Analog Scanning

[2013_junkie]

So first and foremost, I'm completely wet behind the ear with scanning. We have decided to add scanning to a few of our machines, so while we are waiting for them to show up, I've been working with it offline attempting to teach myself. I keep having this pesky issue where I create a linear open scan and there while be just a few random points out of whack.. When I turn point info on it shows something like 1,2,3,4,5,49,6,7. Anyone have any idea why its doing this?? I'd appreciate ANY pointers on programming with scanning!
Capture1.PNG

[JEFMAN]

Download this :
ftp://wilcoxassoc.com/docs/v2017r2ma...ore_manual.zip
and look at the chapter 19 : "scanning your part", maybe it can help...

[KIRBSTER269]

Is this one single scan? Maybe you don't have enough crossings? maybe bump up your crossings regenerate your hits, and see if that changes anything. Big gap between 27 and 29, also 48 and 50

[2013_junkie]

The big gaps is what Im trying to rectify...

[bigtallanddopey]

The big gaps is what Im trying to rectify...

The gaps could well be an edge offset in the path definition tab. You don't have surface edges turned on (the black lines on the model) but I would bet where the gaps are there would be a split line in the CAD model. Set the edge offset to zero and see if that helps, if edge offset is zero then there is something else going on.

[2013_junkie]

The edge offset is set at zero. Its discouraging since I don't really know my way around the scanning aspects of pc-dmis so I'm purely guessing as to why its generating these screwy scans. I recreated the scan, made sure edge offset was at zero, added some crossings, and weird gaps, in different places..

[bigtallanddopey]

Scanning can be a pain, sometimes it does things that you cannot understand why. You are doing the right thing though by practicing offline, just keep playing around with the settings and see how each one affects the scan in different ways.

Maybe you can post screen shots of the different tabs when you are setting up the scan and maybe we can see something obvious. There could also be an issue with the CAD itself, I have had it where there has been a tiny gap in the CAD surface and the scan will continue on the "inside" of the part.

[2013_junkie]

Here is the screen shots of the scan window. Im going to have one of the cad guys look at the model closely to see if there is any issues with it.

[bigtallanddopey]

In the execution tab try changing the boundary type to plane instead of sphere.

I am guessing the program is in inches and the tolerance under findnoms is ok, if it was mm 0.1 would be way too low. Eiether way I don't think that affects anything when offline.

[KIRBSTER269]

As bigtallanddopey suggested change that to Plane, also you have 2 crossings. You don't have the Graphic Tab displayed so I'm assuming you have select Box for CAD selected? Crossings are the times it lifts up and starts again, if you are wanting a continuous scan change that to 1. Make sure surface is selected when you do this. As my example shows when Model isn't highlighted

scan.JPG

[2013_junkie]

So, I've tried all of the recommendations and am still getting weird scans. FYI I'm trying to build these scans using the quick scan feature, think that could be causing my problem?

[bigtallanddopey]

Possibly, ive never used it so cannot comment on what its like. . try just going insert > Scan > linear open.

[KIRBSTER269]

Open a Linear scan open
step1.JPG

choose your path on the model

step2.JPG

Next go to the Graphics tab check select under CAD Controls

step3.JPG

Highlight every bit of the surface around or close by your line

step4.JPG

Generate Path
step5.JPG

[2013_junkie]

Kirb, I did it exactly like you said, multiple times, and Im still getting the same result. I'm going to go check in with one of the cad ppl to see if they found any issues with the model. Starting to think that's what it has to be!

[2013_junkie]

So, If I set it to linescan the skips stop happening, its only doing it when its set to nullfilter..... Odd

[KIRBSTER269]

did you set your Crossings to 1? that is odd

[KIRBSTER269]

If they sent you a Step file, Tell them to give you an IGES. or vice versa. see if it does it again.

[2013_junkie]

Oh man, for some reason it reverted back to sphere with 2 crossings, once I changed it back to plane with 1 crossing, re picked all the cad, it generated it properly. I think you may have solved it!

[JEFMAN]

Checking the cad can be done by measuring some vector points... Just a thought...

[2013_junkie]

I was thinking there may have been some tiny voids in the cad??

[JEFMAN]

Like Kirbster said, depends on the format...
I've more confidence in step than IGES, because of those "holes" (voids !)

[2013_junkie]

As far as models go, I've been with this company for 12 years and cad models have only become available to us within the last year, so I'm a little shaky with the ins and outs of models. We have also been strictly touch trigger (technically still are) but we have some 2.5 deg heads with analog scanning attachments on order. Hence why Im stumbling through the scanning side of pc-dmis. Probably going to take the scanning class but the bean counters expect us to hit the ground running when this stuff comes in so Im trying to get a jump on things.
So, say I want the scan to collect more point data, what do I adjust?