[CIG-MC] Internal heating with CitcomS

[Eh Tan]

Hi Brad,

There might be three reasons for your observation.

First, as Shijie said, the Rayleigh number in CitcomS is scaled by the 
radius of the earth, not by the thickness of the mantle. As a result, a 
Cartesina model with the same Rayleigh number as a spherical model with 
convect ~10x more vigorously (0.45^-3).

Second, if you turn on VMIN and VMAx, the viscosity field is cliped 
between visc_min and visc_max. Your viscosity might already reach 
visc_min, so further temperature increase won't decrease your viscosity.

Third, the most likely reason why your temperature never goes above 1, 
CitcomS has a Lenardic-type filter on the temperature field. The filter 
will clip the temperature above 1 and below 0, while keeping the total 
energy conserved. The purpose of the filter is to suppress numerical 
artifact from the advection term, but the filter is not physics based. 
The filter is only documemted in the manual appendix and is probably one 
of the few "gotchas" in the code. For the upcoming CitcomS release, the 
filter is turned off by default. For CitcomS-2.2 or earlier, you can 
turn the filter off by adding these lines in your input cfg file:

[CitcomS.solver.tsolver]
filter_temp = off


Hope this help.

----
Eh Tan
Staff Scientist
Computational Infrastructure for Geodynamics
2750 E. Washington Blvd. Suite 210
Pasadena, CA 91107
(626) 395-1693
http://www.geodynamics.org


Shijie Zhong wrote:

>Hi Brad,
>
>Can you provide more details? Even if the bottom is insulated, for a given Q, internal temperature will eventually reach a constant where surface heat flux balances total internal heating. 
>
>Did you use regional spherical option for CitcomCU? Cartesian and spherical formulations in CitcomCU scale Q differently.  
>
>Shijie
>
>
>Shijie Zhong
>Department of Physics
>University of Colorado at Boulder
>Boulder, CO 80309
>Tel: 303-735-5095; Fax: 303-492-7935
>Web: http://anquetil.colorado.edu/szhong
>
>
>---- Original message ----
>  
>
>>Date: Fri, 05 Oct 2007 15:23:24 -0700
>>From: Bradford Foley <bfoley at usc.edu>  
>>Subject: [CIG-MC] Internal heating with CitcomS  
>>To: cig-mc at geodynamics.org
>>
>>I'm using CitcomS-2.2.2 and am having a problem with internal heating.  When I using an insulating bottom boundary and internal heating, the mantle will only heat up to a point and then stay at the same temperature.  This is most obvious when I'm using temperature dependent viscosity, as the system doesn't start convecting, and I can see the mantle stop heating up.  When I use Q=10, the mantle reaches a max temperature of about 0.75 and then stays there.  I tried using a higher Q, Q=30, but here the mantle heats up to a temperature of 1, with a small boundary layer to 0 at the top, and stays in this configuration.  These results are different than what I got using CicomCU, where the mantle would heat up beyond 1 and eventually start convecting, while here it appears to stop heating up and doesn't reach a point where convection starts. 
>>
>>-Brad Foley 
>>_______________________________________________
>>CIG-MC mailing list
>>CIG-MC at geodynamics.org
>>http://geodynamics.org/cgi-bin/mailman/listinfo/cig-mc
>>    
>>