Russell,
Lots of good advice above. I've spent a fair amount of time trying to make my cooling system as efficient as possible. This is what I've learned with respect to the radiator:
I can't figure out how to enter Greek letters, so the following will need to suffice:
- Q: The amount of rejected heat; this is what you're looking to increase.
- A: Area of radiator core; this is limited by what will fit, although you can get creative and angle the radiator, use slim side tanks or use multiple radiators.
- DT: Delta T is the temperature difference between the coolant and the air flowing through the radiator. There isn’t a lot that you can do to affect this DT term (you can affect DT elsewhere as discussed below). However, as you point out, putting an A/C condenser in front of the radiator preheats the air which decreases DT.
- K: Radiator heat conduction coefficient; this is where a high-quality radiator helps. K will depend on the quality of the material, thickness of the core, the design of the fins, turbulator, etc. Note that while you want laminar flow through much of the cooling system you want turbulence inside a heat exchanger. Why? Because laminar flow results in the fluid in contact with the fins will cool and DT will decrease. Turbulence ensures that hot fluid is constantly in contact with the fins and therefore increases DT. In addition, as coolant flows through a row it cools which reduces DT. Thus, if you were to look at a color-coded heat rejection image it would show a gradient. For this reason, for a given core size, it is preferable to have shorter rows because that results in a higher average DT. This is the same reason why it is advantageous to plumb multiple heat exchangers in parallel rather than in series.
- M: mass flow rate of air through the radiator.
M is directly related to the mods that you’re currently making. When sitting in traffic M is dependent on your fans and having a shroud with a good seal to the radiator. As Joel points out you want some space between the fan blades and the core. I bought a custom radiator/shroud solution from C&R so I don’t have any insight on how to determine what that space should be.
When the car is in motion M is more complicated. M will be driven by the pressure differential between the front and rear of the core. There are multiple approaches to increase the pressure differential, but in all cases you’ll want to ensure that all air entering the radiator inlet is forced through the core. Many cars, including mine, use a diverging/converging duct. With this approach the inlet is smaller than the core (the air diverges) and the outlet is slightly larger than the inlet, but smaller than the core (the air converges). When air diverges it slows down which increases pressure and when it converges it speeds up which decreases pressure. If you’re wondering how well this works, just look at an airplane wing. The airfoil speeds up the air on the top to decrease pressure.
Diverging/Converging duct. The orientation and geometry for a 40 or SL-C would be different, but the concept is the same
If you’re wondering why the outlet is slightly larger than the inlet, the logic is simple. If you’ve ducted properly, everything that goes in must come out. However, the radiator has rejected heat to the air which has expanded it. The outlet’s size is increased to accommodate that expansion.
Before I understood these basics, I would have been inclined to maximize the size of the outlet which might have resulted in the exact opposite of what I intended. This seems to happen a lot to me when it comes to aero
One other helpful tool to improve M is the Gurney flap. You can install one on the leading edge of the radiator outlet to create a low-pressure zone. It would be easy to add a temporary one with a piece of aluminum to see if it helps. If so, it would tell you that you had an issue with mass flow and you could keep it or make other changes to increase the pressure differential.
I have seen some people claim that you don’t want to move coolant too quickly through the radiator because it won’t have enough time to reject heat. This is BS. Yes, it is true that the longer it’s in the radiator the more that it will cool. However, as it cools DT decreases and heat transfer efficiency decreases. When you're above your target temp you want to move coolant as fast as you can short of causing a cavitation problem.
Properly bleeding your system doesn't remove air bubbles created by your system. You may want to consider a coolant swirl pot to deaerate the coolant. They are common on race cars and they are inexpensive (or you can make your own). Carrol Smith states
“I still consider the water system de-aerating swirl pot (as described in PREPARE TO WIN) to an absolute necessity on any racing car” and here’s a pic from one of his books:
My last point is that the only place where I’ve seen a V presented as “aerodynamic” is for a trailer. Sure, it’s better than a flat rectangle, but that’s a low bar. If possible, put a little round on the V.
I’ve spent a fair amount of effort doing CFD and my car will have a diverging/converging duct with careful ducting, attention to the inlet and outlet sizes and a molded Gurney flap. I'm using a Pierburg brushless pump and a custom expansion tank and coolant swirl pot. Hopefully it all works as intended!