Upcoming Games

No games to display

Full list
Add a game

Upcoming Events

No events to display

Two failures on the same signal?

You are here: Home > Forum > Simulations > Released > Fenchurch St > Two failures on the same signal?

Page 1 of 1

Two failures on the same signal? 08/12/2019 at 21:49 #122152
Albert
Avatar
1313 posts
In the attached save, two signal failures in the East Ham area were reported but in F11 there is only one failed signal - 157. After the first failure it had only a green aspect and that is still the case. No trains were running on the Down line when the second report came in, however there was a train coming from platform 8 on the Up at that time.

Did 157 fail twice (if that is possible?)

Post has attachments. Log in to view them.
AJP in games
Log in to reply
Two failures on the same signal? 09/12/2019 at 02:59 #122161
GeoffM
Avatar
6274 posts
Albert in post 122152 said:
In the attached save, two signal failures in the East Ham area were reported but in F11 there is only one failed signal - 157. After the first failure it had only a green aspect and that is still the case. No trains were running on the Down line when the second report came in, however there was a train coming from platform 8 on the Up at that time.

Did 157 fail twice (if that is possible?)
I believe that is possible, yes. Each lamp can fail independently.

SimSig Boss
Log in to reply
Two failures on the same signal? 14/12/2019 at 16:27 #122277
Albert
Avatar
1313 posts
Okay, but on Fenchurch Street it seems like some items are more prone to failure than others and fail multiple times in one game. I had two signal failures on signal 157, two TCFs on TADB and three on TCTP in one game of Fenchurch Street. Is there a bug in the sim that prevents certain items from failing so the failures get 'concentrated'?
AJP in games
Log in to reply
Two failures on the same signal? 14/12/2019 at 22:12 #122279
GeoffM
Avatar
6274 posts
The probability of the same signal failing is the same probability as a different signal failing. In reality a signal that changes aspect more frequently will have a higher probability than one that changes less often.
SimSig Boss
Log in to reply
Two failures on the same signal? 16/12/2019 at 10:47 #122295
clive
Avatar
2736 posts
Albert in post 122277 said:
Okay, but on Fenchurch Street it seems like some items are more prone to failure than others and fail multiple times in one game. I had two signal failures on signal 157, two TCFs on TADB and three on TCTP in one game of Fenchurch Street. Is there a bug in the sim that prevents certain items from failing so the failures get 'concentrated'?
Not per se.

The sim author tags each item as "can fail" or "can't fail" (sim authors: I'm talking about setting a location). I'm pretty sure all objects in Fenchurch Street can fail apart from those that are on the fringe and belong to another signal box. I'll try to find a moment to check.

Each time a track circuit is occupied, a signal changes aspect, or a set of points move, a random number is chosen and, if it's small enough (based on the slider position), that TC, signal, or points fails. So places that are busy are more likely to fail. But each aspect change has the same probability of failure irrespective of which signal it is or whether it's failed before: "the dice have no memory".

Beyond that, it's just how the dice happen to fall. Any statistician will tell you that you need to do hundreds of trials before you can say there's bias in there. You notice the unusual cases and not the ones where the failures are scattered.

(I remember a points failure in Cambridge at a meet which was the worst possible place in the entire sim. I still remember it after all those years. But I *know* those points aren't treated specially in the sim data.)

Oh, one thing about points and track circuits: if you have a multi-ended set of points, such as a crossover with the two ends co-acting or the 6-ended points in King's Cross and Euston throats, those will have two or six times the probability of failing because each end can fail separately. Some displayed track circuits are actually two or more real TCs in the sim logic (or in the real world); those will also have a higher probability for the same reason. I would have to check the Fenchurch Street data to see if that applies to the ones you mention.

Log in to reply
Two failures on the same signal? 16/12/2019 at 16:25 #122308
GeoffM
Avatar
6274 posts
clive in post 122295 said:
Oh, one thing about points and track circuits: if you have a multi-ended set of points, such as a crossover with the two ends co-acting or the 6-ended points in King's Cross and Euston throats, those will have two or six times the probability of failing because each end can fail separately. Some displayed track circuits are actually two or more real TCs in the sim logic (or in the real world); those will also have a higher probability for the same reason. I would have to check the Fenchurch Street data to see if that applies to the ones you mention.
To clarify, the probability of each end of a point failing is the same: it's the combined effect of multiple ends that has the "bad day" effect.

SimSig Boss
Log in to reply
Two failures on the same signal? 17/12/2019 at 09:32 #122318
kbarber
Avatar
1708 posts
But of course there are always some points (or signals or TCs for that matter) that are known to be problems. BK283 points (now 2052, down main towards P7 at Barking) were a classic example. In those days, there were facing points in the down main and switch diamonds in the up main. And they were always failing, to the despair of the local S&T. Perhaps small wonder the switch diamonds gave way to a fixed set when the line was resignalled. And as for North Pole Junction...!!!

(Incidentally, the North Pole example shows that things don't only happen with a change of state. There was one particular TC there that would happily spend an entire night shift randomly failing (creating a block failure from Mitre Bridge in the process) then clearing again, without a train anywhere in sight, sometimes 8 or 10 times in the course of a shift. In the end it was only exorcised when they rewired the box completely in, IIRC, 1983/84.

Log in to reply
Two failures on the same signal? 18/12/2019 at 00:35 #122324
Edgemaster
Avatar
332 posts
kbarber in post 122318 said:
In the end it was only exorcised when they rewired the box completely in, IIRC, 1983/84.
Ghosts I tell you! Ghosts!

Twitter
Log in to reply
Two failures on the same signal? 18/12/2019 at 09:54 #122326
kbarber
Avatar
1708 posts
Edgemaster in post 122324 said:
kbarber in post 122318 said:
In the end it was only exorcised when they rewired the box completely in, IIRC, 1983/84.
Ghosts I tell you! Ghosts!
It did feel like it on nights sometimes. Sitting reading, with the gas lamp hissing quietly in the background. A soft 'clunk' from under the box. Glance up to see the TC in advance of the up home from Mitre Bridge showing occupied; a glance at the block instrument confirms the needle is now at 'Train on Line'. It's not worth disturbing Mitre Bridge; we all know the drill. Sometimes (extremely rarely), there's another 'clunk' some time later and the TC clears; now put the book down, turn the block to TOL, wind out the Welwyn release and drop the needle, before Mitre Bridge has a train to offer. But usually the next thing is when The Mitre offers the next train, at which point I get him on the phone to tell him the [OBSCENE] track circuit has failed again and accept under Reg 25(a)(iii). Whereupon (and most often before he sends Train Entering Section) the track decides to clear.

The Sleep & Tea spent many hours at North Pole trying to fix that blessed track. (I suspect the workload frustrated many a scheme to rationalise the S&T gangs around the North London though, so they had nothing to complain about in the end.) So not all ghosts are unfriendly :-)

Log in to reply
The following users said thank you: WesternChampion, as1981