Non-destructive testing techniques to help the restoration of frescoes

S. Sfarra1*, C. Ibarra-Castanedo2, D. Ambrosini1, D. Paoletti1, A. Bendada2, and X. Maldague2

1Las.E.R. Laboratory, University of L’Aquila, Department of Industrial and Information Engineering and Economics (DIIIE), Via G. Gronchi 18, I-67100, L’Aquila (AQ), Italy 2Computer Vision and Systems Laboratory, Laval University, Dept. of Electrical and Computer Engineering, 1065, av. de la Médecine, Quebec City, Québec (QC), Canada

*stefano.sfarra@univaq.it; phone ++39 (0) 862 434362; fax ++39 (0) 862 431233.

Abstract

Among the several issues to be considered during fresco’s restoration, the understanding of the effectiveness of the intervention, the identification of the main chemical elements used in previous restorations and the attention to weak areas of the building structure, adjacent to frescoes, are of paramount importance. This work describes an integrated, non-destructive testing approach focusing on these three main issues. In particular, two frescoes of Giacomo Farelli are analyzed herein. These artworks were affected by a strong earthquake in 2009, which had a heavy impact on several cultural heritage objects in L’Aquila (Italy), including on the Santa Maria della Croce di Roio Church (1625), where these two frescoes are located. One of the frescos, which underwent a restoring before the quake, was previously tested by Electronic Speckle Pattern Interferometry (ESPI) before and after the restoration. These previous results are compared with new measurements carried out after the 2009 earthquake using InfraRed Thermography (IRT). The combined approach, ESPI-IRT, clearly highlighted that the structure of the frescoes was significantly affected by the earthquake, since the old subsurface cracks, restored before 2009, were once again evident after the earthquake. In addition, the presence of a subsurface niche containing an ancient statue, also detected by means of InfraRed Thermography, might contribute to increase the severity of the damages. Finally, the joint examination of these frescoes using Near-InfraRed (NIR) Reflectography and X-Ray diffractometry was crucial to confirm the presence of a radioactive chemical element in the wall painting.

Keywords: Electronic Speckle Pattern Interferometry (ESPI); InfraRed Thermography (IRT); Near-InfraRed (NIR) Reflectography; fresco; earthquake; restoration.

1. Introduction

The aging process in a work of art can have different effects depending on several factors such as the original materials, the surrounding environmental conditions and, above all, on poorly performed restorations or even on restorations that were adequately accomplished. In particular, the surface of frescoes, interacting with the environment, can be modified over time. Temperature and humidity variations can cause the appearance of micro-cracks or anomalous strains inside the structure itself. Air pollutants lead to biological and chemical processes, which, in conjunction with some atmospheric physical parameters, can cause a deterioration of materials [1]. The life span of frescoes strongly depends on the condition of the wall as well, the most important aspects to be considered being structural stability, moisture, and acidity. Many causes may decrease the wall load-bearing capacity. Regarding the structure, distinctions should be made between cracks due to poor foundations and insufficiently strong structural elements, or to some external factors such as war, fire, or excessive loads (e.g. an earthquake). In a simplified way, defective foundations, disturbed load transfers, and insufficient cohesion are the main causes of structural instability. Defective foundations and disturbed load transfer may lead to cracks that, in the worst case, continue to grow resulting in the eventual collapse of the building. In case of insufficient cohesion, the structure will exhibit numerous hairline cracks caused by slight differences in settling, thermal loads, or other factors [2].The monitoring of infrastructures with respect to integrity and stability, as well as the detection of damages, gives rise to a specific field of Engineering, namely Structural Health Monitoring (SHM), which is widely treated in literature [3-12]. When dealing with historical buildings, SHM can be a “daunting task” [13] because of the lack of information, possible high costs and the many constraints imposed by the conservation issues and the architecture itself.

Wall paintings are extremely complex artworks since they are an integral part of buildings and their conservation is strictly linked to the monument and to the complex interaction between outdoor and indoor conditions; therefore, fresco diagnostics needs to be performed in situ. Moreover, wall paintings are very heterogeneous and cover large surfaces. All the above remarks make the investigation of real frescoes a difficult task.In frescoes diagnostic literature, much work has been devoted to pigment analysis as well as to the chemical and/or biological study of the fresco surface. A relatively smaller attention has been given to the structural diagnostics of wall paintings. At present, this mainly relies on visual inspection (both in normal and raking light) and on the so-called tapping technique, an empirical method in which the restorer knocks on the surface and then listens to the different pitches of sound. The tapping, though useful in many cases, has however important drawbacks: fully based on the restorer’s expertise, it has poor repeatability and high costs. Furthermore, it is time consuming and cannot be easily applied to produce a full documented record of the diagnostics results.Attempts to translate the tapping technique to a more systematic procedure have been made since the 1990s [14]. This “acoustic” approach led to the introduction of the laser Doppler vibrometer technique [15-19] and of a new method based on the measurement of absorbed acoustic energy [20].Optical methods, based on holographic interferometry, have been widely applied as diagnostic tools in the conservation field [21]. These methods provide precise information about the localization and size of a great variety of defects; however, their use can be difficult for non-optically skilled operators and/or in routine inspections. Often holographic interferometry techniques are too complex and expensive for routine use, moreover the mechanical stability required to perform the measurements is a heavy constraint. Therefore, holographic interferometry is not suitable for in situ diagnostics (unless pulsed lasers are used), which is usually the case when dealing with frescoes.Many shortcomings of holographic interferometry can be overcome, at the expense of a slight decrease in sensitivity and image quality, introducing optical techniques such as Electronic Speckle Pattern Interferometry (ESPI), speckle decorrelation and shearography [22-24]. These new techniques are simple, portable, safe, and easy to use by non-specialists for routine in situ inspections. Recently, a new method, called Mirror Micrometry, capable of detecting fresco surface displacement in the m range [25] and terahertz spectroscopic imaging for the evaluation of underdrawings and paint layers embedded within the wall [26], were proposed.Generally speaking, each technique has its own advantages and disadvantages, however, the drawbacks of a single diagnostic technique can be partially overcome by integration; therefore the combination of different techniques, sometimes in the same portable equipment, is now largely used. In this paper an integrated approach is proposed for the examination of wall paintings based on traditional (visual inspection, tapping, NIR, XRD) as well as innovative (ESPI, IRT) techniques. The use of different methods and instruments, as well as the integration of the techniques, is also reported. Emphasis is given to extending monitoring to the structure considering not only the wall paintings but also the surrounding structures.New results can be achieved by using nowadays traditional techniques performed with new performance and characteristics such as high resolution, portability and versatility. The results obtained could be arranged, if integrated into a multidisciplinary approach, in order to define and design the conservation and the restoration of the work.The case of Giacomo Farelli fresco “The discovery of the Statue of Our Lady” in the Santa Maria della Croce di Roio Church, is unique in its kind because it combines problems of structural instability, identified as insufficient cohesion between the fresco layers during the restoration performed in 1994, with an excessive mechanical load induced by the 2009 earthquake, which reopened the previously restored cracks.The correlation between the thermographic results after the earthquake, and the ESPI measurements, before the 1994 restoration, is very important. The non-destructive testing by ESPI after the restoration, confirms the good results obtained only in the immediacy of the intervention. The structural instability previously mentioned is probably due to a buried architectural structure near to the fresco and beneath to it, detected by IRT. Pulsed phase thermography (PPT) and principal component thermography (PCT) algorithms also allowed to detect some new cracks due to the earthquake mechanical stress, confirmed by NIR both on “The discovery of the Statue of Our Lady” and on “The episode of mare’s knees” frescoes.The Sanctuary of Santa Maria della Croce was built in 1625, as an extension of a small chapel dedicated to St. Leonard (1221).The interior of the new Church, in the Baroque style, remained unfinished for several years and the façade was completed only in 1673. The friendship between the city of L’Aquila and Naples allowed an extensive exchange of artists that brought in L’Aquila influences from the Neapolitan school, including the Maestro Andrea Vaccaro (1598-1670), whose pupil was Giacomo Farelli, active in L’Aquila since the second half of the seventeenth century [27-29].In 1667, Farelli executed two frescoes on the sides of the main altar depicting the two highlights of the Virgin Statue presence in the place where previously, as already mentioned, there was a chapel.

