Impacts of Design Parameters on the Thermal Performance of a Macro-Encapsulated Phase-Change-Material Blind Integrated in a Double-Skin Façade System

Abstract

Double-skin façades (DSFs) are promising sustainable design elements of buildings. However, they are prone to overheating problems in warm seasons due to high outdoor temperatures and intense solar radiation. Although phase-change material (PCM) blinds have proved to be effective at enhancing the thermal performance of DSFs, the impacts of the design parameters are crucial to the overall thermal performance of the system. This study focused on analyzing the impacts of design parameters on the thermal performance of a ventilated DSF system, which consisted of a macro-encapsulated phase-change material (PCM) blind with an aluminum shell. A simulation study was conducted using ANSYS Workbench FLUENT software, and the temperature distributions of the integrated system were compared with different blind tilt angles and ratios of cavity depth to blind width. The results show that both the blind tilt angle and ratio of cavity depth to blind width had a significant influence on the thermal performance of the DSF system. For instance, lower air-cavity temperatures within the range of 37~40 °C were achieved with the PCM blind at tilt angles of 30° and 60° compared with other selected tilt angles (0° and 90°). In terms of the cavity depth to blind width ratio, a ratio of 2.5 resulted in a lower air-cavity temperature and a better thermal performance by the DSF. With the optimal blind tilt angle and cavity depth to blind width ratio, the integrated DSF and macro-encapsulated PCM-blind system can reduce the cavity temperature by as much as 2.9 °C during the warm season.

1. Introduction

The building sector accounts for over 32% of final global energy consumption [1] and about 70% of total CO₂ emissions [2], with building envelopes playing a crucial role in the energy consumption process [3]. For instance, transparent building façades are widely applied due to their aesthetic appreciation and lighting advantages [4]. However, they can contribute up to about 40% of the total heat gain in buildings [5]. To this end, double-skin façade (DSF) envelopes have been studied and implemented, but they also tend to experience overheating issues in warm seasons [6]. There is, therefore, the need for the design parameters and material selection of DSFs to be carefully considered before implementation [7].

In recent years, many researchers have shown the potential of phase-change materials (PCMs) in reducing overall energy consumption by incorporating them into building fabrics and heating, ventilation, and air-conditioning (HVAC) systems [8,9]. For example, Yasiri et al. [10] analyzed the thermal performance of an integrated PCM roof and an external wall system and achieved energy consumption reductions of 25% and 40%, respectively. Kishore et al. [11] evaluated the energy-savings potential of integrated PCM buildings and obtained up to a 47.2% reduction in their annual heat gains. Mokhberi et al. [12] investigated the thermal-regulation-enhancement ability of PCM-filled void formers (VFs), which were incorporated into the structures between floors in multi-story office buildings. The results showed that the integrated PCM structure was able to perform better thermally by reducing heat transmission by 36.74% compared with a standalone insulation-filled structure. Rathore et al. [13] investigated a building envelope integrated with macro-encapsulated PCM in a tropical region and obtained a maximum temperature reduction of 4.3 °C in the indoor peak temperature throughout the year. Integrating phase-change materials (PCMs) into glazing envelopes can also enhance the thermal insulation of building envelopes, leading to improved energy efficiency and thermal comfort [14]. Li et al. [15] developed a thermal model for analyzing glass-envelope buildings incorporated with multiple PCM layers. Their results demonstrated that the PCM layers were able to reduce internal surface temperatures by 1.50 K and 3.54 K under peak heat-flow rates of 1.26 W/m² and 2.62 W/m², respectively.

While incorporating PCMs into glass walls can improve indoor thermal comfort and energy performance, it may also reduce indoor daylight levels compared with traditional shading devices. To address this, integrated PCM-blind systems have been explored in previous studies. Weinlöder et al. [16,17] conducted experimental investigations and found that the light transmittance in integrated PCM double-skin façades can reach 0.4. Furthermore, the inclusion of PCM in a south-facing façade reduced heat loss and solar gains by approximately 30% and 50%, respectively. Alawadhi et al. [18] used finite element methods to simulate an integrated PCM window-blind system and reported a 23.3% reduction in heat flow, as well as a reduction in the peak indoor air temperature, ranging from 0.2 °C to 4.3 °C. Nevertheless, the integration of PCM blinds into DSFs appears to be a promising strategy for enhancing thermal performance. Silva et al. [19] evaluated the potential of an integrated PCM-blind system in a DSF and obtained reductions in the maximum and minimum temperatures by 6% and 11%, respectively. In our previous works [20,21], we developed a micro-encapsulated PCM-blind system and evaluated it in a DSF. The results showed an encouraging average air temperature of below 35 °C and an average air-cavity temperature reduction of 2.2 °C.

