Integrating vacuum insulation panels in building constructions: an integral perspective
The Authors
Martin J. Tenpierik, Climate Design and Environment Group, Faculty of Architecture Urbanism and Building Sciences, Delft University of Technology, Delft, The Netherlands
Johannes J.M. Cauberg, Climate Design and Environment Group, Faculty of Architecture Urbanism and Building Sciences, Delft University of Technology, Delft, The Netherlands
Thomas I. Thorsell, Division of Building Technology, School of Architecture and the Built Environment, The Royal Institute of Technology (KTH), Stockholm, Sweden
Acknowledgements
The authors are thankful to the Energy Conservation in Buildings and Community Systems Program (ECBCS) of the International Energy Agency (IAE), which initiated the High-performance Thermal Insulation in buildings and building systems (HiPTI) project, which made this research possible.
Abstract
Purpose – Although vacuum insulation panels (VIPs) are thermal insulators that combine high thermal performance with limited thickness, application in the building sector is still rare due to lack of scientific knowledge on the behaviour of these panels applied in building constructions. This paper, therefore, seeks to give an overview of the requirements for and the behaviour of VIPs integrated into building components and constructions. Moreover, the interaction between different requirements on and properties of these integrated components are discussed in detail, since a desired high quality of the finished product demands an integral approach regarding all properties and requirements, especially during the design phase. Therefore, the importance of an integral design approach to application of VIPs is shown and emphasized in this paper.
Design/methodology/approach – To achieve this objective, the legally and technically required properties of VIPs and especially their interrelationships have been studied, resulting in a relationship diagram. Based on these investigations of thermal- , service life- and structural-properties have been selected to be studied more elaborately using experimental set-up for structural testing and simulation software for thermal and hygrothermal testing.
Findings – Two relationships between requirements or properties were found to be of principal importance for the design of façade components in which VIPs are integrated. First, thermal performance requirements strongly interact with structural performance, principally through the edge spacer of this façade component. A high thermal performance requires minimization of the thermal edge effect, in most cases reducing the structural performance of the entire panel. Second, an important relationship between thermal performance and service life has been recognised. The operating phenomenon mainly governing this interaction is thermal conductivity aging.
Originality/value – Most research in the field of vacuum insulation until now has been directed towards gaining knowledge on specific properties of the product, especially on thermal and hygrothermal properties. The relationships and interactions between these properties and the structural behaviour, however, have been neglected. This paper, therefore, addresses the need for an integral design (and study) approach for the application of VIPs in architectural constructions.
Article Type:
Research paper
Keyword(s):
Vacuum devices; Integration; Building specifications; Thermal efficiency; Structural design.
Journal:
Construction Innovation
Volume:
7
Number:
1
Year:
2007
pp:
38-53
Copyright ©
Emerald Group Publishing Limited
ISSN:
1471-4175
Introduction
Owing to sustainability and environmental care, it is desirable and due to international treaties and protocols, it is mandatory that CO2-emmisions, partly resulting from the generation of primary energy with fossil fuels, be reduced drastically (8 per cent reduction of greenhouse-gas emissions in 2008-2012 with reference to 1990 for the Netherlands according to Kyoto Protocol) (UNFCCC, 1997). Since, approximately 40.7 per cent of the primary energy generated in the European Union (EU-15) in 1990 is applied for buildings and building-related processes, of which slightly more than 50 per cent is used for building heating in both the residential and commercial sector (Simmler et al., 2005), a reduction in heat transmission through building envelopes can contribute significantly to the desired reduction in CO2-emissions. One way of reducing this heat transmission is improving the thermal performance of building envelopes by increasing the insulator thickness in case of conventional insulation materials. This, however, results in very thick façade constructions or components, having a number of disadvantages concerning an unfavourable net-to-gross floor and height ratio, an increased complexity of building joints and details, an impeded possibility for designing slender constructions and an increased environmental impact owing to additional material use. A more innovative method to increase the performance of the building shell would be to use a more efficient insulation material, like a vacuum insulation panel (VIP).
