Answer to 2002 Part A Q2 - This One I Missed Out

[alexgoei]

Hello Peter,

For some reason I missed out submitting my answer for 2002 part A Q2 on the Points Question.

At that time I was taking some time mulling over my answer for which I found so few entries wondering if I have missed out anything.

I have tried indicating point-to-point on this table as well as certain assumptions.

Please let me know if my swinging overlap entry is correct.

Look forward to your reply and comments

Thank you and Regards

[PJW]

2002 part A Q2 Points Question.

I have tried indicating point-to-point on this table .

It is certainly a case for classic point-to-point; however I do wonder whether you'd be best off not attempting to show (as I think you are a little confused by it) but merely use that recognition as a prompt for getting the general locking correct.

If you do want to get the details of point-point correct, first consider the combinations
211N 212N valid
211N 212R invalid
211R 212N valid
211R 212R valid

Then think how one point can tell whether it can change from a valid to another valid combination.

Imagine yourself as 211.
Since the start position must be valid, if 211N then we definitely know 212 is N. Hence no restriction for 211N to R.
However if 211 starts R, then for it to go N it must check that 212N (since if 212R then an invalid combination would result); this is the point-point: 211 R to N requires 212N

Then imagine yourself as 212.
If 212N then it doesn't matter where 211 is; both combinations are valid.
However for it to go R then it must make sure that 211R (as 211N would give the invalid combination); hence there is point-point: 212 N to R requires 212R.

Note that the point to point locking affects each point just in the one direction and the two lockings are complementary. What we are saying is that "one or other or neither" may direct trains onto the diamond BUT NOT BOTH. The first one that does so prevents the other from so doing.

So you did get it right BUT then you went and spoilt it by finding erroneous stuff as well that rather detracted.

[PJW]

answer for 2002 part A Q2 Points Question.

I have tried indicating certain assumptions.

You have stated that the joint DE/DC is foul (you'd be better refering to "moves over 212R" rather than naming CD).
I know it looks like it, being opposite the DE/CD joint. However I am sure that it isn't; based on the illogcality of providing DC as a separate track circuit if it really were foul. Either it would have all been the same track or the joint positioned so that it could be clear- it has no reason to exist if it does not prove clearance.

However although I say that your assumption is wrong, by putting it on then I fully understand why you put DC in the track locking etc so recording the assumption was 100% the correct thing to do.

[PJW]

my answer for 2002 part A Q2 Points Question.

I found so few entries wondering if I have missed out anything.

There wasn't much.
Dead TCs-
211N to R should be prevented by CD unless occupied by a parallel move so hence should require "(CD or 212R)"
211R to N should not be locked by CD.
You are right that 211R to N should be locked by (DE or 212N) EXCEPT that by including the point-to-point the move of 211 requires 212N ANYWAY so the track entry is actually superfluous because of that decision.

Similar comments apply to 212: the locking relative to CD should only be affecting 212 from R to N and be "(CD or 211N)".

Route calling: remember that you need to show the calling of the "lead" points!
You have 134B(M) calling 212R, you have said that for 212 to go R you need 211R so therefore you should have made 134B(M) call 211R; otherwise it just ain't going to work is it...... so there are several of these to think about.
You'd still want to put this calling in, whether or not you actually put in the point-point locking.
If you don't show point-point then you need also to show the route locking after 134B(M) on 211's CT; if you do show point-point then you don't [since 134B(M) holds 212R and 212 holds 211R]

[PJW]

my answer for 2002 part A Q2 on the Points Question.

Please let me know if my swinging overlap entry is correct.

A reasonable attempt for 212 N>R.
There are the two "extra" things in the overlap via 212R
a) CD track
b) 211R (doesn't actually give useful flank since unidirecional so perhaps a bit arguable)
In this particular case the 211R is not a swinging overlap control since you decided to put it in as point-point anyway. Not convinced whether you thought this through and therefore correctly decided not to have a column for points in the swingin overlap controls or whether you completely overlooked and hadn't realised. As an examiner you'd be given the benefit of the doubt BUT do remember that in the general case the controls are TRACKS, POINTS, OPPOSING ROUTES.
You got the three routes associated with that O/L (though the track listed after 138A(M) looks rather like DB rather than DG so be careful!)


However you should also have made similar entries for 212R>N as in this direction of swing then there is the extra track DC within that O/L.

Also be aware that if you had not included point-point then 211 could have been N and 212R at the time of the routes demanding the O/L being set; hence there would need to be a call of 211 N>R by those routes if 212N. Conversely if 211 were locked N (e.g. by point switch) then those same routes would need to call 212R>N instead.
Obviously with the calling there would need to be the relevant locking to hold the points in that position until the overlap had timed out.

Hence I suppose a lot of the complication in this question was the swinging overlap; looking at it the decision to implement point-point would save quite a bit! You don't even have to make those routes call 211 N>R when 212R since with point-point 211 couldn't be N initially and the locking of course comes essentially for free.

So yes in the circumstances not a bad attempt at all- but I do wonder whether this was a bit more by luck than judgement; only you will know whether you'd be able to tackle a more typical swinging overlap.......... Don't worry about them excessively though; it was a significant part of this unusual paper but generally speaking far less significant in comparison with other things

regards,

PJW

[alexgoei]

Hello Peter,

Thank you for your reply.

It was a fairly unusual layout and I tried to tie it back to the point to point exercise that you previously provided (the one with the left hand junctions etc.).

I will go through it tonight after work and if I do have some more clarification will get in touch with you.

Thank you for replying while on holiday.

Regards