VD - Villes Durables

Maîtrise durablE de la Fissuration des InfraSTructures en bétOn – MEFISTO


Sustainable control of mechanical cracking of concrete infrastructure

Improved predictive models describing the cracking of concrete structures for different types of mechanical loadings

Lifetime management of civil works, from the design stage or during the service-life, is a major concern of public or private infrastructure owners. Evaluation and prediction of structures performance involves the control of concrete cracking. The National Project CEOS.fr (Design and evaluation of special structures: cracking - shrinkage) was launched in 2008 to answer these questions. The MEFISTO project, scientific branch of the National Project, aims to develop and validate new ways of predicting cracking of reinforced concrete structures under various loadings (thermal, hydric and mechanical couplings).<br /><br />The development of numerical tools allows a better estimate of the behavior of structures in terms of design (evolution of standards, codes) and reassessment (increase of structures service-life). Technology transfer between the results obtained within this ANR project and expectations of engineers in Civil Engineering is a constant objective in the MEFISTO project.<br />

To improve the prediction of cracking in structures, two major ways are considered:

- Continuous methods
These approaches are based on the use of constitutive equations based mainly on the concrete damage / plasticity theories that do not describe a crack in a discrete manner but rather represent a degradated area. The objective is to use all information available to represent the position of macro-cracks and crack openings in a post-processing tool for finite element simulations (2D or 3D) but also for approaches called «simplified« (approaches like multi-fiber beams)

- Mixed methods
Mixed approaches introduce cracking at the center of the simulation. By the joint use of continuous and discrete approaches, the methods developed take the benefits of both approaches. The challenge is to make two models coexist in the same calculation by making them work in parallel.

In addition to these two approaches, special attention is paid to the consideration of uncertainty in predicting the opening and spacing of cracks. Reliability-tools were developed to be coupled with mechanical models. The difficulties of couplings, the computation times of all models are important points towards the future use of the tools in an industrial context.

Due to the strong links with the PN CEOS.Fr, the project takes benefits of the experimental results obtained on real size structures (several meters) made under industrial conditions. The first numerical simulations carried out show that the models are able to adequately represent the behavior of the cracked structures tested (static and thermo-hygro-mechanical loadings). The computational times are also an important point for the use of reliability approaches.

Indeed, for a proper use of the tools developed by consulting firms, it is necessary to find a compromise between accuracy and computation time for economic reasons and complexity.

Since the beginning of the project, about twenty scientific papers were produced (conference or peer-reviewed articles). These articles focus on scientific developments of models but also on applications. In the common framework CEOS.fr / MEFISTO, a international extension of the project was initiated by the creation of a cycle of workshops CONCRACK (3 meetings to date). These events have helped to provide the experimental results and compare model results of over 20 international scientific teams.

Urban elements, such as buildings, networks, various infrastructures, are mainly built with concrete. In the general context of the sustainable development, it is requested from the concrete structures, reinforced or prestressed, to ensure the various structural functions over increasingly long periods by optimizing the quantities of materials necessary and for a reduced maintenance. Today, the needed performances expected from structures go well beyond the simple mechanical resistance of material. Among the awaited functions, much are related to the intrinsic quality of material, its adequacy with the environment to which it is subjected and designed lifespan of the structure. Nevertheless, it is an essential parameter which will condition the behaviour towards deformability, air or water tightness, durability and sometimes safety: the cracking of reinforced concrete. The control of this cracking is of a major stake because it will ensure the main functionalities of the structures throughout their life cycle. From this point of view, the prediction of the mechanical behaviour and cracking of reinforced, and/or prestressed, concrete elements are essential to describe the performance and to predict the durability of the structures concerned. This prediction is complex because it depends at the same time on the specific characteristics of material in its environment but also of the design and the mechanical performance of the structure itself. Standard codes approache these aspects by the use of methods resulting from very traditional experiments which often answer very badly the problems arising from the sustainable development in general and structures with specific functionalities in particular. It results from it that to maintain their functionalities (safety, sealing, esthetics...), many buildings and structures must be rehabilitated with large expenses because of this insufficient control of cracking. In this context the purpose of the project MEFISTO is, by being based on the most advanced scientific knowledge field, to develop and validate new ways for the prediction of reinforced concrete structures cracking under various stresses (Thermal, coupled Hydriques and Mechanical).

Project coordination

Bruno Capra (PME (petite et moyenne entreprise))

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.



Help of the ANR 913,466 euros
Beginning and duration of the scientific project: - 36 Months

Useful links

Explorez notre base de projets financés



ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter