| F.N. Cogswell conference report: FLOW PROCESSES IN COMPOSITE MATERIALS | |
CONFERENCE REPORT
This meeting, organized by the British Society of Rheology in association with the British Composites Society, was held at Brunel University, London, UK, on 17-18 May 1988. About 65 people attended with most of the delegates coming from university groups including a significant contingent from Europe, Scandinavia and the United States. It was a disappointment to find that there were very few delegates from the user side of the industry, but those who did attend indicated that both the formal presentations and the informal discussion periods addressed many areas of real industrial need. It was a pleasure to find that some of the university delegates were not currently working in this field but recognizing the potential rapid evolution in composites processing technology, attended the meeting in order to prepare themselves to contribute to this area. The technical session was divided into two groups: continuous fibre composites and short fibre moulding materials.
Continuous fibre composites
Three papers addressed the issue of the permeability of resin through a fibrous bed. Short (Plymouth Polytechnic), studying vacuum impregnation methods, emphasized that flow processes were far from uniform with the resin seeking, and following, the easy flow path with consequent jetting and instability. Morrison (Surrey) addressed the question of bleeding excess resin from prepreg layers, using the “Springer" model as a basis. Having made direct measurements of permeability using classical methods, she was disappointed to find that the predictions of the model were in error by more than an order of magnitude, but declined to use adjustable parameters to accommodate this discrepancy. Killesreiter (Clausthal Zellerfield), who was unable to present his paper in person, used an analogy drawn from the field of enhanced oil recovery to point out the potential significance of surface tension effects. It is clear that in the study of permeability of composite materials, an issue so critical to the quality of the finished product, we cannot rely simply on D’Arcy’s law. What is needed to replace this is unclear, but considerable effort is being directed to resolving this matter.
Two contributions discussed the methods of measuring the rheology of preimpregnated product forms. Groves (ICI) presented oscillatory rheometry data for continuous-fibre-reinforced thermoplastic melts. By presenting the data as the apparent Maxwell viscosity as a function of maximum shear rate, he obtained a normalization of experiments at different strain amplitude, and excellent agreement with steady flow data. Rogers (Nottingham) reviewed the interpretation of such experiments, pointing out that the data actually reflected a compound deformation involving both axial and transverse intra-ply shear. By using off-axis measurements, Rogers demonstrated that it should be possible to resolve between these two. The tools are clearly being sharpened and substantial bodies of data should become available in the near future.
In a pair of papers again dealing with thermoplastic systems, Wheeler (Aberystwyth) and Barnes (ICI) addressed the transverse flow process. Wheeler demonstrated how the transverse flow viscosity could be measured from such experiments and that it increased with deformation. Barnes discussed the factors influencing this limiting deformation and traced the microstructural factor controlling this effect to slight twisting of the inextensible fibres. Transverse flow is perhaps the simplest deformation process for composite materials and it may be a useful tool for qualitative evaluation of the "flowability" of different materials.
O’Neill (Nottingham), who studied the interply slip processes of plies of different orientation, concluded that such flows, which are clearly apparent during the formation of moulding with simple curvature, do not induce any reorganization of the fibre orientation. While one can have sympathy for his disappointment at this negative result, the designers and moulders among his audience gave an audible sigh of relief: only by such rigorous theoretical studies can full confidence be established.
In a joint paper Zahlan (ICI) and O’Neill (Nottingham) explained the mechanisms of spring forward, by which angular shapes distort due to the difference in thermal contraction in the plane and through the thickness of a composite laminate. An analytical solution, finite element analysis and experiment all showed that a right angle section narrowed during cooling from the moulding temperature by about 2°. The use of finite element tools then enabled the prediction of spring forward phenomena in double curvature structures, with satisfactory agreement with experiment. The designer can now accommodate those phenomena.
There were two contributions on fabrication technology. Mallon (Galway) presented his experiences with polymeric diaphragm forming of high performance thermoplastic composites including structures of single curvature, double curvature and a prototype trailer body moulding, including reversed curvature at the wheel arches. His paper emphasized the process variables which must be closely controlled in order to make high quality products with this versatile technology. Batch (Minnesota) outlined his work to model the pultrusion process, emphasizing the interaction between compaction of the fibres and permeation by the resin. Such papers chart significant advances in understanding and pave the way for improved process design.
Short fibre compoites
There were two provocative theoretical contributions. Balasubramanyam (Aberystwyth) charted the shape of a flexible fibre in a flowing melt, emphasizing the importance of stretching flows. Many model systems rather arbitrarily presume that reinforcing fibres are rigid: this contribution showed theoretically how they should deform and was warmly supported by comments on experimental results from the floor. Folkes (Brunel) drew together a variety of observations which suggest that, for high molecular weight polymers, thin layers (10 μm) are very much more mobile than the material in bulk. Many speakers described systems in which the resin layers were 2 μm or less but where bulk properties were assumed. Clearly, precisely defined experiments and rigorous understanding are required in both these areas if the technical implications are to be fully appreciated.
Several papers considered the extrusion or injection moulding processes for short- and long-fibre-reinforced thermoplastics. Bevis (Brunel) recorded the latest developments in multi live feed injection moulding where, by the use of two sets of opposed gates and reciprocating flow through the mould during cooling, it is now possible to create a laminate structure in an injection moulding including an effective cross-ply and even quasi-isotropic organization. Modelling of fibre orientation during flow was presented by Henry de Frahan (Louvain-la-Neuve). The precise modelling is so far only limited to two-dimensional studies, but the problems which must be surmounted to obtain a full rigorous three-dimensional solution were identified. Potts (Brunel) described work which had been initiated by Westland Helicopters. His main conclusion was that to study the rheology of fibre-reinforced materials appropriate to injection moulding applications, it is essential to use an injection machine as pre-plasticizer and to deliver the very high rates appropriate. Studies with conventional rheometry are of limited value because the pre-plasticization history is a major source of fibre attrition. Nevertheless, conventional capillary extrusion studies indicate that a significant level of fibre attrition occurs in the entrance region of the capillary where the flow is dominated by extension. Fibre attrition was also the subject of a study by Franzen (Gothenburg) using conventional compounding extruders. With conventional systems, attrition rapidly produces short fibre lengths independent of the length of the fibres fed to the machine: typical fibre lengths being 0.25 mm for carbon and 1 mm for metallic fibres. Analysis of length distribution leads to the postulation that fibre breakage is predominantly in the middle of the fibre rather than random, and is initiated by compression buckling. Longer fibres are retained in the new generation of Verton materials, where even the most dramatic compounding left a substantial fraction of fibres measuring more than 1 mm in length. These long fibre moulding materials were the subject of a detailed study by Gibson (Liverpool) who, like Potts, used a screw injection machine for pre-plasticization. Gibson emphasized the importance of stretching flow with such materials and presented a method of analysing converging flow for a complete range of die angles, predicting and observing a minimum in pressure drop for convergence angles of about 20°. For such materials it is not enough to study only the shaping process: the whole flow history must be taken into consideration.
Speciality material systems
Fibre-reinforced plastics are only one member of the class of composite materials. Raines (London School of Pharmacy) described the extrusion of composites of drugs and their carriers. Microcrystalline cellulose is the principal carrier and the rheology is markedly sensitive to the various grades available. Baker (Royal Ordnance) dealt with composite materials based on nitrocellulose and nitroglycerine. The safety constraints imposed by working with such energetic composites had led to the development of impressive remote sensing techniques. There was no difficulty in establishing a common rheological language between these speciality systems and fibre-reinforced plastics: the inclusion of these papers in the conference served as a reminder that the latter are relatively simple composites.
General oberservations
Following the tradition of meetings of the British Society of Rheology, the sessions were accompanied by lively discussion periods extending, in at least one case, past 2 am. The hospitality offered by Brunel was of a very high order, adding to the enjoyment of delegates. Although no proceedings for this conference are to be officially published, authors have been invited to submit their papers to the journal Composites, and Butterworth Scientific has agreed to publish as many papers as possible together in a single issue around November 1988.