- Email: [email protected]
Analysis of building facade defects using infrared thermography: Laboratory studies
Author’s Accepted Manuscript ANALYSIS OF BUILDING FACADE DEFECTS USING INFRARED THERMOGRAPHY: LABORATORY STUDIES Elton Bauer, Elier Pavón, Eva Barreira, Eliane Kraus www.elsevier.com/locate/jobe
PII: DOI: Reference:
S2352-7102(16)30021-3 http://dx.doi.org/10.1016/j.jobe.2016.02.012 JOBE105
To appear in: Journal of Building Engineering Received date: 27 August 2015 Revised date: 5 February 2016 Accepted date: 25 February 2016 Cite this article as: Elton Bauer, Elier Pavón, Eva Barreira and Eliane Kraus, ANALYSIS OF BUILDING FACADE DEFECTS USING INFRARED THERMOGRAPHY: LABORATORY STUDIES, Journal of Building Engineering, http://dx.doi.org/10.1016/j.jobe.2016.02.012 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ANALYSIS OF BUILDING FACADE DEFECTS USING INFRARED THERMOGRAPHY: LABORATORY STUDIES Elton Bauer1*; Elier Pavón1; Eva Barreira2;Eliane Kraus1 1
Post-Graduate Program in Structures and Civil Construction, Civil and Environmental
Engineering Department, University of Brasília - Campus Universitário Darcy Ribeiro Asa Norte, 70910-900 Brasília/DF, Brazil 2
Faculty of Engineering, Dept. of Civil Engineering, Univ. of Porto, Laboratory of
Building Physics, Rua Dr. Roberto Frias s/n, 4200-465 Porto, Portugal *
Corresponding author. [email protected]
Abstract One of the main problems facing the inspection of facades by means of passive infrared thermography, in order to identify and evaluate defects or pathologies, is the definition of the most appropriate moment to perform the inspection. Most studies focus on the Delta-T value as a criterion for the identification and evaluation of defects. Nevertheless, the identification of defects will primarily depend on the heat flux, since the way defects appear on the thermogram is determined by their type and by the direction of the flux. This study set out to analyse this problem by studying the behavior of the Delta-T and contrast functions in defects in ceramic tiles and mortar. Accordingly, test samples were made with induced defects during the manufacturing process, to simulate the detachment problems that occur with ceramic tiles of different thicknesses and cracks of different depths. The test samples were assessed during direct and inverse heating cycles. The results showed that it is possible to detect differences in behavior when varying the thickness of the ceramic tiles and the depth of the cracks by considering the Delta-T and contrast functions. In consequence, the inspection moment should be defined according to the behavior of the gradient temperature, the type of the defect and the direction of the heat flux.
Keywords: Infrared thermography; Façade; Delta-T; Thermal contrast; Temperature gradient; Pathologies 1
1. Introduction Infrared thermography, involving an analysis of surface temperatures, is able to identify internal anomalies since heat transportation is affected by the presence of faults, inclusions, moisture, and other occurrences. These anomalies change the heat flux pattern and thus affect the resulting surface temperature, relative to the surface temperature of a defect free surface [1]. Although internal defects can be detected, infrared thermography is considered to be applicable to surfaces since it more easily identifies anomalies close to the surface. The possibility of identifying anomalies will not depend only on the characteristics and depth of the defects. Elements or components, in thermal or hygroscopic equilibrium are difficult to study with thermography [2]. The thermal flux in plan elements is obviously not one-dimensional. However, the thickness being smaller, it is observed that in this direction occur the fastest answers of the alteration of the temperatures. That way it is possible to study these heat flux in a simplified form which permits the identification of defects and anomalies. The gradient temperature between the external and internal surfaces (facade), as well as the direction of the heat flux that passes through the wall, will determine when a defect will be visible and how it can be identified in a thermogram. It can be characterized as being either a cooler or warmer zone. The heat flux in a building envelope obviously is neither static nor constant throughout the day, and is closely dependent on the ambient temperature, solar heating, and cooling of the facade. The performing of a thermographic inspection for delamination or detachment detection it is often recommended at the moment of the day when the values of the heat flux in the facade are greater, that is, when the greatest wall temperature gradient values appear (this study was conducted in stone panels) [3]. A study of the solar trajectory in order to identify the potential effects of sunlight and shade, which will define the heat flux and, consequently, the conditions for the detection of anomalies, is of great importance. The use of sun charts for this assessment is simple and effective but must be completed with priority to the inspection [4]. There are different conditions to consider when identifying anomalies that depend on orientation, making it necessary to study both schemes, direct and inverse heat flux, in order to define the analysis criteria for each inspection. In such cases, it is even more 2
important to conduct a pre-analysis of the heat flux schemes. The aim is to understand the heat flux conditions in the facade, and the behavior of defects in different directions of flux. Applying thermography, it is possible to identify and map the areas with defects. The forms in which have been analysed the defects in facade elements (ceramic tiles and mortar) in laboratory studies are presented in the next items.
2. Theoretical background Different forms of analysis are used to identify anomalies with infrared thermography. The considered to be the most important are the visual temperature differences in the thermogram (used in qualitative thermography) and the Delta-T (used in quantitative thermography). Qualitative thermography is based on a relatively simple analysis to identify hot spots and cold spots of the thermal image by color difference, that is, a visual analysis. The analysis is made by a comparison with standard situations, and the evaluation depends heavily on the expertise of the evaluator [5]. Quantitative thermography is used to classify the importance of an anomaly, being necessary to obtain as accurately as possible the temperature of the target object. It is necessary then measure and appropriately define the thermographic parameters (eg emissivity, reflected temperature). Graphical representations and contrast functions have also been used in other studies as complementary tools for the identification or quantification of defects [6,7]. The thermogram visual differences have been used to enable the identification of cold and hot areas in building facades, thermal bridges, presence of structural elements, like beams and pillars, different materials like mortar and brick, failures in the insulating materials [8], as well as common pathologies like cracks [9,10] and moisture [11,12]. The most commonly used criterion in quantitative thermography is associated with the Delta-T between the area with defects and the defect free surrounding area [13]. However, the use of this criterion does not lead to a conclusive assessment when identifying anomalies in facades [6,14]. It is necessary to consider the gradient temperature and to determine the conditions under which anomalies in the thermograms can be better identified [15]. The study developed by Bauer et al. [1] proved that it is possible to quantify the defects created in a sample, thus exhibiting excellent reproducibility in the three cycles that were evaluated. Despite the differences in the 3
temperatures recorded by the two different thermographic cameras in the study, the Delta-T, quantitative criterion used to evaluate the defect, showed very small differences between the two cameras, proving that it can be used for the evaluation of damage and defects in facades as long as the variations in the temperature gradient are considered in the analysis process. Studies conducted during the day to evaluate defects in facades have been carried out considering the influence of the variation of exterior and interior air temperatures. De Freitas et al. [16] studied polymeric plaster detachment on a facade, obtaining different Delta-T values throughout the day. They concluded that the best period for the inspection was during the hours of the greatest exposure to sunlight (at noon). Edis et al. [17] detected moisture problems from the heat gain of the facade during the daytime and nighttime. These authors recommend carrying out the inspection during the night because the Delta-T, between the moist area (with defects) and dry area (without defects), is stable for a longer period. In this study, the magnitude of Delta-T was also higher around noon. One of the main problems facing the thermographic inspections of facades is when to get the thermal imaging [18]. Thermographic inspections are normally based on finding the thermogram with the highest Delta-T (the defect will always be more visible when the value of the Delta-T is the highest). However, in these situations, the physical limits of the defects are not always well identified, hindering the precise delimitation of the anomalies. Delta-T not only depends on the temperature gradient, as the type of the defect, its size, and the depth at which it exists can modify the Delta-T’s behavior [13]. In studies of small thickness defects, located at a shallow depth relative to the surface, the value of Delta-T tends to be constant after the initial heating moments at the laboratory study, as occurred in the studies developed by Bauer et al. [1] and Freitas et al. [16]. The identification of the optimum time (moment) to perform a thermographic inspection to identify and evaluate defects is a problem which has been studied in cases of active and passive thermography, when a specific heat flux is induced in a sample. In addition to a visual and Delta-T analysis, in active thermography, contrast functions are used as an auxiliary tool for defect identification. In some specific material studies, the value of the contrast function can indicates the presence of the defect [19]. However, in studies of building elements in which the defects are larger, it is expected that this function can 4
be used to determine the optimal time to identify the defects. The thermal events that occur within, an area incorporating defects, and a defect free normal area, change with time because the heat dissipation varies with time, among other variables. Thus, the selection of the most adequate thermogram , at a particular moment in time, is required in order to extract and define the boundaries of the local defects. The selection criteria for this thermogram is generally based on a thermogram with the highest thermal contrast. This contrast can corresponds to the differential temperature of the defects over the differential temperature from a free defect area [20]. Maldague [13] used the contrast function called ‘Standard Contrast’ (Cs), which is calculated according to Equation 1. The contrast is the temperature variation in the area with defects, relative to the evolution of the temperature of the closest defect free area.
() =
( )! ( " ) # ( )!# ( " )
(1)
Where Cs(t) is the ‘Standard Contrast’, $% () is the temperature of the defective area at time t, $% (& ) is the temperature of the defective area at time t0, $ () is the temperature of the defect free area at time t, $ (& ) is the temperature of the defect free area at time t0, and & is the start time (beginning of the cycle). A similar contrast function (standard temperature contrast) was used by Nowak and Kucypera [21] to study the presence of concealed materials inside the walls in laboratory. This function contrast is the temperature variation at any selected point on surface of the material under testing relative to the temperature variation at surface point over homogeneous area prior to thermal stimulation. According to Nowak and Kucypera [21], the ‘Standard Contrast’ (Cs) is independent of the type of material being tested; it is dimensionless and oscillates around a constant value close to one in the steady state, such that it is possible to compare the results of various experiments, as was done by Bauer et al. [14]. Vavilov [19] used a contrast function called ‘Running Temperature Contrast’ (C) in the study of sensitivity and noise in thermographic measurements. Basically, this contrast function is obtained from Equation 2, being determined during the periods of heating. These values are generated for each time and correspond to the evolution of the Delta-T (temperature of the region incorporating defects minus that of the defect free region), in 5
relation to the evolution of the average temperature of the sample. With this evaluation, it should be possible to identify the optimum times at which to identify anomalies.
