A microstructures generation tool for virtual ply property screening of hybrid composites with high volume fractions of non-circular fibers – VIPER

Abstract

Within the framework of computational micromechanics (CMM), a simulation toolset is being developed to predict the mechanical behavior of fiber-reinforced polymers from the measured properties and spatial distribution of the different phases and interfaces in the composite. Towards this end, a numerical methodology is proposed herein for the generation of 2D periodic microstructures with arbitrary fiber geometries. A major advantage of the approach presented in this work is the ability of generating high volume fractions of non-circular fibers very efficiently. The underlying algorithm is based on the minimization of fiber overlapping by dynamic translation and rotation of the fibers until intersections are eliminated. The randomness of the microstructures obtained is assessed by means of multiple spatial descriptors.

A computationally efficient approach for generating RVEs of various inclusion/fibre shapes

“…This drastic reduction 𝑡 𝑔𝑒𝑛 may be majorly due to explicit gradient evaluation, which otherwise has to be calculated using computationally expensive numerical methods. Recently, [25] has reported RVE generation times for non-circular inclusion shapes. In comparison with this study, for generating the RVE of approximately 2070 elliptical and rectangular inclusions, the present algorithm takes 35 and 20 s, respectively, while [25] has reported 69 and 86 s. Higher 𝑡 𝑔𝑒𝑛 of 3D RVEs, compared to 2D RVEs, is due to an increased number of inclusions for a given RVE size and volume fraction.…”

“…In a more recent line of research [13,24,25], the required number of inclusions are randomly initialized, then inclusion overlaps are removed using an iterative procedure. Pathan et al [13] used L-BFGS-B constrained optimization algorithm to minimize/eliminate the inclusion overlaps.…”

“…Though it can generate RVEs with higher inclusion volume fractions, very high computational times are reported. Herráez et al [25] considered inclusions as rigid bodies where overlaps are removed using repulsive force and momentum. This model, [25], generates RVEs of non-circular cross-sections with higher volume fractions in relatively shorter times.…”

“…Herráez et al [25] considered inclusions as rigid bodies where overlaps are removed using repulsive force and momentum. This model, [25], generates RVEs of non-circular cross-sections with higher volume fractions in relatively shorter times. In this work, [25], RVE overlap is quantified using the area of inclusions intersection, which may not be easily determined for arbitrary inclusion shapes.…”

“…This model, [25], generates RVEs of non-circular cross-sections with higher volume fractions in relatively shorter times. In this work, [25], RVE overlap is quantified using the area of inclusions intersection, which may not be easily determined for arbitrary inclusion shapes. Inclusion overlap checking, in general, is non-trivial for non-circular or non-spherical shapes, especially when they are arbitrarily aligned in the space [26,27].…”