search

SPECFEM / specfem2d

SPECFEM2D simulates forward and adjoint seismic wave propagation in two-dimensional acoustic, (an)elastic, poroelastic or coupled acoustic-(an)elastic-poroelastic media, with Convolution PML absorbing conditions.
GNU General Public License v3.0
191 stars 147 forks source link

Weird results from example Rayleigh_wave_with_crack #290

Closed wanglongqi closed 9 years ago

wanglongqi commented 9 years ago

I am new to this software. I start running examples in the EXAMPLES folder. I find following weird behavior of the program.

In Rayleigh_wave_with_crack example, if you change nproc = 4, the outputs are different with nproc = 1.

For np=1, I get following output: https://youtu.be/ztzRk1ccUpc , For np=4, I get following output: http://youtu.be/wv5ekGsURaQ .

I think the results should keep the same for different np. Do I make a mistake here?

komatits commented 9 years ago

That example is the only one that cannot run in parallel (i.e. with MPI), because I use a trick inside the code to include a mathematical crack. Thus always run it with nproc = 1. I'll add a note in the README file of that example.

Dimitri.

On 04/17/2015 07:18 AM, WANG Longqi wrote:

I am new to this software. I start running examples in the |EXAMPLES| folder. I find following weird behavior of the program.

In |Rayleigh_wave_with_crack| example, if you change |nproc = 4|, the outputs are different with |nproc = 1|.

For |np=1|, I get following output: https://youtu.be/ztzRk1ccUpc , For |np=4|, I get following output: http://youtu.be/wv5ekGsURaQ .

I think the results should keep the same for different np. Do I make a mistake here?

— Reply to this email directly or view it on GitHub https://github.com/geodynamics/specfem2d/issues/290.

Dimitri Komatitsch CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics, UPR 7051, Marseille, France http://komatitsch.free.fr

wanglongqi commented 9 years ago

@komatits Thanks for your explanation!