Adding New Composition Fields to a Restart Computation

Dandoug · July 10, 2019, 7:25pm

Hello,

I’m using ASPECT’s viscoplastic material model and I would like to run a simulation for a length of time with only one composition field. After reaching a certain point, I would like to restart the simulation but have an additional composition field introduced. I haven’t found a way to do this and I’m not sure if you can, but I thought I would ask! I appreciate the help.

Regards,
Daniel

tjhei · July 11, 2019, 1:21am

We haven’t implemented that yet. In fact, we disallow it explicitly:

github.com

geodynamics/aspect/blob/master/source/simulator/checkpoint_restart.cc#L223


ia >> pressure_normalization;
AssertThrow (pressure_normalization == parameters.pressure_normalization,
             ExcMessage ("The pressure normalization method that was stored "
                         "in the checkpoint file is not the same as the one "
                         "you currently set in your input file. "
                         "These need to be the same during restarting "
                         "from a checkpoint."));


unsigned int n_compositional_fields;
ia >> n_compositional_fields;
AssertThrow (n_compositional_fields == parameters.n_compositional_fields,
             ExcMessage ("The number of compositional fields that was stored "
                         "in the checkpoint file is not the same as the one "
                         "you currently set in your input file. "
                         "These need to be the same during restarting "
                         "from a checkpoint."));


std::vector<std::string> names_of_compositional_fields;
ia >> names_of_compositional_fields;
AssertThrow (names_of_compositional_fields == parameters.names_of_compositional_fields,
             ExcMessage ("The names of compositional fields that were stored "

This wouldn’t be impossible to address, but one would need to figure out how to initialize the new field (provide an “initial condition”). Of the top of my head I also don’t know a good way to change the FE_System (the deal.II object that determines the Finite Element we use).
I think it would be much easier to have the compositional field from the start of the computation (and maybe just ignore its values).

jbnaliboff · July 12, 2019, 10:43pm

Hi Daniel,

I agree with Timo that the most straightforward approach is to include all compositional fields from the beginning, but alter the parameter file and/or source code so that they are not always “used”.

Can you provide a bit more information about the specific functionality/process you are interested in? Happy to help with the implementation!

Cheers,
John

Topic		Replies	Views
Restarting form a checkpoint, adding a compositional field after ASPECT	4	160	December 13, 2023
Resetting elastic stresses/compositional field values ASPECT	2	268	April 7, 2023
Function for initial composition model not working in ASPECT release mode ASPECT	4	251	March 1, 2022
Problem with restarting 3D model ASPECT	11	410	September 2, 2022
Viscoplastic material model with compositional fields not switching ASPECT material-model	2	760	January 18, 2021

Adding New Composition Fields to a Restart Computation

Related topics