Abstract

The paper presents an ultrasonic method for characterizing the degradation stage of the cover concrete of structures using high-frequency ultrasonic waves (0.5 to 1 MHz).

A first sensitivity study was conducted in our laboratory. Mortar and concrete samples were submitted to chemical degradation by immersion into an ammonium nitrate solution for periods of 15, 30 and 45 days.

The stage of degradation of material has been characterized by ultrasonic velocity and attenuation, and also by porosity and mechanical measurements. The results of the different measurements are well related. The porosity of the degraded layer seems to be approximately constant. The broadband ultrasonic attenuation is found to be the most sensitive to evaluate degradation stage, especially at high frequency.

The present goal of the research is to develop an on site measuring apparatus. A special device is developed to characterise concrete cover by using dispersion and attenuation characteristics of surface waves. This device is at present validated by measurements on laboratory samples, composed of two-layer degraded/sound concrete or mortar. The results enable us to assess the sensitivity of the device to evaluate the degraded layer depth and the rate of degradation. Further measurements will be performed in the conditions i.e. on site bridges and walls.

Introduction

This non-destructive method for characterizing the degradation stage of the cover concrete of RC structures has been carried out in three directions [Ammouche, 2003] : (i) high frequency transducers to determinate the optionel wedge material/transducer\coupling combination for generating surface waves, (ii) tests in-situ to validate the choice of the right combination, (iii) study on two-layer samples by velocity and attenuation measurements.

High frequency ultrasonic surafec wave method

Fig 1: Experimental setup for Rayleigh wave inspection.

Mortar and concrete samples are testing using Rayleigh waves. This samples have been submitted to chemical degradation at different periods.

To generate the surface waves, compression and shear wave transducers are associated to a wedge [Viktorov, 1967].

A higher frequency range (0.5-1 MHz) seems to be highly sensitive to the detection of the cover degradation [Goueygou, 2002].

Two transducers with 0.5-1 Mhz central frequency are combinated with three materials for the wedge : teflon, hard PVC and plexyglass.

These materials are chosen because their sound velocity is lower than the wave velocity in mortar or concrete.

The wedge is inclined at the critical angle qc :

c₁ is the compression or shear wave velocity in the wedge
c₂ is the shear wave velocity in mortar or concrete

Fig 2: Positions of the receiver-transducer.

The receiver transducer is placed at 5 distances from the emitter-transducer, the first position is at 3 centimeters. The distance between each position of the receiver-transducer is 1 centimeter.

The samples have been chemically attacked by ammonium nitrate (NH₄NO3) for different periods : 15, 30 & 45 days.The samples will be characterised by attenuation and velocity measurements.

Testing of Surface wave transducers

These series of test will allow us to compare the teflon wedge/coupling and the material wedge\coupling gelD combination. The criters of choice are the nature of the wedge material and coupling.

The frequency measurements show that the coupling gel D is less attenuating than the SWC.The figure 3 shows that the coupling gelD gives the best results on velocity.

Fig 3: Comparaison between coupling SWC and gelD with the teflon wedge at 0.5Mhz.

A difference of velocity between sound and degraded mortar at 30 days is observed. This velocity difference is more significant with the coupling gel D.

At 0.5Mhz & 1Mhz, the different mortar sample are caracterised by decreasing velocities (Fig.4).

Fig 4: Tests on sound and degraded mortar at 0.5 & 1Mhz with the teflon wedge/coupling gelD.

The coupling gelD will be used for the following tests.

Different materials are compared for the analysis of the wedge combinations.

Shear waves are using with the wedge to generate surface waves for the tests on sound mortar (Fig.5).

Fig 5: Tests on sound mortar with three wedge combinations at frequency 0.5 & 1Mhz,coupling gelD.

Compression waves are using with the teflon wedge to generate surface waves on sound concrete and shear waves with the hard PVC wedge (Fig.6).

Fig 6: Tests on sound concrete with three wedge configurations at frequency 0.5 & 1Mhz.

In the sound mortar, the best results are given by the wedge teflon/gel D combination for velocity and attenuation.

The combination wedge teflon/gel D gives the best results for velocity and frequency in sound concrete. This combination has the advantage to get the lowest absorption.Therefore a significant difference is observed for the frequencies at 0.5Mhz and 1Mhz (Table 1).

A low difference of velocity is observed for the transducers at 0.5Mhz and 1Mhz.Finally, the best results are obtained with the teflon wedge/coupling gel D combination.

Study on two-layers samples

The teflon wedge/ longitudinal wave transducer combination is now tested on laboratory samples, composed of two-layer degraded/sound concrete and mortar.

Two series of samples have been made. One with a chemically degraded slabs and the second with a different W/C ratio.

Chemical degradation :
Samples of mortar and concrete have been made, size 30*30*15 cm. Then each sample has been cut in the length direction in order to take a slab which will be degraded. After degradation, this slab is bonded with cement on the original sample (Fig.16). Samples with different thicknesses were made.

Fig 16: Chemical degradation samples set-up.

Variable W/C ratio:
Samples of mortar and concrete size 30*30*15 cm are made with a W/C ratio equal to 0.5.

On the upper and lower face a new layer of mortar or concrete with a different W/C = 0.8 is poured on origin sample. Only the thickness of this layer will change (Fig.17).

Fig 17: Two-layer with different W/C ratio set-up.

With one sample, two series of tests can be conducted

Finally, we have :

The 4mm mortar and 5mm concrete layers have been chemically degraded for 21 days, the other layers for 31days.

At present, only the following samples have been tested with transducers of frequency 0.5Mhz and 1Mhz (Tab.3).

The receiver transducer is placed between five and eight positions from the emittertransducer,depending on the sample type.

