True position

[[Cl...]]

I have a cylindrical part with 6 equally spaced threaded bosses. Datum -A- is the small center cylinder (ZX).

Datum -B- (Y) is the smaller center boss face. Clocking planar (Z) with the first cylinder (Intersection 1) at 12 O'clock.

There is a position callout (6x) to A B. Would that be a best-fit bore pattern allow translation and rotation for all axis?

 

Edited September 11

[[Ma...]]

Edit:

If datum A is Cylinder 7 ( ID ) and datum B is that plane close to Cylinder 7 caption, then you can freely rotate in Y axis, but no translation.

Edited September 11

[[Cl...]]

Like this?

[[Ma...]]

I am not sure - you should have XZ ( or ZX ) and not XY

[[Ch...]]

Martin is correct.   should be XZ or ZX.  

Another tip is to use 'view tolerance' as the best fit method.  View tolerance will more closely match a hard gage than what GauB

Edited September 11

[[Cl...]]

I've only had to use this option a couple of times in my 25 years. I needed a little refresher. Thanks!

[[Cl...]]

I understand that I'm rotating about Y, but I'm slightly confused about the "Shape of Zone" because the deviations are in X&Y. Not Z.

[[Ma...]]

I think there is known this problem, that you can not change that to XZ. What i don't know is how to correct that.

Perhaps using rotated alignment and every intersection shown in this alignment.
You have to watch nominals.

[[Cl...]]

I ended up rotating 60° to each hole. It worked out.

[[An...]]

Please sign in to view this username.

, what does "View Tolerance" even mean in this context?  I haven't used that method because I didn't understand it.  Thanks for explaining that it resembles hard gaging.

[[Je...]]

Please sign in to view this username.

, I accidentally placed the previous comment under your profile, since you were still signed in to the OI543 in the QA Lab. Zeiss will not allow me to delete the comment, because we're past the immutable 60-minute window.

My regret for this mistake is sincere.  However, something felt delightfully powerful about masquerading as the almighty Andrew Block.

Please sign in to view this username.

, the above comment about "View Tolerance" was from me.

[[Ch...]]

Please sign in to view this username.

Okay, let's talk view tolerance.  

Gauss will minimize the square root of the sum of the squared deviations.  Most of the time this is good enough, but it can sometimes reject a good part. 

View tolerance works differently; it iteratively tries to accept the part based on the tolerance just like you would with a hard gage.

This knowledge will save someone embarrassment.  It sucks when you call a part bad, then someone slides it right onto a functional gage and looks at you and says WTF?

This PDF has been posted on here several times, but I don't think everyone has fully digested pages 17 and 20.  

Algorithms.pdf

Edited September 11