An interesting Japanese research published in 2014 shows a comparison between the lateral resistance tests undertaken on single sleeper and tests on multiple sleepers.

I strongly recommend to read the full research, still available for free at the date of writing: Numerical method for evaluating the lateral resistance of sleepers in ballasted tracks – ScienceDirect. I’ll be selective in my summary and I’ll simplify a lot below. I might very well be wrong!

One vs many – the unlikely winner!

They tested several types of sleepers that presumably should have, due to their more “grippy” shape, a greater lateral resistance.

The results for single sleeper tests show a huge difference between the more complex shaped sleeper and the normal box shape (up to 60% increase in lateral resistance).

However, as the number of sleepers was increased, the lateral resistance of the group, ended up being lower than what was measured for single sleeper tests, and the difference between the excitingly shaped sleepers and the boring box we use, was less and less significant.

See below three graphs that speak for themselves – all credit to the above linked article authors – these images are uploaded on this blog for ease of reference.

a – tests on normal shaped sleepers

b – tests on winged sleepers

last – summary graph

In this last graph we see that the 20mm winged shaped sleeper on one sleeper test showed a (track?) lateral resistance about 60% higher than the normal rectangular one. 20mm?! Yes! They are scaled. I told you, read the article.

On multiple sleeper tests, the shappy (sic!) sleepers are still stronger but only by a maximum of approx 20%.

One sleeper vs multiple test comparison shows that for the winged shaped sleeper, the lateral resistance decreased by approx 80%! (second graph). The normal sleeper lost only 15-20%. (First graph).

For three of the complex shaped sleepers the lateral resistance for single sleeper is almost twice of the one measured for multiple sleeper tests.

So the apparent revolutionary shaped sleepers proved not to be as revolutionary after all, when tested on multiple sleeper panel, closer to a real life usage. Their one sleeper resistance almost halved.

What does this mean?

How can one sleeper be stronger than 3 or 7 or 16?

Are we on the verge of a new railway track order? Do we need only one sleeper on a line?

No, my dear reader.

These tests show that the weak link in the track system is not the sleeper but something else.

When we increase the number of sleepers in the test setup, that weak link reveals more and more its weakness and shows a more accurate resistance of the track system.  You can’t truly compare sleepers just by doing single sleeper tests.

In the UK we have recently seen the results of some very interesting and promising tests on one sleeper setup. We need to see tests on multiple sleepers to trust fully the results.

Don’t bet on winners when the race is with one sleeper per team.

Where hump?

Oh, you noticed, dear reader, that these results and so many other lab tests results we can find online don’t seem to match the shape of this graph, shared in a previous article, for a site test undertaken by British Rail in 1973:

Where is that strong resistance hump we see before tamping and after 34 days of traffic, you rightly ask?!

The Force is strong with this one!

Although the lab tests on one or multiple sleepers might show some interesting results the ultimate test is the reality check of the site.

There is a track buckling video you can find on various channels and also linked in one of my previous articles. I’m not sure of who owns the right to it, I’m sharing it here for information – happy and honoured to quote the rights owner.

My understanding is that this is from one of the so many buckling tests undertaken by the US DOT Volpe Center, well known from their CWR research.

Interesting here is that there is practically no ballast shoulders and there are lateral resistance plates installed in the fourfoot, at every four or five sleepers.

When the buckling occurs we see that the LRPs still stay in place and not move together with the sleepers. So, in this test at least, I don’t think the lateral resistance plates are the ones that fail but the screws that fix them onto the sleepers.

Even if I would have the results of this test – which I unfortunately don’t have, yet (but you know me 😉 ) – I doubt that the ultimate lateral resistance measured here is the actual maximum resistance these annoying yellow things can provide.

My personal opinion is that the lateral resistance force is site proven the strongest with this one: