[overture] Re: deforming grid PDE solver

  • From: jacob@xxxxxxxxxxxx
  • To: overture@xxxxxxxxxxxxx
  • Date: Thu, 26 Jul 2012 00:49:38 -0400

Hi Bill,

Hmm I agree, though I had thought move2.C was a bit complicated and wasn't
sure what exactly it was doing, so I more used deform.C to make my guess
at how to deform the grids...But yeah I can take a pass at making a test
code which is a bit simpler...maybe I'll try advection-diffusion first
with explicit and then with implicit stepping, since updating the solver
should "probably" be trickier. For the deforming boundary I'll probably
first use some kind of spline that deforms in a pre-defined way and try to
build a HyperbolicMapping off of it...if that doesn't work I'll back-pedal
and try something simpler...


On Wed, July 25, 2012 11:47 pm, Bill Henshaw wrote:
> Hi Jacob,
>     Thanks for the code segments. You have made some good progress,
> although
> I do see some problems with what you are doing.
> Recall:
>     GOLDEN RULE 3 of programming -- "It always saves time in
> the long run to have a simple test code".
> Corollary of Golden Rule 3: you may hate this rule but unfortunately
> it is really true.
>     Unfortunately I have not created a simple deforming grid
> PDE solver code. I think that the first step is for us to create
> such a code. I think the easiest way to do this is to merge
> deform.C and move2.C so that we solve an advection diffusion
> equation with twilight-zone on a deforming grid. Perhaps you
> want to make a first attempt at doing this? Otherwise I can
> try to put something together but I'm not sure when I can get
> to it.
> ...Bill

Other related posts: