It is different than FDTD. It is not iterative. It looks like FEM method. But I don't know exactly what the difference is? This is what I am asking? Both methods generate matrices and take the matrix inverse at the end to solve the diff. eqn. Both methods use grids (nodes). So what is the difference?
hi there
i know that the FDFD "Finite difference frequncy domain" is based on solving a differential equations at a single frequency, hence no time dependant in the equations.
Added after 1 minutes:
FDTD
* Finite Difference Time Domain (FDTD) is fast, accurate, easy to implement, and intuitive.
* FDTD is best for wideband scattering in non-uniform media.
* Special means have been developed for dispersive media, such as soil.
FDFD
* Finite Difference Frequency Domain (FDFD) is also good for non-uniform media.
* FDFD handles frequency dependent media easily.
* FDFD is particularly useful for signal processing forward models
* FDFD is computationally expensive (or is it?).
FEM and FDFD are both frequency domain methods where the solution is obtained at a single frequency. The underlying equation they solve are the Maxwell's equations under time harmonic case.
On the other hand, FDTD which is a time domain method solving maxwell's equation under the time dependent case.
FEM and FDFD are very similar except that FDFD has limited applicability due to the regular hexahedral elements used unlike FEM (which can have tetrahedral or hexahedral elements in general). So the mesh for FDFD looks like the cartesian mesh for FDTD and hence the stair-case approximation of the geometry boundary is necessary.
I work on FDFD and FDTD , really i find that FDFD gives the problem less complexity, cuase, in FDFD we solve the problem ignoring the time factor, which means :no need for the source and no need for the ABC, in FEM , it is a general method, and it can be used in both 3-D an 2-D, we know that FD(FD&TD) is complicated in 3D.