A Review of Transparent Insulation Material (TIM) for Building Energy Saving and Daylight Comfort

: Improving the energy efficiency of building is key strategy in responding to climate change and resource challenges associated with the use of fossil fuel derived energy. The characteristics of the building envelope play a decisive role in determining building operation energy. Transparent Insulation Materials (TIMs) add to the strategies that may be used to sustain these improvements: they can reduce heat loss by providing high thermal resistance while effectively transmitting solar energy and contributing to the luminous environment. In this review, key types of TIMs and their characterisation in terms of both thermal and optical behaviours are introduced as well as the benefits that may be realised through their application to buildings. Relatively few studies exist regarding the performance of window systems incorporating TIMs. To provide a clear picture of how to accurately predict the performance of TIM integrated window systems, this paper also explores the literature around window systems incorporating complex interstitial structures, as these share many of the same characteristics as TIMs. The experimental and numerical methods used to evaluate the thermal and optical characteristics of complex window systems are summarised and this body of research provides potential methods for tackling similar questions posed in relation to the performance of window systems with TIMs. Finally, this review introduces a method that permits the prediction of the combined thermal and daylight behaviour of spaces served by TIM integrated window systems. The results of using this methodology show that using TIMs over a conventional window system offers a range of benefits to the occupants of buildings. Thus, this review offers a workflow that may be used to assess and analyse the benefit of applying TIMs for building energy saving and daylight comfort in buildings subjected to varying climate conditions.

Energy efficiency requirements are continuously improved through stricter stipulations in building regulations [3][4][5]. In the UK's Building Regulations Part L: Conservation of fuel and power [6][7][8][9], the current edition (2013 edition with 2016 amendments) introduced a reduction in target emission rate of 25% as compared with the previous edition (2006) [10]. This was accompanied by a decrease of target U-values for building components, increase in the envelope airtightness and increased requirements to control unwanted solar gains, avoid to excessive summer temperatures [11].
In response to evolving standards of building regulation and to satisfy the growing expectations of occupant comfort, the performance of building envelopes has undergone sustained improvement over a period of several decades [12][13][14]. Increasing the thermal resistance of the building envelope is one of the most effective approaches to reduce heat loss (or gain) and in so doing reduce the required energy consumption for maintaining a desired level of indoor thermal comfort [15]. Applying conventional insulation products (e.g. expanded polystryrene (EPS), extruded polystyrene (XPS) and fibre products) to the building envelope is a common and mature practice to increase thermal resistance [1]. Novel insulation technologies such as Vacuum Insulation Panels [3] and aerogel [16] can provide the required thermal resistance using layers that are much thinner than conventional insulation materials and hence reducing the thickness of the building structure. Amongst the various components that form building envelopes, window systems, which can be responsible for as much as 60% of the total energy consumption of a building [17], are exceptionally important elements. This is because windows systems contribute to both heat gain and heat loss through the building envelope and also determine daylight distribution and daylight availability [18]. A strategy that involves the integration of Transparent Insulation Materials within a double-glazing unit offers the potential to deliver combined improvements in thermal, solar and daylight performance [18][19][20][21]. However, relatively few studies exist regarding TIMs' building application in this manner as most of the previous research of TIM concentrates on its application in solar collectors and in solar walls. Additionally, detailed analysis of the balance between the thermal resistance and solar transmittance of TIMs and the impact these have on building performance has less been rigorously explored and only a few studies have been conducted to analyse their energy efficiency when subjected to varying climate conditions [21]. There is also a role in improving systematic review that summarises the key experimental and numerical methods for evaluating the thermal and optical characteristics of window systems containing TIM.
This paper presents the current research status into TIMs. The studies reviewed have been traced back to the 1960s when the use of Transparent Insulation Material (i.e. honeycomb structures) in a flat-plate solar collector to suppress convective heat transfer was first reported [22][23][24]. From 1970s to mid-2000s, a number of studies, which experimentally or/and theoretically investigated TIM's ability to suppress natural convection were undertaken and are summarized in this review . In the past decade, aerogel, which is a typical homogenous TIM structure, has attracted lots of attention and research interest in the field of building energy efficiency [3,16,[50][51][52][53][54][55][56]. However, numerical simulation techniques and sophisticated measurement techniques, which have witness fast growth during the past decade, have seen less application in the research of TIM. Thus, this review also introduces up-to-date methods that can be used to characterise the thermal and optical performance of many TIMs, especially when applied for use in windows. With a focus on how to quantify the benefits of applying window integrated TIM to buildings, this review describes the use of dynamic daylight and energy simulation methods to predict building performance and concludes the need for more accurate daylight and energy prediction methods. The review ends by describing recent work that couples thermal and daylighting analysis to allow comprehensive evaluation of buildings employing glazing integrated TIM systems. The method may be seen as offering a first step in developing design strategies that seek to balance thermal and luminous environment of spaces served by glazing integrated TIM systems [15].

