Manufacturing Study of Thin-ply Composite Prepregs within Automated Fiber Placement(AFP)

Abstract

Composite aerospace structures offer high specific mechanical properties as compared to traditional materials, yet composite materials tend to lack sufficient damage tolerance. In recent years there has been a focus on thin-ply pre-preg materials, which have shown an improvement with respect to micro-cracking and fatigue resistance as compared to the standard-ply-thickness materials. Large scale automated composite manufacturing processes, such as automated fiber placement (AFP), are necessary to produce large structures reliably and affordably. However, thin-ply materials require more layers to reach the same thickness, increasing the number of passes by the AFP machine. The overall production time is thereby increased, increasing the importance to maximize lay-down rate and minimize machine downtime. Thin-ply materials have not yet seen widespread adoption in AFP manufacturing in part due to a lack of study and development that leads to uncertainty associated with their manufacture. The result is that thin-ply materials lack the same repeatability, consistency and quality as standard-ply-thickness composites in the AFP process. In this work, efforts to identify key material and AFP processing parameters of thinply materials that result in high quality, consistent and repeatable feed of material through the AFP head and onto the structure are presented. A single spool, in-house designed machine is used to mimic the AFP process using a 70 gsm thin-ply material. The material-process parameters and interactions are studied using a design of experiments (DoE) approach. The factors being studied fall into three main categories: material parameters, slitting and spooling parameters, and AFP process parameters. Some of the factors being studied include out-time, backing paper vs. double-backing paper, and pre-preg tape tension. The DoE program is supplemented with finite element models (FEM) of the relevant manufacturing parameters.