ACTUALS NOT UPDATING

[Matthew D. Hoedeman]

Yep, once again, V3.7MR3, opened up the program that does an iterative alignment (manual & DCC) and ran the entire thing. NONE of the actuals updated for the measurements I made. After it was done running, I had to type in the feature ID for the first dimension then they all updated to the actuals. Nice work B&S/Wilcox/Hexagon. Thank the stars I only have to do these once in a blue moon!

[John Kugler]

Had the same type of issue yesterday. I was running a program parts at a time and it didn't update the dimensions like it should until I ran it through at one time again.

[Matthew D. Hoedeman]

Only happens to me when the iterative alignment is involved. Makes ya wonder what could be the problem, eh? I have never had it happen with regular alignments.

[tking]

I don't trust alignments made by the system for me. I choose to do my own, deciding the order and features. The system may get it right but I can't defend my work unless I KNOW how I got there. Nice catch, Matt.

TK

[rebeldude]

I haven't had that problem yet, but it is something we should all keep an eye on.

[Matthew D. Hoedeman]

Here's some more info that will muddy the water even more. I wrote down a bunch of numbers (deviations), one set for each type of point the program checks. It checks VECTOR, EDGE, and CIRCLE (all auto features). Here are the results:

VECTOR POINT: All deviations reported from the run matched the actuals from the run

CIRCLES: All deviations reported from the run matched the actuals from the run

EDGE POINTS: The "T" deviation matched the actuals but NOT the "S" deviation. Unfortunately, these EDGE points were on a flat, square surface (0,0,1) and there is NO place on the part to check any with a 3-D surface so I can not investigae it. The XY deviations for the edge point matched the actuals but NOT the Z deviations (surface vector direction), it was wrong, just like the "S".

So, some update correctly, some only update HALF correctly. Real sweet, eh?

[rebeldude]

Sounds like our checking software needs to be checked by some checking software software

[Scott Appleyard]

Maybe Wilcox posted the openings for CMMi people on CMMTalk!