On 1/15/21 10:46 AM, Animesh Rastogi IIT Gandhinagar wrote:
I am using a modified version of Code Gallery Quasi Static Compressibility
<https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdealii.org%2Fdeveloper%2Fdoxygen%2Fdeal.II%2Fcode_gallery_Quasi_static_Finite_strain_Compressible_Elasticity.html&data=04%7C01%7CWolfgang.Bangerth%40colostate.edu%7C3f03e1b7a92342c0331e08d8b97d7f35%7Cafb58802ff7a4bb1ab21367ff2ecfc8b%7C0%7C0%7C637463295948040000%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=wukJxVr%2BMJ0%2BGymh3b5cYRkeUH%2BFCzH2qsBBiujG2h8%3D&reserved=0>
program to compute eigenvalues and eigenvectors of my system matrix. I have
used ARPACK solver, however, I did not get the expected results. Bruno
suggested me to go for PETSc/Slepsc solver to compute the eigenvalues. For
this, I would need to convert the dealii data structures to the PETSc data
structures using PETSc wrappers.
All objects that relate to the linear system you are solving (or to the
eigenvalue problem you are solving). So, matrices and vectors.
I looked at the Step-17 and Step-36 for this. However, I am unable to
understand how I can change my current dealii data structures to the one using
PETSc wrappers. I am also confused regarding what all variables do I need to
convert using PETScWrappers. Also, for the system matrix, do I need to use
PETScWrappers::MPI::BlockSparseMatrix or PETScWrappers::BlockSparseMatrix.
That depends on whether you want to run your problem on multiple machines
using MPI. If you've been using ARPACK, then I suspect that you're not using
MPI. So then PETScWrappers::BlockSparseMatrix.
This finite strain code already assembles the system matrix in parallel using
multi-threading. What changes do I need to do in the assembly if I go on to
use PETSc data structures.
None, really. The functions filling matrices should work on PETSc data
structures just fine.
Best
W.
--
------------------------------------------------------------------------
Wolfgang Bangerth email: [email protected]
www: http://www.math.colostate.edu/~bangerth/
--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see
https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/dealii/5000cf44-3431-18a6-67a1-32beea9131a3%40colostate.edu.
