Abstract We present a curated dataset comprising three-dimensional direct numerical simulations of Kolmogorov flow across multiple distinct parameter regimes. These simulations cover a wide range of Reynolds numbers and excitation modes, capturing diverse flow behaviours. The dataset is designed to support research in turbulence modelling, particularly for machine learning-based model development. Each case includes raw velocity fields and a Python-based interpolation code is provided for the interpolation of the velocity fields onto a structured grid in a format compatible with ParaView, accompanied by comprehensive metadata for ease of use. This open-access dataset aims to facilitate advances in turbulence modelling, a continuing and vital challenge in fluid mechanics research.
Kovács et al. (Thu,) studied this question.