Flexible artificial heart valves were used as an example to illustrate analysis and design procedures that are applicable to flexible implants in general. Static loading, fatigue loading, and material deterioration are all taken into analytical consideration. Leaflet designs that included three thicknesses of six biomedical polymers (silicone rubber, polyurethane, nylon, polyethylene, Teflon, and Mylar) were investicated.

For design purposes the maximum static loading was taken to be the mitral valve loading, as determined by systolic blood pressure during physical exercise, for a case that would include 95% of the population. Material properties and leaflet shape were obtained from current literature. A geometrically nonlinear finite element method of solution was used to obtain membrane leaflet stres ses. The maximum two-dimensional leaflet stresses were simplified to equivalent uniaxial stresses with the use of von Mises' criterion of failure. These equivalent stresses were then compared with allowable polymer tensile strengths. Static loading was not found to be critical in design.

The fatigue loading was approximated by two discrete blood pressure levels (normal and sleeping) in the aortic valve and three discrete blood pressure levels (normal, sleeping, and exercise) for the mitral valve. The resulting complex cyclic stresses in the valve leaflet were simplified to an equivalent uniaxial, completely reversed, alternating stress by means of Sines' equation. The Palmgren-Miner theorem of cumulative damage was used to predict mean design life. Fatigue life was determined to be the critical design parameter for silicone rubber and Teflon.

Material deterioration was analyzed by means of structural reliability. The discrete random leaflet loading produces discrete random levels of applied stress. The tensile strength resistance of the materials was considered to be a random variable that follows a logarithmic normal distribution which is decreasing with time and consistant with current literature. The probability of failure due to tensile strength deterioration was calculated and plotted as a function of time. Varying coefficients of variation of material strength (0.1, 0.2, and 0.5) were considered for each leaflet thickness. Material strength deterioration was found to be the controlling design parameters for polyethylene, polyurethane, and nylon.

It was therefore shown that many structural failures in flexible artificial heart valves can be analyzed in a rational manner. Similarly, this approach can be applied to the structural analysis arid design of other flexible implants.