[steamer_ctn]

Resistive Forces

The resistive force is made up of all the resistances impacting on the train, and these include, rolling resistance (on level) or starting resistance if train is stationary, grade resistance, curve resistance, wind resistance, tunnel resistance, etc. These values are all summed together to calculate the total train resistive force.

The grade resistance is by far the dominating factor in the total resistance forces curve. For example, in the preliminary testing, the grade resistance for a car on the 1.38% grade is approximately 4000lbf, whereas the rolling resistance at the train speeds indicated in the NS document are approximately 300lbf. This is the reason why railways spent so much time and effort in determining the tonnage (load) ratings for different locomotives in different track sections. Too much load could slow the train down, or even worse cause it to stall on a particular grade. Either of these events would have major performance impacts on the railway, and ultimately cost the railway money in delays.

When looking at particular track sections a railway company considers the following gradients in each track section:

i) Ruling gradient - this is the steepest gradient in the section, and the train would need to climb this grade at desired speed determined by timetable requirements. Based on this gradient and the speed required, a tonnage load would be calculated for this section.

ii) Helper (Bank) gradients - these were steep, often short sections of steep gradient, where it was deemed more economical to add a helper locomotive to assist the train up this grade rather then reducing the tonnage load in the section just for this grade.

iii) Momentum gradients - in more recent times railways are wanting to increase their efficiencies of operation, and hence are keen to increase the tonnage ratings. As a consequence they are now reviewing the ruling gradients to see whether this can be reduced in a section by using momentum. In other words, if a train can get a "run up" at a gradient then its momentum will carry it up the gradient. This means that larger tonnages can be assigned, however it is then important to make sure that a train is not prevented from using its momentum. For example trains would not be able to be stopped at certain locations as they may not get going again, or alternatively may not be able to climb a particular gradient.

Of interest it can be seen by studying the NS document, a train in dry weather conditions can haul a load of approximately 9300 tons up the 1.38% gradient, whereas in wet conditions the same locomotives can only haul a load of approximately 7200 tons up the same gradient. (The reduction in loading is due to reduced track adhesion of wet track)

Thus to create an "all weather" tonnage rating, NS selected the lower value, and these values would be published in employee timetables as appropriate. For those interested in seeing an example of railway tonnage ratings refer to this page - see Lake Div ETT. Snow (icy weather conditions) have different impacts yet again on the train.

For testing purposes the FCalc utility has been used to set all the locomotive and wagon rolling resistance values.

Initial testing will consider the following cases (some refinements in testing may occur during testing) -
i) Dry weather - tonnage 9500 tons - balancing speed = 9.0 mph

ii) Dry weather - tonnage 7200 tons - balancing speed = ? mph (somewhat faster)

iii) Wet weather - tonnage 9500 tons - balancing speed = N/A

iv) Wet weather - tonnage 7200 tons - balancing speed = 10.7 mph

[steamer_ctn]

Tractive Forces

As described above the remaining balancing force to be considered is the Tractive Force which is produced by the locomotive.

In order to ensure the best accuracy for the test, the SD70ACe locomotive was selected as a datasheet published by a locomotive manufacturer was available. Wherever possible manufacturers data should always be used as the ideal preference over general data quoted on the INTERNET. When researching for this project it was noted that various data sources were available, however there was a difference in the data between some INTERNET sites, so the question became, which data does one select.

For the purposes of this test a BASIC ENG file configuration was used, however an ADVANCED configuration using the tractive force curve based upon the manufacturers could have also been used.

In setting the locomotive accurately the following key parameters need to be considered:

i) Force - the starting (MaxForce) value and the maximum continuous force, and the accompanying speed, are important values as they provide accurate reference points for the tractive force curve. The values in these ENG file parameters should be populated with the relevant details from the datasheet - MaxForce, MaxContinuousForce, ORTSSpeedOfMaxContinuousForce. These force values represent the force applied to the wheels of the locomotive.

ii) Power - Diesel locomotives were considered to be able to produce a constant amount of power between the speed of maximum continuous force, and the maximum speed of the locomotive. OR uses this power value to "cap" the tractive force value to a maximum at the speed of the locomotive. Hence we need to calculate an "apparent" power value from the information above. If this value is too high, then it will make the locomotive to appear overpowered, similarly if it is too low then the locomotive may appear under powered. Again this represents the power applied to the wheels of the locomotive. Hence MaxPower needs to be configured correctly.

iii) Adhesion - In broad terms adhesion places a limit on the amount of force that can be applied to the wheels. Too low an adhesion force will cause the locomotive to slip. Adhesion force is determined by the weight on the locomotive wheels, and the coefficient of adhesion based upon the track conditions, ie wet, dry, icy, contaminated by leaves, etc. Sometimes the manufacturers data provides an adhesion reference point linked to one of the above force values ( in i) ) above. Typically DC locomotives tended to have adhesion values below 0.33 for dry rail conditions, whereas AC locomotives with more modern slip control software can have values greater then 0.4.

To correctly set the adhesion it is necessary to set the following parameter - ORTSDriveWheelWeight. In addition this it is necessary to calculate the "apparent" adhesion by dividing the tractive force by the adhesive weight (ie the value placed in the ORTSDriveWheelWeight parameter). Once this value has been calculated, and if it is greater then o.33, then the Curtius-Kniffler (C-K) equation can be adjusted to achieve the correct value of adhesion for the locomotive. Adjust the C value in the equation until the adhesion coefficient is obtained.

Where no adhesion value is quoted in the manufacturer's data, it can be assumed that the starting force value can be used to calculate the coefficient of adhesion, and the C-K curve adjusted to this value at 0 mph.

Preliminary testing has shown the following:

i) So far some minor changes have been necessary to the tractive force calculation in OR (these changes should be present in the current unstable version of OR)

ii) That reasonably close results against the NS test document can be achieved

It is hopped to share the ENG and WAG file configurations set up as described in the above notes for reference, and further test outcomes.

[R H Steele]

Thank you for the "Resistive Forces" and "Tractive Forces" posts...most helpful.
Just so I understand where these two are placed in the file?
ORTSSpeedOfMaxContinuousForce --- eng section of ENG file?
ORTSDriveWheelWeight -- wag section of ENG file? Have either of these terms have been published in the manual?

[steamer_ctn]

Both of these values go in the engine section of the file, as suggested on this page.

This information is not in the manual, as all these changes are still in the Unstable (and MG I believe) version, and have not been released into the Testing and Stable versions yet.

[R H Steele]

Thank you, I had not checked your pages, I see it now.
Regards, Gerry

[Lamplighter]

steamer_ctn, on 02 February 2020 - 06:40 PM, said:

To correctly set the adhesion it is necessary to set the following parameter - ORTSDriveWheelWeight. In addition this it is necessary to calculate the "apparent" adhesion by dividing the tractive force by the adhesive weight (ie the value placed in the ORTSDriveWheelWeight parameter). Once this value has been calculated, and if it is greater then o.33, then the Curtius-Kniffler (C-K) equation can be adjusted to achieve the correct value of adhesion for the locomotive. Adjust the C value in the equation until the adhesion coefficient is obtained.

Is it really necessary to interfere with the original constants of the Curtius-Kniffler formula and change the value of C (0.161)? The parameter ORTSCurtiusKniffler (A B C D) currently used by OR has a correction value of D. See. Open Rail Manual (8.1.3 Adhesion of Locomotives – Settings Within the Wagon Section of ENG files), but also here

[steamer_ctn]

Lamplighter, on 03 February 2020 - 04:41 AM, said:

Is it really necessary to interfere with the original constants of the Curtius-Kniffler formula and change the value of C (0.161)? The parameter ORTSCurtiusKniffler (A B C D) currently used by OR has a correction value of D. See. Open Rail Manual (8.1.3 Adhesion of Locomotives – Settings Within the Wagon Section of ENG files), but also here

Thanks for pointing this out, looking at some comments in the code, it doesn't appear to be a "direct" multiplier for the adhesion coefficient (as I was suggesting was required), but rather as a means to improve the stability of the advanced adhesion model.

I will try and dig into this a bit further.

[ErickC]

edwardk, on 30 December 2019 - 03:28 PM, said:

I have to wonder if this work should have been released under a different release category. We have to remember that when OR was just released, maxPower would have reflected the hp minus the efficiency rating. People would have started updating maxPower with the full hp once they discovered OR took care of this in the code. Now, people are expected to go through another round of hp changes. There are probably issues that still exist with the physics that need to be found and fixed.

One question is this. If realistic parameters are used, how close is the train operation to real life operation?

The problem goes deeper than that. The changes were made without asking the most important question: "how are the content developers actually building things?" This would have led to the natural first step of stopping by the file library at t-s or asking developers for samples, then examining and observing. This is important because a major false assumption was made: that locomotives either use only the basic parameters, or they use only the advanced parameters. A cursory examination of currently-available content makes it clear that this is not the case.

When you begin to examine content that takes advantage of the ORTS diesel block, you'll find that little of it actually uses the tractive effort curves for the simple fact that our educated guesses about proprietary data that we don't have access to aren't going to be any better than the curves used by OR's basic model. The obvious solution is a hybrid approach that uses the new parameters to control the engine, which fixes a lot of sound, fuel flow, and power problems, and lets OR handle the power to the rail. This has so far produced results that are pretty accurate for pre-microprocessor locomotives - by "pretty accurate," I mean "closer than any of the other sims can do without serious scripting." I am not the only developer using this approach, and I suspect that it's for similar reasons across the board.

I mean, what am I going to do, hire someone to hack EMD's corporate headquarters and hope they still have data from locomotives built 70 years ago?

We never really thought that it would mean that our locomotives would become immobile, which is what has happened, because it didn't figure in our minds that one of the improvements to the diesel physics model would be the imposition of a "basic or advanced" dichotomy between the basic and advanced parameters, which is what has happened here. You could argue that developers can issue patches, and, while that's certainly true, it's kind of unrealistic. Think about all of the files that would need to be updated in all of the places where they exist - sure, I can issue a patch for my include files over at t-s, but then I have to also send Ian updates for every single locomotive pack that's up for sale. And that's just me - TS has a much bigger back catalog than I do. This is completely separate of the fact that users aren't going to want to download patches for every single addon they have on-hand.

I think that the new parameters are extremely useful when the data can be found. I think that a lot of it is available for newer GE locomotives, for example. But it might be wise, long-term, to substitute the basic tractive effort curves any time the OR ones are missing, whether the locomotive uses the basic engine parameters or the ORTS diesel block.

[steamer_ctn]

ErickC, on 04 February 2020 - 06:20 AM, said:

The problem goes deeper than that. The changes were made without asking the most important question: "how are the content developers actually building things?" This would have led to the natural first step of stopping by the file library at t-s or asking developers for samples, then examining and observing. This is important because a major false assumption was made: that locomotives either use only the basic parameters, or they use only the advanced parameters. A cursory examination of currently-available content makes it clear that this is not the case.
...................................

I don't quite understand the point.

Currently there are two independent ways of configuring the tractive force curve of a diesel locomotive.

In the first method, the force and power parameters can be set, and OR will calculate a tractive force curve from these values. The MaxPower value will cap the maximum continuous force calculated, so if this is set high the locomotive will be over-powered, if it is set low then it will be under-powered. As far as I am aware OR has used (and is still using) this methodology since it was first released.

The second method is to "plot" the tractive force curves in the ENG file, and in this case these curves will override the internal calculation by OR of the tractive force and use the values specified in the curves.

ErickC, on 04 February 2020 - 06:20 AM, said:

I think that the new parameters are extremely useful when the data can be found. I think that a lot of it is available for newer GE locomotives, for example. But it might be wise, long-term, to substitute the basic tractive effort curves any time the OR ones are missing, whether the locomotive uses the basic engine parameters or the ORTS diesel block

This should currently be the case as far as I am aware.

Do you have a situation where this is not the case?

[ErickC]

Absolutely - anyone running locomotives made by either me or Tyler Bundy is going to find them completely immobile after updating to any version of OR that uses the new diesel model. What we both did was use all of the OR diesel parameters but the tractive effort curves. In previous builds of OR, the sim subtracted about 20% of the total output as specified in the diesel power tab, and applied the rest to the rail using the basic parameters. What seems to be happening now is that next to no power is being applied to the rail at all. Both of us still have both the old MSTS MaxPower entries plus the MaximalPower entry in the ORTS diesel block, as well as diesel power tabs. We both lack the new ORTSDieselEngineMaxPower parameter, since our physics files predate it. What we are also both lacking is the tractive effort curve (I can't speak for Tyler, but I don't have access to the right kind of data), which I suspect is what the current diesel model is not happy about. It's entirely possible that either or both of us borked something else. For my part, I'm perfectly happy to send sample locomotives along if it might be helpful.