|
This article was first published in 2008.
|
Anyone can whack on a front spoiler - either as part of a body kit or a more
subtle design positioned under the car near the front axle line. But will it
work? In this story we test whether some trial design front spoilers actually provided
downforce. All the testing was done on the road and some fascinating results
were gathered.
The Car
While there are many aspects of my '98 Lexus LS400 that I like, one
shortcoming of the car is that in crosswinds it lacks directional stability.
This is especially noticeable at high speed on freeways, where its behaviour can
only be described as poor. Many cars that have low drag coefficients suffer from
a lack of directional stability (and the big Lexus is extremely slippery), but
that's no excuse!
So the modification task was to try to improve stability in crosswinds, and
this was to be achieved by increasing front downforce (or reducing front lift,
if that proved to be occurring).
It was never the intention to fit a huge body kit, but instead to perform
some subtle - but hopefully effective - tweaks. Highest on the list was an
active front spoiler that automatically lowered itself at speed. Many of the
country roads around where I live are bumpy and with the relatively soft
suspension of the Lexus, a permanently mounted low front spoiler would hit the
road many times a day. Also, the entrance to my driveway is very steep - the car
can already drag its front bumper lip on the bitumen...
The spoiler could be positioned back from the leading edge of the bumper,
being effectively a flat plate that pivoted down from the undertray.
Well, that was the first idea...
Measuring the Changes
A wind tunnel is normally used to measure changes in lift and downforce.
Inside the tunnel, the car is placed on very sensitive electronic scales that
measure the effective weight of the car at all four corners. If the wind tunnel
is moving lots of air past the car and the front wheels then push down less
heavily, front lift is occurring. If in the same conditions the wheels push down
more heavily than normal, downforce is occurring.
Of course, since there are springs in between the body (which is generating
the aero forces) and the wheels and tyres (which are transmitting these forces
to the road) you can also measure downforce/lift by measuring the ride height.
If the ride height is lower, it's because the body is being pushed downwards. If
the ride height is greater, it's because the body is being lifted by the
airflow. You can even get an idea of the forces involved by seeing how much
weight you need to add (or how much lift you need to apply) to get the body to
move up or down by the same amount.
So how come everyone doesn't measure what's aerodynamically happening to
their cars by measuring ride height at speed? Two reasons. Firstly, you need to
have a sensor mounted on the suspension. That can be as easy as a slide
potentiometer or as complex as an LVDT. Use the slide pot and it's not hard to
rig up a device yourself - although it's still a fiddly job.
Secondly, as each bump is met, the ride height varies all over the place.
Sorting out the average ride height can be quite hard to do when the ride
height's always changing!
However, in this case we were able to get excellent, repeatable
results.
The Lexus has standard suspension height sensors front and rear - these are
used as inputs into the automatic headlight levelling system. By unplugging the
headlight levelling ECU and measuring the resistance across a pair of wires
coming from the front sensor, the relative front ride heights could be read off
as varying resistances.
A Fluke 123 Scopemeter was used to log these readings - the Scopemeter has a
very fast sampling rate and importantly in this application, calculates the
average (and also records the highest and lowest readings) for the whole of the
logged period. In addition, it shows on its TrendPlot graph the actual ride
height pattern of the vehicle.
This ride height measuring system worked brilliantly - the 'step' change in
ride height when the driver sat in the car was clearly observable, and the
difference in the recorded trace between freeway and back roads was obvious.
In this screen dump from the Scopemeter you can see the logged result of a
23-minute drive. The high trace movements are indicative of a high ride height
(ie suspension droop) and the low trace movements show a low ride height
(suspension bump). As you can see, early in the drive the road was fairly smooth
- the car was on a freeway. However, the road then got much rougher.
Aerodynamic testing at speed was carried out in the following way. Firstly
the car was accelerated to the required speed and then held at that speed. Once
the speed had stabilised, the logging was activated. The car was then held at
the speed for several kilometres and then prior to slowing at the end of the
run, the logging was halted. The testing was carried out on a smooth, straight
and nearly flat freeway.
