Unclassified Adiabatic Quadrapole Modes

[ehsan]

Hi Rich,
I tried to calculate the adiabatic frequencies of l=0 to l=2 for a 10 Msun MESA model close to ZAMS.
I find several modes that GYRE fails to assign any mode order at all, so all n_p, n_g and n_pg are set to zero!
Here is an example:

Code: Select all

Root Solving
       l    n_pg     n_p     n_g                 Re(omega)                 Im(omega)                       chi  n_iter        n
       2       0       0       0    0.5029387793852281E-01    0.0000000000000000E+00    0.8411050160727212E-04       6     1419
       2       0       0       0    0.5109837460093900E-01    0.0000000000000000E+00    0.1143441520165077E-03       5     1419
       2       0       0       0    0.5192911250752998E-01    0.0000000000000000E+00    0.6735667924063605E-04       5     1419
       2       0       0       0    0.5278724685448197E-01    0.0000000000000000E+00    0.5658706752306604E-04       5     1419
       2       0       0       0    0.5367395997163085E-01    0.0000000000000000E+00    0.8686201106068521E-04       5     1419
       2       0       0       0    0.5459034983928796E-01    0.0000000000000000E+00    0.6167403809574609E-04       6     1419
       2       0       0       0    0.5553735763155947E-01    0.0000000000000000E+00    0.5916306390782498E-04       5     1419
       2       0       0       0    0.5651617873050672E-01    0.0000000000000000E+00    0.1042561029776355E-03       6     1419
       2       0       0       0    0.5752926444438590E-01    0.0000000000000000E+00    0.6807525912155008E-04       5     1419
       2       0       0       0    0.5857816736447558E-01    0.0000000000000000E+00    0.1129203850357106E-03       6     1419
       2       0       0       0    0.5966355136266139E-01    0.0000000000000000E+00    0.1236958965239650E-03       6     1419
       2       0       0       0    0.6078513464517546E-01    0.0000000000000000E+00    0.1955926533308987E-03       6     1419
       2       0       0       0    0.6193818608959987E-01    0.0000000000000000E+00    0.2416056524385476E-03       5     1419
       2       0       0       0    0.6310761907929491E-01    0.0000000000000000E+00    0.4973061910505123E-03       5     1419
       2       0       0       0    0.6422801163579051E-01    0.0000000000000000E+00    0.1451862165064680E-02       6     1419
       2       0       0       0    0.6515121924034814E-01    0.0000000000000000E+00    0.2419944995245070E-02       6     1419
       2       0       0       0    0.6613223728394522E-01    0.0000000000000000E+00    0.1306954755020589E-02       6     1419
       2       0       0       0    0.6738515367971330E-01    0.0000000000000000E+00    0.4523011821164654E-03       6     1419
       2       0       0       0    0.6877615717577129E-01    0.0000000000000000E+00    0.1899283448800337E-03       6     1419
       2       0       0       0    0.7025292777979089E-01    0.0000000000000000E+00    0.7699034560920360E-04       6     1419
       2       0       0       0    0.7180528290910176E-01    0.0000000000000000E+00    0.3362248381389475E-04       6     1419
       2       0       0       0    0.7343327566735494E-01    0.0000000000000000E+00    0.2696347232159763E-04       6     1419
       2     -14       0      14    0.7514082861878150E-01    0.0000000000000000E+00    0.8922371565661055E-05       5     1419
       2     -41       0      41    0.7693149334011459E-01    0.0000000000000000E+00    0.6267735862435640E-05       6     1419
       2     -40       0      40    0.7881115975689040E-01    0.0000000000000000E+00    0.4924978000411270E-17       6     1419

As you notice, there is also an outlier mode with n_pg=-14
I have attached the input model and my GYRE inlist to reproduce this result.
I use GYRE 2.3 development version. Let me know if I can provide extra information

Code: Select all

gyre_ad [2.3]
=============

Compiler         : GCC version 4.9.0 20130408 (experimental)
Compiler options : -I /Applications/mesasdk/include -march=corei7-avx -mcx16 -msahf -mno-movbe -maes -mpclmul -mpopcnt -mno-abm -mno-lwp -mno-fma -mno-fma4 -mno-xop -mno-bmi -mno-tbm -mno-avx2 -msse4.2 -msse4.1 -mno-lzcnt -mno-rtm -mno-hle -mno-rdrnd -mno-f16c -mno-fsgsbase -mno-rdseed -mno-prfchw -mno-adx -mfxsr -mxsave -mxsaveopt --param l1-cache-size=32 --param l1-cache-line-size=64 --param l2-cache-size=6144 -mtune=corei7-avx -fPIC -mmacosx-version-min=10.7.4 -mno-avx -mno-bmi2 -O2 -std=f2008 -finit-real=snan -ffpe-trap=invalid,overflow,zero -fbacktrace -fopenmp
OpenMP Threads   : 3
MPI Processors   : 1

Thanks for the support.
Regards
Ehsan.

[rhtownsend]

Hi Ehsan --

This behavior arises because the low-frequency l=2 modes in your example are below the atmospheric gravity-wave cutoff frequency. As a consequence, these modes have formally-complex eigenfrequencies; but gyre_ad only looks for real eigenfrequencies, and therefore isn't able to properly converge on a solution.

I've updated the development version of the code to print out a warning when such cases are encountered.

cheers,

Rich