Page 2 of 2 FirstFirst 12
Results 11 to 16 of 16

Thread: Mullan Timetable consists??

  1. #11
    Join Date
    Feb 2013
    Location
    known universe
    Posts
    2,319

    Default

    Quote Originally Posted by ngc427 View Post
    Attachment 81523

    Attachment 81524

    There are screenshots of my world and activities folder. I installed the route straight from the installer, just to a different directory from the default and haven't changed anything with it other than making & adding the timetable mode, so the only difference should be a different root folder name.
    Well, everything looks correct...you'll have to do some detective work to find out where OR is getting this path >>>
    C:\ORTS Profiles\TSMullanPass\Routes\Routes\Activities\Wor ld
    next time it happens...PM me a log.
    Cheers, Gerry
    It's my railroad and I'll do what I want! Historically accurate attitude of US Railroad Barons.
    Forever, ridin' drag in railroad knowledge.


  2. #12

    Default

    Quote Originally Posted by geepster775 View Post
    Running the H-PASNTW, the run starts with a non-clearing jackpot at the shorty Avon siding. 2 EB's waiting on a WB that doesn't fit between switches, plus 2 more WB's queued at West and East Elliston unable to go farther. With all opposing traffic somewhere east of Garrison, player had the main all the way through Missoula.

    Changing the EB and WB paths to eliminate passing paths at Avon, the sim started with the jackpot gone and all WBs were spread out west of Garrison. The first WB was approaching Bonner, so the player gets crossed over at DeSmet and runs out of track authority at the radio-controlled (non-signalled) crossovers at the west throat of Missoula yard, kinda sets up a passing loop west of yard and will not allow player to pass through yard area on north (siding?) track. I have doubts the 'dispatcher' as its currently coded is capable of recognizing long loop zones passing thru so many switch nodes along the way. Probably more suited to path them as one way current-of-traffic zones.

    One other suggestion is don't start the paths quite so close to the end of track. The area is on a grade and trains roll backwards after entering sim and releasing brakes. By the time the GE's load up and start to pull, the train has rolled back and is soooo close to bouncing off the end of track 'bumper', which resets the physics to hard stop again. Cant throttle up too fast to 'lean on em' to avoid bounce off 'bumper' or you break knuckles.
    Thank you for the traffic jam report and suggestion. I'll get a fix for the timetable out sometime later tomorrow and give the route another shakedown to make sure that it runs smoothly.

    Quote Originally Posted by R. Steele View Post
    Well, everything looks correct...you'll have to do some detective work to find out where OR is getting this path >>> next time it happens...PM me a log.
    Once I get the patch underway I'll investigate the file path issue shortly after and get back to you with a log file. I have a few ideas of what could possibly be going wrong and I'll see if I can get them sorted out.

  3. #13
    Join Date
    Jun 2009
    Location
    Post Falls, ID
    Posts
    1,102

    Default

    Quote Originally Posted by geepster775 View Post
    Running the H-PASNTW, the run starts with a non-clearing jackpot at the shorty Avon siding. 2 EB's waiting on a WB that doesn't fit between switches, plus 2 more WB's queued at West and East Elliston unable to go farther. With all opposing traffic somewhere east of Garrison, player had the main all the way through Missoula.

    Changing the EB and WB paths to eliminate passing paths at Avon, the sim started with the jackpot gone and all WBs were spread out west of Garrison. The first WB was approaching Bonner, so the player gets crossed over at DeSmet and runs out of track authority at the radio-controlled (non-signalled) crossovers at the west throat of Missoula yard, kinda sets up a passing loop west of yard and will not allow player to pass through yard area on north (siding?) track. I have doubts the 'dispatcher' as its currently coded is capable of recognizing long loop zones passing thru so many switch nodes along the way. Probably more suited to path them as one way current-of-traffic zones.

