Superadiabatic reduction (linear) timescale limiter

[matthiasfabry]

Very similar to #1006, with the same motivation, only the activation function of the limiter is linear, rather than an exponential logistic-like supression.

[matthiasfabry]

the 40M_carbon_burning_envelope_issues developer test case gets to core collapse with superad_reduction and superad_reduction_use_turnover_limit = .true., superad_reduction_turnover_vc_floor_frac = 1d-3. Final terminal output:

stop because fe_core_infall > fe_core_infall_limit
    0.1000155301E+08    0.1000000000E+08

       1928   9.859765   8245.831   5.520181 -99.000000  22.759131   1.609574   0.000000   0.000000   0.559157   5.028935   2186     14
-1.0699E+01   9.859765   2.729020  17.744054  14.166626 -99.000000  21.149558   0.014715   0.000000   0.019628   4.105381      2
 5.1916E+06   8.983535   6.077671 -99.000000  26.042204 -11.221828  18.117593   0.000000   0.000000   0.000000  0.533E+00  max increase
                                 Fe_core        1.7881089532589030
                          fe_core_infall      100.0155300787535992
                               rel_E_err    5.9575515939319427D-12
                       log_rel_run_E_err       -7.4666874178951153

save LOGS/profile25.data for model 1928
                                         solver retries          14

                                    varcontrol dt limit           3
                                  max increase dt limit         799
                                    burn steps dt limit           6
                                          dX/X dt limit           2
                                       Lnuc_He dt limit           6
                                        lgTeff dt limit          26
                                      lgT_cntr dt limit         704
                                      XHe_cntr dt limit          64
                                      XNe_cntr dt limit           8
                                       XO_cntr dt limit          19
                                      XSi_cntr dt limit         245
                                        dX_nuc dt limit          46

 model saved to final.mod
save photos/x00001928 for model 1928
termination code: fe_core_infall_limit

all values are within tolerance
 png/conv_001928.png/png
 png/Grid1_001928.png/png
__________________________________________________________________________________________________________________________________________________

       step    lg_Tmax     Teff     lg_LH      lg_Lnuc_tot Mass       H_rich     H_cntr     N_cntr     Y_surf   eta_cntr   zones  retry
  lg_dt_yrs    lg_Tcntr    lg_R     lg_L3a     lg_Lneu     lg_Mdot    He_core    He_cntr    O_cntr     Z_surf   gam_cntr   iters  
    age_yrs    lg_Dcntr    lg_L     lg_LZ      lg_Lphoto   lg_Dsurf   CO_core    C_cntr     Ne_cntr    Si_cntr  v_div_cs       dt_limit
__________________________________________________________________________________________________________________________________________________

       1928   9.859765   8245.831   5.520181 -99.000000  22.759131   1.609574   0.000000   0.000000   0.559157   5.028935   2186  14
-1.0699E+01   9.859765   2.729020  17.744054  14.166626 -99.000000  21.149558   0.014715   0.000000   0.019628   4.105381      2
 5.1916E+06   8.983535   6.077671 -99.000000  26.042204 -11.221828  18.117593   0.000000   0.000000   0.000000  0.533E+00  max increase
                                 Fe_core        1.7881089532589030
                          fe_core_infall      100.0155300787535992
                               rel_E_err    5.9575515939319427D-12
                       log_rel_run_E_err       -7.4666874178951153

[matthiasfabry]

Doh! I'll remove these asap.

[Debraheem]

Since this branch and Matteo's branch are nearly identical, and Matteo has developed his branch even further toward a production ready form: If there are no objections I would like to merge this branch into Matteos (see #1006), and we can make the linear or logistic form of the limiter an option.