Cycle-to-cycle variation effects on turbulent shear stress measurements in pulsatile flows

Abstract

Accurate evaluation of turbulent velocity statistics in pulsatile flows is important in estimating potential damage to blood constituents from prosthetic heart valves. Variations in the mean flow from one cycle to the next can result in artificially high estimates. Here we demonstrate a procedure using a digital, low-pass filter to remove the cycle-to-cycle variation from turbulence statistics. The results show that cycle-to-cycle variations can significantly affect estimates of turbulent Reynolds stress and should be either eliminated or demonstrated to be small when reporting pulsatile flow results.

Eulerian and Lagrangian description of a pulsatile, partially turbulent flow

“…In all cases, u' maximum is found within the central third of systole (the same of maximum flow rate); but in the main jet at h=20 mm, and in the secondary jet further downstream. This result agrees well with the previous literature (Tiederman et al 1988).…”

“…Radial mean velocity profiles agree very well in global shape with results reported by other authors (Phillips et al 1980;Chandran et al 1983;Tiederman et al 1988).…”

“…Whichever is the reason, it is a major cause for wrong evaluation of turbulence intensity (Tiederman et al 1988).…”

“…Several methods have been proposed to solve this problem; some from internal combustion engine researchers (Liou and Santavicca 1985;Obokata et al 1987) and also in the biomedical engineering literature (Tiederman et al 1988). Even though the philosophy underlying these works is partially useful, the specificity of each situation obliges to look for particular procedures giving a good discrimination among different sources of velocity variance.…”

A Comparison of Flow Field Structures of Two Tri-Leaflet Polymeric Heart Valves

“…26 Velocities were phase-averaged within each 20 ms time windows (each cardiac cycle of 860 ms was divided into 43 time intervals) for subsequent analysis. Approximately 21,500 measurements were made at each location, an average of approximately 500 measurements within each 20 ms time window.…”