Fluid Researchers Gathered in Syracuse for DFD Meeting
New work in simulation techniques, turbulence, and the fluid dynamics of physical oceanography were among the highlights of the 1996 fall meeting of the APS Division of Fluid Dynamics, held 24-26 November in Syracuse, New York. More than 800 contributed papers were presented, in addition to several invited lectures and a mini-symposium on low-temperature superfluids. The meeting also featured the 14th Annual Gallery of Fluid Motion, an exhibit of contributed photographs and videos of experimental fluid dynamics. Outstanding entries, selected for originality and their ability to convey and exchange information, will appear in the September 1997 issue of Physics of Fluids.
Over the last 10 years, direct numerical simulation (DNS) of turbulence has emerged as a powerful technique in this area of research. According to Parviz Moin of Stanford University and NASA- Ames, recipient of the 1996 Fluid Dynamics Prize, DNS is particularly viable for low-to-moderate Reynolds numbers because of the disparity in the range of scales. For complex turbulent flows with higher numbers, large eddy simulation (LES) is experiencing a breakthrough with the advent of the dynamic subgrid scale procedure in 1990, a new approach to turbulence modelling where one actually computes, rather than prescribes, the model coefficients. However, outstanding issues remain concerning the constitutive equations for large eddies and the effects of numerical and subgrid scale modeling errors on large-scale turbulence statistics.
The application of numerical simulations in fluid dynamics has become commonplace, but questions still remain about the relation of such simulations to physical phenomena, as well as to rigorous solutions of partial differential equations, according to Richard Peskin of Rutgers University. In a Monday afternoon session, he discussed the role of symbolic computation (such as computer algebra and related paradigms) as a pre-processing tool to aid in establishing the correctness and completeness of numerical solutions, as well as comprehending numerical results, using the example of the dynamics of a two-dimensional cylindrical wake flow.
Computational aeroacoustics (CAA) involves the numerical simulation of the generation and radiation of sound by unsteady flows. Noise predictions prior to the advent of high performance computers were based on acoustic analogies, which are not readily applicable to problems with complex geometries. Initially the CAA technique was applied to the development of algorithms for discretization and boundary treatments. More recently, scientists at Penn State University have attempted to apply CAA methodologies to more practical problems, necessitating the use of parallel computers to adequately produce the three-dimensional unsteady simulations.
Purely Elastic Instabilities
Purely elastic instabilities in viscometric flows are instabilities that are present in the absence of fluid inertia, because they occur solely as a result of the elasticity of the flowing fluid. Research activity in this area has burgeoned over the last decade because these flow instabilities have practical implications for rheometry. Scientists at Stanford University have demonstrated that such work also has application in understanding new instabilities in nonviscometric flows as well, using local linear stability analysis to theoretically suggest that flow is inelastically unstable for all eccentricities. The Stanford team then conducted flow visualization experiments of the viscoelastic flow between eccentric cylinders, using a solution of high molecular weight polybutene dissolved in a viscous solvent. A related study examined the occurrence of the phenomenon in recirculation flows.
Researchers at Cornell University have concluded that flow-induced microstructure has a strong influence on the rheology of suspensions of non-Brownian fibers and thermal, electrical and mechanical properties of injection-molded composite materials. Donald Koch and his colleagues applied slender-body theories and simulations to describe the hydrodynamic interactions among the fibers, as well as to predict their properties. They then investigated the dynamic evolution of the microstructure during flow, finding that at modest concentrations, fibers change their orientation due to hydrodynamic interactions mediated by the fluid. At higher concentrations, direct mechanical contacts among the fibers control both the microstructure and the thermal and rheological properties of the material.
Paths to Transition in Open Flow Systems
With minor exceptions, transition to turbulence in open flow shear layers is forced by external disturbances which enter the system across the inflow and lateral boundaries, according to Eli Reshotko of Case Western Reserve University. In turn, the shear layers respond by converting these environmental disturbances into internal disturbances by a receptivity process which filters forcing motions. Reshotko found that the phenomenon of transient growth is also subject to such a process, leading to numerous transition scenarios dependent on the nature and intensity of external disturbances penetrating the system.
On Tuesday afternoon, Robert Kerr of the National Center for Atmospheric Research in Boulder, Colorado presented numerical evidence consistent with analytic bounds on singular behavior in the three-dimensional incompressible Euler equations, extending the finding to viscous, turbulent dynamics. According to Kerr, in viscous flows the properties of the antiparallel vortex interaction calculations which show singular behavior can be used to identify the three steps by which full- developed turbulence might form from smooth initial conditions. While admitting that taken alone, none of the flows he cited as examples approach the Reynolds numbers needed for convincing evidence, he noted that in each case he observed (1) formation of vortex sheets and suppression of singular behavior, followed by (2) a strong increase in peak vorticity, and finally (3) a peak in enstrophy.
Fluid Dynamics in Physical Oceanography
On Monday afternoon, Jack Whitehood of the Woods Hole Oceanographic Institute reviewed current research in the fluid dynamics of physical oceanography. "The ocean is the most massive fluid body in contact with human kind, and understandably its behavior covers an immense range of length and time scales," he said, adding that the largest and longest time scales are linked to ideas about the ocean's evolution. While research indicates that scales governing temperature and salinity and heat transfer laws may play a role in climate issues, more work needs to be done in this area, as well as in research on eddy flux mechanisms, which are only partly understood.
According to Whitehead, much is known about the ocean's general circulation. Vorticity conservation laws govern a rotating spherical shell of water, and these are also manifest in today's oceans. Large boundary layer currents are found the western sides of basins and at the equator, and wind and driving have been observed to result in circulation patterns, ventilated regions, and constant potential vorticity gyres. At shorter length and time scales, fronts, jets and mesoscale eddies are numerous, balanced between rotational and inertial effects, as well as the possibility of friction, dissipation and mixing.