To provide a comprehensive overview of the current state of research in this field, Table 1 summarizes the applicable seasons, encapsulation methods, and key design parameters for phase-change-material (PCM)-integrated building envelopes. Despite their potential, several challenges persist in the implementation of these systems, as discussed below.

Existing research has predominantly focused on PCM performance under winter conditions or within specific temperature ranges. While PCM-integrated blinds have demonstrated effectiveness in mitigating overheating by absorbing excess heat from the double-skin façade (DSF) cavity during peak warm seasons, their performance during mid-season conditions—when intermittent overheating may still occur—remains underexplored. This research gap underscores the need for further investigation into PCM system behavior across a wider range of climatic conditions to ensure year-round efficacy.

The selection of an appropriate PCM encapsulation method presents another critical challenge. Various techniques, including direct mixing, immersion, shape stabilization, macro-encapsulation, and micro-encapsulation, have been investigated [22]. While microencapsulated PCMs offer certain advantages, they are limited by high costs and potential supercooling effects. In contrast, macro-encapsulation has emerged as the most widely adopted method in construction applications [23]. For instance, Li et al. [24] reported a 13.88% reduction in energy consumption fluctuations during extreme cold conditions using macro-encapsulated PCM blinds within a sunspace. Similarly, Rehman et al. [25] observed a 67.3% reduction in energy associated with interior window surface temperatures when employing a PCM shutter system. Nevertheless, macro-encapsulation is not without drawbacks, including low thermal conductivity and potential leakage issues [26], though these can be partially mitigated through advanced sealing techniques.

Following the selection of macro-encapsulation for PCM-blind development, the choice of shell material and optimization of the design parameters become pivotal in enhancing thermal performance. Prior studies have frequently utilized opaque plastic materials for encapsulation; however, their low thermal conductivity restricts the overall system efficiency. For example, Abdulmunem et al. [27] employed polyvinyl chloride (PVC) panels, while Atiganyanun et al. [28] utilized silicon dioxide particles encapsulated via the sol-gel process. Notably, research on high-thermal-conductivity materials, such as aluminum, for macro-encapsulation remains limited [29,30]. Further exploration in this area is warranted, as materials with superior thermal conductivity could substantially improve PCM systems’ performance.

Table 1. Summary of the applicable season, encapsulation method, and design parameter of PCM-integrated envelopes.

Season/ Temperature	Encapsulation Method	Shell Material	Design Parameter	Reference
Winter	Macro-encapsulated	Aluminum alloy	-	[24]
Winter	Attach to the floor panel	-	Thickness	[31]
10 °C	Macro-encapsulated	Polymer	-	[32]
Winter and summer	Macro-encapsulated	Concrete wall	-	[33]
Summer	Macro-encapsulated	Polyvinyl chloride (PVC)	Dosage of PCM	[27]
Spring and autumn	Attach to the ceiling panel	-	Panel volume	[34]
27 °C	Macro-encapsulated	High-density polyethylene (HDPE)	Length and thickness	[25]
17 °C~30 °C	Macro-encapsulated	Dendritic glass	Geometrical configuration	[35]
40 °C	Micro-encapsulated	Silicon dioxide	Incorporation of PCM	[28]
65 °C	Micro-encapsulated	-	Dosage of PCM	[36]

Table 1. Summary of the applicable season, encapsulation method, and design parameter of PCM-integrated envelopes.

Season/ Temperature	Encapsulation Method	Shell Material	Design Parameter	Reference
Winter	Macro-encapsulated	Aluminum alloy	-	[24]
Winter	Attach to the floor panel	-	Thickness	[31]
10 °C	Macro-encapsulated	Polymer	-	[32]
Winter and summer	Macro-encapsulated	Concrete wall	-	[33]
Summer	Macro-encapsulated	Polyvinyl chloride (PVC)	Dosage of PCM	[27]
Spring and autumn	Attach to the ceiling panel	-	Panel volume	[34]
27 °C	Macro-encapsulated	High-density polyethylene (HDPE)	Length and thickness	[25]
17 °C~30 °C	Macro-encapsulated	Dendritic glass	Geometrical configuration	[35]
40 °C	Micro-encapsulated	Silicon dioxide	Incorporation of PCM	[28]
65 °C	Micro-encapsulated	-	Dosage of PCM	[36]

In summary, while PCM-integrated building envelopes hold significant promise for energy efficiency, key challenges persist concerning seasonal adaptability, encapsulation methods, and design optimization. It can be seen from the reviews that existing studies have mainly focused on evaluating the thermal performance of PCM-integrated systems with varying positions of the PCM layer or different types of PCMs. Furthermore, critical design parameters of an integrated PCM-DSF system, such as the cavity depth to blind width ratio, require further investigation to optimize the thermal performance. There are few works that have thoroughly discussed the impacts of design parameters (like the ratio of the cavity depth to blind width) on the thermal performance of macro-encapsulated PCM blinds in DSFs. Addressing these limitations will be essential to fully realizing the potential of PCM systems in sustainable building-energy management.

