Computational Generation of Virtual Concrete Mesostructures

Abstract

Concrete is a heterogeneous material with a disordered material morphology that strongly governs the behaviour of the material. In this contribution, we present a computational tool called the Concrete Mesostructure Generator (CMG) for the generation of ultra-realistic virtual concrete morphologies for mesoscale and multiscale computational modelling and the simulation of concrete. Given an aggregate size distribution, realistic generic concrete aggregates are generated by a sequential reduction of a cuboid to generate a polyhedron with multiple faces. Thereafter, concave depressions are introduced in the polyhedron using Gaussian surfaces. The generated aggregates are assembled into the mesostructure using a hierarchic random sequential adsorption algorithm. The virtual mesostructures are first calibrated using laboratory measurements of aggregate distributions. The model is validated by comparing the elastic properties obtained from laboratory testing of concrete specimens with the elastic properties obtained using computational homogenisation of virtual concrete mesostructures. Finally, a 3D-convolutional neural network is trained to directly generate elastic properties from voxel data.

Impact of External Mechanical Loads on Coda Waves in Concrete

“…The polycarboxylatether-based superplasticizer was added to the mixing water to improve the concrete’s workability and reduce the cement paste viscosity to enhance the sensor-to-concrete coupling condition. The characteristic concrete parameters were additionally modeled [ 23 ], particularly considering the concrete model’s damage evolution [ 24 ]. After production, the specimens were cured at 20 °C under water for 7 days and restored to 65% RH and 20 °C.…”

“…The immediate strain , together with a plastic term , is complemented by two terms that account for shrinkage and creep for a constant compressive stress during the loading period between and . Equation ( 2 ) [ 23 ] summarizes the total strain ( Figure 3 ) for a load application at at constant environmental conditions. …”

Numerical Simulation-Based Damage Identification in Concrete

“…This section presents a brief summary of the entire workflow illustrated in Figure 1. First a realistic synthetic concrete specimen was generated using the in-house Concrete Mesostructure Generator (CMG) code [31]. Next, a uniaxial compression simulation was performed on the synthetic concrete specimen by means of the Discrete Element Method [29].…”

A Statistical Finite Element Method Integrating a Plurigaussian Random Field Generator for Multi-scale Modelling of Solute Transport in Concrete

“…In order to simulate multi-scale solute transport in concrete, a representation of both micro-and meso-structure is required. Whilst such representations can be obtained from images, it can be advantageous to generate such structures virtually (Holla et al 2021). For cementitious materials, the most commonly employed approach to obtaining a virtual micro-structure is to use a cement hydration model such as HYMOSTRUC3D (van Breugel 1995;Koenders 1997) or CEMHYD3D (Bentz 2005).…”

“…A key challenge in such models is the representation of the complex shapes exhibited by aggregate particles, with a number of models idealising their shape as spherical or ellipsoidal (Zheng et al 2009;Li et al 2016). In order to generate more realistic concrete meso-structures, a number of techniques for capturing the complex shape of aggregate particles have been proposed (Qian et al 2016;Zhang et al 2018;Holla et al 2021). In the Anm model, developed by Qian et al (2016), aggregate particle shapes are represented using spherical harmonic functions.…”