S4E3

Abstract

Constitutive models are of fundamental importance to many biomedical problems. Biological tissues such as arterial tissue are often comprised of collagen fibers embedded in a ground matrix, and the fibers are usually dispersed in orientation. Existing structure-based constitutive models such as the widely used Gasser–Ogden–Holzapfel (GOH) model usually separate a general fiber distribution into two/several fiber families. However, the number of fiber families in some tissues (e.g., aortic valve leaflets) may be difficult to identify. Moreover, the number of fiber families may also vary for the same kind of tissues (e.g., arterial tissue) from different positions. Applications of the GOH model to tissues with unclear number of fiber families may be physically inconsistent. This talk will present a novel unified-fiber-distribution (UFD) model for biological tissues which considers the fibers as a unified distribution, rather than separating the fiber distribution into two/several fiber families. The UFD model with 4 material parameters obtained better performance of modeling biomechanical behaviors of arterial tissues than the widely used GOH model with 5 parameters. The talk will also present a new 2nd kind of Poisson effect which characterizes the deformation coupling of materials under biaxial tension. The 2nd kind of Poisson effect explains why the UFD model with fewer parameters exhibits better performance than the GOH model. The UFD model can be applied to biological tissues of which the fiber distributions and the number of fiber families are not known. The 2nd kind of Poisson effect is a useful and essential supplement to the traditional Poisson effect and can be applied to investigate the accuracy of nonlinear constitutive models.

Abstract

Leonardo da Vinci reported the first ever mechanical test of an iron wire. Few hundred years later, Galileo Galilei took a giant leap forward by defining the concept of strength and performing rigorous mechanical characterization experiments. Over the past centuries, a variety of mechanical testing methods have been established for characterizing different mechanical properties of materials. However, accurate and easy-to-use methods for characterizing materials at small scales and/or arbitrary shapes are still largely missing.

In this talk, I will present two methodologies I have developed to measure fracture properties of materials at small scales and arbitrary shapes. The first one involves light to enable contactless mechanical testing, i.e., testing without any physical interaction of the user with the test specimen and the testing device or instrument. The second methodology focuses on the use of machine learning approaches to bypass the need for time-consuming simulations so as to link experimental data with mechanical properties. I will conclude by highlighting the importance of applying these methods in various research areas and sectors from energy storage, to optical component fabrication, to water and wastewater treatment.