Flux Vector Drive - Synchronous speed

[W@ttson]

Hey guys,

Just wanted to get a confirmation of the following understanding of the capabilities of a flux vector drive utilizing Closed Loop Vector control.

So with a FVD with encoder feedback you can get full torque at zero RPM with a speed range of something in the range of 1500:1. Really good stuff. The drive utilizes a decoupled magnetizing current and torque producing current to produce the necessary torque required by the load. Though the theoretical capability curve shows a flat line for Torque vs speed from 0RPM to base speed is it really true that the flux vector drive can output FLT at synchronous speed. With direct across the line starting / control the induction motor requires a certain amount of slip and the design of the machine (A,B,C,D) will dictate the torque speed curve. The induction motor requires some slip to produce a torque (rotor magnetic field is trying to catch up to the stator magnetic field and thus produces are torque to accomplish this).

In summary, can a 900RPM Design B squirrel cage motor actually provide FLT at 900RPM when coupled with a FVD which utilizes closed loop feed back or is this just theoretical and in actuality field performance dictates that some slip is still required.

Hope jraef gets to see this question

 

[Jraef]

It's true that you must have slip in an AC induction motor to produce torque, and it is also true that you begin to lose torque as you go over base speed, because you run out of voltage to keep the V/Hz ratio correct. However, in the NEMA design world, motors are expected to deliver "nominal" rated power at +-10% of design voltage. So a motor designed for use on a 480V system is designed around a "utilization" voltage rating of 460V. A VFD rectifies the incoming AC to DC and uses that DC to create the PWM output that, to the motor, appears to be an AC sine wave. The RMS voltage of that sine wave cannot be higher than what the DC bus allows, meaning you can't get more voltage out than voltage in. But if you program the VFD to put out 460V, and the line source was 480V, technically you have a tiny bit more to play with.

So if your motor has 3% slip, and you want to run it at full synchronous speed, you are actually going to need to TELL that motor to run at 103% speed. That will, ostensibly, reduce your V/Hz ratio by roughly 3%, resulting in a proportionate loss of torque. But in reality, the rated torque of that motor is SUPPOSED to be available down to -10% of line voltage, in which case the V/Hz ratio would be much less than what you will see.

Bottom line, at just 3% over base frequency, I would not be concerned over any potential significant loss of motor capacity if it is critical for you to run it at the base synchronous speed. And I doubt you would need full Flux Vector Control with an encoder feedback to attain it, Sensorless Vector Control will likely be fine (although you don't say what this is for).

 

[W@ttson]

It's true that you must have slip in an AC induction motor to produce torque, and it is also true that you begin to lose torque as you go over base speed, because you run out of voltage to keep the V/Hz ratio correct. However, in the NEMA design world, motors are expected to deliver "nominal" rated power at +-10% of design voltage. So a motor designed for use on a 480V system is designed around a "utilization" voltage rating of 460V. A VFD rectifies the incoming AC to DC and uses that DC to create the PWM output that, to the motor, appears to be an AC sine wave. The RMS voltage of that sine wave cannot be higher than what the DC bus allows, meaning you can't get more voltage out than voltage in. But if you program the VFD to put out 460V, and the line source was 480V, technically you have a tiny bit more to play with.

So if your motor has 3% slip, and you want to run it at full synchronous speed, you are actually going to need to TELL that motor to run at 103% speed. That will, ostensibly, reduce your V/Hz ratio by roughly 3%, resulting in a proportionate loss of torque. But in reality, the rated torque of that motor is SUPPOSED to be available down to -10% of line voltage, in which case the V/Hz ratio would be much less than what you will see.

Bottom line, at just 3% over base frequency, I would not be concerned over any potential significant loss of motor capacity if it is critical for you to run it at the base synchronous speed. And I doubt you would need full Flux Vector Control with an encoder feedback to attain it, Sensorless Vector Control will likely be fine (although you don't say what this is for).

The application is for something similar to hoist and crane. So the governing characteristic is the starting torque which we will need to be able to have the FLT at 0 RPM. Usually if we are not using a Flux vector drive we strive to use a NEMA Design D with a 900RPM motor. The running speed is usually around 860RPM due to slip or if a wound rotor motor with a thyristor drive is used the slip can go up to 25%. For this one particular job we are using a FVD and we have a total time elapsed constraint and a different portion of the system will be taking a bit longer than anticipated. I got the question of "well if theoretically with an FVD you can get FLT from 0 to base speed then could we run the motor at 900RPM instead of the 860RPM that we have set just to be consistent with the non VFD operation.

I believe that past base speed the torque drops off at a [1/w^2] right?

 

[Jraef]

The application is for something similar to hoist and crane. So the governing characteristic is the starting torque which we will need to be able to have the FLT at 0 RPM. Usually if we are not using a Flux vector drive we strive to use a NEMA Design D with a 900RPM motor. The running speed is usually around 860RPM due to slip or if a wound rotor motor with a thyristor drive is used the slip can go up to 25%. For this one particular job we are using a FVD and we have a total time elapsed constraint and a different portion of the system will be taking a bit longer than anticipated. I got the question of "well if theoretically with an FVD you can get FLT from 0 to base speed then could we run the motor at 900RPM instead of the 860RPM that we have set just to be consistent with the non VFD operation.

I believe that past base speed the torque drops off at a [1/w^2] right?

Well, the PEAK torque capability drops by the square of the voltage change, but the CONTINUOUS torque varies proportionally. PEAK torque in this case is the Break Down Torque (BDT) of the motor. If it is a Design B motor, that's going to be between 200-220% of FLT, compared to LRT of 150-160%. So if you are designing for LRT, the FVD can actually provide you with BDT from that motor, which is starting out MORE than you are designing for. So although technically your BDT might be reduced to 94% of what it COULD have been, it's still going to be far more than you need.

