[ErickC]

I was just thinking about this at work today. I was thinking about adhesion and tractive effort, and how locomotive wheel slip control systems work with the load regulator to maximize the amount of power that can be applied to wheels without slippage, or, with modern wheel slip control systems, with a precise amount of slip to maximize tractive effort. Of course, modelling either the load regulator or the wheel slip control system to any real degree of fidelity would necessitate modelling the other, i.e., you'd have to do both at the same time, and probably create a basic governor model as well. The benefit, of course, would be a more accurate representation of a locomotive, and probably the pruning of a whole lot of parameters that attempt to model the combined effects of these systems. A prime example is the tractive force curves. The maximum force that a locomotive can produce is purely a function of power, transmission efficiency, and speed, related in this formula:

TE = ( HP x transmission efficiency x 375 ) / ( Speed in MPH )

At low speeds, of course, this is much greater than the starting tractive effort, which is a function of weight and adhesion. Under these conditions, the wheels will slip, and wheel slip control systems unload the electrical system in order to maximize traction. This is what the TE curves seem to attempt to emulate.

Anyway, as far as locomotives in the US go, even first-generation diesels had simple wheel slip control systems, like IDAC, that worked by modulating the field excitation. Improvements were made over the years, and, of course, modern systems use ground radar and digital computers to maximize adhesion with wheel creep. One important thing to remember is that the wheel slip system type is completely independent of the traction motor type, and is often modular. For example, Super Series was used solely on DC locomotives, and both Mod3 and EM2000 were used on both AC and DC locomotives from the factory. These are all systems that use microprocessors and ground radar to control wheel creep and maximize adhesion. AC power has nothing to do with it. There are also retrofit excitation modules for the Dash-2 locomotives that replace the original WS10 module and add microprocessor traction control, meaning that these locomotives gain some of the benefits of the later systems. Obviously, as a direct module replacement, this system lacks ground radar and its added benefits.

You could probably model most control systems fairly accurately by adding a relatively small number of parameters, more or less as outlined in your original post:

( 0 ) Automatic sand only (even the earliest electrical anti-slip systems automatically add sand)
( 1 ) Anti-Slip Brakes
( 2 ) Basic electric anti-slip - this should model an early system like EMD's IDAC, which detects the change in resistance from the wheels spinning, automatically applies sand, and momentarily cuts power to all wheels without changing engine speed - we would need a parameter to control how fast the load can be dumped and re-applied
( 3 ) Microprocessor Control without radar - detects wheel slip conventionally but processes the load changes with microprocessors (think of the BOA-WS module that retrofits Dash-2 locomotives)
( 4 ) Microprocessor Control with radar - modern systems, such as Super Series, Mod3, or EM2000, that use a combination of microprocessors and ground radar to very precisely manage wheel slip for maximum adhesion

I am unclear on whether or not the microprocessor retrofit systems that lack ground radar have creep control, but the materials do suggest that they create an increase in traction above conventional systems, just not as much as the systems with ground radar - that is, a bump of 25% rather than 33%. Does the advanced adhesion model already take into account the increase in adhesion resulting from wheel creep? In that case, the microprocessor with radar system only needs to very precisely maintain wheel creep at 11% or so. Because it would give us a much more accurate representation of tractive effort, and would eliminate the tedious work of calculating the tractive force curves, I think it's a great idea. What would also be great is if we could get sound variables for wheel slip percent, load percent, and fuel flow percent, which would allow sound designers to more accurately model wheel creep or slip, traction motor volume, and exhaust volume.

[Weter]

What do I want to add:
1. The wheel-slip (or was it bad *eng-file)model, at versions of 2020 beginning, upset me, as I increased throttle on some EMU too fast, so WS started, and remained for a FEW SECONDS AFTER I turned controller off(I.E. completly cut off TM's current), so, Why?

2. I remembered variants of increasing "adhesion weight"
-to mechanical rise non-driving wheel(s) of steamer, or
-to magnetize driving wheels of electric loco(ЧС7)
As well, as automatic turning sanders on, when emerency brake has triggered.

