machined to cast step

[[Jo...]]

I have a large casting that gets a machined diameter. It calls for a minimum step from machined diameter to cast diameter. The machined diameter has decent form around .003", while the casting may be .060" out of round. They will never be perfectly concentric.. I have played around with curve distance and can't get it to work. I suspect it's because it is 360 degree curves. So maybe I can use runout and a formula ?

Any suggested methods appreciated.

[[Ch...]]

I believe that "Size>Curve Dimensions>Curve Distance" is what you want. Then select the "minimum value" option.

you have a place to insert your inner and outer curve

[[Ch...]]

I also feel like you could take the min circumscribed of the machined circle and the max inscribed of the cast circle, then check concentricity of those two features. Actually, more like position to a single datum or ISO concentricity instead of ASME concentricity, but who has time to go down a concentricity rabbit hole.

[[Ch...]]

Pro Tip: If you measured them as circles but want to use the Curve Distance characteristic, then just recall feature points the circles into curves.

[[Jo...]]

Please sign in to view this quote.

The average diameter is correct, but I get crazy numbers for max, min. I'm using the plane distance evaluation because they are not in the same plane, otherwise it would be a 3d distance.

[[Jo...]]

Please sign in to view this quote.

That's how I did it.

[[Jo...]]

Please sign in to view this quote.

I suspect the vectors are the problem, both circles are outside measurements instead of inside/outside.

[[Ch...]]

I'm trying to find time to play with this but people think they need stuff.

I'm thinking if we could use "construction>projection" to project each curve/circle onto the same plane, then maybe the Curve Distance characteristic would work. Would just have to watch out for projection error.

[[Jo...]]

Please sign in to view this quote.

Curve distance has that option built in. It works if I recall id circle to od circle. I haven't had the opportunity to attend curve training, so it's trial and error.

[[Jo...]]

Please sign in to view this quote.

Also, projecting a circle, or polar caliper measurement will use theo circle construction, unlike curve using individual data points, if I understand correctly.