[copperpen]

http://www.trainsim....on-Model-Issues says it all from one persons point of view. It is not clear whether this refers to steam, or other traction forms.

[Z31SPL]

Alright so I see many topics about Advanced adhesion model and wheelslip. Not sure if this is common knowledge or if we are at this point yet in development, but has there been any group consensus as to what settings in the options result in a fairly realistic simulation of wheelslip and train handling?

[disc]

Advanced adhesion simulation has problems, but the developer of that code, left the project more than one year ago, and it seems noone understands the math in axle.cs.

[steamer_ctn]

As a result of seeing a number of posts expressing concerns about the accuracy of the operation of the advanced adhesion model, I am proposing to investigate, and fix if possible some of the more obvious errors. Some of the issues may include those raised by t_rail in the link contained in the first post of this thread.

As many have already indicated (in other threads) the original developer of the adhesion model is no longer actively involved, so I am not sure how far this work will progress. I am expecting it to be a slow process as the logic of the model needs to be understood before code changes can be undertaken.

In order to ensure that there is a good benchmark to test OR code changes against, a test model based upon the BR Class D16/2 has been set up. This locomotive has been selected as an extensive test report was produced by BR, and currently a copy of this is available online.

Information about the locomotive, as well as a link to the test model can be found on this page.

The model has four versions of the locomotive to allow for comparison:
i) MSTS - Orig - standard MSTS ENG file supplied with the locomotive
ii) MSTS - Mod - small changes to the above file to allow operation with the test stock.
iii) OR - Basic - a basic ENG file that any modeller hopefully could set up based upon readily available information
iv) OR - Advanced - an ENG file using "advanced" features that requires detailed information that may not be readily available without significant research

Throughout the testing process it is contemplated that the test model will need to be adjusted as testing brings to light the need for any changes.

The initial testing will probably focus on the "Load Hauling Performance" of the above OR models in simple and advanced adhesion mode. The report contains information of the balancing speeds for a set train load on different gradients.

[longiron]

Team,

I've reached out to Matej directly with a request to assist in helping OR developers understand his advanced traction model. Let's see if he responds.

chris

[disc]

On another forum a guy told that the instability of the advanced adhesion model (wheelslip value oscillation) is caused the wheelslip change values are not multiplied by the frametime. My tests with Monogame Openrails confirms this, as in Monogame version the fps is higher and much more stable, so the wheelslip values are more stable too.
Also with stable FPS the power loss (engine force vs axle out force) was much less, so that's probably caused by the same problem but hidden by the filtered axle out force.

[steamer_ctn]

I have started to make some changes to the advanced adhesion model to try and correct some of the stability problems reported by various people. This instability appears to manifest itself as reductions in tractive force.

The first series of changes is in #3556, which includes some minor code syntax corrections, and adjustment to the friction calculation code for fog and precipitation.

As indicated above (in post #4) I have created a test locomotive whose performance can be confirmed against a BR test report. (See the link in that post for the locomotive, wagons, etc).

To test the difference in the code changes I have run tests with this test locomotive (OR-Basic) up the 1 in 80 gradient on the CTN test route. According to the BR test report, a 300 ton load should achieve a balancing speed of approximately 37.5mph on this slope for this locomotive.

Tests can be done running the above test in a pre-#3556 version, and #3556 to see the difference.

NB: the "Adhesion proportional to rain/snow/fog" must be selected.

It is also recommended for testing purposes that the following settings are used:
i) Adhesion factor correction - 100%
ii) Adhesion factor random change - 10%

[ATW]

I see now that the num wheels value in the engine section plays a partial role now in ORTS. I have notice a possible anti slip control for the whole consist if a leader unit is equipped with a value of 0 in the engine section of NumWheels where no units slip nor is there any F5 HUD Force Info showing but units are sure pulling especially the leader at 100% power in Run 8 starting a heavy train.

Still more tests an code work but I like the start of this progress.

[disc]

ATW, on 09 June 2016 - 08:37 PM, said:

I see now that the num wheels value in the engine section plays a partial role now in ORTS. I have notice a possible anti slip control for the whole consist if a leader unit is equipped with a value of 0 in the engine section of NumWheels where no units slip nor is there any F5 HUD Force Info showing but units are sure pulling especially the leader at 100% power in Run 8 starting a heavy train.

Still more tests an code work but I like the start of this progress.

I don't see any change with anti slip, and numwheels, nor in game, nor in the changed code.

[ATW]

Yes there is not really much any notice changes... it's just I set both sections of my NumWheels on a leader to 0 an saw a change where I even don't include Antislip lines.