2D scan following a 3D surface

[Matthew D. Hoedeman]

I've been 'fighting' this for over 25 years. And in that time, I can't believe that a solution hasn't been added to Pcdmis, that or I'm not looking in the right place.

3D surface on a trim detail. The TRIM edge is always straight up and down, as in, NO Z VECTOR. But there is only 6 to 15mm of this trim edge at the top of the detail. I want a scan that will follow the 3D shape and maintain the 2D-ness (NO Z VECTOR) of the trim edge. Perimeter scan, NOPE, wants to go all the way around that strip of trim edge, I only want/need ONE PASS mad-way between the surface on the top and where the trim edge stops without making tons of little scans along the edge every place it changes height. ONE SCAN PASS, not a bunch of little ones. I am doing straight little scans now, and I'm getting mismatch where one ends and the next one begins. Don't know if you can see the red line in the middle of the orange surface, but that is what I want "automatically"

[SingularitY]

This is something I would normally use a perimeter scan for. Is there a reason why this isnt working for you? You say it wants to go all the way around the trim edge. Do you mean it wants to follow the entire perimeter and end where it began?

[SingularitY]

20200123_190545.jpg I'm sure you know this, but this is where my placement for a perimeter scan would be. If the perimeter wants to go the opposite way than intended, I select "Inner Bound" and recalculate.

[Matthew D. Hoedeman]

This is something I would normally use a perimeter scan for. Is there a reason why this isnt working for you? You say it wants to go all the way around the trim edge. Do you mean it wants to follow the entire perimeter and end where it began?

yes, starts (for example) middle-center-top, then wants to go all the way to the right, then down, then all the way to the left, then up, then back to where it started. And, in addition to that, every time it starts to move, it actually errors out the controller and stops everything.

[Matthew D. Hoedeman]

Just tried it in offline mode, and it only did the one edge, unlike EVERY FREAKING TIME when I tried it ONLINE and it went all the way around the thing, top, bottom, & both ends.

[JEFMAN]

Can you try to generate the scan offline, then export points and read the file as a freeform scan ?

[Matthew D. Hoedeman]

Can you try to generate the scan offline, then export points and read the file as a freeform scan ?

I could try to make (not run) and export and import, but no more detail to work on to see if it works or not.

[Matthew D. Hoedeman]

This is something I would normally use a perimeter scan for. Is there a reason why this isnt working for you? You say it wants to go all the way around the trim edge. Do you mean it wants to follow the entire perimeter and end where it began?

AND, it did screwy things in 3 places. TWO of them I can almost understand, but NOT the one that the image is in wireframe and not a solid view. Well, I was going to upload 2 more images, but the forum won't let me add any more. But you can see that it, for no reason I can tell, jumped vertically when the surface is smooth

[SingularitY]

That's odd. I've actually never tried to program a perimeter scan online so I dont know those quirks yet. I've seen something similar before but lowering my perimeter scan CAD tolerances fixed that. But I doubt this is the same issue. Are those screwy areas what the scan looks like on the CAD or only after the part is ran?

[Matthew D. Hoedeman]

That's odd. I've actually never tried to program a perimeter scan online so I dont know those quirks yet. I've seen something similar before but lowering my perimeter scan CAD tolerances fixed that. But I doubt this is the same issue. Are those screwy areas what the scan looks like on the CAD or only after the part is ran?

no part, they are all back on the die, can't do any real life testing until the next time crap like this comes up, so that is the generated scan path. The other screwy areas were where that thumb-nail dip in the top form is, it made triangles at each 'corner'.

[Matthew D. Hoedeman]

I'm sure you know this, but this is where my placement for a perimeter scan would be. If the perimeter wants to go the opposite way than intended, I select "Inner Bound" and recalculate.

yep, that's what I did (online yesterday), and it wanted to start in the top-middle of the edge.