A bunch of little spheres filling the Sphere

Altair Forum User
Altair Forum User
Altair Employee
edited October 2020 in Community Q&A

Hi, 

 

I have a sphere with 3 mm Radius, and I would like to fill around its surface a bunch of little spheres (close together but not touching each other)  which could coat the whole surface of 3 mm sphere uniformly.  lets say each little sphere has a radius of 0.5e-4 as attached in .cfx file. Or more clearly, filling out the volume of 3 mm Radius sphere with little spheres uniformly but not touching with each other (basically forming back the volume and shape of 3 mm Radius sphere  by little spheres). Could someone guide me what would be the easiest way to do that?   

 

Thanks 

Unable to find an attachment - read this blog

Tagged:

Answers

  • JIF
    JIF
    Altair Employee
    edited January 2018

    Hi Alzaabi,

     

    What part of the problem are you stuck at? I would think the problem consists of three parts:

    1. Determine the points on the sphere. This is a mathematical (packing) problem and simply searching with google for 'equispaced points on sphere' should give you more than enough to get something going. I have seen articles about this, but I have never implemented any of these.
    2. Determine the number of points for your radius (big sphere and small sphere).
    3. Put it all together in a Lua script in CADFEKO to create the spheres.

    I could also have added another item - hope you don't run out of resources to perform the simulation. It sounds like a problem that could be quite expensive to simulate, depending on the distance between the spheres (elements should be smaller than the gap between them usually), the number of spheres and the number of elements required to accurately represent the spheres.

     

    Please post your solution on the forum when you have completed it. It sounds like something that could look interesting. I'm also interested to know what you are trying to simulate - I'm guessing RCS reduction, but for that I would probably rather use periodic boundary conditions (PBC).

  • Altair Forum User
    Altair Forum User
    Altair Employee
    edited February 2018

    Hi Alzaabi,

     

    What part of the problem are you stuck at? I would think the problem consists of three parts:

    1. Determine the points on the sphere. This is a mathematical (packing) problem and simply searching with google for 'equispaced points on sphere' should give you more than enough to get something going. I have seen articles about this, but I have never implemented any of these.
    2. Determine the number of points for your radius (big sphere and small sphere).
    3. Put it all together in a Lua script in CADFEKO to create the spheres.

    I could also have added another item - hope you don't run out of resources to perform the simulation. It sounds like a problem that could be quite expensive to simulate, depending on the distance between the spheres (elements should be smaller than the gap between them usually), the number of spheres and the number of elements required to accurately represent the spheres.

     

    Please post your solution on the forum when you have completed it. It sounds like something that could look interesting. I'm also interested to know what you are trying to simulate - I'm guessing RCS reduction, but for that I would probably rather use periodic boundary conditions (PBC).

    Hi JIF, 

     

    Thanks for your valuable comment!

     

    Actually I am trying to compare the RCS between the big sphere and the little spheres. They both have different dielectric properties but all the little spheres should have same volume as the big sphere at the end. So I need only to simulate the total little spheres and big sphere separately. 

     

    To reduce the complexity  of the simulation, I think I only need the little spheres to be distributed uniformly on the surface of the big sphere instead of filling them inside the big sphere.

     

    I am searching for something that could fulfill this requirement in FEKO. 

     

    Thanks,