[CIG-SHORT] How does Pylith number vertices on fault

[Brad Aagaard]

Hongfeng,

The v_fault in the error message refers to the internal Sieve label for 
a vertex on the fault. It does not directly correspond to the numbering 
of vertices in the mesh or fault in the output or the input mesh. For 
example, Sieve does not numbering vertices differently from cells. In 
fact, it labels cells first and then labels vertices. This should be 
transparent to the user except in diagnosing error messages that expose 
the Sieve labels.

The error message you are getting suggests the solver tolerances are not 
setup properly. You are getting a very small opening and small 
compressive normal traction, which means the linear solver tolerance was 
not tight enough or the friction zero tolerance was not large enough.

Regards,
Brad


On 10/04/2012 07:39 AM, Hongfeng Yang wrote:
> Hi Developers,
>
> I am tracking time history of variables on particular vertices on the
> fault. However, I found the vertex ID in Pylith (v_fault) is larger than
> the total number of vertices on the fault. How can I convert the value
> of v_fault back to the vertex location on the fault surface?
>
> For example, I got the following fault opening error message
> WARNING! Fault opening with nonzero traction., v_fault: 4154, opening:
> 1.03716e-10, normal traction: -0.0042952
> mpinemesis: faults/FaultCohesiveDyn.cc:382: virtual void
> pylith::faults::FaultCohesiveDyn::integrateResidual(const
> pylith::topology::Field<pylith::topology::Mesh>&, PylithScalar,
> pylith::topology::SolutionFields*): Assertion `fabs(tractionNormal) <
> _zeroTolerance' failed.
> [0]0:Return code = 0, signaled with Aborted
>
> v_fault points to vertex ID 4154. The total number of vertices on the
> fault is 1472, however.
>
> Thanks,
>
> Hongfeng
>