Therefore, this study aims to investigate a macro-encapsulated PCM blind with an aluminum shell for DSF integration. The thermal performance and impacts of the design parameters of the integrated system are discussed to fill the gaps in existing knowledge on macro-encapsulated PCM blinds with an aluminum shell incorporated into DSFs. Firstly, numerical models were developed for a ventilated DSF with a macro-encapsulated PCM blind. The temperature distribution in the integrated system with various design parameters (blind tilt angles and cavity depth to blind width ratio) was simulated using ANSYS Workbench 2020 R2 FLUENT software. The optimal values of these crucial design parameters were identified through simulation and analysis. This study provides useful information for designers and researchers aimed at the optimization of PCM blinds incorporated into envelopes to achieve satisfactory thermal performances with optimized design parameters.

2. System Description and Development

2.1. System Description

Figure 1A illustrates a schematic of the configuration of the macro-encapsulated PCM integrated into the double-skin façade (DSF) system. The DSF operates under natural ventilation driven by buoyancy effects, with an air inlet at the base and an outlet at the top. The proposed PCM blind is positioned within the DSF air cavity at a tilt angle to optimize solar radiation absorption. Each blade of the blind consists of an aluminum shell encapsulating paraffin-based PCM. During the overheating hours (with high air-cavity temperatures), the PCM in each blade absorbs the excessive heat in the DSF cavity during the melting process. Subsequently, as the air-cavity temperature decreases, the stored heat is released via solidification and, ultimately, dissipated through the naturally ventilated DSF cavity.

Figure 1B shows the experimental test rig and installation spots for data collection. To obtain the surface temperature of the integrated system, a total of four PT100 sensors were installed on the external and internal surfaces of the outer and inner glass panes of the DSF, while fourteen PT100 sensors were mounted on the upper and lower surfaces of each blade in the blind system. To obtain the air-cavity temperature in the DSF, K-type thermocouples were fixed at appropriate locations. The DSF was naturally ventilated with air openings at the top and at the bottom of the external and internal glass wall. The developed PCM blind was installed in the DSF cavity with an original tilt angle of 30° for receiving the maximum amount of solar radiation.

2.2. PCM-Blind Development

Based on the design concept of the integrated macro-encapsulated PCM-blind system, the process of the system’s development is depicted in Figure 2. The procedure includes material selection, material characterization, and PCM-blade development. Firstly, historical weather data (ambient temperature and solar radiation) and air-cavity temperature data on a real DSF system was collected in 2022 and analyzed to identify a suitable PCM melting temperature. The analytical hierarchy process (AHP) method was adopted to select candidate PCMs. The selection criteria included thermo-physical properties, stability, safety, and cost of commercially available paraffin materials for developing the PCM blind for the DSF system. Then, during the material characterization stage, differential scanning calorimetry (DSC) tests were conducted on the candidate PCM samples to obtain the heat of fusion and the specific heat of the PCM during the melting and solidification processes. The specifications of the selected PCM are summarized in

Table 2. The specifications of the selected PCM.

Material		Property
Melting point (°C)		35 ± 2 °C
Heat of fusion (kJ/kg)		200 kJ/kg
Flash point (°C)		130 °C
Specific heat capacity (kJ/kg·°C)	Solid	2.12 kJ/kg·°C
	Fluid	1.98 kJ/kg·°C
Thermal conductivity (W/m·°C)	Solid	0.32 W/m·°C
	Fluid	0.30 W/m·°C
Density (kg/m3)	Solid	865 kg/m3
	Fluid	853 kg/m3

. Finally, the PCM blade was developed by utilizing a hollow rectangular aluminum-alloy tube for macro-encapsulation of the PCM. The calculated amount of melted PCM was injected into the aluminum-alloy tube, and, once solidified, the ends were sealed with epoxy adhesive and left undisturbed for over 48 h.

