alignment problem

[d.evans]

I have a part that has a Cyl-type hole (1.130") in the center (Datum C), an odd shaped outer edge with a straight section (Datum B) and a flat bottom surface (Datum A).

There is a hole pattern consisting of 4 small .098" holes around the the center 1.130 Datum C hole.

I measured the 1.130" Datum C hole as an auto cyl, the flat Datum A as a several hits on the surface plate itself constructing a Plane from those hits. Datum B is the straight portion of the outer edge. I took a few hits and made a line. Did my alignment Level to A, rotate to B, origin at C for x and y axes.

The .098" holes callout though has a Feature control frame that goes .098m|Cm|A|B.

This says to me, level to C, rotate to A, translate to B. Is this thinking right?

My problem, if this is right is that A and C are on the same axis. Can you level to the C cylinder, and rotate to A plane when they are BOTH Zplus?

I tried both doing the TP dimensions checking the "use datums" since the first datum has a MMC condition as well as the feature itself.

Then I tried constructing a new alignment that matches the callout, so I don't have to check "use datums", but that's where I'm stuck. C and A are both Z plus. Can you level to Z plus and then select another feature which rotates to Zplus about Zplus?

PCDMIS 3.7mr3, NOT using cad for this one.
I'll try to get a pdf for you guys/gals to view...

[d.evans]

here's the print

thanks for any help.

[Matthew D. Hoedeman]

The callout is screwed up, they put the C(M) in first because they wanted the hole pattern to be checked to C, but can't understand that it is already aligned to C from the original alignment, basically, whoever made the print is an idiot. Do your TP to the ABC alignment, DO NOT TRY to use the (M) for the C datum since the USE DATUMS part of TP will do nothing but screw up your report (by 'sliding' C around to get the best feature location).

[d.evans]

I thought the same thing

(not that the guy is an idiot), but that the print isn't right. thanks.

[Wes Cisco]

Have you considered the likelyhood that the engineer who designed this part has an incomplete understanding of GD&T? That is an unfortunately common occurance. I would level to the cylinder, rotate to Datum B and set origins as you did previous, x&y on Dat C and z on Dat A. You learn to either make a stink and get clarification for the stuff that isn't right, or you just do what needs to be done and get one with it. I usually put a note in my report about how I lined up and why when it is different from the FCF. HTH

P.S. I would not try rotating to Datum A after having leveled to the Cylinder. I think the engineer is saying Datum C is most important, but that Datum C and the hole patern should be perpendicular to Datum A, he just doesn't know how to do it properly.

[Farmer]

I'm no expert with G,D&T but shouldn't you level to A, zero to B and rotate to C?

[Jeffrey Lovely]

Level to A
Rotate to B
Origin to C (INTERSECTION) AT A/C

[Jeffrey Lovely]

Level to A
Rotate to B
Origin to C (INTERSECTION) AT A/C

A few added thoughts or questions to see if part will funtion
with the above alignment
as the housing gets smaller does that allow more tolerance
and should perpidicularity be applied per LMC? to -C-
what gets bolted together
if housing is to large will it bind and if crooked can bind even mor
These may have been the designers thoughts

maybe add perpidicularity and explain why, if this is an in house produced part
Datum A is still 3 points in my book)

[d.evans]

P.S. I would not try rotating to Datum A after having leveled to the Cylinder. I think the engineer is saying Datum C is most important, but that Datum C and the hole patern should be perpendicular to Datum A, he just doesn't know how to do it properly

When I tried to rotate to Datum A, its almost as if the program shrugged its shoulders.

Normally when I'm doing an alignment, I can see the slight shift of the features occuring on the display, adjusting for what is happening in my alignment utilities window, but when I tried leveling to C and rotating to A, the A part seemed to make no 'shift' on the screen. Also the dimensions all came back EXACTLY nominal on EVERY feature... (usually a sign that something is not quite right somewhere) not even a tenth of a .001" off in any direction.

I figure, if the holes are located to within .006" TP on a plastic part without the help of MMC on the C datum, it must be really well made.

When I dimensioned the holes per Matt's suggestion, they were al rosy.