Analysis of contact area in a continuous application-and-peel test method for prepreg tack

Abstract

The relationship between prepreg tack and the degree of intimate contact (DoIC) between prepreg and a rigid substrate was explored in the context of a continuous application-and-peel test method. Tack for a unidirectional prepreg tape was characterised for different surface combinations and varying test parameters (material feed rate, temperature) at a constant compaction pressure. Application of the prepreg to a transparent rigid substrate (glass), was carried out at matching test conditions to the prepreg tack measurements. Optical microscopy was utilised to acquire images of the contact area at the prepreg-glass interface. Image analysis of the micrographs enabled quantification of the contact area. The time- and temperature-dependent viscoelastic behaviour of the resin was explored directly on the prepreg using oscillatory parallel plate rheometry, and time-temperature superposition was applied to construct both tack and DoIC master curves. The shifted DoIC data showed that true contact area increases with decreasing shifted feed rates, until maximum contact area is achieved. Similarly, tack increases with decreasing shifted feed rates. However, at a critical feed rate, the bond failure mechanism switches from adhesive to cohesive failure. In cohesive failure, tack decreases with decreasing feed rate despite the high levels of DoIC.

Introduction

In Automated Material Placement (AMP), robotic machinery is used to place layers of unidirectional fibre tape pre-impregnated with a partially cured thermoset resin, known as a prepreg and commonly made from carbon fibre and epoxy resin, on a tool surface to lay up laminates. This type of process is typically used for the manufacture of large composite components with low geometrical complexity, in particular for aerospace applications, e.g. wings or fuselage sections.

AMP process variants are Automated Tape Laying (ATL), where prepreg tapes have widths in the order of several inches, or Automated Fibre Placement (AFP), where widths are smaller than one inch. In AFP, material deposition rates are typically lower than in ATL, but narrow tapes are more suitable for deposition along curved paths. For both variants, process parameters such as deposition rate, temperature and applied pressure can be adjusted. Effective use of AMP is reliant on the understanding of prepreg material characteristics in order to identify optimal deposition parameters with respect to speed, quality of bond and absence of defects. Once a laminate is laid up with the required number of layers and orientations, it is processed in an autoclave at high temperature (to induce resin cure) and pressure (for consolidation). Thus, prepreg materials are required to have sufficient adhesive properties to maintain a prescribed geometry prior to autoclave processing.

The lay-up performance, for a given set of process parameters, is strongly related to the level of adhesion, or tack, between (1) the prepreg and the tool surface, (2) successive prepreg plies and (3) the prepreg and the placement head roller. While high tack between prepreg tape and roller may result in prepreg sticking to and wrapping around the roller, insufficient tack between plies and/or between prepreg and tool contributes to the formation of defects in the lay-up, such as wrinkling and bridging [1]. It is essential to maintain sufficient adhesion prior to curing to reduce the likelihood of defect formation that may affect the final properties of the cured component. Characterisation of prepreg tack is therefore critical for the optimisation of AMP processes [1].

An approach to understanding tack phenomena is investigation of the evolution of the contact area between the prepreg and its mating surface. One of the earliest studies of prepreg tack by Gillanders et al. [2] explored the correlation between tack strength (defined as the tack force per unit probe area) and contact area using an inverted probe technique in which the probe compression and retraction phases were controlled. Images of the contact area between prepreg and a glass substrate were acquired after compression. The true contact area of the bond created under varying conditions of contact time and pressure was determined from the images. For the prepreg examined, a strong correlation between tack strength and contact area was observed.

The relationship between tack strength and contact area has been widely investigated to shed light onto bonding and debonding mechanisms in pressure sensitive adhesives (PSAs) [[3], [4], [5]]. Common experimental methods for investigating PSAs are the probe tack test and the peel test. In the case of probe tests, real-time direct observation of the deformation process of a PSA film was realised using the inverted probe technique proposed by Lakrout et al. [3]. This makes use of a mirror positioned at 45° to the contact plane allowing visual observation of the contact of the PSA film bonded to a glass slide. This experimental setup was able to record (1) the true contact area between the probe and the film and (2) the cavitation and fibrillation processes during the debonding of the film from the glass surface. The contact area patterns associated with the debonding processes could then be related to specific points on the adhesive’s stress-strain curve. Horgnies et al. [4] employed a peel test to observe the fibrillation of a PSA bonded to a glass substrate under controlled force and time. Here, the contact area formed was captured on optical micrographs, and fibrillation was observed from the side during the peeling stage, enabling the authors to confirm that debonding energy increases with increasing contact area.

In both of these prior studies, the bond is formed as a separate process to the debonding stage. Recently, a continuous application-and-peel method of measuring tack force was developed by Crossley et al. [6] as a means of obtaining a measure of bond quality of relevance to AMP processes. In this method, a prepreg tape is pressed onto a substrate and peeled in the same process at a constant rate and temperature. Using this method and a process of time-temperature superposition (TTS) based on resin viscoelasticity, Crossley and co-workers [7] were able to produce tack mastercurves in which a maximum tack force can be observed at the transition between adhesive and cohesive failure of the bond. The method was later used by Endruweit et al. [8] on unidirectional carbon fibre/epoxy prepregs to study the effect of different surface combinations, material out-time, humidity and compaction force, and by Smith et al. [9] for a study of woven carbon fibre/epoxy prepreg tack subjected to varying material out-times.

In an effort to study inter-ply void formation in out-of-autoclave prepregs, Helmus et al. [10] proposed an experimental setup comprising of a vacuum bagging arrangement using a glass tool, with a camera located underneath the tool to acquire macroscopic images of the contact area between the prepreg and the transparent tool. The images were post-processed and binarised to analyse the contact patterns as a function of time, up to the end of the curing cycle. This technique provided a quantitative measurement of the evolution of the contact area during the manufacturing process over a large sample area.

The aim of this work is to employ the established contact area imaging and processing techniques in the context of the continuous application-and-peel tack measurement method for prepregs. The goal is to investigate the role of contact area on tack in a test method relevant to AMP processes. The continuous application-and-peel tack testing fixture [7] is employed to create a contact between prepreg and a rigid glass substrate at a range of temperatures and rates, but without the subsequent peel stage. The interfaces between glass substrates and prepregs are then imaged using optical microscopy, and image processing is applied to obtain measurements of the true contact area. Complementary measurements, which include the continuous peel stage, are carried out under identical conditions to determine prepreg tack on both rigid substrates and on prepreg substrates. Both the tack and the contact area data are reduced to master curves using TTS, with parameters determined from the resin viscoelastic behaviour.