Unitized Cavity Wall Design and Implementation

Nomenclature

T Temperature

t Time

x_i Spatial Coordinate i

ρ Mass Density

U_i Fluid Velocity Along Coordinate i

p* Modified Pressure

μ Dynamic Viscosity

μ_t Turbulent Viscosity

g_i Gravity

ϐ Coefficient of Thermal Expansion

k Turbulent Kinetic Energy

ω Dissipation of Turbulent Kinetic Energy

_k Turbulence Production

P_Kb Buoyancy Turbulence Production of k Model

P_ϵb Buoyancy Turbulence Production of ϵ Model

F₁ Blending Function

1.0 Introduction

The design of unitized aluminum framed curtain wall borrows many lessons from architectural history. The first aluminum framed systems were motivated largely by efficiency of production and the use of extruded aluminum as light durable material that possessed strength to span large areas with slender mullions. The resulting “glass” walls provided an enticing esthetic with broad exterior views and access to natural daylight. Despite these advantages, innovation did not occur without considerable challenges. Among those are the high inherent heat conduction of aluminum alloy, the “greenhouse” tendency of glass enclosures to induce significant solar gain and propensity of lightweight and stiff diaphragms to transmit sound. It is only with the passage of many years that increasingly elaborate solutions to these problems have been established. The conductive nature of aluminum was addressed by the advent of polymer “thermal breaks” placed between aluminum profiles and proliferation of insulated glass which provides an insulating air cavity between glass lites. Solar control was addressed with coloration, reflective coatings and eventually visually transparent metal depositions tuned to reject irradiation while still transferring visual light. Adaptation of acoustic engineering specifications such as sound transmission class, and later outdoor-indoor transmission class, fostered an appreciation amongst designers for the thickness of glass lites, the ratio of thickness between inboard and outboard lites, gas cavity thickness and the use of viscoelastic materials such as PVB to improve sound damping.

Despite this progress the appetite for performance has not diminished and the evolution of technology continues to influence architectural practices. Curiosity naturally leads us to question what might be accomplished with multiple façade layers? In fact, this inquiry is merely an echo of the sort of inquisition designers have considered for more than a century. The use of shutters for privacy and security predates the use of window glass and they are still sometimes used in conjunction with glass for the same purposes. Attentive benefits of the additional façade layer include increased thermal resistance and sound attenuation. Later, the use of storm windows became common place to achieve these benefits without disrupting daylighting. Dual-layered curtain wall systems, commonly referred to as double skinned façades (DSFs) are simply an extension of this philosophy applied to unitized curtain wall systems. When deployed effectively they have been shown to provide better thermal and acoustic performance than is possible with conventional single layer systems.

The merits of DSFs are weighed in various publications. For the purpose of this paper it suffices to observe that DSFs have been implemented on various projects throughout the world and are no longer considered to be an architectural novelty. A common implementation, that will serve as the focus of this study, is to produce single unitized assemblies which are glazed at both their interior and exterior extents with a larger internal cavity. This approach is illustrated in Figure 1. Shading devices are often included between the inner and outer glazing. Contractors responsible for furnishing systems of this nature must entertain special considerations throughout the process of design and implementation to ensure the successful outcome of project objectives. This paper explores these topics for the sake of conveying practical knowledge.

Figure 1 Unitized Cavity Wall

During the planning stages of a given project care must be taken to consider:

• Design
  • Performance Goals
• The Influence of Environmental Conditions that are Unique to a Given Project
• Integration of the Façade with Building Facilities
• Logistics
  • Storage and Handling of Materials During Manufacturing
• Shipment, Installation and Protection of Installed Units

2.0 Initial Design

The most common reason for implementing a cavity wall design is to improve the acoustic performance of the façade. Attentively, harnessing the added benefit of enhanced building energy efficiency typically also receives focus. These requirements are commonly conveyed in the form of ratings governed by engineering standards. For instance, Sound Transmission Class (STC) [1] and Outdoor Indoor Transmission Class (OITC) [2] values are commonly specified based on data from ASTM E90 [3] testing. Furthermore, energy performance is commonly expressed in requirements for fenestration U-Factor, Solar Heat Gain Coefficients (SHGC) and Visual Light Transmittance (V_t). In the United States, these values are commonly determined by numerical simulation and testing in accordance with NFRC100 [4].

