Splitters are a popular modification for racers; stop by a
forum like Grassroots Motorsports and you’ll find thread after thread on adding
a splitter to an existing car. Road cars? Not so much, and I suspect most people who screw an eBay lip to the front of their cars do so for looks. But I wonder, could a
splitter on a road car do anything beneficial, in terms of improved stability
or reduced drag, or…something else? Should I fit one to my car? Let’s find out.
Initial Testing
I mocked up a splitter last year out of corrugated plastic.
This isn’t the best choice of material—ideally, you’ll want something stiffer
and able to be bolted to the car—but I tried to design it in such a way that I
could still get useful information from it. I folded the coroplast over itself,
so it’s fairly stiff, and I wrapped it over the existing (short) splitter so
that, when taped down on the top and bottom, it doesn’t flex much and any force
the splitter develops should be transferred to the car, at least in part.
My first two tests of the splitter relied on measuring changes in steering angle in a crosswind with and
without it fitted to the car. Both tests showed that there might be a slight
benefit to directional stability from adding a splitter—much to my surprise,
because of the shortcomings in this mockup I described above.
Recently I decided to revisit a splitter, after seeing some
interesting flow patterns on my car.
Tuft Testing
When I conducted an initial tuft test of my Prius, I noticed that the flow at the front of the car had a curious
feature: the tufts on the stock splitter pointed forward and wrapped around
under the car. Adding the splitter extension—which sticks out about 200
mm—completely changed this flow. The recirculation is gone with the extension
in place:
Further, look at the tufts on the license plate and just
above the cooling air opening. Without the extension, they point down or out to
the side, indicating a stagnation line (the point where the air splits between
going up and over or down and under the car body) somewhere around there or
slightly above. With the extended splitter in place, tufts in those areas point
straight up or up and to the side—the stagnation point has moved down.
To continue my investigation, then, I decided some pressure
testing was in order.
Pressure Testing
I first measured pressures across the splitter itself to see
what kind of force could be developing there. Taping a pressure patch to the
top side of the extension and another directly below it on the bottom, I found
the following pressure differences between the two (averaged in opposite
directions at 80 kph), measured at the center, then outside, then a point
midway between:
|
|
Center
|
Middle
|
Outside
|
|
Extended
Splitter
|
+320 Pa
|
+290 Pa
|
+190 Pa
|
Taking the average of these across the splitter and using
its measured area of 200 mm x 1200 mm gives a downward force on the splitter of
around 60 N or 6 kgf. The change in lift of the car as a whole might be more or
less than this, due to changes in pressure elsewhere—as we will see in just a
moment—but these pressures at least give me an indication that the splitter may
reduce front lift, and change the pitching moment (the tendency of the
car to rotate around its y axis, i.e. nose up or down) and lift balance.
To find out if this extension was improving radiator flow by
sending more air through the grill, I measured pressures across the cooling
package, toward the bottom of the lower
engine coolant radiator. These results were particularly important to me, since
I’m getting this car ready for a road trip that will take me across the hot
Southwest, and anything that will improve
cooling airflow might be worth implementing permanently before I leave:
|
|
Standard
|
Splitter
|
|
Radiator/AC
Condenser
|
+120 Pa
|
+120 Pa
|
Okay…no change. After seeing this, I measured the pressure
in the engine bay compared to a pitot tube static reference and leaving the
sensing tube behind the radiator in place: -30 Pa with the splitter extension,
-20 Pa without. That means the pressure on the front face of the radiator is
actually lower with the extended splitter on the car—but because the
pressure in the engine bay is lower as well, the difference across the
radiators is the same. This is not what I expected after seeing those tufts
pointing straight back into the grill opening!
Next, to see if the splitter and its associated change in
stagnation height are affecting pressures elsewhere, I measured pressures on
the rear window and rear diffuser using a pitot tube static reference:
|
|
Standard
|
Splitter
|
|
Rear Window
(upper)
|
-40 Pa
|
-30 Pa
|
|
Rear Window
(lower)
|
0 Pa
|
0 Pa
|
|
Diffuser (front)
|
-70 Pa
|
-60 Pa
|
|
Diffuser (rear)
|
-40 Pa
|
-50 Pa
|
How about that—the splitter extension definitely has an
effect on flow at the back of the car. The pressure is lower at the top of the
window without it but the same at the bottom of the window/in front of the
spoiler. Underneath the car, the front of the diffuser (you can see where I
placed the pressure patch in the photo above) develops a lower pressure without
the extension, along with a higher pressure at its rear. In mathematics-speak,
this is a stronger pressure gradient, or a larger change in pressure
(here, positive—it’s increasing front-to-back) across the same length of
diffuser panel without the splitter extension than with it.
 |
| The pressure gradient is the slope of each line. Steeper line = stronger gradient. |
This indicates that the diffuser works more effectively (N.B. a diffuser is a thermodynamic device whose purpose is to increase pressure in a fluid flow at the expense of its velocity) without the
extended splitter on the front of the car, perhaps because less air is going
under the car with the long splitter in place and lower stagnation point. With
only the stock, short splitter, the diffuser has greater suction at its forward
end and higher pressure at its outlet.
