Experimental analysis of insulating materials using Guarded Hot Box - Preliminary results

The containment of energy consumption in the construction sector strongly depends on the envelope, which is mainly responsible for heat loss in buildings. Thus, great attention should be paid to the selection of thermally-performing materials. In this work, preliminary results of three different configurations of walls in real size have been compared, conducting the analyses with the heat flow meter method inside the Guarded Hot Box apparatus, capable of guaranteeing repeatable and controlled conditions. The analyses were carried out with two types of heat flow meters, characterized by different sensitivity. The side of the wall facing the hot chamber has been insulated with rock wool for all the experiments, while the other side has been insulated first with expanded polystyrene (EPS) with graphite, then with hemp, and finally with cork. The results showed that the sample with the best thermal behaviour is the first one, i.e., the wall with EPS with graphite, characterized by a transmittance value between 0.148 W/m2K and 0.153 W/m2K. The other two configurations, characterized by the use of natural materials, showed worse performance with conductance values about 20% higher than EPS. The percentage differences between the two heat flux sensors for the experimental campaigns ranged from 2.8% to 4.4%.


Introduction
Energy use in buildings increased from 115 EJ in 2010 to almost 135 EJ in 2021, representing 30% of global final energy consumption.As emerges from these data from the International Energy Agency's September 2022 report [1], it is extremely important to invest in energy efficiency interventions in the building sector to align with the Net Zero Emissions scenario by 2050 [2].The continuous attention paid to building energy efficiency has led to the development of increasingly complex construction solutions capable of satisfying the requirements of current regulations in force.As the complexity of building components increases, it is necessary to grow the investigation criteria and develop measurement and analysis systems as accurately as possible.In this context, therefore, research and experimentation assume a particular and strategic importance.
In the last years, the Guarded Hot Box apparatus is attracting a lot of attention among researchers.This instrument allows the measurement of the transmittance of building components by recreating controlled thermal conditions inside its chambers.Various studies have been conducted on the applicability of the GHB method for calculating thermal parameters in construction.
Chowdhury et al. [3] used a Guarded Hot Box to explore the thermal characteristics of walls and roofs commonly used in Indian residential construction.Bai et al. [4] integrated the numerical analyses carried out with the finite element method with experimental analyzes conducted in Hot Box to study the thermal properties of a wall made of hollow shale blocks.Trgala et al. [5] have experimentally analyzed various building envelopes potentially suitable for thermal insulation.The measurements were carried out under dynamic conditions in the cold chamber of the Hot Box.The study shows that the accuracy of the results seems to increase with the length of the measurements.Ferrari et al. [6] tested the thermal performance of four different walls.For the study, a single climatic chamber (considering the laboratory room where the chamber was placed as the indoor space) was used.The measurements were conducted under dynamic conditions.de Rubeis et al. [7] presented a novel integrated measuring and control system for hot box experiments.In this work, the thermal parameters of a doubly insulated X-lam wall sample were measured by keeping the thermal conditions inside the chambers under control, thanks to the use of 135 probes which constantly recorded the temperature values during the 72 hours of measurements.The results obtained experimentally with the heat flow meter method (HFM) showed a difference of less than 20% from the theoretical values.In another study, Nardi et al. [8] evaluated the thermal parameters of a sample wall placed in a Guarded Hot Box.The results obtained through a thermographic investigation were compared with heat flow meter measurements and theoretical values.
The aim of the present work is to compare the thermal performance of three wall samples with real dimensions of 300 cm x 300 cm, using two different heat flow meters (Hukseflux HFP01 and HFP03) inside the Guarded Hot Box, under repeatable and controlled conditions.The side of the wall facing the hot chamber has been insulated with rock wool for all the experiments carried out, while the other side has been insulated first with expanded polystyrene (EPS) with graphite, then with hemp, and finally with cork.Thanks to the Hot Box, known thermal conditions were imposed, with the aim of having a temperature difference that can give rise to an appreciable thermal flow.In this specific case, stationary analyses were carried out to understand which of the three wall configurations was the most performing.

