Measuring Radiator Flow

Measuring radiator airflow is easy and cheap.

For almost as long as I’ve had the Prius, I have blocked off part of the grill opening—the theory being that reducing flow through radiators reduces the drag associated with pushing air through them, what we call cooling drag.

On my truck, my first throttle-stop tests showed that covering the grill openings completely reduced drag around 4%. Obviously, we can’t drive around like that all the time; most manufacturers today use shutters in front of radiators that can open or close to allow airflow when needed and reduce drag when not.

Most of them use internal shutters, but a few (such as Ford and Hyundai) have started using visible, external shutters. This is the front end of the Ford Mach-E; the slats rotate to an open position when cooling air is needed.

I didn’t want to get that involved in designing an active system so I blocked part of the truck’s grill with plastic covers, leaving the center part of the upper grill open.

But what does blocking the grill openings like this do to the amount of airflow across the radiators? I know it can reduce drag, but does it adversely impact cooling system performance? And if so, how much?
 
Under Pressure
 
The flow of air through a radiator depends on the pressure on its front side and pressure on its rear; the difference between these is what we call a “differential pressure” (or “gauge pressure”), and it can be measured quite easily with some tubing and a manometer. Greater differential pressure (with respect to the back side) means more flow; smaller differential pressure means less flow. And negative differential pressure? Well, that means air is flowing across the radiator in the opposite direction it’s supposed to!

 
To measure the differential pressure, I ran tubing to the front and back sides of the radiator (truck) or radiators (Prius). Like mine, your car probably has a cooling fan behind the radiator; if there’s enough room, you can snake the tube around it like I did on the truck but if not, placing the tube end somewhere near the radiator outlet is good enough. This side of the radiator will serve as the reference for comparison of the front pressure.

Tape the second tube to the front of the radiator:

Then drive the car and record the pressures; I used a test speed of 80 km/hr. A digital manometer that will calculate average pressure over a recorded interval is helpful for this testing, since you might find that the turbulence in the engine bay makes the pressure readings jump around a lot (especially on older vehicles—the truck was much worse in this respect than the Prius). I did runs in opposite directions and then averaged the results.
 
The Prius has an interesting arrangement of heat exchangers, with the inverter coolant radiator sitting above the engine coolant radiator (behind the upper "mustache" grill openings), and behind them the larger AC condenser. I measured pressures on both the upper (inverter) and lower (engine) radiators. The truck has a conventional single radiator—AC was an option but this DLX trim doesn’t have it.
 
Results

	Open	1	2	3	1+3	1+2	2+3
Upper (inverter)	+110 Pa	+70 Pa	+60 Pa	+40 Pa	+20 Pa	-	+10 Pa
Lower (engine)	+80 Pa	+50 Pa	+30 Pa	-	-	+20 Pa	0 Pa

 
Surprise! Blocking any part of the grill opening reduces flow through both upper and lower radiators by at least 35%. That is significantly more than I expected. The grill block I’ve been using on this car is equivalent to number 2; in that configuration, flow through the inverter radiator is cut in half while flow through the engine radiator is reduced by more than 60%. That probably means the grill block was reducing drag, but in the interest of inverter and engine longevity (Priuses of this generation have a reputation for eating inverters and head gaskets), I’ve gone ahead and removed the grill block permanently. You’ll have to decide what balance of drag reduction and cooling capacity you’re comfortable with on your own car.
 
Now, how about the truck? I tested on the same road and in similar conditions as when I did the Prius tests. I expected the results to be similar. Were they?

	Open	1	1+2
Center (engine)	+40 Pa	+40 Pa	0 Pa

 
In a word: no. Unlike the Prius, the truck has no ducting or baffling around the front side of the radiator, no engine undertray, a lot more space in the engine bay, and large gaps between the grill, bumper, lights, and lower valance. I mentioned that the instantaneous pressure readings jumped around a lot more too; I thought there was something wrong with the manometer at first so I moved the reference around but got exactly the same results:

It turns out that flow through the truck’s radiator is just that much worse than on a more modern, optimized system like the Prius. And the two small upper grill blocks were blocking flow across the radiator almost completely! I’ve removed them but left the lower block in place. Next I’ll try sealing gaps on the front end and putting baffles around the radiator to see how much I can improve flow across it.
 
Radiator flow testing is simple, cheap, quick, and enlightening. Try it on your car!