Edge Scan

[jonny]

I am trying to perform an edge scan, but the machine keeps crashing when I do so. I believe the main reason is my inexperience with a CMM, so I hope that somebody in this forum can help me out. I have tried a couple of different ways to do this, all unsuccesful. I first select 'edge' as my hit type, then I establish my boundaries by hitting as close as possible to the leading edge, then as close as possible to the middle of the edge, then as close as possible to the finish point of the edge. I then go through and change the 'init', 'perm', etc. settings that pop up when you select 'edge' for the hit type. But when I click create after I am done, the probe goes off to a surface that is nowhere near the edge that I am trying to measure. What am I doing wrong? I have looked in the manual, and it only speaks in very vague terms about how to create an edge scan. Thank you in advance...

Jonathan

[\v/inston]

insert a liner scan - insert/scan/liner scan. your first hit is the starting point, the second hit is the direction and the third is the ending point. there are ajustments for hit spread and other things like probe comp (ask matt about that)

also it depends on what you are trying to accomplish, if you are dimensioning a profile then you will need nominal data for each hit. if you are trying to reverse engineer a trim line then you just need to do the scan an output it to iges.

[thecoot]

If you are using Auto Features Edge point , pick the surface perpendicular to the edge you are after first to get your sample hits.

Excuse the cheesy sample


Coot

[Matthew D. Hoedeman]

Time was, when you were creating scans, that Pcdmis would actually give you good directions on what to hit and in what order to hit them, right on the screen at the lower-left corner, now sometimes you get it, sometimes you don't. Check out the help file on scans and the edge scan and make sure that you are taking all the hits and in the correct order. I would bet you are not taking enough hits or in the correct order. You have to sample the surface as well as the edge, not just hit start, direction and end. AND, since this is a scan with edge points, you will need 3 surface sample hits and 2 edge sample hits. This will give you the correct surface vector as well as location and it will give you the correct edge vector so that all the probe comp is done correctly. So, you will be taking 6 hits for each and every point of data it gathers. WOW, is that a lot of stuff or what? This is the reason I don't do edge scans, 6 hits for 1 point of data. It takes FOREVER and then some. But, if you don't have a choice, you do what you gotta do.

[jonny]

thanks everyone