Crack propagation and failure in idealized bicuspid leaflets under pulsatile flows using a coupled peridynamics–lattice Boltzmann–immersed boundary framework
Sarah Davidson, Ya Zhang, Yonghao Zhang, Sina HaeriFluid–structure interaction becomes substantially more complex when the solid can fracture: failure alters the geometry exposed to fluid loading, while the modified flow redistributes the forces driving further damage. Multi-leaflet systems such as biological and artificial valves are a canonical setting for studying this fluid–breakable-structure interaction, since damage, perforation, or detachment progressively degrades their flow-control function. Breakable bicuspid leaflets are simulated with a strongly coupled bond-based peridynamics–immersed boundary–cascaded lattice Boltzmann framework, varying the Young's modulus E and critical stretch ratio sc to quantify how fragility alters leaflet motion, flow features, and function. When damage is allowed, behavior falls into four outcomes: normal operation, partial failure, failure during opening, and failure during closure, with the most common mode being failure during closure where near-wall fracture initiates on the outlet side and leads to detachment from the wall. For sc≳0.35, normal operation is recovered and leaflet displacement in the flow direction remains consistent, although damage can still reduce leaflet range of motion and alter the flow. Low E and sc cause failure during opening under slight deformation, and partial failure yields asymmetric dynamics with a vortex behind the remaining leaflet and a reverse jet due to loss of sealing. At fixed E, sc chiefly sets failure timing; E chiefly sets failure mode and damage severity. However, no sweeping statements link damage levels to E or sc and the failure mechanism must be considered before judging the significance of any parameter. These results link specific fracture modes with distinct flow features in a two-dimensional brittle bicuspid leaflet. However, these findings are primarily qualitative and cannot be used directly to predict in vivo behavior or the mechanical response of soft tissue.