Introduction

French national experts and practitioners of Non Destructive Evaluation have decided to contribute, under the auspices of AFGC (French Association of Civil Engineering) and COFREND (French Confederation for Non Destructive Testing), to write a book explaining what can be the interest of NDE for assessment of reinforced concrete structures, and how the NDE methods can be implemented and their results can be used.

This decision followed a statement:

• on one hand, built heritage is ageing, its structural condition may be deficient or not precisely known and safety requirements (or needs for economical rational decisions) are increasing,
• a lot of methods can be used (and are sold) telling that they give data useful for assessing the structural condition,
• for many complex reasons, building owners are often disappointed once they have asked for NDE (and paid for them), since they do not know how to handle the data they have paid for.

The aim of the Working Group is to write a State of the Art, and guidelines on the use of NDE (when, where, how, for what...) and on the use of NDE results for answering the real building owners questions.

1. Working process - objectives

The Working Group has been created in September 2000. This paper will describe the methodology we have followed, the questions we have tried to answer and it will present the content of the book to be issued.

The guidelines are designed such as to serve as a practical help for the various professionals involved in the structural assessment process:

• buildings owners and managers, whose needs remain generally very down to earth and who are not always civil engineering experts and never NDE specialists,
• practitioners of Non Destructive Evaluation, whose results of measurements programmes cannot always be used as such by engineers who have to put a diagnosis or reassess the structure,
• experts of diagnosis/assessment who must handle data from various sources (structural history, owners'requirements, NDE results, softwares...) and try to draw a final and synthetic decision.

It is the reason why the Working Group was built with an equilibrated participation of professionals from these different fields: building owners, companies using NDE methods, experts of structural assessment, researchers (from universities or others institutes)... All meetings (about 15 during these 3 years) were plenary meetings, where all topics have been addressed in common. Regular exchanges have also been facilitated by creating a website (open to Working Group members, but also to external people interested in our work, after a simple agreement). This website contained all practical information (meeting minutes f.i.), but also the working versions of all sections of the book. This made possible the effective work in common, each section of the written book being under the responsibility of a co-ordinator, but with the practical help of several other participants.

From a scientific point a view, the first key question was to decide what kind of problem we intended to address. It was decided to focus on questions related to structural assessment of REAL structures, i.e. structures which:

• are in a real environment (atmospheric conditions, loads),
• are full-scale,
• are made of a real material (opposed to "lab" material, like microconcrete, unreinforced concrete, finite size and regular shape samples, moisture controlled...),
• have often an (at least partially) unknown history...

The building owner wants his structure to fulfil given requirements in this context and asks if NDE can help him in assessing the structural condition. Thus, NDE must work with the same constraints and must give data which are of practical interest for the given problem.

Since we wanted to focus on methodology in practical cases better than on theoretical or laboratory results, it was decided, as far as possible, to illustrate the speech by many examples of on-site real problems.

It was also decided, at this preliminary stage, to focus on reinforced concrete, avoiding any scattering on other building materials (prestressed concrete, masonry). Exceptions were considered only when some methods have had few field applications on RC concrete and when we thought that the same method might apply in concrete.

We have decided to tackle important questions like:

• the lack of information on the structure (history, plans, conditions of use...) and its effects on decisions,
• the varying environmental conditions and others factors of bias and their effects on measurements,
• the sensitivity and reproducibility of measurements,
• the adequacy between: (a) the performed measurement and the property/object that it is intended to detect/characterise, (b) the link between this property/object and the function the structure has to fulfil.

It was chosen to divide the book into three main sections, each of them quoting many examples, taken as far as possible from real case studies.

Section A replaces the NDE within the general frame of maintenance/assessment, lists the practical questions NDE has currently to answer and shows how these questions can be, in practice, answered more or less satisfactorily.

Section B, after a general review of basic physical principles of the methods, is devoted to present 9 families of techniques, with several variants. For each family, real case studies are detailed, such as to discuss efficiency, field of use, precautions, limits...

Section C tries to draw a general synthesis. A defect/method matrix is proposed and the helpfulness of approaches which combine several methods is illustrated on several cases. Finally, it is shown how NDE results can be used for numerical assessment or coupled in a broader strategy, including long term structural monitoring.

This work is based on the expertise of the members of the Working Group as well as that existing in the international literature. This large amount of references is not quoted here but can be found in the book.