 

[W@ttson]

Well, the PEAK torque capability drops by the square of the voltage change, but the CONTINUOUS torque varies proportionally. PEAK torque in this case is the Break Down Torque (BDT) of the motor. If it is a Design B motor, that's going to be between 200-220% of FLT, compared to LRT of 150-160%. So if you are designing for LRT, the FVD can actually provide you with BDT from that motor, which is starting out MORE than you are designing for. So although technically your BDT might be reduced to 94% of what it COULD have been, it's still going to be far more than you need.

Yes, we look to the drives overload capabilities at starting. Typically, in the heavy duty mode, the drive will be able to output 150% FLT for 60 Seconds or 200% FLT for 3 Seconds. We typically need anywhere between 150% to 180%. After the initial overload we slide into anywhere between 60% of FLT to FLT.

Sorry I got my constant horsepower region confused with the field weakening region. At in the Constant horsepower region the torque drops off in the following relationship T_rated*(w_rated/w_rotor) and in the field weakening region its a T_rated*(w_rated/w_rotor)^2...

So just to summarize the take away from all of this. Induction motors always need a slip, even with a flux vector drive, however, with the flux vector drive the slip isn't exactly indicative of the amount of torque output since the torque producing current can be varied independently of the magnetizing current.

I think I need to look back into some literature on slip frequency and how it ties into things.

Thank you for the responses.

 

[Jraef]

...
So just to summarize the take away from all of this. Induction motors always need a slip, even with a flux vector drive, however, with the flux vector drive the slip isn't exactly indicative of the amount of torque output since the torque producing current can be varied independently of the magnetizing current. ...

There you go, you got it, although rather than saying "independently" (which is not actually inaccurate), I instead like to say "without affecting the magnetizing current." Although they are "independent" in that you have totally separate regulator algorithms in the drive, the magnetizing current is not one we play with much, we basically give it what it needs, no more. But the separation allows us to play with the Torque producing current component at will and not send the motor into saturation, which is what happens when you DON'T have the ability to separate them. That then means we can make the motor deliver BDT in an instant without worrying about putting the motor into immediate saturation and losing control. Fun stuff, although honestly, I don't get a chance to use it much any more. 99.9% of the applications out there don't need that sort of precision. Hoists though, yeah baby!

I did the hoists and cranes for the Boeing 777 plant in Everett, Washington when they built it in the 90s. 62 bridge cranes and 8 "carriers" that rode on them. The carriers are like upside down tanks, complete with turret (no cannon or armor though), each one with 4 x 40 ton hooks to lift and hold the aircraft. When we released the holding brakes, we were not allowed even 1mm of movement.

 

[Besoeker]

Hey guys,

Just wanted to get a confirmation of the following understanding of the capabilities of a flux vector drive utilizing Closed Loop Vector control.

So with a FVD with encoder feedback you can get full torque at zero RPM with a speed range of something in the range of 1500:1. Really good stuff. The drive utilizes a decoupled magnetizing current and torque producing current to produce the necessary torque required by the load. Though the theoretical capability curve shows a flat line for Torque vs speed from 0RPM to base speed is it really true that the flux vector drive can output FLT at synchronous speed. With direct across the line starting / control the induction motor requires a certain amount of slip and the design of the machine (A,B,C,D) will dictate the torque speed curve. The induction motor requires some slip to produce a torque (rotor magnetic field is trying to catch up to the stator magnetic field and thus produces are torque to accomplish this).

In summary, can a 900RPM Design B squirrel cage motor actually provide FLT at 900RPM when coupled with a FVD which utilizes closed loop feed back or is this just theoretical and in actuality field performance dictates that some slip is still required.

Hope jraef gets to see this question

Synchronous speed is not an especially useful term when it comes to variable speed drives i.e. it's not one fixed predetermined speed. If you supply your nominally 900 rpm motor with higher than mains frequency it can run faster than 900 rpm. That then puts it in what is sometimes referred to as the constant power region.

 

[W@ttson]

There you go, you got it, although rather than saying "independently" (which is not actually inaccurate), I instead like to say "without affecting the magnetizing current." Although they are "independent" in that you have totally separate regulator algorithms in the drive, the magnetizing current is not one we play with much, we basically give it what it needs, no more. But the separation allows us to play with the Torque producing current component at will and not send the motor into saturation, which is what happens when you DON'T have the ability to separate them. That then means we can make the motor deliver BDT in an instant without worrying about putting the motor into immediate saturation and losing control. Fun stuff, although honestly, I don't get a chance to use it much any more. 99.9% of the applications out there don't need that sort of precision. Hoists though, yeah baby!

I did the hoists and cranes for the Boeing 777 plant in Everett, Washington when they built it in the 90s. 62 bridge cranes and 8 "carriers" that rode on them. The carriers are like upside down tanks, complete with turret (no cannon or armor though), each one with 4 x 40 ton hooks to lift and hold the aircraft. When we released the holding brakes, we were not allowed even 1mm of movement.
View attachment 15368

That is really fascinating! Thank you for sharing that. Yeah when you are dealing with handling of such big things, there is no room for error!

I looked through my notes and got the equation for the slip relation if anyone is interested:

S*w_rated = (r_r/L_r)* (I_torque/I_mag), where

S= slip (unitless)
w_rated = rated speed in rad/s
r_r = rotor resistance
L_r = rotor inductance
I_torque = Torque producing current
I_mad = is the magnetizing current