۱- Introduction

۲- Development of the consolidation model

۳- Parametric study

۴- Variability modelling of simple layups

۵- Discussion

۶- Conclusions

References

Abstract

Consolidation of a prepreg layup to a target thickness is critical in order to achieve the required fibre volume fraction and dimensions in a composite part. Experiments show that different processing conditions lead to different levels of compaction and variability in the thickness. This paper presents an analysis of processing conditions and their effects on consolidation of thick composite components. A model that accounts for both percolation and squeezing flow is employed to study two toughened prepreg systems – IM7/8552 and IMA/M21. This paper analyses the significance of the process parameters on the thickness of prepregs and its variability. The analysis of different layups and processing conditions suggests several strategies to control target thickness and its variability. The IMA/M21 prepreg system was found to have lower variability due to its toughening mechanism. The presented results provide a better understanding of the composite manufacturing and can be used to provide an informed choice in design for manufacture of composite structures.

Introduction

Increasing effort to reduce the cost and design time of composite components has resulted in adopting digital manufacturing, where each stage of the process, e.g. draping [1] or resin infusion [2], is simulated numerically. Simulation of the manufacturing processes makes it possible to predict the final geometry of a part, its performance and to optimise the manufacturing strategy to mitigate possible defects such as wrinkles [3]. Modelling of the consolidation process predicts the final thickness of the part and can help to predict possible defects arising from non-uniform compaction e.g. ply waviness or out of tolerance. Previously, such predictions were focused on predicting the final shape and thickness of the various composite parts e.g. corner parts [4–۶]. However, some of the defects have a stochastic nature [7] i.e. their location and severity is not predetermined only by the geometry or processing conditions but arise from the stochastic nature of material properties and variations in conditions. In the light of these uncertainties, even manufacturing of a flat laminate can result in an inconsistent thickness between the trials, which can particularly be an issue for thick laminates. The tolerance on the final thickness of a laminate for industrial applications can be as tight as ± ۰٫۳ mm on a part of any thickness. Such tolerance means that the coefficient of variation of thickness of a 20 mm laminate should be as small as 0.5%. The problem of reducing the variation of the thickness is discussed and analysed in this paper. Difficulties with deviation of the as-manufactured thickness from the value set by design is currently resolved by shimming, which increases parts weight, or adding sacrificial plies and then machining to the target value, which gives additional costs to the manufacturing. Defects related to the prepreg compaction can also arise when a composite part is manufactured using two rigid tools. First, an over-thick lay-up can prevent tooling closure or tool stand-off preventing full consolidation of parts of the laminate. Second, manufacturing of a tapered laminate can generate moderate waviness, even when a suitable design of tool geometry is chosen as shown in Fig. 1. In the case of an over-thick lay-up, if the design is implemented in the absence of accurate data or models of compressibility of the prepregs, then the thicker part of a tapered laminate generates severe wrinkles as shown in Fig. 2. These examples show that predicting the thickness to which the laminate can be compacted is important for avoiding defects in composite manufacturing. More precise manufacturing can help to reduce or completely avoid these difficulties.