[mesa-users] Removing mass from an irradiated planet

[Bill Paxton]

Hi Francesca,

Here's what I've found so far.

First thing to notice is that the terminal output says 'lgL' is limiting the timesteps.
__________________________________________________________________________________________________________________________________________________

       step    lg_Tcntr    Teff       lg_LH     lg_Lnuc     Mass       H_rich     H_cntr     N_cntr     Y_surf     X_avg     eta_cntr  zones retry
   lg_dt_yr    lg_Dcntr    lg_R       lg_L3a    lg_Lneu     lg_Mdot    He_core    He_cntr    O_cntr     Z_surf     Y_avg     gam_cntr  iters bckup
     age_yr    lg_Pcntr    lg_L       lg_LZ     lg_Psurf    lg_Dsurf   C_core     C_cntr     Ne_cntr    Z_cntr     Z_avg     v_div_cs     dt_limit
__________________________________________________________________________________________________________________________________________________

       1400   4.038200    983.548 -29.853591 -29.853591   0.000337   0.000337   0.710000   0.001008   0.270000   0.646773  23.268099    661      1
   5.306519   0.188806  -1.007574 -99.000000 -31.251076 -13.000000   0.000000   0.270000   0.009360   0.020000   0.245956  26.707130      2      0
 7.1834E+09  12.765914 -11.688580 -99.000000   7.000000  -3.609342   0.000000   0.003442   0.002099  2.000E-02  1.073E-01  0.571E-10           lgL




So let's look at the history of lgL.  Turns out it is dropping okay, but then goes unstable down around 10^-12 Lsun.

Next, copy star/test_suite/debugging_stuff_for_inlists to your inlist and turn on info about errors.

      report_ierr = .true. ! if true, produce terminal output when have some internal error

That gives lots of messages like this:

P_eqn_surf L <= 0          -1   -2.1304059896371673D+21
 eval_equ_for_solver: ierr after eval equations
 eval_equations: eval_equ returned ierr          -1
 newton failed for hydro
 hydro_newton_step returned ierr          -1

Go to star/private and grep for 'P_eqn_surf L' -- that says it is in hydro_eqns.f, so open that and find it.

It is in PT_eqns_surf -- the surface equations for P and T where the model must match the results of the atm module.
But before the routine even calls the atm, it does a sanity check on the surface luminosity.
It assumes that a negative surface L is bogus, so it raises an error -- and that in turns causes a drop in timestep.

Usually a negative surface L is a clear signal that the solver has gone off the rails.
But you have a very unusual situation here with an old (cold) planet getting strong irradiation.
So perhaps this something other than a "standard" convergence problem.
Do you think the model could legitimately develop a negative surface L because of the irradiation?
The poor mesa/star code is going to be very confused by this!  (e.g., what's Teff for negative L?)

At this point I'm stuck for the moment.  Perhaps someone knows how to deal with this sort of situation.
It may be that using a stellar evolution code to model planets runs into problems in cases like these.
Or maybe there is a simple fix that the planet modelers have figured out.   Anyone have ideas?

Cheers,
Bill










On Sep 4, 2014, at 1:49 PM, Francesca Valsecchi wrote:

> Hi Bill, 
> 
> 
> Thank you for the quick response!
> 
> Ok, I re-named the models model1.mod --> model5.mod for ease of reference:) 
> I am attaching the following files:
> 
> - model4.mod (created in step 4)
> - model5.mod (created in step 5 before MESA starts having troubles). Step 5 takes only a few minutes though, in case you want to try to re-run it from the beginning.
> 
> 
> 
> Thank you very much for the help with this!
> 
> Fra
> 
> 
> <model4.mod><model5.mod>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20140904/3f4991b1/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: log_L.pdf
Type: application/pdf
Size: 21763 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20140904/3f4991b1/attachment.pdf>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20140904/3f4991b1/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: log_L.pdf
Type: application/pdf
Size: 21670 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20140904/3f4991b1/attachment-0001.pdf>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20140904/3f4991b1/attachment-0002.html>