It is fundamental to study the thermal behaviour in all architectural constructions throughout their useful life, in order to detect early deterioration ensuring durability, in addition to achieving and maintaining the interior comfort with the minimum energy consumption possible. This research has developed a methodology to assess the thermal behaviour of façades in heritage buildings. This paper presents methodology validation and verification (V & V) through a laboratory experiment. Guidelines and conclusions are extracted with the employment of three techniques in this experiment (thermal sensors, thermal imaging camera, and 3D thermal simulation in finite element software). A small portion of a homogeneous façade has been reproduced with indoor and outdoor thermal conditions. A closed chamber was constructed with wood panels and thermal insulation, leaving only one face exposed to the outside conditions, with a heat source inside the chamber that induces a temperature gradient in the wall. With this methodology, it is possible to better understand the thermal behaviour of the façade and to detect possible damage with the calibration and comparison of the results obtained by the experimental and theoretical techniques. This methodology can be extrapolated to the analysis of the thermal behaviour of façades in heritage buildings, usually made up of homogeneous material.
In order to optimize the available energy resources, it is necessary to analyse the buildings energy consumption, since the building industry has a significant weight in the consumption of resources (energy and raw materials). Thermal response of the enclosures is a very important factor in this regard. On the other hand, the thermal study of a façade can give information regarding its pathological state. This paper develops a methodology that analyses heritage building façades in a thermal way with a low economic cost.
It is necessary to consider multiple constructive and environmental factors in thermal response of façades and in energy losses that occur through them. Constructive design of façades and possible pathology or alteration of the building materials are the most important factors. A correct arrangement of the façade layers and good carpentry are highlighted among the constructive factors. Green urban infrastructures [
Every day there is greater interest in collecting data from buildings [
Some failures in building envelope occur often during the construction process [
The deterioration monitoring would give an early warning of incipient problems that allow the planning of the maintenance programs, minimizing the relevant costs [
The use of data monitoring systems together with improved service-life prediction models leads to additional savings in life cycle costs [
To facilitate more cost-effective data collection for a wide variety of important building environmental and operational parameters is necessary to search tools [
On the one hand, a low-cost Arduino-based microcontroller has been employed in this research, because it is a cheap, flexible, and programmable open-source system, with easy-to-use hardware and software components [
On the other hand, there are several modalities [
Nondestructive techniques such as temperature sensors and thermography facilitate the study of building materials without damaging the building, especially its façades. Most historic buildings have been built with different types of stone. Usually, these types of buildings have thick façades composed of one or two materials. In the case of stone walls, this material covered the entire thickness of the wall. Different materials can be alternated, as, for example, the case of brick with earth or rammed earth walls, if the façade was formed by other materials. The state of deterioration of a stone monument is characterised by the type, intensity, and extent of the damage. Determining the location of the forms of deterioration has been demonstrated to be a highly appropriate research method for preventive conservation.
Thermography may propose some difficulties if a quantitative approach is intended [
The main objective of this research is to develop a specimen in the laboratory that allows obtaining real data. The specimen consists in generating a thermal flow through a homogeneous stone material. Temperature data were recorded using Arduino-based software and hardware designed by the authors. Temperatures are confronted with thermographic technology. Finally, the data obtained help to calibrate and verify a three-dimensional model of finite elements, which allows obtaining a greater amount of data and a more global estimate of the thermal behaviour. From this specimen, this work proposes a methodology to evaluate the thermal behaviour of façades in historic buildings and to detect pathology.
This research presents a methodology applicable to any material and historical building with the objective of evaluating its envelope thermally.
