[CIG-SHORT] problems applying initial stress conditions

[Brad Aagaard]

Lucas,

Your output settings for the elastic material component are missing

cell_filter = pylith.meshio.CellFilterAvgMesh

Without this setting, you are getting output at all of the quadrature 
points for each cell rather than just a single average value for each 
tensor component. ParaView doesn't know how to deal with so many values 
so it is probably just assuming there is 1 value per cell. It also 
doesn't know how to display 48 components (8 quad points times 6 
components), so it won't let you pick a tensor component. As a result, 
the data within ParaView is garbled.

Regards,
Brad


On 06/12/2012 08:00 AM, Lucas Abraham Willemsen wrote:
> Hi,
>
> After completing the Pylith tutorials I tried to model some small problems by adapting the tutorial files. When doing this I ended up with some problems I can not explain.
>
> I wanted to model a fully elastic block cut in half by a rate and state friction fault. The entire block has the name 'elastic'. I try to apply an initial compressional stress of 10 MPA in the x direction. I then impose a y-velocity on both the negative and the positive x faces of the block in order to get a shear stress on the fault. This attempt is given the name 'test5_dynamic.cfg' (please ignore the comments, they are remnants from the 'step14' tutorial).
>
> The biggest problem is that the initial stress of 10 MPa is not applied to the entire block with the name 'elastic'. Instead, it is applied to a strange F-shaped region which is not defined in my cubit scripts (which included in 'mesh' folder). I believe that because of this reason the fault slips uniformly instead of sticking and slipping. A figure of the strange F-shape is included in the archive I attached to this email.
>
> In order to investigate the problem I removed the slip rate on the block while keeping the initial conditions. The name of this experiment is 'test7_dynamic.cfg'. The stress situation at t = 0 still had the weird F-shaped region to which the 10 MPa was applied. In the next time steps this strange F-shaped region did not change shape. Some parts of the block had no stress on them while others (in the F-shape) had 10 MPa. I am puzzled by what is going on. I would be very grateful if someone could show me where I made a mistake.
>
> The file showing the stress situation has the name 'testX-dynamic-stressstrain.xmf' in which X is either a 5 or a 7 depending on which problem is run.
>
> Cheers,
> Lucas
>
> P.S.
> Paraview does not allow me to investigate the components of the stress tensor for some reason. I can only see one value for 'stress'.
>
> (I could not include the HDF5 output files without making the attachment excessively large)
>
>
>
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short