[Fablok]

Well, this is how we probably flew, you can say, the main electrical circuits in electric locomotives of Polish production. And here is the basic data. For the 4E series and its improved version 303E. In the photo EU07-001 at the PaFaWag factory in Wrocław.1965

https://i.ibb.co/ZV07rM8/11913983-932172780175925-4032459192482557681-n.jpg

And here is the corrected version, type 303E produced by Poznań HCP.

https://upload.wikimedia.org/wikipedia/commons/0/08/EU07-472_10.05.2005_Warszawa.jpg


Producer
Pafawag Wroclaw (4E) 001-244
HCP Poznan (303E) 301-543

Years of construction
1964–1974 (4E)
1983–1992 (303E)

Axle arrangement
Bo'Bo'

Service weight
80t (4E)
83.4t (303E)

Length
15,915mm (4E)
16,235 mm (adapted for the installation of an automatic coupler) (303E)

Width
3038mm

Height
4343mm

Wheel diameter
1250mm

Traction
electric

Engine type
EE541A

Number of engines
4

Supply voltage
3kV DC

Continuous power
2000kW

Hourly power
2080kW

Maximum tractive effort
280 kN

Gear Ratio
79:18

Maximum operating speed
125 km/h

Axle load on the rails
20t (196kN) (4E)
20.85 t (204 kN) (303E)

brake system
Oerlikon

The EU07 electric locomotives and their ET22 and ET41 derivatives are unfortunately dying. These are the last moments to see them still in motion. And the matter is made worse by the fact that from October 1, after the amendment of the penal code, it will be forbidden to photograph railways in Poland due to the war in Ukraine and espionage. But the greatest breakthrough was made by the grandfather of these locomotives, i.e. the English Class 83. EU06 was created from it and thanks to the purchase of the license we have these locomotives. Unfortunately, the EU06 has not been used for a long time. They spent their whole lives in mine and their locomotive depot in Krakow. But in order to reach them at the end of their lives, instead of prestigious and long trains, they traveled with freight trains. ;( Well, after these descriptions, I hope that we can do something in our Open Rails to celebrate millions of kilometers made in Poland :)

Grandfather EU06-03 first time in Poland :) December 1961

https://i.ibb.co/dfQx3ft/EU06-03.jpg

[Weter]

Yes, ORTS, like some other train sims, allows us to reproduce legendary rolling stock in virtual world for still having it existent.
It would be great, if Poland would assign some funding for natural locomotive samples museefication. Or, at least if some private collections would exist, as it is in Britain, where are some volunteers societies, established for preservation of some most legendary rolling stock. The DPS, for instance. www.thedps.co.uk

[Jean-Paul]

Fablok, on 28 August 2023 - 10:40 AM, said:

Gentlemen. I bring up this topic, because thanks to you we have beautifully recreated the Oerlikon type brakes, I would like to suggest to our specialists whether it is possible to divide the throttles so that they work in a given setting as a series of engines and then in parallel. Polish electric locomotives with resistance start have just such control. For example, the position of the controller from position 0 to position 28 is driving on resistances in a series system. Position 29 is shunting, operated by a separate lever with 6 positions. And from position 30 to 43 is driving on a series motor system and 44 is shunting. What exactly am I talking about? A 4-engine locomotive has 2 ammeters, one for serial driving and two for parallel driving. 3 axle locomotives have 3 ammeters for series, series-parallel and parallel running. So, would it be possible to divide the work of ammeters and add shunt. For example, as an additional controller, which would be operated with, for example, Shift+D and Shift+A keys. Below is a video of the locomotive driving and what it looks like in reality.

