- This is test one. The Catesby Tunnel in
Northamptonshire, England used to be for trains. It was built by the Victorians in 1897, but fell into disrepair after
the line that went through it was closed in the '60s. It was abandoned for many years, but now it's a vehicle testing facility. You know a wind tunnel? This is a more realistic version of that where the air isn't moving,
but the car definitely is. - So Catesby Tunnel is
2.7km long. It's got a cross section of 42m²,
perfectly straight. It's quite an amazing feat considering the Victorians built it
without a laser level. Obviously the surface
of the earth is curved, but Catesby Tunnel is flat. So if you put a laser line along it, we're within 3mm of
being off that laser line. But when we first inquired about
repurposing Catesby Tunnel, it was totally derelict. They'd ripped up the track,
they'd taken away the ballast, they'd destroyed the drainage. So at the far end, the tunnel was actually
knee-deep in water. There were no facilities
or utilities here. There were bats living within the tunnel that we had to tread carefully around. So at the far end of the tunnel,
they've actually got their own "bat hotel". When we test outside, we've got the wind, we've got the rain,
we've got other vehicles, we've got lumps and bumps in the road. We want to remove those and
make this controlled a condition as possible so that the test is only seeing the changes
that we're looking for. - We're doing a "coast-down test" here. Julian got the car up to
130km/h, very quickly(!), and then he took
his foot off the accelerator and just let it coast in neutral. This tunnel's long enough
that there'll be a couple of minutes before he
has to touch the brakes. From that, you can work out
the drag and friction forces, which is really useful data
for car manufacturers. - Because Catesby Tunnel is a
bit different to a normal wind tunnel, we have to collect data on board. We can't use GPS, because we can't see the
satellites in the sky. The approach that we've come
to is an image-based system. - There's 1200 tags now in the tunnels and they are 5m apart,
and they've been surveyed to within a few millimeters,
each position. And then we measure the time it takes to go between each tag, and that's picked up by
this high speed camera that's working a hundred times a second. And that gives us a very high accuracy for speed of about a hundredth
of a kilometer an hour. The tunnel is a fantastic place
for this kind of measurement, because there is no extraneous sunlight and sunlight is very
difficult to cope with when you are working with infrared. - But I said that this was test one, and it was, I wasn't lying. But like any video about cars, you didn't see it from just one angle. There were camera shots
from all around the tunnel to get different angles
and keep it interesting. There were shots from
cameras that were attached to the car with suction cups for dynamic shots from
almost impossible angles. But this is a proper glossy car video. We don't want to see those
suction cup cameras from outside, and beside, we've only got two of them, and we don't want to
see the camera operator in the back either 'cause he's not there. There's got to be a little
bit of unreality to this that no one will notice once
it's all edited together. So when I said this is test one,
that is actually true. This right now is test one, but the shots from outside,
we haven't filmed them yet. We're going to have to go back and forth in this tunnel 15, 20 times before we've got every angle we need to make this look interesting. So I had a thought. Do those cameras actually
affect the results? If you've got a camera on
the outside of your car, how much more fuel are you actually using? How's that for timing? - Good timing. Spot on! - Typically, we go out and back for a test and average over both of those sections. We have a turntable at each end. The turntable takes about
20 seconds to turn around. We don't have access at
the far end of the tunnel, so we have to turn around inside. It saves doing an Austin Powers
37-point turn at the far end. - So exactly the same test the other way. - Yes. - Okay. Go for it. - In order to get the best accuracy, the driver has to be very consistent and he has to do the same thing
time after time. You have to drive in a very straight line. So any deviation, that
adds to the distance and the drag that the car's experiencing. You can start off at
pretty much any speed, as long as it's above the
start of your coast-down speed. You want to let the car settle a bit, because the moment you
put it into neutral, the car will start to rock backwards
and forwards a little bit, and all these little things all add up. - Is that a garage door opener? - It is.
- Just for the turntable? - Yeah.
- That's nice. - Smart. - I feel like this needs the music
from "Thunderbirds." - So the whole idea of Catesby Tunnel is to create constant
atmospheric conditions. We shut the doors at both ends and we seal the tunnel so that there's no wind coming through it. Because we're surrounded by earth, it actually means that the
temperature within the tunnel is within one or two
degrees all year round. We had the Aston Martin
Valkyrie in the tunnel, we've had cyclists in
here during drafting, which is fairly unique, something that you can't
do in other places. One of the biggest issues is what happens if there's
an accident in the tunnel and how do we get to them? So throughout every test,
we have a fire safety team on site. They're there to respond
if there is an accident. - There is a very slight
gradient to this tunnel and while I'm sure some
of the marketing team would be very happy if
it was perfectly flat, turns out there is a
little bit of an advantage to that gradient, because drag increases
with the square of speed, but friction increases linearly. So if you have one test
going very slightly downhill and one test going very slightly uphill, in theory you can
separate those two factors out from each other with a lot of maths that I do not understand. - The Catesby Tunnel surface
is an asphalt surface, typical of an A-road that
you would see outside, but just finished to as high
tolerance as we can get. - I do like how you stick to one side of the tunnel,
like it's a two-way road. - It just seems wrong barreling that fast down the center of a road. - Yeah. - And also, for consistency, for the aerodynamic wake. - Oh of course, it'll
be different depending on where you are in the tunnel. So I've been kept in the
dark for about 20 minutes now while you wrangle data. What do we have? - Well, this is an example of
one of the coast-down tests. The dark blue line is the speed. So that's us accelerating
rapidly up to 130, then pop it into neutral and
then coast down to about 40. - And then brake so you
don't hit the barrels. - And then brake so I
don't hit the barrels. - So we had four or five runs
with a clean car. - Yeah. - We had five, six runs with the GoPros in varying configurations. So we've got some sort of
drag from some sort of camera. - Yes, exactly.
- Alright. - And you can see, from 130
to 120 kilometers an hour, it took 7.04 that time and 7.03 that time. So a hundredth of a second difference.
- That's so close. - You wouldn't get that outside. - Yeah.
- It just wouldn't happen. And then with the GoPros on, from 130 to 120, took 6.98 and 6.97. Again, hundredth of a second difference. - Yep.
- Incredible. But there is a very noticeable difference between having a clean car and when you've got the
GoPros on the outside. It's measuring just a small difference
of about 0.8% between the two. - And does that map to fuel usage? - Yes.
- And this is not the most scientific test. We did a couple runs, the
GoPros were all over the place, but there is enough drag from them that you can measure it in there? - Yes.
- And there would be some effect on fuel. - Yes. Now of course, we were using
the ideal car for this test, because the weight doesn't
change from test to test. - Right, it's electric.
- Whereas, in a fossil fuel car, of course,
you've got to make sure that you don't let that tank go too empty. And there's lots of things
you have to take into account. - Yeah. So there's roughly 0.8%
more drag on the car from sticking a couple
of GoPros somewhere. - Yes, absolutely. - That's the closest I've got
to science in a long time. Thank you very much. - No problem. Thank you.
