Roundness from constructed circle

[Sekhar]

Hi team,

I construct a circle from auto cone and derive the roundness from that circle.
Is that correct?
one of my Engineer says, extract the geometry dimensions(roundness, straightness etc) form the constructed feature is wrong.

[Quality ish]

I believe a constructed feature is always going to be assumed to be perfectly round/straight/flat etc. If you were to construct a circle using a cylinder and a cone (measuring where a drill point makes contact with the ID of the hole, perhaps), you could get the roundness from the measured cylinder, but I would think that the constructed feature would not necessarily be an accurate depiction of the part.

[JEFMAN]

Hi team,

I construct a circle from auto cone and derive the roundness from that circle.
Is that correct?
one of my Engineer says, extract the geometry dimensions(roundness, straightness etc) form the constructed feature is wrong.

If you construct a circle from a cone (at a given height / diameter), the roundness should be 0, because the constructed circle cut the associated cone (perfect) by a plane (perfect also).
If you want dimensioning the roundness of a circle along the cone, you have to measure some lines around the cone, then intersect them by a plane (generic plane is enough), and construct a circle from intersect points.
In this case, the result will be more accurate than measuring a circle at the same height with vector points because of the drifting errors of cmm. A little defect on the height of the target gives also a defect on the radius, so the roundness is wrong !

[Sekhar]

Got it.

Thanks Ish and Jef