Jim-
jumping ahead of your cartooning effort, it's my sense you're
NOT looking for "An" issue but rather a
collection of "issues"! the suggested cartoon was recommended as both a device to collect relevant design data
AND to change the focus from things to the
system. finned-tube heat exchangers aren't hard and there's a great deal of material available on how they work. the hard part of heat transfer is cramming it into a small irregular space and then attempting to provide it minimum air flow; under those circumstances it comes down to details AND attention to detail! from your problem statement heat exchanger water flow rate AND inlet air temperature are fixed. that leaves the product of core area and air flow rate as the only
cooling system design degree of freedom....................... heat transfer goes up with increasing air flow rate and/or increasing core area (assuming fixed coefficient of core heat transfer
AND equal air flow rate through the core across the core's face). smoke, oil, or CFD will confirm your current installation is getting little, if any, dynamic pressure recovery from behind the wing. with little or no dynamic pressure recovery to promote air flow through the heat exchanger core, you're left with only the pressure differential from the fan. it's been my experience there's two problems with fans: air flow drops off with pressure (either back pressure OR pressure drop through the core)
AND they have to be spaced away from the heat exchanger core to assure something approaching equal air flow across the face of the heat exchanger core. the attached collage shows experimental hardware used to explore the cooling system design trade space for a FFord. the experimental hardware implementation is dated but the cooling system concepts examined are still applicable today! to help with the decimal point, for a 1600cc Kent I'm currently using a single 16" fan that flows north of 3000cfm using a 2" space between the core face and the heat exchanger core with a 1/2" edge radius. a simple descriptive geometry solution for a spaced circle to rectangle transition is more than adequate. it's my experience that a sealed transition is
not negotiable! without one it's my opinion that best-on-best you're likely to be using only half the projected area of your fan of your heat exchanger core which clearly isn't the answer!
Art
artesmith@earthlink.net