Another Datum Question

[RussL]

We have a part.
Hole is toleranced true position to D, E & F
Where Datum D is the surface the hole penetrates (about 2" X 2")
Datum E is the centerline of the hole being toleranced along with nine other holes (A hinge centerline about 60" long), we best fit the ten holes to define E
F is a surface perpendiclar to the Datum D surface, rotated about the hinge centerline (D - 90 degrees).
D & E are theoretically normal, but when leveling to a 2" X 2" surface, the hinge line (E) can be out considerably over a 60" length.
They already have a separate perpendicularity call out from D to E.
They basically want the hinge holes in line and the surface perperndicular to it for the head of a bushing.
I suggested asking the customer to make E the primary Datum, this will more than likely never happen though.
How would you align this part?

[Tested]

Can you put a picture of it up so we can see better what your asking.

[Jan d.]

Russ, what version of the software do you run (do you mind filling out the User profile)? This is not difficult to do using V4.1.

You'd level to D and then just measure E (as you do) and then measure F. Just create a dimension using the V4.1 XactMeasure GD&T and it will give you the right answer. No more leveling required.


Jan.

[RussL]

I'm hoping I attached a .jpg correctly.
We use V3.7 MR-2 - XP SP2
We had leveled to E and inspected each individual hole to E
We used the individual perpendicularity call outs (not shown on sketch) for perpendicularity of D to E.
Problem is, again, we are not inspecting to print (model).

[Tested]

Russ,
I'm sure that they think they know what they want, but have little idea of how to explane it. I would agree with you that leveling on -D- is not going to work unless every thing is so perfect as to be ground. If your able to get a good part I truly believe it will be a fluk and nothing that you can repeat on a regular bases. I don't think you can run the same part twice and hold any kind of repeatability. What you have done is correct, and what you would ask for would help very much. It looks as if you already knew the answer.

[Jan d.]

I agree. Poor choice of datums. E ought to be the main datum. This part will be real hard to bring in tolerance. They better make D absolutely perfect. Any small deviation may have a huge effect over 60".

Also, make sure you take many points on D. The more points on D, the less effect 1 bad point can have on the orientation of your plane. Therefore giving E a better chance to be within toerance.


Jan.