Vacuum-Insulated Glazing Assessment by CFD Modeling and Laboratory Measurements

Abstract

This paper concerns measurements and CFD simulations of vacuum-insulated glazing (VIG), which consists of two glass panes separated by a narrow gap from which air has been removed. Distancers, e.g., in the form of small balls, are inserted into this gap every few centimeters to prevent the glass from deflecting. In the first part, simulations of two-pane VIG thermal transmittance with the Ansys Fluent program are described, resulting in thermal transmittance of VIG without the network of distancers equal to 2.18 W/(m²K) and with the distancers equal to 2.29 W/(m²K). The influence of the supports on the thermal transmittance of VIG is also determined. The CFD results show that the supporting balls increase the two-pane VIG thermal transmittance by about 0.15% with respect to the glazing without the distancers. Then, VIG is analyzed both numerically and tested in two measurement stands. Firstly, the tests are performed in a guarded hot-plate apparatus, according to the EN ISO 8302 standard. The two-pane glazing with one low-emissivity coating has a measured thermal transmittance equal to 1.75 W/(m²K). Other measurements were undertaken in the calorimetric chamber equipped with the hot-box apparatus. The results of the numerical assessment are then compared to the measurements of the existing three-pane vacuum-insulated glazing with two low-emissivity coatings, the same as simulated. The procedure follows the EN ISO 8990 standard. Measurement results of 1.10 W/(m²K) are compared to the simulation results of VIG thermal transmittance equal to 1.09 W/(m²K). A satisfactory agreement is reached. Additionally, this paper considers a new correction coefficient to thermal transmittance according to standard EN 673 in order to achieve a proper calculation of vacuum-insulated glazing in the center-of-glass region. The authors propose to use an adjustment coefficient of 1.05 when calculating the thermal transmittance of vacuum-insulated glazing without taking into account convection in the vacuum space and the thermal influence of distancers.

1. Introduction

Windows in buildings constitute approximately 10–15% of the external area, but they contribute approximately 30% to the building’s heat loss [1]. Therefore, the search for competitive solutions in highly insulated windows technology is quite an important issue in building energy efficiency. Vacuum-insulated glazing (VIG) is becoming an increasingly attractive alternative to windows in energy-efficient buildings. A review of VIG technology with future prospects was given by Cuce and Cuce [2] as well as by Peng et al. [3].

Many researchers have examined the topic of VIG, which consists of two window panes at a distance of about 0.2 mm and a vacuum gap between them [4]. In order to maintain such a small distance in the space between the panes, a network of distancers has to be introduced into the gap. The distancers can be of different shapes/geometry and materials and can have different arrangements. Zhu et al. [1] indicated the importance of increasing the pillar resistance for creating effective thermal insulation while maintaining the vacuum space. They stated that the effects of various pillar parameters (thermal conductivity, height, shape, and spacing) influence the thermal performance of VIG. They also took into consideration the need to analyze the edge effect, including seals and frames. The researchers analyzed different sealing techniques: the solder glass edge sealing method, the vacuum chamber edge sealing method, and the pump-out edge sealing method [5,6]. Memon et al. [7] proposed an edge seal made of Sn-Pb-Zn-Sb-AlTiSiCu composite to protect the vacuum. Bao et al. analyzed pressures in air cavities on both sides of the vacuum cavity and introduced an interlinking pipe between the two gaps to protect the vacuum against a failure caused by additional differential pressure from both exposed surfaces and bending stresses [8]. Another issue with VIG concerns the distancers (support pillars), which can make glass prone to damage during production, transportation, and installation. Riedel et al. [9] proposed the automatically classified damage by the trained and tested deep learning model using convolutional neural networks. In VIG technology, sealing is also important. Authors indicate that the glass edge is a key point in developing a high-quality VIG [10]. One of the interesting propositions concerns the usage of activationless getters in the form of powders or granules of Ca0.35Li0.45Mg0.20 inside the vacuum gap [11]. Another proposition takes into account the separation of the panes by offset grids made, for example, of frit-coated fibers fused with the glass panes in a vacuum. Such a solution creates independent cells, increasing redundancy in sealing [12].

