Profile on opposing cardguides

[[Mi...]]

Hey guys, I'm having problems on a profile call out. I have groups of cardguides with a profile callout of .008 on opposing rails. What I'm doing is dragging a line auto path down each rail and keeping the filter outlier at spline, low-pass and setting the outlier elimination to Include Adj. Points/ number 5. I then create a plane by recalling the feature points of the 2 lines and finding the profile of that plane. It works fine in the simulation but not working so well on the Contura. These parts supposedly checked fine in manufacturing so they don't want to red tag the lot. Is there something wrong with what I'm doing?

[[Me...]]

If you're number are coming close but a little out of tolerance, you're strategy is probably OK. Is productions Datum structure the same as yours? That could cause a different output. If you try single space-points you may be able to find the offending area easily.

[[Mi...]]

they are close, some just in tolerance most are out. I tried every possible datum combination but they don't help. The thing is, I even tried setting an alignment off of the cardguides themselves. Again, in simulation the profile comes out zero. When I run the program on the machine the part is out. I tend to believe the parts may be wrong but checking the cardguides to the cardguides should measure good so I started doubting myself. We are going o have the original inspector run his program for in process operations to see what he did. I was just wondering if what I was doing could possibly be obviously wrong.

[[Me...]]

Did you try flatness on the plane recalled from the line points? Just a thought.

[[Mi...]]

I just checked the flatness on a few of the planes and they are all within .003 .

[[Me...]]

One more try and i'm out. I've never tried recalled points to a plane, and checked with profile, but since you're using two scanned lines, did you try just a feature plane with both those scanned lines included within the feature? I'm out.

[[Mi...]]

yeah, that's what I did. A feature plane and recall feature points. I recalled the 2 lines.

[[Aa...]]

Try this:
For each guide plane, select that plane for the primary (and only) datum. Then right-click on the plane symbol in the DRF. Evaluation method will probably say "Outer Tangential Element" or "LSQ Feature", depending on your Calypso settings. Switch that to "Minimum Feature".

If your filter and outlier elimination settings were put in place at the characteristic level (instead of the feature level), you'll need to put them in place here, as well.

BTW, I haven't seen the print and don't know what GD&T standard it's drawn to, but there's a chance those profiles must be satisfied simultaneously. Just a word of caution.

[[Mi...]]

ok, I went to characteristics and found the first plane profile. Opened it and removed all datums and entered that plane as the Primary datum, right clicked and changed to Min Feature. Without any other datums it refuses to calculate, I get "error with references". The GD&T callout is profile/.008/A/B/C 2surfaces,11plcs. -A- is the top surface, -B- is a motherboard locating pin hole (cylinder) and -C- is a locating oval. I wonder if I use, min.feature and the called out ABC datums but when I right click the Primary datum(-A-) if I can go to "Constraint" and choose the plane as Feature reference 1???

[[Mi...]]

well... I tried that stuff, basically all of it... and I got no joy. It just didn't seem to want to come in. Then I ran the program from yesterday that isolated the cardguides but ignored the datums and the part measured good. There is something screwy here and I think it stems from the way the parts were checked, approved and run on the machine floor. I have no control there so they can do the final inspections, I am washing my hands of it. But still, I would like to know if what I originally was doing is correct. I think it makes sense and it worked fine on the simulation, lol!

[[Aa...]]

Oh, so there's a DRF on the print? That changes everything.
I would say the DRF is the culprit. Specifically, the "oval slot" you have for the planar rotation is likely causing the issue.

[[Mi...]]

that's the first "cheat" I used. Since the machine floor is milling in the -B- and -C- datums to the outside of the assembly I set the outside of the assembly to control the planar rotation and set -B- as my X0Y0 as per print. Personally, I would have come off the cardguides to put in the datums and that was how I programmed the machine programs but after it leaves my desk I have little control.