A METHOD AND APPARATUS FOR THE THERMOGRAFIC DETECTION OF THE THERMOHYGROMETRIC CONDITIONS OF VAST SURFACES FIELD OF THE INVENTION

The present invention relates to a method and a thermographic ancillary equipment for the detection of the thermo-hygrometric conditions of porous means by using infrared thermography and, more precisely, to a method for the thermographic detection of the thermo-hygrometric conditions of environmental parameters and vast surfaces, such as buildings, frescos and the like. STATE OF THE ART In civil buildings and in monumental and historical buildings, including works of art on walls such as frescos and the like, it is known that the evaporation process, with the migration of salts within the material, represents the main reason for the deterioration of the surfaces. Specifically, these aspects are of vital importance in the conservation of the artistic and historical heritage. In the analysis of the moisture of such surfaces, the key points are the knowledge of the water content in the material forming them and the vapour exchange between such a surface and the surrounding atmosphere.

As a consequence thereof, the separate measurement of the hygrometric quantities within the material or in the air is not sufficient to know and quantify the whole process. Indeed, only in theory may the mere evaluation of the content of humidity of the atmosphere surrounding the surface be sufficient in the case in which the internal atmosphere and all of the surfaces have the same thermal level, a condition which is rarely achieved as a matter of fact.

Therefore, in actual cases, two kinds of monitoring are required: one directed to know the evolution of the microclimatic state of the air, and one to know the corresponding features of the materials and especially of the surfaces. For instance, in a crowded environment, it may happen that the concentration of vapour increases as well as subsequently the dew point, or that the temperature of a surface decreases by diffusion of the heat towards the outside. In the case in which the temperature of a surface is below the dew point, condensation occurs in this surface. Therefore, the risk of surface condensation may only be evaluated by monitoring at the same time the temperature of the air and the object and by continuously

detecting the boundary conditions, even when the latter is not immediately visible with a naked eye, because the water is, for instance, absorbed within the pores of the surface under analysis. Indeed, to estimate the evaporation flow of the surface it is required that the content of vapour in the air is determined, as well as the air speed and the atmospheric pressure.

To this day, there is a main difficulty in carrying out such analyses, due to the fact that the measurements of the water content are performed on few points. The measurement of the moisture is generally complex because it is an indirect measurement, quantities such as the mass, the electrical conductivity, the propagation speed of elastic waves or microwaves and the temperature as well may be employed as an indicator of the presence of moisture.

On the other side, the use of infrared (IR) thermography is known to qualitatively identify areas in which humidity accumulates. IR thermography is an optical measurement method, that in virtue of a radiometric calibration allows obtaining a time sequence of maps of temperature, provided that some parameters are known, such as the emissivity of the surface and the instantaneous and angular radiance, or that the effects thereof are negligible. However, the complexity of the structure of a building and the dependence of its thermal history from the climatic conditions and from the actual use make the interpretation of the thermograms difficult and not always reliable.

The IR thermography may potentially be used for the detection of the moisture contained in the surfaces and the main modes of use are based on the following physical phenomena: a) the selective absorption of the infrared radiation (optical method); b) the influence of the water content on the thermophysical properties of the porous means (active thermal method); c) evaporation cooling (passive thermal method).

For the above purpose, a thermographic camera is used, which "images and records" the distribution of the surface temperature of an opaque object, such as for instance the wall of a building. The field of view depends on the objective used and on the shooting distance. The thermal and spatial resolution depend on the features of the particular thermographic camera used. The spectrum working band may vary

and is generally in the range between 1.3-2.5 μm, 3-5.5 μm, 8-13 μm or 8-9 μm, but it may be a part of these if appropriate filters are used.

In general, these techniques have shown to not always be reliable and sensitive enough for the porous means and are complex to use. In any case, the method according to the present invention is included in the "c" category.

Finally, also the mathematical models are very difficult to employ because of the coupled influence between the heat transfer and the vapour and water flows in the material. SUMMARY OF THE INVENTION It is the object of the present invention to overcome the above-said limitations by providing a method for the thermographic detection of the thermo-hygrometric conditions of vast surfaces, which employs IR thermography and simplified mathematical models of the evaporative processes, to study the hygrometric conditions of such surfaces and the surrounding air at the same time. According to a first aspect of the method of the present invention, there are provided means to measure the main thermo-hygrometric quantities related to the atmosphere, which affect the evaporation of the surfaces, said measuring means being arranged near the wall to be analysed.

According to a second aspect of the method of the present invention, there are provided means for the thermographic detection, which are made to cooperate with said measuring means in order to determine and record the time evolution of the environmental temperature, the relative humidity and the air speed in one or more points near the surface to analyse. Furthermore, the content of moisture of the solid surface is evaluated spoiling peculiarities of porous materials drying. According to a third aspect of the method of the present invention, the evaporation rate is evaluated without performing traditional measurements, by means of thermographic detections, which are processed through automatic and reliable data reduction algorithms. To this purpose, new specific parameters are used, which are defined as: "evaporation thermal index" (ETI), and its time derivative (dETI/dt, also indicated as ETI). According to another aspect of the method of the present invention, a thematic map of the critical moisture content (w _c ) on the inspected surface is made. Specifically, w _c represents a fundamental value in the activation of

the degradation processes triggered by moisture and represents a peculiar and extremely important datum from the practical point of view.

Therefore, the present invention provides a method to measure the main thermo-hygrometric quantities related to the atmosphere and solid porous bodies, which affect the evaporation from the surfaces using accurate surface temperature measurements by IR thermography and an apparatus related to the actuation of such a method according to the accompanying claims. ADVANTAGES

Advantageously, according to a method suggested by the present invention, the thermo-hygrometric conditions of each point included in vast surfaces may be evaluated in a totally non invasive or destructive manner, as well as may the boundary conditions with the same experimental apparatuses and at the same time. Therefore, the accuracy of the measurement may be considerably improved, thus obtaining images, which are immediately interpretable to identify the causes of the abnormal accumulation of moisture.

Another advantage of the method of the present invention is given by the rapid analysis provided by an optical sensor in comparison with any other contact measurement instrument, and by the same absence of errors due to the contact and to possible damages of the surface. A further advantage of the method of the present invention is given by the option of using simplified and automatic analysis models for the quantitative evaluation of the physical phenomenon of coupled heat and mass transfer associated to the corresponding thermo-hygrometric flows. The clear advantage of this approach is the possibility to very simply obtain the required information, by maintaining a degree of accuracy comparable to the traditional technical measurements (indicatively ± 10 %).

Another advantage of the method of the present invention resides in that it may be used autonomously, thus guaranteeing ease of use and reducing the measurement error; or with a mode combined with other instrumentation. The latter allows to couple the measurements obtained by traditional measurements, by means of which a recording may be obtained for very long time intervals (months or years) to few optical detections at predetermined times. This allows to give a virtually

continuous spatial distribution, as well as it being possible to estimate not only the amount of moisture in the walls, but also and above all to easily identify the cause of its accumulation. The method therefore consists in a very useful instrument in the diagnosis of the dampness, especially in monumental buildings and the estimate of the recovery time.

DETAILED DESCRIPTION OF THE INVENTION

A detailed description will now be provided of the thermografic detection method for the thermo-hygrometric conditions of vast surfaces of the present invention, given by way of non-limitative example, referring to the accompanying drawings, in which:

Figure 1 shows a device for the determination of the temperature of the air, the relative humidity and the dew point by IR thermography, which is part of the apparatus of the present invention;

Figure 2 is a plot showing the measurement of the air temperature and temperatures detected on a variety of wet surfaces with the device in figure 1 , as the relative humidity (RH) varies;

Figure 3 is a graph which shows the measurement of the relative humidity detected with the device in figure 1 using data of fig.2 and compared with a state of the art reference hygrometer; Figure 4 schematically shows a device for the detection of the gradient of the humidity exchange ratio between wall and atmosphere by means of IR thermography;

Figure 5 schematically shows a device for the detection of the surface thermal exchange coefficient and the estimation of the speed of the air by means of IR thermography;

Figure 6 shows to the plot of the evaporation rate, measured with the gravimetric method on samples of "S. Marco" brick displaying different thickness, as a function of time;

Figure 7 shows the evolution of the specific evaporation versus time, measured by means of the gravimetric method on samples of S. Marco bricks displaying variable thickness;

Figure 8 shows the standard measurements of the evaporation rate of the

same samples of S. Marco brick in Figure 6 and Figure 7, according to RILEM 25-

5432/1980 standards, as a function of the average remaining moisture content;

Figure 9 shows different phases of the evaporating process versus time;

Figure 10 shows the temperature difference between air and wet surface (δT _as ) measured by thermography on S. Marco brick samples, with various thickness;

Figure 11 shows the trend of the evaporation rate measured according to the RILEM standard and estimated by the Evaporation Thermal Index simplified method versus the residual average moisture content;

Figure 12a shows the trend vs. time of function K (t) for the S. Marco brick, and the proportionality coefficient KQ,

Figure 12b shows the values of ER and ETI as T _a and RH vary for samples of S. Marco brick;

Figure 12c shows the variation of the value of Evaporation Thermal Index for brick samples, yellow tuff and white tuff measured with T _a =20°C, RH=40% and different i/ _a ;

Figure 12d shows the experimental values of the coefficient Ko for various samples of S. Marco brick displaying different thickness and other materials, as T ₃ , RH and v _a vary;

Figure 13 shows the estimate of the evaporation flow versus time computed for two samples of S. Marco brick of different thickness using different approximations of the K function, with respect to the standard measurement;

Figure 14a represents the function (dETI/dt) in time, for samples of S. Marco brick having different thickness;

Figure 14b shows the of the characteristic times versus thickness, obtained from Figure 14a and the linear fitting parameters;

Figure 15 shows the absolute evaporation versus time measured by gravimetry and calculated using ETI per the factor Ko, for samples of S. Marco brick having different thickness;

Figure 16a shows the specific evaporation versus the average remaining content of water (w) parametric for relative humidity;

Figure 16b shows the specific evaporation of samples made of different building materials, as a function of the average remaining moisture content;

Figure 16c shows the specific evaporation of samples of S. Marco brick displaying different thickness versus the average remaining moisture content;

Figure 17 shows another embodiment of the apparatus according to the invention. At first there will be described some fundamental aspects of the operation, the method of the present invention is based on. Measurement of the thermohvgrometric state of the humid air

The first aspect of vital importance in the study of the thermohygrometric conditions relates to the so-called boundary conditions. That is, the thermo- hygrometric parameters of the humid air surrounding the porous means must be detected, as they affect the thermo-hygrometric processes occurring between the surface and the surrounding air. Therefore, according to the method of the present invention, there is provided the optical measurement of these quantities with a time continuity and strategic spatial position, by using IR thermography combined with a special new device which is an object of the present invention and is intended for the detection of the air temperature, the relative humidity and the dew point.

Figure 1 represents this device, from now on designated as IRpsicro. The operation principle thereof is based on the well-known technique of psychrometric measurement, but is carried out by means of a radiometric detection of the surface temperatures, which are made equivalent to the classical thermohygrometric measurements, which are defined as dry bulb temperature (T _a ) and wet bulb temperature (T _b ). Therefore, by analysing a thermogram it will be possible to know the main microclimatic parameters at the same time as the distribution of temperature on the examined surface. It must be specified here that the positioning and the distance of this accessory device, which in general is preferably positioned near the wall subjected to the detection, is especially important.

The IRpsicro essentially consists of two surfaces S _a and S _b , as shown in figure 1. First of all, it will be possible to determine the temperature of the air (T _a ) directly from the average temperature thermographically measured on the surface S _a . This was made of a plastic material film covered with a paint having a known emissivity which is near to a unit value, in the sensitivity spectrum band of the thermographic

camera. S _b consists instead of a special hydrophilic hose, which absorbs by capillarity the demineralised water contained in the underlying vessel, in which it is immersed below. The hose is placed within a constant section channel and is rearly lapped by a forced air flow having an average speed of 5 m s ^"1 . These surfaces will be placed in a significant region in relation to the humidity exchanges between surface and wall.

This detection allows determining the microclimatic conditions with a suitable spatial distribution. The temperature measurement accuracy is potentially better by 0.1 K, as the emissivity of this surface is known. Obviously, the precision depends on the particular thermographic apparatus used. In the case the instrument is employed outdoors it will be advisable to place a reflecting screen consisting of a thin aluminium sheet so as to screen the direct solar radiation, otherwise incident on S _a and S _b .

As far as the relative humidity measurement is concerned, it has already been specified that the measurement of the average temperature on the surface S ₃ provides with a very good precision the temperature of the humid air. It may also be considered that this value coincides with the dry bulb temperature of a psychrometric measurement. Similarly, the average temperature thermographically detected on S _b represents the wet bulb temperature 7 _b . The channel is appropriately provided with baffles so as to allow an appropriate air flow which is on the average uniform on the surface S _b , though neither affecting S _a nor the wall. The temperatures T _a and 7 _b , may be respectively defined as the temperature to which the water vapour is taken in conditions of equilibrium of convective heat and mass exchange with air subjected to forced convection. In other terms, it may be stated that T _b (detected on the wet surface) is the lowest temperature to which air may be cooled by the evaporation of water, when it is air providing the vaporisation latent heat required for the change of status.

It must be noted here that the wet bulb temperature obtained with the psychrometer in natural convection, moderately differs from this. However, there is a decrease in the precision of the measurement and some mathematical reductions adopted in the equations describing the thermal exchange are not allowed. Furthermore, to achieve the maximum precision, the psychrometric quantities are

measured considering the atmospheric pressure (P), when this is different from normal. Therefore, the measurement of the relative humidity (RH) at an atmospheric pressure (P=101000 [Pa] = 1.013 [bar]) is calculated by using the average temperatures T ₃ and T _b and by using the classic psychrometric relations, such as for instance the following equations (1).

