Quasi-Newton methods for partitioned simulation of fluid-structure interaction reviewed in the generalized Broyden framework: code and data

These files accompany the publication
N. Delaissé, T. Demeester, R. Haelterman and J. Degroote. Quasi-Newton methods for partitioned simulation of fluid-structure interaction reviewed in the generalized Broyden framework. Archives of Computational Methods in Engineering, Vol. 30, 3271-3300, 2023. doi: 10.1007/s11831-023-09907-y
In this work, the performance of multiple quasi-Newton methods are compared in terms of memory requirements and computational time. The results are generated for the well-known flexible tube example case, using the open-source code CoCoNuT. This code, developed at Ghent University, is Python-based and has the capability to couple existing solvers, both open-source and commercial solvers.
This archive consists of the following files.

coconut.tar.gz: the specific CoCoNuT version used (sep-2022), including the Python flow and structure solvers for the flexible tube and modifications for monitoring memory requirements
compare_coupling_algorithms.tar.gz: the scripts to set up the cases and perform the calculations and post-processing
results.tar.gz: the generated result data
For requirements to run CoCoNuT, refer to the documentation. Additionally, the Python package guppy3 is required for monitoring the memory use. In this work the data were generated with Andaconda3-2022.05 and the package guppy3-3.1.2.
Before running the provided scripts, make sure the parent directory of the "coconut" folder is added to the PYTHONPATH. The calculations can be started with "python run.py". For the cases which names contain "_m" followed by a number, e.g. "_m100", the number refers to the number of discretization points on the interface. The cases with suffix "_c" are distinct from those without, as they don't perform the time consuming memory monitoring and are therefore used for measuring computational time.