() =
' ( )!*' ( ) ( )
(2)
Where () is the ‘Running Temperature Contrast’, $+ () is the temperature of the defective area at time t, $,+ () is the temperature of the non-defective area at time t, and T(t) is the sample excess temperature (evolution of the sample temperature during heating phase). Bauer et al. [14] employed this modified function to analyse the behavior of defects in ceramic tiles by finding close values in the heating and cooling cycles. The objective is to discuss, based on laboratory studies, which the conditions that allow better visualization and identification of anomalies in facade inspections, using passive thermography. Thus, this paper discusses the problem of selecting the optimum time at which to capture a thermal image during a thermographic inspection, based on a Delta-T behavior analysis and Cs and C contrast functions for both of the most common problems addressed by facade studies: the detachment of ceramic tiles and cracks in rendering mortar.
3. Experimental development The experimental development of this study are presented in sequence, whereas since the development of the test specimens, the description of apparatus and equipment used, and the methodology used in the treatment of the results.
3.1 Design and production of the test specimens For this study, specimens were developed using a board, which had previously been cast in cement and sand mortar. Two kinds of specimens were developed, each incorporating a different type of defect. In the first case (specimen 1), ceramic tiles with a defect in the adhesive mortar (lack of material in the adhesive mortar) were fitted (Figure 1-a). The voids are closed by mortar which is applied between ceramic tiles and the tiles edges. The defect can also be defined as a lack of material in the adhesive mortar which 6
simulates an anomaly in the contact between the ceramic tile and the base (detachment of facade tiles). That lack of material consists of a void space under the central area of the ceramic tiles (see Figure 1) with the dimensions of 15.5 x 4.0 cm and a thickness of 0.2 cm (thickness of the adhesive mortar). Three tiles of different thicknesses were employed (specimen 1) being called Ceramic A, Ceramic B and Ceramic C (thickness of 0.75 - 0.40 - 0.65 cm, respectively). Some important properties are given in Table 2. In the second case (specimen 2), superficial cracks of different depths were placed into the mortar board (Figure 1-b) with dimensions of 15.0 x 0.2 cm and a depth of 0.5 - 1.5 - 2.5 cm, corresponding to cracks A, B and C, respectively. The materials used to manufacture the specimens were (Figure 1): mortar (M1) consisting of cement and artificial sand for the production of the 50 x 25 x 4 cm board, adhesive mortar (M2) and ceramic tiles. The mortar properties can be observed in Table 1 and the properties of the ceramic tiles in Table 2. Emissivity values of the tiles and mortar board were determined in order to insert these data in the process of thermograms. It is not the purpose of this study to discuss how the emissivity may affect the results and variables. It is observed that the ceramic emissivity values are very close (0,80 – 0,84) and similar to those obtained in the study of Barreira et al. [22], which compared the differences between the methods for measuring emissivity.
3.2 Description of the test For this study, two different heating cycles were implemented. One was a direct heating cycle, in which the boards were heated up from the front surface, with the heated surface and the infrared camera measurement coinciding (Figure 2-a). In direct heating, the heat enters through the studied surface (coated with ceramic or cracked), simulating the situation at which the facade receives solar radiation and the heat penetrates from outside to inside of the building through the facade. The other cycle involved inverse heating, where the boards were heated from the opposite surface (from behind the mortar board), without the heated surface and the measurement from an infrared camera coinciding (Figure 2-b). In inverse heating, the heat enters through the opposite surface and leaves through the studied surface (coated with ceramic or with cracks), which
7
simulates the moment at which the heat leaves the inside of the building and moves outside. In the direct cycle, two heating devices were used (Figure 2-a), which allows the infrared camera to be placed as perpendicular to the studied board as possible, thus avoiding the problem of reflections from the heating device. In the inverse cycle, only one heating device was used (Figure 2-b). The red edge (Figure 2) corresponds to the analysed surface (coated with ceramic or cracked). In both cycles, the specimen was heated for 120 minutes. The analysed and processed results were obtained during the heating. For each experiment, the calibration parameters (emissivity, ambient temperature, relative humidity, and reflected apparent temperature) were measured and input into the infrared camera. The ambient temperature and relative humidity were measured with Meterlink MO297. The reflected apparent temperature was determined using the method based on corrugated aluminium sheet, following the procedure described in ASTM E1862 [23]. The emissivity was determined using a standard black tape, according to the ASTM 1933 [24] procedure. To capture the thermograms, the infrared camera used was the FLIR T400, temperature range of -20 to 120°C, 2% accuracy, spectral range from 7.5 to 13 μm, resolution of 320 x 240 pixels, 25° lens and IFOV of 1.36 mrad. The camera was programmed in order to obtain each termogram every 4 minutes continuously during the experiment. Thermocouples were used to determine the temperature gradient between the faces of the mortar board and to validate the temperature data. The temperature gradient was used to define the phases in which the heating takes place under different conditions throughout the test. A Pico USB TC-08 data logger was connected to a computer throughout the test to collect the readings from the thermocouples (every 2 min) (Figure 2).
3.3 Methodology of collecting and processing data Each experiment had begun with the determination of the parameters to be inserted into the infrared camera, which were: temperature, relative humidity, reflected apparent temperature and emissivity. During heating, the gradient was determined by the
8
difference, in modulus, between the thermocouple temperature on the heated face and the temperature of the opposite face (mortar board). Temperature data captured by thermocouples are automatically saved on the computer during the experiment. Thermal images are stored on an SD card to be processed. Image processing was performed using the Flir-Tools software and Flir-QuickReport, both with free version. The emissivity values are adjusted on each thermogram based on previously determined values in laboratory (Tables 1 and 2). The value of each temperature refers to the average value of the marked area in Figure 3. The analysis was performed separately for each ceramic tile and crack. In each thermogram captured of specimen 1, the following values were determined (Figure 3-a): the average temperature of the area with defects (Td), the average temperature of the neighboring area, which is of the same size but is defect free (Tnd), the average temperature of the area of each ceramic tile (A, B and C) (Ta). In each thermogram captured of specimen 2, the following values were determined (Figure 3-b): the value of the average temperature of the crack (Td), the average temperature of the surrounding area without the crack (Tnd) and the average temperature of the coating (board) were determined (Ta). From the temperatures of the thermograms, obtained during the heating cycle (direct and inverse), it was first determined the evolution of the Delta-T, the same being calculated according to Equation 3. A second analysis involved the determination of C(t) (Equation 2), now named Thermal Running Contrast (TRC) which is calculated according to Equation 4. The final analysis involved the calculation of Cs(t) (Equation 1), now named Thermal Contrast (TC), calculated according to Equation 5. The respective equations are described below.
Delta-T= Td (t) -Tnd (t)
(3)
Where Delta-T is the temperature difference, Td (t) is the temperature of the defective area at time t, and Tnd (t) is the temperature of defect free area at same time t. $- =
' ( )!*' ( ) . ( )!. ( " )
.100 (4)
Where TRC is the Thermal Running Contrast, Td (t) is temperature of the defect area at time t, Tnd (t) is the temperature of defect free area at time t, Ta (t) is the board 9
temperature (specimen 2) or tile temperature (specimen1) at time t, Ta (t0) is the board temperature (specimen 2) or tile temperature (specimen1) at time t0, and t0 is the start time of the cycle. $ =
' ( )!' ( " ) *' ( )!*' ( " )
(5)
Where TC is the Thermal Contrast, Td (t) is the temperature of the defect area at time t, Td (t0) is the temperature of defect area at time t0, Tnd (t) is the temperature of defect free area at time t, Tnd (t0) is the temperature of defect free area at time t0, and t0 is the start time of the cycle.
4. Results and discussion The results are shown and discuss the behavior observed in the gradient, the evolution of the surface temperature of the specimens (tiles, mortar), the evolution of the Delta-T and the behavior of the contrast function (TRC and TC) for both cycles. The aim is to understand the behaviors observed associating them with the field conditions, discussing the conditions for best visualization of defects.
4.1 Gradient temperature – heat phases The heating regime and the heat transport between the faces of the boards affect the identification of anomalies in the thermograms. At each reading (every 4 minutes), the temperature distribution, resulting from heat flux, is identified by thermograms. It is necessary to identify if the variations result from the heat flux variations or are derived from changes caused by the anomalies. It is necessary to study the evolution of the temperature gradient and, also the evolution of the surface temperature of the board in order to understand the different phases which occur in the test. The average temperature behavior of the studied surface and the temperature gradient during the heating cycle (direct and inverse) are shown in Figure 4 for the specimen 1, and in Figure 5 for the specimen 2. It can be seen that, in the four cases (Figures 4 and 5), the heating step can be divided into three phases. In the Phase 1, the average surface temperature and the temperature gradient rapidly increase. In the Phase 2, the average surface temperature increases and the temperature gradient is approximately constant. Finally, in the Phase 3, the average surface temperature and the temperature gradient are 10
approximately constant. The temperature gradient analysis is intended only to define the phases of heating cycle. The gradient is not used for a specific heat flux analysis or for a particular study of the defects (cracks, tile detachment).
4.2 Delta-T behavior The study of the evolution of the Delta-T is one of the first analyses of the thermographic study. The initial aim is to visualize and locate defect in the study area [12]. Figures 4 and 5 show the Delta-T results obtained during the heating stage of the study of the defects in ceramic tiles (Figure 4) and the samples with cracks (Figure 5).
4.2.1 Delta-T behavior in ceramic detachment (specimen 1) The Delta-T arises when there is a change of heat flux caused by internal defect. This modification is observed in the temperature change on the surface, which are captured in the thermograms. The dimensions, geometry, and default position, among other aspects, affect this change in temperature. In Phase 1 (direct cycle, Figure 4-a), is observed a strong growth of the Delta-T being achieved the highest values at the end of this phase, as shown in Table 3. The thermograms at 4 and 12 min (Figures 6-a, 6-b) shows the behavior of this phase observing the differences in temperature resulting from heating. Continuing the heating, in the Phase 2 is observed gradient stabilization and continuous increase in the surface temperature of the specimen 1 (Figure 4-a). Delta-T drops sharply due to the lower superficial temperature growth of the defect area (Td) during the heating (Phase 2).The increase in the system temperature leads to an increase in the internal air temperature (detachment voids), perhaps intensifying the heat transport, which allows an increase of the heat flux in the local defect area.Thus, the Delta-T between both areas decreases with time. Meanwhile, in this phase, the gradient temperature remains relatively constant (Phase 2). In the Phase 3, where the temperature gradient and the average temperature of the surface are approximately constant, Delta-T also remains constant. The values can be observed in Table 3 and Figure 4-a. In Figures 6-c and 6-d it can be seen that there is little difference in the thermograms 100 and 120 minutes, demonstrating few changes that occur at temperatures in Phases 2 and 3.
