PSI - Issue 79

Mansi Gupta et al. / Procedia Structural Integrity 79 (2026) 259–265

260

E-mail address: bhowmiksonali@nitrkl.ac.in impact the crack tortuosity and trajectory, thereby the overall failure behavior (Trawi ń ski et al. (2018)). The classical experiments and numerical models treat concrete as a homogeneous material. Therefore, in order to thoroughly understand the fracture behavior, the multiphase characteristics of concrete need to be analysed. The mesoscale simulations represent the largest length scale where pronounced concrete heterogeneities can be modelled while maintaining the computational efficiency. In recent years, these simulations have emerged as an effective and reliable method for investigating concrete fracture, while considering its heterogeneity at the same time (Wang et al. (1999), Kwan et al. (1999)). At mesoscale, concrete can be explicitly modelled as a mixture of mortar and aggregates. The key advantage of mesoscale modelling is its ability to represent interfacial transition zone (ITZ), which has lower strength and stiffness in comparison to surrounding mortar, making it a susceptible cite for crack development and ultimate failure (Kargari et al. (2023)). The low density and high porosity of ITZ facilitates the microcrack propagation and is often the primary location for failure (Chen et al. (2024)). Mesoscale modelling provides crucial insights into the strain localization, stress distribution and damage evolution (Carrara et al. (2018)). At the same time, it enables the derivation of macroscopic load-deformation responses; making it a robust and efficient approach. There are two main steps in mesoscale modelling: mesostructure representation and subsequent numerical analysis. The mesostructure representation can be done either by digital image-based approach or by parameterization modelling (Thilakarathna et al. (2020)). In digital image-based approaches, the concrete specimen is scanned using X-ray computed tomography (XCT) and the images are processed using image processing techniques to generate 2D or 3D mesostructure (Yang et al. (2019)). In the second approach, the mesostructure is generated by algorithms representing the mesostructure geometry parametrically. In this approach, care must be taken for aggregate shape, spacing and the particle size distribution. Fuller curve is generally used to represent the particle size distributions in the mesostructure geometry in various 2D or 3D simulations (Wang et al. (2015), Wriggers and Moftah (2015)). For the numerical analysis, the methods include lattice model, discrete element method, and finite element method (Thilakarathna et al. (2020)). In numerical simulation, the cohesive zone model (CZM) is generally adopted for capturing the potential cracking in the mortar or ITZ. The CZM is phenomenological damage framework widely used in fracture studies, capturing fracture using cohesive law based on traction-separation relationship (Wang et al. (2015), Trawi ń ski et al. (2016)). The damage evolution in CZM is governed by various softening laws that aptly capture the fracture behavior in quasi brittle materials like concrete (Roesler et al. 2007). The pioneering work by Hillerborg et al. (1976) proposed a fictitious crack model for concrete with linear softening law. Subsequently various softening laws like bilinear (Petersson (1981), Bazant and Pfeiffer (1987)), and exponential (Gopalaratnam (1985)) were applied to accurately represent the fracture of plain and reinforced concrete. In this study, the effect of aggregate shape and volume has been studied on the fracture behavior of concrete using mesoscale modelling. The mesostructure generation is done by following the methodology of Wang et al. (2015) and 2D circular, elliptical and polyhedron aggregates were modelled. The FE software, Abaqus/CAE (Manual (2013)) was used for numerical simulations. The effect of aggregate shape and volume fraction was studied on the fracture of geometrically similar concrete beams under three-point bending. 2. Procedure for mesoscale modelling 2.1. Experimental details The fracture behavior in concrete has been studied for the set of three-point bend beams from the existing literatures (Zhu et al. (2024), Trawi ń ski et al. (2016), Trawi ń ski et al. (2018), Bhowmik and Ray, (2019)). The beam geometry with dimensions 320 x 80 x 40 mm, notch 8 x 3 mm, and effective span length of 240 mm was used for the modelling of mesostructure with different aggregates shape and volume fractions (Zhu et al. (2024), Trawi ń ski et al. (2016), Trawi ń ski et al. (2018)). The dimensions for geometrically similar beams were taken from the experimental data of Bhowmik and Ray (2019), with notch to depth ratio of 0.2, and span to depth ratio of 4. For the sets of experimental beams are simply supported with displacement loading at the mid span. The details of the beam dimensions are provided in Table 1. The plain concrete beams are composed of ordinary Portland cement, aggregates (coarse and fine) and water. The aggregates particles with size less than 2 mm was include as a part of mortar matrix and is not modelled in the mesoscale geometry (Zhu et al. (2024), Trawi ń ski et al. (2016)).