Volumetric Heat Source to Mass Heat Source

[andrewdnolan]

In the surface boundary process function (Surface_Processes in src/elmer_UDF/SurfaceBoundary.f90) I calculate the volumetric heat source (Q_lat) based on Eqn. (9) from Wilson and Flowers (2013):

thermal-structure/src/elmer_UDF/SurfaceBoundary.f90

Line 528 in a6041db

Q_lat = (1 - r_frac) * (rho_w/h_aq) * L_heat * f_dd * SUM(PDD)

My initial attempts to prescribe Q_lat as prescribing a volumetric source (J/m^3) along the surface boundary conflicted with the Dirichlet B.C. based on air temp and in-effect Q_lat was ignored by Elmer. Previously as solution this problem, I prescribed Q_lat as a volumetric heat source one node below the surface, which worked but was nonphysical.

Instead I've edited the code to convert Q_lat to a mass heat source (J/kg) which can be used to augment the Dirichlet B.C. for enthalpy along the free surface. See Eqn. (6) from Licciulli et al. (2019) for an example of this approach.

I'm unsure what density should I use to convert Q_lat from a volumetric (J/m^3) to mass (J/kg) heat source. I'm currently just using the density of water:

thermal-structure/src/elmer_UDF/SurfaceBoundary.f90

Line 550 in a6041db

Surf_Enthalpy % values (Surf_Enthalpy % perm(n)) = Q_lat/rho_w + H_surf

but I'm maybe this should be surface density?