Higher Order Panel Method / BEM Current State Overview

Your descriptions largely confirm what I believed (I think you're right).

I'm not 100% on a couple (possibly overly nuanced) points but I'm looking into it more before I ask an overtly stupid question.

-------------------------------------------------------------------------

In your original post you asked:

I've actually got a very old commercial aerospace panel code that does this, but it was written in 1989 in Fortran 77 and therefore most likely does not represent the "current state." Was there a particular application you wanted to explore?

The reason I have it is that it was given to me by my boss when I was an intern to see if it might be applicable to the analysis of planing hulls for a computer program called POWERSEA my boss wrote (I added some code to is as an intern).

 

What's the impact on accuracy?

 

None, or improved accuracy.
The P,Q functions in Michell's integral (although calculated numerically) are
exact for piece-wise parabolic sections and waterlines. That means a Wigley
hull requires to be specified at only 3 stations and 3 waterlines. Methods that
don't employ some of the quadrature tricks require 100 stations and waterlines
for good accuracy. Some brute force methods (and even "industrial strength"
methods in QUADPACK) won't converge at all.

 

Thanks. Now you're making me feel very old. I haven't done much with panel codes since around 1989, and Fortran 77 was the primary language I used. I'll send a Private Message.

 

Would it be useful to have a Rhino plugin that prepares Michlet input files?

That wasn't my intent!

 

I have never seen or used Rhino, so I can't comment.
(I know Delftship produces Michlet inout files.)

To be honest, I don't do much with Michlet any more. I am more involved with
programs that use both the local-field and "wavy" components of Havelock
sources and vortices: Michlet uses only the "wavy" single integral and it does a
reasonable job with that, so I regard it as a completed project.

 

BEM - Unsteady BEM

Thanks everbody for helping me out. Could I ask another question. I am also trying to convert my code from steady to unsteady. I have read Katz Low speed aerodynamics and it is a good reference. It suggest that the only modification I require is to impose time dependent boundary condition on the geometry of wing and in the end by using unsteady Bernoulli equation to calculate Cp. I have the boundary condition with me but my it is interms of angular velocity. Now if I am trying that my wing section is translating forward and also pitching up and down then I shud have a B.C. like this:

(∂∅)/∂n= (V_0+Ω×r)∙n


where V_0 is the translational vel of the wing section and Ω×r is the contribution due to pitch/rotation.

if I am using combined source/ doublet method, I will get

σ= -(V_0+Ω×r)∙n

where σ stands for source strength

and for pure sinsuoidal pitch doing up and down with translational forward, I will get

Ω×r= θ_dot*zi ̂- θ_dot*xk ̂

where θ_dot is angular velocity, x and z are the x and z distances in i and k directions.

My problem is this θ_dot. How can I introduce this in my code. Or is there is anyother way to simulate the translation and pitching of the wing section in my Boundary condition.