4 motors

https://www.youtube....h?v=NQL9LUK89z0

6 motors

https://www.youtube....h?v=R-97Dq4z7dk

Hi, Fablok !
I suscribe with enthusiasm to your remark ! Concerning electric locomotives, OR neglect completely changes in coupling modes of motors. We must know that electric locomotives were a "relative specificity" of European railways. In USA, for instance, if CM & SP ("Milwaukee") was a pioneer in 3kV DC electrification, electrifications were rare, and often realized around AC current, with a predominance of "direct motors" (as, in Europe, Germany and Switzerland) in which coupling changes are of course unknown. But in France (my country), Belgium, Netherlands, Poland, Italy, Spain.... DC current was the rule before 1950-1955. And even with AC electrifications, locomotives used generally rectified current feeding DC motors (generally without coupling changes, serie/parallel being the rule). After 1990-1995, we saw a generalization of triphased alimentations, with asynchronous or more rarely synchronous motors. And MSTS, like OR, just offers a "notch" efficiency which corresponds to this particular case... i.e. effort and power are strictly proportional to % throttle.
In the long story of OR, introduction of "PowerCurves" was a decisive evolution : it allows a modulation of tractive effort which is not linear. Realistically, increasing of speed weakens effort as you mention it in one of your more recent posts. Coupling changes, notches on resistors and "shunting (field weakening)" may be represented in these curves, by defining appropriated curves.
But, that's sure, some problems remain. First, a fictive coupling change has no influence on amperage, and "shunting" no more... In fact, that's really not the case, and that's the reason why coupling changes speeds and "shunting" speeds must be selected with an extreme care, to maintain intensity of current at reasonable values. To resume, even with a very realistic power curve, it's always possible to do anything you want (like, for instance, starting @ 100 % throttle) without any consequence.
Surprisingly, I think we must go on the Diesel side to understand what could be improved. I speak only of "Diesel-electric", the tricky case of hydraulic, hydromechanic and mechanic Diesels remaining incompletely resolved (sorry, but improvements are still far to be perfect...). First models of american Diesels (E & F-series, first -GP, Alco, Baldwin, and so one...) could be regarded as DC electric locomotives, with the important difference that electric current being produced by a Diesel with adjustable power, there was no need for resistor-fitted notches. These first generation engines used coupling changes, and of course field weakening by shunting to obtain - with success - a relatively linear power profile, following selected notch value (by tradition, eight). If at beginning, coupling changes and levels of field weakening were manual, EMD, Alco, GE... choosed to make these transitions automatic : it was the case from EMD F7 models (optional on F3), Alco "Centuries", etc... Process was a fair electro-mechanical realization : at a given speed, electric contacts were modified and realized change (of coupling or "shunting"). From 1965, DC generators were replaced by alternators in 2nd generation locomotives (GP38, 40, SD40, etc...), and coupling change was less necessary... But weakening field levels remain, of course, because traction motors remain DC. And like with electric engines, at constant speed, when field weakens, intensity (amperage) growths !
So, the problem of a perfect realism is almost the same as with electric engines : it's possible to have realistic power curves, but A-meter will never show realistic values. The only simple case concerns post-1995 Diesel, fitted with AC motors (AC44, SD70 MAC, etc...) for which "MSTS" or OR model is correct.
Diesel are here, in fact for another reason : from one model of locomotive to other, if "tradition" of 8-Notches governor has been maintained over time, we must know that, first, real "position" in terms of maximal power varies from one builder to other, and from one model to other. Alas,"MSTS-fans", the traditional "0 - 0.125 - 0.25 - 0.375 - 0.5 - 0.625 - 0.75 - 0.875 - 1" is completely wrong !!. The eight notches of a governor are associated to a fixed RPM of Diesel motor, and depend of this one... Just a question : did you really succeed to maintain 10 mph in a yard and "solo" with "0.125" notch ? I don't... That's because first notch is often in reality something like 0.05 or 0.06... If you practice Multiple Units - almost the rule in the USA - no problem if your units are the same... But what if they are different, with different repartition of "8-Notch" governor ? In reality, MU connectors are simplistic : when an impulsion is given on leading engine to go to Notch 1, it's tranmitted to 2nd, 3rd, etc... units on which Notch 1 is triggered, with its own effect on concerned engine. For instance, if @ Notch 4, first engine is @ 600 rpm and 35 % power, second @ 850 rpm and 48 % power, in reality that would be the case. In OR, calculation makes a strange mix in which all engines will be @ 35 % power (value of leading engine), and for trail engines, @ rpm not provided by builder...
So, I come to my first conclusion : if a progress must be accomplished with OR, it would be to dissociate "Notch" concept from a power ratio. I suggest this evolution : Notches are defined in .eng files as
EngineControllers (
Throttle ( 0 8 1 0
NumNotches ( 9
Notch ( 0 0 Dummy )
Notch ( 1 0 Dummy )
Notch ( 2 0 Dummy )
Notch ( 3 0 Dummy )
Notch ( 4 0 Dummy )
Notch ( 5 0 Dummy )
Notch ( 6 0 Dummy )
Notch ( 7 0 Dummy )
Notch ( 8 0 Dummy )
)
)
And in a adapted "MotivePower.inc" or what you want, Notches are associated in the case of Diesel with a fixed RPM, and with a PowerCurve definition. In the case of electric locomotives, associated with a range of intensities, determinated by a MaxIntensity varying following speed, coupling and field weakening. Work seems simple with Diesel engine ("Throttle" looks similar, in fact, with "Gear" item, no ?), less with electric ones...!
To illustrate how a PowerCurve can be realistic (in present configuration), I join a sample realized for my friend Nicolas Schou, which created a wonderful french "2D2-9100", with three couplings ("Série", Série-Parallèle", and "Parallèle"), each coupling fitted with not less than 7 levels of field-weakening. Sixty notches + neutral, what makes driving more difficult than a steam engine one. This locomotive is downloadable here : http://download.acti...410_Sounds.zip. PowerCurves were realized with builder and SNCF datas.
I hope sincerely that this theme will lead to an interesting development !
Have a nice day !
Cheers,
Jean-Paul

