Nowadays, environmental considerations are leading manufacturers to integrate more and more composites based on plant fibre reinforcements into their productions. However, predicting the behaviour, characteristics, and feasibility of parts remains complex and depends on fabric behaviour on one hand, and on the other hands of the resin injection. This study proposes to focus on dry fabric behaviour.
Indeed, dry reinforcements are made by weaving tows, composed by fibres, in different architecture whose elements influence the final behaviour of the fabrics. In addition to this assembly, various other factors influence the fabrics behaviour, such as the conditions under which the fibres are grown and extracted [1]. In order to study the reinforcement behaviour, this study proposes to perform a simulation at the mesoscopic scale in the case where the strands used are flax roving. This scale enables to study the reinforcement with a certain precision while keeping a calculation time relatively correct.
To do this, a geometrical model of the roving and a suitable behaviour law are required. Since there is no pre-existing law of specific behaviour to flax roving, an hypoelastic behaviour law of an equivalent transverse isotropic homogeneous material already used in the simulation of roving of synthetic fibres will be used and adapted to the specificities of flax. To do this, roving will be studied experimentally under different loads and particularly in tensile test.
Some studies on roving made with flax fibres exist [2,3] but are still to be developed. Indeed, these rovings are particular as the fibres used are discontinuous and are in reality assemblies of fibre bundles called technical fibres. Most studies therefore focus on unit fibres (which make up the bundles), whose properties can be similar to those of glass fibre.
The aim of this presentation is to study the tensile properties of a 520tex flax roving supplied by the company Depestele. The multi-scale assembly as well as the particularities related to flax fibres induce a high variability of tensile properties or many factors can influence such as humidity, temperature or linear density [4]. This implies that a statistical study must be conducted.
The results of this experimental study will enrich the behaviour law and identify the necessary coefficients. In order to validate the simulation, a specimen from the test campaign will be used, using images obtained by digital microscopy, to reproduce a geometric model taking into account the variability of sections and to validate the simulation.