2.1 Acoustic Performance

STC and OITC values can only be accurately determined by testing since the engineering standards invoked do not include a formal calculation procedure to complement test-derived data. Efforts using software to predict these values must be performed at the discretion of the designer and will achieve varying degrees of success in accordance with necessary assumptions. Traditional curtain wall systems, which implement insulated glass units (IGUs) benefit from the fact that test data is readily available from manufacturers for numerous make ups. It is reasonable to assume as rough approximation that STC values for a framed assembly will fall 3-6 dB below the rating for the glass. Since cavity wall systems include a large interstitial airspace, there is far less data available to motivate predictions. In the event that such data is not available for a proposed assembly, and calculations must be relied upon, commercial programs such as INSUL [5], produced by Navicon Engineering, can provide rough approximations. Such results are expected to be accurate within +/- 5 dB for simple glass make ups. If the targeted performance for a given project cannot be compromised, the complete glass make up should be pre-tested in the same laboratory that will be used for project specific samples to ensure that a conservative rating is obtained. It is generally true that the performance achieved by cavity wall systems will exceed that of conventional designs due to the effect of mass-air-mass resonance which occurs at increasingly lower frequencies as the gap width is increased. On the order of thickness typically associated with a double-skin system the resonance value falls below the range of interest associated with STC or OITC values. Other established principles of acoustic design apply in the same manner that they would for single-leaf make ups. Generally, increasing the mass of the system, varying the thickness of lites used in the inner versus outer leaf, and employing laminates will improve sound attenuation. A consideration unique to double wall design, may stem from the use of operable lites or vents in either leaf. Fixed exterior venting may significantly reduce the sound attenuating properties of the system.

2.2 Thermal Performance

The energy efficiency of a proposed construction lends itself more amiably to numerical prediction than acoustic performance as a consequence of the equations solved. In order to solve acoustic problems with a very high degree of accuracy both vibration within structural elements and acoustic pressure variations in air, both of which occur at very fine length scales, must be solved as a coupled problem. While this is technically possible, the process is cumbersome requiring vast computer resources that are not practically available to most designers. Most practitioners would agree that it exceeds the conventions of the industry. Compounding this difficulty, solutions are sensitive to aspects of a given design that are difficult to measure such as clearance between gaskets and complicated airborne pathways that commonly exist in mullion cavities. Heat transfer and fluid-flow simulations are, in some cases, inherently less burdensome and supported by a longer history of being incorporated into engineering standards. For example, the venerable NFRC 100 procedure for determining fenestration U-factors, employs a specified calculation procedure that is performed using Finite Element Analysis software.

2.2.1 Thermal Simulation

Thermal simulation of cavity wall systems requires an appreciation for the legacy of existing engineering standards and nature of the wall systems for which their assumptions are tailored. For instance, the Therm software program developed evoked by NFRC100 allows routine engineering calculations to be performed in a time efficient manner by discretizing a single governing equation of heat transfer (1). This equation assumes that heat transfer is governed by diffusion and that heat transfer which occurs as a result of fluid movement need not be explicitly solved for. Instead, the effect of fluid motion is merely assumed to contribute to the diffusive process by means of observing an elevated “effective conductivity,” k_eff per the method described in ISO15099 [6] of S.M. ElSherbiny et. Al. [7]. This quantity is fictious since fluid does not flow along temperature gradients. It is deemed to represent an acceptable assumption because the fluid regions simulated are typically, “confined” such as narrow gaps in insulated glazing units or small frame cavities where fluid motion is not a powerful heat transfer mechanism, limiting the resulting error. Furthermore, it is appropriate to consider the historical context corresponding to which the software was initially developed which would have precluded significant advances in the performance of desktop computers we currently enjoy. It would have been difficult for those responsible for adopting the assumption to foresee practical fluid-flow simulations as a reasonable requirement of an engineering standard.

Steady state equation of diffusive heat transfer

Numerical simulation of fluid flow, commonly referred to as computational fluid dynamics (CFD), observes a broader set of governing equations (2,3,4 & 5). The resulting predictions appreciate convective heat transfer as consisting of both, the transfer of heat along gradients in temperature due to the conductivity of the fluid and the bulk motion of energy carried by fluid particles as they move.

Reynolds-Averaged Navier Stokes Equations with Buoyancy

Pseudo-Conservation Equations of Turbulence Production and Dissipation according to the Shear Stress Transport Model – [7] Mentor (1994)

Conservation of Mass

Conservation of Energy

Clearly Equations 2-5 complicate the process of obtaining a solution when compared to Equation 1; however, modern commercial CFD programs have enabled the task to be dealt with more readily than could have been imagined twenty or even ten years ago. Since Equations 2-5 are non-linear some care is required in ensuring that accuracy is achieved. Elaboration is owed to the fact that these equations represent an incompressible fluid. Since free convection inherently involves temperature induced density gradients in the presence of gravity it is correct to question how they would be appropriate. This is accounted for by including a source term to account for these variations on the pressure gradient as a function of Temperature (6). This approximation is credited to Boussinesq and is typically accurate since free convection is generally not associated with high velocities where density variations would affect the inertia of the fluid. Other modifications are necessary to account for the nature of turbulence.

A highly detailed discussion of the mathematical model would be voluminous and distract from the broader focus of this paper. Sparing information that would not aid in providing a conceptual understanding of the equations it suffices to describe each one generally. Equations 2 & 4 are fundamental equations of fluid flow that are derived by observing Reynold’s Transport Theorem, which allows the dynamic behavior of individual fluid particles to be applied to those flowing through a control volume, thereby establishing equations of conservation of momentum (eq. 2) and conservation of mass (eq. 4). These equations are well described in most introductory textbooks on the subject of fluid mechanics. The equations of turbulence (eq. 3) cannot be derived from, “first principles” in a similar manner. They are a mathematical model representing very small fluctuations in velocity that cannot be practically resolved by the grid of elements employed for the numerical approximation. The flow velocities U_i shown in equation 2 are mean velocities when the turbulence model is employed. The fictitious quantity μ_t is derived by a method of substitution. It is first assumed that the fluid velocity consists of both a mean velocity U_i and a harmonic fluctuating component U_i’ such that the velocity at any time is V_i = U_i + U_i’. Substitution of this quantity into (eq 2.) and time averaging each equation results in cancelation of many of the alternating terms since the time average of a harmonic wave is zero. Therefore, the only remaining fluctuating terms can be regarded as contributing to pseudo shear stresses represented by the fictitious quantity μ_t. The study of computational fluid mechanics has produced many approximations of this artificial quantity for the sake of allowing turbulent flow problems to be solved with conventional computers. Their accuracy has proven to be sufficient for solving a wide range of practical problems in the fields of aerospace and mechanical engineering. The model chosen for this paper is the Shear Stress Transport (SST) model proposed by Mentor [8]. It is a “hybrid” two-equation model where μ_t is calculated from the rates of turbulence production _k and turbulence dissipation ω. These equations are determined by fit with experimental data. The SST model is a hybrid model in the regard that it blends a different two-equation models as a function of proximity to walls. In the bulk stream, away from walls, the venerable k-ϵ model of Launder and Spaulding is employed for the sake of efficiency; however, near walls the k-ω model of Wilcox is used which provides higher resolution of turbulence quantities. It is well known that gradients of velocity and turbulence quantities are exceptionally high near walls. Blending of these models is achieved with a blending function F₁. Lastly, Eq. 5 is included to account for conservation of thermal energy within the flow for the sake of establishing temperature distributions and local rates of heat transfer.

It is reasonable to question whether the foregoing discussion is merely academic and whether it is significant to the broader discussion of cavity wall design? Comparison of model outcomes for the temperature field predicted within a cavity wall enclosure provides glaring cause for designers to appreciate the resolution provided by the CFD model. Since temperature stratification is the result of free convection, which is not a diffusive process, using a program like Therm to calculate the temperature distribution within the cavity may result in significant errors.

To illustrate this point, an example problem is presented here which models heat transfer in a closed cavity wall. The solution domain and boundary conditions are shown in Figure 2a. Three cases were considered. The first two employed CFD to solve the governing equations. Specifically, Ansys CFX commercial software were used. In the first case, turbulent free convection was specified employing equations 2 through 5. In the second case, a simpler model of laminar free convection was observed. The equations solved are similar to those of the previous case, but without the turbulent viscosity production and dissipation equations 3a & 3b and μ_t = 0. It is appropriate to consider both models since transitional turbulence will occur in a transient problem both temporally and spatially. The final case employs FEA using Therm software. Therm employs the steady state equation of heat transfer [1] with an effective conductivity predicted per the method of of S.M. ElSherbiny et. Al. [7].

