Hey everybody,
I have recently started calculating oscillation frequencies of red giant stars with gyre. For this I proceeded as follows: I used a stellar model (which I obtained from Mesa) for a red giant with a mass of 0.6*Msun. For the frequency scan I used freq_min = 3 microhertz and freq_max = 4 microhertz since the frequency of maximum oscillation power v_max = 3.47 microhertz for the star I want to investigate. As a solution I get only gravity modes with thousands of nodes. I would have expected p modes because the convective envelope is large in stars with less than one solar mass. So all in all my question is why I only get g modes through the gyre calculations. Thanks for the help.
kind regards
Gyre_Beginner
Only g modes calculated i red giant star

 Posts: 6
 Joined: Tue Nov 12, 2019 4:49 am
Only g modes calculated i red giant star
 Attachments

 evo_Msun=0.6.mesa
 (664.95 KiB) Downloaded 55 times

 gyre_Msun=0.6.in
 (1.58 KiB) Downloaded 62 times
Re: Only g modes calculated i red giant star
How exactly did you estimate νmax?Gyre_Beginner wrote: ↑Thu Dec 05, 2019 10:16 amFor the frequency scan I used freq_min = 3 microhertz and freq_max = 4 microhertz since the frequency of maximum oscillation power v_max = 3.47 microhertz for the star I want to investigate.

 Posts: 6
 Joined: Tue Nov 12, 2019 4:49 am
Re: Only g modes calculated i red giant star
Hello Warrick,
regarding the v_max I refer to the paper "Models of red giants in the CoRoT asteroseismology fields combining asteroseismic and spectroscopic constraints" (link: https://arxiv.org/pdf/1505.01529.pdf). On page 3 there is a tabular containing information about red giants like v_max, the large frequency spacing, ect. The star I am investigating in this tabular is alpha Boo. So far my challenge was to find a suitable frequency range that I can scan with gyre. In the paper "GYRE: an opensource stellar oscillation code based on a new Magnus Multiple Shooting scheme" (link: http://www.astro.wisc.edu/~townsend/res ... s/gyre.pdf) a star with a mass of 1.5 Msun was modeled by Gyre and therefore the scanning interval of the frequency was chosen so that it loosely correspond to the frequency of maximum power v_max. My approach was the same, so I started to search for publications which contains v_max of red giant stars.
kind regards
Gyre_Beginner
regarding the v_max I refer to the paper "Models of red giants in the CoRoT asteroseismology fields combining asteroseismic and spectroscopic constraints" (link: https://arxiv.org/pdf/1505.01529.pdf). On page 3 there is a tabular containing information about red giants like v_max, the large frequency spacing, ect. The star I am investigating in this tabular is alpha Boo. So far my challenge was to find a suitable frequency range that I can scan with gyre. In the paper "GYRE: an opensource stellar oscillation code based on a new Magnus Multiple Shooting scheme" (link: http://www.astro.wisc.edu/~townsend/res ... s/gyre.pdf) a star with a mass of 1.5 Msun was modeled by Gyre and therefore the scanning interval of the frequency was chosen so that it loosely correspond to the frequency of maximum power v_max. My approach was the same, so I started to search for publications which contains v_max of red giant stars.
kind regards
Gyre_Beginner
Re: Only g modes calculated i red giant star
What's the radius of your stellar model compared to the entry for α Boo in that table? How would that affect νmax?
I'm deliberately baiting you into computing νmax of your model for the sake of your own understanding. That is, I don't think it's helpful for me to just tell you what νmax is for your stellar model but I'll give you a clue that it isn't 3.47 μHz. Try using the scaling relations by Kjeldsen & Bedding (1995) to estimate νmax. These are the same relations from which equations 1 and 2 are derived in the first paper you linked.
I'm deliberately baiting you into computing νmax of your model for the sake of your own understanding. That is, I don't think it's helpful for me to just tell you what νmax is for your stellar model but I'll give you a clue that it isn't 3.47 μHz. Try using the scaling relations by Kjeldsen & Bedding (1995) to estimate νmax. These are the same relations from which equations 1 and 2 are derived in the first paper you linked.

