Design Software

smjmitchell wrote...

I agree there is no way that you can accurately simulate the behaviour of
an
airplane without the use of derivatives.

That's not my point. My point is that if you're going to trust
the results of _any_ design and analysis software you better
understand what it can do and how it does it. Particularly if
your design steps outside the bounds of what's 'typical' and thus
might violate some of the assumptions that went into the software.
Kinda what you said early on, I think.

All the "real" simulators work this way.

Quite a few of very good sims use the base parameter in something
like a table lookup, and avoid stability derivatives for anything
but linear analysis.

Dave 'Bode' Hyde

I wrote:

If you're doing a non-derivative
design and want some hope at all of being close...

smjmitchell ...

I agree there is no way that you can accurately
simulate the behaviour of an airplane without
the use of derivatives.

If your statement above was in respose to my
"non-derivative" statement, I wasn't referring
to stability derivatives, I was referring to a design
that's derived from a prior design, i.e. evolutionary.
Accurate simulation without the use of stability derivatives
is easy if you have estimated aero data. Dynamic derivatives
make it easier, but often times damping derivative estimates
are so inaccurate you might as well not use them anyway.

But having said all this, a simulation of a lightplane-envelope-type
airplane is not required before building one. Personally
I'd spend the time trying to get good aero estimates
or hard data (truck testing, R/C, etc) than in tweaking a
sim...and you need the hard data to do real tweaking anyway.

For home-design type stuff I use a CAD package(*) - designCAD right
now, but I've been know to use AutoCAD. Aero analysis is back-of-the
envelope. Structures so far is TLAR, but before I cut metal I
will have a professional FEA done by a non-advocate. I haven't
gotten to the engines yet :-)

The stuff I use at work is a little more detailed and somewhat
more accurate (and far more specialized), but the results are
usually pretty close to the home-done level stuff, which is
where we start anyway.

(*) and I barely scratch the surface of its capability.

Dave 'engineer, professional and amateur' Hyde

But having said all this, a simulation of a lightplane-envelope-type
airplane is not required before building one. Personally
I'd spend the time trying to get good aero estimates
or hard data (truck testing, R/C, etc) than in tweaking a
sim...and you need the hard data to do real tweaking anyway.

I agree ... there is no point in trying to simulate the flight
characteristics of a light airplane via a simulator. You gotta make a
choice. Do you want to be a scientist and analyse the thing to death or do
you want to build and fly it in a reasonable time. Most light aircraft are
evolutionary in design and that together with conventional design practises
(there are huge amounts of data out there in the public domain) eliminates
the need for all this advanced analysis. After all the vast majority of the
certified light planes currently flying were designed on drawing boards,
using simple analysis methods and emphirical data without CAD, panel codes
and FEA. Also all these advanced tools have NOT really resulted in any
significant advances in the state of light plane performance, safety,
styling, cost etc. You simply don't need all this stuff - well OK some of it
is nice to have - but lets be honest you can do the job without it and
really for a amateur designer it would be better to forget it. By the time
you learn the software, do the analysis, puzzle over the results, redesign
etc and then build the thing you will be an old man. Better just to dust off
the old drawing board and get on with the job using simple conservative
calculations and comparative design methods.


Quite a few of very good sims use the base parameter in something
like a table lookup, and avoid stability derivatives for anything
but linear analysis

True but that is doing the same job as a derivative ... it is providing a
relationship between some state of the airplane (i.e. alpha, beta, control
deflections etc) and the forces acting on it (X,Y,Z,L,M,N). A derivative is
linearised where as the look up table approach can include the
nonlinearities.