| description abstract | Applications for porous fibrous materials range from electrochemical substrates to web reinforcement in polymeric composite materials. The details of local load transfer are studied in a class of cost-effective, stochastic fibrous networks used in battery applications, which form the substrate for a composite electrode. The connectivity of these materials is quantitatively related to modulus and strength, and detailed results of different simulations approaches in approximating material construction are discussed. In Part I, we discuss microscale assumptions, including beam type, nodal connections and equivalence of models to more physically realistic models. Simulation of large networks is computationally intensive, and show low-strain, nonlinear behavior even when comprised of elastic elements when failure criteria (here, strength-of-materials) are applied to produce sequential rupture of beams and nodes. Strategies for effective simulation of these materials requires detailed analysis of the simplest assumptions which can be made at the microscale which produce acceptably realistic response. We show that simple Euler-Bernoulli beam elements can be used to effectively model such materials, even when segment lengths in a network are very small. Moreover, connections comprised of simple torsion springs produce realistic behavior, and can mimic more realistic junctures by adaptation of the linear solution to a compliant zone model. In Part II of this work, we demonstrate the effect of model selection on full network behavior, and also discuss issues of connectivity at the scale of the porous material rather than element-by-element. This work points toward use of simple constructions to model complex behavior, and may ultimately provide insight into modeling of a large class of porous materials. [S0094-4289(00)01704-7] | |