Materials and Methods
In this work, experimental analyses were conducted to compare three wall configurations characterized by the use of different insulating materials on the outer surface of the wall: i) EPS, ii) hemp, iii) cork.

Methodology
The present study has a twofold objective: i) to analyze the thermal characteristics of the insulating materials under examination, ii) to compare the results obtained with two different heat flow meters.
The methodology used for this study can be divided into two macro-phases.
The first phase focuses on the creation of the three wall samples (S1, S2, S3), after the insulating materials selection through a scientific literature review.
The second phase, i.e., the analysis phase, focuses on the study of the thermal behaviour of the three wall configurations through theoretical and experimental analyses conducted in Hot Box using the HFM method.

Creation of the wall samples
The specimen has actual dimensions of 300 cm x 300 cm and is made up of a double-insulated X-lam loadbearing element known as CLT (Cross Laminated Timber).For the first experiment, the internal surface of the wall was insulated with rock wool, while the external surface was insulated with EPS mixed with graphite.The wall was sandwiched between double layers of plasterboard (Figure 1a).For the second experiment, EPS with graphite was replaced with hemp panels, keeping the thickness unchanged (Figure 1b).In the last cycle of measurements, the insulating layer on the cold side was replaced with cork panels, keeping the thickness unchanged (Figure 1c).

Analysis phase
The analyses of the three samples were carried out via theoretical and experimental approaches.

Theoretical approach
The theoretical approach was carried out under steady and one-dimensional conditions.
The calculation of the total resistance (  ) has been carried out through Ohm's law, expressed by the Equation (1), based on the analogous electrical circuit: where: - , and  , are the internal and external surface resistances [m 2 K/W], equal to 0.13 m 2 K/W and 0.04 m 2 K/W respectively, as indicated by EN ISO 6946 standard [26]; - , is the conductive thermal resistance of each layer [m 2 K/W].
Table 1 shows the various layers of the three samples with their thermal properties.Therefore, based on Equation (1) and on the thermal properties of the samples' layers, the total thermal resistance calculated for the samples is 5.69 m 2 K/W for (S1), 4.96 m 2 K/W for (S2) and 4.90 m 2 K/W for (S3).
Known the total resistances of the walls, the thermal transmittance was calculated using Equation (2): Finally, by imposing a temperature difference of 20 °C between the inner and the outer surfaces of the walls, the heat flux passing through each sample was calculated using Equation (3).
Table 2 shows the thermal transmittance and the heat flux values calculated for the three samples.

Experimental approach
The Hot-Box used for the experimental analyses is the GHB (Guarded Hot-Box) type.In the Hot-Box with a guard chamber, the metering box is surrounded by a chamber that has the control function (guard box).A detailed description of the apparatus can be found in previous works [8,27,28].By using the GHB approach, it is possible to set and control the temperatures within the chambers, guaranteeing repeatable and controlled conditions.At the same time, it allows the evaluation of the influence of boundary conditions on the thermal behaviour of the analyzed building component.Through the use of electrical resistances and the introduction of thermal energy into the hot chamber, the temperature was brought up to the set point value of 20 °C.In the cold chamber, on the other hand, the thermal conditions were regulated by a refrigeration unit which allowed the set point value of 0 °C to be reached.During the experimental phase, the HFM method was used.The analyses were carried out using two different heat flux probes, installed on the internal surface of the specimen, two temperature probes, one of which applied to the cold side wall and the other to the hot side wall, and a data logger.The recommendations provided by ISO 9869 are followed [29].The measurements were carried out using Hukseflux HFP01 and HFP03 heat flow meters, whose characteristics are summarized in Figure 2. Hukseflux HFP01 is one of the most commonly used heat flow meters, while the HFP03 can be considered the higher sensitivity version of the HFP01 probe.Although based on the same technology, the two sensors differ mainly in sensitivity, accuracy, and measurement area (equal to 8 x 10 -4 m 2 for HFP01 and 64 x 10 -4 m 2 for HFP02).
The schematization of the experimental setup and the main characteristics of the measuring instruments are shown in Figure 2. specified in ISO 9869 standard [29], was utilized to process the data previously collected.In detail, the conductance (Λ) and thermal resistance (R) values have been obtained utilizing Equations ( 4) and ( 5), respectively: - is the progressive sum of the differences between the internal and external surface temperatures [°C]; is the progressive sum of the density of the heat flux [W/m 2 ].Finally, the transmittance value (U) was calculated according to Equation (2).

