Evaluation of the thermal and structural performance of potential energy efficient wall systems for mid-rise wood-frame buildings

Energy and Buildings 82 (2014) 416–427

Contents lists available at ScienceDirect

Energy and Buildings journal homepage: www.elsevier.com/locate/enbuild

Evaluation of the thermal and structural performance of potential energy efficient wall systems for mid-rise wood-frame buildings Hadia Awad a , Mustafa Gül a,∗ , Hamid Zaman a , Haitao Yu b , Mohamed Al-Hussein a a b

Department of Civil & Environmental Engineering, University of Alberta, 9105 116th St., Edmonton, AB, Canada T6G 2W2 Landmark Group of Builders, 9765 54 Ave., Edmonton, AB, Canada T6E 5J4

a r t i c l e

i n f o

Article history: Received 3 April 2014 Received in revised form 18 June 2014 Accepted 15 July 2014 Available online 25 July 2014 Keywords: Energy-efficiency Thermal performance Building envelope Load-bearing capacity Light wood-frame Mid-rise construction

a b s t r a c t Approximately 30% of energy use in Canada is consumed in buildings, specifically space heating, which can be reduced by constructing thermally-resistant building envelopes. This study evaluates potential innovative energy-efficient wall systems for mid-rise (four to six storeys) wood-frame buildings in terms of thermal and structural performances. Regarding the thermal resistance performance, four wall systems are developed, installed in a full-scale test house, and examined, along with a baseline wall system, using the field data collected. The selection of the wall systems is based on current practice, structural analysis, pre-fabricability, and energy-efficiency. Several sensors are installed on each wall system to measure temperature, heat flow, and relative humidity. In addition, structural tests are conducted to determine the compressive loading capacity of the tested wall systems for both concentric and eccentric loads, where full-scale panels are constructed and tested in laboratory. As a general finding, all the tested wall systems achieved ASHRAE’s minimum assembly R-value recommendation of RSI 3.45, where the I-joist wall system had the highest R-value, while the conventional wall system had the highest load-bearing capacity. This paper recommends I-Joist wall systems for their higher energy efficiency suggesting more future research on efficient end connections to achieve consistent structural performance. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Buildings account for 30–40% of the total primary energy use and 24% of the generation of greenhouse gases globally [1,2]. If current trends continue, by 2025, buildings worldwide will be the largest consumers of global energy, using as much power as the transportation and industrial sectors combined. Recent studies have found that improving energy-efficiency in buildings is the least costly way to reduce a large quantity of carbon emissions [2]. Meanwhile, the building sector offers significant potential to reduce primary energy use and CO2 emissions [3] through such measures as reduced heating demand, increased efficiency of the energy supply chain, and greater use of renewable resources for materials and fuels [4]. Reductions in the specific energy demand of buildings and increased use of renewable energy are important measures of climate change mitigation [5]. For this reason, energy-efficiency

∗ Corresponding author. Tel.: +1 7804923002; fax: +1 7804920249. E-mail addresses: [email protected] (H. Awad), [email protected] (M. Gül), [email protected] (H. Zaman), [email protected] (H. Yu), [email protected] (M. Al-Hussein). http://dx.doi.org/10.1016/j.enbuild.2014.07.032 0378-7788/© 2014 Elsevier B.V. All rights reserved.

in buildings is now a primary objective of energy policy at the regional, national, and international levels [6]. Several strategies can be used to realize this potential, including the increased incorporation of energy-efficiency requirements into building standards, such as requirements that specify minimum energy-efficiency for buildings. Wood-framing is the preferred approach for the building envelopes of low-rise residential facilities and commercial buildings in North America, being that wood-frame building envelopes are lightweight, easily built, durable, renewable, and lower in embodied energy than most alternative building materials [7,8]. The Canadian Wood Council [9] has stated that this is due to the proven performance of properly designed and built wood-frame buildings, which have historically provided strong and lasting housing. Building on the success of low-rise wood-frame buildings, in British Columbia, wood-frame construction solutions for mid-rise buildings have been developed and refined in recent years, leading to more sustainable communities and affordable housing solutions that can positively change the face of North American cities and provide more multi-unit buildings for a fast-growing population. Internationally, the conditions for market growth of multi-storey construction seem to be most favorable in Sweden, the UK, and Germany [10].

