Grant: Discovery Grants Program - Individual

Incompressible zero-pressure-gradient boundary layer over a smooth flat-plate and incompressible flow through a smooth pipe are two fundamental building blocks of modern fluid mechanics. In the meantime, they are also the limiting cases and calibration benchmarks of many complicated engineering flows such as the boundary layer on an aircraft wing. Experiments have provided indirect evidence for the organization of large-scale vortex structures in the form of hairpin packets in the boundary layer and pipe flow. Theoreticians also hypothesized the existence of hairpin forest. Perplexingly, neither dominant hairpin packet nor hairpin forest could be detected from previous incompressible turbulent boundary layer and pipe flow simulations. We have recently achieved a breakthrough on direct numerical simulation (DNS) of flat-plate boundary layer with a natural appearance of young hairpin forest (Wu and Moin, Journal of Fluid Mechanics, Vol. 630, 2009). Top experts of the world's turbulence and boundary layer communities have expressed strong interest and have called for prompt targeted in-depth investigations (Marusic, Journal of Fluid Mechanics, Vol. 630, 2009). Here, we propose to perform fundamental investigations on the statistics and coherent structures of the flat-plate boundary layer. We also propose to investigate the large coherent structures of spatially-developing pipe flow from laminar through transition to fully-developed turbulent state. The proposed research will provide decisive and faithful resolutions to an array of pipe flow and flat-plate boundary layer questions that have been distilled from decades of experimental and theoretical investigations by many renowned groups in the world. Specific issues that will be examined include the kinematic and dynamic characteristics of hairpin forest, the physical mechanism of the very-large-scale motions indirectly observed in experiments, as well as a critical comparison of coherent structures obtained from the present DNS and previous particle image velocimetry experiments. This research will also impact the methodology of turbulence simulation.