Because of their very good insulating properties, VIPs, originally developed for consumer goods like refrigerators, have recently been introduced to building practitioners. These panels consist of a nano-structured open-celled core material, which is after evacuation tightly sealed into a low permeation barrier envelope commonly consisting of a multilayered metallized polymer film or a multilayered aluminium foil. The notion of a pack of coffee clearly comes into picture when thinking of VIPs. Suppression of gas conduction (and convection) and reduction of radiative heat transfer reduces the centre-of-panel thermal conductivity of a VIP to between 3 × 10−3 W m−1 K−1 and 5 × 10−3 W m−1 K−1. This approximately is a factor 5-12 lower than the thermal conductivity of conventional thermal insulators. As a result, a VIP of only 20 mm can ideally replace a layer of mineral insulation or polyurethane foam (PU) insulation of 185 or 120 mm, respectively.
Several investigations into the properties of VIPs for buildings have already been conducted. Most of these investigations, however, mainly addressed the specific properties of VIPs under laboratory, thus non-applied, conditions, like thermal conductivity aging and service life estimations (Simmler et al., 2005; Simmler and Brunner, 2005; Caps et al., 2001; Schwab et al., 2005a, b, c, e), or ideal centre-of-panel thermal performance sometimes including thermal edge effects (Binz et al., 2005; Ghazi Wakili et al., 2004; Rath, 1989; Schwab et al., 2005d; Thorsell and Källebrink, 2005). Studies into the effects of integrating VIPs in building constructions, however, are less frequent and less elaborate. Some studies tried to extract guidelines and recommendations for a successful planning and execution of the application of VIP in building constructions from exemplar projects in some cases including measurements after installation (Binz et al., 2005). Beside these exemplar projects, which are mainly located in Switzerland and Germany, some investigations into the thermal edge effects of façade component edge spacers or building constructions which include VIPs have been performed as well (Bundi, 2003; Nussbaumer et al., 2005; Nussbaumer et al., 2006). Some façade element manufacturers, finally, have also developed a number of standard VIP-integrated façade panels that can be applied in curtain wall façades. These façade panels are based on typical double-glazing panels.
None of these investigations or developments, however, has investigated the actual application of VIPs from a thorough building scientific point of view, combining building physical and structural behaviour. Since, VIPs form a complex system of materials, application in building façades is not without risks regarding its thermal, hygric, hygrothermal and structural behaviour and performance. Integration must thus be performed very meticulously, considering all relevant properties and requirements. This paper, therefore, shows that an integral approach to VIP application in buildings is required to optimize overall performance and to prevent unnecessary risks. A second objective of this paper concerns giving a first overview of some of the important aspects relating to integration of VIPs into prefabricated building components. It thus not elaborates on thermal, hygrothermal and structural behaviour in detail, but solely discusses those aspects necessary to understand the relationships between several properties and requirements.
To achieve these objectives, this paper first continues with a discussion on the demands on VIPs applied in building constructions and investigates the interrelationships among these requirements. In this discussion, two interrelationships are shown to be of special importance for VIPs: the interaction between thermal and structural requirements and the behaviour and interaction between thermal properties and service life. Based upon this discussion, thermal properties, service life behaviour and structural aspects are selected to be investigated in more detail in the subsequent sections. Numerical simulations with software tools using energy and moisture balance techniques (Trisco v. 10.0w and Delphin v. 4.5) have been applied to calculate thermal bridge effects and temperature and relative humidity fields inside building components. Based upon some of these results, it is argued that panels larger than at least 0.5 m2, but preferably even larger, should be applied if the importance of thermal bridging needs to be reduced and the service life of VIPs needs to be increased. Besides, it is discussed that in building components the service life of a VIP can be increased by keeping the relative humidity inside this component stable at a value as low as possible. To investigate the structural performance of VIPs, flexion tests and tensile strength tests in the flatwise plane according to ASTM C 393 (1992) and ASTM C 297 (1992), respectively, have been applied on several VIPs and VIP-integrated building panels. Based on preliminary results, it is finally shown that VIPs have the potential to be applied in structural sandwich components as well.
Requirements for VIPs integrated into building constructions
The application of VIPs (Figure 1) in building constructions puts specific requirements on the VIP, in most cases heavier than the requirements coming from traditional VIP applications, like refrigerators and mobile organ boxes due to longer service life expectations. Based upon the European Construction Products Directive (European Council, 1988), six different product-related requirements on building products in general can be distinguished. To this set of requirements one additional product-related requirement specific for VIPs needs to be added, resulting in the following list of prerequisites
- structural requirements (mechanical resistance and stability);
- fire protection requirements (safety in case of fire);
- requirements regarding hygiene, health and environment;
- application safety and fitness for use (safety in use);
- acoustical requirements (protection against noise);
- thermal requirements (energy economy and beat retention); and
- service life requirements.
