[Gro.Bi]

I have set up a pool with two storage-paths and two access-paths. But in the simulation only one storage-path is used.
Due to the document 'Changes to Timetable Mode May 2017' (found on ET on this link: link) I guess that it might be possible to use multiple storage paths in pools. I have created them (inclusive access-paths) as follows:

1^st access path:


1^st storage path:


2^nd access path:


2^nd storage path:


the train's path at its terminal station:


In my timetable I have some trains defined to start in the pool, to dispose in the pool and to be retreived from the pool, but only one storage path is used.
When the second train arrives at its terminal station and is to be disposed in the pool, it remains on its platform instead, even though the storage path's capacity would be sufficient to store it.
I have found examples for pools on ET and had a look at them, but it seems these pools have only one storage path.

Regards,
Gro.Bi

Attached File(s)

•  OpenRailsLog.txt (13.58K)
  Number of downloads: 186

[roeter]

Multiple storage paths are most certainly possible, I am using plenty of those on my own route.
Could you upload the pool's definition file, please? Just want to check that the definitions are correct.

Regards,
Rob Roeterdink

[Gro.Bi]

Here is the pool-file. My above statements are concerning the pool 'Nsk-P'. While uploading I had to rename the file to '.txt'.

Regards,
Gro.Bi

Attached File(s)

•  lollandsbanen.pool-or.txt (962bytes)
  Number of downloads: 191

[roeter]

There's nothing wrong with pool definition, I think the problem is somewhere else.
In the pool definition the maxunit is set to 2. This means that the 2nd train is set to join the 1st train on the first storage track. That's how pools work - the storage tracks are filled to capacity, starting at the first defined track, and only when that track is full will the 2nd storage track be used etc.
The access to the pool is protected by signals. It could well be that these signals do not support call-on. So the 2nd train cannot get to the pool because the signal will not allow it to move onto occupied track.
You can test this by setting maxunit to 1, this will force the second train to go to the (unoccupied) 2nd storage track.
If that works, you will need to look at the signal script and ensure it will allow a train to call-on into the siding.

Regards,
Rob Roeterdink

[Gro.Bi]

Thanks for the advices.

I have investigated and my result is as follows:

• #maxunits: The strange fact is that #maxunits;2 seems to have no effect, as all 4 trains starting in the pool are stored on one storage path, even though it obviously exceeds the capacity of the siding:
  
  
  
  In my timetable 4 trains are to start in the pool, whereof 2 trains have the 1st storage path and the other 2 have the 2nd storage path in the #path line; therefore I expect the 4 trains are located 2 per storage path each. Changing #maxunits to 1 makes no difference, as this line seems to be ignored (I found some lines #maxunits; line ignored in the Log-File).
  
  
• signals: I guess it's a good advice to look for a reason, but as I have only little knowledge about signalling, it will take some time for me to find a solution.

Separate from that I have two general questions about pools:

• for which OR version is the line #maxunits intended? I found it on the Internet https://open-rails.r...turntable-pools but I'm afraid it might be unsuitable for my currently used 1.3.4 version.
  
• is it obligatory to use callon in pools?

Regards,
Gro.Bi

[roeter]

 Gro.Bi, on 10 September 2021 - 08:08 AM, said:

#maxunits: The strange fact is that #maxunits;2 seems to have no effect, as all 4 trains starting in the pool are stored on one storage path, even though it obviously exceeds the capacity of the siding.

That's exactly why #maxunits was introduced.

Quote

for which OR version is the line #maxunits intended? I found it on the Internet https://open-rails.r...turntable-pools but I'm afraid it might be unsuitable for my currently used 1.3.4 version.

To be honest I have no idea when exactly this was introduced. The problem is that my own 'home' version has very little relation to any of the online versions, so I cannot check the version I am using.

Quote

is it obligatory to use callon in pools?

Many signal systems do not allow AI trains to move into occupied track. If that is the situation here, pools just won't work.

The reason is that most signals always check on block_state() ==# BLOCK_CLEAR. As said, that prevents AI trains to move into occupied track.
The complex way to solve this is indeed to use TRAINHASCALLON(), which is always true if the route leads into a pool.
The easy way: for the route leading into the pool, replace block_state() ==# BLOCK_CLEAR by block_state() !=# BLOCK_JN_OBSTRUCTED. This will allow access to the occupied siding unconditionally, but it is very unlikely that any other train except those heading into the pool will be set to run into those sidings.

Ofcourse, if the same type of signal is used all over the place, or if there is no specific route setting for these sidings, things get quite complicated. In that case, placing a 'dummy' signal (type NORMAL otherwise it won't work) at the entry to the siding which is always clear could also do the trick but that can sometimes lead to problems.

Regards,
Rob Roeterdink

[Weter]

Nice screenshot.

Just in pair of words, I had a success at defining a pool with two storage paths last summer. That worked.

As I heared, the train "counts" 30m or less from the front side of switch, which is marked by "red pole" at route editor, and game logic can't see the limitation pole of given track beyond switch.
So, sadly, looking to the picture, you have posted, I can suppose, the game logic is considering Nsk 13 able to fit for 4 units, so only 5-th and further will be directed to Nsk 14

[Gro.Bi]

 Weter, on 10 September 2021 - 01:29 PM, said:

So, sadly, looking to the picture, you have posted, I can suppose, the game logic is considering Nsk 13 able to fit for 4 units, so only 5-th and further will be directed to Nsk 14

I have tested with 5 units and the result is exactly as predicted. Therefore it seems that the #path line for trains starting in pools has no effect, as OR places the train on any vacant location within the pool, beginning on the first listed siding.

So there is no possibility for me to solve the situation as per now, except I would find an OR version that accepts the #maxunits line.

Regards,
Gro.Bi

[Weter]

Gro.Bi said:

#path line for trains starting in pools has no effect, as OR places the train on any vacant location within the pool, beginning on the first listed siding.

That literary was said in manual. Let's see:
11.4.7.6
$static /pool command description:

Manual 11.4.7.6 said:

The path must be a storage path as defined for that pool. Note that the train may be placed on one of the other storage paths as defined for that pool, this is defined through the pool logic.

[Gro.Bi]

 Weter, on 11 September 2021 - 01:35 AM, said:

That literary was said in manual. Let's see:
11.4.7.6

Sorry, I have overlooked that. But it's the explanation about starting trains in pools and how they are distributed on the storage paths. So my question about that is redundant now.

Regards,
Gro.Bi