Low Inertia Motor 2

[FeX32]

Hey Guys,

I'm working on a project where I have determined I have tight requirements on inertia. I cannot reveal too much about the project, but I can say that I am driving a bunch of motors from one main motor. The main motor provides a strong torque and large inertia to force the system into a certain state.
Now, my issue is the sizing the smaller motors that are being driven by the system. These motors will need to provide a sinusoidal torque that I can control regardless of what speed the system is rotating at (0-200Hz).
The system will go through a theoretically calculated and simulated resonance. This is why I need the motors to be at an specific inertia.
I am a little concerned about the motor capabilities. For example, is it reasonable to ask a motor which is spun (by means of a shaft connected to another motor) to provide a sinus torque that sometimes opposes the spinning torque and sometimes adds to it? (at a frequency of up to 200Hz) I have simulated everything theoretically, and mathematically it functions as predicted.
So I would like some advice on selecting a motor based on for example:
Inertia: 2 Kg-cm^2
Torque: 5Nm nominal with 20Nm peak (up to 3000 rpm)
Speed: 0 - 3000 rpm.

Any advice or comments is appreciated.

Cheers,

 

[FeX32]

Btw, closest I found was this:

http://www.baldor.com/products/servomotors/n_series/bsm_nseries_spec.asp?Catalognumber=BSM80N-275AF

But torque output it a 7Nm to low.

Cheers,

 

[mikekilroy]

To answer your basic 'can this be done' question, absolutely.

without looking far, I see Kollmorgen has the motor you want. just a quick once over search found B-206-B with 6.62/20 nm cont/peak, 2.5kg-cm2, 2800rpm. You look closer and you will find what you want in those specs:

http://www.kollmorgen.com

But a bigger issue is your simulation may be flawed based on your description, and those motor specs wont get you want you think anyway. If you mean to have this motor go 0-3000rpm, back thru 0rpm to -3000rpm then back thru 0 - all in 5msec (200hz) on sinewave profile, and repeat forever, you will need well in excess of 50nm peak, and the continuous of ANY sinewave torque is .707*peak, so over 35nm continuous.

I cannot see how you would simulate a sinewave and expect continuous of 5 and peak of 20; even if your speed range on the sinewave was NOT 3000rpm peak-peak, the continuous WILL be .707 x peak. So if 20nm is correct (from not running +/-3000rpm sine), then your continuous is 14nm.

LAstly, you may consider electronic gearing (electronic line shaft) instead of the mechanical scheme to follow a master motor through likely a complex differential gearbox.... Most servo drives will run electronic line shaft from a simple master encoder input for no additional cost....

http://www.KilroyWasHere<dot>com

 

[FeX32]

Thank you for your reply Mike! Much appreciated.

If you mean to have this motor go 0-3000rpm, back thru 0rpm to -3000rpm then back thru 0 - all in 5msec (200hz) on sinewave profile, and repeat forever, you will need well in excess of 50nm peak, and the continuous of ANY sinewave torque is .707*peak, so over 35nm continuous.

How did you compute the 50Nm? Are you considering that the motor must accelerate up to these speeds by itself?
I only need the motor to produce a torque while it is being accelerated by another motor that is connected to it.
Sort of like a generator, but also with a precise torque profile. Is this possible?

I'm not sure I understand you perfectly. The type of torque profile I am looking for the motor to generate as a function of its own rotation (rotated by means of another motor only in the positive direction) is a profile like this:

https://dl.dropboxusercontent.com/u/18923918/t_q.png

It's max is 20 and mean is 5 Nm. And it will be forcibly rotated by another motor from 0 to 3000rpm. My main concern at first was the ability of the motor to change torque quickly, which considering the plot it must do at 200Hz if it is forced to spin at 3000rpm.

Please let me know any thoughts, and if my motor spec is reasonable.

Thank you

 

[FeX32]

Oh now I think you meant that the motor must accelerate from 0 to 3000 rpm then brake back to 0 in 5ms? I don't need these capabilities.
I believe I only need a torque control. It will be sped up by the other motor and slowly ramped up from 0 to 3000 rpm.
While this is being done I only want that torque profile generated by this smaller motor, regardless of speed. And it will not be required to accelerate itself up to and speed.
I hope this is reasonable.

Thank you!

 

[mikekilroy]

ok, given you are not going to share details of the app for whatever reasons, I will accept your comments at face value. So it appears the task is to believe you have a infinite power source motor that rotates your motor anywhere from 0 to 3000rpm, and irregardless of what torque this smaller motor puts out, the speed does not vary. OK, so be it.

So all you want is a servo capable of outputing that torque profile anywhere from 0 to 3000rpm.

I see continuous motor rating required as 5nm+/-15nm sine profile so 5nm+30nm*.707= 26nm

Absolute Peak torque output is 20nm though.

Even though your driving motor has infinite power, zero impedance, no losses, you still have a reason to have a motor with 2kg-cm2 max. OK, we use that as a given to not question also.

Lastly, you ask if you can do this torque profile on a servo motor of this possible rating. The answer is yes since you have no requirement to actually change speed or position, only output this torque profile.

A typical servo drive of this size today will have a current loop bandwidth of 1500hz A GOOD one will be up 3000hz plus. So your 200hz response is no problem.

If on the other hand there is ANY spec requirement for veloity or position changing along with this profile, you likely will NOT be able to make it happen.

If you confirm the 26 nm continuous requirement along with 2kg-cm2 spec then there is a small chance that it can be done with a very special servo; you would have to get involved with a company who specializes in this kind of motor to get a real quote. As such a company, I say you have not given enough information yet for a quote.

http://www.KilroyWasHere<dot>com

 

[FeX32]

Hi Mike,

Thanks again for your response. Info is much appreciated.

For IP issues I would rather not reveal exactly what it is I am doing on the forum.
However, if you are willing to help more I can provide much more information via email.
If you consent, I can contact you via your company email if that's alright?

I can confirm these specs you mentioned. The inertia is the most important requirement. This is why I am having trouble finding a motor. There are some that are close to what I want, but no cigar yet. I was thinking if it may be possible to provide a larger voltage bus to run a motor that has the required inertia but lesser rated torque, to possibly provide the torque I need for a short period of time. Say a couple of minutes.
Thanks again,

 

[FeX32]

"5nm+/-15nm sine profile so 5nm+30nm*.707= 26nm"

I'm trying to understand this. Correct me if I am wrong, I see 20Nm*.707 +5 = 19Nm Continuous.
Which is interesting since the true peak is 20 Nm, which I thought is usually 4* continuous.
Thanks,

 

[itsmoked]

FeX32 I recommend you do contact Mike. He's a wealth of info and support. In the interest of full disclosure, if it's not immediately obvious from his website, he is a Kollmorgen rep.

He recently helped me thru two very complicated motion projects.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[FeX32]

Thanks Keith.
It does look like Mike can help me out greatly. It seems that I may need a custom motor if possible.
Otherwise I may have to redesign or compromise torque or a little inertia. Resonance is key in my setup.