Target points on edges

[RLD89]

I have part with complex surface with 4 target points for primary and 4 edges, no two opposing edges.
I want to use two target points on each of two opposite, but non-opposing edges to create a compound secondary datum to use as a center plane. Is this acceptable?

 

[drawoh]

RLD89,

I take it you are using datum targets.

You need three targets to define a rigid surface. You need two targets to define your secondary face or edge, and one target to define your tertiary datum.

If I were using datum targets to locate a centre line, I would apply two targets to one side of the feature, and one to the opposite side. Is the width of your part sloppy? I consider datum targets to be a strategy for locating inaccurate parts like weldments and castings. I avoid fixturing to inaccurate features of size. Are you going to specify your datum at MMC/MMB?

Have you a sketch or a drawing?

--
JHG

 

[RLD89]

drawoh,
Thanks for your help. The part width is well controlled. Background to clarify: I am using 3 target points on the OML for primary. I have another part to center in the opening and the important thing is to have equal gaps on all four sides between the parts. This is not a tough task in itself; only calling out the correct GD&T is (at least for me).
The lines on each two edges opposite edges are parallel on the OML, but the edge surfaces are non-planar and non-opposing. I created 2 target points on each of the opposite longer edges as a secondary B-C. One of the other edges has a single datum target to establish the tertiary D.
Everything is RFS so no modifiers required. The 1994 specification does not allow a line to be a datum, but a point, axis or plane is. If I could use the lines as Datums and they could become the compound secondary, I could avoid target points altogether. It seems like a lot of work as an end around not to having a feature of size to create a center plane to. I suppose, this is not really what target points were invented for. I may be overcomplicating it all, but I am trying to stay within the rules.
Please correct my logic.

 

[drawoh]

RLD89,

ASME Y14.5M-1994 allows nothing other a feature to be a datum. Two datums can be used to define a plane, a line or an axis, as per Figures[ ]4.20 and[ ]4.21 in the standard. You are doing nothing like that. My suggestion of three datum targets for the secondary datum is a much better idea. My rule of thumb is that if a feature of size datum is ten times as accurate as the features toleranced from it, I don't need mess with MMC, and work out the design of fancy inspection fixtures.

--
JHG