    One other suggestion is don't start the paths quite so close to the end of track. The area is on a grade and trains roll backwards after entering sim and releasing brakes. By the time the GE's load up and start to pull, the train has rolled back and is soooo close to bouncing off the end of track 'bumper', which resets the physics to hard stop again. Cant throttle up too fast to 'lean on em' to avoid bounce off 'bumper' or you break knuckles.
    Avon strikes again... I have run into numerous problems with this siding. I had once leaned too hard on the statement in the OR Manual that indicates train lengths are taken into account by the OR dispatcher in setting up meets... I think if you have the very simplest situation - one westbound and one eastbound, and no other trains on the route - then the sim is smart enough to hold trains back at a longer siding. But it appears that the OR dispatcher is incapable of planning "far enough" ahead, i.e. looking not just at the immediate impending meet, but on 1, 2, even 3 trains out to identify potential non-fitter situations. In addition to that, the OR dispatcher is not smart enough to realize that if it sets up a meet between, for example, an EB fitter and a WB non-fitter, that it cannot send a 2nd EB train behind the 1st unless there is sufficient clear length and signal permissions for the 2nd east to be able to squeeze in behind the 1st.

    Long story short - With the current intelligence of the OR dispatcher, a timetable mode is going to be extremely difficult to manage on this type of route unless you set passing paths to be at ONLY locations where the longest trains can both clear. That's nearly impossible if you have a double-length empty grainer in the mix since they are non-fitting at a large proportion of the sidings on the Mullan route. For the sake of reducing the number of issues, you may be forced to not include any double empties in the timetable (the 230 car empty grain consist), and providing at least an hour of spacing between any trains in the same direction (and probably eliminate Avon as a passing location, as Geepster did.)

    More to the original topic of the thread: ngc, let me know if you ever want to develop a bit more realistic timetable. I have a word doc that I wrote up describing in detail the train symbols and their daily frequency. The symbols seen in the default Mullan consists are only a fraction of all the symbols that traverse MRL...
    Last edited by PerryPlatypus; 04-14-2021 at 09:08 PM.
    ~Sean Kelly~

    MRL Mullan Pass for ORTS
    https://www.trainsimulations.net/mullanpass

  4. #14

    Default

    Quote Originally Posted by PerryPlatypus View Post
    Long story short - With the current intelligence of the OR dispatcher, a timetable mode is going to be extremely difficult to manage on this type of route unless you set passing paths to be at ONLY locations where the longest trains can both clear. That's nearly impossible if you have a double-length empty grainer in the mix since they are non-fitting at a large proportion of the sidings on the Mullan route. For the sake of reducing the number of issues, you may be forced to not include any double empties in the timetable (the 230 car empty grain consist), and providing at least an hour of spacing between any trains in the same direction (and probably eliminate Avon as a passing location, as Geepster did.)

    More to the original topic of the thread: ngc, let me know if you ever want to develop a bit more realistic timetable. I have a word doc that I wrote up describing in detail the train symbols and their daily frequency. The symbols seen in the default Mullan consists are only a fraction of all the symbols that traverse MRL...
    Thank you for the feedback as well.
    I'm willing to get the timetable as prototypical as possible and if you don't mind sending the word doc my email and PMs are open. I try to get as accurate as possible with my timetable modes though research that I do, but sometimes it falls short of the real operations because of the scarcity of information out there about smaller railroad operations without knowing anyone who works at the railroad. I'll most likely issue the new version of the timetable with expanded train symbols with the rest of the fixes to the path files sometime later this week.

  5. #15
    Join Date
    Sep 2006
    Location
    .
    Posts
    2,629

    Default

    There is no rush for fixes, let folks poke around a bit more.

