SPECFEM3D_Cartesian simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra (structured or not).
I encountered a problem when reading external tomography files. The problem was not present in 8/11/2014 devel, but it is present in 11/11/2014 devel. I have a very simple example where I read in three external tomography files for the homogeneous halfspace example. This requires modifying the materials_file to have -1, -2, -3 for each element, and then the nummaterial_velocity_file is this:
But my impression is that the old format still applies and should work. If anyone has suggestions or is willing to try this simple example, please let me know. I can send the three xyz tomo files (they are small) and also the modified materials_file. Thanks.
I encountered a problem when reading external tomography files. The problem was not present in 8/11/2014 devel, but it is present in 11/11/2014 devel. I have a very simple example where I read in three external tomography files for the homogeneous halfspace example. This requires modifying the materials_file to have -1, -2, -3 for each element, and then the nummaterial_velocity_file is this:
2 -1 tomography elastic homogeneous_halfspace_tomo_dx02dy02dz01_ztop0.xyz 1 2 -2 tomography elastic homogeneous_halfspace_tomo_dx04dy04dz04_ztop10.xyz 1 2 -3 tomography elastic homogeneous_halfspace_tomo_dx16dy16dz08_ztop50.xyz 1
(The Par_file setting is MODEL = default)
The exit in the recent version occurs in model_tomography.f90:
Error specified tomographic material id was not found
In this case it arises because imat = 3 and imaterial = 1 (only in the 11/11/2014 version).
output_mesher.txt suggests that it is successfully reading these files, as it shows this:
./DATA/tomo_files/homogeneous_halfspace_tomo_dx02dy02dz01_ztop0.xyz number of grid points = NX_NY_NZ: 57132
./DATA/tomo_files/homogeneous_halfspace_tomo_dx04dy04dz04_ztop10.xyz number of grid points = NX_NY_NZ: 14700
./DATA/tomo_files/homogeneous_halfspace_tomo_dx16dy16dz08_ztop50.xyz number of grid points = NX_NY_NZ: 300
But that is the end of output_mesher.txt, since the exit occurs there.
There have been several changes in model_tomography.f90 between these versions: https://github.com/geodynamics/specfem3d/commits/devel/src/generate_databases/model_tomography.f90
But my impression is that the old format still applies and should work. If anyone has suggestions or is willing to try this simple example, please let me know. I can send the three xyz tomo files (they are small) and also the modified materials_file. Thanks.