[overture] Re: smoothed polygon

  • From: Bill Henshaw <henshaw@xxxxxxxx>
  • To: "overture@xxxxxxxxxxxxx" <overture@xxxxxxxxxxxxx>
  • Date: Thu, 22 Dec 2011 10:58:29 -0800

Hi Alsessandro,
The smoothed polygon will not in general be exactly symmetric. What I suggested would make it more symmetric. You are correct that if the grid is not exactly
symmetric (and the interpolation points too) then the lift may not be zero.

...Bill


Alessandro Orchini wrote:
Good to know that the smoothed polygon does not pass through the vertices, but that doesn't fully answer my question.

What I need to know is if the smoothed polygon mapping is symmetric or not. I'm asking this because I run simulations on a thin-flat-plate (created with the smoothed polygon mapping) in a uniform wind, and i found that even at angle of attach = 0, the lift force and moment are not zero, and the flat plate reaches equilibrium at an angle different from 0.

Also looking closer to the mapping i see that the radial lines above and belove the plate seems to be a bit shifted, and i'm guessing that this "broken symmetry" can cause an unbalance in the moment evaluation and a rotation...

Thanks,

Alessandro




------------------------------------------------------------------------
Date: Fri, 16 Dec 2011 12:52:12 -0800
From: henshaw@xxxxxxxx
To: overture@xxxxxxxxxxxxx
Subject: [overture] Re: smoothed polygon

Hi Alessandro,
   The smoothed polygon does not actually pass through the vertices.
I suggest that you start the polygon on the middle of the bottom edge
and maybe add another vertex on the middle of the top edge, even though
the polygon is flat there (or to be even more symmetric add a few vertices along the
bottom and top edges.

...Bill
Alessandro Orchini wrote:

    Thanks Bill, that was really helpful.

    Actually I've another question about the smoothed-polygon mapping.
    I usually create them with a routine similar to the following one:

    ***********************************
      smoothedPolygon
        vertices
       5
    0. -0.025
    1.0 -0.025
    1.0 0.025
    0. 0.025
    0. -0.025
    n-dist
        fixed normal distance
        -1.0
    lines
        250,150
        boundary conditions
        -1, -1, 1, 0
    n-stretch
        2. 10. 0.00001
     exit
    ***********************************


    and I think that the leading edge should be in the origin. Is that
    so, or the smooth function makes the mapping a bit "longer" and it
    is somehow shifted?

    I do appreciate your help,

    Alessandro





Other related posts: