Some questions about transmission with 3-phase ac motors

[darwins]

I have some questions about transmission with 3-phase motors and variable frequency supply.

The basic principle says as the supply frequency increases the speed of the motor(s) increases.
So you accelerate by increasing the supply frequency.
The supply frequency also sets a "maximum" speed in the sense that if the synchronous speed is exceeded then the motor becomes a "generator".

Presumably then the maximum frequency supplied to the motor is such that the synchronous speed would be at (or slightly above) the maximum design speed of the train. Also speed setters can work by fixing a maximum frequency.

What is it that sets the rate of increase of frequency during acceleration?
Is it fixed depending on power 'notch'?
Or is there some feedback mechanism based on axle speed / rotational speed of the motor?
Or is there some feedback mechanism based on rail speed / ground speed?

Presumably it is the same variable frequency supply that supplies all traction motors on one vehicle (or even group of vehicles).

How do the above answers influence slipping with such trains?

Supply frequency would limit the maximum rotational speed of any driven wheel, but that is not so much use if the synchronous speed is much higher than the rail speed.

If increase in frequency is based on actual rotational speed then wheels could presumably still rotate up to maximum speed with locomotive barely moving.

If increase in frequency is based on rail speed/ground speed either measured by radar or perhaps in a multiple unit train by measuring the speed of an unpowered axle then presumably supply frequency could be used to limit rotational speed and reduce slipping.

How does this all work in practice? Can traction with ac motors experience uncontrolled slipping in the same way as traction with dc motors or does it work differently?

 

[edwin_m]

Not sure I fully understand this. By "supply frequency" are you referring to the AC frequency being fed into the motors, not the 50Hz frequency of an AC traction supply?

 

[darwins]

Not sure I fully understand this. By "supply frequency" are you referring to the AC frequency being fed into the motors, not the 50Hz frequency of an AC traction supply?

Yes. I am referring to the frequency supplied to the motors.

 

[ac6000cw]

Presumably it is the same variable frequency supply that supplies all traction motors on one vehicle (or even group of vehicles).

It depends... if you need maximum tractive effort in all conditions e.g. for heavy freight, then having 'inverter per axle' control of tractive effort is best since each axle can experience different railhead conditions (e.g. due to weather) and weight (from weight transfer effects caused by axle torque). But this is the most expensive option, so for less demanding duties sharing an inverter across several axles is very common.

If increase in frequency is based on rail speed/ground speed either measured by radar or perhaps in a multiple unit train by measuring the speed of an unpowered axle then presumably supply frequency could be used to limit rotational speed and reduce slipping.

Correct - it's the very tight control of axle rotational speed that AC-drives can achieve which allow locos and MUs to operate at higher ratios of tractive effort to weight than is possible with series-wound DC motor drives. Plus 3-phase induction motors are more robust and generally smaller and lighter for equivalent power/torque capability.

Can traction with ac motors experience uncontrolled slipping in the same way as traction with dc motors

AFAIK, basically no, but of course the traction control system can force wheelslip a controlled fashion by design. (Normally the maximum tractive effort point is when an axle is rotating slightly faster than required by true ground speed - that is what causes the controlled slip/wheelcreep screaming noise when a freight loco is 'on the limit' climbing a gradient...)

 

[boiledbeans2]

Note that motors are usually operated with a torque command, not speed command, i.e. the driver notch command is translated to x% of the full torque. The motor speed will depend on the load.

So in theory, if the motor is very powerful and produces a very high torque (and/or there is low friction), it could cause wheel slip.

Also for an induction motor under load, the motor speed will always be slower than the synchronous speed, and this is the motor slip (not the same as wheel slip). And the presence of motor slip produces torque.

 

[AM9]

I have some questions about transmission with 3-phase motors and variable frequency supply.

The basic principle says as the supply frequency increases the speed of the motor(s) increases.
So you accelerate by increasing the supply frequency.
The supply frequency also sets a "maximum" speed in the sense that if the synchronous speed is exceeded then the motor becomes a "generator".

Presumably then the maximum frequency supplied to the motor is such that the synchronous speed would be at (or slightly above) the maximum design speed of the train. Also speed setters can work by fixing a maximum frequency.

What is it that sets the rate of increase of frequency during acceleration?
Is it fixed depending on power 'notch'?
Or is there some feedback mechanism based on axle speed / rotational speed of the motor?
Or is there some feedback mechanism based on rail speed / ground speed?

Presumably it is the same variable frequency supply that supplies all traction motors on one vehicle (or even group of vehicles).

How do the above answers influence slipping with such trains?

Supply frequency would limit the maximum rotational speed of any driven wheel, but that is not so much use if the synchronous speed is much higher than the rail speed.

If increase in frequency is based on actual rotational speed then wheels could presumably still rotate up to maximum speed with locomotive barely moving.

If increase in frequency is based on rail speed/ground speed either measured by radar or perhaps in a multiple unit train by measuring the speed of an unpowered axle then presumably supply frequency could be used to limit rotational speed and reduce slipping.

How does this all work in practice? Can traction with ac motors experience uncontrolled slipping in the same way as traction with dc motors or does it work differently?

The frequency of the ac generated by the traction electronics is determined by the instantaneous rotational speed of the motors with a bit added on to cause a phase lead which provides the torque. The actual speed of them is also monitored for safety reasons but not via the traction system. AS has been said above, the notch setting is just a proportion of the maximum allowed torque, and depending on the rolling resistance (usually quite low), the gross weight of the consist (i.e. tare plus payload), the gradient and wind resistance, a given notch setting will give a particular acceleration or eventually speed. Similarly, electro-regenerative braking does the reverse, applying generated current from the rotating motors which has to be converted eventually to slightly phase lead the 50Hz at the transformer (or rise above the 750VDC at the collectors shoes). Actual line speed can be measured on a non-motored wheel, avoiding any wheelslip issues. There's probably too much detritus in the four-foot to make doppler radar effective for speed measurement.

 

[RichardAsh1981]

There's probably too much detritus in the four-foot to make doppler radar effective for speed measurement.

I'm fairly sure BR Class 60 were described when new as using Doppler radar for speed sensing at very low speeds - something like it kicked in below 1 mph in order to allow the creep (limited slip) control to work reliably (on a vehicle with all axles motored of course). If it worked in MGR coal trains there were presumably solutions to detritus in the four-foot!

 

[edwin_m]

I'm fairly sure BR Class 60 were described when new as using Doppler radar for speed sensing at very low speeds - something like it kicked in below 1 mph in order to allow the creep (limited slip) control to work reliably (on a vehicle with all axles motored of course). If it worked in MGR coal trains there were presumably solutions to detritus in the four-foot!

I think the GM Super Series creep control fitted to the 59s used Doppler radar, and don't the 66s have something similar? This is separate from the slow speed control, as it allows better adhesion at any speed by introducing a very small amount of controlled slip when high tractive effort is needed at any speed.

All these classes have DC motors, showing that a fine degree of control is possible with this technology. It's probably easier with asynchronous AC motors, although as mentioned the motor turns slightly slower (or faster when braking) than the speed represented by the frequency it is being fed with.

There are also synchronous motors, where as the name suggests the motor will turn at the same speed as the rotating field created by the frequency it is fed with. I guess this might allow even more precise control of rotation speed. The French where quite keen on these for a while, including some TGVs and the "Sybic" locomotive, but they've never really caught on for rail traction. I believe they are now making a bit of a comeback, made possible by new materials that allow permanent magnets to be produced that are powerful and durable enough for this application.

 

[Ken H]

One huge benefit of AC motors is no commutator, so less maintenance. That justified retro fitting AC motors to some cl 455's. To make capacity at Wimbledon Depot.
If you have constantly variable frequency, what about interferance with signalling?

 

[AM9]

One huge benefit of AC motors is no commutator, so less maintenance. That justified retro fitting AC motors to some cl 455's. To make capacity at Wimbledon Depot.
If you have constantly variable frequency, what about interferance with signalling?

