# Summary

Dataset from a turbulent channel flow simulation at friction Reynolds number Re_tau = 20,000. The modelling approach considers a single large eddy simulation (LES) combined with an array of non-space-filling quasi-direct numerical simulations (QDNSs). The dataset supports the paper by Sandham et al. (Accepted):

* Neil D. Sandham, Roderick Johnstone, Christian T. Jacobs (Accepted). Surface-sampled simulations of turbulent flow at high Reynolds number. International Journal for Numerical Methods in Fluids.

The LES was performed with a structured grid of 120 x 120 x 90 solution points. Each wall comprised 4 x 4 QDNS blocks (giving a total of 4 x 4 x 2 QDNSs), with each QDNS being performed with a 32^3 structured grid.

# Dataset

The dataset can be found in the directory `les-qdns-parallel-Re20000-4x4` within the `les_qdns_parallel_Re20000.tar.gz` file, which comprises the following files and directories:

* `*_dns.dat`: Mean velocity and mean RMS velocity components, averaged over all of the QDNS blocks, from k=1 to k=Nz where Nz is the number of points in the z-direction. This is in plain text format and can be read with a standard text editor.

* `*_les.dat`: Mean velocity and mean RMS velocity components from the LES, from k=1 to k=Nz where Nz is the number of points in the z-direction. This is in plain text format and can be read with a standard text editor.

* `les/fields_in.bin`: Final solution fields (all three velocity components and pressure) in binary format from the LES. This can be read with, for example, a Python program using the `FortranFile` class from the `sympy.io` module.

* `dns_lower_x_y/fields_in.bin`: Final solution fields (all three velocity components and pressure) in binary format from the QDNS block on the lower wall with identifier (x,y). This can be read with, for example, a Python program using the `FortranFile` class from the `sympy.io` module.

* `dns_upper_x_y/fields_in.bin`: Final solution fields (all three velocity components and pressure) in binary format from the QDNS block on the upper wall with identifier (x,y). This can be read with, for example, a Python program using the `FortranFile` class from the `sympy.io` module.

# Acknowledgements

This work was supported by the following grants:

* ExaFLOW: Enabling Exascale Fluid Dynamics Simulations (European Commission Horizon 2020 project grant 671571).
* The UK Turbulence Consortium (EPSRC grant EP/L000261/1).

The authors also acknowledge support of iSolutions at the University of Southampton and the use of the Iridis 4 cluster.

# Copyright statement

Copyright (C) 2016, 2017 Neil D. Sandham, Roderick Johnstone, Christian T. Jacobs