Search

The response of masonry structures to explosions can be hardly investigated relying only on numerical and analytical tools. Experimental tests are of paramount importance for improving the current comprehension and validate existing models. However, experiments involving blast scenarios are, at present, partial and limited in number, compared to tests under different dynamic conditions, such as earthquakes. The reason lies on the fact that full-scale blast experiments present many difficulties, mainly due to the nature of the loading action. Experiments in reduced-scale offer instead greater flexibility. Nevertheless, appropriate scaling laws for the response of masonry structures under blast excitations are needed before performing such tests. We propose here new scaling laws for the dynamic, rigid-body response and failure modes of masonry structures under blast loads. This work takes its roots from previous studies, where closed-form solutions for the rocking response of slender blocks due to explosions have been derived and validated against numerical and experimental tests. The proposed scaling laws are here validated with detailed numerical simulations accounting for combined rocking, up-lifting, and sliding mechanisms of monolithic structures. Then, the application to multi-drum stone columns is considered. In particular, we show that, whilst the presence of complex behaviors, such as wobbling and impacts, similarity is assured. The developments demonstrate their applicability in the design of reduced-scale experiments of masonry structures.

International audience; We investigate the dynamics of inverted pendulum structures under fast-dynamic excitations arising from an explosion. We model blast actions using established empirical models and best-fit interpolations of existing experimental tests. We focus attention on pure rocking response mechanisms. We present moment balance equations and overturning conditions. Inspired by previous works in the frame of earthquake engineering, we derive new analytical, closed-form solutions for the rocking response and the overturning domain of slender blocks due to explosions.The analytical findings and assumptions are validated through existing experimental tests and detailed three-dimensional numerical simulations, which consider the full interaction between the blast waves and the structure. We show that unilateral rocking response and overturning are predominant mechanisms compared to sliding and up-lifting.Finally, we develop design charts to be used as a straightforward decision making tool for determining the critical stand-off distance between the explosive source and the target in order to prevent overturning. Our model can be easily applied for the design of protective devices to preserve artefacts, secure buildings and humans, as well as for devising energy absorbing systems based on rocking.

International audience; Machine Learning methods and, in particular, Artificial Neural Networks (ANNs) have demonstrated promising capabilities in material constitutive modeling. One of the main drawbacks of such approaches is the lack of a rigorous frame based on the laws of physics. This may render physically inconsistent the predictions of a trained network, which can be even dangerous for real applications. Here we propose a new class of data-driven, physics-based, neural networks for constitutive modeling of strain rate independent processes at the material point level, which we define as Thermodynamics-based Artificial Neural Networks (TANNs). The two basic principles of thermodynamics are encoded in the network's architecture by taking advantage of automatic differentiation to compute the numerical derivatives of a network with respect to its inputs. In this way, derivatives of the free-energy, the dissipation rate and their relation with the stress and internal state variables are hardwired in the architecture of TANNs. Consequently, our approach does not have to identify the underlying pattern of thermodynamic laws during training, reducing the need of large datasets. Moreover the training is more efficient and robust, and the predictions more accurate. Finally and more important, the predictions remain thermodynamically consistent, even for unseen data. Based on these features, TANNs are a starting point for data-driven, physics-based constitutive modeling with neural networks. We demonstrate the wide applicability of TANNs for modeling elasto-plastic materials, using both hyper-and hypo-plasticity models. Strain hardening and softening are also considered for the hyper-plastic scenario. Detailed comparisons show that the predictions of TANNs outperform those of standard ANNs. Finally, we demonstrate that the implementation of the laws of thermodynamics confers to TANNs high degrees of robustness to the presence of noise in the training data, compared to standard approaches. TANNs ' architecture is general, enabling applications to materials with different or more complex behavior, without any modification.

International audience; Masonry structures are often characterized by complex, non-planar geometries. This is also the case for historical and monumental structures. Here we investigate the dynamic behaviour of non-standard, curvilinear masonry geometries, such as vaults, subjected to blast loading.We use the Discrete Element Method (DEM) for modelling the dynamic structural response to explosions. The approach allows considering the detailed mechanical and geometrical characteristics of masonry, as well as the inherent coupling between the in- and out-of-plane motion.The proposed modelling approach is validated with existing experimental tests in the case of planar masonry geometries, walls. The DEM model well captures the dynamic response of the system and the form of failure within the masonry structure.Then the response of a curved masonry structure subjected to blast loading is investigated. The influence of various micro-mechanical parameters, such as the dilatancy angle, the tensile strength and the cohesion of the masonry joints on the overall dynamic structural response of the system is explored. The effect of the size of the building blocks is also studied.Finally, the common DEM assumption of rigid blocks is assessed through detailed comparisons with simulations involving deformable blocks.

International audience; The response of museum artefacts and statues subjected to deliberate explosions of moderate intensity is investigated and their vulnerability assessed. The study focuses on the most predominant failure mechanisms, namely overturning and fracture due to the tensile stresses developed by the impact of shock waves. The rocking response is investigated relying on the existing knowledge and theory of inverted pendulum structures subjected to earthquake loading. An analytical, established approach for determining the overturning domain, developed in a previous study, is used to investigate the critical stand-off distance between the target and the explosive in order to avoid toppling. The proposed analytical model is adopted by defining appropriate correction parameters to consider the real geometry of the museum objects. We assess the overturning domain of some emblematic statues of high aesthetic and cultural value, and namely: Michelangelo's David, Farnese Hercules, Aphrodite of Milos, Athena Giustiniani, Laocoön and His Sons, and Belvedere Torso. Finally, direct damage due to the high tensile stresses is investigated and computed for different target geometries. For the dimensions and explosive quantities herein considered, overturning is found to prevail over direct material damage for targets with regular geometry.