Looking back through some old M E s came across a formula to aid arriving at the correct interference.
A = (1.00625 D) + 0.5 where A is the interference in Thous, and D is in inches.
Presumably A would be microns, and D would be mm if working in Metric units.
Seems a bit large to me, being overcautious.
0.001" on 1.000" steel into cast iron seems to work OK for me, so maybe 1/1000 on diameter might be an adequate rule of thumb..
Anyone like to confirm?
Howard