Generating a scan on a section cut/surface intersection?

[vpt.se]

I am trying to generate a linear open scan on the intersection of two surfaces, but it seems I can't.

Consider the following:
Capture.JPG

I want to generate a defined scan on the "ridge" where the two surfaces meet, but however I do it, I only get the upper or lower surface.
I created a section cut and wanted to use that curve to place my start/end points, but it refuses to select on the curve.
Tried holding SHIFT and click, the curve is highlighted but there is no points placed.

Got any tips?

[NinjaBadger]

I'm not sure I follow. You want a scan on the ridge where the left hand arrow points to (scanning into the screen as shown in the image)? What are the two arrows on the right showing?

[Matthew D. Hoedeman]

I think it is a continuation of the same shape as it goes around a part, think looking at an L <----- with this view

[NinjaBadger]

Ah right I get it.

[JEFMAN]

I think the problem in this case is that the vectors don't exist along the curve.
Getting them could be calculating the average of the upper vector and the lower one, and then input them in a xyz file with the right coordinates.
However, I don't think that the cmm drift allows to measure it accurately...

[NinjaBadger]

Agreed on both count. Also if the part deviates from CAD (which it will) you're not going to be scanning right on the 'peak' anyway.

[vpt.se]

This is my though as well as to why selection on the curve isn't possible. There are simply no vectors.

[Matthew D. Hoedeman]

he could use (GASP!) a cylinder probe, but I don't know if it will work. I tried (ONCE, as a trial) to use just a plain extension as a cylinder probe to scan a 'flat' piece of sheet metal, but the SP25 stuff sure didn't like that! I calibrated a probe, then replaced it with the extension that was longer, and even though it was being done at the correct 'height' on the extension, it just moved wonky as heck. It wouldn't even follow a straight line without a bunch of crappy movement.

[BIGWIG7]

have you tried cross section scans not following the ridge but creating scans perpendicular to it? you'll just have to create a lot more scans than one continuous scan that follows the ridge.

[vpt.se]

Nope, seems tedious ;P

[vpt.se]

Appreciate the replies, I might have to choose another weapon for this particular task, like laser scan it.

At least I can stop trying to get PC-DMIS to place 1/D/2 points on the section cut curve...

[JEFMAN]

Appreciate the replies, I might have to choose another weapon for this particular task, like laser scan it.

At least I can stop trying to get PC-DMIS to place 1/D/2 points on the section cut curve...

With a laser scan, I believe that the best way is scanning with the laser line parallel to the edge (I was told it's the better way to do it).

Maybe a REVO head could do the job, unfortunately, it doesn't work on Hex cmm's, and they didn't bought yet any company which invent something like it ..... !

[Matthew D. Hoedeman]

With a laser scan, I believe that the best way is scanning with the laser line parallel to the edge (I was told it's the better way to do it).

Maybe a REVO head could do the job, unfortunately, it doesn't work on Hex cmm's, and they didn't bought yet any company which invent something like it ..... !

With the Romer, at training, I was told it is 'better' if the laser line 'goes around' the corner (radius), as in, if you are scanning a radius, the laser line should "look" like a circle, if ya know what I mean.

[JEFMAN]

With the Romer, at training, I was told it is 'better' if the laser line 'goes around' the corner (radius), as in, if you are scanning a radius, the laser line should "look" like a circle, if ya know what I mean.

2 teachers = 2 methods...
It was with a tracker + T-scan.... On a squared edge (no radius).

[Matthew D. Hoedeman]

2 teachers = 2 methods...
It was with a tracker + T-scan.... On a squared edge (no radius).

Maybe it is the type of scanner, this is a built-in laser on a Romer arm. Square edge the same, 'wrap' the shape. But then, Hex/BnS training is pretty spotty at times. The 'teacher' learned more from me at Romer training about the software than I learned from her.

[vpt.se]

The 'teacher' learned more from me at Romer training about the software than I learned from her.

Is there anyone on this earth that wouldn't?