All the previously mentioned references took into consideration some of VIG issues. However, none of the research mentions the discussion concerning legal issues related to windows with VIG. Manufacturers of windows with VIG have difficulties with obtaining the CE mark for their products since there is no harmonized standard covering this type of window glazing. In order to obtain the CE mark and place the product on the market in mass production, the manufacturer must conduct a European technical assessment and obtain a European reference document. Manufacturers who want to enter the new market of VIG in windows must overcome the problem of the size of glass, seals, and additionally, use of low-emissivity coatings on glass surfaces. Unfortunately, they have no standard concerning VIG thermal transmittance calculations, which are obligatory in window technical assessment. Therefore, there is a need to develop an extension of European standards to include VIG as one of the types of glazing in windows. In this research, the authors used typical VIG used by manufacturers of energy-efficient buildings. This type of glazing has a vacuum gap between the panes supported by metal balls with a diameter of 0.2 mm, arranged in an array every 50 mm. The authors analyzed the thermal transmittance of VIG glazing in the center-of-glass region. They also analyzed the calculation procedure of thermal transmittance of typical glazing given in standard EN 673 [13], which does not take into account VIG. The authors in this paper propose an adjustment coefficient to the standard EN 673 concerning the thermal transmittance of VIG calculations. This amendment to the standard is strongly needed by the window manufacturers.

This paper is organized in this way: firstly, the preliminary measurements of surface emissivity are described (Section 2). With known emissivities, numerical simulations of heat transfer through the VIG are described in order to state how much the distancers influence the thermal transmittance of the glazing (Section 3). Then, the guarded hot-plate measurements (Section 4) and calorimetric chamber measurements (Section 5) are given again to check the thermal transmittance of the real specimen with a vacuum gap and low-emissivity coating. The last Section 6 summarizes the results and proposes an amendment to the standard EN 673.

2. Surface Emissivity Measurements

Because of the fact that only radiation occurs in the glazing gap between the glass panes, surface emissivities on all glass surfaces have to be known. In this chapter, the emissivity measurement results are presented.

2.1. Measurement Method—Emissometer

In this work, the emissivities were measured according to the measurement procedure of ASTM C1371 [14]. This method makes use of an emissometer, a device that comprises a 50-millimeter-diameter circular head, electrically heated to 82 °C. During the tests, a 4.3 mm air gap maintained by a surrounding plastic cylinder separates the sample and the head. This instrument incorporates a differential heat flow gauge with two sensor pairs: one black–opaque (high-emissive) and the other with low-emitting gold treatment (Figure 1).

The differential sensor outputs the voltage difference between the two thermopile pairs. This measurement enables the assessment of the radiative heat exchange. The first pair measures both conductive and radiative heat flux, while the second evaluates only the conductive heat flux, as the thin air gap minimizes convection. By analyzing the radiative heat flow and the surface temperatures (sample temperature maintained near ambient), the sample’s emissivity can be determined. The equipment requires substantial heat dissipation through the analyzed object to maintain a temperature difference between the sensor and the sample. Consequently, it is most accurate for highly conductive materials rather than thermal insulators.

Figure 1 shows the measurement of the glass emissivity with the emissometer; on the right-hand side, the calibration samples are also reported: high (black) and low (gray) emissivity.

2.2. Measurement Results—Emissometer

Two samples were tested: floated glass emissivity as well as the emissivity of the low-e coating. The floated glass had a surface emissivity of 0.84, while the emissivity of the glass with low-e coating was equal to 0.45, which will be mentioned in the following paragraphs.

3. Numerical Simulations of Thermal Transmittance of VIG

3.1. Method and Description of Calculation Examples

The numerical simulations of the repeating element in the glazing away from the edge seal have been performed in the Ansys Fluent 2024 R1 program. The program uses the finite volume method to solve the heat transfer steady-state derivative equations. The part of the analyzed geometry (Model 1) is given in Figure 2. The second analyzed model differs only in the absence of the distancer (Model 2).

3.2. Numerical Model Settings

The geometry was calculated using two-dimensional calculations with the axisymmetric arrangement. This simplification lets the 2-D calculations be, in fact, 3-D.

