[CIG-LONG] Re: Cumulative strain history in Gale et alia

[Shijie Zhong]

Benchmark is obviously important for all the codes. I was curious what type of analytic solutions Christoph was referring to. Is it for the stokes flow driven by a buoyancy at a single harmonics? or is it for a sinking stokes sphere. For the stokes' flow with linear rheology (the only case where analytic solutions are possible), Citcom does a very job in benchmarks, and I suspect that GALE should be able to do the same.  

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: Wed, 10 Oct 2007 09:08:04 -0700 (PDT)
>From: Walter Landry <walter at geodynamics.org>  
>Subject: Re: [CIG-LONG] Re: Cumulative strain history in Gale et alia  
>To: cig-long at geodynamics.org
>
>Christoph Schrank <schrank at geology.utoronto.ca> wrote:
>> I also noted some pretty long run times in the Newtonian
>> free-falling slab benchmarks for reasonable resolutions. I did not
>> use a layer of air on top though because Gale provides a free
>> surface. The time steps had to be decreased significantly in order
>> to obtain reasonable velocity fields. However, the recent results
>> closely resemble those of FLAC and I2VIS. We are still tweaking the
>> Stokes flow models (in 2D) and comparing them to runs in Ellipsis
>> 2D. Those two match quite well, however, we are still deviating
>> considerably from the analytical solution (generally overestimating
>> the velocity). But I need more models to explore the computational
>> parameter space properly. I suggest it would be possible to share
>> some images/plots via this list?!
>
>Yes, please.  I would be very interested.  If Gale differs from an
>analytic solution, that would be a bug.
>
>Cheers,
>Walter Landry
>walter at geodynamics.org
>_______________________________________________
>CIG-LONG mailing list
>CIG-LONG at geodynamics.org
>http://geodynamics.org/cgi-bin/mailman/listinfo/cig-long