2.3. Data Collection and PCM Selection

Data collection was conducted in an experimental test rig placed on the roof of an office building at the University of Shanghai for Science and Technology, Shanghai, which is in a hot-summer and cold-winter region of China. Two stages of data collection were carried out, including one week of data collection (21~27 June 2022) during the overheating season on the aluminum-blind-integrated DSF for the PCM selection and data collection on a typical day (25 October 2022) during the warm season on the PCM-blind-integrated DSF for the boundary condition setup and numerical model’s validation.

Figure 3 presents the variations in the DSF air-cavity temperature with an aluminum blind from 21 to 27 June 2022. In general, the air-cavity temperature fluctuated between 25 °C and 45 °C, with the peak occurring at around 11:00 a.m. every day. Based on the average air-cavity temperature’s cumulative time, the longest duration was in the 25~30 °C range, followed by 30~35 °C, 40~45 °C, and 35~40 °C. Therefore, paraffin was selected for preparation of the PCM blind, and its technical specifications are listed in Table 2.

3. Numerical Model and Procedure

3.1. Heat Transfer Model

The heat transfer process of the integrated PCM blind and DSF system is demonstrated in Figure 4. The solar radiation that reaches the external glass wall can be divided into direct and diffuse solar irradiance [38]. A proportion of the solar radiation enters the room, while the remaining proportion is absorbed by the DSF-PCM-blind system. Long-wave radiation and convective heat transfer occur between the external glass wall and the ambient environment. The PCM blind, which is positioned in the DSF’s cavity, receives the solar radiation that penetrates through the external glass wall. Convective heat transfer between the air cavity and PCM blind, long-wave radiation between the adjacent blades, and long-wave radiation between the PCM blind and glass walls also take place in the DSF’s cavity [39]. The extra heat trapped in the DSF’s cavity can be exhausted by the natural ventilation.

In the establishment of the heat transfer model, we referred to our previous research [38], and it can be expressed by Equations (1)–(7).

$$Q_{net}=Q_{sol}-Q_{refl}-{Q_{conv}}_{\_o}-Q_{conv\_i}$$

(1)

$$Q_{\mathrm{sol}}=Q_{abs}+Q_{tra}+Q_{refl}$$

(2)

$$Q_{abs}=\alpha \times A_1\times G_{sol}$$

(3)

$$Q_{tra}=\tau \times A_1\times G_{sol}$$

(4)

$$Q_{net}=Q_{abs}+Q_{tra}-{Qconv\_i}_{conv\_o}$$

(5)

$$Q_b=Q_{tra}+Q_{rad\_b1}+Q_{conv\_b1}$$

(6)

$$Q_b=k_p{\displaystyle \frac{d{T}_p}{dx}}\left|x=0\right.$$

(7)

where Q_net is the net solar radiation, Q_sol is the total solar radiation, Q_refl is the reflected solar radiation, Q_{conv_o} is the convective heat transfer on the outer mglass pane, Q_{conv_i} is the convective heat transfer on the inner glass pane, Q_abs is the absorbed solar radiation of the DSF system, Q_tra is the transmitted solar heat radiation, Q_b is the absorbed heat of the blind, Q_{rad_b1} is the radiative heat transfer on the blind, Q_{conv_b1} is the convective heat transfer on the blind, G_sol is the total solar irradiance, α is the absorption coefficient, $\tau$ is the transmittance of the DSF outer glass skin, A is the area of the glazing wall (m²), k_p is the heat transfer coefficient of the PCM blind [$\mathrm{W}/\left(\mathrm{m}·\mathrm{K}\right)$], and T_p is the surface temperature of the blind (K).

In terms of numerical solutions for simulating the thermal behavior of PCMs, the apparent heat capacity method and the enthalpy method are commonly used [40]. The apparent heat capacity method treats the latent heat of the phase transition as a substantial apparent heat capacity within a relatively narrow temperature range, making it less suitable for modeling PCM systems with broader phase-change temperature zones. In contrast, the enthalpy method offers a superior solution for complex coupled heat exchange problems involving multiple media and boundaries. It establishes a unified energy equation across the entire region, including liquid and solid phases, as well as phase transition interfaces. It calculates the distribution of the thermal enthalpy, allowing for the delineation of phase-change interfaces [41], and eliminates the trouble of tracking phase-change interfaces or separating solid and liquid phases for independent treatment. With its advantages of high accuracy and computational efficiency, the enthalpy method is particularly well-suited for modeling PCM blinds in DSF systems and is shown in Equations (8)–(10) [42].

$$\rho c{\displaystyle \frac{\partial T}{\partial \tau }}=\nabla ·(\lambda \times \nabla T)$$

(8)

$$\rho {\displaystyle \frac{\partial H}{\partial \tau }}=\nabla ·(\lambda \times \nabla T)$$

(9)

$$(\rho ,\lambda )=\left\{\begin{array}{c}{\lambda }_s,{\rho }_s H<H_s\hfill \\ {\lambda }_l,{\rho }_l H>H_l\hfill \end{array}\right.$$

(10)

where H is the latent heat (J), H_s is the latent heat of the solid PCM (J), H_l is the latent heat of the liquid PCM (J), $\rho$ is the density ($\mathrm{kg}/{\mathrm{m}}^3$), T is the surface temperature (K), $\lambda$ is the thermal conductivity [$\mathrm{W}/\left(\mathrm{m}·\mathrm{K}\right)$], ${\lambda }_s$ is the thermal conductivity of the solid PCM [$\mathrm{W}/\left(\mathrm{m}·\mathrm{K}\right)$], ${\lambda }_l$ is the thermal conductivity of the liquid PCM [$\mathrm{W}/\left(\mathrm{m}·\mathrm{K}\right)$], ${\rho }_s$ is the density of the solid PCM ($\mathrm{kg}/{\mathrm{m}}^3$), and ${\rho }_l$ is the density of the liquid PCM ($\mathrm{kg}/{\mathrm{m}}^3$).

3.2. CFD Model

CFD (computational fluid dynamics) is a commonly used method for simulating the thermal performance of a PCM-integrated system, since it can provide more detailed and accurate information compared with other methods [43]. This study utilized ICEM preprocessing, the pressure-based solver in ANSYS Workbench FLUENT software, and TECPLOT post-processing for analysis. The following considerations were made prior to the simulation:

• Only one-dimensional conduction was considered within each phase-change blade, while the convective heat transfer was neglected;
• The airflow within the DSF’s cavity was treated as a two-dimensional incompressible fluid;
• Long-wave radiation between the DSF-PCM-blind system and the surrounding environment were not considered;
• Viscous dissipation within the fluid flow was ignored by treating all physical parameters as constants, except for the fluid density;
• Density was only considered in relation to the momentum equation’s volume forces, while remaining constant in all other terms;
• Heat transfer through the DSF system to indoor heat gains was negligible.

3.2.1. Governing Equations

The Boussinesq approximation was used. The buoyancy effects caused by temperature differences were adopted to enhance the convergence of iterative computations. The density term in the gravity equation is represented by Equation (11), as follows:

$$\rho ={\rho }_0[1-\beta (T-T_0)]$$

(11)

where $\rho$ is the density of the fluid corresponding to $T$ ($\mathrm{k}\mathrm{g}/{\mathrm{m}}^3$), ${\rho }_0$ is the fluid density corresponding to $T_0$ ($\mathrm{k}\mathrm{g}/{\mathrm{m}}^3$), $T_0$ is the ambient temperature (°C), and $\beta$ is the thermal expansion coefficient.

The realizable k-ε model was chosen as the turbulence model, since it has been proven to be able to simulate heat transfer and airflow within DSF cavities [44]. For the radiation model, the discrete ordinates (DO) model was selected, since it can account for non-gray radiation, scattering effects, and radiation in semi-transparent media, which makes it suitable for modeling the DSF-PCM-blind system [45].

The PCM blind was simulated using the enthalpy method, and the energy equation can be described by Equations (12) and (13).

$$h=h_{ref}+{\int }_{T_{ref}}^T C_{P_{p}dT}$$

(12)

where h is the specific enthalpy (J/kg), h_ref is the specific enthalpy at the reference temperature (J/kg), $T_{ref}$ is the reference temperature (°C); $h_{ref}$ is the enthalpy at the reference temperature ($\mathrm{J}$); and $C_{P_p}$ is the specific heat capacity at a constant pressure ($\mathrm{J}/(\mathrm{k}\mathrm{g}·\mathrm{K})$.

The concept of the liquid fraction was identified in Fluent to identify the melting/solidification processes of the PCM, which is described by Equation (13).

$$\beta =\left\{\begin{array}{cc}0\hfill & T<T_S\hfill \\ {\displaystyle \frac{T-T_s}{T_l-T_s}} \hfill & T_s<T<T_l\hfill \\ 1\hfill & T>T_l\hfill \end{array}\right.$$

(13)

where $\beta$ is the liquid fraction, T is the PCM temperature (K), $T_S$ is the solidification temperature of the PCM (°C), and $T_l$ is the melting temperature of the PCM (°C). When 0 < $\beta$ < 1, the PCM presents a liquid–solid two-phase state.

3.2.2. Boundary Conditions

In terms of the boundary condition setups, the weather data collected on a typical day (25 October 2022) with high air-cavity temperatures were used. The inlet airflow velocity was obtained by two hot-wire anemometers. Climatic parameters including ambient temperature and wind speed were collected using a weather station. An Agilent-DAQ-973A data logging device was employed for data collection. All the sensors and equipment were calibrated before the experiment.

The boundary condition setups are demonstrated in

Table 3. The boundary conditions used in the model.

Boundary	Type	Value
Outdoor curtain wall	Wall	16 W/(m2·K)
Outdoor vent (upper)	Pressure inlet	0 Pa
Indoor curtain wall	Wall	3.6 W/(m2·K)
Indoor vent (lower)	Pressure outlet	0 Pa
Upper and lower wall	Wall	Adiabatic
Blind surface	Coupled	Transient state

. The convective heat transfer coefficient between the external glass surface and the outdoor air was set at 16 W/(m²·K), and the convective heat transfer coefficient between the internal glass wall and the indoor air was 3.6 W/(m²·K), which refers to the Chinese standard [46]. The upper and lower walls were insulated. The pressure—inlet and pressure—outlet were set up as the DSF inlet and outlet boundary conditions. The wall boundary conditions were set for the external and internal glass walls. The other walls were set up as adiabatic walls. Unsteady-state simulations were conducted with a time step of 1 s. For the temperature boundary condition’s setup, outdoor temperatures were fitted to the polynomial functions (Figure 5), and the profile files were created as UDFs and imported into FLUENT. For the solar radiation’s setup, both direct and diffuse radiation data, collected at 1800s intervals using the data logger, were used (Figure 5) to create a UDF file, which was imported into FLUENT. The specifications for the air, aluminum alloy, and glass are presented in

Table 4. Specifications of the material used in this model.

Material	Density (kg/m3)	Specific Heat Capacity (kJ/kg·°C)	Thermal Conductivity (W/m·K)	Absorption Coefficient	Reflectance
Air	1.225	1006.43	0.0242	-	-
Aluminum alloy	871	871	202.4	1200	0.82
Glass	670	670	0.76	1000	0.1

.

3.2.3. Independence Test

After the geometric model was developed, unstructured meshes were generated utilizing ICEM software. Four mesh resolutions were tested (100,895-, 203,067-, 301,655-, and 404,064-cell grids) to ensure the solution’s independence. The baseline mesh consisted of polyhedral cells with prism layers near the walls, refined iteratively based on velocity gradient thresholds. In Figure 6a, the simulated temperature profiles along the centerline in the DSF cavity at Y = 55 cm for the four different grid sizes were compared. The differences among the grid sizes with 200,000, 300,000, and 400,000 grid cells were minimal, with an average error of less than 1%. Therefore, the grid size consisting of 301,655 cells (Figure 6b) was selected for further simulation as a compromise between simulation accuracy and computational resources and time.

3.2.4. Model Validation

The model’s validation focused on the DSF system, since validations using room-level quantities (e.g., indoor air temperature) may lead to much higher uncertainty [47]. Therefore, the simulation’s results for the average air-cavity temperature were compared with the collected data on the PCM-blind-integrated DSF system on 25 October 2022, as presented in Figure 7. It is obvious that the trends of the measured and simulated average air-cavity temperatures are identical. The maximum values of the simulated and measured average air-cavity temperature are 40.6 °C and 40.5 °C, respectively.

In order to prove the reliability of the numerical models, the validation approaches in the ASHRAE Guidelines 14-110 [41] and the Measurement and Verification Guidelines 3.0 of the U.S. DOE [48] were adopted. According to these guidelines, the values of the mean bias error (MBE) and the root mean square error (RMSE) should be within the ranges of −10%~10% and −30%~30%, respectively. The MBE and RMSE can be calculated using Equations (14) and (15). By calculating the MBE and RMSE of the experimental data and the simulation results for the air-cavity temperature, it was found that the MBE value was 2.15%, and the RMSE value was 7.5%. The results indicate that both the MBE and RMSE values were within the recommended range suggested by the guidelines, which proved the reliability of the numerical model for simulating the DSF and PCM-blind system.

$$MBE={\displaystyle \frac{{\sum }_{i=1}^n (M_i-S_i)}{{\sum }_{i=1}^n M_i}}$$

(14)

$$RMSE=\sqrt{{\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n} (M_i-S_i)}$$

(15)

where $S_i$ is the simulated temperature, $M_i$ is the measured temperature, and $n$ is the number of comparison values.

