European Journal of Mechanics B/Fluids 58 (2016) 109–116 Contents lists available at ScienceDirect European Journal of Mechanics B/Fluids journal homepage: www.elsevier.com/locate/ejmflu Alternative interpretation of the Superpipe data and motivation for CICLoPE: The effect of a decreasing viscous length scale Ricardo Vinuesa a,b,∗ , Richard D. Duncan b , Hassan M. Nagib b a Linné FLOW Centre, KTH Mechanics, Osquars Backe 18, Stockholm, Sweden b Department of Mechanical, Materials and Aerospace Engineering (MMAE), Illinois Institute of Technology (IIT), Chicago, IL, 60616, USA article info Article history: Received 21 October 2015 Received in revised form 11 February 2016 Accepted 31 March 2016 Available online 19 April 2016 Keywords: Experimental limitations Wall-bounded turbulence Pitot tubes Superpipe CICLoPE abstract Pressurization and cryogenic conditions have been used in some experiments to change the kinematic viscosity ν of the flowing gas by many orders of magnitude in order to achieve high Reynolds number conditions in facilities of limited size. This leads to a substantial reduction of the viscous length scale ℓ ∗ = ν/u τ , as in the so-called Princeton ‘‘Superpipe’’ experiments. We demonstrate that the limited dimensions of the facilities and probes can lead to inaccuracies in the near-wall measurements for increasing Reynolds number. Specifically, a lack of accurate wall-normal probe positioning is simulated using three different datasets of wall-bounded turbulent flows. Relatively large errors in the overlap region parameters are observed for position errors of small physical magnitude that become greatly amplified in wall units as ℓ ∗ is reduced. This offers an alternative interpretation to some of the key findings reported by the Superpipe team, such as the increasing lower limit of the logarithmic region y + log,min , the existence of a power law region between the wall and the logarithmic layer, and the ‘‘mixing transition’’ phenomenon in wall- bounded turbulence. © 2016 Elsevier Masson SAS. All rights reserved. 1. Introduction High Reynolds numbers in wall-bounded turbulent flows are often reached by increasing the characteristic length and/or ve- locity scale. Both can usually be modified in boundary layer ex- periments, since the characteristic length is the distance from the leading edge x, and the characteristic velocity is the freestream ve- locity U ∞ . However, the characteristic length is fixed by the size of the facility in most fully-developed channel and pipe flows (as the height of the channel H and the pipe diameter D respectively), with only the bulk velocity U b and kinematic viscosity ν variable. To avoid compressibility effects, some experiments overcame the limited physical dimensions through manipulation of ν . In these experiments, the facility is pressurized well above atmosphere to reduce the value of ν significantly, so that very high Reynolds num- ber conditions can be obtained. This technique was used in the Superpipe experiments, where Reynolds numbers based on bulk velocity and pipe diameter up to Re D = 35 × 10 6 were achieved [1,2]. The mean velocity profile in this experiment was first stud- ∗ Corresponding author at: Linné FLOW Centre, KTH Mechanics, Osquars Backe 18, Stockholm, Sweden. E-mail address: rvinuesa@mech.kth.se (R. Vinuesa). ied by Zagarola and Smits [1], who presented measurements for Reynolds numbers from Re D = 0.031 × 10 6 to 35 × 10 6 using a 0.9 mm diameter probe. They reported the existence of two over- lap regions: a power law for 60 < y + < 500 and a logarithmic region for 600 < y + < 0.07R + when Re D > 0.4 × 10 6 . Here, R = D/2 is the pipe radius and y + is the wall-normal coordinate normalized by the viscous length scale ℓ ∗ = ν/u τ . The friction ve- locity u τ = √ τ w /ρ (where τ w is the mean shear stress at the wall and ρ is the fluid density) is used to normalize the mean velocity U , to form U + = U /u τ . For the classical law of the wall [3] or ‘‘log law’’ given by: U + = 1 κ ln(y + ) + B, (1) they reported the values κ = 0.436 and B = 6.15 for the constants. McKeon et al. [2] performed new experiments in the same Reynolds number range using a 0.3 mm diameter probe, and analyzed both sets of data using new static pressure corrections [4]. Their study showed significant changes from the previous analysis: a power law region was found for 50 < y + < 300, and the log region was found for 600 < y + < 0.12R + when Re D > 0.2 × 10 6 . The values of the constants were modified to κ = 0.421 and B = 5.6. Nagib and Chauhan [5] recently showed that the Superpipe experiments exhibit an increasing lower limit of http://dx.doi.org/10.1016/j.euromechflu.2016.03.010 0997-7546/© 2016 Elsevier Masson SAS. All rights reserved.