PDA

View Full Version : Relating to Academics



nickm85
01-13-2005, 07:14 PM
I'm a mechanical engineering student trying to tie a cnc router into something academic so that I can build it for an independent study and get credit, any ideas? Apparently I can' just build something for the sake of building it, which I think is the main reson for building anything! So far I've come up with measuring the strain on the rails but that's about it. Are there any common problems that people have that could quantitatively measured and analyzed? Thanks for any suggestions...

Al_The_Man
01-13-2005, 07:38 PM
I guess it ties in with the mechanical engineering of a CNC machine, that is the calculation of motor sizing and any reduction required based on the desired acceleration/deceleration of the machine in order to size the motor based on the industry recommended practice of sizing the servo motor so that the inertia to load ratio is less than to 10:1 or less than 3:1 for a stepper, to become familiar with this concept there are two links that may help http://www.electromate.com/technicalsupport/?c=kollmorgansoftware
and register on this site and download the AC servo handbook at www.motiononline.com
hope this helps.
Al

TinkerDJ
01-13-2005, 11:29 PM
I'm an electrical engineering student and with my project. I am building a stepper motor driver. For full credit it has to be pic controlled. My biggest thing is not so much actually making a working design and model so much as having to explain the working of each part and how it relates to what we have taken so far. Like Al said wouldn't a good idea be to build and write about the torque and innertia invovled as the machine moves. Also when i was upgrading before going into engineering i got an A on a report i did simply on the vectors involved in the operation of the cnc machine. That was more about how the program gcode is translated into movement. If you can get your prof to agree then you may be able to get some funding or at the very least use it to get free parts from suppliers.

OCNC
01-13-2005, 11:56 PM
You could study the relative efficiencies of different mechanical drive systems, specifically acme lead screw vs. ball lead screw vs. rack and pinion vs. toothed belt (tiiming belt). Calculate torque losses, accel/decel limitations and determine speed limiting parameters. Also include cost benefit analysis and maintainence issues.

Chris

lerman
01-14-2005, 02:52 AM
I'm a mechanical engineering student trying to tie a cnc router into something academic so that I can build it for an independent study and get credit, any ideas? Apparently I can' just build something for the sake of building it, which I think is the main reson for building anything! So far I've come up with measuring the strain on the rails but that's about it. Are there any common problems that people have that could quantitatively measured and analyzed? Thanks for any suggestions...

A question recently arose on this site about the advisability of using a linear encoder on an axis or a rotary encoder on a leadscrew axis versus a rotary encoder on a servo. It was suggested that backlash between the servo and the encoder might cause unstable or "clunky" behavior (my words).

So, I would suggest:

1 -- Analyze the behavior of systems using (a) a rotary encoder on the servo, (b) a linear encoder on the axis, AND (d) a combination of both. Presumably, you should be able to refer to existing literature for (a), at least.

2 -- Build a mechanical system of at least one axis with adjustable backlash and both a rotary encoder on the servo and a linear encoder on the axis. Instrument the system so as to measure both its static and dynamic behavior.

3 -- Write software to drive the mechanical system. Tune it so as to optimize its behavior in all three cases listed in 1, above.

Document and publish your findings. Publish the software and make it available as open source to this community.

So, what do you think? Can I con you into doing my work? :-)

Ken

arvidb
01-14-2005, 09:12 AM
For the rotary/linear encoder report, make a difference between systems that have a separate speed sensor (e.g. tachometer) connected to the motor as a part of the servo, and systems that is using the encoder for both speed and position feedback.

Arvid

turmite
01-15-2005, 01:34 AM
You could study the relative efficiencies of different mechanical drive systems, specifically acme lead screw vs. ball lead screw vs. rack and pinion vs. toothed belt (tiiming belt). Calculate torque losses, accel/decel limitations and determine speed limiting parameters. Also include cost benefit analysis and maintainence issues.

Chris


Chris this is an excellent idea since he likes to build for the sake of building. Using your idea he can now has to build at least three different machines for a side by side comparison. How cool is that!

Mike

nickm85
01-15-2005, 09:28 AM
Thanks for the help. I like the idea for comparing efficiencies, especially since I'm not much on writing software. I'll have to see if I can get something approved. Thanks again