[CIG-SHORT] RuntimeError: Determinant of Jacobian is smaller than minimum permissible value (1e-06)!

[Lucas Abraham Willemsen]

Hi,

I'm trying to help someone set up a Pylith problem in which a crack is opening up in an elastic material. I generated a 2D block in cubit with a small interface on the bottom, representing the crack. The plan is to apply a tensile stress on the bottom half of the crack interface, and see how the top half opens up. But after generating the mesh and creating a simple Pylith script I get the following error:

RuntimeError: Determinant of Jacobian (6.25e-10) for cell 0 is smaller than minimum permissible value (1e-06)!

After googling for a while I found two possible explanations for this error:

1: applying wrong quadrature to elements of the mesh:  http://www.geodynamics.org/pipermail/cig-short/2010-October/000660.html
2: distorted elements in mesh: http://www.geodynamics.org/pipermail/cig-short/2010-October/000660.html

I consistently use 2nd order quadrature points in my tri3 mesh, just like what is done in the bar_shearwave example for instance. I also use FIATSimplex consistently. So I have no reason to believe it is reason number 1. Reason number 2 also seems unlikely to me. I adapted the mesh so that all elements have similar size. The mesh looks good and there are no elements with extreme aspect ratio's. I have attached a screenshot showing the mesh. The orange line at the bottom is the highlighted crack interface.

If it is not too much trouble, could someone please take a look at it? It could very well be something trivial but I'm not able to pinpoint the cause.

best,
Lucas

(P.S. the reason for applying only a tensile stress to the bottom of the crack is that the top part of the crack is supposed to represent intact rock. I was planning to use a cohesive value for the top part of the crack, representative of intact rock. But the manual says that the shear strength goes to zero when the stress on a fault interface is tensile (tensile cutoff). Does this mean that using cohesion on the crack is not the way to represent the growth of the crack in intact rock?)



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geodynamics.org/pipermail/cig-short/attachments/20120727/e15ed2c5/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: openingcrack.zip
Type: application/x-zip-compressed
Size: 52818 bytes
Desc: openingcrack.zip
Url : http://geodynamics.org/pipermail/cig-short/attachments/20120727/e15ed2c5/attachment-0001.bin 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screenshot-Cubit 13.2.png
Type: image/png
Size: 369665 bytes
Desc: Screenshot-Cubit 13.2.png
Url : http://geodynamics.org/pipermail/cig-short/attachments/20120727/e15ed2c5/attachment-0001.png