Building main grid,#geometry_objects_read

45 views
Skip to first unread message

Dave Bédard

unread,
Mar 2, 2025, 6:01:13 PMMar 2
to gprMax-users
Hi everyone,

I'm trying to use #fractal_box and Pepinski mixing model. Building main grid take a long time, so as recommended, I used #geometry_objects_write to build my model in a HDF5 file. The input file Affaissement_figure23.in is attached and used to build my model.

I made another input file to run my simulation with #geometry_objects_read (Affaissement_figure23.in), but the simulation is two times longer as the grid is built two times instead of one (13 minutes total, 2 min for the simulation vs 6 minutes total, 2 min for the simulation). I thought #geometry_objects_read was not supposed to build the model on every iteration.

I also have some #material with negative conductivity in the material text file created by #geometry_objects_write.  

I don't know what I'm doing wrong, can someone help me?

Thanks,

Dave
Affaissement_figure23_geometry_materials.txt
Affaissement_figure23b.in
Affaissement_figure23.in

Craig Warren

unread,
Mar 3, 2025, 9:31:56 AMMar 3
to gprMax-users
Hi Dave,

Did you also use the argument --geometry-fixed? If you don't gprMax will build the model for every model run.

Kind regards,
Craig

Dave Bédard

unread,
Mar 3, 2025, 12:09:13 PMMar 3
to gprMax-users
Hi Craig,

No, I didn't use the -geometry-fixed argument. It's much faster now, I don't have to build the geometry every time and I don't need to write geometry object.

Thanks for your quick reply.

Best regards,

Dave
Reply all
Reply to author
Forward
0 new messages