RE: Parameter Estimation

 

> I'm an aeronautical engineering student, and I need to 
> construct a FDM for an aircraft of my choice (I don't know 
> yet), for an end of career project. I currently know of two 
> sources for aircraft derivatives estimation:
> 
> - Digital Datcom
> - Roskam
> 
> I was wondering if someone could tell me of more methods for 
> estimation, and information about its precision and 
> aplicability would be greatly appreciated too. Thank you in advance.


Well, as the champion for Digital DATCOM, and a practicing Aero Engineer,
allow me to say a few words on the subject.

First off, I hope that you meant the end of your SCHOOL career ;-}

I am going to assume that in the construction of a FDM for your aircraft,
you have to meet a particular goal, such as a particular performance level
(climb rate, speed, distance, etc), and that you are going to be running the
calculations more than once. Also, I'm going to assume that you are going to
be interested in more than just a couple of data points, but will include
takeoff and landing configurations, high alpha up to and including stall, as
well as controllability.

Under these assumptions, I'd go with the DATCOM, of course (and I'm not
being paid to say that, either). Should you manually crank through the
equations in Roskam, perkin and Hague, et. Al., and you find that your
aircraft is not controllable enough, and that you need to length the
fuselage, you'd have to recrank all those calculations. You also might make
a mistake somewhere along the way. So, why reinvent the wheel here?  Use the
tools that you available to you.

DATCOM can be highly accurate. I have proven that to myself against a couple
difference aircraft, comparing DATCOM results to actual aircraft flight test
results. (I'm looking forward to doing more, I just don't have the time).
Should you not be satisfied with the outcome of your model, you can easily
change design parameters and recalculate your model in a matter of seconds
instead of hours. Furthermore, DATCOM+ spits out data tables in the new
JSBSIM format, which is just about to hit the streets.

Concerning X-Plane, well, I haven't done a LOT of research on it, because I
didn't like the way it did things. If I remember correctly, it does a CFD
(Computational Fluid Dynamics) analysis on your aircraft, and apparently
does a decent job with a minimal number of panels, I think it was less than
10. What I didn't like about it was that I was attempting to match my
computational model with an actual aircraft (I build flight simulators for a
living), so the adjustments required in order to meet a data point is not
really clear. Sometimes you actually have to cheat a little and fake a
coefficient table in order to meet a test point. Doing that with a CFD model
is a little harder.

Okay, if you aren't sold yet, let's throw out a few more ideas, hopefully to
triger some ideas for your design. You probably aren't doing the simpliest
aircraft out there, so you are probably going to need flaps, trailing edge
at least, and depending on the aircraft, leading edge flaps. How about
spoilers or speed brakes for a fast little biz jet?  Are your fellow
students doing simple little aircraft without flaps, spoilers, or
speedbrakes? If so, you can make a big impact by doing more. Want even more?
DATCOM+ calculates the left and right flaps separately, due to a pet
interest of mine. I heard a hteory that the Boeing 737 that crashed in
Pittsburgh in ... Umm... 1993(?) was caused by a flap contro horn that
broke, allowing one flap to retract. What would the effects of that be?
Wow, thing about it. That aircraft would be uncontrollable. I'd like to get
some time history data of that crash and test out my theory.

Okay, I'm not being paid by the word here.

Good luck in whatever you choose. Let us know how you are doing, and if any
of us can help. There are some wonderful people in this environment.

Bill Galbraith
billg (at) holycows.net
www.holycows.net/datcom/



_______________________________________________
Flightgear-flightmodel mailing list
[email protected]
http://mail.flightgear.org/mailman/listinfo/flightgear-flightmodel
2f585eeea02e2c79d7b1d8c4963bae2d