Read Numerical modeling of ground penetrating radar (GPR) for investigation of jointing defects in brick masonry structures text version

NDTCE'09, Non-Destructive Testing in Civil Engineering Nantes, France, June 30th ­ July 3rd, 2009

Numerical modeling of ground-penetrating radar (GPR) for the investigation of jointing defects in brick masonry structures

Rani HAMROUCHE*1, Gilles KLYSZ1, Jean-Pierre BALAYSSAC1, Stéphane LAURENS1, Jamal RHAZI2, Gérard BALLIVY2, Ginette ARLIGUIE1 Université de Toulouse ; UPS, INSA ; LMDC (Laboratoire Matériaux et Durabilité des Constructions) ; 135, avenue de Rangueil ; F-31 077 Toulouse Cedex 04, France 2 GRAI (Groupe de Recherche en Auscultation et en Instrumentation) University of Sherbrooke Sherbrooke (Québec) Canada J1K 2R1 * : corresponding author e-mail : [email protected] [email protected]

1

Abstract

The objective of this work is to use Ground Penetrating Radar for the inspection of brick masonry structures and in particular seeking for deep jointing defects. As a first approach, a numerical modelling of a GPR antenna with a central frequency of 1.5 GHz is used to define the sensitivity of radar waves to detect jointing defects. A specific algorithm for the processing of the simulated signal was developed to locate the defects. A parametric study using this algorithm allowed the setting of limits in terms of size, orientation and depth of the sought defect. Early results are promising and show that jointing defects must able to be detected at the surface and in-depth.

Résumé

L'objectif de ce travail consiste à utiliser le radar de type GPR pour l'auscultation des structures en maçonnerie de briques et en particulier à rechercher des défauts de jointoiement en profondeur. Comme première approche à ce travail, un modèle numérique d'une antenne radar GPR couplée de 1.5 GHz de fréquence centrale est utilisé pour définir la sensibilité des ondes radar à la détection des défauts de jointoiements. Un algorithme spécifique de traitement des signaux simulés a été développé pour localiser les défauts. Une étude paramétrique mettant en oeuvre ces deux algorithmes a permis de définir des limites en termes de taille, d'orientation et de profondeur du défaut recherché. Les premiers résultats sont prometteurs et montrent que des défauts de jointoiement doivent pouvoir être détectés en surface et en profondeur.

Keywords

Non-Destructive Testing, masonry structures, GPR, dielectric properties, Signal processing.

1

Introduction

Nowadays in France, 4577 masonry bridges can be counted, followed by reinforced concrete bridges with 3974 (total bridges 24 856) [1]. In the masonry structures, the degradation of mortar jointing creates a weakening which can induce important mechanical damages to the structures. In fact, the mortar plays the role of binder and ensures continuity between the bricks. The deterioration of this continuity is therefore a risk in regard with the

NDTCE'09, Non-Destructive Testing in Civil Engineering Nantes, France, June 30th ­ July 3rd, 2009

integrity of the masonry structure. The causes of this degradation can be mainly mechanical, chemical or biological. Mechanical causes can induce cracking of the mortar whereas chemical causes can lead to a lixiviation or a swelling of the mortar. In the case of biological causes the consequences can be an acid attack of the mortar. Several Non Destructive Testing (NDT) techniques are currently used to assess the condition of structures but few of these techniques are applied to masonry structures. This is especially true with regard to the use of ground penetrating radar (GPR) technology for the detection of voids in these structures [2]. The objective of this study is to contribute to the development of the GPR for the detection of voids by using numerical simulations. The aim is to check the sensitivity of this technique to the jointing defects, and to validate a signal processing algorithm able to identify, localize and size these defects. In this study, we proceeded to the electromagnetic characterization of bricks and lime mortar, the goal is to integrate the characteristics of each component in our model, and get the brick mortar interfaces that match the reality of the masonry, and then try to observe their mutual influence in masonry, followed by simulations of the electromagnetic wave propagation in a masonry wall, with and without voids. The wall is considered either as an homogeneous medium or as an heterogeneous medium in the simulation. Algorithms to construct 2D images of the wall from the collected simulated signals are also proposed and tested in the two configurations; this study is the first stage of a long study on the improvement of algorithm for signal processing masonry.

