# When a 20 m rail is 20 m long?

Perhaps I’m splitting hairs here, but it is a fair question to ask:

When a 20 m rail is 20 m long?

Please, have your say and feel free to comment below, after voting!

And this is not a trap question like “Which weighs more: 1 kg of steel rail or 1 kg of feathers?”.

Later edit: By the 20 m length it is implied a standard rail length, not a rail cut to suit a plain line welding, where, as Steve commented, doesn’t really matter. It’s too late now to change the question – which I realise now it might be misleading for readers used with standard rails measured in feet – nor to amend the answers … it is not only one correct answer, hence I’ve allowed multiple selection. Sorry for changing the voting rules during the vote.

What if we are talking about a standard 60 feet long rail? When is a 60 feet rail 60 feet long?

#### When a 20 m rail is 20 m long? Follow up:

The rail delivered at a standard length (i.e. 20.000 m, 60ft /18.288 m) has well-defined conditions of length measurement. An implicit condition is that, when measured, the rail should be un-clamped, free of compression or tension forces – stress free. Hence “At the stress free temperature” counts as a correct answer.
But this is not enough.
We need also a figure, a reference temperature, because a “20 m” rail, free of stress at 0 °C, is 7 mm shorter than the same rail, free of stress at 30 °C. Obviously, in this context by temperature we understand the temperature of the rail and not the ambient temperature.

##### CEN member and affiliated countries – European Norms

For the countries where the European Norms for railway track are implemented as national standards, the rail length standard measurement condition is defined by EN 13674-1:2011 – Railway applications – Track – Rail Part 1: Vignole railway rails 46 kg/m and above. Table 9 of this norm states:
The given rail lengths apply for + 15 °C. Measurements made at other temperatures are to be corrected to take into account expansion or contraction of the rail.
For switch and crossing rails, the same rule is defined in EN 13674-2:2006 + A1:2010.

In the United Kingdom this EN 13674-1 standard supersedes BSI BS 11 – Specifications for railway rails, which, at least in the 1985 edition, had this rule for standard rail length:
The lengths shall be based on a rail at a temperature of 20 °C (68 °F).
Daniel Pyke’s comment: BS11 is not really superseded, it still has to be kept (and was revised in 2015), because some of the rail profiles are still used and not in the EN standard. So this temperature rule might still be in place for the British specific rail profiles, not covered by EN 13674-1.

##### United States – AREMA Standards

The AREMA Manual for Railway Engineering refers to a temperature of 60 °F (15.6 °C) when specifies standard rail lengths.

Probably, around the world, other national standards specify for standard rail length measurement other rail temperatures than the ones quoted here. Therefore the correct full answer depends on where you are and even on what type of rail you measure …

Every time a track standard specifies a rail length, that length is assumed measured at a standard rail temperature, on a stress free rail. This includes the rail dimensions for switch and crossing trackworks. The practical significance of this convention can be considered together with the measuring tolerances allowed by the standards. In EN 13674-1 these tolerances are:

Length Measurement tolerance
both ends drilled
≤ 24 m
24 m ≤ 40 m
40 m to 60 m
60m
± 3 mm
± 4 mm
± 10 mm
± 20 mm
undrilled or one
end drilled
± 1 mm/per metre of rail
For special purpose undrilled rails the length tolerance is ± 6 mm up to 24 m and ± 10 mm for > 24 m rail.

These tolerances come from the measurement method (tape measurement usually) and not from any standard rail temperature allowance. For example, a 25 m steel measuring tape has a ± 2.6 mm tolerance (EU Class I).

A 10 °C variation from the standard temperature consumes entirely the measuring tolerance for the rails shorter than 40 m. The tolerance is defined to compensate a measurement uncertainty and not cover a rail temperature variation – a parameter easy to measure and consider.

## 3 thoughts on “When a 20 m rail is 20 m long?”

1. I guess I have a bit of an advantage with this as my old job was looking after the quality of new rails, one parameter of which is the length. My requested change to length specification to the standard even made it into the European rail standard! There are in fact 3 different answers that I’m aware of!

Like

2. Steve Hooper says:

From a pragmatic view, does it really matter ?
If being welded then rail is cut at current temperature and welded if jointed then joint gap set as per current temperature.
With 67C degrees difference giving length change of 15mm which is extreme and highly unlikely, I don’t see any issue.
But I do get your question, so at what ambient temperature is it first cut at 20m exact?

Like

3. Alan Babaee says:

I liked very much the way you asked the question and I try not to overthink about it. I believe when a rail has its neutral temperature and there is no stress in it, we can expect to have a rail with its right length.

Like

This site uses Akismet to reduce spam. Learn how your comment data is processed.