PSI - Issue 28

Mohamed Ali Bouaziz et al. / Procedia Structural Integrity 28 (2020) 393–402 Author name / Structural Integrity Procedia 00 (2019) 000–000

394

2

traditional material manufacturing processes. Mechanical characterizations of fabricated structures are key for their good use in industrial applications. The fracture energy is an important mechanical property in materials engineering . The good understanding of mechanical features of materials requires their characterization at different scales. In the case of additively manufactured materials, many experimental and numerical approaches have been developed for the characterization at local scales. Different works have focused on the development of approaches of kinematic field measurements (Ciccotti et al. 2010; Clair et al. 2011; Djouda et al. 2020; Marae-Djouda et al. 2020; Marae-Djouda et al. 2018; J. Marae Djouda et al. 2017; Marae Djouda et al. 2018; Petit, Montay, and Hild 2014; Roux, Réthoré, and Hild 2009). Digital image correlation (DIC) appeared like one of the most mature experimental technique because of its ease of use in a wide range of materials including AM materials. It allows the measurement of kinematic fields to be performed at very local scales (i.e. micrometric or nanometric scales). Based on these kinematic fields around a crack, Cherepanov–Rice's integral (Rice 1968) was used in some works to directly calculate the energy release rate from strain fields around a crack. In the experimental implementation of DIC for the characterization of kinematic fields and fracture parameters, many studies were developed. In order to measure the crack resistance curves of cross-ply CFRP composite laminates,Catalanotti et al. (2010) developed a method based on the measurement of kinematic fields using DIC. In order to evaluate the J -integral for a power-law hardening material, Yoneyama et al. (2014) used three different methods; the path integral method, the domain integral method and the least squares method with the Hutchinson, Rice and Rosengren (HRR) fields (Rice 1968), (Rice and Rosengren 1968), (J. W. Hutchinson 1978). The study of Yoneyama et al. showed that these methods enabled the J -integral to be evaluated with good accuracy (Lanzillotti et al. 2019; Yoneyama et al. 2014). Some hybrid methods were also developed (Barhli et al. 2017; Hareesh and Chiang 1988; Mathieu et al. 2013; Nishioka et al. 2001). The measured displacement fields were used as boundary conditions for finite element analyses. Becker et al. (2012) evaluated the J -integral from the measured crack displacement field, which they coined JMAN. This method was applied to AISI 316L stainless steel. Excellent agreement with other fracture characterization techniques was achieved when testing JMAN on elastic, elastoplastic and quasi-brittle materials (Becker et al. 2012; Becker, Marrow, and Tait 2011). The advantages put forward by the authors of the works cited above over other methods are geometric freedom, i.e. the method can be applied to any type of tested sample (Barhli et al. 2017; Catalanotti et al. 2010; Yoneyama et al. 2014). It is also insensitive to inelastic strains close to the crack tip (Becker et al. 2012; Becker, Marrow, and Tait 2011). However, to the authors’ knowledge, the implementation of the direct experimental use of DIC or hybrid methods for fracture parameter evaluations was not performed yet in the specific case of AM polymeric materials. It is expected that J -integral evaluations using directly the measured crack displacement field methods and/or hybrid method are likely to be transferable to AM polymeric materials. In this study, a combined computational and experimental investigation was performed to study the fracture behavior of additively manufactured ABS materials. A micro single edge notch tension (µSENT) specimen was used. Measured displacement and strain fields using DIC (Djouda et al. 2020; Djouda et al. 2020) were coupled with finite element simulations to evaluate fracture mechanics parameter (i.e., J -integral) and to measure crack length extension. Two methods were investigated. On the one hand, FE calculations were run with ABAQUS using measured displacements as boundary conditions. The crack tip position was found by minimizing the error between computed and measured displacement fields. The J -integral was assessed with the built-in interaction integral (Brocks and Scheider 2001). On the other hand, MATLAB scripts were developed to calculate the J -integral using measured kinematic fields. 2. Experimental set-up and protocol A  SENT specimen was 3D printed with the dimensions shown in Figure 1(a). The notch was also made by AM. The notch dimension meets ASTM E1820 (ASTM-E1820–11 2011) and ASTM D6068 (ASTM-6068 2013) standard recommendations, namely, specimen width and initial crack length. The specimen was printed by adding molten layers of Acrylonitrile Butadiene Styrene (ABS) using a Makerbot replicator 2X. The sample was printed as a solid part, in a flat [+/-45°] orientation (Figure 1(b)). In order to ease surface functionalization, the specimen was mechanically polished. The thickness was reduced from 6 mm to 3 mm. White speckles with micrometric dimensions were deposited using an airbrush (average diameter equal to 20 µm) (Marae-Djouda et al. 2020).