11
The behavior discussed explains many of the results obtained in different surveys, in that the greatest Delta-T values have been observed in the late morning after receiving direct radiation [16], observing the detachment of the plaster coating; Edis et al. [17], observing moisture. In such cases, the Delta-T values grows along with the average surface temperature, such that the greatest differences are observed in the late morning, in the same way as in Phase 1 of the direct heating cycle. The greatest Delta-T values were obtained for ceramic B, ceramic C, and ceramic A, respectively, i.e., the minimum ceramic thickness (Figure 1) produced the highest DeltaT values during the entire heating process (Table 3). As the thickness of the ceramic falls, the closer the defect is to the surface, such that larger Delta-T values are generated, as was observed by Cheng et al. [25]. When heat is applied from the rear of the board (inverse heating) and is then dissipated from the ceramic coated surface, the maximum values of the Delta-T module (Figure 4b) are obtained at the end of the heating cycle (Table 3). Figure 7 shows the thermograms of the inverse cycle. Because of the change in direction of heat flux, now Td is smaller than Tnd. The differences between the Delta-T in the three ceramics (Figure 4) increase over time in Phases 1 and 2, reaching a steady state only at the end of the heating period, when the average surface temperature and the gradient temperature between the faces are approximately constant (Phase 3). Note that, during the direct and inverse heating, when the average surface temperature and the gradient are approximately constant (Phase 3), Delta-T also remains constant. Similar to direct cycle, the greatest Delta-T (module) values were obtained for ceramic B, ceramic C, and ceramic A, respectively (Figure 4-b). These conditions are not usually found during facade inspections, only at night and early in the morning is when the lowest temperature variations appear. Different researchers recommend these times to perform thermographic inspections to identify some flaws, because that's when the Delta-T remains more stable [16].
4.2.2 Delta-T behavior in cracks (specimen 2) The crack is a defect which does not have confined air. The heat exchanges take place freely with the environment around. In the test, the difference in the depth of cracks is
12
significant and grows in the order of A, B, and C respectively: 0.5 cm, 1.5 cm and 2.5 cm. The width of the cracks is 0.2 cm. During the direct cycle (Figure 5-a) of the Phase 1, the Delta-T (module) exhibits strong growth (Table 4), in Phase 2 Delta-T (module) falls, and in the third stage, it remains constant with small values
(Figure 8). The Delta-T (module) is greater with the
increasing depth of the crack, i.e. it grows in the A, B, C respectively. One explanation for this may be associated with the temperature of each board layer of mortar on heating. For deepest fissures (C), the temperature in the layer at a depth of 2.5 cm is lower than the surface temperature (Phase 1, heating). Besides, the crack has a temperature which is influenced by the internal temperature of the layer (2.5 cm). As the temperature increases (Phases 2 and 3), Delta-T (modulus) decreases reaching low values at the end of the cycle (Table 4). In the inverse cycle (Figure 5-b), Delta-T
increases in the sequence A, B, C,
respectively. The Delta-T grows continuously in Phases 1 and 2, being stable in the Phase 3 (Table 4) (Figure 10). The largest Delta-T on the deepest crack can be explained by analyzing the direction of heat flux. Once the flux is inversed, the deeper the fissure, the higher the temperature in this layer. Thus, Td is larger than Tnd (Tnd is measured at the surface). This behavior suggests that infrared thermography could be used as a tool to assess the level of degradation (associated with the depth of the crack). The cracks in the laboratory study are wide (2 mm), which makes the discussions be restricted to this condition. These cracks can be detected by a visual inspection, but without an in-situ measurement it is not possible to accurately assess its importance, which may be possible, on a preliminary basis, with infrared thermography.
4.3 Contrast function behavior The study of contrast functions is intended to obtain parameters that can be applied to identify the conditions for better visualization of defects in thermograms. The TC study varies according to the direction of heat flux. For the case of no defect visualization, TC will be equal to 1. When Td > Tnd, TC will have values greater than 1. Inverting the direction of the heat flux, where Td < Tnd, TC is lower than 1. The TC will be significant if it occurs in the function a peak higher or lower than 1. 13
The TC presents the principal values at the start of Phase 1 to the specimen 1, in both direct and inverse cycles (Figures 10-a, 10-b). Apparently when the Delta-T increase, at the start of heating, is accompanied by a sharp increase in TC (direct cycle). Indeed, TC increases with increasing Delta-T rate of change in Phase 1. When this change rate decreases, the TC begins to approach 1 (Phase 2). Observing the thermograms in time of 12 min (Figure 6-b) and 16 min (Figure 7-b) is clearly the best visualization of defects in Phase 1. For the specimen 2, is observed in direct cycle that TC values are less than 1 in Phase 1 (Td < Tnd) (Figure 11-a). Clearly the lowest TC values (<1) occur in the first part of Phase 1, namely, has a high Delta-T rate of change in these early stages. In the inverse cycle, there is a peak of TC (> 1), reaching values greater than 3.0 (Table 6) in Phase 1 (Figure 11-b). The TRC basically analyzes the relation between the Delta-T and the growth of sample temperature throughout the heat cycle. This allows comparing situations more precisely where the samples were differently heated as is the case of the direct and inverse cycles. It is important to identify the moment when the peak contrast occurs. Figures 12 and 13 show the Thermal Running Contrast (TRC) results. In Tables 7 and 8 is shown the differences between the maximum values of TRC on ceramics tiles and cracks. Vavilov [19] investigating effects of noise on the TRC mentions a peak value in the TRC at the beginning of the heating cycle. It was noted that the TRC behavior is similar to the TC, with peaks at the beginning of the heating (Phase 1) and a tendency to stabilize at the end of the phase (starting from the middle of Phase 2). This indicates that both functions are the optimum values obtained during the interval time where the gradient (heat flux) grows faster. As the gradient increases by the rapid increase in surface temperature, it can be assumed that the limits of the defects appear better defined, in the intervals of time where the growth of the surface temperature is higher. The thickness of the ceramics and the depths of the cracks cause differences in the values of the two contrast functions (TC and TRC) throughout the heating stage (Figures 10, 11, 12 and 13). The smallest ceramic thickness and the greatest crack depth correspond to the case of TC (Table 5 and 6) with the values furthest from one and in the case of TRC with the highest module values (Table 7 and 8). Maldague [13] observes that for heating, the highest value of Delta-T is inversely associated with the depth of the defect and directly with the thermal resistivity. This same tendency may be 14
associated with the behavior of the TC and TRC. The two functions of contrast showed similar behavior in both types of defects studied, the peak values were obtained in the early stages of heating. After studying the Delta-T behavior and contrast functions during the three defined phases, it can be proved that, to analyse these types of defects (detachment and cracking), it is necessary to understand the behavior of the average surface temperature and the gradient temperature in order to determine the optimum moment at which to identify and characterize the defects. Note that the peak value of Delta-T (Table 3 and 4), TC (Table 5 and 6) and TRC (Table 7 and 8), as well as the moments where the functions are constant, varies according to the type of defect and direction of heat flux (from 1.05 to 3.31°C in the case of the ceramic coating and from 0.72 to 2.33°C in the case of cracks). For these reasons, it is important to conduct a preliminary study of the behavior of the facade temperatures and solar radiation before performing a thermographic inspection, it is recommended a simulation of the behavior of the heat flux in the facades. By associating the identification and evaluation of the defect with the moment at which the boundaries of the defects are more visible (the contrast value is higher) and the Delta-T values are maximum, it can be assumed that the facade inspection, for detecting ceramic detachment problems and cracks, should be done at the moment of the day at which the temperature gradient and average surface temperature exhibit the strongest growth. In those moments the values of Delta-T, TC, and TRC are highest. If the goal is to compare the defects (comparing differences of thickness or depth of cracks), it is necessary to look for the moments of maximum gradient.
5.
Conclusions
The aim of this study is to identify the best conditions for viewing and identificating detachments and cracks in thermographic inspections. Thus the experimental study, simulating detachment and cracks in heating cycles, led to the following main conclusions. The evolution of the Delta-T with heating depends on the type and also the dimensions and geometry of the detachment. Each behavior is associated with the default perturbation degree in heat flux, and how it is detected in thermograms. It was observed 15
that the thickness of the tiles directly influences the Delta-T values. The Delta-T (module) is inversely proportional to the thickness. In the case of cracks, the evolution of the Delta-T is dependent on the depth in both the direct and the inverse cycle. Also, how the surface temperature shown in thermogram (hotter or colder) it directly depends on the direction of the heat flux and is different from detachment. It is necessary to examine the other cracks behavior conditions (geometry, width, and type) in order to obtain more specific patterns in a more precisely way to assist in identifying anomalies in these inspections. For a given type of defect, Delta-T behaves differently depending on the direction of the heat flux. It is possible to detect detachment differences in the behavior associated with the thickness of the ceramic, and in the depth of the cracks by using Delta-T and the contrast functions, which will probably enable the comparison of defects of the same type in different areas of the facade. For this reason, it is very important to be aware of the direction of the heat flux during the thermographic inspection. In the studied defects, the maximum values of Delta-T and the contrast functions did not always coincide in terms of time. Accordingly, it can be concluded that the determination of the optimum inspection moment requires the analizys of the defect type and the temperature gradient behavior. The contrast function had the maximum values obtained during the interval time when the gradient grows faster. As the gradient increases by the rapid increase in surface temperature, it can be assumed that the limits of the defects appear better defined in the intervals of time where the growth of the surface temperature is higher. It is suggested and recommended to conduct a preliminary study of the behavior of the facade temperatures and solar radiation before performing a thermographic inspection, it is also recommended performing a simulation of the behavior of the heat flux in the facades.
Acknowledgements The Material Testing Laboratory of the University of Brasilia for access to inspection database. CAPES and CNPq for research support in providing grants. 16
References [1]
E. Bauer, V.P. de Freitas, N. Mustelier, E. Barreira, Infrared thermography – evaluation of the results reproducibility, Struct. Surv. 31 (2015) 181–193. doi:dx.doi.org/10.1108/SS-12-2012-0039.
[2]
E. Barreira, V.P. de Freitas, Evaluation of building materials using infrared thermography, Constr. Build. Mater. 21 (2007) 218–224. doi:http://dx.doi.org/10.1016/j.conbuildmat.2005.06.049.