The methodology presented in this paper uses three techniques. One technique gives theoretical results and the other two obtain real and experimental results. The first technique is based on the elaboration of a finite element model with simulation software. This gives the theoretical results of the experiment, that is, the temperature reached by the material at each point with current conditions if it were homogeneous and it had no damage or failures. One of the techniques that provides experimental results is the use of temperature sensors that record the temperature data along the thickness of the façade. And the other experimental technique used is the thermal camera, which allows mapping temperatures on the exterior surfaces. The methodology basically consists of analysing the variation of temperature data obtained theoretically and experimentally. If the data are very similar, the façade is healthy. However, if the data change significantly in any area, there is pathology. This paper shows practical examples to understand the detection of cracks, mass faults, detachments, and so on.
A façade specimen, a model, has been constructed to develop this research.
The model is a box of 73 × 63 × 36 cm3. The exterior is made of chipboard that allows rigid and stable walls. A solid light concrete block has been placed inside the box. This block is 25 × 25 × 60 cm3 and it is wrapped by thermal insulation of rock wool 4 cm thick, except on two of its faces. The heat flows from the inside of the box to the outside through the block of lightweight concrete. The characteristics of the materials used can be seen in Table
Properties of the used materials.
ID material | Material | Density | Specific heat | Thermal conductivity |
---|---|---|---|---|
(Kg/m3) | (J/Kg°K) | (W/m°K) | ||
#1 | Chipboard | 750 | 1600 | 0.24 |
#2 | Rock wool | 40 | 800 | 0.035 |
#3 | Solid light concrete block | 400 | 1000 | 0.1 |
Specimen to perform the specimen without the top cover.
Section of the specimen with the location and sensors nomenclature.
A piece of rock wool has been placed inside the air chamber between the heat source and the concrete block. This piece is in the central third in the plan view. This is done to avoid direct radiation to the block and, in this way, only the physical phenomena of conduction and convection will appear. These two phenomena appear in the envelopes of buildings.
Table
Temperature sensors used.
# | Type | Observations |
---|---|---|
T01 | DS18B20 | Outside temperature. |
T02 | DS18B20 | Outside film coefficient temperature. |
T03 | DS18B20 | Outside edge temperature (top). |
T04 | DS18B20 | Outside edge temperature (bottom). |
T05 | K | 5 cm from outside edge temperature. |
T06 | DS18B20 | Intermediate zone temperature. |
T07 | K | 5 cm from inside edge temperature. |
T08 | DS18B20 | Inside edge temperature. |
T09 | DS18B20 | Inside film coefficient temperature. |
T10 | DS18B20 | Inside temperature (up). |
T11 | DS18B20 | Inside temperature (bottom, protected from the source). It controls the heat source’s on/off switching. |
It is noteworthy that the indoor sensors T9, T10, and T11 have been protected against the heat source, since it also radiates infrared energy that skews the temperature reading.
The main material of the specimen is a Ytong block. It is a cellular concrete and, therefore, very light that combines resistance and thermal insulation. The main chemical composition of this block is silica sand (70%), cement (14%), blowing agent (0.05%), and also water. The distribution of cells for a block with 500 kg/m3 of density is macrocells (50%) and microcells (30%) [
The board used in this research has been the Arduino MEGA based on the ATmega328P. It contains everything needed to support the microcontroller; it is simply connected to a computer with a USB cable or powered with an AC-to-DC adapter or battery to get started [
Custom sensor expansion boards can be developed to directly plug into the standardized pin-headers of the board. They enable the microcontroller to connect to several sensors [
This ensures that power supply will be continuous during the experiment and batteries will not be depleted in several days or weeks.
A program compiled in C++ language has been made for data collection. This program is responsible for data collection, publication, and temperature control inside the test.
The application builds an array of objects, thermometers, which take the temperature, store it, and publish it on both the serial port monitor and HTML by creating a local web server, which allows real-time access to the situation of the test through a local web page.
The structure of the program is based on objects, classes. To do this, it defines a pure virtual class called Thermometer, which defines all common data and methods from acquisition to publication. This main class is inherited by classes that are defined according to different types of sensors, Dallas DS18b20 and K probes, through the MAX 31856 digitalization and amplification module.
