GD&T question - Position of that slot

I believe that you need a second datum, to get the right orientation.
With the cylinder as a primary alone, the part below would be ok (IMO)



So if you add a B on a flat side (upper plane for the perpendicularity or mid plane on both sides ?), you should fix this problem (or A as a plane and B on the cylinder ?)​

Thanks Jef, yes that situation is what i thought and i want to avoid.
If i understand well this FCF could do the job :



I just wonder what would be the difference if i invert A and B in the FCF ?
Anyway i keep going, as i don't necessarily want the same tolerance for the form and location, i'll have to use composite FCF.

In this case, I would create the FCF with B|A instead of A|B.
If B is the primary, then A becomes the projection of all hits of the cylinder on a plane perp to B, to construct a max inscribed circle.
If A is the primary, then you have to search a mid plane containing the cylinder axis and minimizing the tangent external width (Maybe PC-DMIS can do it, but seems to be complicated ?)
IMO

PC-Dmis (using the geometric tolerance command) would handle A|B fine. Cylinder A would constrain 4 degrees of freedom (2 rotation and 2 translation), plane B would constrain the third rotational degree of freedom, leaving translation along the axis of A unconstrained. Depending on the design intent, additional GD&T controls may be required - for example a tertiary datum to constrain translation along the axis of A and a perpendicularity of the slot walls to the uppermost surface (or parallelism to B).

Thanks, yeah that's also how i see it. Anyway the part showed is only a representation, the real part is a bit more complex and the slot design too, the engineer has dimensionned the slot position on the A axis with a tight linear dimension, i think we will let it like this for clarity.

The part showed was only a representation of the real part, which has 2 holes and slots. At first sight, i thought the holes were not on the same plane, but after a closer look they are !
So maybe it could simplify the GD&T, here is what it looks like:



Are you OK that including A-B in the FCF will simulate a plane in that case ?

Or, as the hole is currently dimensionned with a "2x", do you think it is allowed to do something like this ? A would be a plane in that case if we take in consideration that this is both holes.
But yeah i admit it would not less clear than the one above.

Disregard the concept of using Symmetry entirely. Symmetry is defined as a single PLANAR tolerance zone and cannot apply to multiple axis.

Per your last post upper pic, (producing one datum definition of A-B of bores): If the two bores are twisted between each other (not parallel) it will entirely mess up the datum control. Hyphenation of a datum should only occur if the two datums are COAXIAL, this instance has only one axis in common between the two datums... so hyphenating them would be an erroneous GD&T definition.

I think the best bet is to define each bore as A and B datums, then tolerance the 2x slots as you have with profile to [A] , as neil.challinor explained.
Of course, don't forget to tolerance and/or basic and FCF the distance between the A and B datum cylinders.

Hm OK, i'm not on the machine to try all of that, but i see.
Anyway the bores should not necessary be aligned, the important thing is that each slot is aligned to its bore.

So we finally toleranced the drawing like this:



That should do the job. Sure the slot dimensions are basic and their position on the remaining axis are constrained (linear dimensions).
Thanks for your help guys.

Hey, this week i've got some parts so i tried to measure them as defined above.
But it seems it is a bit more tricky, the results are not the one expected
it seems the datum order also have its importance here.

As reminder, B is the cylinder where the slot goes into, C is the mid plane of external faces, D is a plane // to C, passing by B.

For testing, i just took some points on each side of the slot and created SCN1.

Here are the results:



I tried to locate the slot only in different ways (B only / BC / CB / D) and also took the distance of each points from the cylinder B.
I'm not able to say which result is correct !
The Z distances of points seems good so the results of the form location should be OK.

Here is what my real part looks like:



And here is what the analysis shows:



Questions:

- Why are the results pos.3 et 4 different ?
- The point with the largest deviation is 0.025 from the cylinder, i would expect the result of the form position to be 0.050, right ?
- Why is the analysis showing me some points out of tol while their distances form the cylinder looks OK ?
- What other dimensionning could i use to locate that slot from that cylinder ?

1/ Results are different because the datum system is different... ! In the first case, the cylinder is the primary, the secondary have to be parallel to it and minimizes the max distance, the pos 4 defines the mid plane as primary, and calculate a axis parallel to it.
2/ It seems you use ASME and ISO, so the result is Max - min, and not 2*abs(max) (IMO)
3/ can you post the code ?
4/ ...?