[Mesa-users] Nonphysical behavior of outermost envelope of rapidly rotating stars

Larry Molnar lmolnar at calvin.edu
Mon Jul 27 16:14:36 EDT 2020


Hi Pablo,

I have run some more cases for M = 1 solar mass to show more precisely the strange behavior.

  1.  I am now saving out every profile so the plots are not choppy.
  2.  I used smaller steps in Omega/Omega_crit initial to show the suddenness of the onset of the behavior.
  3.  I redid one case using the wind controls you suggested.

I attach the inlist used for the final case. M10RotExampleA.txt contains the controls used to make the nonrotating ZAMS model from which the other models start. M10RotExampleB.txt contains the controls for the main sequence evolution for a specific initial value of Omega/Omega_crit.

I attach three plots to show the behavior.

  1.  HighRes_HR.png: shows initial rotation ratios of 0.39, 0.40, and 0.41 differ from each other marginally while 0.42 makes a large excursion. The case labeled 0.421 is actually 0.420 again but with the wind controls turned on. These do not seem to have changed the output at all.
  2.  HighRes_L-age.png: shows the luminosity with age is very similar in all cases, a sign that only the outermost layer participates in the excursion.
  3.  HighRes_OmegaOmegaC-age.png: shows the rotation ratios at the photosphere with age. The last case not to make the excursion is OmRat_intial = 0.41. It has a maximum value of OmRat = 0.678, which is nowhere near the breakup value.

In summary, this nonlinear behavior during main sequence hydrogen burning for an intermediate rotation rate seems to me very unexpected, although I can't say what exactly is going wrong. It is far enough from breakup rotation that details of the wind do not seem relevant (which is consistent with the lack of change when those parameters are turned on). This also seems to contradict the results given by earlier versions of MESA (as seen, for example, in the HR tracks of Gossage et al. (2018, ApJ, 863, 67). Their Figure 1 (lower left corner) shows cases for OmRat_inital = 0.50 with normal behavior.

Any additional thoughts would be appreciated.

Larry



From: Larry Molnar <lmolnar at calvin.edu>
Sent: Saturday, July 25, 2020 2:56 PM
To: Pablo Marchant <pamarca at gmail.com>
Cc: mesa-users at lists.mesastar.org
Subject: Re: [Mesa-users] Nonphysical behavior of outermost envelope of rapidly rotating stars

Hi Pablo,

I did include the inlists. (They were renamed as .txt files for convenience on a Windows machine.)

I was aware of the default w_div_wcrit_max = 0.9d0, but the problem set in at the smaller value of 0.7, so it was not clear that this flag should be active.

Not every step is being output.

Larry

________________________________
From: Pablo Marchant <pamarca at gmail.com>
Sent: Saturday, July 25, 2020 3:02 AM
To: Larry Molnar <lmolnar at calvin.edu>
Cc: mesa-users at lists.mesastar.org <mesa-users at lists.mesastar.org>
Subject: Re: [Mesa-users] Nonphysical behavior of outermost envelope of rapidly rotating stars

Hi Larry,

it's usually best if you include your inlists as well. Note that although the fits are good up to Omega_Psi/Omega_cri~0.999, there's still a cap on the rotational corrections that by default is set lower than this, with the option
w_div_wcrit_max = 0.9d0
essentially all values that are a function of w are capped to this. This is just an inherent necessity from the explicit calculation of these corrections.

Now, given this cap a model can go past Omega_Psi/Omega_cri=1 because the centrifugal forces are capped in this way. The workaround we have for this is an implicit method to boost wind strength when approaching critical, that boosts it just enough so that the star remains at the edge. The test case high_rot_darkening does this, and you can turn it on with the set of options

  ! implicit wind scheme when critical rotation is reached
    max_mdot_redo_cnt = 200
    min_years_dt_for_redo_mdot = 1d-5
    surf_w_div_w_crit_limit = 0.99d0
    surf_w_div_w_crit_tol = 0.01d0
    rotational_mdot_boost_fac = 1d10
    rotational_mdot_kh_fac = 1d10
    mdot_revise_factor = 1.1
    implicit_mdot_boost = 0.1

in your controls. This will boost the wind in such a way the star remains at omega/omega_c<0.99.

