Abstract

In this contribution, some results obtained by applying sonic tomography on historic masonry structures are presented. Three different case histories are described. The investigated masonry portions date back to different ages, from the XIII to the XVII Century, and belong to different kind of structures, i.e. towers, churches, defensive town walls. They are all characterized by non-homogeneous masonry, made of mixed stone and/or bricks, with possible presence of multiple leaves, incoherent infill, voids, poor mortars, and lack or weak connection among the elements. The results show that sonic wave transmission velocity can give useful information for evaluating the consistency and the state of conservation of this kind of masonry. Moreover, when sonic tests and tomographies are applied before and after repair interventions able to modify the physical and mechanical properties of masonry, such as grout injections, they can give useful information on the effectiveness of the strengthening technique applied.

Introduction

Sonic tests are a completely non-destructive method. They consist in transmitting stress waves, within the frequency range of acoustic waves (20 Hz÷20 kHz), in construction materials such as concrete, masonry, wooden elements, etc.

The travel time of stress waves through homogeneous and isotropic solids is proportional to the dynamic elastic modulus, Poisson's ratio and density of the medium [Ref 1]. In the case of masonry, due to its intrinsic heterogeneity and anisotropy and to the large number of existing typologies, the velocity of sonic waves cannot be directly related to the properties of masonry (strength, stiffness). However, some works show that, for a given masonry typology, a fair good relationship can be found between the values of sonic waves velocity and the mechanical parameters of masonry, above all the modulus of elasticity [Ref 2]. Moreover, it has been already demonstrated that sonic tests can be usefully applied to perform local analyses and to get qualitative information on the consistency of masonry [Ref 1, Ref 3, Ref 4, Ref 5].

Besides direct (through the thickness of the wall) and indirect (on the same face of the masonry wall) measurement of sonic waves velocity between two points, it is possible to deepen the analysis by carrying out sonic tomographies. In this kind of test, the measures of sonic pulse velocity are combined along different ray-paths on a cross section of masonry, and are subsequently processed in order to define mean values of velocity on each portion of the wall section itself. It is also possible to carry out amplitude tomographies of masonry sections [Ref 6]. It is known, in fact, that not only the analysis of sonic velocity, but also of wave shape, amplitude and frequency, obtained by FFT elaboration, can be useful to qualify existing masonry walls [Ref 5].

By means of sonic tomographies it is possible to detect the internal morphology of structural elements. They allow identifying the possible layering of the sections (multiple-leaves characterized by different composition), the presence of voids, cavities and general anomalies in density [Ref 6]. If carried out in the same wall sections before and after a consolidation intervention (e.g. grout injections), they allow evaluating the difference in consistency before and after the repairs, i.e. they can be used to control the effectiveness of the intervention [Ref 1, Ref 2, Ref 7, Ref 8].

Moreover, sonic tomographies applied in different areas of a building, can be used to evaluate cross-sectional properties of the tested masonry elements, in order to properly model the structure [Ref 9]. This use of sonic tomographies is very promising, even if a limitation to the extensive application of the method on a single building derives from the fact that sonic tests are still highly time-consuming.

In real case studies, it is advisable to apply other non destructive or minor destructive tests to cross-check the qualitative results of sonic tests. Double flat jacks, to quantify the mechanical properties of the investigated masonry portions, direct inspections, fiberscope investigations on bore-holes or existing cavities, visual analysis of extracted cores, carried out before and after the consolidation interventions, can be performed. Moreover, direct surveys during the repair interventions allow comparing and establishing a connection between the strengthening phases and the sonic tomography results.

Sonic tests equipment, procedure and data processing

The equipment to perform sonic tests consists of impulse hammer to initiate the stress waves and piezoelectric accelerometer to measure the vibrations of the wall, resulting from the propagation of the waves. The hammer is provided with three tips with different hardness, to be used according to the surface condition of the tested masonry wall and the desired energy and frequency content of the impulse.

Hammer and accelerometer are connected to an amplifier and an analog-to-digital converter, coupled to a laptop computer, in order to view in real time and store both the generated impulse waves and the propagating pulse waves. In the presented case studies, the sampling frequency used is about 120000 samples/s. The trigger time limit is 5 s, pretrigger recording time is about 10 ms and, after the waves are generated, the signals are recorded for a total time of about 20 ms.

The software for data acquisition, travel times measurement (defined as the time interval between initiation of the stress wave and reception by the accelerometer) and data storing has been purposely developed in LABVIEW^â. The software for tomographic data analysis [Ref 10] was developed in Visual Basic 6. It is based on the theoretical non rectilinear propagation of elastic waves.