Update: After some offline discussion about lift and stagnation height, I decided to go back and measure pressures on the front of the car with and without the splitter extension in place. I took the average of bi-directional readings in five locations on the centerline at 80 kph, shown below without splitter, with splitter, and the change:
Pressures on the hood are lower with the extension in place, probably a result of sending more air over the car. Stagnation pressure (measured on the license plate) did not change, and the readings on the engine undertray, curiously, got lower at the front but increased near the center.
 |
| Pressure measurement location, close to the center of the engine undertray. The pressure here was much higher than I expected in both configurations—at atmospheric (without splitter) or above (with it). |
Coastdown Testing
Finally, I did some coastdown testing using the method I developed earlier this year to see if I could measure a change
in drag with the splitter. Ten runs were made in a single direction in quick
succession on a calm day, 5 with the splitter extension taped to the front of
the car and 5 without. This time, I recorded the speedometer display with my
phone and later went through the video frame by frame for greater accuracy than
using a stopwatch. The results, measured in seconds:
|
|
Standard (110-100 kph)
|
Splitter (110-100 kph)
|
Standard (100-90 kph)
|
Splitter (100-90 kph)
|
Standard (110-90 kph)
|
Splitter (110-90 kph)
|
|
1
|
8.11
|
8.03
|
9.87
|
9.13
|
17.98
|
17.16
|
|
2
|
8.91
|
8.92
|
9.07
|
9.94
|
17.98
|
18.86
|
|
3
|
8.92
|
8.04
|
9.13
|
9.87
|
18.05
|
17.91
|
|
4
|
8.10
|
8.90
|
9.88
|
9.20
|
17.98
|
18.10
|
|
5
|
8.11
|
8.91
|
9.95
|
9.13
|
18.06
|
18.04
|
|
mean
|
8.43
|
8.56
|
9.58
|
9.45
|
18.01
|
18.01
|
As before, I did some confidence (statistical) testing to
check these data rather than just compare the simple averages. All tests returned very high p-values, which means these results
do not show a change in drag; any change resulting from the splitter
extension is too small to measure with this method. Note that this does not
mean the splitter isn’t affecting drag—only that I can’t say one way or the
other based on this test, and if the splitter is increasing or decreasing aerodynamic drag, that change is likely small.
Decision Time
Now that I’ve taken the time to measure all this, I can make
an informed decision as to whether or not I should install a permanent
splitter.
Keeping in mind my goals (low drag for best
fuel efficiency, good directional stability and low
lift without increasing drag) and taking into account the various tests I’ve
now completed—steering angle and crosswind stability, flow visualization
through tufts, pressure measurement, and coastdown—I’ve decided not to
install a splitter on the Prius. While a longer permanent splitter could
improve stability and reduce front lift, it appears to do so at the expense of reducing
the effectiveness of the rear diffuser and does not appear to reduce drag. And
it doesn’t come with a secondary benefit such as improving cooling airflow. So,
I won’t build one for my car.
Your car and goals will be different. If you have previously
modified your car by guessing at what air does around it or merely imagining
what effect a particular shape change will have (or worse yet, pasting a “template”
over its side-view profile because some rando on the internet told you to),
STOP. Yes, testing and investigating takes more time—but
it is infinitely more rewarding and interesting because it is real, not
imagined. Remember, you’re trying to get a picture of what happens to the immensely
complex airflow over your car after making a change to its shape. Use your
brain; figure out what sorts of tests will show you what you want to know; make
informed decisions based on real results; leave the speculating and guesswork
behind. Aerodynamics isn’t magic or illusion, it isn’t paint-by-numbers, and it
sure as hell isn’t simple—but it can be illuminated through straightforward and
inexpensive testing. It is a field with many unknowns waiting to be uncovered,
unexpected phenomena waiting to be brought out of the mists of confusion, and
surprises around every corner. Go try it yourself and see what you can
discover.
Comments
Post a Comment