Recently, there has been a great interest in developing and maturing natural fibre composites for structural applications. Natural fibres derived from plants such as flax and hemp have the potential to compete with traditional glass fibres as reinforcements in polymer matrices, due to good specific properties (stiffness-to-density ratio). The perspective of using natural fibres is to have a sustainable, biodegradable, CO2-neutral alternative to glass fibres. However, so far, it has not been possible to take full advantage of the natural fibre properties when using them for composite applications. Several challenges have to be addressed and solved, many of which pertain to the fact that the fibres are sourced from a natural resource: 1) Inconsistent properties, depending on plant species, growth and harvest conditions, and fibre extraction techniques. 2) Strength values of composites are lower than expected based on tests of single fibres. 3) Compared to continuous glass fibres, natural fibres are relatively short, which makes it difficult to achieve an optimized fibre architecture. 4) Natural fibres are hydrophilic, meaning that they do not bond well with standard polymer matrix systems, most of which are hydrophobic. The present ph.d. thesis is primarily concerned with challenges 2 (unexpected low strength of composites) and 3 (optimization of fibre architecture). Reasons for the lower than expected strength of natural fibre composites are investigated by performing X-ray tomographic microscopy during tensile tests of small composite specimens. With this technique, 3D images can be obtained with spatial resolution < 1 µm. By studying the 3D microstructure of the composite specimens at a number of arrested load steps, a number of damage mechanisms have been identified: (i) Interface splitting cracks typically seen at the interfaces of bundles of unseparated fibres, (ii) matrix shear cracks, and (iii) fibre failures typically seen at fibre defects. The three damage mechanisms initiated at about 50, 75 and 90% of the failure stress, respectively. After harvesting the plants, the fibre bundles in the plants are extracted, and separated into individual fibres. If this separation is not complete, bundles consisting of 5-15 fibres will remain among the fibres. Important insight was gained on the significance of avoiding bundles of unseparated fibres. It was found that such bundles are likely to result in fibre/matrix debonding cracks, which can lead to ultimate failure by large splitting cracks. Also, the fibre bundles were observed to have a tendency to fail across the entire cross section of the bundle. This will lead to a large stress concentration, which can result in specimen failure. Since individual natural fibres are relatively short (50-70mm), they are traditionally spun into fibre yarns in order to be able to handle the fibres. However, spinning the fibres effectively equates to introducing a large amount of fibre misalignment, which decreases the composite stiffness properties. Through development of a model based on the geometry of a yarn with fibre twisting and yarn helicity, the relation between fibre misalignment and composite tensile stiffness was examined. The model incorporates a ±-stiffening effect, similar to what is used in laminate theory. Experimental studies were performed with composites fabricated from yarns with different amounts of fibre twist and yarn helicity. By fitting the proposed model to the experimental data, good agreement was obtained. From the model predictions, it was found that yarn helicity is actually more detrimental than fibre twisting with regards to composite stiffness. Finally, studies are performed on the fracture toughness of natural fibre composites. Initially, a novel approach is proposed for calculating the fracture toughness from data obtained from double cantilever beam tests. The developed approach is based on determination of the curvatures of the beams during the tests and it is not necessary to have any knowledge of the layup sequence, or stiffness and thickness of individual layers. This is especially beneficial for complicated/unknown beam layups. It was proposed that the beam curvatures are determined using strain gauges. After developing the approach, it was used to determine the fracture toughness of flax/PLA (polymer based on lactic acid) specimens made from yarns with different twisting angles. It was found that a high twisting angle greatly decreases the fracture toughness of the composite, such that specimens made with yarns with no fibre twisting were more than 10 times tougher than specimens with a high degree of twisting. Thus, based on the work in the present ph.d. thesis, it is found that achieving a method for separating the fibres completely without damaging them, is im- portant for optimizing the composite strength. Furthermore, it is found that achieving a good fibre alignment is important for both the composite stiffness and the composite fracture toughness. These suggestions for manufacturers of natural fibre composites, are presented with an overall purpose of contributing to optimizing natural fibre composites for load-bearing usage.
Main Research Area:
Sørensen, Bent F., Madsen, Bo, Lauridsen, Erik Mejdal