Numerical Investigation of Effects of Camlock System on Thermal Conductivity of Structural Insulated Panels

: Structural insulated panels (SIPs) are widely used in residential and commercial buildings due to their reliable mechanical and thermal performance. However, using framing members and nails to join SIPs causes thermal bridging across the insulation and thus increases heat losses from the building envelope. Alternatively, SIPs joined with embedded camlock systems can overcome this issue. In this paper, a parametric study of the effects of the camlock system material and SIP geometric design on the thermal performance of SIP walls was investigated using a multi-scale ﬁnite element modeling approach. The model considers the structural design details of the camlock system. In addition, the effects of the SIP materials, SIP thickness, and the number of camlock systems per unit area on the through-thickness thermal conductivity of the SIP walls are examined. It was found that the SIP thickness is a dominating factor inﬂuencing the thermal performance of the SIP. The through-thickness (overall) thermal conductivity of the SIP wall increases linearly with the increase in the number of camlock systems used per unit area. However, it rises exponentially with the increasing SIP thickness. The reduction in the overall R-value of the SIP caused by the camlock system embedded in the SIP did not exceed 13.8%.


Introduction
Residential energy consumption accounts for more than one-third of global electrical energy consumption [1][2][3]. Consequently, interest in green construction materials has grown, particularly for the exterior walls of buildings, which separate and protect structures from the climate. Reducing heat transfer via outside walls is critical without sacrificing structural integrity. These factors should be considered early on during the design process, and they have been the focus of much of the research on advanced construction technologies [4].
Structural insulated panels (SIPs) are energy-efficient building systems composed of a rigid, lightweight insulating material sandwiched between two outer structural skins. The insulation core is typically composed of insulating foams such as polyurethane (PUR), expanded polystyrene (EPS), or extruded polystyrene (XPS). The outer skin is usually composed of sheet metal, magnesium-oxide boards, cement-fiber boards, plywood, or oriented strand boards (OSBs). Owing to their high thermal insulation performance, SIPs are extensively used as building envelope materials in extreme climate regions, such as the Arctic [5][6][7]. According to Harries et al. [8], SIP walls have nearly a 42% higher R-value than typical timber-framed walls. SIP-built houses save about 14% energy use compared to studframed houses [9] and 60% compared to brick-concrete structure houses [10]. According to Lu and Memari [11], the thermal performance of SIP walls depends mainly on the insulation thickness and panel size. They estimated the annual energy savings from implementing 3.5 in and 5.5 in SIP walls compared to standard framing walls to be 3.08% and 7.62%, respectively. Another advantage of SIPs is the panels' easy and quick installation process, which, in turn, reduces installation costs [12]. According to Mullens and Arif, installing SIPs on walls and roofs saves up to two-thirds of the on-site framing labor [13].
However, the thermal insulation performance of any SIP envelope is vulnerable to the panel's joints, where air and moisture infiltration and thermal bridging may occur [14]. Abdou [15] examined the influence of a joint sealant on the thermal performance of an interlocking SIP envelope composed of fiber-reinforced plastic. The absence of a joint sealant reduced the SIP wall's overall thermal resistance (R) due to the air leakage. This reduction ranged from 5% to 46%, depending on different factors. Wyss et al. [16] and Kayello et al. [17] published similar findings. They conducted experimental studies to determine the airtightness and thermal performance of SIP joints subjected to pressure differences that may cause joint openings. They used oriented strand boards and expanded polystyrene for the SIPs and wood for the joints.
Qiang Du et al. [18] conducted an experimental study on the thermal transmittance of different types of spline joints commonly used in SIP walls. They also performed a FE numerical parametric study to examine the influence of insulation thickness, nail dimensions, and spacing on the thermal performance of SIP walls. They reported that the thermal transmittance increases significantly with the increase in the size of the spline.
A SIP joint has less thermal resistance than the SIP center owing to thermal bridging at the framing members. Therefore, the overall R-value of the SIP wall is reduced; this reduction is estimated to be 10-20% depending on the size and geometry of the joining mechanism [19]. As a solution, the camlock system enhances the thermal performance of SIP walls by providing outstanding tight connections to the SIPs. A cam locks system is a fastening mechanism used to connect two SIPs. It comprises two halves, one containing a pin and the other having a cam, which rotates by using a key to connect or disconnect the SIPs. The use of the camlock system eliminates the need for frame members and nails, which are responsible for thermal bridging. However, there exists a significant gap in the literature on the thermal performance of SIPs with embedded camlock systems.
Therefore, the objective of this paper is to investigate the thermal performance of the SIP walls joint by a camlock system which has never been considered in the previous literature. Furthermore, the thermal performance of the alternative joining method is considered, namely the surface spline; thus, the thermal performance-through-thickness thermal conductivity-of both options can be compared and discussed. In order to achieve the objective of this study, a multi-scale finite element (FE) model was developed and validated to predict the through-thickness thermal conductivity of SIP walls. The model results were firstly verified by experimental results obtained by other researchers [18] where they used a hot box apparatus. Then, utilizing the powerful commotional capabilities of the model, the effects of different measures on the thermal insulation performance of the SIP walls joined by camlock systems were investigated. Specifically, the thermal conductivity of the camlock system, the thermal conductivity of the SIPs, the thickness of the SIPs, and the number of camlock systems per unit area were all examined. The results provide comprehensive data on these various design options that can guide future research and development.