2

Evaluation of dielectric properties of bricks and mortar

Three bricks of dimension (42x28x5cm) are used for the characterization of dielectric properties. The measurements are performed on the bricks in natural condition without preconditioning (volumetric moisture of bricks is 0.08%). Apart from the measurement face, all sides are covered with an aluminum foil to ensure a good reflection of radar waves at the boundaries of the bricks. The first position of the moving receiver is at 16.2cm from the fixed position of the transmitter, followed by 10 steps moving of 1 cm. We thus obtain 11 signals corresponding to the various positions of the receiver. For the lime mortar, 3 slabs (50x25x5 cm) were made. The measurement device is the same as for the bricks. Measurements are made on the slabs after fully carbonation (accelerated process) and stabilization of weight. The composition of the mortar is shown in Table 1. This composition is chosen because it's the most used in masonry structures in France [3]. Table 1. Composition of mortar

Component Lime Sand (0/5) Water Quantity 400 Kg/m3 1025 liters 275 liters

GPR measurements were carried out using a SIR-2000 GPR system, with two 1.5 GHz coupled antennas. Each antenna is made up of a transmitter and a receiver in order to obtain a variable reflection angle. The evaluation of electrical conductivity and dielectric permittivity is performed using the FDTD numerical model of a GPR antenna developed by Klysz [4]. In these simulations, the size of bricks and mortar slabs were respected, the purpose of the estimation is to fit the electromagnetic parameters until to obtain a good correlation between simulated and measured signals [5]. The results presented in fig.1 show a correlation between measured signal and simulated signal on the bricks. The example presented here corresponds to signals recorded at the position 11 of the receiver, so at a distance of 26.2 cm. The

NDTCE'09, Non-Destructive Testing in Civil Engineering Nantes, France, June 30th ­ July 3rd, 2009

determined electromagnetic parameters are a relative permittivity 'r = 3, and a conductivity = 0.02 S / m.

Figure 1. Simulation versus measurement In the case of mortar slabs, the electromagnetic parameters are a relative permittivity 'r = 8, and a conductivity = 0.065 S / m.

3

Simulation of the propagation of electromagnetic waves in masonry walls

We present here 2D simulations of the propagation of electromagnetic waves in masonry walls (327 cm length and 58 cm thickness). The size of the bricks is the same as the one previously mentioned (28 cm x 5 cm) and the thickness of lime mortar joint is 2 cm. In the first simulation, the wall is considered as homogeneous, i.e. bricks and mortar joints have the same values of 'r = 5.5 and = 0.042. These values were obtained by averaging the electromagnetic characteristics of both the brick and the lime mortar. For the second simulation, the electromagnetic properties of bricks and joints are taken equal to the values previously found ('r = 3 and = 0.02 S/m for brick and 'r = 8 and = 0.065 S/m for the joints in lime mortar). During the measurements, the position of the transmitter is fixed in the middle of a brick, and the receiver will take 28 positions separated with one centimeter, and with the first one at 13.9 cm from the transmitter (Fig. 2). Voids (2 cm thickness and 5 cm length) were introduced in the wall parallel to the measurement surface (fig. 2), the choice of this model of bricks configuration is to simplify this first phase of our study, it will be followed by a parametric study on the limits of our algorithm by varying the bricks configuration, the size of voids and the distance between them. The results of simulations for the homogeneous structure are shown in Fig.3. These illustrations are taken at different times during wave propagation: before the contact of the wave with void, after the contact with void and finally at the contact with the bottom of the wall. In the case of a homogeneous medium, any discontinuity in this medium will create a reflection that will be clearly visible in the GPR data. In the Fig.3, we can clearly see the wave reflection caused by the void and the bottom of the wall.

