Testing and Improving Stability: Part 2

Fences are commonly affixed to racing cars; road cars, not so much. What will they do here? Only one way to find out.

Last time, I wrote a little about the difference between steady and dynamic directional stability in car aerodynamics. I only scratched the surface of this very complicated subject, but I was able to find, through testing, the effect of adding rear fins on my Prius’ steady stability (as measured by steering wheel angle) in a constant crosswind.
 
After finding that I could add an X-gauge to my Scangauge computer that would read the car’s steering angle sensor, I decided to try some more tests to see what various modifications do to the crosswind stability. Fortunately, here in the Midwest it is often quite windy, giving lots of opportunity for testing.
 
I first spent a few weeks driving around with the STA X-gauge displayed and noted that when the car is going straight down a level, straight road, it reads +1.6° (that is, the wheel is turned slightly to the left, toward the road centerline). This might be due to road crowning, or the steering wheel having been slightly off-center the last time I had the car aligned, or differences in ride height side-to-side, or any number of reasons. The important thing is I know what the “center” reading looks like on the display so I can relate changes in steering angle to that measurement and not 0.0°.
 
For each modification below, “average” is what the display read over most of the test section, and “max” the highest value I saw on the gauge. Because we’re dealing with atmospheric wind here, there’s some unsteadiness baked in, but I tried to be as consistent as possible with keeping the car centered in the lane.
 
As I wrote previously, STA reports steering angle to the nearest multiple of 1.6°. Because of this, there is a “smoothing” effect as smaller changes in the wheel’s rotation don’t show up on the gauge. I’ve found that this makes it easier to read and record steering angle than my previous method of a phone taped to the wheel. But that also means that these reported angles are less accurate than the more sensitive angle measurements from the phone.
 
Spoilers
 
First up, after fins: two of the spoilers I previously pressure-tested. Winds on this day were approximately 20 mph out of the southwest, which means the crosswind component was smaller than I would have liked since all the roads here are on a grid aligned with the four cardinal directions. Testing at 55 mph on an E-W aligned road, I found something curious: the steering angle required to keep the car going straight was smaller heading east (crosswind from the right) than west (crosswind from the left). This might be a result of the fact that the car, for whatever reason, must be corrected to the left in the absence of wind; with crosswind from the right, it’s being pushed toward the center of the road—in the direction it would normally have to be corrected by steering—and with wind from the left, the opposite is true. So more steering angle is required when the wind comes from the left, and less when it comes from the right.

I forgot to take pictures during this test, but this gives you an idea what each spoiler looks like. The lip strip is significantly shorter and smaller than the Hellcat but, as near as I can estimate from pressure changes, still reduces lift about half as much as the bigger spoiler.

I almost didn’t test the lip spoiler, since it’s so small I didn’t think it would do anything. But at the last minute I changed my mind and threw it in the car. That was fortunate because my assumption was wrong!
 

	Standard	Lip Spoiler	Hellcat Spoiler
WB average	+4.8°	+3.2°	+4.8°
WB max	+6.4°	+4.8°	+6.4°
EB average	0.0°	0.0°	0.0°
EB max	-1.6°	-1.6°	-1.6°

 
That’s right, the tiny lip spoiler actually showed a slight improvement in the car’s ability to track straight in a sidewind compared to the large Hellcat spoiler or none at all. Where those two required a steering angle of 3.2° (W)/1.6° (E), the lip spoiler reduced that to 1.6° (W) with no change in the other direction.
 
Fences and Splitter
 
I cut out some cardboard fences to approximate a common hood/fender shape on modern cars. You may have noticed that a lot of cars have a sharp edge along the top of the fender and a scalloped hood. Does this kind of shape impact stability, I wondered?

Here seen on the current Honda Accord; notice the prominent fender line that slopes down to the hood.

Mine are sort of an exaggerated version of that, fences running along the hood cut line a few inches tall.

I tested similar fences positioned along the edges of the back window and the roof over the rear doors.

Finally, I made a quick splitter out of corrugated plastic to see if it would make any difference.

On a very windy day—winds were out of the SSE at 25+ mph—I tested on the same E-W road as before. I’m not sure if it’s because the wind was closer to a true crosswind, but this time the steering angles in both directions were closer to the same. (It might be that this changes with yaw angle; this test was closer to the same yaw angle in each direction than before). As before, I measured steering angle with nothing added to the car first, then with each modification alone:

	Standard	Hood Fences	Hatch Fences	Roof Fences	Splitter
WB average	+4.8°	+6.4°	+4.8°	+4.8°	+4.8°
WB max	+6.4°	+8.0°	+6.4°	+6.4°	+6.4°
EB average	-1.6°	-3.2°	-1.6°	-1.6°	0.0°
EB max	-3.2°	-4.8°	-3.2°	-4.8°	-3.2°

 
Looking at the chart, two things stand out: first, the hood fences have a pretty clear negative impact on stability on this car; second, the splitter might be improving things. I tried to pay attention to how “jittery” the car felt on each run—that is, how much I had to minorly correct the wheel, how light the front end felt, etc.—and there was a pretty clear difference in feel between the hood fences and the splitter. Those fences made the front end feel squirrely, while the splitter had more of a composed, secure feel and required hardly any small corrections of the wheel. I’m planning to test the splitter further, but I would be very surprised if it isn’t measurably reducing front lift. Not bad for something I threw together in ten minutes!
 
Caveats
 
This sort of testing shows that it is possible to affect the directional stability of a car by modifying its aerodynamics, but remember that this is really just showing us one case in one set of conditions: straight roads with a constant crosswind. The steering angle is just one parameter of a complex system, which includes the tires’ slip angle; the compression or extension of the suspension at any corner; the load on each tire; the suspension’s natural frequency, front and rear; roll resistance, front and rear; changes in camber or toe with suspension movement; the longitudinal and vertical position of the center of pressure; the rolling, pitching, and yawing moment; the rolling, pitching, and yawing derivatives—to name just a few.
 
Because the system is so complex, and because I can’t easily measure these other parameters in real time, I will have to rely on subjective testing for further evaluation. On your car, you might test on a closed track or autocross course and see if lap times improve or if the handling is more to your liking; on mine, I’ll implement some of these modifications (specifically, the lip spoiler and front splitter) to see if the car “feels” better on the road (since I use this car on track and in autocross exactly once a year, how it performs on the road in everyday driving is much more important to me).
 
Also keep in mind: just because a modification worked or didn’t work on my car doesn’t mean it will on yours; air flows differently around objects of different shapes. Remember, you must test if you want to know. So go test!