Understanding headways: Signet Weekend 2003 Handout

[nthomso3]

Peter,

Whilst looking through your handout given at the Signet weekend re. 2003 Module 2 headway calcs I got a little confused.

I understand your working for the non-stopping headway and that you are using the calculated spacing as a guide to whether the stopping headway can be achieved, however in your diagram you have shown train 2 approaching a YY aspect, which has just stepped up from a Y, as train 1 clears the overlap of the platform exit signal. Should it not be the case that this train should be approaching a green aspect, just stepped up from YY when train 1 clears the overlap?

If my thinking is correct then train 2 must traverse another full signal section, 1000m @ 33.3m/s, so another 30s must be added to the time taken. The total time is still within the 300s specified but has a lot less contingency.

Forgive me if I am being a bit stupid here but I was of the impression that headways should be calculated from G to R and not from YY.

Neil

[Peter]

Think about the effect of the aspects seen on the way the driver would drive his train. IF he is due to stop in the station, and intends to have a "normal" sort of brake application to achieve that, he will start braking at normal braking away from the platform. Up to that point he has been on green (he only sees the YY because he is catching up the stopping train which is about to move away) and hence until the sighting point of the YY the presence of the train ahead is of no consequence to him. The brake application at this point does not affect his time to clear the section becasue he has to brake for the station anyway.

I hope I have managed to explain that clearly.

Peter (the other one)

[nthomso3]

But if train 1 has only just cleared the overlap, train 2 has already passed a YY aspect in the signal prior to it's position in the diagram so could not have been running on greens?

[Peter]

But if train 1 has only just cleared the overlap, train 2 has already passed a YY aspect in the signal prior to it's position in the diagram so could not have been running on greens?

Lesson number 1 for self - look at the diagram before answering. I see what you mean given the number of signals shown. Not quite sure what PJW was meaning.

[PJW]

But if train 1 has only just cleared the overlap, train 2 has already passed a YY aspect in the signal prior to it's position in the diagram so could not have been running on greens?

Lesson number 1 for self - look at the diagram before answering. I see what you mean given the number of signals shown. Not quite sure what PJW was meaning.

Mea culpa- you are right. The dangers of not drawing out all the diagram and trying to do against time.

The critical section is generally the one where the train actually dwells in the platform and thus the one that I showed at Yellow and therefore the brakes would first be applied when sighting the signal that I haven't drawn at the left end of the diagram.

However the "traditional" assumption is that the signal prior to the station need only be displaying Y (and thus the previous signal YY) - see the series of the small (imperial equivalent of A5 size) "Green books", certainly where the "headway" speed significantly less than the "maximum permissible". Drivers were expected to "know" that the signal spacing was generous with the braking distance that they actually required. This is not behaviour we would encourage (or even tolerate) today and so I attempted to make more realistic by reflecting "defensive driving", but as you have correctly pointed out did not take this consideration far enough.

Perhaps insisting on a Green at that approach signal, whilst technically "pure", may in practice lead to a more pessimistic view of practical headway than is reasonable. A driver seeing the YY braking a bit to kill the speed and then coasting to be ready to brake again at the next signal would instead see it as YY and so continue to coast only gradually reducing speed until the Y on the signal protecting the station is sighted. So yes the driving of train 2 is being affected by train 1 but the time "lost" would not have been so much as might first be supposed.

So yes I am sure that I would have lost a few marks, but actually I think not that many. Do look at the mark allocation- we don't get much of a clue for the 2003 paper but in 2007 there were only 20 marks for all calculations and determining signal spacing- so if we said for simplicity: 5 for braking, 5 for non-stop headway, 5 for stopping and 5 for actually deciding and explaining what form of signalling to use, then there are not that many on offer at all. OK the split may not be exactly even, put I can't see that stopping headway could possibly be more than 8 marks. So of those how many have I got; I have shown the concept, have performed the calculations with some explanation and got a reasonable answer so surely I must have got 75% of them. Therefore perhaps I have lost 2 marks.

So absolutely right, well done for noticing, it shows you are understanding. One "advantage" of taking the very pure approach will be that you'll probably shift your solution from the typical "is it just do-able with 3 / do I need 4 aspects?" grey area where many of the IRSE layouts seem to be designed for. Can save some agonising and need to transition 3 to 4 and vice versa BUT you give yourself more signalling to show on your layout in the limited time.

Also think carefully about the 3 aspect case with stopping headway and see how much improvement in headway is theoretically achieved by 4 aspects; assuming too much defensive driving suggests very little difference in the solutions, yet in practice we all know empiracly that 4 aspects help a lot. This really is what I was discussing above in the situation where a driver DOES see a restrictive aspect, slows down a bit but then the separation between the trains stabilises and they get another YY. What this really tells us is that there is only so much that can be done reasonbably graphically and via calculation; in reality both the brake rate and acceleration rate are far from constant. Nowadays we use modelling packages using computers (and eve these are far from perfect algorithms); the purpose of having in IRSE exam is so that people have an idea of the CONCEPT rather than really getting the accurate answer.

Perhaps what I should have done was explain a little more re the assumptions and approximations I was making. There is a lesson there; also keep focussed on whether the effort to do anything is likely to be worth the improved marks- always monitor your "marks per minute" Key Performance Indicator.

Thanks for pointing this out,

regards
PJW