[WeiW]

Do we have a sub-filter scale model of scalar variance in converge? I searched the manual and udf example templates, but I only found modeling of mixture fraction variance (\tilde{Z^{''2}}), velocity variance, and G-eqn variance. What I need is a scalar (total enthalpy, sensible enthalpy) variance, which can be modeled by either a transport equation model, or an algebraic model (like dynamic Smagorinsky model). I would like to to know if we already have such models in converge so that I can directly use it. These models are pretty basic in LES simulation, so I would think a well-integrated software like converge may have abundant built in models.

[Dan.Maciejewski]

CONVERGE closes the sub-grid-scale (SGS) turbulent diffusion and turbulent conductivity with the modeled turbulent viscosity and appropriate non-dimensional numbers, e.g. turbulent Schmidt and Prandtl numbers. If you are interested using a more sophisticated model for sub-grid terms, you could use User-Defined Functions (UDF) to implement those.

However, in LES as we refine the refine the grid, these sub-grid terms become smaller as the resolved field captures these details, which is desirable. Adaptive Mesh Refinement (AMR) in CONVERGE allows the solver to refine the mesh only where needed to resolve the turbulent effects. Thus with AMR, CONVERGE can resolve the turbulent effects (enhanced diffusion of momentum, energy and species) where needed without creating an overly refined mesh everywhere, which is slow to solve.