I plan to simulate heat transfer within an exchanger. For energy conservation in transfer fluid, there is a term: 4h/D*(Tf – Tw). I intend to involve this term as an extra source term in the collision step. Since it is dependent of time and space(temperature of tube wall), I wonder if there is a convenient and efficient way to do this.
I have paid attention to the field concept used in openlb, my idea is to first define a relative field for fluid area after we calculate the wall temperature, then we can calc the convective term in the collision step within fluid domain, the difficulty lies in how to define a field variable of flow direction and bind it both at fluid and wall in the same abscissa.
Honestly, I’m even not sure whether this idea would work out, especially in parallel mode.