These classes receive the inheritance of the main class, Thermometer. They also define the specific methods of data acquisition according to external libraries of free license. These libraries contain the predefined classes of the specific hardware use of each sensor. There are more than fifty different types of sensor whose deployment into practical devices facilitates long-term monitoring of structural changes, reinforcement corrosion, concrete chemistry, moisture state, and temperature [
The Dallas temperature sensor DS18B20 has been used for the main sensors. It can be powered from data line. Power supply range is 3.0 V to 5.5 V. This type of sensor measures temperatures from −55°C to +125°C with ±0.5°C accuracy from −10°C to +85°C. The DS18B20 Digital Thermometer provides 9- to 12-bit (configurable) temperature readings which indicate the temperature of the device. Information is sent to/from the DS18B20 over a 1-Wire interface, so that only one wire (and ground) needs to be connected from a central microprocessor to a DS18B20. Power for reading, writing, and performing temperature conversions can be derived from the data line itself with no need for an external power source. Because each DS18B20 contains a unique silicon serial number, multiple DS18B20s can exist on the same 1-Wire bus. This allows for placing temperature sensors in many different places. Applications where this feature is useful include HVAC environmental controls, sensing temperatures inside buildings, equipment, or machinery, and process monitoring and control [
K-type probes have been used to test the data obtained by DS18B20 sensors. These probes can perform measurements below 0°C.
A Dual MAX31856 thermocouple breakout board has been used to connect the K-type temperature sensors to the board. It has 19-bit temperature resolution, handles all thermocouple types (K, J, N, R, S, T, E, and B), and allows readings as high as +1800°C and as low as −210°C depending on thermocouple type, and a line frequency filtering of 50 Hz and 60 Hz is included.
Using a higher resolution external analog-to-digital convertor would provide better readings; however, since the thermistor has an accuracy of ±0.05°C in optimal conditions, the level of precision from the 10-bit ADC is sufficient [
In this work, up to 11 temperature sensors have been used. The variations recorded by these sensors over time are known and they are used to corroborate their accuracy and to validate this research. For this reason, all temperature sensors have been calibrated. Calibration is the result of comparing the data obtained by sensors with that obtained by high quality thermometers in a test. In this way, an affordable measurement system can be used in tests, which were previously carried out with expensive thermometers. Calibration has consisted in placing all sensors in a receptacle to test the temperature at different values. The temperature values observed in all thermometers have been very similar. The temperature differences between the sensors and the reference thermometers were lower than ±0.05°C and therefore it was not necessary to correct the data obtained.
Data of all temperature sensors are saved in real time on a micro-SD card in txt format, which can easily be imported into any spreadsheet software. Thus, it is not necessary to connect the sensors to a computer for the long period that the test can last. Micro-SD memory cards that are sold today have a huge storage capacity. Sensor and time data are stored as plain text in a comma-delimited format and each data point consists of only a few bytes of data, allowing storing billions of measurements [
The micro-SD memory card used on the board has been able to store the data received by the 11 sensors, every 5 minutes, for days, and it has no storage problems. A web page has been created to check the temperature of each sensor in real time. However, the web page was very simple and it was not linked to a Mysql database to store these data.
A FLIR B335 camera has been used for this research generating thermographic images with 320 × 240 pixels of resolution. It has a temperature range between −20 and +120°C and less than 50 mK NETD sensitivity. Further processing of the thermographic images has been done with the FLIR QuickReport software. The colour palette of these pictures can be modified, as well as the temperature range and the distance to the object (usually 1 meter). Also, the maximum, minimum, and average temperature of the studied areas can be calculated. Finally, the temperature assigned to each pixel of the image is exported in Excel format.
The authors have evaluated the contributions on thermal comfort for traditional façades of buildings [
Many other previous studies have already established a link between infrared thermography and the detection of defects in stone materials, although in these studies the thermographic data for different points of the walls are interpreted by means of graphs [
The thermal pattern of a material largely depends on its characteristics (thermal diffusion, porosity, density, etc.). The possibility of being able to clearly visualise the defects of a particular material depends on the difference between the thermal characteristics of the material and the absence of homogeneity [
The emissivity value in this study is 0.95 as the default value, and so we believe that the results obtained from the thermographic measurements are reliable. Moreover, emissivity is very similar for nonmetallic materials [
The research has been completed with a finite element model of the specimen using the commercial program ANSYS Mechanical v.15, a finite element software used in engineering and architecture able to study multiple variables simultaneously [
This software allows perfectly simulating the studied case and calculating the thermal flow in each point. For this purpose, a mesh size of 0.5 cm has been used. The element type is Solid 278, a simple three-dimensional parallelepiped of 8 nodes, because it is the element that best fits in the calculations and for the model geometry. The element has a 3D thermal conduction capability. The element has eight nodes with a single degree of freedom, temperature, at each node. The element is applicable to a 3D, steady-state or transient thermal analysis. However, in [
Figure
Temperature in the vertical and horizontal section of the finite element model.
The temperature in all nodes of the finite element mesh (every 0.5 cm) is shown in this figure. Inside the box the heat source raises the temperature 30°C above the outside temperature. Outside, the ambient temperature at the selected time is about 15°C. This generates a temperature gradient between inside and outside. It can be seen very clearly in Figure
A more adequate understanding of the temperature gradient that reflects the finite element model is achieved in this figure.
Figure
The heat exchange between the air and the wall is a complex phenomenon. Each mobile molecule of air strikes the static molecules of the wall material and exchanges with them some of its vibration. Temperature is the mean of the states of vibration.
In the interface between solid and gas, very complex phenomena occur, depending on the air velocity and the horizontal or vertical position of the solid. These phenomena also depend on the temperature of the materials. The balance of these phenomena is simulated rudely by the film coefficient of convention.
The thermography of the solid helps to set the film coefficient, by comparing the temperatures of the simulation with those obtained in the thermography. The heat flow generated by the heat source produces a convection phenomenon of the air inside the box. A heat transfer occurs between the fluid and the surfaces. Here the thermal boundary layer between both elements is very important. This boundary layer is linked with temperature gradients in the fluid caused by the presence of a surface at different temperature. When forced convection occurs, the values of the film coefficient vary between 25 and 70 approximately depending on the air velocity and material, in this case, concrete [
Film coefficients used in the calculations.
Flux | |||||
---|---|---|---|---|---|
Horizontal | Vertical | ||||
Upward | Downward | ||||
Out | In | Out | In | Out | In |
4 | 35 | 4 | 35 | 4 | 35 |
The work presented in this paper is based on a specimen carried out in the laboratory with a homogeneous stone material (Figures
Once the specimen is set up, the heat source is switched on, in this case, a bulb of 60 W. From that moment, all data generated by the thermal sensors is recorded on an SD card. The test lasted several days, generating a high number of temperature records every five minutes, day and night. Outside temperature has been fluctuating, although the values have been maintained around 15°C inside the laboratory because it was winter. The inside temperature of the housing has also been kept constant with a temperature 30°C above the outside. The T11 thermometer has controlled the bulb’s on/off switching. When the temperature difference is not equal to 30°C compared to outside, with a small margin of ±0.5°C, an order is sent to turn the heat source on or off.
In addition, during the experiment, different thermographic pictures were also captured. The temperature of all faces has been recorded from different points of view. This allows comparing temperature data and verifying them more accurately.
A three-dimensional model of the experiment with finite elements was developed to simulate the specimen after obtaining sufficient laboratory data. Data from sensors and thermography has been used to calibrate this 3D theoretical model with finite elements.
The comparison of the data obtained by the three techniques allows validating and verifying (V & V) the theoretical model. Once the finite element model has been validated, it is possible to extrapolate thermal analysis to buildings façades in heritage. An exhaustive check of the temperatures that are reached in different points of a façade can be realized. Extrapolations or modifications in the boundary conditions can be made to know how the material would behave in those suppositions.
