 |
New 90 head design.
Currently I am in the process of designing a
complete new head. Amongst others the design goals are:
 | Better linearity in the head
when putting in control inputs. Most heads on the marked are very
non-linear and even asymmetric. |
| Virtually no interaction
between roll and nick. |
| Virtually no interaction
between pitch and roll and pitch and nick. |
| Virtually no interaction
between forward speed and roll trim. |
| More stable head. More
stable in the hover, and more stable in fast forward flight. |
| More responsive head. More
agile when roll or nick is provided to the head. |
| Programmable head such that
people can choose their own setup with reference to e.g. Delta-3
offset, or Bell-Hiller mixing ratio, or Flybar Feedback, or Flybar
Input or Cyclic Input or etc. This allows end-users that know what
they are doing to easily tradeoff between certain aspects (e.g.
stability v.s. load on servo's v.s. agility of the head v.s. etc. |
| Lighter design. Some
designs are way over dimensioned and too heavy. |
In order to do so I first designed a very conventional head as a
starting point. At this point in time I have already fully designed a
rather conventional head with very linear and symmetric
response. This head is already better than most commercially available
heads.
The reason for designing this conventional head first is to be able
to only replace a small set of parts on the head first for testing. By
doing so I can verify my mathematical models and simulations on the
dynamics of the head and the interaction of design options. This helps
to speedup the design (less prototypes required) for the final head
and builds up confidence in the mathematical models.
To give you an idea of this process I have provided below some
pictures of the head modeled in 3D, some animations of the control
input simulations as well as some stress analysis.
Note that the head is a parameterized 3D model. This means that I
can very easily change any parameter. I can use this to fit it to any
helicopter I like with very little effort. I use the Raptor 90 as a
test bed at this point in time.
|
| The Conventional head Here you have an exploded view of
the conventional head (+- 130 parts modeled in 3D with actual
measurements).
Some simple aspects that play a role in the design of
a head to give you an idea on the design space (there are many more
not listed here :-)):
|
Distance between blade bolts -> introduces more or
less blade lead / lag and thus changes blade stability and servo
load. |
|
Bell / Hiller mixing ratio -> changes the feedback
loop of the flybar, changes stability, roll rate and the amount of
energy that goes in the flybar. |
|
Flybar control input gain -> can buy you extra
stability at the cost of additional power loss in the flybar system. |
|
Delta-3 offset -> changes stability and agility.
Changes stability in fast forward flight. |
|
Airfoil of paddles -> changes aggressiveness of the
flybar (stability). |
|
Weight of paddles - changes stability in the flybar,
changes the stability of the helicopter and the agility. |
|
Position of paddles on the flybar (chord wise) ->
changes stability of the flybar, load on servos. |
|
Position of the paddles on the flybar (length wise)
-> changes stability and agility at the cost of power loss in the
flybar. |
|
Stiffness of the flybar -> changes the
aggressiveness of the flybar. |
|
Length of the main blades -> changes tip speed of
the blade, agility of the helicopter and the energy loss in the main
blades. |
|
Chord of the main blades -> changes the agility of
the main blades and the energy loss in the main blades. |
|
Airfoil of the main blades -> changes the lift and
the stability and the efficiency of the blades at different angles
of attack. |
|
Weight of the main blades -> changes the agility,
changes the stability, changes the autorotation and changes the
loading of the engine. |
|
Center of gravity of the main blades (length wise)
-> changes the stability, agility and autorotation as well as the
engine loading. |
|
Center of gravity of the main blades (chord wise) ->
changes the lead / lag and thus blade stability and servo load. |
|
Torsion stiffness of the main blades -> changes the
blade stability and the agility of the helicopter. |
|
Bending stiffness of the main blades - changes the
stability and agility of the helicopter. |
|
Blade spindle dampers or flap dampers -> changes
agility and stress on the head and helicopter. |
These are just some of the aspects that play a role. I
listed them as said just to give you an idea of the complexity of
designing a new head. There are many interacting design decisions to
be made.
The final head design is likely to have a flybar below
the main blades. This simplifies the control layout considerably.
Here is the current conventional head:
Here you have it in 360 degrees view (takes a while to
load):
|
Here you can see the pitch movement (takes a while to
load):
If you want to see
some more simulations, then please click here (Having them all on one
page would make the loading too slow).
Here you can see some stress analysis pictures:
This first picture shows the absolute cumulative bending
upwards due to roll input. The spindle is locked on the utmost left-hand
side and the force applies on the bolt hole in this simulation.
Cumulative bending as a function of Roll input (mm).
In this simulation you can see the absolute cumulative
stretch due to the centripetal force. Here you can see that the stretch
starts from where the bladeholder becomes thinner (light blue part). Due
to the "stretch upon stretch" you can see that just after the bolt hole
there is the maximum displacement due to the strain.
Bladeholder stretch due to centripetal force of the
main blade pulling (mm).
Here you can see the stress in the bladeholder. You can
clearly see that the stress is the highest near the bolt hole and where
the bearing sits to take the thrust load.
Stress in the bladeholder (mN/mm2).
Here you can see the strain in the bladholder due to the
stress in the bladholder. The strain is the relative stretch due to the
stress.
Strain in the bladeholder.
Here you can see again the strain in the bladeholder.
Strain in the bladeholder. Note that strain where the
blade bolt pulls, and where the bearing rests inside the bladeholder.
Non manufactured parts
Some pictures of parts that I just modeled for fun, I
will never manufacture these :-).
The paddle. I will not manufacture these since
SAB already manufactures whatever paddles are required.
The links I used for modeling. I will not
manufacture these since it is too expensive to create a mold for them.
They do have an upgrade aspect to them though. I explicitly modeled them
such that you can use them both sides. This design would require a very
expensive mold though to manufacture them.
