Issue 48

C.A.C.P. Coelho et alii, Frattura ed Integrità Strutturale, 48 (2019) 411-418; DOI: 10.3221/IGF-ESIS.48.39 412 However, the main disadvantages of those materials are related with their poor compression and transverse properties [1, 2]. In this context, literature reports a high sensitivity of these materials to low-velocity impact events that occur easily in- service or during the maintenance activities. These type of events are responsible by damages that are difficult to detect [3, 4] and, simultaneously, responsible by significant reductions of the residual mechanical properties [5–8]. Studies about composite laminates subjected to low-velocity impact events on the through-thickness direction are abundant in the literature, especially in terms of damage characterization [9-12], compression-after-impact [13], multi-impacts [14], environmental effects [15-17] and numerical investigations [18, 19]. On the other hand, literature also presents strategies to improve their impact performance with resource to hybridisation [20-22] or using nano-enhanced resins [23–27]. However, the impact characterization in all these works was performed in composite plates. The analysis of laminate cylindrical shells, are less disseminated in the open literature, but this topic should be empathized due to the complexity and design of many advanced structures. Gong et al. [28], for example, developed an analytic solution to predict the response of laminate shells subjected to impact loads. This solution included both contact deformation and transverse shear deformation and it was used to study the effects of different impact conditions and shell size, as well as the curvature´s effect, on the contact force and central deflection of the shell. Results were compared with those reported in the literature and good agreement was found. An experimental study was developed by Kistler [29] and he noticed that the geometry strongly influences the impact response. Stiffer structures have higher impact strength, smaller centre deflections and shorter contact time. A three- dimensional eight-node non-conforming element with Taylor’s modification was used by Zhao and Cho [30] to analyse the interlaminar stress distribution, initial damage pattern and progressive failure on laminate composite shells subjected to impact loads. The stiffness effect was also studied by Arachchige et al. [31, 32] and the results shown that the impact load increases with increasing stiffness while the contact time decreases. Regarding the velocity of impact, it was verified a direct relationship with the contact load. Kistler and Waas [33] studied the impact behaviour of cylindrical graphite/epoxy panels with different thicknesses, curvatures and support conditions. They concluded that increasing the thickness leads to increased stiffness and, consequently, higher impact force as well as lower deflection and contact time. Krishnamurthy et al. [34, 35] studied the damage and the impact response of cylindrical graphite/epoxy shells using the finite element method. According with the study developed, higher mass values of the impactor are responsible by the increasing of the contact time and the damage occurred under the impact point. A nonlinear finite element analysis of impact response and impact-induced damage in curved composite laminates subjected to transverse impact by a foreign object was developed by Kumar [36]. It was possible to conclude that the impact response is significantly dependent on the shell curvature. Therefore, this study intends to improve the knowledge related to the impact response of hybrid composite cylindrical shells. Combining two or more fibre types, hybridisation is a promising strategy to toughen composite materials, and a better balance of the mechanical properties is obtained relatively to non-hybrid composites. For this purpose, laminates with the same number of layers, but composed by different type of fibres, were manufactured and conveniently characterized in terms of static and impact strength. Both loading modes were tested with the same boundary conditions, where the curved edges of the test specimens were free while the straight edges were supported. E XPERIMENTAL PROCEDURE aminate composite cylindrical shells were manufactured by hand lay-up and overall dimensions are shown in Fig 1. A system of SR1500 epoxy resin and a SD2503 hardener standard, both supplied by Sicomin (Châteauneuf-les- Martigues, France), was used with six layers, all in the same direction, of bi-directional woven fabrics to produce the specimens. Three different typologies were investigated with the following stacking sequence: 6C; 2C+2K+2C and 2C+2G+2C, where the “number” represents the number of layers used and C = Carbon (taffeta with 196 g/cm 2 ), K = Kevlar (taffeta with 281 g/cm 2 ) and G = Glass (taffeta with 205 g/cm 2 ) fibre layers. Carbon bi-directional woven fabrics were supplied by CIT (Legnano, Italy), while the Kevlar woven fabrics were supplied by DuPont (Richmond, USA) and glass fibre woven fabrics by Porcher Industries Germany (Erbach, Germany). The system was placed inside a vacuum bag for 24 hours and a maximum pressure of 0.5 mbar was applied for 9 hours in order to maintain a constant fibre volume fraction and an uniform laminate thickness, beyond to eliminate any air bubbles existing in the laminate. According the supplier’s datasheet, the post-cure was carried out in an oven at 60ºC for 16h. The T g of the resin is about 70ºC. Low-velocity impact tests were performed using a drop weight-testing machine IMATEK-IM10 (Old Knebworth, United Kingdom). More details of the impact machine can be found in [37]. An impactor diameter of 10 mm with a mass of 2.826 kg was used. As shown in Fig. 2, the impact will occur at the centre of the samples with free support of the curved edges while the straight edges are bi-supported. The impact energy used was 5 J, which corresponds to an impact velocity L