[3]
F. Cerdeira, M.E. Vázquez, J. Collazo, E. Granada, Applicability of infrared thermography to the study of the behaviour of stone panels as building envelopes, Energy Build. 43 (2011) 1845–1851. doi:http://dx.doi.org/10.1016/j.enbuild.2011.03.029.
[4]
A. Frota, Geometria da Insolação, Geros Ltda, Sao Paulo, 2004.
[5]
D. Paoletti, D. Ambrosini, S. Sfarra, F. Bisegna, Preventive thermographic diagnosis of historical buildings for consolidation, J. Cult. Herit. 14 (2013) 116– 121. doi:10.1016/j.culher.2012.05.005.
[6]
M. Fox, D. Coley, S. Goodhew, P. De Wilde, Time-lapse thermography for building defect detection, Energy Build. 92 (2015) 95–106. doi:10.1016/j.enbuild.2015.01.021.
[7]
R. Usamentiaga, P. Venegas, J. Guerediaga, L. Vega, J. Molleda, F.G. Bulnes, Infrared thermography for temperature measurement and non-destructive testing., Sensors (Basel). 14 (2014) 12305–48. doi:10.3390/s140712305.
[8]
S.F. Larsen, M. Hongn, Termografía infrarroja en la edificación: aplicaciones cualitativas, Av. En Energías Renov. Y Medio Ambient. 16 (2012) 25–32.
[9]
I. Martínez, A. Castillo, E. Martínez, M. Castellote, Physico-chemical material characterization of historic unreinforced masonry buildings: The first step for a suitable intervention, Constr. Build. Mater. 40 (2013) 352–360. doi:10.1016/j.conbuildmat.2012.09.091.
[10] F. Bisegna, D. Ambrosini, D. Paoletti, S. Sfarra, F. Gugliermetti, A qualitative method for combining thermal imprints to emerging weak points of ancient wall structures by passive infrared thermography – A case study, J. Cult. Herit. 15 (2014) 199–202. doi:10.1016/j.culher.2013.03.006. [11] J.R. Kominsky, J.S. Luckino, N.H. Street, T.F. Martin, Passive Infrared Thermography — A Qualitative Method for Detecting Moisture Anomalies in Building Envelopes, Tedford Pond. (2007) 1–11. [12] E. Edis, I. Flores-Colen, J. de Brito, Quasi-quantitative infrared thermographic detection of moisture variation in facades with adhered ceramic cladding using principal component analysis, Build. Environ. 94 (2015) 97–108. doi:10.1016/j.buildenv.2015.07.027. 17
[13] X. Maldague, Theory and Practice of Infrared Technology for Nondestructive Evaluation, Wiley Series, Michigan, 2001. [14] E. Bauer, E.K. Castro, E. Pavon, A.H.S. Oliveira, Criteria for application and identification of anomalies on the facades of buildings with the use of passive infrared thermography, in: V.P. Freitas (Ed.), 1st Int. Symp. Build. Pathol., Porto, Portugal, 2015: p. 12. [15] E. Bauer, E.K. Castro, A. Hildenberg, E. Pavon, Critérios para a aplicaçao da termografia de infravermelho passiva como tecnica auxiliar ao diagnóstico de patologias em fachadas de edifícios, Rev. Politec. (Instituto Politec. Bahia). 26 (2014) 266–277. [16] S.S. de Freitas, V.P. de Freitas, E. Barreira, Detection of façade plaster detachments using infrared thermography – A nondestructive technique, Constr. Build. Mater. 70 (2014) 80–87. doi:10.1016/j.conbuildmat.2014.07.094. [17] E. Edis, I. Flores-Colen, J. De Brito, Passive thermographic detection of moisture problems in façades with adhered ceramic cladding, Constr. Build. Mater. 51 (2014) 187–197. doi:10.1016/j.conbuildmat.2013.10.085. [18] E. Edis, I. Flores-Colen, J. de Brito, Passive Thermographic Inspection of Adhered Ceramic Claddings: Limitations and Conditioning Factors, J. Perform. Constr. Facil. (2012) 258. doi:10.1061/(ASCE)CF.1943-5509.0000365. [19] V. Vavilov, Noise-limited thermal/infrared nondestructive testing, NDT E Int. 61 (2014) 16–23. doi:10.1016/j.ndteint.2013.09.002. [20] W.L. Lai, S.C. Kou, C.S. Poon, W.F. Tsang, C.C. Lai, Characterization of the deterioration of externally bonded CFRP-concrete composites using quantitative infrared thermography, Cem. Concr. Compos. 32 (2010) 740–746. doi:10.1016/j.cemconcomp.2010.03.008. [21] H. Nowak, M. Kucypera, Application of Active Thermography for Detecting Material Defects in the Building Envelope, in: Inframtion Proc., 2010: pp. 1–12. [22] E. Barreira, E. Bauer, N. Mustelier, V.P. de Freitas, Measurement of materials emissivity – Influence of the procedure, in: 13th Int. Work. Adv. Infrared Technol. Appl., Pisa, Italy, 2015. [23] ASTM E1862-97 (2010), Standard Test Methods for Measuring and Compensating for Reflected Temperature Using Infrared Imaging Radiometers, West Conshohocken, Pennsylvania, United States, 2010. doi:10.1520/E186297R10.2. [24] ASTM 1933-99, Standard Test Methods for Measuring and Compensating for Emissivity Using Infrared, West Conshohocken, Pennsylvania, United States, 1999.
18
[25] C.-C. Cheng, T.-M. Cheng, C.-H. Chiang, Defect detection of concrete structures using both infrared thermography and elastic waves, Autom. Constr. 18 (2008) 87–92. doi:10.1016/j.autcon.2008.05.004.
Fig. 1. Manufactured specimens: (a) mortar board with ceramic tiles and the defect in the adhesive mortar (specimen 1); (b) mortar board with superficial cracks (specimen 2). Fig. 2. Distribution of equipment for direct heating cycle (a) and inverse heating cycle (b). Fig. 3. Areas used in the determination of parameters thermographic: (Td) average temperature of the area with defects; (Tnd) average temperature of the defect free area; (Ta) average temperature of total area. Fig. 4. Average temperature of surface, gradient temperature and Delta-T in Specimen 1 during direct heating cycle (a) and inverse heating cycle (b). Fig. 5. Average temperature of surface, gradient temperature and Delta-T in Specimen 2 during direct heating cycle (a) and inverse heating cycle (b). Fig. 6. Thermograms of the specimen 1 during the direct heating cycle (detachment). Fig. 7. Thermograms of the specimen 1 during the inverse heating cycle (detachment). Fig. 8. Thermograms of the specimen 2 during the direct heating cycle (cracks). Fig. 9. Thermograms of the specimen 2 during the inverse heating cycle (cracks). Fig. 10. TC behaviour for Specimen 1 during direct heating cycle (a) and inverse heating cycle (b). Fig. 11. TC behaviour for Specimen 2 during direct heating cycle (a) and inverse heating cycle (b). Fig. 12. TRC behaviour for Specimen 1 during direct heating cycle (a) and inverse heating cycle (b). Fig. 13. TRC behaviour for Specimen 2 during direct heating cycle (a) and inverse heating cycle (b).
19
Table. 1. Properties of mortars used in the specimens. Materials Mortar (M1) Adhesive mortar (M2)
Compressive strength 14MPa 6MPa
Density
Emissivity
1690kg/m3
Water absorption for capillarity 2.4kg/m².min1/2
1600kg/m3
3.4kg/m².min1/2
---
0.91
Table. 2. Properties of ceramic tiles used in the specimens. Materials Ceramic A Ceramic B Ceramic C
Density 2200 kg/m3 1700 kg/m3 1800 kg/m3
Water absorption 1.4% 2.2% 1.8%
Emissivity 0.84 0.80 0.82
Table. 3 Behavior of Delta-T in the specimen 1. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (°C) (°C) Ceramic A 1.05 1 0.10 3 Direct Ceramic B 2.72 1 0.82 3 cycle Ceramic C 1.58 1 0.58 3 Ceramic A -1.88 3 -1.88 3 Inverse Ceramic B -3.31 3 -3.31 3 cycle Ceramic C -2.41 3 -2.41 3
Table. 4 Behavior of Delta-T in the specimen 2. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (°C) (°C) Crack A -1.09 1 -0.02 3 Direct Crack B -1.30 1 -0.26 3 cycle Crack C -1.70 1 -0.47 3 Crack A 0.72 3 0.72 3 Inverse Crack B 1.15 3 1.15 3 cycle Crack C 2.33 3 2.33 3
Table. 5 Behavior of TC in the specimen 1. Peak values and values in the constant range. Peak Phase Constant Phase value observed range observed 20
Direct cycle Inverse cycle
Ceramic A Ceramic B Ceramic C Ceramic A Ceramic B Ceramic C
1.28 1.66 1.32 0.73 0.70 0.58
1 1 1 1 1 1
1.00 1.04 1.02 0.91 0.87 0.85
3 3 3 3 3 3
Table. 6 Behavior of TC in the specimen 2. Peak values and values in the constant range. Peak Phase Constant Phase value observed range observed Crack A 0.63 1 1.00 3 Direct Crack B 0.54 1 0.99 3 cycle Crack C 0.38 1 0.98 3 Crack A 2.33 1 1.04 3 Inverse Crack B 2.83 1 1.05 3 cycle Crack C 3.40 1 1.08 3
Table. 7 Behavior of TRC in the specimen 1. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (%) (%) Ceramic A 21 1 0.44 3 Direct Ceramic B 57 1 3.30 3 cycle Ceramic C 29 1 2.29 3 Ceramic A -26 1 -7.57 3 Inverse Ceramic B -39 1 -13.34 3 cycle Ceramic C -34 1 -9.71 3
Table. 8 Behavior of TRC in the specimen 2. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (%) (%) Crack A -36 1 -0.12 3 Direct Crack B -46 1 -0.99 3 cycle Crack C -63 1 -1.89 3 Crack A 42 1 2.84 3 Inverse Crack B 93 1 4.54 3 cycle Crack C 245 1 9.19 3
21
HIGHLIGHTS Criteria for identifying detachments and cracks with infrared thermography inspections. Analysis of Delta-T and contrast functions to study facade defect (detachment and crack). Influence of surface temperature
and gradient temperatures behaviors on defect
visibility. Differentiated Delta-T behavior according to defect type and heat flux direction. Use of direct and inverse heat flux in the assessment of detachments and cracks with infrared thermography
22
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
PII: DOI: Reference:
S2352-7102(16)30021-3 http://dx.doi.org/10.1016/j.jobe.2016.02.012 JOBE105
To appear in: Journal of Building Engineering Received date: 27 August 2015 Revised date: 5 February 2016 Accepted date: 25 February 2016 Cite this article as: Elton Bauer, Elier Pavón, Eva Barreira and Eliane Kraus, ANALYSIS OF BUILDING FACADE DEFECTS USING INFRARED THERMOGRAPHY: LABORATORY STUDIES, Journal of Building Engineering, http://dx.doi.org/10.1016/j.jobe.2016.02.012 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ANALYSIS OF BUILDING FACADE DEFECTS USING INFRARED THERMOGRAPHY: LABORATORY STUDIES Elton Bauer1*; Elier Pavón1; Eva Barreira2;Eliane Kraus1 1
Post-Graduate Program in Structures and Civil Construction, Civil and Environmental
Engineering Department, University of Brasília - Campus Universitário Darcy Ribeiro Asa Norte, 70910-900 Brasília/DF, Brazil 2
Faculty of Engineering, Dept. of Civil Engineering, Univ. of Porto, Laboratory of
Building Physics, Rua Dr. Roberto Frias s/n, 4200-465 Porto, Portugal *
Corresponding author. [email protected]
Abstract One of the main problems facing the inspection of facades by means of passive infrared thermography, in order to identify and evaluate defects or pathologies, is the definition of the most appropriate moment to perform the inspection. Most studies focus on the Delta-T value as a criterion for the identification and evaluation of defects. Nevertheless, the identification of defects will primarily depend on the heat flux, since the way defects appear on the thermogram is determined by their type and by the direction of the flux. This study set out to analyse this problem by studying the behavior of the Delta-T and contrast functions in defects in ceramic tiles and mortar. Accordingly, test samples were made with induced defects during the manufacturing process, to simulate the detachment problems that occur with ceramic tiles of different thicknesses and cracks of different depths. The test samples were assessed during direct and inverse heating cycles. The results showed that it is possible to detect differences in behavior when varying the thickness of the ceramic tiles and the depth of the cracks by considering the Delta-T and contrast functions. In consequence, the inspection moment should be defined according to the behavior of the gradient temperature, the type of the defect and the direction of the heat flux.
