Hello everyone, I plan to use FEniCS to solve a coupled steady-state heat transfer problem for a three-layered system consisting of solid-gas-solid layers, where the gas gap does not have a mesh. The solid regions will be modeled using the Fourier equation, and the gas gap will be modeled using the Ross-Stoute interstitial heat transfer model. Currently, there are two approaches:
1.Global Solution: The entire model will be represented as a single mesh file (without meshing the gap). The challenge lies in combining the Fourier equation and the interstitial heat transfer model into a unified governing equation.
2.Partitioned Solution with Boundary-Coupled Iteration: The two solid regions will be separately meshed into two mesh files (without meshing the gap), and solved independently. The boundary coupling will then be carried out iteratively. The challenge here is that the author is only familiar with the partitioned solution algorithm for the solid-gas two-layer system:
The partitioned solution for the solid-gas two-layer system involves assuming the temperature distribution on the coupled boundary, solving for the solid, obtaining the heat flux density and temperature gradient at the coupled boundary, and then solving for the gas region based on interface continuity conditions to obtain a new temperature distribution on the coupled boundary. This process is repeated until convergence.
I am currently considering the second method and have two questions:
1.If I extend the partitioned solution approach for the two-layer system to a three-layer solid-gas-solid system, what would be the specific procedure?
2.If I use the partitioned solution approach, do I need to create two separate mesh files (without meshing the gap), or can I create a single mesh file and solve each region separately and then iteratively couple them during the solution process?
I hope to receive guidance from the seniors. Thank you!