The thermal conductivities of glass and steel were considered equal to 1.0 W/(mK) and 50.0 W/(mK), respectively. Soda–lime glass thermal conductivity is given in the standard EN 673:2011. Glass in Building—Determination of Thermal Transmittance (U Value)—Calculation Method was used. It is the standard value when calculating glazing examples made of glass panes. Because of the fact that in well-made vacuum glazing, the heat flow rate through the vacuum space due to gaseous conduction is negligible [15], it was assumed that only radiation occurs in this space. The surface emissivity of the glass panes was fixed at 0.84, according to the measurement results in the previous paragraph.

The boundary conditions were set according to the EN ISO 10077-2 standard [16]. The heat transfer coefficients on the external and internal surfaces were assumed to be 25 W/(m²K) and 7.7 W/(m²K), respectively, with external and indoor air temperatures fixed at 0 °C and 20 °C, respectively. The top boundary surface was assumed adiabatic, while the bottom surface was stated as axisymmetric. Then, the calculations were performed for the geometry from Figure 2 after rotation along the x-axis (bottom edge), creating the cylinder of 8.2 mm in height: 4 mm external pane, 0.2 mm vacuum with distancer, 4 mm internal pane.

The settings for the finite volume CFD model for the heat transfer through the VIG model are listed in

Table 1. CFD model settings.

Solver	Stationary
Discretization scheme	Energy	Second order upwind
Radiation model	Discrete Transfer Radiation Model (DTRM) for diffuse surfaces with ray tracing—all pane surfaces are radiating surfaces with given emissivity—the “ray tracing” technique can predict radiative heat transfer between surfaces without view factor calculations.
Relaxation factor for energy	0.75
Residuum for energy	1 × 10−16

below [17].

The finite element mesh was a refined mesh with an element size of 0.01 mm. The mesh quality indicators, maximum aspect ratio, and skewness were equal to 2.5 and 0.45, respectively [17,18,19].

3.3. Numerical Simulation Results

Two cases were calculated: the VIG model with and without a distancer. The aim of these two calculations was to check the influence of the distancer on the heat flow through the glazing. The edge effect was neglected in this case, and only the center-of-glass calculations were performed.

The temperature distributions for two glazing examples, with the distancer (Model 1) and without the distancer (Model 2), are presented in Figure 3 and Figure 4, respectively.

After comparing Figure 3 and Figure 4, it can be seen that the influence of the metal distancer is small and only local in the strict neighborhood of the distancer (Figure 2). The temperature distribution in Figure 4 for the model without the distancer is as for one-dimensional, undisturbed heat exchange.

The calculation results obtained in the Ansys Fluent program indicate that in the model with the distancer, heat losses are only slightly higher. They can be seen in

Table 2. CFD model calculation results.

Quantity	Model 1	Model 2
Heat exchange area, A, m2	7.85·10−5	7.85·10−5
Temperature difference, ΔT, K	20	20
Total heat transfer rate, Φ, W	0.00360	0.00342
Total thermal resistance of the model, RT, m2K/W	0.437	0.459
Thermal transmittance of the model, U, W/m2K	2.290	2.178
Total thermal resistance of VIG according to EN 673, RT, m2K/W		0.448
Thermal transmittance of VIG according to EN 673, U, W/m2K		2.23

It should be mentioned that thermal transmittance of typical 2-pane glazing filled with 16 mm gap of argon/air mixture 90%/10% with no low-e coatings has the thermal transmittance of U = 2.63 W/(m²K), calculated following the EN 673 standard. As can be seen in Table 2, the reduction in thermal transmittance leads to energy savings of about 15% for 2-pane glazing.

.

The results show that in the model with the distancer (Model 1), thermal transmittance is only 5.2% higher compared to the model without the distancer (Model 2). Because of the fact that the two models took into account only a close vicinity of the distancers, not the whole repeatable element, in the vacuum glazing with a network of distancers every 50 mm, the influence of the distancers is even lower, as it reaches 0.15% comparing to the glazing without the distancers.

