skip to main content

This content will become publicly available on November 1, 2023

Title: Evolution of Force Chains Explains the Onset of Strain Stiffening in Fiber Networks
Abstract Fiber networks are the primary structural components of many biological structures, including the cell cytoskeleton and the extracellular matrix. These materials exhibit global nonlinearities, such as stiffening in extension and shear, during which the fibers bend and align with the direction of applied loading. Precise details of deformations at the scale of the fibers during strain stiffening are still lacking, however, as prior work has studied fiber alignment primarily from a qualitative perspective, which leaves incomplete the understanding of how the local microstructural evolution leads to the global mechanical behavior. To fill this gap, we studied how axial forces are transmitted inside the fiber network along paths called force chains, which continuously evolve during the course of deformation. We performed numerical simulations on two-dimensional networks of random fibers under uniaxial extension and shear, modeling the fibers using beam elements in finite element software. To quantify the force chains, we identified all chains of connected fibers for which the axial force was larger than a preset threshold and computed the total length of all such chains. To study the evolution of force chains during loading, we computed the derivative of the total length of all force chains with respect to more » the applied engineering strain. Results showed that the highest rate of evolution of force chains coincided with the global critical strain for strain stiffening of the fiber network. Therefore, force chains are an important factor connecting understanding of the local kinematics and force transmission to the macroscale stiffness of the fiber network. « less
Award ID(s):
Publication Date:
Journal Name:
Journal of Applied Mechanics
Sponsoring Org:
National Science Foundation
More Like this
  1. The extracellular matrix provides macroscale structural support to tissues as well as microscale mechanical cues, like stiffness, to the resident cells. As those cues modulate gene expression, proliferation, differentiation, and motility, quantifying the stiffness that cells sense is crucial to understanding cell behavior. Whereas the macroscopic modulus of a collagen network can be measured in uniform extension or shear, quantifying the local stiffness sensed by a cell remains a challenge due to the inhomogeneous and nonlinear nature of the fiber network at the scale of the cell. To address this challenge, we designed an experimental method to measure the modulus of a network of collagen fibers at this scale. We used spherical particles of an active hydrogel (poly N-isopropylacrylamide) that contract when heated, thereby applying local forces to the collagen matrix and mimicking the contractile forces of a cell. After measuring the particles’ bulk modulus and contraction in networks of collagen fibers, we applied a nonlinear model for fibrous materials to compute the modulus of the local region surrounding each particle. We found the modulus at this length scale to be highly heterogeneous, with modulus varying by a factor of 3. In addition, at different values of applied strain, wemore »observed both strain stiffening and strain softening, indicating nonlinearity of the collagen network. Thus, this experimental method quantifies local mechanical properties in a fibrous network at the scale of a cell, while also accounting for inherent nonlinearity.« less
  2. We study the effect of inter-fiber adhesion on the mechanical behavior of cross-linked ran- dom fiber networks in two dimensions. To this end, we consider networks with connectiv- ity number, z , below, at, and above the isostaticity limit of the structure without adhesion, z c . Fibers store energy in the axial and bending deformation mode and the cross-links are of freely rotating type. Adhesive forces lead to fiber bundling and to a reduction of the total volume of the network. The degree of shrinkage is determined as a function of the strength of adhesion and network parameters. The mechanical response of these struc- tures is further studied in uniaxial tension and compression. The stress-strain curves of networks without inter-fiber adhesion exhibit an initial linear regime, followed by strain stiffening in tension and strain softening and strain localization in compression. In pres- ence of adhesion, the response becomes more complex. The initial linear regime persists, with the effective modulus decreasing and increasing with increasing adhesion in cases with z > z c and z < z c , respectively. The strain range of the linear regime increases signif- icantly with increasing adhesion. Networks with z > z c subjectedmore »to tension strain-stiffen at rates that depend on the adhesion strength, but eventually enter a large strain/stress regime in which the response is independent of this parameter. Networks with z < z c are stabilized by adhesion in the unloaded state. Beyond the initial linear regime their tangent modulus gradually decreases, only to increase again at large strains. Adhesive interactions lead to similar effects in compression. Specifically, in the z > z c case, increasing the adhe- sion strength reduces the linear elastic modulus and significantly increases the range of the linear regime, delaying strain localization. This first investigation of the mechanics of cross-linked random networks with inter-fiber adhesion opens the door to the design of soft materials with novel properties.« less
