ENVIRONMENTAL THERMAL COMFORT OF A REDUCED MODEL USING CROSS-LAMINATED TIMBER

The challenges for the use of the cross-laminated timber (CLT) system in the Brazilian agricultural market are significant. This study evaluated the thermal performance of fiber cement tiles associated with a CLT non-conventional structure compared to those of ceramic, fiber cement and aluminum roof tiles based on following thermal comfort indexes (i.e., black globe humidity index (BGHI), radiant heat load (RHL) and specific enthalpy) using physical conventional models of reduced-scale rural facilities under summer conditions. The non-conventional CLT model comprised closing walls and a lining that form a self-supporting structure with few air inlets. This model presented reduced thermal comfort indexes compared to the other conventional roofs. Moreover, the CLT model has an average black globe temperature (Tbg) of 32.9 °C, which was lower at all times compared to those


INTRODUCTION
Thermal sensation (i.e., cold or hot) depends on the building material properties, including its surface temperatures and the facility of heat exchange with the medium. When exposed to solar radiation and other sources from the outside, tiles and walls cause hot environments internally, which consequently produces thermal discomfort. Traditional roofing materials are already relatively well known in buildings and are commercially available. However, new and unconventional materials must be evaluated to assess their thermal performance in agricultural facilities.
A constructive technology, called cross-laminated timber (CLT) or cross-cut laminated wood, has begun attracting interest in Brazil because of its industrialization and sustainable characteristics, also considering the country's great lumber potential. Adekunle & Nikolopoulou (2018), Leiva et al. (2018), and Kollár & Štefko (2016) evaluated buildings using the CLT technique, aiming at thermal comfort in urban environments occupied by people; they obtained satisfactory results. However, studies on the use of this technology in agricultural facilities are scarce.
Engenharia Agrícola, Jaboticabal, v.40, n.4, p.413-419, jul./aug. 2020 In this context, this study aims to evaluate the thermal performance of a fiber cement roof associated with the CLT structure compared to those of clay, fiber cement, and aluminum roof tiles in reduced models of rural facility based on the thermal comfort indexes of BGHI, RHL, and effectiveness under the conditions of the summer season temperature.

