[CIG-SHORT] PyLith problems

[Matthew Knepley]

On Dec 6, 2007 2:40 PM, Jiangning Lu <jlu at chandler.mit.edu> wrote:
> Hi Brad,
>
> I'm using the same boundary conditions in this model with the
> example/3d/tet4/dislocation case.
>
> Using the same mesh,I also run another problem successfully without a fault
> specified. From this can I say that the material ids are correct? In other
> words, specifying a fault will not change the material id settings?

The fault creates a saddle point, which has a 0 block on the diagonal.
This makes
ILU crap out. You can either switch to another PC, -petsc.sub_pc_type=sor, or
you can tell PETSc to shift, -petsc.sub_pc_factor_shift_positive_definite.

  Thanks,

    Matt

> Thanks,
>
> Jiangning
>
>
> Quoting Brad Aagaard <baagaard at usgs.gov>:
>
> > Jiangning-
> >
> > The zero pivot suggests that you have a singular matrix. This usually means
> > that the problem is not sufficiently constrained (some sort of rigid body
> > motion is allowed). This is what why the asm preconditioner doesn't work for
> > the two-cell examples. I think it would also show up if you have cells that
> > are not associated with any material (i.e., the ids don't match), so that the
> > stiffness matrix contains a bug of zeros.
> >
> > Brad
> >
> >
> > On Thursday 06 December 2007, Jiangning Lu wrote:
> >> Hi,
> >>
> >> When I tried to run one of my model by setting pc_type = asm, I got the
> >> following error:
> >> ......
> >>  -- Solving linear equations.
> >> [0]PETSC ERROR: --------------------- Error Message -------------
> >> [0]PETSC ERROR: Detected zero pivot in LU factorization
> >> see
> >> http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html#Ze
> >> roPivot! [0]PETSC ERROR: Zero pivot row 61695 value 2.29519e-19 tolerance
> >> 2.93442e-13 * rowsum 0.293442!
> >> [0]PETSC ERROR: Petsc Development Version 2.3.3, Patch 8, unknown HG
> >> revision: unknown
> >> [0]PETSC ERROR: See docs/changes/index.html for recent updates.
> >> [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
> >> [0]PETSC ERROR: See docs/index.html for manual pages.
> >> [0]PETSC ERROR:
> >> ------------------------------------------------------------------------
> >> [0]PETSC ERROR:
> >> /chandler/home/jlu/scratch1/pylith_workarea/nemesis/trunk/mpinemesis on a
> >> linux-gnu named elsasser by jlu Thu Dec  6 12:50:16 2007
> >> [0]PETSC ERROR: Libraries linked from
> >> /chandler/home/jlu/scratch1/pylith_workarea/petsc-dev/lib
> >> [0]PETSC ERROR: Configure run at Wed Nov 28 15:24:59 2007
> >> [0]PETSC ERROR: Configure options --with-sieve=1 --with-clanguage=C++
> >> --with-shared=1 --download-f-blas-lapack=1 --with-boost=1
> >> --download-boost=1 --with-chaco=1 --download-chaco=1 --with-mpi=1
> >> --download-mpich=1
> >> --with-mpi-shared=1
> >> [0]PETSC ERROR:
> >> ------------------------------------------------------------------------
> >> [0]PETSC ERROR: MatLUFactorNumeric_Inode() line 1495 in
> >> src/mat/impls/aij/seq/inode.c
> >> [0]PETSC ERROR: MatLUFactorNumeric() line 2276 in
> >> src/mat/interface/matrix.c [0]PETSC ERROR: PCSetUp_ILU() line 564 in
> >> src/ksp/pc/impls/factor/ilu/ilu.c [0]PETSC ERROR: PCSetUp() line 788 in
> >> src/ksp/pc/interface/precon.c [0]PETSC ERROR: KSPSetUp() line 234 in
> >> src/ksp/ksp/interface/itfunc.c [0]PETSC ERROR: PCSetUpOnBlocks_ASM() line
> >> 224 in src/ksp/pc/impls/asm/asm.c [0]PETSC ERROR: PCSetUpOnBlocks() line
> >> 821 in src/ksp/pc/interface/precon.c [0]PETSC ERROR: KSPSetUpOnBlocks()
> >> line 158 in src/ksp/ksp/interface/itfunc.c [0]PETSC ERROR: KSPSolve() line
> >> 347 in src/ksp/ksp/interface/itfunc.c [0]PETSC ERROR: preallocateOperator()
> >> line 190 in
> >> unknowndirectory/solver_embed.cpp
> >>
> >> When I changed the setting to pc_type = jacobi, the error did not show up,
> >> but the result looks not right either.
> >> Any thoughts?
> >>
> >> Thanks!!
> >> Jiangning
> >>
> >> ---
> >> Jiangning Lü
> >> 54-521     x3-6389
> >> Geophysics
> >> EAPS, MIT
> >> johnnylu at mit.edu (????
> >> _______________________________________________
> >> CIG-SHORT mailing list
> >> CIG-SHORT at geodynamics.org
> >> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
> >
> >
> > _______________________________________________
> > CIG-SHORT mailing list
> > CIG-SHORT at geodynamics.org
> > http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
> >
>
>
>
> ---
> Jiangning Lü
> 54-521     x3-6389
> Geophysics
> EAPS, MIT
> johnnylu at mit.edu (????
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>



-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which
their experiments lead.
-- Norbert Wiener