Energy dissipation scaling in uniformly sheared turbulence.

The rate of turbulent kinetic energy dissipation in spatially developing, uniformly sheared turbulence is examined experimentally. In the far-downstream fully developed region of the flow, we confirm that the dissipation parameter C(ɛ) is constant. More importantly, however, we find two upstream regions where this parameter could be scaled with the local turbulent Reynolds number as C(ɛ)=ARe(λ)(α); the exponents in these two regions are, respectively, α=-0.6 and 0.5. The observed changes in scaling laws are explained by consideration of structural changes in the turbulence.

Relative dispersion of a passive scalar plume in turbulent shear flow.

Relative dispersion of a passive scalar plume was investigated in uniformly sheared, nearly homogeneous, turbulent flow with Reλ≈150 using planar laser-induced fluorescence. Mean concentration maps were determined both in the laboratory frame and in a frame attached to the instantaneous center of mass of the plume cross section. The distance-neighbor function had a shape that was compatible with Richardson's expression. The mean square particle separation, two estimates of which were found to be nearly identical, had a streamwise evolution that was consistent with Richardson-Obukhov scaling with a Richardson's constant of g=0.35. Batchelor scaling was also consistent with a wide range of the results.

Adaptation of a rabbit myocardium material model for use in a canine left ventricle simulation study.

The myocardium of the left ventricle (LV) of the heart comprises layers of muscle fibers whose orientation varies through the heart wall. Because of these fibers, accurate modeling of the myocardium stress-strain behavior requires models that are nonlinear, anisotropic, and time-varying. This article describes the development and testing of a material model of the canine LV myocardium, which will be used in ongoing simulations of the mechanics of the LV with fluid-structure interaction. The model assumes that myocardium deformation has two extreme states: one during which the muscle fibers are fully relaxed, and another during which the muscle fibers are fully contracted. During the second state, the "total" stresses are assumed to be the sum of "passive" stresses, which represent the fully relaxed muscle fibers, and "active" stresses, which are additional stresses due to the contraction of the muscle fibers. The canine LV myocardium is modeled as a transversely isotropic material for which material properties vary in the fiber and cross-fiber directions. The material behavior is considered to be hyperelastic and is modeled by a strain-energy density function in a manner that is an adaptation of an approach based on measurements of the stress-strain behavior of rabbit LV myocardia. A numerical method has been developed to calculate suitable parameter values for the passive material model using previous passive canine LV myocardium stress measurements and taking into account existing physical and numerical constraints. In the absence of published measurements of total canine LV myocardium stresses, a method has been developed to estimate these stresses from available passive and total rabbit LV myocardium stresses and then to calculate active material parameter values. Material parameter values were calculated for passive and active canine LV myocardium. Passive stresses calculated using the model compare well to previous stress measurements while active stresses calculated using the model compare well with those approximated from rabbit measurements. The adapted material model of the canine LV myocardium is deemed to be suitable for use in simulations of the operation of both idealized and realistic canine hearts. The estimated model parameter values can be easily revised to more appropriate ones if measurements of active canine LV myocardium stresses become available. The extension of this material model to a fully orthotropic one is also possible but determination of its parameters would require stress-stretch measurements in the fiber and both cross-fiber directions.

Numerical simulations of blood flow in artificial and natural hearts with fluid-structure interaction.

This article describes two ongoing numerical studies of fluid-structure interaction in the cardiovascular system: an idealized pulsatile ventricular assist device (VAD), consisting of two fluid chambers separated by a flexible diaphragm; and blood flow and heart wall motion during passive filling of a canine heart. Simulations have been performed for the VAD and compared with the results of a previous study and to our own preliminary experimental results. Detailed measurements of the flow field in the VAD model and additional simulations are in progress. Preliminary simulations using both an idealized model of the natural heart as well as a realistic model have identified the limitations of the current numerical methods in dealing with large three-dimensional deformations. Ongoing research aims at extending the range of simulations to include large deformations and to incorporate an anisotropic material model for the heart wall to account for the muscle fibers.

Kinematics of the ankle joint complex in snowboarding.

Because snowboarders are known to injure their ankles more often than Alpine skiers, it has been postulated that stiffer snowboard boots would provide better protection to the ankle than current soft boots do. Snowboarders are also known to injure their front ankle more often than their back ankle, presumably because of the asymmetrical rotations of the ankles due to asymmetrical binding adjustement. To test these hypotheses, we measured the kinematics of the feet and legs of 5 snowboarders wearing soft boots and stiffer step-in boots during snowboarding maneuvers using an electromagnetic motion tracking system. The results were expressed in anatomically relevant rotations of the ankle joint complex, namely dorsi-/plantar flexion, eversion/inversion, and internal/external rotation. The measured ankle rotations show differences in the movement patterns of the front and back legs. Step-in boots were shown to allow less dorsiflexion, eversion, and external rotation than softer boots, possibly explaining why they are associated with a lower rate of fractures of the talus than soft boots.