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

Using light wood-frame structures for constructing mid-rise buildings means that there is higher compressive load bearing on the wall systems. This needs to be addressed by increasing the strength of the walls by reducing stud spacing. The use of more lumber per panel to increase structural performance, however, is accompanied by a significant decrease in thermal resistance performance, since wood has a relatively high thermal conductivity compared to insulation materials, resulting in a thermal bridge between the indoor and outdoor environments. A study by Berggren and Wall [5] has shown the increasingly prominent role of thermal bridges in transmission heat transfer calculations to improve a building’s energy performance. However, our research aims to optimize the structural and thermal performance of the selected wall systems taking into account anticipated future building code requirements for both thermal efficiency and structural capacity. ASHRAE-90.1 [11] has specified energy performance requirements for buildings for different climate zones. Climate zones are determined on the basis of Heating Degree Days (HDD). An HDD is a measure of how cold a location is over a period of time relative to a base temperature of 18 ◦ C [11]. Edmonton, Alberta, where the test house is installed, has a recorded range of 5000 to 6000 HDD. In climate zones of 5000 to 7000 HDD, the range within which Edmonton falls, ASHRAE-90.1 [11] recommends a minimum assembly R-value (RSI) for wood-frame buildings of R-19.6 (3.45).

417

humidifier in order to maintain an average indoor temperature of 22 ◦ C and a relative humidity of 50%. The selected wall systems are: two engineered-wood I-Joist (TJI 230) wall stud systems at spacing of 304.8 mm and 406.4 mm (12 in. and 16 in.), respectively; two 38 mm × 140 mm (2 in. × 6 in.) staggered wall stud systems at spacing of 304.8 mm and 203.2 mm (12 in. and 8 in.), respectively; and a baseline wall system which consists of 38 mm × 184 mm (2 in. × 8 in.) wall studs at a spacing of 304.8 mm (12 in.). The period of data collection is April, 2012 to April, 2014. The structural performance test focuses on one type of each wall system, (i.e., an I-Joist, a staggered, and a conventional wall system). The objective of the structural performance test is to determine the compressive load capacity of the tested wall systems for both concentric and eccentric types of loads. Full-scale panels with dimensions of 1.2 m × 2.4 m (4 ft × 8 ft) are designed for testing that conform to the specifications of actual wall construction. In order to prevent axial buckling at the wall studs, mid-blocks are fixed between the wall studs, thus increasing the wall panels’ loading capacities. As recommended by the American Society for Testing and Materials (ASTM) [12], each type of test is confirmed by testing three like specimens. The primary objective of the study is to select a wall system that strikes an optimal balance between energyefficiency (high R-value) and load-bearing capacity for mid-rise construction. 2. Thermal performance of wall systems

1.1. Research objectives and scope In the extreme weather conditions of Canada, and North America in general, there is a demand to create thermally-resistant building envelopes. This study aims to develop energy-efficient wall systems for mid-rise construction that can, thermally, mitigate heat loss, and, structurally, increase load-bearing capacity compared to conventional wall systems. Regarding thermal resistance performance, which is quantified in terms of R-value, four different wall systems are developed, manufactured, and installed in a full-scale test house; exposed to the natural exterior climate; and examined for their long-term thermal resistance performance, along with a baseline wall system. The selection of the wall systems is based on specific considerations: current practice; preliminary structural analysis; pre-fabricability; material availability; expected energy efficiency; and population growth, which necessitates more multi-unit buildings. The general objective of this research project is to identify promising exterior wall systems that can be expected to meet future building code requirements for thermal performance while providing the required structural capability for mid-rise woodframe construction. The specific goals of the study are as follows: (a) Evaluate the long-term thermal performance of the selected exterior wall systems exposed to natural outdoor climatic conditions and to controlled indoor conditions; (b) Evaluate the structural performance of the selected wall systems with laboratory tests; and (c) Characterize the temperature, air, and moisture (hygrothermal) response of these wall systems in extreme weather conditions and to indoor conditions of humidity and pressure. 1.2. Research methodology The testing methodology is conducted by installing several sensors on each wall system, including heat flux sensors, thermocouple sensors, and humidity sensors. The extreme weather conditions in Alberta are challenging, even though the indoor environmental quality is maintained by installing aheater, thermistor, and

