Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Pointwork- Signalling the Layout
#1
Attachment is one animated Powerpoint (ideally 2003+) slide giving some hints re identifying Normal lie / numbering of points and the track joints whch are required in and around pointwork.


Attached Files
.ppt   Point lie and IRJ.ppt (Size: 808.5 KB / Downloads: 595)
Reply
#2
Good PPT PJW. Shame it was a pain to open. Can we save this in another format such as PPT97 or OpenOffice please?

Thanks, Jerry
Le coureur
Reply
#3
A handy bite-sized presentation!
Thanks
Reply
#4
Quote:Very useful presentation Sir. Thank you.


Attached Files
.pptx   Point lie and IRJ.pptx (Size: 748.36 KB / Downloads: 10)
Reply
#5
Hi,
Please could you explain why you paired the switch diamonds, 723a/b but did not include 722 points in the same point identity. I would have made that set a triple ended set of points with the same identity as they are all required to be normal for parallel moves, and all reverse for moves from 724 approaching 720?
Regards
Mark
Reply
#6
You would be perfectly right to do so; indeed it is the traditional thing to do, much of Network Rail's existing signalling is like that and the examiners would probably expect just that as a solution (which I have made clear in the Module 2 IRSE Study Pack).

The reason why I showed as I did is that I was following NR's most recent standard; I am afraid that like many of the newer ideas I cannot provide a convincing rationale; perhaps I am getting "long in the tooth" myself. To be charitable,
1. there is less "additional cost" when the interlocking is computer based rather than requiring additional relay hardware to implement (and indeed on systems such as Ebilock / SIMIS-W each point end is actually a separate entity within the geographical interlocking anyway, however the signal application engineer numbers their Signalling Plan)
2. there is possibly a potential benefit in failure mode situations.
Personally however, if I dictated policy then I would not have changed; I believe that there is still "cost" and I think that almost all the advantages can be achieved by retaining numbering as one three ended set PROVIDED the single lead (722) had a separate detection input to the interlocking than the switch diamond ends.

I believe that NR often sets policy on not much more than a whim and does not properly assess all the options- it correctly identifies that there are problems with some existing 3-ended installations and that they would be overcome by setting a maximum of two ends per point number, yet fails to realise that there is a "middle way". Read "Modern Railways" magazine re "boiling frogs" and the constant ratcheting up of industry costs because of the loss of ability to balance various factors to come up with the most appropriate solution relying upon engineering judgement ....better stop before I get too political and upset my current employer!

For the exam then, feel free to number as you describe, but if you do so then do not claim that your layout is adopting "current NR standards". Perhaps "British Mainline practice, circa 2000" is a better thing to quote for this and other issues.

Hope this resolves your query- if not then post further follow up,

PJW


MarkN Wrote:Hi,
Please could you explain why you paired the switch diamonds, 723a/b but did not include 722 points in the same point identity. I would have made that set a triple ended set of points with the same identity as they are all required to be normal for parallel moves, and all reverse for moves from 724 approaching 720?
Regards
Mark
Reply
#7
PJW Wrote:For the exam then, feel free to number as you describe, but if you do so then do not claim that your layout is adopting "current NR standards". Perhaps "British Mainline practice, circa 2000" is a better thing to quote for this and other issues.

Three is not outlawed by current NR standard. 11600 part F3.5.1 says (in the bit about devices operated together having the same number) "Such and arrangement should, wherever practicable, be restricted to two point ends in order to simplify testing and corrective maintenance" further saying that you can go to three ends if it is "particularly expedient" and if the infrastructure controller agrees.

I agree with Peter that an arbitrary ceiling of two is a bit low, clearly a quick look at the SMTH permutation table for four, five or six common ends for the detection testing matrix starts to tell you this is a bit OTT.
Reply
#8
Thankyou.
I agree with the points that you present. If we are taking regard of the onerous nature of the SMTH necessary after failure then I feel that the solution originally offered was best.
Thinking about the triple ended set, it's long time to get back into service problem could be mitigated against by having split detection on each end, which would result in one set of control logic out, & separate detection on the way in.
Another note, might it be a good idea to mention the type of interlocking which this is being designed for?

Relay Interlocking, the aim is to reduce the number of control lines (cores) while still being safe and meeting the customer specification.

CBI, depends on application but may force you to have single ends (Ansaldo ACC)

SSI is only capable of 2 points identities per module, and so would prefer points to be paired where possible to reduce the impact on the interlocking. ie less point modules.
Reply


Forum Jump:


Users browsing this thread: 2 Guest(s)