TU Delft Turbulent Wake Dataset

The TU Delft turbulent wake dataset is a classic benchmark in fluid mechanics, built from years of wind-tunnel experiments and simulations at Delft University of Technology. Covering cavity flow, cylinder wake and square-cylinder wake, it is dominated by experimental measurements (PIV/LDA) and serves as a reference for validating CFD solvers and surrogate-model accuracy.

What's included

• Lid-driven cavity (2D/3D): classic case, Re = 1,000 ~ 100,000
• Cylinder wake: time-averaged and fluctuating data at various aspect ratios and Re
• Square-cylinder wake: velocity, pressure and turbulent kinetic energy at various depth ratios
• Plane mixing layer: velocity profiles and vorticity

Specs

• Data types: experimental (PIV/LDA) + LES
• Variables: instantaneous & time-averaged velocity, TKE, Reynolds stress
• Format: ASCII / CSV with run parameters
• License: CC BY 4.0 (attribution to TU Delft required)

Use cases: CFD solver validation (OpenFOAM / ANSYS Fluent); turbulence-model evaluation (k-e vs k-w SST vs LES); surrogate accuracy on turbulence; teaching experimental methods and post-processing.