[edwardk]

steamer_ctn, on 27 October 2016 - 03:25 PM, said:

Does this mean that the Mass() statement in the ENG file was not reading the correct locomotive weight?

What was the Mass() statement in the ENG file?

In the ES44DC, the Mass is 186.8t. Can your advanced adhesion process use metric ton? All I know is that I converted the metric ton to lbs before using it in ORTSDriveWheelWeight(411894). What you see is the initial metric ton value converted to lbs. Once read in, its converted to kilograms.

Edward K.

[steamer_ctn]

edwardk, on 27 October 2016 - 05:22 PM, said:

Can your advanced adhesion process use metric ton?

It will use any valid mass units (internally within OR it will use kg).

edwardk, on 27 October 2016 - 05:22 PM, said:

In the ES44DC, the Mass is 186.8t.

Because the units "t" is included in the parameter, the 186tonnes will be converted to kg for OR use.

edwardk, on 27 October 2016 - 05:22 PM, said:

All I know is that I converted the metric ton to lbs before using it in ORTSDriveWheelWeight(411894). What you see is the initial metric ton value converted to lbs. Once read in, its converted to kilograms.

As there is no UoM indicated in this parameter, it will be read in directly as kg (assuming that this is the way that the parameter statement appears in the ENG file). Thus if this value is the locomotive weight in pounds, then the weight has effectively been increased by 2.2 times to 411894 kg. Naturally doubling the adhesive weight of the locomotive will significantly improve the slip capability of the locomotive (but may impact the realism).

The parameter needs the units to be specified, ie ORTSDriveWheelWeight(411894.0lb)

[edwardk]

ORTSDriveWheelWeight(411894.0lb)? What I initially used here is almost correct then? What about the extreme wheel slipping when on a 1.3%+ grade when this parameter is not used? The initial Mass() is used. I realize there is wheel slippage, but what I experienced would not be considered normal. Is it possible the adhesion value you are using is too low? I believe you using 0.4f for snow. 0.8f is being used for simple adhesion.

If you have one of SLI's Scenic Sub routes, try running it in the middle. At this moment, I am unable to come with an area for you to test.

Edward K.

[copperpen]

This is what the au_ctn test route is for. It gives a common ground that everyone can use to test and verify results. It can be found at Peter's site, coalstonewcastle. Just click the link in his post and search for the route.

[steamer_ctn]

edwardk, on 28 October 2016 - 02:55 AM, said:

ORTSDriveWheelWeight(411894.0lb)? What I initially used here is almost correct then?

If the ORTSDriveWheelWeight is missing from the ENG file then the full weight of the locomotive will be used as the adhesive weight. In some instances, say for locomotives with non-driven wheels, the adhesive weight will actually be less then the full weight of the locomotive. Thus it will provide a "best case" slip performance.

edwardk, on 28 October 2016 - 02:55 AM, said:

What about the extreme wheel slipping when on a 1.3%+ grade when this parameter is not used? The initial Mass() is used. I realize there is wheel slippage, but what I experienced would not be considered normal. Is it possible the adhesion value you are using is too low? I believe you using 0.4f for snow. 0.8f is being used for simple adhesion.

What is normal?

The only information provided so far is a grade and the fact that it is a diesel locomotive hauling the load (ES44DC).

What is the load that is being hauled? Often the load needed to be derated for train operation in ice and snow conditions. What is the expected load that the ES44DC can haul up a 1.3% grade? Is this for dry or icy conditions? Has this information come from a working timetable or railway operational manual?

edwardk, on 28 October 2016 - 02:55 AM, said:

If you have one of SLI's Scenic Sub routes, try running it in the middle. At this moment, I am unable to come with an area for you to test.

As suggested by Copperpen I prefer to utilise the test route on the Coals to Newcastle site as this has a number of consistent gradients of sufficient length to give a good indication of the performance. It is also freely available for all users, and this thus eliminates any issues with users obtaining a copy, which can be an issue if the stock and route is payware, or difficult to obtain.

My other preference is to ensure that all the rolling stock has been "optimised" for operation in OR to ensure consistent performance. I have used this approach for testing some of the recent changes that I did to the adhesion model. See this page for info demonstrating this approach, and the "standard" model that I used. I have set the rolling stock up to provide a consistent performance in OR.

If you wish to explore this further, please set up a test scenario that we can all use to confirm the agreed expected performance.

[edwardk]

steamer_ctn, on 28 October 2016 - 11:52 PM, said:

If the ORTSDriveWheelWeight is missing from the ENG file then the full weight of the locomotive will be used as the adhesive weight. In some instances, say for locomotives with non-driven wheels, the adhesive weight will actually be less then the full weight of the locomotive. Thus it will provide a "best case" slip performance.


What is normal?

The only information provided so far is a grade and the fact that it is a diesel locomotive hauling the load (ES44DC).

What is the load that is being hauled? Often the load needed to be derated for train operation in ice and snow conditions. What is the expected load that the ES44DC can haul up a 1.3% grade? Is this for dry or icy conditions? Has this information come from a working timetable or railway operational manual?


