The problem a lot of people have is that they try to get the cart before the horse. You
first have to place six INDIVIDUAL holes into the overall component, three on each side. The overall component can have functional datums assigned. You have done that by identifying two features as datum A (planar) and datum B (axis). If the "clock position" of the 3-hole pattern (one pattern on each side of part) is important then you'll need to establish an orientation datum (tertiary) to constrain rotation about datum axis B. If that constraint is not necessary then only two datums will suffice (A and B). I have always said that (in most cases) we don't design datums... we design features that may become datums for downstream features to reference from. First you have to put some holes (6) in a big honkin part! Then, you can choose one or more of those existing holes to act as datum references to control other features. Take a look at my suggestions on the attached marked copy of your pdf file. You'll see that I locate the 3-holes on each side of the part to a common datum framework (A, B, T). I then select the individual hole on one side to act as a datum reference for the location of the opposing hole (coaxiality).
The previously proposed two single-segment TP control shown in the center of your file called "Design" is an incorrect usage. If the upper FCF constrains the orientation to datum A to Ø0.05 the lower FCF tolerance becomes incoherent. The lower FCF can only further constrain the upper control (be less that 0.05 tol in lower FCF).
My thoughts, anyway!
Attachment:
design_more.pdf [83.57 KiB]
Downloaded 261 times