Abstract

This paper describes the general approach followed in France as part of the research project "Damage Evaluation of Concrete Cover & Assistance to Diagnosis and repair of Civil Engineering Structures" supported by the RGCU (Réseau Génie Civil et Urbain). The project is grouping together nine laboratories in partnership. The work program includes auscultations on both in-service structures and laboratory specimens using high frequency ultrasonic method, radar, and resistivity measurement. The in-field investigations consist in the auscultation of different bridges situated in France. For laboratory tests, the investigations are carried out on three types of concrete specimens (60x60x12cm), corresponding to three strength classes C25,C40 and C60. The concrete specimens are submitted to different solicitations: mechanical loading, physical and chemical ageing (natural and controlled drying, accelerated carbonation and chloride penetration). This document focuses essentially on the general approach (experimental program) and presents some first results.

1. Introduction

The use of NDT methods in Civil Engineering is well developed to evaluate the material characteristics. With NDT it is possible to localize degraded areas to identify degradation mechanisms. But the scale of study is limited to the scale of constitutive materials. For the main mechanism of degradation of reinforced concrete structures that is the corrosion of steel reinforcement, it is possible to enhance the scale of analysis. NDT of concrete cover allows the positioning of the structure in its life cycle because concrete cover controls the penetration of aggressive agents into concrete, and because the knowledge of this ingress can be related with the development of the structural degradation.

Nine French laboratories have been grouped to conduct researches to develop NDT methods for analysing concrete cover, aiming particularly to improve the gain offered by the simultaneous operating of various methods as high frequency ultrasonic method, radar and resistivity measurements. More than the three authors, the paper is the report of a collective work, and must be particularly acknowledged the young researchers, at Bordeaux J.F. Lataste, at Lille A. Fnine and at Toulouse G. Klysz.

2. Research team

The research is conducted by several laboratories and companies located in North of France (LML, ECL, SOVEP), at Toulouse (LMDC, GETEC, LRPC, ONERA), at Bordeaux (CDGA), at Nantes (LCPC), and at Arles (LERM).

In the North of France, the Laboratory of Mechanics of Lille LML is associated with Ecole Centrale de Lille (Laboratory of Acoustics) and the SOVEP engineering company to develop ultrasonic method. The LML is a research unit associated with CNRS (Centre National de la Recherche Scientifique) UMR 8107 working on fluid mechanics, solid mechanics and civil engineering. The team associated to the present project is specialised in the mechanical analysis and the life cycle assessment of degraded RC structures.

The Electronics -Acoustics Group of IEMN (Institut d'électronique et de microélectronique du Nord) UMR CNRS 8520, is located in Ecole Centrale de Lille.

SOVEP is an engineering company specialised in urban structures assessment, repair and rehabilitation. It contributes to the project by logistic helping.

At Toulouse, Laboratory for building Materials and Constructions Durability LMDC is a Research Unit, attached to both Université Paul Sabatier, Toulouse III (UPS) and Institut National des Sciences Appliquées de Toulouse (INSA Toulouse). The researches are multi-disciplinary and rely on the techniques of the materials science and include development of new materials, characterisation and enhancement of the performances and durability, environmental and economical concerns. Moreover, a technology transfer department has been shaped to propose expertise and short-term research facilities to industry, notably to SMIs and SMEs.

ONERA (Office National d'Etudes et de Recherches Aérospatiales), the French aerospace research agency, is a public scientific and technical establishment with both industrial and commercial responsibilities. Toulouse agency employs 400 persons and about 100 students.

GETEC is a company specialised in evaluation and management of civil engineering structures, which employs 25 persons. It contributes to the project by logistic helping.

LCPC is a State research organization working for the national and the local authorities in connection with professionals involved in civil engineering, transport, urban engineering and environment. LCPC has 540 employees, from which 200 engineers and researchers, and about 50 to 60 Ph-D students divided into three sites, Paris, Nantes and Marne-La-Vallée near Paris.