Additional calculations were also performed in accordance with the EN 673 standard. In these calculations, the thermal resistance associated with convection inside the vacuum space was neglected. It was assumed that only thermal radiation occurs in space. The calculation results of the thermal resistance of the glazing as in Model 2, as well as thermal transmittance, are given in Table 2 in the last two lines. The results obtained according to the standard and numerical calculations agree within 2.5%.

4. Guarded Hot-Plate Measurements of the Vacuum Glazing

4.1. Measurement Method—Guarded Hot Plate

The University of Perugia’s Laboratory of Thermotechnics houses a “single-specimen” guarded hot-plate apparatus with a 50 cm × 50 mm frontal area (Figure 5). This apparatus features a main heating element comprising a central square measurement area and a surrounding guard ring. The measurement area delivers a specific power through resistors fed with direct current, while the guard ring is maintained at the same temperature by the control system.

A second guard hot plate lies beneath the main heating element, sandwiched between insulating panels. This plate is also maintained at the same temperature as the central area by the control system. This configuration aims to establish a one-dimensional heat flow within the measurement zone.

The specimen is positioned between the main heating element and a water-cooled metal plate (cooling unit).

The apparatus incorporates 34 shielded J-type thermocouples:

-

17 thin-wire thermocouples are located within the measurement zone;

-

eight thin-wire thermocouples are positioned within the ring guard near its boundary with the measurement zone;

-

eight immersion thermocouples are placed within the cooling unit in direct contact with the specimen;

-

one shielded thermocouple is inserted within the bottom guard hot plate.

The sensors can be categorized into two groups:

• thermal imbalance monitoring: this group of sensors monitors the temperature difference between the measuring area and the surrounding guard zones;
• average temperature measurement: this group measures the average temperature of both the hot and cold faces of the specimen, enabling the calculation of the average cross-temperature difference.

Sensors responsible for maintaining the temperature balance between the apparatus components are connected to a control panel. This panel houses six temperature controllers, each capable of operating in both PID (Proportional–Integral–Derivative) and ON/OFF modes. The temperature monitoring and measurement are handled by three Data Acquisition (DAQ) systems. Each DAQ system has four channels, allowing for the connection with twelve sensors in total. The power supplied to the central zone is measured using a professional digital multimeter.

The measurement protocol foresees seven steps:

-

Step 1 (sample conditioning): the vacuum glazing is conditioned for 24 h between 105 °C and 110 °C in a climatic chamber for humidity evaporation, as prescribed by Standard EN 12664:2002 [20]. The conditioning is considered effective if the panel weight difference between two conditioning periods is lower than 0.1 kg/m³ or 0.01% by volume.

-

Step 2 (thickness measurement): the thickness of the sample is measured on its four sides, and the mean value is registered;

-

Step 3 (environmental conditions): the room temperature and relative humidity are registered, fixing them to values close to 20 °C and 50%, respectively;

-

Step 4 (panel setup): the lateral edges of the sample are covered with tape, with the aim of preventing humidity from entering the panels themselves;

-

Step 5 (pressure set): the upper pressure is set to 3000 Pa;

-

Step 6 (temperature set and probes positioning): the hot-plate temperatures are set to 60 °C (hot side) and 15 °C (cold side). Three temperature probes are positioned on both the hot and cold sides of the sample to avoid the influence of any thermal resistance effect linked to the contact with the hot and cold plates. The average values of the three thermocouples for each side, T_h and T_c, are used for this calculation;

-

Step 7 (measurement): the thermal conductivity is retrieved according to the following equation:

$$\mathsf{\lambda }={\displaystyle \frac{\mathrm{q}\mathrm{l}}{\mathrm{A}\left({\mathrm{T}}_{\mathrm{h}}-{\mathrm{T}}_{\mathrm{c}}\right)}},$$

(1)

where q is the heat flow per time unit, W; A is the area of the sample facing the plates, m², and l is the specimen thickness, m.

4.2. Measurement Results—Guarded Hot Plate

The aim of this part of this research was to measure the thermal resistance of the tested vacuum glazing. The specimen consisted of two 4-millimeter-thick glass panes separated by a vacuum gap with a thickness of 0.2 mm. One of the panes in contact with the vacuum gap had a surface emissivity equal to 0.45, and the other pane surfaces had an emissivity of 0.84. The value of surface emissivity of clear uncoated glass is given in standard EN 673. It is equal to 0.84, as it was measured and given in Section 2. The glass surface can have some coatings that change the surface emissivity. The lower the surface emissivity, the lower the radiation heat transfer occurs in the gap between the panes. The same applies to the solar heat gain, which is reduced both in winter (negative effect) and summer (positive effect), with low-e inner coatings. The authors used the glass with a low-emissivity coating of 0.45. Figure 6 presents (a) the view of the vacuum valve in the corner of the glazing (diameter of about 7 mm), (b) one of the 0.2 mm distancers in the glazing grid (the caliper is set at 1 mm distance), and (c) the side view of the glazing.

The results of guarded hot-plate measurements, together with the calculation results according to the EN 673 standard, neglecting gas convective resistance and supports in the vacuum gap, are given in

Table 3. Guarded hot-plate measurement results.

Quantity	Value	Unit
Equivalent thermal conductivity, λeq	0.020	W/(mK)
Thermal resistance of the specimen, Rλ	0.400	m2K/W
Total thermal resistance of the specimen, RT	0.570	m2K/W
Thermal transmittance of the specimen, U	1.75	W/m2K
Total thermal resistance of VIG according to EN 673, RT	0.647	m2K/W
Thermal transmittance of VIG according to EN 673, U	1.54	W/m2K

.

In order to compare the measurement result with the calculation result, the absolute percentage error (APE) was calculated following the following formula:

$$\mathrm{A}\mathrm{P}\mathrm{E}={\displaystyle \frac{\left|{\mathrm{U}}_{\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{s}}-{\mathrm{U}}_{\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{c}}\right|}{{\mathrm{U}}_{\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{s}}}}\cdot 100$$

(2)

in which U_meas, W/(m²K), stands for glazing thermal transmittance achieved from measurements and U_calc, W/(m²K), stands for glazing thermal transmittance achieved from calculations according to EN 673 standard, as given before.

As can be seen in Table 3, the absolute percentage error of VIG thermal transmittance is equal to APE = 12%, following Equation (2). It should be mentioned that in the calculations concerning the EN 673 standard, the distancers are not taken into account; this is why the measurement result is higher than calculated with the standard. It should be noted also that in the calculations following the standard, the center-of-glass region is taken into account (one-dimensional calculations).

The results again show the potential of VIG in energy savings. The typical two-pane glazing filled with a 16 mm gap of argon/air mixture 90%/10% with one low-e coating of surface emissivity equal to 0.45 has the thermal transmittance of U = 2.10 W/(m²K), calculated following the EN 673 standard. Comparing this value to Table 3 results, the reduction in thermal transmittance leads to energy savings of about 27% for the analyzed two-pane glazing.

5. Calorimetric Chamber Measurements of the Vacuum Glazing

5.1. Measurement Method—Calorimetric Chamber

After the guarded hot-plate measurements, the calorimetric chamber measurements were performed in Cracow University of Technology, Thermal Engineering Laboratory (Figure 7 and Figure 8).

The calorimetric chamber is a room divided into two parts: a warm room with controlled conditions (simulating internal air conditions) and a cold room (simulating external air conditions) separated by the surrounding panel into which a measured specimen is inserted. The important part of the apparatus is the hot-box system installed in the warm chamber. It is equipped with radiators and fans to provide the amount of heat that flows through the sample at given internal and external air temperatures. The hot-box apparatus is additionally equipped with two sets of thermocouples on both sides of its walls. All measuring devices are handled by the Data Acquisition system DAQ. The measurement is taken from several hours of controlled steady-state conditions when the temperatures on both sides of hot-box walls are equal, which indicates no heat transferred through the hot box, i.e., all heat generated by the radiators is transferred through the specimen, surrounding panel and the flanking losses between them. The steady-state temperatures of air, baffle, surrounding panel, and reveal on both sides of the tested specimen are measured along with the input heat power into the hot box [17]. The measurements meet the demands of the EN ISO 8990 [21] and EN ISO 12567-1 [22] standards. The standards demand at least three hours of steady-state conditions with the measured temperatures unchanged within ±1% and the input heat power ±0.5%. The control and acquisition data system was set to achieve this goal using a PID controller. The measurements were taken every 5 s and registered in the system.

