Elsevier

Composite Structures

Volume 180, 15 November 2017, Pages 63-74
Composite Structures

Exploring damage kinetics in short glass fibre reinforced thermoplastics

https://doi.org/10.1016/j.compstruct.2017.07.096Get rights and content

Abstract

In situ SEM tensile tests are performed to shed more light on the onset and damage evolution in the shell layer of a short glass fibre reinforced polyamide 6.6 (SGFRP) composite obtained by injection moulding. Damage mechanisms are studied in three different loading directions including 0°, 45° and 90° with respect to the mould flow direction (MFD). The development of damage is monitored until total failure at different scales of observation. Qualitative results indicate that the orientation of tensile specimens with respect to the mould direction determines to a large extent the nature of involved damage mechanisms. Interfacial damage is by far the leading damage mechanism. Quantitative investigation further indicates multi-stage damage kinetics, which demonstrates an asymmetric behaviour with respect to sample orientation. One to two main directions of damage growth are identified as responsible for the failure properties of the studied composite depending on sample orientation.

Introduction

Scanning electron microscopy (SEM) provides high resolution micrographs of surface details at different length scales. SEM scans are generally performed post-mortem (after sample fracture) or after destructive surface preparation protocols. In the case of Short Glass Fibre Reinforced Polymers (SGFRPs), post-mortem SEM micrographs have long been used for inspection of local fibre dispersion, cohesive quality of fibre/matrix interface and ductility or brittleness of fractured facets [1], [2], [3], [4]. A new trend of SEM utilisation in a less constraining manner has emerged to monitor SGFRPs damage kinetics. In particular, combining SEM and quasi-static mechanical load allows correlation between micro-scale changes and macro-scale loading levels. However, a compromise is required between the spatial resolution and the extent of the region of interest. Martyniuk et al. [5] used SEM with in situ radial tensile load in the case of single fibre tests. The extent of the area of analysis was limited to the transverse fibre section to probe the initiation and the evolution of fibre/matrix debonding. Sato et al. [6], [7], [8] postulated a five-step damage scenario based on local SEM scans during interrupted in situ 3-point bending load of a PA66GF30 (30 wt% of glass fibres in polyamide 6.6) composite. The zone of analysis was limited to the surface region under tensile loading, which included a limited population of fibres. Arif et al. [9] applied a similar experimental procedure to PA66GF30 samples conditioned at three different humidity levels. The authors were able to extrapolate Sato’s scenario to include the influence of humidity-level on the preponderance of damage mechanisms. Schoßig et al. [10] imposed interrupted in situ tensile load to notched PPGF20 (20 wt% of glass fibres in polypropylene) and PB-1GF20 (20 wt% of glass fibres in polybutylene-1) specimens. The authors used environmental SEM to monitor local damage mechanisms taking place at a localised area in front of the macroscopic fracture tip. SEM observations were then correlated with acoustic emission results and macroscopic load levels to chronologically classify damage events taking place in depth of the composite material. Bourmaud et al. [11] applied sequential SEM in situ tensile load to compare the dispersion of short glass fibre and vegetal flax fibre in polypropylene and their consecutive effects on damage kinetics. The authors mainly focused on qualitative surface changes such as fringes along the crack surfaces in an extended area of analysis captured at a micrometric spatial resolution.

Complementarily to these studies, literature review shows that the effect of the angular deviation between uniaxial load direction and the Mould Flow Direction (MFD) on damage mechanisms has not been adequately investigated based on SEM in situ experiments.

Indeed, due to the fountain flow effect during injection moulding of regular forms [12], SGFRPs are known to have, in depth, a five-layer arrangement of fibres reported as the skin-shell-core structure. The ratio of the shell layers by the total thickness of the thin plate is the most significant: up to 75% of the total thickness according to Ref. [13]. Moreover, the fibres in these layers are preferentially aligned in the MFD. Consequently, the macroscopic tensile behaviour depends mainly on the angular deviation between MFD and the load direction. The highest macroscopic tensile performance including strength and stiffness is obtained when the material is loaded in the MFD. On the other side, the lowest tensile mechanical performance corresponds to specimen orientation in the direction perpendicular to the MFD.

For this purpose, this work is intended to shed more light on the influence of specimen orientation on damage kinetics at the microscopic and macroscopic scales. In accordance with the study of Bourmaud et al. [11], interrupted in situ SEM tensile experiments are performed here to monitor onset and evolution of damage in the shell layer of an injection moulded PA66GF35 composite. In addition, with consideration of the work of Arif et al. [9], most damage mechanisms are detectable at the highest humidity conditioning-levels. Thus, to have more insight about surface damage mechanisms and their corresponding characteristic sizes, we limit our study to humidity-saturated samples. Reliability of interrupted in situ tensile loading is tested for three cutting orientations. The qualitative and quantitative assessment of damage kinetics is discussed with regards to macroscopic load levels and surface observations of damage results.

Section snippets

Experimental procedure

The same grade of the PA66GF35 composite material previously used by the authors in Ref. [14] is selected. Rectangular plates are firstly prepared using injection moulding process. In this study, a large injection-gate with a typical width of 120 mm is used to reduce the turbulence of the melt material at the entrance of the mould’s cavity as shown in Fig. 1a. Consequently, an extended homogeneous in-depth flow is generated during the filling step, which results in a skin-shell-core structure

Mechanical response

Fig. 4 compares the ex-situ and in-situ tensile responses for the attempted orientations. Ex situ experiments confirm the effect of cutting orientation with respect to MFD (Fig. 4a). The higher performance is obtained for Θ = 90° when most fibres are aligned with the loading direction. It is worth noting that the angle 45° does not lead to a symmetrical trend in a sense that the associated tensile trend is closer to the loading configuration normal to MFD (Θ = 0°). From the ex-situ experiments,

Conclusions

Damage evaluation in the shell layer of PA66 thermoplastic reinforced with 35 wt% glass fibre shows a major influence of the interfacial debonding following a three-stage kinetics involving onset, growth and percolation. SEM observations of details as small as 0.5 µm confirm damage onset taking place beyond the elasticity stage (stress levels above 50 MPa) and influencing significantly the evolution of the irreversible strain. A great nonlinear influence of the sample orientation is put forward.

Acknowledgements

The authors thank Ghislain LOUIS and Damien BETRANCOURT from Civil and Environmental Engineering department (Mines Douai, France) for their assistance and significant guidance for conducting the in situ SEM tests.

Cited by (0)

View full text