Turbulent drag reduction using rigid microfibers

Using a combination of direct numerical simulation (DNS) and experiment, the mechanism by which rigid fibers cause turbulent drag reduction has been investigated. The fibers are assumed to have a length smaller than the scale of the spatial variations of the velocity field and negligible inertia; these "microfibers" include rigid biopolymers and colloids. Fiber extra stresses are modeled using a constitutive equation related to the moments of the fiber orientation.; Using DNS of a fiber drag-reduced turbulent channel flow, the effect of fiber rheology and the drag reduction mechanism are determined. Drag reductions of up to 26% are obtained, with semi-dilute, non-Brownian fibers being most effective. This result suggests that additive elasticity is not necessary for drag reduction. Fiber stresses are correlated to inter-vortex, biaxial extensional flow regions, with the spatial gradients of these stresses generating opposition forces that destroy nearby vortices. The cross-flow plane stress components are shown to be responsible for the drag reduction effect, with the largest contribution from small fluctuations in the x-z plane shear stress. The dynamics of stress generation are investigated using conditional statistics of fiber stress and orientation along Lagrangian pathlines in the drag-reduced flow. Fibers are shown to align in the cross-flow plane, generate stress and realign in the flow direction in a process taking approximately 4 strain units, defined using the strain rate along the fiber. The stress fluctuations end because nearby vortices are weakened or destroyed and not because of relative motion of the vortices and fibers.; Simulations of polymer-fiber mixtures in the turbulent channel flow, motivated by observations of synergistic behavior in the literature, were performed. A synergy effect was only obtained in a small computational domain in which the turbulence structure is substantially modified. Lastly, drag reduction using a rigid rod-like polymer was experimentally and numerically investigated in a zero pressure gradient turbulent boundary layer. Using particle image velocimetry, drag reductions of up to 15% were observed using injection of the polysaccharide scleroglucan. The turbulence statistics obtained using DNS compare favorably to the experimental results, with differences due to the homogeneous concentration distribution assumed in the DNS.