Jump to content

Feature Graphics, CAD Evaluation and PiWeb


---
 Share

Recommended Posts

I was hoping someone might be able explain some things that seem odd when using Piweb graphics and evaluation in CAD and perhaps how to correct them.

Here are my questions:
1) The attached shows a PiWeb graphic of a Curve plot. Notice the line on the horizontal right side. Then the next picture below that is CAD eval. The line is not there. The report has the profile out 7.9mm on a feature that is only 4mm wide. Why don't the 2 correlate?
2) The next screed shot (3rd down) is Feature actual and nominals. Again no sign of the stray PiWeb line and also I'm wondering why gap between the nominal and actual. Magnification is set to one.
3) The 4th screen shot shows the actuals (vectors?) tipping down. Is it correct that this is an indication that the wall being measured is tipped and this is the direction of the vectors? This is a plastic part so it is probable.

Thanks in advance,

profile graphics.pdf

Link to comment
Share on other sites

Please sign in to view this quote.

1) I've seen quirky results like this and I don't give them much regard.

In your scan strategy, try activating pre/post scan (0.1) which will help eliminate some of these bad points caused by initial probe contact. If you don't have room, you can go into evaluation and limit the evaluation to not include the beginning points.

2) Measured points show at probe center.

3) Are you using a star probe on XXT sensor? If so, this tilt is very common but nothing to worry about.
Link to comment
Share on other sites

This is very common. I deal with this all the time.
I tried setting up pre/post in curve strategy window, but it's doing nothing ( unless you go to element editor and manually turn on masking points ).

My best experience is to go to curve eval window and manually select to mask few points from start and end ( depends on scanning speed ).
When i used masking as for basic elements ( element editor and enabling masking ), then masked points were evaluated as zero deviation - they should be skipped. It acted wierd.

And if you have more curves recalled into one big, then manual masking is your only option to mask pre/post points.
Link to comment
Share on other sites

 Share

×
×
  • Create New...