At Bordeaux, C.D.G.A. (Center for the Development of Applied Geosciences) is a University laboratory. It studies problems in the fields of geology, geotechnics and civil engineering. It is involved in researches on strategies of assessment and maintenance of the built heritage. It develops methods, derived from subsurface geophysics, to improve the structural assessment in reinforced concrete buildings. Electrical resistivity measurements, which are usual techniques for corrosion diagnosis, have been implemented for damage assessment of structures. The same can be told about infrared thermography whose applications in environmental geotechnics have yet proved their interest.

LERM (Laboratoire d'Etudes et de Recherches sur les Matériaux) is a truly independent private-owned laboratory. LERM achieves studies and applied research in construction materials and environmental fields. Its activity involves physical, chemical and mineralogical analysis (aggregates, binders, concretes, etc.), as well as in-situ tests (non destructive testing methods). Diagnosis studies on ancient and moderns structures and formulation/characterisation of new types of concrete (high performance concrete, self compacting concrete, etc.) are among the issues of interest of the laboratory. LERM employs 43 persons and contributes to several research projects like electrical resistivity measurements with the CDGA and radar antennas development.

3. Description of NDT methods

3.1 Ultrasonic method
The use of low-frequency ultrasonic waves (50 kHz) is frequent to inspect concrete structures. This level of frequency is not adequate to detect the defects of the concrete cover because these defects are of a little size (1 to 100 mm). In a previous study [Ould-Naffa, 2002] it has been demonstrated that the use of a higher frequency range (0.5-1 MHz) allowed a sensitive detection of the cover concrete degradation. The measured acoustic parameters were pulse velocity and attenuation of compression (P), shear (S) and Rayleigh (R) waves. In the tests conducted in the laboratory, the degradation had been chemically accelerated. To correlate the results, the characterization was complemented by porosity and Young's modulus measurements. A good correlation between velocity decrease, attenuation increase and porosity was established.

To develop the method and fit it to in-field measurements, in a second time only the Rayleigh waves are selected as allowing the measurement on the surface of the concrete by one face. Rayleigh waves are generated using the wedge method. Several transducer/wedge combinations are tested [Matthews, 1996], in order to estimate the dispersion of R waves. Dispersion curves are observed for various degradation times and correlated with the degraded layer profile.

The increase of the depth of the degraded layer induces a global decrease of phase velocity and an increase of dispersion. The analysis at different frequencies could be used to evaluate the transition zone between degraded and sound material.

3.2 Radar method

Fig 1: Propagation of electromagnetic waves in reinforced concrete.

The radar principle is based on the propagation of electromagnetic impulses through a material. The impulse is currently centred on one frequency, which is currently higher than 1 GHz for concrete sounding. Ground coupled antennas are generally used because of higher sensibility and accuracy. A radar transmitter (T) stimulates the propagation of several electromagnetic waves. The first wave is the direct wave (S1 on Fig. 1a), propagating along the air-material interface in the subsurface with a velocity determined by the upper part of this subsurface. Next, wave reflections caused by changes in dielectric electromagnetic properties of the medium are recorded by the receiver (R) during time (for example on figure 1 S2 is the reflection by the reinforcement).

n : wave speed
c: wave speed in air
e'_r : dielectric constant

The recorded signal for each impulsion is generally referred to a scan or a trace, which relies amplitude variation versus time on a vertical axis (Fig. 1a). The time is those required by the signal to travel first to the reflector and to come back to the receiver. For each emitter position the recorded signals are represented in a shades of grey mode. After juxtaposition, a time-section along the profile of the antenna movement (Fig. 1b) is obtained. One of the most common applications is the exploration of the reflected signal by the reinforcement to provide information on its depth. Indeed if the wave speed can be calculated by an estimation of dielectric constant (see relation 1), depth can be defined.

Since a few years, some developments focus on the deterioration of the wave propagation [Laurens, 2002], [Garciaz, 2001]. Indeed, the dielectric properties are modified by some concrete properties such as water or chloride content. So the propagation factors (speed, attenuation) are affected by changes in concrete properties [Soutsos, 2001].