Keywords: Infrared thermography; Façade; Delta-T; Thermal contrast; Temperature gradient; Pathologies 1
1. Introduction Infrared thermography, involving an analysis of surface temperatures, is able to identify internal anomalies since heat transportation is affected by the presence of faults, inclusions, moisture, and other occurrences. These anomalies change the heat flux pattern and thus affect the resulting surface temperature, relative to the surface temperature of a defect free surface [1]. Although internal defects can be detected, infrared thermography is considered to be applicable to surfaces since it more easily identifies anomalies close to the surface. The possibility of identifying anomalies will not depend only on the characteristics and depth of the defects. Elements or components, in thermal or hygroscopic equilibrium are difficult to study with thermography [2]. The thermal flux in plan elements is obviously not one-dimensional. However, the thickness being smaller, it is observed that in this direction occur the fastest answers of the alteration of the temperatures. That way it is possible to study these heat flux in a simplified form which permits the identification of defects and anomalies. The gradient temperature between the external and internal surfaces (facade), as well as the direction of the heat flux that passes through the wall, will determine when a defect will be visible and how it can be identified in a thermogram. It can be characterized as being either a cooler or warmer zone. The heat flux in a building envelope obviously is neither static nor constant throughout the day, and is closely dependent on the ambient temperature, solar heating, and cooling of the facade. The performing of a thermographic inspection for delamination or detachment detection it is often recommended at the moment of the day when the values of the heat flux in the facade are greater, that is, when the greatest wall temperature gradient values appear (this study was conducted in stone panels) [3]. A study of the solar trajectory in order to identify the potential effects of sunlight and shade, which will define the heat flux and, consequently, the conditions for the detection of anomalies, is of great importance. The use of sun charts for this assessment is simple and effective but must be completed with priority to the inspection [4]. There are different conditions to consider when identifying anomalies that depend on orientation, making it necessary to study both schemes, direct and inverse heat flux, in order to define the analysis criteria for each inspection. In such cases, it is even more 2
important to conduct a pre-analysis of the heat flux schemes. The aim is to understand the heat flux conditions in the facade, and the behavior of defects in different directions of flux. Applying thermography, it is possible to identify and map the areas with defects. The forms in which have been analysed the defects in facade elements (ceramic tiles and mortar) in laboratory studies are presented in the next items.
2. Theoretical background Different forms of analysis are used to identify anomalies with infrared thermography. The considered to be the most important are the visual temperature differences in the thermogram (used in qualitative thermography) and the Delta-T (used in quantitative thermography). Qualitative thermography is based on a relatively simple analysis to identify hot spots and cold spots of the thermal image by color difference, that is, a visual analysis. The analysis is made by a comparison with standard situations, and the evaluation depends heavily on the expertise of the evaluator [5]. Quantitative thermography is used to classify the importance of an anomaly, being necessary to obtain as accurately as possible the temperature of the target object. It is necessary then measure and appropriately define the thermographic parameters (eg emissivity, reflected temperature). Graphical representations and contrast functions have also been used in other studies as complementary tools for the identification or quantification of defects [6,7]. The thermogram visual differences have been used to enable the identification of cold and hot areas in building facades, thermal bridges, presence of structural elements, like beams and pillars, different materials like mortar and brick, failures in the insulating materials [8], as well as common pathologies like cracks [9,10] and moisture [11,12]. The most commonly used criterion in quantitative thermography is associated with the Delta-T between the area with defects and the defect free surrounding area [13]. However, the use of this criterion does not lead to a conclusive assessment when identifying anomalies in facades [6,14]. It is necessary to consider the gradient temperature and to determine the conditions under which anomalies in the thermograms can be better identified [15]. The study developed by Bauer et al. [1] proved that it is possible to quantify the defects created in a sample, thus exhibiting excellent reproducibility in the three cycles that were evaluated. Despite the differences in the 3
temperatures recorded by the two different thermographic cameras in the study, the Delta-T, quantitative criterion used to evaluate the defect, showed very small differences between the two cameras, proving that it can be used for the evaluation of damage and defects in facades as long as the variations in the temperature gradient are considered in the analysis process. Studies conducted during the day to evaluate defects in facades have been carried out considering the influence of the variation of exterior and interior air temperatures. De Freitas et al. [16] studied polymeric plaster detachment on a facade, obtaining different Delta-T values throughout the day. They concluded that the best period for the inspection was during the hours of the greatest exposure to sunlight (at noon). Edis et al. [17] detected moisture problems from the heat gain of the facade during the daytime and nighttime. These authors recommend carrying out the inspection during the night because the Delta-T, between the moist area (with defects) and dry area (without defects), is stable for a longer period. In this study, the magnitude of Delta-T was also higher around noon. One of the main problems facing the thermographic inspections of facades is when to get the thermal imaging [18]. Thermographic inspections are normally based on finding the thermogram with the highest Delta-T (the defect will always be more visible when the value of the Delta-T is the highest). However, in these situations, the physical limits of the defects are not always well identified, hindering the precise delimitation of the anomalies. Delta-T not only depends on the temperature gradient, as the type of the defect, its size, and the depth at which it exists can modify the Delta-T’s behavior [13]. In studies of small thickness defects, located at a shallow depth relative to the surface, the value of Delta-T tends to be constant after the initial heating moments at the laboratory study, as occurred in the studies developed by Bauer et al. [1] and Freitas et al. [16]. The identification of the optimum time (moment) to perform a thermographic inspection to identify and evaluate defects is a problem which has been studied in cases of active and passive thermography, when a specific heat flux is induced in a sample. In addition to a visual and Delta-T analysis, in active thermography, contrast functions are used as an auxiliary tool for defect identification. In some specific material studies, the value of the contrast function can indicates the presence of the defect [19]. However, in studies of building elements in which the defects are larger, it is expected that this function can 4
be used to determine the optimal time to identify the defects. The thermal events that occur within, an area incorporating defects, and a defect free normal area, change with time because the heat dissipation varies with time, among other variables. Thus, the selection of the most adequate thermogram , at a particular moment in time, is required in order to extract and define the boundaries of the local defects. The selection criteria for this thermogram is generally based on a thermogram with the highest thermal contrast. This contrast can corresponds to the differential temperature of the defects over the differential temperature from a free defect area [20]. Maldague [13] used the contrast function called ‘Standard Contrast’ (Cs), which is calculated according to Equation 1. The contrast is the temperature variation in the area with defects, relative to the evolution of the temperature of the closest defect free area.
() =
( )! ( " ) # ( )!# ( " )
(1)
Where Cs(t) is the ‘Standard Contrast’, $% () is the temperature of the defective area at time t, $% (& ) is the temperature of the defective area at time t0, $ () is the temperature of the defect free area at time t, $ (& ) is the temperature of the defect free area at time t0, and & is the start time (beginning of the cycle). A similar contrast function (standard temperature contrast) was used by Nowak and Kucypera [21] to study the presence of concealed materials inside the walls in laboratory. This function contrast is the temperature variation at any selected point on surface of the material under testing relative to the temperature variation at surface point over homogeneous area prior to thermal stimulation. According to Nowak and Kucypera [21], the ‘Standard Contrast’ (Cs) is independent of the type of material being tested; it is dimensionless and oscillates around a constant value close to one in the steady state, such that it is possible to compare the results of various experiments, as was done by Bauer et al. [14]. Vavilov [19] used a contrast function called ‘Running Temperature Contrast’ (C) in the study of sensitivity and noise in thermographic measurements. Basically, this contrast function is obtained from Equation 2, being determined during the periods of heating. These values are generated for each time and correspond to the evolution of the Delta-T (temperature of the region incorporating defects minus that of the defect free region), in 5
relation to the evolution of the average temperature of the sample. With this evaluation, it should be possible to identify the optimum times at which to identify anomalies.
