[CIG-LONG] turning off elasticity in SNAC

[Michael Kaplan]

Hi,
  Im interested in exploring the effects of elasticity on lithospheric
dynamics and consequently am looking for a way to use purely viscous, purely
plastic, and visco-plastic rheologies in SNAC.  Also, to compare SNAC and
GALE it would be helpful to use the same rheologies.  Using SNAC's Maxwell
plug-in and setting elastic moduli orders of magnitude larger than the
viscosity is problematic because it creates a very short maxwell relaxation
time, which SNAC enforces must be greater than the time step.
  Switching dtType from a constant time step to dynamic time stepping allows
the simulation to run, but produces seemingly unreasonable results.  It
seems with dynamic time stepping that only the first time step is less than
the maxwell relaxation time, with all later time steps unconstrained by the
relaxation timescale.  For reasons that I dont understand, this seems to
create simulations for which all the deformation is localized to the
boundaries which have a non-zero kinematic BC.
  One solution I plan on testing is to create a new plug-in for a purely
viscous rheology, but any other solutions or comments would be greatly
appreciated.
Thanks in advance for the help!

-Michael Kaplan
USC Geodynamics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geodynamics.org/pipermail/cig-long/attachments/20100723/6a7ade3e/attachment.htm