Test Setup
The experimental work of Qiang Du and his colleagues [18]-who used a hot box apparatus according to ASTM C1363-11 [20] in order to measure the thermal transmittance of SIP wall samples of size 1.5 × 1.5 × 0.165 m 3 -was used in this study for validation. This apparatus consists of two boxes, one chilled to sub-zero temperature (called the cold box) and one kept near room temperature (called the hot box). The wall sample is positioned in an insulated test frame which is placed between the hot and cold boxes. Two heat flow meters are glued to the SIP wall sample from the side of the hot box. The thermal resistance (R-value) of the SIP wall sample is then determined as follows: where R is the thermal resistance of the SIP wall sample (m 2 K/W), Q is the quantity of heat received by the SIP wall surface in the hot box (W), A is the surface area of the SIP wall (m 2 ), and T hot and T cold are the surface temperatures facing the hot and cold boxes, respectively. The heat received by the SIP wall surface in the hot box was measured in the thermal bridge areas of the SIP wall using two heat flow meters. The hot and cold surface temperatures were measured using four thermocouples, two for each side.
The thermal transmittance U, in (W/m 2 -K), can be calculated from the R-value as follows: Thermal conductivity is then determined as follows: where k is the thermal conductivity of the SIP wall sample (W/m-K) and L is the thickness of the SIP wall sample in meters.

SIP Wall Samples
The SIP used in the experimental work of Qiang Du and his colleagues [18] was made of 143 mm of an EPS sandwich between two layers of OSB with a thickness of 11 mm. The total thickness of the SIP wall samples was 165 mm. They first examined a single SIP without any joints. The second sample was a SIP wall with a joint in the middle of the SIP wall. The SIPs were connected by a surface spline joint. The dimensions of the nails used in the surface spline joints were 3.2 mm in diameter and 60 mm in length, with a constant spacing of 150 mm as outlined in the international building code [21]. Figure 1 illustrates a schematic of the SIP wall connected by surface spline and embedded camlock systems and shows the unit cells of each wall. The unit cell can be defined as a regular and repetitive pattern that forms the structure of the wall by repeating itself along the X and Y spatial planes. Figure 2 shows the top view schematics of the unit cell of the SIP wall connected by surface spline and embedded camlock systems. Figure 3 shows a transparent model of the camlock system revealing the interior structure.

Geometrical Modeling
Creating a single SIP wall model containing the minute details of the camlock system will result in a complex model with a huge number of nodes and consequently a very long computational time. Therefore, to enhance the accuracy of the model and to reduce the computational time, a two-scale finite element homogenization methodology ( Figure 4) was employed in this study. This technique was used to predict the equivalent thermal conductivity of a SIP wall connected by embedded camlock systems. At the first scale, the thermal conductivities of a camlock system were determined along the X-, Y-, and Z-directions using a camlock model. In the second scale, the SIP wall was modeled using the SIP wall unit cell. The camlock system in the SIP wall unit cell was replaced with a homogenous solid cuboid. The conductivities obtained from the first scale (the camlock model) were assigned to this solid homogenous cuboid. The through-thickness thermal conductivity of the SIP wall and the effect of various factors that may reduce the thermal insulation of the wall were predicted. All of the finite element modeling was performed using the 6.14 ABAQUS software package [22].  turer's sheet [23]. The properties of air are assigned to th turer's sheet [23]. The properties of air are assigned to th

Camlock Model
The camlock model was a 90 × 94 × 16 mm hollow cuboid with 1 mm-thick walls. It featured two cams on the left half and one pin on the right half, as shown in Figure 3. A hexagonal wrench hole was used to rotate the cams to lock them with the pin. Additional details regarding the dimensions of the camlock system can be found on the manufacturer's sheet [23]. The properties of air are assigned to the interior cavities of the camlock finite element model. The finite element mesh of the camlock model comprised 324,918 nodes and 1,835,261 tetrahedral elements of type DC3D4.

