Exploring mechanical damage in fascia: Experiments and advanced constitutive modeling approaches

Abstract

Biological tissues exhibit complex structures that necessitate mechanical models incorporating details of their key components and the physical processes occurring within the material. Our objective is to enhance the understanding of damage mechanisms in fibered tissues through mechanical testing. This includes conducting uniaxial tensile tests on fascia beyond physiological stretch limits and developing two constitutive models to describe damage and rupture. These models integrate both phenomenological and microstructural perspectives.

Two perpendicular directions, corresponding to the two families of collagen fibers, were compared: the longitudinal direction, characterized by greater stiffness, and the transverse direction. The mean Cauchy rupture stress (${\sigma }_r$) was reported as 16.67 for the longitudinal direction and 4.76 MPa for the transverse direction, with a significant difference observed between them ($p$-value $<$ 0.05). Similarly, a significant difference in stored strain energy was found between the two directions ($p$-value $<$ 0.05) between directions, being in longitudinal equal to 1.33 $\text{N}\cdot \text{mm}/{\text{mm}}^3$ and 0.49 in transversal one. However, rupture stretches (${\lambda }_r$) did not exhibit a significant difference ($p$-value $>$ 0.05) with values of 1.17 and 1.22 for the longitudinal and transverse directions, respectively.

In this study, a hyperelastic constitutive model for fascia was modified to incorporate damage effects into the strain energy function. Additionally, an extended version of a microstructural damage model was developed to effectively replicate the experimental data. The proposed damage models successfully captured the stress–strain behavior and accurately represented the damage process. The coefficient of determination $R^2$ for the fitted data ranged from 0.616 to 0.973, except for Sample IV, which exhibited an $R^2$ value of 0.251 when using the phenomenological model. In all cases, the microstructural model provided a more accurate fit compared to the phenomenological model, with $R^2$ values ranging from 0.748 to 0.927.