The investigated section of masonry is divided into square pixels and the tomographic reconstruction gives a mean value of sonic velocity for each of these masonry wall portions. The dimension of the mesh is of about 20 cm and depends on the wavelength (i.e. frequency) of the propagating stress waves. The frequency of the waves, therefore, influences the accuracy of the method and should be controlled on site, in order to improve the resolution of the tests [Ref 6]. Finally, refined tomographic renderings can be obtained with many different programs, such as Matlab 5^â, Surfer 7^â and others.

The traveltime tomographies shown in the following were obtained by means of the equipment used and the software developed at the University of Padua.

Analysis of masonry morphology and state of conservation

In the following, some case studies concerning the application of sonic waves tomography for the analysis of masonry morphology and state of conservation are presented. In the first case, it was possible to detect the presence of three leaves in the masonry of a bell tower. The thickness and composition of the three layers was checked by means of other minor destructive tests. The results of different tests showed a very good agreement. In the second case study, sonic tomographies on a church façade allowed detecting areas of masonry in different states of conservation. It was also possible to characterize a badly cracked portion of masonry.

Santa Giustina Bell Tower (Padua)
The bell tower of the Benedictine Monastery of Santa Giustina (Fig. 1, left) was originally built in XIII century to be about 40 m tall. During the reconstruction of the monastery in the XVI-XVII centuries (Fig. 1, middle, shows the current plan of the complex), the height of the bell tower was almost doubled, preserving the original structure and erecting new floors.

Fig 1: Santa Giustina Monastery: external view of the bell tower (left); plan (middle), tested section of the bell tower (right).

In recent years, a very dangerous crack pattern, characterized by both very large and long cracks developing for the entire length of the lower tower portion and smaller and spread cracks related to the behaviour of masonry under constant compressive loads, was observed. Some urgent temporary measures were undertaken to counteract the serious structural damage, due to the fact that the most ancient portion of the bell tower did not have adequate mechanical characteristics to resist the doubled loads. The interventions consisted of local repairs, such as limited rebuilding of the main cracks, bed joint reinforcement and grout injection, and in the construction of a steelwork for the horizontal tying of the tower and the transversal confinement of masonry [Ref 11].

Inspections were mainly performed after the construction of the steel structure and allowed to characterize the morphology of the wall section. A horizontal section of the lower portion of the tower, 140 cm thick, was investigated with sonic tomography. The tested section had three available sides for transmitting/receiving the sonic waves (Fig. 1, right) and this allowed for an accurate tomographic reconstruction of the section itself. In the same portion of masonry also radar tests and corings were performed.

The sonic tomography showed the presence of an external layer characterized by the highest sonic velocities (from about 1800 to 2200 m/s). In fact, the external leaf of the tested wall was made of trachyte stone. The masonry section was then characterized by mean values of velocity varying from 1400 to 1800 m/s, and by an inner core characterized by the lowest values of sonic velocity (1000-1400 m/s, see Fig. 2, left). The results of the echo-radar test, carried out with an antenna with centre frequency between 500 and 900 MHz, showed the presence of several discontinuities, in correspondence of the areas characterized by different sonic velocities, most likely caused by the layering of the masonry section (see Fig. 2, middle). Finally, a core was extracted in the position shown in Fig. 2, right. The bore-hole was drilled starting from the inner side of the bell tower, until the stone layer was reached. It was possible to observe that the first part of the extracted core is constituted by an entire and solid piece of brickwork, while the second part is made of loose and poor material, used to fill the masonry wall. The two portions exactly corresponded, when superposed to the sonic tomography, to areas characterized by good values of sonic velocity (from 1400 to 1600 m/s), and to the inner core characterized by the lowest values of sonic velocities. The morphology of the tested wall was subsequently further verified by using endoscopy.

Fig 2: Sonic tomography on Santa Giustina Monastery Bell tower (left), comparison with the results of Ground Penetrating Radar (middle), comparison with an extracted core (right).

A very good agreement between the results of sonic tests and radar, coring, endoscopy, was thus obtained. The sonic tomography alone, however, was able to give useful and effective information about the morphology of the wall section.

Carmine Church (Padua)
The static behaviour of the Carmine Church façade in Padua was analyzed in 2001, before carrying out some conservation works. The analysis of the church was organized into five general steps: historical documentation, geometrical and critical survey, on-site testing with ND and MD methods, FE modelling, monitoring of cracks with removable extensometers.

