Tertiary datum --- A point, a line or a surface?

[SeasonLee]

From the print attached, you will find the primary datum is a surface, the big hole is a secondary datum and the tertiary datum is a quadrant point of a curve.

1. Can we use a point (quadrant) as a datum, the curve is not a FOS.
2. How to set up the axis on the CMM on this case.

Thanks for all of your comments

SeasonLee

 

[pmarc]

SeasonLee,

1. You can use a quadrant as a datum feature. To do so you should specify datum target point or line for datum feature C. Location of such datum target should be dimensioned with basic dims. in relation to datums A and B and shown possibly on two views of the drawing. I am not sure if I understood you well, but datum feature C will not establish any datum axis in this case. One datum plane is established by datum feature A (or in fact by datum feature simulator of datum feature A), and two remaining datum planes are established by the axis of datum feature B. Datum feature C simply constrains rotation (clocking) of these two datum planes derived from B.

2. Are you asking for a method of setting up the axis of the datum feature hole B on CMM?

Additional comment: I think one important thing is missing on your drawing; you have not specified any geometric relationship between datum features A and B. I would put perpendicularity control of datum feature B relative to A to have it clearly defined.

 

[SeasonLee]

Thanks pmarc.

The datum system will allow designer to do two important things, first is the functional relationships and the second is measurement instructions, and so where is the axes of the hole? I am using the vertical straight side as Y-axis on the CMM measurement, the print is not clear on the datum system, I am wondering is it the right way to measure this part.

I can fully understand the datum C will constrain the last degree of freedom rotation, but why not to choose the vertical straight edge as the tertiary datum?

Thanks

SeasonLee

 

[SeasonLee]

I am also interested to know are there any examples on Y14.5 shows by using datum target as tertiary datum.

Thanks

SeasonLee

 

[pmarc]

SeasonLee,

There is a whole paragraph 4.24 in Y14.5-2009 standard that describes usage of datum targets. You can find there examples where datum targets are used for tertiary datums.

You can pick vertical straight face or any other face as a tertiary datum as long as it constrains remaining rotational degree of freedom. But you should keep in mind that function of a part is a priority when choosing datum features. So you should pick a face of a high importance for component functioning.

 

[TheTick]

Ask your inspectors to give this the sniff test. While technically "legal", a curved edge is a precarious place for a datum. You will never get a dependable measurement.

 

[HGMorgan]

As shown per dwg, -C- can only be a tangent pt on the radius, parallel to -B- @ -A-. A fixture could simulate the datums and help set up the CMM inspection axis.
Or, this can be set up @ CMM by:
1. Taking pt's on the radius common to -C-
2. Using resulting radius Center Point offset to -B- as defined by Basic dimentions.
(I do have a problem with that alignment: the radius is less than 90 degree's & near formed area, not good. The radius next to Limit Pt K would be better).

The long vertical edge common to Limit Pt K would seem to be a good candidate for a Datum Targe -C-, however, that edge has more tolerance, so we don't want to use it. The small hole would also seem to be a candidate, but it is only about 1/3 the total length, so we don't want to use that either.

This part will probably need Best Fit ran in order to see actual condition.

 

[SeasonLee]

Thanks HGMorgan

The PT-K is on the vertical edge, if we choose PT-K as the datum point on Datum -C-, it will be almost the same as by using this vertical edge as tertiary datum.

Thanks

SeasonLee

 

[axym]

SeasonLee,

I agree that Datum C is not specified in a Y14.5-compliant way and its meaning is not clear.

If it were me inspecting this on a CMM, I would avoid trying to derive a center point or axis from the curve. The uncertainty associated with this would be large, and there is very little chance that the center of that curve has anything to do with clocking the part. I would find the extreme point of the curve (point with the largest +X value) and clock to that.

As pmarc says, the function of the part is a priority when choosing datum features. When this part is put into contact with its mating parts or assembly fixture, exactly how is the part's last rotational degree of freedom constrained? Which part feature makes contact, and exactly what does it make contact with? If we know these things, then we can specify the tertiary datum properly and inspect the part the "right" way.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[SeasonLee]

Evan, Thanks

I agree with what you said, a datum system will aid in making repeatable dimensional measurements, I don’t think datum -C- makes it repeatable.

We are just an OEM parts supplier, sorry for I don’t have much information about the mating parts and its function. I noticed this issue when we were doing the 1st article sample inspection, and all dimensions looks good if we choose the vertical edge as the tertiary datum.

Thanks

SeasonLee