[CIG-LONG] Conservation of Mass

[Louis Moresi]

For your information cig-long readers, Walter, Louise and I had a
conversation around this issue two weeks ago.

I don't think there are any supporters of Q1Q1 for this particular problem
left any more. Underworld defaults to Q1P0 but provides recovered deviatoric
stress and pressure quantities which mitigate the checkerboard problem. This
can be configured quickly and is at least well-understood and tested.

Q2P1 would be useful to have next. It is implemented in underworld (and
therefore gale) but there are some shortcomings / bugs left to iron out with
the current integration scheme for PIC and (my guess is that) there are some
outstanding issues with multigrid and some of the processing plugins. I am
not sure about the implementation of the stress / pressure recovery for
these elements although they are likely to be less important.

Walter was very keen to get this particular element working in gale before
he steps back from active development.

Louis


On Sat, Sep 24, 2011 at 20:16, laetitia le pourhiet <
laetitia.le_pourhiet at upmc.fr> wrote:

>  Hi Brian and the others,
>
> There is not much you can do about that I gave up trying to do such models
> with Gale.
> It is now established Q1Q1 elements are not suited for this kind of
> problems.
> What you can do is  to
> 1) Remove the free surface/sky to work with delta rho instead of rho
> 2) Increase the resolution.
>
> Basically you should only use these elements for extended thin shit
> problems.
>
> The old Q1P0 had other problems, but they could be used to model
> drip/subduction problems.
>
> If CIG still wants to spend some time/money on Gale
> It would be usefull to have a switch between elements, like in UW so that
> depending on the model set up we can choose the best suited element.
>
> hope this was helpfull
>
> Laetitia
>
>
> On 9/24/11 12:44 AM, Brian Wilson wrote:
>
> Hi all,
>
> I'm working on a model of lithospheric drip and despite having zero flux on
> the boundaries, I am getting downward velocities with no compensating
> upwelling, which would seem to violate the divergence free condition. Has
> anybody else had a similar problem and/or figured out a solution? I've
> attached a copy of my input file if anyone wants to look at it; everything
> is scaled as follows: distance: km, mass: 10^15 kg, time: m.y.
>
> Thanks,
> Brian
>
>
> _______________________________________________
> CIG-LONG mailing listCIG-LONG at geodynamics.orghttp://geodynamics.org/cgi-bin/mailman/listinfo/cig-long
>
>
>
> --
> *Dr Laetitia Le Pourhiet*<\n> MCF - Institut des Sciences de la Terre
> Paris - UPMC<\n> tel : +33 1 44 27 58 83<\n> web :
> http://perso.numericable.com/laetitia.le.pourhiet/index.html<\n>
>
> _______________________________________________
> CIG-LONG mailing list
> CIG-LONG at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-long
>
>


-- 
------------------------------
<http://picasaweb.google.com/lh/photo/gQ7_To7dX8sI0sEe2ty0AaO6j9vpN3ZPghWjdy2iDOg?feat=embedwebsite>Professor
LouisMoresi
louis.moresi at monash.edu
GoogleProfile <http://www.google.com/profiles/louis.moresi> | Busy or
not ?<http://www.moresi.info/pages/lm-calendar>
Underworld Geodynamics Software <http://www.underworldproject.org/>
Mobile phone: +61 4 2850 1907 ☎
------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geodynamics.org/pipermail/cig-long/attachments/20110924/ecf7b807/attachment.htm