How does a CMM use 2 axis to establish a datum plane

[ringman]

Imagine a part with 2 axes which are not in line, but offset. These are used in a callout as primary and secondary datum features. How does the CMM accomodate this condition?

The axes are depicted on the drawing as parallel. Manufacturing would not guarantee paralllelism, (in my opinion), but rather skewed liines.

 

[dingy2]

Ringman:

From the information given, I can see taking multiple points around the primary datum cylinder to create the primary axis using a CMM. It would create an average cylinder size and axis. This might be a bit like performing cylindricity using a CMM.

The secondary datum would be used for orientation around the primary. I am going to assume that both are in the RFS state rather than MMC.

I would take a number of points around a segment on datum feature B in the center along its length. This is the best way if the part was not really parallel as reflected on the drawing.

Depending upon your software, one should now be able to create a plane using axis on datum A and the center of datum feature B for orientation.

Hope this helps.

Dave D.

http://www.qmsi.ca

 

[PaulJackson]

ringman,

Are these feature axes parallel but "offset" like a hand crank (end to end) or like the rungs of a ladder.

Paul

 

[ringman]

Pauljackson

They are offset like a crank.

Thanks

 

[PaulJackson]

ringman,

Dave’s advice is right-on. I wrongly jumped to a different picture of the problem thinking that both of the “offset cylinders” were declared as the primary datum feature demonstrating the hazards of scanning (not carefully reading) the question.

Stopping rotation from the “offset cylinder” is the same as it would be for one that is not “offset.” since most CMM’s are highly capable of scrutinizing projected locations and orientations. Once the primary axis is established by the “cylindrical” primary datum feature there are only two degrees-of freedom remaining left to constrain…translation along the axis and rotation about the axis. So if you assume that the secondary datum feature is “straight but not exactly parallel to the primary and if your assumption is true” then probing a best-fit circle in the middle of its depth will yield the best estimate of the secondary’s “actual mating envelope” orientation from the primary. If your assumptions about the secondary’s form are untrue then the estimated orientation will not be so accurate. If its depth is shallow compared to the primary then the estimate of its orientation will be less prone to error…likewise if its depth is greater than that of the primary the probability of error will be greater.

Paul

BTW... hi dave!

 

[dingy2]

Paul

What is this? You agreeing with me? Must be around Christmas or something.



Dave D.

http://www.qmsi.ca