Abstract

The development of robust and computational competitive numerical techniques for the physically non-linear analysis of concrete structures is the main objective of this research project. To support the numerical computations, hybrid and mixed finite element models will be used. The performance of meshless methods will also be assessed.

Since the decade of 90 the research team has been involved in the development of effective alternative hybrid and mixed finite element formulations. An extensive validation program and performance evaluation allowed the identification of the most relevant properties of these formulations. It made also possible the recognition of the models that possess the characteristics that allow them to overcome some of the limitations associated with the use of the classical finite element formulations. In this context, one of the most powerful tools is the hybrid-mixed stress model, due to the flexibility in the selection of the approximation functions and to the accuracy that can be achieved in the computation of the stress field. More recent it is the involvement of the group in the development of meshless methods; very encouraging results have already been obtained with the use of the Moving Least Squares technique.

After establishing the properties of these alternative formulations, the following challenge consisted in extending their application into the analysis of problems closer to Engineering practice. In a first step, elastoplastic models for the analysis of plane structures have been developed. After that, continuum damage models for the physically non-linear analysis of concrete structures have been studied and implemented. The research group has acquired some experience in the development and use of this type of models. The research in this area evolved in the framework of a research project already concluded (POCTI/ECM/33066/2001) and the main results have been presented in a PhD thesis.

The physically non linear analysis of concrete structures is still a challenge and corresponds to an application field where the high performance alternative formulations can be conveniently explored. The fracture processes associated to the propagation of defects strongly influence the structural behaviour of concrete structures and may be modelled using continuous or discrete approaches. The constitutive models able to reproduce diffuse micro-cracking in a continuum without creating a real discontinuity in the material lie in the first category. The Continuum Damage Mechanics is the most popular continuum approach. One strong drawback associated to the application of pure continuous models lies in the fact that fragile rupture is frequently governed by the growth of a dominant crack. The second approach is suited for these cases. A discrete approach corresponds then to introduce in the model a displacement discontinuity with a cohesive constitutive behaviour. Although both approaches continue to deserve the scientific community attention, the current trend is clearly to combine the two methodologies. In such a way, the analysis is initiated with the application of a continuum damage model. From a given damage value or from a critical size of the strain localization bandwidth, a displacement discontinuity is introduced in the model and the evolution of the process is modelled with a cohesive crack law.

The project is organized in two tasks. Task 1 is centred in the extension and the generalization of the damage models developed in the framework of the PhD thesis and the research project mentioned above. In a first stage, coupled plasticity and damage models will be implemented and assessed. Then, the models will be generalized in order to make possible the consideration of dynamic loadings. Finally, the first steps towards the generalization of the existing models to 3D analysis will be undertaken. The objective of Task 2 is the study and implementation of cohesive fracture models for concrete analysis. In this context it is important to explore the features of the hybrid and mixed finite element models in the accurate representation of singularities and discontinuities. The first goal of this task corresponds to the development of numerical tools to simulate the formation and the evolution of the discrete cracking. In a second stage adequate transition techniques to move from a pure continuous analysis based on damage mechanics models to a discrete analysis based on the consideration of cohesive crack models will be studied, implemented and validated.

For the development of each one of the research tasks, the launching of a doctoral program is planned. For the development, calibration, validation and performance evaluation of the new models it is important to ensure the support of foreign external consultants, with the purpose of complementing the experience and expertise of the research team, and to strengthen the cooperation links with the respective institutions.