  3. We present a study of the mechanical behavior of planar fibrous mats stabilized by inter-fiber adhesion. Fibers of various degrees of tortuosity and of infinite and finite length are considered in separate models. Fibers are randomly distributed, are not cross-linked, and interact through adhesion and friction. The variation of structural parameters such as the mat thickness and the mean segment length between contacts along given fibers with the strength of adhesion is determined. These systems are largely dissipative in that most of the work performed during deformation is dissipated frictionally and only a small fraction is stored as strain energy. The response of the mats to tensile loading has three regimes: a short elastic regime in which no sliding at contacts is observed, a well-defined sliding regime characterized by strain hardening, and a rapid stiffening regime at larger strains. The third regime is due to the formation of stress paths after the fiber tortuosity is pulled out and is absent in mats of finite length fibers. Networks of finite length fibers lose stability during the second regime of deformation. The scaling of the yield stress, which characterizes the transition between the first and the second regimes, and of the secondmore »regime's strain hardening modulus, with system parameters such as the strength of adhesion and friction and the degree of fiber tortuosity are determined. The strength of mats of finite length fibers is also determined as a function of network parameters. These results are expected to become useful in the design of electrospun mats and other planar fibrous non-cross-linked networks.« less
  4. Abstract Many engineering materials are made from fibers, and fibrous assemblies are often compacted during the fabrication process. Compression leads to the formation of contacts between fibers, and this causes stiffening. The relation between the uniaxial stress, S, and the volume fraction of fibers, φ, is of power law form. The derivation of this relation based on micromechanics considerations takes as input the structural evolution represented by the dependence of the mean segment length of the network, lc, on the current density, ρ (ρ is defined as the total length of fiber per unit volume of the network). In this work, we revisit this problem while considering that the mean segment length should be defined exclusively by fiber contacts that transmit load. We use numerical simulations of the compression of crimped fiber assemblies to show that, when using this definition, ρ∼1/lc2 at large enough strains. Purely geometric considerations require that ρ∼1/lc, and we observe that this applies in the early stages of compaction. In pre-stressed networks, the density–mean segment length scaling is of the form ρ∼1/lc2 at all strains. This has implications for the relation between stress and the fiber volume fraction. For both ρ versus lc scalings, S∼(φn−φ0n), wheremore »φ0 is the initial or reference fiber volume fraction; however, n = 3 when ρ∼1/lc and n = 2 for ρ∼1/lc2. These predictions are compared with experimental data from the literature.« less
  5. Abstract Soft bioinspired fiber networks offer great potential in biomedical engineering and material design due to their adjustable mechanical behaviors. However, existing strategies to integrate modeling and manufacturing of bioinspired networks do not consider the intrinsic microstructural disorder of biopolymer networks, which limits the ability to tune their mechanical properties. To fill in this gap, we developed a method to generate computer models of aperiodic fiber networks mimicking type I collagen ready to be submitted for additive manufacturing. The models of fiber networks were created in a scripting language wherein key geometric features like connectivity, fiber length, and fiber cross section could be easily tuned to achieve desired mechanical behavior, namely, pretension-induced shear stiffening. The stiffening was first predicted using finite element software, and then a representative network was fabricated using a commercial 3D printer based on digital light processing technology using a soft resin. The stiffening response of the fabricated network was verified experimentally on a novel test device capable of testing the shear stiffness of the specimen under varying levels of uniaxial pretension. The resulting data demonstrated clear pretension-induced stiffening in shear in the fabricated network, with uniaxial pretension of 40% resulting in a factor of 2.65 increasemore »in the small strain shear stiffness. The strategy described in this article addresses current challenges in modeling bioinspired fiber networks and can be readily integrated with advances in fabrication technology to fabricate materials truly replicating the mechanical response of biopolymer networks.« less