Up to 10 m of height, the façade of the church is still made of the original 14^th Century masonry wall. This portion, that is 95 cm thick, is made of a mixed masonry, with layers of bricks and stones linked together with irregular mortar joints. The external surface of the wall is characterized by the presence of some loose elements and of relevant vertical cracks. Conversely, the upper part of the façade was destroyed by an earthquake in 1491 and was consequently rebuilt with more regular brickwork during the 16^th Century. Exfoliation and scouring of the materials, due to environmental factors, is the most evident damage in this portion of the façade.

A sonic test campaign was carried out in the lower part of the façade, at about 6 m of height. The area across one of the main vertical cracks was investigated (Fig. 3, left). Subsequently, single and double flat jack tests were carried out in the same area and in others, in order to measure the level of stresses and to evaluate the mechanical properties of the different portions of the façade. Direct sonic tests were carried out on a 6x7 grid (20x25 cm) of transmitting/receiving points. On the second and the fifth row and on the second and the fifth column of acquisition points, also sonic tomographies were performed (Fig. 3). The horizontal tomographies were passing, therefore, across the crack, while the vertical tomographies were made at a distance of 80 cm and 20 cm from the crack.

Fig 3: Survey of the investigated portion of the façade (left), scheme of a horizontal and a vertical sonic tomography (middle, right).

The sonic pulse velocity measured during the direct tests varied between 1000 and 2000 m/s. The values were lower on the fifth and sixth column, that actually were those nearby the crack (their mean velocity values were of 1200 and 1400 m/s, while on the other columns they were from 1400 to 1700 m/s). The velocity also showed a decrease towards the lower part of the tested area. The mean values of sonic velocity on the fifth and sixth row, for example, were equal to 1355 and 1504 m/s, while on the other rows the mean values were, respectively, 2009, 1905, 1835, 1720 m/s.

The tomographies showed the same distribution of velocities, confirming the results obtained with direct tests. The lower horizontal tomography, in fact, was characterized by smaller values of sonic velocity than the upper horizontal tomography (average values of velocity on the sections of about 1050 and 1450 m/s, respectively; see Fig. 4, left). Both the vertical tomographies were characterized by smaller values of velocity on the lower part of the tested wall sections (see Fig. 4, middle and right). These results could be related to the presence of many irregularities and loose elements in the lower part of the tested area, as can be also seen in the survey (Fig. 3, left).

Beside that, some of the ray-paths of the horizontal tomographies crossed the crack. The results, in fact, clearly show that the sonic velocities are layered parallel to the possible direction of propagation of the crack in the wall thickness. In the upper horizontal section, for example, the values of velocity vary from 1800 m/s at about 90 cm from the crack, to about 1100 m/s in correspondence of the crack (Fig. 4, left above). The difference in masonry consistency can be also detected from the vertical tomographies. While the mean value of velocity is about 1500 m/s for the tomography at 80 cm from the crack, the mean value on the section close to the crack decreases to about 1150 m/s (Fig. 4, middle and right). The results show, in general, that the average conditions of the wall are fair good, according to the main classifications available in literature [Ref 3, Ref 4]. Low values of sonic velocity (around 1000 m/s) allow defining areas of masonry in bad conditions. In particular, the clear and strong decrease of velocity in the wall sections and in the areas of wall sections close to the crack, point at that crack as a significant phenomenon that interests the façade in its entire thickness.

Fig 4: Results of the sonic tomographies: upper (left above) and lower (left below) horizontal tomographies; vertical tomographies at 80 cm (middle) and at 20 cm (right) from the crack.

Control of the effectiveness of grout injection

As it has been largely described in literature [Ref 1, Ref 3, Ref 4, Ref 5] and has been shown with the two previous examples, sonic tests and above all sonic tomographies are able to detect flaws and differences in brick and stone masonry properties. Therefore, they have been proposed for almost twenty years as inspection tool to check the effectiveness of those repair techniques that modify the state of consistency of masonry [Ref 3, Ref 7]. In particular, grout injections should operate by filling the voids and flaws and by consequently increasing both the density and the mechanical characteristics of the injected masonry. If grouting interventions are effective, they should be thus identifiable with the use of the sonic method. Other repair techniques could be also controlled by means of sonic tests, to be carried out before and after the intervention. Joint repointing, for example, produces a local (close to the wall surface) improvement of masonry characteristics. It should be thus possible to check its effectiveness by means of aimed sonic tests, e.g. indirect tests on the repointed wall face.

However, the discussion is here limited to the use of sonic tomographies to check the effectiveness of grout injections. In the described case study, it was also proposed a method for surveying the grouting phase. It was thus possible to relate the increase in sonic velocity with the quantity of injected grout.

