I'm trying to set up the problem of deriving the thermodynamics of hadronic matter. I know how to proceed in the case of an effective description such as mean field (Walecka/linear sigma model) but I'm trying to start from QCD. I do not wish to actually obtain the partition function, but I want to at least set up the complete problem.
In the mean field approximation, we usually write baryon and meson fields and baryon/meson couplings as the interactions, meaning that we also bypass the problem of describing the hadrons as QCD bound states. On the extreme oposite, I know how to proceed in the case of pure quark matter (it's just the "QCD Lagrangian").
How do I write the full Lagrangian? How do I represent the hadrons and the hadron-hadron strong (well, and EM) interaction without shifting to an effective approximation? Can it be done with quarks and gluons as the basic degrees of freedom?
I don't have any compromise with the "practical usefulness" of the result and I'm not particularly interested on the peculiarities of the dynamics of the bound states.