SIP Unit Cell
A SIP wall unit cell was built based on the periodic features of the camlock system in the SIP wall. The SIP unit cell consisted of a small cuboid embedded inside a larger cuboid, as shown in Figure 3. The larger cuboid represents the SIP, whereas the smaller cuboid represents the camlock system. The orthotropic thermal conductivities derived from the camlock model were assigned to the small cuboid that represented the camlock system. Multiple SIP unit cells were constructed with various dimensions to investigate different parameters: the wall thickness, number of camlock systems per square meter, thermal conductivity of the camlock system, and thermal conductivity of the SIP. The examined thermal conductivities of the SIP were 0.05, 0.1, 0.15, 0.2, 0.25, and 0.3 W/m-K.

Boundary Conditions
The hot box experimental setup was used to measure the thermal conductivity (k) of the specimen, with Fourier's law: where Q is the total heat flux from the higher to the lower temperature. A and L are the cross-sectional area and thickness of the sample, respectively. T Hot and T Cold represent the surface temperatures on the hot and cold boxes, respectively. This setup can be simulated using finite element software through steady-state thermal analysis. The boundary conditions of this steady-state thermal analysis involve setting two different temperatures (T Hot and T Cold ) on two opposite faces perpendicular to any spatial direction, while the remaining faces are insulated. These boundary conditions result in a constant unidirectional heat flow (Q) toward the surface at a lower temperature (T Cold ). The value of Q can be extracted from the model and inserted in Equation (4) to determine the thermal conductivity. Table 1 lists the input thermal properties of EPS, OSB, air, and galvanized steel for the nails and camlock systems. Six different materials were examined as material options for the camlock systems. These materials are polypropylene, stainless steel, steel, galvanized steel, brass, and aluminum. The thermal conductivities of these materials are also listed in Table 1.

Validation of the FE Model
The steady-state FE thermal analyses of the single SIP without any joint along the Z-direction yielded a through-thickness thermal conductivity of 0.0419 W/m-K, with a marginal error of 2.3% lower than that experimentally obtained by a hot box apparatus, i.e., 0.0429 W/m-K [18]. The experimental and simulated through-thickness thermal conductivities of the affected area unit of the surface spline were 0.048 W/m-K [18] and 0.0473 W/m-K, respectively. Note that the experimental measurements were conducted twice for each type, and the comparison here was carried out with the mean values. The negligible difference between the experimental and simulated results may be due to heat transfer effects by convection inside the experimental hot and cold boxes, which were not considered in the FE model. However, the FE model results are in excellent agreement with the experimental results; accordingly, the used FE model is deemed satisfactorily validated.

Camlock Model
The orthotropic thermal conductivity values of the camlock model were predicted by repeating the steady-state thermal analysis along all three directions (X, Y, and Z). Similar thermal conductivities are expected in the X-and Y-directions as the structure of the camlock model is similar along both these directions. Table 2 summarizes all the resulting thermal conductivities derived from the camlock model for the six different materials with varying thermal conductivities. Polypropylene has the lowest thermal conductivity, three orders of magnitude lower than aluminum, with the highest thermal conductivity. Figures 5-7 present the heat flux contour maps of the camlock model with polypropylene and aluminum in the X-, Y-, and Z-directions, respectively. Similar to thermal conductivity, the heat flux flowing in the polypropylene camlock model is three orders of magnitude lower than aluminum.

