How is secondary Datum "Anchored" to Primary Datum?

[TopPocket]

Hi, this is an offshoot question from my previous thread but it's quite general so more likely to have an answer.

When defining datums like in the drawing below I understand Datum B to be created as a virtual axis, perfectly perpendicular to Datum A (unlike the datum feature).

However I am unsure as to what point is used to anchor the axis. Is it the point it intersects Datum A or is it the mid point of Datum B or something else?

 

[TheTick]

Datum B is effectively the center line of a cylinder. It is in its own space, as defined by the cylinder, independent of datum A. In 3D reality, it almost certainly does not intersect datum A. The perpendicularity frame in no way mandates the intersection in 3D.

 

[TopPocket]

Oh it was my understanding that having Datum B coming off the perp frame was a way of indicating it should be exactly perp to A and if it came off the dimension it would be the axis of the cylinder (green dashed line). But you're saying it will be the axis of the cylinder regardless?

If that is that case is there a way to define what I've labelled Datum B?

 

[TheTick]

Wait... I read this whole thing wrong. I thought datum A was the other cylinder.

 

[TheTick]

But still, datum B has no "anchor". It exists unto itself, but controlled within .05 perpendicularity. There is no other location control.

The answer you seem to be seeking ultimately is dependent on whatever control frames reference these datums.

 

[Burunduk]

TopPocket,
I assume you are working according to ASME Y14.5 (from a brief look at the other thread, I've seen that you use some ASME terms).
Datum B is the axis of the smallest circumscribed perfect cylinder that contacts the actual datum feature B and is exactly perpendicular to datum A.*
The only relationship between datum plane A and datum axis B is an orientation relationship (basic 90°). There is no and can't be a location relationship between them, therefore you don't and can't "anchor" the axis the way you ask about.
By the way, datum B is derived as described above, regardless if the datum feature symbol attaches to the size dimension or the perp. feature control frame.

*Assuming both features are referenced in a feature control frame as |A|B|

 

[TopPocket]

Okay thanks for your replies.

My use of the word "anchor" might be a touch misleading but I believe your reply indicates the answer is the higher of the two Datum B axis in my first sketch.

The axis of the smallest perfect cylinder, that is perp to Datum A, that can fit around datum feature B will be positioned between the contact points of said cylinder.

In other words Datum A defines the orientation of the axis of Datum B and the datum feature cylinder defines the position of the axis.

 

[Burunduk]

In other words Datum A defines the orientation of the axis of Datum B and the datum feature cylinder defines the position of the axis.

That is correct. Note, however, that the datum that sets the orientation for the other datum depends on the datum precedence order in the feature control frame. What was described above is correct for the |A|B| datum reference frame, as shown for the position tolerance in your other thread about the same part.
If the reference was to |B|A| then datum B would be the axis of an independent cylinder which is the smallest that can contain the actual physical datum feature B without any other restrictions. Datum A would be a plane perfectly perpendicular to datum axis B that also has to contact the actual surface of datum feature A.
If each datum was referenced separately in a different feature control frame, then each datum would be independent of the other and act as a primary, for each separate datum reference frame. You may know this already, but just to make sure.

 

[TopPocket]

This is very well explained - thank you.