[CIG-SHORT] problem getting fault rupture (slip weakening)

[Brad Aagaard]

Lucas,

I wouldn't be surprised if you need to adjust more than just the length 
scale. Pay attention to the time scale as well. I think a very important 
question you need to figure out is what is the order of magnitude of 
displacement and slip you expect to see. If the slip is expected to be 
somewhere near 1.0e-8, then the length scale for nondimensionalization 
is going to need to be rather small. Your discretization size would be 9 
orders of magnitude greater, so you would be trying to resolve a wide 
range of scales, which is very tough to do computationally. Those types 
of problems often require adaptive mesh refinement and other multi-scale 
techniques.

Until your nondimensionalized parameters give solution and Jacobian 
values of similar order, I think you shouldn't focus too much on solver 
performance, because it could be misleading. Nevertheless, it should get 
better with the right choice of scales.

Brad


On 4/25/13 2:18 PM, Lucas Abraham Willemsen wrote:
> Hi Brad,
>
> I will try to look at the length scales and see if I can find some appropriate nondimensionalizing values that will work with cell dimension of 10 meter and slip weakening distance of 5e-3 m . Although the preliminary investigations I did were not very hopeful when I restricted my nondimensionalization changes to the length_scale variable.
>
> When I use 1 meter as slip weakening distance the linear solves go fast so I guess the conditioner number of the linear system is good and that the nondimensionalization scheme gives values with similar orders of magnitude. But even though linear convergence is great, the nonlinear iterations do not converge. Do you know of ways to improve nonlinear convergence as well?
>
> thanks,
> Lucas
> ________________________________________
> From: cig-short-bounces at geodynamics.org [cig-short-bounces at geodynamics.org] on behalf of Brad Aagaard [baagaard at usgs.gov]
> Sent: Thursday, April 25, 2013 14:49
> To: cig-short at geodynamics.org
> Subject: Re: [CIG-SHORT] problem getting fault rupture (slip weakening)
>
> Lucas,
>
> I re-read your email and now I see that you did try a larger
> slip-weakening slip distance. I am not familiar with the problem you are
> solving and don't have time to look into it.
>
> I suspect you still have some scaling issues. I would start by writing
> down the order of magnitudes of the stresses, tractions, BC,
> displacements, and slip and rates you expect to see. These are the
> driving scales in your problem. Select your nondimensionalization
> parameters so that the nondimensional quantities of these fields are
> within at least few orders of magnitude of each other. The further apart
> the nondimensionalized fields are, the more ill conditioned the problem
> is going to be.
>
> Regards,
> Brad
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>