After getting the data using the three techniques for the healthy block, three types of damage usual in heritage are applied to the block and the experimental data with sensors and thermal camera are taken again. The variation of these results with respect to the theoretical ones allows detecting the damage. Three types of usual damage on historic buildings façades have been induced in the solid light concrete block.
Damage 1 represents a continuous crack with 0.5 cm of thickness and 5 cm of height and its depth covers the entire thickness of the block (25 cm). Damage 2 involves a mass loss of the material 5 cm in diameter and 10 cm deep from the inner face. Outside of the solid light concrete block nothing is seen with the naked eye. Damage 3 represents a flake, that is, a piece of the block with dimensions of 7.5 × 15 cm2 which has been detached.
Figure
Types of deterioration in the solid light concrete block.
As explained in the previous section, the micro-SD memory card recorded from the beginning the temperature and the following information: its sensor number, its serial number, the date, time, and whether the heat source is on or off. Table
Example of data recorded by temperature sensors.
Sensor |
Data | Hour | Temperature |
---|---|---|---|
T01 | 2017-03-21 | 11:27:29 | 16.875 |
T02 | 2017-03-21 | 11:27:29 | 17.938 |
T03 | 2017-03-21 | 11:27:29 | 18.375 |
T04 | 2017-03-21 | 11:27:29 | 18.375 |
T05 | 2017-03-21 | 11:27:29 | 22.266 |
T06 | 2017-03-21 | 11:27:29 | 26.313 |
T07 | 2017-03-21 | 11:27:29 | 34.727 |
T08 | 2017-03-21 | 11:27:29 | 45.063 |
T09 | 2017-03-21 | 11:27:29 | 45.122 |
T10 | 2017-03-21 | 11:27:29 | 45.688 |
T11 | 2017-03-21 | 11:27:29 | 43.188 |
In Figure
Temperature through the solid light concrete block.
In Figure
Thermographic pictures from different points of view of the specimen.
Figure
Standard picture and thermography one of the specimen.
The methodology carried out in this research is intended to be used in historic buildings. Infrared thermography has been widely used in historic buildings (Figure
Infrared thermography applied to a historic building.
The corners are also reinforced with this type of stone. The main part of the wall is composed of a rammed earth wall. The top of the corner is built with brick. It corresponds with the bell tower. This nondestructive technique is very useful in this type of buildings in heritage. It allows knowing the temperature of inaccessible points with a conventional thermometer and acquiring data from multiple points and detecting injuries and/or humidity.
The model is calculated when the specimen geometry is entered into the calculation software, the corresponding finite elements are generated, and the relevant boundary conditions are applied. In Figures
Analytical comparison of temperatures obtained on the exterior face of the solid light concrete block, by finite elements and thermography.
Point | Finite elements (°C) | Thermography |
Variation |
---|---|---|---|
P1 | 17.7298 | 17.8 | +0.40% |
P2 | 17.6851 | 17.9 | +1.22% |
P3 | 17.4555 | 17.5 | +0.25% |
P4 | 17.1746 | 17.2 | +0.15% |
Temperatures comparison on the exterior surface of the solid light concrete block, by finite elements (a) and thermography (b).
Temperature in the section of the solid light concrete block.
Figure
Table
In Figure
In Figure
Table
Comparison between theoretical and sensors results.
# | Finite elements (°C) | Sensors |
Variation |
---|---|---|---|
T01 | 16.875 | 16.875 | 0.0 |
T02 | 18.337 | 17.938 | −2.2 |
T03 | 18.466 | 18.375 | −0.5 |
T04 | 18.679 | 18.375 | −1.6 |
T05 | 22.617 | 22.266 | −1.6 |
T06 | 27.668 | 26.313 | −4.9 |
T07 | 34.526 | 34.727 | 0.6 |
T08 | 43.227 | 45.063 | 4.2 |
T09 | 43.981 | 43.375 | −1.4 |
T10 | 45.700 | 45.688 | 0.0 |
T11 | — | 43.188 | — |
Once the block has been damaged, the data storage is carried out again. This is in a later time, when the outside temperature is different. Due to this, it is necessary to simulate the theoretical model again with the new boundary conditions and to make new thermographic captures.
