[mesa-users] dlnkappa/dlnT at const P, not rho?

[Josiah Schwab]

Hi Lewis,

> I have tried multiple times to get it in that form, with no luck. 
>
> I considered using the triple product rule: (dlnkappa/dlnT)_p = - (dlnp/dlnT)_kappa / (dlnp/dlnkappa)_T
>
> But then how do I get (dlnp/dlnT)_kappa from the code? I’m running in circles with it!

Schematically you likely want to use

  (df/dy)_z = (df/dx)_y (dx/dy)_z + (df/dy)_x

which would give you

   (dkap/dT)_P = (dkap/dRho)_T (dRho/dT)_P + (dkap/dT)_Rho

and then you just need to get (dRho/dT)_P from the stuff in the eos.
The triple product might be handy there.

Josiah