Re: [Flightgear-devel] Sky dive free fall

[leee <le...@sp...>]

On Saturday 30 May 2009, Curtis Olson wrote:
> On Sat, May 30, 2009 at 12:02 PM, Vivian Meazza  wrote:
> >  I don’t see much of a problem with a suitable jumper. Making
> > it a ballistic object with drag and mass would be easy. But a
> > more realistic FDM … Hmmm
>
> An accurate FDM would be immensely complex considering all the
> possible poses a human can achieve.  But perhaps something
> simplistic could be worked up using the arms and legs as control
> surfaces.  My focus right now is not so much getting accurate
> free fall dynamics, but to get a nice jumper model and then just
> hack up some sort of dynamics with approximately the right
> lift/drag ratio for someone is a stable free fall pose.
>
> The goal would be to get approximately the right fall rates and
> timings so that there is training value in solving problems and
> overcoming various combinations of faults with in a realistic
> time frame.  It's still only a very partial simulation but
> hopefully a step better than just sitting around in a circle
> talking through various scenarios.
>
> The next step would be to have a canopy that could be configured
> to have various problems opening up and be able to draw that
> somehow from the perspective of the sky diver, and perhaps have
> some appropriate dyanmics for partially tangled or partially
> inflated chutes?  Obviously there's endless variabiltiy and high
> fidelity in all respects would be crazy to try to achieve, but it
> would be interesting to take a few small steps forward and see
> how far we can get.
>
> Best regards,
>
> Curt.

Is lift actually much of a factor in free-fall skydiving?  I thought 
it was mostly just varying drag, in which case the FDM might be 
simpler than you think.  For example, the YASim FDM assigns drag to 
extended gear elements, which are located at specific points on the 
aircraft and so act at those points.  It might be possible then to 
hack the YASim FDM about a bit to remove the need to solve for 
cruise and approach conditions and just use the gear drag bits, 
which could then be placed and animated to simulate a human body.

LeeE