[Getdp] result all zero

Christophe Geuzaine Christophe.Geuzaine at ulg.ac.be
Tue Oct 30 09:36:35 CET 2001


Lin Ji wrote:
> 
> Hi, Christophe,
>     I am not trying to solve the original wave equation in frequency domain.
> I am solving the Helmholtz equation
>         u_xx + u_yy + (\tau^2 / c^2)u = 0
> directly where the subscript _xx denotes the 2nd partial derivative to x and
> \tau corresponds to a specific frequency. I had the following formulation for
> the above equation:
> 
> Formulation {
>   { Name hhltz; Type FemEquation;
>     Quantity {
>       { Name u; Type Local; NameOfSpace Hgrad; }
>     }
>     Equation {
>       Galerkin { [ tau_c*tau_c/mu[] * Dof{u} , {u} ];
>                  In Omega; Integration I1; Jacobian JVol;  }
> 
>       Galerkin { [ -Dof{d u} , {d u} ];
>                  In Omega; Integration I1; Jacobian JVol;  }
> 
>       Galerkin {  [ f_bc[] , {u} ];
>                  In Gamma_f; Integration I1; Jacobian JSur;  }
>     }
>   }
> }
> 
> where mu[] = c^2 and tau_c denotes the central frequency and f_bc[] is the
> Neumann boundary condition. They are all defined in the Function object.

Ok. I tried your formulation on a simple rectangular mesh, and I get a
solution... Maybe your mesh is wrong?

-- 
Christophe Geuzaine

Tel: 32 (0) 4 366 37 10    http://geuz.org
Fax: 32 (0) 4 366 29 10    mailto:Christophe.Geuzaine at ulg.ac.be