MATERIAL AND METHODS
This study was conducted in the experimental area of Construction and Ambience in the Center of Agroveterinary Sciences of the Santa Catarina State University -Lages campus (Santa Catarina, Brazil), which has geographical coordinates of 27° 49′ latitude south and 50° 20′ longitude east, an average altitude of 940 m, and a Cfb type of climate (i.e., mesothermic humid or temperate according to the Köppen Climate Classification System). Normally, the area does not have a well-defined dry season, with January as the warmest month (average annual temperature: 24 °C). Present rain mainly comes in the summer season. Moderate winds prevail in the northeast direction.
Four models, which were constructed in a distorted reduced scale according to the similitude theory, were used (Sampaio et al., 2018;Jentzsch et al., 2013) on 1:10 and 1:2 scales in the horizontal and vertical dimensions, respectively, with internal measures of 3.00 m length, 1.40 m width, and 1.50 m ceiling height.
Aside from allowing the testing of different construction configurations, using the reduced physical models based on the similitude theory to evaluate different roofing materials has other advantages, including low costs of materials and labor (Akamine & Passini, 2017).
The models were based on the usual dimensions of poultry sheds (i.e., east-west direction and built away from each other to avoid shadowing). The lateral models (northsouth direction) were closed with two layers of transparent plastic to greenhouse formed by a 100 μm film. Other possible openings, which were mainly in the roof, were closed with polyurethane foam to avoid maximum thermal exchanges by the wind. The surroundings were formed by thin grass and gravel. A three-model structure was formed by bricks covered with mortar, concrete floor, and two-story roof of a wood structure. The fourth model was formed using the CLT system. Figure 1 displays all the models used in this study.
TfcCLT -corrugated fiber cement tiles mounted on the structural lining of the CLT formed by a self-supporting structure; TCN -ceramic tiles; TFCN -corrugated fiber cement tiles; and TAL -corrugated aluminum tiles. The CLT comprised three boards with 2.0 cm thickness each, which were crossed and glued with reactive polyurethane and sewn in the desired size to form a self-supporting structure of 6.0 cm thickness ( Figure 2). Note that this is the first study that used CLT plates in agricultural construction and so the CLT plates had three layers, which is the minimum number required by the production technique for the plates to be compatible with the mechanical resistance necessary for distorted-scale constructions.  respectively. The roof structure was built using timber with an average inclination of 25% of tiles. The external air velocities were obtained by a portable digital thermo-anemometer (Model Itan-720), which presented an air velocity scale between 0.4 and 45.0 m.s −1 , accuracy of ±3% reading + 0.2 m.s −1 , and resolution of 0.1 m.s −1 . The variations of the temperature and the humidity of the air and the black globe temperature (Tbg) inside and outside the models were obtained with a digital black globe thermometer (Model ITWTG 2000/Instrutemp). The globe of this instrument was formed by a hollow brass sphere with 75.0 mm diameter. The wet bulb temperature (Twb) was obtained with a sling psychrometer (rotating handle).
The instruments were placed in the geometric center of each reduced-scale model , Silva et al., 2015b at 0.75 m floor height, as well as in the external environment. This height was chosen because it showed the geometric center of the model, thereby determining a comparison by higher heat (Figure 3). The measurements were performed in 13 typical days during the summer season. The data were collected in January (10 days) and February (3 days) of 2019 based on the higher specific enthalpy values obtained at hourly intervals between 9:30 a.m. and 5:30 p.m. The days chosen for comparison had few or no incidence of clouds and rains and corresponded to the days of maximum heat. The BGHI, RHL, and Tbg indices were used to evaluate the thermal environments. The external enthalpy of the air (H) (Silva et al., 2015b) was calculated to determine the days with the highest "caloric content," which were considered as the days of greatest discomfort.
The BGHI, ε = BGHI (environment) − BGHI (type of tile)/BGHI (environment) − BGHI (aluminum tile) formula was used to calculate the effectiveness (ε) of each type of tile in relation to the aluminum tile. This index was used by Silva et al. (2015, a, b) to compare the roof tiles and the thermal comfort.
Randomized block design (daily hours), days as a repetition, and four treatments (roof type) were used to compare the thermal comfort indices. A no-shade external environment was used as the control. The data obtained were used for the analysis performed by means of the Ftest of the analysis of variance and Tukey's test at 5% significance. TfcCLT -corrugated fiber cement tiles mounted on the structural lining of the CLT formed by a self-supporting structure; TCN -ceramic tiles; TFCN -corrugated fiber cement tiles; and TAL -corrugated aluminum tiles. The means followed by different letters in the schedules indicate the statistical difference (P < 0.05) for each time by Tukey's test. The average values of the BGHI (i.e.,82.80,83.17,84.61,and 85.98) were registered for the TfcCLT, TCN, TFCN, and TAL treatments, respectively. The TfcCLT and TCN treatments were statistically equal between 11:30 am and 05:30 pm. The aluminum tile exhibited a lower BGHI value than that of the fiber cement tile. The BGHI registered in the external environment (i.e., no-shade environment) was 89.20, indicating the importance of roofing for thermal comfort. These results, in whole or in part, present a tendency and/or values similar to those obtained by Silva et al. (2015, a, b), Carneiro et al. (2015), and Akamine & Passini (2017) for conventional roofs.

