International Symposium (NDT-CE 2003)Non-Destructive Testing in Civil Engineering 2003
|Start > Contributions >Lectures > Thermography:|
Association of Infra-red thermography and thermal impedance applied to the detection of empty spaces under concrete slabsOSTROWSKI Christopher, ANTCZAK Emmanuel, DEFER Didier, DUTHOIT Bruno
Laboratoire d'Artois Mécanique et Habitat, équipe MTI, Faculté des Sciences Appliquées, Technoparc
Futura, 62400 Béthune cedex, France
Infrared-thermography is an emergent technique for non-destructive testing in Civil Engineering. This method directly senses the infrared radiation, that the investigated structure emits, and is consequently able to detect the differences of temperature on the surface. Detection of subsurface defects by means of IR-Thermography is possible, since empty spaces modify the heat flow trough the system, generating surface temperature dissimilarities. This kind of approach must be associated with another assessment technique to obtain reliable results. The characterisation trough thermal impedance is a way to obtain information by means of non-destructive testing. This methodology and processing make it possible to display prominent subsurface defects.
In the field of the non destructive testing of buildings, the thermal methods are emergent. In non-destructive testing of buildings, thermals methods emerge. In addition to NDT, several other NDE methods are available, among them the so-called big five: radar, acoustics, electric, electromagnetic and x-ray radiography. The methods are standardized in North America. The datas are crossed to ensure a best precision in results. The recent appearance of infra-red thermography allowed the thermal methods to involve in the inspection of buildings. It is very gravitational from its use facility (standard video camera), it is carried out without contact and profits of a presentation from the results in the form of thermal picture. Behind this simplicity apparent, an interpretation of results requires much experiment and knowledge from the parameters influencing of the measurement .
The thermal impedance  has been used for several years in our laboratory to characterize materials and their evolution . From the experimental point of view, it's a method based on a use of fluxmetric sensors , robust, reliable and light, adapted to an in situ use. A processing data of flow and temperature is necessary to reach the required parameters. The combination of both methods is arouses double interest. The first lies in the fact that we use two thermal methods based on different transfers. A thermal impedance, where the sensor is in contact with the system, is based on a study of the transfers by conduction. The infra-red thermography, which is a remote measurement, associating the radiative transfers. The second interest is from the point of view of the results. Thermography brings qualitative datas, which are complementary with impedance study, which under certain conditions, allows to treat a quantitative aspect of the problem. At the moment, thermography is used to valid the observed results of remote measurement.
Technics of measurement
The site of experimentation
This is translated in radiometric equation used for this kind of inspection by this following expression :
From the imposed natural excitations, the system answers according to its configuration. By our experiments, we know that the privileged moments for thermographic inspection are at the end of the day, when the medium leaves a significant energy period of storage, that is to say early the morning when the medium restores its energy with its environment. It's necessary to privilege measurements in covered weather to neglected direct solar radiation and to know that measurements in rainy weather are ineffective because of homogenization of the surface temperatures. In our study's case, it was necessary to make certain assumptions to measurements on surface, to homogeneity of facings and to uniformity of the surface temperature. By considering the distance separating the concrete facings from the point of measurement like its surface compared to emitted solar flow, these assumptions are completely admissible. Certain constraints of measurement were imposed: a minimum of excitation is necessary to excite the system, measurements must be carried out in covered weather to avoid the reflection of the direct solar radiation, and with the configuration of the ground, the erection had to be made on the opposite bank with limited number of stations. The aim of measurements is to obtain a thermogram of all plates to locate the defects in under-face of the concrete facings.
The instrumentation installation allows to raise the datas of flow and temperature on the facing surface. We can thus study the energy interactions between a system and it microclimatic environment. The system response to the natural requests will depend on its nature. For experimental point of view, the used sensors are of fluxmetric type. Their design is based on the technology of printed circuits. This make it possible to have light sensors of low thickness (some tenth of mm) and of variable sizes and forms. Their particularly robustness intends them for in situ measurements, whatever the atmospheric conditions. Simultaneous measurement in the same plan of flow and the temperature is obtained by a plated thermocouple integrated into the sensor whose dimensions are: 25 cm x 25 cm x 0.03 cm. They are positioned on the surface of concrete facing by joining lubricate. The acquisition of data was carried out by an autonomous power station with a statement of data all 120s during several days. The positioning of sensors on slab was conditioned by thermographic statement. It was decided to instrument supposed healthy zones and supposed deteriorated zones. This will make it possible to compare, after data processing, experimental impedances obtained by measurements, with theoretical impedance of a medium defined as being healthy.