3.3 Seismic method
In view of quantifying the penetration depth of cracks a seismic method that uses surface waves is being tested [Hévin, 1998a]. The physical principal relies on the penetration depth of the surface waves that is of the order of the wavelength. Small wavelengths will be reflected by the crack whereas long wavelengths will pass below the crack. The crack can be seen as a filter to surface waves whose cut-off frequency is related to its depth. This technique has been validated on concrete test slabs that can be considered homogenous with regards to the wavelengths used when no crack are present (wavelengths typically larger than 5cm) [Lataste 2003]. A simple empirical formula between the crack depth and the cut-off frequency has been obtained in the case of thick slabs or more precisely when the surface waves are not dispersive (when their velocity is independent of the wavelength). It is complementary to the normalized method [BS 1986] that uses compression waves.

3.4 Electric method
From an electrical point of view, concrete is a composite material whose compounds can be describe as:

1. a solid phase purely resistive (aggregates),
2. a solid phase which participates to conduction through its porous structure (cement matrix) and being the source of ions found in the third phase,
3. the liquid phase, i.e. interstitial solution.

The various damages which undergo the concrete during its life generally lead to porosity variations: the cracking of concrete can be seen as preferential ways for fluids flow within material, and physical and chemical damages (as leaching or carbonation for example) are inferring porosity change [Claisse, 1999]. To measure electrical resistivity leads to assess concrete conditions.

A measurement device has been then developed to exploit this sensitivity. Pattern and dimensions, as well as its associated measurement process, are defined to be adapted to on site investigation on reinforced concrete structure. Measurement realised from the surface are totally non destructive, and allow to characterize surface material properties. The drawing of electrical resistivity and electrical anisotropy maps or profiles, leads to assess properties variations linked to damage presence.

4. Capacities and limits of the NDT methods

4.1 Ultrasonic method
The correlation between velocity, attenuation and porosity is at present validated for strong degradations as ammonium nitrate attack and particularly on mortar. In real conditions and for concrete representative of civil engineering structures, the method and particularly the signal analysis must be adapted and improved. The quality of the transmitted pulse is important, and in a first time the best coupling of transducer with solid edge must be designed. The sensibility of the method to moisture content, to cracking density, and to granularity of concrete must also be studied.

4.2 Radar method
Some previous studies show the sensibility of direct wave to moisture gradient within cover concrete. In the project presented in this paper, the sounding of cover concrete by mean of radar direct wave is carried out. In radar sounding, two kinds of velocity measurements are generally carried out when the antenna offset can be raised, Wide Angle Reflection Refraction (WARR) and Common Middle Point (CMP) measurements [Tillard, 1991]. WARR acquisition consists of increasing the distance between the antennas stepwise while one remains stationary.

Both antennas are simultaneously moved apart on either side of the middle point of the profile in case of CMP measurement. By exploiting these methods it is possible to evaluate speed of electromagnetic waves (Fig. 2). Nevertheless in the case of reinforcement very closed to the surface it can be difficult to separate the material wave and the rebar reflection.

Fig 2: Schematic WARR-measurement. The material wave can be identified as a wave with a linear move out starting from the origin of the x-t plot.

4.3 Seismic method
The precision of the determination of the depth of surface opening cracks is around 10% if the concrete can be considered as a non-dispersive media. If the concrete is dispersive no extensive study exists that gives a simple relation between the crack depth and the surface waves filter cut-off frequency. Anyhow the effect of the crack is clearly visible and comparative study can still be carried out.

In case of cracks that are partially closed, the surface waves method will give a global depth that will be larger that the depth of the first contact but smaller that the total length of the crack if it was completely open. It is nonetheless an enhancement with regards to the method that uses compression waves for the later give the depth of the first contact.

Furthermore it has been shown that the filtering of the surface waves by a crack is not perturbed by the presence of water filling [Hévin, 1998b].

4.4 Electric method
Electrical resistivity measurements on concrete structures (in this project) are used to assess electrical properties of concrete (linked to its porosity and sensitive to crack presence). Works are engaged on two ways: The first axis is dealing with the characterisation of surface properties of concrete. The sensitivity of electrical measurement to material porosity gives information on the integrity of concrete (and its variations) on different zones. In itself this data cannot be exploited, but can be coupled to complete results obtained with other methods.

The second work axis is on the cracks characterisation. Resistivity and anisotropy variations are exploited to assess cracks intensity. Indeed, variations are linked to depth, opening and humidity of the crack, which are information that can be coupled with acoustic method to be refined [Lataste, 2003].

