Thermal Technical Analysis of Lightweight Timber-Based External Wall Structures with Ventilated Air Gap

: Lightweight timber-based structures are an increasingly common part of envelopes of new buildings due to increasing requirements for their energy performance. In addition, due to the fact that wood is a sustainable material, it can be assumed that the share of these structures in civil engineering will continue to increase. The subject of this article is the thermal analysis of timber-based lightweight structures under winter conditions to expand information about thermal processes in these structures. This article deals with the lightweight timber-based external wall structures with a ventilated facade and a double-skin roof structure. Experimental temperature measurements inside the structures and ventilated air gaps are used to perform the thermal analysis. By comparing experimental and theoretical data obtained by performing numerical simulation, it was shown that for achieving an ideal match of numerical simulations and measured physical properties it is necessary to take into account not only external temperatures affecting these structures, but also other factors such as solar radiation and heat emission into the cold night sky. In the case of the external walls with ventilated facade, the beneﬁt of a ventilated air gap has been demonstrated in relation to smaller temperature ﬂuctuations that affect the structures.


Introduction
Today, the development of building materials and new compositions of building structures is closely connected to the demands for improving the energy efficiency of buildings [1][2][3][4]. Another advantage of wood-based constructions is their sustainability. This is mainly due to the fact that almost forty percent of all energy in Europe is consumed by buildings [5,6]. It is for this reason that the share of timber-based structures has been growing in recent years as part of buildings envelope. Much research abroad deals with thermal technical parameters of these structures [7][8][9][10], as well as in the Czech Republic [1,11]. The timber-based building construction systems can be divided into light skeletal structures, heavy skeletal structures, wooden panel structures, and timber and log structures [12]. Of these the skeletal structures or panel structures are the most commonly used types in the construction of new buildings. Hybrid structures that combine a reinforced concrete or steel load-bearing structure and prefabricated wood-based panels are also very interesting. The main advantage of these structures is their greater usability in the area of multi-story buildings. The advantage of all these structures is usually their smaller thickness while achieving better thermal technical parameters, which is due to the relatively light load-bearing structure of wood or wood-based materials [13][14][15][16]. However, despite the many benefits that these structures offer, many investors are still leaning towards using traditional masonry structures. However, based on a previous study [17], it was shown that 46.3% of the Czech population prefers wood as a building material. The trend of increasing the share of wooden buildings has been known in the Czech Republic since 2005, but it can be assumed that the share of wooden buildings will still increase. This paper deals with lightweight double-skin envelopes falling under a category of panel structures with a skeleton formed by I-beams. Attention is focused on double-skin structures. The principle of double-skin structures is based on a ventilated air gap. The design of a ventilated air gap is important for the proper functioning of such structures so that they meet the above requirements. However, the Czech legislation does not pay close attention to the design of a ventilated air gap. There is currently no standard available for the design of ventilated facades. The closest usable is the CSN 73 1901 standard [18], but it is primarily intended for the design of ventilated air gaps in double-skinned roof structures. The paper monitors thermal and humidity parameters of double-skinned structures, both for a facade and a single pitch double-skinned roof. Many researchers have been investigating the influence of ventilated air gaps on thermal technical parameters of structures [19][20][21][22]. Suarez et al. studied the benefits of ventilated facade in the Continental Mediterranean climate with cold winter and hot summer. The analysis was performed using the numerical simulation uses a commercial Computational Fluid dynamic code (CFD). The created numerical model has been validated with infrared measurements of the exterior surface of the actual facade. The research points mainly to energy savings in the summer with higher solar radiation and surrounding air temperature [23]. Serra et al. present the results of an experimental campaign on a climate facade with mechanically ventilated air gap. The energy efficiency of the facade has been evaluated considering the ability to pre-heat the ventilation air in the winter season, and the ability to remove part of the solar load during the summer season [24]. When using common calculation methods of energy efficiency of structures, it is accepted that the temperature in the ventilated gaps of envelopes corresponds to the exterior temperature. Susanti et al. demonstrate that due to various heat exchange process the temperatures of gaps may not be equated to the exterior temperature, especially in the case of double-skin roof [25]. The results of experimental research [26] demonstrated that the absolute error in the calculations of the average daily temperature of the ventilated air gap and external surface of the thermal insulation layer does not exceed the limits of 1.21 • C and 2.71 • C. The influence of doubleskinned structures on the energy performance of buildings is dealt with by experts all over the world, and some research is focused on climatic areas with temperate climates [27][28][29]. In this research, results of experimental heat measurements within the compositions of the monitored structures and the parameters of the internal and external environment were used in order to evaluate the influence of ventilated air gaps on the thermal technical behavior of lightweight timber-based cladding. A similar experimental study on heat transfer through a lightweight building envelope wall under real, atmospheric boundary conditions was performed in this work [30]. Attention is focused on envelopes and roof structures. A numerical simulation of varying temperatures occurring in an envelope was performed and then compared with the data obtained by measurements, thus proving that the temperature distribution in the lightweight envelope corresponds, with a slight variation, with the results of the numerical calculations. Many studies deal with the assessment of the suitability of the use of numerical calculation methods for the prediction of heat transfer through the structures [8,31,32].

Analyzed Lightweight Envelopes
For the purposes of a temperature analysis, lightweight envelopes of a passive experimental house were chosen; see Figure 1. This house is located in Ostrava, Czech Republic.  The official name of this building is the "Research and Innovation Center of the National Wood Processing Cluster". This experimental building was built in 2012 and thanks to its exceptional nature, allows a long-term monitoring and evaluation of physical-technical properties of building structures and the internal environment of the entire building under real external conditions [11]. The technical equipment of the building includes a superior regulation of the designed heat sources with the possibility to utilize them for research and educational purposes. The system enables to measure all necessary quantities, flows, powers, and thermal energy. The experimental building is designed as a timber prefabricated panel structure. The structural solution of the building is evident in Figure 2. External wall structures in contact with the outside air are made of prefabricated panels, which were manufactured under stable climatic conditions in a production hall. Prefabricated panels are formed by a load-bearing structure based on wood, between which thermal insulation is inserted. The outer and inner surface of the panel is thermally insulated too. During the production of these panels, hygrothermal sensors were inserted into their structure, which permits long-term monitoring of these parameters. The loadbearing elements of the external walls and roof are made of Steico wooden I-beams. External wall structures are insulated with Steico wood fibre insulation, placed between the beams as well as on the structures from the outside and inside. The cladding of the individual layers of the external wall structures is made of Fermacell gypsum fiber boards. All assessed structures meet the values of the heat transfer coefficient Upas according to ČSN 730540-2 [33] recommended for passive buildings.
Two types of double-skin structures were chosen for the thermal analysis, namely: • External walls with ventilated facade, • Double-skin roof structure.
There are two designs of the external walls with ventilated facade. The first one consists of a 20-mm-thick air gap and exterior wood cladding. The cladding is made of wooden slats spaced 30 mm apart, so the cladding does not form a continuous surface. The official name of this building is the "Research and Innovation Center of the National Wood Processing Cluster". This experimental building was built in 2012 and thanks to its exceptional nature, allows a long-term monitoring and evaluation of physicaltechnical properties of building structures and the internal environment of the entire building under real external conditions [11]. The technical equipment of the building includes a superior regulation of the designed heat sources with the possibility to utilize them for research and educational purposes. The system enables to measure all necessary quantities, flows, powers, and thermal energy. The experimental building is designed as a timber prefabricated panel structure. The structural solution of the building is evident in Figure 2. The official name of this building is the "Research and Innovation Center of the National Wood Processing Cluster". This experimental building was built in 2012 and thanks to its exceptional nature, allows a long-term monitoring and evaluation of physical-technical properties of building structures and the internal environment of the entire building under real external conditions [11]. The technical equipment of the building includes a superior regulation of the designed heat sources with the possibility to utilize them for research and educational purposes. The system enables to measure all necessary quantities, flows, powers, and thermal energy. The experimental building is designed as a timber prefabricated panel structure. The structural solution of the building is evident in Figure 2. External wall structures in contact with the outside air are made of prefabricated panels, which were manufactured under stable climatic conditions in a production hall. Prefabricated panels are formed by a load-bearing structure based on wood, between which thermal insulation is inserted. The outer and inner surface of the panel is thermally insulated too. During the production of these panels, hygrothermal sensors were inserted into their structure, which permits long-term monitoring of these parameters. The loadbearing elements of the external walls and roof are made of Steico wooden I-beams. External wall structures are insulated with Steico wood fibre insulation, placed between the beams as well as on the structures from the outside and inside. The cladding of the individual layers of the external wall structures is made of Fermacell gypsum fiber boards. All assessed structures meet the values of the heat transfer coefficient Upas according to ČSN 730540-2 [33] recommended for passive buildings.
Two types of double-skin structures were chosen for the thermal analysis, namely: • External walls with ventilated facade, • Double-skin roof structure.
There are two designs of the external walls with ventilated facade. The first one consists of a 20-mm-thick air gap and exterior wood cladding. The cladding is made of wooden slats spaced 30 mm apart, so the cladding does not form a continuous surface. External wall structures in contact with the outside air are made of prefabricated panels, which were manufactured under stable climatic conditions in a production hall. Prefabricated panels are formed by a load-bearing structure based on wood, between which thermal insulation is inserted. The outer and inner surface of the panel is thermally insulated too. During the production of these panels, hygrothermal sensors were inserted into their structure, which permits long-term monitoring of these parameters. The load-bearing elements of the external walls and roof are made of Steico wooden I-beams. External wall structures are insulated with Steico wood fibre insulation, placed between the beams as well as on the structures from the outside and inside. The cladding of the individual layers of the external wall structures is made of Fermacell gypsum fiber boards. All assessed structures meet the values of the heat transfer coefficient U pas according tǒ CSN 730540-2 [33] recommended for passive buildings.
Two types of double-skin structures were chosen for the thermal analysis, namely: • External walls with ventilated facade, • Double-skin roof structure.
There are two designs of the external walls with ventilated facade. The first one consists of a 20-mm-thick air gap and exterior wood cladding. The cladding is made of wooden slats spaced 30 mm apart, so the cladding does not form a continuous surface. The second design of the ventilated façade consists of a 40-mm-thick air gap and an external cladding made of Cetris cement-bonded particleboard. Again, the boards do not form a complete continuous surface as they are spaced from each other by 9 mm gaps. The designs of the facade cladding is evident in Figure 3. The second design of the ventilated façade consists of a 40-mm-thick air gap and an external cladding made of Cetris cement-bonded particleboard. Again, the boards do not form a complete continuous surface as they are spaced from each other by 9 mm gaps. The designs of the facade cladding is evident in Figure 3. The double-skinned roof structure design is similar to the external walls structure design, as part of its composition consists of a 60-mm-thick ventilated air gap. The exterior side of the structure is finished by Steico Universal thermal insulation layer. In this case, the outer surface of the thermal insulation is provided with a safety diffusion foil.

Experimental Measurements
The monitored building is equipped with a measuring system, thanks to which many physical quantities have been continuously recorded since the building was put into service. Measuring devices of TR instruments spol. s.r.o. are used for these long-term measurements incorporating type Rotronic sensors. The measured data are recorded in a time interval of 15 min, using a multi-channel dataTaker DT80G Geologger measuring control panel. For measuring of hygrothermal parameters inside of the structures ROTRONIC HC2-CO4 RH/t sensors are used. For measuring of thermal parameters on external and inner surface of structures there are used ROTRONIC TG 68-60 Pt1000 and TG 7 Pt1000 sensors. For measuring of internal and external air temperatures there are used RO-TRONIC Pt1000 sensors. Temperatures and relative humidity measuring sensors were built into the exterior wall structures already in the prefabrication phase. Location of sensors for continuous measurements inside the exterior walls and the roof, including the compositions of the individual structures are shown in Figures 4 and 5. For the purpose of this research, Arexx devices with TSN-TH70E and TSN-33MN sensors were installed inside the air gaps of external walls. The measured data were recorded in a time interval of 45 s. The accuracy of the measuring devices used is given in Table 1. Only certain time periods were selected for the purposes of this analysis, which are subsequently evaluated. The double-skinned roof structure design is similar to the external walls structure design, as part of its composition consists of a 60-mm-thick ventilated air gap. The exterior side of the structure is finished by Steico Universal thermal insulation layer. In this case, the outer surface of the thermal insulation is provided with a safety diffusion foil.

Experimental Measurements
The monitored building is equipped with a measuring system, thanks to which many physical quantities have been continuously recorded since the building was put into service. Measuring devices of TR instruments spol. s.r.o. are used for these long-term measurements incorporating type Rotronic sensors. The measured data are recorded in a time interval of 15 min, using a multi-channel dataTaker DT80G Geologger measuring control panel.  Table 1. Only certain time periods were selected for the purposes of this analysis, which are subsequently evaluated.    The measurements were performed in January, during which the lowest outside temperatures were recorded. Data obtained by measurements on the inner and outer surfaces of the structures, external and indoor climatic data, and data obtained by measurements inside the air gaps are used for the analysis. The measurement results were used to analyze the effect of air gaps on the structures and also to define the boundary temperature conditions for the purpose of numerical simulation of heat propagation in the exterior wall.    The measurements were performed in January, during which the lowest outside temperatures were recorded. Data obtained by measurements on the inner and outer surfaces of the structures, external and indoor climatic data, and data obtained by measurements inside the air gaps are used for the analysis. The measurement results were used to analyze the effect of air gaps on the structures and also to define the boundary temperature conditions for the purpose of numerical simulation of heat propagation in the exterior wall.  The measurements were performed in January, during which the lowest outside temperatures were recorded. Data obtained by measurements on the inner and outer surfaces of the structures, external and indoor climatic data, and data obtained by measurements inside the air gaps are used for the analysis. The measurement results were used to analyze the effect of air gaps on the structures and also to define the boundary temperature conditions for the purpose of numerical simulation of heat propagation in the exterior wall.

Numerical Simulation of Temperature Curves in Lightweight External Wall Structures
Numerical calculations of thermal processes in structures were used in the research. In common building practice, analytical methods are usually used for thermal engineering Sustainability 2021, 13, 378 6 of 15 assessment of structures [34]. However, they are not suitable for detailed evaluation of temperatures in lightweight perimeter structures due to their design, where the structures include systematic thermal bridges. In this case, it is necessary to perform thermal engineering tasks using numerical calculation methods. The output of numerical calculations is a temperature field that describes the temperature distribution in the evaluated model of the construction. These are determined by converting the partial differential equations defining the heat conduction to the system of linear equations and their subsequent solutions [35]. Thus, as in the case of analytical methods, the numerical solution is based on the Law of Energy Conservation, which is applied to thermal engineering phenomena such as Fourier's Law. Heat dissipation through conduction is characterized by a heat flux q that is directly proportional to the temperature gradient. This dependence on stationary heat conduction is expressed mathematically by the first Fourier's law [35]: where q is density of heat flow (W·m −1 ), λ is thermal conductivity coefficient (W·m −1 ·K −1 ), θ is the time temperature ( • C), and x is the coordinate of the point (m). If the temperature does not change over time, as in the case of steady (stationary) heat conduction, then for heat conduction, assuming the independence of the thermal conductivity coefficient λ on the temperature and direction of heat propagation, the socalled Fourier's partial differential equation can be defined for three-dimensional heat conduction in the Equation (2), analogous to one-dimensional and two-dimensional heat conduction. Further, in the case of unstabilized (non-stationary) heat conduction, the fact that the temperature is a function of time θ = f (t), enters the relationships. Here, the socalled second Fourier's law applies, analogously for one-dimensional and two-dimensional heat conduction: where a is thermal conductivity coefficient (m 2 ·s −1 ), θ is the temperature ( • C), t is the time, and x, y, and z are the coordinate of the point (m). For the purposes of the research, the Finite Elements Method (FEM) was used. The Finite Elements Method is a variation method, the basic principle of which is the division of a continuous region into a set of separate sub-regions, the so-called finite elements [36,37]. The detected parameters are then determined at the individual nodal points of these elements. The principle of the method is also the conversion of the partial differential equation into a system of linear algebraic equations for the desired potential values at the nodes of the finite elements. Since this is an iterative method, first, an anticipated solution is selected at the beginning of the calculation, and then the solution to the equations is sought using successive iterations (refinements). The iteration process stops at the moment when a predetermined inaccuracy is reached, which is the difference between the observed temperatures of the given and the previous iteration round at all nodes [38]. When computers are used, it is possible to choose very low values of inaccuracies and for this reason these methods are sometimes incorrectly referred to as "exact". In the case of manual calculation, it would not be possible to achieve such accuracy due to the time-consuming nature of the calculations.
As part of the research, numerical calculations of the courses of heat variations in the structures were performed and these were then compared with the actual measured data. To be able to perform such comparison, it is necessary to carry out numerical calculations under non-stationary boundary conditions, i.e., under a state corresponding to the actual behavior of exterior and interior temperature boundary conditions, which are under normal circumstances never stationary but change over time [39]. Unlike stationary tasks, the temperature distribution in the design at each calculation step is dependent on the temperature distribution in the structure in the previous step.
In the case of non-stationary problems, according to Fourier's Second Law, is the physical magnitude, which influences the temperature distribution in the construction under the influence of the variable boundary conditions, called the coefficient of thermal conductivity a. This quantity expresses the rate of balancing of the temperatures in the substances and depends on the coefficient of thermal conductivity λ, the density ρ, and the specific heat capacity of the substance. Therefore, in order to create a computational model for non-stationary heat conduction tasks, it is also necessary to specify the physical properties of the materials in addition to the geometry itself. For non-stationary calculations, it is therefore necessary to define these material characteristics, and variable boundary conditions, as a function of time and also the initial state of temperature distribution in the construction. This fundamentally affects the results of the calculation itself. The initial state can be set by determining the temperature field by means of a stationary calculation with the input of boundary conditions so that the resulting temperature field closes as closely as possible to the actual temperature distribution in the structure at the beginning of the non-stationary simulation.

Theoretical Analysis of Temperature Curves in Lightweight Exterior Wall Structures
The theoretical analysis included only a selected detail of the external wall with a ventilated air gap. The numerical simulation was performed for a time interval of 48 h with a time step of 600 s. The ANSYS R16.0 program [40] was used for the simulation of the temperature conduction, which is capable of solving non-stationary thermal-technical tasks.
The detail of the construction model was created in accordance with the actual dimensions of the structure composition. Individual materials were assigned thermal-technical properties declared by the manufacturer of the building materials used; for the purposes of simulation, these declared properties were adjusted to the so-called calculated values, which took into account the degree of moisture absorption of materials. After creating the model, it was necessary to define the boundary conditions. Boundary conditions used were those of Newton's type II, which are defined by the temperature of the ambient air on the inside and outside of the structure and the heat transfer coefficient on the inside and outside of the structure. The indoor and outdoor air temperature was entered as a non-stationary boundary condition. These values were determined by experimental measurements. The heat transfer coefficients on the inside and outside of the structure were chosen as stationary according toČSN 73 0540-3 [41], which may lead to a certain calculation error, as it is a variable. The calculation thus does not take into account the effect of air temperature and its flow at the surface of the structure on the outside and inside on the thermal resistance during heat transfer. The temperature distribution in the structure, which corresponds to the temperature field determined on the basis of the average boundary conditions of the previous 48 h was chosen as the initial state of the simulation.

Temperature Measurements Inside the Structures and Their Comparison with Numerical Simulation
As already mentioned, a numerical simulation of the temperature fields was performed for a characteristic section of the external wall with a ventilated air gap. The result of the numerical simulation is the distribution of temperatures in the structure under loading by the actual temperatures of the indoor and outdoor environment obtained by experimental measurements. Graphical outputs of temperature fields are shown in Figure 6. These figures illustrate the change in temperature distribution in the evaluated detail during the simulation. These figures illustrate the change in temperature distribution in the evaluated detail during the simulation. Based on the results of numerical simulations, a comparison of actually measured temperatures within the composition of the external wall structure and the results of numerical simulations of temperature fields using non-stationary boundary conditions was performed. This comparison was made for both the location of the thermal bridge and for the axis between the thermal bridges. The average differences of resulting temperatures based on experimental measurements and numerical simulation of five positions in the evaluated details are then shown in Table 2. The results are given as the average for the first 24 h of simulation, followed by the average for the subsequent 24 h of simulation and finally by the average for the entire simulation time, i.e., 48 h. When comparing the results, it is also necessary to take into account the accuracy of measuring devices and to this end the measurement deviations are given in Section 2.2, when these measured data were entered as external and internal boundary conditions and then the measured data inside structures again compared with simulation results. Larger differences between the measured and calculated values were found at the site of the thermal bridge, the result of which is related to the deformation of the temperature field at the thermal bridge site. The established smallest temperature difference of 0.3 °C and the highest temperature difference of 1.3 °C can be considered as very good results. Larger temperature differences were found closer to the outside of the structure, which correspond to the actual effects acting on the structure from the outside, such as sunlight, air flow, etc. These effects were not included Based on the results of numerical simulations, a comparison of actually measured temperatures within the composition of the external wall structure and the results of numerical simulations of temperature fields using non-stationary boundary conditions was performed. This comparison was made for both the location of the thermal bridge and for the axis between the thermal bridges. The average differences of resulting temperatures based on experimental measurements and numerical simulation of five positions in the evaluated details are then shown in Table 2. The results are given as the average for the first 24 h of simulation, followed by the average for the subsequent 24 h of simulation and finally by the average for the entire simulation time, i.e., 48 h. When comparing the results, it is also necessary to take into account the accuracy of measuring devices and to this end the measurement deviations are given in Section 2.2, when these measured data were entered as external and internal boundary conditions and then the measured data inside structures again compared with simulation results. Larger differences between the measured and calculated values were found at the site of the thermal bridge, the result of which is related to the deformation of the temperature field at the thermal bridge site. The established smallest temperature difference of 0.3 • C and the highest temperature difference of 1.3 • C can be considered as very good results. Larger temperature differences were found closer to the outside of the structure, which correspond to the actual effects acting on the structure from the outside, such as sunlight, air flow, etc. These effects were not included in the numerical simulation, only outdoor temperature was used as a marginal outside condition. However, choosing a structure with a ventilated facade should eliminate these effects. A graphical display of the course of the temperature variations was created for greater clarity; see Figures 7 and 8. Figures 7 and 8 show the results of measurements and numerical simulations over 48 h. The results show a difference between the experimental and theoretically determined temperatures in the given cross-sections of the structure. It is clear from the figures that the simulation did not take into account the effect of solar radiation, when significantly higher temperatures were experimentally found in the afternoon in position 5, i.e., in the position on the outside of the structure (in this case in front of the ventilated air gap). This phenomenon also manifested itself in position 4 in the experimentally obtained data, but with a certain time response, when the measured temperatures rose slightly in contrast to the calculated temperatures. In other positions, this phenomenon is no longer apparent.    The results show a difference between the experimental and theoretically determined temperatures in the given cross-sections of the structure. It is clear from the figures that the simulation did not take into account the effect of solar radiation, when significantly higher temperatures were experimentally found in the afternoon in position 5, i.e., in the position on the outside of the structure (in this case in front of the ventilated air gap). This phenomenon also manifested itself in position 4 in the experimentally obtained data, but with a certain time response, when the measured temperatures rose slightly in contrast to the calculated temperatures. In other positions, this phenomenon is no longer apparent.

Evaluation of Experimental Measurements of Ventilated Air Gaps
In addition to the above-mentioned measurement of temperatures inside structures, the results of which were compared with the results of numerical simulation, other measurements were performed to monitor the effect of ventilated air gaps in these structures. These measurements were also performed in January, and included an external wall with a ventilated façade structures and a double-skinned roof structure; a description of the compositions and the specification of the location of the measuring devices is given in Section 2.2. Figure 9 depicts graphically the temperature profiles on the external side of the structures. Since these are double-skinned structures that are being evaluated, the temperatures in question are those found inside the air gap that directly affect these structures.
