- If some big industrial
company or research lab wants to prototype a new road surface
or road building method, how do they test that for the long term? Roads need to last decades. How do they simulate hundreds
of thousands of vehicles rolling over their new road for years, without actually doing that? The answer is something called an "accelerated pavement testing facility". And there's only a few in the world. This is one of the largest.
It was built in 1984. It's in the west of France. And it's better known
as the fatigue carousel. - The fatigue carousel
consists of a central tower, with hydraulic motors
which run the machine. And then it has four arms, and outer wheels which load the pavements. A ring is 120 metres in length. And the test section is six metre wide. So this machine allows to
simulate, in a few months, 10 or 20 years of
traffic on the real road. For a long period, the machine was used to validate
French pavement materials. And now we are trying to
find new applications. Recently, we had a test
with bio-binders, to replace bitumen, because petroleum will,
in some time, disappear. And now there is also a lot of interest, for charging electric
vehicles by the road. [machine hums] - You can tell how heavy those arms are, by how much those tyres
squash at the bottom. Oh-- [machine noise, increasing in pitch] It's a good noise. - The tests are done
with the outer wheels, at loads corresponding to heavy vehicles. But we can also install on
each arm, single wheels, dual wheels or also
two axles or three axles. When the machine runs, we can also move the wheels laterally,
to reproduce the effect of real traffic. And you can also change the loads. On a single axle, we can go from
4.5 tonnes to about 8.5 tonnes. So sometimes we can overload the pavement,
to accelerate deterioration. - It takes a few minutes
to creep up to speed. And it's kind of a little
bit unnerving standing here. Because, it started off as, "oh, it's very slow.
oh, this is all very safe." And now it's-- now there's a lot of
metal coming towards me! Faster and faster. - This machine can run
at about 100km/h. The usual testing speed
is 70km/h, because we don't want
to go to the maximum, but we can go to 100km/h. - I thought this was up to speed. And I think it's still getting faster. I've got dust in my eye from the wind. Ah! - We instrument, usually, the soil
and also the pavement layers. So in the soil we put, generally, our displacement transducers,
our vertical strain transducers, to measure the vertical displacement. And then in the bituminous layers, the most common instrumentation
is strain gauges. We use, also temperature sensors, because temperature
variation is very important, for the pavement response. And we use, also,
accelerometers or geophones, to measure the vertical acceleration, and then to get the vertical displacement. - Testing on this carousel doesn't mean just for a
few hours or a few days. A test here can last months,
the carousel constantly rotating, with the wheels travelling at anything
up to 100km/h, with a thousand-horsepower
engine powering them, so that years' worth of pavement
stress and road damage can be built up in weeks. There are actually three different
circular test tracks here, and the carousel can be taken
apart and moved between them. That way, while a test is happening
over here for a few weeks, another track can be torn up and rebuilt
for the next set of experiments. And the tracks can be
divided into sections, to run several different tests at once. One of the tracks is also watertight, for testing how pavements behave
on wet or soaked ground. And I really hope that the
size and scale of this thing is coming across on camera, because I've had to change my microphone
to the windproof one, because of how much turbulent air
this thing generates. Those arms are twice my
height. They are massive. It is genuinely unnerving
to be standing here, not able to see this with
just this noise behind me. - There are some limitations, of course,
because it's an accelerated test. We cannot control the climate conditions. And then there is ageing of
the bituminous materials. They become more brittle.
And obviously we cannot simulate that, because we have to do the
test in several months. We include one reference section
with a material that we know. We can make a comparison
with a classical material, and correct the results
by some coefficients, to adapt to real conditions. - It's tempting to think
that in the 21st century, we can test everything by
computer modelling and simulation. But those simulations have to be
grounded in truth somewhere. You can't build those models
without real-world data. And when you're dealing
with public safety and billion-euro infrastructure projects, sometimes there is no substitute
for just building something real. [carousel whirring]
