EFFICIENT NUMERICAL ANALYSIS BASED ON PERTURBATION METHOD FOR THE EFFECT OF WAVINESS IN CFRP

The paper proposes a numerical multiscale homogenization method for carbon fiber reinforced composites, where fiber alignment is disturbed due to unintended imperfections (fiber waviness). Imperfection is introduced as input to the calculation, and the calculation is always done using idealized perfect geometry. This has a distinct advantage in numerical calculations where the same mesh can be used for a series of imperfect geometries. The calculation is based on second-order perturbation method in order to capture the anisotropy which is the feature of CFRP. The proposed method is validated by comparing the results to those of a conventional calculation. The results demonstrate that the proposed method can accurately capture the stress distribution when the amplitude of the imperfection is small.

A multiscale approach to estimate the cellular diffusivity during food drying

Theoretical models for food drying commonly utilize an effective diffusivity solved through curve fitting based on experimental data. This creates models with limited predictive capabilities. Multiscale modeling is one approach which can help transition to a more physics-based model minimizing the empirical information required while improving a model’s predictive capabilities. However, to enable an accurate scaling operation, multiscale models require diffusivity at a fine scale (microscale). Measuring these properties is experimentally costly and time consuming as they are often temperature and/or moisture dependent. This research conducts an inverse analysis on a multiscale homogenization food drying model to deduce the temporal diffusivity of intracellular water. A representation of the real cellular water breakdown was considered and appropriate assumptions to represent its cellular heterogeneity, in relation to time, were investigated. The work uncovered that a linear decrease in intracellular water content could be assumed and thus a function for its diffusivity was developed. The proposed function is in terms of sample temperature and intracellular water content opening the possibilities to be applied to various food materials.

A New Estimate for the Homogenization Method for Second-Order Elliptic Problem with Rapidly Oscillating Periodic Coefficients

In this paper, we will investigate a multiscale homogenization theory for a second-order elliptic problem with rapidly oscillating periodic coefficients of the form ∂ / ∂ x i a i j x / ε , x ∂ u ε x / ∂ x j = f x . Noticing the fact that the classic homogenization theory presented by Oleinik has a high demand for the smoothness of the homogenization solution u 0 , we present a new estimate for the homogenization method under the weaker smoothness that homogenization solution u 0 satisfies than the classical homogenization theory needs.

Elastic Constants Prediction of 3D Fiber-Reinforced Composites Using Multiscale Homogenization

This paper presents a multi-scale-homogenization based on a two-step methodology (micro-meso and meso-macro homogenization) to predict the elastic constants of 3D fiber-reinforced composites (FRC). At each level, the elastic constants were predicted through both analytical and numerical methods to ascertain the accuracy of predicted elastic constants. The predicted elastic constants were compared with experimental data. Both methods predicted the in-plane elastic constants “ E x ” and “ E y ” with good accuracy; however, the analytical method under predicts the shear modulus “ G x y ”. The elastic constants predicted through a multiscale homogenization approach can be used to predict the behavior of 3D-FRC under different loading conditions at the macro-level.