The analysis of the data revealed that in the case of the roof structure, there are significant drops in temperature in the ventilated air gap compared to the outside air temperature. The subcooling of the air inside the air gap occurs in connection with the cold radiation of the night sky [22,42]. This phenomenon has a negative effect on the heat loss of the roof structure at night, when the structure is burdened with a larger temperature gradient. The maximum difference found between the outside air temperature and the temperature inside the air gap is 5.5 °C. Regarding the temperature balance within the observed period, the average air temperature inside the air gap was −2.2 °C, while the average outdoor air temperature was −0.8 °C. The cold air inside the air gap then has a

Evaluation of Experimental Measurements of Ventilated Air Gaps
In addition to the above-mentioned measurement of temperatures inside structures, the results of which were compared with the results of numerical simulation, other measurements were performed to monitor the effect of ventilated air gaps in these structures. These measurements were also performed in January, and included an external wall with a ventilated façade structures and a double-skinned roof structure; a description of the compositions and the specification of the location of the measuring devices is given in Section 2.2. Figure 9 depicts graphically the temperature profiles on the external side of the structures. Since these are double-skinned structures that are being evaluated, the temperatures in question are those found inside the air gap that directly affect these structures.
The analysis of the data revealed that in the case of the roof structure, there are significant drops in temperature in the ventilated air gap compared to the outside air temperature. The subcooling of the air inside the air gap occurs in connection with the cold radiation of the night sky [22,42]. This phenomenon has a negative effect on the heat loss of the roof structure at night, when the structure is burdened with a larger temperature gradient. The maximum difference found between the outside air temperature and the temperature inside the air gap is 5.5 • C. Regarding the temperature balance within the observed period, the average air temperature inside the air gap was −2.2 • C, while the average outdoor air temperature was −0.8 • C. The cold air inside the air gap then has a direct effect on the temperature of the outer surface of the lower skin of the roof structure. It is different in the exterior double-skin walls when the temperatures in the ventilated air gap are demonstrably higher, on average by 0.5 • C in the case of the Cetris cladding and by 1.1 • C in the case of the timber cladding. This fact has, among other things, a positive effect on the energy balance of the external wall structures. direct effect on the temperature of the outer surface of the lower skin of the roof structure. It is different in the exterior double-skin walls when the temperatures in the ventilated air gap are demonstrably higher, on average by 0.5 °C in the case of the Cetris cladding and by 1.1 °C in the case of the timber cladding. This fact has, among other things, a positive effect on the energy balance of the external wall structures.    direct effect on the temperature of the outer surface of the lower skin of the roof structure. It is different in the exterior double-skin walls when the temperatures in the ventilated air gap are demonstrably higher, on average by 0.5 °C in the case of the Cetris cladding and by 1.1 °C in the case of the timber cladding. This fact has, among other things, a positive effect on the energy balance of the external wall structures.   As in the case of Figure 9, the effect of the radiation into the clear night sky [22,42] is noticeable. The most cooled part of the construction is the roof structure. Very low temperatures on the surface of the lower skin of the roof structure are affected by the temperature of the air inside the air gap, which may fall below the ambient air temperature at night. This can be seen in the previous Figure 9, when the temperatures inside the air gap at night and morning were significantly lower than the outside air temperature. In the case of the external walls with a ventilated air gap, it is clear that the structures are effectively protected against external influences, and this can be stated on the basis of a more stable course of temperature variations.

Discussion
The subject of this research was winter thermal analysis of the timber-based lightweight external wall structures. The aim of the paper is to expand information about thermal processes in these structures, whose main advantage is their sustainability. The evaluated structures are part of the experimental building constructed in the passive standard. This paper focuses on double-skinned structures, i.e., structures with a ventilated air gap. The attention is specifically focused on the external walls with a ventilated facade and a doubleskin roof structures. The data obtained by experimental measurements of temperatures inside these structures including temperatures inside ventilated air gaps are used for the analysis, and the temperatures of the indoor and outdoor environment were also used for a comprehensive evaluation.
