Solution strategies based on discrete crack framework for short and long-term analysis of reinforced concrete structures

Abstract

Reinforced concrete is prevalently used in the construction industry. However, the simplified models in existing design guidelines usually result in conservative solutions that do not explore the full nonlinear capacity. While discrete modelling techniques can provide detailed modelling capabilities, their computational complexity and intensive data requirements have limited their adoption in practical designs. This thesis extends the application of discrete modelling techniques to address different nonlinear mechanisms such as cracking, crushing, and bond-slip typical of reinforced concrete structures to finally deploy them in a proposed AI-assisted design framework. A new reinforcement model with an embedded slippage behaviour is proposed as to enable flexibility in the modelling of reinforcement layouts. To benefit the full capabilities of the discrete formulation in complex nonlinear simulations, a new robust solution-finding algorithm based on the maximisation of energy dissipation is proposed. The resulting computational framework accounts for crack initiation and propagation, reinforcement elongation and slippage, and concrete crushing. The individual models in a discrete framework are coupled together to address the interrelated behaviours like tension-stiffening, strain-softening, and time-dependent. The solidification theory is employed to account for creep and shrinkage in sustained loading conditions. Furthermore, a new damage model is proposed to address the viscoelasticity inside the fracture zones. After verification, the model is put into use by generating an extensive dataset to train a surrogate machine-learning model for providing optimised design solutions exploiting the full nonlinear capacity. The proposed framework demonstrates enhanced predictive capabilities and efficiency in the nonlinear analysis of reinforced concrete structures, highlighting the potential to provide performance-based design solutions with optimised targeted material use.