[mesa-users] Vink_wind_eta

Fri Jun 5 14:06:13 EDT 2015

There is some paper by van Loon who has put observational constraints at the transition between cool star winds (where the mass loss mechanism involves molecules and dust) and hot winds (based on radiation pressure on lines) at log Teff = 3.65, if I can remember correctly. I spent some time to find the reference, but wasn't able to do so. But 10,000K sounds a bit high for this transition.

I am not sure I like the "recipe" in the parameter names. While it is true that these are very uncertain prescriptions, many of them are based on a lot of laborious and serious work, done in the best way possible at the time. So calling them "recipe" is maybe a tad to casual?

Best, Falk.

PS: For AGB stars I believe a good mass loss addition for experiments would be the pulsation period based math loss that the Australian groups use, based - I believe - on Vassiliadis & Wood (1993)

________________________________
From: Aaron Dotter <aaron.dotter at gmail.com>
Sent: June 4, 2015 5:40 PM
To: Zsolt Keszthelyi
Cc: MESA users group
Subject: Re: [mesa-users] Vink_wind_eta

Hi Zsolt,

Thanks for this clear proposal.  It looks good to me.

My only suggestion is that, in keeping with MESA tradition, we let the user set the transition temperature between hot and cool.  It could be 10,000K in controls.defaults.

Something like wind_hot_cool_transition_T = 1e4 !K

Aaron

On Fri, Jun 5, 2015 at 12:19 AM, Zsolt Keszthelyi <Z.Keszthelyi at campus.lmu.de<mailto:Z.Keszthelyi at campus.lmu.de>> wrote:
Hi Bill, MESA users,

My suggestion would be to make the partitioning clear between radiative line driven winds (of hot massive stars) and dust driven winds (of cool giant stars).

As Mathieu said: the wind recipes are constrained for given Teff ranges, and I would suggest to use ‘hot winds’ above 10 kK and ‘cool winds’ below that. (In this sense thus my recommendation is basically to use the partitioning from the ‘Dutch wind’ but already at a higher level.)

1. Cool stars

I would not touch the existing RGB/AGB schemes and partitioning but it would sound reasonable to me to include them under a control ‘cool_wind_recipe’.

>From the current list the following prescriptions may be allowed to be used in this scheme: 'Reimers', 'Blöcker', 'de Jager', 'van Loon', 'Nieuwenhuijzen'.

2. Hot stars

I would prefer to call ‘hot_wind_recipe’ the scheme where you set in your wind description for hot stars until 10 kK.

>From the current list the following prescriptions may be allowed to be used in this scheme: ‘Kudritzki’, ‘Vink’.

I would recommend changing here only the naming but not the functionality: Vink_wind_eta --> Vink_wind_scaling_factor and Kudritzki_wind_eta --> Kudrizki_wind_scaling_factor (see previous emails).

A remark: Investigations including correction for clumping in the winds of hot stars may agree with a decreased Vink rate by a factor 2-3. This would imply Vink_wind_scaling_factor = 0.3 – 0.5.

3. Wolf-Rayet stars

This scheme might be called ‘WR_wind_recipe’ and could be allowed for Teff  > 10 kK and surface H mass fraction < 0.4, with the prescriptions from Gräfener & Hamann and Nugis & Lamers.

I suggest to change here as well: Gräfener_wind_eta –> Gräfener_wind_scaling_factor + add Nugis_wind_scaling_factor. (The ‘Nugis wind’ exists under the Dutch scheme but not separately thus far.)

Besides the switches between the 3 (hot, cool, WR) schemes the interpolation may also be considered.

Between hot and cool:

In a 30 Msun test model I ran, the Vink rates already gave smaller Mdot values than the ‘cool_wind_recipes’ (e.g., de Jager, Nieuwenhuijzen) at around 16 kK, hence it might be reasonable to start the interpolation already at this point – in contrast to the ‘Dutch wind’ where the interpolation is at 12 kK (but perhaps with a larger dT range, say dT = 3-4 kK). An other alternative I tested for this issue is, e.g., to include the *second* bi-stability jump (due to the recombination of Fe III to II in the wind) in the Vink recipe. However, it gives high Mdot values (almost 10^-4 Msun/yr) at high temperatures (17 kK) which might not be fully in agreement with derived mass loss rates. It has indeed been suggested that the jump in Mdot might be smaller but this is very-very challenging to constrain... (My understanding is that the Geneva code uses the second bi-stability jump and the Bonn code uses the interpolation instead, see e.g., Groh et al. 2014 and Brott et al. 2011.)
Perhaps the best for the users would be to keep the interpolation (either as it is or shifted to Teff = 16 kK) but add a wind routine (which I simply called Vink2) that contains the second bi-stability jump. Then users could choose which method they prefer (with the second jump the interpolation to cool winds will be small – but it depends on, of course, at which Teff it starts).

Maybe someone would have suggestions for the transition between hot-WR and cool-WR?

It is evident that in this partitioning the ‘Dutch wind’ as it is now, could be used in the following way:

hot_wind_recipe = Vink
Vink_wind_scaling_factor = 1 (default)

cool_wind_recipe = de Jager/van Loon/Nieuwenhuijzen
de Jager/van Loon/Nieuwenhuijzen_wind_eta = 1 (default)

WR_wind_recipe = Nugis & Lamers
Nugis_wind_scaling_factor = 1 (default)

These are obviously just some ideas and I would be very happy to read comments from the others.

Cheers,

 Zsolt

> Von: Bill Paxton <paxton at kitp.ucsb.edu<mailto:paxton at kitp.ucsb.edu>>
> An: Zsolt Keszthelyi <Z.Keszthelyi at campus.lmu.de<mailto:Z.Keszthelyi at campus.lmu.de>>
> Kopie: MESA users group <mesa-users at lists.sourceforge.net<mailto:mesa-users at lists.sourceforge.net>>
> Datum: Wed Jun 03 20:21:35 CEST 2015
>
> Hi Zsolt,
>
> Would you be willing to put together a specific proposal for changes to "wind_eta"
> controls and send it to mesa-users for everyone to consider.
> Once we get agreement in detail, I'll make the changes to the code.
>
> Thanks,
> Bill
>
>
>
>
> ------------------------------------------------------------------------------
> _______________________________________________
> mesa-users mailing list
> mesa-users at lists.sourceforge.net<mailto:mesa-users at lists.sourceforge.net>
> https://lists.sourceforge.net/lists/listinfo/mesa-users
>

------------------------------------------------------------------------------
_______________________________________________
mesa-users mailing list
mesa-users at lists.sourceforge.net<mailto:mesa-users at lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/mesa-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20150605/8fa9eb38/attachment.html>