As part of the analysis of temperatures inside the composition of the external wall structure with a ventilated façade, a comparison of the results of the experimental measurements and numerical simulations of heat diffusion in the building structure was performed. The numerical simulation was performed using ANSYS software for non-stationary boundary conditions. Temperatures determined by experimental measurements for indoor and outdoor environments were included in the calculation as non-stationary boundary conditions. To perform the numerical simulation, a characteristic section of the structure of the evaluated external wall was chosen. The model of the structure was made for thermal technical parameters of materials used in the composition of the evaluated structure. Numerical simulation was performed for an interval of 48 h. Subsequently, the temperatures were taken at the places of the characteristic sections, which correspond to the places where the measuring devices are located in the monitored structure. These results were compared with the outputs of the experimental measurements, where the average temperature deviations obtained by numerical simulation and measurement were determined; see Table 2. A graphical comparison of the results is shown in Figures 7 and 8. The results show that the deviations between numerical simulation and the measurement are very small, especially in the place of the structure in the axis between the beams, where there is no deformation of the temperature field. The temperature differences for the 10 evaluated positions in the characteristic section of the structure range from 0.1 to 1.4 • C. With regard to possible inaccuracy of the experimental measurement it is noted that these values are very low and a method of using numerical simulation of temperature fields can be regarded as a suitable tool for the prediction of thermal processes in lightweight external wall structures. At the same time, however, when comparing the results, the error of numerical calculation is also noted, which arose by simplifying the boundary conditions that affect the structures, when only the temperature in the interior and exterior was entered without taking into account other influences such as solar radiation. Particularly the effect of solar radiation is evident in Figure 7, when the thermal response of the structure to solar radiation at temperatures determined experimentally is apparent. Accordingly, it is necessary to take this effect into account to achieve greater numerical calculations accuracy-a point that will be the subject of further research.
The second part of the paper deals with the influence of the air gaps on the thermaltechnical parameters of external wall structures. This section interprets only the results of the experimental measurements. The results of the experimental measurements for the evaluated lightweight external wall structures are shown in Figures 9 and 10. The results of these experimental measurements brought to light interesting findings. It has been shown that the external wall with a ventilated façade has a positive effect on the structure itself, as it is not subjected to temperature fluctuations, as is the case with an external wall with contact insulation, which is directly exposed to the external environment. The frontal façade effectively protects the external wall structure from low temperatures in the winter, which at the same time allows enhanced removal of moisture from the structure due to the higher temperatures inside the air gap. This is not the case with the evaluated double-skin roof structure, when a fundamental phenomenon was recorded, namely the subcooling of the air inside the ventilated air gap in the winter as the result of the radiation of the roof surface into the clear night sky. At very low ambient temperatures, temperatures lower by more than 5 • C were found inside the air gap compared to the outside air temperature. Subcooling of the surface of the inner skin is taking place at the same time. A cause of this phenomenon is also an insufficient air flow inside the ventilated air gap due to its inadequate thickness, which was the case of the monitored structure. The behavior of temperatures inside ventilated air gaps is the subject of further research. Follow-up research should also include the use of the simulation numerical methods, which have been evaluated as suitable to be used for the simulation of thermal processes in structures, taking into account other recorded influences that in addition to outdoor temperatures affect structures noticeably as well.

Conclusions
The subject of this research was the thermal analysis of timber-based lightweight external wall structures under winter conditions. The evaluated structures are part of the experimental building constructed in the passive standard, which is located in Ostrava, Czech Republic. The building is designed as a timber prefabricated panel structure and thanks to its exceptional nature, allows a long-term monitoring and evaluation of physicaltechnical properties of its structure. This paper focuses on its double-skinned structures, i.e., structures with a ventilated air gap. Attention is specifically focused on external walls with a ventilated facade and a double-skin roof structure. The data used by experimental measurements of the temperatures inside these structures and temperatures taken inside the ventilated air gaps are used for the analysis; to accomplish a comprehensive evaluation the temperatures of the indoor and outdoor environment have been taken into account as well. Numerical methods for calculating the course of temperature variations in structures under non-stationary boundary conditions are shown to be a suitable tool for predicting temperature processes in the lightweight external wall structures; however, all factors that affect these processes must be taken into account. In the case of structures with ventilated air gaps, the influence of heat emission into the cold night sky as well as the influence of solar radiation were monitored as part of the experimental measurements and should become part of further thermal technical analyzes using numerical simulations.