Cittadella Town Walls (Cittadella, North of Padua)
The Cittadella Town Walls were built in 1220. They have a diameter of about 450 m and along their perimeter there are 32 towers, at interaxis of about 40 m (Fig. 5, left and middle). The wall is about 13-14 m high and 220 cm thick. The external leaves of the wall are made of alternate layers of pebbles and bricks (Fig. 5, right). The inner and thicker part of the wall is built as an "opus concretum", with horizontal layers of pebbles connected by thick layers of mortar. The periodical monitoring of the structure revealed an increasing out-of-plumb towards the external side of the walls. In the most critical area, it reached an amount of 3.5°, e.g. about 80 cm of horizontal displacement on a height of 13÷14 meters.

Fig 5: Aerial view of Cittadella (left, the red square is the strengthened area); view of a tower from inside the town (middle); texture of the external leaf of the walls (right).

The superficial foundations and the presence of an embankment on the internal side of the walls, reaching a height of even 360 cm on the external ground level, were the main causes for the out of plumb.

The restoration works consisted in positioning concrete struts on the external part of the walls, in reinforcing the towers' foundations, and finally in re-creating the external embankment to balance the trust of the internal one. Before carrying out this overall intervention, some repairs consisting in local rebuilding of cracks and grout injections (from the ground level to 3.2 m of height) were made. The main aim of the local repairs was of consolidating the masonry, which was characterized by the presence of some large vertical cracks and of macroscopic and distributed voids in the inner leaf [Ref 12, Ref 13].

Direct sonic tests on a 6x6 grid (60x30 cm) of acquisition points and a vertical tomography were performed before and after the injections to evaluate their effectiveness. Bore-hole corings and endoscopies were also carried out to cross-check the results of sonic tests.

The sonic velocities on the vertical tomography before injections were between 690 and 1030 m/s (see Fig. 6, left, mean value of about 830 m/s), values that correspond to bad condition of masonry (v<1000 m/s, [Ref 3, Ref 4]). After injection, all the velocities increased and their mean value was equal to 1110 m/s. The highest value of velocity was equal to 1390 m/s (Fig. 6, middle). These values, considering also the large thickness of the wall and the great number of joints crossed by sonic waves, reveal fair good condition of masonry. Moreover, the mean increase of velocity of about 35% on the section (Fig. 6, right), showed that the grout injections were effective in consolidating properly the masonry wall.

Fig 6: Results of the vertical sonic tomography: original conditions (left); after grout injection (middle); percentage increase of sonic velocities (right).

The highest increases in sonic velocities were generally localized in areas previously characterized by the lowest values, thus revealing an effect of homogenization of the wall section operated by the injection. The unique exception is constituted by the upper-central area of the section, (see Fig. 6, right), where a very small increase of velocity was detected. On the top of this portion of masonry, in fact, there was a tunnel excavated inside the walls provoking a large outflow of grout, which was thus not employed for the actual strengthening of masonry.

Fig 7: Comparison between the quantity of injected grout (brown) and the increase of sonic velocity in direct tests (blue) (left); substantial presence of grout verified by means of endoscopy (right).

The injection phase, in fact, was accurately checked in order to estimate the quantity of injected grout and detect the grout leakages from injection holes and from cracks in the wall. After carrying out the sonic tests and tomographies, it was possible to relate the highest increase in sonic velocity to the highest quantity of injected grout. Where this relationship was not clear, the differences in velocity could be related to the detected propagation of the grout inside the wall (Fig. 7, left).

In the lower part of the section, that showed the highest increase in sonic velocity (from 40 to 60%), a borehole 150 cm deep was drilled and a visual analysis of the extracted core, together with an endoscopic investigation of the hole, were carried out. These analyses confirmed the presence of grout distributed along the hole (Fig. 7, right) and the core.

Final remarks

By the presented case studies, it is possible to draw some conclusion on the use of sonic tomography. The obtained maps of velocities allowed qualifying the state of conservation of masonry, detecting the section morphology and the presence of defect, flaws and voids. Due to the low frequency used, the resolution of the method is limited and it was not possible to detect in detail the morphology of the tested areas. However, sonic tomographies gave an overall description of the low velocity areas and detected the presence of multiple leaves in masonry.

Sonic tomographies applied before and after grout injections were able to detect the change in masonry consistency. The effect of grout injections in terms of homogenization of the wall conditions was clearly revealed. By surveying the intervention phase, it was possible to find a fair good agreement between the increase in sonic pulse velocity and the quantity of injected grout. It was thus confirmed that sonic tomographies are a useful tool to control the effectiveness of masonry repairs and the extent of grout penetration.

