Hi all,
I am trying to simulate an injection process in a doublet system of EGS. To simulate the initial phase I choose to use a HT process and I want to use the pressure and temperature distribution obtained in initial phase as the initial condition of my THM injection process. How can I achieve it? I haved tried to use the last vtu result file of HT process but it is a linear mesh vtu so the mechanical process cannot be simulated. However, use a quadratic mesh to simulate the HT process also failed. Maybe there is a better method. Can anyone help me?
You don’t necessary need a quadratic mesh for THM if I remember it correctly. Could you try to use the linear elements mesh for THM while changing the order of the ‘displacement’ process variable to 1?
2 Likes
Hi, @dmitri.naumov ,I follow your advice and it works. The simulation can run but it doesn’t turn out the way I want. I use the way that I mentioned in my post, use the last vtu result file of HT process as the input mesh file in my THM process file, but the pressure and temperature distribution seems not to be read as the initial condition. I think the symbol <mesh></mesh> just read the grid, while other data arrays won’t be input. So far the problem haven’t been solved.
OK, you are getting closer. In you project file you have to set the initial temperature and pressure field to be read from the mesh: Have a look on the example here for pressure_ic:
Same has to be done for temperature.
Post your prj file if there is still an error.
– d
Thanks a lot. I read the file you provide and the initial condition seems good. But I have another question while I have read other files which also use MeshNode type for parameters’ definition.
https://doxygen.opengeosys.org/d0/df3/ogs_ctest_prj__parabolic__componenttransport__variableneumannboundary__vdbc_input__prj
I notice that when the parameter is used as a boundary condition, a certain mesh name will be written in the definition. However, when it is used as an initial condition, there will not be a certain mesh name. Does the MeshNode type parameter for initial condition use the field_name in the first mesh we write at the beginning of prj file by default when there are more than one mesh input? Or we can also add a certain mesh name for the initial condition parameter’s definition?
Right, the first mesh is implicitly used if nothing else is specified. For the boundary values the mesh on which the parameter is defined must be specified.
You can add more meshes (like in the component transport example you link) and use them to get certain values for the parameters. This meshes must provide a bulk_element_id and bulk_node_id CellData and PointData fields (created by identifySubdomains tool), s.t. OGS knows their relationship to the first (primary/bulk) mesh.
I see. Thank you very much!
Hi, @dmitri.naumov
I have a new question now. I want to extract the distribution of the pressure at the bottom line of the initialization result as the boundary condition of the THM injection process. How can I achieve that? I have tried the constructMeshesFromGeometry tool, but it can only extract the single line mesh without any distribution of the variables. I think it probably can be done in Paraview, but my model is 2D so some extract filters don’t work fine. Can you give me some advice?
The post-processing is not done in OGS. In Paraview one could use the PlotOverLine filter to extract information along a line, but there are other ways too.