Microlab 1: Convergence study

Copy the test suite $MESA_DIR/star/test_suite/1.5M_with_diffusion somewhere outside $MESA_DIR.

In the &star_job section of inlist_1.5M_with_diffusion, add

write_profile_when_terminate = .true.

filename_for_profile_when_terminate = 'LOGS/mdcX_tdcY_nomaxdt/profile_mdcX_tdcY_nomaxdt_Xc010.data'

where you replace X and Y with the actual values that you are using.

In the &controls section of inlist_1.5M_with_diffusion, add

star_history_name = 'history_mdcX_tdcY_nomaxdt.data'
log_directory = 'LOGS/mdcX_tdcY_nomaxdt'
time_delta_coeff = Y

If you study the effect of changing time_delta_coeff, then also add in the same &controls section

set_min_D_mix = .true.
min_D_mix = 1d2 ! minimal chemical diffusion coefficient in cm^2/s

For both cases, change

xa_central_lower_limit_species(1) = 'h1'

xa_central_lower_limit(1) = 0.1

D_mix_ignore_diffusion = 1d10 ! Don't do diffusion where Dmix > 10^10 cm^/s.
mesh_delta_coeff = X
max_years_for_timestep = 0    ! no maximum
max_model_number = -1         ! no maximum

Each person at a table select a different value for mesh_delta_coeff between 0.2 and 2.0, or a value for time_delta_coeff between 0.05 and 2.0. Set the other one equal to 0.5.

In the history_columns.list, add

surface mg24

Compile and run the model. Afterwards, plot (with MESA Explorer) the following quantities. (The files are in the LOGS directory.)

For the effect of mesh_delta_coeff:

radius vs. brunt_N2 (profile) and 2. center_h1 vs. mass_conv_core (history)

For the effect of time_delta_coeff:

log_Teff vs. log_L (history) and 2. center_h1 vs. surface_mg24 (history)

Compare the results with those of the others at your table. How are these quantities affected? What resolution would be sufficient?