This research is based on field measurements, which, as advanced by Straube et al. [13], are useful primarily because they expose building components to the whole range of conditions experienced by enclosures in service. This means that the boundary conditions are, by definition, realistic in magnitude, rate, sequence, and probability of occurrence. Simulation software is used as a complementary component of this research to estimate the thermal performance of the selected wall systems. Consequently, the results from in-situ data collected are validated by comparing them to simulation results. 2.1. Wall systems selection The selection of potential wall systems for field and lab testing in this project is based on a number of considerations. Traditional stick-frame construction practice uses 38 mm × 89 mm (2 in. × 4 in.) or 38 mm × 140 mm (2 in. × 6 in.) lumber studs placed at a spacing governed by the type of wall (load-bearing or nonload-bearing) and the amount of load acting upon it [14]. However, load-bearing walls in mid-rise buildings, being subjected to greater vertical and horizontal loading, are likely to have closely spaced studs, ultimately making them less energy-efficient. Our study thus explores the potential of using wall systems comprising engineered-wood I-joists and staggered studs, rather than conventional stud configurations. Engineered-wood I-joists have higher flexural strength and are usually used as horizontal members in floors, thus making them a good candidate to perform well against horizontal loads (such as wind). Moreover, they are expected to have better thermal resistance than traditional lumber studs because of their relatively thinner web. The consideration behind selecting staggered studs is to break the thermal bridging created at the stud locations. It is understood that such configurations may not be suitable for load-bearing walls because the studs that are not connected to sheathing can buckle against their weaker axis. In addition, in order to compare these engineered I-joist and staggered stud options with conventional practice, a 38 mm × 184 mm (2 in. × 8 in.) conventional stud configuration is selected. Use of 38 mm × 184 mm (2 in. × 8 in.) lumber instead of 38 mm × 89 mm

418

Table 1 Description of the wall systems selected for thermal performance testing.

Plan Insulation (inside-out)

Top plates

Wall System 2 (WS2)

Wall System 3 (WS3)

Wall System 4 (WS4)

Wall System 5 (WS5)

TJI 230 304.8 mm

TJI 230 406.4 mm

Staggered 38 × 140 mm (2” × 6”) 304.8 mm

Staggered 38 × 140 mm (2” × 6”) 203.2 mm

Conventional 38 × 184 mm (2” × 8”) 304.8 mm

12.7 mm Gypsum Board 6 mm polyurethane 50.8 mm air gap 139.7 mm fiberglass 50.8 mm SPF 9.525 mm OSB 38 × 234.95 mm

12.7 mm gypsum board 6 mm polyurethane 50.8 mm air gap 139.7 mm fiberglass 50.8 mm SPF 9.525 mm OSB 38 × 234.95 mm

12.7 mm gypsum board 6 mm polyurethane – 139.7 mm fiberglass 50.8 mm SPF 9.525 mm OSB 38 × 184 mm

12.7 mm gypsum board 6 mm polyurethane – 139.7 mm fiberglass 50.8 mm SPF 9.525 mm OSB 38 × 184 mm

12.7 mm gypsum board 6 mm polyurethane – 139.7 mm Fiberglass 50.8 mm SPF 9.525 mm OSB 38 × 184 mm

Note: TJI 230 is a product of Weyerhaeuser NR Company. Source: www.woodbywy.com.

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

Stud type Spacing

Wall System 1 (WS1)

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

(2 in. × 4 in.) or 38 mm × 140 mm (2 in. × 6 in.) ensures a wall system with the higher structural strength required for mid-rise construction [15]. Other important building elements are the oriented strand board (OSB) sheathing and drywall siding, a water resistive barrier in which the OSB sheathing plays an important role as a structural element, while the siding improves the hygrothermal performance of the wall system [16]. Table 1 provides details of the selected wall systems. 2.2. Field testing A 3 m × 6 m (approximately 10 ft × 20 ft) test house (Fig. 1) is assembled using the selected wall systems, which are placed sideby-side along one side of the house. The side of the test house comprising the selected wall systems is oriented toward the north direction for the current set-up in order to avoid direct solar exposure. The test house is located at the University of Alberta south field at 127 Street and 20 Avenue SW (Heritage Valley Area, near Ellerslie Road in southwest Edmonton, latitude 53.41◦ N, longitude 113.54◦ W). A total of 41 sensors are installed inside the test house (24 thermocouple sensors, 10 heat flux sensors, and 7 relative humidity sensors). Fig. 2 shows the wall system set-up with respect to the north direction, the positions of the different sensors, as well as the HVAC system used inside the test house. In order to maintain even distribution of temperature and relative humidity, a fan is installed in the middle of the test house [17].