Transparent Insulation Materials
Transparent Insulation Materials, which seek to offer the simultaneous resistance to heat flow and facilitate the transmission of light, are almost always assembled with at least one transparent cover (e.g. glazing pane) and typically occupy the air cavity between the panes of a double pane glazing unit [57]. TIMs are generally classified under four categories according to the structure of the TIM layer and its relation with the glazing cover, as shown in Figure 1. The categories are: (a) glazing-perpendicular structure, (b) glazing-parallel structure, (c) mixed structure and (d) homogeneous structure. There are three typical types of glazing-perpendicular structure: 1) capillary, 2) honeycomb, and 3) parallel slat array [57].
Glazing-perpendicular TIMs (Figure 1 (a1)-(a3)) divide the air cavities into small cells that run across the cavity, perpendicular to the glazing surfaces. The cell walls provide additional viscous resistance to the onset of free convection and interfere with the thermal radiation transferred from one pane of the double glazed unit to the other. Both of these effects increase the thermal resistance of the glazing system [58]. Glazing-parallel structures (Figure 1 consist of a number of glass or plastic layers arranged parallel to the glazing panes. Increasing the number of layers reduces heat loss, but results in a reduction in solar and visible transmittance due to the increased reflectance and absorbance provided by the additional layers [57]. Mixed TIMs (Figure 1 (c)) are combinations of glazing-perpendicular and glazing-parallel structures. This structure can achieve more effective convection suppression than the glazing-perpendicular or glazing-parallel structures, but see reduced daylight and solar radiation transmission. Homogenous TIMs (Figure 1 (d)) are receiving increased attention, with recent research into silica aerogel offering a translucent material that can occupy the air cavity of a double glazed unit (Figure 1 (d)). The high porosity of aerogels results in a very low thermal conductivity (e.g. 0.005~0.02W/mK) [54]. There are two types of silica-based aerogels: monolithic silica aerogel (MSA) and granular silica aerogel (GSA).
MSA can be clear enough to see through but its high cost and fragility currently limits commercial uptake [59]. GSA has a low solar transmittance (less than 0.5) due to bulk scattering effects and is commonly used in skylights [55,56]. Currently, the glazingperpendicular TIMs and aerogels are receiving the greatest amount of attention in the fields of both research and industry. The appearance of a TIM varies with the structural geometry and the material from which the TIM is constructed. Figure 2 shows several typical designs of TIMs with glazingperpendicular structures and aerogels. It is possible to see through most of the honeycomb and monolithic silica aerogel TIMs, although at certain viewing angles disruption of the scene beyond can occur (see Figure 2

