Are there similarities between quasi-static indentation and low velocity impact tests for flax-fibre composites?

https://doi.org/10.1016/j.indcrop.2021.113840Get rights and content

Highlights

  • Performing Low-Velocity Impact (LVI) and Quasi-Static Indentation (QSI) tests.

  • LVI/QSI comparison for epoxy laminates reinforced with 6 layers of flax twill fabric.

  • Analysing load-displacement data and visible damage for LVI/QSI at four energy levels.

  • Examining acoustic emission to assess internal damage chronology in QSI specimens.

  • Similarities in load-displacement, absorption energy and damage mechanisms.

Abstract

Flax-fibre composites are increasingly used as a replacement of classical synthetic composite materials. Due to the good energy absorption properties of flax fibres, they represent a promising alternative in structures susceptible to low velocity impact (LVI) damage. However, this type of dynamic loading is complex, expensive to perform and not necessarily easy to fully investigate. A simpler way to tackle this problem consists in investigating quasi-static indentation (QSI) tests, but this alternative remains relatively under-researched for natural fibre composites. Thus, this paper aims at providing a comparison between both types of loading to facilitate the later analysis and modelling of flax fabric laminates submitted to LVI. Six layers of a flax 2/2 twill fabric were used as reinforcement for epoxy laminates made through vacuum infusion. Specimens were then submitted to instrumented LVI and QSI tests at comparable energy levels, with a 1.5 %–3.9 % difference only. Load-displacement curves and visible damage were first analysed and compared between both test types. Then, the internal damage within QSI specimens were investigated using acoustic emission (AE). Our findings showed good analogies between both testing methods in all the stages of damage development. Great similarities were found in load-displacement curves (in shape, stiffness and peak load), in energy absorption capacity (at 5 and 10 J) and in visible damage. Actually, the differences between QSI and LVI remain low, i.e. 2.1 % for linear stiffness, from 0.2 to 5.6 % for peak load and less than 7% for the proportions of absorbed energy. Comparison of the QSI damage analysed from the AE data with LVI results from literature suggested similar mechanisms and onset sequences. These results revealed that QSI monitoring could provide characteristic indications on the damage evolution of flax-fibre woven composites during an LVI test.

Introduction

Plant fibres have been increasingly used as composite reinforcement in the last decades. Despite their inherent variability that may impact the final properties of composites (Baley et al., 2020; Haag et al., 2017; Le Gall et al., 2018), they exhibit many advantages over their synthetic counterparts, such as lower environmental impact, good acoustic insulation and vibration damping, high specific mechanical properties, low cost and safe handling (Amiri et al., 2017; Correa et al., 2019; Dicker et al., 2014). Flax-fibre composites, in particular, are commonly used for non-structural or semi-structural applications. Recently, several research works have addressed the development of high performance biocomposites for structural applications (Baley et al., 2020; Le Duigou et al., 2019; Zuccarello et al., 2018). In this respect, the properties of flax fibres make them suitable for use as reinforcement of composites subjected to low velocity impact (LVI). One common test method used to assess the LVI performance of plant-fibre composites is drop-weight impact, which is very similar to a real impact scenario (Agrawal et al., 2014; Sutherland, 2018).

Despite these advantages, literature specifically addressing the impact resistance of composite laminates with long continuous natural fibres submitted to a drop weight impact is not extensive (Muneer Ahmed et al., 2021). Some studies have been conducted on woven composites based on jute fibres (Dhakal et al., 2014) or hemp fibres (Scarponi et al., 2016), and on unidirectional, cross-ply and woven flax fabric composites (Awais et al., 2020; Bar et al., 2020; Bensadoun et al., 2017; Liang et al., 2015; Ravandi et al., 2017; Sy et al., 2018). The type of fibre architecture can have an effect on the impact resistance. Bar et al. (Bar et al., 2020) found that flax plain woven composites were better than UD composites, mainly due to interlaced structure of plain woven fabric. Sy et al. (Sy et al., 2018) reported that cross-ply flax/epoxy laminates exhibited higher penetration threshold energy and impact toughness compared to their unidirectional counterparts. The damage progression during LVI may also evolve. Liang et al. (Liang et al., 2015) investigated the fracture mechanism of quasi-isotropic flax/epoxy composites and found that delaminations occurred first, at low energy levels, followed by the development of intra-laminar transverse cracks resulting from fibre failure. From an X-ray micro-computed tomography study, Miqoi et al. (Miqoi et al., 2021) suggested a damage scenario for the impacted woven composite. They stated that matrix cracking first appeared on the un-impacted surface and propagated along the yarns in a transverse and longitudinal path. When the energy was sufficiently high, it developed into delamination which propagated between the damaged yarn and the perpendicular yarn just below.