As suggested by Copperpen I prefer to utilise the test route on the Coals to Newcastle site as this has a number of consistent gradients of sufficient length to give a good indication of the performance. It is also freely available for all users, and this thus eliminates any issues with users obtaining a copy, which can be an issue if the stock and route is payware, or difficult to obtain.

My other preference is to ensure that all the rolling stock has been "optimised" for operation in OR to ensure consistent performance. I have used this approach for testing some of the recent changes that I did to the adhesion model. See this page for info demonstrating this approach, and the "standard" model that I used. I have set the rolling stock up to provide a consistent performance in OR.

If you wish to explore this further, please set up a test scenario that we can all use to confirm the agreed expected performance.

If you have Scenic Sub 2, run the What Goes Up activity. The only part of the activity is that you must wait more than a few minutes for the approaching AI train. As far as optimizing my trains, I have been doing this already so its not that. If it does turn out to be a train setting issue, I will let you know.

Edward K.

[steamer_ctn]

edwardk, on 29 October 2016 - 03:57 PM, said:

If you have Scenic Sub 2, run the What Goes Up activity. The only part of the activity is that you must wait more than a few minutes for the approaching AI train.

I do not have any payware content, so I can't run any tests using it.

edwardk, on 29 October 2016 - 03:57 PM, said:

If it does turn out to be a train setting issue, I will let you know.

Ok, I will wait to hear from you again if any further investigation is required.

[edwardk]

Peter,

I believe I found the pattern. I started off with 2 ES44DC locomotives and slowly started adding loaded coal cars until I started experiencing wheelslip. At this point, freight tonnage was 2398.6stn and the overall weight was 2810.4stn. I added one more locomotive which was the SD40-2. Keep in mind that adding the additional locomotive should have prevented wheelslip. It did not. At this point, freight tonnage is still 2398.6stn, but the overall weight did go up to 3005.5stn. It looks as if only the total weight is being calculated, leaving out the additional hp and tractive effort.

I extended the testing by replacing the SD40-2 with the same category locomotive and the tests were a bit more positive. I still experienced wheelslip if the throttle setting was too high. Keep in mind that I experienced this in an activity using 2 ES44DC locomotives and 2 SD40-2 locomotives. Although the SD40-2 is only rated at 3000hp and the tractive effort is lower, the combined hp and tractive effort should have helped, but did not. So the pattern appears to involve overall weight and type of locomotives used at the front, leaving out overall hp and tractive effort. If the adhesion parameters you are using for snow is close to real life, I have to wonder if this is a bug in your area or somewhere else.

Note: I never did test with a locomotive at the rear.

Edward K.

[edwardk]

Repeated first test with the SD40-2, but added SD40-2 on the rear. It looked as if this was going to work, but still experienced wheelslip. The only thing different is that sanding actually helped here where as before, it did not. The use of sanding prevented wheelslip. Any possibilities the adhesion value for snow a bit too low? Of course this is still not factoring in possible physics issues.

Addition Information: Not including this test, wheelslip was sporadic in the 1% grade range but was in full effect by time I reached 2.4% grade.

Edward K.

[steamer_ctn]

Some comments and thoughts:

edwardk, on 30 October 2016 - 12:33 PM, said:

I still experienced wheelslip if the throttle setting was too high.

This makes sense as the higher the throttle setting, the higher the tractive effort applied to the wheels, and thus the more change of slip occurring.

I agree that adding an extra locomotive should reduce the chances of slip as the tractive effort to start the train is now spread across a "greater" adhesive weight.

I am wondering if different locomotives have different slip thresholds, i.e. the throttle setting for the first locomotive may cause the helper locomotives to exceed their slip thresholds, and thus they could be slipping whilst the first one is not. (I would need to review the code to determine if the slip indication only comes off the first locomotive).

edwardk, on 30 October 2016 - 12:33 PM, said:

If the adhesion parameters you are using for snow is close to real life, I have to wonder if this is a bug in your area or somewhere else.

At this stage I believe that the adhesion parameters are close to the realistic values.

As you have indicated, I was not the original developer for the advanced adhesion model so I have only made some small changes so far as I am still not fully across the complete logic for it.

edwardk, on 30 October 2016 - 12:33 PM, said:

It looks as if only the total weight is being calculated, leaving out the additional hp and tractive effort.

Based on a quick test that I have done, I am wondering if this might be a pointer to the issue.

Can you do the following tests, and confirm the results?

i) Run a double headed light engine combination (2 locos only). When running them, watch the Motive (Force) column in the FORCE INFORMATION HUD.
My results indicate that both locomotives show a motive force against them, thus both loco are contributing to the pulling power of the train.

ii) Run a double headed train (2 or more locos) with a full load, and again watch the Motive force.
My results suggest that the second (or subsequent locos?) does not contribute any Motive force to pulling the train. This would explain the type of results that you are getting, as adding extra locomotive does not appear to increase the pulling power of the train.

Can you repeat these tests in v1.1 to see if it is a recently introduced problem, or has been in existance for a while?

Can you try the same tests with a steam locomotive.