The locations I've seen sand drags are to create friction upon the wheels of a runaway - to slow its forward motion and to limit the effect of it hitting the bufferstop at speed; as that could risk deflecting the runaway into the running line(s)
Images of Clarkson, Hammond and May. "I mean - what could possibly go wrong?"
They were looking for something a bit more high-performance than a traditional sand drag (quote from source above).
According to the report linked off that page, the foam beds were actually built and tested. I'm pretty sure I've seen a film of one of these.
I agree it doesn't really make sense. If I'm reading the report correctly, 100mm was only used for the last test in section 1.1, where if the acceleration was constant it would have been 1.5m/s2. For the highest speed test with 50mm sand depth it would have been 0.97m/s2. So it's possible a lesser depth would have given them the 0.6m/s2 they were looking for, although the train's mass would have been another factor (20 tonnes tested against a requirement for 193 tonnes).
Space is presumably a concern for this solution to be practical around the network.
Sand drags or similar may be provided where a double track converges to single. These are places where a SPAD could lead to a head-on collision at a high closing speed, but also may be used by trains with passengers so the derailment after a SPAD could also be hazardous. Many of these are former double track so the sand drag (or some alternative) can continue on the old formation for as far as considered necessary. Some may have been removed since TPWS was introduced, which addresses the same hazard in a different way.
They were my first thought on seeing the thread.
My late father was involved in some testing when he worked at London Transport (as it then was) involving 1938 tube stock and District line (CO/CP) stock being driven at 15-20mph-ish into sand drags at Upminster depot. He took cine film of it and it was fascinating to watch. The trains ploughed in and derailed/rode up but stayed upright and pretty much on course. See this video from 12:27 onwards.
Not entirely accurate.
While I agree that buffer stops should not be driven into during normal train operation, if they really just existed to mark the end of the line they would have much simpler construction.
They are intended to provide an extra protection (quite literally a backstop) if the non-normal situation arises where a train doesn't stop before reaching one. As such they are subject to risk assessment which considers the likelihood and severity of consequences.
That's a light engine touching the buffers at well below walking pace, then presumably applying a certain amount of tractive effort to confirm they move. If used in anger, the buffers would only be able to apply a similar force to that for their distance of travel (it's possible the force increases over that distance). Consider what speed a Class 56 could accelerate a train to over a distance of a couple of metres, and the same train hitting the stops at above that speed will run them through their travel and hit some kind of hard stop after that.
Sometimes you get both. First a red light on a stick. Then something more substantial.
The ones at Glasgow Central use water which goes into a reservoir. Up until about 30 years ago it wasn't unusual for loco-hauled trains to touch or even slightly compress the buffers at Glasgow to make maximum use of the available platform length.
Might not be so relevant for units, but it’s still permitted to close on to buffer stops for operational requirements.
As was I
I remember when all the castings broke during a cold snap in 1996. This was Platform 9:
