how is direction determined on AC traction motors?

[daveklepper]

The modern AC motors used in railway service are non-sychronous "induction" motors.  They do not have slip rings or commutators and do operate on the induction principle by use of the hystoresis effect.  In a synchronous induction motor, the rotating aluminum or copper bars are parallel to the motor shaft.  In railway motors, the bars are slanted.  There is always slip between the rotating field and the rotating bars, with the slip greater with increasing load.  The rotating field is slow at start-up, so the frequency of the AC current is low.

[Erik_Mag]

gregc

what are salient pole motors, what is their benefit over non-salient pole motors and are AC traction motors that type of motor?

If your familiar with the small DC motors used in model trains, the armature windings are on what would be called salient poles in a AC synchronous motor or alternator. These are almost invariably used on machines with greater than 4 poles, but are rarely used on large machines running at 1800 or 3600 RPM in a 60Hz world due to the high centrifugal force (centripetal acceleration). The salient pole itself will provide some torque without any field excitation, but the torque is not a simple sine of the 'angle' between the field and the center of the pole. Angle is in quotes as the electrical 'angle' is the physical angle multiplied by the number of pole pairs on the machine (60 Hz, 3600rpm 1 pole pair, 1800rpm 2 pole pairs, 1200rpm 3 pole pairs...).

In a synchronous machine, the rotor spins at the same speed and same direction as the rotating field, with the angle between the rotor varying with the torque produced. If the rotor position lags the rotating field, the machine will be acting as a motor generating mechanical work, if the rotor is leading the field, the machine will be acting as a generator absorbing mechanical work and generating electrical power.

[rdamon]

For those of us who have lost the attention span for reading :)

[Flintlock76]

Strange.  I thought directional control on AC motors was handled like this.

https://www.lionelsupport.com/media/servicedocuments/70-8659-250.pdf

[Overmod]

gregc

1) the rotating magnetic field is cutting thru the rotor "bars" inducing a current.   presumably that current is flowing in the opposite direction on a bar on the opposite side of the rotor as that bar moves in the opposite direction thru the same magnetic field...

Not necessarily.  As I understand the construction, all the bars in the rotor are massively shorted together at both ends.  The rotating field can be thought of as essentially annular, with the active part in the air gap between stator pole faces and rotor, so there is very little magnetic effect from the field affecting a given bar on the bars diametrically across from it.  I do not have a reference expressly proving this but have always assumed that the 'return current' in the rotor is through the bars and structure not affected by close magnetic-field induction.  This should be easy to confirm with reference to a good text.

2) that current creates a magnetic field in the rotor.

Technically true, but the only part of the rotor you want to consider at this point is the particular part of the bar structure in which the current of interest, and hence the associated magnetic field of interest, is being induced.  Only the magnetic fields associated with the currents actually being induced in excited bars will contribute to the motor action...

...  the rotor magnetic field rotates both due to the changing stator field (is this correct?) as well as the rotation of the rotor the rotor magnetic field interacts with the rotating stator magnetic field generating a torque.    (????)

(????) is so right!  Remember I said look only at the bars?  Think of the rotor ONLY as the magnetic fields produced by the bars.  You could almost treat this like the fields in magnetic gears, with the  'lobes' of the rotating magnetic field then interacting with them.  Do not bother with some "rotor magnetic field".  There are bars being pushed, and these are mechanically constrained to go in a circle that keeps them in the active part of the stator rotating magnetic field, and the mechanical constraint produces the motor torque...

so my difficulty is seeing how the rotating fields both induce a current and remains (oppositely) aligned with the rotor magnetic field to produce torque.

This is one of those counterintuitive things like how charges in cloud and ground produce lightning.  As it happens the effect of the rotating field is not to produce the counterpart of back EMF but to induce magnetism (albeit indirectly) which then interacts with the generating field itself.  This becomes important because... 

3) and then there's the rotor turning inside both magnetic fields producing BEMF resisting the flow of current.

As Lita Ford sang: STOP RIGHT THERE!  This ain't a DC motor; the power is coming FROM the rotating field AGAINST the bars, and there is no back EMF against a fixed field as in a DC machine.  If you look at the permanent-magnet analogue you won't see back EMF in the rotor; you may see induced magnetism (and thence current) in the stator windings but that doesn't affect the motor action, only the potential difference across the stator windings that is driving the current producing the rotating magnetic field.  Someone from Bell Labs would be sketching all the vectors involved by this point; that can't be me.  There have to be "Bell-Labs grade" videos by now that illustrate this without dumbing it down... 