Today the technique is still in development. The accuracy of measurement needs to be improved, particularly for investigation on very dry concrete [Lataste, 2001].

5. In-situ investigations

The different methods have been operated at a first time on the Empalot Bridge near Toulouse. The first investigations results are presented in other paper [Klysz, 2003]. At this stage of the work the aim of these investigations was to define the limits of each technique in regard to in-situ environmental constraints and to test the different method combinations. Even if improvements are necessary one can notice that all the techniques are able to provide reliable results on degraded areas, either for the detection of damages like cracks or surface degradations or for the detection of wet areas relied to the corrosion of the reinforcement. Moreover in-situ investigations have emphasised the first combinations. As some limits have been well highlighted, these first investigations will permit to orient the laboratory developments.

Some other structures have already been chosen near Lille for the final validation of the methods. The scope is to carry out a blind investigation on some damaged areas with all the methods. Then a destructive diagnosis carried out by the structure manager will allow the non-destructive evaluation assessment.

6. Laboratory tests

In parallel with in-situ investigations, laboratory tests aim to well-define the capacity, sensitivity and limits of each NDT method in the context of more "controlled" conditions (material history, composition and micro-structural characteristics, chemical and hygrometric conditions, etc.). Theses investigations will also focus on the coupling between several methods in order to confirm preliminary in-situ investigations as well as to orient upcoming ones.

6.1 Concrete types and specimens manufacture
Three types of concrete corresponding to different levels of strength and performance are used to manufacture laboratory specimens: C25, C40 and C60. C25 class corresponds to a class of concrete widely used for usual building applications. The mix C40 is enough representative of concrete class extensively used in civil engineering structures. The choice of C60 is related to the recent great development of HSC (High Strength Concrete) for civil engineering structures. Concrete compositions are given in Table 1.

The samples series C25 and C60 are manufactured by LERM while the series C40 is produced by LMDC.

Most samples are little slabs of size 60x60x12 cm which corresponds to minimal dimensions required by the Radar method. Some additional samples 10x10x40 cm for freeze-thaw experiments are also prepared. A total number of 63 specimens (35 samples C25, 17 samples C40 and 11 samples C60) are today manufactured. Part of these samples (39 samples) is reinforced by steel bars Æ12 mm with a spacing of 15 cm. The thickness of the concrete cover is fixed to 25 mm.

6.2 Characterisation on fresh and hardened concrete
Characterization on fresh concrete includes slump, air content, and fresh density measurements.

On hardened concrete, mechanical, physical and micro-structural analysis are planned to a better identification of the studied materials.

The concrete compressive strength (fc₂₈) is measured on three cylinders 11x22 cm made at each concrete fabrication. After 28 and 90 days of humid cure, cores are extracted from control slabs to perform the following measurements at three or five levels across specimen height of each concrete class:

• Porosity accessible to water (AFREM procedure).
• Oxygen permeability (AFREM procedure).
• Free water content.
• Pore size distribution by mercury intrusion.

6.3 Type of degradation
After three months of conservation in laboratory conditions, the samples are exposed to aggressive chemical ambiences, mechanical loading or controlled rate of saturation:

• exposure to chloride,
• accelerated carbonation,
• mechanical cracking by flexural loading,
• freeze-thaw cycles.

The detailed experimental program and sample repartition within the laboratories are presented in Table 2.

The objective is to produce different aged samples allowing the precise analysis of the response of each NDT method and its abilities in controlled conditions. Each site will have its own slabs to calibrate its own method. Then each team will apply its own method on the degraded slabs made by Toulouse and Lille.

Among the numerous questions to be treated:

• how micro-structural and/or chemical modifications resulting from concrete cover carbonation or chloride penetration can affect the response of NDT methods (ability and sensitivity) ?
• what's the influence of moisture conditions ?
• what's the influence of damage (cracks, microcroacks, their geometry and extent) ?
• how the presence of steel bars can affect the response ?
• what complementary information can be reached from coupling several NDT methods ?