() =
' ( )!*' ( ) ( )
(2)
Where () is the ‘Running Temperature Contrast’, $+ () is the temperature of the defective area at time t, $,+ () is the temperature of the non-defective area at time t, and T(t) is the sample excess temperature (evolution of the sample temperature during heating phase). Bauer et al. [14] employed this modified function to analyse the behavior of defects in ceramic tiles by finding close values in the heating and cooling cycles. The objective is to discuss, based on laboratory studies, which the conditions that allow better visualization and identification of anomalies in facade inspections, using passive thermography. Thus, this paper discusses the problem of selecting the optimum time at which to capture a thermal image during a thermographic inspection, based on a Delta-T behavior analysis and Cs and C contrast functions for both of the most common problems addressed by facade studies: the detachment of ceramic tiles and cracks in rendering mortar.
3. Experimental development The experimental development of this study are presented in sequence, whereas since the development of the test specimens, the description of apparatus and equipment used, and the methodology used in the treatment of the results.
3.1 Design and production of the test specimens For this study, specimens were developed using a board, which had previously been cast in cement and sand mortar. Two kinds of specimens were developed, each incorporating a different type of defect. In the first case (specimen 1), ceramic tiles with a defect in the adhesive mortar (lack of material in the adhesive mortar) were fitted (Figure 1-a). The voids are closed by mortar which is applied between ceramic tiles and the tiles edges. The defect can also be defined as a lack of material in the adhesive mortar which 6
simulates an anomaly in the contact between the ceramic tile and the base (detachment of facade tiles). That lack of material consists of a void space under the central area of the ceramic tiles (see Figure 1) with the dimensions of 15.5 x 4.0 cm and a thickness of 0.2 cm (thickness of the adhesive mortar). Three tiles of different thicknesses were employed (specimen 1) being called Ceramic A, Ceramic B and Ceramic C (thickness of 0.75 - 0.40 - 0.65 cm, respectively). Some important properties are given in Table 2. In the second case (specimen 2), superficial cracks of different depths were placed into the mortar board (Figure 1-b) with dimensions of 15.0 x 0.2 cm and a depth of 0.5 - 1.5 - 2.5 cm, corresponding to cracks A, B and C, respectively. The materials used to manufacture the specimens were (Figure 1): mortar (M1) consisting of cement and artificial sand for the production of the 50 x 25 x 4 cm board, adhesive mortar (M2) and ceramic tiles. The mortar properties can be observed in Table 1 and the properties of the ceramic tiles in Table 2. Emissivity values of the tiles and mortar board were determined in order to insert these data in the process of thermograms. It is not the purpose of this study to discuss how the emissivity may affect the results and variables. It is observed that the ceramic emissivity values are very close (0,80 – 0,84) and similar to those obtained in the study of Barreira et al. [22], which compared the differences between the methods for measuring emissivity.
3.2 Description of the test For this study, two different heating cycles were implemented. One was a direct heating cycle, in which the boards were heated up from the front surface, with the heated surface and the infrared camera measurement coinciding (Figure 2-a). In direct heating, the heat enters through the studied surface (coated with ceramic or cracked), simulating the situation at which the facade receives solar radiation and the heat penetrates from outside to inside of the building through the facade. The other cycle involved inverse heating, where the boards were heated from the opposite surface (from behind the mortar board), without the heated surface and the measurement from an infrared camera coinciding (Figure 2-b). In inverse heating, the heat enters through the opposite surface and leaves through the studied surface (coated with ceramic or with cracks), which
7
simulates the moment at which the heat leaves the inside of the building and moves outside. In the direct cycle, two heating devices were used (Figure 2-a), which allows the infrared camera to be placed as perpendicular to the studied board as possible, thus avoiding the problem of reflections from the heating device. In the inverse cycle, only one heating device was used (Figure 2-b). The red edge (Figure 2) corresponds to the analysed surface (coated with ceramic or cracked). In both cycles, the specimen was heated for 120 minutes. The analysed and processed results were obtained during the heating. For each experiment, the calibration parameters (emissivity, ambient temperature, relative humidity, and reflected apparent temperature) were measured and input into the infrared camera. The ambient temperature and relative humidity were measured with Meterlink MO297. The reflected apparent temperature was determined using the method based on corrugated aluminium sheet, following the procedure described in ASTM E1862 [23]. The emissivity was determined using a standard black tape, according to the ASTM 1933 [24] procedure. To capture the thermograms, the infrared camera used was the FLIR T400, temperature range of -20 to 120°C, 2% accuracy, spectral range from 7.5 to 13 μm, resolution of 320 x 240 pixels, 25° lens and IFOV of 1.36 mrad. The camera was programmed in order to obtain each termogram every 4 minutes continuously during the experiment. Thermocouples were used to determine the temperature gradient between the faces of the mortar board and to validate the temperature data. The temperature gradient was used to define the phases in which the heating takes place under different conditions throughout the test. A Pico USB TC-08 data logger was connected to a computer throughout the test to collect the readings from the thermocouples (every 2 min) (Figure 2).
3.3 Methodology of collecting and processing data Each experiment had begun with the determination of the parameters to be inserted into the infrared camera, which were: temperature, relative humidity, reflected apparent temperature and emissivity. During heating, the gradient was determined by the
8
difference, in modulus, between the thermocouple temperature on the heated face and the temperature of the opposite face (mortar board). Temperature data captured by thermocouples are automatically saved on the computer during the experiment. Thermal images are stored on an SD card to be processed. Image processing was performed using the Flir-Tools software and Flir-QuickReport, both with free version. The emissivity values are adjusted on each thermogram based on previously determined values in laboratory (Tables 1 and 2). The value of each temperature refers to the average value of the marked area in Figure 3. The analysis was performed separately for each ceramic tile and crack. In each thermogram captured of specimen 1, the following values were determined (Figure 3-a): the average temperature of the area with defects (Td), the average temperature of the neighboring area, which is of the same size but is defect free (Tnd), the average temperature of the area of each ceramic tile (A, B and C) (Ta). In each thermogram captured of specimen 2, the following values were determined (Figure 3-b): the value of the average temperature of the crack (Td), the average temperature of the surrounding area without the crack (Tnd) and the average temperature of the coating (board) were determined (Ta). From the temperatures of the thermograms, obtained during the heating cycle (direct and inverse), it was first determined the evolution of the Delta-T, the same being calculated according to Equation 3. A second analysis involved the determination of C(t) (Equation 2), now named Thermal Running Contrast (TRC) which is calculated according to Equation 4. The final analysis involved the calculation of Cs(t) (Equation 1), now named Thermal Contrast (TC), calculated according to Equation 5. The respective equations are described below.
Delta-T= Td (t) -Tnd (t)
(3)
Where Delta-T is the temperature difference, Td (t) is the temperature of the defective area at time t, and Tnd (t) is the temperature of defect free area at same time t. $- =
' ( )!*' ( ) . ( )!. ( " )
.100 (4)
Where TRC is the Thermal Running Contrast, Td (t) is temperature of the defect area at time t, Tnd (t) is the temperature of defect free area at time t, Ta (t) is the board 9
temperature (specimen 2) or tile temperature (specimen1) at time t, Ta (t0) is the board temperature (specimen 2) or tile temperature (specimen1) at time t0, and t0 is the start time of the cycle. $ =
' ( )!' ( " ) *' ( )!*' ( " )
(5)
Where TC is the Thermal Contrast, Td (t) is the temperature of the defect area at time t, Td (t0) is the temperature of defect area at time t0, Tnd (t) is the temperature of defect free area at time t, Tnd (t0) is the temperature of defect free area at time t0, and t0 is the start time of the cycle.
4. Results and discussion The results are shown and discuss the behavior observed in the gradient, the evolution of the surface temperature of the specimens (tiles, mortar), the evolution of the Delta-T and the behavior of the contrast function (TRC and TC) for both cycles. The aim is to understand the behaviors observed associating them with the field conditions, discussing the conditions for best visualization of defects.
4.1 Gradient temperature – heat phases The heating regime and the heat transport between the faces of the boards affect the identification of anomalies in the thermograms. At each reading (every 4 minutes), the temperature distribution, resulting from heat flux, is identified by thermograms. It is necessary to identify if the variations result from the heat flux variations or are derived from changes caused by the anomalies. It is necessary to study the evolution of the temperature gradient and, also the evolution of the surface temperature of the board in order to understand the different phases which occur in the test. The average temperature behavior of the studied surface and the temperature gradient during the heating cycle (direct and inverse) are shown in Figure 4 for the specimen 1, and in Figure 5 for the specimen 2. It can be seen that, in the four cases (Figures 4 and 5), the heating step can be divided into three phases. In the Phase 1, the average surface temperature and the temperature gradient rapidly increase. In the Phase 2, the average surface temperature increases and the temperature gradient is approximately constant. Finally, in the Phase 3, the average surface temperature and the temperature gradient are 10
approximately constant. The temperature gradient analysis is intended only to define the phases of heating cycle. The gradient is not used for a specific heat flux analysis or for a particular study of the defects (cracks, tile detachment).
4.2 Delta-T behavior The study of the evolution of the Delta-T is one of the first analyses of the thermographic study. The initial aim is to visualize and locate defect in the study area [12]. Figures 4 and 5 show the Delta-T results obtained during the heating stage of the study of the defects in ceramic tiles (Figure 4) and the samples with cracks (Figure 5).
4.2.1 Delta-T behavior in ceramic detachment (specimen 1) The Delta-T arises when there is a change of heat flux caused by internal defect. This modification is observed in the temperature change on the surface, which are captured in the thermograms. The dimensions, geometry, and default position, among other aspects, affect this change in temperature. In Phase 1 (direct cycle, Figure 4-a), is observed a strong growth of the Delta-T being achieved the highest values at the end of this phase, as shown in Table 3. The thermograms at 4 and 12 min (Figures 6-a, 6-b) shows the behavior of this phase observing the differences in temperature resulting from heating. Continuing the heating, in the Phase 2 is observed gradient stabilization and continuous increase in the surface temperature of the specimen 1 (Figure 4-a). Delta-T drops sharply due to the lower superficial temperature growth of the defect area (Td) during the heating (Phase 2).The increase in the system temperature leads to an increase in the internal air temperature (detachment voids), perhaps intensifying the heat transport, which allows an increase of the heat flux in the local defect area.Thus, the Delta-T between both areas decreases with time. Meanwhile, in this phase, the gradient temperature remains relatively constant (Phase 2). In the Phase 3, where the temperature gradient and the average temperature of the surface are approximately constant, Delta-T also remains constant. The values can be observed in Table 3 and Figure 4-a. In Figures 6-c and 6-d it can be seen that there is little difference in the thermograms 100 and 120 minutes, demonstrating few changes that occur at temperatures in Phases 2 and 3.
