Void consolidation of thermoplastic composites via non-autoclave processing
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
High quality continuous fiber reinforced thermoplastic composite materials can be obtained with autoclave processing that utilizes vacuum and high levels of positive pressure to achieve very low void content materials. Due to the high cost of autoclave, there is a strong demand to develop cost effective out of autoclave (OOA) alternatives that provide equivalent properties. Oven vacuum bag (OVB) processing is inherently low cost and scalable to large part manufacturing. One of the most important criteria to achieve autoclave equivalent properties is void reduction. Preliminary consolidation experiments achieved void free thin AS4/APC2 carbon PEEK thermoplastic laminates in the OVB processing. However, the methods and the mechanisms for void reduction are not fully understood. The goal of this research is to develop a fundamental understanding of the void reduction mechanisms and develop process models that can be used to design and optimize the OVB processing cycle for thermoplastic laminates of varying thickness and in-plane dimensions. ☐ The microstructure of the voids in AS4/APC2 carbon PEEK thermoplastic prepreg is critical to the initial condition of void reduction and the understanding of the void reduction mechanism. Spherical voids were assumed based on the 2D characterization in the literature. In this study, high resolution 3D X-ray micro-CT was employed to obtain 3D information of voids in a large volume of prepreg tape. Representative properties including void content, geometry and dimensions were obtained through statistical study. The statistical representative volume element (SRVE) of the void microstructure was determined. The majority of the voids were found to be rod-like with aspect ratio from 2 to 150 and finite lengths smaller than 1 mm. The voids are encapsulated in the prepreg tape and do not form porous pathways along the fiber direction. ☐ Due to the rough surfaces of the thermoplastic prepreg tapes, porous interlayer regions are formed when prepreg tapes are stacked, which is another important property that may affect the void reduction. With a 1D flow experimental set-up, the product of the in-plane air permeability and thickness (Kh) of a wide range of lay-up configurations of AS4/APC2 before and after processing were investigated. The interlayer permeability from the raw tape lay-ups is about 1-2 orders higher than the reported permeability of some partially impregnated thermosetting prepreg designed for vacuum bag processing. The interlayer permeability exhibits directional dependency, and 2D rotation transformation matrix can be used for predicting the Khs of different lay-ups. Different lay-ups show different extent of reduction with the increasing temperature and dwell time during processing. Fiber –fiber contacts limit the contact between layers and prevents significant drop of the permeability during processing. ☐ With the encapsulated voids and highly viscous resin in the AS4/APC2 prepreg tape, and vacuum bag pressure, the driving force for void reduction can only be provided by reducing the internal pressure of voids through the removal of air from the void. Through the consolidation experiments of 72-layer thick laminates, a combined mechanism that air diffusion from the void through a single layer prepreg followed by interlayer air evacuation is validated to be the key mechanism to achieve low void content in thick laminates via OVB processing. Analytical models for single layer diffusion and interlayer air flow are developed to simulate the air removal during OVB processing. With the inputs of the consolidation experiments, the modeled results are in line with the experimental results, further validating the combined mechanism for void air removal. ☐ The air flow time through the interlayer region highly depends on the interlayer permeability and the part in-plane dimensions. Maintaining the interlayer permeability during the heating ramps of the processing is critical especially for large parts. A model based on the time and temperature dependent elastic-viscoelastic-viscoplastic behavior of the PEEK resin is established to describe the deformation of the interlayer region formed by the resin rich surface region of the prepreg tapes. The modeled strains are compared with the strain calculated from the measured Kh, and the difference is discussed with the microstructure of the prepreg tape and the surface characteristics of the tapes from processed lay-ups. It is found that the deformation of the interlayer is initially governed by the resin response. Fiber-fiber contacts then supports the local stress and result in a strain plateau, preventing further deformation and reduction of the interlayer permeability. ☐ With the air removal models and the interlayer material response, parametric studies were conducted to investigate the effects of tape properties, processing conditions and part sizes. Lower initial void content, high level of vacuum, and relatively high heating rate are favorable to reduce the processing time. Low temperature dwell cycle may be included in the heating ramp to optimize the processing cycle. Based on the combined mechanism for air removal, very thick section large parts can be OVB processed (75 m parts in 8 hours with no thickness limitations based on the design of the processing cycle).
Description
Keywords
Applied sciences, Modeling, Out of autoclave, Permeability, Processing, Thermoplastic composites, Voids