Simulation of Debulking in Resin-Saturated Composite Prepregs Using a Pore-Pressure Cohesive Zone Model

Abstract

Aircraft composite parts are commonly manufactured from resin-saturated thermoset pre-impregnated plies laid-up over a rigid tool and consolidated and cured in an autoclave. In many applications, autoclave consolidation alone is not sufficient to remove the air, or bulk, that has been entrapped within the laminate during layup and may lead to the formation of voids in the final part. Entrapped air at ply interface can also participate in the formation of ply wrinkles, especially in curved laminates. High bulk content in curved sections is associated with an excess of ply length material that builds up during the layup. During compaction, this excess ply length may be squeezed into a tighter radius, which might lead to ply buckling and formation of wrinkles. Wrinkles and voids may significantly affect structural integrity of composite structures and increase rejection rates in the production cycle. Debulking, or vacuum consolidation, has been a common practice extensively used by aircraft manufacturers to reduce the amount of air entrapped during the layup of composite parts prior to curing. Yet, the underlying physical principles governing formation of defects during debulking are not well understood. This work is part of the Office of Naval Research (ONR) project “Physics-Based Composite Process Simulation†that seeks to fill the gaps in understanding the physical principles governing the formation and evolution of manufacturing defects. Recently, the authors of this work have introduced an approach for modeling debulking in resinsaturated prepregs where cohesive elements enriched with pore-pressure degree of freedom are inserted at ply interfaces, using simulation concepts originally developed for the analysis of hydraulic fracturing in geomechanics problems. Further improvement of the methodology combined the pore-pressure cohesive zone model with cohesive contact to capture the tacky behavior of the interface. This work presents a verification of the ability of the method to capture the physics involved during debulking at the coupon scale using a two-ply carbon/epoxy IM7/8552 debulk test specimen with a seeded wrinkle. The methodology is then further extended and verified at the element scale for the first time for simulation of debulking in a rotor blade grip test element.