Fig 18: Surface wave velocity for two-layer chemically degraded mortar.

The velocity decreases when the thickness increases for the both frequencies.

The velocities are rather close.

The tests on the two-layer different W/C ratio mortar and two-layer chemically degraded concrete have not been analysed yet.

The results will be done later.

In-situ device

This teflon wedge/coupling gelD and honey combination has been used in real conditions on a bow-string bridge built in 1915-1916 in Toulouse (France) [Klysz,2003].

This bridge has four spans of 45 meters length each. One of them (n°1) has been rebuilt on 1952 (Fig.7).

The arcs of the spans are supported by vertical and inclined ties for spans 2,3 and 4, by vertical ties for span 1.

The ties are strongly reinforced and on certain ones the coating is very thin, even absent.Visible degradations result by concrete delamination with apparent corroded steels.

Fig 7: View of the bridge with the different spans.

Fig 8: Different views of the degraded intrados and extrados span n°1.

The intrados of the deck is also degraded in several parts:

• corrosion of steel (foto a,b & c)
• sealing defects (foto b & c)
• vertical and horizontal cracks (foto a & b)
• intrados with cracks of different depth (foto b)
• water infiltration (foto b & c)

First investigations on the intrados allow the choice of areas more or less degraded in order to be able to test the method and compare them (Fig.9).

Two ties of the span 1 and one tie of the span 4 were chosen :

• tie 8 face a, tie 10I face b & c on span 4.
• tie 9 face a,b & c on span 1.

The receiver transducer is placed at five distances from the emitter-transducer, the first position is at five centimeters.It's the same operations as we did before in the high frequency method.The distance between each position of the receiver-transducer is 1 centimeter. Five positions are tested.

Fig 9: Views of the span n°1 & 4.

Fig 10: Measurements line on the ties (green).

After having carried out this profile of measurements, the emitter is moved 10 centimeters along the profil and a new profile of measurements is acquired with five displacements of the receiver (Fig.10).

The coupling gel D was used on the ties with a transparent sheet for photocopier.

On the intrados of the deck of the span n°1 (wet and damaged), we chose to work on an overall length of 1.80 meters corresponding to 18 profiles with each one 5 different receiver position (Fig.11). This length covers the area where the reinforcement is apparent with delaminations and cracks. On the intrados of the deck of the span n°4 (more healthy), ten profiles were recorded (Fig.11).

Fig 11: Measurements line of the intrados.

Fig 12: (a) View of emitter and receiver transducers. (b) recording equipments.

When the maximum coupling is obtained, the operator records the signal on the scope. Then the signal is transmitted to the computer via an acquisition software.

In the ties, it's not easy to differenciate the velocities.

Fig 13: Velocy profiles of the differents ties.

The acoustic parameters don't evolve as wait.

Velocities in the ties of the span n°4 should be faster than the ones of the span n°1 (damaged and older).

These results can be explained by the fact that the investigation depth is near to the thickness of coating and by the humidity masking the degradation.

Fig 14: Velocities of the different spans.

On the intrados of the spans, the coupling honey was used for adherence reasons with a transparent sheet for photocopier.

In the cracked zone, the difference between the spans is more evident : surface wave velocity average of 2684m/s in the span n°4 compared to 2196-2236m/s in the span n°1.

The healthy (more recent) and damaged (older) spans are then differenciable.

As before, the central frequency measurement does not enable us to differenciate the damaged and healthy parts with frequency analysis (Fig 15 a & b).

a)

b)

Fig 15: (a) Frequencies of the different ties. (b) Frequencies of the different spans.

The results are not corresponding to our waitings. High-frequency ultrasonic attenuation is highly sensitive to damage and degradation. But in these tests, it was very difficult to observe it.The surface wave seems to be trapped into the coating.

Conclusions

The wedge teflon/coupling gelD combination seems to be the best choice, giving the best results in all the configuration.

This combination allows the differenciation between sound and degraded material.

The surface wave velocity decrease with degradation.

Therefore, this combination has the advantage to get the lowest absorption.

At frequency 0.5Mhz and 1Mhz, velocities are rather close for degraded and two-layer samples.

The tests on the two-layer samples show that the velocities decrease when the thickness of the first layer increase.

At 0.5Mhz and 1Mhz, the velocities are rather close.

In situ, the difference between the older (assumed damaged) span and the recent (assumed healthy) is visible.

More of the in-situ results were different of our waitings : high frequency seems to be sensitive to damage and degradation. The surface waves seem to be into the coating.

References

1. [Viktorov, 1967] I.A Viktorov, Rayleigh and Lamb waves, Plenum Press, New York, 1967.
2. [Goueygou, 2002] M.Goueygou and al., Assesment of broadband ultrasonic attenuation measurements in inhomogeneous media, to appear in Ultrasonics International, 2002.
3. [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
4. [Young S. Cho, 1999] Young S. Cho & Feng-Bao Lin, Spectral analysis of surface wave response of multi-layer thin cement mortar slab structures with finite thickness, Structural faults+repair 99.
5. [Goueygou, 2002] M.Goueygou and al, Measurement of Ultrasonic attenuation and Rayleigh wave dispersion for testing concrete with subsurface damage, Munich, October 2002.
6. [Klysz, 2003] Klysz G.,Balayssac JP, Dérobert X., Aubagnac C., Evaluation of cover concrete by coupling some non-destructive techniques Contribution of in-situ measurements, NDT-CE International Symposium Proceedings Berlin, September 16-19 2003
7. [Ammouche, 2003] Ammouche A., Garciaz J.L., Buyle-Bodin F. (2003), Contribution of coupling non-destructive methods for the diagnosis of concrete structure, NDT-CE International Symposium Proceedings, Berlin, September 16-19 2003