One has to realize that there are
huge compromises in trying to mimic the forces during flight with a 3DoF
platform. For example, simulating fore-aft or sideways sustained forces can
only be achieved by tilting the platform to let gravity act on the occupant.
However the tilting action to achieve the desired angle can be unwanted in
itself, thus generating false force cues. To minimize the false cue, the
tilting action has to be slowed down, but this then results in time delays
of the desired force cue. Thus a compromise must be made between delays and
false force cues. It is also important to realize that the occupant position
with respect to the platform rotation point will affect the magnitude
in which the false force cues are felt. Generally, the closer the pilot's CG
is to the platform rotation point, the better.
Driving the heave motion would seem straightforward, but in fact is quite
complicated. During flight, vertical acceleration is very dynamic. The
limited platform heave range will often result in saturated force cues. This
needs to be overcome by applying some sort of dynamic compression and
high-low pass filtering. It should also be noted that MSFS does not output
very accurate vertical accelerations. X-plane performs much better in this
respect.
The type of airplane, weather settings, your way of flying, they all come
into play when tuning the platform. Most likely you will need different BFF
config settings for different planes and conditions. Below you
can download my current config files that I use now: Note that these both
use the independent rig setting. I use these files in combination with
the GWbasic program as shown on this page.
For small GA airplanes: Small-GA.bff
For bigger commercial A/C: Boeing737.bff
The fine-tuning of the motion cues is very tricky and time
consuming. The video's show the basic platform motion versus flight
conditions, but making your brain "believe" is another matter entirely.
Motion can be convincing, but when it's wrong, it actually makes the feeling
of flight less convincing. More is definitely not better. In trying to make
your brain "believe", it is absolutely necessary to use a fully enclosed
platform, without any external clues. Wide scenery screens will help a lot
in fooling your brain. If you feel nauseous, you are doing something wrong,
unless you're flying through a thunderstorm.
During the extensive testing phase, I have come to realize that the best way to achieve the correct motion cues is by checking them out separately. To simplify things, I have almost always used the Cessna 172 in light turbulence weather conditions, doing traffic patterns, and a Boeing 737 for testing the bigger commercial aircraft.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~Under construction~~
Slow heave motion is not easily
felt. The heave motions need to be fast. (<0.5sec or so) therefore
heave range need to be a much as possible, to avoid maxing out. The
sensitivity of Ian's software needs to be tweaked a lot, and will vary for
different planes.
The platform response is quite fast, therefore the low pass filter in the
heave can be set at lowest value possible that will not produce
oscillations.
Examining heave flight data also shows that FS2004 has some limitations
with respect to vertical accelerations on ground. Touch-down heave does not
seem to produce realistic results; It is far too much. We believe that FS
treats the whole airplane as a solid structure, without any gear suspension.
This would explain the steep high peaks during touch-down and riding over
grass land. This can be compensated by editing the .BFF file settings:
;When aircraft is on ground
VaScaleg=0.5
;When aircraft is in air
VaScalea=1.0
This will reduce the on-ground heave excursions by half.
~to be continued~