NDTCE'09, Non-Destructive Testing in Civil Engineering Nantes, France, June 30th ­ July 3rd, 2009

Figure 2. Joint defect parallel to the measurement surface

Figure 3. Joint defect parallel to the measurement surface (homogeneous structure)

Figure 4. Joint defect parallel to the measurement surface (heterogeneous structure)

NDTCE'09, Non-Destructive Testing in Civil Engineering Nantes, France, June 30th ­ July 3rd, 2009

Fig. 4 shows the same simulation as in Fig.3 but for a heterogeneous medium made up of two materials with different electromagnetic properties, (brick and mortar). These materials generate multiple reflections that will blur the reflection of the wave that is induced by the void. This makes difficult to interpret, and thus distinguish the reflections due to the interfaces to those due to the defects. A specific treatment of the received signals is required to extract the information relating the presence of the void from the signals.

4

Principle of the algorithm developed for the reconstruction of image and results

Each point of the structure is considered as a possible reflector. The time required for the wave to propagate directly from the transmitter - reflector - receiver is calculated by assuming that the propagation velocity is constant in the medium. A time windowing allows to extract the part of the signal related to the reflection. This process is repeated for each signal obtained at different positions of the receiver and the time amplitudes are stacked. An image is built on which the reflector points are linked to the higher amplitudes. The results of the application of our algorithm on the different simulations are presented in Fig.5.

Figure 5. Case of the homogeneous wall: simulated B-scan (left) result given by the algorithm (right), In Fig.5, the case of the homogeneous wall is quite simple due to the uniformity of its structure, this makes a clear signal. Indeed, a fast analysis of the simulated B-scan shows the direct wave (first reflection from the right), followed by the reflection of the void and by the reflection on the bottom of the wall. The void is clearly represented with dimensions close to its real dimensions, as well as its real position in the wall. The case of the heterogeneous wall is more complex due to the multitude of interfaces between brick and mortar, which makes the signal difficult to interpret. In the results presented in Fig.6, the void in the heterogeneous wall is almost as clearly represented as in the case of the homogeneous wall (Fig.5) with dimensions close to the real dimensions of the void. Concerning the accuracy of the localization in depth, the difference can be explained by an error in the estimation of the wave velocity in the medium. This error comes from the arrangement of bricks and mortar (difference between vertical and horizontal arrangements).

NDTCE'09, Non-Destructive Testing in Civil Engineering Nantes, France, June 30th ­ July 3rd, 2009

Figure 6. Case of the heterogeneous wall: simulated B-scan (left), result given by the algorithm (right)

5

Conclusions

The purpose of this study was to enhance the use of GPR for the detection of voids in masonry structures. The work presented in this paper aimed to proceed to a sensitivity study by means of numerical simulations. These simulations enabled to better understand the propagation of the radar waves in this type of structure and analyze the effect of the voids. A signal processing algorithm has been developed and its application on simulated signals gave encouraging results as it permitted a good localization of jointing defects. To validate this algorithm, a full-scale wall was built by integrating jointing defects with different sizes and orientations.

References

1. Ponts en maçonnerie (Campagne d'évaluation 2005), Image de la Qualité des Ouvrages d'Art, SETRA, juillet 2006. 2. D. M. McCann and M. C. Forde: Review of NDT methods in the assessment of concrete and masonry structures, NDT & E International Volume 34, Issue 2, March 2001, Pages 71-84. 3. Domède Nathalie. Méthode de requalification des ponts en maçonnerie. PhD thesis, Institut National des Sciences Appliquées de Toulouse, France 2006. 4. G. Klysz, X. Ferrieres, J.P. Balayssac and S. Laurens: Simulation of direct wave propagation by numerical FDTD for a GPR coupled antenna, NDT & E International Volume 39, Issue 4, June 2006, Pages 338-347. 5. G. Klysz, J.P. Balayssac and X. Ferrières: Evaluation of dielectric properties of concrete by a numerical FDTD model of a GPR coupled antenna--parametric study, NDT & E International Volume 41, Issue 8, December 2008, Pages 621-631.

Information

Numerical modeling of ground penetrating radar (GPR) for investigation of jointing defects in brick masonry structures

6 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

1262351


Notice: fwrite(): send of 209 bytes failed with errno=104 Connection reset by peer in /home/readbag.com/web/sphinxapi.php on line 531