LAUSR

Abstract Flowability of thermoplastics reinforced with natural fibres represents a major problem for the injection moulding of these composites. Flowability of natural fibre polypropylene composites is evaluated using the injection moulding in an Archimedean spiral mould. The influence of the following process parameters is studied in terms of flowability: injection temperature of 180–220 °C, injection pressure of 500–1000 bar and mould temperature of 25 and 80 °C. Additionally, different fibre shapes are investigated namely; non-branched straight cellulose fibres and branched hemp fibres with low flexural stiffness. Chopped cellulose fibres length of 0.5 and 1.5 mm are tried. Fibre contents of 10 and 30 wt% are also studied. The results showed that the effect of the investigated parameters namely; Pressure, temperature and mould temperature is significant on the flowability in a descending order. The homogeneity of fibre weight content along the injection moulded spirals represents an important quality feature. Therefore, samples are drawn at even distances from the spiral length. The cut samples are dissolved to extract fibres and calculate the fibre content for each segment. Fibre content was found to be variable and the deviation from the nominal value where measurements can reach up to almost 30% deviation. The results show that the phenomenon of fibres’ separation takes place by low pressure and low injection temperature. The deviation in fibre content is found to follow a cyclic trend. The frequency of the cyclic behaviour increases with more temperature and pressure but with lower variation amplitude. At low pressure and temperature; change in fibre length behaves in a similar pattern like fibre content. Fibre length is measured by QICPIC system (dynamic image analysis). It is found that the increase in the measured fibre length is accompanied by an increase in the measured fibre content. However, more understanding, based on fundamentals of fluid mechanics, is needed for establishing robust simulation models.