Minimum Coordinate feature

[[Gr...]]

Attached is a picture of an alignment notch that is giving me trouble.
The drawing says the bottom of the notch needs to be .063 +/- .005" from my 'Circle 1 top'
My 'Circle 1 top' is the origin of the base alignment, and has a roundness of .0002

Using result elements:
Manually creating a 'Point1' by using Recall Feature Points on the deepest recorded point in 'Circle notch 37.1' and using: getActual("Circle 1 top").radius- getActual("Point1").coordPolHeight
gives me 0.0703

getActual("Circle 1 top").radius - ( getActual("Circle notch 37.1").coordPolRadius- getActual("Circle notch 37.1").radius)
gives me 0.0698
(pretty close, but smaller arcs / form problems may cause problems)

Using a Minimum Coordinate feature on 'Circle notch 37.1'
getActual("Circle 1 top").radius- getActual("Minimum Point circle notch 37.1").coordPolHeight
gives me 0.0611


How do I get the Minimum Coordinate feature/characteristic to work correctly?

pic7.jpg

[[De...]]

I have always wished there was better documentation on this construction.

The documentation says "Use the Minimum Point construction to compute the coordinates of the point at which negative deviation from feature geometry is at its maximum."

I know when the feature selected is a plane, minimum coordinate will always give you the lowest point in relation to face normal of the plane - NOT in relation to the Base Alignment. So with a plane facing the Z+ direction the minimum coordinate is in the Z- direction, but with a plane that is in the Z- Direction the minimum coordinate is in the Z+ direction.

I have gotten consistent results using minimum coordinate with a plane.

From the way the documentation reads using this on a circle should give you the point on the circle where the radius from center line is the smallest, however in practice this does not seem to work.

I have not been able to achieve the expected results when attempting to use this on a cylinder or circle feature.

[[Gr...]]

Well darn. I did not know it was a form thing.
It really seems like there should be find the outermost/innermost point on a radius/curve/surface.
Especially for functional fit purposes.

I do not have curve, but maybe this is a curve thing?

[[Gr...]]

So much of Calypso seems to be arcane knowledge that there is no way anybody could possibly know or find out without having somebody else that knows telling them.
The rest is split between functionalities that we SHOULD have, and functionalities that seem like Charlie Brown Christmas tree creation hand-waving.
http://1.bp.blogspot.com/-qu4KQXcrm8M/V ... cbtree.gif

[[De...]]

Its hard to tell what the part looks like so its a bit hard to know what approach to take.

If the notch is in a face, you could probably get an intersection of the circle and the face and then use the Z location of the intersection.

If you use the Z location however, make sure you make a fully defined alignment and assign it inside of the intersection construction, this will make the resulting Z location be reported in relation to the alignment.

Otherwise the result will be in relation to your base alignment probably won't be in relation to the relevant features unless they happen to make up your base alignment.

[[Gr...]]

Derek, I have no idea what you are getting at. What is the point of getting the intersection? Or its Z location? What I want would nominally be the distance in the X (but actual location will not be perfectly @ Z=0, and other situations will not always be on the axis).
My part has the notch on a cylinder, but it would be useful to also learn about it being on a plane.
Included is a picture of the same notch, but with surfaces.

pic8.JPG

[[De...]]

I completely misinterpreted what I was seeing in your original picture.

Is circle1 top that you are refering to the OD of the part?

[[Gr...]]

yep

[[Ro...]]

Ive had mixed results using MIn & Max features, even an established program will sometimes give me a bad result that can be verified via secondary gaging.
In this situation since you are already probing the thru hole and the notch, i would use a theoretical line and 2 intersections to get the distance you are looking for.
probe each feature at the same "Height" from the face.
create a 2d line by recalling the thru hole and the 37.1 notch
Intersect1 is the 2d line & the notch circle (be sure to select the correct result, as there will be 2)
Intersect2 is the 2d line & the thru hole (be sure to select the correct result, as there will be 2)

Min distance is 3d Simple distance between intersect 1 & intersect2
or if you want distance from the center of the thru hole to the notch that can be done just as easy.

forum5.JPG

[[Er...]]

Use the new caliper function, with outer tangential on the notch. That way you don't have to go to hell and beyond with creating theoretical elements.

[[Ro...]]

3 items is Hell & Beyond?

forum6.JPG

[[Er...]]

300% 😃

And the section might not even touch the right point.

[[No...]]

On a circle the min/max coordinates give you the cordinates of the point with the min/max radial deviation to the calculated best fit element (I don't remember if you can choose between actual and nominal as reference). BUT this point could be anywhere on the circle, so if you use it to measure the depth of the notch in +X the results are purely random.

In case you don't have 2017 with the caliper function, there's another workaround:
Recall the points of the notch to a line and set its a1/a2 nominal angles so it runs orthogonally to your measuring direction. Then use the min coordinate of the line. But for this to work it's mandatory to align the notch to Z=0.

[[Er...]]

"Minimum: point with the smallest distance to the other feature." - Based of measured points, not calculated geometry. That's how I understood the OP's question.

Since that point is the "bottom of the notch" radial to his origin?

But I might be wrong. 🙂