4. Results and Discussion

4.1. System Performance on Typical Meteorological Day

Figure 8a presents the temperature contours of the DSF at the following various times: 8:00 a.m., 10:00 a.m., 12:00 a.m., 14:00 p.m., 16:00 p.m., and 18:00 p.m. on 25 October 2022. The air-cavity temperature first increased, peaked at 12:00 a.m., and then decreased as time passed, which corresponds with the outdoor temperature variations. Due to the presence of the PCM blind, the air-cavity temperature near the external glass skin could maintain a stable temperature range within 30~35 °C, which is much lower than the temperature range of the DSF with an aluminum blind.

Figure 8b shows how the liquid fraction of the PCM changes within the blades over time, and Figure 8c displays the daily average liquid-phase fraction of the PCM blade. The melting process of the PCM occurred from 8:30 a.m. to 19:30 p.m. A gradual rise in the liquid-phase fraction could be witnessed from 8:30 a.m. to 13:00 p.m., which demonstrates that the macro-encapsulated PCM was able to absorb excessive heat in the DSF cavity, maintaining the cavity temperature within the desired range of 37~39 °C during the melting period. From 13:30 p.m. to 19:30 p.m., the liquid-phase fraction steadily decreased to 0 along with the solidification process. During this process, the PCM continuously released the absorbed heat and delayed the temperature drop in the DSF cavity. The highest liquid-phase fraction value was 0.86, which occurred at 13:30 p.m., which indicates that an appropriate quantity of the PCM was used.

4.2. Influence of the Blind Tilt Angle on the System’s Thermal Performance

In order to reveal the influence of the blind tilt angle on the system’s thermal performance, the following four angles were considered: 0°, 30°, 60°, and 90°. Figure 9 displays the temperature distributions within the DSF cavity at 8:00 a.m., 10:00 a.m., 12:00 a.m., 14:00 p.m., 16:00 p.m., and 18:00 p.m. When the blind tilt angle was 0°, the DSF air-cavity temperature was obviously higher than the system with other angles. The maximum air temperature reached 45 °C, which is probably due to its worst solar shading effect allowing more direct solar radiation to gather in the DSF cavity. When the blind tilt angle was 90°, a large temperature difference could be found between the external and internal glass skins. The surface temperature of the external glass skin was about 47 °C, while the temperature of the internal glass skin was 35 °C due to an effective shading effect. In general, the DSF with blind tilt angles of 30° or 60° was able to maintain the air-cavity temperatures within the range of 37~40 °C.

Figure 10 demonstrates the average air-cavity temperature profiles of the integrated macro-encapsulated PCM blind and DSF system with various blind tilt angles. The temperature trends are consistent with the blind tilt angles. The 0° PCM blind had a notably higher peak temperature at 45.2 °C, while the system with blind tilt angles of 30° or 60° maintained a stable air-cavity temperature within the range of 38~40 °C. Although the 90° PCM blind also achieved a lower air-cavity temperature of about 36 °C, this would weaken the indoor lighting to a large extent. Based on the above analysis, both the 30° and 60° tilt angles are favorable choices for a PCM blind in a DSF cavity.

Figure 11 and Figure 12 show the profiles of the average liquid fraction of the PCM blind with different blind tilt angles. The phase-change process occurred between 8:00 a.m. and 18:00 p.m. for the PCM blind with different angles, with the peak values of the liquid fraction occurring around 13:00 p.m.~14:00 p.m. in the afternoon. In descending order of blade tilt angles (90°, 60°, 30°, and 0°), the corresponding peak liquid fraction values are 1, 0.93, 0.86, and 0.61, respectively. When the PCM was solid, the liquid fraction was first 0 and then gradually increased as it melted into a liquid state. A liquid fraction between 0 and 1 indicates a liquid–solid mixed-phase state. With a tilt angle of 90°, it took roughly an hour for the PCM to entirely melt into a liquid state. For the PCM blind with tilt angles of 60° and 30°, the liquid fractions of the PCM were similar, reaching peak values near 1.00 after sufficient heat storage (peaking at 0.93 and 0.86, respectively), indicating an appropriate PCM selection and quantity. The PCM blind with a tilt angle of 0° only achieved a maximum liquid fraction of 0.61.