The use of complementary ND and MD test, such as surveys, direct inspections, fiberscope investigations, etc., confirmed the results obtained and the reliability of sonic tomographies applied both for the analysis of masonry morphology and conditions and as inspection tool to check the effectiveness of repairs. As shown in the presented case studies, also the comparison of sonic tomographies with the results obtained by means of direct sonic tests, can give useful information for the analysis of data.

This kind of test still needs to be improved and calibrated with mechanical parameters of masonry and on the bases of the different masonry typologies. However, it is a useful tool for a completely non-destructive evaluation of masonry walls.

Acknowledgement

Engg. Daniele Facchin and Nicola Monteforte are gratefully acknowledged for their support in the data acquisition, processing and interpretation. The sonic tests on the Carmine Church and the Bissara Tower were carried out and analyzed also with engineering students Camilla Levorato and Matteo Biasiolo. The radar tests on Santa Giustina Bell Tower were carried out by Eng. Giuseppe Lenzi.

References

1. Epperson G.S., Abrams D.P. (1989) "Non destructive evaluation of masonry buildings", Advanced Construction Technology Center, Doc. N. 89-26-03, Urbana Illinois, October 1989
2. Riva G., Bettio C., Modena C. (1998) "Valutazioni quantitative di caratteristiche meccaniche di muratura in pietra esistenti mediante prove non distruttive", Materiali e Strutture, "L'ERMA" di Bretschneider Ed., n° 1
3. Berra M., Binda L., Anti L., Fatticcioni A. (1992) "Non destructive evaluation of the efficacy of masonry strengthening by grouting techniques". Proc. International workshop: Effectiveness of injection techniques for retrofitting of stone and brick masonry walls in seismic areas; Polytechnic of Milan, March 1992; pp. 63-70.
4. Forde M.C., Birjandi K.F. & Batchelor A.J. (1985) "Fault detection in stone masonry bridges by non-destructive testing", Proc. 2^nd International Conference Structural Faults & Repair, Engineering Technics Press, Edinburgh, pp. 373-379.
5. Abbaneo S., Berra M., Binda L. (1996) "Pulse velocity test to qualify existing masonry walls: usefulness of waveform analyses". Proc. 3^rd Conference: Nondestructive evaluation of civil structures and materials; Boulder, Colorado, September 1996; pp. 81-95
6. Valle S., Zanzi L., Binda L., Saisi A., Lenzi G. (1998) "Tomography for NDT applied to masonry structures: Sonic and/or EM methods" In "Arch bridges", A. Sinopoli Editor, Balkema, Rotterdam, pp. 243-252
7. Riva G., Bettio C., Modena C. (1997) "The use of sonic wave technique for estimating the efficiency of masonry consolidation by injection", Proc. 11th International Brick/Block Masonry Conference, Shangai, China, October 1997, pp. 28-39
8. Schuller M., Berra M., Fatticcioni A., Atkinson R., Binda L. (1994) "Use of tomography for diagnosis and control of masonry repairs" Proc. 10^th International Brick/Block Masonry Conference, Calgary, Canada, July 1994, pp. 438-447
9. Bongiovanni G., Celebi M., Clemente P., (1990) "The Flaminio obelisk in Rome: vibrational characteristics as part of preservation efforts", Earthquake Engineering and Structural Dynamics, Vol. 19, pp.107-118
10. Monteforte N. (1998) "Applicazione della tomografia sonica nei problemi di diagnosi strutturale". Graduation Thesis (in Italian), Tutor C. Modena. University of Padua, 1997-1998
11. Modena C. (1997) "Criteria for cautious repair of historic buildings", ", Evaluation and Strengthening of Existing Masonry Structures, L. Binda & C. Modena Ed., RILEM, pp 25-42
12. da Porto F. (2000). "Indagini sperimentali sull'efficacia di tecniche di consolidamento di murature storiche in pietra". Graduation thesis, Tutor Prof. C. Modena, University of Padua.
13. Modena C. (2002) "Sistema fortificato di Cittadella. Intervento di consolidamento statico su un tratto di cinta compreso tra Porta Treviso e Porta Bassano", in "Lavori in corso, nuovi progetti a Cittadella", edited by Associazione Architettando, Stampa Grafiche Antiga, Aprile 2002
14. Valle S., Zanzi L., Binda L., Saisi A., Lenzi G. (1998) "Tomography for NDT applied to masonry structures: Sonic and/or EM methods" In "Arch bridges", A. Sinopoli Editor, Balkema, Rotterdam, pp. 243-252