[mesa-users] pseudo-Lagrangian scheme in MESA

[Jean-Claude Passy]

Dear all,

I have some conceptual questions regarding the numerical scheme used in MESA for mass loss/accretion.
MESA (among other codes) is said to use a pseudo-lagrangian approach in the outer region of the star to solve the stellar equations. My questions are the following:

	1) Does it mean that a) the equations are written and solved as a function of q (so there is an extra term in the partial time derivative); or b) the equations are still written in the Lagrangian form and some rescaling is applied (Section 6.6 in Paxton et al. 2011)?

	2) Why doing that? It must be for numerical reasons, and I’d love to get the details.

	3) Where are the relevant pieces of code doing that?


Hope that someone can clarify things for me.

Cheers,

JC