11
The behavior discussed explains many of the results obtained in different surveys, in that the greatest Delta-T values have been observed in the late morning after receiving direct radiation [16], observing the detachment of the plaster coating; Edis et al. [17], observing moisture. In such cases, the Delta-T values grows along with the average surface temperature, such that the greatest differences are observed in the late morning, in the same way as in Phase 1 of the direct heating cycle. The greatest Delta-T values were obtained for ceramic B, ceramic C, and ceramic A, respectively, i.e., the minimum ceramic thickness (Figure 1) produced the highest DeltaT values during the entire heating process (Table 3). As the thickness of the ceramic falls, the closer the defect is to the surface, such that larger Delta-T values are generated, as was observed by Cheng et al. [25]. When heat is applied from the rear of the board (inverse heating) and is then dissipated from the ceramic coated surface, the maximum values of the Delta-T module (Figure 4b) are obtained at the end of the heating cycle (Table 3). Figure 7 shows the thermograms of the inverse cycle. Because of the change in direction of heat flux, now Td is smaller than Tnd. The differences between the Delta-T in the three ceramics (Figure 4) increase over time in Phases 1 and 2, reaching a steady state only at the end of the heating period, when the average surface temperature and the gradient temperature between the faces are approximately constant (Phase 3). Note that, during the direct and inverse heating, when the average surface temperature and the gradient are approximately constant (Phase 3), Delta-T also remains constant. Similar to direct cycle, the greatest Delta-T (module) values were obtained for ceramic B, ceramic C, and ceramic A, respectively (Figure 4-b). These conditions are not usually found during facade inspections, only at night and early in the morning is when the lowest temperature variations appear. Different researchers recommend these times to perform thermographic inspections to identify some flaws, because that's when the Delta-T remains more stable [16].
4.2.2 Delta-T behavior in cracks (specimen 2) The crack is a defect which does not have confined air. The heat exchanges take place freely with the environment around. In the test, the difference in the depth of cracks is
12
significant and grows in the order of A, B, and C respectively: 0.5 cm, 1.5 cm and 2.5 cm. The width of the cracks is 0.2 cm. During the direct cycle (Figure 5-a) of the Phase 1, the Delta-T (module) exhibits strong growth (Table 4), in Phase 2 Delta-T (module) falls, and in the third stage, it remains constant with small values
(Figure 8). The Delta-T (module) is greater with the
increasing depth of the crack, i.e. it grows in the A, B, C respectively. One explanation for this may be associated with the temperature of each board layer of mortar on heating. For deepest fissures (C), the temperature in the layer at a depth of 2.5 cm is lower than the surface temperature (Phase 1, heating). Besides, the crack has a temperature which is influenced by the internal temperature of the layer (2.5 cm). As the temperature increases (Phases 2 and 3), Delta-T (modulus) decreases reaching low values at the end of the cycle (Table 4). In the inverse cycle (Figure 5-b), Delta-T
increases in the sequence A, B, C,
respectively. The Delta-T grows continuously in Phases 1 and 2, being stable in the Phase 3 (Table 4) (Figure 10). The largest Delta-T on the deepest crack can be explained by analyzing the direction of heat flux. Once the flux is inversed, the deeper the fissure, the higher the temperature in this layer. Thus, Td is larger than Tnd (Tnd is measured at the surface). This behavior suggests that infrared thermography could be used as a tool to assess the level of degradation (associated with the depth of the crack). The cracks in the laboratory study are wide (2 mm), which makes the discussions be restricted to this condition. These cracks can be detected by a visual inspection, but without an in-situ measurement it is not possible to accurately assess its importance, which may be possible, on a preliminary basis, with infrared thermography.
4.3 Contrast function behavior The study of contrast functions is intended to obtain parameters that can be applied to identify the conditions for better visualization of defects in thermograms. The TC study varies according to the direction of heat flux. For the case of no defect visualization, TC will be equal to 1. When Td > Tnd, TC will have values greater than 1. Inverting the direction of the heat flux, where Td < Tnd, TC is lower than 1. The TC will be significant if it occurs in the function a peak higher or lower than 1. 13
The TC presents the principal values at the start of Phase 1 to the specimen 1, in both direct and inverse cycles (Figures 10-a, 10-b). Apparently when the Delta-T increase, at the start of heating, is accompanied by a sharp increase in TC (direct cycle). Indeed, TC increases with increasing Delta-T rate of change in Phase 1. When this change rate decreases, the TC begins to approach 1 (Phase 2). Observing the thermograms in time of 12 min (Figure 6-b) and 16 min (Figure 7-b) is clearly the best visualization of defects in Phase 1. For the specimen 2, is observed in direct cycle that TC values are less than 1 in Phase 1 (Td < Tnd) (Figure 11-a). Clearly the lowest TC values (<1) occur in the first part of Phase 1, namely, has a high Delta-T rate of change in these early stages. In the inverse cycle, there is a peak of TC (> 1), reaching values greater than 3.0 (Table 6) in Phase 1 (Figure 11-b). The TRC basically analyzes the relation between the Delta-T and the growth of sample temperature throughout the heat cycle. This allows comparing situations more precisely where the samples were differently heated as is the case of the direct and inverse cycles. It is important to identify the moment when the peak contrast occurs. Figures 12 and 13 show the Thermal Running Contrast (TRC) results. In Tables 7 and 8 is shown the differences between the maximum values of TRC on ceramics tiles and cracks. Vavilov [19] investigating effects of noise on the TRC mentions a peak value in the TRC at the beginning of the heating cycle. It was noted that the TRC behavior is similar to the TC, with peaks at the beginning of the heating (Phase 1) and a tendency to stabilize at the end of the phase (starting from the middle of Phase 2). This indicates that both functions are the optimum values obtained during the interval time where the gradient (heat flux) grows faster. As the gradient increases by the rapid increase in surface temperature, it can be assumed that the limits of the defects appear better defined, in the intervals of time where the growth of the surface temperature is higher. The thickness of the ceramics and the depths of the cracks cause differences in the values of the two contrast functions (TC and TRC) throughout the heating stage (Figures 10, 11, 12 and 13). The smallest ceramic thickness and the greatest crack depth correspond to the case of TC (Table 5 and 6) with the values furthest from one and in the case of TRC with the highest module values (Table 7 and 8). Maldague [13] observes that for heating, the highest value of Delta-T is inversely associated with the depth of the defect and directly with the thermal resistivity. This same tendency may be 14
associated with the behavior of the TC and TRC. The two functions of contrast showed similar behavior in both types of defects studied, the peak values were obtained in the early stages of heating. After studying the Delta-T behavior and contrast functions during the three defined phases, it can be proved that, to analyse these types of defects (detachment and cracking), it is necessary to understand the behavior of the average surface temperature and the gradient temperature in order to determine the optimum moment at which to identify and characterize the defects. Note that the peak value of Delta-T (Table 3 and 4), TC (Table 5 and 6) and TRC (Table 7 and 8), as well as the moments where the functions are constant, varies according to the type of defect and direction of heat flux (from 1.05 to 3.31°C in the case of the ceramic coating and from 0.72 to 2.33°C in the case of cracks). For these reasons, it is important to conduct a preliminary study of the behavior of the facade temperatures and solar radiation before performing a thermographic inspection, it is recommended a simulation of the behavior of the heat flux in the facades. By associating the identification and evaluation of the defect with the moment at which the boundaries of the defects are more visible (the contrast value is higher) and the Delta-T values are maximum, it can be assumed that the facade inspection, for detecting ceramic detachment problems and cracks, should be done at the moment of the day at which the temperature gradient and average surface temperature exhibit the strongest growth. In those moments the values of Delta-T, TC, and TRC are highest. If the goal is to compare the defects (comparing differences of thickness or depth of cracks), it is necessary to look for the moments of maximum gradient.
5.
Conclusions
The aim of this study is to identify the best conditions for viewing and identificating detachments and cracks in thermographic inspections. Thus the experimental study, simulating detachment and cracks in heating cycles, led to the following main conclusions. The evolution of the Delta-T with heating depends on the type and also the dimensions and geometry of the detachment. Each behavior is associated with the default perturbation degree in heat flux, and how it is detected in thermograms. It was observed 15
that the thickness of the tiles directly influences the Delta-T values. The Delta-T (module) is inversely proportional to the thickness. In the case of cracks, the evolution of the Delta-T is dependent on the depth in both the direct and the inverse cycle. Also, how the surface temperature shown in thermogram (hotter or colder) it directly depends on the direction of the heat flux and is different from detachment. It is necessary to examine the other cracks behavior conditions (geometry, width, and type) in order to obtain more specific patterns in a more precisely way to assist in identifying anomalies in these inspections. For a given type of defect, Delta-T behaves differently depending on the direction of the heat flux. It is possible to detect detachment differences in the behavior associated with the thickness of the ceramic, and in the depth of the cracks by using Delta-T and the contrast functions, which will probably enable the comparison of defects of the same type in different areas of the facade. For this reason, it is very important to be aware of the direction of the heat flux during the thermographic inspection. In the studied defects, the maximum values of Delta-T and the contrast functions did not always coincide in terms of time. Accordingly, it can be concluded that the determination of the optimum inspection moment requires the analizys of the defect type and the temperature gradient behavior. The contrast function had the maximum values obtained during the interval time when the gradient grows faster. As the gradient increases by the rapid increase in surface temperature, it can be assumed that the limits of the defects appear better defined in the intervals of time where the growth of the surface temperature is higher. It is suggested and recommended to conduct a preliminary study of the behavior of the facade temperatures and solar radiation before performing a thermographic inspection, it is also recommended performing a simulation of the behavior of the heat flux in the facades.
Acknowledgements The Material Testing Laboratory of the University of Brasilia for access to inspection database. CAPES and CNPq for research support in providing grants. 16
References [1]
E. Bauer, V.P. de Freitas, N. Mustelier, E. Barreira, Infrared thermography – evaluation of the results reproducibility, Struct. Surv. 31 (2015) 181–193. doi:dx.doi.org/10.1108/SS-12-2012-0039.
[2]
E. Barreira, V.P. de Freitas, Evaluation of building materials using infrared thermography, Constr. Build. Mater. 21 (2007) 218–224. doi:http://dx.doi.org/10.1016/j.conbuildmat.2005.06.049.
