1
00:00:10,493 --> 00:00:16,822
Hello, in this four-minute
video we are going to see how

2
00:00:16,822 --> 00:00:23,828
Heliciel simulates in CFD a wind
turbine and a tidal turbine,

3
00:00:23,828 --> 00:00:31,513
and how to read what the calculation
confirms and what it does not.

4
00:00:32,743 --> 00:00:41,298
Here is a wind turbine 4 metres
in diameter modelled in Heliciel;

5
00:00:41,298 --> 00:00:51,171
it has 3 blades and turns at 166.8 rpm
in a wind of 5.5 metres per second.

6
00:00:51,221 --> 00:00:57,287
By the blade element method, the
designer calculates 231 newtons

7
00:00:57,287 --> 00:01:02,785
of thrust on the blades and
35.9 newton-metres of torque.

8
00:01:04,962 --> 00:01:08,361
In the CFD bench, the
flow lines slow down as

9
00:01:08,361 --> 00:01:11,911
they cross the rotor and
spread out behind it:

10
00:01:11,911 --> 00:01:16,896
it is the signature of a machine
that takes energy from the wind.

11
00:01:16,946 --> 00:01:21,566
Facing the wind, the air presses
on the blade; on the back,

12
00:01:21,566 --> 00:01:27,360
it is sucked, above all towards the
blade tips, where the effort gathers.

13
00:01:27,410 --> 00:01:34,764
The vertical cut shows the wake: a wide
zone of slowed wind, wider than the rotor,

14
00:01:34,764 --> 00:01:37,033
stretching far behind it.

15
00:01:37,033 --> 00:01:42,478
A second wind turbine placed
there would receive less wind.

16
00:01:42,528 --> 00:01:48,812
Across the axis, the disc of slowed
wind is already wider than the rotor.

17
00:01:50,909 --> 00:01:57,437
The bench finds 221 newtons of thrust on
the blades, 4 percent from the designer:

18
00:01:57,437 --> 00:02:03,481
it is the effort that loads the mast
and the foundations, and it is solid.

19
00:02:03,531 --> 00:02:07,372
The torque, however, comes out much lower,

20
00:02:07,372 --> 00:02:12,677
and the bench says so in its
alerts: torque not reliable.

21
00:02:12,727 --> 00:02:15,062
On a lightly loaded wind turbine,

22
00:02:15,062 --> 00:02:19,307
the torque is a small difference
between two large efforts;

23
00:02:19,307 --> 00:02:25,251
it needs a very fine mesh to describe the
boundary layer on the back of the blades.

24
00:02:25,301 --> 00:02:29,647
We therefore keep the designer's
power for production,

25
00:02:29,647 --> 00:02:32,947
and the CFD for the thrust and the wake.

26
00:02:35,030 --> 00:02:38,841
Let's go under water. Here is a tidal

27
00:02:38,841 --> 00:02:43,476
turbine 16 metres in
diameter with 2 blades,

28
00:02:43,476 --> 00:02:48,112
in a tidal current of
2.4 metres per second.

29
00:02:48,162 --> 00:02:51,916
Sea water weighs nearly a
thousand times more than air;

30
00:02:51,916 --> 00:02:56,830
the mesh is nevertheless built with
the same buttons, in a few minutes.

31
00:02:56,880 --> 00:03:00,905
Facing the current, the water
presses on the whole blade,

32
00:03:00,905 --> 00:03:03,660
with a maximum along the leading edge.

33
00:03:03,710 --> 00:03:08,563
Behind the rotor, the slowed water
stretches over several diameters:

34
00:03:08,563 --> 00:03:12,822
this is the question of spacing
between machines in a farm,

35
00:03:12,822 --> 00:03:14,844
which CFD allows us to ask.

36
00:03:16,940 --> 00:03:19,931
With the rotor really turning,

37
00:03:19,931 --> 00:03:26,411
the bench calculates 454,000 newtons
of thrust on the structure,

38
00:03:26,411 --> 00:03:30,239
hub included, that is about 46 tonnes:

39
00:03:30,239 --> 00:03:36,182
the figure that sizes the mast,
the base or the anchoring.

40
00:03:36,232 --> 00:03:40,004
The bench's alerts confirm
that this thrust is reliable,

41
00:03:40,004 --> 00:03:42,834
a robust integral over the whole surface;

42
00:03:42,834 --> 00:03:44,873
the torque, on the other hand,

43
00:03:44,873 --> 00:03:50,244
would require resolving the circulation
around the blade profiles more finely.

44
00:03:52,771 --> 00:03:58,973
With Heliciel, the CFD of an energy-
capturing machine is read in two steps:

45
00:03:58,973 --> 00:04:01,949
the alerts first, the figures next.

46
00:04:01,999 --> 00:04:07,355
We could now study the spacing of
two rotors in the same current.