4.3. Influence of the Cavity Depth to Blind Width Ratio on the System’s Thermal Performance

Figure 13 compares four different ratios of cavity depth (X) to blind width (L), which were X = 1.5 L, X = 2 L, X = 2.5 L, and X = 3 L. The temperature distributions in the DSF cavity with X = 1.5 L, 2 L, and 2.5 L were quite similar. The average air-cavity temperature remained relatively stable between 10:00 a.m. and 14:00 p.m., fluctuating within the range of 35~42 °C. The stable air-cavity temperature resulted from the effective heat absorption by the PCM blind, which also corresponds to the temperature profiles in the DSF cavity. However, in the DSF cavity with a cavity depth to blind width ratio of X = 3 L, the air-cavity temperature fluctuated significantly between 10:00 a.m. and 16:00 p.m. At 14:00 p.m., the air near the external glass skin and the blind’s surface reached about 45 °C, while the air around the internal glass skin exceeded 40 °C. This indicates that the PCM blind struggled to reduce the cavity temperature effectively. This is probably due to the fact that the dimensions of the PCM blind had reached their upper limit for absorbing extra heat in a DSF cavity with a comparably large depth. Additionally, the larger DSF cavity size would result in a lower airflow velocity, which may weaken the convective heat transfer on the blind’s surface.

According to Figure 14, the differences among the air-cavity temperatures near the internal glass skins with different cavity depth to blind width ratios were relatively small. As the cavity depth to blind width ratio increased, the air temperature near the external glass skin rose. Specifically, for X = 1.5 L, 2 L, and 2.5 L, the air-cavity temperatures were maintained below 42 °C, which indicates the effectiveness of the PCM blind in temperature adjustment. However, for X = 3 L, the front-side cavity temperature peaked at 46 °C, suggesting that the PCM blind struggled to efficiently absorb excessive heat from the air cavity.

Based on the above analysis, Figure 15 compares the average air-cavity temperatures between the original experimental setup and the optimized design parameters. It demonstrates that adjusting a cavity width ratio of X = 2.5 L could effectively reduce the air-cavity temperature and eliminate the overheating effect during the daytime mid-season. The maximum temperature may decrease by 2.9 °C and the average cavity temperature could drop by 1.2 °C during the time period from 7:30 a.m. to ~21:30 p.m.

5. Conclusions

This study was aimed at investigating the influence of design parameters on the thermal performance of a macro-encapsulated PCM blind within a DSF system. Numerical models were developed, and CFD simulations were conducted using ANSYS Workbench FLUENT software. Experimental data were then collected from a DSF system equipped with both aluminum blinds and PCM blinds for validation of the model. The temperature distributions within the DSF cavity were compared across various blind tilt angles and cavity depth to blind width ratios. Based on these findings, optimized design parameters for the integrated system are proposed. The main conclusions are summarized as follows:

• The PCM blind was able to absorb the excessive heat in the DSF cavity for an average duration of 5.9 h, providing the smallest range of temperature change, which indicates this system can be utilized to optimize thermal environmental impacts based on real-time conditions;
• The air temperature at different positions for the previous DSF test ranged from 39 to ~40 °C. Both the surface temperature of the PCM blind and the air-cavity temperature in the DSF could be maintained within the range of 37~40 °C by optimizing the design parameters. The optimal design parameters can improve the thermal performance of the macro-encapsulated PCM-blind integrated into the DSF system;
• The lowest PCM-layer surface temperature can be achieved with a blind tilt angle of 30° in the DSF system. Moreover, the blind tilt angles of 30° and 60° are recommended, since they demonstrated better temperature adjustment effects in the DSF cavity, with cavity temperature stabilized in the range of 38~40 °C;
• The cavity depth to blind width ratio of X = 2.5 L was recommended as optimal, since it could effectively reduce the air-cavity temperature and eliminate the overheating effect during the daytime in the warm season.

This study demonstrates that macro-encapsulated PCM blinds with aluminum shells can serve as an effective passive solution for mitigating overheating in double-skin façades (DSFs) within China’s hot-summer, cold-winter climate zone. However, several limitations remain, and potential avenues for further research are outlined below. Firstly, the thermal performance of the optimized macro-encapsulated PCM and DSF system across the entire year warrants further investigation, particularly under varying seasonal conditions. This would provide a more comprehensive understanding of its effectiveness in both extreme and transitional climate phases. Additionally, a direct comparison between the performance of the macro-encapsulated system and a micro-encapsulated system is necessary to assess the relative advantages and challenges of each approach, especially in terms of efficiency, cost, and long-term durability. Furthermore, a life-cycle assessment (LCA) and cost analysis of the PCM-blind system are critical for evaluating its economic feasibility and sustainability. Such an assessment would offer valuable insights into the practical implications of adopting this technology in building design and construction, particularly for large-scale or long-term applications.