Search

Direct numerical simulations of turbulent channel flow at a shear Reynolds number of R e * = 360 in smooth and rough channels have been performed. Made of irregular undulations, surface roughness was such that the ratio of the channel half-height to the root mean square roughness height is equal to 48, and the root mean square and the maximum crest and trough heights are equal to 7.5 and 23 wall units, respectively. The simulation results confirm that turbulence in the outer layer is not directly affected by the rough surface. The roughness effects on the turbulent stresses, the mean momentum balance, and the budget of turbulence kinetic energy are confined to the layer between 0 and 25 wall units; beyond which the profiles collapse with those for smooth channels. In the roughness sublayer, the peak value of the streamwise normal stress is reduced, while the spanwise and wall-normal components are increased. The largest increase is for the Reynolds shear stress, resulting in a significant increase in the turbulence production near the wall, even though the velocity gradient is decreased. The kinetic energy budget shows that turbulence production dominates the mean viscous diffusion of turbulence kinetic energy, and both mechanisms are balanced by turbulent dissipation. The friction factor using the Colebrook–White correlation calculated by specifying the sand–grain roughness equal to the root mean square of the roughness height predicts the friction velocity and the bulk velocity accurately. The streaky structures that exist near smooth walls were observed to be broken by the roughness elements, leading to a denser population of coherent structures near the wall, which increases the velocity fluctuations. The coherent structures developed in the roughness layer do not seem to penetrate into the outer layer. [ABSTRACT FROM AUTHOR]

This paper presents Direct Numerical Simulation (DNS) results of a turbulent water flow in a channel at Reτ = 400 laden with 0.25 mm diameter air bubbles clustered near the wall (maximum void fraction of α = 8% at y+ ~ 20). The bubbles were fully resolved using the level set approach built within the CFD/CMFD code TransAT. The fluid properties (air and water) were kept real, including density, viscosity, and surface tension coefficient. The aim of this work is to understand the effects of the bubbles on near-wall turbulence, paving the way towards convective wall-boiling flow studies. The interactions between the gas bubbles and the water stream were studied through an in-depth analysis of the turbulence statistics. The near-wall flow is overall affected by the bubbles, which act like roughness elements during the early phase, prior to their departure from the wall. The average profiles are clearly altered by the bubbles dynamics near the wall, which somewhat contrasts with the findings from similar studies [J. Lu and G. Tryggvason, "Dynamics of nearly spherical bubbles in a turbulent channel upflow," J. Fluid Mech. 732, 166 (2013)], most probably because the bubbles were introduced uniformly in the flow and not concentrated at the wall. The shape of the bubbles measured as the apparent to initial diameter ratio is found to change by a factor of at least two, in particular at the later stages when the bubbles burst out from the boundary layer. The clustering of the bubbles seems to be primarily localized in the zone populated by high-speed streaks and independent of their size. More importantly, the bubbly flow seems to differ from the single-phase flow in terms of turbulent stress distribution and energy exchange, in which all the stress components seem to be increased in the region very close to the wall, by up to 40%. The decay in the energy spectra near the wall was found to be significantly slower for the bubbly flow than for a single-phase flow, which confirms that the bubbles increase the energy at smaller scales. The coherent structures in the boundary layer are broken by the bubbles, which disrupts the formation of long structures, reducing the streamwise integral length scale. [ABSTRACT FROM AUTHOR]