Assessing of the vulnerability of civil engineering structures under seismic loading is one of the central points of the reduction of the impact of the seismic hazard on common residential buildings, security-related civil buildings, or industrial plant buildings involving a “special risk” such as the nuclear industry. Possible failures built up an entry in the probabilistic seismic hazard analysis, for both building and equipment placed therein. Vulnerability must be analysed during the design stage and also during periodic reassessments, where a detailed analysis is required: geometry of building structures, material behaviour…
The R&D axes are:
- Understanding of the physical phenomena concerning the structures behaviour under seismic loading (nonlinearities, local or overall scale complex material constitutive relations, damping stemming from several dissipative physical phenomena…), based on an experimental approach (structural elements, models representing fully 3D structures) by means of the TAMARIS platform at CEA and at different scales modelling.
- Development and validation of models, notably used as a predictive tool in probabilistic approaches (conditional probability of failure and fragility curves), based on data obtained by tests to identify damage criteria and sensitivity to the models and constitutive relations used. Numerical simulations requiring nonlinear transient analyses are generally computation time consuming; that is why modelling at a global scale is developed from the local description aiming at process the structures of complex industrial buildings.
The current research actions are mainly interested in modeling the behavior of concrete and reinforced concrete under seismic loading. Indeed, these materials are essential constituents of many civil engineering structures.
Models for concrete under seismic loading
During the last fifty years, the scientific community has become aware not only of interest but also the difficulties of formulating constitutive models of concrete under cyclic loadings. In the 80s, the first uniaxial model of concrete appears. Although still widely used today, they are not sufficient when local information (crack width, tortuosity, etc..) are sought. Recently, a constitutive framework to take into account the main dissipation mechanisms was developed. This has been recognized as having desirable properties. However, an identification of the unilateral effect (crack closure effect) should be conducted. Issues related to unilateral effect criterion, its description and its formulation remain currently open. To provide some answers is clearly an expectation of the scientific community and an industrial needs.
In the framework of SEISM institute, research is conducted on the identification of unilateral effect. It is based on a dual approach combining discrete model and continuous model. Indeed, the discrete approach should help to provide local information on the mechanism of crack closure such that the criterion of reclosing (yield stress, deformation threshold, energy threshold, etc..). The set of identified information will then be incorporated in a continuous model in order to make computation on structures of large scale.
This work will improve the understanding of nonlinear phenomena and lead to a macroscopic model of concrete behavior.
Modeling of reinforced concrete under seismic loading
A major component of civil engineering construction is reinforced concrete. Modeling the behavior of reinforced concrete introduced an additional level of complexity with respect to the modeling of concrete. Indeed, it is necessary to take into account the interaction between steel and concrete and the degradation of steel. The aim of various work within SEISM institute is the development of structural element models integrating these mechanisms, eg energy dissipation(Christelle Combescure’s Thesis – EDF) and providing local information such as cracking (Ejona Kishta’s Thesis – CEA et ENS Cachan).
For reasons of computational efficiency, the developed models do not take into account all dissipative phenomena. To improve them, it is possible to enhanced the numerical damping introduced in the simulations. The experimental and numerical characterization of damping of structures under earthquake is part of the objectives of the institute SEISM (Romain Crambuer’s Thesis – ENS Cachan).