Post-impact flexural tests on jute/polyester laminates monitored by acoustic emission

Abstract

A number of industrial sectors, including automotive, building and leisure, have recently shown interest on natural fibers, as reinforcement for low-weight, environmental-friendly materials. The introduction of natural fiber composites in components for large volume applications requires studies aimed at predicting their in-service behavior, in particular fatigue and impact testing. Some recent investigations have raised concerns on the performance of these materials, especially when subjected to a high number of cycles, as an effect of a long-service life, in harsh environmental conditions [1‐3]. The use of nondestructive testing techniques, such as acoustic emission, as monitoring methods during mechanical tests, proved useful in previous studies on mechanical characterization of natural fiber composites [4, 5]. Two approaches are used to predict impact damage on laminated composites reinforced with artificial fibers [6]. The first one is based on estimating the overall size of impact-damaged area and considering stress distribution in the region surrounding the impact point, whilst the second one is aimed at detecting the appearance of the first matrix crack, then the study of the initiation and propagation of delamination. When dealing with plant-fiber composites, both approaches are viable, at least in principle: however, a number of difficulties can be perceived. First, the measurement of impact-damaged area has been revealed to be particularly difficult, as an effect of the fibers becoming loose and suffering early debonding around the impact point, even at low stress [7]. Second, the study of impact damage initiation is based on the assumption that the laminate shows limited presence of defects prior to impact and that the direction of impact, whether mono- or bi-dimensional, determines the damage propagation mode. However, in biological materials, such as plant fibers, the combined presence of stronger and weaker parts is a natural layout, aimed at obtaining the maximum possible impact resistance. In other words, natural materials work effectively through the limited and controlled occurrence of defects [8]. In plant fiber reinforced composites, because of the larger dimensional variability of fibers and fiber bundles, defects can more easily lead to disruptions in the laminate geometry than in glass fiber reinforced composites [9]. In addition, the capability of detecting defects and especially measuring their level of criticality is crucial to arrive to the possible production of components with plant fibers reinforcement. In this context, the acoustic emission (AE) technique can present some interest, for its capability of providing data from real-time monitoring during post-impact testing [10]. In the present work, plain woven jute fabric/polyester plates (100 mm×100 mm), manufactured using a resin transfer moulding (RTM) process with a 60% wt. reinforcement content, have been impacted and then subjected to cyclic post-impact three-point bending tests. The impact energy was changed, by varying the mass of the hemispherical drop-weight steel impactor, whilst maintaining the impact velocity constant at 2 m/s (±5%). Impact tests were conducted on a CEAST Fractovis impact tower fitted with an anti-rebound device. The plates were divided into eight categories with five specimens in each category. One group of specimen was not impacted, while the remainder were impacted at energies of 5, 7.5, 10, 12.5, 15 and 20 J, and the last one was impacted to penetration according to ASTM D3763, always using a hemispherical impactor tip diameter 12.7 mm. Depending on the impact energy, the total mass of the impactor varied: 2.53 kg (5 J), 3.89 kg (7.5 J), 4.99 kg (10 J), 6.26 kg (12.5 J), 7.54 kg (15 J), and 9.97 kg (20 J). The penetration energy measured on these laminates was equal to 25 (±2.2) J. On the impacted specimens, a 25.4 mm diameter indentor was used to undertake cyclic three-point bend tests, using an Instron 9400 universal testing machine, and following the scheme in Fig. 1. A loading rate of 4 mm/min was used. This condition was the