mick viper - 19 September 2008 05:07 PM

Great info...thanks. Some of things you're saying sound
like the Reynolds number; is that correct or is that
something else entirely.

One more thing I was wondering about. Why hasn't anyone
in F1 tried wings that are forward swept, like the X-29A
plane? Wouldn't that increase downforce without increasing
drag?

The forward-swept wings of the X-29 were made that way for a
different reason than those of the typical Formula One car.

When an airplane goes supersonic, a shock wave builds up around
the wing, starting at the leading edge. If the wing is straight,
the shock wave builds up along the entire wing at the same time.
This mean it loses both lift and control in transsonic flight at
the same time. That makes it shake like crazy and it is
impossible to control. Planes approaching that speed in WWII
crashed because the pilots couldn't pull them out of dives.

It was the Germans, experimenting with the Messerschmidt Me262,
to hit on the idea of swept wings, rather than straight ones.
They discovered that sweeping the wing meant the sonic shock wave
started building at the inner end of the wing (nearest the
fuselage), then slowly developed along the span. That reduced the
shaking and kept the aelerons working through the sound barrier.

The X-29 did something similar, but swept the wings forward
to bring the strongest part of the shock wave close to the
fuselage. It also made the plane dynamically unstable. It is so
touchy that it needs a computer control system to make constant
minor adjustments to the controls to keep it from going out of
control in every direction at the same time, shaking itself to
bits and crashing. Or all of them at once.



The X-29

Formula One cars don't go anywhere hear the speed of sound, so
for the most part, their wings are straight spans, like
a "Hershey Bar." Some teams, like Ferrari, did try "bow-tie"
wings, where the cord in the middle was shorter than that at the
outer ends to reduce drag. In Ferrari's case,(see below)
they pinched the trailing edge.



Tyrrell 20, on the other hand, swept the leading edge. Again,
I suspect, to reduce drag by concentrating most of the airflow
and trailing edge turbulence in one place, rather than all along
the trailing edge of the wing.

Click here and here for the only photos I could find of that Tyrrell rear wing.

mick viper - 14 October 2008 12:11 PM

That's the Tyrell with a "V" shaped rear wing. Interesting. Seems like there would be an advantage to this in terms of greater length at the leading edge of the wing or more surface area. But of course I'm never right!

So what about the Reynolds number as relates to F1? Please expound.

snip....

Reynolds number, defined as the length times the velocity divided by the kinematic viscosity, is a dimensionless number useful in the study of aerodynamics. To take one simplified example, let's suppose you're testing a new F-1 wing in a small (inexpensive) wind tunnel. Actually, it's not only that the tunnel is less expensive, but the small models are cheaper too. If you operate the model at the same Reynolds number as the full scale item, the reactions can be expected to be similar and accurately scalable. So a half scale wing operated at double the normal velocity (speed) should produce accurate values for things like the ratio of lift relative to drag. Note that the actual forces won't be the same, but the ratios will be accurate and therefore should yield good data. It should be noted that wind tunnels exist which operate at higher or lower than atmospheric pressures in order to achieve the desired Reynolds number when testing scale models. I don't know if the F-1 guys test in such tunnels or not, but I wouldn't be surprised if they do.

The Reynolds number is also useful for predicting when and where laminar(smooth)or turbulent (disturbed) airflow will exist on the surface something like a wing. At sea level, you can expect a Reynolds number of around one million per foot at 100 mph. In other words, the Reynolds number one foot from the leading edge of an F-1 wing will be one million at 100 miles an hour, 250,000 6 inches back at 100 mph, 2 million one foot back at 200 mph, and so on.

The transition Reynolds number (where laminar flow changes to turbulent flow) for typical wings can vary from around 200,000 for normal surfaces and perhaps nearing one million for VERY smooth and properly contoured surfaces; however, any laminar flow existing at Reynolds numbers that high can be expected to be unstable and subject to detaching from the surface at the slightest disturbance. That, of course, increases drag.

