[CIG-SHORT] RuntimeError: Invalid fault mesh: Too many vertices of an element on the fault

Mon Jan 7 20:44:39 PST 2013

It is difficult to insure you will not have 3 nodes of a triangle or 4 nodes of a tet all sitting on a concave boundary (fault). The data structure we use in LAGriT to specify faces (edges in 2d triangles) is an integer list where the first integer is the global element number and the second integer is the local face number (tri: 1,2,3 or tet 1,2,3,4)

Carl

________________________________
From: cig-short-bounces at geodynamics.org [cig-short-bounces at geodynamics.org] on behalf of Charles Williams [C.Williams at gns.cri.nz]
Sent: 06 January 2013 20:53
To: surendra at caltech.edu
Cc: cig-short
Subject: Re: [CIG-SHORT] RuntimeError: Invalid fault mesh: Too many vertices of an element on the fault

Hi Surendra,

Dealing with this sort of geometry is a bit tricky.  PyLith needs to know which face (or edge in 2D) lies on the fault.  When all of the fault vertices lie on the fault, it's not possible to figure this out.  We used to have an alternative method for specifying faults with LAGriT-generated meshes.  There was an option to provide faces (or edges) rather than just vertices.  I'm not sure if this option still exists or not.

As an alternative, if there are only a few places where the fault has a sharp bend or kink, you could put in an artificial boundary perpendicular to the fault that bisects the kink.  I'm not sure, but I think this should force Cubit to generate triangles that don't have all 3 vertices on the fault.  Let me know if this works.

Cheers,
Charles

On 4/01/2013, at 9:32 AM, surendra at caltech.edu<mailto:surendra at caltech.edu> wrote:

Hi Charles,

I have a similar error message in 2D.  My geometry is a surface with bunch of spline curves inside.  I zoomed in on the fault where this error occurred.  I don't see how there could be problem forming a cohesive cell in such case.  Any kink for triangle elements will have this issue.  Mesh quality is fine everywhere.  Maximum condition number is ~1.5.

Any thoughts?

thanks,
Surendra

----
Traceback (most recent call last):
  File "/Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/lib/python2.7/site-packages/pylith/apps/PetscApplication.py", line 65, in onComputeNodes
    self.main(*args, **kwds)
  File "/Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/lib/python2.7/site-packages/pylith/apps/PyLithApp.py", line 103, in main
    mesh = self.mesher.create(self.problem.normalizer, interfaces)
  File "/Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/lib/python2.7/site-packages/pylith/topology/MeshImporter.py", line 120, in create
    self._adjustTopology(mesh, faults)
  File "/Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/lib/python2.7/site-packages/pylith/topology/MeshGenerator.py", line 127, in _adjustTopology
    interface.adjustTopology(mesh, firstFaultVertex, firstLagrangeVertex, firstFaultCell)
  File "/Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/lib/python2.7/site-packages/pylith/faults/faults.py", line 367, in adjustTopology
    def adjustTopology(self, *args): return _faults.FaultCohesive_adjustTopology(self, *args)
RuntimeError: Error occurred while adjusting topology to create cohesive cells for fault 'flt61'.
Invalid fault mesh: Too many vertices of an element on the fault
application called MPI_Abort(MPI_COMM_WORLD, -1) - process 0
/Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/bin/nemesis: mpirun: exit 255
/Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/bin/pylith: /Users/surendra/Desktop/pylith-1.8.0-darwin-10.6.8/bin/nemesis: exit 1
-------

<flt61.png><flt61_nodes.png>

On Jul 29, 2012, at 6:12 PM, Charles Williams wrote:

Hi Hongfeng,

In the past, this sort of error has occurred when there is a lot of curvature in the fault in combination with poor mesh quality.  I'm assuming you're using a tetrahedral mesh.  In that case, having too many vertices on the fault indicates that all 4 vertices for a cell lie on the fault.  In your case, this is probably indicative of a poor quality mesh.  For example, a very flat cell could have all 4 vertices on the fault.  When that happens, it's no longer clear how to split the cell faces to form a cohesive cell.  You will need to improve the mesh quality to remove any cells with this sort of problem.

Cheers,
Charles

On 30/07/2012, at 1:03 PM, Hongfeng Yang wrote:

Hi all,

I am running dynamic rupture simulations on a fault mesh with realistic
geometry. The simulation was done on a mesh with certain discretization
size (say d1). However, when I decreased the discretization size of the
mesh (say d2), PyLith stopped after reading the mesh due to an error,
Invalid fault mesh: Too many vertices of an element on the fault

I check the mesh quality with grid size of d2 and find there are a lot
of poor quality elements. But I do not understand how there could be too
many vertices of one element on the fault surface? What does this error
message mean, and what is causing the problem?

Thanks,

Hongfeng

--
Postdoc Investigator
Woods Hole Oceanographic Institution
Dept. Geology and Geophysics
360 Woods Hole Rd, MS 24
Woods Hole, MA 02543

_______________________________________________
CIG-SHORT mailing list
CIG-SHORT at geodynamics.org<mailto:CIG-SHORT at geodynamics.org>
http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short

Charles A. Williams
Scientist
GNS Science
1 Fairway Drive, Avalon
PO Box 30368
Lower Hutt  5040
New Zealand
ph (office): 0064-4570-4566
fax (office): 0064-4570-4600
C.Williams at gns.cri.nz<mailto:C.Williams at gns.cri.nz>

_______________________________________________
CIG-SHORT mailing list
CIG-SHORT at geodynamics.org<mailto:CIG-SHORT at geodynamics.org>
http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short

Charles A. Williams
Scientist
GNS Science
1 Fairway Drive, Avalon
PO Box 30368
Lower Hutt  5040
New Zealand
ph (office): 0064-4570-4566
fax (office): 0064-4570-4600
C.Williams at gns.cri.nz<mailto:C.Williams at gns.cri.nz>

Notice: This email and any attachments are confidential. If received in error please destroy and immediately notify us. Do not copy or disclose the contents.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geodynamics.org/pipermail/cig-short/attachments/20130108/02634139/attachment.htm 