Torque of wound rotor motor with vfd drive 3

[Barry1961]

I have very limited information on the application, so I apologize in advance.

A company has a 1960's era 480vac 50hp 1200rpm wound rotor motor that has been recently re-wound. The windings are supposed to be inverter duty. This motor drives a set of horizontal plate rolls and will be run forward and reverse starting slow and ramping to full speed in .5 to 1 second. It will run at full speed for 3-5 seconds then will be manually reversed.

The plan is to short the rotor and run it with a VFD, Hitachi SJ300 sensorless vector. Since this will be starting under full load there is some concern about starting torque/current. I would guess they could increase the low speed/voltage torque boost since they will never run for any length of time at low speed.

I have never done this myself and don't know if there are any changes to starting torque when running a WRM this way.

Thanks in advance for any tips.

Barry1961

 

[DickDV]

In a start/stop/reverse application like this, I would prefer to leave the lowest resistors in the rotor circuits. If you short the rotor circuits, you will usually end up with a motor with such "tightness" that the currents required for speed changes become unmanageable.

If you find that the slip with one resistor is too high, you can usually have your motor shop people short the rotor in the field. So, you have the option of going all the way if you try the one resistor arrangement and it doesn't work out.

I'd hedge my bet on this one and at least try it.

 

[aolalde]

High torque with reduced inrush current is one of the main advantages of a wound rotor induction motor (WRIM). By short circuiting the rotor winding, this advantage is lost and the motor becomes like a squirrel cage induction motor or SCIM with very low rotor resistance. Now the motor has high inrush current and low starting torque.

Ramping this motor with a low Volts/Hertz start could improve the performance as compared to starting it with 60 HZ, but still the torque could be too low.

The times and reversions described look quite stringent for any motor application and will be more severe as the value of the load inertia increases.

My feeling is that this application as described will have lots of problems.

 

[Barry1961]

Could you give me a ball park guess on percent reduction in starting torque? I hate to ask with limited information but just to get an idea...50%....10%??? Also, how would leaving in the last resistor affect the start torque, again just a ball park?

Barry1961

 

[DickDV]

Leaving the last resistor in the circuit would increase the starting torque and reduce the starting current.

The motor at all speeds would have higher slip from no load to full load but the sensorless vector drive can compensate for that if set up properly.

Even with the last resistor in the circuit, I would size the drive so it has 200% of motor FLA for 5 seconds. If the rotor was shorted, you would probably need 300% for 5 seconds.

 

[rbulsara]

I am with aolalde.

It really depends what starting torque is needed. If the original wound rotor was installed to minimize inrush because of high starting torque requirement, you will have problems. If the wound rotor was primarily there to vary speed, not necessarily for starting high torque, it may work.

 

[sreid]

A question about some of the concerns that have been raised. If one were starting with a clean sheet of paper today wouldn't the design use a squirrel cage induction motor and a vector VFD? What would make shorting the wound rotor windings so the motor looks like a squirrel cage motor "wrong"? Inrush current and and low speed torque boost can be controlled by the acceleration ramps and/or the V/F curves. What am I missing?

Also, doesn't shorting the rotor mean you get rid of the slip rings/brush gear?

 

[Barry1961]

We are going with a fixed resistor and the next larger VFD for increased current. The customer can reduce the torque required by decreasing the amount of bend per pass.

The motor shop that re-wound the motor said they had converted WRM's before and suggested a 8 ohm resistor for this paticular motor. Time will tell.

Thanks, Barry1961

 

[jraef]

For sreid,
"What would make shorting the wound rotor windings so the motor looks like a squirrel cage motor "wrong"? Inrush current and and low speed torque boost can be controlled by the acceleration ramps and/or the V/F curves. What am I missing?"

