Strange Problems on Horseshoe Curve...

[TVRRMAN]

ok guys and Gals, I've got a real brain buster and was wondering if anyone else has experimented with this and encountered this:

For experimenting purposes, I went about and have tried to create the infamous three and four track meets that could and occasionally did happen up at Horseshoe Curve. I'm following PRR-Specific Rules (also something breached apparently with the free roam scenario that has you rail-fanning the curve).

in the included image, a triple meet is about to occur. What isn't pictured is the said "Consist tilt" derailment that occurred directly after the passenger train got beside the coal drag heading upgrade on track three.


Consist excessive stretching...


Derailment (shown for illustrative purposes only)


Anyone encountered this? Any easy solutions to the Consist tilt derailments? (NOTE: happens on both Legacy mode and in TS2012 mode)

Thanks in advance for your insights / opinions...

[Kali]

How heavy are the passenger cars, and what is the chance of derailment? I've had some really bizarre results with trying to make weightless stock I can throw on the back of anything to log performance data with.

[TVRRMAN]

How heavy are the passenger cars, and what is the chance of derailment? I've had some really bizarre results with trying to make weightless stock I can throw on the back of anything to log performance data with.

Kali,

Weight on the tested stock is as follows:
BM70m / BM70na RPOs: 33 tons /Chance of Derailment: 0.4375
Tested RSC Stock: 49.895 tons / Chance of Derailment: 0.4375 (amfleet - AmCafe cars)
GN Chair Car: 56.0 tons / Chance of Derailment: 0.4375

During all tests, the RPO's and RSC's own Amfleet and AmCafe cars showed the same results during all three or four way meets. During meets with traffic with an empty track in between, no derailments were encountered.

[Kali]

Okay, I don't think there's anything we can do about that then... more creaky core code.

Well maybe the superelevation is making hitboxes collide, that's the only thing I can think that we can change.

[MikeK]

I have experienced this many times on many different routes. Trains will just derail for reasons that wouldn't apply in the real world.

Were you using dynamic brakes on one of the trains that derailed? I have made a tiny 20 car train derail on straight track just by using 20% dynamic brakes.

As Kali says, there is nothing much we can do about it, that is the current state of the game.

[TVRRMAN]

I have experienced this many times on many different routes. Trains will just derail for reasons that wouldn't apply in the real world.

Were you using dynamic brakes on one of the trains that derailed? I have made a tiny 20 car train derail on straight track just by using 20% dynamic brakes.

As Kali says, there is nothing much we can do about it, that is the current state of the game.

Mike,

As far as I can tell, the dynamics were not in use. The scenario this test was being conducted in was a cloned and modified version of the Horseshoe Curve scenario that has you railfanning up at Horseshoe Curve. All trains were under AI control with me only doing the watching. I was only alerted to the derailments when the scenario abruptly ended. Should the units be in dynamic mode though, I suspect it should have been decipherable.

I believe the same problems also occur with the RSC Amfleet and Amcafe cars also, as I had the same derailment happen but without anything beside or near the cars. That consist was a six car cut powered by a P42DC, and made up of four RSC Amfleet Coaches, Amfleet Cafe, and Another Amfleet passing as the Business class car.

[GaryG]

The dynamics on many locos are set with the dynamics giving far too much braking. I've found that for those, keeping in the 1%-12% range gives relatively realistic braking effort. As you're seeing, 20% on some locos is enough to actually stop a train without applying train brakes - enough force being applied from the front to do exactly what you're seeing.

GaryG

[Kali]

I find the dynamics on the F7s unusuably powerful, especially given it's only the lead engine... given I use maybe 5% at most, a 20x reduction and fixing to pass down the consist would still probably be a bit much :p something to try on a rainy afternoon anyway.

[Kali]

http://www.youtube.com/watch?v=YE2JNFHyPJg

I do have to wonder if that sort of thing is editor only, or if there really is collision detection out to that sort of range when the game is running.

[ryandevl]

Not related to the consist stretching or derailments, but...

I was attempting the "people power" scenario on HSC and all was well until right after Cresson station. A Stop signal blocks the path onward, and doesn't change. After 10 minutes I became impatient and left, because after looking at the 2D map, I could see that the following signal after that one is marked Clear and there is no AI traffic or any obstructions that should cause this Stop signal. Anyone else encounter this problem?

[Kali]

RW has always had some signalling issues, although the latest one seems to have a lot more - try hitting tab next time, it'll let you pass a red in a permissive manner.