[CIG-SHORT] 2D Simulations

[Brad Aagaard]

On 9/14/11 9:42 AM, James Moore wrote:
> I’m trying to approximate an infinitely long strike slip fault… which is
> essentially a 2D anti-plane problem. Hence my thoughts regarding using
> periodic domains to make the 3D case resemble a stretched 2D case with
> no end effects.

For an infinitely long fault in the y direction, you would not expect 
any motion in the x direction, so you can pin this DOF using a Dirichlet 
BC applied to the +y and -y faces in addition to the +x and -x faces. 
That said, you can't overlap the fault with the Dirichlet BC so this 
does result in some small scale edge effects.

Brad