Questions About Turbulent Pipe Flow Quasi-3D Tutorial
Dear All, I have some questions about Turbulent Pipe Flow Quasi-3D Tutorial: * When I run this tutorial with DELL Workstation with 12 core the analysis finished at 335000th time step but in the xml file it is written that NumSteps=2000000. Is this a random number or is there a way to calculate this number of time steps? * In this problem Quasi-3D approach is used but there is no information about the length of the pipe. At first I thought this is a periodic problem but there is also no periodic boundary condition in the xml file. Can you please clarify this point? o Moreover, when I converted the last *.chk file to Tecplot's dat file I only see the front face of the pipe (2D) how can I add the third dimension to see the contours in 3D as shown in the link below: + http://www.nektar.info/wiki/3.4/UserGuide/Examples/IncNavierStokesSolver/Tur... o Is it possible to use Quasi-3D approach for all axisymmetric geometries like constricted 3D pipe as below: * In the Geometry it is said that D=1. Is it unitless otherwise what is its unit mm or meters? And finally regardless of the tutorial I want to ask that: * Is it possible in Nektar++ to use the solution of a mesh as an initial condition of another mesh? Regards, Kamil
Hi Kamil, On 02/07/14 14:04, Kamil ÖZDEN wrote:
* When I run this tutorial with DELL Workstation with 12 core the analysis finished at 335000th time step but in the xml file it is written that NumSteps=2000000. Is this a random number or is there a way to calculate this number of time steps?
Did the simulation complete successfully (and write out a final .fld file), or did it produce an error message? Do you set the FinTime parameter in the session file? The most intuitive way to use these parameters is to prescribe the time-step you want and the total simulation time (FinTime), and then set NumSteps=FinTime/TimeStep. This removes any ambiguity.
* In this problem Quasi-3D approach is used but there is no information about the length of the pipe. At first I thought this is a periodic problem but there is also no periodic boundary condition in the xml file. Can you please clarify this point?
The pipe is discretised in the streamwise direction using a Fourier expansion, which is naturally periodic. The length of the pipe is specified using the LZ parameter.
o Moreover, when I converted the last *.chk file to Tecplot's dat file I only see the front face of the pipe (2D) how can I add the third dimension to see the contours in 3D as shown in the link below: + http://www.nektar.info/wiki/3.4/UserGuide/Examples/IncNavierStokesSolver/Tur...
Not sure why you are not seeing a 3D file. If it is not too large can you forward to me personally (not the mailing list) the .xml file you are using, along with the last .chk, and I will investigate.
o Is it possible to use Quasi-3D approach for all axisymmetric geometries like constricted 3D pipe as below:
At present, we do not have the cylindrical coordinate system implemented which would be required for this (although we are developing the initial implementation for this feature at the moment).
* In the Geometry it is said that D=1. Is it unitless otherwise what is its unit mm or meters?
The incompressible Navier-Stokes equations are solved in a non-dimensional form, which depends how you choose to define the Reynolds number for your problem.
And finally regardless of the tutorial I want to ask that:
* Is it possible in Nektar++ to use the solution of a mesh as an initial condition of another mesh?
If the geometry is the same, but the mesh is different, you can project the solution of one mesh onto another and provide this as an InitialCondition for the second simulation. For this, you should use the FieldConvert utility. See the "Interpolating one field to another" on https://www.nektar.info/wiki/3.4.1/UserGuide/Tutorial/FieldConvert Cheers, Chris -- Chris Cantwell Department of Aeronautics Roderic Hill Building Imperial College London South Kensington Campus London SW7 2AZ Email: c.cantwell@imperial.ac.uk www.imperial.ac.uk/people/c.cantwell
participants (2)
-
Chris Cantwell
-
Kamil ÖZDEN