[iain kennedy]

Hi all,
I have a eumach 1000 machine with a heidenhain Itnc 530 controller. When I machine a square pocket, the X axis dimension is 0.04mm to 0.05mm smaller than the Y axis. We have just gone through calibration with a laser interferometer, and found the greatest error to be about 0.008mm. A similar problem occurs when machining a circular pocket. Any thoughts on what could be causing these errors? I tend to think this is a software issue as our calibration results are pretty good.
Thanks in advance,
Iain

[the_gentlegiant]

Hello iain,

What you have would be classified more as an error of form then of size.

This problem can be caused by any number of things, and some form of this error exists on just about any machine.

How rigid was your setup?
How fast was your feed rate?
What was your finish cut parameters?
Is the error the same along the whole length of the machined surface?
is the error the same when performed in different areas of the work envelope? Meaning is it localized to one area?
Is the error directly in line with the axis or at 45 degrees to the axis? 45 degrees to the axis is a sign of the machine being out of square.

A laser test is only good for setting backlash and discovering screw error compensation amounts. It does little to test your machine as an electro-mechanical system that it is. You would need to do Ball Bar testing to discover errors in squareness, servo lag, servo mismatch, all that kind of stuff. Any of these things can cause size and form errors. Incorrectly determined and set backlash compensations can also cause these errors.

If this is a vertical machining center you may have more axis flex in the X direction then Y. In the Y direction your forces will be directed at the vertical column. Especially in a negative Y move. I think machines are naturally stiffer in the Y direction. Sort of like on a CNC lathe running a right hand tool facing down, which causes the cutting forces to lift the turret off the ways. I believe you tend to get better rigidity running left hand tools right side up to drive the cutting forces into the turret and therefor into the ways.

If you have a rigid setup and substantial enough tool, are keeping your finish cut to the minimum and at a relatively slow speed to minimize flex, and are still having this issue, then at least you've eliminated poor test parameters from the equation. As with any machine, the faster and harder you run it,the more perfectly balanced the entire system as a whole needs to be to get near perfect results.

Get a Ball Bar Test done if you really want to see what's going on with your machine as a whole system.

[iain kennedy]

Hi the_gentlegiant,
Thank you for your reply, yes, a ball bar test is next on the list but finding someone with one near me is proving to be difficult.
regards,
Iain

[iain kennedy]

The finish cut depth was 0.05mm, the feed rate 0.07 per tooth in aluminium.
Regards,
Iain

[the_gentlegiant]

The finish cut depth was 0.05mm, the feed rate 0.07 per tooth in aluminium.
Regards,
Iain

Hi iain,

That's more of a scratch then a cut. How are you even measuring something so shallow? I though you might have cut maybe 13mm deep with perhaps 0.07 per tooth with a 13mm end mill or something.

Unless you're leaving out a ton of other details, that finish cut doesn't make sense at all.

Sorry to hear you can't find a Ballbar tester in your area. Figured you probably knew about the process.

[iain kennedy]

The aluminium plate is 20mm thick, the square pocket is 100mm x 75mm through. The 0.05mm was the depth into the wall, so, basically machining the pocket from 100mm x 75mm, to 100.1mm x 75.1mm
Hope that explains things a little better.
Regards,
Iain

[the_gentlegiant]

That's better.