>> "flat finder" system. Which is the tipically accuracy of commercial
>> mask-aligners in the alignment of mask to wafer flat ?
> The bigger problem is, commercial wafers have at least a 1 degree
> mis-alignment between main flat and crystallographic plane. I am
> guessing that is much more than the tolerance of any aligner, and
> hence it is your bottleneck. You might need to etch test trenches
> along the edge, before alignment.

I have the possibilty to measure flat off-axis with x-rays.
By the way idea of trenches is nice.