[mesa-users] Bug: variable order in other_mlt

[Warrick Ball]

Hi,

I've been continuing occasional work on implementing Canuto, Goldman & 
Mazzitelli's (1996) theory of convection.  While hunting for a bug of my 
own, I noticed a slip in the template for other_mlt.  This could only 
possibly have affected your results if you added your own MLT routine by 
copying null_other_mlt.

First, here's the template for other_mlt given on lines 92--102 of
$MESA_DIR/star/other/other_mlt.f90 (leading whitespace removed):

subroutine null_other_mlt(  &
     id, k, cgrav, m, r, T, rho, L, P, &
     chiRho, chiT, Cp, Cv, csound, X, opacity, grada,  &
     gradr_factor, gradL_composition_term, &
     alpha_semiconvection, semiconvection_option, &
     thermohaline_coeff, thermohaline_option, &
     dominant_iso_for_thermohaline, &
     mixing_length_alpha, alt_scale_height, remove_small_D_limit, &
     MLT_option, Henyey_y_param, Henyey_nu_param, &
     prev_conv_vel, max_conv_vel, dt, g_theta, tau, MLT_dbg, &
     mixing_type, mlt_basics, mlt_partials1, ierr)

Here's what the actual function call looks like on lines 936--946 of 
$MESA_DIR/star/mlt/private/mlt_info.f90 (leading whitespace removed):

call s% other_mlt(  &
     s% id, k, cgrav, m, r, T, rho, L, P, &
     chiRho, chiT, Cp, Cv, csound, X, opacity, grada, &
     gradr_factor, gradL_composition_term, &
     alpha_semiconvection, semiconvection_option, &
     thermohaline_coeff, thermohaline_option, &
     dominant_iso_for_thermohaline, &
     mixing_length_alpha, alt_scale_height,
     remove_small_D_limit, &
     MLT_option, Henyey_y_param, Henyey_nu_param, &
     prev_conv_vel, max_conv_vel, g_theta, dt, tau, MLT_dbg, &
     mixing_type, mlt_basics, mlt_partials1, ierr)

In particular, here are the second last lines of each:

     prev_conv_vel, max_conv_vel, dt, g_theta, tau, MLT_dbg, &
     prev_conv_vel, max_conv_vel, g_theta, dt, tau, MLT_dbg, &

The order of dt and g_theta appear to be backwards in null_other_mlt.  I 
found the bug in r7624 but I just checked out r8118 and the bug is also 
present (but the other_mlt call is now on line 1085).

It's entirely possible that this doesn't actually affect anyone, but I 
figured I should report it anyway.

Cheers,
Warrick





------------
Warrick Ball
Postdoc, Institut für Astrophysik Göttingen
wball at astro.physik.uni-goettingen.de
+49 (0) 551 39 5069