I managed to get the PRL&P boxcar through a tight curve

[dvdhoven]

Hi,

Last weeks I am investigating why cars don't want to go through tight curves. Most European tramway cars. Sometimes successfull and sometimes not.

But lately I was testing the very nice boxmotor OWP 103 - very much thanks Tim for this and other cars; of course thanks to all others for the great Traction stuff - and built a consist of this motor and 2 PRL&P boxes and a PRL&P caboose.
I started on the Mesaba route at Gilbert. There is a very tight curve and I couldn't get the boxes through the curve.
Well I have experimented some time and come with these adjustments to the wag file and now I get them through that curve:

InertiaTensor ( Box (3.366m 4.172m 11.5m) )

I changed that 11.5 from the original value 10.72 because I had a very strange effect. After changing the r0 from the coupler to 0.4, I get through the curve but then the cars were close to each other. To keep some distance, I adjusted the InertiaTensor.

This are the new values for the Coupling:
Coupling (
Type ( Automatic )
Spring (
Stiffness ( 2e7N/m 2e7N/m )
Damping ( 0.2e7N/m/s 0.2e7N/m/s )
Break ( 5.1e7N 5.1e7N )
r0 ( 0.4 0.4 )
)

I think that only the r0 values are really important.
Here are the original values:
Coupling (
Type ( Automatic )
Spring (
Stiffness ( 5e6N/m 0 )
Damping ( 1e6N/m/s 0)
Break ( 2e7N 0 )
r0 ( 15cm 1e9 )
)

The break values are that high because one time, I got stuck in curve and with a bang the coupler broke and only the motor run away.

Dick van den Hoven
http://forums.flightsim.com/dcforum/User_files/3c3739ad2f412aac.gif

[tryagin]

Hi Dick,

Glad you like the OWP Boxmotor! (I call 'er "Old Maud"). I was reading your post and have some numbers for you. I found these to work quite satisfacorily for the tight tram curves, like XTracks 30degree points and curves:
Type ( Automatic )
Spring (
Stiffness ( 2e7N/m 0 )
Damping ( 0.2e7N/m/s 0 )
Break ( 1e7N 0 )
r0 ( 0.1 1e9 )
)
CouplingHasRigidConnection ()
)
Buffers (
Spring (
Stiffness ( 1e7N/m 0 )
Damping ( 1e6N/m/s 0 )
r0 ( 0m 1e9 )
)
Centre ( 0.5 )
Radius ( 1 )
Angle ( 0.5deg )
)
I made an interurban flatcar that wouldn't go through those curves, and noticed how easily 2 or3 of Matt Austin's Milwaukee 1100's in a train negotiated these curves. So, I copied the values from his 1100 eng file over to the C&LE flatcar wag file- and 'voila' - it works! So, now I make sure any car that will be going through city streets will have these values.
I hope this'll help get you through any tight spots in future! Glad you like the traction stuff! There's more fun on the way!

Cheers!

Tim

[dvdhoven]

Hi Tim,

Just that r0(0.1 1e9) wasn't working!
I started with the 1100 values, but they don't work for this particular car and on my machine.
So I changed it to 0.5. After that the cars had some gap between the couplers, but they go through the curve. But then, big suprise! The cars slammed to each other after the curve without any space.
So I altered the InertiaTensor and that worked fine, after the going through the curve no slamming.
Then I changed the r0 to 0.4 and the couplers had no space.

My personal experiences are that sometimes you get something working and after a day, you start MSTS and the whole thing isn't working! Frustrating!

My personal thoughts for these subjects is the following:
What is involved:
- The Bounding box. I have tried a bounding box that includes the couplers, but now I see that a bounding box without couplers is working fine.
And I noticed that Bounding Boxes that overlap in curves, causes the troubles that the cars won't run.
- The r0. Decreasing this value, gives initially more gap between the cars
- The InertiaTensor. The spacing between the cars, can be influenced by this parameter, both in starting position and on the run.

And some other point,
I noticed that the trucks on that flat car are really on the end of the car, so less overhang of the car. Your box car have more overhang and so the coupler must be longer to get through the curve. And the coupler can be made longer by the r0 variable.


Dick van den Hoven
http://forums.flightsim.com/dcforum/User_files/3c3739ad2f412aac.gif

[tryagin]

Hi Dick,

Very interesting data you have here and worth a serious consideration. Thank you for it! I must admit, these numbers tend to fry my brain, and any light shed on what they mean is greatly appreciated.
You're so right about the fickleness of MSTS. See my comment in "CA&E Update", reply 11. yah, fickle as a weather-vane!

Thanks,

Tim