Simple Fixes to Improve Physics (not necessary for gameplay)

[NDORFN]

Sounds good. Several games have a "realism" or "difficulty" level, or user-selectable realism options.
Submarine simulators in the Silent Hunter series, for example.

This doesn't change physics calculations, it changes parameters.

For example, you could have a setting for "realistic brakes" or "strong brakes". In the latter case, you would just get amazing brake shoes with a very high friction coefficient yet no overheating.

Another example would be "realistic brake recharge rate" or "fast brake recharge rate". You just get a better compressor.

All these idea's could be achieved without the involvement of RSC with a couple of simple batch files.

[Chacal]

Looks like we're getting somewhere.
I suppose the batch file would have to apply the same parameters to all of a player's engines.

[NDORFN]

Yeah. There'd be a bit of work involved and we'd have to own all available engines between us, but I'm pretty sure we do anyway.

[Rich_S]

At first I was trying to update the physics of North American diesel locomotives and Mike even added my automatic air brake fix to his RW_Tools. I was also working on a quick fix for the dynamics, were you simply divided the default RSC value by 10 and used that number as your starting point for Maxforce along with setting the CylinderPressureForBrakeToCutPower value to zero. The biggest problem is, every time RSC updates the game, your work gets overwritten. Yes, you can restore from backups, but what if the RSC update contained a crucial fix for a known bug? I believe the only real solution is to have RSC update it's files with prototype values. Seriously, what is the point of having the best looking train game in the world, if I can coast down Horseshoe with 80 loaded hoppers and never have to touch the brakes, or I can pull 100 loaded hoppers up Cajon in Notch 3?

At this point I'd suggest all those people working on physics updates, start working together as a team. Once you as a group have identified all problems that exists and solutions to the problems contact RSC and see if they are willing to work with you updating the physics for North American locomotives and rolling stock. If they are not, then the only other option would be to work with Mike Simpson and see if he could add routines to RW_tools that would update the various files, much like the routines that were added to Route_Riter to update MSTS physics. The only downside to this second method is, those routines would have to be run after each RSC update

Regards,
Rich S.

[Chacal]

Rich, we already have to reapply several mods after an update or even a cache verify in Steam. This is easily automated, Machinist has postedhis batch file and I have my own. Reapplying all mods is now only a matter of double-clicking an icon and waiting 20 seconds.
This is why the idea of a batch file for physics is great: it is easy to automate.

You are correct about the danger of overwriting crucial RSC updates. But it is relatively easy to make changes inside blueprint files, which is what RW Tools does. The problem with RW Tools is that it is not automated (I am not aware of command-line parameters for it) , you have to do it manually every time. Worse yet, you have to remember how to do it every time. In the past I have made batch files using VBScript that apply modifications inside xml files for a different game, so I know it is doable.

In the end you are quite right about the need to ask RSC to apply those changes to its rolling stock. In theory it is a simple problem: fix the core physics calculations in the game engine, then fix the rolling stock parameters in blueprints. But we all know they have their own priorities and plans and we should not wait for that to happen.

[MontanaRails]

I agree with what Chacal and Rich_S have said.

On the topic of Donner Pass track, I think I've improved the ride quality a little more.

1) What it does: The grade transitions are pretty nice now, but all the curves with super elevation are still kinda rough going in/coming out. Make these changes (back up the old file first!) and see if it improves the ride any - it seems to more gradually introduce the superelevation into curves, instead of it being abrupt:

File Location: C:\Program Files (x86)\Steam\steamapps\common\railworks\Assets\RSC\DonnerPass\RailNetwork\Track

File Name: DonnerPass.bin

What to Change: Make sure MaxCantAngleDegrees reads "2", Change CurveToAnglePercent to read "5", Change LineUnevenness to read "25" (optional, just makes the track a little rougher, more like jointed rail). Also, about halfway down the file you'll find a section called SuperElevation - I cant figure out what this section does, as the input value seems to make no difference. Its default value is "150" To be in line with all the other routes with Super Elevation, I set this to read "0".
--------------------------------------------
Next, it seems like all of the Donner pass Locos had their Cab Camera latency settings stiffened a lot (probably in an effort to downplay the rough superelevation transitions). Remember to back up the files (RW_Tools will automatically back up).

2) What it does: In order to get more cab sway/head sway, make these changes to every Donner Pass locomotive's Cab View

Sample File Location: C:\Program Files (x86)\Steam\steamapps\common\railworks\Assets\RSC\DonnerPass\RailVehicles\Diesel\SD40-2\Default\Cabview

Sample File Name: SD40-2 Cab camera.bin

What to Change: Make sure the bottom section reads as follows:

Code: Select all

               <verticalDirectionalFreedom d:type="sFloat32" d:alt_encoding="0000000000000040" d:precision="string">2</verticalDirectionalFreedom>
               <horizontalDirectionalFreedom d:type="sFloat32" d:alt_encoding="0000000000000840" d:precision="string">3</horizontalDirectionalFreedom>
               <SpringXCoefficient d:type="sFloat32" d:alt_encoding="000000A09999B93F" d:precision="string">0.1</SpringXCoefficient>
               <SpringYCoefficient d:type="sFloat32" d:alt_encoding="000000E051B8BE3F" d:precision="string">0.12</SpringYCoefficient>
               <SpringZCoefficient d:type="sFloat32" d:alt_encoding="000000E051B8AE3F" d:precision="string">0.06</SpringZCoefficient>
               <DampingXCoefficient d:type="sFloat32" d:alt_encoding="00000080976E923F" d:precision="string">0.018</DampingXCoefficient>
               <DampingYCoefficient d:type="sFloat32" d:alt_encoding="000000E051B88E3F" d:precision="string">0.015</DampingYCoefficient>
               <DampingZCoefficient d:type="sFloat32" d:alt_encoding="00000040E17A843F" d:precision="string">0.01</DampingZCoefficient>

I've run several default scenarios after these changes, and they all seem to work as intended. If the consensus is that these are good additions, I'll send them to RSC (so they hopefully make it into a patch).

[dfcfu342]

Just changing to prototype values is not a solution based on my adjusting of values. Either the physics equations listed in the Wiki are no longer used or the game engine does some strange rounding. I found adding prototype weights to some covered hoppers caused them to fall off the tracks as soon as the scenario started it was quite comical to watch actually.

Friction values also cannot be altered much either. I found prototypical values caused an overflow in the calculations which caused "inverse drag" that accelerated the train rather than slow it uphill and downhill it would slow the train causing the need to use power downhill.

I'm almost satisfied with the adjustments I've made to values but they are only prototypical for loaded cars. Emtpy cars will behave as if they're HO scale models but I prefer to run my trains heavy and loaded so that's the way they're set up

[NDORFN]

How do you guys decide what's realistic? Do you try some value then make a comparison to a video or something?

[dfcfu342]

How do you guys decide what's realistic? Do you try some value then make a comparison to a video or something?

Cab ride videos are a great source for getting a "feel" for what it should be like and Al Krug's site is great for back ground information and explanation of the theory behind railroading physics.

http://www.alkrug.vcn.com/home.html

[NDORFN]

Wow the train forces calculator is interesting... I'm gonna put it to the test.

[Rich_S]

How do you guys decide what's realistic? Do you try some value then make a comparison to a video or something?

I can't speak for others, but my modifications come from my personal experience at work.