    I also tried to run the empty coke train U-RBOLAU, it was 3rd in a parade of 3 trains leaving Missoula eastbound with MISLAU leading and the Q-SSEATL in the middle. With close spacing I was beginning to see some of the same faults that we saw years ago when running the Rockslide activity for SLI Feather River. In that activity, you were third behind 2 other AIs, the first AI got away, the second AI immediately ahead of you had an emergency stop, during which the player pulled up right behind its caboose. When the stopped AI finally got moving and cleared the next siding interlocking, the dispatcher routed the player train in and out of the siding needlessly. The AI ahead was still moving in between the two switches, and the dispatcher saw the siding as the clearer track over a (everso slightly) greater distance and sent the player on that routing instead. So the player ended up going into and out of sidings with no opposing trains to meet.

    For some reason, the stack train ahead of the player kept to the main at Bonner, but got routed through the derail-equipped McQuarrie storage siding (a ballast train holding siding, no power switches) which slowed its progress down, which resulted in the player's empty coke train getting sent into Bonner CTC siding to meet nobody. I don't know why McQuarrie siding (adjacent to a rock quarry) would have a passing path set for it, it shouldn't, its not a CTC siding used for meets. Are there other storage sidings set up with passing paths in error like that?

    And I was a bit wrong earlier, there really aren't two main tracks (or main/siding) skirting the south side of Missoula yard, there is only one, along with some type of running track that starts about halfway, so the best you can do is have a quasi-passing loop on both the west side and east side. If you are not on the main, you end up going straight into the yard on the other track. I don't know how many BNSF trains run through vs actually get yarded waiting on a crew, but maybe we are trying to shove thru too many non-stops than the track layout is actually designed to handle.

  6. #16

    Default

    Quote Originally Posted by geepster775 View Post
    There is no rush for fixes, let folks poke around a bit more.

    I also tried to run the empty coke train U-RBOLAU, it was 3rd in a parade of 3 trains leaving Missoula eastbound with MISLAU leading and the Q-SSEATL in the middle. With close spacing I was beginning to see some of the same faults that we saw years ago when running the Rockslide activity for SLI Feather River. In that activity, you were third behind 2 other AIs, the first AI got away, the second AI immediately ahead of you had an emergency stop, during which the player pulled up right behind its caboose. When the stopped AI finally got moving and cleared the next siding interlocking, the dispatcher routed the player train in and out of the siding needlessly. The AI ahead was still moving in between the two switches, and the dispatcher saw the siding as the clearer track over a (everso slightly) greater distance and sent the player on that routing instead. So the player ended up going into and out of sidings with no opposing trains to meet.

    For some reason, the stack train ahead of the player kept to the main at Bonner, but got routed through the derail-equipped McQuarrie storage siding (a ballast train holding siding, no power switches) which slowed its progress down, which resulted in the player's empty coke train getting sent into Bonner CTC siding to meet nobody. I don't know why McQuarrie siding (adjacent to a rock quarry) would have a passing path set for it, it shouldn't, its not a CTC siding used for meets. Are there other storage sidings set up with passing paths in error like that?

    And I was a bit wrong earlier, there really aren't two main tracks (or main/siding) skirting the south side of Missoula yard, there is only one, along with some type of running track that starts about halfway, so the best you can do is have a quasi-passing loop on both the west side and east side. If you are not on the main, you end up going straight into the yard on the other track. I don't know how many BNSF trains run through vs actually get yarded waiting on a crew, but maybe we are trying to shove thru too many non-stops than the track layout is actually designed to handle.
    Fixed the McQuarrie siding issue, and also found that the Phosphate siding is also set up as a passing track even though it's not fully signaled + derail equipped like McQuarrie.
    I also fixed the issue with U-RBOLAU, my solution was to bump the starting time for the MISLAU back an hour, which changed the traffic pattern to fix the incorrect routing issue from the AI dispatcher.
    I have an interesting idea with trains stopping for crew changes at the yards where applicable, the other day I was playing around with wait points in the path editor and might incorporate that into the timetable somehow. I also noticed it allowed for semi-complex switching actions, which might make for an interesting dynamic of actually having to double up your train for the MISLAU or possibly even doing some locals in between the timetable action.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
-->