Name | Modified | Size | Downloads / Week |
---|---|---|---|
slice_steady_state.avi | 2016-05-01 | 8.4 MB | |
midplane.avi | 2016-05-01 | 16.0 MB | |
readme.txt | 2016-04-25 | 1.6 kB | |
BBKJ_par.sh | 2016-04-25 | 955 Bytes | |
3D_P_4.cpp | 2016-04-17 | 21.8 kB | |
Totals: 5 Items | 24.4 MB | 0 |
The above code solves the 3D Heat conduction equation using Flux based solution methodology on a Staggered Cartesian grid. Presently, the Initial condition is taken as 0. The Boundary condition is Dirichlet on all the walls. The Volumetric Heat generation is taken as 0. Just take care in the choice of time step. Will upload with CFL condition asap. The memory has been allocated contiguosly which, for the ease of memory access. Parallelized as SPMD using MPI. Instructions on running the Parallel program: NP corresponds to the number of process. So, if you are going to use 2/4/8/16... process, change the value accordingly. To compile the program: Type the command: /opt/mpich2/gnu/bin/mpicxx -o Compiled_code.out 3D_P_4.cpp (Note : /opt/mpich2/gnu/bin/ corresponds to the path where, mpicxx command is available) If you want to submit your job using Grid Engine Scheduler, download BBKJ.par file and change the job name, Number of process in the line - #$ -pe mpich 16 (Here 16 corresponds to 16 process). and the directory path where the .out file is located. (MYPROG="/home/sagar/home/FINEGRID/Compiled_code.out") Save the file in the same directory as the .out file and type the command: qsub BBKJ_par.sh Your job has been submitted with the desired number of cores. For any problems/errors/queries: Please drop a mail at : gb.arivazhagan@gmail.com gb.arivazhagan@iitb.ac.in For tutorials on Message Passing Interface: (MPI) Please do visit : http://mpitutorial.com/tutorials/ You need Tecplot software to postprocess the results. Enjoy! Have fun!