Benchmark : drying out of a layer

Before trying to imrpove our application to make it works for different cases, we look at another benchmark, which has an analytical solution. The description and the results can be found on this page.

In this benchmark, the temperature is constant, and therefore at each step we just have to solve the moisture equation. To prevent the application from attempting to solve the heat equation, we implemented an option hm.heat that we set to false at the execution (the option hm.moisture does the same thing for the moisture equation, but is not needed for this case).

The command to run the simulation is:

mpirun -np 6 ./feelpp_hm_heat_moisture --config-file cases/drying-layer/drying-layer.cfg --ts.time-step=3600 --ts.time-final=3600000 --hm.export=all --hm.heat=false --hm.picard=true

Less parameters are involved in this application comparing to the first benchmark and those which are common to both the two cases have different values. The values of the coefficients are different of the first benchmark, and there are less of them. To make the dinstinction in the code, we added macros around the lines about the previous case, which is not the best way to proceed. The next step of the internship is to improve the code, in order to allow the model parameters using a JSON file.