As Standard
Testing with the car is standard trim gave these results:
Speed |
70 km/h |
100 km/h |
130 km/h |
140 km/h |
Average Front Ride Height
(expressed in ohms of sensor resistance) |
1839 |
1839 |
1838 |
1832 |
(The lower the resistance reading, the lower the ride height.)
The trend is downwards as speed rises - since (very roughly) about 4 ohms = 1
mm of ride height, you can see that the average ride height has dropped by 1-2mm
as the speed has gone from 70 - 140 km/h. (That might seem a trivial amount but
it is indicative of some downforce occurring at the front of the standard
car at speed. Most cars develop lift.)
The Spoiler
The next step was to make a simple drop-down front spoiler that could be
manually placed in the 'active' position. This comprised a flat piece of
aluminium sheet 150 x 80mm which was hinged along the rear edge. By means of
adjusting a tethering cable, its angle (and so how low it extended) could be
changed. It was placed a little forward of the front axle line. (Incidentally,
it wasn't always this battered - but in testing I had it as low as possible,
with inevitable results!)
However, on-road testing soon showed that when it was deployed downwards, the
aerodynamics headed in the wrong direction!
| |
Standard |
Trial Front Spoiler |
Speed |
70 km/h |
100 km/h |
130 km/h |
140 km/h |
100 km/h |
140 km/h |
Average Front Ride Height
(expressed in ohms of sensor resistance) |
1839 |
1839 |
1838 |
1832 |
1870 |
1883 |
This graph makes the difference clearer.
It's obvious that with the spoiler
in place, the front ride height is greater. In fact, at 100 km/h the front is
being lifted by up to 7mm. Again that doesn't sound much - until you try to lift
the car by the same distance and see how much effort it takes!
It's likely that with the spoiler positioned back from the front of the car,
aerodynamic pressure was being brought to bear on the forward part of the front
undertray, creating lift. In this case the reduction of airflow under the car
was more than offset by the build-up of pressure under the car, forward of the
spoiler. So this approach certainly didn't work.
The Splitter
I then tried a splitter - a
flat piece of material projecting forwards from the lowest point of the frontal
bodywork. This has the potential to capture air trying to flow under the car
from the stagnation point (the point at which flow separates to flow over and
under the body) and so create some downforce.
Rather than use aluminium, the trial splitter was made from thin particle
board. The next decision was whether to have it projecting across the full width
of the lower bumper/spoiler, or just across the middle part. The problem was
that the factory rubber edging located at each front corner was already a bit
damaged (by being scraped on the ground, mostly coming into my driveway) and so
if the splitter worked, these bits would need to be renewed or revised. But
removing these rubber extension pieces and then placing the splitter right
across the car would involve changing two factors simultaneously - so the
'centre-only' splitter was trialled first. It seemed likely that if a small
splitter worked, a larger one would work better.
Testing was done a little differently - because it was a new day and the
baseline ride height may have changed a little, multiple tests were done at
three speeds, each with the splitter in place. Even if it worked really well, at
low speeds the splitter wouldn't be developing any downforce, so the 70 km/h figure
was used as the baseline.
Speed |
70 km/h |
100 km/h |
140 km/h |
Average Front Ride Height
(expressed in ohms of sensor resistance) |
1856 |
1852 |
1854 |
As can be seen, the ride height stayed almost the same, irrespective of
speed. The splitter wasn't developing the lift that was seen previously with the
spoiler, but by the same token it wasn't developing any measurable downforce,
either... (Incidentally, the pattern of a lower ride height at 100 km/h when
compared with both 70 and 140 km/h was repeated on many runs - perhaps the front
airflow was varying in its behaviour?)
Conclusion
This is a real world story - and the result isn't positive. Without reducing
front ground clearance, I now can't see any way of improving front downforce.
Perhaps a very fancy active front spoiler that folds itself up flat against the
undertray and then concertinas down at the leading edge might do it - but how
would you make one of those?
Measuring average ride height is an excellent way of assessing lift and
downforce variations, especially on cars that have relative soft springing. But
the trouble with good testing of this sort is that it can show when mods
aren't working - which is always a bit of a downer!
|