 Posts: 6
 Joined: Tue Nov 12, 2019 4:49 am
Re: Only g modes calculated i red giant star
Hello Warrick,
thank you very much for your answere. I have used the scaling relations from the paper Kjeldsen & Bedding (1995) with the radius and effective temperature of my stellar model which leads to much better results. I could now find both p and g modes, because I now know in which frequency range I have to search (as you said the v_max in my previous calculations was chosen wrong). I have now investigated eigenfunctions of some calculated modes for a red giant (Msun = 0.6) which brings me to my current question: Plotting x=r/R against Re(xi_r) from the file mode.00001 (corresponding to n_pg = 31) leads to the oscillation curve below. Unfortunately this curve dont make any sense for me. Since I am plotting a g mode, I would expect the oscillation to take place much closer to the core of the star. My previous understanding was that g modes do not propagate into the convective envelope which is pretty large in stars with fewer than one solar mass. So as I said I wouldnt expect g mode oscillations take place so close at the surface. Can someone please show me the mistake I am doing. I thank you all for your help.
kind regards
Gyre_Beginner
thank you very much for your answere. I have used the scaling relations from the paper Kjeldsen & Bedding (1995) with the radius and effective temperature of my stellar model which leads to much better results. I could now find both p and g modes, because I now know in which frequency range I have to search (as you said the v_max in my previous calculations was chosen wrong). I have now investigated eigenfunctions of some calculated modes for a red giant (Msun = 0.6) which brings me to my current question: Plotting x=r/R against Re(xi_r) from the file mode.00001 (corresponding to n_pg = 31) leads to the oscillation curve below. Unfortunately this curve dont make any sense for me. Since I am plotting a g mode, I would expect the oscillation to take place much closer to the core of the star. My previous understanding was that g modes do not propagate into the convective envelope which is pretty large in stars with fewer than one solar mass. So as I said I wouldnt expect g mode oscillations take place so close at the surface. Can someone please show me the mistake I am doing. I thank you all for your help.
kind regards
Gyre_Beginner
 Attachments

 mode.00001.txt
 (192.91 KiB) Downloaded 60 times

 g mode_n_pg = 31.png (9.57 KiB) Viewed 1825 times

 gyre_Msun=0.6.in
 (1.63 KiB) Downloaded 61 times
Re: Only g modes calculated i red giant star
At this point, your questions aren't really about GYRE but about stellar oscillation theory in general. You may want to consult some course notes or your advisor. I recommend the notes by Jørgen ChristensenDalsgaard. In Section 5.3.3, he discusses mixed modes in the star η Boo. Fig. 5.19 (eigenfunctions of a mixed mode) is hopefully enlightening.
Anyway, the mode you have plotted is neither purely a gmode nor purely a pmode. It is a mixed mode. In the nonconvective core, the mode behaves like a gmode; in the envelope, it behaves like a pmode. If you look closely at x<0.01, you'll see the gmodelike pulsations. gmodes are more prominent in horizontal displacement, so try plotting "Re(xi_h)" instead and see what you find.
Note that n_pg is basically n_p  n_g, where n_p is the number of nodes in the pmode cavity and n_g is the number of nodes in the gmode cavity. The fact that n_pg is overall negative doesn't mean this is a gmode. You should find that both n_p and n_g are nonzero (although it isn't recorded in the mode file), indicating that the mode has both glike and plike behaviour.
Anyway, the mode you have plotted is neither purely a gmode nor purely a pmode. It is a mixed mode. In the nonconvective core, the mode behaves like a gmode; in the envelope, it behaves like a pmode. If you look closely at x<0.01, you'll see the gmodelike pulsations. gmodes are more prominent in horizontal displacement, so try plotting "Re(xi_h)" instead and see what you find.
Note that n_pg is basically n_p  n_g, where n_p is the number of nodes in the pmode cavity and n_g is the number of nodes in the gmode cavity. The fact that n_pg is overall negative doesn't mean this is a gmode. You should find that both n_p and n_g are nonzero (although it isn't recorded in the mode file), indicating that the mode has both glike and plike behaviour.

 Posts: 6
 Joined: Tue Nov 12, 2019 4:49 am
Re: Only g modes calculated i red giant star
thank you for your help Warrick. That answere helps me a lot
greetings
Gyre_Beginner
greetings
Gyre_Beginner