Attached File(s)

•  SNCF_2D29102.eng (58.7K)
  Number of downloads: 5

[Jean-Paul]

Weter, on 04 September 2023 - 10:06 AM, said:

Thank You, Jean-Paul:
I've just learned some things' I didn't know before.
But, ORTS diesel () block allows to define not only tractive force curves, but notch-to-diesel rotation speed dependencies already.
As well - for torque and fuel consumption.

Hi, Weter !
Yes, of course, torque and RPM are included in "motive".inc. But, what if when two Diesels in "MU" have not the same Notches distribution ? For instance, first with :
Throttle ( 0 1 0.125 0
NumNotches ( 9
Notch ( 0 0 Dummy )
Notch ( 0.125 0 Dummy )
Notch ( 0.25 0 Dummy )
Notch ( 0.375 0 Dummy )
Notch ( 0.5 0 Dummy )
Notch ( 0.625 0 Dummy )
Notch ( 0.75 0 Dummy )
Notch ( 0.875 0 Dummy )
Notch ( 1 0 Dummy )
)
)
And second with :
Throttle ( 0 1 0.125 0
NumNotches ( 9
Notch ( 0.000 0 Dummy )
Notch ( 0.050 0 Dummy )
Notch ( 0.110 0 Dummy )
Notch ( 0.230 0 Dummy )
Notch ( 0.350 0 Dummy )
Notch ( 0.480 0 Dummy )
Notch ( 0.600 0 Dummy )
Notch ( 0.800 0 Dummy )
Notch ( 1.000 0 Dummy )
)
)
Knowing that's the first which fixes throttle's value, when on Notch 7, for instance, I'll get 0.875 value for both engines, which is not correct for second one (0.8 on Notch 7). Second unit will give approx. 10 % of supplemenatry effort, which doesn't exist in reality, because each value of Notch determinates a discrete value of RPM (and MaxPower). In reality, when 1st is on Notch 7, second too, what means first is at 87.5 %, and second at 80 %. Not a very big problem, I know, but not totally correct.
Thank you for your interest !
Cheers,
Jean-Paul

[pschlik]

There's a lot here that would be great to have in OpenRails, and some serious improvements to the (diesel) electric motor model have been on my mind, but this is far from a 1 patch improvement. There's a lot to be done to get from where we are now to the ability to support the EU07 or similar, so it'll have to be step by step.