And just a last comment, I don't know if you're outputting every step in your simulations, but your tracks look a bit choppy before the craziness ensues. This could be a sign of poor timestep resolution.

Cheers

On Fri, Jul 24, 2020 at 10:34 PM Larry Molnar <lmolnar at calvin.edu<mailto:lmolnar at calvin.edu>> wrote:

I am running the latest MESA based on the revisions described in Paxton et al. (2019): MESA R12778, SDK 20.3.2, on a Linux operating system.



I am interested in rotation of main sequence stars with intermediate mass. With the improved estimates of geometrical factors, the ratio OmRat = Omega_Psi/Omega_crit should be accurate up to 0.999. However, when we ran simple test cases with rotation turned on near ZAMS, we found spurious excursions in the H-R diagrams for OmRat_initial >= 0.4. I attach a pair of inlist files we used to get to ZAMS and then to turn on rotation from that point. We ran cases with initial mass of 1.0 and 2.0 Msun and with a range of OmRat_initial up to 0.6. I attach HR diagrams for each mass showing a set of curves for different initial OmRat. I also attach plots of luminosity vs age and OmRat of the photosphere vs. age for each mass.



For the initial mass = 2 solar masses case, the excursion in the H-R diagram occurs for initial values of OmRat > 0.4. The excursions begin at the end of hydrogen burning in the convective core, a time when the OmRat should jump up slightly. However, in the misbehaving cases OmRat jumps above unity to nonphysical values. The stellar luminosities remain normal with time, though. So it seems the only part of the stellar structure that is wacky is the outermost envelope. Once the star begins to go up the giant branch, everything is normal again.



For the initial mass = 1 solar masses case, a similar set of excursions occurs although the trigger is slightly different. There is no convective core to stop burning. The trigger instead is when the OmRat reaches ~0.7. As before, everything returns to normal again once the star begins to go up the giant branch. I include a table of values below for the case M_init = 1.0 and OmRat_init = 0.5. The errant behavior is in profile numbers 7-26.



Any ideas on what is going on in the code (or if there is something different I should do in the control file) are welcome.



Larry Molnar



Profile

age

R1

Temp.

Lum.

