IGES Problems Eliminated

The report from PC-DMIS says the part is oversized

PC DMIS report

But PC-DMIS "nominal" points compared to the CAD model
by our software from inside the CAD system show something else

"Nominal" points from PC DMIS

How does the part actually look?

Actual points from CMM

Best fit results. Actual part is smaller.

After best fit

Profile tolerance 0.100 mm

Units - whiskers are deviation x 10

green whiskers - in tolerance
green curve - nominal geometry

red whiskers - missing material
red curve - lower spec limit

blue whiskers - extra material
blue curve - upper spec limit


The Part:	Small (25.0 mm diam) die insert.
The Task:	Inspect the profile of the part with 0.100 mm tolerance. This company used IGES file to inspect this part in PC-DMIS. The first picture represents comparison between the "nominal" points generated by PC-DMIS and the nominal CAD model. We used our own SPV to evaluate how far the "nominal" points are from the CAD surface inside Unigraphics®. The second image shows the results as reported by the third party software. Then we checked and best-fitted the actual center-of-probe points to the CAD model inside Unigraphics®.
The Results:	PC-DMIS reports the part being oversized despite the fact that it was recut number of times. In reality the part is undersized, with a mean of -0.037 mm after best fit and 34 (out of 179) points below the lower tolerance of -0.050 mm. The real reason is that the CMM software did not interpret the IGES file properly and all "nominal" hits where generated off the actual surface anywhere from -0.058 mm all the way down to -0.111 mm. On the other hand the NC cutter path was generated inside UG. Is IGES any good? Sometimes, but when there is a problem you wouldn't even know.
Highlights:	Eliminate loss of accuracy and reduce scarp by using integrated inspection.