Results
Two tests were conducted on each wall configuration: the first test was performed using the HFP01, the second using the HFP03.Table 3 shows the values obtained from the experimental analyses.Table 3. Results of the experimental campaigns.The uncertainty analysis and propagation of uncertainty in the conductance values were carried out following Holman's method [30], using Equation ( 6): 1/2 (6) where: -  is the uncertainty in the conductance value; -  ,   , and   , are the uncertainties in the independent variables q,  , and  , .
Figure 3 shows the results obtained using the two different heat flow meters for each of the wall configurations.The graphs contain the measurements made in the last 72 hours of each measurement cycle.A comparison of the transmittance values obtained with the two HFM for each sample shows that HFP03 always gives slightly lower values than HFP01, with percentage differences ranging from 2.8% to 4.4%, as shown in Figure 4.

Conclusions
The aim of the present work was to compare the thermal performance of three wall samples with real dimensions of 300 cm x 300 cm, characterized by the presence of different insulating materials: i) EPS with graphite, ii) hemp, and iii) cork.The analyses were conducted using two heat flow meters (Hukseflux HFP01 and HFP03) in laboratory conditions thanks to the Guarded Hot Box apparatus.The starting point of the study was the knowledge of the materials' characteristics, the instruments' technical specifications, and the theoretical values, which made possible a comparison with the results obtained through the experimental analyses.
The sample that showed the best thermal behaviour is the first one, i.e., EPS with graphite.However, the results obtained for the other two wall configurations, i.e., hemp and cork, are promising, also considering that these materials are natural, recyclable and with low environmental impact.Moreover, the comparison of the two measuring instruments, characterized by different measurement area, sensitivity, and accuracy, showed that the more sensitive instrument (i.e., HFP03) always provided slightly lower conductance values than those obtained with HFP01.
This preliminary study offers several research points.The analyses of insulating materials already present on the market are an excellent basis for undertaking a line of research aimed at discovering innovative solutions that also consider unusual insulating materials under development, if they demonstrate that they have the necessary requirements to make a significant gap in the energy field.Furthermore, the measurement system could also be deepened and improved.The next step will be integrating the analyses already carried out in this work with dynamic regime analyses, to simulate the energy behaviour of the building as close as possible to the real behaviour.

Figure 2 .
Figure 2. Experimental setup's configuration, sample wall's detail (dimensions in millimeters), and technical specifications of the measuring instruments.The measurements had a total duration of three weeks.They were carried out from 7 March 2022 at 1:00 p.m. to 14 March at 10:00 a.m. for the first sample (S1), from 16 March at 1:00 p.m. until 23 March at 11:30 a.m. for the second (S2) and from 28 March at 1:30 p.m. to 4 April at 8:50 a.m. for the last sample (S3).The data logger acquired the values with a sampling interval of 10 minutes.The progressive average method,

Figure 4 .
Figure 4. Transmittance values obtained with the HFP01 and HFP03 and relative percentage difference.

Table 1 .
Properties of the samples' layers from cold chamber to hot chamber.

Table 2 .
Theoretical transmittance and heat flux values of the three samples.

Table 4
illustrates the percentage differences between the theoretical and experimental transmittance values for the tests conducted.

Table 4 .
Comparison between theoretical and experimental transmittance values (values in [W/m 2 K]).