Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Error compensation in Machine tools, Software

Status
Not open for further replies.

simasa

Mechanical
Sep 25, 2007
30
0
0
DE

I am planning to implement a dynamic error compensation system for a CNC Milling Machine to improve its accuracy. Machine is controlled by Sinumeric 840 controller. The Real 3D-trajectory of the tool will be measures on line.


Tasks may include:
Conversion of NC-code (currently only G01) into a 3D
trajectory (reference tool trajectory)
Comparing the reference tool trajectory with online
measured trajectory
Calculating the path deviations (predicted errors)
Recursive software compensation of predicted errors
while generating a modified NC Code

These tasks would be performed in a software way on a separate real time computing unit (Matlab based), and modified path will be fed to the machine controller through an effective interface.

Any hint's/guideline concerning software aspects of the tasks would be really appreciable. Any contacts related to existing software packages or fictional libraries would also be welcome.

Thanks in advance
simasa
 
Replies continue below

Recommended for you

Much of what you are trying to compensate out is from machine deflection created by load. These loads include positioning of the machines mass, cutting tool length and diameter, and spindle torque.

Some of these deflections are non-linear and some become vibrations. When we looked at doing this in the late 1990’s we thought the best method was to treat it as an axis feedback error rather than modifying the commanded position. Our approach was going to use lasers measuring through mirrors as the encoder to the control. The machine was to be a traveling column design with the lasers originating from the work piece table to account for yaw in any direction. The yaw control would require at least a 5 axis control, precise rigid high speed positioning and may have required additional PID loop work.

Ed Danzer
 
1) Put the machine in a 69 degree F temperature controlled room.

2) Use temperature controlled Ball Screws (coolant through the ball screw center). Better to use linear motors. Water cool the forcers.

3) Use high resolution, high accuracy linear scales for position feedback (Think Heidenhain zerodur Glass Scales). Use the "Flash" Interpolator Boards. Still use rotary encoders on the drive motors for velocity feedback.

4) "Survey" all axes for pitch, roll, yaw and translation slide errors (or use really high accuracy slides). Still survey and compensate for secondary errors. (see Renishaw or Heidenhain for Laser Machine Survey Tools).

5) Use PID control with 100% velocity and acceleration Feed Forward. Use digital Bi-Quad Filters to get you through the first resonant mechanical frequency.

6) Flood cool the cutting work piece using temperature controlled cutting fluid.
 
Status
Not open for further replies.
Back
Top