Hello Guest it is October 22, 2019, 07:42:01 AM

Author Topic: Eccentric machining based on rotarty probing data?  (Read 1107 times)

0 Members and 1 Guest are viewing this topic.

Offline nicad

  •  40 40
    • View Profile
Eccentric machining based on rotarty probing data?
« on: March 22, 2017, 10:24:49 AM »
I'm not sure where to start with but I need to start somewhere! :)
The need: I'm chucking a tube in a lathe with C-axis control. The tube is round on the OD and the ID, but the ID can be off-center by up to 0.01".
I'm needing to grind a short section of the OD to be concentric with the ID. I need this process to be automated as each tube is slightly different and I'll be doing hundreds.

I'm envisioning some procedure where a probe is run into the tube, then over to check the ID at angle 0, then rotate the C-axis to angle 90, probe again, then angle 180, etc.. to get 4 readings to know the center of the ID.
Then the fun part-- still automated within G-code or macro or brains-- slowly rotate the C-axis while grinding the OD true to the ID by moving the X-axis back and forth in a sine wave to "follow" the ID bore based on the probing data. The offset in the Y-axis as it rotates is not critical as it's not off far enough to make enough difference.

A good example I've found is this: https://www.youtube.com/watch?v=FpP7iTKuWpw
I'll be grinding so the spindle C-axis rpm can be as slow as it needs as the X-axis keeps up.

Any ideas how to accomplish this??

Offline Graham Waterworth

  • *
  •  1,890 1,890
  • Yorkshire Dales, England
    • View Profile
Re: Eccentric machining based on rotarty probing data?
« Reply #1 on: March 30, 2017, 06:10:06 PM »
Forget chucking it, put it between centres and then the ID runs true and you correct the OD

Without engineers the world stops

Offline nicad

  •  40 40
    • View Profile
Re: Eccentric machining based on rotarty probing data?
« Reply #2 on: April 02, 2017, 05:00:25 PM »
I wish.. but I'm feeding in a 90" long tube, trimming the end, grinding 1" length of OD to be concentric to the ID, doing some other processing and then cutting about 6" length and starting over.
I figured out the steps and know it can be done using trig math functions in Mach. Brief outline:
-Probe at 90* increments to find the center of the ID.
-OD center is already known (CL of spindle) so run trig math to figure out the angle of the vector between the two centers.
-Rotate the spindle to the angle of the vector so that the offset is at the max in the X axis.
-Run a for loop counting off each degree for C (or whatever resolution needed) and command a move for X to move as a Cos function of the command degree and initial measured vector length.

However, I've decided to approach it from a hardware level and have designed a floating 3-jaw pneumatic chuck. I'll run a mounted ball bearing up into the ID of the tube, feed it over while spinning until it forces the tube to run true on the ID, then clamp the chuck. Will be much faster overall and eliminates potential probe failure and any X-axis backlash issue. ;)