Figure 2 (a) Solution Domain and Boundary Conditions

Figure 2 (b) Temperature Distributions and Heat Transfer Predicted by CFD and FEA Models

The outermost boundary is assigned a uniform conductance h_e = 3.6 [hr-sf-F/Btu] corresponding to a laminated outer light and 15 MPH exterior wind. The inner boundary is assigned h_i = 1.8 [hr-sf-F/Btu] which represents a 3/8” lite with natural convection occurring along the interior of the cavity wall. Upper and lower most extents of the solution domain are adiabatic. The exterior temperature is 0 [F] and the interior temperature is 70 [F]. The dimensions of the cavity are 6’-2” in height by 9-5/8” in depth.

The color contour diagrams on the left side of figure 2b indicate the CFD results, whereas the result predicted by Therm is shown on the right-hand side. It is readily apparent that results of the latter model are invariant in along the vertical coordinate, whereas substantial temperature stratification is correctly resolved by the former CFD results. Figure [3] depicts the temperature distributions along the inner surface of the outer light of glass. It is noteworthy that the glazing temperature near the base of the cavity is over-predicted by the Therm model which significantly impairs its use for certain types of analysis including assessment of condensation resistance. For the CFD cases, perturbations in the temperature distribution are resolved by the laminar model which corresponds to the presence of large eddies. These flow features are not predicted by the turbulent model of free convection which produces smoother variations. The temperature field otherwise compares similarly over the total height of the wall cavity.

The ratio of heat transfer occurring at the inner and outer leaf was simulated until steady state heat transfer was confirmed. This condition was found to occur at approximately 15 minutes of simulated time for both CFD cases. The rate of heat transfer q converged to a value of approximately 18.64 Btu/hr-ft² for the SST case and 18.01 Btu/hr-ft²-F for the laminar case. Initially, heat flows through the exterior lite at a greater rate since the initial condition was specified to be the interior temperature. As air within the cavity cools and mixes heat transfer begins to occur at the inner lite until the rates of heat transfer at each lite are equivalent and a measure of steady state heat transfer is achieved. The effective thermal conductivity of the cavity is henceforth calculated from the steady state rate of heat transfer, the width of the cavity and difference in average temperature calculated from each lite. These values are found to be k_{CFD, SST} = 0.27 Btu/hr-ft-F and k_CFD,laminar = 0.26 Btu/hr-ft-F_{, SST} whereas the Therm model produces a value k_T = 0.28 Btu/hr-ft-F per the temperatures calculated at each boundary and the effective conductivity model of S.M. ElSherbiny et. Al. [7] which is based on experiments. It can subsequently be inferred that use of the effective conductivity model is appropriate for approximating heat transfer across the wall cavity, it does not reliably predict temperature distribution within the cavity when the mode of heat transfer is modeled as pure conduction. Furthermore, it suggests that the use of CFD can reliably approximate the experimentally determined rate of heat transfer measured by the technique of S.M. ElSherbiny et. Al. [7], which is a measure of validation.

Figure 3 Quasi-Steady State Temperature Distribution Along Outer Lite Predicted by CFD and FEA Models

2.2.3 Condensation Resistance

Determining the temperature distribution within the façade cavity is critically important since the space is generally not tempered and will be colder than the interior of the building during heating months. It is during this period that condensation is most likely to occur. While effective, heating the cavity space throughout the winter mitigates the insulating benefit of employing a double façade and is both complex and expensive to implement. Furthermore, it is impractical to seal the space off from interior humidity due to joints in the system and the fact that long term resilience of vapor impermeable sealants, such as butyl, are difficult to guarantee. Thus, from a practical standpoint, it is common that the façade cavity must be regarded as being interior-exposed, sharing a dew point with the building interior. The reduction in temperature caused by separation of the façade cavity from the interior enhances condensation risk.

The project environment imposes a juncture on the design direction that is most appropriate to pursue. Either the façade cavity will remain warm enough that interior condensation is not a concern. Otherwise, a strategy for condensation mitigation needs to be enacted. This fact highlights an important aspect of cavity wall design which is that there is no, “one size fits all” approach to system design. For instance, highly complicated mechanical designs may not be a cost efficient solution in warm climates and passive systems, with no humidity control, would not be appropriate for cold climates. Clearly, the interior humidity associated with the form of occupancy, glazing and other variables factor into this assessment.

