[aspect-devel] Aspect melt_fraction

Mon Mar 17 10:04:41 PDT 2014

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Good idea. I've been imagining something similar, but more kludgey.
Right now I'm trying to use one of the compositional fields as the
melt fraction solution variable, and using this material model derived
melt fraction as a sort of melting rate. Then I'm applying an
additional solver to it after each stokes step. I have the impression
I might hit a dead end before I get this working though, and I'd
definitely prefer to just do it "right" the first time. I'm aiming to
implement the solution from Spiegelman, 1993 for flow through a
deformable porous medium.

I'll check out your branch today for ideas. Starting next month this
is going to be a higher priority for me so hopefully I can at least be
on the right track by then and avoid duplicating your effort.

I agree that if this will end up as a feature in the main branch we
should decide together how to proceed.

Cheers,
Jonathan

On 03/17/2014 04:09 AM, Juliane Dannberg wrote:
> Dear all,
> 
> I agree that it is a great idea to track the melt fraction more
> easily. On the way to implementing active melt migration (i.e.
> porous flow), my postprocessor was only a preliminary way to
> calculate the melt fraction. However, we should think about how we
> want this to be implemented in the end. I imagined that the melt
> fraction (or porosity) will be handled in a way comparable to the
> temperature or composition (as a solution variable), and in the
> material model only the melting rate/ rate of solidification will
> be calculated. If someone needs the melt fraction only as output
> quantity, the postprocessor will still be there, but if active melt
> transport is switched on, the melt fraction output is generated by
> default, just like for the temperature/pressure/...
> 
> I already started implementing something like this in my own
> Aspect branch (Jonathan, if you want to have a look: 
> http://aspect.dealii.org/websvn/listing.php?repname=Aspect+Repository;
>
> 
- --> branches --> j-dannberg). Until now, melt fraction is only handled
> as any other compositional field for purposes of advection, but
> there is the material model 'latent heat', where latent heat of
> melting is already generated and material properties can depend on
> the melt fraction.
> 
> So if the Aspect community thinks we should add such a feature to
> the main branch soon, maybe we should figure out together how we
> want to implement this.
> 
> Best, Juliane
> 
> 
> On 03/16/2014 10:39 PM, Louise Kellogg wrote:
>> I agree, this seems like a useful addition to the main branch.
>> 
>> On Mar 14, 2014, at 1:00 PM, Magali Billen <mibillen at ucdavis.edu 
>> <mailto:mibillen at ucdavis.edu>> wrote:
>> 
>>> I think it would be great to have a module for calculating
>>> melt fraction in the main branch. This doesn't come up for
>>> every problem, but there's often a problem that you're working
>>>  on and then part way through you realize that melt fraction
>>> might be helpful. If we are looking ahead to more realistic
>>> theologic models, for example, being able to have rheology
>>> depend on melt fraction would require calculating melt fraction
>>> first. Magali
>>> 
>>> On Mar 14, 2014, at 12:04 PM, Jonathan Perry-Houts wrote:
>>> 
> Hi all,
> 
> I'm starting work on melt migration in ASPECT, and the first step
> was to make melt fraction a more easily tracked quantity.
> 
> In December Juliane made a nice postprocessor to visualize the
> amount of melt generated in the model based on Katz, 2003 paper.
> The algorithm is great but I don't like the fact that it's
> hard-coded in to the postprocessor. I think it would be more useful
> to make melt fraction a quantity calculated by the material model
> and have the postprocessor display it in the same way viscosity and
> density are displayed.
> 
> This also allows material properties to be easily derived from the 
> melt fraction without calculating the same thing twice for each 
> timestep. If anyone has a better suggestion for dealing with melt
> I'd be happy to hear it.
> 
> I'm not sure that other people would be interested in these
> changes enough to put them in the main branch, but I figured I'd
> share them with the mailing list in case anyone does find them
> useful. The changes are relatively unobtrusive, except for breaking
> Juliane's melt_fraction visualization postprocessor, but that
> functionality can be easily replicated by including the same
> algorithm in the material model.
> 
> I can also send out an example material model which calculates
> melt based on Juliane's algorithm if people are interested.
> 
> The modified files are: include/aspect/material_model/interface.h 
> include/aspect/postprocess/visualization/melt_fraction.h 
> source/material_model/interface.cc <http://interface.cc> 
> source/postprocess/visualization/melt_fraction.cc
>>>> <interface.h><melt_fraction.h><interface.cc 
>>>> <http://interface.cc>><melt_fraction.cc>_______________________________________________
>>>>
>>>> 
Aspect-devel mailing list
>>>> Aspect-devel at geodynamics.org
>>>> <mailto:Aspect-devel at geodynamics.org> 
>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>>>
>>>
>>>> 
- --------------------------------------------------
>>> Professor of Geophysics & UCD Chancellor Fellow Chair, Geology
>>> Graduate Program Earth & Planetary Sciences Dept., UC Davis 
>>> Davis, CA 95616 2129 Earth & Physical Sciences Bldg. Office
>>> Phone: (530) 752-4169 http://mygeologypage.ucdavis.edu/billen/ 
>>> --------------------------------------------------
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> _______________________________________________ Aspect-devel
>>> mailing list Aspect-devel at geodynamics.org
>>> <mailto:Aspect-devel at geodynamics.org> 
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>>
>>
>>> 
********************************************
>> Louise Kellogg Professor, Dept. of Earth & Planetary Sciences One
>> Shields Avenue, University of California, Davis, CA 95616 Phone:
>> (530)752-3690 ********************************************
>> 
>> 
>> 
>> 
>> 
>> 
>> _______________________________________________ Aspect-devel
>> mailing list Aspect-devel at geodynamics.org 
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>
>> 
> 
> 
> _______________________________________________ Aspect-devel
> mailing list Aspect-devel at geodynamics.org 
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
> 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBAgAGBQJTJysmAAoJEGe6xJ1FYRpR3xgH/2IiEGv6gKHfUx4sNaKsgVfS
tYW0ESahT90UFsHGcSP0adh89MVysOBdc0wxoWMSxr+aA99FpQ8gRklWk18aPMs7
RwRL71JnLhv4e+TnpAx+oxaJNfAmD0JgAjR/uhMrL9S+gGWye6wFkxMELt+zfSiX
RB4n7ePQXmLsBnzFzLXFpv4dhJhmOsDZrY1G+dZy/VjFh6I69OxrOvfQWL60d3oq
hTadeGdrYdvLzxGK5m8IJJybVU5J6CGcJztBNHitCOHS+77PCju6vkgz8kJ4C7QK
JPDgQst9Bybs8KT2OQFSi9JW39XyAJqspe4B4D6VaK+5WacSzAgJyDA6tg/gvDM=
=GBcC
-----END PGP SIGNATURE-----