Torque Required to move a System 1

[spggodd]

Morning!

I have a problem I can't crack..

I am trying to work out the torque required to get a system moving from its static position.
The system consists of various gearboxes.

Practically I could fit a bar to one of the shafts and apply load until it turns and work it out this way but for now the system isn't build so I need to way of doing this by calculation.

Any ideas?

Thanks

 

[tygerdawg]

The strictly analytical method is to go through the entire train of gears, calculate all of the rotational inertias of the gears and shafts and bearings and couplings, select your required time to accelerate those rotational inertias up to a desired speed, and apply the "Torque = Inertia*RotationalAcceleration" formula. Too much work.

The practical method is to analyze your final load (rotational & translational motions) for peak torque. Peak torque = T-accel + T-stiction + T-friction + T-gravity + T-runningTorque + T-aDozenOtherThings. Then work it back through the gear ratios of your gear train for the required input torque. Add a factor of safety for fudge ("torque is cheap, use plenty of it"), then get on with more important things.

I always recommend that one should websearch & download the "Smart Motion Cheatsheet" in PDF form, also to peruse the engineering section of any of the available gearmotor manufacturer's catalogs. SEW-Eurodrive's English-language Engineering Guide (from their German site, not the US site) used to be one of the most rigorous analysis methods I've ever seen. Not sure this can be downloaded anymore, though. Could check out the others like Dodge, Falk, Nord, etc.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[cloa]

Doesn't friction come into this too.

 

[Jboggs]

My gut tells me you can do all the calculations you want, and in fact should, but the bottom line is that its going to take whatever it takes because you cannot account for EVERY variable factor in your calculations. You're going to have to do some real world tests. If you can't build the whole system, build a low-cost prototype for testing purposes only. Or maybe build just a small part of the system and extrapolate your test results to forecast the characterisitcs of the whole system. Or see if you can find some existing similar mechanism somewhere nearby that you can personally investigate.

 

[spggodd]

Tygerdawg, this looks good I will have a go at doin it in the way you describe. I will also check out those references.

JBoggs:

The motor I'm using is quite powerful and will easily overcome the torque of the system and will become quite unmanageable, what I need to do is work out the torque required to start the current system so that I can select a suitable dynamometer to increase the torque of the system to a controllable level. Hope that makes sense?!

 

[zekeman]

"Practically I could fit a bar to one of the shafts and apply load until it turns and work it out this way but for now the system isn't build so I need to way of doing this by calculation."


I'm puzzled.

If the system isn't built, why do you need all those gearboxes?

Maybe one may suffice.

 

[spggodd]

The system includes 3 gearboxes, 2 are the test specimens and the other is simply to change the direction 90 degrees.