3. I agree, that we had to choose more principial models, so there's no so much difference between some of variants in context of influence, or behavior at ORTSimulation process.
There are more lacks among other sides of train controls modelling, which are rather more unprototipic at present time.
However, all said here, certanly hAs sence too.

[darwins]

Wheel Slide Prevention - WSP

This film about the three step braking system gives a good explanation of how WSP functions on a modern electric multiple unit.

Three Step Brake

Some points from this are

Speedometer is measuring wheel rotation speed (not actual speed as it does in OR!) - so when the speedo descends towards zero when the wheels start to slide and rises as it spins again.

You also get to see the brake cylinder pressure needle moving - although the driver has selected step 3 which should give about 3 bar in the brake cylinder, you can see the pressure dropping to 2 bar or 1 bar and rising again.

In this example the WSP functions for both service braking and emergency braking. There is a WSP override that can be used in emergency braking only, but if used will result in a time-penalty before the brakes are released.

In some countries WSP operates only for service braking and not for emergency braking.

[Weter]

In OR it rises to 150, if wheelslip occurs.

[darwins]

 ErickC, on 21 June 2020 - 03:26 AM, said:

You could probably model most control systems fairly accurately by adding a relatively small number of parameters, more or less as outlined in your original post:

( 0 ) Automatic sand only (even the earliest electrical anti-slip systems automatically add sand)
( 1 ) Anti-Slip Brakes
( 2 ) Basic electric anti-slip - this should model an early system like EMD's IDAC, which detects the change in resistance from the wheels spinning, automatically applies sand, and momentarily cuts power to all wheels without changing engine speed - we would need a parameter to control how fast the load can be dumped and re-applied
( 3 ) Microprocessor Control without radar - detects wheel slip conventionally but processes the load changes with microprocessors (think of the BOA-WS module that retrofits Dash-2 locomotives)
( 4 ) Microprocessor Control with radar - modern systems, such as Super Series, Mod3, or EM2000, that use a combination of microprocessors and ground radar to very precisely manage wheel slip for maximum adhesion

Which of any of these are possible in OR at the present time?

It would be nice to have them all working at some stage. UK was not so keen on sand as USA, so we had some locos built with ( 1 ) and ( 3 ) that were not provided with sand at all.

I assume ( 0 ) and ( 1 ) are not yet possible.

Is there yet any way to put ( 2 ) into an eng file with or without the automatic sanding?

What about ( 3 )?

So far as I can see none of ( 0 ) to ( 3 ) would require a deviation from the default ORTSCurtius_Kniffler ( ) parameter, although possibly some examples of ( 3 ) might increase the "virtual adheasion".

Then finally we come to ( 4 ) for which higher starting TE should be possible without wheelslip. I can put my locomotive data into the Curtius-Kniffler spreadsheet from Peter's CTN website and then add

ORTSCurtius_Kniffler ( 7.5 44 0.25 0.7 )

into the wag section of my eng file. This alone does not seem to make any difference to the locomotive performance. I assume I need to add other parameters. Can someone please give me a working example of how to put this into an eng file. (I would expect the current to the traction motors to be controlled, to prevent slipping, but I would not expect to physically see the throttle move on the driver's desk.)

[darwins]

So after some trial and error:

ORTSAdhesion ( ORTSCurtius_Kniffler ( 7.5  44  0.25  0.7 ) )

in the wag section of the eng file will give me the desired adhesion.

AntiSlip ( 1 )
ORTSWheelSlipCausesThrottleDown ( 1 )

in the eng section reduces power to the traction motors, by physically reducing the throttle to the next lowest available notch on the throttle
as conditions improve the throttle remains in the lower position and the system does not attempt to restore the power demanded by the driver

For the notched throttle with 8 notches the system prevents the slip from by cutting the throttle from 100% to 88%

For a continuously variable throttle the system was unable to prevent the slip. Following an intial cut to 99% and a few seconds of slipping power was then cut down to 24%.

Wonder what happens with a lot of notches, like a tap-changer?