Amplitude modulation of velocity and wall shear stress in turbulent channel flow with hemispherical roughness

"Near-wall turbulence cycles tend to be self-sustained in low Reynolds number wall-bounded turbulent flows. However, large-scale structures that reside in the outer region of the flow grow more energetic as Reynolds number increases and impart strong ""footprints"" on the near-wall turbulence. In addition, recent experiments in high Reynolds number smooth-wall turbulent boundary layers have shown that these outer-layer, energy-containing large-scale events amplitude-modulate (AM) the small-scale activities in the near-wall region. While smooth surfaces are ideal for the theoretical modeling of wall turbulence, rough surfaces are more widely encountered in practical engineering problems. The presence of roughness elements greatly modifies the drag, heat-transfer and aerodynamic characteristics of surfaces, which involves strong interplay between the pressure and viscous effects but is difficult to measure experimentally in the intermediate vicinity of rough surfaces. In this regard, this dissertation focused on performing direct numerical simulations (DNS) of turbulent channel flow with hemispherical roughness on both walls, which fully resolve all relevant turbulent scales and allow accurate measurements of the velocities and wall shear stresses. Though the analysis of amplitude modulation is first studied over smooth surfaces, outer-layer similarity between the smooth- and rough-wall turbulent flows suggests that this analysis is also applicable to rough walls and present DNS data showed enhanced effects of AM within the roughness sublayer, compared to the smooth-wall baseline. The physics-based predictive models, pioneered by Mathis et al. (2011, 2013) using the framework of AM, were extended to predict all three velocity components and wall shear stress in the present rough-wall flows. The anisotropy of wall turbulence was also incorporated which enabled us to produce improved predictions of the statistics of correlated velocities that the original model predicted unsatisfactorily. The dynamical responses of wall shear stress to outer large-scale structures in rough-wall flows were further investigated in terms of roughness-cell-averaged statistics. In particular, it was shown that the variance and roughness-cell dimensions are connected by a power-law-based scaling relation, which can be useful in advancing turbulence modeling in the context of ""wall-modeled"" large-eddy simulations by incorporating various surface attributes in more practical engineering applications."

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Theoretical & Applied Mechans

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2020

Author dc:creator

• Wu, Sicong

Contributors dc:contributor

• Pantano, Carlos
• Fischer, Paul
• Christensen, Kenneth T
• Villafane, Laura

• Turbulent boundary layer
• Turbulence modeling

Statement dc:rights

• Copyright 2020 Sicong Wu

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/107889

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/107889