[/quote] again, while this may be easier (for some) to abstractly visualize at speed, i believe a clear explanation of what happens when the motor is barely turning at startup would explain a lot![/quote]Well, look.  You have a rotating magnetic field going around a ring of transverse, end-shorted aluminum bars.  As the field strength increases at a particular bar, a current is induced in it.  This produces a magnetic field, and that magnetic field interacts with the field that induced it.

We assume that the ring of bars is made so it is free to turn inside the stator if released.  However many bars are being affected at a particular moment... it is a multiple of three for three-phase applied current... are all being impelled around in the same direction, so unsurprisingly there is a net torque whether or not the rotor is physically turning.

Now we come to something more interesting: 

perhaps at 0 rpm, a rotating stator magnetic field induces a rotating magnetic field in the rotor which remain appropriately aligned to produce torque but then if the frequency of the AC current thru the stator field depends on the rotor rpm (preventing wheel slip), then that frequency at 0 rpm can't be arbitrary, otherwise, wouldn't the torque be be changing direction?

See what happens when you mistake the forest for the trees?

There has to be continuous movement for the magnetic field to induce a current.  So if the rotating magnetic field were rotating precisely at rotor speed, there is nothing to generate current in those little bars, and therefore no magnetic action.

Now, if the stator field isn't rotating at all, to be 'synchronous' with the locked rotor... its phase angle could be set to produce 'best torque' on the rotor if the rotor were magnetic. We could do this just fine with permanent magnets, and some AC motors do.  We could do it with DC through oriented windings in the rotor, and some AC motors can.  But in this class of motor, the one rotating field is doing it all and therefore there has to be both continuous relative motion, to keep the bars magnetically energized, and enough phase-angle correlation to produce 'best' torque on the excited rotor as it happens to be oriented when it became locked.  This is going to involve some calculable lower developed torque than a 'perfect' magnetic synchronism (I can't say how it's proportional to the slip because I'm not a motor engineer with reference materials, but there are people here who will give you the applicable formulae).

It would naively appear that you would modulate the rotating field at the slowest possible rate because of the I2R losses every time a given stator pole is energized and the faster we do that in nonsuperconductive windings the more irrecoverable heating we waste our power on.  Now if we modulate the inverter power to where we 'know' interaction is strongest (and there are ways we can assess this other than by reference to a rotary encoder) we can increase the overall power transfer to the rotor over time, but at the potential expense of higher torque ripple...

 

[gregc]

my understanding is there are two or three different things occuring simultaneously.

1) the rotating magnetic field is cutting thru the rotor "bars" inducing a current.   presumably that current is flowing in the opposite direction on a bar on the opposite side of the rotor as that bar moves in the opposite direction thru the same magnetic field

2) that current creates a magnetic field in the rotor.   the rotor magnetic field rotates both due to the changing stator field (is this correct?) as well as the rotation of the rotor

the rotor magnetic field interacts with the rotating stator magnetic field generating a torque.    (????)

so my difficulting is seeing how the rotating fields both induce a current and remains (oppositely) aligned with the rotor magnetic field to produce torque.

3) and then there's the rotor turning inside both magnetic fields producing BEMF resisting the flow of current.

 

again, while this may be easier (for some) to abstractly visualize at speed, i believe a clear explanation of what happens when the motor is barely turning at startup would explain a lot!

perhaps at 0 rpm, a rotating stator magnetic field induces a rotating magnetic field in the rotor which remain appropriately aligned to produce torque

but then if the frequency of the AC current thru the stator field depends on the rotor rpm (preventing wheel slip), then that frequency at 0 rpm can't be arbitrary, otherwise, wouldn't the torque be be changing direction?

 

[Overmod]

timz

I've never seen the answer to that -- how many RPMs does the field rotate to get maximum TE at a standstill?

The answer is not that immediate, as the idea is to modulate the change 'at the bar' to keep current flowing but not so much that the bar melts, then have the resultant that pulls on each bar STATIONARY at closest approach, just as the locked rotor is stationary.  In practice you can't separate the two functions.

In practice you will likely fire the inverters so the field rotates around the slip differential speed, and presumably design the motor cooling so neither the stator current nor the induced rotor-bar circuit paths overheat.  I would expect the 'computer' would carefully watch this and derate as necessary based on what the motor sensors tell it.

I would be surprised if Dave Goding doesn't know the answer in detail for EMDs, including any special provisions when you use one inverter per truck.  He will certainly know where to go for the answer if he needs more data to give a full enough reply.

[timz]

No one knows how fast the field rotates, as the locomotive starts?

[Overmod]

timz

(Overmod says the field doesn't rotate, which sounds like a semantic quibble. If nothing rotates, effectively at least, how does the motor know which way to start turning?)

He seemed to be having trouble with the idea of a rotating field that continually 'rotated' faster than the rotor it was driving, so I tried a different way to look at the magnetic physics ... including an analogy to marquee lights which he 'got'.

In a sense the 'rotating' field is only a resultant; it does not detach from the stator and spin on its own dragging the rotor along.  But it is an attractive and intuitive way to explain the effect.  I have no difficulty imagining how a rotating field produces slip, either, but gregc seemed to, so I tried another way.

Obviously as with Balaam's-ass metastability, the rotor will tend to turn in the direction it interacts with the stator magnetic fields more strongly ... moment by moment.  This is not something like an unshaded-pole clock motor that needs to be turning to 'know' which way to run... or, really, that needs to have special start windings cut in.

Didn't Tesla lecture fairly lucidly on how this kind of polyphase  induction motor worked?

[timz]

gregc

what's the frequency and phase(!) of an AC traction motor when starting at 0.000 rpm?

I've never seen the answer to that -- how many RPMs does the field rotate to get maximum TE at a standstill?

(Overmod says the field doesn't rotate, which sounds like a semantic quibble. If nothing rotates, effectively at least, how does the motor know which way to start turning?)

[Overmod]

gregc

I thought that's what I said??

Well, then you understand it.  Next!  

I believe I understood the concept of a rotating field...

Then you have no difficulty seeing how that field induces current in the shorted stator bars.  And you have a right hand if the effects of the rotating field on that induction might seem confusing...

Is the current induced in the rotor DC or AC?

you did not answer this question and i believe it would be very insightful.

I did answer it... I told you to think about it.  Since you say you understand the rotating field, it is trivial to map its effect in the rotor bars, and only slightly less trivial to do so with the rotor in relative motion.  

[gregc]

Overmod

An important thing to realize, though, is that this is an 'automatic action'; the slip changes in proportion to 'the speed difference that induces whatever magnetic field to produce output torque balancing load at speed'.

i thought that's what i said.   ??

Overmod

The field itself is not physically 'rotating'; it is the resultant of the fields induced by the three phases in their respective windings, and only has the net effect of handed rotation as seen by a rotor pole.

i believe i understood the concept of a rotating field

each pole on the stator is connected to one of the 3 phases of power.   so the magentic field builds on one pole, the next in sequence and then decreases.   yes like the marqui lights.   with 3 phases there is a definite direction

greg

Is the current induced in the rotor DC or AC?

you did not answer this question and i believe it would be very insightful.

 

Overmod

Field changes even locked-rotor, inducing a current in the bar and hence a time-dependent magnetic field.

you couldn't use a few more technical terms.

 

 

again, those Bell Lab discussions made things simple to understand

[Overmod]

gregc

yes.   I understand that slip will increase as the load on the motor increases resulting in more torque and conversely decrease as load decreases.

An important thing to realize, though, is that this is an 'automatic action'; the slip changes in proportion to 'the speed difference that induces whatever magnetic field to produce output torque balancing load at speed'.  That's a lot of word salad, but you see the point that the action is self-correcting based only on relative change between the 'rotating field' and the magnetic field generated by the current that was induced in the shorted (or resistively loaded) rotor element by that same changing magnetic field ... in the house that Jack built. 

I'm reading that the magnetic field rotating at a faster rpm than the rotor.

Step back a bit to understand what is happening here. The field itself is not physically 'rotating'; it is the resultant of the fields induced by the three phases in their respective windings, and only has the net effect of handed rotation as seen by a rotor pole.  The alternation of stator current will produce induction in an adjacent rotor pole, and the resultant magnetic field interaction between stator and pole will produce attraction or repulsion and, since the rotor pole bar is built into a rotor, this will produce torque on the rotor and this will be 'handed' in the direction the overall stator field acts.

Now, the stator fields are changing in the same general sense that marquee lights induce perception to 'follow' them, even though each light considered by itself is just turning on or off.  So if you had pre-magnetized rotor poles they would be attracted to the resultant, and try to 'lock in' to where the attraction pulling them around is greatest.  But the shorted poles of these motors aren't pre-magnetized; in fact they're designed not to have residual magnetism or hysteresis.  So something has to create the magnetic field in the rotor poles first, before any mutual attraction or repulsion can even occur, and for that to be done by the same three AC fields there has to be some mutual change.

