I think it's an area calculation, in which they set barrel cross-section and the area af the "wall" of a cylinder that would fit in the gap (if you will imagain that). Assuming this is the case...

pi*R^2 = pi*2R*G

Where R = radius and G = gap

deviding by pi and one factor of R gives "G" on one side gives

R = 2G

1/2R = G

Or, in terms of diameter

1/4*2R = <b>1/4D = G</b>

But, now, we are stuck wondering "is this truely sufficent?". After all, compressable fluid flows are complex beasties, and surely can't be modled perfectly by a simple area calculation.

For this reason, I'd err a bit on the edge of caution, and use a bit more gap than the calculation gives you. It doesn't affect preformance as much as some would try to imply, or at least the GGDT doesn't think so...

Registered users: Bing [Bot], Exabot [Bot], Google [Bot], Google Adsense [Bot], Yahoo [Bot]