The (leagal) requirements for EM emissions and immunity of the traction and infrastructure respectively ensure compatibility. That is a fundamental issue of all modern systems. The use of axle counters goes some way to making that easier to achieve. The spread of ETCS will assist in providing a uniform environment for trains to operate in.

 

[ac6000cw]

I think the GM Super Series creep control fitted to the 59s used Doppler radar, and don't the 66s have something similar? This is separate from the slow speed control, as it allows better adhesion at any speed by introducing a very small amount of controlled slip when high tractive effort is needed at any speed.

For all-axles-motored, high tractive effort locos, I think having some way of measuring ground speed is vital to deal with 'all axles slipping' situations.

All these classes have DC motors, showing that a fine degree of control is possible with this technology.

Agreed.

It's probably easier with asynchronous AC motors, although as mentioned the motor turns slightly slower (or faster when braking) than the speed represented by the frequency it is being fed with.

It's the steepness of the torque versus rotational speed curve that makes asynchronous AC induction motors better at wheelslip control (from https://electricalacademia.com/induction-motor/torque-speed-characteristics-induction-motor/ ):

'T' = rotor torque, 'Sync speed' = the rotating electromagnetic field frequency = supply frequency, 'Rated speed' = Sync speed less the 'motor slip'. If the axle slips, the motor rotational speed increases rapidly and the torque will decrease rapidly, thus self-correcting the slip.

 

[boiledbeans2]

All these classes have DC motors, showing that a fine degree of control is possible with this technology. It's probably easier with asynchronous AC motors, although as mentioned the motor turns slightly slower (or faster when braking) than the speed represented by the frequency it is being fed with.

I'm sure they had the "final generation" of DC motors - the SEPEX (Separately Excited) DC motor - before AC motors became the norm. SEPEX are the "final generation" DC motors because they needed power electronic devices to work efficiently. But soon after, they realised they could use power electronic devices to directly control an AC motor instead, and that was the end of the DC motor.

In fact, modern AC drives try to ape the SEPEX motor in terms of control.

Quote from wikipedia: https://en.wikipedia.org/wiki/Vector_control_(motor)#Technical_overview

In vector control, an AC induction or synchronous motor is controlled under all operating conditions like a separately excited DC motor.[21] That is, the AC motor behaves like a DC motor in which the field flux linkage and armature flux linkage created by the respective field and armature (or torque component) currents are orthogonally aligned such that, when torque is controlled, the field flux linkage is not affected, hence enabling dynamic torque response.

== Doublepost prevention - post automatically merged: 31 Mar 2022 ==

There are also synchronous motors, where as the name suggests the motor will turn at the same speed as the rotating field created by the frequency it is fed with. I guess this might allow even more precise control of rotation speed. The French where quite keen on these for a while, including some TGVs and the "Sybic" locomotive, but they've never really caught on for rail traction. I believe they are now making a bit of a comeback, made possible by new materials that allow permanent magnets to be produced that are powerful and durable enough for this application.

In electric vehicles, permanent magnet synchronous motors (PMSM) are the norm today.
For trains, it hasn't really caught on, though I'm not sure why. The high-speed AGV sets operated by Italo is one notable example with PMSM (as well as the record breaking TGV V150). But strangely, the next generation TGV (Avelia Liberty) reverts back to induction motors (according to the specs on Wiki).

Another example is Toshiba, who retrofits older trains in Japan with PMSM to extend their life.

== Doublepost prevention - post automatically merged: 31 Mar 2022 ==

It's the steepness of the torque versus rotational speed curve that makes asynchronous AC induction motors better at wheelslip control (from https://electricalacademia.com/induction-motor/torque-speed-characteristics-induction-motor/ ):

(diagram here)
'T' = rotor torque, 'Sync speed' = the rotating electromagnetic field frequency = supply frequency, 'Rated speed' = Sync speed less the 'motor slip'. If the axle slips, the motor rotational speed increases rapidly and the torque will decrease rapidly, thus self-correcting the slip.

Note that the diagram shown is only valid for a single voltage and flux. In dynamic operation, these 2 terms in general would not be a constant.