Because of the unknown convective and radiation fractions of surface thermal resistances on both specimen sides, the measurements have to be divided into two stages. Firstly, two calibration panels (thin and thick) with known thermal resistances are being measured in order to achieve surface resistances on both sides of the panels. Then, the specimen with the unknown thermal transmittance is measured.

5.2. Calibration Panel Measurement Results—Calorimetric Chamber

Two calibration panels had total thicknesses of 28 mm and 68 mm and were built of polystyrene covered on both sides with 4 mm glass. The panels had the dimensions of 1100 mm × 1100 mm. Both of them were mounted on the surrounding panel with a reveal depth of 40 mm on the warm side according to the EN ISO 12567-1 standard. Figure 9 presents the view of one of the calibration plates mounted on the surrounding panel in the calorimetric chamber. Each of the calibration panels was tested under three conditions: internal (warm) temperature of 20 °C and external (cold) temperature of 0 °C, −10 °C and 10 °C with the fan settings fixed. The results of the calibration measurements are given in

Table 4. Calibration panel measurement results.

		Calibration Panel 1 (Thickness 28 mm)				Calibration Panel 2 (Thickness 68 mm)
		Meas. 2	Meas. 1	Meas. 3	Meas. 2	Meas. 1	Meas. 3
θce	°C	0.58	10.74	−9.39	0.61	10.72	−9.47
θse,b	°C	0.60	10.73	−9.37	0.61	10.69	−9.45
θse,ca	°C	1.79	11.37	−7.60	1.10	11.00	−8.80
θse,p	°C	0.98	10.97	−8.72	0.83	10.80	−9.17
θse,sur	°C	0.90	10.90	−8.99	0.90	10.90	−9.04
θci	°C	19.76	19.87	19.65	19.82	19.88	19.78
θsi,b	°C	19.46	19.71	19.20	19.67	19.81	19.57
θsi,ca	°C	16.97	18.60	15.41	18.86	19.50	18.23
θsi,p	°C	17.17	18.75	15.77	18.90	19.50	18.42
θsi,sur	°C	19.55	19.89	19.29	19.62	19.90	19.50
Φin	W	37.25	37.25	55.43	19.14	8.97	27.89
νi	m/s	0.1	0.1	0.1	0.1	0.1	0.1
νe	m/s	4.4	4.0	4.5	4.5	4.2	4.6

The following nomenclature was given in the table: θ_ce, θ_ci—air temperatures in cold and warm chamber; θ_se,b, θ_si,b—cold and warm baffle surface temperatures; θ_se,ca, θ_si,ca—calibration panel surface temperatures on the cold and warm sides; θ_se,p, θ_si,p—reveal panel surface temperature on the cold and warm sides; θ_se,sur, θ_si,sur—surround panel surface temperature on the cold and warm sides; Φ_in—input power in the hot box; ν_e, ν_i—air temperature on the cold and warm side of the calibration panel.

.

It should be mentioned that Pt100 resistance sensors were installed in the cold chamber with a measurement uncertainty of 0.1 K. The warm chamber and hot box were equipped with thermocouples T-type with a measurement uncertainty of 0.5 K. The measurement uncertainties of dimensions and input power in the hot box were equal to 0.0005 m and 0.3 W, respectively.

The measurements let us sketch the regression curves using the least-squares method: surround panel thermal resistance in terms of its average temperature (Figure 10), as well as total surface resistance and convective fraction in terms of density of heat flow rate (Figure 11 and Figure 12, respectively).

Once the regression curves from Figure 10, Figure 11 and Figure 12 are known, the second stage of measurements concerning the tested specimen can be started. It is assumed that the surface resistances and convective fractions are similar during calibration panel measurements and specimen measurements with the same fan settings.