Based upon this set, it is possible to investigate the (inter)relationships between these requirements for prefabricated building components, resulting in the scheme shown in Figure 2
First, there exists an important relationship between the thermal performance of a VIP-incorporated building component on the one hand and its structural performance on the other hand. Whether this relationship is mainly determined by the properties or behaviour of the VIP itself, or by the edge spacer construction of the component, primarily depends on the mechanical behaviour of the component, i.e. does it structurally act as a sandwich construction or as an “edge-spacer” construction, e.g. double-glazing (Figure 3). In prefabricated building components, VIPs until now have only been applied in edge-spacer constructions, like cavity-filled double-glazing panels or building panels derived from double-glazing, not considering the sheet based VIPs sometimes called vacuum insulating sandwiches (Willems, 2003).
In these edge-spacer constructions, structural behaviour is completely determined by a combined action of this edge spacer and the outside or top facing. Depending on the stiffness ratio and the character of the joint between both elements, the component can be typified structurally as a framework or as a configuration of non-connected plates. As a consequence, both the top facing and the edge spacer need to be dimensioned accordingly, in most cases resulting in additional thermal losses at the panel edge. Proper detailing and choosing the right materials can on the one hand minimize the influence of this thermal bridge significantly, but also influence the structural performance of the component as a whole. For sandwich constructions with a VIP core, however, no structural edge spacer is required, on the one hand resulting in the possibility of reducing thermal bridging due to this edge spacer, but on the other hand imposing additional structural requirements on the VIP itself. Both examples already indicate the significant interaction between thermal requirements and performance on the one hand and structural requirements and performance on the other hand.
Second, the relationship between thermal performance and VIP service life is important. The service life of a VIP is typically defined as the elapsed time from the moment of manufacturing until the moment the thermal conductivity of the core material, λ
c, has increased to some critical value, λ
lim, often set at a value of 8 × 10−3 W m−1 K−1 (Simmler et al., 2005) or at 11 × 10−3 W m−1 K−1 (ASTM C 1484, 2001). This core material thermal conductivity is thus not a constant property but increases over time because atmospheric gases and water vapour continuously permeate through the low permeation barrier envelope. This process is called thermal conductivity aging. Since, the thermal performance of a VIP façade component or a VIP-integrated building construction primarily depends on the centre-of-panel thermal conductivity combined with a geometry-based multiple of the linear thermal transmittance of the component edge
Thermal performance
The thermal performance of any insulation material depends on how well heat transfer is prevented within the material. The total heat transfer in any porous material can be divided into three different heat transfer processes: heat transfer via radiation, via the solid skeleton of the core and via gas inside the material. Further, the transport through the gas can be divided into gas convection and gas conduction. Gas convection arises from bulk movements of the gas and gas conduction is the interaction between molecules in the gas. A VIP is, therefore, evacuated to suppress both convection and gas conduction within the core. The small pore size in the core in conjunction with this low pressure restricts gases from any bulk movement (convection) within the remaining gas. Besides, to prevent gas conduction the mean free path (the average distance travelled by a particle between interactions with other particles) of the gas molecules has to be larger than the pore size of the core material, implying that the smaller the pore size, the higher the pressure allowed until gas conduction occurs. Figure 4 shows this dependency of thermal conductivity on pore gas pressure for several potential VIP core materials. As can be seen, fumed silica is among the insulation materials for which the jump in thermal conductivity occurs at a relatively high pore gas pressure, making it a suitable option for VIPs. For this material, a pressure of approximately 100 mbar is allowed before the centre-of-panel thermal conductivity has increased to approximately 8 × 10−3 W m−1 K−1. Conduction within the core on the other hand is limited by the geometric shape of the particles in the material. Glass fibres, for example, are tubular, while most powders have spherical particles leading to a very small area of contact between the particles. Since, conduction needs to pass through this area of contact, the total core conduction will be very small. Finally, all surfaces emit and receive radiation, the amount of which depends on the temperature, surface properties and geometric positioning in regard to other surfaces. This radiative exchange is minimized in VIP cores by added opacifiers or scattering molecules.
A VIP is not a single material; it is a system of materials with a core material and another material or composite wrapped around the core functioning as a gas barrier. Independently of what material is used in the barrier or the core, the barrier will create a thermal bridge when wrapped around the edge of a panel, as shown in Figure 5. The packaging design used today with flexible envelopes, heat-sealed at the short edges of the panel, creates one flange on at least two ends of the panel. When this flange is bent towards any of the surfaces the thickness of the highly conductive metal or metallization is tripled as a consequence increasing the thermal bridge effect.
This thermal bridge can increase the overall heat transfer by a factor of two compared with the centre-of-panel value of VIP with barriers made of metallized polymer composites or polymer and foil combinations (Glicksman, 1991; Ghazi Wakili et al., 2004). To calculate a correct mean thermal transmittance, or U-value, for a vacuum panel, the edge and corner effects must be considered, as can be done with the following formulation: Equation 1 in which U m (W m−2 K−1) is the mean thermal transmittance per area panel, U c (W m−2 K−1) is the thermal transmittance in the centre of the panel per area panel, A (m2) is the area of the calculated panel, Ψ edge (W m−1 K−1)) is the linear thermal transmittance of the edge, l (m) is the perimeter of the calculated panel and X corner (W K−1) is the point thermal transmittance in the corners.
Figure 6 shows how the mean U-value is affected by the size and shape of the panel. The solid graph shows the relation for a square panel, hence both sides are of equal length, while the broken line shows the relation for a panel for which one side is five times longer than the other side and finally the dotted line shows the relation for a panel for which the relation between the sides is 1 and 10. These calculations were carried out with a linear thermal transmittance of 0.006 W m−2 K−1 as measured by Ghazi Wakili et al. (2004) for a good panel design available on the market today.
To have such a low heat loss over the edge, however, metallized polymer barriers have to be used with very thin metallized layers. Thin metal layers, however, lead to higher risk of cracks and pinholes resulting in a higher permeation rate of gases through the envelope. It is a trade-off between thermal performance and, in the end, service lifetime, unless some specially designed edge element like the serpentine edge proposed by Thorsell and Källebrink (2005) is used. As has been shown, it is of more importance to choose large panels than to choose square panels. As long as the chosen panel is not smaller than 0.5 m2, the added heat loss because of edge effects will be limited to less than 20 per cent for metallized polymer barrier films.
Any thermal bridge in a building component would be of greater relative importance if the surrounding insulation where to be replaced with super insulation, even though the overall performance would be improved. In this case larger surface temperature differences along the surface would have to be expected than before the substitution. Such differences can lead to thermal discomfort for the inhabitants of the building. Any unavoidable thermal bridge should, therefore, be designed thoroughly so that the temperature gradient over this thermal bridge is as small as possible and so that the risk of moisture problems is minimized. Also the built-in thermal bridges of the VIP must be taken into consideration when designing VIP-integrated building constructions, especially in respect to moisture risk and surface temperatures since the edge loss will in most cases add to any other adjacent thermal bridge.
For a thick layer of VIP, one solution to minimize thermal bridging can be to use two or more thinner layers of panels which are positioned in such a way that the joints of the panels are displaced. An outer wall of a building, however, usually needs to have a certain minimum thickness to hold any necessary load bearing structures. This means that there is space to combine vacuum panels with traditional insulation on at least one side. Since, additional insulation will lower the temperature gradient over the panel, the importance of the heat loss along the edge will decrease, as shown in Figure 7. The broken line shows added insulation on one side of the panel while the solid line shows added insulation on both sides of the vacuum panel.
Functional service life
The functional service life of a VIP incorporated building element or component mainly depends on the thermal conductivity aging of the VIP inside, since the properties of the remaining elements used for the component do not seriously deteriorate over time with respect to their thermal behaviour, if the component is designed properly. The definition of the functional service life, t SL, of a VIP-integrated building component is, therefore, more-or-less equal to the service life definition of the VIP itself. Two different service life definitions can be distinguished, the first of which is the most widely used with respect to vacuum insulation.
Service life definition 1
The elapsed time from the moment of manufacturing until the moment the thermal conductivity of the core material, λ
c, has increased to some limiting value
Service life definition 2
The time that expires from the moment the panel is manufactured until the moment the time-averaged thermal conductivity of the core material equals some critical value, λ critical; or the service life has expired if: Equation 3
With this second definition, it is possible to take into account that the thermal conductivity increase rate varies over time due to a non-linear correlation between thermal conductivity and pore gas pressure (Figure 4) as well as due to a (theoretically) non-linear increase of pore gas pressure with time, at least over longer periods of time. This second definition results in the possibility to set a limit for the total heat loss through the building construction including thermal bridges during the entire VIP service life, while the first definition just limits the instantaneous heat loss.
The increase in thermal conductivity, or actually the increase in pore gas pressure and water content, is principally determined by the core-material properties as shown in Figure 4, the presence of getters and desiccants, the initial vacuum and water content, the envelope permeance for atmospheric gases and water vapour, the panel dimensions and the environmental conditions regarding temperature, relative humidity and atmospheric pressure. Since, the first four factors are product-related and cannot be influenced by architects and building engineers, especially the last two factors are interesting from the perspective of construction design. Both will be discussed in more detail in this paragraph.
For the effect of panel dimensions on service life, both the thickness and the ratio of panel perimeter length to surface area need to be considered, since both factors influence the volume in which gases can be stored and the envelope or edge area through which gases can permeate. To discuss the effect of panel dimensions on service life in more detail, we must distinguish between the effect of water vapour and the effect of atmospheric gases on the service life. Since, the permeance of high barrier films for oxygen and nitrogen is very low (in the order of 5 × 10−4 cm3 m−2 day−1 for oxygen for a triple layer metallized film) (Hanita Coatings, 2004), oxygen and nitrogen permeation primarily occurs through the envelope seam, implying that both the panel thickness and the panel perimeter length (or actually the ratio of perimeter length to surface area) are important for the service life. With respect to water vapour, however, permeation through the surface area and through the envelope seam are of the same importance for typical panel sizes resulting from the higher barrier film permeance for water vapour (in the order of 0.01 g m−2 day−1 for a triple layer metallized film) (Hanita Coatings, 2004). This implies that especially for larger panels the thickness is important, while for small panels both the thickness and the panel perimeter length (or the ratio of perimeter length to surface area) are equally important. In general, thick and large panels have a longer expected service life than thin and small panels. So, in the design process of building constructions with VIPs, panels need to be designed as large as possible with a ratio of panel perimeter length to surface area as high as possible, also because of the reduction of the thermal edge effect.
The environmental operating conditions regarding temperature, relative humidity and atmospheric pressure are the second service-life influencing factor that can be affected by building designers. In standard laboratory tests to determine the VIP service life, some standard steady-state test conditions have been used for measurements conducted at EMPA, ZAE-Bayern and NRC in Switzerland, Germany and Canada, respectively, (Simmler et al., 2005). These test conditions can be classified into two categories: 80°C and 80 per cent RH, 30°C and 90 per cent RH, 65°C and 90 per cent RH for extreme conditions to determine upper limits and 23°C and 50 per cent RH for average conditions to be used for, e.g. quality assurance measurements. These measurements give estimates of the service life of a VIP under laboratory conditions, which are, however, not similar to actual conditions in building constructions.
To investigate the service life of a VIP in a building component or construction, it is important to take actual temperature and relative humidity fluctuations inside this component into account. Based on simulations or measurements of the relative humidity and temperature inside a building construction over one year, laboratory determined VIP service lives can be corrected for fluctuations in the partial water vapour pressure around the VIP by means of an average load due to water vapour calculated as: Equation 4 in which t (s) is time, A i (m2) is a surface area (i=1: front surface; i=2: back surface; i=3: panel sides), A tot (m2) is the total panel surface (enveloping) area, p v (Pa) is the partial water vapour pressure surrounding the VIP and p v;0 (Pa) is a reference partial water vapour pressure. For a vapour-tight façade component under Dutch climatic conditions, numerical simulations and calculations according to equation (4) have resulted in an average load due to water vapour of 0.51 and 0.46 for a south facing façade component and a north facing façade component with an average U-value of 0.2 W m−1 K−1, respectively, if the temperature on both the inside and the outside of the component equals the outside climatic temperature (Tenpierik and Cauberg, 2005). If the temperature on one side of the panel is kept constant at 20°C, while the temperature at the other side varies according to the climatic conditions referred to, this annual average load because of water vapour increases to 0.80 and 0.78, respectively. For climates with higher average temperatures and for façade elements in which the internal relative humidity follows the external relative humidity more closely (completely vapour-open façade components), this relative load will be even higher. So, by designing a building construction in such a way that the relative humidity in this component is kept low and relatively stable, expected service lives of the VIP can be prolonged to periods of over 20-25 years, which is currently the expected service live of VIPs with a triple layer metallized film envelope under standard laboratory conditions (23°C and 50 per cent RH).
Structural performance
Façade components with or without VIPs can generally be divided into two main categories based on their structural action: “edge-spacer” constructions and sandwich constructions (Figure 3). In the first component, structural action is completely performed by the facings in combination with the edge spacer, while in the latter component the facings together with the VIP contribute to this action. Thus, while for edge-spacer components the mechanical properties of a VIP are not relevant, these properties need to be known for sandwich components.
Table I, therefore, presents results of three and four-points flexural tests according to ASTM C 393 (1992) conducted on 20 mm thick evacuated and non-evacuated panels, the values of which can be compared to textbook values of the Young's modulus of alternative sandwich core materials. Typical core materials for sandwich panels are expanded polystyrene, extruded polystyrene (XPS) and PU that have flexion moduli of 1-3, 6-10 and 3-8 MPa, respectively. The Young's modulus of an evacuated VIP is thus significantly higher than the flexural modulus of alternative sandwich core materials. Since, however, the component flexural stiffness, is primarily determined by the Young's modulus of the component facings and their area moment of inertia, the importance of the VIP flexural modulus is rather small.
As a consequence, this principally results in the requirement of a minimum distance (or actually a minimum area moment of inertia) between the facings mainly depending on the flexural modulus and thickness of these facings. In this respect VIPs have a disadvantage over conventional thermal insulators as core material for sandwich panels, since thermal requirements result in a larger thickness of conventional materials relative to VIPs, at least for equal thermal resistances. As a consequence, if the thermal properties of vacuum insulation are to be exploited completely, resulting in façade components with a very thin core of for example just 2.0 cm thick (R-value≈5.0 m2 K W−1), facings with a high flexion modulus and of sufficient thickness need to be selected. If, for example, for these sandwich components glass facings with a modulus of elasticity of 70,000 MPa are used, 3 mm thick glass facings on both sides would structurally suffice for a component of 1.2 × 1.8 m2 and a wind load of 2.0 kN m−2. In these calculations the effect of shear in the core material is included; shear in the adhesive layer between component facing and VIP, however, is not considered. This example clearly shows that it is theoretically possible to construct a VIP-integrated sandwich panel with sufficient stiffness to be applied as façade component.
Until now, however, the limiting factor for a successful production of sandwich panels with a vacuum insulation core has particularly been shear in the adhesive between facings and VIP (Figure 8), resulting in too large panel deflections. Not only flexion tests according to ASTM C 393 (1992), but also tensile strength tests in the flatwise plane on sandwich panels with different adhesives according to ASTM C 297 (1992) have shown that this adhesive layer, failing at normal stress levels between 0.05 and 0.11 N mm−2 for test samples of 110 × 110 mm2, is the weakest chain in the sandwich component system due to insufficient adhesion between glue and polyester (PET) top layer of the metallized low permeation barrier film.
Moreover, structural safety and product liability during the service life of a VIP-integrated component may be a serious issue. Table I shows that damaged VIPs are less stiff than undamaged panels, while the ultimate flexural strength of both panels is more-or-less equal. This indicates that the pressure difference caused by the internal vacuum has a significant influence on the Young's modulus but not on the strength of the panel. For practical purposes thus, in the case of a perpendicular to surface loaded VIP, a loss of vacuum will increase the deflection of the panel with a factor of approximately 1.7, but will not cause the panel to fail directly.
Since, however, the core material Young's modulus does not significantly affect the flexion stiffness of the building component as a whole, additional safety precautions are not required because of this reduced stiffness. Indirect failure, however, may become imminent if the component is improperly designed thus slipping out of its grooves. Moreover, wind suction on the façade component could tear loose the low permeation barrier film from the core material in absence of a pressure difference over this film in case of foil damage. Additional safety precautions might, therefore, be desirable. Such a safety mechanism could simply be a thin but strong non-metallic reinforced tape at the sides of the component.
Conclusions
This paper has shown the importance of applying an integral approach to designing building components and constructions in which VIPs are integrated. Especially, two important interactions between performance requirements on complete building components, i.e. thermal insulation requirements versus structural performance and thermal insulation requirements versus functional service life, have been investigated thoroughly, since they can be considered of special interest for VIP-integrated building components. Based upon this holistic view, a number of recommendations for designing such panels can be derived.
