Journal of Composite Materials

Maximization of mode I critical delamination fracture toughness (G _IC) in cross ply graphite–epoxy laminates is studied using Design of Experiments and Taguchi arrays. The novel methodology proposed allows continuous and discrete factors to be considered simultaneously. Main and interactive factor effects on G _IC are evaluated. Selection of factors is based on design and material criteria; the ones considered are: width, length, thickness, stacking angle and stacking sequence. A Fractional Factorial experiment is used to reduce the number of tests. Double Cantilever Beam specimens are used for all mode I experiments. Statistically significant factors are culled and a response equation (RE), based on the importance of each factor, is established. The RE is then employed to optimize G _IC within the given set of restrictions. The effect of interfaces on delamination propagation is also investigated.

Low cycle fatigue (LCF) behavior of composite laminates is investigated along with the associated micromechanical damage. At the ply level, fatigue damage is described by a damage-mode-sensitive model, which is based on a family of S–N curves that are tailored for various types of damage. Since the LCF conditions may involve high loads reaching up to 90% of the material ultimate-strength, the S–N curves reflect unique LCFfeatures such as non-cyclic and bi-linear corrections due to high property-degradation rates. The LCFcharacterization of laminates is based on authors’ previous research and other relevant studies. The proposed fatigue model is implemented in a laminate and finite element analysis, and a case of a notched laminated plate is analyzed to examine accumulation of fatigue damage at the ply level in the plate. The effective fatigue-damage is predicted by the new fatigue model and compared with experimental results.

The present study considers filling of honeycomb type cores with foam to produce sandwich constructions. The potential benefits of this approach are enhancement of damage resistance, and ability to process honeycomb type sandwich structures through cost-effective vacuum assisted resin transfer molding (VARTM). As weight penalty is incurred in complete filling of honeycomb cells with foam, an alternative approach to reduce weight is partial filling of the cells, without losing the advantage of VARTM processing of the core. Two cores are considered, a polyurethane foam for full filling of honeycomb cells, and syntactic foam for partial filling, in conjunction with carbon–epoxy facesheets. Their impact response was investigated under low and high velocity impact (LVI and HVI respectively). For both cores, the foam filling was found to provide confinement to the cells. The resistance to penetration, energy absorbed and damage modes in LVI and HVI were a function of core stiffness, extent of filling and number of facesheet plies. The results illustrate that partial syntactic foam filled sandwich plate (with reduced weight penalty in comparison to full filling) can provide LVI response improvement in the order of 56% increase in peak load, and for HVI about 74% improvement in ballistic limit.

The effects of moisture and temperature on high strain rate responses of S2-glass–vinyl ester woven composites have been studied in this work. Compressive properties and micro-structural damage progression under high strain rate loading have been investigated using the Split Hopkinson Pressure Bar technique. The compressive stress pulse is applied through the thickness direction (TD) as well as through the filler direction. The stress strain responses at strain rates ranging 490–1470s⁻¹ under various environmental conditions have been investigated and the relevant failure modes have also been identified by optical and scanning electron microscopic examinations.

The results indicate that moisture and temperature degrade the compressive failure strength under high strain rate loading. The level of such degradation is seen to vary with strain rates and loading direction. In general, the failure strength and failure strain of both dry and wet samples are seen to be increased under high strain rate loading. The strain rate sensitivity of the woven composites is more significant if the failure is dominated by the resin materials. High temperature tests were conducted at temperatures ranging 23–204°C. The maximum rate effect is observed below the glass transition temperature, T _g (121°C) specifically for TD specimens. Microbuckling of the fiber, kinking and fiber splitting are the major failure modes for the specimens loaded in filler direction. The specimens loaded in the thickness direction mostly failed under matrix cracking.

Sandwich construction using foam cores is used in applications, and an understanding offailure loads is an important part ofthe design process. Tests on sandwich beam specimens with embedded strain gages show that the simpler analysis procedures can be in error by large factors, while an elasticity solution from the literature gives good agreement with the experiments. Comparisons with laboratory tests on sandwich bend specimens with carbon–epoxy faces and polyurethane foam cores ofseveral densities show that the elasticity solution, in conjunction with criteria for failure of the faces and yield of the core give good agreement with the data, and show that the mode of failure and the load at failure could be predicted. Optimization ofsandwich structures for strength/weight is considered, and it is shown that the laboratory strength data support the theoretical optimization results.