RESULTS AND DISCUSSION
Considering BGHI = 75 as the limit value for the environmental thermal comfort of most animals cited in several studies, the tiles roofs presented a low thermal performance in hotter schedules.  used polypropylene linings on fiber cement and ceramic roofs and concluded that this technique contributed to the reduction of thermal discomfort in environments, regardless of the tile type. This was similar to the result found in the present study using the CLT lining. Figure 5 shows the development of the RHL index as a function of the schedules.
TfcCLT -corrugated fiber cement tiles mounted on the structural lining of the CLT formed by a self-supporting structure; TCN -ceramic tiles; TFCN -corrugated fiber cement tiles; and TAL -corrugated aluminum tiles. The means followed by different letters in the schedules indicate the statistical difference (P < 0.05) for each time by Tukey's test. The average values of the RHL were of 498.87, 502.59, 509.63, 519.08, and 639.12 W.m −2 for the TfcCLT, TCN, TFCN, TAL, and AMB treatments, respectively. This index responded similarly to the BGHI. The RHL in the no-shade environment was high. These results show a tendency and/or values similar to those obtained by Silva et al. (2015, a, b) and Akamine & Passini (2017).
The BGHI index is an important indicator of the environmental conditions; however, the thermal changes caused by the wind near the roofs and even inside the models should be avoided. Occurrence of such thermal changes may distort the results normally expected for the BGHI as compared to the RHL index, which considers the air velocity combined with other variables. According to Guimarães et al. (2014), animals kept in environments with a larger thermal amplitude may show a compromised performance because of having to adapt to the temperature variations in a short time. Figure 6 shows the development of the Tbg index as a function of the schedules. The Tbg index behavior during the day was similar to those of the BGHI and RHL indices, verifying again that the CLT registered the lowest values in the environment with roof.
TfcCLT -corrugated fiber cement clean tiles mounted on the structural lining of the CLT formed by a self-supporting structure; TCNceramic clean tiles; TFCN -corrugated fiber cement clean tiles; and TAL -corrugated aluminum clean tiles.  TfcCLT -corrugated fiber cement tiles mounted on the structural lining of the CLT formed by a self-supporting structure; TCN -ceramic tiles; TFCN -corrugated fiber cement tiles; and TAL -corrugated aluminum tiles. The external specific enthalpy (H) of the air (Silva et al., 2015b) was calculated to determine the days with the highest "caloric content," which are considered as the days of greatest discomfort. The models were closed; hence, the specific enthalpy inside these environments was larger in comparison to the external enthalpy of the air. For this variable, the CLT roof presented a greater enthalpy in schedules of greater heat. This result is attributable to the self-supporting characteristic of this roof structure, which also works as a lining that stores heat. This advantage cannot be observed with the ceramic roof.
Compared to aluminum, the roof tiles with the same thermal performance presented a thermal effectiveness (ε) equal to 1. The roof tiles with a better performance showed values higher than 1, whereas those with the worse performance had values lower than 1. The highest values of effectiveness were found in the TCN (1.89), TFCN (1.44), and TfcCLT (2.01) treatments. These values are similar to those obtained by Silva et al. (2015, a, b), albeit with minor differences, because the discrepancy has several causes, such as type, thermal characteristic, and state of conservation of the tiles in question, external air circulation in the instruments located inside the models, and the characteristics of the local climate at that moment.
The discussions and comparisons in this study can be further developed if more specific publications in the literature related to the thermal comfort of the CLT for rural use are available.

CONCLUSIONS
In terms of thermal comfort, conventional roofs were ranked herein as follows from best to worst: ceramic, fiber cement, and aluminum. The reduced-scale model with a fiber cement roof associated with the CLT showed reduced indices of thermal comfort compared to the other conventional roofs. The CLT showed potential for use in rural buildings due to the thermal comfort it presented and its advantage of being self-supporting. However, challenges still remain, among which is the need for investments, new research, and further knowledge about this material and other aspects, particularly, the acceptance of this technological innovation in the rural environment. Moreover, all evaluated models did not provide the desirable thermal comfort throughout the day for intensively reared animals, requiring the use of environmental control systems aimed at optimal thermal indices inside the premises.