The pressure variations are generally negligible at sea-level, but if the correction for the possible pressure variations is required, the final relation for the psychrometric measurement of the relative humidity is:

17.677λ _{m = m} 6λ l2 e^ -P _{sl {} T^ T _b )β^66(l ₊ 0.00n5 T _b ) ^

6.112 e ^r ° ^+243°5

The validity of the measurements carried out by this new method is demonstrated by the plots shown in Figure 3, detected in a climatic chamber and compared with two hygrometers: respectively, with a capacitive probe (Vaysala, HMP112Y), with a dew point sensor (General Eastern, 1200 APS) and the psychrometer incorporated in the chamber itself. The different curves shown in Figure 2 have been obtained by using different materials for the manufacturing of the surface S _b . In the graphs, each step corresponds to a value of relative humidity set in the climatic chamber so as to respectively correspond to 20%, 40%, 60% and 80%, by maintaining the temperature set point of the chamber at 20 ⁰ C. From the graphs of Figure 2 the time constant for the IRpsicro may also be determined, which is mainly connected to the thickness of the material forming the surface S _b . The small deviations between the values detected by IRpsicro and those set in the climatic chamber are due to the non-uniformity of the conditions within the cell. The comparison of the wet temperature surfaces (T _b ) and the dry (T ₃ ) one by means of IRpsicro is measured inside a climatic chamber on different materials to setup the most suitable S _b surface.

Furthermore, in Figure 3 there are shown the values of RH obtained by equation 1a and compared with the reference psychrometric measurements. There

may be noted a systematic underestimation of about 6% with respect to the true value, due to the reduced speed of the air (equivalent to 3.5 m s ^"1 , so as not to perturb the operation of the cell). Also in the case of the use of IRpsicro in natural convection, through a simple process of calibration the constants allowing to obtain very high values of accuracy for the measurement of the relative humidity are determined, with the best secondary instruments. The calibration constants a=0.99, 5=5.96, have been calculated for a linear calibration. The orange curve in figure 3 is obtained by using these values.

Measurement of the saturation pressure and partial vapour pressure The pressure of the saturated vapour (e _s ) may be determined by the Murray relation:

P _s = 0.61078 e(17.269 T _a /237.3+ T ₃ ), in which T _a is the air temperature, expressed in ⁰ C.

Another relation is Magnus' relation. The saturation pressure (Ps) may also be calculated with the following equation (2):

17.677; i ^> , = 6.112e ^r - ⁺²⁴³ - ⁵ (2)

Similarly, the same equation allows calculating the vapour pressure of water

(P _w ) corresponding to the temperature of the wet bulb 7 _b , expressed in ⁰ C. The actual vapour pressure is given by the following relation: p = p _ιl , ~?,(r _α - r ₄ )o.ooo66(i+ o.ooi i5 r _A ) _i where the pressure P _st , corresponds to Ps corrected for the amsl level of the measurement and is expressed in millibars and the temperature in kelvins. Measurement of the dew point It must be noted that even though some parameters are dimensionally a temperature, they may express an equivalence with the phase changes of the humidity content (condensation/evaporation), which imply variations in molar masses. From this point of view the approach employed, which is based on the thermographic measurements finds confirmation in the theory and the practice commonly used. Therefore, on the basis of the measurements of T _a and T _b , carried out on the surfaces S _a and S _b , the dew point (Td) may be calculated by means of equation (3).

- ^)

P(T - T _b )

Therefore, by analysing a thermogram which contains IRpsicro therein, the temperature of the surface will be known at the same time as the dew point, and therefore the risk of condensation may be evaluated for each point. Measurement of the exchange ratio

A fundamental purpose of the environmental measurements is represented by the need to identify the interactions which occur between environment and the body. For instance, a measurement of the gradient of the exchange ratio (o) made near a wall may indicate if and when this is liable to condensation or evaporation. For instance, if on the wall the level of vapour exchange is greater than in free air at a certain distance (Ao <0) it means that the local enrichment is due to the evaporation by the wall; if instead it is smaller (δo >0), it means that the condensation has taken away molecules from the air closest to the wall.

In order to carry out this measurement, there is provided a second embodiment of the device, now identified as PsicrolR and shown in Figure 4.

More precisely, according to this second embodiment, the same material and technology employed in the above-described first embodiment is employed to manufacture the surfaces S _t> i and S _b 2- The crucial point consists of the distances and size of the evaporating surface. For this purpose, there is provided that the evaporating surfaces are arranged parallelly to the wall and at a distance of 5 mm for S _b i and a 100 mm for S _b2 . The two surfaces are obviously contained in the visual field of the thermographic camera and are such as to represent a minimum obstacle for the natural circulation of the air adjacent to the examined surface.

A surface of the S _a type must be placed near the surface S _b2 so that the temperature of the air in that area is also measured.

Therefore, the exchange ratio is calculated and expressed as usual in grams of water evaporated per kilogram of humid air, on the basis of the measurements of T _a and T _b , by using equation (4)

7.65 T _{h β} = 38oi.5 lo— ^• -i-ow ^χ i<rf(r. -r _t ⁾

7.65 T, 7.65 r _t

P -6.l l2 lO 2 ^λ 43 ⁱ .-1"2+ ⁺ r ⁱ _A * . x. 1lAθ- ^~ 4* P(T _a - T _b ) 10243.12+7λ

(4)

Finally, the gradient is calculated by approximation from the difference δo=o ₂ - o-i, and is obtained by applying equation (4) respectively to the average temperature detected on the surfaces: S _b2 , S _a e S _b i , S _a . Measurement of the surface thermal exchange coefficient

As will be better disclosed hereinafter, a simplified model of the thermal exchange, allows to use the thermographic detection for the determination of the surface thermal exchange coefficient. A third embodiment of the IRpsicro device, which is designated as AnemolR, is employed to carry out this measurement, as shown in Figure 5. According to this embodiment, the accessory device is adapted to detect the surface exchange coefficient and/or the estimate of the speed of the air by means of IR thermography.

More precisely, this third embodiment of the device provides that the same consists of a reservoir containing demineralised water, an evaporating surface, saturated with water by capillarity (S _b ), connected to the reservoir, a waterproof surface with a high emissivity pigment (S ₃ ) and a high radiometric contrast calibration ruler. Unlike the first embodiment which represents the base version, in this case the water reservoir will be closed and have a regular section (Av [m ² ]), with a thin wall in a plastic material and a low section/height ratio, so that the detection of the variation of the volume of water is precise enough, on the basis of a level measurement (/ [m]). The variation of volume will be δ\/=/\ _w δ/. The instrument sensitivity will increase inversely with the cross section Av [m ² ]), whereas the autonomy will increase proportionally to A _w times the height [m ³ ].

Therefore, if both a good autonomy and a high resolution of the measurement need to be guaranteed, the reservoir will be refilled when the level will reach a minimum value, with a system which may be manual or automatic.

According to an example of embodiment of a prototype of the above-said device, a cylindrical reservoir having a diameter of 6 mm and a working lenght of a 100 mm has been made. This configuration provides a resolution of 0.056 [g/mm].

The evaporating surface (S _b [m ² ]) displayed a base perimeter of 20 mm and a height of 50 mm.

An improvement aiding the measurement of the level consists in the activation at regular intervals of a heating circuit, which coincides with the ruler divided into millimetres which will slightly heat the outside of the reservoir (about 1 °C). The heating may, for instance, be generated by a normal printed circuit to which a low voltage current pulse is provided (generated for instance by a normal dry cell) just before carrying out the thermographic reading. In this manner, as the water contained in the reservoir has a high thermal inertia (C=(p Cp λ) ⁰³ [W m ^'2 K ^"1 s ^"αδ ], the surface temperature of the reservoir will increase more above the level of the water because of the energy provided by Joule effect.

It must be noted here that, in the case in which the same device is to be used as an infrared psychrometer, a small fan will have to be fitted on top of the wet surface. In this manner, when the fan is switched on, the relative humidity will be measured and, when it is switched off, the speed of the air will be measured.

As far as the physical principle is concerned, for the measurement of the convective thermal exchange (/?) [W πf ² K ^"1 ] by the above-said third embodiment of IRpsicro or anemolR, equation (5) shows that in steady state, most of the energy required to support the water evaporation from the surface of the hydrophilic material, forming S _b of IRpsicro is provided by the air flow which laps it, which is at a temperature T _a .

Therefore the specific thermal flow (Q) [W nrf ² ] provided by the environmental adduction will be given by the product of h and the difference in the temperature between air and surface. In turn, this flow is equivalent to the amount of vapour generated per second (ER) [ g m ^"2 s ^"1 ] multiplied by the latent heat of the water (CL).

(5) A ^{L K α b}} (T _α - T _b ) {T _α - T _b )S _b δt {T _α - T _b )S _b δt

The coefficient h thus plays a fundamental role, as will be explained in greater detail hereinafter and is an important parameter to be determined.

In this regard, it must be noted here that for the purpose intended by the method of the present invention, C _L may be considered a known constant (2.257 10 ⁶

[J Kg ^"1 ]), as well as the density of the water at an ambient temperature p _w « 1025 [kg

rrf ³ ]. Whereas the thermographic system measures both T ₃ [K] and T ₅ [K], i.e. the temperature of the respectively dry and wet surface. Finally, the moderate variations of temperature involved allow to linearise h because the convective component of the surface thermal exchange prevails on the radiative component. Equation (5) therefore shows that from the measurement of the variation of the level of the distilled water contained in the reservoir, coefficient h may be derived by measuring T ₃ and T _b . The measurement of the level is carried out by the thermographic system, as water has a very high thermal capacity and therefore a vertical temperature gradient is generated outside the reservoir just at the surface of the water, each time the surface of the reservoir is thermally stressed, even because of small variations of T _a .

Furthermore, in the case a low thermal resolution thermographic camera is used, the thermal gradient imposed on the surface with the further heating device may be stressed as shown in figure 5. The algorithm shown hereinafter, which automatically processes the thermographic pictures, is used for the detection of the level, the calculation of the vapour flow and the calculation of the convection coefficient.

Furthermore, it must be specified that the measurement is also easily carried out in the dark and that it is not required for the reservoir to be transparent. The data contained in the thermographic sequence are processed with the following algorithm implemented in a Matlab ^® environment:

1 - Segmentation of the area corresponding to the reservoir, to the wet and dry surface and measurement of the corresponding average temperatures;

2 - Calculation of the average by columns on the area S ₃ ; 3 - Low-pass spatial filtering with a median filter;

4 - Calculation of the spatial derivative along the column vector for the wet surface;

5 - Identification of the level, expressed in pixels, obtained from the distance of the row index of the minimum of the derivative with respect to the coordinate of the bottom of the container;

6 - Determination of δ/, from the transformation in metrical units of the level of water in the reservoir by means of the calibration with a displayed ruler;

7 - Calculation of the mass of evaporated vapour M=(pw A _N δ/) [g];

8 - Calculation of the time derivative approximated with the difference between the datum calculated on thermograms subsequent in time;

9 - Calculation of h by means of equation (5), CL, pw , A _w , S _b being known and T ₃ e T _b being measured on the thermograms.

In the case the fluctuations of the temperature of the air allow a passive detection of the level of the water, the thermal stress device will not be required as shown in figure 5. In this case, a simple way to make the calibration ruler consists in placing by the side of the surface S ₃ (in the visual field of the thermographic camera) a stripe of reflecting aluminium (S _r ) having known length, delimitated at the ends by edges of black opaque paint.

Finally, it must be noted that the energy inputs, as for instance, solar radiation may significantly affect the evaporation processes and are to be taken into account, especially if the detection is carried out outdoors. The arrangement of a diffused reflector comprised of a creased aluminium foil and added to IRpsicro allowing to evaluate this parameter and shadow the surfaces (see figure 5). Estimate of the speed of the air

The evaporation phenomena are strongly affected by the intensity of the speed of the air (v _a ). For the purpose, an anemometric sensor placed near the examined wall provides this value. Figure 17 shows a reference wet surface, added to an aluminium cavity, working as a reference for the air temperature measurement and equipped with a portable data logger recording main environment quantities, for the needed time. This calibrated reference allows to enhance at easy the measured temperatures by thermography to 0.1 °C. The frame 1 supports IRpsicro targets 3 (one of which is shown in enlarged scale in the centre of the figure) for the measure of the boundary conditions, including air speed, the close up of one target 3 and the reference IRpsicro for the temperature calibration 2 equipped with a data logger for the long run recording of the environment conditions. Alternatively, the anemometer is not required, because the estimation of the speed of the air (v _a [m s ^"1 ]) may be obtained by equation (6) which derives it from the value of h in a laminar regime of natural convection, i.e. when the movement of the

air only occurs by gravitational and by thermal field effects between the surface and fluid. Literature reports that laminar motion normally occurs for a speed of the air lower than 6.5 [m s ^"1 ].

In general, the transfer of heat from a surface which is thermally conditioned and placed in an air flow depends on all of the characteristics of the fluid (density p, viscosity μ, thermal conductibility λ, temperature of the fluid T ₃ , pressure P, specific heat Cp, etc.) and on all of the parameters of the flow. Fortunately, for low speeds, i.e. those concerned by the present application, natural convection becomes fundamental instead. As the temperature of the evaporating surface remains constant and that of the fluid varies very little, a simplified relation is determined between speed of the air and variation of the surface temperature. Therefore, an heuristic relation may be used to estimate v _a , among those in the literature, such as for instance equation (6) set forth below:

0,8 _u _ 1057(1,352 + 0,0192r _α )v _α

It is known that for very low air speeds (lower than 0.3 [m s ^"1 ]), equations (6) are not very accurate. Furthermore, the usage of AnemolR generally requires a thermographic equipment with a high spatial resolution.

For the purpose above, an alternative technique is provided by using an algorithm employing "artificial neural networks" (ANN) by means of a special calibration. The "back-propagation" architecture for neural networks is the simplest and is used in this case. It consists of a first layer of neurons, which contains the input units, the second layer contains the "hidden" units that process information and transfer it to the following output units. Each neuron of a layer is connected to all units of the previous layer, but has no connections with neurons of the same layer. The signal propagates in one direction from the input to the output through the hierarchy of the intermediate layers.

The dynamics of the system is represented by two laws: an activation law, which updates the state of the neurons; a learning law, which modifies the strength of the connections. ANN therefore processes data by providing an answer similar to the

behaviour used for the training phase. In the learning phase, the information passes from an input neuron to an output neuron ("Forward Propagation"). The network stores the experiences in the form of thresholds and weights and initialises the procedure again. This procedure is continuously repeated by using a great number of models. A complete model consists of input data provided by speed measurements carried out with accurate anemometric instruments, in controlled conditions. The learning cycle is completed when all of the models have passed the network. The sequence of the models within a cycle is chosen randomly. The training is interrupted, when many cycles one after the other have provided no improvements and the results are satisfactory. When the learning process has been completed, the ANN is ready for its "true" processing. Therefore, after having "fed" the actual data, the expected estimates will be obtained. The training step for the ANNs may be fairly long, but the processing speed is high.

For the purpose, according to the method of the present invention, through a series of preliminary measurements a network may be trained to determine v _a . The temperature difference measured on the S _b surfaces, at actual air speed and on a reference S _b , where v _a >5 m s ^"1 is used as the input signal. Furthermore, it is possible to take into account the different temperatures at which the air may be. The training must include the RH whole range, as well. The training is obtained by comparison of the measurement given by a reference anemometer. The iterative process leads to the minimisation of a function of positive cost, which measures the difference between answers provided by the neural network, fed with the data of IRpsicro and the reference anemometer.

Actually, T ₃ , RH and the speed of the air is locally monitored by a network of IRpsicro held by the frame, as the picture 17 shows. Just one of these targets 3 must be devoted to the humidity measurement, the others to T _a and v _a ,

The answer provided by the IRpsicro sensor is a function of its time constant which, given the constructive characteristics, is therefore not suited for turbulent type flows. Conversely, IRpsicro acts as a low pass filtering of the data and it is indicated for analysis of stationary or slowly-varying flows as needed in the study of the evaporation phenomena, in the environmental study.

Monitoring of the evaporation processes

According to the present invention, the evaluation of the dampness within the solid material is also provided. It is known that the thermo-hygrometric phenomena are extremely complex when the concentration of water starts to differentiate between the damp surface and the adjacent atmosphere or within the porous material. Indeed, up today there is not an ultimate monitoring procedure.

According to the method of the present invention, it is also based on the analysis of particular conditions and characteristic of porous materials, which are significant and important in the practice, when, some of these phenomena prevail on the others. In this manner, simple and robust models may be used to estimate with good accuracy the fundamental quantities in the evaluation of the risk of degradation of the surfaces, triggered by moisture.

Furthermore, all of the measurements are provided by IR thermography, by using the devices and the previously described procedures. Hence the imaging capability of IR thermography is better spoiled, showing specific patterns associated with main moisture pathologies and clearly indicating the causes.

The processes involved are distinguished in dynamic or stationary. In our case, the choice of an appropriate time scale, allows to apply equilibrium equations to processes and balances, which strictly speaking would be prohibitively complex for an analytical description. Therefore, some parameters are introduced as macroscopic indicators allowing to describe the thermohygrometric processes acting as quasi-stationary processes.

The evaporation mainly concerns the most superficial layers of the porous material, progressing in time, so that the deeper layers are concerned only when the environmental conditions allow this. The main factors, which affect evaporation may listed as follows:

- the concentration of the water vapour in the atmosphere; the higher the concentration, the lower the evaporation;

- the concentration of other substances in the air; - the temperature of the water (in a porous means, it depends on the temperature of the surface and the air, and the higher the temperature, the greater the evaporation);

- the- flow rate of the air (in an open environment, it is connected to the speed of the boundary layer). If the air stills over the surface, the concentration of water will tend to saturate, thus progressively reducing evaporation. If air freely circulates, this is not the case. Furthermore, the evaporation will tend to further increase, because the air molecules in motion have a greater kinetic energy;

- intermolecular forces, which tend to retain the water molecules adhering to the solid surface; the greater the surface tension, the less evaporation will occur. All of these factors are considered in the method of the present invention and the effects thereof on the parameters used are evaluated to quantify the thermo-hygrometric phenomena. Therefore, the results of the new method are compared with those measured in the traditional way.

The Evaporation Rate "ER" [g s ^"1 ] is defined as the net water vapour time derivative, i.e. the flow rate density, which evaporates from the surface (A) of the porous means in the time interval δf.

Equation (7) quantitatively expresses the evaporation rate and the calculation thereof carried out on the basis of a sequence of measurements of the mass of material (M) in time, where t _\ and fn are consecutive moments in time. In this manner, if a constant time interval is used, At, the weight variation AM will be multiplied by the constant MA At.

Figure 6 shows the evaporation rate measured with the gravimetric method on samples of "S.Marco" brick having different thickness (δz) in constant environmental conditions (T _a =20°C, RH=40%). As may be noted, there is a dependence of the measured values in time as a function of the thickness of the material. This is due to the non-uniformity of the concentration of water within the sample, during the drying. The methodology suggested is potentially capable to also study this aspect, but for the main aim of estimating the average content of water the measurements will have to be made independent from the thickness. This result is made easier when the thickness of the material is relevant.

Specific Evaporation

The specific evaporation or evaporation density is defined as E(t)=(M _s -M)/A [g m ^"2 ], which provides the amount of water evaporated from the surface and is given by the decrease in weight, i.e. the difference between the mass of material at a certain

5 moment in time and the mass of the same, saturated with water (M ₈ ), normalised on the evaporating surface.

Figure 7, shows the evolution of the specific evaporation as a function of time, measured with the weight method on samples of S. Marco brick having variable thickness (between 3, 12, 28 e 45 mm). As may be noted in the figure, during the 10 saturated phase the relation is linear. The slope of the line, independent from the thickness of the material, varies with the boundary conditions. Furthermore, it may be noted in the figure that the evaporation of the total amount of the water contained in the material, obviously depends on the volume of the sample and is given by the area subtended by the curves in Figure 6, i.e. by the integral thereof.

15 However, it is more advisable to express the evaporation parameters as a function of the moisture of the material instead of time. Commonly, ER and E are defined as a function of the average remaining content of water (w) .expressed as a

% of the volume (>ιv) or of the mass {w _M ) with the relations (8), knowing that Pw ^W 1 [g cm ^"3 ]; and /W _d ". mass of the material with a physiological moisture content:

— o 2n0 W M,, = ^■ — ^ ¹ IOO _n o _r r W _v v = ^W _p ^ _v - MOO = W

Figure 8 and Figure 9 show the standard measurements of the evaporation rate of the same samples of S. Marco brick having various thickness (3, 12, 28 and 45 mm) of Figure 6 and Figure 7, according to RILEM standards 25-5432/1980, as a function respectively of the average remaining content of humidity and of time.

25 The critical content of humidity measured according to the RILEM standard

The critical moisture content (w _c ) is an extremely important parameter for the evaluation of the water content of a porous material and the method focuses on its measurement. In practice, w _c represents the transition threshold between the condition near saturation and the progressive desiccation. The practical importance

of a rapid and reliable technique as that described, capable of identifying the possible regions of the evaluated surface which are above or around this critical point, may easily be understood.

In the case in which the moisture content of the wet surface initially corresponds to the saturation condition, the evaporation will proceed at a rate which is close to maximum. This process is mainly controlled by the boundary conditions of the atmosphere. However, the evaporation is also affected by the cohesion forces and by the surface tension. For a determined material, the flow of evaporating water will therefore be a function of the microclimatic factors, such as: air speed, energy available for the surface, shortage of the vapour pressure.

The evaporation generally proceeds at a constant rate in this first step of the desiccation process, at least in so far as such microclimatic conditions are maintained. Therefore, having reached the critical moisture content, if the water inflow is less than the evaporation rate, the moisture content will progressively decrease until the pores of the material start to dry up, obviously starting from the surface. When the average remaining content of water will be below w _c , the evaporation flow is supported by the diffusion of water within capillaries, progressively decreasing the concentration of water of increasingly deeper layers. Therefore, in this second phase it will be the characteristics of the material which determine the intensity of the evaporation flow from the surface, which will very rapidly decrease. This peculiarity of porous material is spoiled by this invention.

In this manner, by analysing the surfaces, the evaporation process may be distinguished in a phase in which it occurs at the maximum degree possible, as the surface porous are saturated with water in a liquid phase and then, after having overcome w _c , with a progressively decreasing intensity. This method is focused to the detection of this particular condition, which is linked to a moisture content of interest for the analysed material, instead of an arbitrary moisture value.

Finally, a third phase may be distinguished in which the actual desiccation of the material will be reached, or rather when the so-called physiological water content (vi/ _f ) is reached. The transition to the second and third stage occurs when the porous material in not capable of providing enough water to the surface to satisfy the evaporation requirement. In this case, the phenomenon is controlled by the hydraulic

conductivity of the porous means. As the wall dries up, the evaporation plane shifts inwards and, accordingly, the net evaporation decreases.

The shape of the graph curve in this second phase is closely connected to the type of material (for instance, bricks tend to lose more water by evaporation than tuff, which drains faster).

Hereinafter, there will be provided an analysis on some experiments carried out in a laboratory on samples of various porous materials frequently used in historical buildings, in accordance with the standards. The samples have been previously characterised, for the determination of the evaporation rate and the critical content of humidity, as a function of time and water content.

In Figure 9, the results of one of these tests are represented in detail showing the critical moisture content and the three phases of the evaporation rate, in time, for a sample of S. Marco brick, according to Rilem standards. As may be noted in the figure, the critical times corresponding to the achievement of the critical moisture content and the transition between the previously described three phases of the evaporation process are highlighted therein.

These tests allow to provide greater illustrative clarity on the operation of the suggested thermographic methodology, for the determination of the content of moisture and obviously, provide a validation of the results obtained. Thermographic technique for the estimation of the evaporation flow ER

It is known that the evaporation considerably cools the surface, furthermore in this case, the energy required is mainly provided from the surrounding environment. From this, there follows that the thermohygrometric conditions may be evaluated by using simplified models, based on the estimate of the heat needed for the phase change. In practice, it is assumed that the thermal conduction only has a moderate effect. In this manner, the flow of evaporating mass may be estimated from the measurement of a parameter, which is based on the temperature gradient between the evaporating surface and the air outside the boundary layer.

In Figure 10 there are shown thermal gradients between air and humid surface (δT _as ) which have been measured thermographically on wet samples of S. Marco brick displaying various thicknesses, placed in a climatic cell, with constant boundary conditions, considered standard (7 _a , 20 °C, RH 40%). In particular, the figure shows

the time evolution of the difference in temperature of the samples and the temperature of the air. Some samples have been contained within a box (low air speed), except for two samples which have been placed outside the container (average air speed V ₃ =U m s ^"1 ). The tests have been performed in parallel with the evaporation measurements, according to the RILEM standards, displayed in the previous figures.

From the comparison between Figure 10 and Figure 6 the validation of the hypothesis at the root of the suggested measurement methodology appears immediate. Many measurements performed by setting different values for the boundary conditions (T _a , RH, v _a ) have shown that the evaporated water mass and the thermal gradient δT _as vary in a similar manner. For instance, Figure 10 shows that as the air speed increases, the increase in the evaporation rate becomes a greater temperature gradient between wet surface and air.

These tests allow to quantify the relations between the hygrometric state detected in accordance with the standard and the detections performed with the suggested method. The Evaporation Thermal Index, the proportionality coefficient Kn, the function K(t)

Evaporation and condensation are isothermal processes which are accompanied by the transfer of latent heat, respectively towards the outside of the water body and towards the surface of the water. They are affected by the amount of water vapour in the air near the evaporating surface, which in turn affects the latter. The latent heat of evaporation, CL [J Kg ^"1 ], relates the specific heat flux, (Q) [W πf ² ], and the evaporation flow rate ER [g rrf ² s ^"1 ] in the following manner:

Q = ^ = C ₁ - ER ^ h(T _α - T _s )+~(T _: - T _s )=> ER ^ k- AT^ ₍₁₀₎

where λ is the thermal conductivity of the material and h is the surface heat exchange coefficient.

Equation (10) establishes that the heat subtracted by evaporation is provided by the adductive thermal exchange, given by the sum of the sensible heat of the air, exchanged by conduction and of the radiative heat flux, plus the conduction of the material itself. But in general, the conductive flux is lower by at least one order of magnitude with respect to the convective flux. Therefore, it is reasonable to consider

the conductive heat flux as a fraction of the thermal adduction in air. Therefore, the ratio between the surface exchange coefficient h [W m ^"2 K ^"1 ] and the latent heat C _L may be concentrated at first guess in a proportionality coefficient k. The coefficient k also considers the contribution given by the conductive thermal flux. Generally, it must be reminded that the adductive/conductive flux fraction depends on the particular material and varies in time with the evolution of the desiccation.

As far as the radiative heat flux is concerned, the same is not very significant for the inner surfaces of the buildings, whereas it must be evaluated for the external surfaces. For the purpose, IRpsicro provides the existence of a third functional surface S _r , which displays a high degree of diffused reflection and is particularly suitable to indicate the extent of the radiating flow, similarly to what is schematically shown in Figure 5.

A parameter designated as Evaporation Thermal Index may also be defined as ETI = (T _a -T _s )/T _a , where T ₅ is the temperature of the wet surface, whereas T _a is the temperature of the air (expressed in Kelvin degrees). This adimensional parameter establishes a proportionality between the evaporation rate and the thermal gradients, determined with thermographic detections.

Therefore, a proportionality coefficient may be established Ko = ER/ETI [g m ^"2 s ^"1 ] = k/T _a , which allows to obtain the evaporation rate, once the evaporation thermal index ETI has been measured. Figure 11 shows this aspect, where there are compared the trend of the evaporation rate and its estimate, obtained by means of the method set forth, both of them as a function of the average remaining moisture content. More precisely, Figure 11 shows the evaporation rate measured according to the RILEM standard and estimated with the Evaporation Thermal Index times K ₀ vs. the moisture content. The measurements have been performed in standard conditions and low air speed on samples of S. Marco brick of different thickness. As may be noted in the figure, a substantial agreement is deduced as well as a systematic nature of the shifts, especially for the lower values of the moisture content. This is due to the fact that as the evaporation decreases the weight of the conductive thermal flow assumes an increasing percentage. Furthermore, it must be reminded that the thermographic measurement is strictly superficial, whereas the weight measurement is of the volumetric type and the concentration of water varies

with depth.

When required, an improvement may be obtained by using the function K(t) instead of the coefficient K ₀ . Indeed, Figure 12a shows the the function K(t) vs. time and the coefficient K ₀ for a sample of S. Marco brick having a thickness of 28 mm. The experimental tests have shown that K has a regular trend which may be approximated with two straight lines, which intersect in a critical point t _m , as will shortly be shown. Therefore, a constant value Ko may be assigned to K for the whole saturation step and an increasing value, which may be approximated with a linear function K=m t in the interval between f _m up to ft. It is worth mentioning, the value of Ko and the slope m of the line have been experimentally determined, discovering that mainly depend on the porosity of the material. It is although interesting to point out that it has been shown that for many building materials, in which the average porosity is around values of 40í50%, they vary little (Ko^δ, /77*3 [g m ^"2 s ^"2 ]). It has also been ascertained that K very slightly depends both on the temperature of the air and on the relative humidity. Indeed, as it may be noted in Figure 12b, there are shown values of ER and ETI as T _a and RH vary for samples of S. Marco brick. The environmental conditions considered in the diagram are those that may most easily be encountered in practice. For both of these parameters, a linear relation of the evaporation rate and of the ETI parameter has been found in the saturation phase, as the relative humidity and the temperature of the air vary. Therefore, the ratio ER/ETI remains practically constant and it is required to apply corrective functions, only for very low values of the moisture content (a case which will be discussed below).

As far as the dependence of K on the material is concerned, it is substantially related to its porosity {ψ), which when known allows to calculate K ₀ from the following empirical relation:

K ₀ = -2.65 iog(ψ)+9.6 (11)

The dependence of K on the air speed is explicitly given by the surface exchange coefficient (h) according to equation (12), which follows from (10):

With reference to the air velocity effect, Figure 12c shows the variation of the

value of the Evaporation Thermal Index for samples of different building materials, that is: 2 different kinds of brick, yellow and white tuff (measured with T _a = 20°C, RH = 40% and different v _a ). As may be noted in the figure, the evaporation increases in a virtually identical manner, for different materials having similar porosity when the air speed goes from virtually 0 to about 3 m/s.

Finally, Figure 12d shows the experimental values of K ₀ for various samples of S. Marco brick displaying different thickness and other materials, as the temperature, the relative humidity and the air speed are set to different values.

As may be noted, in the figure only building materials characterised by low or very low porosity such as limestone or marble are significantly distinguished.

Furthermore, there is detected a dispersion of the values which is poorly affected by the boundary conditions and moderately affected by the different thickness of the brick samples. This may be ascribed to the particularity of the suggested method that only focuses on the surface phenomena and it is not relevant for real measurements, where materials have a considerable thickness. Finally, it must be reminded that bricks are not homogeneous materials, thus the dispersion of the results must be considered typical of the variability of microscopic characteristics which are normally encountered in in situ measurements.

Therefore, by using K and ETI the evaporation flow may be estimated with good precision on the basis of thermographic detections.

Figure 13 shows the estimate of the evaporation flow versus time for two samples of S. Marco brick displaying different thickness (with boundary conditions of

T _a =20 ⁰ C, RH=40%, v _a =Z m s ^"1 ). The evaporation flow, measured with a gravimetric method and different estimates of the same, obtained by the relation ER = K ETI, are shown in the figure.

These tests confirm the underestimation of the flow by using the relation especially in the lower evaporation phase.

An extremely satisfactory result is instead achieved with the function K = K ₀ per t _\ < t < t _m K = 3K ₀ /(t _d -t _m ) per f _m < t < t _f

The continuous calculation of the coefficient K shows, instead, unacceptable instabilities of the III phase. These phenomena are natural consequences of the

approximations assumed, but it must be reminded that the measurements involved are mainly interesting for high moisture content (I and Il phase). Furthermore, a high sensitivity of K has been detected as the air speed varies, and therefore the evaluation of v _a is possible, as previously described. Time derivative of the Evaporation Thermal Index

It is required to specify here the critical times t _u t _c , t _m , ft. To do this, it is particularly useful to perform a time derivative of the Evaporation Thermal Index (dETI/dt).

Figure 14a shows the function (SETI/dt) vs. time, for samples of S. Marco brick having different thickness (and with r _a =20°C, RH=40%). Starting from the sample saturated, the function becomes negative at time t _c , i.e. when the critical moisture content is reached, and the peak corresponds to time t _m , whereas the derivative is again zero at time t _d ; t _\ and ft are respectively the initial and final time of the test.

Therefore, during desiccation a point of inflection will be reached, which coincides with time t _m where a marked peak is noted for any material. As the third phase of desiccation is reached, the derivative goes to zero again at time t= t^

Once again the dependence may be noted of the values measured in time as a function of the thickness of the material. However, in following Figure 14b the regularity of the variation of the characteristic times related to the various phases of the evaporation process may be noted. For instance, for the S. Marco brick the linear fitting of critical times measured for different sample thickness has been evaluated with the least-squares method, which allows to predict the time of the critical moisture point (t _c ) as the thickness varies (δz) with the following equation: f _c =12225 δz+18626 Despite of the understandable deviation of the value corresponding to the 3 mm thick sample, the correlation coefficient is very good (R=O.9959). The same may be done for other characteristic times and other materials.

Therefore, the true critical moisture content virtually assignable to a material having no thickness may be simply calculated. Furthermore, the function K may be constructed, as previously described.

In brief, the function (dETI/dt) identifies the characteristic times of evaporation, emphasising them. These times characterise the Il phase (shown in Figure 9), i.e. the

most interesting state from the point of view of the diagnosis of surfaces deteriorated by a moisture excess. This approach spoils a general characteristic of porous materials. Furthermore, for multilayer materials, the function {dETIIdt) is useful in the evaluation of the concentration of the moisture in depth. Estimation of the desiccation Pit)

The following relations show that the amount of evaporated water E(t) may be calculated by integrating the Evaporation Rate (ER) in time:

Figure 7 shows a plot of the specific evaporation vs. time. E(t) proceeds linearly vs. time, until critical time t _c is reached. Therefore, in the saturation phase, the following expression applies:

where E _s and M ₅ are respectively the evaporation and the mass of the material at the maximum water content.

A function which is complementary to evaporation is desiccation or decrease in the specific weight loss, defined as the water mass contained per unity wall surface and as a function of time, as from the following expression: For small enough δt, the following relations applies:

At this point, however, these functions may be used to estimate the value of E(t) and D(f) through thermographic measurements instead of weight loss measurements. Indeed, equations (13) and (14) may also be applied to an estimate obtained on the basis of ETI and the function K, as previously shown.

In this manner, the net evaporation of a unit surface may be calculated. For instance, Figure 15 shows the trend in time of the absolute evaporation calculated by using equation (14) and the factor K ₀ for samples of S. Marco brick having a thickness of 12 and 28 [mm].

As may be noted in the figure, the agreement of the results is good, even using for the calculation the simplified constant value for the function K(O=Ko.