7. Conclusions

Each of the four methods is at present fitted to the analysis of the concrete cover of reinforced concrete structures. Their accuracy is dependent on physical conditions of the cover, particularly moisture content, cracking density, roughness of surface, and presence of coating. The coupling of the four methods should increase the accuracy of the evaluation, since a first in-situ experiment has been promising. Laboratory tests should be confirming this original approach.

Acknowledgements

The French Ministry of Research, the French Ministry of Public works, the Region Nord-Pas de Calais, the FEDER of EU, are greatly acknowledged for their financial support. The doctoral studies of A. Fnine and G. Klysz are financed by RGCU (réseau Génie Civil et Urbain), which is greatly acknowledged.

References

1. [Andrade, 2000] Andrade C., Alonso C., Arteaga A., Tanner P., 2000, Methodology based on the electrical resistivity for calculation of reinforcement service life, 5th CANMET/ACI, Barcelona (S), pp. 899-915.
2. [BS 1986] British Standard, Recommendations for measurements of velocity of ultrasonic pulses in concrete, Testing concrete, BS 1881: Part 203, 1986.
3. [Claisse, 1999]Claisse PA., El-Sayad H., Shaaban IG., 1999, Permeability and pore volume of carbonated concrete, ACI Materials Journal, vol 96 n°3, pp 378-381.
4. [Garciaz, 2001]: J.L. Garciaz et J.L. Perrin, Application du radar géophysique pour l'évaluation des variations de permittivité et de résistivité des matériaux de surface - Conception d'une antenne spécifique et traitement des données, Journées scientifiques ENDGC2, Méthodes d'évaluation non destructives pour le génie civil, Bordeaux, 8 et 9 novembre 2001.
5. [Hévin 1998a] Hévin G., Abraham O., Pedersen H.A., Campillo M., Characterisation of surface cracks with Rayleigh waves: a numerical model, International Journal of Non-destructive Testing and Evaluation, 31(4), pp289-297, 1998.
6. [Hévin 1998b] Hévin G., Utilisation des ondes de surface pour l'auscultation des structures en génie civil: application à la caractérisation des fissures de surface, Thése de l'Université Joseph Fourier, Grenoble, 1998.
7. [Klysz, 2003]: Gilles Klysz, Jean-Paul Balayssac, Xavier Dérobert, Christophe Aubagnac, Evaluation of cover concrete by coupling some non-destructive techniques - Contribution of in-situ measurements, Int. Symposium NDT-CE, Berlin, Allemagne, 2003.
8. [Lataste, 2001] Lataste JF., Sirieix C., Frappa M., Breysse D., 2001, Electrical resistivity measurement applied to Non Destructive Testing in Civil Engineering: Influence of the humidity, 7^th EEGS, Birmingham (UK), September 2-6 2001.
9. [Lataste 2003] Lataste J.-F., Abraham O., Breysse D., Sirieix C., Sensitivity to crack parameters of two non destructive techniques: electrical resistivity measurements and acoustic methods, Int. Symposium NDT-CE, Berlin, Allemagne, 2003.
10. [Laurens, 2002]: S. Laurens, J.P. Balayssac, J. Rhazi, G. Arliguie, Influence of concrete moisture upon radar waveform, RILEM, Materials and Structures, Vol.35, N°248, 2002, pp.198-203
11. [Matthews, 1996] M.C. Matthews, V.S. Hope, C.R.I. Clayton, The use of surface waves in the determination of ground stiffness profiles, 1996, Proc. Instn Civ. Engng 119, pp. 84-95.
12. [Ould-Naffa, 2002] S. Ould Naffa, M. Goueygou, B. Piwakowski, F. Buyle-Bodin, Detection of chemical damage in concrete using ultrasound, 2002, Ultrasonics 40, pp. 247-251.
13. [Soutsos, 2001]: M.N. Soutsos, J.H. Bungey, S.G. Millard, M.R. Shaw,A. Petterson, Dielectric properties of concrete on radar testing, NDT&E International, Vol. 34, 2001, pp.419-425
14. [Tillard, 1991]: S. Tillard, Evaluation des performances de la technique radar en prospection géophysique, Doctorat de l'Université Louis Pasteur de Strasbourg, 1991