Non destructive evaluation of the structures for damage prediction and optimisation of the follow-up – EvaDéOS

The project is divided into three operational tasks. The first (Task T2) aims to define and formalize the needs of the structure managers and to translate them into reliable criteria. It must also identify relevant real structures where, in connection with the activities of the task T3, it will extract indicators which will be used as inputs for the predictive models of task T4. T3 task includes the necessary developments, mainly in the field of non-destructive evaluation, but also destructive or semi-destructive. It is first necessary to correct the degradation of predictive models, and then to evaluate the pathological indicators to compare with those calculated by the models of degradation. The main difficulty of this task lies in the development of a methodology to take into account the gradients of structure indicators by non-destructive techniques. Then the task T4 focuses on the development of probabilistic models to be used in a reliability context to facilitate the decision making. There are two types of models, first of all those for predicting the corrosion initiation and those dedicated to the prediction of the volume of corrosive products. Finally, task T5 task aims at developing a methodology for decision support in a reliability approach, on the basis of criteria defined by managers and in relation with the results of predictive models.

One of the first important results of the project was to gather around the same problem different experts of structure monitoring. It was also possible to propose a unique inquiry to the structure managers involved in the project in order to identify their practices and their needs. This questionnaire has been designed to be translated in terms of decision criteria. Another interesting and promising result consisted in proposing a specific methodology to optimize the measurement campaigns by means of non-destructive testing methods during the auscultation of the structures, in terms of localization and number of measurements. The project will also quickly provide new results in terms of sensitivity of non-destructive techniques versus indicators of reinforced concrete not studied previously (carbonation and gradients of moisture). On degradation models a synthesis of existing models has been completed and a sensitivity analysis of these models to the variability of the input data has provided very promising results. A benchmark of the most relevant models will be started very soon bank to test and compare their effectiveness. Finally in the same time, methods for time-variant reliability will be applied to the selected models. Finally, the methodology which will be used for optimizing the monitoring of the structures has been defined, regarding the specificities of the different structure managers.

The project will develop a framework for preventive monitoring of reinforced concrete structures. It will propose recommendations for the implementation of non-destructive methods for the prevention of corrosion of steel reinforcement. It will also help to identify the most reliable degradation models distinguishing engineering models and more sophisticated models and by comparing them versus their effectiveness, but also their cost in terms of number and quality of inputs. Finally it will also illustrate the potential of reliability methods such as time-variant. Finally, it will show how it is possible to update the results of predictive models through results provided by non-destructive testing methods.

1. C. GOMEZ-CARDENAS, ZM. SBARTAÏ, JP. BALAYSSAC, V. GARNIER, D.BREYSSE, Optimal spatial sampling for Non-destructive testing measurements, CFM, Bordeaux, 26-30/08/2013

This communication shows how it is possible to optimize the measurement campaigns during the evaluation of a structure

2. De Larrard T., Duprat F., Bastidas-Arteaga E., Schoefs F., Effects of Climate Variations and Global Warming on the Durability of RC Structures Subjected to Carbonation, 11th International Conference on Structural Safety & Reliability, ICOSSAR 2013, 16-20/06/2013, New York, USA

This communication shows how the effects of climatic variations of the model predictions can be taken into account .

3. Rakotovao N., Bastidas E., Schoefs F., De Larrard T., Duprat F., Sensibilité des modèles simplifies de chloration du béton vis-à-vis de leur capacité à propager des incertitudes, CFM, Bordeaux, 26-30/08/2013

This commication describes the sensitivity of models versus the variability of the inputs.

4. R. Décatoire, T. De Larrard, T. Yalamas, F. Schoefs, S-M. Elachachi (2014) Evaluation stochastique de la profondeur de carbonatation et actualisation des prédictions, 8ème Journées Fiabilité des Matériaux et des Structures (JFMS 2014), Aix-En-Provence, France

This paper describes the procedure allowing the actualisation of the model predictions (Task 5),

5. Détermination du profil de teneur en eau dans le béton par inversion des mesures de permittivité au moyen de sondes capacitives, M. Fares, Y. Fargier, G. Villain, X. Dérobert, O. Coffec, S. Palma-Lopes, Congrès Diagnobéton, 19-20 mars 2014, Toulouse

This communication describes an original determination of moisture gardients in concrete (Task 3)

The purpose of this industrial research project is to propose a global methodology for following the status of reinforced concrete structures, which includes the following phases: condition assessment, diagnosis and prognosis of degradations, and decision-making. The linking between the phases, which is also an objective of the project, is expected to be achieved by bringing together the partners specializing in each phase and harmonizing their actions with respect to the goals and needs of the owners/managers. The approach as a whole, comprising evaluation, prediction, and optimization of inspection and maintenance strategy, is based upon probabilistic methods associated with Non Destructive Testing. The partnership gathers together most of those having an interest in the follow-up process, from the owners/managers or their representatives to the specialists of NDT and experts in decision-making. The target structures comprise road bridges (although the methodology can be extended to other types of transport), electricity supply infrastructures and residential buildings. In order to set reasonable bounds on the project, only the corrosion of reinforcement will be considered, which remains a major concern for the deterioration of reinforced concrete structures. Unlike other projects that have focused on this degradation, EvaDéOS is turned towards a preventive approach, intervening when the degradation is not yet revealed but the necessary conditions are present for its more or less short-term onset. This preventive approach is an original asset of the project. Assessing the condition of structures where damage is not yet visible implies using and possibly improving NDT methods. An attractive challenge for strengthening their potential is to prove the value of their contribution in the framework of the project, where the probabilistic prediction approach requires the spatial and time variability of numerous data to be provided. The selection and development, if appropriate, of degradation models suited to a probabilistic context is another main objective of the project. Firstly, attention will be paid to models predicting the ingress of aggressive agents leading to the depassivation of reinforcement and thus to the initiation of corrosion. Secondly, models predicting the propagation of corrosion and the expansion of corrosion products will be considered. Sophisticated models, based upon the Finite Element Method, but also simpler end-user models employed in daily engineering practice will be under consideration and their combined use will also be looked into. The aim is that the outputs of the predicting models will be incorporated into degradation indices, previously defined with respect to the needs and practices of owners/managers, and thereafter utilized in the decision-making process. A methodology will be proposed on the basis of decision criteria chosen upstream by the owners/managers, so as to optimize the dates of inspection and maintenance operations, taking into consideration the probabilistic prevision of degradations, the costs involved and the updating of the observation data.