Hi Everyone

The attached pdf shows a simplified version of a plastic part I am working on. This part mates with a pcb with optical components which needs to be located precisely relative to features on the plastic part. As you can see in pdf I am planing to use the top of the screw bosses as targets for datum A (primary), none controversial I hope. In creating datum B (secondary) & datum C (tertiary) I would ideally like to use the conical (because of draft) location bosses next to the screw bosses. This is because these features are the nearest things I have to a direct link to the mating pcb, if I use other features I feel that I will be de-emphasizing the function of the component for ease of drawing & inspection. Options I have so far thought of are as follows:

Option 1 (easy, but suspect this is not GT&D legal)
Use the axes of 2 location bosses to create datum B (secondary) then use the axis of the back location boss to create datum C (tertiary)

Option 2 (fiddly, but maybe is GT&D legal)
The top of the location bosses are small (dia 2mm) flat circles I could place datum target in the centre of the circles to create the datums (2 targets for datum B, 1 target for datum C etc...). If this approach is 'legal' would it be best to position the targets using GT&D or 'normal' dimensions?

If anyone could give me a pointer as to which of the above 2 options (or maybe a different option) is the best way to proceed it would be much appreciated.

Thanks in advance.

First, use all 4 boss tops as Datum targets A.
Second, use all four tapered posts, at a set diameter as datum target lines B1 thru B4. In the DRF, reference A and B(M). All d.o.f. are eliminated, and it reflects the actual functionality of your workpieces. This works with either '94 or '09.

Jim Sykes, P.Eng, GDTP-S
Profile Services www.profileservices.ca
TecEase, Inc. www.tec-ease.com

MechNorth, thanks for the response much appreciated.

1/ Using all 4 boss tops as Datum A targets - Agreed

2/ Using 'target datum lines' sounds interesting & had not occurred to me. However, I could do with a bit more detail as to how the datum lines are positioned on the part. I am guessing from your description that this solution work something like this:

• 'use all four tapered posts' = This puzzles me (so I suspect that I am misunderstanding), why do I need all 4 posts? Surly 3 would be enough, 2 for the secondary datum & 1 for the tertiary datum.
• 'at a set diameter' = I am guessing this is equivalent to a set hight up the conical location posts?
• 'datum target lines B1 thru B4' = Laying a datum target line across two location posts at a dimensioned height?
• 'In the DRF, reference A and B(M)' - Understood no problems with this, I assume that '(M)' is MMC

Am I reading your response correctly or have I got it totally wrong?

Any further clues to the ignorant gratefully received.

Basically right, Perp.
Use all 4 posts because you don't know which posts will actually make contact with the holes of the PCB. If you arbitrarily select only two or three, then you have a significant chance of not being right.
Using all 4 gives you a pattern of datum features establishing the datum, without priority of one over any others. This is reflective of how the part actually mates. The primary datum (A) takes away 3 dof. The pattern of datum features eliminates all 3 remaining dof at once.
I would suggest putting the datum target lines at the same height as the primary datum target areas. The simulator for the 4 posts would represent the 4 holes in the PCB; the 4 lines of contact should be at the MMC size of the holes in the PCB (worst-case fit).
(M) is indeed MMC.

Jim Sykes, P.Eng, GDTP-S
Profile Services www.profileservices.ca
TecEase, Inc. www.tec-ease.com

While I totally agree with Jim on datum features selection, I have some difficulties in visualizing 4 datum target lines simultated at MMC size representing worst-case size of the holes in the PCB. Are you, Jim, thinking of applying basic diameter dimension to that target lines equal to the MMC size of holes? If yes, can B be really referenced at (M) in a FCF if this will not be MMC of datum features but MMC of countepart features?

I made some assumptions about the designed fit of the posts within the pcb holes, specifically that at worst case they are line-to-line fit (i.e. the MMC of the hole is the same as the MMC of the post). That's common with molded plastic part design because of the shrinkage. Pmarc is right that setting a "basic" [BSC] size (equal to the MMC size of the holes) in the DRF would be better, but it is only available as of the 2009 release. To do it in the '94 standard, the basic diametric size of the datum target line could be dimensioned on the drawing, and the datum reference could just reference the datum-B rather than B(M); people may try to interpret a "tolerance" onto the size of the datum target line though, as this was not directly discussed in the '94 standard.

Jim Sykes, P.Eng, GDTP-S
Profile Services www.profileservices.ca
TecEase, Inc. www.tec-ease.com

Hi quick note to say thank for the information. I have used the 'datum target line across 2 location posts' suggestion on the 2 posts in the middle of the part. Using a datum target line like this is new to me, all the book examples I have seen show the target line crossing a 'more or less' continuous surface not 2 separate features, this will be useful In the end I had to go for point targets on the 2 location posts @ the back because of other things (not shown on drawing) going on in that area. But, as suggested, I have used a 'cage' of datums to lock down all 4 location posts. Thanks again.