Modelagem eletromagnética 2.5-D de dados geofísicos através do método de diferenças finitas com malhas não-estruturadas
We present a 2.5D electromagnetic formulation for modelling of the marine controlledsource electromagnetic (mCSEM) using a Finite Diference frequency domain (FDFD) method. The formulation is in terms of secondary fields thus removing the source point singularities. The components of the electroma...
| Autor principal: | MIRANDA, Diego da Costa |
|---|---|
| Grau: | Tese |
| Idioma: | por |
| Publicado em: |
Universidade Federal do Pará
2018
|
| Assuntos: | |
| Acesso em linha: |
http://repositorio.ufpa.br/jspui/handle/2011/10223 |
| Resumo: |
|---|
|
We present a 2.5D electromagnetic formulation for modelling of the marine controlledsource
electromagnetic (mCSEM) using a Finite Diference frequency domain (FDFD)
method. The formulation is in terms of secondary fields thus removing the source
point singularities. The components of the electromagnetic field are derived from
the solution of the magnetic vector potential and electric scalar potential, evaluated
in the entire problem domain that must be completely discretized for the use of
the FDFD. Finite difference methods result in large sparse matrix equations that
are efficiently solved by sparse matrix algebra preconditioned iterative methods. To
overcome limitations imposed by structured grids in the traditional FDFD method,
the new method is based upon unstructured grids allowing a better delineation of
the geometries. These meshes are completely adaptable to the models we work with,
promoting a smooth design of their structures, and may only be refined locally in
regions of interest. We also present the development of RBF-DQ method, (radial
basis function differential quadrature) which makes use of the technique of functions
approximation by linear combinations of radial basis functions (RBF) and the
technique of differential quadrature (DQ) for approximation of the derivatives. Our
results show that the FDFD method with unstructured grids when applied to geophysical
modeling problems, yield improved quality of modeled data in comparison
with the results obtained by traditional techniques of FDFD method. |