Train Rail 5

[Ginger]

I am looking at the use of 30lb/yd light rail to construct a temporary train track (610mm gauge)in a tunnel we are constructing.

The site team wish to run locomotives and muck wagons with 3.5 tonne axle loads on the rails with fabricated steel sleepers located at 1.0m centre to centre.

My question is: How do I analyse the rail? I have looked at it as a continuous beam, supported at the sleepers (assumed as knife edge supports), with a series of point loads applied vertically representing the wheel loads. These wheels are moved along the rail in a train to produce worst case bending and shear.

I have also applied notional horizontal loads to the rail to represent nosing or lurching of the train on the rail. These combined vertical and horizontal effects seriously overload the rail. The results makes me think that my method of analysis is incorrect.

How should I analyse the rail??

Regards

Andy Machon

 

[whymrg]

Check "AREA Manual for Railway Engineering". They have chapter on unloading pits. In this chapter they specify rails and distances that can be supported by this rail.

I did not work with this for 4 years, but that what I remember.

Good luck.

 

[Ron]

Andy,
Loading as a beam supported only by the sleepers might be giving you some of the problem. Bending is mitigated by the elastic response of the bedding material.

I did something similar for "a major theme park" in Florida and similarly, but with much higher loadings, for "an un-named space agency" to transport a "launch vehicle". I applied the wheel load directly to the bedding material. Once the vertical strain in the bedding material was known, I could adjust the loading appropriately in the rail to do more accurate bending calcs. This comes from the fact that the max. vertical deflection in the bedding will limit the max. deflection in the beam, thus controlling the bending stresses (but not shear stresses!) Be careful with this approach to check web buckling.

Ron

 

[sven]

Andy,

I believe that your method is basically correct, but for a "construction" design to work you will have to modify the usual conservative structural assumptions. Also, I don't agree that the subgrade will provide any significant support unless it is extremely stiff and then you wouldn't need sleepers.
Try changing the point loads to short distributed loads and the spans to the clear distance between the supports. Also, for a temporary structure, you can usually live with up to 30% "overstress" in the rail. Even so, I think 1m spacing may be a little too far for 30lb rail.
A good reference for rail design is "Modern Railway Design" published in 2001 in Holland.

 

[austim]

Hi, ginger.

Your current assumptions appear to assume no deflection of the sleepers under load. Even if the sleepers will bear directly onto a rock floor that may be a little too severe. If you will have ballast under them, then it is
far from realistic.

On every ballasted rail track that I have ever seen, the sleepers deflect quite a lot under load. I always understood that for bending stresses, the conventional anlysis (before the days of simple computer analyses for continuous beams on discrete elastic supports) was to treat the track as a beam on a continuous elastic foundation. (Very much as Ron has indicated above).

If your high stresses are in negative bending over the sleepers, then you should get real reduction in your estimated bending stresses from some consideration of spring supports instead of knife edge.

Mind you, the possibilities remain that the rail you are trying to use is just too light for the job, or that your span between sleepers is too large for the rail.

 

[Ginger]

Hi guys.

Thank you all for your valuable input, I appreciate the advice. Just to clarify things as I may have misled you; the rails are seated on steel sleepers which are the shape of an orange segment, when viewed in elevation. These sleepers sit in the invert of circular segmentally lined tunnel and carry the rails at 200mm above invert level. There is no ballast. The rail track is used to service the tunnel boring machine with new segments and to cart away the tunnelling spoil.

Based on the advice given, I will try Sven's approach, although I suspect Austim may have hit the nail on the head in his last sentence!!

I'll be back !!

Best regards

Andy

 

[sound789]

My favourite GINGER,

We design such type of "rail-sleeper on ballast" with the help of the Reynold's Handbook on Reinforced concrete to get quick but slightly conservative solutions.

 

[idly123]

Well go back to the Book on "Railway Engineering - By saxena" it gives a very good insight iin to ur problem
raj Raj