Polymer matrix composites (PMC) have emerged as viable replacements for metallic materials in the aerospace, automotive, and marine industries. They can be designed for reduced weight, increased mechanical strength, and cost-efficient processing. A major drawback to PMCs is their susceptibility to delamination as a result of poor composite fracture toughness. Researchers have explored various routes to improving composite toughness, ranging from improving polymer matrix toughness to novel designs in fiber weaving. One promising approach to improve composite toughness has been interlayer toughening; a technique that involves the addition of discrete layers in between composite plies. Early research for interlayer toughening focused on improving epoxy prepreg composites. Interlayers were comprised of thermoplastic films, rubber particles, or only resin. In general, it was found that increased interlayer thickness allowed for fuller development of a plastic zone ahead of the crack tip, thus improving composite toughness. Additionally, the presence of a second phase in the interlayer (i.e. rubber particle) induced energy dissipating mechanisms such as increased plastic deformation, particle bridging, and microcracking that further increased toughness. Drawbacks to these systems include poor adhesion of the matrix and interlayer material and control of interlayer thickness.

In this work, we used nonwoven, fibrous mats as interlayers in vinyl ester matrix composites processed through vacuum assisted resin transfer molding (VARTM). Electrospinning was used for the interlayer mat fabrication. The fibers in these mats were randomly oriented and had diameters of approximately 1 ƒÝm. This approach had distinct advantages over the aforementioned interlayer-toughened systems. First, the mats act as placeholders for the interlayers during processing. The thickness and uniformity of the interlayer can be controlled through electrospinning prior to composite lay-up and resin infusion. The control of interlayer thickness and uniformity is generally a problem for RTM composites containing particulate interlayers due to the forces caused by resin flow. Another benefit is the general dispersion of the second phase within the interlayer. For our system, resin fills the pores within the mat to form a co-continuous resin-fiber morphology. Particulate interlayers have shown a concentration gradient from the point of resin infusion to the exit.

Polysulfone (PSF) was chosen as the fiber material due to its toughening capabilities. Initial work showed that the addition of the electrospun interlayers reduced modulus and interlaminar shear strength and became more pronounced as the interlayer thickness increased. ESEM analysis of delaminated composites showed the presence of voids throughout the interlayer. These voids were attributed to the partitioning of styrene between the fibers and matrix phase. Prior to cure, styrene diffuses into and swells the PSF fibers. During cure, the reactive matrix draws styrene contained in the fibers back into the matrix and leaves voids at the fiber-matrix interface. This can be further exaggerated during post-cure, in which styrene can vaporize in the void space and react into the matrix.

A major obstacle in this work was to improve the interfacial behavior at the fiber-matrix interface to eliminate voids. The objective was to restrict styrene from diffusing into the fiber without restricting the resin from wetting the fibers completely. We developed a two-step technique that was useful in modifying the fiber surface. First, we used oxygen plasma to affix oxygen-functional groups like hydroxyl, carbonyl, and carboxyl on the fiber surface. Next, treatments with organofunctional silanes were used to form a polysiloxane shell of desired reactivity to the fiber surface. The silanes investigated were vinyltrimethoxysilane (VTMS) and methyltrimethoxysilane (MTMS). The MTMS was chosen to study the effects of a non-reactive polysiloxane shell on the ability to limit styrene diffusion into the fiber. VTMS was selected to ascertain the benefits of having a fiber that reacted to the matrix. It was found that the role of silane organofunctionality was the most important aspect for eliminating voids. The non-reactive polysiloxane showed little resistance to styrene diffusion in the fibers. Void formation in these interlayers was reminiscent of the untreated PSF fibers. When the polysiloxane had a reactive group attached to its structure, voids were eliminated at the fiber-matrix interface by reducing the driving force for styrene withdrawal from the fiber.

The fiber-matrix interface strongly influences properties in PMCs. The composites described here are further complicated by the fact that two fiber-matrix intefaces exist. Therefore, we first studied the role of embedding PSF fibers with different surface treatments into vinyl ester on the storage modulus {E'), glass transition temperature (Tg), and fracture toughness (G1c). The MTMS- and VTMS-treated fibers successfully improved G1c from 110 „b 10 J/m2 to 440 „b 10 J/m2 and 400 „b 30 J/m2. It is likely voids around the MTMS-treated fibers induced an additional crack-pinning mechanism that improved toughness slightly more than the VTMS-treated system. The voids, however, significantly reduced the E' of the MTMS system to 2.29 GPa. The VTMS system actually increased from 2.90 GPa to 3.00 GPa. Tg was not affected by fiber addition.

Composites containing untreated, MTMS-treated, and VTMS-treated interlayers were compared with respect to mode I fracture toughness, interlaminar shear strength (ILSS), flexural strength, flexural modulus, and Tg. Additionally, these composites were compared to rubber-modified composites to determine the overall effectiveness of interlayer toughening. In general, voids in the interlayers for the untreated and MTMS-treated interlayers proved to reduce ILSS, flexural strength, and flexural modulus without much improvement in G1c. The VTMS-treated interlayer proved to give the best combination of properties. G1c was improved from 740 „b 80 J/m2 to 950 „b 80 J/m2 while ILSS, flexural strength, and flexural modulus were virtually unaffected. When compared to a rubber-toughened composite, fracture toughness is not as significantly improved; however, the retention of strength and thermal properties is much better.