# Thread: Flatness with Datum Alternative

1. ## Flatness with Datum Alternative

Hello All,

I am in need of controlling a part and the method I keep coming back to is Flatness in relation to a Datum. Of course I know that flatness should not be used with a datum so there may be an alternative to what I am trying to do, I just don't know it! (My computer is not allowing me to post a document or image so I'll have to opt for the million words version)

My part is essentially a rectangular thin bar with an axle hole towards one end. I wish to control the part so that when the part is held by washers about a diameter a little larger than the axle hole the tip of the part does not bend or twist away from the straight edge / axis created by these washer contact points.

My thought was to make this theoretical washer contact point a Datum Target area and then specify that the remainder of the surface must be flat within .004 in relation to this datum area. Note that the datum and the surface being controled by the datum are along the same plane...

If I specify just flatness or straightness on the part there is a chance that only the area around the axle hole will bend up slightly (.003 for example) so that by GD&T the part is in tolerance but when the part is held onto the axle hole it bends once then the length of the part deviates way off at an angle.

I could dimension the edge to be perpendicular with the axle hole however the part is very thin I don't want to create a challenging measuring situation. I am hoping to build a fixture that simulates the axle and clamps the part similar to how the washers would. A vision system could then very easily tell me if either side bent away from this straight edge / axis created by the datum area against the fixture.

Previously our drafters called out a datum target area "A" and then called the length of the part "Area X". A leader was then used to say "Part must not bend or twist .004 away from Datum A".

Apologies again for the lengthy detail and thank you for any input!

2. Welcome to EE!

Sounds like you need a parallelism of the surface relative to the datum target surface area. Orientation tolerances when applied to a plane also limit the flatness of that plane to the Orientation tolerance specification. So, if you have a parallelism of .004 relative to the Datum target area "A" the flatness of the surface will not exceed .004.

3. Thanks Kelly!

That was the way I was starting to lean so thank you for the clarity. I did not remember that parallelism controlled the flatness as well.

Thanks again!