Hydrodynamic performance of mechanical prosthetic heart valve

Each year, more than 10,000 operations aimed at replacement of diseased heart valves by prosthetic devices are carried out in North America alone. Physiological compatibility, structural integrity and favorable hemodynamics represent three important criteria governing the design of a prosthetic heart valve. The thesis studies fundamental fluid characteristics of three widely used mechanical heart valve configurations, namely, the Starr-Edwards, Bjork-Shiley and St. Jude. In the beginning, the pertinent available literature is reviewed which clearly points out limitations concerning the available information and their presentation. This is followed by a detailed description of design, construction, calibration and Instrumentation of a steady flow glycerol-water solution tunnel and a pulsatile flow cardiac simulator. The former test facility is ideally suited for testing heart valves under fully open condition during which the maximum flow and pressure (energy) losses occur. The latter simulates the transient condition over a typical cardiac cycle rather precisely. Highly sensitive Barocel pressure transducing system, magnetic flowmeters, laser-doppler anemometer, and a microprocessor controlled waveform generator together with sophisticated data acquisition and processing system makes the facility ideal and unique for the purpose. Finally, the results of three distinct series of experiments with prosthetic valves using: (i) the steady flow glycerol-water solution tunnel; (ii) the steady flow in the cardiac pulse duplicator; and (iii) the pulsatile flow cardiac simulator; are presented and discussed. The significant contribution of the project lies in the fundamental data on pressure drop and its partial recovery; velocity profile, turbulence intensity, shear stress and their decay downstream; both in steady and pulsatile flow conditions. The results provide a comprehensive picture, fundamental insight and physical appreciation as to the hydrodynamic performance of prosthetic heart valves which would serve as reference for future development. Emphasis throughout is on the use of proper nondimensional parameters to make the information independent of test facilities, flow velocities, size of the models, etc., which should represent a welcome step forward. It would make comparison of results obtained by different investigators using different test-facilities possible. Based on the results following general conclusions can be made: (a) Nondimensional pressure drop and discharge coefficient results suggest the Starr-Edwards configuration to be fluid dynamically superior. (b) There is a significant and rapid recovery of pressure in the wake which depends on the Reynolds number and size of the downstream section. In the present study it was found to be as large as 24%! Hence, considering pressure drop immediately across a heart valve as a measure of its performance, as widely reported in literature, can be misleading. (c) The Starr-Edwards prosthesis has a relatively lower value for the maximum velocity and turbulence intensity and their rapid decay in the wake compared to the Bjork-Shiley and St. Jude valves. (d) Adjustment of parameters characterizing the cardiac network affect details of the cardiac cycle. (e) At the onset of systole all the valves show negative flow until the valve-closure is complete. The Starr-Edwards valve has the largest negative flow rate as well as the longest duration until its closure while the St. Jude valve shows the smallest amount of reversed flow over the shortest time. Negative flow is a significant parameter since the loss in volume must be compensated either by increasing the heart rate or the stroke volume. (f) All the valves show a decrease in Cp with an Increase in the Reynolds number. Thus the valve performance improves at higher Reynolds numbers. The degree of improvement depends on the valve configuration and is relatively smaller for the ball and cage geometry. (g) During the pulsatile flow study, the maximum velocity recorded for the Starr-Edwards valve, at a given downstream location, is essentially the same as that observed during steady flow case. On the other hand, the turbulence Intensity is distinctly lower. Similar trends were observed for the other two configurations. In general, the peak velocity and turbulence intensity for the St. Jude valve are smaller than those for the Bjork-Shiley case. (h) For the Starr-Edwards prosthesis, the sticking character of the ball may substantially alter the pressure-flow rate relation. The thesis ends with several recommendations for future work which are likely to be rewarding.Applied Science, Faculty ofMechanical Engineering, Department ofGraduat