 Density and temperature stability
The density and temperature stability (i.e. the maximum temperature below which the material is stable) of TIMs are also determined by the material from which they are composed, along with their structural geometry. Table 1 illustrates the density, cell size, and maximum resistance temperature of several common TIMs.
Typically, glazing-perpendicular TIMs with capillary and honeycomb have a density of around 20~50kg/m 3 [60]. Silicon dioxide (SiO2, amorphous quartz) forms the basis of most aerogels; they are made up of approximately 96% air with the remainder being an openpored silica structure [61]. The densities of monolithic silica aerogel lie between 35 00kg/m 3 [60]. Those aerogels used in TIM designed also for effective transmission of solar energy are produced using supercritical drying and have density of between 100~150kg/m 3 [60]. and Padmapriya [42], Plazer [37,38] and Arulanatham and Kaushika [62] analysed the optical properties of glazing-perpendicular TIMs using numerical simulation. Kaushika and Sumathy [29] summarised the research carried out by Hollands et al. [48], Kaushika and Padmapriya [42] and Arulanatham and Kaushika [62], stating that the total beam radiation transmittance at incidence angle θ, τ(θ), for TIMs based on honeycomb or parallel slats can be expressed as: where E is the fraction of the cross-section area occupied by cell wall material. The transmittance of the wall material, τE(θ), and the transmittance of the cells, τc(θ), is calculated by considering the reflection, absorption and the transmission occurring at the cell surfaces.
Wong [57] summarised numerical investigations conducted by Symons [47], Plazer [44] and Plazer [37,38] and concluded that, derived from the summation of all individual rays transmitted or reflected at the cell walls, τ(θ) can be expressed in a simplified form as: where φ is the azimuth angle; ρ and α are the reflectance and absorbance at the cell wall, respectively, and n is the average number of cell wall interactions for the incoming light beam.
It is possible to express n using ݊ ൌ ‫ܣ‬ ή ‫݊ܽݐ‬ ߠ for square honeycomb cells, where A represents the width of the square. For circular honeycomb cells, ݊ ൌ ‫ܣʹ‬ ή ‫݊ܽݐ‬ ߠ may be used where A represents the cell diameter.
In order to validate the theoretical predictions of τ(θ), Symons [47], Plazer [44], and Plazer [37,38] also conducted experimental studies using a spectrometer and an integrating sphere. A detailed exploration of this experimental method can be found in section 2.4.1.
Wong et al. [57] concluded that the experimental results support the results from numerical calculation. The value of τ(θ) for honeycomb, parallel slat and V-slat TIM over incidence angle (θ) in the range of 0 ~ 60°is higher than 0.9 as shown in Figure 3. Buratti and Moretti [56,61,63] have also conducted a series of measurements for the light and solar transmittance of various aerogels (i.e. MSA and GSA) by employing a spectrophotometer (shown in Figure 4 (b)) along with an integrating sphere (shown in Figure   4 (c)), which is used to detect the diffused light transmitted through the TIM sample. The measured wavelength dependent transmission coefficient of MSA on its own and aerogels sandwiched within double glazed units is illustrated in Figure 4 (a) [61]. Monolithic aerogel offers comparatively high transmittance in the near-infrared range when compared with float glass. The results also show that there is a difference between the measured total transmittance determined using an integrating sphere and the measured direct transmittance determined without the use of an integrating sphere, thereby confirming the diffuse nature of aerogel.

 Thermal properties of TIM
The thermal behaviour of glazing-perpendicular TIMs has been investigated both numerically using finite difference methods and experimentally using hot-plate and hot-box experimental methods.
Numerical studies were primarily used to provide a detailed understanding of the convective, conductive and radiative heat transfer within TIMs [25,[32][33][34]. Arulanantham et al. [34] used a finite difference method to explore the convective stability of horizontal and inclined air cavities enclosed by square honeycomb and concluded that convection suppression was affected by aspect ratio (A=L/d, where L is the thickness of the honeycomb and d is the width of the square), thickness of cell wall material (ߜ) and the inclination of cells (ߚ) as shown in Figure 5. Kaushika et al. [33] reported that a honeycomb structure offered better thermal performance for glazing units installed in a horizontal orientation but when units are inclined to more than 30°to the horizontal, parallel slat structures showed better thermal performance. Kumar and Kaushika [25] found that honeycomb structures with aspect ratio in the range between 10~15 were suitable for convection suppression in air cavities of depth 5~20cm for temperature differences between 20~120°C. Arulanantham and Kaushika [32] used numerical methods to study combined conductive and radiative heat transfer through a honeycomb structure in a solar collector and stated that the heat loss coefficient was determined by the interaction between the emissivity of honeycomb cell material and the plates type (i.e. black end plate or selective end plate) of the collector in which the honeycomb is employed.

Figure 5: Schematics of (a) square honeycomb (b) section of a square honeycomb cell (adapted from [25])
Experiments have also been conducted to measure the thermal conductance of TIMs [27,36,38,43]. In order to evaluate the thermal conductance of capillaries, TIMs of various thickness under different tilt angles, a hot-plate apparatus consisting of a hot (i.e. 230°C) and a cold (i.e. 5~60°C) copper plate, heat flux meters and temperature sensors were utilised by Goetzberger [43] and Platzer [36,38]. The thickness of the samples varied from 10mm to 120 mm and the tilt angle was varied within the range of 0~180°. They also used their experimental results to validate their numerical calculation and an approximate 3.6% deviation between measurement and prediction was found when the TIM was tilted within the range of 0~55°. Suehrcke et al. [27] also used a guarded hot-plate apparatus to measure the heat transfer coefficient across TIMs with both corrugated sheet and honeycomb structure, made from cellulose acetate (CA) film with an emissivity of 0.65. They concluded that 30 mm thick CA honeycomb is more effective than 30 mm thick corrugated sheet in reducing heat loss: the heat transfer coefficient for the honeycomb was found to be 3.51W/m 2 K against 4.02W/m 2 K for the corrugated sheet. They also compared their heat transfer coefficients gained from hot-plate measurement with those of numerical simulation, and found the results agreed well with a difference of less than 3%.
Buratti and Moretti [56] used a hot-box apparatus to investigate the thermal transmittance of window systems integrating monolithic aerogel and granular aerogel and found that the windows tested with monolithic silica aerogel offered a lower U-value (0.63 W/m 2 K) than granular aerogel (1W/m 2 K).

Recent work conducted by Sun et al. [20] used a validated Computational Fluid
Dynamic (CFD) model to investigate the thermal properties of double glazing unit with integrates parallel slat TIM with different cell aspect ratios (i.e. the ratio between slat spacing and the thickness of the PS-TIM layer), slat thickness and slat thermophysical properties (i.e. conductivities and emissivities). This study reveals that the presence of a PS-TIM structure can not only supress convective heat transfer but also cause a significant reduction in radiative heat transfer. The results show that an overall 35-46% reduction in thermal conductance can be achieved as compared with the same double glazing in the absence of PS-TIM and that the material properties have a more noticeable influence on small cell structures than large cell structures. In their further studies [21], the CFD simulation process was repeated for representative combinations of temperature gradient and mean temperature.
Polynomial regression of the resulting conductance was used to generate an equivalent dynamic thermal conductivity of the PS-TIM layer to characterise their thermal properties.
These values for both PS-TIM and ordinary double glazed units are shown in Figure 6. From a comparison of Figure 6 (b) (c) (d) with (a), it is evident that the difference between each line, which represents thermal properties under varying temperature difference, is significantly diminished for windows containing 10mm pitch PS-TIM and nearly or completely disappear when 7.5mm and 5mm pitch PS-TIM is present. This means that the convective heat transfer through the cavity between two glazing panes has been effectively reduced with 10mm pitch PS-TIM and almost entirely supressed with 7.5mm and 5mm pitch PS-TIM. The reduction of the gradient of each line in Figure 6 (b) (c) (d), which represents the variation of thermal conductance with variation of mean temperature, indicate that the radiative heat transfer has been reduced through the presence of PS-TIM structure. The influence of applying PS-TIM on the window's heat gain has also been researched in [21].
The simulation results show that average reductions in heat gains of approximately 38%, 42% and 46% for the 10mm, 7.5mm and 5mm PS-TIMs respectively were obtained when compared with ordinary double-glazing.

 Acoustic properties
There is a small amount of research exploring the acoustic performance of aerogel based TIMs. It has been determined that window systems with granular aerogel have more efficient acoustic insulation compared with a conventional window system [61]. The measured sound reduction index, R, of double glazed windows with aerogel is higher than that of a conventional double glazed window in the range 100~2000Hz as shown in Figure   7. This is primarily due to the acoustic absorption provided by the aerogel, which is behaving as a porous acoustic material, and the beneficial impact that this has when present in the cavity of multi leaf structures.

Applications of TIM in buildings
There are three main ways to apply TIM in buildings: the first, and probably the earliest application, uses TIM under the front glass cover of a flat-plate solar collector; the second method, which is most often referred to as a solar wall, uses TIM in front of a thermal mass wall or a thermal mass roof; and the third method uses TIM in the air cavity of a window system or glazed façade. The configurations of these applications are presented in Figure 8.  Application of TIM in flat-plate solar collectors Research initiated in the 1960s, primarily focused on the application of honeycomb structures in solar absorbers as a convection suppression device [22][23][24]. This involved theoretical and experimental research on the application of plastic/glass honeycombs to flatplate collectors and has been expanded more recently to include studies on capillaries as well as silica aerogel [26,35,40,43,47,64]. A coefficient, named the effective transmittanceabsorbance product (ταe), taking into account both transmission, τ , of the glass cover and absorption α of the absorber, was used to characterise performance. Wong [57] concluded that the value of (ταe) decreases with increase of incidence angle; as a result, there is a negative influence on the optical performance of the solar collectors [57]. A comparison of (ταe) at various incidence angles θ for flat-plate solar collectors is shown in Figure 9. and 6 cm FEPT honeycomb [46,57] This reduction in collection efficiency needs to be compensated for by reductions in heat loss if the use of TIM is to be worthwhile. For the honeycomb and capillary TIMs made of plastic or glass, Symons [47] stated that the working temperatures of the solar collectors using TIMs could be increased from 90°C to 150°C. Rommel and Wagner's research [35] revealed that honeycomb TIM could offer a thermal conductivity of 0.9W/mK for the cover system at a temperature of over 80°C. Nordgaard and Beckman's research [64] showed that the flat-plate solar collector with the integration of MSA could achieve an efficiency of more than 60% at a temperature below 80°C and annual energy gains were increased by 41%.
As mentioned above, when TIM is integrated into a flat-plate collector, collector thermal efficiency has been significantly improved, thus the system is comparable to a vacuum tube collector [29]; however, some issues were also identified, such as the fragile and bulky nature of TIMs employing glass as the structural material, and lower stagnation temperature of TIMs made from plastic [29].

 Application of TIM in solar walls
Applying TIM in front of a concrete or masonry wall (Figure 8 (b)) allows solar energy to be transmitted and stored in the thermal mass while simultaneously reducing the heat loss back out through the front surface to the external environment. The combination of TIM and thermal mass was originally reported by Kaushika et al. in 1987 [45]. The system they investigated was integrated into a roof and made use of a square cell, air filled honeycomb material laid on the top of a blackened absorber and concrete slab for thermal storage. The honeycomb materials were made from 0.076mm thick polycarbonate films or 1 mm thick acrylic sheets. In their study, τ(θ) was calculated and found to be greater than 0.9 for the TIM made from polycarbonate film and 0.6 for the TIM made from acrylic sheet.
Their results showed that with the presence of honeycomb slabs, the heat collection efficiency of the roof had been significantly improved and the heat loss during night time had been considerably reduced [45]. In their later research applying the same systems under the climatic conditions in Boulder, USA, a positive solar gain was achieved for more than 10 hours during a typical winter day when the thickness of honeycomb was 15cm [39].
The IEA Solar Heating and Cooling Programme, Task 20 [65,66] stated that the TIMs have been used as part of solar wall heating systems installed as part of building renovation measures throughout Europe. The programme also suggested that shading devices and forced ventilation should be applied to these systems in order to avoid overheating problems during summer. Braun [41] tested the performance of demonstration projects using both experimental and numerical methods to understand the space heating potential of southfacing transparently insulated massive walls. In these projects, the TIMs used were based on capillaries and honeycombs. The results showed that this passive solar façade system offers the potential to save 200 kWh per year for every square metre of collector wall. The difference between the measured and calculated temperatures of the absorber and points inside the wall was found to be less than 5°C. Athienitis and Ramadan [30] used an explicit   predicted. The results indicate that the inclusion of PS-TIM systems improved the luminous environment by reducing the hours of over illumination and in so doing resulted in a more uniform illumination of the working plane under all these five climates. Also, as can be seen from Figure 11, cooling is the dominant mechanism through which energy savings were made when integrating PS-TIM into building windows. The largest potential of energy saving among all these cities is Singapore, which is a representative of hot climate [21]. However, the TIM market is still very limited. The obstacles to large-scale marketisation of TIM include restricted access to information [71], imperfections associated with the manufacturing process, and the relatively high costs involved.
It is clear that TIMs have a role to play in reducing the thermal conductance of building envelopes, and improving both solar gain and daylight performance. These three areas of behaviour influence the environment and energy performance of any space separated from the external environment by envelope components with integrated TIMs. Developing the current understanding of these individual performance criteria beyond its current state would assist significantly in developing holistic models that permit a more realistic picture of the nature of the environment and energy demand of buildings integrating TIM.

Thermal and optical investigations of integrating TIM into window systems
As mentioned in section 2.3, although the thermal and optical performance of TIMs applied to windows have seen relatively little research, the concept of using interstitial structures, such as horizontal Venetian blinds, pleated blinds, etc. within the air cavity of a double glazing unit has been widely researched. These structures share many of the same characteristics as TIMs and this body of research therefore provides potential methods for tackling similar questions posed in relation to the performance of window systems with TIMs.
Previous experimental and numerical methods used to thermally and optically characterise these complex window systems are summarised and discussed in this section.

Thermal investigation  Experimental methods
When presented with the need to measure the steady-state thermal properties of a building component under laboratory conditions, the hot-plate method, heat flow meter method and guarded or calibrated hot-box method are the three key approaches that may be used. In these three methods, hot-plate and flow meter methods are suited to the characterisation of homogeneous materials, such as single glazing, insulation materials etc.
while the hot-box method [72][73][74][75][76] is more widely used to measure the overall heat transfer through large, inhomogeneous structures, such as glazing system with frames or interstitial structures [77]. The specimen is mounted between two chambers that are kept at stable hot and cold conditions. The hot chamber serves as a guard to a metering box, which is mounted over the test sample. By maintaining equal temperatures in the hot chamber and the metering box, all the heat supplied to the metering box is assumed to be transmitted through the sample.
Measurement of the energy supplied to the metering box is then used to determine the thermal transmittance of the sample [78]. Chen [81] studied the centre-of-glass U-value of a window with high-reflectivity Venetian blinds by using hot-box configuration. The influence of the blind's slat slope-angles on U-values was evaluated and an empirical correction factor for the window U-value was proposed. Fang et al. [82] made comparative tests of flat vacuum glazing systems between guarded hot-box measurements and finite element model simulations, The resulting flanking loss, overall heat transfer coefficients and mean surface temperatures showed good agreement.
The dynamic thermal properties of window systems measured under more realistic environmental conditions, can be determined in in-situ measurements. This method uses thermocouples and heat flux meters to measure both the temperature gradient between any two surfaces of building components, as well as the heat transfer rate through them, on the site where they are situated [83,84]. There are however, various factors that can significantly affect the accuracy of the measurements taken. These include the type and quantity of sensors used, component location, and extreme ambient conditions [83]. There are also some variables that cannot be controlled during measurements, such as external temperature, wind speed and radiant energy from the sun, which can make measurements impractical [83].
To overcome the uncertainties associated with in-situ measurement methods to simultaneously simplify guarded hot-box apparatus and streamline the process of acquiring thermal transmittance of the sample, Sun et al. [58] used an integrated method to measure the thermal characteristics of a double glazing unit with and without a Venetian blind installed in the air cavity between the two glazing panes. This measurement method followed International Standard ISO 9869-1:2014 [83], using heat flow meters to determine thermal transmittance through the sample under test. The apparatus setup was also informed by International Standard ISO 12567-1:2012 [75], using two chambers that are kept at stable hot and cold conditions to determine the temperature gradient between two surfaces of the sample. Figure 12 shows the external view and schematic cross-section of the test chambers. The measured heat transfer rate for the double glazing unit under various temperature conditions were compared with calculated results based on standard calculation method set out in EN673 [85], and the differences were found to be less than 1%. In addition, the measured results for the glazing system without Venetian blinds were also compared with results obtained from CFD simulation, and the difference was found to be less than 4%.

 Numerical methods
As an alternative to experimental measurement, finite element or finite volume simulations are the most common approach used in order to obtain the thermal properties of window systems with complex interstitial structures, especially when the air flow pattern in the cavity is of particular interest. Computational Fluid Dynamics tools have been widely used to solve the heat transfer problem and investigate the free convection in air cavities of complex glazing systems with integrated shading devices, such as horizontal Venetian blinds, pleated blinds and different configurations of fins [86][87][88][89][90]. Both hot box test [79] and laser interferometry tests [91][92][93] have been used to validate CFD models. Collins et al. [87] used a two-dimensional steady laminar natural convection model to investigate the energy transfer through a window cavity containing louvers. They concluded that the slat tip-to-glass spacing has significant influence on the overall convective heat transfer in a window cavity with a blind. Dalal et al. [88] conducted a CFD simulation to investigate the steady free convection in the cavity of a double were examined to investigate their effects on overall thermal performance. The results from these studies were used to generate empirical correlations for the thermal resistance of double glazing units with interstitial Venetian blinds.

 Experimental methods
The measurement of transmittance and reflectance of a building component can be obtained using a spectrophotometer along with an integrating sphere. A number of previous studies used this methodology to explore various glazing materials and window systems in terms of their respective optical performance. Jonsson et al. [94] quantified the transparency and light scattering performance in the visible range of antireflection glass coatings and an electrochromic foil. Long et al. [95] and Ye et al. [96] used a UV-visible-NIR spectrophotometer (240~2600 nm) to measure the spectral transmittance of a thermochromic film (i.e. VO2). Goia et al. [97], Gowreesunker et al. [98], and Liu et al. [99] investigated the spectral behaviour of double glazing systems with different PCM samples.
Gao et al. [52] investigated the optical properties (transmittance and reflectance between 290 2500 nm) of glazing units incorporating aerogel granules in the air cavity using a 150 mm integrating sphere. Berardi [53] and Buratti and Moretti [56,61,63] measured the optical behaviour of a monolithic aerogel panel on its own and sandwiched between two glazing panes with and without the use of an integrating sphere.
This method of combining a spectrophotometer with an integrating sphere is mainly suited to characterising homogenous glazing systems or materials. For a specific case of a window system that includes a complex interstitial structure, such as TIM, these standard methods can prove inappropriate as the presence of these interstitial structures can result in significant deviation in the directional characteristics of transmitted and reflected flux, as well as variation in the total amount of flux transmitted/ reflected. The measurements cannot capture valuable directional information that is necessary to make accurate predictions of daylight distribution in the room served by these glazing systems.
To retain information relating to the direction of entry of the light to its direction(s) of exit, recent investigators used goniophotometers to capture directional optical properties (i.e. Bidirectional Scattering Distribution Function (BSDF)) for complex window systems [100][101][102][103][104][105]. The BSDF data comprises matrices of coefficients that for light from each incident direction quantifies the proportion transmitted in each outgoing direction. A goniophotometer used either a scanning-based method [100][101][102][103][104][105] or a video-based method [106][107][108][109][110][111] to capture reflectance and/or transmittance distribution in the outgoing hemisphere for a single incidence angle from the incoming hemisphere.

 Numerical methods
In order to conduct cost and time saving parametric studies, numerical methods corresponding to these experimental methods mentioned in the previous section have also been developed and validated. Radiosity methods standardised by 15099 [112] are commonly used to calculate the overall transmittance or reflectance for these complex fenestration systems, such as window systems with integrated Venetian blinds [113][114][115][116][117][118][119][120][121], based on knowledge of the measured optical characteristics of each of the individual system components. However, this method is not suitable for models with highly specular surfaces, as it assumes that all the surfaces are Lambertian reflectors.
Virtual goniophotometers developed and validated by Andersen et al. [106,107] and Boer [122] based on commercial forward ray-tracing simulation tools (TracePro and OptiCad, respectively) can be used to obtain BSDF data for complex fenestration systems based on descriptions of their geometry and optical characteristics. Another validated program, genBSDF (implemented in RADIANCE) [123], has been used to generate BSDF data for glazing systems with integrated TIMs [18]. The generated BSDF data for these glazing systems can be used further in building daylight and energy simulation tools (e.g. RADIANCE and EnergyPlus) for determining their building performance.

Building performance prediction of window systems with integrated TIM
When exploring how a window with integrated TIMs performs under different climatic conditions, building performance simulation is considered ti be an effective way of estimating its potential for energy saving and daylight comfort improvement. Accurate performance prediction depends on precise characterisation of the thermal and optical properties of TIM fenestration systems, in which two-or three-dimensional heat transfer and/or light transmittance might exist due to the presence of the TIM structure.
Wong et al. [137] conducted a building energy performance prediction using ESP-r software to explore the use of a TIM wall and/or TIM glazing on the south facing façades of retrofitted high-rise and low-rise office buildings in London, UK. The simulation results for a full calendar year revealed that applying TIM has the potential to reduce temperature swings during the daytime. Furthermore, the TIM also offers the potential to reduce heating energy load during winter and reduce overheating problems during summer when applied in conjunction with a mass brick wall, shading overhangs and natural ventilation.
With advances in simulation technologies, simulation tools' capabilities to predict building energy and daylighting performance with complex fenestration systems has been significantly enhanced. The challenge related to the use of these tools for modelling windows with integrated TIM is how to establish a bridge between precise characterisation of a TIM window system and implementation of these characteristics in building simulation tools.
Based on the review of potential methods mentioned in previous sections, the authors proposed a comprehensive model to investigate the thermal and optical performance of a window system with TIMs and also how they shape the daylight and energy performance of the buildings to which they are applied to [18][19][20][21]. Through this approach, the dynamic thermal conductance across the glazing system caused by variation of environmental conditions is obtained from Computational Fluid Dynamics simulation [20]. These results inform the construction of empirical equations that may be integrated into EnergyPlus to represent the thermal characteristics of the TIM in building simulation [19,21]. Meanwhile, a ray-tracing technique is used to predict the optical characteristics that were formatted into BSDF data. RADIANCE was used to generate a comprehensive picture of daylight performance when applying window systems with TIMs was also conducted accompanied with dynamic building daylight metrics [18]. The BSDF data are also input into building simulation tool (i.e. EnergyPlus). Then, the developed model is used to obtain relatively accurate building heating, cooling and lighting energy estimates, when glazing systems with Parallel Slats Transparent Insulation Materials (PS-TIMs) are applied within a window [19,21]. The flow chart of each stage is shown in Figure 13.

Conclusion
The review demonstrates that TIMs have a role to play in improving the thermal resistance of building envelopes, adjusting solar gains and affecting daylight performance.
Over the past decades, the thermal behaviours of TIMs have been investigated numerically and experimentally: however much of this work focused solely on their application in solar collectors. In this guise, the application of TIMs has been proven to effectively reduce heat loss and improve solar collector efficiency. However, relatively few studies exist regarding the thermal and optical performance of TIMs contained within the cavities of double glazed windows. Here, the working temperatures, additional natural convection patterns and intensity, are significantly different. There are also the parallel considerations of impact on view (into and out of the building) and opportunities to enhance daylighting strategies that account also for visual comfort.
As relatively few studies exist that seek to effectively and accurately characterise window systems with integrated TIMs and explore their impact on building performance prediction, this review also summarised the experimental and numerical studies for characterising other complex glazing systems with interstitial structures, which share many of the characteristics of TIMs. Their bodies of research provide potential methods for tackling similar questions posed in relation to the performance of window systems with TIMs. Among the research that has taken place into windows with interstitial structures, many studies have been conducted using Computational Fluid Dynamics modelling to evaluate the possibility of reducing free convection and long-wave radiation exchange between the two glazing panes.
These CFD models have been widely validated by comparing the predicted thermal properties with results obtained using experimental tests. The optical evaluations of these complex window systems are also discussed. An advanced method that uses a matrix of Bidirectional Scatter Distribution Functions (BSDFs) to characterise the optical properties of complex window systems, is regarded as a more appropriate for glazing systems with TIMs.
Based on both previous sections, the need for more efficient and accurate daylight and energy prediction methods for characterising window systems with integrated TIM is concluded. The ability to import thermal characterization into tools such as EnergyPlus as well as the ability to couple this to lighting tools such as RADIANCE creates a framework where such studies may be undertaken. In doing so, this review delivers methods to assess the impact TIMs on building performance and explores how these may be used to analyse their energy efficiency when subjected to varying climate conditions. Researchers, designers or engineers who are keen to develop strategies using windows with integrated TIMs that balance energy efficiency and the quality of the luminous environment of spaces can use the methodology outlined in this review to qualify their benefits.
It is worth noting that although TIM has the potential to offer improved performance of daylight distribution and yield energy saving, the designer would have to consider the extent to which they interrupt view out of and in to a building. Meanwhile, the consideration of whether the daylit environment created by windows with integrated TIM, such as uniformity, are either suitable or desirable has not yet been discussed. These represent limitations of this review that require further study. The long term performance of applying TIMs windows in real buildings and their economic and environmental benefit also need further research.