3D finite-volume forward modeling of transient electromagnetic using octree meshes

Jing Xu,Cao HuaKe,Zhou JianMei,Liu WenTao,Li Xiu,Wen YiHao
DOI: https://doi.org/10.6038/cjg2022Q0798
2023-01-01
Abstract:An efficient and stable 3D transient electromagnetic (TEM) simulation algorithm is the theoretical basis for processing and interpretation methods. Accelerating the calculation process and reduce computer memory consumption is particularly important. In this paper, we use the mimetic finite volume algorithm to discretize the time-domain Maxwell's equations on octree meshes. Compared with the staggered mesh, the octree mesh can significantly reduce the number of elements at the boundary with same grid fineness in the core area, and the number of grids at the boundary is more flexible for the simulation of complex geometric boundaries. Compared with unstructured tetrahedral grids, the location of octree cells is more regular. Thus, using octree meshes can reduce the unknown of equations to be solved and the memory consumption. After discretizing the time-domain Maxwell's equations by the mimetic finite volume method, the transient electromagnetic response can be expressed as a matrix exponential function with respect to the initial magnetic field. Using the shift-and-invert (SAI) Krylov subspace model reduction algorithm, a series of TEM modeling results can be obtained by performing only one matrix decomposition and hundreds of reverse substitutions on the displacement coefficient matrix. The algorithm in this paper optimizes the 3D TEM modeling process in both spatial discretization and equation solving. The numerical experiments demonstrated that the method is accurate and efficient.
What problem does this paper attempt to address?