Concentricity Help!

[tking]

It will, it just needs to use the 3D centerline of the datum to check it to, this is why you level to the cylinder. Am I oversimplifying this? Anyway this is what I would do as I understand the requirement to this point.

You gotta love this stuff!

TK

[kbotta]

Here, this is from ASME and will I believe help...

[Underspec]

Can someone please translate this impressive explanation a simpler way??

I feel like my wife is talking again when I read this...

[kbotta]

It means you would need to take 2 points exactly 180 degrees from eachother - or polor opposites - and construct from them a mid point. You would keep doing this until you have a point cloud. Then that point cloud would need to be inside of the zone of tolerance specified inside your feature control frame.
And just how many points exist on your surface?
Maybe you had an old drawing, or someone seen this and knew that concentricity is not an easy inspection method - and changed it to position, which is a much more friendly control. It makes it much more easier for you. Also, If you would post a pic of the drawing in question, we could walk you through this step by step...
just a thought....
Kevin

[Schrocknroll]

Are you outputting concentricity from Pcdmis?
the reason I ask is, it is NOT ASME Y14.5M-1994 compliant.
Pcdmis does not correctly obtain the results for this.
maybe you trying to get coaxiality?
Kev

What does PCDMIS give you instead of concentricity?

[tking]

Picture in your mind the centerline of your datum cyclinder, now draw a cylinder around that centerline with the diameter of your concentricity tolerance zone. The centerline of your feature under test must fall within that.

Helpful?

TK

Kbotta has it more technically correct (per the book), I would check it as I said but at multiple levels throughout its length. OR you can spend a lot more time collecting points (per the book) and this would capture the effect of form errors as well (the total concept of concentricity).

Kudos to Kbotta

TK

[Schrocknroll]

Picture in your mind the centerline of your datum cyclinder, now draw a cylinder around that centerline with the diameter of your concentricity tolerance zone. The centerline of your feature under test must fall within that.

Helpful?

TK

How is that different than true position?

[Jeffrey Lovely]

concentric to another feature - not to datums a-b-c

[kbotta]

Pcdmis gives you a coaxial result, not concentricity.

You must test polor opp points, and get a mid point from each.
every point on the surface. Just how many points?
That is an awful lot of inspection.
Imagine, rotate + pts 180 deg from each other + construct mid pt +dim mid pt
repeat
Than, slide down the axis to the next 'row' of pts. Ridiculous...

or you could write a loop to do this, with probe changes?

Not an easy thing.
They are differant.

Jeffery is partially right. Concentric to another feature (remember - must be RFS) and not to a complete DRF. It is for a surface of revolution (mainly). I wouldn't exactly say it is controling the CL of the feature, more the surface's mid points (ok it would be nominaly the CL)...
kb

[Schrocknroll]

It depends on your customer. Pratt and Whitney's drawing interpretation spec(PWA 360) treats concentricity the same as coaxility and true position, simply using the axis of the feature instead of a median point cloud taken from diametrically opposed points.

Always check with the customer specs first, if applicable.

Kevin,
Thank you for enlightening us as to the correct ASME interpretation of concentricity.

[Underspec]

Funny, when I construct a cylinder from two circles at different depths, the cylinder in the graphics window is facing perpendicular from the plane. Lol.

I've tried everything and my concentricity is still high.

[tking]

I agree with both Kevin and Jeff. My problem with following the full intent of the callout is the "when am I done" part as there is an infinite number of areas to check. It's always got to be discrete opposing points with the middle of the 2 checked back to the tolerance zone. No matter how small my angular or level increment there will be some area not inspected. If I try a feature made of more than the 2 opposing points I still have the best fit centerpoint or centerline not the true middle.

I choose to do the centerpoints of circles at multiple levels all of which must fall within the cylindrical tolerance zone as my best compromise between time and absolute true value. You would have to use your judgement (and design intent) to decide when you have substantially fulfilled the intent of the callout.

Fun callout,huh?

TK

[george frick]

TIR

put one end of the cylinder in a v-block and find the high point of the other cylinder to be checked with a dial indicator. rotate the part 360 and read the total indicator reading. (TIR) this is your concentricity

[Underspec]

Lol George...I would if the tolerance for concentricity wasn't .002".

[kbotta]

put one end of the cylinder in a v-block and find the high point of the other cylinder to be checked with a dial indicator. rotate the part 360 and read the total indicator reading. (TIR) this is your concentricity

Actually, that would not be correct per ASME. You would actually need 2 indicators at opposite pts, with the feature located.....
kb