For a custom designed system, modeling during the initial phases of the project will reveal whether or not the façade cavity is expected to fall below the dew point temperature. This outcome will need to be verified by testing at a later time, so some level of judgment is required when establishing the modeling assumptions. If condensation is predicted without a mitigation strategy then the design will need to include one. Mitigation strategies can be either active or passive as described in Table 1 (a) & 1 (b). Generally speaking, passive systems are simpler to implement on account of the fact that they do not rely on integration with building systems to ensure condensation resistance is achieved. On the other hand, active systems require an energy source and will impact the work of other trades such as electrical, plumbing and/or HVAC systems. For demanding applications, such as exceptionally cold or humid climates and for architectural designs that prohibit building maintenance personnel from directly accessing the façade cavity, moisture control by mean of desiccated air is a common design strategy. Facilities to support this design approach must include additional mechanical pad space for air compressors and desiccant air which is routed through pressurized supply lines to each façade cavity. At each cavity wall unit, a small flow rate of regulated dry air bleeds into each cavity to counter leakage from the building interior.

2.2.4 Heat Gain

Cavity walls are distinguished from conventional wall systems not only by the additional façade layer but also by elements, such as blinds, that are commonly placed within the cavity. Taken in combination, the additional layers of semi-transparent glass and presence of elements which are opaque, and possibly reflective of, solar radiation vastly complicates prediction of cavity temperatures. Before any attempt is made to dissect this complexity, it serves to describe the interaction of solar radiation with a single piece of glass. Even the very best low-iron glass, which is essentially “visually” transparent, does not exhibit this property for frequencies of radiation which fall outside the visible spectrum. If we were blessed with a special form of vision that could observe all frequencies of thermal radiation, the glass would exhibit a tint corresponding to the portion of the infrared band that is either absorbed or reflected away and not transmitted. The fact that glass transmits high frequency thermal radiation but absorbs radiation at lower frequencies has significant consequences when the broader aspect of building design is considered. Radiation emitted from the sun is comprised of a broad spectrum of frequencies and a large portion of these frequencies pass through the glazing unabated. Frequencies within the visible spectrum are apparent as daylight; furthermore, certain frequencies spanning the near infrared and ultra-violet also transmitted within this band. When these rays of radiant energy are incident upon opaque surfaces such as walls, carpeting and furniture they are in part reflected, but also to some degree, absorbed. This energy gain results in the temperature of those materials increasing and consequent re-radiation of the absorbed energy at much lower frequencies, that unlike solar radiation, cannot transmit through the glass. This creates a net “solar gain” within the building which only achieves balance when it is either countered by HVAC equipment, removed by exterior ventilation or when the interior-exterior temperature difference is large enough to drive heat through the walls by conduction.

In a sealed cavity wall the interior space is not necessarily conditioned. Attentive to preceding discussion, since each layer of glazing absorbs a portion of the radiation incident upon it. Opaque elements, which will likely be deployed during sunny weather, are particularly absorbing resulting in a potential for internal heat gain that is unmitigated. Traditional thermal analysis of building envelopes focuses on the total rate of energy that passes through the façade, but does not grant very much attention to the temperature of each layer since the temperatures achieved generally do not pose a significant risk to the resilience of the materials deployed. The design of a cavity wall system; however, must account for these temperatures to ensure that internal components are resilient to withstand the resulting environment. A common example is motorized shades. Since acoustic performance is often a motivating factor for the implementation of cavity walls it is not uncommon to include laminated glass which is known to be sensitive to extreme temperatures.

Tables 2 (a) & (b) convey temperatures predicted by Window software. This software program was developed by Lawerence Berkley National Laboratory and performs 1-dimensional finite element heat transfer calculations considering both radiant and non-radiant forms of heat transfer. Furthermore, the treatment of radiant heat transfer appreciates the semi-transparent nature of glass at different angles of incidence. These calculations assume ASHRAE summer design conditions including 248.2 Btu/hr-ft² incident heat flux with an ASTM standard air mass 1.5 solar spectrum.