SIP Unit Cell
A steady-state thermal analysis was carried out on the SIP unit cell in the Z-direction to predict the through-thickness thermal conductivity of the SIP walls, one containing a surface spline joint and the other one containing camlock systems. In both models, galvanized steel properties were assigned to the nails and camlock systems. The resulting thermal conductivities were 0.04201 and 0.0427 W/m-K for the SIP walls containing a camlock system and surface spline joint, respectively.
In the parametric study, four main parameters were examined. Two of these parameters are related to the camlock system: the camlock materials and density (quantity per unit area). The other two parameters, that is, the thickness and thermal conductivity, are related to the SIP. More than 300 simulations were conducted. The corresponding results are presented in Tables 3-5.

Discussion
The use of the camlock system in joining SIPs reduces the thermal bridges compared to the spline surface. The camlock system has the advantage of not being exposed to the environment.
The camlock thermal conductivity of each material was derived from the camlock model, as listed in Table 2. The through-thickness (Z-direction) thermal conductivity of the polypropylene camlock system was 0.045 W/m-K, which is lower than the previously recorded values of SIP thermal conductivity. Therefore, the through-thickness thermal conductivities of the SIP wall were reduced in all the SIP unit cell models containing a polypropylene camlock system. These results vary between −0.01 to −0.67%. All of the camlock materials, except for polypropylene, increased the through-thickness thermal conductivity of the SIP wall. This increase depended primarily on the thickness of the SIP; it varied by approximately 2-4% in the SIP wall with a thickness of 20 mm contained on the camlock system, as shown in Figure 6. However, the effect of the camlock system on the wall thermal conductivity was reduced with an increase in the thickness of the SIP wall. For instance, the increase in the SIP wall with a thickness of 100 mm did not exceed 0.24%. Similar to the camlock material, the effect of the SIP thermal conductivity was evident for the SIP thickness of 20 mm. However, it was minimal in the thicker SIPs. Unexpectedly, the lower the SIP thermal conductivity, the stronger the effect of the camlock system. For example, the thermal conductivity of the SIP wall increased by 4 and 3.69 percent when an aluminum camlock system was installed in a 20 mm-thick SIP with thermal conductivities of 0.05 and 0.3 W/m-K, respectively.
Furthermore, the effect of the density of camlock systems (the number of camlock systems per square meter) was investigated with an aluminum camlock and variable SIP thicknesses and thermal conductivities. The thermal conductivity of the SIP wall increases with the camlock density. This increase was linear with respect to the density of the camlock system. For instance, a constant increase of 0.24% per camlock was noted under the SIP thickness of 100 mm and thermal conductivity of 0.05 W/m-K. Similar to the results in Table 3, as the SIP thermal conductivity decreased, the effect of the density of the camlock system on the overall SIP wall thermal conductivity increased. Moreover, the effect of the camlock system density on the overall SIP wall thermal conductivity decreased exponentially as the SIP thickness increased. Therefore, it is essential to reduce the number of camlock systems in the SIPs wall, but without jeopardizing the airtightness performance of the wall, especially with reduced thickness SIPs.
The highest reduction in the overall R-value of the SIPs with camlock connections was 3.85% when the thickness of the SIP was 20 mm and the camlock material was aluminum. This reduction goes up to 13.8% in the same condition with four camlock systems per square meter. McIntosh and Guthrie [19] reported that the thermal bridging at the framing members of SIPs joint causes a 10-20% reduction in the overall R-value of the SIP wall. Therefore, from the thermal point of view, it is better to use camlock systems to connect the SIPs.

Conclusions
In this study, steady-state thermal analyses were performed on two-scale finite element models to determine the effects of camlock systems on the thermal conductivity of SIP walls. The main findings of this study can be summarized as follows:

•
The use of the camlock system in joining SIPs reduces the thermal bridges compared to the spline surface.

•
The lower the SIP thermal conductivity, the stronger the effect of the camlock system. • The thicker the SIP, the lower the effect of the camlock system. • Variations in the SIP wall through-thickness thermal conductivity were linear with respect to the camlock density and the thermal conductivity of both the SIPs and the camlock system. • Changes in the through-thickness thermal conductivity of the SIP wall were exponential with respect to the SIP thickness.

•
Optimizing the embedded camlock systems density (the number of embedded camlock systems per unit area) is critical to reduce the thermal conductivity of the SIP wall without jeopardizing the airtightness performance of the wall, especially with reduced thickness SIPs.

•
The reduction of the overall R-value of the SIP caused by the camlock system embedded in the SIP did not exceed 13.8%.

•
Extensive experimental work is needed to evaluate the thermal and mechanical behavior of SIPs with embedded camlock systems.