At the end of this journey

Hello everyone, and welcome to my final post! In this post, I will present the 2 test cases I used to verify the 3-D laminar incompressible flow solver I developed during my project using the FEniCSx computational platform. The test cases refer to the 3-D laminar flow around a cylinder confined inside a channel, with a parabolic inlet profile both for steady and transient state.

In order to validate the results the drag and lift force coefficients were calculated and compared to available experimental data. Different meshes with various numbers of elements were investigated. Furthermore, different basis functions were used. For the steady-state case, we can see in the following figures how the solver is able to accurately predict the coefficient values, and how this accuracy increases for more number of elements.

• 

• 

In the following figure we can see the velocity contours in 2 symmetry planes of the computational domain.

• 

• 

Next, a transient case was simulated, again the accuracy of the solver was verified using the drag and lift coefficient data. 3 different setups were investigated with varying mesh sizes and basis functions orders. In the following videos we can see the time evolution of the velocity magnitude contours for the transient case.

You can check my final presentation for more info!

Before this post ends, I would like to thank my mentor Dr. Ezhilmathi Krishnasamy for helping me throughout this project. I would also like to thank PRACE for providing me with this fantastic opportunity. I learned many new things about HPC systems and parallel programming. I highly recommend it to everyone applying next year. I leave you with a picture with Mathi and my co-workers during the program: Filippo and Alexander during our visit to the AION cluster room.