419

for mid-rise construction. We observe that the extent of heat flux is influenced by the change in outdoor temperatures. For example, in February and March, during which the outdoor temperature varies between −20 ◦ C and 5 ◦ C, the corresponding heat flux values are high, while in May and June, during which the outdoor temperature varies between 5 ◦ C and 25 ◦ C, the corresponding heat flux values are very low. Even the differences between different wall positions are minute compared to at colder times of year. The severe weather conditions in Edmonton, Canada, with outdoor temperatures ranging between −30 ◦ C and 30 ◦ C, make maintaining the indoor temperature a challenge. Fig. 5 shows outdoor wall surface temperatures at stud and cavity of the selected wall combinations, in addition to outdoor air temperature. It should be noticed that, in winter there are slight differences between the readings, even though all the tested walls are exposed to the same outdoor weather conditions, are placed adjacent to one another, and are all north-oriented. For summer days, however, the outdoor temperature is almost equal to the indoor temperature, even though the indoor heating system is still in operation to maintain constant indoor conditions throughout the entire data collection process. Relative Humidity was also measured within the cavities of the selected wall systems in order to demonstrate the hygro-thermal performance of those wall systems in future studies. Fig. 6 shows that the relative humidity within the wall systems, as well as the indoor environment varies between 40 and 80% in winter, while in summer it varies between 40 and 60%.

2.3. Data analysis The ASTM [18] recommends maintaining a constant temperature on the interior side of the wall, and a significant difference between indoor and outdoor temperatures for rapid convergence. Accordingly, a heater, humidifier, and thermistor are installed inside the test house. Fig. 3 shows a sample of hourly average indoor surface temperatures for the selected wall systems. We observe that the indoor temperatures vary between 21 ◦ C and 23 ◦ C. Two sensors are installed inside the test house to measure the east and west ambient temperatures, in addition to a reference temperature sensor inside the data logger (CR 1000). As assumed, the indoor ambient temperature is slightly higher than the surface temperatures of the selected wall systems (approximately 24 ◦ C), which are exposed to a greater extent to energy loss due to the significant difference between indoor and outdoor temperatures. However, the thermal resistance performance of combined wall systems depends upon not only indoor surface temperatures, but the difference in temperature (T, ◦ C), through the entire wall combination, as well as the amount of energy lost through the wall (heat flux, q, W/m2 ), assuming that the heat flows in one direction perpendicular to the wall [19]. The thermal resistance of a wall (R-value) is directly proportional to the difference in temperature between indoor and outdoor wall surfaces (T), and inversely proportional to the amount of energy lost, also called heat flux (q). The increased heat flux of a wall reduces its thermal resistance, leading to higher energy consumption to maintain the indoor air temperature. Fig. 4 shows the heat flux at stud and cavity of the selected wall combinations. The negative value of the measurements indicates the direction of the heat flow from inside to outside. This graph shows that the heat flux through the stud of a wall system is noticeably higher than that of the corresponding cavity. This verifies experimentally the thermal bridging phenomenon resulting from the relatively high thermal conductivity of wood as a main structural element compared to wall insulation materials (SPF and fiberglass). It is hypothesized that decreasing the amount of studs (i.e., increasing spacing) of a wall system leads to higher thermal resistance of the entire wall, taking into consideration its corresponding structural performance

2.4. Calculation of thermal resistance from field data To determine the wall systems’ thermal resistance, this study follows the summation technique described in ASTM Standard C1155-95 [18]. The first step determines the temperature difference between the indoor and outdoor surfaces of the wall systems using the following equation: Ts = Tis − Tos

(1)

where Ts is the temperature difference (◦ C); Tis is the indoor surface temperature (◦ C); Tos is the outdoor surface temperature (◦ C). By comparing Ts of different positions of the selected wall systems, a very slight difference can be observed among the wall systems. The variations in heat flux readings (Fig. 4) are much greater than those for Ts . As mentioned above, the R-value of a wall system is affected by both Ts and q; moreover, the variation of the thermal resistance performance is mainly influenced by variations in heat flux. The data logger installed in the test house collects the data on one-minute intervals. The next step involves averaging the temperature, heat flux, and relative humidity data for every hour—(ASTM recommends this time interval to be less than or equal to 60 minutes), and calculating the estimated thermal resistance using the following equation:

M Re =

K=1  M

 TSK

(2)

q K=1 K

where Re is the estimated thermal resistance (RSI) in K m2 /W; k is the counter for summation of time series data; M is the duration of test for each convergence (hour); q is the heat flux (W/m2 ). As observed in Eq. (2) every hour, an estimated thermal resistance value can be calculated based on the cumulative temperature difference and heat flux values for each wall system. Thus, the

420

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

Fig. 1. Full-scale test house for evaluation of thermal resistance performance. The photograph on the left was taken on January 15, 2013, and the photograph on the right was taken on August 6, 2013.

RSI values reach stability over time; the convergence factors are calculated satisfying the following equation: CRn =

Re (t) − Re (t − n) Re (t)

(3)

where CRn is the convergence factor for the chosen convergence interval (hour) where n = convergence interval (hour) and t = time (hour). In this study, 12 h is chosen as the convergence interval. As the thermal resistance values approach a stable condition, the convergence factors are expected to stabilize at a lower value. ASTM recommends continuing this process until the convergence factor remains within a range of ±0.1 (±10%) for three consecutive periods of convergence intervals. It should be mentioned that in this study we continue this until all sensor locations (stud and cavity) of the selected wall systems reach this convergence limit. Fig. 7 represents the convergence of RSI values over time. Once the convergence is reached, the corresponding RSI values are recorded, and the same data analysis procedure is repeated (for subsequent trials) using the next set of data starting from the time when the convergence criteria was met for the previous dataset. Thus, three sets of RSI are obtained from the collected data. Finally, the mean, standard deviation, and coefficient of variation (COV %) of the obtained RSI values are calculated. As can be seen in Table 2, the COV % of the RSI values of the selected wall systems meet the ASTM goodness of fit criteria (10%) for all of the wall systems. After calculating the RSI values at stud and cavity of the wall systems, the overall panel thermal resistance values are determined. These values are not representative of the entire wall systems since the selected wall systems have varying stud dimensions and spacing. The overall panel thermal resistance (RSI) values are determined using Eq. (4). Fig. 8 plots the calculated assembly RSI values of the selected wall systems. Rwall =

100% %Areastud /Rstud + %Areacavity /Rcavity

(4)

where Rwall is the thermal resistance of the wall system (K m2 /W); %Areastud is the percentage of area of studs in the wall system; %Areacavity is the percentage of area of cavity in the wall system; Rstud is the thermal resistance at stud; and Rcavity is the thermal resistance at cavity. 2.5. Summary of the results Table 3 summarizes the field data results in the form of RSI values at stud and cavity of each wall system separately, and then the assembly R-values of those wall systems. We observe that the RSI values of the selected wall systems are quite close to each other for different months of the year. However, we also acknowledge that there is a slight decrease in RSI values as the weather becomes

warmer (i.e., the RSI in February is higher than in April). The RSI value of the first wall system is assumed to be higher than that of the conventional wall system, since it has a lower thermal bridging effect. However, more extensive research should be conducted to support better understanding of these test results. 3. Structural performance of wall systems The objective of the particular component of the study presented in this section is to determine the load-bearing capacity of both concentrically- and eccentrically-loaded external wall systems. Of the five types of tested walls, three wall systems are selected for the structural testing. The selection of wall systems involves choosing the larger spacing for each wall type (e.g., I-Joist at spacing of 406.4 mm (16 in.) rather than 304.8 mm (12 in.) for both the I-joist and the staggered wall system). It is hypothesized that closer stud spacing provides higher load-bearing capacity. The remaining wall systems may be tested in subsequent phases if needed. The specimens are designed to conform, structurally, to the actual wall systems at the test house. This means that insulation materials not affecting the structural performance of the panels are ignored in the structural testing process. Mid-blocks are fixed along the centerline of the vertical, weak axis, to support vertical studs against buckling, as well as to resemble common practice for mid-rise construction. Since ASTM Standard E72-10 [12,20] recommends that the width of the specimen be 1.2 m (4 ft), 1.2 m × 2.4 m (4 ft × 8 ft) wall specimens are selected for the lab test. According to the ASTM [12], the wall should be tested for different loading types such as (a) compressive, (b) tensile, (c) transverse, (d) concentrated, and (e) racking. At this point in our research, only the compressive performance of different wall systems is addressed. The deformation of the specimen is determined by installing six deflectometers attached to the wall specimen; east and west vertical, east and west in plane, and east and west out-of-plane. Fig. 9(a) shows the position of the deflectometers attached to the wall specimen. The compressive loading is expressed in terms of force, as recommended by Oral et al. [21], while the specimen deformations corresponding to the loading are expressed as displacement (mm). The compression and tension reactions are expressed in negative and positive signals, respectively. Specimen failure, it should be noted, occurs due to either physical damage or the inability to bear load increments (plateau). The observed specimen damages are summarized below: (a) Excessive buckling, resulting in wood stud failure; (b) Excessive bending of the loading head, resulting in the bending of the wall specimen (especially in the case of eccentric load tests);

