Investigation of various techniques for controlled void formation in fiberglass/epoxy composites

Abstract

The effect of porosity in composite materials has been studied for years due to its deleterious effects on mechanical properties, especially matrix dominated properties. Currently there is an increasing use of composites in infrastructure worldwide, for example bridge components, residential and building structures, marine structures such as piers and docks, and large industrial chemical tanks. Most of these applications use fiberglass composites. Unfortunately, most of the published literature has focused on carbon fiber composites, in which fiber diameter and gas-fiber interactions are different than fiberglass composites. Therefore, the present study was undertaken to revisit the effect of porosity but specifically in fiberglass composites. The goal of this experimental study was to implement and evaluate various methods for creating porosity in fiberglass composites in a controlled manner in terms of obtaining repeatable void content, morphology, and location within the laminate. The various methods included using different amounts of autoclave pressure, adding a small amount of water between prepreg layers, and using dry fabric layers to starve the laminate of resin. Ultrasonic C-scan nondestructive evaluation was used to assess the quality of the cured panels, as well as optical and electron microscopy and void content measurements via resin burn-out. The cured panels were mechanically tested using the short beam shear (SBS) method. The results showed that the water spray method proved to be the best in terms of producing noticeably different levels of porosity, although the panels required drying to remove residual water after cure. The voids from all three techniques were either oval or elongated in-plane between the plies, but they were not uniformly distributed in-plane. The use of C-scan proved to be helpful for characterizing overall uniformity of each panel, although the results could not be used to directly compare void content between panels. The use of SBS testing was successful for evaluating void dominated properties in panels with high void content, although it was not very sensitive to coupons with lower void contents. Several interesting observations are offered in this manuscript of the fracture surface details and their relation to the SBS load deflection curves. Overall, it was found that the failure mechanisms were mixed mode and the voids did not serve as failure initiation sites. However, the voids participated mainly in the horizontal propagation of cracks between layers, presumably making it easier when they were intersected by a crack and reducing SBS strength.