Note that Erik brought up providing some of the rotor magnetism via an external current, for example by taking the 'windings connected to resistance' in the prior example and running a current ... call it a DC current of appropriate polarity ... through it.  This will start niftily, but if it synchronizes to the resultant of the stator fields, there won't be any induced rotor current, and the whole of the magnetic interaction will depend upon how the winding produces and projects its magnetic field, more power of a different type and more consumption and resistive heating, etc. -- and you will still be limited by synchronous speed as far as the continuous torque is concerned.

Is the current induced in the rotor DC or AC?

Think about what the rotor bar 'sees' over time in its frame of reference to see the range of answers.  The take-home is that the magnetic field generated by the bar is entirely determined, in one of these motors, by stator-field action as 'seen' by the rotor.

again, discussing the startup situation would be very interesting.

Field changes even locked-rotor, inducing a current in the bar and hence a time-dependent magnetic field.  This will interact with the resultant of the stator field to produce a starting torque.  Note that the induced current is far from constant, though, so it won't just 'sit there and buzz' as the AC stator fields interact in space.

[gregc]

Overmod

It is important to recognize that the slip is continuous;

continuous or constant?

yes.   i understand that slip will increase as the load on the motor increases resulting in more torque and conversely decrease as load decreases.

i'm reading that the magnetic field rotating at a faster rpm than the rotor.

is the current induced in the rotor DC or AC?

again, discussing the startup situation would be very interesting.

[Overmod]

gregc

i've tried to say that I believe the rpm of the rotating magnetic field and the rotor must be the same, but that the rotating field is slightly ahead of the rotor phase.

It is important to recognize that the slip is continuous; the rotor always runs that little bit slower than 'synchrony' with the rotating field visualized as a thing.  Think of it as the amount of "turning power" that is consumed in keeping the rotor bars developing their magnetic field. They cannot BE magnetic without current, and that current is only produced when magnetic-field strength at the bar is changing...

[gregc]

JC UPTON

 S = (Ns - Nr)/Ns    Ns=> synchronous speed Nr=> running speed

i'm not going to understand something quoted out of a text book without some pre-text

i've tried to say that I believe the rpm of the rotating magnetic field and the rotor must be the same, but that the rotating field is slightly ahead of the rotor phase.

if this isn't correct, what are relative the phases of the magnetic fields in the rotor and stator?

[gregc]

Overmod

Remember that a magnetic field only induces current when it is changing.

that's a very good point.   how does it work at startup when the rotor is barely moving?   do the stator fields get pulsed?

Overmod

The slip rings are a convenience to get the induced-current path outside the rotating element to where variable resistance to the induced current can be imposed.

that's helpful.   so torque can be controlled/reduced by adding resistance in the rotor current path instead of reducing field current.   is this a benefit of salient pole design?

Overmod

Note that were the rotating field and the rotor poles to run in sync, there could be no induced current in the rotor; it would then cease to show magnetism and hence torque.

yes.  my understanding is this is why wheel slip is less, if not prevented with AC traction motors.

 

now i'm interested in startup, when a rotor is still and just starting to move

[Overmod]

JC UPTON

Slip definition:    S = (Ns - Nr)/Ns    Ns=> synchronous speed Nr=> running speed              or    S% = (Ns - Nr)*100/Ns

It might be more helpful to him to start with the equation that shows the magnetic field induction as a proportion of relative speed between 'field rotation' and rotor face (or, more appropriate, effective rotor-bar) rotation...

(e.g. E = BL (Vsyn - Vr) or variant as appropriate; Vr being the same 'thing' as "Vm" defined as motion of the conductor relative to the field and measured accordingly...)

so he can see what's producing the numbers in the slip formula.  He is an EE so most of the principles are well known to him.

(This reminds me all of a sudden of a guy at Lockheed, a brilliant engineer, to whom eddy currents were so magical but inscrutable that he attempted to get a classified patent on them...)

[JC UPTON]

AC traction motors are INDUCTION motors.  As such they rely on the slip FREQUENCY to INDUCE a current in the rotor bars, which produces a magnetic field.  This magnetic field IS SLOWER than the rotating magnetic field of the 3 phase AC winding.  Slip definition:

   S = (Ns - Nr)/Ns    Ns=> synchronous speed Nr=> running speed 

            or

   S% = (Ns - Nr)*100/Ns

[Overmod]

Remember that a magnetic field only induces current when it is changing. In a typical 'synchronous' motor two things are happening: the magnetic field change in the stator induces a current in elements of the rotor (and this current develops a magnetic field as it flows) and the changing magnetic field in the stator then reacts with the current and magnetic field it has induced.  The 'slip' is necessary because it represents the relative motion that induces the currents in the rotor elements.