In general, due to thermal bridging of high barrier films enveloping a VIP and due to the importance of edge seams on the total envelope permeance for atmospheric gases, panels must be designed as large as practically possible with a ratio of panel perimeter length to surface area as high as possible. The thermal bridge effect can also be minimized by developing VIPs without edge seams (seams along the surface) or by using an alternative edge design like the serpentine edge (Thorsell and Källebrink, 2005). A reduction of the envelope permeance, however, would require both a barrier film with reduced permeation coefficient and a longer overlapping seam, which is contradictory to the aforementioned thermal performance requirement. A similar contradiction exists for the type of barrier film chosen for a VIP. The best performing high barrier films are thick metal-based foils, as a consequence resulting in long expected functional service lives on the one hand, but in large energy losses through the panel edge owing to the high thermal conductivity of these foils on the other hand. It is, therefore, necessary to find the optimal solution for each application regarding service life and thermal requirements. For typical building applications nowadays, metallized polymer films have a permeance sufficiently low to achieve a VIP service life of over 20 years, which can even be prolonged if the relative humidity inside the building component is kept stable at a value as low as possible.
Not only for thermal performance versus functional service life, but also for thermal performance versus structural performance contradictory requirements exist. Edge spacer constructions do not impose requirements on the mechanical properties and behaviour of VIPs. Structural action is completely fulfilled by this edge spacer, which as a result needs to be designed accordingly, generally leading to high thermal edge losses. Although sandwich constructions have edge spacers to protect the VIP from external impact forces, these spacers, however, do not have a structural function, as a consequence of which they can be constructed thin and can be based on low thermal conductivity materials. The application potential of sandwich components, however, stands or falls with the ability of adhering the component facings onto the VIP properly. Until now, this has been the limiting factor for a successful application of VIP-integrated sandwich components. Another important contradictory effect is caused by a loss of vacuum inside the panel, resulting in an increase in thermal conductivity of a factor of approximately 1-5 depending on the stage of aging. Structural safety is, however, not reduced, notwithstanding a decrease in flexion modulus of the VIP with a factor of approximately 1.7.
As a conclusion, an integral approach to the design of VIP-integrated building components and constructions has the potential to generate design solutions that are optimal for a certain application.
Equation 1
Equation 2
Equation 3
Equation 4
Figure 1Typical VIP with barrier film removed at one corner
Figure 2Interrelationships between different requirements for VIP incorporated building panels
Figure 3Structural action of a sandwich component (a) and an “edge-spacer” component (b)
Figure 4Pore gas pressure as a function of thermal conductivity for different core materials
Figure 5VIP cross-section and resulting heat flows
Figure 6Overall thermal performance of a VIP as function of panel size
Figure 7The influence of adjoining insulation on the linear thermal transmittance (ψ-value)
Figure 8Flexural behaviour of a sandwich without (above) and with (below) shear in the adhesive layer
Table IFlexural mechanical properties of VIPs
References
ASTM C 297 (1992), "Standard test method for tensile strength of flat sandwich constructions in flatwise plane", Annual Book of Standards, American Society for Testing and Materials International, West Conshohocken, PA, Vol. Vol. 15.03.
ASTM C 393 (1992), "Standard test method for flexural properties of flat sandwich constructions", Annual Book of Standards, American Society for Testing and Materials International, West Conshohocken, PA, Vol. Vol. 15.03.
ASTM C 1484 (2001), Standard Specification for Vacuum Insulation Panels, American Society for Testing and Materials International, West Conshohocken, PA, .
Binz, A., Moosmann, A., Steinke, G., Schonhardt, U., Fregnan, F., Simmler, H., Brunner, S., Ghazi, K., Bundi, R., Heinemann, U., Schwab, H., Cauberg, J.J.M., Tenpierik, M.J., Jóhannesson, G.A., Thorsell, Th.I., Erb, M., Nussbaumer, B. (2005), "Vacuum insulation in the building sector. Systems and applications (subtask B)", Final report for the IEA/ECBCS Annex 39 HiPTI-project, .
Bundi, R. (2003), "Vakuumisolierte Paneele. Forschung und Entwicklung im Fassadenbau", Fassade, Vol. 3 pp.19-22.
Caps, R., Heinemann, U., Ehrmanntraut, M., Fricke, J. (2001), "Evacuated insulation panels filled with pyrogenic silica powders: properties and applications", High Temperatures- High Pressures, Vol. 33 pp.151-6.
European Council (1988), "The construction products directive", Council Directive 89/106/CE, Strasbourg, December 21, .
Ghazi Wakili, K., Bundi, R., Binder, B. (2004), "Effective thermal conductivity of vacuum insulation panels", Building Research and Information, Vol. 32 No.4, pp.293-9.
