The Effect of Nylon and Polyester Peel Ply Surface Preparation on the Bond Quality of Composite Laminates

Abstract

The preparation of the surfaces to be bonded is critical to the success of composite bonds. Peel ply surface preparation is attractive from a manufacturing and quality assurance standpoint, but is a well known example of the extremely system-specific nature of composite bonds. This study examined the role of the surface energy, morphology, and chemistry left by peel ply removal in resulting bond quality. It also evaluated the use of contact angle surface energy measurement techniques for predicting the resulting bond quality of a prepared surface. The surfaces created by preparing three aerospace fiber-reinforced composite prepregs were compared when prepared with a nylon vs a polyester peel ply. The prepared surfaces were characterized with contact angle measurements with multiple fluids, scanning electron microscopy (SEM), and x-ray electron spectroscopy. The laminates were bonded with aerospace grade film adhesives. Bond quality was assessed via double cantilever beam testing followed by optical and scanning electron microscopy of the fracture surfaces.The division was clear between strong bonds (GC of 600- 1000J/m2 and failure in cohesion) and weak bonds (GC of 80-400J/m2 and failure in adhesion). All prepared laminates showed the imprint of the peel ply texture and evidence of peel ply remnants after fabric removal, either through SEM or XPS. Within an adhesive system, large amounts of SEM-visible peel ply material transfer correlated with poor bond quality and cleaner surfaces with higher bond quality. The both sides of failed weak bonds showed evidence of peel ply remnants under XPS, showing that at least some failure is occurring through the remnants. The choice of adhesive was found to be significant. AF 555 adhesive was more tolerant of peel ply contamination than MB 1515-3. Although the bond quality results varied substantially between tested combinations, the total surface energies of all prepared surfaces were very similar. Single fluid contact angle measurements/water break tests were therefore not predictive of bond quality, and are recommended against. The multiple fluids used allowed the construction of wettability envelopes, a more detailed look at the surface energy profile. The envelopes of nylon and polyester prepared systems were noticeably different, but while potentially useful for detecting changes or errors in surface preparation of known systems, they were not valid for predicting bond quality in new systems. Ultimately, it was determined that wetting is a necessary but not sufficient condition for bonding.