[CIG-LONG] Side moving in the wrong direction

[Walter Landry]

Cc'ing the list.

nicolas riel <a.nicolas.riel at gmail.com> wrote:
> Hello Walter,
> 
> I'm still working with the code and encountered some problems when running
> a model with simple rheologies and without thermal properties. This model
> (attached file) uses an inflow boundary on the left side with a wrapTop
> condition (with or without a stressBC simulating isostasy on the bottom).
> 
> The step 0 is going well but the step 1 crashes and gives this error :
> 
> "side is moving in the wrong direction".
> 
> Walter, do you think that an air layer with a staticTop instead of a
> wrapTop  would fix this problem ?
> 
> I would like to try it, however I don't know how to set the "maxAirY" with
> the "json" syntax. I think that could be a good idea to add it in the
> manual. This way one can fix the Top boundary with vy=0 and stabilizes the
> model.

I would not recommend using an air layer in general.  An air layer is
only useful if the height gets very small.  For example, if you are
simulating a wedge.  If you do want to use an air layer, you do not
have to do anything special.  "maxAirY" is no longer needed because
HydrostaticTerm is no longer needed.

As for your current input file, you just need to fix the top left corner

  "staticLeftTop": "True"

> Moreover when I try to add the thermal properties for example in the
> Geomod2004 benshmark, the thermal field at step 0 is always rough with
> irregular boundaries.

What do you mean by "rough"?  The bottom is set to 873 while the bulk
is set to 573.  That is what I see.  What do you see?

Cheers,
Walter Landry