torque

[PFF]

Hey Guys,

Been asked to measure torque on a small electric motor. Any suggestions of some sort of fixture I could throw together to do the job? I'd like to be able to measure the torque at different levels of temperature, current and voltage fluctuations. Thought of using sensors etc. but haven't tried this before. Any suggestions would be appreciated.

 

[racookpe1978]

Dead rotation? Or while turning?

While turning at variable speeds?

How big a motor is max size?

 

[iop995]

What kind of electric motor / size?

 

[electricpete]

1 - There are ways to estimate torque with electrical measurements, along with a model of the motor. I assume you're not interested in those.

2 - What are you using to create the load? It may have means to infer the torque. Many motor shops use Dyno for load test... from it's name you can tell is measures torque.

3 - Another approach (which I've never seen done but certainly should work): consider that the electromagnetic torque is equal/opposite on rotor stator... and transferred from stator through the motors feet. Therefore you should be able to deduce torque using load cells under the motor feet. Example calculation:
Motor off, all 4 feet see 25 pounds each. That is motor weight.
Motor on, left pair of feet see 20 pounds each, right pair of feet see 30 pounds each.
Distance between feet is D
Each foot saw change of 5 pounds with moment are about centerline of D/2
Torque = (5+5+5+5)*D/2 = 10*(D/2)

Two notes about apporach 3
1 - Be careful that an individual foot is not capable of transmitting a moment (vs force) through a bolted connection because that would be an error. One way to do that would be to loosen (but not remove, for safety) the motor feet bolts. And only loosen them after the motor is started. Think through the safety aspects for yourself - don't rely on me.
2 - of course you might be able to put a bar under both left feet and another under both right feet so you only need 2 load cells.

=====================================
(2B)+(2B)' ?

 

[waross]

If you are checking the motor without load, spin an aluminum disk. Place a DC electromagnet in proximity to the disk. Vary the torque load by varying the current trough the magnet coil. Measure the reaction on the DC magnet with a load cell or other device of your choice.
Google Prony brake or "de Prony" brake.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[electricpete]

My item 2 - the load (fan or whatever) should be mounted separately from the motor, torque otherwise load casing torque can complicate the picture.

Cool idea from Bill. Very elegant. I might try that myself one of these days in my garage. For homemade version could perhaps also use permanent magnet and vary the distance.

=====================================
(2B)+(2B)' ?

 

[electricpete]

no edit button. sorry

My item 2

should've been

My item 3

=====================================
(2B)+(2B)' ?

 

[waross]

Hi Pete; Consider also changing the radius from the center of the shaft.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Skogsgurra]

There are quite a few principles that are being used for torque measurement. It is a big application in motor testing labs, where SRM, Atlas-Copco, Penta, Scania, SAAB, AVL and Volvo (the ones I work with) use rigs with up to megawatt brakes.
Mostly, a coupling with either magnetic, optic or strain gauge measuring devices are used. Those machines can be very accurate with errors down in the .1 percent range.
For extreme precision, Siemens used to produce so-called Pendelmaschine where the brake (DC or AC motor) was arranged in bearings so that the reaction forces from the rotating motor could be measured statically as a force between motor stator and the support. I do not see any new Pendelmaschine any more. With pendelmaschine, there were problems with highly dynamic load changes (elasticity in brake starts oscillations).
For medium accuracy - one or two percent - it is easier to just measure how much power the brake machine generates. There are lots of losses to be compensated for, but once such a set-up has been calibrated, it is very reliable and easy to use. A DC generator is very easi to use and calibrate, but an AC induction generator is a lot cheaper and sturdier. I did measurements on a little motor this autumn and because of heavy EMI, we tried to use 50 Hz power measurements. We had quite usable results also with that simple technique and I am now testing an Arduino in such an application. I will look for a write-up showing the principles and what results we got. Sorry, I have to "unidentify" it first. Will take some time.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

I show the short version here:

http://www.gke.org/rapporter/files/...ransducer on PWM frequency inverter input.pdf

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[PFF]

Thanks for all the advise! Much appreciated.