search

geodynamics / pylith

PyLith is a finite element code for the solution of dynamic and quasi-static tectonic deformation problems.
Other
154 stars 98 forks source link

Default Boundary Conditions #130

Closed kimtaeho07 closed 5 years ago

kimtaeho07 commented 5 years ago

Is the default boundary condition traction-free? I've noticed that in some of the tutorials, boundary conditions are not specified for the entire domain by the relevant .cfg file. However, it's often that the output shows non-zero displacements along these boundaries. I could not find the information in the manual.

willic3 commented 5 years ago

If I’m understanding your question correctly, the default boundary condition is Dirichlet, and the default Dirichlet condition is zero displacement.

If you don’t constrain a boundary, no condition is applied, and if you do constrain a boundary, only the specified degrees of freedom are fixed.

baagaard-usgs commented 5 years ago

Expanding on Charles's answer...

If you do not specify a boundary condition for one of the boundaries, then it will be traction free. This is generally true when using the finite-element method.

If you do specify a boundary condition, the default type of boundary condition is a Dirichlet boundary condition with zero values.

DianKusumawati commented 9 months ago

Dear Sir @baagaard-usgs @willic3 @kimtaeho07

I would like to ask regarding the zero boundary condition.

If I would like to make an active tectonic model (like the one in the figure below), which is the origin of tectonic loading at the plate interface, with no boundary condition at the periphery/ sides of the mesh; does that mean I should put zero displacement boundary condition at the periphery/ sides of the mesh?

fig

The figure is from Terakawa and Matsu'ura (2009) https://academic.oup.com/gji/article/178/3/1663/1998838

Any guidance will be very valuable. Thank you

baagaard-usgs commented 9 months ago

If you want to drive deformation inside a model that mimic plate motion, then usually you would apply time-dependent Dirichlet boundary conditions on the lateral boundaries that are parallel to plate motion (boundaries with normals parallel to the y direction) and specify the velocity. On the bottom, you would apply a Dirichlet zero-displacement boundary condition in the vertical direction (no vertical motion). On the boundaries with normals parallel to the x direction, you would keep them traction free (no boundary condition).

DianKusumawati commented 9 months ago

Thank you for the comment and suggestion. I will try the boundary condition.