In the wound-rotor induction motor the windings act just like the shorted bars in the 'regular' type as far as having current (and flux) induced in them.  The slip rings are a convenience to get the induced-current path outside the rotating element to where variable resistance to the induced current can be imposed.  The higher the resistance to current flow, the less actual current flow occurs in induction, therefore the less magnetic field to interact with, hence starting control.

Note that were the rotating field and the rotor poles to run in sync, there could be no induced current in the rotor; it would then cease to show magnetism and hence torque.  Of course that would cause it to start slowing down... more so if it is driving some sort of load... and when it slows there starts to be mutual change again, and current is induced in the rotor again, and the magnetic fields can react again...

In a permanently-magnetized rotor, the 'reacting' magnetic field is always present, hence such a motor could be operated at a speed synchronous to the field the rotor 'sees' -- since relative motion to induce rotor magnetism is not necessary.  

(Of course the rotating permanent magnet structure tries to induce back EMF in the stator architecture, so to get faster rotation you need field weakening, paradoxical as that might seem.)

You may find interest in the explanation of introductory material on motor operation in this master's thesis:

http://www.diva-portal.org/smash/get/diva2:601079/FULLTEXT01.pdf

We need a Bell Labs style cafeteria meet to straighten out what the definition of salient poles in different motor design represents.  For many years I have touted the advantages of switched-reluctance motors, which can be thought of in a sense as stepper motors so excited as to produce continuous torque without locking into a step position -- you can see how complex some of the resulting interactions can be.  By now there are surely good explanations of how switched-reluctance motors are best driven; I'm not going to confuse you by babbling on a forum without diagrams and proper equations.  There were some highly interesting YouTube videos showing how motors of this type can produce the effect of smooth continuous rotation with highly unbalanced loads on the rotor, a fascinating thing to see if much of your knowledge of motor and drive kinetics is from experience and observation as mine is.

In my opinion once you have the complicated electronics to be able to run a switched-reluctance motor at high torque, it is an effective traction-motor candidate over what can be a remarkably high speed range.  Up until recently the electronics have been a cost barrier compared to more conventional 'inverter' synthesized-AC drives for synchronous polyphase induction motors.

There are railroad applications that use permanent-magnet traction motors. All the ones I took note of were in high-speed trains, and I do not remember the discussions regarding their specific advantages and disadvantages as cost-effective TMs although those were often discussed.  (As I recall these were all not like model motors with permanent-magnet field, by the way; that is a whole different discussion...)

[gregc]

one thing i miss from working at Bell Labs is how some people were able to explain what i thought were fairly complex concepts so clearly.   some lunch conversations were so much better than some college lectures.

one person would often start a discussion by saying "I don't know if you know this ..." to discuss basics without implying ignorance so that the listener could better understand the topic of discussion.

 

what are salient pole motors, what is their benefit over non-salient pole motors and are AC traction motors that type of motor?

[Erik_Mag]

Greg,

Having had a couple of power systems labs as well as an electric machinery lab...

The pulling out of synchronism is for a synchronous motor, where the torque for a non-salient pole machine is proportional to the cosine of the angle between where the motor is operating with respect to the rotating field and where the rotor produces zero torque. Large synchronous motors using have field windings on the rotor, fed though slip rings in the olden days and now an AC generator and rectifiers mounted on the rotor. The advantage of a controllable field excitation is being able to adjust power factor, along with increased field for when the machine is producing higher torque as either a motor or generator.

Induction motors make use of the slip between the rotating field and the bars on the rotor to induce currents in the rotor bars that interact with the rotating field to produce torque. I got a visceral demonstration of that when putting a thick wall aluminum tube between the pole faces of a 0.25T magnet. With the axis of the tube parallel with the magnetic field, I could move the tube up and down with little resistance. I got a great deal of resistance when trying to rotate the axis of the away from being parallel with the magnetic field due to the eddy currents excited in the aluminum tube. The faster I tried roatating the tube, the stronger the resistance to rotation.

The limiting factor for induction motor torque is where the combination of skin effect and inductance gangs up to limit the in-phase induced current. Thick copper rotor bars will develop maximum torque at a small slip, but very little torque at a large slip. High resistance rotors will generate large torque at high slip, but much less torque at low slip. This also leads to lower motor effiecincy.