A bit of code is present for simulating series wound motors (which has been sitting unused(?) for 10 years), but there still needs to be simulation of how the motors are connected to eachother, how the source (wires/transformer/generator) is connected to the motors, the ability to change how the motors are connected in real time, new control types, even some way to define locomotive performance without the use of tractive force tables (this is a very loaded problem).

This would have to start with connecting the series wound motor code to the simulation and seeing what happens. That's probably a massive can of worms on its own, but still has the potential to boost realism without requiring completely rethinking existing .eng files. That alone wouldn't be enough to recreate an EU07 though. Once that's in place, it starts to become feasible to consider how motors are connected in series/parallel (right now, the electrical simulation isn't deep enough for that to matter) and determine what that means for the voltage and amperage that must be provided by the power source to produce the specified force (at this point you'd still have to work backwards from tractive force curves).

Once there's this simulated motor arrangement, work could finally be done to change that arrangement (and field weakening-looks like the series motor code has been set up with that in mind) in real time to shift the voltage and current demands on the source. I would assume the first pass of transition would be automatic transition, as featured on many diesel electric locomotives, as this would no doubt be simpler than adding new control types for transition. In such a case, users would likely be responsible for defining the speeds at which transitions occur, the type of transition (generator, motor series/parallel, field weakening), and the parameters of the transition (what's the field weakening ratio, are the motors in series or parallel, etc).

All of these steps could reasonably be implemented while still using MaxTractiveForceCurves, with electrical performance being derived from force performance. But once you get to manual transition, it's much more useful to think about requiring user input for electrical parameters (think of curves for generator amps vs volts, a table of resistor settings vs throttle, a fixed voltage at the input in the case of old DC electrics) and instead deriving the forces from that. Without this, either separate tractive force curves would be needed for each transition/field weakening configuration (not feasible) or the transitions would be entirely aesthetic and not affect tractive effort, which would defeat the purpose (this would likely be the case for the first implementation of automatic transition).


So hopefully that makes it clear how this can be done, though requiring some time and willpower.

[darwins]

This was the throttle arrangement on Britain's first main line diesel locomotives:

Notch 1 - Engine 425 rpm - reduced generator excitation

Notch 2 - Engine 425 rpm - full generator excitation

Notch 3 - Engine 600 rpm - reduced generator excitation

Notch 4 - Engine 600 rpm - reduced generator excitation

Notch 5 - Engine 600 rpm - reduced generator excitation

Notch 6 - Engine 600 rpm - reduced generator excitation

Notch 7 - Engine 600 rpm - full generator excitation

Notch 8 - Engine 750 rpm - full generator excitation


The 6 traction motors were permanently connected in three pairs.

There was one stage of field weakening, weak field being applied above 32 mph in Notch 7 and above 42 mph in Notch 8.

Weak field would be automatically disengaged below a certain speed (maybe 20 or 25 mph) of if the controller was moved to Notch 2 or below.

Unloading of the generator was at 83 mph and the traction motors were governed to give a maximum speed of 93 mph.


Later standard throttles kept something in common with this. For the first part of increasing the throttle the engine rpm were maintained at idle speed, whilst the generator field was gradually increased. Once the generator was at full field then the engine rpm was increased until maximum rpm was reached.

[FloBarr]

Hi all,

It's a pleasure to see DC Motors are interresting peoples!

I've made some code modification several years ago (Darwins tested it), and I now use it on a personnal version of OR, on a real cab simulator. At the moment, my code handles Electric and Diesel DC motors locomotives, with (usual) or without notches (for diesel where voltage is set by pneumatical controler), couplings, and field reductions (manual or auto); it also adds new controls, like field reduction, auto throttle increase (Progression Rapide in France), current control and automatic stop/locking of increasing, ... Also added a gearbox setting for this locomotives (used in France on several diesel and electric series)

It's lot harder to set than standard locomotive, as it requires to define motors resistances, voltages, current/torque ratio, ... but results are interresting, and pretty accurate (but not exact, some parameters are not used, like motor temperatur, changing resistance!). Of course, real ammeter and motor voltmeter has been added. But... There is no more torque curve! Torque is calculated from voltage, current, bemf, gearbox ratio...

Flo

PS: sorry for my poor english!

[Csantucci]

Very interesting! It would be nice if you could make your changes public.

[Jean-Paul]

Jean-Paul, on 05 September 2023 - 12:37 AM, said:

Hi, Weter !
Yes, of course, torque and RPM are included in "motive".inc. But, what if when two Diesels in "MU" have not the same Notches distribution ? For instance, first with :
Throttle ( 0 1 0.125 0
NumNotches ( 9
Notch ( 0 0 Dummy )
Notch ( 0.125 0 Dummy )
Notch ( 0.25 0 Dummy )
Notch ( 0.375 0 Dummy )
Notch ( 0.5 0 Dummy )
Notch ( 0.625 0 Dummy )
Notch ( 0.75 0 Dummy )
Notch ( 0.875 0 Dummy )
Notch ( 1 0 Dummy )
)
)
And second with :
Throttle ( 0 1 0.125 0
NumNotches ( 9
Notch ( 0.000 0 Dummy )
Notch ( 0.050 0 Dummy )
Notch ( 0.110 0 Dummy )
Notch ( 0.230 0 Dummy )
Notch ( 0.350 0 Dummy )
Notch ( 0.480 0 Dummy )
Notch ( 0.600 0 Dummy )
Notch ( 0.800 0 Dummy )
Notch ( 1.000 0 Dummy )
)
)
Knowing that's the first which fixes throttle's value, when on Notch 7, for instance, I'll get 0.875 value for both engines, which is not correct for second one (0.8 on Notch 7). Second unit will give approx. 10 % of supplemenatry effort, which doesn't exist in reality, because each value of Notch determinates a discrete value of RPM (and MaxPower). In reality, when 1st is on Notch 7, second too, what means first is at 87.5 %, and second at 80 %. Not a very big problem, I know, but not totally correct.
Thank you for your interest !
Cheers,
Jean-Paul

Hello !
Remaining on Diesel-electrics, this discovery which is not one...And which proves Weter completely right :sign_thanks: I tried to define in "includes" an universal "Notches.inc" for US Diesel, following the 0-0.125-0.25-0.375-0.5-0.625-0.75-0.875-1 rule for US Diesel. By defining in power curves adequate values for notches, considering 0.125 = 1 ; 0.250 = 2, etc... and affecting the real power and RPM for each notch, I obtained an "universal" combination of notches, which may be convenient for all types of US Diesel, fitted with eight notches, taking in account their specificities in terms of RPM and effective power on each notch (sometimes rather different from one builder to other, or one type to other). Only condition is to fix convenient power Diesel for each RPM, and the good value of rpm for each notch.
I sincerely think the old uses of MSTS' conventions left a lasting mark on my mind...!
Thanks agin to Weter !
Have a nice day !
Cheers,
Jean-Paul

[pschlik]

And to interject on the US diesels, the horsepower response to throttle is indeed not linear (it's somewhat under-proportional), nor is the starting tractive effort in each notch (which is actually over-proportional), but I still personally use notch 1 = 0.125, notch 2 = 0.25, etc, as this even spacing of the notches works better for raildriver. You won't notice a difference on keyboard, but on external hardware setting notch 1 = 0.05, notch 2 = 0.1, etc...will make the real-life throttle handle a bit too precise at lower settings. Instead, it's better to reflect the nonlinear power and tractive effort through setting ORTSDieselEngines and ORTSMaxTractiveForceCurves correctly for each notch.