What you are missing is that a Wound Rotor Induction Motor (WRIM) is not the same as a Squirrel Cage Induction Motor (SCIM) to begin with. They end up accomplishing the same tasks, but get there in different ways. In a SCIM, the current vs torque profile is pre-determined by the winding and rotor designs, that is why you have Design A, B, C, D etc. Each design has a different ratio of current vs torque output. With a WRIM, since the rotor windings are brought out through the slip rings, the current vs torque production ratio is infinately variable by virtue of the resistance applied. When you short the rotor, you are not automatically creating an induction motor with a Design B torque pattern, it is more like a Design A. So trying to start it means that you will get less torque production per amp applied at lower speeds compared to using some ammount of rotor resistance. Having a VFD on it can help keep that from being as much of a problem, but you would still not be getting the most out of that motor until you were at or near full speed without seriously oversizing the VFD. It is cheaper to put in some resistors.

Another problem (which should be noted by the OP as well), is that not all WRIMs have rotors designed for the high currents that would occurr when starting with the rotor shorted. If they were designed to be started with resistors and only run shorted at full speed, the rotor current rating may be too low for starting shorted. Again, the VFD will help with this, but may not be able to generate enough torque at low speeds to get moving if it is in current limit.


"Also, doesn't shorting the rotor mean you get rid of the slip rings/brush gear?"

Only if you do it in the rotor itself. Otherwise, you can short them outside.

"Venditori de oleum-vipera non vigere excordis populi"

 

[sreid]

Jaref,

I generally agree with all you are saying but are not most of the motor chacteristics you state based on a constant voltage-constant frequency power source? Things change dramatically when a motor is driven by a variable voltage-variable frequency vector drive.

One can run a NEMA A design at full rated torque at zero speed using a vector drive(you can't in actuallity because there is no fan cooling at zero speed but you can for brief periods).

Can you get 200% of continuous torque at zero speed for short periods? Yes if the Vector drive is sized to do that. The vector drive can hold the rotor flux constant (direct current) and produce torque through additional stator current (quadrature current). The limitation is magnetic saturation in the motor.

Most of a motor's torque speed characteristics are determined by the rotor bar design (for off the line operation). So if a shorted rotor WRIM curves are, say, identical to a SCIM curves, from an electrical black box perspective, they are the same.

Is there a heating difference between the two rotor? If they are both 50 HP rated motors at base speed, there can't be much difference based on rotor heating and motor efficiency.

I'm not trying to be a jerk here. Being a servomotor/servo amplifier designer, I really want to know the reality and if incorrect default assumptioms are being made, to correct them.

 

[DickDV]

sreid, I think that jraef has provided wise analysis of wr motors vs. squirrel cage motors.

What maybe isn't so obvious is that rotor resistance affects the slope of the torque-speed curve in the near-synchronous range as well as changing the starting characteristics.

A wound rotor motor with shorted rotor circuits will likely have too little slip (the torque-speed curve is too near vertical). This makes the current under overload conditions rise too rapidly causing problems for the drive, especially if the drive sizing is a bit skimpy.

Better to leave a little resistance in the rotor, suffer with a little more slip (a good vector drive will compensate for the slip anyway) and have overload currents that aren't so severe.

As I said in my first post above, a shorted rotor produces an excessively "tight" motor which is too difficult to manage properly.

jraef's comment about a NEMA Design A motor is right to the point--it has very little slip but the starting inrush is huge--similar but not as severe as a shorted wound rotor.

 

[aolalde]

With all due respect Sreid, you are contradicting all basic principles of induction motors operation.
Your statement: “Can you get 200% of continuous torque at zero speed for short periods? Yes if the Vector drive is sized to do that. The vector drive can hold the rotor flux constant (direct current) and produce torque through additional stator current (quadrature current). The limitation is magnetic saturation in the motor”.

