Marias Pass: Unfixable

I didn't intend to, but I somehow suckered myself into experimenting with the Marias Pass (I was using MP3) switch node problem at Cutbank and Java East. Using default rolling stock, I put together a train and ran from Shelby to Cutbank:I tried the train with the "corrected" n/s value.I tried the train over it with "5" coupler strength.I tried altering the default timing value.I tried easing over it at 5 MPH.After all of the above: The train uncoupled. (Didn't BREAK the couplers, but still uncoupled almost every car that was eased over the first switch node at Cutbank.)So, I made a backup of the route, went to the approximate halfway point between Shelby and Cutbank (Ethridge) and cut in a short spur on the westbound main. Named it "Ethridge Team Track", then went into the sim and ran the train to Cutbank.Presto: No uncoupling.HOWEVER, doing this ruins almost every path in any activity, and since the .tdb is different after the track mods, you can't "fix" the broke paths by overwriting the previous paths into the Paths folder.So, it's like this: I'm of the opinion that Marias Pass (all current versions) can NOT be fixed to a usable state with the mods we've all thought would fix it. It would HAVE to have the track changes I experimented with tonight, all the activities fixed, then uploaded as an entirely new version of MP so that others could download/write activities for it and share them. Two lousy short spurs would have done the trick, and an interesting route with two personalities (mountain climber or grainger) would be entirely usable. Said two spurs being: An Ethridge Team Track and a Blacktail MoW Track, both on the westbound main w/west connections.And that's really a shame, because there could be some really nice switching activities written for it from Shelby to Cutbank and back. In fact, Marias Pass (3) was where I had intended for all my BN stuff to eventually haunt when I was in the mood.Ah, the things that could have been.Andre Ming

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

RE: Marias Pass: Unfixable, bobs, May-27-03, 01:38 PM, (1) RE: Marias Pass: Unfixable, Coonskin, May-27-03, 04:19 PM, (2) RE: Marias Pass: Unfixable, Swissie, May-27-03, 05:31 PM, (3) RE: Marias Pass: Unfixable, Coonskin, May-27-03, 06:10 PM, (4) RE: Marias Pass: Unfixable, Bing, May-27-03, 10:41 PM, (5) RE: Marias Pass: Unfixable, bobs, May-28-03, 00:47 AM, (6) RE: Marias Pass: Unfixable, Coonskin, May-28-03, 01:49 PM, (7) RE: Marias Pass: Unfixable, dndwilliams, May-30-03, 10:49 PM, (8)

Conferences | Topics | Previous Topic | Next Topic

What you learned is why I rebuilt all the activities that are included in MP30 and 31. Any changes to the track are likely to break the activity paths.However, with coupling values changed from 1 to 5 I have not had problems with uncoupling or breaking. I can see lots of movement in the consist, but it holds together.Bob Siederer
Co-author of Marias Pass 3.1

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON May-27-03 AT 04:21 PM (EDT)

Hi Bob:Guess I could do likewise after modifiying v3, but then I'm stuck with a "MP3a" one-off. Not good for shared activities.I was dumbfounded that I STILL experienced uncoupling at the infamous Duo of Disaster switches even after all of the changes I implemented. Were you able to fix the problem with finality via XTracks on the 3.1 version?I would definitely use 3.1, but I can't reinstall XTracks just yet, for I'm knee-deep building projects that use default track.Curious: How/where did you introduce a switch node on the westbound main before Cutbank and Java East on version 3.1?I still say the Marias Pass route has a lot of fun left in it. As mentioned above, I really like the mulitiple operational personalities of the route: Mountain climber, grainger road, interesting branchline, etc, as well as the different scenery you get on the opposite side of the pass. Quite a hand KUJU dealt us with Marias Pass. Andre Ming

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON May-27-03 AT 05:53 PM (EDT)

