Automatic Acceleration Throttle for electric multiple unit trains and trams
#1
Posted 03 January 2020 - 04:16 AM
Typically this would be controlled by current sensitive relays. The driver would move the controller to maximum and the unit would start in the 'first notch' and 'notch up' automatically as the speed increased (or more accurately as the traction current decreased).
This type of throttle was never represented in MSTS and can not be accurately represented in OpenRails.
A good 'work around' for the acceleration can be made in OpenRails by using ORTSMaxTractiveForceCurves in this case changes in 'notch' are set according to speed and can be calculated to change at a fixed value of Tractive Force (which is more or less equivalent to a fixed current value in OR).
This 'work around' is fairly good for inner suburban trains and trams - where running is accelerate / coast / brake. The limitations make it less suitable for faster non-stopping trains. This is because the real throttle only provided for 'notching up' and not for 'running down'. In order to reduce power the driver would have to return the controller to OFF and then notch up again.
Two variations of this were common in UK from the 1950s to the 1990s.
British Railways:
Five position controller 0 - OFF / 1 - Shunt or Inch / 2 - SERIES / 3 - PARALLEL / 4 - WEAK FIELD
Normally the train would 'notch up' until it reached the selected power.
However if the controller was moved back to a lower setting then notching was halted.
This allowed the train to remain for a short time in a resistance notch before either continuing to notch up or shutting off power to coast.
On d.c. units running for any length of time in that way would cause resistances to overheat, but it could be used to slow down acceleration in low adhesion conditions.
London Transport:
Four position controller 0 - OFF / 1 - Shunt or Inch / 2 - SERIES / 3 - PARALLEL
In this case there was no means of halting 'notching' at intermediate values, acceleration would continue until the highest running notch selected was reached.
There were generally three other switches in the cab that affected the behaviour of the throttle.
Rate Switch - allowed either normal rate of notching up or slower acceleration for conditions of low adhesion.
Weak Field Switch - when OFF position 3 on the controller was full parallel, but when ON notching continued to the weak field setting(s).
Coasting Control - limited maximum speed generally to 30 mph if the weak field switch was off.
I believe the additional switches could only be changed when the train was stationary.
Trello card
#2
Posted 03 January 2020 - 10:42 AM
More about this can be read here: https://library.e.ab...rmer-Kiruna.pdf
#3
Posted 03 January 2020 - 04:07 PM
That method even sometimes reduces chances of string lining depending on Degree of Curve vs how much force the Railhead/Wheels/Draft can take, if stock is legal to negotiate the given curve and what are its limits from straight line over or breaking the rail.
I can tell you I don't like how Gensets 3GS4B are programed in remote operation that do this that they can't pull even in packs vs a single GP15 pulling hard an faster.
#4
Posted 04 January 2020 - 10:57 AM
The other controller is for AC power. It uses a step down transformer with three voltage taps. Each voltage tap has a resistor that can be cut out, so there are a total of six steps. The controller has three positions that correspond to the different voltages. The logic automatically steps through the voltages with and without resistors until the selected voltage is reach. This controller has the ability to step down to a lower voltage without the resistor.
Doug
#5
Posted 04 January 2020 - 12:06 PM
dajones, on 04 January 2020 - 10:57 AM, said:
The other controller is for AC power. It uses a step down transformer with three voltage taps. Each voltage tap has a resistor that can be cut out, so there are a total of six steps. The controller has three positions that correspond to the different voltages. The logic automatically steps through the voltages with and without resistors until the selected voltage is reach. This controller has the ability to step down to a lower voltage without the resistor.
AGNEW, W. A. "Electric Trains - Their Equipment and Operation" is a very detailed work from 1937 dealing with several controllers then in use in Britain, ( all d.c. )
Quote
The idea of "power up" hold and "power down" was very common throughout Europe for a.c. electric locomotives.
There was a good description of it in an earlier thread and a link to a video of a Hungarian loco.
For example British class 81 a.c. locos had 38 transformer taps for full field, plus two steps of field weakening. The controller was
OFFRUN DOWN - automatically goes down through the notches until it either reaches OFF or is moved to HOLD
NOTCH DOWN - moves manually one notch downHOLDNOTCH UP - moves manually one notch upRUN - automatically goes up through the 38+2 notches until it either reaches the maximum or is moved to HOLD
As I understand it this kind of controller does not use a current sensitive relay to decide when to notch up or down but simply takes about 1.5 seconds in each notch before changing up or down. So it is a little different from automatic acceleration in a multiple unit train.
There is also a Trello card for that.
As stated there the idea of an ADVANCE - HOLD - REDUCE type of control can be applied to steam reversers on steam locos as well.
So far OR has not had the ability to include controls of this general type.
