Add functionality for volumetric multiscale coupling to micro-mechanics simulations - #135

I could be able to build the adapter by making these changes to the ConfigReader.cpp

if (config["participants"][participantName]["interfaces"][I]["elements-mesh”]) { interface.elementsMeshName = strdup(config["participants"][participantName]["interfaces"][i]["elements-mesh"].get_value<std::string>().c_str()); }

to

if (config["participants"][participantName]["interfaces"][i].contains("elements-mesh")) { interface.elementsMeshName = strdup(config["participants"][participantName]["interfaces"][i]["elements-mesh"].get_value<std::string>().c_str()); }

Hope this helps.

Thanks @MinhazPX for the suggestion. The build works now. I will look into the Segmentation fault problem you are facing.

While trying to get this to work together with @MinhazPX, I noticed that the data structure stx from nonlingeo_precice.c cannot be modified externally. Doing this corrupts the internal memory, and CalculiX is unable to handle the change when results() are called. One alternative is to modify the data sti instead. But this is problematic, as CalculiX internally uses sti to store the converged stress values. This means the only way to modify stresses from outside is to write a UMAT and provide it to results().

It seems the current way of handling stress won't work.

With reference to the previous comment, I am marking this pull request as draft because there is no clear solution for modifying stresses in CalculiX in sight.

there is no clear solution for modifying stresses in CalculiX in sight.

Sounds related to #111