Maybe not the old concentricity. Certainly not the new dynamic profile.

[3DDave]

Seems like Y14.5 doesn't cover rotational symmetry. I do wonder how something so affecting the dynamics of rotating equipment has been ignored. Seems like if a dynamic profile for rectilinear shapes is added, some consideration for cylindrical coordinate systems deserves inclusion as well.

https://files.engineering.com/getfi...9c-4e8f-b6f1-4df855f5259f&file=coaxiality.pdf

The old concentricity might be interpreted to apply to even numbered features as there are median points between features on opposite sides of an axis; it alluded to "the median points of correspondingly located elements of two or more radially disposed features" but wasn't specific about anything but surface of revolution features.

 

[drawoh]

3DDave,

I take it you are allowing the pointy bits to be sloppy as hell, but precisely identical to each other? As far as I can tell, this was the whole point of the old symmetry specification. I don't understand why I would do this. Is there a quick and dirty way to make five pointy bits identical, even if not particularly accurate?

At some point, you need to specify a rotating and balancing test the part must pass. This does not guarantee that the pointy bits will be identical.

--
JHG

 

[axym]

3DDave,

The old Concentricity characteristic was defined as controlling the "median points of correspondingly located elements of two or more radially disposed features". I would say that this could also be described as controlling the "consistency of radial coordinate in a cylindrical coordinate system" of the correspondingly located elements. The surface elements are defined by intersections with piercing lines that are normal to and pass through the datum axis (which would be the polar axis of the cylindrical coordinate system". Instead of controlling the median points with Concentricity, I believe that a nearly equivalent control could be achieved by controlling the correspondingly located elements with Dynamic Profile. This would definitely be an extension of principle, that is not currently defined or illustrated in Y14.5.

The geometric control shown on your 5-pointed star part seems to be distantly similar if we look at it in the context of the cylindrical coordinate system. Except that it would be the angular "theta" coordinate that is controlled instead of the radial "r" coordinate. Also, the surface elements would need to be defined differently - using intersections of the surfaces with cylinders (or possibly circles) that are centered on the datum axis. This would create sets of 5 correspondingly located elements, that could be controlled with Dynamic Profile. Another extension of principle, to be sure.

I believe that this would define the type of control shown in your figure, where the form of each planar surface is not tightly controlled but the "angular consistency" is. What do you think?

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[3DDave]

The old definition was unclear about what "more" meant. For example, for a hexagon, it is more than two elements but each face has one correspondingly located radial element from which a median point can be determined. This is not the case for an odd number. One could never be certain because the standard never gave any more than a weak phrase.

Dynamic profile requires a linear offset from an existing profile and so would not apply to this case.

Each set of intersections would be a circle - which would just be self-referential.

As to making them the same, but with little concern for accuracy - a punch/die that blanks out the blades one at a time would do that. Or a router and template.

Anyway - just musing about why Y14.5 cannot seem to decide if symmetry in particular and now concentricity are worthwhile controls when situations like this have no obvious geometric characteristics to apply.