[aspect-devel] strain rate corner values and velocity boundary conditions

[Timo Heister]

Dear Anne,

(and back to the list)

> Thanks so much for the patch! I've tried it (only 1 run yet, so not thoroughly) and now I get what we want (top figure instead of bottom figure).

excellent. If nobody else complains about the format, I will check this in soon.

> I've also attached the output file (the model is a bit different from the one I sent you). The inputfile was set to a maximum of 200 nonlinear iterations, but this is an educated guess of the needed number of iterations, usually I use a criterion based on the difference in solution for stopping the nonlinear iterations.

Oh, this made me realize that the "iterated Stokes" solver has no
automatic stopping criterion. I just implemented this. See the new
"Nonlinear solver tolerance" variable.

> Also, while we're on the subject of boundary conditions (I hope you don't mind me asking), it would be very nice to have open boundary conditions (imposing a lithostatic pressure condition for the normal stress). This would be very useful in mantle models as free slip boundaries have been shown to affect mantle flow. I have attached a paper discussing this (Chertova et al. 2012).

Thomas Geenen is one of the authors.

@Thomas: if I understand this correctly, you 1. set the tangential
component of the velocity to zero (Anne, you can do that already) and
2. set the stress on the boundary to the lithostatic pressure. Is that
correct? Are those pressure values set by the user or updated from the
computation?

> Or do you want me to post this on the mailing list?

Yes (and I just did). :-)

-- 
Timo Heister
https://urldefense.proofpoint.com/v1/url?u=http://www.math.clemson.edu/~heister/&k=dsHy%2FVymniCD0osh6tze%2Fw%3D%3D%0A&r=WLGUy1n%2F1Vzsqown7KFPcAEzR5DIdbP771jYw4Spy%2Bk%3D%0A&m=iitgjBqPvMnLhLHo4boyiZtTKZCiZwop1Etgi3Zi0Do%3D%0A&s=9d888a3988062d23c4dc56611b700196ed7200e1fdfd04ad79fcd3a03fef6adb