5.3. Tested Sample Measurement Results—Calorimetric Chamber

The specimen under test was the three-pane vacuum glazing consisting of glass 4 mm, vacuum 0.2 mm, glass 4 mm, argon/air mixture 90%/10% 12 mm, and glass 4 mm. There were two layers of surface emissivity equal to 0.45: one on one pane surface in contact with vacuum and the other in contact with argon/air gap. The other glass surface emissivities were equal to 0.84, as with the float glass. The size of the vacuum glazing was 1230 mm × 1473 mm with a thickness of 24.2 mm. The specimen was located in the surrounding panel with a depth of 40 mm from the warm side, following the window standard EN ISO 12567-1. The measurement results, along with the calculation of glazing thermal transmittance results, are given in

Table 5. Calorimetric chamber measurement and calculation results.

Quantity	Value	Unit
Warm air temperature, θci	19.69	°C
Cold air temperature, θce	0.81	°C
Warm baffle surface temperature, θsi,b	19.36	°C
Cold baffle surface temperature, θse,b	0.81	°C
Warm reveal surface temperature, θsi,p	16.80	°C
Cold reveal surface temperature, θse,p	1.00	°C
Warm surround panel surface temperature, θsi,sur	19.50	°C
Cold surround panel surface temperature, θse,sur	1.10	°C
Heat power input in the hot box, Φin	44.74	W
Density of heat flow rate through the specimen, qsp	20.37	W/m2
Convective fraction—warm side Fci	0.556	-
Convective fraction—cold side Fce	0.813	-
Total surface resistance Rs,t	0.150	m2K/W
Environmental warm side temperature, θni	19.54	°C
Environmental cold side temperature, θne	0.81	°C
VIG thermal transmittance (measured value), Um	1.09	W/(m2K)
VIG thermal transmittance (standardized value), Ust	1.10	W/(m2K)
VIG thermal transmittance uncertainty, ΔUm	0.01	W/(m2K)
Thermal transmittance of VIG according to EN 673, U	1.07	W/(m2K)

.

It should be emphasized that in the calculations according to the EN 673 standard, metal distancers, as well as convection in vacuum space, were neglected. Nevertheless, the calculation and measurement results are in satisfactory agreement. The absolute percentage error, calculated with Equation (2), is equal to APE = 2.7% (less than a 3% discrepancy). The results indicate that the distancers did not increase the heat flow through the glazing to a noticeable extent.

The results also indicate the potential of VIG in energy savings. The three-pane glazing with two gaps filled with 12 mm argon/air mixture 90%/10% and with two low-e coatings with surface emissivities of 0.45 has the thermal transmittance of U = 1.36 W/(m²K). Compared to the value of 1.07 W/(m²K) for VIG, it brings energy savings of 21%.

6. Summary and Conclusions

This paper deals with the problem of distancers inside the vacuum gap and their influence on the thermal transmittance of the vacuum-insulated glazing. Numerical simulations of the VIG with and without the distancer showed that the distancers enlarge the thermal transmittance of the glazing by less than 1% in the center-of-glass region, which was taken into account. Along with the simulations, the calculations of the thermal transmittance according to the EN 673 standard were carried out with the neglect of convective heat transfer in the vacuum gap. The results of the simulations and standardized calculations were in good agreement; the discrepancy was equal to 2.5%. Then, the guarded hot-plate measurements were performed to measure two-pane vacuum-insulated glazing thermal resistance, calculate its thermal transmittance, and compare it to the EN 673 calculation results. They followed the surface emissivity measurements. The last part of this paper considers the calorimetric chamber measurements of calibration panels and three-pane vacuum-insulated glazing. The results showed that neglecting convection in vacuum space and neglecting metal distancers do not cause a big discrepancy between the calculations according to the EN 673 standard and the measurements.

The results show the lack of calculation procedures taking VIG into account and the need to implement it to the standards that are used by the manufacturers in window declaration of performance. The results also indicate that vacuum-insulated glazing has lower thermal transmittance than typical glazing examples, which is the energy-saving potential, and the topic should be further developed.

In conclusion, it is proposed to use an adjustment coefficient equal to 1.05 when calculating the thermal transmittance of vacuum-insulated glazing without taking into account convection in the vacuum space and the thermal influence of distancers.