Reasonable Parallellism Tolerancing for stack of flat plates

[skarf2]

Hello all,

I am working on an array reactor cell, which involves a stack of plates which are bolted together. Flatness and parallellism are highly desirable for uniform compression across the reactor membrane. I am new to geometric tolerancing, and I would like to know what a reasonable parallellism tolerance would be. The plates are 10" square by .5" thick.

Thanks in advance.

J

 

[dgallup]

It all depends on how they are made. If they are just stamped you will have lots of pulldown on one side & burrs on the other. If they are ground you can easily have .001" flatness. We lap parts to .0005 mm flatness (half a micron). Specify what you need, it's up to manufacturing to figure out how to get it. Specifying more than you need will be a waste of money.