1
00:00:10,955 --> 00:00:16,750
Hello, in this four-minute
video we are going to see how

2
00:00:16,750 --> 00:00:23,062
Heliciel simulates in CFD a
5-inch drone propeller in hover,

3
00:00:23,062 --> 00:00:30,409
from the model to the result, in less
than six minutes of calculation.

4
00:00:31,540 --> 00:00:38,364
Here is the propeller of the drone
model supplied with Heliciel:

5
00:00:38,364 --> 00:00:45,936
127 millimetres in diameter, 3 blades,
turning at 25,000 rpm in hover.

6
00:00:45,986 --> 00:00:49,616
Its blade tips move at
half the speed of sound.

7
00:00:49,936 --> 00:00:58,238
The designer calculates 7.02 newtons of
thrust, that is about 720 grams per motor.

8
00:01:00,353 --> 00:01:03,935
Let's open the PRO CFD
menu of the designer.

9
00:01:09,236 --> 00:01:13,796
It sends the propeller to the test
bench exactly as it is drawn,

10
00:01:13,796 --> 00:01:17,714
with the air and the speed
of rotation of the project.

11
00:01:19,790 --> 00:01:23,475
In the bench, we choose
the Level 3 mesh preset,

12
00:01:23,475 --> 00:01:28,772
and tick the two boxes that chain
the calculation after the meshing.

13
00:01:28,822 --> 00:01:35,938
On a desktop PC, the meshing takes one
minute and forty seconds; the first phase,

14
00:01:35,938 --> 00:01:41,959
with the propeller frozen and the
rotation simulated, forty seconds;

15
00:01:41,959 --> 00:01:47,360
the second, where it really
turns, three and a half minutes.

16
00:01:47,410 --> 00:01:51,726
A small propeller is checked
between two bench tests.

17
00:01:53,797 --> 00:01:58,924
Here is the calculated case. On top,
the side that sucks in the air,

18
00:01:58,924 --> 00:02:02,276
the backs of the blades
are in low pressure,

19
00:02:02,276 --> 00:02:06,161
blue towards the tips,
where the speed is highest.

20
00:02:06,211 --> 00:02:12,305
Underneath, the side that pushes the
air, the whole face is in overpressure.

21
00:02:12,355 --> 00:02:18,073
The bench finds 4.88 newtons,
30 percent less than the designer,

22
00:02:18,073 --> 00:02:23,791
for 9 percent more torque: rather
500 grams per motor than 720.

23
00:02:23,841 --> 00:02:28,040
This gap is to be looked
for first on the mesh side,

24
00:02:28,040 --> 00:02:32,320
and the bench adds a warning:
on such a small blade,

25
00:02:32,320 --> 00:02:36,885
a laminar portion is likely
on the front of the profile,

26
00:02:36,885 --> 00:02:40,717
which the turbulence
model does not represent.

27
00:02:42,871 --> 00:02:47,004
The vertical cut shows
the jet: on the right,

28
00:02:47,004 --> 00:02:52,699
the still air is sucked in and
accelerates towards the rotor;

29
00:02:52,699 --> 00:02:59,955
on the left, the jet leaves fast, with
a slower core in the shadow of the hub.

30
00:03:00,005 --> 00:03:03,467
Across the axis, the disc
of the jet is cut into

31
00:03:03,467 --> 00:03:06,641
three sectors by the
imprint of the blades.

32
00:03:06,691 --> 00:03:11,647
The jet is not uniform: fast under
the blades, slow under the hub.

33
00:03:11,647 --> 00:03:15,820
This is the picture to look
at before placing a sensor,

34
00:03:15,820 --> 00:03:19,007
an antenna or an arm under the propeller.

35
00:03:21,113 --> 00:03:25,661
The second phase makes the
propeller really turn,

36
00:03:25,661 --> 00:03:29,189
and confirms the first: 4.94 newtons.

37
00:03:29,239 --> 00:03:33,507
It also gives its rhythm:
1,250 hertz of blade passing,

38
00:03:33,507 --> 00:03:36,533
the high-pitched whistle of the drone.

39
00:03:38,978 --> 00:03:43,014
With Heliciel, CFD is not
reserved for large machines:

40
00:03:43,014 --> 00:03:46,527
a drone propeller is
checked in a few minutes.

41
00:03:46,577 --> 00:03:53,551
We could now compare two pitches, or two
profiles, before printing a single one.