Table 2 (a) indicates temperatures predicted at normal incidence for various cavity wall make ups without shading elements. Table 2 (b) describes temperature variations with a venetian blind centered in the façade cavity with the slats of the shade angled at 45 degrees. It is apparent that values of the solar heat gain coefficient (SHGC), the percentage of solar heat flux that passes through the glazing, is lower when the shade is deployed. This outcome agrees with intuition. A perhaps less intuitive outcome is that while the SHGC values are lower for these cases, the internal cavity temperature is much higher. The tables also address the affect of including an insulated glazing unit as either the inner or outer leaf. While the lowest SHGC is achieved with the insulated glazing unit placed at the inner leaf, the temperatures associated with that case are among the highest. For this reason, it may be sensible to consider placing the insulated glazing in the outer façade layer, especially if it facilitates use of a high performance soft coat with in the IG cavity.

Some degree of discretion is warranted when considering whether or not a cavity wall design is appropriate for a given climate and its design must be tailored accordingly. As has been discussed in the foregoing, extreme cold will likely result in the potential for condensation. Alternatively, in regions of intense solar exposure, heat build up might be a greater concern. Both active and passive solar control strategies are possible as shown in tables 3(a) and 3 (b).

2.3 Structural Design

Cavity wall systems abide by the same fundamental concepts that govern the design of conventional curtain wall systems with a couple of significant exceptions that must be carefully considered.

First and foremost, load sharing which occurs between the lites of glass composing the inner and outer leafs may be significantly more complicated to analyze than is the practice for single insulated glazing units. This is due to the fact that the façade cavity might be quite large and may also be internally vented, externally vented or both to varying degrees. Structural review will need to consider the modes of operation which may occur during extreme wind events. For instance, while a sealed cavity may behave in a similar manner to an insulated glazing unit, significant venting to either the interior or exterior may require either the inner, outer or even possibly both leafs to withstand full structural wind load. In this regard, the operation of vents evokes significant consequences for the resilience of the assembly. Other considerations include the sensitivity of shading systems or any elements placed within the façade to the effects of wind. In certain instances careful analysis may be required for “corner units” if there is any communication across the façade cavity, or if venting for a given wall cavity is shared across a corner.

Second, it is implicit that cavity wall units will be significantly heavier than conventional wall systems imposing greater loads on structure and potentially affecting the design of anchoring systems. When planned for appropriately these requirements are not difficult to accommodate.

3.0 Logistics

It is implicit that manufacturing cavity wall units requires greater forethought than conventional wall systems on account of their increased weight and complexity which requires a longer assembly sequence, and in some cases, additional manipulation with ceiling cranes or other facets. Less obvious manufacturing implications are associated with the handling of materials, long term storage and bunking. These challenges are greatly compounded when units feature a “sealed cavity.”

3.1 Glass “Memory”

It is of paramount importance to carefully consider the potential for “memory” to be imparted on the inaccessible surfaces of the glazing if the façade cavity does not include access for periodic maintenance. Memory is the presence of visual disparities which result from condensation or minor dust accumulations over time whereby patterns, such as cup marks, are either temporarily or permanently apparent. It is important to accept that efforts to mitigate the formation of condensation and dust are not failsafe. Foreseeable circumstances such as outages in space conditioning equipment, errors in handling or maintenance and aging of system components over many years will result in their presence. They will not represent an observable defect when they form in a thin uniformly dispersed layer. Glass will potentially develop memory when it comes in contact with gaskets, material processing equipment such as vacuum cups, is handled by fabricators or is exposed to sealant spills and not properly cleaned. In particular, the use of silicone gaskets or any material which contains silica requires care due to its affinity for glass. Other materials may impart residue that cannot be observed by visual inspection. The glass fabricator should be consulted to determine appropriate glass cleaning procedures that can be incorporated prior to final assembly.

3.2 Storage, Shipping and Installation of Sealed Cavity Walls

Handling and installation of cavity wall unitized-assemblies involves the same inherent process that would be applied to a conventional curtain wall unit; however, the specific procedures at each stage of the process will differ to some degree. The façade cavity poses a vulnerability from a logistics standpoint, because it needs to remain clean during storage, shipping and installation. Furthermore, it may house special equipment such as motorized blind assemblies and other the materials.