Sensors give temperatures in the depth of the block while the thermography takes its superficial data. Both aspects should be studied by comparing them with the theoretical data. Table
Comparison between theoretical and sensors results in damaged block.
Sensor |
Healthy block finite elements (°C) | Damaged block sensors temperature (°C) | Variation |
---|---|---|---|
T01 | 17.500 | 17.500 | +0.0 |
T02 | 18.597 | 19.313 | +3.9 |
T03 | 19.259 | 20.125 | +4.5 |
T04 | 19.433 | 20.188 | +3.9 |
T05 | 23.696 | 27.719 |
|
T06 | 29.370 | 33.438 |
|
T07 | 36.124 | 42.141 |
|
T08 | 45.368 | 49.625 | +9.4 |
T09 | 46.912 | 50.375 | +7.4 |
T10 | 47.500 | 50.000 | +5.3 |
T11 | — | 47.313 | — |
Comparison of the theoretical temperatures and thermographies.
Damage |
Healthy block finite elements |
Damaged block thermography |
Variation |
---|---|---|---|
1 | 18.3286 | 21.9 |
|
2 | 18.6333 | 20.4 | +9.5 |
3 | 18.1529 | 19.5 | +7.4 |
When the solid light concrete block is damaged, the inner temperature goes outwards more easily creating a thermal bridge. This is the main conclusion drawn from Table
Figure
Thermography of the solid light concrete block with damage: (a) type 1, (b) type 2, and (c) type 3.
Table
Thermal behaviour of building façades and heat losses through them must be known. The study of the influence of these energy flows, along with other environmental and climatological factors in the deterioration of the façades, is important. This is crucial in heritage, where buildings have suffered for centuries the inclement weather and their façades have lasted as long as possible. Most of these historic buildings are composed of façades of a homogeneous material, usually stone, but they can also be made of rammed earth or brick.
With the aim of applying this methodology in façades of historic buildings, the following conclusions can be highlighted: Thanks to this methodology, it is possible to understand more complex physical phenomena by considering the three dimensions, such as the nonlinearity of the thermal gradient of some areas. When performing the 3D study instead of the 2D one, the real boundary conditions are analysed and all aspects involved in the thermal behaviour at each point are considered. This methodology allows detecting damage in buildings. This damage is localized when a considerable variation appears between the theoretical results (finite elements) and the experimental ones (thermal sensor and thermal camera) in some area of the façade. Damage is located generally with a thermal bridge, that is, an area with higher temperatures than expected, because when there is some damage the section is depleted and the heat goes out to the outside more easily by this zone. Accuracy of film coefficient values is essential for obtaining results in line with reality in thermal simulation programs. Infrared thermography is a very useful tool to calibrate the different film coefficient values in a building, and from these, the simulation can be done and results and conclusions of all its points can be extracted. This research has shown that it is possible and very useful to design the Arduino-based software and hardware necessary to place 11 or even more temperature probes on the same board. In addition, it is possible to store this information during the days or weeks for which the specimen has lasted. This is a very powerful and economical tool for preventive conservation. Thermal insulation produces a more pronounced thermal gradient due to its thermal conductivity. Meanwhile, the gradient widens in the thickness of the solid light concrete block. The software allows showing the temperature flows through the different materials. In this way, to find out in which areas the largest flows have originated is possible and, therefore, the greatest energy losses are known. This is vitally important for the objective pursued in this research. Specifically, the greatest energy losses have occurred at the corners of the interior space where the heat source is located.
The authors declare that there are no conflicts of interest regarding the publication of this paper.