[shanyeyun]

Originally Posted by EmadTandis

I checked it before and concluded that:
# K.originalPatchField()/(1 - p.weights())/mld.magDelta(p.index())
means original K for a considering domain (not interpolated K) divided by "distance from cell to face for patch". Regarding ".originalPatchField()", from my experience, it returns cell value of K. However you can find it in regionCouplingFvPatchField.C for a better understanding.


This is my idea about effect of regionCoupling patches in solution process:



internalCoeffs_: added to diag() of neighbor cell in the matrix

boundaryCoeffs_: It will be then multiplied by matrixUpdateBuffer_ of the shadow cell and the result is added to matrix.source()

manipulateMatrix(): It will add a new source to matrix.source()

Though both of us agree with ".originalPatchField()", from experience, it returns cell value of K'' However, when I dig deeper the source code, I can not find evidence from the source code. Hope anyone could clarify this based on the source code.

[EmadTandis]

Originally Posted by shanyeyun

Yes,

k= distanceFromFluidCellToSolidCell /

( distanceFromFluidCellToWall / ThermalConductivityAtFluidCell + distanceFromSolidCellToWall / ThermalConductivityAtSolidCell ),


And this equation is actually based on continuous Temperature and continuous heat flux at the wall.
(note: all I don;t consider radiation. )

This is right. It is based on physical law (conservation of energy at wall) and kHar is calculated in such a way as we have a single domain (with variable K) including the interface whose diffusivity is consistent with continuous heat transfer.