OmRat

#

y

Rsun

K

Lsun

0

3.98E+07

0.9203

5239

5.73E-01

0.54

1

4.47E+07

0.9339

5250

5.95E-01

0.5476

2

8.26E+07

0.9425

5252

6.07E-01

0.5534

3

3.17E+08

0.9561

5260

6.29E-01

0.5674

4

1.55E+09

0.987

5268

6.74E-01

0.5929

5

3.14E+09

1.0363

5257

7.37E-01

0.6347

6

4.60E+09

1.1119

5188

8.05E-01

0.7127

7

5.13E+09

1.5238

4457

8.23E-01

1.1583

8

5.28E+09

1.5776

4394

8.34E-01

1.2469

9

5.84E+09

1.6156

4386

8.68E-01

1.2807

10

6.12E+09

1.6364

4381

8.87E-01

1.2986

11

7.48E+09

1.7523

4350

9.88E-01

1.3919

12

8.72E+09

1.8746

4320

1.10E+00

1.4736

13

9.78E+09

1.9916

4293

1.21E+00

1.5268

14

1.03E+10

2.0571

4281

1.28E+00

1.5483

15

1.07E+10

2.1132

4271

1.34E+00

1.5672

16

1.10E+10

2.1657

4261

1.39E+00

1.5817

17

1.13E+10

2.2148

4252

1.44E+00

1.5912

18

1.15E+10

2.2598

4244

1.49E+00

1.5959

19

1.17E+10

2.3052

4236

1.54E+00

1.596

20

1.19E+10

2.3419

4229

1.58E+00

1.5928

21

1.20E+10

2.3749

4223

1.61E+00

1.5869

22

1.22E+10

2.4125

4217

1.65E+00

1.5767

23

1.27E+10

2.5281

4196

1.78E+00

1.5202

24

1.30E+10

2.6287

4181

1.90E+00

1.3909

25

1.33E+10

2.6598

4185

1.95E+00

1.1868

26

1.34E+10

2.6092

4218

1.94E+00

0.982

27

1.35E+10

1.8622

5089

2.09E+00

0.5952

28

1.35E+10

1.7896

5179

2.07E+00

0.5577

29

1.36E+10

1.7977

5175

2.08E+00

0.4921

30

1.37E+10

1.8189

5146

2.08E+00

0.4188

31

1.38E+10

1.8553

5096

2.09E+00

0.3655

32

1.38E+10

1.9031

5045

2.11E+00

0.3262

33

1.39E+10

1.9607

4999

2.16E+00

0.297

34

1.39E+10

2.028

4962

2.24E+00

0.275

35

1.40E+10

2.1052

4933

2.36E+00

0.2575

36

1.40E+10

2.1921

4910

2.51E+00

0.2431

37

1.40E+10

2.2873

4893

2.70E+00

0.2315

38

1.41E+10

2.3882

4881

2.91E+00

0.2223

39

1.41E+10

2.4941

4873

3.15E+00

0.2147

40

1.42E+10

2.6087

4865

3.42E+00

0.2078

41

1.42E+10

2.7251

4858

3.72E+00

0.202

42

1.42E+10

2.846

4851

4.03E+00

0.1968

43

1.42E+10

2.9752

4844

4.38E+00

0.192

44

1.43E+10

3.0965

4839

4.72E+00

0.1882

45

1.43E+10

3.2259

4832

5.10E+00

0.1843

46

1.43E+10

3.3599

4825

5.50E+00

0.1807

47

1.43E+10

3.4945

4817

5.91E+00

0.1775

48

1.44E+10

3.6267

4810

6.33E+00

0.1745

49

1.44E+10

3.7604

4803

6.76E+00

0.1717

50

1.44E+10

3.897

4795

7.21E+00

0.1691

51

1.44E+10

4.0306

4788

7.67E+00

0.1666

52

1.44E+10

4.1611

4780

8.12E+00

0.1644

53

1.44E+10

4.2913

4773

8.59E+00

0.1624

54

1.44E+10

4.4265

4765

9.08E+00

0.1603

55

1.44E+10

4.562

4757

9.58E+00

0.1584

56

1.44E+10

4.6921

4749

1.01E+01

0.1567

57

1.45E+10

4.8267

4742

1.06E+01

0.1549

58

1.45E+10

4.9623

4734

1.11E+01

0.1532

59

1.45E+10

5.0033

4731

1.13E+01

0.1528



Prof. Larry Molnar

Department of Physics and Astronomy

Calvin University

1734 Knollcrest Circle SE

Grand Rapids, MI 49546-4403 USA



E-mail: lmolnar at calvin.edu<mailto:lmolnar at calvin.edu>

Phone: 616-526-6341


_______________________________________________
mesa-users at lists.mesastar.org<mailto:mesa-users at lists.mesastar.org>
https://lists.mesastar.org/mailman/listinfo/mesa-users<https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.mesastar.org_mailman_listinfo_mesa-2Dusers&d=DwMFaQ&c=4rZ6NPIETe-LE5i2KBR4rw&r=rohBEfdOjnrgwB12cgkmE-lSVHI5uTtNMjZ7eHCNBFw&m=TcV9zvWM_qhAe3cKy0vLoh10y3Q8TnJfqy2u4UVgI4w&s=aopYUxFCI8sZ1wKbWCljf4UKIV4oAPGoa0z8dZIwdGk&e=>


--
Pablo Marchant Campos
M.Sc on Astrophysics, Universidad Católica de Chile
PhD on Astrophysics, Argelander-Institut für Astronomie, Universität Bonn
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200727/b613e4ab/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: inlist
Type: application/octet-stream
Size: 771 bytes
Desc: inlist
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200727/b613e4ab/attachment.obj>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: M10RotExampleA.txt
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200727/b613e4ab/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: M10RotExampleB.txt
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200727/b613e4ab/attachment-0001.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: HighRes_L-age.png
Type: image/png
Size: 56657 bytes
Desc: HighRes_L-age.png
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200727/b613e4ab/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: HighRes_OmegaOmegaC-age.png
Type: image/png
Size: 83315 bytes
Desc: HighRes_OmegaOmegaC-age.png
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200727/b613e4ab/attachment-0001.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: HighRes_HR.png
Type: image/png
Size: 89314 bytes
Desc: HighRes_HR.png
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200727/b613e4ab/attachment-0002.png>


More information about the Mesa-users mailing list