[CIG-SHORT] Re: PyLith BC help

[Matthew Knepley]

Always mail the list. Brad and Charles know the most about this problem.
However, we should see text from the exception. Can you run in the debugger
with --petsc.start_in_debugger and see where it fails?

   Matt

On 8/29/07, Rick Bennett <rab at geo.arizona.edu> wrote:
> PyLith User Manual authors,
>
> I am in need of help using Pylith that is time sensitive.   I don't
> think my question is worth an email to the short-term email list.
> But please steer me in the right direction.
>
> I am trying to specify boundary conditions on all sides (except the
> top surface) of a 3d model.  The model mesh and nodesets have been
> created using CUBIT (my journal file is attached in the file named
> mesh3d.jou).  The nodesets have been created such that they do not
> share elements along the edges as can be verified by their empty
> intersections.  Two cfg files are attached.  The BC are specified in
> the second, called earthquake.cfg.   The +y axis defines vertical in
> this model.   I have implemented the bug fix that Brad sent to the
> mail list earlier today (or yesterday) regarding specification of
> up_dir and normal_dir.
>
> Pylith aborts with the following message:
>
>  >> /Users/rab/pylith/pylith-1.0.1-darwin-powerpc/lib/python2.3/site-
> packages/pylith/problems/Formulation.py:151:initialize
> -- implicit(info)
> -- Creating solution field.
> [0]0:Return code = 0, signaled with Abort trap
>
> I am not yet familiar enough with the structure of the code to know
> where the abort takes place.  The error appears to be related to the
> specification of conflicting BC.  But what specifically does pylith
> not like about the BC, and how do I correct them or the nodesets to
> which they refer?
>
> Thanks,
> Rick
>
>


-- 
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