[CIG-CS] Particle problems and Semi-Lagrangian Schemes for Gamr

[Walter Landry]

Back to the list.

Matthew Knepley <knepley at mcs.anl.gov> wrote:
> Of course, you know I like this method. Its the "right thing to
> do"TM However, I think the question to ask is:
> 
>   "How do you guys check the solution?"
> 
> If its something like conservation of mass fraction, you can assure
> people that a semi-lagrangian method with an accurate time-stepper
> will obey that.

I think the only way to solve the problems that I detailed in my
previous email ("Gamr status") is to track interfaces (unless you have
a better idea?).  If we track interfaces well, then we get mass
conservation.  Particles or semi-lagrangian would be used to track
material properties like integrated strain or elastic stress.  So, for
example, to reproduce what CitcomS does, neither is required.

However, we would still need particles for tracers.  But those are not
used when computing solutions, so it is not necessary to have such a
dense array.

Cheers,
Walter Landry
walter at geodynamics.org