Conference Agenda

With the rapid advancements in experimental techniques and the growth in the amount of available data, especially when using digital image correlation and tomography, modelling is facing the challenge of transforming this enormous amount of knowledge into analytical equations that govern the material response. Classical constitutive equations may struggle to capture complex material responses, which is among the reasons why data driven approaches emerged. In this study, we apply a data-driven scheme to the modelling of failure of a quasi-brittle material such as concrete and discuss the difficulties induced when modelling localized failure. For the sake of simplicity, we consider a one-dimensional problem. Synthetic data sets are generated from a bi-linear damage model and exhibit strain softening. As expected, the one-dimensional example of a bar subjected to tension demonstrates that the obtained solutions are sensitive to the finite element discretization. A localization limiter is needed and the implementation of a non-local (integral) model circumvents the difficulty. There is, however, a notable observation in this case: optimal sets of strain, stress, and non-local history variable lie consistently outside the data set and do not converge within the data set upon mesh refinement. Several possibilities for solving this problem are considered, from the enlargement of the data set with non-local effects to the introduction of an additional constraint e.g., following the Lip-field approach. The latter method preserves locality of the constitutive response and it is found to be very easy to implement.

Concrete cracking is analyzed using nonlinear fracture-mechanics-based constitutive models within a finite element framework. In such simulations, the predicted behavior of reinforced concrete is high-ly sensitive to the assumed crack spacing or crack band size, particularly when relatively large finite elements are employed. To alleviate this limitation, the present study introduces an approach in which artificial neural network surrogate models are used to estimate the crack spacing in reinforced concrete structures. Model uncertainties in terms of mean and maximum crack width are evaluated against a database of laboratory tests. The influence of reinforcement layout, geometric simplifications and mesh discretization on these uncertain-ties is examined. Overall, the proposed modelling strategy introduces an advanced tool for assessment of crack widths and mainly crack spacing in reinforced concrete structures at the serviceability limit state.

Reinforced concrete structures traditionally rely on Bernoulli-Euler beam theory, which assumes linear stress distribution. However, in D-regions areas near support or geometric discontinuities these assumptions break down, and more advanced methods are required, Strut and Tie method (STM) is one of the most popular. STM models rely heavily on mechanical judgment of the engineer, to determine optimal configurations, creating barriers to efficient design. This study examines machine learning algorithms to predict internal forces in deep beams with consider as D-regions. A hybrid ML/AI data bases were evaluated using standard formulation, and validated for three ML/AI approaches, linear regression, with and without polynomial expansion (2nd and 3rd order), and Artificial Neural Networks (ANN). The most advanced linear regression model and the ANN model achieved excellent accuracy, which is appropriate for structural design. The results demonstrate that machine learning can effectively automate STM analysis for deep beams, eliminating expert-dependency and enabling efficient structural design more types of D-regions.