Special issue on High Reynolds Number Experiments

Abstract

Why do we need high Reynolds number experiments? This is a question I sometimes ask myself. You may have your own answer to this question, but those people who are doing numerical simulation, theorists and experimentalists should each have their own answer. In this special issue, the leading experts present their new ideas or original experiments in response to this question.Personally, I think that high Reynolds number experiments are necessary to seek novel physics in turbulence. For instance, we do not have much information about the Lagrangian quantities. You can understand this point by reading the article 'Why we need experiments at high Reynolds numbers' by Warhaft. High Reynolds number experiments are also indispensable to reveal the universality of turbulence. One famous example is Kolmogorov's similarity hypothesis; another is the logarithmic velocity profile derived by von Kármán. They become clearly satisfied as Reynolds number increases. But there have been many arguments over these problems even in this century, thus we still have to make an effort to reveal the nature of turbulence.Kolmogorov's idea is based on small scale physics; in this sense, Mouri and Hori's paper 'Vortex tubes in turbulence velocity fields at high Reynolds numbers' is a contribution to understanding how eddy size is defined and scaled. In contrast to the universality in the small scale limit, the large scale anisotropy effect is a key factor in consideringthe local isotropic condition even in grid turbulence. This point is discussed by Kurian and Fransson in 'Grid generated turbulence revisited'.The mean velocity profile over a flat plate in a zero-pressure gradient boundary layer is discussed with the help of a composite profile in 'Criteria for assessing experiments in zero pressure gradient boundary layers' by Chauhan et al. Related important physical quantities are computed, and how they are scaled against Reynolds number is discussed, analyzing the vast experimental database available, which is really interesting. The universal velocity profile is usually discussed through the normalization of the velocity and wall distance with the viscous scales. Therefore, an accurate determination of the wall shear stress is indispensable for this procedure. Zanoun et al tried to measure it by oil film interferometry in the paper 'Refined cf relation for turbulent channels and consequences for high-Re experiments', and Mori et al performed it by using the towing tank in 'Direct total skin-friction measurement of a flat plate in zero-pressure-gradient boundary layers'.In this issue we would like to introduce two new experimental facilities which can help us to tackle unresolved problems in turbulence. Talamelli et al report a response to my initial question.This is systematically mentioned in 'CICLoPE—a response to the need for high Reynolds number experiments', and they presentthe excellent new facility which is under construction. Yoshioka et al used the towing wind facility and simulated the highly complex flow between steady ground and a moving model, which is discussed in 'Measurement of ground effect and boundary-layer transition by towing wind tunnel'. It is a unique experiment. And this facility promises to proceed with high Reynolds number experiments.Finally, I will introduce the spirit of experiment, which is carved on the plate at Ettore Majorana Centenary, Erice, Italy: ''We experimentalists are not like theorists; the originality of an idea is not for being printed in a paper, but for being shown in the implementation of an original experiment. Patrick M. S. Blackett, London 1962.''