Perp of Plane to Bore vs Bore to Plane

[[Da...]]

I am measuring a large disk shaped part with a 14.00" OD, the drawing calls for the perpendicularity of the face to be within .003" of the -A- Bore. The face is a rather large area, approximately 14.00" in diameter, the -A- Bore is 2.410" in diameter and 1.00" deep and centered in the part. I use the large face as my Primary Datum (Spatial) and a circle in the -A- Bore as my Secondary Datum (Planar Rotation).

If I report the dimension as called out in the print (face to bore), the perpendicularity of the Face is anywhere from .008" - .010" but when I switch the features around and check the perpendicularity of the -A- bore back to the Face I am reporting a value between .0003" - .0008".

Could someone please shed some light on this for me or possibly provide an illustration? I have my own ideas why this is happening but I would like some input before I respond to engineering. It seems that when evaluating the perp of the larger area the error is magnified but when evaluating the perp of the short axis of the bore back to a stabilized plane the error makes more sense.

[[Me...]]

It's the size of the features and projected tolerance zone error. The tolerance zone is created by the datum feature. As a general rule, you should check the shortest feature to the longest feature for perp. Any error in the smaller feature will be projected and used to create the tolerance zone for the longer feature.

Robert

[[Me...]]

VERY primitive example...

Perp.jpg

[[Ca...]]

I've noticed the same issue with some parts I have been checking. Is this "legal" to report perpendicularity of the bore to the plane even though the print calls out plane to bore?

[[Ja...]]

Please sign in to view this quote.

Do your face and bore share the same unit normal vector? I assume in Z+ direction?

If so, how are you using one for Spatial Rotation and the other for Planar Rotation?

[[Me...]]

Please sign in to view this quote.

Technically you need to report what's on the print. It's up to the engineer to understand the difference. Realistically, you need to ask yourself and/or the engineer a few questions...

1) Is the part geometry any different depending on which way you measure it? No

2) Will the part function in exactly the same way depending on which way you measure it? Yes

I will check it both ways if it's failing the print callout. I'll let the engineer know that I will be rejecting good parts because of the callout order on the print, if it passes with the order of features flipped. That's usually all it takes. A lot of engineers don't really think about the tolerance zones created by the features and given the information they almost always will change it.

If not, the price just went up!

Robert

[[Me...]]

Actually, each method is quite different. When applied to a surface, orientation tolerance zones (perpendicularity, parallelism, angularity), also control the flatness of the feature. It could be that the form error is consuming the majority of the allowed perpendicularity tolerance. If flatness error is to be excluded from the evaluation, the tangent plane modifier should be specified in the feature control frame.

[[An...]]

See attached.

Contribution_26_07_2019.pdf

[[De...]]

If you want to understand what is going on - do the following in whatever cad system you have at hand.

Starting at the same vertex, Draw two lines perpendicular to one another.

Make a vertical line the length of your bore.

Make the horizontal line be the half the length of the plane.

Now rotate both lines by some angle. Lets say 0.1 degrees.

If the bore is shorter, and the plane is longer, there will be more movement at the end of the line representing the bore.

If the bore length and the face are the same, the movement would be equal.

If the bore is longer and the face smaller, the line representing the bore will have more movement.

This is what the CMM is 'seeing', the checks are not the same, a short diameter being used as a datum for a large face will almost always give bad results due to the fact the a minimal amount of error in the diameter will cause a large amount of error at the outside edges of the face.

[[Ja...]]

If I may add to what Derek mentioned - If you’re using the plane to set spatial, then Calypso is going to assume your bore is to be perp to it. So when it goes to scan the bore (in this case you scanned it as a circle), it’s going to change the scan path to ensure that it’s perp to what Calypso perceives as the actual orientation. The result? – It’s going to check good. If it isn’t, there’s something wrong with the CMM or there’s a significant amount of form error somewhere. But that’s a moot point anyway because -A- isn't being picked up correctly. -A- is a 1” long bore and must be scanned as such. That’s going to give you the true axis (hopefully) of -A-. The reason I say hopefully is because CMMs in general don’t like poor bore length to diameter ratios (short bore lengths compared to the diameter of the bore). If you see a ratio of less than 2:1 (bore length twice as long as the diameter) you need to be more careful in scanning and make sure you apply proper filters and outliers (which should be done anyway). And as others have pointed out, report it as called out on the print. However there’s nothing that says you can’t have an additional callout on your report doing it the other way – just make sure it’s plainly stated as such to eliminate confusion.