Another type of loading that creates damage in laminates and resembles LVI is Quasi-Static Indentation (QSI). A QSI test consists in applying on the material a transverse load perpendicular to the indented surface via a hemispherical indenter. Whereas the impactor of an LVI drop tower test is in free fall before hitting the surface of the composite, the indenter of a QSI test is brought into contact with the surface prior to the test. Thus, both testing methods are comparable in their working principle, except that one is dynamic in nature (LVI) while the other is quasi-static (QSI). However, an LVI corresponds to an impact event in which the contact time of the impactor on the material surface is long compared to the propagation time of the stress-wave induced by the impact, making it close to static loading (Andrew et al., 2019). Consequently, several authors have compared LVI with QSI, particularly for carbon-fibre/epoxy laminates, sometimes recommending the use of indentation to analyse and better understand impact damage mechanisms (Nettles and Douglas, 2002; Saeedifar et al., 2018; Serna Moreno and Horta Muñoz, 2020; Spronk et al., 2018; Wu et al., 2020).

As a matter of fact, the implementation and instrumentation of LVI drop tower tests is often challenging. These tests require the use of special equipment (drop tower). They usually have a short duration, making it hard to investigate the damage sequence. The roughness of an impact limits the use of recording devices such as acoustic emission sensors. Moreover, load-displacement curves may be hard to read and interpret due to the presence of oscillations that result from the dynamic nature of the test. Conversely, QSI tests can be carried out on a universal testing machine, requiring as additional equipment only an indenter and a specific equipment for fixing the specimen. Additionally, QSI maximum displacement can easily be monitored to investigate the damage sequence. Finally, the acoustic emission technique can be implemented safely, low acquisition rates suffice, and load-displacement curves are exempt of oscillations.

Nevertheless, the question arises regarding the potential use of QSI test in complement to LVI test. Some authors did not find any significant differences between both tests whereas others reported non-negligible dissimilarities. The study conducted by Nettles and Douglas (Nettles and Douglas, 2002) on quasi-isotropic carbon/epoxy laminated plates showed no distinct differences between QSI and LVI tests based on the maximum applied transverse load. Likewise, Suresh Kumar et al. (Suresh Kumar et al., 2017) reached the same conclusion on quasi-isotropic glass/epoxy, glass/basalt/epoxy and glass/carbon/epoxy composite laminates. Their results indicated that there were no significant differences with regard to the dent depth, back surface crack size and load-deflection behaviour. In particular, the changes in peak contact force and residual deformation were similar. Saeedifar et al. (Saeedifar et al., 2018) also found that the general behaviour of two quasi-isotropic carbon/epoxy laminates under QSI and LVI tests showed great similarity. However, they reported two differences: about 10 % maximum difference of the delaminated area and a significant increase in the critical load corresponding to the initial delamination growth for the LVI tests compared to the indentation tests. In (Wu et al., 2020), the QSI results for two carbon fibre braided composites were in good agreement with LVI tests before the peak load. After the peak load, the load measured in QSI was slightly higher than the impact load in LVI test. As a result, it was concluded that QSI tests for braided laminates could be used to analyse the damage onset and development during an impact event. In a paper of Zhang et al. (Zhang et al., 2015), the LVI and QSI tests on carbon/bismaleimide laminates resulted in a similar delamination shape and a similar general trend of delamination size throughout the thickness direction. In conclusion of their study, the authors claimed that using QSI-induced damage to replace LVI-induced damage made it possible to assess approximately equivalent strength in static compression, which was not recommended for compressive fatigue strength. The finding is completely different in another study on carbon/epoxy and glass/polyamide-6 composites based on cross-ply or quasi-isotropic stacking sequences (Spronk et al., 2018). Differences between both tests were found on the load-displacement response and were significant for the glass/polyamide-6 composite due to the constituent rate-dependency. Although some characteristics were relatively similar, the conclusion came down to the fact that the LVI and QSI test methods cannot be exchanged for material characterisation, according to the authors. Some differences in the first slope of the load-displacement curve were also found on [±45]4s carbon fibre laminates since the laminate under LVI was 36 % stiffer than in the QSI test (Serna Moreno and Horta Muñoz, 2020). However, similar levels of internal energy were found in the most notable events during the loading process. The results of Goodarz et al. (Goodarz et al., 2019) suggested that the limit of applicability of the quasi-static analysis for the dynamic problem of aramid/epoxy plain-weave laminates (with nanocomposites interlayers) was at impact energy level before beginning plate penetration. Zulkafli et al. (Zulkafli et al., 2020) investigated the effects of stacking sequence of hybrid cross-ply banana/glass fibre reinforced polypropylene composites on QSI and LVI. By comparing the damage assessment of the QSI and LVI specimens, the difference observed was located at the fracture level, since the LVI specimen was more brittle than the QSI specimen. According to the authors, this can be explained by the sudden impact applied to the specimens. It should also be noted that the damage area of the LVI specimens was much larger than the QSI specimens.

