From: RegDave@aol.com
X-PH: V4.4a@mx4.cso.uiuc.edu
Date: Mon, 13 Jul 1998 22:02:07 EDT
To: m-selig@uiuc.edu
Subject: Re: Airfoil DataBase
X-Mailer: AOL 4.0 for Windows 95 sub 170

Prof. Selig,

...snip...stuff remove - mss980715

I've attached the executable file with this note (I run McAfee and
Norton anti-virus and am not aware of any "bugs" infecting my
machine). It should work on most of the *.dat files in your
database. I use this for making foam core templates up to 9.5 inch
long as either a single or double template.  It should run under Win
3.1 and Win 95 since it's VB 3.0 It's set for 800x600 resolution and
does not auto-scale to anything else (Sorry, I just do this for fun!).

Just a note on some of the strange commands - the R-Series airfoil is
simply a generalization of the NACA 4 digit sections allowing the
camber and thickness to be any value at any location. Gets really
weird results if something outrageous is specified. It's a holdover
from the first version of this code and not very useful.

The airfoil coordinates are fit with a cubic spline at 1000 points
(upper and lower surface) so it goes exactly through whatever data
points are given. Uses a square root term for the first three data
points to get the LE radius right.

It requires that the coordinates are in standard 2-column format but
it can be LE to TE or TE to LE. Eppler format preferred but it can
also handle fixed spacing with the number of entries identified in the
second line of the file.  Actually, it just ignores that when detected
and reads the two columns and sorts things out as if it were in Eppler
format. If the airfoil name isn't in the first line of the file, it
uses the filename for that. Actually, it looks for a non-numerical
value in the first non-blank character of the first line.  If it finds
one, it assumes that's the name. If not it closes the file and then
re-opens to reset the file pointer. Kind of crude but it works. So if
someone uses a strictly number designation for an airfoil, it will
crash.

The skin thickness correction for templates uses the tangent at the
local data point to develop an orthogonal projection of the skin
thickness. So it's not just a vertical shift. This preserves the
proper skin thickness around the LE.

The TE can be thickened and this simply adds a linear thickness ramp
from LE to TE divided equally between the upper and lower surface.

Oh yes, almost forgot. The airfoil analysis that's presented in the
plot simply sorts numerically through the arrays for max camber,
thickness and position(s). It then calculates the moment coefficient
using a numerical integration of the Abbott and Von Doenhoff
equations. Not great but it's simple and fast. Uses Simpson's
rule. Tried Newton's but since it's a cubic fit, Simpson's gives the
best accuracy anyway.

I've not built any error trapping into this so in the compiled version
it just disappears if it hits a read error. Not a problem here at home
but also not great for other users. Maybe someday that will be fixed.

If this is useful and anyone wants the source code, I'd need to send
that in a zip file.

Thanks, 
- Dave Register