Mechanical and Civil Engineering Seminar
Canonical wall bounded turbulent flows have received a lot of attention because of their inherent theoretical interest, but also because of their practical importance. Certain features of these flows can be obtained from first principles, such as the total shear stress distribution in channel flows, other features can be argued to be valid at high Reynolds numbers (such as the logarithmic mean velocity distribution or the constant stress region). Other features still under debate are the variation of the near wall peak of the variance of the streamwise fluctuation velocity with Reynolds number and the existence of an outer peak of the variance at high Re as well as the value of the von Kármán constant. To give affirmative answers to these and other questions very high Reynolds number data are needed. DNS can give a lot of details unattainable for experiments, but is limited to low Re, whereas Large Eddy Simulations (LES) can provide high Reynolds numbers but needs modelling of the near wall region. The problem for experiments is that it is hard to reach high Reynolds numbers and still retain sufficient spatial resolution and/or to have long enough sampling times for statistically converged results. In this talk some new results extrapolated from simulations and experiments about high Re behaviour are discussed, common measurement technique fallacies are exposed, as well as the type of earthbound facility that is needed to be able to resolve some of these outstanding questions are introduced.