Sealed-cavity systems which rely on a pressurized air supply to regulate humidity are particularly vulnerable, since they typically feature very limited access for cleaning. In some cases the unit would need to be de-glazed to address accidental fouling of the glass. Neither the storage facilities in which the units are stored, nor at the jobsite, where the units are exposed to weather, are humidity-controlled environments. Furthermore, it could be months or even years, before completed assemblies receive supply air from the pressurized system. In order to prevent condensation during this time period special storage procedures are required. Within the shop, the use of a temporary supply air system is possible. Alternatively, a calculated quantity of desiccant can be placed within each assembly to act as a temporary buffer. The latter option requires that the units be designed to house the desiccant, typically contained in bags, securely and out of sight and in a location where they can be conveniently accessed for replacement. The use of desiccant provides and added advantage of protecting the assemblies after they leave the shop and potentially even after the building is commissioned assuming that the desiccant is unsaturated at the time of unit “hook up.” Shipping and installing the units without desiccant is possible, but requires favorable weather conditions and for pressurized air systems to be ready for use at the time of installation. This sometimes impractical from a standpoint of trade-coordination. If desiccant is employed it may need to be replaced on site at some point during the building construction. Since environmental conditions are variable, a good quality assurance program will involve monitoring the humidity and temperature within select assemblies throughout storage and installation.

During shipping the orientation of bunked units may be dictated by features of the assembly such as roller blinds, desiccant or other features. Blinds, even when fully retracted, may need to be restrained to prevent damage. This can be achieved by shipping the units upright or potentially including guides for the retracted assembly. Alternatively, blinds may be secured temporarily by various methods and freed during installation. The orientation of opening to façade cavity must be carefully sealed against ingress from rain water, in particular any location where standing water could accumulate along gaskets, access panels or pressure equalization orifices requires special attention.

Once on site units should be staged in a dry area. It is important to recognize air and water barriers inherent to the design of the wall system may not guard against water incursion when the envelope is incomplete. In particular, the head of each assembly may need temporary protection to ensure that rain water, construction debris and other contaminants are not swept from the floor above onto the unfinished construction, especially if mullion cavities and unsealed-joints provide a pathway for incursion to occur.

Conclusion

Design, storage, installation and protection of cavity wall units has been described in the preceding sections. In review, it is important to carefully consider the use environment of a given design to ensure that the resulting construction will achieve acoustic goals, perform in an energy efficient manner and mitigate potential problems resulting from internal condensation, heat buildup and wind load if the design features vents. Traditional methods of calculation performed in accordance with standards and industry conventions may not be appropriate for use if their assumptions were based on simpler single wall systems. Assembly and storage of unitized assemblies requires great care to ensure that the internal surfaces of the cavity are not fowled or contaminated in a manner that will produce marks due unavoidable dust build up and condensation. Furthermore, stored units must be stored with desiccant or other means to control cavity moisture if the internal cavity cannot be easily accessed for cleaning. Shipment and site installation requires additional measures to secure internal components such as blinds and to protect installed units from water shed at each floor line before the building is fully closed in.

References

1. ASTM E 413 , Classification for Rating Sound Insulation, ASTM International, West Conshohocken, PA, 2022
2. ASTM 1332, Standard Classification for Rating Indoor-Outdoor Sound Attenuation, ASTM International, West Conshohocken, PA, 2022
3. ASTM E90, Standard Test Method for Laboratory Measurement of Airborne Sound Transmission Loss of Building Partitions and Elements, ASTM International, West Conshohocken, PA, 2022
4. [4] ANSI-NFRC 100, Procedure for Determining Fenestration Product U-factors, National Fenestration Ratings Council, Greenbelt, Maryland, United States, 2022
5. [5] Insul 10.0, Sound Analysis Software, Navicon Enginering, Fullerton, CA, 2023
6. [6] ISO15099, Thermal performance of windows, doors and shading devices, International Organization for Standardization, Geneva, Switzerland, 2003
7. [7] El Sherbiny, S.M., Raithby, G.D. and Hollands, K.G.T., Heat Transfer by Natural Convection across Vertical and Inclined Air Layers. Journal of Heat Transfer, 104, 96-102.,1982
8. [8] F. R. Menter, Two-Equation Eddy-Viscosity Turbulence Models for Engineering Applications, AIAA Journal, Vol. 32, No. 8, pp. 1598-1605, 1994