Civil and Environmental Engineering Department, University of Virginia, Charlottesville, VA 22904-4742, United States.
J Mech Behav Biomed Mater. 2013 Dec;28:291-308. doi: 10.1016/j.jmbbm.2013.08.001. Epub 2013 Aug 24.
The Particle Swarm Optimization algorithm driven by a homogenized-based model is employed to target the response of three types of heart-valve chordae tendineae with different stiffening characteristics due to different degrees of waviness of collagen fibril/fiber bundles. First, geometric and material parameters are identified through an extensive parametric study that produce excellent agreement of the simulated response based on simplified unit cell architectures with the actual response of the complex biological tissue. These include amplitude and wavelength of the crimped chordae microstructure, elastic moduli of the constituent phases, and degree of microstructural refinement of the stiff phase at fixed volume fraction whose role in the stiffening response is elucidated. The study also reveals potential non-uniqueness of bio-inspired wavy microstructures in attaining the targeted response of certain chordae tendineae crimp configurations. The homogenization-based Particle Swarm Optimization algorithm, whose predictions are validated through the parametric study, is then shown to be an excellent tool in identifying optimal unit cell architectures in the design space that exhibits very steep gradients. Finally, defect criticality of optimal unit cell architectures is investigated in order to assess their feasibility in replacing actual biological tendons with stiffening characteristics.
基于均匀化模型的粒子群优化算法被用于研究三种具有不同纤维卷曲程度的心脏瓣膜腱索的响应特性，这些卷曲程度导致了不同程度的胶原纤维束变硬。首先，通过广泛的参数研究确定了几何和材料参数，这些参数产生了基于简化单元结构的模拟响应与复杂生物组织实际响应之间的极好一致性。这些参数包括卷曲腱索微结构的振幅和波长、组成相的弹性模量以及在固定体积分数下硬相的微观结构细化程度，其在变硬响应中的作用得到了阐明。该研究还揭示了在实现某些腱索卷曲结构的目标响应时，生物启发式波浪形微观结构的潜在非唯一性。通过参数研究验证的基于均匀化的粒子群优化算法随后被证明是一种在设计空间中识别具有非常陡峭梯度的最优单元结构的优秀工具。最后，研究了最优单元结构的缺陷临界性，以评估其用变硬特性替代实际生物腱的可行性。