First if the flux is constant you have no induction in the rotor winding. (“rotor flux constant direct current” ? that is a contradiction). The induced voltage in the rotor
e= -N d? / dt

The force developed in a conductor is the cross vectorial product of the air gap flux density and the rotor current. F = I * L x B
F = force experienced by a rotor conductor, a vector. (N, Kgf, Lbf )
I = conductor carried current, a scalar (Amperes).
L = a vector of magnitude of the conductor length in the direction of the current.
B = the magnetic flux density whose direction is given by the flux lines, a vector (T, kLines/sqin, gauss).

As you can see there is no additional stator current that produces force in the rotor conductors.

The only thing that the most sofisticated “vector drive” can do is to feed a voltage at certain frequency to the motor leads. I agree that by making a combination of V/HZ and timing a drive can get more convenient motor responses, but the motor basic operation phenomena does not change.

 

[sreid]

Aolalde,

I am using the terms "direct current" and "quadrature current" in Vector Drive terminology (d-q control) where direct current is the standard phrase for "The stator current that fluxes the rotor" and quadrature current is "The stator current which produces torque."

 

[DickDV]

sreid, everywhere I've been, "the stator current that fluxes the rotor" has been called "magnetizing current".

"the stator current which produces torque" has been called "magnetizing current".

If those terms work in your environment, it might be less confusing to use them instead.

On the other hand, this board has worldwide exposure and it may well be that "direct" and "quadrature" currents are meaningful where you are. I'm not saying either is right. Simply trying to avoid misunderstanding.

 

[edison123]

jraef's neat post gets a star.

 

[sreid]

Well, I can certainly see the confusion in terms and to be absolutely correct I should have said "Direct Axis Current" and "Quadrature Axis Current" but this is often shortened to direct current and quadrature current when talking about field oriented control. A reference on the web;

http://www.analog.com/library/analogDialogue/archives/34-06/dsp/DSP.pdf

See second paragraph on the second page.

 

[mc5w]

If you are doing repetitive reversing you do need 1 or more rotor resistors. Repetitive reversing is NOT a good application for a VFD.

The minimum frequency for your motor could very easily be 15 Hertz. A lot of people forget that at low frequencies induced rotor voltage is low and the only rotor design is design D because of the low voltage. If there is a way to program your VFD to act a a 15 Hertz soft start then as a VFD then do things that way.

I one time set up a steel coil straightener so that at low demand the sonar that sensed the steel coil loop would do on-off control causing the VFD to act as a 15 Hertz soft start. At higher demand the VFD had a flat maximum of 90 Hertz so that I would not get a zero loop gain condition. You DO NOT EVER want to reach absolute maximum torque or speed because when you do your loop gains all drop to zero.

 

[Marke]

I think that I must be missing something here.

I see a number of recommendations that resistance be left in the rotor circuit. Why?

Adding resistance in the rotor circuit will alter the speed torque curve of the motor. It will increase the slip frequency at which maximum torque occurs.

The motor is to be used with a vector drive, so the motor is never going to "start" as in operate at high slip. The operation is always going to be under low slip conditions (less than the slip frequency of maximum torque)

As the drive is a vector drive, it will be able to track the motor loading. Increasing the slip of the motor by adding resistance will reduce the speed accuracy of the motor relative to the demand speed and will reduce the efficiency of the system. I do not believe that it will improve the braking operation required for rapid reversal of the motor.

As the drive is a sensorless vector drive, there may need to be some voltage boost at very low speed.
I beleive that the important point for this application is the rapid reversal will require either a regenerative drive (active front end) or adequate braking resistors to be fitted.

When using an VFD with a motor, the motor should never be "started". It is always driven from zero speed. Very low slip motors operate well on drives, but not on starters. I think that we need to separate the two modes of operation.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[mc5w]

But you may have to size your drive to continuously carry the locked rotor current of the motor or 50% of locked rotor in a frequent reversing application without rotor resistance.

Also, your motor is NOT intended to operate without resistance when starting or plugging. Your motor also is NOT really meant to operate off of a VFD.

If you really think about it, what you really need is a servo drive and motor with a rather hefty dynamic braking resistor.