No abstract

Parametric studies were performed using finite element analysis (FEA) to learn how material and surface properties of polypropylene (PP) affect scratch behavior. Three-dimensional FEA modeling of scratching on a PP substrate with a spherical-tipped indenter is presented. Three different loading conditions, that is constant scratch depth, constant normal load, and linearly increasing normal load, are adopted for this parametric study. From the FEA findings, it is learned that Poisson's ratio has a negligible effect on scratch performance, whereas raising the coefficient of adhesive friction induces a significantly larger residual scratch depth and tangential force on the scratch tip. Increasing the Young's modulus of a material does not necessarily improve its overall scratch performance. On the other hand, modifying the yield stress of a material has a major impact on scratch resistance as a higher yield stress reduces the residual scratch depth. From this numerical effort, it is concluded that the yield stress and coefficient of adhesive friction are the most critical parameters to influence the scratch performance of a material. Analyses also suggest that the general trend in the parametric effect of the above four parameters on scratch behavior is independent of the applied normal load level.

This paper focuses on understanding the progressive failure behavior of woven composites. Five weaves, i.e. plain, 4-, 5-, 8-harness satin and twill, are considered. Rather than developing a new progressive failure analysis approach, the focus is placed on comparing the damage behaviors of the various weaves predicted by the selected failure criterion and property degradation model. The loading conditions include uniaxial tension and compression. The discussions focus on (1) the effect of the woven architecture on the predicted progressive failure behaviors (2) the similarities and difference in the damage initiation and evolution mechanisms between the plain and satin weaves and (3) the sensitivity of the predicted progressive failure behavior to assumptions about geometric nonlinearity and the property degradation model. The results have shown that the weave architecture (i.e. weave pattern) has significant effects on the predicted progressive failure behaviors even if the composites have the same overall fiber volume fraction, tow waviness and tow cross-section. Comparisons of the damage initiation and evolution mechanisms in the plain and 4-harness satin weaves indicate significant similarities in the damage behaviors in the comparable regions for the two weaves. The results also show that the predicted response of low waviness composite, which is more commonly seen in most structural applications, is quite insensitive to the assumed property degradation model. This imposes difficulties in validating a model by comparison with test results for low waviness composites.

Delamination growth caused by local buckling of a delaminated group of plies was investigated. Delamination growth was assumed to be governed by the strain-energy release rates GI, GII, and GIII. The strain-energy release rates were calculated using a geometrically nonlinear, three-dimensional, finite element analysis. The program is described and several checks of the analysis are discussed. Based on a limited parametric study, the following conclusions were reached:1. The problem is definitely mixed-mode. In some cases GI is larger than GII; for other cases the opposite is true. 2. In general, there is a large gradient in the strain-energy release rates along the delamination front. 3. The locations of maximum GI and GII depend on the delamination shape and the applied strain. 4. The mode I11 component was negligible for all cases considered. 5. The analysis predicted that parts of the delamination would overlap. The results herein did not impose contact constraints to prevent overlapping. Further work is needed to determine the effects of allowing the overlapping.

No abstract

Techniques were developed and described for performing three-dimensional finite element analysis of plain weave composites. This paper emphasizes aspects of the analysis that are different from analysis of traditional laminated composites, such as the mesh generation and representative unit cells. The analysis was used to study several different variations of plain weaves that illustrate the effects of tow waviness on composite moduli, Poisson's ratios, and internal strain distributions. In-plane moduli decreased almost linearly with increasing tow waviness. The tow waviness was shown to cause large normal and shear strain concentrations in composites subjected to uniaxial load. These strain concentrations may lead to earlier damage initiation than occurs in traditional cross-ply laminates.

General formulas for obtaining boundary conditions for micromechanics analysis of textile composites are developed based on the concept of equivalent subcell, which is the smallest region that has to be modeled. The equivalent subcell is identified by exploiting the periodicity and symmetries exhibited in the material microstructures and loading. The conventional formulas for the full unit cell of periodic structures and the formulas for the subcell are unified by introducing the “subcell vector” d and a constant vector R that accounts for the mismatch in the local displacement perturbations between two equivalent subcells. The usually lengthy derivation process required to obtain boundary conditions has been significantly simplified by shifting the effort from “deriving” to substituting the parameters established for the equivalent subcells into the formulas. The applications of these formulas are illustrated by generating boundary conditions for the smallest subcells of 4-harness satin weave and plain weave.

A finite element implementation of the transient nonlinear Nernst-Planck-Poisson (NPP) and Nernst-Planck-Poisson-modified Stern (NPPMS) models is presented. The NPPMS model uses multipoint constraints to account for finite ion size, resulting in realistic ion concentrations even at high surface potential. The Poisson-Boltzmann equation is used to provide a limited check of the transient models for low surface potential and dilute bulk solutions. The effects of the surface potential and bulk molarity on the electric potential and ion concentrations as functions of space and time are studied. The ability of the models to predict realistic energy storage capacity is investigated. The predicted energy is much more sensitive to surface potential than to bulk solution molarity.