Glicksman, R. (1991), "Two-dimensional heat transfer effects on vacuum and reflective insulations", Journal of Thermal Insulation, Vol. 14 pp.295-310.
Hanita Coatings (2004), High Barrier Laminate for Vacuum Insulation Panels, Hanita Coatings, Kibbutz Hanita, January, .
Nussbaumer, T., Ghazi Wakili, K., Tanner, Ch. (2006), "Experimental and numerical investigation of the thermal performance of a protected vacuum-insulation system applied to a concrete wall", Applied Energy, Vol. 83 No.8, pp.841-55.
Nussbaumer, T., Bundi, R., Tanner, Ch., Muehlebach, H. (2005), "Thermal analysis of a wooden door system with integrated vacuum insulation panels", Energy and Buildings, Vol. 37 pp.1107-13.
Rath, D. (1989), Wärmetransport in evakuierten Dämmsystemen, Universität Dortmund, Dortmund, .
Schwab, H., Heinemann, U., Beck, A., Ebert, H-P., Fricke, J. (2005a), "Permeation of different gases through foils used as envelopes for vacuum insulation panels", J. of Thermal Env. and Bldg. Sci., Vol. 28 No.4, pp.293-317.
Schwab, H., Heinemann, U., Beck, A., Ebert, H-P., Fricke, J. (2005b), "Dependence of thermal conductivity on water content in vacuum insulation panels with fumed silica kernels", J. of Thermal Env. and Bldg. Sci., Vol. 28 No.4, pp.319-26.
Schwab, H., Heinemann, U., Wachtel, J., Ebert, H-P., Fricke, J. (2005c), "Predictions for the increase in pressure and water content of vacuum insulation panels (VIPs) integrated into building constructions using model calculations", J. of Thermal Env. and Bldg. Sci., Vol. 28 No.4, pp.327-44.
Schwab, H., Heinemann, U., Beck, A., Ebert, H-P., Fricke, J. (2005d), "Prediction of service life for vacuum insulation panels with fumed silica kernel and foil cover", J. of Thermal Env. and Bldg. Sci., Vol. 28 No.4, pp.357-74.
Schwab, H., Stark, C., Wachtel, J., Ebert, H-P., Fricke, J. (2005e), "Thermal bridges in vacuum-insulated building facades", J. of Thermal Env. and Bldg. Sci., Vol. 28 No.4, pp.345-55.
Simmler, H., Brunner, S. (2005), "Vacuum insulation panels for building application. Basic properties, aging mechanisms and service life", Energy and Buildings, Vol. 37 pp.1122-31.
Simmler, H., Brunner, S., Heinemann, U., Schwab, H., Kumaran, K., Mukhopadhyaya, Ph., Quénard, D., Sallée, H., Noller, K., Kücükpinar-Niarchos, E., Stramm, C., Tenpierik, M.J., Cauberg, J.J.M., Erb, M. (2005), "Vacuum insulation panels. Study on VIP-components and panels for service life prediction of VIP in building applications (Subtask A)", Final report for the IEA/ECBCS Annex 39 HiPTI-project, .
Tenpierik, M.J., Cauberg, J.J.M. (2005), "Vacuum insulation panels in building facades: moisture and temperature conditions during insolation", in Jóhannesson, G.A. (Eds),Proceedings of the 7th Symposium on Building Physics in the Nordic Countries, The Icelandic Building Research Institute, Reykjavik, 13-15 June, pp.937-44.
Thorsell, Th.I., Källebrink, I. (2005), "Edge loss minimization in vacuum insulation panels", in Jóhannesson, G.A. (Eds),Proceedings of the 7th Symposium on Building Physics in the Nordic Countries, The Icelandic Building Research Institute, Reykjavik, 13-15 June, pp.945-52.
United Nations Framework Convention on Climate Change (2005), Kyoto Protocol, United Nations, Bonn, available at: http://unfccc.int/ (accessed 18 July), .
Willems, W.M. (2003), "Zur Dauerhaftigkeit ausgewählter Vakuum-Dämmsysteme", Proceedings of VIP-Bau. Vakuum Isolations Paneele. Evakuierte Dämmungen im Bauwesen, Universität Rostock, Rostock-Warnemünde, 10-11 July, pp.P1-P14.
Corresponding author
Martin Tenpierik can be contacted at: m.j.tenpierik@tudelft.nl