POLYFLOW: Solve an extrusion problem by cascaded approach using csv file
There are situations when its not possible to simulate the entire geometry at one go because of mesh and memory restrictions. In extrusion/coextrusion problem sometimes the die is so long that its hard to keep mesh count low.
In such cases one can run the case in cascaded manner i.e split the geometry into 2-3 small parts. Run the first part and export the field variable in csv format from outlet of first part and apply at the start of the second part as boundary condition. This approach can be then be repeated for subsequent sections.
Following are the steps needed to export csv file from CFD-POSt and then apply at the starting grid of second geometry:
1) Read the cfd post file (cfx.res) in CFD-POST
2) Export data from the outlet boundary in csv format. In the CFDPOST panel mesh information is already selected. You just need to select all the field variables which are needed for imposing boundary condition at the inlet of second geometry. For example, if the second simulation is generalised newtonian non-isothermal flow select temperature and three velocity components in the export panel of CFD-POST
3) Open the .csv file in Excel and edit it in following format in any text editor or in MS-excel:
1; .3078385E+01; .0000000E+00; .1466668E+02;
2; .3078372E+01; .0000000E+00; .1500000E+02;
3; .3068895E+01; .2415273E+00; .1466668E+02;
4; .3068882E+01; .2415263E+00; .1500000E+02;
5; .3040485E+01; .4815655E+00; .1466668E+02;
6; .3040472E+01; .4815635E+00; .1500000E+02;
7; .2993329E+01; .7186346E+00; .1466668E+02;
8; .2993316E+01; .7186317E+00; .1500000E+02;
9; .2927718E+01; .9512732E+00; .1466668E+02;
10; .2927706E+01; .9512693E+00; .1500000E+02;
The field data need to be arranged in the above format . Also a new colum needs to be created for numbering.
4) save the file again in CSV format.
4) Import the new csv file in your second simulation and apply velocity and temperature using CSV profile. The name of the label should be same as the one entered in the csv file. For example here we are using label "TEMPERATURE" to indicate temperature profile. So in polydata during case set up use the label for temperature as "TEMPERATURE".
One drawback of this approach is we don't know the boundary condition at the exit of part 1 to part n-1. This will induce some error in the solution.