total run out on a datum 1

[DJ Nelson]

We have a step pin the tolerancing seems to be incorrect.
See Datum A on the Attachment they call out total run out of .0015 on datum A.
in sec 9.2 one part in the standard that resembles our pin we have it seems as if you can not applied Total runout to the datum.
In sec 9.4 the standard it gives another example were if you have multiple datums a significant length or axis separation.
We think they should maybe change the call out to either straightness or cylindricity.

 

[greenimi]

What matters is whether concentricity is important to control a feature.

3DDave,
I would say it is not important in the S-cam case. Not sure why the median points are "more important" than the S-cam surface.
And you haven't answered (at least not directly) my questions above. I still failed to see why concentricity is preffered to profile in the S-cam case proposed. I don't mind to use the S-cam as a poster child to argue the needs for concentricity in ASME, but some strong arguments should be provided to justify those needs.

 

[greenimi]

I also found this old thread (from almost 10 years ago)

https://www.eng-tips.com/viewthread.cfm?qid=364973

Still not convienced specially now that the dynamic profile modifier is out (to circumvent pmarc's issue below)

Personally I would keep directly toleranced dimensions as far away from the oval/elliptical shape as possible. If one is interested in controlling its size and form within the same tolerance, profile callout applied to the true (basic) contour should be the first choice. If one is interested in separating its size from form... then in my opinion current state of the art does not offer solely graphical/symbolic tool(s) to accomplish this.

 

[3DDave]

They control different things. Concentricity is to control relative variation. Profile controls individual variation.

I can't picture this for you.

There are strong arguments for not having dynamic profile, like lack of actual applications.

 

[greenimi]

Concentricity is to control relative variation. Profile controls individual variation.

Yes, but I guess for the S-cam you need the relative variation on the surface/ feature and not relative variaton on the median points (as the concentricity offers, median points are derived from pairs of surface points. In the definition of Concentricity, the points are intersections of piercing lines passing through and normal to the datum axis).

I am still not understanding why we care (in the S-cam provided example)about median points? Why the relative variation should be rooted from the median points and not from the feature/surface itself?

 

[3DDave]

Median means equal disposition about the center. You want the contacting parts to have equal movement at the same time.

There is no control that says a profile on one side shall have exactly the same profile on the other side.

Concentricity does this. Profile does not.