This paper is concerned only with one-directional transfer. q(w) and f(w) represent the temperature and flux vectors in the access plane. The input impedance Ze depends in a general way on the thermal system characterised by a transfer matrix and the output boundary conditions. The impedance of the system can be easily calculated in details by using the thermal quadripole theory. The material is considered to be a semi-infinite medium defined by a characteristic impedance Zm:
The healthy medium will be regarded as being continuous since thermophysical characteristics of the concrete and of the ground are relatively close and the contact between two materials is assured. In this case the system will be supposed to be semi-infinite, with an effusivity b. It is admitted that any system drawing aside of this configuration will be regarded as being deteriorated.
Calculation of experimental impedance
(where have ai and bi are the parameters of the linear filter)
By transformation in the field of Z one arrives at:
One then obtains the transfer function of the system by:
If one poses Z = e j2pfTe (with Te the step of time) then one brings back to the field of Fourier where one finds the experimental impedance system.
According to the defined procedure, the first stage of experimentation consists to take facings thermal picture to obtain a cartography of the surface apparent temperatures. Figure 2 presents a picture of one part of a concrete facing and the corresponding thermographic representation.
The thermographic statements were carried out in end of afternoon in covered weather. The apparent temperatures of surface reveal variations of about 2°C. Consider that we leave a period of a few days without precipitations, we can estimate that the concrete water content is relatively homogeneous, which cannot explain such surface variations. In the same way, the facing shows a rather regular quality surface, consolidating assumption made on material emissivity. One can thus estimate that the results observed are mainly due to sub-surface heterogeneities, caused either by the undermining presence or by different qualities of grounds or by irregularity of facing in under-face. In all the case it has a relationship with degradation. Possible variations of emissivity couldn't generate differences in so consequent temperatures. So the thermographic representations let suspect great degradations on all slabs. We will be limited in the continuation of work to a given representative zone. The next step consists in coupling the thermographic method with a contact method of fluxmetric type. The provision of sensors is conditioned by raised by surface cartographies. The principle is to lay out the sensors according to apparent temperatures observed, this process is represented figure 3. We chose to test the adjacent zones presenting the most consequent variations in temperature, in a range of values varying to 1°C until 2°C.
Connected to a power acquisition station autonomous, the sensors give the temperatures statements and the flows carried out over several days. Figures 4 and 5 present the type of data collected.
The curves reveal the preponderance of the component day/night. The richness of the excited frequencies and the duration of the raised signal will determine the depth of inspection. In our case of study, it is about 20 cm.
From the recorded data, one carries out the calculation of the experimental impedances as specified in analysis of impedance. The aim is to calculate each impedance and to compare them with a theoretical impedance corresponding to a healthy medium, that one will consider as being semi-infinite. The curves obtained are represented figure 6. The beam called curves 1 represents the impedances calculated starting from the measurements taken on the zones with approximately 13°C (zone 1) on the thermogram. This we strongly draw aside from the beam called curves 2 which gathers the measurements taken on the other zones at lower temperatures (zones 2) and the semi-infinite model. Considered that the curves not sticking to the semi-infinite model reflect zones of alteration, one notes that corresponding to the zones of the highest temperatures which can explain by the presence of one cavity in under-face of facing. This assumption joined the one had been done at the beginning on the global behavior of the system. The various positioning of the curves confirm the presence of the discontinuities supposed by infra-red thermography. The next step of work will consist in a development of a model allowing the quantification of the detected defect.
The thermographic methods and thermal impedance showed a great complementarity. For experimental point of view, they are particularly adapted to in situ measurement because of their use ordinary. The data processing is more consequent, but deepened, it will make it possible to quantify observed defects. The limits of the method lie in the depth of control. This one depends on the richness of the signal, in particular low frequency, which obliges to carry out tests of long duration under natural excitations.