Rich,
If you go to
http://www.scribd.com/Lew%20Merrick/shelf you will find a document
O-Ring Seal Application that will walk you through sizing the gland for your o-ring. It is written using inch-based measure, but that should not be too much of a problem. The biggest problem I see in this application is sizing the piston for stability. With large bore piston cylinders you generally need to make the piston long enough to prevent "tipping" as it moves. If you get this wrong, you have stick-slip stick-slip stick-slip out the kazoo!
General purpose air cylinders are normally designed with a nominal .001 inch (0.025 mm) diametral allowance (clearance). The piston would be toleranced as +.000/-.002 inches (+0/-0.05 mm) from the nominal (clearance allowed) diameter. The cylinder bore would be toleranced +.002/-.000 inches (+0.05/-0 mm) from the nominal diameter. The piston OD and sidewalls of the gland(s) could have a finish as coarse as 32 micro-inches (80 micron). The cylinder bore and all the bottom face of the gland(s) should have a finish of 16 micro-inches (40 micron) to insure sealing. You would probably want to design the o-ring groove to provide 12%-15% compression of the o-ring (assuming nitrile o-ring material).
If you use aluminum for either or both your piston and cylinder, you
will want them anodized (sulfuric anodize is just fine, chromic anodize should be avoided). If your operating pressure is 100 psi, then your
design pressure should be 150 psi. If you are using 6061-T6 aluminum for the cylinder, the allowable yield stress is 40,000 psi (MIL-HNDBK-5J). This would set your minimum wall thickness (at the thinnest part of the cylinder) to: t = Pr/s = (150 psi)(inside radius in inches)/(40,000 psi). If you use some other material for the cylinder, substitute in the maximum allowable yield stress for that material (you can e-mail me at tangent@olympus.net if you need such values) for the "s" in the equation.
Does that help?