Andre,with regard to Cutbank, if it is the two fly-overs East of the yard: Did you check the area around and the ramp of the grade crossing itsself, too? IIRC, that grade crossing there can also be responsible for uncoupling cars, I remember earlier posts regarding some bb values given to that "toy-like" ramp for the cars in the default version of MP, but actually I thought I was then told by somebody it had been corrected in MP3.1 (can't find the exact post, but another one that also might be a hint at http://www.train-sim.com/dcforum/DCForumID6/9131.html. I've occasionally had cars uncouple either direction over that spot, and just the fact that you suffer an uncoupling instead of a broken coupler hints that there might be more to the problem in that particular area than "just" the old node problem - which also exists, but which should at least allow passage of trains with modified coupler stregth and coupler damping values. When taking a close look at a slow moving train that was made of cars such as some default equipment whose bb goes close to the ground (Orient Express carriages it was that I used, if I'm right), it was apparent that after juggling over the offending node, the cars somewhat again got stuck on that ramp of the grade crossing, until the leading car uncoupled, or when trying to run over that spot at higher speeds (limit is 30mph, I think, but already 20 or so would do it), the cars simply derailed and piled up. With regard to Java East: I haven't had problems there after the coupler damping modifications, expect for some jolting that triggered the warning in the activity evaluation that the cargo might have been damaged. Still, I agree fully with you that some flyovers or sidings would help with the node issue:1) Halfway between Shelby and Cutbank, with Ethridge being the "natural" location,
2) Halfway between Cutbank and Piegan / Pardue (another 15 mile stretch with no nodes!)
3) Halfway between Summit and Java East, with Blackfoot being the ideal choice.Side remark: As you probably now, there is a fifth "version" of the MP route available, as one earlier version of MP, I think it was MP 2.0, has been remodelled into JBJ or Sweden1, an electrified ficticious Swedish heavy ore haul route. In that lies a nice piece of irony: The authors that did all the changes from MP2.0 to Sweden1 were obviously unaware of the node problem. Although they did many changes to the track layout, including aditional spurs, sidings and fly-overs on dual track sections almost every few miles, they did leave exactly the three troublesome stretches of track , named above, as they were... So for example while they created a lot of new track around Essex and Essex Pit, the node problem between Java East and the balloon track still persists, same with Cutbank either side...Lukas a.k.a Swissie
(Edit for Spelling)

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Swissie:I'm not intimately familiar enough with the east side of Marias to realize there was another node problem between Cutbank and Piegan. I did notice the bb's of the default equipment seems to be quite deep. As for the westbound uncoupling at Cutbank, mine appeared to be happening at the switch node. i.e, immediately after the switch's points but before the road.Again, this has me puzzled, for I WAS able to get a train over Java East after the m/s corrections.If it weren't for messing up any downloaded/uploaded activities, I would go ahead an attempt to fix the route with switches where needed... and pray I wouldn't blow up the .tdb/signal/track object interface in the process.I may do that anyway for my own personal use... ONE OF THESE DAYS. Right now, I really need to be doing some other things besides tinkering with MP.As I've said, I really like Bob S's version 3 of the route... he's fixed the dipsy-doodles dramtically. (As well, I like the idea that v3 uses default track exclusively... perhaps enhancing the possibility of having Bob's version of MP in MSTS2.) Plus, I like the scenery, the diversity, along with its potential. I could geniunely get into an early-BN version of this route, as well as running some GN on it... just these infernal node issues. Andre Ming

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Try the Coupler Break value up over 7 - some put it at 9. Increase the DerailBuffer Force value to 1000kN. I run all kinds of trains through the Cutbank Crossover Curse, with no broken couplers, derails or unscheduled uncouplings. Per numerous prior threads, it appears to be a train length problem. Speed has nothing to do with it. Increasing values per above I've gone through there at ten miles over, no problem. Just don't look too close at what is happening at the crossover, it will raise your hair. The cars are bounncing back and forth like yuour Uncle Ed playing "Lady of Spain" on his accordian. I probably run two or three trains a week through Cutbank, always make it.Bing

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON May-28-03 AT 00:50 AM (EDT)

We didn't introduce any new nodes in MP31. What I did do was replace the 10 degree switches with 5 degree versions in many places, including the Cutbank crossovers and one side of the Cutbank bridge. The hope at the time was that this would help, but it didn't. It does, however, look much more realistic. Almost all of the mainline switches and crossovers were updated in the same way.I do recall fixing a boundary problem with a bridge somewhere, but I don't remember where...probably somewhere near Glacier.It is possible that replacing the 10 degree switches with 5 degree made some of the bad spots a little better, but not because of a node change. It would only be because of some other bounding box problem.Just to confuse the issue, I didn't use any track that is XTracks-unique. The reason XTracks is required is that it accounts for the sections that were missing from the original MSTS distribution, including all of the right hand manual switches that were missing. Those have been used throughout MP31. XTracks wasn't on my system when I developed MP31...I had manually added those few track sections to the tsection.dat. The simplest thing for everyone else is to just install XTracks, which I too have since done.
Bob Siederer
Co-author of Marias Pass 3.1

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Bobs:That sounds encouraging on the MP3.1 vs MSTS2 speculations. Good to hear. I really like this route, hope it can make it into MSTS2 in some form or another.BTW: Your signalling efforts are wonderful. Me, a complete AI traffic newbie, am creating my first involved activity, yet your signals cause the AI trains to perform as expected and designed. Great fun.Thank you so much for your efforts with Marias Pass.Andre

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Hi Andre,
I too had a similar problem and by doing things very slowly found out that the uncoupling occurred at the east side of the first road crossing beyond the switch. Went into RE and found out that it was a short section of Xtrack that joined the regular track section just a couple of feet inside of the east side of the roadway. Haven't had the time to figure out how to fix it yet (too much grass to cut now that spring has arrived plus other problems). May get back to it when fall comes and things slow down a bit.
Dick Williams

Alert | IP

[ Home | MSTS Forum | MSTS FAQ | File Library ] Copyright © 2002 by Nels Anderson. All Rights Reserved.