Hi,

In order to help our student design teams and for a project for the freshman, I've been working on a parameterized part file, that can be modified from a table to allow for a large variety of aircraft with a small amount of user input. The airfoils have been presenting problems. The best solution I've come up with so far is to parameterize each point, with respect to chord line, thickness, etc as appropriate, then make them into power copies. Once produced, using them is quick and painless. The problem with this is that it's very time consuming to write two formulas for each point on the airfoil, and as a result, my available catalog of airfoils is rather small. There has to be a better way. I've thought of having it source the point coordinates from a design table, but that would require each airfoil data set to have the same number of points, which isn't always possible or adequate. Can anyone in the industry tell me how their company does it? Any additional ideas or suggestions would be very much appreciated. Thank you.

-Trevor

One way to attack this is to use a fog law to define your equation, then use a loop to create the points.

One reason fog laws are cool is that you can define discontinuous functions.

For example, a fog law defining the camber line for a NACA 4 digit series airfoil looks like this:

Formal Real parameters are x and y

Real parameters in the tree are P_Law and M_Law

the body of the law is:

if ( x <= P_Law )
{
   
     y =   (M_Law /(P_Law **2)) * ( 2 * P_Law   * x - x**2)
}
else
{
   
     y = (M_Law / ((1 - P_Law ) ** 2)) * (( 1 -(2 *P_Law ) ) + (2 * P_Law  * x) - x**2)
}

This describe the curve you want. To create the points you can use a KWA loop.

One input to the loop is your fog law, here locally named MCL.

A second input is LawStep which is the step in x along the chord line.

the body of the loop is:

Point_$i$ isa GSMPoint
{
    PointType = 0;
    TypeObject isa GSMPointCoord
    {
        X = $i$ * LawStep * 1in;
        Y = MCL.Evaluate($i$ * LawStep) * 1in;
        Z = 0in;
    }
}

Which generates all the points with all the formulas in one shot.

Finally a spline can be generated through all the points by starting with a Parameter Curve (f(x) tool -> Add new parameter of type curve) and assigning the formula:

spline(CamberPoints->Query("GSMPoint","" ) )

I am not sure my fog law is actually correct but the results look right.  I got the formula from here:  www.aerospaceweb.org/question/airfoils/q0041.shtml

I attach my experiment for your reference.

CliffJohnson,

Thanks for your reply. I had tried the law curves before, and couldn't get it to work. I'll give it another shot. Ultimately though, most airfoils do not have defining equations.

knowledgeds,

Correct me if I'm wrong, but in order to use a design table, the points would already have to be defined in a sketch, with parameters relating their x and y coordinates. For example, the nth point would have a (xn*c,yn*c) value, with c being the chord line, and x&y being the distance for a unit airfoil. The x&y values from 0 to n would be parsed from the table, but there would be a fixed number of points available in the sketch. If an airfoil has fewer points, some points in the sketch will have zeros, which will impact the spline. If it has more, it won't be fully defined.

Thanks for your help

-Trevor

knowledgeds,

Thanks, I'll give that a try.  Is it possible to then parametrically modify that pattern? For it to really be useful, it would need to parameterize each point with respect to the chord line, so that once the airfoil is in Catia, it can be modified in context with respect to changes in chord line, thickness, etc as appropriate.