3D and Some Other Things Missing from the Theory of Massive Star Evolution

This is a sketch of a 321D approximation for stellar convection which is nonlocal, and thus has nonzero fluxes of KE (to be published in more detail elsewhere). Boundary conditions are discussed in a fluid dynamics context (i.e., predictions for overshoot, semiconvection and entrainment are analyzed). We plan to add this as an option to MESA. Inclusion of KE fluxes seems to help resolve the solar abundance problem (Asplund 2009). Smaller cores may ease the explosion problems with core collapse supernova simulations.