H. Cao, P. Schmidt, J. Davidson, Edwards Lifesciences, Irvine, CA
Conventional life prediction is based on deterministic approach where a pass/fail result is arrived by comparing the load carrying capacity of the structure to the actual service loads. The conservativeness of the design is often expressed in terms of a safety factor. Because of the variety of assumptions made in the derivation of the material fatigue propety and in the analysis of the service stresses for the actual device, the computed safety factor does not convey an accurate picture of the robustness of the design. This presentation presents a typical case for predicting the durability performance of stentless heart valve wireform using statistical analysis where design robustness can be assessed by reliability. Bending fatigue testing was conducted using square cross-sectioned beam specimens in a simulated physiological environment. Tests were conducted at 30 Hertz for more than 600 million cycles. The fatigue life data was analyzed with Weibull statistics. An equivalent strain description of the mean strain effect is investigated. The statistical distribution parameters were obtained by using a maximum likelihood estimation methodology. Our research showed that the fatigue life data of Nitinol could be described by Weibull distribution. Constant reliability and constant life diagrams are developed, providing a practical methodology for life analysis involving the application of Nitinol in implantable devices where long-term structural integrity and reliability are desired.
Summary: Conventional life prediction is based on deterministic approach where a pass/fail result is arrived by comparing the load carrying capacity of the structure to the actual service loads. The conservativeness of the design is often expressed in terms of a safety factor. Because of the variety of assumptions made in the derivation of the material fatigue propety and in the analysis of the service stresses for the actual device, the computed safety factor does not convey an accurate picture of the robustness of the design. This presentation presents a typical case for predicting the durability performance of stentless heart valve wireform using statistical analysis where design robustness can be assessed by reliability using Weibull statistical life distribution.