[3]
F. Cerdeira, M.E. Vázquez, J. Collazo, E. Granada, Applicability of infrared thermography to the study of the behaviour of stone panels as building envelopes, Energy Build. 43 (2011) 1845–1851. doi:http://dx.doi.org/10.1016/j.enbuild.2011.03.029.
[4]
A. Frota, Geometria da Insolação, Geros Ltda, Sao Paulo, 2004.
[5]
D. Paoletti, D. Ambrosini, S. Sfarra, F. Bisegna, Preventive thermographic diagnosis of historical buildings for consolidation, J. Cult. Herit. 14 (2013) 116– 121. doi:10.1016/j.culher.2012.05.005.
[6]
M. Fox, D. Coley, S. Goodhew, P. De Wilde, Time-lapse thermography for building defect detection, Energy Build. 92 (2015) 95–106. doi:10.1016/j.enbuild.2015.01.021.
[7]
R. Usamentiaga, P. Venegas, J. Guerediaga, L. Vega, J. Molleda, F.G. Bulnes, Infrared thermography for temperature measurement and non-destructive testing., Sensors (Basel). 14 (2014) 12305–48. doi:10.3390/s140712305.
[8]
S.F. Larsen, M. Hongn, Termografía infrarroja en la edificación: aplicaciones cualitativas, Av. En Energías Renov. Y Medio Ambient. 16 (2012) 25–32.
[9]
I. Martínez, A. Castillo, E. Martínez, M. Castellote, Physico-chemical material characterization of historic unreinforced masonry buildings: The first step for a suitable intervention, Constr. Build. Mater. 40 (2013) 352–360. doi:10.1016/j.conbuildmat.2012.09.091.
[10] F. Bisegna, D. Ambrosini, D. Paoletti, S. Sfarra, F. Gugliermetti, A qualitative method for combining thermal imprints to emerging weak points of ancient wall structures by passive infrared thermography – A case study, J. Cult. Herit. 15 (2014) 199–202. doi:10.1016/j.culher.2013.03.006. [11] J.R. Kominsky, J.S. Luckino, N.H. Street, T.F. Martin, Passive Infrared Thermography — A Qualitative Method for Detecting Moisture Anomalies in Building Envelopes, Tedford Pond. (2007) 1–11. [12] E. Edis, I. Flores-Colen, J. de Brito, Quasi-quantitative infrared thermographic detection of moisture variation in facades with adhered ceramic cladding using principal component analysis, Build. Environ. 94 (2015) 97–108. doi:10.1016/j.buildenv.2015.07.027. 17
[13] X. Maldague, Theory and Practice of Infrared Technology for Nondestructive Evaluation, Wiley Series, Michigan, 2001. [14] E. Bauer, E.K. Castro, E. Pavon, A.H.S. Oliveira, Criteria for application and identification of anomalies on the facades of buildings with the use of passive infrared thermography, in: V.P. Freitas (Ed.), 1st Int. Symp. Build. Pathol., Porto, Portugal, 2015: p. 12. [15] E. Bauer, E.K. Castro, A. Hildenberg, E. Pavon, Critérios para a aplicaçao da termografia de infravermelho passiva como tecnica auxiliar ao diagnóstico de patologias em fachadas de edifícios, Rev. Politec. (Instituto Politec. Bahia). 26 (2014) 266–277. [16] S.S. de Freitas, V.P. de Freitas, E. Barreira, Detection of façade plaster detachments using infrared thermography – A nondestructive technique, Constr. Build. Mater. 70 (2014) 80–87. doi:10.1016/j.conbuildmat.2014.07.094. [17] E. Edis, I. Flores-Colen, J. De Brito, Passive thermographic detection of moisture problems in façades with adhered ceramic cladding, Constr. Build. Mater. 51 (2014) 187–197. doi:10.1016/j.conbuildmat.2013.10.085. [18] E. Edis, I. Flores-Colen, J. de Brito, Passive Thermographic Inspection of Adhered Ceramic Claddings: Limitations and Conditioning Factors, J. Perform. Constr. Facil. (2012) 258. doi:10.1061/(ASCE)CF.1943-5509.0000365. [19] V. Vavilov, Noise-limited thermal/infrared nondestructive testing, NDT E Int. 61 (2014) 16–23. doi:10.1016/j.ndteint.2013.09.002. [20] W.L. Lai, S.C. Kou, C.S. Poon, W.F. Tsang, C.C. Lai, Characterization of the deterioration of externally bonded CFRP-concrete composites using quantitative infrared thermography, Cem. Concr. Compos. 32 (2010) 740–746. doi:10.1016/j.cemconcomp.2010.03.008. [21] H. Nowak, M. Kucypera, Application of Active Thermography for Detecting Material Defects in the Building Envelope, in: Inframtion Proc., 2010: pp. 1–12. [22] E. Barreira, E. Bauer, N. Mustelier, V.P. de Freitas, Measurement of materials emissivity – Influence of the procedure, in: 13th Int. Work. Adv. Infrared Technol. Appl., Pisa, Italy, 2015. [23] ASTM E1862-97 (2010), Standard Test Methods for Measuring and Compensating for Reflected Temperature Using Infrared Imaging Radiometers, West Conshohocken, Pennsylvania, United States, 2010. doi:10.1520/E186297R10.2. [24] ASTM 1933-99, Standard Test Methods for Measuring and Compensating for Emissivity Using Infrared, West Conshohocken, Pennsylvania, United States, 1999.
18
[25] C.-C. Cheng, T.-M. Cheng, C.-H. Chiang, Defect detection of concrete structures using both infrared thermography and elastic waves, Autom. Constr. 18 (2008) 87–92. doi:10.1016/j.autcon.2008.05.004.
Fig. 1. Manufactured specimens: (a) mortar board with ceramic tiles and the defect in the adhesive mortar (specimen 1); (b) mortar board with superficial cracks (specimen 2). Fig. 2. Distribution of equipment for direct heating cycle (a) and inverse heating cycle (b). Fig. 3. Areas used in the determination of parameters thermographic: (Td) average temperature of the area with defects; (Tnd) average temperature of the defect free area; (Ta) average temperature of total area. Fig. 4. Average temperature of surface, gradient temperature and Delta-T in Specimen 1 during direct heating cycle (a) and inverse heating cycle (b). Fig. 5. Average temperature of surface, gradient temperature and Delta-T in Specimen 2 during direct heating cycle (a) and inverse heating cycle (b). Fig. 6. Thermograms of the specimen 1 during the direct heating cycle (detachment). Fig. 7. Thermograms of the specimen 1 during the inverse heating cycle (detachment). Fig. 8. Thermograms of the specimen 2 during the direct heating cycle (cracks). Fig. 9. Thermograms of the specimen 2 during the inverse heating cycle (cracks). Fig. 10. TC behaviour for Specimen 1 during direct heating cycle (a) and inverse heating cycle (b). Fig. 11. TC behaviour for Specimen 2 during direct heating cycle (a) and inverse heating cycle (b). Fig. 12. TRC behaviour for Specimen 1 during direct heating cycle (a) and inverse heating cycle (b). Fig. 13. TRC behaviour for Specimen 2 during direct heating cycle (a) and inverse heating cycle (b).
19
Table. 1. Properties of mortars used in the specimens. Materials Mortar (M1) Adhesive mortar (M2)
Compressive strength 14MPa 6MPa
Density
Emissivity
1690kg/m3
Water absorption for capillarity 2.4kg/m².min1/2
1600kg/m3
3.4kg/m².min1/2
---
0.91
Table. 2. Properties of ceramic tiles used in the specimens. Materials Ceramic A Ceramic B Ceramic C
Density 2200 kg/m3 1700 kg/m3 1800 kg/m3
Water absorption 1.4% 2.2% 1.8%
Emissivity 0.84 0.80 0.82
Table. 3 Behavior of Delta-T in the specimen 1. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (°C) (°C) Ceramic A 1.05 1 0.10 3 Direct Ceramic B 2.72 1 0.82 3 cycle Ceramic C 1.58 1 0.58 3 Ceramic A -1.88 3 -1.88 3 Inverse Ceramic B -3.31 3 -3.31 3 cycle Ceramic C -2.41 3 -2.41 3
Table. 4 Behavior of Delta-T in the specimen 2. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (°C) (°C) Crack A -1.09 1 -0.02 3 Direct Crack B -1.30 1 -0.26 3 cycle Crack C -1.70 1 -0.47 3 Crack A 0.72 3 0.72 3 Inverse Crack B 1.15 3 1.15 3 cycle Crack C 2.33 3 2.33 3
Table. 5 Behavior of TC in the specimen 1. Peak values and values in the constant range. Peak Phase Constant Phase value observed range observed 20
Direct cycle Inverse cycle
Ceramic A Ceramic B Ceramic C Ceramic A Ceramic B Ceramic C
1.28 1.66 1.32 0.73 0.70 0.58
1 1 1 1 1 1
1.00 1.04 1.02 0.91 0.87 0.85
3 3 3 3 3 3
Table. 6 Behavior of TC in the specimen 2. Peak values and values in the constant range. Peak Phase Constant Phase value observed range observed Crack A 0.63 1 1.00 3 Direct Crack B 0.54 1 0.99 3 cycle Crack C 0.38 1 0.98 3 Crack A 2.33 1 1.04 3 Inverse Crack B 2.83 1 1.05 3 cycle Crack C 3.40 1 1.08 3
Table. 7 Behavior of TRC in the specimen 1. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (%) (%) Ceramic A 21 1 0.44 3 Direct Ceramic B 57 1 3.30 3 cycle Ceramic C 29 1 2.29 3 Ceramic A -26 1 -7.57 3 Inverse Ceramic B -39 1 -13.34 3 cycle Ceramic C -34 1 -9.71 3
Table. 8 Behavior of TRC in the specimen 2. Peak values and values in the constant range. Peak Constant Phase Phase value range observed observed (%) (%) Crack A -36 1 -0.12 3 Direct Crack B -46 1 -0.99 3 cycle Crack C -63 1 -1.89 3 Crack A 42 1 2.84 3 Inverse Crack B 93 1 4.54 3 cycle Crack C 245 1 9.19 3
21
HIGHLIGHTS Criteria for identifying detachments and cracks with infrared thermography inspections. Analysis of Delta-T and contrast functions to study facade defect (detachment and crack). Influence of surface temperature
and gradient temperatures behaviors on defect
visibility. Differentiated Delta-T behavior according to defect type and heat flux direction. Use of direct and inverse heat flux in the assessment of detachments and cracks with infrared thermography
22
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13