Old Concentricity Issue

[djayne]

I've posted about this before and never resolved the issue. I re-search the threads again but, can't really find what will make this work. It sounds pretty simple - Measure two bores, one of which is the datum, and find the concentricity between the two. I sent my original code to the guy who trained us on PCDMIS several years ago and he told me it should work. I get alot of wild and inconsistant readings. I check one part ten times last Friday and got a 1.1mm difference between readings.
So, my question is: What would be the most reasonable way to tolerance this dimension?
I have tried measuring as two cylinders, two circles, individual points to construct one cylinder - I have also changed up datum structures, projected circles to midplanes, constructed circles to datum planes and even tried the programs on 3 different B&S CMMs. When I measure the parts on the LK with LKDMIS it seems to dimension consistantly and within reason. I have also just tried to use the TP of one bore to the other which seems to dimenion within reason. Any suggestions - I will try just about anything (program wise).
Here is a picture of the part in mm : RUA Picture.pdf

[craiger_ny]

It sounds like you have tried a bunch of stuff but how about this? Are you measuring one circle in one alignment leveling scenario and then measuring the other circle in another leveling scenario? Even if both leveling scenarios level to features that are nominally the same their vectors are never going to be perfectly the same. This will skew you concentricity. If you are leveling to a feature then measuring a circle, then measuring a circle under the same leveling scenario then disregard what I just said and make sure what you are leveling to has a vector that coincides to your circle. For example if you are leveling to a face who's actual (not nominal) perpendicularity is off to one (or both) of the bores you are measuring you will have problems as well. I guess I am asking if your leveling situation is sound. Of course measuring 2 cyliders should have skirted that, I'm probably not helping you here I'll stop now................just a thought in case you didn't pursue it.

Craig

[djayne]

I might have a level problem. I level off of one of the bore faces and that bores center point for 0,0,0. Then I use the angled bore for rotation per blueprint. I have tried to level to a line between the 2 bores which gave me almost the exact same situation as the first. I never noticed at the bottom of the threads that it lists similar threads. I read one that talked about taking identical hits and using midpoints - that seems like something PCDMIS should be able to work out for you.

[Peter Warcholyk]

"djayne"

After reviewing your picture.pdf file, I don't understand what you are after.
It would appear from your picture that you have 3 bores.

Do you have cad data available? If so, could you not measure the three bores, using surface sample hits and then do an iterative alignment.

If possible please try to clarify. Hope that this provides you with another possible idea.

ZydecoPete

[djayne]

I'm interested in the concentricity of the two in-line bores. I'm working on trying to match the hits to be identical in the front and back bore now.

[kbotta]

Sounds more like coaxiality...

[djayne]

Sounds more like coaxiality...

I agree...My bosses want the printout to read exactly how the blue print calls it out. I'm trying to locate an electronic blueprint to post the section - If I can't, I'll scan it off and post.

[djayne]

Here is the Drawing

Here is the portion of the drawing with the concentricity call out on it : RUA CONCENTRICITY2.PDF

[bboyd]

Are those bushings pressed into your main component or is it a machined component?
We measure a similar part with bushings, problem we ran into was they weren't always pressed in perpendicular to each other so when you compared the 2 the numbers were all over the place

[cmmguy]

I see 5 similar threads below on this topic. Kevin, can you please kill the word concentricity. Can someone go to the next ANSI or ISO meeting and have that work permenantly stricken from the vocabulary

[djayne]

Oh yeah!

Are those bushings pressed into your main component or is it a machined component?
We measure a similar part with bushings, problem we ran into was they weren't always pressed in perpendicular to each other so when you compared the 2 the numbers were all over the place

That's another thing - They are so worried about the concentricity before assembly and care nothing about the part after assembly except press-in and width of part. The Concentricity I'm trying to measure is just two through bores before bushings.

[bboyd]

I don't see why it wouldn't repeat as long as you are referencing the same plane/face for each bore.

[kbotta]

CMMGUY,
I would love to bury that word. Believe me. When I tell "them" it is NOT concentricity that they want... "they" think I am crazy - and I do not have a clue. Then it is back to the old section 5.12 page 144 routine. You'd think I'd have it memorized by now - well - actually I am pretty close ...
Kev

[kbotta]

Do you have a ISO drawing, or a ASME drawing?

Judging from the datum feature symbol(s), it looks to be a 94 drawing. Is it ASME?

Cause, if it is - you would want to check the coaxiality (or position) of the one bore the one identified in the FCF. I do not believe the intent here is a true "concentricity" evaluation (99.9999% sure it is not) (we can look at this further, if you really want - but it does not even look to be a surface of revolution....)

If it is a ISO drawing, than you are still checking coaxiality - or position.

We need to know what "language" the drawing is speaking...

Kev

[Wes Cisco]

We need to know what "language" the drawing is speaking...

Kev

Could be Woolkie, but I would be willing to bet is actually a heavily accented dialect of Mathrusian.