Unsteady Fluid Mechanics

Note: an updated version of the dataset and its description is expected to be made available soon. 

Fluid flows are systems governed by the conservation equations of fluid mechanics. In their standard form, these express the conservation of mass, momentum, and energy. This dataset presents such a flow over a cylinder undergoing an imposed motion in the cross-flow direction. The resulting dynamics are challenging to model due to the inherent nonlinearity present in the Navier-Stokes equation, the autonomous component present in these dynamics as well as synchronization, entrainment, and lock-in effects. The dataset contains multiple types of imposed motions (sinesweep, sine, multisine) on the cylinder, and the velocity, pressure, and vorticity of the flow in the wake of the cylinder are measured on a 31 x 31 grid. Furthermore, the kinematic pressure along the cylinder surface as well as the drag and lif force components are also provided. This enables studying a wide range of challenging identification problems.

A detailed description of the considered system and dataset can be accessed through this link. The datasets itself is available for download here in .mat format (1.2GB)

Special thanks to Jan Decuyper, Tim De Troyer, and Mark C. Runacres for making this dataset available.

An unsteady fluid flow in the wake of a moving cylinder. 
Figure source: G.I. Beintema, Data–driven Learning of Nonlinear Dynamic Systems: A Deep Neural State–Space Approach, PhD Thesis, Eindhoven Universtiy of Technology, 2024.


Please refer to the Unsteady Fluid Mechanics benchmark dataset as:

J. Decuyper, T. De Troyer, and M.C. Runacres, Canonical systems for unsteady fluid mechanics: A new family of nonlinear benchmarks, Technical Report, FLOW, Vrije Universiteit Brussel, Belgium, 2024.

Furthermore, a detailed description of the CFD simulations and validation against literature results can be found in:

J. Decuyper, Nonlinear state-space modelling of the kinematics of an oscillating cylinder in a fluid flow, PhD thesis, Vrije Universiteit Brussel, 2017.