Following from definitions of porosity (ψ) and specific mass of material (p) and water (pw), a relation between porosity, density and D is established:

JL- K hem< ψ

and it may be assumed that:

M _s = (p + p _w - ψ)V

consequently AM ₀ = V _p p _{W j} AM ₀ = V - ψ- p _w

Therefore, the specific desiccation of a unit surface may also be expressed as a function of w: (δz = sample thickness, in mm)

p _w - V p _w - δz- A

D(t) = 10 - δz ■ w => E(t) = -10 - δz- w + D ₀ ( _{1 5} ) Equation (15) is important as it allows to express the specific desiccation proportionally to the average remaining content of water (w) and therefore to invert these relations. This function is shown in Figure 16a and shows a perfect linearity. In this manner, the value of w may be obtained on the basis of the thermographic data, by using the functions calculated up to now. It must be specified here that the effects of the variations of the boundary conditions have been measured by means of the shifts of the interpolating lines, obtained by least squares regression of the experimental data. For instance, Figure 16a shows the variations due to the variation of the relative humidity between 20 and 80%. Altogether, it may be noted that these values have shown to be poorly affected by the variations of the ambient temperature, by the relative humidity and the air speed. In any case, if required, the corrections are easily applicable on the basis of

the curves shown in Figure 16. On the other side, the proportionality changes much with the porosity of the material and the possible variation in the volume of the samples, as shown in the following Figures 16b and 16c.

Figure 16b shows the specific evaporation of samples of different building materials, as a function of the average remaining moisture content.

Figure 16c shows the specific evaporation of samples of S. Marco brick having different versus the average remaining moisture content. More precisely, Figure 16b refers to different samples of building material: S. Marco brick (SMB), yellow brick (Ybrk), white tuff (WTuff), yellow tuff (AhTuff), Pietra Serena limestone (PtrS) and marble (Mrb). It should be noted that all of the materials having similar porosity behave in a similar manner. Pietra Serena stone and marble, which have much lower porosity, make an exception.

Most important,, from Figures 16a,b,c it may be noted that D ₀ may be easily obtained, i.e. the total content of water at saturation per surface unit (M ₅ -M _d )IA. Indeed, the intersection with the ordinate corresponds to w = 0 and gives the mass of the saturated sample. The maximum value of w is obtained for D=O instead and corresponds to the maximum possible content of water.

Otherwise, the above-mentioned relations allow to estimate the porosity of the material as follows:

_Ύ _ A _{O ψ =} ^δ M lθδz[mm] V ₍₁₆₎

In conclusion, once the functions plotted in fig.16 are experimentally evaluated for a particular material, then the moisture content and the drying process are evaluated by means of the measure of the paramenter ETV and the boundary conditions, both given processing thermograms. Detailed description of the method of the present invention

According to the present invention, for the general application of the thermographic method for the thermohygrometric monitoring of the surfaces to be examined, two procedures may be provided: a) initially a segmentation of the inspected area on the basis of the critical moisture content value and, then b) quantifying the moisture content.

The first goal will be carried out looking at the variation of the evaporative process that is linked to the critical moisture content trough the concept of the Evaporation Thermal Index time derivative. In order to have a practical and robust algorithm, not affected by the way how the evaporation is enhanced this result is achieved by means of a statistical analysis of the surface temperatures variation both in space and time. The procedure comprises the recording of a sequence of thermographic images which is synchronised with the stressing of the surface carried out by increasing the air speed with a flow parallel to the surface itself.

In particular, the areas above, close or far away the critical moisture content are highlighted with the technique of the Principal Component Analysis (PCA).

Therefore, to carry out this segmentation, the following steps are provided:

1) programming a sequence, acquiring a thermogram about every few seconds (for instance 10-15 s);

2) placing within the visual field "IRpsicro" and/or "AnemolR"; 3) starting the recording;

4) after a few images have been recorded (e.g. 3 images), operating a forced ventilation system parallel to the surface, with an average speed between 3 and 5 m/s;

5) continuing the recording, for a duration of about a few minutes (orientatively 2-5) and possibly taking a digital photograph of the surface;

6) processing data with the proprietary or appropriate software for the thermographic camera so as to extract the average temperatures on the surfaces of IRpsicro and process them as described in the first part, thus determining the time profiles of T ₃ , RH, v _a ; 7) carrying out the analysis of the PCA on the sequence, after having subtracted to all the images the average thermogram obtained before the switching on of the vents;

8) composing in the three planes of a RGB image the second, third and fourth components of the PCA; 9) examining the first principal components and the resulting image identifying the areas just below Wc as the most intense, up to Wb, whereas those above Wm will have lower values, while the areas almost dry has the

minimum signal value.

The above described method may also be applied by appropriately exploiting possible environmental heating/conditioning systems. Another solution that may be feasible, especially for outdoor surfaces, employs natural ventilation. For the estimation of the moisture content of a surface, one or more IRpsicro and/or AnemolR are placed at the selected area and the thermographic apparatus is positioned so as to frame them as shown in the principal figure.

The procedure therefore consists in the following steps:

1) calculating the main microclimatic parameters (T ₃ ,, RH, T^, v _a , P _sat , δo as set forth in the first part);

2) measuring T _s and calculating ETI(t);

3) evaluating K and the function K(w) as a function of the material and the speed of the air (v _a );

4) deriving ETI'(ή and determining characteristic times; 5) calculating the specific evaporating flow E(t);

6) calculating the specific desiccation D(t);

7) calculating w as a function of D(t), D ₀ ;

8) estimating porosity;

9) verifying coefficient K and the possible adjustment of the results starting from item (3) again;

10) possibly correcting T ₃ , RH;

11) Producing the final data.

The above mentioned procedure could be used at first in this complete process into a controlled conditions for the material characterization. The functions extracted from the whole process, worked out on the actual boundary conditions are ready for a fast tests belonging to a real inspection. In this case, just values of K ETI and boundary conditions computed at that time are needed. The ER measurement is done by input K ETI values into plots as in figure 13. The same approach could be used for evaluating the residual moisture content or to estimate the drying process. For instance, in the case of the evaluation and control of the surface condensation risk on a concerned surface, the following procedure of steps is provided by applying the devices and the method of the present invention:

IRpsicro is at first arranged in the visual field of the thermographic system at a distance of about 10 cm from the wall.

Then, the thermographic system is placed and focused on the region to be analysed. In this condition, the spatial distribution of the surface temperature and the air temperature, equal to that detected on surface S ₃ , is measured.

With the measured data, the relative humidity of the air is determined (by means of equation 1 and the corresponding dew temperature is determined by applying equation 3).

At this point, by arranging an isotherm around the calculated value for T _d , the surface at risk of condensation will be highlighted.

Furthermore, it is convenient to take a digital photograph of the area and record it with the thermogram or at least identify it by means of a laser pointer. The extension of the area at risk of condensation may possibly be measured. For this purpose the frame 1 shown in the figure 17 allows to easily register visual and IR images and eventually to assemble a unique mosaic image when optical limitation do not let to a complete view of the surface. This embodiment is clearly visible on both IR and visual band, indicating the centre of any target 3. It is easy to calibrate both mosaic images into metric units, because the distance between the optical centres in each IRpsicro target 3 is known. Furthermore, because only some or just one of the IRpsicro targets 3 is submitted to a forced ventilation of about 5 m s ^"1 achieved through the supporting pipes, or an air compressed reservoir or by means of a fan, RH is measured there; it is also possible to measure local v _a values on the others IRpsicro targets 3, according to the described procedure, spoiling ANN. T ₃ is measured on each IRpsicro targets 3.