Wheel Covers and Base Pressure

Several weeks ago, I wrote about an unusual base pressure reading while testing various spoiler angles on my car, a 2013 Toyota Prius. Specifically, the pressure at the center of the trunk, at 80 kph, was 0—no relative difference from atmospheric! This is unusual because there should be negative differential pressure on the rear surface of a car (its base), where a turbulent wake forms behind it. What was going on?
 
Winds were calm that day, a rarity here in the American Midwest, which may have caused the odd pressure reading. I also had no wheel covers on the car since I had removed the wheels to have new tires mounted a few days before. This last observation led me to conduct some tests to see if the wheel design could be influencing base pressure.
 
This isn’t an outlandish idea, even if it appears so at first. Just this year Dr. Thomas Wolf, former head of aerodynamics at Porsche, published a paper detailing testing of wheel designs on the Porsche Cayenne which found that the pumping action of the rear wheels strongly influences base pressure and can reduce drag coefficient as a car goes faster. Could some interaction between the wheels and base pressure be happening on my car? What is the relationship between base pressure and wheel design, or is there no discernible influence when measured on the road?
 
Let’s find out.
 
Test 1: Constant Speed
 
At a constant 80 kph, I measured pressure at a high position (center of license plate--Location 1) and low position (lower part of the bumper cover--Location 2) on the base of my car with no wheel covers, stock front wheel covers, front and rear stock covers, and front stock plus rear full covers.

Top: bare wheel. Middle: Factory wheel cover. Bottom: Full cover.

I averaged runs in opposite directions, using correct rounding and significant figures. The results:

So, not much difference—but there may be slightly higher base pressure at the center of the trunk with wheel covers, and the rear wheel covers appear to have more effect than the front on that upper reading, whereas adding the front covers raised pressure on the lower part of the bumper but it went back down after adding the rear covers. Interesting.
 
Test 2: Increasing Speed
 
This led me to design another test. This time I would leave the stock front covers in place and change the rear, from the bare wheel to stock cover to full cover. I measured pressure on the base at the center of the license plate at speeds increasing from 80 kph to 110 kph, in increments of 10 kph. As expected, the negative pressure increases in magnitude with speed, correlating with an increase in drag as the car goes faster.

What’s interesting to note here is that both the stock covers and full covers might be increasing base pressure. Because of the resolution of my manometer, I can’t confidently state that the covers are increasing base pressure, only that they might. But two observations bolster this claim: first, I found the same increase in the upper base pressure reading in my previous test and, second, the increase was fairly consistent across speeds and directions.
 
What About the Weird Reading?
 
Of course, all this also tells me something about that 0 Pa base pressure reading that prompted me to test further: it was probably just a fluke. In no other tests since have I measured 0 Pa at that location, in any configuration of the car.
 
My best guess at what was going on blames Mother Nature. I try to test on days when windspeeds are 10 mph or less, which still rules out a lot of days here in the windy Midwest. But the day I conducted the spoiler test with that odd reading, winds were calm—as in, little to no wind. It may very well be that in those conditions, with 0°-yaw flow, there is indeed no difference between atmospheric and local pressure at the center of the base on a 3^rd-generation Prius but, with the rarity of windless days here, I will likely never have an opportunity to verify that. So for now, it remains only a guess.
 
I’ll go ahead and leave the stock wheel covers on the car unless I measure something in the future that suggests they’re affecting drag or lift in some other way. Unfortunately, I’m limited by my measuring equipment at the moment, but if I figure out a way to measure pressure more sensitively, I might revisit these tests. In the meantime, on to other things after this little side trip.