I have a part that needs to be parallel within .0008" to datum 'A'
We cannot fixture datum A to the granite as it has a boss, we use three ground blocks to fixture the part on three points, then sweep with an indicator.
Old timey engineer says that is the wrong way to check it, and throws a pair of micrometers on it and gets .0005".
He's checking thickness correct? Not parallelism. At this point I don't know how else to explain it to engineering...
Cool, is sitting it on three, ground blocks, and sweeping an indicator on the opposite side, the proper way to check it, without throwing it up sideways on a CMM..?
The datum should be a perfect plane, tangent to the material on the free side of it.
Sweeping with the indicator measure only the parallelism of the plane from a datum passing through the blocks, not the real datum (except if the blocks ha exactly the same height, and cover the whole surface)
It sounds like you are checking it the best way possible with the means you have available. I assume if you were able to check it with a CMM you would be. If you are able to use a CMM setting it up to check it the way
JEFMAN stated is the way to check it
The BEST way to check the parallelism is on a certified surface plate setting on blocks. If flatness of -A- is critical then use 4 blocks, set on -A- and indicate -A-. 3 blocks are "OK" for parallelism but not flatness.
Think of it as a 3 legged chair which won't rock but seems flat and parallel, but can be much unparalleled. Hope it is being ground!