This Computation Fluid Dynamic (CFD) model of a sugar beet clamp has been developed using the OpenFOAM open source software. The table below gives links to posts that describe the development of the model.
NB: owing to this work being under development for publication, the publication on this website of the details of the model development have stopped. The final model count is closer to 80, with many of them run multiple times. The expected manuscript submission date is end 2022. At the latest, the paper will be in the thesis this work is a part of, with a defense date set for 26 May 2023. That said…
The github repository for these models are public. You need an OpenFOAM installation to run the model, ParaView to inspect the results, and a fair bit of specialised knowledge to modify the model (properly) …
The physical model is fairly idealised, but also has a fairly good resolution and is built on some sound experimental data.
Some of the key features of the model include:
- separate momentum and energy equation solvers. The momentum equation is solved first, with the resulting fields passed to the energy equation solver. This fits the available data (15 minute interval averages), and allows the slower/ small dt momentum equation to be solved over a shorter period, and the fast / large dt energy equation to be solved over the entire 15 minute period, all while the whole system runs as quickly as possible.
- separate energy equations for fluid and solid. We have good data for the fluid phase (air), but it is the solid phase we really care about. Also, it is possible that the temperature loggers deep in the experimental clamp we had are actually in the solid phase: they were driven in after the construction of the clamp. Over much of the domain, the temperature of the two phases seems very similar.
- inlet and outlet swap. By having a symmetrical domain, changes in the wind direction can be accounted for by just choosing the appropriate mesh for the given 15 minute interval step.
- cover region. This is pretty broad (15 cm) to accommodate a coarse grid. The actual cover is 0.11 cm thick.
- Ts at the clamp edge. To ensure no odd conduction in the solid phase (Ts) at the edge of the clamp region from outside of the clamp, the Ts field in the open region is allowed to follow the Ts in the clamp region.
TABLE: Steps in the model development, including solver, case files, tags of the important features, and links to related blog posts.
| Topic | Solver | Case files | Tags and Notes | Post | |
| 15 | Uncovered clamp full test run | pimpleFoam and Own: clampPimpleFoam v5 | clamp_15 | Allrun. Subcases. incompressible, turbulent, transient. blockMesh, topoSet, setFields, mapFields. Velocity, pressure, temperature. Porous media. | Link |
| 14 | Implementing a strategy for dealing with precision issues | pimpleFoam and Own: clampPimpleFoam v5 | clamp_14 | Allrun. Subcases. incompressible, turbulent, transient. blockMesh, topoSet, setFields, mapFields. Velocity, pressure, temperature. Porous media. | Link |
| 13 | Dealing with precision issues by removing p and U from solver. | Own: clampPimpleFoam v5 | clamp_13 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous media. | Link |
| 12 | Implicit T in inter-phase heat transfer | Own: clampPimpleFoam v4 | clamp_12 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous media. | Link |
| 11 | Convection only. A number of other small updates are noted. | Own: clampPimpleFoam v3 | clamp_11 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous media.https://www.meran.se/sugar-beet-clamp-cfd-modelling-a-strategy-for-dealing-with-precision-issues/ | Link |
| 10 | Coarser mesh and adjust time step | Own: clampPimpleFoam v3 | clamp_09 clamp_10 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous media. | Link |
| 9 | Convective Heat Transfer Coefficient as a field | Own: clampPimpleFoam-2 to clampPimpleFoam-3 | clamp_08 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous media. | Link |
| 8 | Updated 3D ventilation | Own: clampPimpleFoam v2 | clamp_vent_07 | incompressible, turbulent, transient. blockMesh, snappyHexMesh, topoSet, setFields. Velocity, pressure, temperature. Porous medium. | Link |
| 7 | Time variant inlet conditions on velocity and temperature | Own: clampPimpleFoam v2 | clamp_07 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous media. | Link |
| 6 | Modelling a porous membrane | Own: clampPimpleFoam v2 | clamp_06 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous media. | Link |
| 5 | Adding second energy equation to pimpleFoam | Own: clampPimpleFoam-1 to Own: clampPimpleFoam-2 | clamp_05 | incompressible, turbulent, transient. blockMesh, topoSet, setFields. Velocity, pressure, temperature. Porous medium. | Link |
| 4 | Adding energy equation to pimpleFoam | pimpleFoam to Own: clampPimpleFoam-1 | clamp_04 | incompressible, turbulent, transient. blockMesh, topoSet. Velocity, pressure, temperature. Porous medium. | Link |
| 3 | Moving to a transient solver | pimpleFoam | clamp_03 | incompressible, turbulent, transient. blockMesh, topoSet. Velocity, pressure. Porous medium. | Link |
| 2 | 3D ventilated clamp | simpleFoam | clamp_vent_02 | incompressible, turbulent, | Link Link* |
| 1 | Adding accurate Darcy and Forchheimer coefficients* | simpleFoam | clamp_02 | incompressible, turbulent, steady state. Velocity, pressure. Porous medium. *NB: kinematic velocity for the fluid was wrong in the initial version of this example. It has been updated in the case files and in a new post. | Link Link* |