In summary, some of the available conclusions on a QSI/LVI comparison are contradictory and there is no real consensus on whether both tests are equivalent. Moreover, to the best of the authors’ knowledge, such a comparison for non-hybrid natural-fibre laminates has never been the main subject of investigation in any study. Up until now, such comparisons have been conducted for the sole purpose of providing a reference in the very few studies that deal with hybrid laminates (Jusoh et al., 2017; Malingam et al., 2018). This lack of consensus and data has prompted us to carry out our own investigations on the similarities between LVI and QSI for woven flax/epoxy laminates, with a double aim: (i) providing indications and advice to researchers and industrials who would be thinking of replacing LVI with the cheaper and more convenient QSI testing method; (ii) laying the foundation of our future work, which will consist in gaining a deeper insight into impact damage mechanisms and in proposing analytical and numerical models of LVI. For this purpose, flax twill-weave fabric laminates were manufactured using vacuum infusion process and then subjected to impact and indentation tests. Next, the obtained load-displacement curves were analysed separately and concomitantly for different levels of energy. The LVI/QSI comparison was also conducted on the absorbed energies in relation to the total energy. Finally, post-impact images and data obtained from a detailed analysis of the AE signals were used to study the damage occurring within the impacted and indented laminates.

Section snippets

Material and manufacturing process

A 2/2 twill fabric of flax untwisted rovings, with a surface weight of 360 g/m², was supplied by Depestele Group and used as reinforcement for our composites. Rectangular-shaped samples were cut out of the fabric roll to the dimensions of 350 × 400 mm² and stacked on top of each other to form a 6-layer preform. All plies were oriented in the same direction. The preform was then impregnated with the matrix via vacuum infusion as depicted in Fig. 1. The matrix consisted of epoxy resin SR 8100 and

Concepts and terminology

Data obtained from LVI and QSI tests is commonly represented as load-displacement curves in which the load applied to a specimen by the impactor or indenter is plotted against its vertical displacement. A typical QSI curve is shown in Fig. 3a. All the remarks and definitions regarding this figure also apply to the results of LVI tests. The curve is divided into two parts according to the evolution of the displacement values. Whereas they increase in the first part due to the indenting of the

Conclusions

Low velocity impact (LVI) and quasi-static indentation (QSI) tests have been performed on flax-epoxy woven laminates to investigate the similarities between both types of mechanical loading. Specimens were tested at four different energy levels (5, 10, 15 and 20 J), with minimal differences in energy values, below 3.9 %, between QSI and LVI. Similarities were found in load-displacement curves, energy absorption capacity and visible damage. Actually, the differences between QSI and LVI remain

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Adélaïde Leroy: Conceptualization, Methodology, Investigation, Writing - original draft. Daniel Scida: Conceptualization, Methodology, Investigation, Writing - review & editing. Emile Roux: Conceptualization, Writing - review & editing. Franck Toussaint: Conceptualization, Supervision, Writing - review & editing. Rezak Ayad: Conceptualization, Supervision, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to gratefully acknowledge the urban community “Grand Reims” and the “Université de Reims Champagne Ardenne” for their financial supports to the BIOIMPACT project in which this work is conducted.

References (42)

Cited by (10)

  • How to transform lignin into a useful component of flax fiber for composite materials

    2023, Industrial Crops and Products
    Citation Excerpt :

    The main arguments for using natural fibrous fillers instead of synthetic materials are, first of all, their availability and renewability of raw materials, low cost and biodegradability, as well as the stability of linear dimensions, corrosion resistance, and harmlessness to human health (Ramesh et al., 2017, 2020; Girijappa et al., 2019; Sahu et al., 2021; Prakash et al., 2022). Along with that flax-fiber composites are commonly used for non-structural or semi-structural applications (Leroy et al., 2021). The main problem is the insufficient level in their impact resistance, low tensile and compressive strength (Li et al., 2020).

  • Damage evolution in quasi-statically indented alumina based oxide/oxide ceramic matrix composites: An experimental investigation

    2022, Ceramics International
    Citation Excerpt :

    While a number of studies have looked at the “resultant state” of damage in CMCs due to QSI and/or low energy impacts [5–12], to date there has been no study that looked at the evolution of damage in time during the indentation or impact event. Multiple studies in PMCs have looked at the similarities in material response between LVI and QSI [30–35]. The key difference between the two modes of inducing damage is that while LVI is always considered a dynamic process, QSI tests are performed at such a low speed that the dynamic effects are negligible.

  • Influence of hybridization on the mechanical and dynamic mechanical properties of aramid/S2-glass hybrid laminates

    2022, Materials Today Communications
    Citation Excerpt :

    QSI results have also been used to study damage mechanism and to predict the dynamic behavior [4]. Leroy et al. [13] performed low velocity impact (LVI) and QSI tests on flax/epoxy laminates. The specimens were tested at four different energy levels (5–20 J), with minimal differences in energy absorption (below 3.9%) between QSI and LVI.

View all citing articles on Scopus
View full text