2. Non Destructive Testing within the frame of structural maintenance/assessment

Section A starts with two introductory chapters that define the words which are often used in the field of structural IMR (inspection - maintenance - repair) approaches: monitoring, measurement, diagnosis, evaluation, assessment... The different stages leading to a final forecast are described and the contribution of NDE is pointed at.

Fig 1: How different stages are articulated.

NDE methods appear as complementary of other approaches. Once it has been decided to use them, comes the question of the relevancy of the technique: what is the relation between the measured parameters and the diagnosis/forecast? These questions are often asked in a general way, like:

• "how long can I use this bridge without restraining the traffic ?",
• "what part of this building must be repaired in priority ?",
• "what is the level of safety if I do nothing ?"

and if we try to find some answers, these questions must be analysed.

This section of the report tries to list in different chapters the various objects/properties for which NDE can be useful (and what are used today in practice) :

• in terms of material properties : strength, cracking (of various shape and magnitude, due to various mechanisms...), homogeneity,
• in terms of geometrical assessment : limits of layers, depths, voids, heterogeneities, inclusions, interfaces, reinforcements,
• in terms of tightness (for bridges, pipes, of liquid tanks) : nature and magnitude of leakage, way followed by the fluids...
• in terms of lifetime estimate, since the true question can often be reduced to expected lifetime. For this last case, models must be used to assess the future evolution of properties and NDE methods would aim to evaluate the parameters requested by the model (for instance cover or diffusion coefficients...).

3. Physics and methods

Section B is centred on the NDE methods that are currently in use in the field of civil engineering.

It starts with a summary of the basic physical phenomena that intervene in methods using wave propagation and based on very low frequency phenomena. The objective is to give the reader an insight in the physics of the method to help him make appropriate technical choices when confronted to a problem. For instance, if a void is to be detected, his attention is drawn towards void size and void depth but also to wavelength in case of wave propagation. As far as possible the physical phenomena that are common to several methods are described only once in a synthetic manner, as in figure 2, that shows in a unique graph wavelengths for concrete and steel irrespective of the wave type. It is also aimed to help the reader to put aside methods that obviously will not perform well in the context of civil engineering applications. This chapter is the only one in Section B that does not rely on practical cases.

Before going into chapters dedicated to each method, all the requirements for the proper transfer of laboratory measurement techniques to in situ validated and efficient ones are described. The many changes from lab to in situ measurements survey are emphasised: limited interventions time, access difficulty, post interpretation... This chapter is particularly important in the case of NDE for civil engineering because almost no method is normalised, and research activity in this field is dynamic in designing new methods. At last, attention is drawn towards training and qualification level of the manpower.

Fig 2: wavelength classically used in NDE in civil engineering.

The NDE methods are then gathered in six families and reviewed, respectively seismic, electromagnetic, infrared, nuclear, electric and optic methods. The chapter's structure for each family is similar. The application fields are described together with the measured observable with regards to the parameters that are deduced. Then the most common encountered application difficulties are listed including calibration, bias... For each method practical examples are given to illustrate those points. Those NDE results are furthermore always presented in the more general context of the diagnosis of the structure itself.

A detailed case studies for each method was not feasible (f.i. surface waves techniques are only briefly presented). Anyhow for all the methods that have known in situ applications, a synthetic form is filled. Each form has the same structure and ends with a set of three marks, awarded after the group consensus, about the adequacy, availability and efficiency of the technique.

4. From measurements to results

The last section synthesises the knowledge about the possibility (and difficulties) to deduce, from (good) data obtained after a NDE survey, a given structural assessment. This section contains the following chapters:

4.1. From measurement to assessment
Questions of sampling and statistical representativity, repeatability and measurement noise are tackled. The influence of various influence parameters (such as temperature or water content) is discussed.

4.2. Abilities of the various methods for diagnosis
This chapter contains synthetic tables where pathologies and methods are crossed: what methods are able to detect / localise / quantify, what objects / properties can be found by which method (objects and properties are the items presented in Section A in relation with the diagnosis to be put on the structure).

The level of relevance of the method is discussed in relation with the expressed needs. This is illustrated on some typical objects / properties. Recommendations are given about the more efficient methods for given families of object / properties.

4.3. Complementarity of NDE methods
On three different examples (detection and localisation of rebars, prestressed ducts and development of corrosion), it is explained for what and how several NDE methods can be coupled. It is shown how a second method can confirm results given by a first one, or give the result with a better accuracy, or give a result the first method is unable to obtain.

