Profile calculation

[mfzaring]

Any insight into why it is using 2x the worst deviation for a form only profile? Bascially it is not allowing the tolerance band to move, which is the intent of the check. I assumed the form only would use a worst pos plus worst neg. The example below shold be in tol and shows in based on the meas value, yet still shows out of tol using math of a form and location profile. Any help would be greatly appreciated!
prof.PNG

[brian.freeman]

Is there a specific tolerance for this profile? I ask because you show 0.0000 as your tolerance. As for 2X max deviation, it isn't , max deviation in positive and -(Negative) = total deviation. Per AMSE 14.5.1M

(Didn't see the + tol) Plus, .0004 = +.0002 / -.0002 measured -.00025 (.00005 OOT)

[Mike Ruff]

Profile is defined as the max deviation minus the min deviation (for ASME Y14.5, not ISO). So... .00014 -(-.00025) = .00039. The measured value of .00040 vs. .00039 is due to rounding error

[VinniUSMC]

This is not true. ASME (currently) says only that there is not 1 number that may be used to report profile. You should be reporting a min and max value. PC-DMIS interprets that as max-min. And the next revision of ASME apparently is changing to match ISO. So, it will be a moot point for newer revisions.

[JEFMAN]

I would use minimax vector, instead of least square... (I'm not at the cmm, maybe I confuse with surface form)

[KIRBSTER269]

Well you're doing Form Only, So Size doesn't matter, Lets say you are you are using a CAD model, taking hits on the perfect part, since there is nothing controlling orientation or size of the feature, technically based of the results of your hits it creates it's own feature, Then to whatever the best shape it has come up with, it took the highest hit and lowest hit and added them together, you are doing a bilateral profile, you are out .00005, That means based off of all your hit results and the best shape that PC-DMIS could come up with, if you show your Max and Min like Jeffery suggest you'll most likely see Max .000144<<<(could be) and -.000254<<<(could be) in the Min. meaning anything that this number in the negative it is out of tolerance.

Correction YOU ARE NOT DOING A Unilateral PROFILE my bad <<<<<<<<<<<<<<


Capture.JPG




But your MIN is reading -.00025 which is not in your +.00020 or -.00020

[mfzaring]

Miin/max instead of least squared fixed this issue. Thanks JEFMAN . Found in the help index that min/max uses ASME 14.5 which is what I wanted

[KIRBSTER269]

I think I see the problem with this who haw stuff. Look at the Y Column and also the K Column. Here's one of my scans his or hers is below. I'm telling mine to press in the Y plus direction and my numbers are pretty much staying the same. I would have to assume that this is a Plane in the Y direction, since all the Y's are close to Zero and these are random points on this surface which none of them are square hits, so all in all, .00005 out of tolerance maybe correct calculated by a bunch of crappy hits. The Bottom IJK's are all over the place

Capture.JPG

Capture2.JPG

[JEFMAN]

It seems to be a nice curve... The red one is the xyz + ijk/10
courbes.JPG

[KIRBSTER269]

yeah but is it suppose to be? a lot of info missing yes?

[KIRBSTER269]

I don't know Jeffery .00009 that's in inches. almost .0001 in the plus and the -.00007 close to -.0001 that's a tenth in each direction. That's not much of a curve. that's pretty flat That would be .002286 mm and -.001778 mm, in your world

[JEFMAN]

KIRBSTER269 :Hi Kirby, you're right, some infos are missing...
I think te problem is "form only about YPLUS" in the title.
The curve and the vectors show a curve in the Y plane (moves along x and z), but the form shouldn't be measured in this work plane. I'm not at the cmm to check, but I think it doesn't make sense...