Figure 1 a, b

Figure 1a represents the discovery of the Statue in a niche of the Three Saints Wood in Ruvo di Puglia, while Figure 1b represents the effort made by farmers to lift up the mare.The frescoes, which have a surface of approximately 5 m2, were probably made using the typical technique at Farelli’s time, known as the buon fresco [30]. Detailed explanations about this construction technique are given in the next section.The fresco in Figure 1a has been tested by ESPI, IRT, NIR and X-ray diffractometer methods, whilst the fresco in Figure 1b was inspected only by thermographic and reflectographic methods. Figure 2a and Figure 2b show, respectively, a niche before and after the 2009 earthquake. As will be seen in Figure 19, the void containing a statue discovered during the initial restoration (near the fresco shown in Figure 1a) had been identified prior to the earthquake by IRT.

Figure 2 a, b

2. Masonry cracking and technical execution of the frescoes analyzed

Stone masonry is a traditional form of construction practiced for centuries in the regions where stone is locally available. It is still found in old historic centers, often in buildings of cultural and historical importance, and in developing countries where it represents an affordable and cost-effective housing construction. This construction type is present in earthquake-prone regions of the world, such as Mediterranean Europe and North Africa, the Middle East, India, Nepal, and other parts of Asia. The World Housing Encyclopedia contains nine reports describing stone masonry housing construction practices. The reference in Italy is WHE Report 28 and for the area studied in this work the structural walls are made of the following: a) rubble stone in mud/lime mortar or even without mortar, b) two exterior wythes of larger stones with rubble infill in mud/lime mortar, often without through stones that should connect the exterior wythes [31-33].The most important factors affecting the seismic performance of these construction typologies are: a) the strength of the stone and mortar, b) the quality of construction, c) the density and distribution of structural walls, d) wall intersections and floor/roof-wall connections. Stone masonry construction frequently show poor seismic performance. Poor quality of mortar is the main reason for the low tensile strength of rubble stone masonry. Timber floor and roof structures are usually not heavy and therefore do not include large seismic forces. However, typical floor structures are made of timber joists that are not properly connected to structural walls. An evidence of this type of structure that was installed below the fresco named “The episode of mare’s knees” in order to support it, is reported in section 6 (Results and discussion). These structures are rather flexible and are not able to act as rigid diaphragms. Due to their large thickness, stone masonry walls are rather heavy and induce significant seismic forces.Delamination and disintegration of the masonry are damage patterns typical for walls built with two exterior wythes and rubble infill in weak mud mortar with many air voids. Out-of-plane failure can occur when the connections between the exterior and interior walls are inadequate. When the connections between the perpendicular walls are strong, the wall shear capacity can be exhausted, thus causing typical shear cracks to develop [34-37].Since the stone wall can be considered as the first and inner layer in a fresco painting, and in the conservation purposes the knowledge of incipient and invisible flaws, as well as the understanding of how the presence of the support cracks or discontinuities alter the movements of the painted surface are key points in the restoration field, both frescoes were inspected by IR Thermography in the front and rear part [38].Cracking encompasses relatively narrow fissures 0.04 – 1.27 [cm] in masonry units, mortar, or grout alone or in any combination. It may be caused by a variety of conditions, such as improper material choices (primarily the wrong masonry unit/mortar combination), wrong field installation (plugged weep holes or hinged/slide masonry anchors loaded with mortar to the point of impairing the free movement), oversized anchors, structural settlement, etc [39-40].Cracking is the most common and most visible form of cladding problem, in a range encompassing everything from cosmetic defects to total failure. Sometimes the location, orientation and form of the crack give an indication of the cause of the failure. Masonry deterioration may start from something as insignificant as small cracks in some of the individual units (often a minor esthetical rather than structural problem), then progresses through cycles of thermal movement alone, or in combination with freezing and thawing cycles, to wider cracks over a larger area or through cycling loads, which sooner or later may become structurally significant, especially if not mapped and monitored early in the process [41-42].Taking into account these considerations, the cracks that afflict the interior of the Church were totally mapped after the 2009 earthquake (Figure 3).

Figure 3

Historical notes on the frescoes analyzed in this work were found reading two ancient books written by hand [43-44]. Although in this case the main cause of the cracks formation on the frescoes was due to structural instability and to architectural buried structures, it is important to provide details about the technical execution of the buon fresco in order to understand that this type of artworks must be considered as a composite material (Figure 4).

Figure 4

Lime putty is the main ingredient of the buon fresco painting. Preparation of painting surfaces for fresco involves the applications of plaster of increasingly finer texture. The first step is the heating (calcinations) of the limestone (calcium carbonate, CaCO3) at 800 – 900 °C to make porous lime (calcium oxide, CaO). Excess water acts as a lubrificant so that the crystals slide easily over one another [45]. Historically, lime was slaked in pits or troughs over a period of at least six months to obtain lime putty of the desired consistency. Artisans in Michelangelo’s time used plaster aged for as long as ten years. Fresco plaster itself is made from the slaked lime and varying portions of sand or marble dust. Generally, walls are plastered with several layers of such fresco plaster in order to decreasing proportions and particle size of sand.Hardening of the fresco plaster on the wall includes several simultaneous physical and chemical processes: the absorption of water into the wall, evaporation of water from the surface, and the carbonation of the slaked lime by carbon dioxide, CO2.Fresco plaster coats are made of high calcium lime putty and “aggregate” which, most commonly is washed river sand, marble meal, volcanic tuff or a combination of it. The proportion of the mortar or plaster mix generally is: 1 part lime putty and 2 parts aggregate (sand) or 5 parts lime putty and 8 parts aggregate (sand). The latter was the mix preferred by Neapolitan school [46].Washed rived sand is the best aggregate for making a fresco plaster; it is clean from impurities such as silica, dust, clay, organic particles, and salts. This sand is also most likely to be of a right angular shape needed for “proper interlocking”. Traditionally, there are six distinctive fresco plaster coats (from last to first): a) intonaco and intonachino or skim coat: final plaster coat on which the actual painting is done; b) sinopia: final preparatory drawing on the arriccio on which intonaco is applied. It was normally executed in red ochre; c) arriccio or brown coat: smooth, sand finish coat on which sinopia is applied; d) float coat: smooth plaster coat, base for arriccio coat. This coat is fine levelled and floated with a large wooden float. Usually, this would be a “conventional” stucco; e) rough plaster coat made with somewhat coarse sand; f) scratch coat: the initial and one of the most important coats. This coat is applied directly to the wall and then scratched with “tooth edged” trowel [30].