4.4. Contribution of laboratory tests and prototypes
Laboratory specimens and prototypes (which can be on site or laboratory prototypes) can bring useful information. Their contribution is detailed in specific cases.

Because of their history, their materials properties, the boundary conditions or loading are well-controlled, laboratory specimens and prototypes can be of a great help when calibrating methods, or when their sensitivity (or field of application) has to be carefully defined. Many examples are discussed, such as the careful study of the practical ability (and sensitivity) for radar reflectometry to estimate the depth of prestressed ducts, or for impact echo to detect voids.

4.5. Contribution of semi-destructive methods
Since, on one hand, the boundary between non destructive methods and semi-destructive methods is varying, and, on the second hand, semi-destructive methods are often used as a quick mean of assessment, it is discussed how they can best contribute to the structural assessment, in addition to NDE methods. Their help for calibrating some parameters, for assessing directly some mechanical properties (when NDE methods are usually more focussed on physical properties), or for estimating spatial variations in the material is discussed.

4.6. Coupling NDE with monitoring and surveying
First, the aims of monitoring are given, thus it is shown what are its links with NDE methods, either before NDE use (kind and location of NDE measurements are defined after a monitoring stage) or after it (kind and location of monitoring are defined after the NDE methods have given a first estimate of the structural functioning).

The specific contribution of monitoring with regard to structural evaluation(speaking about how the structure "works") is discussed on examples.

4.7. Structural and functional assessment
In this chapter, the scale is changing. Its aim is to reinforce the link with the questions asked by the building owner. These questions are often formulated in terms of priority, level of safety, costs... NDE methods contribute to the assessment which can be either absolute or relative. Since, for many reasons, an absolute assessment is very difficult to draw (many unknowns remain, even after a wide investigation), a relative assessment is often more efficient (and suffices to take decisions):

• NDE methods can contribute to a "zoning" (i.e. spatial relative assessment), telling what parts are the most deficient,
• NDE methods can contribute to a time relative assessment, by comparing the data obtained at different periods on the same structure. This must be coupled to models of time evolution of the properties, to conclude about the relative urgency of works,
• NDE methods can contribute to a relative assessment between "twin structures", in a given asset (f.i. a series of similar bridges).

Depending on the kind of the required assessment, the weaknesses of some methods (f.i. difficulties for obtaining absolute values) can be cancelled. However, the spurious effects of some influence parameters (like temperature or humidity) remain and are a main obstacle. These points are illustrated on examples and the contribution of NDE methods to management strategy is shown.

4.8. Structural re-assessment
Estimating the remaining lifetime or the reliability of an existing structure is often a challenge, because of changes in regulations or in loading applied to the structure, or due to uncertainties on loading history. This estimation usually relies on modelling that requires data describing the present state as well as parameters for time-evolution models. NDE methods contribute by giving information that can help this re-assessment. It is shown, on specific examples, how NDE methods should be used to gather the more useful data, and how the structure can be re-assessed afterwards.

Conclusions

After three years work the AFGC-COFREND Working Group has produced a book that gives the present state of the art of non-destructive evaluation of concrete structures in France. An originality of this compendium is to clearly privilege a presentation anchored in the application field. It is expected that buildings owners and managers as well as NDE practitioners and diagnosis experts will find useful information as NDE has been positioned according to their different and respective points of view.

This book is thus a step towards a better integration of NDE in the life of structure. It fits the general trend that, due to economic pressure, asks for life extension of present and future structures under high security level for the users.

Acknowledgements

The authors have to thank all contributors of the Working Group who, by their regular participation have contributed to the success of this work, namely C. Abdunur, E. Antczak, J.P. Balayssac, S. Bonnet, F. Buyle-Bodin, J.M. Caussignac, B. Chau, J.L. Chazelas, B. Collin, C. Constantopoulos, P. Cote, D. Defer, X. Derobert, B. Fargeot, D. François, J.L. Garciaz, V. Garnier, B. Godart, C. Hugot, J.F. Lataste, D. Leroy, D. Marlot, C. Oglaza, G. Olivier, O. Paris, R. Poquet, V. Robert, P. Roenelle, C. Sirieix, F. Taillade and some others whose list would be too long but who have helped us during this three years project.

References

1. It has been chosen to not quote for some of the many references that have been used by the Working Group. All information is available on our website at the following address: http://listes.lcpc.fr/listes/afgc-cofrend