[Retheion]

Hi

I'm trying to size up my expectations and also what to buy for my AMAT25LV CNC project, and I'm prepearing an excel sheet with calculations for all kinds of stuff related to loads, friction, speeds, times, forces, acceleration and inertia.
Though I do not fully understand how to use all the numbers, I have gotten some numbers out. I understand that if these numbers are correct, they are still the best of what can be done, and it's wise not to expect the machine to perform this way.
Basically I'm asking if anyone has experience with this stuff, and if this looks correct.

X-axis for example:
I'm calculating with a 0,353Nm brushed DC servo motor. Coupled via a 2:5 ratio pulley and belt arrangement. A 700mm long 1605 ballscrew, 200kg of total weight on the table 0,2 friction coefficient from dovetails. 0,00003kg/m^2 inertia on the motor. 0,0000354kg/m^2 inertia for the ballscrew (I'm excluding pulleys and belt at the moment). For the motor I have expected 3750rpm, from the datasheet it has a no load speed of 4700rpm, and the general knowledge seems to be that the motor can handle about 80% of that speed while still maintaining 80% of the torque.

This setup gives me 125mm/s for full speed.
About 500N in linear force from the servo motor via the pulley.
Which in turn gives me an acceleration time with a 200kg load of 0,05seconds, and a distance of 3,125mm. This is without the inertia of the motor and ballscrew, but as far as I can tell from my numbers just the drive assembly inertia would give me about 2100rev/s^2 of angular acceleration. Full ballscrew revolutions of 1500rpm within 0,012seconds!
All these calculations are done with the constant torque of the motor, and this all seems too good to be true, is it?

-Marius

[sajurcaju]

I strongly recommend building some simple model of your axis, complete with 200kg weight. You will learn a lot from this.
I did not dig into the calculations.

Steve

[Retheion]

Thanks for the reply, that's a nice tip. However I'm trying to do the math so that I wont have to buy any of the parts until I'm pretty confident in the calculations and the setup. I'll get a better feel of everything once I get the mill, still waiting for the shipment.

-Marius

[RCaffin]

Seriously doubtful about the use of dovetails. Linear Bearing rails would be far, far better.

I suspect your accelerations numbers are a long way off. You may find that to get those figures you will need to stick 50 A into the motor for a second. The motor might (or might not) take it, but the driver sure won't. Ah well, it will accelerate as fast as it does.

Cheers
Roger

[Retheion]

The linear bearings might come in the future. Now I'm not trying to argue. But if the numbers are calculated somewhat correct, and the motor needs more amps to pull it off, it could use more right? The times are from using the constant torque and amp values, so It should have some spare torque right? Up to the peak value ofcourse.
Anyways, you are correct, it will accelerate as fast as it does, and probably there are alot of factors I haven't accountef for. Even if I get a third of those accelerations, I'd be a happy man :-)


Sent from my iPhone using Tapatalk

[RCaffin]

Oh sure, all of the above.
It's just that I would find building with linear bearing rails a lot easier than trying to machine up dovetails - especially long large ones on the X axis.

Cheers
Roger