For a typical F-1 wing, it should pay off for the leading edges and first 12 inches of the wing to be kept very clean and smooth. Aft of that, the surface condition becomes less important since turbulent flow is more stable remains attached much better than laminar flow. This turbulent boundary layer is almost certainly much thicker than any surface roughness on the typical wing more than one foot aft of the leading edge.

If I were designing an F-1 wing, part of my research would be in the area of wing profiles which would exhibit good lift/drag ratio characteristics if and when they enjoy laminar flow on the leading edges, but which would not deteriorate too badly when these same surfaces are covered with sticky rubber bits, stone dings, etc. I'm not privy to F-1 aero work, but I can almost guarantee they look into things like this because relative to aircraft, the typical F-1 wing operates at pretty low Reynolds numbers simply because they're smaller and operate at moderately slow speeds.

Mozella - 15 October 2008 08:43 AM

mick viper - 14 October 2008 12:11 PM

That's the Tyrell with a "V" shaped
rear wing. Interesting. Seems like there would be an
advantage to this in terms of greater length at the leading
edge of the wing or more surface area. But of course I'm
never right!

So what about the Reynolds number as relates to F1? Please expound.

snip....

Reynolds number, defined as the length times the velocity
divided by the kinematic viscosity, is a dimensionless
number useful in the study of aerodynamics. To take one
simplified example, let's suppose you're testing a new F-1
wing in a small (inexpensive) wind tunnel. Actually, it's
not only that the tunnel is less expensive, but the small
models are cheaper too. If you operate the model at the
same Reynolds number as the full scale item, the reactions
can be expected to be similar and accurately scalable.
So a half scale wing operated at double the normal velocity
(speed) should produce accurate values for things like the
ratio of lift relative to drag. Note that the actual forces
won't be the same, but the ratios will be accurate and
therefore should yield good data. It should be noted that
ind tunnels exist which operate at higher or lower than
atmospheric pressures in order to achieve the desired
Reynolds number when testing scale models. I don't know
if the F-1 guys test in such tunnels or not, but I wouldn't
be surprised if they do.

The Reynolds number is also useful for predicting when and
where laminar(smooth)or turbulent (disturbed) airflow will
exist on the surface something like a wing. At sea level,
you can expect a Reynolds number of around one million per
foot at 100 mph. In other words, the Reynolds number one
foot from the leading edge of an F-1 wing will be one million
at 100 miles an hour, 250,000 6 inches back at 100 mph,
2 million one foot back at 200 mph, and so on.

The transition Reynolds number (where laminar flow changes
to turbulent flow) for typical wings can vary from around
200,000 for normal surfaces and perhaps nearing one million
for VERY smooth and properly contoured surfaces; however, any
laminar flow existing at Reynolds numbers that high can be
expected to be unstable and subject to detaching from the
surface at the slightest disturbance. That, of course,
increases drag.

For a typical F-1 wing, it should pay off for the leading
edges and first 12 inches of the wing to be kept very
clean and smooth. Aft of that, the surface condition
becomes less important since turbulent flow is more
table remains attached much better than laminar flow.
This turbulent boundary layer is almost certainly much
thicker than any surface roughness on the typical wing
more than one foot aft of the leading edge.

If I were designing an F-1 wing, part of my research would
be in the area of wing profiles which would exhibit good
lift/drag ratio characteristics if and when they enjoy
laminar flow on the leading edges, but which would not
deteriorate too badly when these same surfaces are covered
with sticky rubber bits, stone dings, etc. I'm not privy
to F-1 aero work, but I can almost guarantee they look
into things like this because relative to aircraft, the
typical F-1 wing operates at pretty low Reynolds numbers
simply because they're smaller and operate at moderately
slow speeds.

An interesting post.

Another way of using like Reynolds Numbers to test a wing
shape would be to test in something other than air. I've
heard of small-scale aircraft shapes being tested in high
speed hydraulic chambers instead of wind tunnels. The
Reynolds numbers were the same, even though the models were
radically different sizes.

Such is the joy of testing a shape according to a number
that is independant of the actual size of the object.
But eventually you still have to test the full sized
wing/car/aircraft/boat/whatever. That's when you see
how accurate your mathematics and scale model tests
really are.