A compromise was the wound rotor motor,  where the winding connections were brought out via slip rings. For starting, a high resistance would be put in series with the windings, which would then be cut out as the motor picked up speed. The starting resistances could also be used as a speed control, and this was the method used by the GN 3 phase electrics as well as the N&W and VGN phase converter locomotives.

[gregc]

1% slip (phase difference) is not the same as "1% faster" and of course not the same a wheel slip.   i think i remember that day in motors class where we used a strobe light to see the phase of the rotor and watched as sync broke when we exceeded the load limit.

and 1% ahead, presumable 3.6deg, in phase requires knowing the exact phase of the rotor, not just the rpm.

what's the frequency and phase(!) of an AC traction motor when starting at 0.000 rpm?

i'm guessing when starting, an AC traction motor is like a stepper motor.

[Overmod]

gregc

but presumably there's a phase (angular location) of the stator field and the armature that must be synchronized such that if the armature increases rpm, wheel slip, torque drops or even changes direction. i wonder how the phase of the armature is measured to adjust the field rotation, mechanically or electrically.

If you look up 'slip' in a discussion of synchronous three-phase motors you'll find an explanation.  The point at which the rotor spontaneously 'locks in sync' (I.e. runs as a synchronous motor with stable peak torque' is with about that 1% slip rather than perfect following.

This is with self-induced magnetism in the rotor, which is essentially passive; no active control of the rotor is required.

The speed control is regulated with respect to the feedback signal from rotor speed, not phase-angle change; if you were to measure the rate of 'rotation' of the effective stator field you would find it faster by the slip...

[gregc]

3-phase AC answers the direction question

the article said "The variable frequency drive creates a rotating magnetic field which spins about 1% faster than the motor is turning".    not sure they didn't mean it's a few degrees ahead of the rotor.

the amount of lag affects torque

but presumably there's a phase (angular location) of the stator field and the armature that must be synchronized such that if the armature increases rpm, wheel slip, torque drops or even changes direction.

i wonder how the phase of the armature is measured to adjust the field rotation, mechanically or electrically.

[Overmod]

gregc

needs to match phase of armature fields? so armature with commutator (not a squirrel cage type design) i thought a benefit of AC motors was no commutator

Why would following what is essentially a rotating magnetic field require any active current through the rotor?

Granted, if you don't see how magnetism is induced in the rotor 'bars' you may wonder how the rotating field acts on them, let alone produces very high torque.  But any good motor textbook will explain this... or ought to.

I take it you understand how the arrangement of stator windings produces the 'rotating field' effect.  A variant of this can be used for BLDC motor control, where the effective magnetic field producing rotary torque is controlled many times a second to be at the best angle between stator and rotor poles -- here the stator-induced field does not 'rotate' but is switched to remain effectively ahead of the (magnetic) rotor in the direction of rotation...

I suspect if you want a real headache, look up control algorithms for torque-producing switched-reluctance motors... but when this is done right, some of the real-world operating results can be frankly surprising.

[timz]

gregc

i thought a benefit of AC motors was no commutator

You were right.

[gregc]

daveklepper

By the computer-controlled sequence of the PHASING of the variable-frequency AC to the multiple field coils around the motor.

needs to match phase of armature fields?

so armature with commutator (not a squirrel cage type design)

i thought a benefit of AC motors was no commutator

[JC UPTON]

To expand on Dave’s explanation a little…

Rotation direction in all three phase AC motors is determined by the ELECTRICAL rotation.  If one swaps two (of the three) phases, that reverses the electrical rotation & hence the motor rotation.

Dave is correct about the inverter managing this function in inverter fed motors, which is true for all AC motors, not just AC traction motors.  Instead of physically reconnecting the power leads, the inverter software controls the power electronics to perform the reversal.

[daveklepper]

By the computer-controlled sequence of the PHASING of the variable-frequency AC to the multiple field coils around the motor.  The rotating armature consists of slanted aluminum or copper bars. The inverters that convert the rectified-to-DC electricity bsck to AC at variable frequencies produce different phases to each of the multiple field coiis, and the sequence deternines direction.

In an inside-out wheel-motor, the "armiture" is a fixed. non-rotating circle of field coils, and the rotating bars are on the outside.  All railroad locomotives (diesel and straight electric) and nearly all MUs use the first type of AC motor.  The wheel-motor is in use mainly in low-floor European trolley buses, dual-mode buses, battery buses, and airport plane-side buses and in several experimental ultra-light light-rail cars.