Region definition

[Dimitri Balasoiu]

Hello everyone,

I have a question about sfepy's definition of regions.

I want to define a region Omega_1, wich is a subregion of Omega. So I type :

regions = {
    'Omega' : 'all',
    'Omega_1' : ('vertices of group 1', 'cell', Omega),
}

, where group 1 is a group of vertices.

Question : wich field element P1 will be created over the region Omega_1 ?

More precisely, I have a triangle (A,B,C), with node A in group 1, but the other ones outside of group 1.

Will the finite element wich is worth 1 at A, and 0 at B and C be in my field element basis ?

In wich file are the basis functions for the fe fields defined ?

Dimitri

[Robert Cimrman]

Hi Dimitri,

On 04/24/2016 08:48 PM, Dimitri Balasoiu wrote:
> Hello everyone,
>
> I have a question about sfepy's definition of regions.
>
> I want to define a region Omega_1, wich is a subregion of Omega. So I type :
>
> regions = {
> 'Omega' : 'all',
> 'Omega_1' : ('vertices of group 1', 'cell', Omega),}
>
>
> , where group 1 is a group of vertices.
>
>
> Question : wich field element P1 will be created over the region Omega_1 ?
>
> More precisely, I have a triangle (A,B,C), with node A in group 1, but the
> other ones outside of group 1.
>
> Will the finite element wich is worth 1 at A, and 0 at B and C be in my
> field element basis ?

If the field was defined in Omega_1 (assuming it is the default, volume, field
and not a surface field), it will use only the entire cells in Omega_1 - so no,
cells that only touch the region in a vertex or edge/face are not in the field.

But you can define a field over the whole domain Omega, and integrate over
Omega_1. What do you try to achieve?

>
> In wich file are the basis functions for the fe fields defined ?

See sfepy/discrete/fem/poly_spaces.py.

r.

[Dimitri Balasoiu]

I want to compute the energy of a fractured medium.

I want to minimize an energy of the following type :

\int _Omega1 |\nabla u_1|^2 + \int _Omega2 |\nabla u_2|^2 + Energy(fracture).


How would I specify the domain of integration ? I couldn't find it in the documentation.


Dimitri

[Robert Cimrman]

On 04/25/2016 12:01 PM, Dimitri Balasoiu wrote:
> I want to compute the energy of a fractured medium.
>
> I want to minimize an energy of the following type :
>
> \int _Omega1 |\nabla u_1|^2 + \int _Omega2 |\nabla u_2|^2 +
> Energy(fracture).

Can u_1 and u_2 have the same field?

>
> How would I specify the domain of integration ? I couldn't find it in the
> documentation.
>

Check [1] - the integration is over a region.
r.

[1] http://sfepy.org/doc-devel/users_guide.html#syntax-of-terms-in-equations