#6
Posted 04 January 2020 - 02:40 PM
NickonWheels, on 04 January 2020 - 01:16 AM, said:
Sort of, its just automatic governor systems on modern freight trains removing unnecessary acceleration to save fuel and emission standards... Its like the accelerometer governing the Engine or Traction to not exceed its set limit where for example your train can't gain the set limit of no more then 15MPH every 60 seconds. The engines wont rev up traction any faster but you will still get power and speed climbing slowly just not any faster then set. Its like running ORTS in autopilot trying to make AI not be so impatient or too fast heavy on the throttle. Doesn't really have to do with slowing down deceleration.
PTC (Positive Train Control) has similar future in some units that are automated with the click of a button.
#7
Posted 01 February 2020 - 02:46 PM
darwins, on 04 January 2020 - 12:06 PM, said:
https://trello.com/c...pe-of-throttles
(This controller can be faked in OR by the work around of changing power according to speed depending - but of course it should change power with time.)
Please let me know where to find the work around mentioned in the Trello card. For example, on a class 40, the field changes at approx. 24mph increasing the amps at that speed.
Useful topic: https://www.railforu...-divert.119164/
This post has been edited by AlanColey: 01 February 2020 - 02:51 PM
#8
Posted 01 February 2020 - 11:29 PM
AlanColey, on 01 February 2020 - 02:46 PM, said:
(This controller can be faked in OR by the work around of changing power according to speed depending - but of course it should change power with time.)
Please let me know where to find the work around mentioned in the Trello card. For example, on a class 40, the field changes at approx. 24mph increasing the amps at that speed.
Useful topic: https://www.railforu...-divert.119164/
Hi Alan
The workaround that I was referring to above was one that I had applied to British Electric Multiple Units that have resistance notches.
The same can be used to fake the performance of diesel electric locomotives - including field weakening steps.
To do this you do need to know at what speeds the changes such as field weakening and offloading occur.
Graphs like chart 9 and chart 12 in the BR test report for 10203 should help.
Although I note in the charts a smooth curve is given for tractive effort versus speed - so current change would appear smooth in OR - fortunately the relevant speeds are highlighted so it may be possible to "saw tooth" the changes in force in order to model the changes in current.
BR 10203 (direct ancestor of Class 40) is modelled on CTN using tractive force curves - you might like to have a look at this and see what could be done to highlight the changes due to field weakening.
I am shortly going to try to make some replacement OR files for British diesel electric locos.
One problem for me with trying to include field weakening and offloading points would be getting the relevant prototype information.
I think that it may be possible for me to find the relevant information for classes 08, 15, 20, 24, 28, 31, 40, 45, 47, 50, 55 and 56 at the NRM - with curves for other classes being assumed similar.
If you have prototype information for any others (including HST) please let me know.
#9
Posted 05 July 2025 - 12:04 PM
In a discussion on the Train Sim Safe House website, Gwyd has suggested that he may try to script a throttle controller that cannot be "notched down" and must be returned to OFF an notched up again to a lower setting. This would be a very useful solution until such time as we are able to model automatic acceleration fully.
https://tsforum.foru...outh-east-route
#10
Posted 05 July 2025 - 11:41 PM
I already have a working beta for a throttle controller that covers most rheostatic DC motors, with automatic notch up, and with or without notch-down individually for each position. However, this still is only focused in rheostatic control, but it lacks the flexibility to cover other locomotive types. We have to do some minor modifications to allow more parameters to be controlled, besides starting resistance, motor arrangement and field weakening.
Another missing topic is to have a proper simulation of the DC motor, but until that is available we can use the tractive force curves.
#11
Posted 06 July 2025 - 12:52 AM
Whilst there were many variations, the basic system invented by Frank J Sprague was extremely widespread on electric multiple unit trains, metro trains and tram cars for almost 100 years from 1890s to 1980s.
It sounds like the model you are working on will cater for this.
The basic controller has just four positions:
OFF - Returns all connections to starting position
SHUNT - Motors connected in series with all resistances in circuit - for slow speed shunting only
SERIES - Automatically cuts out resistances one at a time until motors are in series with no resistances in circuit
PARALLEL - Transitions to Parallel with all resistances in circuit and then cuts them out one at a time
Cutting out of resistances and connection of circuits, either by electro magnetic control, electro pneumatic control or pneumatic cam control could only be done in one direction = acceleration. To decrease power the controller would have to be returned to OFF, power cut and a lower setting chosen.
In some cases if you moved the controller back before the final setting was reached the notching up would continue, in others if you moved the controller back then the mechanism would stop notching until you moved it forward again. (Potentially leaving you in a notch that was not a "running notch" if you did not take care!)
Two variations of this used in UK were
OFF / SHUNT / SERIES / TRANSITION (or First Parallel) / PARALLEL
I am not at all sure of the purpose of that as First Parallel was clearly not a "running notch".
and
OFF / SHUNT / SERIES / PARALLEL / WEAK FIELD
for more modern trains.
Notching was generally done using current dependent relays, which as you say can for the moment be represented with tractive force curves. Although in the early 20th century Siemens built at least a few trains where the notching up was time dependent rather than current dependent.
Hopefully whatever you are planning will allow ORTSLabels to be included in throttles.
Given there is no detailed modelling of electrical circuits at the moment, then what you are planning to use for dc resistances should work equally well for ac tap changer technology.
The equivalent of the above on older ac multiple units in UK was
OFF
1 - lowest voltage tap for shunting
2 - notches up automatically to half voltage tap
3 - notches up automatically to full voltage tap
4 - notches up automatically to full voltage tap then applies field weakening
Automatic notching should then equally deal with tap changers on locomotives too where the throttle usually has
OFF
RUN DOWN
NOTCH DOWN
HOLD
NOTCH UP
RUN UP
I think some French Open Rails locos already have a script for this!
#12
Posted 06 July 2025 - 03:24 PM
Some notes.
Here DC EMU controllers comprise of "group switcher" (which is a camshaft, actuating individual contractors in defined order) and some separate contractors, with own coil or EP drive for actuation. Group switcher is propelled by two EP cylinders or diaphragm pushers, turning shaft on steps to notch-up direction only. Trams has servo-motor-propelled shaft, but it is driven through Malta Cross gear, so has stepped movement too.
Therefore, when handle is returned to zero, line contractor cuts power and shaft turns full revolution to zero step again. Independent contractors are for switching to parallel and field weakening, so when acceleration is needed to be stopped, it's possible to move handle one notch back, but if from parallel to series, notable jerk will occur, so it's strongly recommended to return handle to zero in such cases. If field weakening to parallel - that's not so jerky.
Also, there's low acceleration button, which changes current thresholds, which causing notching-up - that's for slippy conditions, as Darwin told.
Sometimes there are two additional notches between series and parallel, which are used to control group switcher directly, when automatic system is failed.
That's like AC -fashioned systems: one notch-up/hold.
Hence, group switcher deals with switching-off rheostat's sections mostly, contractors are changing motors arrangement and shunting excitation coils with resistors for field weakening, while independent group switch, called reversor, swaps plus and minus for changing rotation direction, and, when by amid brakes are present, it also assembles cirquits for performing that. Most often, reversor has no arc extinguishers, but comprises of "drum" and contact fingers, so it acts, when current is cut off only.
#13
Posted 09 July 2025 - 09:25 AM
Can such behaviour be replicated?
Here is a video of what i wish to achieve about the automatic acceleration/deceleration
Watch the throttle and the notch indicator. The notch indicator is the analog one on the far left.
https://www.youtube....h?v=pWmSByoI54o
#14
Posted 09 July 2025 - 10:09 AM
There are quite many steps of voltage predefined values made, so no need to re-arrange motors from series to parallel and back, so no huge group switches, dealing with giant currents needed. Resistors are still used, however, as long as motors are series-excitated DC machines, so strong stator field still restricts rotation speed.
Back to voltage control, there are 2 contact scheme types mainly: hi-voltage switching, either low. High has that benefit, that currents are small, while power is in high voltage. So, controller has smaller contacts, often more compact, but efficient insulation needed. Low voltage scheme dealing with higher currents, so contacts are bigger, but insulation can be simpler. Some additional devices are needed to make scheme more stable, such as reactor coils and so on.
Usually, no automatic acceleration systems on mainline locomotives, but driver notches manually up and down, referring with a permeates reading. EMUs have similar looking controller, as DC ones, but I didn't interest, how it works.
As far, as many long-time(all time) running positions, driver's handle usually has no all of them, but some variations, as described earlier: "Quick Off" "Auto off" "One Down" "Coast" "One Up" "Hold" "Auto Notching-up".
Advanced locos had automatized control with numerous handles and maybe wheel: current reference, speed reference, tractive force reference.
Later, electronic controls, controlling variable-frequency drives (power invertors) made system more sophisticated, flexible, smart, and capable to computer-control, which is a step to autopilot.
#15
Posted 09 July 2025 - 10:38 AM
Yes i am aware of the complexities.
For example, at this locomotive the throttle works this way if i'm not mistaking:
- Step 1 / ON
- Step 2 / Notch 1 Traction
- Step 3 / Notch 2 Traction
- Step 4 / Notch 3 Traction
- Step 5 Forward / Automatic increase traction notch by notch with 0,4s delay between (Notch 4 - Notch 40)
- Step 6 Forward-Backward / Holding position for notches (either acceleration or deceleration)
- Step 5 Backward / Automatic decrease traction notch by notch i guess with the same time delay between (Notch 40 - Notch 4)
- Step 7 / Notch 4 Weak Field 1 n%
- Step 8 / Notch 5 Weak Field 2 n%
- Step 9 / Notch 6 Weak Field 3 n%
This is what i recall.