As your airfoil is convex the hull() certainly works. Try this:

// Draw a NACA airfoil

function yt (x,c,t) =
t/0.2c(0.2969sqrt(x/c)-0.126(x/c)-0.3516pow(x/c,2)+0.2843pow(x/c,3)-0.1015*pow(x/c,4));

module airfoil (c,t,m){
p  = concat([ for(i=[m-1:-1:0])
[ cii/(m-1)/(m-1),  yt(cii/(m-1)/(m-1), c, t) ] ],
[ for(i=[0:m-1])
[ cii/(m-1)/(m-1), -yt(cii/(m-1)/(m-1), c, t) ] ]);

rotate([90,0,0]) linear_extrude(0.01)
polygon( p );
}

module blend_airfoils(c1,t1,c2,t2,y,n){
hull (){
airfoil(c1,t1,n);
translate([0,y,0]) airfoil(c2,t2,n);
}
}

blend_airfoils(70,0.1,50,0.2,200,20);

--
View this message in context: http://forum.openscad.org/Trying-to-incorporate-equation-based-shapes-tp3116p16616.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

Or even better:

module blend_airfoils(y){
hull (){
children(0);
translate([0,y,0]) children(1);
}
}

blend_airfoils(200) {
airfoil(70,0.20,30);
airfoil(45,0.1,30);
}

--
View this message in context: http://forum.openscad.org/Trying-to-incorporate-equation-based-shapes-tp3116p16620.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

Using @Parkinbot's approach. Something like this:

use <naca4.scad>
use <Naca_sweep.scad>
num_points = 100;
airfoil1 = vec3D(airfoil_data(0015, 110, num_points), 0); // NACA, length,
numsamples
airfoil2 = vec3D(airfoil_data(1733, 110, num_points), 20);
airfoil3 = vec3D(airfoil_data(2418, 110, num_points), 40);
// airfoil4 = T_(0,0,80, vec3D(airfoil_data(0015, 130, num_points)));
data = [airfoil1, airfoil2, airfoil3];
sweep(data);
http://forum.openscad.org/file/n16625/2vv5rg5.jpg
The advantage being that airfoil_data() takes a 4-digit NACA number directly
and includes all the non-symmetrical ones.

@Parkinbot does use a different approach which allows you to rotate,scale,
and translate the curves prior to assembly. See airfoil4 for putting the
translation outside the vec_3D().
So making twisted blades like boat propellers is possible in a somewhat
complex description. See his thingiverse for excellent examples.

So to be slightly clearer:

1. airfoil_data() takes the special 4 digit NACA rep which you can read
   about here:

• https://en.wikipedia.org/wiki/NACA_airfoil
• http://www.pdas.com/naca456pdas.html  - would be great to do ALL of these.
  and extracts the fields for chord ratios etc  and then creates the curve as
  pairs of [X,Y] coordinates.

2. The Vec3D() function turns the pairs into triplets (I used optional Z arg
   in first 3 airfoils)

3. Further transforms can be applied (I used Translate which is T_() in
   @parkinbot's approach, in airfoil4)

4. The resulting list is a transformed sequence of coords suitable for
   sweeping or skinning together.
   @parkinbot's Naca_sweep.scad includes a function sweep() to do this but the
   skin() from elsewhere also works.

The missing part is a splined curve through the airfoils.
To do this you need to play with @Parkinbot's spline functions - referenced
above in his post...

I believe you could call this code approach idiosyncratic but it's effective
at describing a complex series of swept curves in a concise way.

IMHO incorporating these ideas into a module that was consistent across
openSCAD (such as the MCAD module which seem slike it has many unmerged pull
requests) and using the same approach everywhere means we could all benefit
from a NACA lib. (and spline interp, and lofting, etc... All old threads it
would be great to actually resolve in the current design as well as
openSCAD2...)
YMMV, as usual

--
View this message in context: http://forum.openscad.org/Trying-to-incorporate-equation-based-shapes-tp3116p16625.html
Sent from the OpenSCAD mailing list archive at Nabble.com.