[step-manufacturing] Tool wear demonstration - discussions with NIST folks

[Martin Hardwick]

Hi All,
I met with a bunch of my NIST colleagues yesterday afternoon to discuss 
technical support for the tool wear demo at our upcoming June meeting. 
We have some interesting options.
At the Wednesday telecon, Larry Maggiano mentioned that one of his 
colleagues talked with NIST's Eric Stanfield about using one of our 
measurement systems to measure actual tool wear. I'm working with Eric 
to figure out how to do this with minimal turnaround.
We have a dynamometer and can hook it up to the machine tool in a couple 
of modes. One mode is to mount the part atop the sensor and continually 
measure forces. Another is to mount a sacrificial piece of the same 
metal as we're cutting atop the sensor, placed off to the side, and 
periodically take cuts while measuring forces for a short period.
The dynamometer is an "AC" device, meaning that the signal is 
proportional to the change in force, and a time integration is done in 
the support electronics to turn the change in force over time into a 
force measurement. The sensor is thus good for high-frequency, 
short-term measurements. This argues for the periodic sacrificial cut 
mode. Another argument for the periodic mode is that it's easier to set 
up. The part can be clamped onto the table as usual, without regard to 
the sensor cables etc.
I propose that we set up the demo so that we cut the boxy part as usual, 
and include periodic breaks from cutting, say after each setup, during 
which a simple well-defined cut is done while we measure forces.
Regardless of the mode, we still want to incorporate the force 
measurements into the software somehow. How do we plan on doing this? 
One way is to simply compare what forces are expected, based on a model 
(from Boeing?), and congratulate ourselves when they match. Another way 
is to bring the measured forces into a model for a learning-type 
application.
I think using the dyno will be simple, but coming up with a software 
interface to bring the force data into ST-Machine will be the hard part. 
NIST can write the DAQ software to get the forces and log them out into 
whatever format is convenient.
What do you all think?
--Fred
Frederick M. Proctor
Group Leader, Control Systems Group
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.steptools.com/pipermail/step-manufacturing/attachments/20100122/03ebd1a3/attachment.htm>