The aim of this thesis is to assess the feasibility of integrating nanoparticles into glass fiber (GF) reinforced isotactic polypropylene (iPP) composites via existing thermoplastic processing routes, and to investigate whether this results in significant improvements in the mechanical properties of the final composites. A longer term aim will be to extend the approach to the preparation of hybrid composites with added non-structural functionality. However, the nanoparticles that have provided the focus for the present project, montmorillonite layered silicates (MMT) and nanocarbons, were chosen for their potential as structural reinforcing elements. A melt-spinning grade and a film grade of iPP were used to prepare iPP-based nanocomposite precursors in the form of melt-spun fibers and extrusion-calendered films respectively. Long glass fiber (LGF) and glass mat thermoplastic (GMT) composites were then compression-molded from co-woven, co-wound and intercalated semi-finished products. The processing behavior and structural performance of the resulting composites are discussed in terms of the matrix morphology and its influence on the matrix rheological and mechanical properties, and interactions between the matrix and the reinforcing fibers. The nanocomposites were prepared by either (i) combined solvent and melt-mixing or (ii) direct melt-mixing. Combined solvent and melt-mixing was more suitable for dispersing carbon nanofibers (CNF), which tended to agglomerate. With iPP/MMT, both routes gave a mixed intercalated-exfoliated morphology with an MMT interlayer spacing up to 57 % greater than in the as-received MMT. However, direct melt-mixing was considered to be better suited to industrial requirements, and also more convenient for laboratory scale preparation. Melt-compounded iPP/MMT injection moldings showed a monotonic increase in stiffness with increasing MMT content, a 40 % increase in tensile modulus being measured at 13.5 wt% MMT, for example. The tensile strength, on the other hand, reached a maximum 10 % increase over that of the pure iPP at about 3 wt% MMT, but fell off at higher MMT contents. iPP/MMT and iPP/CNF fibers were melt-spun using a laboratory-scale industrial spinning line. Processability was consistent with the melt rheology, the maximum MMT content for which fiber spinning was possible being about 5 wt%. The MMT platelets were aligned with the fiber axis over the whole range of MMT loadings and fiber draw ratios. MMT particle aspect ratios of about 150 were observed by TEM in this case, i.e. greater than in the as-compounded iPP/MMT, for which the particle aspect ratios were about 50. An aspect ratio of 150 was found to be consistent with micromechanical modeling of the observed increases in fiber stiffness with MMT content, which reached 170 % for iPP/2 wt% MMT fibers melt-spun with a draw ratio of 1 and a drive-roll velocity of 360 m/min. The tensile strength again reached a maximum at about 3 wt% MMT. The thermal stability of the fibers, determined by thermal mechanical analysis, also increased on MMT addition, the onset of extensive fiber creep shifting from 90 °C for pure iPP fibers to about 110 °C for iPP/1.1 wt% MMT fibers. Moreover, significantly reduced shrinkage was observed in the presence of MMT, which is a potential advantage for textile based composite processing. In the case of iPP/CNF fibers, limited particle orientation and the presence of aggregates in all the formulations led to relatively poor tensile properties, about 50 % lower than those obtained with iPP/MMT fibers melt-spun with the same filler content (4 wt%) and processing conditions. iPP/MMT therefore provided the main focus for subsequent work. iPP/MMT films were produced by extrusion-calendering. Partial orientation of MMT platelets in the melt flow direction resulted in anisotropic stiffness and strength. However, both the tranverse and axial stiffnesses increased with MMT content, with improvements of up to 75 % at 5.9 wt% MMT with respect to those of the pure iPP films, consistent with Halpin-Tsai predictions for composites containing oriented platelets with the observed aspect ratio of 50. The fracture resistance of the films, determined using modified essential work of fracture (EWF) tests, was likewise strongly dependent on the testing direction. For axial crack propagation, the EWF decreased monotonically with MMT content, but for transverse crack propagation, it reached a maximum at about 3 wt% MMT. This was attributed to orientation dependent cavitation and crack deviation in the presence of the MMT particles. Rheological measurements indicated increases in the low shear rate melt viscosity by up to two orders of magnitude on MMT addition to the iPP, with a potentially significant influence on the impregnation behavior of composite preforms. For co-woven and co-wound LGF-matrix fiber preforms, impregnation was highly dependent on the fiber bed geometry. For 40 vol% glass fiber co-wound composites, the porosity increased from about 7 vol% for a pure iPP matrix compression molded at 0.6 MPa, to 14 vol% for a iPP/3.4 wt% MMT matrix compression molded at 1.8 MPa. However, the more intimate mixing between the fibers obtained in co-woven preforms led to more consolidated composites in each case (below 2 vol% porosity) with no filtering of the MMT particles. In modeling the impregnation kinetics of glass mat thermoplastic composites (GMT) based on iPP/MMT, the matrix was therefore considered to behave as a continuum and, for simplicity, to show Newtonian behavior. Consistent results were obtained, but in the presence of MMT, higher impregnation times were predicted than observed experimentally in model GMT preforms, owing to the nonlinear response of the nanocomposites. Moreover, under experimental conditions corresponding to the industrial process (0.2 MPa at 200 °C), iPP/5.9 wt% MMT-based hybrid composites were fully impregnated (porosity between 11 and 17 vol%) and the glass mat completely relaxed after about 30 s of compression molding, which is consistent with typical GMT industrial process cycle times (20 - 60 s). Fully consolidated 30 wt% GF - hybrid GMT specimens were prepared for mechanical testing by compression molding at 2 MPa and 200 °C for 10 min, resulting in a porosity of about 2 vol%. At room temperature, the flexural modulus and strength of iPP/MMT-based GMTs increased monotonically with MMT addition, the increases reaching about 45 % and 33 % respectively at 5.9 wt% MMT. The increase in the flexural modulus on MMT addition was greater than predicted on the basis of a conventional rule of mixtures taking into account glass fiber orientation and aspect ratio, and the measured Young's moduli of the matrix. This was tentatively attributed to improvements in the flexural properties of the matrix in the presence of the MMT higher than those observed in tension. Increases in the flexural modulus and strength were also observed at 50 °C and 90 °C, but were less marked than at room temperature. The impact strength of the hybrid composites decreased with increasing MMT content at room temperature owing both to the decrease in matrix fracture resistance inferred from the data for the precursor films, and to an improved fiber-matrix interface as reflected by SEM observations, and argued to be due to the presence of coupling agents. Given that the presence of MMT was shown not to have serious consequences for impregnation, taken as a whole these results are considered highly promising for the implementation of nanocomposite matrices in GMT, and to establish the general feasibility of producing hybrid fiber reinforced thermoplastic nanocomposites by conventional processing routes.
EPFL2007
