- 1

- 0

running machine without workpiece, reads up to about 100 gram cm tooling imbalance

(frequent tooling changes- so some inherent imbalance will always exist)

adding workpiece reads up to 2000 gcm as resultant vector of tooling and workpiece.

subtracting tooling vector from resultant basically gives 'workpiece vector'- or so I thought...as tooling is in different plane than workpiece, the 'resultant' we are measuring has 3D effects coming into play- but with single plane (single point radial movement) feedback cannot resolve any dual plane issues...

I'm a PLC/CNC guy, and have really gotten in over my head on this project- just getting the trig formulas in for a three point circle took like 8 pages of 'ladder logic' (PLC's arent as number friendly as CAD) best repeatability I can achieve is about 150 gcm when rotating workpiece to various angles (without fixture comp, was closer to 500 gcm- oddly even more than worst tooling imbalance)

trying to visualize the forces at play here- hoping a physics/math guru can shed some light on a hypothetical set of numbers:

Note the height above the bearings cannot be detected by the machine, but I know roughly what height center of mass for tooling and workpiece heights are...

**picture a disk that measures 100 gcm imbalance at 90 degree position, located 10" above the machine bearings.**

picture another disk that measures 2000 gcm imbalance at 180 degree position located 5" above machine bearings

combine the two forces, what is the resultant vector in purely radial direction (should work out as PLC currently calculating)

picture another disk that measures 2000 gcm imbalance at 180 degree position located 5" above machine bearings

combine the two forces, what is the resultant vector in purely radial direction (should work out as PLC currently calculating)

I'd like to plug these numbers in and confirm my logic is correct(vectors drawn in CAD say its OK, but maybe I'm extracting them the wrong way?)

then the big question-

**what is the true 3dimensional vector, and what kind of formula would you use to find it.**

(thinking about adding dual plane feedback, but no idea AT ALL as to how to interpret the data).