Most of mechanisms at the origin of cracking in concrete occur in the cement paste. Its microstructure evolves rapidly at early-age with the formation of hydration products, the decrease of the total porosity, the reduction of the pores size and the increasing of the tortuosity. The development of mechanical properties, the time-dependent behavior and the transport properties of concrete rely directly on the evolution of the cement paste microstructure. The objectives of this GP are to discuss the most relevant ways of modelling the microstructure development of cement pastes, report recent advances and make comparative reviews of models.
The large variety of cement, additions, etc. as well as the cost and time/devices required to the realization of experiments renders difficult to optimize the concrete mix regarding specifications (material properties, impact on global warming). Virtual testing is an interesting alternative. Starting from the description of the cement paste microstructure (GP2.a), upscaling of material properties must be undertaken up to the concrete scale. The objectives of this GP are to discuss the most relevant ways to deal with virtual testing, report recent advances and make comparative reviews of numerical strategies (analytical homogenization, finite element calculations, etc.).
The prediction of cracking in reinforced concrete structures depends not only on the material properties of concrete but also on the mechanical boundary conditions, climate conditions and geometry. Macroscopic models capable of describing time-dependent behavior, transport phenomena and cracking are thus required. This is important in order to optimize the materials and construction techniques and to interpret data in monitored structure in a relevant way. These types of multi-physics macroscopic models may be suitable for use in structural design offices (e.g. embedded into finite element software packages). The objectives of this GP are to discuss the most relevant ways to deal with macroscopic modelling, report recent advances and make comparative reviews of modelling strategies (balance equations, mechanics and thermodynamics of porous media, simplified approach, etc.).
Many degradation mechanisms are associated with the movement of ions through concrete. Therefore modelling of transport processes is essential to determine appropriate service-life of concrete structures. The anticipated outcome of this GP is to establish a recommendation for appropriate models for predicting transport properties of concrete and modelling approaches to simulate chloride transport through concrete. The specific goals of this GP are:
(a) establishing link between models at different scales starting from scale of cement paste microstructure (GP2.a) to determine the transport properties of concrete in conjunction to work of GP2.b;
(b) Using this links between microstructure and properties of concrete to simulate the chloride transport in concreto;
(c) testing predictability of different models for simulating lab scale and field scale chloride transport experiments.
A fundamental tool for assisting development, understanding and comparison of both micro-scale and macro-scale models mentioned in the aforementioned GPs are benchmarking tests. It will be possible to evaluate the relative performance of numerical simulation approaches and software (including research and commercial applications) applied to selected real examples of material and structural behavior. Both blind and subsequent non-blind stage of benchmarking have to be included, in relation to the round-robin testing realized in the WG1. The objectives of this GP are to develop relevant strategies for establishment of benchmark studies and compare/analyze the results of different participants in order to propose validation tests and recommendations.