Search

A three-dimensional and pulsatile blood flow in a human aortic arch and its three major branches has been studied numerically for a peak Reynolds number of 2500 and a frequency (or Womersley) parameter of 10. The simulation geometry was derived from the three-dimensional reconstruction of a series of two-dimensional slices obtained in vivo using CAT scan imaging on a human aorta. The numerical simulations were obtained using a projection method, and a finite-volume formulation of the Navier-Stokes equations was used on a system of overset grids. Our results demonstrate that the primary flow velocity is skewed towards the inner aortic wall in the ascending aorta, but this skewness shifts to the outer wall in the descending thoracic aorta. Within the arch branches, the flow velocities were skewed to the distal walls with flow reversal along the proximal walls. Extensive secondary flow motion was observed in the aorta, and the structure of these secondary flows was influenced considerably by the presence of the branches. Within the aorta, wall shear stresses were highly dynamic, but were generally high along the outer wall in the vicinity of the branches and low along the inner wall, particularly in the descending thoracic aorta. Within the branches, the shear stresses were considerably higher along the distal walls than along the proximal walls. Wall pressure was low along the inner aortic wall and high around the branches and along the outer wall in the ascending thoracic aorta. Comparison of our numerical results with the localization of early atherosclerotic lesions broadly suggests preferential development of these lesions in regions of extrema (either maxima or minima) in wall shear stress and pressure. Keywords: Aortic Arch, Biofluid Mechanics, Scientific Computing, Direct Simulation, Atherosclerosis

Highly unsteady three-dimensional flows in curved pipes with significant variation of flow geometry and flow parameters are studied. Using improvements in computational efficiency, detailed knowledge concerning flow structures is obtained. The numerical solutions of the Navier-Stokes equations have been obtained with a variation of the projection method, and the numerical method was enhanced by new algorithms derived from the physics of the flow. These enhancements include a prediction of the flow unsteady pressure gradient based on fluid acceleration and global pressure field corrections based on mass flow. This new method yields an order of magnitude improvement in the calculation's efficiency, allowing the study of complex flow problems. Numerical flow simulations for oscillating flow cycles show that the curved pipe flows have a significant inviscid-like nature at high values of the frequency parameter. The shape of the velocity profiles is strongly influenced by the frequency parameter, whereas the influence of variations on the pipe cross-sectional area is shown to be rather weak. For large values of the frequency parameter the flow history strongly influences the low mass flow part of the cycle leading to highly unusual velocity profiles. The wall shear stress is studied for all the flows calculated. Our results show that wall shear stress is sensitive to area constrictions, the frequency parameter, as well as the shape of the entrance profile. [DOI: 10.1115/1.1400748]

Equations of compressible fluids are solved by using the particle method in accordance with a grid-free numerical technique. The time-dependent one-dimensional phenomena derived from the linear wave equation solutions are simulated numerically by the study model. Steady shock wave and interactions between the linear waves and the nonlinear shock waves are also simulated. The experiment also simulates compressible fluids by the particle method. The particle method is also modified to solve the Burgers equation and the equations of finite waves, with and without the diffusion terms.

Gill rakers have been assumed to serve as either (1) a sieve that accumulates particles larger than the mesh size or (2) a sticky filter that retains particles encountered via a hydrosol filtration mechanism. However, our video endoscopy (30 - 250 Hz) in the oral cavities of freely-swimming fish during pump suspension feeding demonstrates that the rakers of species belonging to three families (gizzard shad, Dorosoma cepedianum; goldfish, Carassius auratus; ngege tilapia, Oreochromis esculentus) do not function as a sieve or as a sticky filter. The gill rakers of pump suspension-feeding Sacramento blackfish (Orthodon microlepidotus) have been hypothesized to serve as a barrier to fluid flow, directing particle-laden water to the sticky oral roof where particles are trapped (Science 251: 1346-1348, 1991). However, our use of computational fluid dynamics to simulate flow in the vicinity of rakers at intermediate Reynolds numbers (250 - 500) shows that the rakers of pump suspension-feeding fishes are leaky, never serving as a barrier to flow. Based on these results, we propose a new model of filtration in suspension-feeding fishes. Supported by NSF grants to SLS and AYC.

The suction feeding fish generates a flow field external to its head that draws the prey towards the mouth. To date there are very few empirical measurements that characterize the fluid mechanics of suction feeding, particularly the temporal and spatial patterns of velocity and pressure of water in front of the fish. To characterize the flow in front of suction feeding bluegill (Lepomis macrochirus), measurements with high spatial (1 mm) and temporal (500 Hz) resolution were made with Particle Image Velocimetry (PIV) (3 individuals, 45 feedings). A novel method of tracking individual seed particles was also used to determine the parcel of water that the fish ingests. A Computational Fluid Dynamics (CFD) model, which calculates pressure everywhere in the flow and can be used for parametric studies with relative ease, was used to complement the measurements. PIV and particle tracking show that water is drawn from all around the mouth. Fluid velocity decreases exponentially with distance from the mouth and is only significant (>5% mouth speed) within roughly 1 mouth diameter (D) of the fish. CFD shows that generated suction pressure decays even more quickly than velocity, increasing to within 5% of ambient at 0.5 D. CFD parametric studies also exhibit similarity of flows for varying combinations of ram, suction, and mouth size, suggesting some generality of the empirical results. Comparison of time resolved fluid velocity measurements at a constant distance in front of the fish (5 mm, 0.5 D) in gape kinematics shows that peak fluid speed occurs at about 80% of peak gape. The magnitude of peak fluid speed during each strike decreases as a function of time to peak gape ([R.sup.2]=0.87), demonstrating a relationship between kinematics and fluid dynamics. Supported by NSF IBN-0326968.

Suction feeding is the most commonly used mechanism of prey capture in fishes and other aquatic vertebrates, but little is known about the temporal or spatial patterns of water flow generated outside the fish's head during suction feeding or how this flow influences prey capture performance. Using computational fluid dynamics techniques, we developed a three-dimensional computer model of suction feeding that we used to explore (1) the spatial distribution of flow velocity in front of the mouth, (2) the nature of the bow wave that is created in front of a forward swimming predator, and (3) the relative importance of drag and acceleration reaction in generating the forces that prey experience as a result of these flows. The velocity of water toward the mouth of a suction feeding fish diminishes greatly within a very short distance from the mouth of the fish, approximately proportional to the inverse of distance to the third power. This distance increased by only modest amounts when the fish greatly increased the suction velocity at its mouth opening. Forward body movement creates water movement away from the mouth of the fish (a bow wave) that must be overcome by suction in order to yield a net movement of water toward the mouth of the fish. We propose a new function for jaw protrusion: it ameliorates this effect of the bow wave on prey capture by moving the mouth opening (the suction source) out of the influence of the bow wave. During suction feeding, acceleration reaction force may dominate the total forces exerted on prey (depending on flow conditions and the size of the prey), suggesting that both water velocity and acceleration are important determinants of suction feeding performance.