Ah, much clearer now !

So you don't know where your maximum is. I think I would try a binary search method first, so I am searching a matrix of fairly constant element number, but of smaller and smaller areas. Or did I miss something ?

What sort of beam pattern have you got. Elliptical ? How many side lobes ?

Still don't see why you need four piezos though..
I am interested in electro-optics and adaptive optics anyway, so this is fun!

Steve