Perp of cylinder

[[To...]]

For a cylinder, I assume I can only do perpendicularity of the axis, but the drawing appears to be leaning towards the surface. Also, I can't fit the surface between 2 parallel planes, perpendicular to A (end face at bottom). Before I reach out to the customer, should I just assume the axis of the cylinder? Y14.5-1994
. perp of cylinder.pdf

[[Jo...]]

I'm confused by the graphic. So is Datum -A- the short cylinder ? and the highlighted surface a plane ?

[[To...]]

Please sign in to view this quote.

Please sign in to view this quote.

Highlighted surface is the cylinder.

[[Jo...]]

I see the problem now.. concerning data, could you scan "plane" using a circle path around the hole, then recall the data points into a nominal cylinder ? Maybe constrain it ? I don't see how you could establish a good vector with that size radius. Perpendicular could be 2 axis to another cylinder, correct ?

[[To...]]

Left view is rotated 180°

[[An...]]

Please sign in to view this quote.

Use this "unflat" plane for planar rotation (datum B) and then
evaluate the angle-deviation of a Least Squares-Plane.

The reference length?
Has to be defined by the designer.

[[Da...]]

The surface highlighted does not appear to be the cylinder shown on the left view. Is there a second giant cylinder for the highlighted face?

[[To...]]

After thinking about this, I personally don't think perpendicularity can be legitimately applied to a surface of a cylinder however; I do believe I can measure lines on the surface, parallel to the cylinder's axis and validate whether they are all within spec or not. Any takers?

[[To...]]

Please sign in to view this quote.

Dave, trust me. The view on the left side is looking in the -Z direction but the view needs to be rotated 180° around Z

[[Ke...]]

Please sign in to view this quote.

Before I got to this part of the post, I was thinking to myself:"Self... I wonder if they are trying to control the form of the surface, by treating the cylinder as a plane..."

[[Ma...]]

I think - all that matters here is axis on img on right side in direction left-right in view. So your lines does not cover this. I would try a circle ( less masked points ) or measure a few points from up to down to avoid hole and masking points on short line.
Even that Plane can be good point. But be carefull how you create that virtual plane.

[[To...]]

I just realized I missed something that adds to the mystery. On the model, the surface on the opposite side of B is a plane.

[[Ma...]]

Then plane is a way to go

[[To...]]

Please sign in to view this quote.

Please explain process of using a plane. If I recall the points from a circle or cylinder, the LSQ plane will have natural form error that will take away from the perpendicularity result.

[[Ma...]]

Now I am at home - tommorow I'll can send pictures.

Basicly start with extracting a cylinder from model. Then make a circular scan path around a hole ( aka with same center as hole ).
Then you can create a plane with recall points and check recalculate nominals ( still no measured data ).
Plane is modified according to cylinder form. Change plane from teroetical to measurable a make same circular scan path.
This should be enough for measure parallelism and perpendicularity. If your CAD model view from Calypso is from Z- and X+ is left-right then perpendicularity for Y axis is not needed - only angle X/Z

[[Da...]]

Ok, I am missing something here. -B- is a plane being held perpendicular to -A- which is a plane, why are we talking about a cylinder? Is it because there is not enough room to get points on -A-?

[[Ma...]]

Please sign in to view this quote.

Because Base B is cylinder - not a plane

[[Da...]]

Please sign in to view this quote.

If B is the cylinder it can't be perpendicular to A, only parallel.

[[To...]]

Please sign in to view this quote.

David, A is a plane on the bottom. If you scroll thru the previous posts, you will a drawing.

[[Da...]]

Tom,
I agree that A is the bottom plane. I am thinking that B is the plane on top. The cylinder at B can't be perpendicular to A, but the plane is.

[[To...]]

Please sign in to view this quote.

Cylinder B has a space axis of +Z, so it is definitely perpendicular to the plane on the bottom.

perp of cylinder.png

[[Jo...]]

I'm completely befuddled by this. I just don't see how this cylinder axis is perp to A. It is definitely parallel to A. I also don't see how the highlighted surface in an earlier snip is a cylinder. It looks like a plane to me. As I see this callout, the feature that is datum feature B must be perpendicular to the datum established by datum feature A within .002". I don't see where the cylinder comes into play at all.

[[To...]]

Please sign in to view this quote.

HA! Yes, it looks like a plane but as you can see from the feature window in my last post and the other parts of the drawing, it's a cylinder.

[[Jo...]]

I see that now. Since the callout is applied to the surface and not associated with the dimension, it applies to the surface and not the axis. What is missing is the note EACH ELEMENT. This would mean that each linear element of that surface needs to be perp to the datum. If the axis was the desired control then a diameter symbol would have to be in the feature control frame. Either way, the callout is incorrect so we just have to take a guess or ask a question.

[[Ma...]]

Tom, you can restrict radius for cylinder B and use cartesian perp. That will remove error from second axis.
I tried with a plane, but i didn't tested it. Two ways
- create a plane and set circular strategy around a hole
- create circular cut on cad model, make 2D curve ( possibly correct vectors ), measure, then recall points to plane ( but this can have a little second axis error )