Experimental Analysis of a Synchronous AC Motor and Fan Load

[CasualKilla]

Hi guys, I am looking to calculate the moment of inertia (J) and viscous friction (B) for a prototype motor connected to a fan load. The motor is powered with a constant voltage 3 phase supply and a torque sensor is connected in-between the motor and fan (figure 1). The torque sensor can give transient data of both torque and speed.

 

figure 1:

 

I have not tested the fan, but research seems to indicate the fan torque-speed curve will follow a quadratic curve like the one shown below.

 

figure 2:

 

I am assuming the bearing friction should have a fairly constant value B, however, the air resistance will cause an effective B that is dependent on the speed. By assuming the torque-speed to follow a perfect quadratic trend, I can convert the differential equation into a non-linear form with 3 unknown constants (eq1 -> eq4).

 

figure 3:

 

I have an idea to simply take 3 time points and on the start-up or shut-off transient response, and then solve equation 4. But I have concerns about the validity of equation 4 and the possibility of torque ripple from the motor.

 

I have 2 questions:

 

1) Is my model correct, or correct enough (this is engineering after-all).

2) How would I calculate these values experimentally with the given setup in figure 1.

Edited April 21, 2015 by CasualKilla

[studiot]

1) I will have to think about your model.

 

2) You don't calculate experimentally you measure.

 

If you want to measure friction and MOI or other parameters, you can do mechanically exactly the same as you would do to take measurements.

Add a load in parallel and measure the difference.

 

The mechanical load can be a flywheel of known properties.

The input can be measured and the time to run down from one speed to another, after removal of the drive, can be measured.

[Sensei]

figure 1:

Typical ammeter will not work with AC, but DC.

And will show 0 A, even though current is flowing (1/100sec one direction, 1/100sec other direction).

 

Check whether yours ammeter can work with AC, if you really need this data..

Edited April 22, 2015 by Sensei

[EdEarl]

 

wikipedia

 

A synchronous electric motor is an AC motor in which, at steady state,[1] the rotation of the shaft is synchronized with the frequency of the supply current; the rotation period is exactly equal to an integral number of AC cycles.

Speed variation as load changes indicates you do not have a synchronous motor. Synchronous motors change speed with a change in AC frequency. You may have an induction motor.

Edited April 22, 2015 by EdEarl

[Enthalpy]

Probably a squirrel cage motor, the standard choice for fans and blowers, as these need no big torque at slow speed.

 

Every (non broken) multimeter I've seen could measure AC currents.

 

Losses: don't suppose that bearings oppose a constant torque. They have viscous losses too, often as strong as the others. This results from grease and oil, which can also produce quadratic loss torque. In addition, motors have their own blower for cooling, which produces a quadratic torque too - probably not to be discerned from the load in your experimental task.

 

Torque ripples are small at a three-phase motor - as opposed to a permanent-magnet commutator motor, or a single-phase squirrel cage. It's a good reason to prefer three-phase motors. You can neglect them in the experiment. If you didn't want to, you could measure the speed over an integer number of shaft turns.

Edited April 22, 2015 by Enthalpy

[CasualKilla]

Probably a squirrel cage motor, the standard choice for fans and blowers, as these need no big torque at slow speed.

 

Every (non broken) multimeter I've seen could measure AC currents.

 

Losses: don't suppose that bearings oppose a constant torque. They have viscous losses too, often as strong as the others. This results from grease and oil, which can also produce quadratic loss torque. In addition, motors have their own blower for cooling, which produces a quadratic torque too - probably not to be discerned from the load in your experimental task.

 

Torque ripples are small at a three-phase motor - as opposed to a permanent-magnet commutator motor, or a single-phase squirrel cage. It's a good reason to prefer three-phase motors. You can neglect them in the experiment. If you didn't want to, you could measure the speed over an integer number of shaft turns.

In that case, would it be better to model the bearing friction torque as Tbearing = constant * (w^2) ?

 

I plan on doing a shut-off test, but I am curious, will there still be a reading on the torque sensor when there is no Motor applied torque? If so, what is the value of this torque reading, I have included a model below. We assume the shaft is turning at a initial constant speed w.

 

 

 

Edited April 26, 2015 by CasualKilla

[Enthalpy]

How you model bearing losses also depends on how much freedom you have on this project...

 

Depending on the conditions, their torque can depend mainly on w⁰, w¹ or w². This changes with the load, the speed and the kind of lubricant. Skf's website has nice explanations on that topic. The motor's cooling fan (if it has one) is also a main contribution to motor losses.

 

If you have this latitude, I'd suggest that you measure the motor's losses without the load, for instance through its unpowered deceleration, or by coupling a know inertia like a metal disk. The speed vs time curve will permit you to identify the w⁰, w¹ and w² contributions. The same operation but with the intended load will then separate what results from the load.

 

Though, school projects often make assumptions (like: motor loss torque as w⁰, fan torque as w²) for simplicity.

[imatfaal]

!

Moderator Note

 

Milbaduya

 

There is no need to copy and paste the words of another members post just to show your agreement. You can explicitly quote a post and express your opinions and add to it - or you can use the reputation system to plus rep the post. Or both. What you have done seems a bit like plagiarism - although I hope that was not the intention.

 

Thanks - no need to respond to this moderation; report the post if you are unhappy.