Virtual testing methodology to predict the mechanical behavior of collagen hydrogels from nanoarchitecture

Collagen-based hydrogels are three-dimensional, cross-linked structures capable of mimicking the extracellular fibered matrix of biological tissues, making them particularly well-suited for biomedical applications. These hydrogels typically exhibit highly non-linear mechanical behavior, which strongly depends on their internal nanostructural characteristics - an interconnection that remains poorly understood. The aim of this work is to combine high resolution imaging with a multiscale in silico structural model to virtually reproduce the mechanical behavior of a widely used collagen-based hydrogel, using solely its nanoarchitecture as input. The real fiber structure of the hydrogel was originally quantified at the nanometer scale using state-of-the-art microscopy, specifically, focused ion beam-scanning electron microscopy (FIB-SEM). In silico shear tests were then performed on the reconstructed collagen matrix to compute, through a multiscale approach, its homogenized mechanical response, including the energies and stresses developed by the fibers during the tests. Different samples of the hydrogel were also mechanically characterized by means of rheological tests to fit the model and show the feasibility of the methodology. The in silico simulations successfully captured the detailed mechanical interactions between fibers as well as the experimental non-linear mechanical behavior of the hydrogels. Results also highlight the relevant role of the bending energy throughout the entire range of deformation analyzed. This methodology provides a framework to elucidate the structure-mechanical behavior relationship of fiber network topologies, and can be applied to predict mechanical response of both native tissues and biomaterials based exclusively on their fibered nanostructures.