Design optimization of a cardiovascular stent with application to a balloon expandable prosthetic heart valve

[Display omitted] •A stent optimization strategy is proposed for an expandable prosthetic heart valve.•Surrogate models are used for numerically estimated response quantities of interest.•Failure is considered via a ductile fracture criterion calibrated for two materials.•Intuition provided via multi-objective optimization and plots of the Pareto frontier.•A single optimized design is proposed via aggregate objective function optimization. A cardiovascular stent design optimization method is proposed with application to a pediatric balloon-expandable prosthetic heart valve. The prosthetic valved conduit may be expanded to a larger permanent diameter in vivo via subsequent transcatheter balloon dilation procedures. While multiple expandable prosthetic heart valves are currently at different stages of development, this work is focused on one particular design in which a stent is situated inside of an expandable polymeric valved conduit. Since the valve and conduit must be joined with a robust manufacturing technique, a polymeric glue layer is inserted between the two, which results in radial retraction of the valved region after expansion. Design of an appropriate stent is proposed to counteract this phenomenon and maintain the desired permanent diameter throughout the device after a single non-compliant balloon dilation procedure. The finite element method is used to compute performance metrics related to the permanent expansion diameter and required radial force. Additionally, failure due not only to high cycle fatigue but also due to ductile fracture is incorporated into the design study through the use of an existing ductile fracture criterion for metals. Surrogate models are constructed with the results of the high fidelity simulations and are subsequently used to numerically obtain a set of Pareto-optimal stent designs. Finally, a single design is identified by optimizing a normalized aggregate objective function with equal weighting of all design objectives.