If you don't know the characteristics of the NACA curved-divergent inlet for its proper design, I would advise you to instead use a stagnation/pitot inlet, that is, one which projects out from the cowling pointing forward and with a streamlined aft body. See Hoerner's "Fluid Dynamic Drag" for some inlet shapes and their drag coefficients. You might also want to look at the NACA parallel-wall submerged duct which has a square outlet that is much easier to interface to a round duct. It also has better pressure-recovery characteristics than the curved-divergent inlet at duct velocities greater than half of the free-stream velocity. Look at your oil-cooler's spec sheet from the manufacturer, pick the cooling flow from the middle of the graph of heat-rejection vs flow, size the inlet for that flow at cruise, then add 25% to the area. Say  the center of your spec sheet graph shows 300 cfm/ 5cfs, and your cruise speed is 230 mph. Then your inlet area is 5 * 144 / (230 * 22/15) = 2.13 square inches. Increase that by 25% to get 2.67 sq. in. 1 5/8 " square, 2 5/16" X 1 3/16" rectangle or 1 7/8" diameter round, all with nice, rounded lips.  Then provide a diverging duct from that inlet to the oil cooler to slow the flow and increase the pressure, and then a converging duct from the cooler to the cowl outlet, pointed toward the rear, with an outlet area of about 3-3.5 sq in to get the flow back up to freestream velocity.

Paul,

 

My friend with a Lancair 360 and he uses a race car inlet for oil cooler air on the side of his cowling - it is shaped like an angled rearward offset funnel (yes, just a funnel).  He also has to shut off the air during winter since the cooling flow is too much.  He does not need sophisticated exit plenums - just submerged shark gill exits since his cooler is mounted on the cowl itself.  There are so many way to skin the cat - just ask mine......