Bug? TI &T single track issue?

[AP]

If you set up a single track line with Time Interval with Telegraph signalling between two termini, and then from the same end, dispatch two trains a little over 10 minutes apart, where the first is a slow freight train, the second a fast passenger train, something odd happens.

If both trains are dispatched on clear signals (which at 10 minutes apart, they are) both should run at full speed. And eventually the fast should catch the slow and they should crash - unless the player creates e.g. a siding the slow train can pull into just prior to that point, as I was trying to do.

However, what actually seems to happen, is that the second train only runs full speed until a point midway into its journey then arbitrarily starts slowing down. Which the second train should not do, since it is operating under a clear signal.

I could understand the second train slowing down if it sighted the first train (825m / 7 tiles ahead or closer, as the videos tell us). But the second train seems to be slowing down for no good reason 20+tiles back from the first train - acting on knowledge it should not have.

A second side issue is that the Stop signal preceding my Siding 1 seems to show caution, not clear, when the second train reaches it, owing to the passage of the first train which has turned off the main line within the last 5 minutes. Surely it should know the next train is trying to reserve a different route and show clear? Or do I need to use a Junction signal in this application to get a clear signal to be given to allow the overtaking train to proceed at full speed for the rest of its journey?

Save file attached.

[jamespetts]

This is not a bug: what is happening here is that the train is slowing down for a signal protecting a junction. In the time interval working method, every signal protecting a junction is treated as being at caution and having a speed limit of half line speed. A signal's speed limit is disapplied only when the train passes another signal.

[AP]

This is not a bug: what is happening here is that the train is slowing down for a signal protecting a junction. In the time interval working method, every signal protecting a junction is treated as being at caution and having a speed limit of half line speed. A signal's speed limit is disapplied only when the train passes another signal.

So if I did not have signals on the main line at all, in the vicinity of the siding (i.e. limited signals at the siding to Station Signals on the siding yard itself), the second train would pass at full speed?

Alternately I could put a second Stop Signal at the exit of the loop the second train uses, and that would let it speed up again? Or would the loop exit points count it as another Junction Signal?

Or would both ways work?

[jamespetts]

So if I did not have signals on the main line at all, in the vicinity of the siding (i.e. limited signals at the siding to Station Signals on the siding yard itself), the second train would pass at full speed?

No, since then the signal at the station itself would be protecting the junction at the siding. The signal that protects the junction is the last signal before a junction, no matter how far before. If you removed the siding entirely, the train would continue at full speed throughout. Similarly, if you added another signal after the siding, the train would return to full speed after passing that signal.

The important thing to remember is that only a signal where there are no intersections of any sort between it and the next signal can be passed at full speed in the time interval (with or without telegraph) working method.

[AP]

Ok, thanks,  I think I understand.

The signal that protects the junction is the last signal before a junction, no matter how far before.

That last part seems a bit odd, if you think about it. I don't think it's intuitive to have it so a signal 200 tiles away is considered protecting a siding, in an era when signals can't be more than 10 tiles from a mechanical signalbox. 

It's going to lead to lots of "why are my trains going so slowly". Perhaps they should just slow for the points,  rather than the signal

I also think you might want to be careful of terms,  junction implies divergence of major routes, I wouldn't have considered minor station pointwork or sidings as junctions.

[jamespetts]

Ok, thanks,  I think I understand.  That last part seems a bit odd, if you think about it. I don't think it's intuitive to have it so a signal 200 tiles away is considered protecting a siding, in an era when signals can't be more than 10 tiles from a mechanical signalbox. 

It's going to lead to lots of "why are my trains going so slowly". Perhaps they should just slow for the points,  rather than the signal

I also think you might want to be careful of terms,  junction implies divergence of major routes, I wouldn't have considered minor station pointwork or sidings as junctions.

The last signal before a junction is the one that protects it - that is quite a fundamental concept of signalling. The historical records that I found are clear that signals protecting a junction in the time interval working method never showed clear: only danger and caution, so what we already have is the closest reasonably obtainable simulation to reality as documented.

[AP]

Why are trains slowing 25 tiles ahead of a signalled junction, when distant signals can be placed at most 10 tiles from a signalbox, and are therefore almost always considerably closer than 25 tiles from the signal they protect (for a single passing loop with a signal box, distant signals usually end up about 7 tiles ahead of the stop signal, since 7*2=15, leaving 5 tiles between for a 3-tile passing loop+ pointwork.

The trains seem to be slowing earlier for junctions than they need to for stop signals?  If they can stop in 7 tiles, there is no reason for them to slow down so early.

--

I agree that logically the only signal protecting a junction can be the last one preceding it. But I'm not clear that extending the speed limitation so far from the junction, in the event the player forgets to put signals in proximity to it, is helpful, since the risk is local to the pointwork. You should consider setting a radius for it?

Making every turnout a "junction" is slightly unbalancing since trains could happily run over some pointwork at full speed, it was only facing points that were risk items for derailments, trailing pointwork was fine. But I can't see how you might distinguish the two.

[AP]

Hi James

I had another go at signalling this one. The second (fast) train still doesn't speed up to full speed again after the siding, in spite of passing a stop signal after the siding exit pointwork, and their being another stop signal at the east terminus protecting the pointwork there.

What is stopping it going full speed for the second half of its journey?

[jamespetts]

That seems to have been caused by a bug, which I have now fixed. Thank you for the report.

[AP]

No trouble, thanks James for the help.