Study on the Accuracy of Structural and FSI Heart Valves Simulations

• Published in: Cardiovascular engineering and technology 2018-12, Vol.9 (4), p.723-738
• Main Authors: Luraghi, Giulia, Migliavacca, Francesco, Rodriguez Matas, Josè Fèlix
• Format: Article
• Language: English
• Subjects: Accuracy; Biomechanical Phenomena; Biomedical Engineering and Bioengineering; Biomedicine; Cardiology; Computer simulation; Damping; Domains; Elastic Modulus; Engineering; Finite Element Analysis; Finite element method; Heart; Heart Valve Prosthesis; Heart valves; Heart Valves - anatomy & histology; Heart Valves - physiology; Hemodynamics; Humans; Kinematics; Mathematical analysis; Mathematical models; Models, Cardiovascular; Numerical Analysis, Computer-Assisted; Patient-Specific Modeling; Physiological effects; Predictive Value of Tests; Prosthesis Design; Reproducibility of Results; Simulation; Stress, Mechanical; Valve leaflets
• ISSN: 1869-408X; 1869-4098
• DOI: 10.1007/s13239-018-00373-3

Purpose The performance of heart valves, either native or artificial, can be evaluated by means of finite element analyses, either from a structural or a fluid–structure interaction (FSI) point of view. The latter captures the coupling between the valve leaflets and the blood in a more realistic way. The selection of the appropriate finite elements approach for the model is the first and fundamental step to achieve accurate simulations. The aim of this work is to investigate the influence of the type, formulation, size, and shape of the elements in heart valves simulations. Methods The effects related to the choice of the finite elements-shell or solid- in structural and FSI simulations were analyzed. In particular, the analysis of grid convergence on both the structure and fluid domains, the influence of the element typology, formulation and damping factor in an idealized three-leaflets valve model loaded with physiological pressure conditions were investigated. Results Stress values and valve kinematics results confirmed the importance of performing a proper verification process for selecting the most appropriate elements with the optimal accuracy to computational cost ratio. Conclusion In this regard, our results indicate the quadrangular shell with reduced integration and viscous hourglass control to be the best choice to model heart valves. If a solid discretization is required, quadratic hexahedral elements with full integration are also acceptable. Finally, our results show that the damping coefficient needs to be carefully selected in order to smooth out the high frequency modes of the structure without introducing excessive numerical artificial viscosity.