3. Electronic Speckle Pattern Interferometry (ESPI)

A diffusely scattering surface illuminated by laser light appears covered by a pattern of bright and dark spots, or speckles, distributed randomly in space. This occurs because neighbouring microscopic elements that constitute the object surface produce random differences of the optical path for the scattered light. At any point, therefore, scattered waves arrive from many of these elements simultaneously and, as they are highly correlated, their instantaneous amplitudes add coherently. However, as the phases are random they may provide at any point constructive (bright speckle) as well as destructive (dark speckle) interference [47]. Considered by the majority of holographers as the stain of coherent illumination, this “annoyance” can still provide information about the surface characteristics as well as its displacement.The combination of the imaged speckle pattern and a reference wavefront to produce a phase-referenced speckle pattern can be used in a way similar to holographic interferometry. The idea of linking this speckle interferogram with a TV camera has the advantage of setting these techniques apart from holography. The ability to capture images by a video system combined with enhancements achieved by subsequent electronic processing allowed the development of a new, more efficient method, called electronic speckle pattern interferometry, for using in situ.The electronic speckle pattern interferometry (ESPI, also called TV-holography) was developed as a method of producing interferometric data without using traditional holographic recording techniques [48], [49], [50]. Practically, ESPI can be viewed as the combination of holography and speckle interferometry, the holographic film being replaced by a CCD camera as the recording medium [51]. Obviously, the photosensor of the TV camera is not suitable for optical reconstruction of the hologram; therefore, the reconstruction process is performed electronically and visualized on a monitor.The signal picked up by the TV camera is converted into a corresponding video signal by the camera scanning action. This video signal is electronically processed through an intermediate recording medium (commonly a

frame grabber) before being displayed on a TV screen, so that the variations in the texture of the speckles are converted into a variation of brightness. This image is entirely equivalent to a holographic reconstructed image and possesses the same interferometric sensitivity.In this way, the specklegrams are recorded under ambient parameters variations. Some irregularities of the corresponding correlation fringes allow to detect the presence of detachments or cracks.In this section is given only a brief practical description of the portable system used for the measurements; the full theory is well covered in the literature [29], [52]. Figure 5 schematically shows the experimental configuration for in situ diagnostics. Other configurations can be retrieved in literature [53], [54], [55], [56]. A laser diode (P = 15 mW) is used as the light source. The laser diode junction is temperature stabilized by a thermo-electric controller. Furthermore, the laser system includes a very low-noise current driver. The laser source is coupled to a polarization maintaining single mode fiber.

Figure 5

The laser light emitter by the output end of the optical fiber directly illuminates the surface under investigation. A small part of the spherical wavefront, caught by a prism, serves as reference beam. A beam-splitter cube reflects the reference light onto the CCD sensor, where the object image is focused by an objective lens. An interference filter, centred at the laser diode wavelength, is introduced in front of the imaging lens to prevent external light from reaching the CCD camera. The images captured by the CCD camera are recorded by a PC-based frame-grabber. Besides, the PC system is equipped with temperature and relative humidity transducers. The short exposure time (1/30 s) of the ESPI technique provides a greater stability and tolerance of external disturbances with respect to traditional holographic techniques; therefore, the ESPI system can be used outside the research laboratory [52]. To obtain a good quality ESPI interferogram, the system, as well as the surrounding environment, must remain stable during the exposure time to at least λ/8. A solution which can mitigate these problems, consists of assembling all components in an optical head mounted on a high stiffness tripod [57]. However, the effect of residual relative rigid displacement between the object and the optical head will be to reduce the fringe contrast, which can be partially recovered by image enhancement techniques, as shown in this work.

4. Near-Infrared (NIR) Reflectography

When exposing a painting to a broad-band light source (from ultraviolet to the far IR), part of the radiation will be absorbed by, another fraction of the radiation will be transmitted through and the rest will be reflected from the incident surface, depending on the radiation wavelength and the material being radiated. For instance, a visible camera will capture the light (in the visible spectrum 0.35 to 0.75 μm) reflected from the painting surface, providing information about colours and textures. The NIR part of the radiation, which contains practically no thermal emissions, can penetrate thin layers of painting before being reflected back to the surface from a non-absorbing media such as the preparation surface (usually made of chalk and gypsum) and will eventually be absorbed by carbon based (or other absorbing) elements, if present. Most of the oil paints used for panel painting (usually linseed oil with inorganic suspended oxide or mineral salt pigments) are transparent to NIR light, whilst carbon derivatives (graphite and charcoal) are opaque in this spectral region [58].The transparency in the NIR band is a complex function of the optical characteristics of (1) the pigment colour (with brown and gray being in general more transparent than some light colours, whilst black is the most opaque), (2) the underdrawing material, (3) the paint layer thickness (typically a fraction of millimetre), and (4) the detector wavelength (transparency increases between 1.0 and 2.5 μm for different configurations, generally showing a peak near to 2 μm) [58-61]. A NIR camera can be used to reconstruct two-dimensional (2D) images, i.e. reflectograms, of the reflected light under the painting layers. Interesting applications include the detection of guiding sketches and signatures (opaque to NIR radiation) drawn by the artist prior to the application of painting layers; the detection of hidden paintings (painters often used a previously painted canvas or changes their mind during the painting progression), the monitoring of the restoration processes required on aging cultural heritage artworks, and the detection of intentional and unintentional alterations.NIR reflectography has been studied since the 1930s. At the beginning, photographic films were used. Although NIR photography works are interesting, restrictions on the spectral band (0.7 to 0.9 μm) and time delays (due to film development) limited the wide spread of the technique (an interesting NIR photography investigation can be found in [62]. It wasn’t until the 1960s, after the work of Van Asperen de Boer, that Vidicon cameras (0.9 to 2.0 μm) first and digital cameras (1.1 to 5 μm) later, began to be used routinely by many recognized art Museums [63-66]. The next generation of NIR reflectography systems is the multi-spectral (up to 14 spectral bands) single-point scanners, which considerably diminish the effects of optical and geometrical non-uniformities with respect to multi-detectors arrays [59, 67]. The use of modified commercial NIR cameras, with suitable lenses, filters, and light sources, is still a popular alternative for artwork inspection given its easiness of operation

compared to single-point scanners. In this work, a CMOS camera (Canon 40DH, with a 22.2 x 14.8 mm sensor – 10 megapixel resolution and 0.38-1.0 m spectral sensitivity), with a visible cut-off filter to limit the spectrum in the range 0.7 to 0.9 m was used for reflectography inspection.

5. Square Pulse Thermography (SPT)

Active thermography is a well-known NDT technique, allowing fast inspection of large surfaces that has been extensively investigated [68, 69]. Data acquisition is fast and straightforward, as illustrated in Figure 6.

Figure 6

A long-wave infrared camera (ThermaCAM S65HS produced by FLIR, 7.5-13 m, 320x256 pixels) was employed for thermographic testing.In the square pulse configuration the specimen surface is submitted to a long square heat pulse (from a few seconds to several minutes), and the temperature raise and decay is registered using an infrared camera and stored as a 3D matrix composed by N thermograms, where x and y are the spatial coordinates, and t is the time [70, 71].SPT data is generally processed to improve defect visibility and to perform quantitative characterization of defects. In this paper, we propose to use principal component thermography and pulsed phase thermography to process SPT data.

5.1 Principal Component Thermography (PCT)

Singular value decomposition (SVD) is an eigenvector-based transform that forms an orthonormal space, which is close to principal component analysis (PCA), with the difference that SVD simultaneously provides the PCAs in both row and column spaces. The SVD of an MxN matrix A (M>N) can be calculated as follows [72]:

A=URVT (1)

where U is a MxN orthogonal matrix, R being a diagonal NxN matrix (with singular values of A present in the diagonal), VT is the transpose of an NxN orthogonal matrix (characteristic time). Hence, in order to apply the SVD to thermographic data, the 3D thermogram matrix representing time and spatial variations has to be reorganised as a 2D MxN matrix A. This can be done by rearranging the thermograms for every time as columns in A, in such a way that time variations will occur column-wise while spatial variations will occur row-wise. Under this configuration, the columns of U represent a set of orthogonal statistical modes known as empirical orthogonal functions (EOF) that describes spatial variations of data [73]. The first EOF will represent the most characteristic variability of the data; the second EOF will contain the second most important variability, and so on. Usually, original data can be adequately represented with only a few EOFs. Typically, a 1000 thermograms sequence can be replaced by 10 or less EOFs.

5.2 Pulsed Phase Thermography (PPT)

Pulsed phase thermography (PPT) is another attractive technique, in which data is transformed from the time domain to the frequency domain using the one-dimensional discrete Fourier transform (DFT) [74, 75]:

Fn=Δt ∑k=0

N−1

T (kΔt ) exp(− j 2π nk / N )=Ren+ j Imn (2)

where j is the imaginary number (j2= -1), n designates the frequency increment (n=0,1,…N), t is the sampling interval, and Re and Im are the real and the imaginary parts of the transform, respectively. In this case, real and imaginary parts of the complex transform are used to estimate the amplitude A, and the phase [76]:

An=√Ren2+Imn

2 and

φn= tan−1 (Imn

Ren) (3)

The DFT can be used with any waveform (e.g. transient pulsed thermographic profiles). Phase profiles for surface temperature are anti-symmetric, providing redundant information in both sides of the frequency spectra. In the following, only the positive part of the frequency spectra is used whilst the negative frequencies can be safely discarded. The phase is of particular interest in NDE given that it is less affected than raw thermal data by environmental reflections, emissivity variations, non-uniform heating, and surface geometry and orientation.

These phase characteristics are very attractive not only for qualitative inspections but also for quantitative characterization of materials.

6. Results and discussion

To guide the readers through the large number of information given in this section, a diagram that summarizes all the main steps relative to the NDT campaigns is reported in Figure 7.

Figure 7

First of all, the fresco “Discovery of the Statue of Our Lady” shown in Figure 1a, has been inspected by ESPI technique during the 1994 restoration. The fresco was interlaced by micro-cracks, evidence of forces at work in the wall. In addition, comparing the IRT results coming from the year 2009 (June) and the year 2012 (July), it is possible to reveal a grows of the subsurface cracks (Figure 14d). In this case, the indication reported above about the cracks mapping in the early stage of the process has been useful during the data analysis. Without removing the fresco from its normal exhibition environment, the whole surface was analyzed by ESPI technique; only some meaningful results are reported here. The map of some detachments between upper layers and the support is presented in Figure 8. Tap test was also used in order to confirm the ESPI results.

Figure 8

After the 2009 L’Aquila earthquake, the fresco was studied by square pulse thermography, focusing the attention on the same area inspected in 1994, where subsurface micro-cracks were found by ESPI. The area studied by ESPI is marked in Figure 8 and in Figure 9 (green colour), while the area studied by square pulse thermography is underlined in Figure 9 (yellow colour).

Figure 9

Because of the subtraction nature of ESPI measurements, the interferograms contain no visible details of the object. As the restorers need to know the exact location of the defects on the artwork, digital techniques can be used to solve this problem. For this reason, it is necessary before realizing the ESPI interferogram, to capture the details of the object by illuminating the object itself with incoherent light. This image must be realized by the video camera of the ESPI system, located in the same position used for recording the speckle pattern. Subsequently, by edge-detection processing, it is possible to obtain the edge map of the object under investigation [77]. By means of a digital addition to the ESPI interferogram, is obtained an exact location of the cracks in relation to the fresco (Figures 10, 12). The area within the green rectangle measures 20 cm2 and was subjected to a brief thermal irradiation using a 250 W infrared lamp at a distance of approximately 1 m. Specklegrams were recorded continuously during the cooling process.

Figure 10

Abrupt deviations along the trend of some correlation fringes are evident. These discontinuities correspond to the presence of cracks. Figure 11 shows a deformation map of the selected region. This three-dimensional representation of the cracking area was obtained using the method proposed by Kreis [78].

Figure 11

After a restoration it is important to monitor over time the state of the artwork in order to take further decisions. Until now, the examination has been primarily a matter of experience; the use of the portable ESPI system described above allows taking decisions about the conservation of the fresco during each phase of the restoration, taking into account factors such as the microclimate. For this purpose, some tests were made at the same location after that the fresco has been restored. The interferogram reported in Figure 12 was realized under the same experimental conditions of the previous one (Figure 10). Figure 12, which corresponds to the area reported in Figure 8, shows how the cracks and the surrounding area have been repaired.

Figure 12

Figure 13 shows some results obtained by IRT after the earthquake (June 2009); it can be said that, apparently, the same cracks shown in Figure 10 and supposedly repaired in 1994, reappeared due to the earthquake of April 2009. The choice of using IRT instead of ESPI technique after the earthquake, was mainly suggested by security considerations: IRT can capture a large area in a single frame and, compared to ESPI, is quicker and does not require a scaffold. These are key issues, taking in account that, because of the earthquake, the Santa Maria della Croce Church is a restricted area.

Figure 13 a, b, c, d

Figure 13b is a phasegram obtained by pulsed phase thermography on the whole data sequence, i.e. including heating and cooling phases. It is possible to see in this figure different kinds of cracks: the same detected before the restoration, and also some surface features such as painting traces (note the hand at the centre). PCT was applied to the heating and cooling phases, separately. Results are shown in Figures 13c and 13d, respectively. During the heating phase, which lasted 180 s, the thermal signature of surface and near the surface features is very high. Hence, features in Figure 13c are located at the surface or close to it, for instance, the lady’s hand and cloths painting traces near the centre of the image. In the case of the cooling phase, deeper features are seen since heat has the time to propagate through the material. Focusing the attention on Figure 13b, it is possible to note an interesting subsurface crack with a rectangular shape marked by red arrows. Since the first crack detected by ESPI and subsequently by IRT was located at the bottom left corner of the rectangle, the analysis using IRT technique were repeated working with different heating times and considering this marker as a key point. In the first one (July 2012), an IR lamp (2 KW) was used as in the previous test but for a longer time (1968 s) in order to verify whether possible inhomogeneous loads inside the wall may have been responsible of the damage during the 2009 earthquake (Figure 14).

Figure 14 a, b, c, d

In the second test (August 2012), a propane gas heater was employed to inspect both the front and the back side of the wall of the “The discovery of the Statue of Our Lady” fresco, in an attempt to detect a buried window inside the wall structure. The front and back sides of the fresco were heated for 40 minutes, and the cooling down was processed by PCT. Processed results relative to the front side are presented in Figure 15.

Figure 15 a, b

It is possible to see a warmer area in the center top of Figure 15a (EOF1), showing a contrast difference between the supposed buried architectural structure (white area) and the surrounding area (dark area) as well. A more clear indication of the buried window can be seen in Figure 15b, where quasi-parallel cracks can be observed. The back side results are shown in Figure 16. Given the limited area available to carry out the inspection, only a relatively small part was tested. Still, it is possible to observe a horizontal pattern (highlighted by arrows) that resembles a group of stones or bricks forming a line, which is supposed to correspond to the bottom edge of the ancient window.

Figure 16 a, b, c, d

The experiment to investigate both the front and the rear side of a real fresco using IRT for confirming the presence of an architectural buried structure, was approached in this work for the first time to the best of our knowledge.Comparing the 2009 PCT results (Figure 13 c) with the 2012 PCT results (Figure 14 c) it is possible to observe further deterioration of the cracks passing through the central area inspected, which might be related to the presence of the buried window.In addition, it is interesting to note the presence of different subsurface cracks, shown in Figures 16 c, d, not detected by naked eyes. The latter result explains the importance of applying IRT technique during the restoration of an architectural heritage seriously damaged by an earthquake. In fact, the visual inspection of the left aisle (Figure 3) revealed only shallow cracks and not deep cracks (Figure 16 c, d). These two results should be combined and taken into account in the restoration phase.In order to detect underdrawings and show more clearly the surface defects that occurred after the earthquake, the NIR technique was also used for the same area analyzed in June 2009 by IRT (Figure 17).

Figure 17

As can be seen in Figure 17, no underdrawings were detected. Still, the use of NIR reflectography enhances the surface and subsurface cracks, as we compare this result with the image of the same area shown in Figure 13a. The emissivity of a pigment, highlighted with a black rectangle, is enhanced by NIR technique. This pigment was analyzed in a laboratory and identified as white lead. The X-ray diffractometer result is shown in Figure 18.

Figure 18

In fact, white lead has a measurable radioactive trace with appropriate tools, which can roughly determine the age and disappears completely after a minimum period of 160 years. Drying is fast enough and produces a very elastic film; robustness cannot be reached with any other white, but tends to lose coverage, over the years. It tends to darken, due to the action of hydrogen sulphide traces in the air; in fact, oxidized, it turns into lead oxide, brown in colour. This tendency to become dark is much more evident when it is used with binders (mural painting, tempera) and in the presence of moisture [79]. Probably, the buffered niche (Figure 2a and 2b) adjacent to the fresco (Figure 3), identified by a raw thermogram (Figure 19) during another IRT campaign (2006), was crucial in the mechanical vibration caused by the earthquake, as much as the buried window (Figure 15 b).

Figure 19

In our experience, the integration between different NDT techniques, such as optical and thermographic, has furnished interesting results for samples with real and fabricated defects [80-85]. The possibility of comparing data previous to and subsequent to a mechanical shock for an artwork, as in the case of “Discovery of the Statue of Our Lady” fresco, it is unique in its kind.The radiation source during the reflectographic and thermographic inspections consisted of 2 kW halogen lamp (STAR), which provides both a wide spectrum radiation (including the NIR band) and heat stimulation for IRT technique. Instead, to inspect the niche two infrared lamps 250 W (OSRAM SICCATHERM) were used in reflection mode.In “The episode of mare’s knees” fresco shown in Figure 1b, it was important to integrate PCT and NIR results in order to produce a map of the surface and subsurface cracks (Figure 20). This map could be very useful for the restorers in order to try to understand the more appropriate way to repair this fresco.

Figure 20 a, b, c, d

In this case, NIR technique was also important to reveal a retouching probably made during old restorations or by the author himself. In fact, as can be seen in Figure 21, an original character was incorporated into the mantle of another character.

Figure 21 a, b

In a second experiment realized in order to define more clearly the subsurface cracks marked by red lines in Figures 20 b, c, the IR camera was placed perpendicular to the “The episode of mare’s knees” fresco and a new IR data set was acquired. Seeing the Figure 3, readers can understand that the two frescoes discussed in this work were painted inside the same room and one in front of each other. Checking an ancient text [86] it is possible to realize that Farelli painted “The discovery of the Statue of Our Lady”, but it is not sure that he is the author of “The episode of mare’s knees” fresco. In fact, the signature and the date (1667) are only reported on the first one.In addition, Figure 22a is very interesting if compared with the wall’s typology of “The discovery of the Statue of Our Lady” fresco (Figure 15 a, b). This is a first clue that could indicate the building of the walls that divide the two aisles to the room (Figure 3) in two different time periods. The second clue is given below.

Figure 22 a, b, c, d

These IR results were obtained working with: a) a frame rate of 1 Hz, b) a 2 KW IR lamp, c) a heating phase of 1940 s and a cooling phase of 1256 s.Four subsurface spots having a rectangular geometry were detected in Figure 22a (red dashed ovals). Taking into account their position (on the highest or lateral part of the fresco), these thermal anomalies could be due to

reinforcement elements of the wall. The NIR result (Figure 22d) was arranged with the EOF6 result (Figure 22b) to define a cracks’ map reported in Figure 22c. EOF4 clearly shows the four subsurface dark spots.Since this part of the Church includes the small chapel dedicated to St. Leonard (1221), it is plausible to assume that they are lead clamps (manufactured by the builders) poured in the molten state inside some artificial voids of the stones [87]. In addition, it is impossible to detect these anomalies by NIR technique (Figure 22d). At this point, we can say that they are deeper than the subsurface cracks.As explained elsewhere [88], the wall section frequently hides a complicated technique of construction in two or three leaves of different thickness; these leaves are connected in different ways or sometime are not connected at all. In the present case, the reinforcement elements previously described connect adjacent stones frontally. Anyway, the characteristics of masonry as a composite are frequently unknown nowadays, since the knowledge about the construction techniques was lost decades ago, i.e. when the new construction materials as steel and concrete became in use [89].The proposed method of inspecting firstly the fresco by an IR lamp in order to detect the shallow configuration of the upper layers (applying both the thermographic and the reflectographic methods), and secondly using a propane gas heater in order to reveal the deeper texture of the wall (working in the first step on the front, and in the second stage on the rear side) was also used on “The episode of mare’s knees” fresco.The EOF2 result coming from the first step applying SPT is shown in Figure 23b.

Figure 23 a, b

The latter figure confirms that applying as heating source hot air for several minutes (40 minutes), it is possible to reveal the construction technique of the wall (note the shape of the stones at the center of the figure). In addition, the assumption about the role of the lead clamps is confirmed after having detected the shape of two stones at the upper right corner (highlighted by dark circles) of the same figure. The other two clamps located at the upper edge of the fresco (Figure 22 a, c), are not now identified because outside the field of view, while the fourth clamp is still detectable although the shape is not well defined (Figure 23b).Finally, “The episode of mare’s knees” fresco was inspected by IRT (heating: 630 s, cooling: 630 s) from the rear side (Figure 24a). In this case, both the raw thermogram capture at t = 630 s and the EOF2 result show some interesting subsurface structures. For example, Figure 24b exhibits three horizontal buried structures (A, B and C). Structure A is completely covered by plaster and it is at higher level compared to the visible stone beam on the right side. This beam seems to be thicker if connected with the A and C subsurface structures. In the same way, the subsurface beam A was also detected by PCT (Figure 24c).

Figure 24 a, b, c

The latter figure clearly reveals a paraboloid-shaped structure (marked with a white dotted line in Figure 24a). This structure forms a cusp with the second paraboloid-shaped structure visible to the naked eyes (marked with a dark dotted line in Figure 24a). The junction is located between the subsurface beams, named: A (on the left), B and C (on the right). It appears as a dark spot in Figure 24c and is slightly visible in Figure 24a. The non conjunction between the A, B, C, and the visible beams is surprising, and may indicate, taking into account the shape cusp detected, a modification of the ceiling during the time.The horizontal subsurface structure detected in Figure 16b (rear side of “The discovery of the Statue of Our Lady” fresco) is not present in this case. In fact, by using the propane gas heater in order to heating the front side of “The episode of mare’s knees” fresco, another buried window was not revealed (second clue) in the thermographic images.

Conclusion

Many current methods of optical testing are not widely available for the examination of frescoes because laboratory facilities are required. The satisfactory operation of the TV holographic system described in this work is illustrated by its application to the analysis of Farelli’s fresco named “The discovery of the Statue of Our Lady” situated inside Santa Maria della Croce di Roio Church: the system is very simple to use in situ. The same considerations can be done for square pulse thermography used outside the laboratory, with the advantage that this technique is able to investigate a wider area with less time, and is able to detect large voids in a wall that could affect the surface stability during an earthquake (Figure 19). In order to confirm the presence of subsurface cracks, the NIR technique was very useful both to detect a pentimento of the author in the second fresco of the Church named “The episode of mare’s knees”, and to characterize, together with X-ray diffractometer technique, the presence of a radioactive chemical element in “The discovery of the Statue of Our Lady” fresco.

This work confirms that ESPI and the infrared vision may be considered generally for all types of work where one can foresee a degrading reaction to ambient stimuli or mechanical stress (e.g. earthquake); these stimuli generally cause internal detachments and cracks due to structural instabilities (Figure 14d). Another consideration is the importance of establishing a monitoring program following the restoration of a fresco of great historical value. In fact, the same cracks supposedly repaired by the 2004 restoration (Figure 12) were reassessed following the earthquake of 2009 (Figure 13b), at least for the area highlighted with a green rectangle (Figure 10). This part of the fresco was inspected by optical, NIR and thermographic NDT techniques in different times (Figures 13, 14).Comparing the cracks revealed inside this area from June 2009 to August 2012, it is possible to say that the buried window detected by PCT in Figure 15 acts as a point load rather than a distributed load.Exacerbating the situation there is the void caused by the door located below “The discovery of the Statue of Our Lady” fresco, and the void inside the wall due to a buffered niche (Figure 3).The same considerations inherent the first fresco (Figure 1a) relatively to the joint approach between the infrared and optical NDT techniques before and after a catastrophic event, could suggest the use in the future of ESPI technique for the second fresco (Figure 1b), currently investigated only by infrared vision. This work enhances the possibility of reveal subsurface cracks [90] and architectural buried structures also by SPT-PCT, and confirms the literature data inherent the better place (in front of the object instead of tilted respect to it) where to put an IR camera in order to clearly detect the subsurface cracks [91]. However, comparing the results coming from the inspections reported in Figures 20 and 22, it is possible to obtain more details about the subsurface defects in composite structures [83].In addition, the use of propane gas to heat the structures for a longer period was found worthy, providing indications of the presence of the supposedly buried window beneath “The discovery of the Statue of Our Lady” fresco in both rear and front sides of the fresco. Furthermore, it was possible to have a better idea of the nature of the building structure, such as stones patterns. This integrated result can be easily interpreted by the restorer who will focus on restoration of these artworks.

Acknowledgements

The authors want to thank the support of the Soprintendenza per i Beni Ambientali, Architettonici, Artistici e Storici per l’Abruzzo (Italy) for granting permission to carry out experiments on the Farelli’s frescoes in the Santa Maria della Croce di Roio Church and for providing us the result of X-ray diffractometer. Figures 1, 5-13, 18: Courtesy of WIT Press from (WIT Transactions on The Built Environment Volume 118, 2011, p. 784). The authors would also like to thank Prof. Lilliana Genova for the comments on the English language.

References

[1] Mora, P., Mora, L., Philippot, P.: Conservation of wall paintings, Butterworths, London (1984).[2] Arendt, C.: The role of the architectural fabric in the preservation of wall painting. In: Proceedings of a

symposium organized by the Courtauld Institute of Art and the Getty Conservation Institute, London (1987).

[3] Farrar, C.R.; Worden, K. (Eds.): Theme Issue ‘Structural health monitoring’, Phil. Trans. R. Soc. A 365, 299–632 (2007).

[4] Farrar, C.R.; Worden, K.: An introduction to structural health monitoring, Phil. Trans. R. Soc. A 365, 303–315 (2007).

[5] Brownjohn, J.M.W.: Structural health monitoring of civil infrastructure, Phil. Trans. R. Soc. A 365, 589–622 (2007).

[6] Sohn, H.; Farrar, C.R.; Hemez, F.M.; Czarnecki, J.J.; Shunk, D.D.; Stinemates, D.W.; Nadler, B.R.: A Review of Structural Health Monitoring Literature: 1996–2001. Los Alamos National Laboratory Report, LA-13976-MS (2003).

[7] Chang, P.C.; Flatau, A.; Liu, S.C.: Review Paper: Structural Health Monitoring, Structural Health Monitoring 2, 257-267 (2003).

[8] Mukhopadhyay, S.C. (Ed.): New developments in sensing technology for structural health monitoring, Springer-Verlag, Berlin Heidelberg (2011).

[9] Vedalakshmi, R.; Thangavel, K.: Reliability of electrochemical techniques to predict the corrosion rate of steel in concrete structures, Arab. J. Sci. Eng. 36, 769-783 (2011).

[10]Meier, U.: Carbon Fiber Reinforced Polymer Cables: Why? Why Not? What If?, Arab. J. Sci. Eng. 37, 399-411 (2012).

[11]Gjørv, O.E.: Durability of concrete structures, Arab. J. Sci. Eng. 36, 151-172 (2011).

[12]Breen, J.E.: Techniques for improving durability of post-tensioned concrete bridges, Arab. J. Sci. Eng. 37, 303-314 (2012).

[13]Anastasi, G.; Lo Re, G.; Ortolani, M.: WSNs for structural healtyh monitoring of historical buildings, Proc. IEEE 2nd Conference on Human System Interactions, 21-23 May 2009, Catania (Italy).

[14]D’Ambrogio, W.; Mannaioli, A.; Del Vescovo, D.: Use of FRF measurements as a non-destructive tool to detect detachment of frescoes, Proc. 12th International Modal Analysis Conference, Honolulu (1994).

[15]Castellini, P.; Paone, N.; Tomasini, E.P.: The laser Doppler vibrometer as an instrument for non-intrusive diagnostics on works of art: application to fresco painting, Opt. Las. Eng. 25, 227-246 (1996).

[16]Castellini, P.; Esposito, E.; Legoux, V.; Paone, N.; Stefanaggi, M.; Tomasini, E.P.: On field validation of non-invasive laser scanning vibrometer measurement of damaged frescoes: experiments on large walls artificially aged, J. Cult. Herit. 1, S349-S356 (2000).

[17]Del Vescovo, D.; Fregolent, A.: Assessment of fresco detachments through a noninvasive acoustic method, Journal of Sound and Vibration 284, 1015–1031 (2005).

[18]Del Vescovo, D.; Fregolent, A.: Theoretical and experimental dynamic analysis aimed at the improvement of an acoustic method for fresco detachment diagnosis, Mechanical Systems and Signal Processing 23, 2312-2319 (2009).

[19]Collini, L.; Garziera, R.; Mangiavacca, F.: Development, experimental validation and tuning of a contact-less technique for the health monitoring of antique frescoes, NDT&E Int. 44, 152-157 (2011).

[20]Calicchia, P.; Cannelli, G.B.: Detecting and mapping detachments in mural paintings by non-invasive acoustic technique: measurements in antique sites in Rome and Florence, J. Cult. Herit. 6, 115-124 (2005).

[21]Paoletti, D.; Schirripa Spagnolo, G.: Interferometric methods for artwork diagnostics, in: E. Wolf (Ed.) Progress in Optics, Vol. XXXV, Elsevier, Amsterdam (1996).

[22]Rastogi, P.K. (Ed.): Digital Speckle Pattern Interferometry & Related Techniques, Wiley, New York (2000).

[23]Ambrosini, D.; Paoletti, D.: Holographic and speckle methods for the analysis of panel paintings. Developments since the early 1970s, Reviews in Conservation 5, 38-48 (2004).

[24]Hinsch, K.D.; Zehnder, K.; Joost, H.; Gülker, G.: Monitoring detaching murals in the Convent of Müstair (Switzerland) by optical metrology, J. Cult. Herit. 10, 94-105 (2009).

[25]Zehnder, K.; Voûte, A.: Monitoring detaching murals in the Convent of Müstair (Switzerland) by mirror micrometry, J. Cult. Herit. 10, 493-500 (2009).

[26]Jackson, J.B.; Mourou, M.; Whitaker, J.F.; Duling, I.N.; Williamson, S.L.; Menu, M.; Mourou, G.A.: Terahertz imaging for non-destructive evaluation of mural paintings, Opt. Comm. 281, 527–532 (2008).

[27]Antonucci, E.: Roio e il suo Santuario. Editrice Eco, S. Gabriele dell’Addolorata (TE) (1986).[28]Murri, F.: Roio e la sua storia. Editrice La Fonte, Pescara (1989).[29]Paoletti, D.; Schirripa Spagnolo, G.: The potential of portable tv holography for examing frescoes in situ,

Stud. Conserv. 40, 127-132 (1995).[30]Merrifield, M.P.: The art of fresco painting in the middle ages and the renaissance. Dover Publications,

Inc., U.S.A. (2004).[31]World Housing Encyclopedia, A joint project by EERI and IAEE, www.world-housing.net, accessed

September 08, 2012.[32]Binda, L.; Modena, C.; Baronio, G.; Abbaneo, S.: Repair and investigation techniques for stone masonry

walls, Construction and building materials 11, 133–142 (1997).[33]Lutman, M.: Stone masonry construction, www.world-housing.net/wp-content/uploads/2011/06/Stone-

Asonry_Lutman.pdf, accessed September 08, 2012.[34]Penazzi, D.; Valluzzi, M.R.; Cardani, G.; Binda, L.; Baronio, G.; Modena, C.: Behaviour of historic

masonry buildings in seismic areas: lessons learned from the Umbria-Marche earthquake, Proc. 12 th Int. brick/block masonry conf., Madrid, Spain (2000).

[35]Binda, L.; Gatti, G.; Mangano, G.; Poggi, C.; Sacchi Landriani, G.: The collapse of the civic tower of Pavia: a survey of the materials and structure, Masonry International 6, 11–20 (1992).

[36]Binda, L.; Baronio, G.; Gavarini, C.; De Benedictis, R.; Tringali, S.: Investigation on materials and structures for the reconstruction of the partially collapsed cathedral of Noto (Sicily), 6 th Int. conf. structural studies, repairs and maintenance of historical buildings, STREMAH 99, Dresden, Germany (1999).

[37]Binda, L.; Baronio, G.; Gambarotta, L.; Lagomarsino, S.; Modena, C.: Masonry constructions in seismic areas of central Italy: a multi-level approach to conservation, 8NAMC, Austin, USA (1999).

[38]Schirripa Spagnolo, G.; Guattari, G.; Grinzato, E.; Bison, P.G.; Paoletti, D.; Ambrosini, D.: Frescoes diagnostics by electro-optic holography and infrared thermography, NDT.net 5, 2000.

[39]Hendry, A.W.: Masonry walls: materials and construction, Construction and Buildings Materials 15, 323-330 (2001).

[40]Abrams, D.P.: Strength and behaviour of unreinforced masonry elements, Earthquake Engineering – Tenth world conference, Rotterdam (1992).

[41]De Felice, G.; Giannini, R.: Out-of-plane seismic resistance of masonry walls, Journal of Earthquake Engineering 5, 253-271 (2001).

[42]Griffith, M.C.; Vaculik, J.; Lam, N.T.K.; Wilson, J.; Lumantarna, E.: Cyclic testing of unreinforced masonry walls in two-way bending, Earthquake Engineering & Structural Dynamics 36, 801-821 (2007).

[43]Signorini, A.: La Diocesi di Aquila descritta ed illustrata, Vol. 1. Stabilimento tipografico Grossi, L’Aquila, Italy (1868).

[44]Montorio, S.: Lo Zodiaco di Maria ovvero le dodici province del Regno di Napoli, come tanti segni, illustrate da questo sole per mezzo delle sue prodigiosissime immagini, che in esse quasi tante stelle risplendono, Napoli, Italy (1715).

[45]Hale, G.: The technique of fresco painting. Dover Publications, Inc., U.S.A. (2000).[46]Mayer, R.: The artist’s handbook of materials and techniques. Viking, U.S.A. (1991).[47]Goodman, J.W.: Statistical properties of laser speckle patterns. In: Dainty, J.C. (ed.) Laser speckle and

related phenomena 9, p. 9. Springer, Berlin (1975).[48]Butters, J.N.; Leedertz, J.A.: Holographic and video techniques applied to engineering measurements, J.

Meas. Cont. 4, 349-354 (1971).[49]Macovski, A.; Ramsey, S.D.; Schaefer, L.F.: Time-lapse interferometry and countering using television

systems, Appl. Opt. 10, 2722-2727 (1971).[50]Schwomma, O.: Osterreichisches Patent 298 830 (1972).[51]Ennos, A.E.: Speckle interferometry. In: Dainty, J.C. (ed.) Laser speckle and related phenomena 9, p.

203. Springer, Berlin (1975).[52]Lokberg, O.J.; Malmo, J.T.: Long-distance electronic speckle pattern interferometry, Opt. Eng. 27, 150-

156 (1988).[53]Paoletti, D.; Schirripa Spagnolo, G.: Application of fibre optic digital speckle interferometry to mural

painting diagnostics, Meas. Sci. Technol. 4, 614-618 (1993).[54]Kato, J.; Yamaguchi, I.; Ping, Q.: Automatic deformation analysis by a TV speckle interferometer using

a laser diode, Appl. Opt. 32, 77-83 (1993).[55]Zanetta, P.: Optical techniques applied to measurement in Art, Ph.D. Thesis, Loughborough University

of Technology (UK) Technical note ISEI/IE/2777/94, Commission of the European Communities, Joint Research Center, ISPRA Site (1994).

[56]Albrecht, D.: Electronic speckle pattern interferometry instruments development, optimization and applications, Ph.D. Thesis, Loughborough University of Technology (UK) Special Publication No. I.99.40, Commission of the European Communities, Joint Research Center, ISPRA Site (1999).

[57]Thomson, W.T.: Theory of vibration with application. Prentice-Hall, New York (1972).[58]Dinwiddie, R.B.; Dean, S.W.: Case study of IR reflectivity to detect and document the underdrawing of a

19th century oil painting. In: Proc. SPIE - Thermosense XXVIII (2006).[59]Fontana, R.; Bencini, D.; Carcagni, P.; et al.: Multi-Spectral IR reflectography. In: Proc. SPIE - Optics

for Arts, Architecture and Archeology (2007).[60]Walmsley, E.; Metzger, C.; Delaney, J.K.; et al.: Improved visualization of underdrawings with solid-

state detectors operating in the infrared, Stud. Conserv. 39, 217–231 (1994).[61]Van Asperen de Boer, J.R.J.: Reflectography of paintings using infrared vidicon television system, Stud.

Conserv. 14, 96–118 (1969).[62]Desneux, J.: Underdrawings and pentimenti in the pictures of Jan Van Eyck, The art Bullettin 40, 13-21

(1948).[63]Van Asperen de Boer, J.R.J.: Infrared reflectograms of panel paintings, Stud. Conserv. 11, 45-46 (1966).[64]Rijksmuseum: Infrared reflectography. http://www.rijksmuseum.nl/aria_encyclopedia/00047885?

lang=en. Accessed 21 May 2009.[65]The Cleveland Museum of Art: Infrared reflectography.

http://www.clevelandart.org/exhibcef/picassoas/htmlk/7464063.html. Accessed 21 May 2009.[66]The art Institute of Chicago: Examination techniques.

http://www.artic.edu/aic/conservation/revealingpicasso/exam_infared.html. Accessed 21 May 2009.[67]Obrutsky, A.E.; Acosta, D.: Reflectography, An NDT method for images diagnosis.

http://www.ndt.net/article/wcndt2004/pdf/optical_techniques/740_obrutsky.pdf. (2004). Accessed 14 January 2012.

[68]Maldague, X.P.V.: Theory and practice of infrared technology for non-destructive testing. John Wiley & Sons, New York (2001).

[69]Balageas, D.L.; Krapez, J.C.; Cielo, P.: Pulsed photothermal modeling of layered materials, J. of Appl. Phys. 59, 348-357 (1986).

[70]Vavilov, V.; Kauppinen, T.; Grinzato, E.: Thermal characterization of defects in building envelopes using long square pulse and slow thermal wave techniques, Res. Nondestruct. Eval. 9, 181-200 (1997).

[71]Arndt, R.W.: Square pulse thermography in frequency domain as adaptation of pulsed phase thermography for qualitative and quantitative applications in cultural heritage and civil engineering, Infrared Phys. & Technol. 53, 246-253 (2010).

[72]Rajic, N.: Principal component thermography for flaw contrast enhancement and flaw depth characterization in composite structures, Compos. Struct. 58, 521-528 (2002).

[73]Marinetti, S.; Grinzato, E.; Bison, P.G.; Bozzi, E.; Chimenti, M.; Pieri, G.; Salvetti, O.: Statistical analysis of IR thermographic sequences by PCA, Infrared Phys. & Technol. 46, 85-91 (2004).

[74]Maldague, X.P.; Marinetti, S.: Pulse Phase Infrared Thermography, J. Appl. Phys. 79, 2694-2698 (1996).[75]Ibarra-Castanedo, C.; Maldague, X.: Pulsed phase thermography reviewed, QIRT J. 1, 47-70 (2004).[76]Meola, C.; Carlomagno, G.M.: Recent advances in the use of infrared thermography, Meas. Sci. Technol.

15, 27-58 (2004).[77]Castleman, K.R.: Digital Image Processing. Prentice Hall, U.S.A. (1996).[78]Kreis, T.: Digital holographic interference-phase measurement using the Fourier-transform method, J.

Opt. Soc. Am. A 3, 847-855 (1986).[79]Piva, G.: L’arte del restauro – Il restauro dei dipinti nel sistema antico e moderno – Third Edition. Hoepli

Ed., Milano (1988).[80]Bendada, A.; Sfarra, S.; Ambrosini, D.; Paoletti, D.; Ibarra-Castanedo C.; Maldague, X.: Active

thermography data processing for the NDT&E of frescoes. In: Proc. 10 th International Conference on QIRT, Quebec, Canada (2010).

[81]Ibarra-Castanedo, C.; Sfarra, S.; Ambrosini, D.; Paoletti, D.; Bendada, A.; Maldague, X.: Subsurface defect characterization in artworks by quantitative pulsed phase thermography and holographic interferometry, QIRT J. 5, 131-149 (2008).

[82]Sfarra, S.; Theodorakeas, P.; Ibarra-Castanedo, C.; Avdelidis, N.P.; Paoletti, A.; Paoletti, D.; Hrissagis, K.; Bendada, A.; Koui M.; Maldague, X.: Importance of integrated results of different non-destructive techniques in order to evaluate defects in panel paintings: the contribution of infrared, optical and ultrasonic techniques. In: Proc. SPIE - Optics for Arts, Architecture and Archeology, Münich, Germany (2011).

[83]Sfarra, S.; Ibarra-Castanedo, C.; Ambrosini, D.; Paoletti, D.; Bendada A.; Maldague, X.: Integrated approach between pulsed thermography, near-infrared reflectography and sandwich holography for wooden panel paintings advanced monitoring, Russ. J. Nondestruct. 47, 284-293 (2011).

[84]Sfarra, S.; Theodorakeas, P.; Ibarra-Castanedo, C.; Avdelidis, N.P.; Paoletti, A.; Paoletti, D.; Hrissagis, K.; Bendada, A.; Koui M.; Maldague, X.: Evaluation of defects in panel paintings using infrared, optical and ultrasonic techniques, Insight 54, 21-27 (2012).

[85]Sfarra, S.; Ibarra-Castanedo, C.; Ambrosini, D.; Paoletti, D.; Bendada, A.; Maldague, X.: The use of optical and infrared techniques for the restoration of the frescoes damaged by the earthquake: a case study - the fresco of Giacomo Farelli in the Church of Santa Maria della Croce di Roio (L’Aquila, Italy) . In: Brebbia, C.A.; Binda, L. (ed.) Structural studies, repairs and maintenance of heritage architecture XII - Twelve international conference on structural repairs and maintenance of heritage architecture, STREMAH 12, pp. 627-642, WIT Press, USA (2011).

[86]Manuelli, G.: Prima visita pastorale, Italy (1931/1934).[87]Breymann, G.A.: Costruzioni civili. Di Baio s.r.l. Ed., Stoccarda (1853).[88]Binda, L.; Saisi, A.: Application of NDTs to the diagnosis of historic structures. In: NDTCE’09 – Non-

Destructive Testing in Civil Engineering, France (2009).[89]Binda, L.; Penazzi, D.; Saisi, A.: Historic masonry buildings: necessity of a classification of structures

and masonries for the adequate choice of analytical models. In: 6th Int. symposium on computer methods in structural masonry – STRUMAS VI, Italy (2003).

[90]Sham, F.C.; Chen, N.; Long, L.: Surface crack detection by flash thermography on concrete surface, INSIGHT 50, 240-243 (2008).

[91]Ibarra-Castanedo, C.; Sfarra, S.; Ambrosini, D.; Paoletti, D.; Bendada, A.; Maldague, X.: Diagnostics of panel paintings using holographic interferometry and pulsed thermography, QIRT J. 7, 85-114 (2010).

FIGURE CAPTIONS

Figure 1 a) Fresco: “The discovery of the Statue of Our Lady” by G. Farelli (1667) before the restoration of 1994, b) Fresco: “The episode of mare’s knees” supposedly painted by G. Farelli (1667)

Figure 2 a) Niche: before the 2009 L’Aquila Earthquake, b) Niche: after the 2009 L’Aquila Earthquake

Figure 3 Cracks map of the Santa Maria della Croce di Roio Church, Scale 1:100

Figure 4 Cross-section of an ancient fresco – 3D modeling (layers)

Figure 5 ESPI – Experimental set up. BS: beam splitter cube; CD: current driver; H: humidity sensor; IF: interference filter; T: temperature sensor; TEC: thermo-electric controller

Figure 6 Experimental setup for square pulse thermography

Figure 7 Diagram showing all the main steps relative to the NDT campaigns

Figure 8 Drawing of the fresco “The discovery of the Statue of Our Lady”, with highlighted zones of detachment, and the area studied by the integrated approach

Figure 9 Fresco, “The discovery of the Statue of Our Lady” by G. Farelli (1667) after the earthquake of 2009, with marked the areas studied by ESPI (green rectangle) and by IRT (yellow rectangle)

Figure 10 False colour specklegram of the area marked in green in Figure 8, after digital image processing

Figure 11 Three-dimensional representation of the deformation in the cracking area shown in Figure 10

Figure 12 The same area of Figure 8 after restoration. Specklegram after digital image processing: filtering, enhancement and in false colour

Figure 13 Active thermography results (June 2009), a) Photograph of the inspected area, b) phasegram obtained by pulsed phase thermography; and principal component thermography results for, c) the heating phase and, d) the cooling phase of the sequence

Figure 14 a) Photograph of the inspected area with highlighted the cracks detected in 2009 by white colour, while the cracks revealed in 2012 are marked by yellow colour; PCT results (July 2012): b) EOF1, c) EOF3, d) EOF4 with a summary crack’s mapping during three years (2009 – 2012)

Figure 15 PCT results (August 2012), a) EOF1, b) EOF3

Figure 16 a) Photograph of the inspected area marked by black dotted rectangle of the “The discovery of the Statue of Our Lady” fresco – rear side; PCT results (August 2012): b) EOF1, c) EOF2, d) EOF3

Figure 17 NIR result of the area analyzed by IR Thermography reported as (yellow) rectangle in Figure 13a

Figure 18 X-ray diffractometer result of the particular pigment highlighted by a black rectangle in Figure 17

Figure 19 Raw thermogram captured before the 2009 earthquake, during the cooling stage. The inspected area is shown in Figure 2a

Figure 20 a) Photograph of the inspected area of “The episode of mare’s knees” fresco, b) EOF1 heating, c) EOF2 heating, d) NIR result

Figure 21 a) Photograph of the inspected area of “The episode of mare’s knees” fresco, b) NIR result: pentimento

Figure 22 Fresco: “The episode of mare’s knees” supposedly painted by G. Farelli (1667): a) EOF3, b) EOF6, c) EOF4, d) NIR result

Figure 23 a) Photograph of the inspected area of “The episode of mare’s knees” fresco – front side, b) EOF2

Figure 24 a) Photograph of the inspected area of “The episode of mare’s knees” fresco – rear side, b) raw thermogram at t = 630 s, c) EOF2

FIGURES

a) b)

Figure 1

a) b)Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

(a) (b)

(c) (d)

Figure 13

(a)

(b)

(c) (d)

Figure 14

(a)

(b)

Figure 15

(a)

(b)

(c)

(d)

Figure 16

Figure 17

Figure 18

Figure 19

(a) (b)

(c) (d)

Figure 20

(a) (b)

Figure 21

(a)

(b)

(c) (d)

Figure 22

(a)

(b)Figure 23

(a)

(b)

(c)

Figure 24

AB

A

C