sun get mighty hot.

So I don't think a delta of 80C is grossly overstated - but even if it's only 50 or 60 it's still quite a lot of stress.

I think it's likely you'll see these extremes in a few places in the world. Anyone done any rail temp measurements?

John

 

[schlebb]

Hi guys:

I went to school in Germany, and was fascinated by the fact that railway rails did not have any expansion joints!

The explanation from my professor - they are laid in a very gentle "S" curve. Thus, the S - shape is shortened and elongated with temperature variations.

--- no clackety-clack on trains---


regards,


schlebb

 

[harrisj]

Schlebb: If sancat and tstanley are right - and I think they are - then your professor was wrong.

If rails shuffle sideways twice a day they'll soon lose their alignment, cant and other desirable caharacteristics. Not to mention gauge problems under bridges, tunnels, platform edges etc.

No, the rails must stay in the same place, and I think the stress just sits there in the rail and is reacted by the ties in the ballast - like the man says.

John

 

[rhodie]

Thought this might be interesting:

http://southern.railfan.net/ties/1967/67-5/field.html