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

Fig. 2. Plan of the test house showing the position of the selected wall systems, sensors, and indoor heater and humidifier.

421

422

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

Hourly Average Indoor Temperatures (February-June 2013) 27 WS1-Stud

Temperature (°C )

26

WS1-Cavity WS2-Stud

25

WS2-Cavity 24

WS3-Stud WS3-Cavity

23

WS4-Stud 22

WS4-Cavity WS5-Stud

21

WS5-Cavity 20

T-Ambient-E T-Ambient-W

Fig. 3. Hourly average indoor temperatures from thermocouple sensors attached to the wall systems’ surfaces.

Hourly Average Heat Flux (February-June, 2013)

5 WS1-Stud WS1-Cavity

Heat Flux (W/m2)

0

WS2-Stud WS2-Cavity

-5

WS3-Stud WS3-Cavity -10

WS4-Stud WS4-Cavity

-15

WS5-Stud WS5-Cavity

-20

Fig. 4. Hourly average variation of heat flux readings attached to the interior wall systems. Table 2 RSI at stud and cavity of the selected wall systems and its corresponding standard deviation and COV%. WS1

Trial 1 Trial 2 Trial 3 RSI (K m2 /W) St. Dev. COV%

WS2

WS3

WS4

WS5

Stud

Cavity

Stud

Cavity

Stud

Cavity

Stud

Cavity

Stud

Cavity

3.53 3.54 3.53 3.53 0.01 0.16

5.50 5.52 5.50 5.51 0.01 0.16

4.34 4.34 4.33 4.34 0.01 0.15

7.71 7.73 7.71 7.72 0.01 0.11

2.11 2.12 2.11 2.12 0.00 0.10

7.15 7.18 7.16 7.16 0.01 0.16

3.30 3.31 3.30 3.30 0.00 0.07

5.66 5.68 5.67 5.67 0.01 0.15

2.01 2.01 2.01 2.01 0.00 0.12

5.32 5.32 5.31 5.32 0.01 0.16

Table 3 Summary of the R-values over different months from February to April, 2013. Field data

RSI (Feb.–Jun., 2013) Assembly RSI (Feb.–Jun., 2013) RSI (Feb., 2013) Assembly RSI (Feb., 2013) RSI (Mar., 2013) Assembly RSI (Mar., 2013) RSI (Apr., 2013) Assembly RSI (Apr., 2013) RSI (May, 2013) Assembly RSI (May, 2013) RSI (Jun., 2013) Assembly RSI (Jun., 2013)

WS1

WS2

WS3

WS4

WS5

Stud

Cavity

Stud

Cavity

Stud

Cavity

Stud

Cavity

Stud

Cavity

3.41 4.52 3.46 4.59 3.40 4.51 3.41 4.62 3.45 4.71 3.46 4.75

4.9

5.02 8.48 5.11 8.59 5.01 8.47 4.84 7.97 4.78 7.78 4.70 7.62

9.59

2.19 5.66 2.21 5.70 2.19 5.65 2.12 5.49 2.12 5.51 2.12 5.51

7.31

3.48 5.19 3.51 5.20 3.48 5.18 3.37 5.01 3.36 5.01 3.35 5.01

5.85

2.2 4.93 2.22 4.99 2.20 4.92 2.12 4.73 2.09 4.66 2.08 4.62

5.99

4.99 4.89 5.05 5.16 5.21

9.70 9.59 8.95 8.71 8.51

7.35 7.31 7.10 7.14 7.13

5.84 5.84 5.65 5.66 5.65

6.07 5.98 5.74 5.65 5.60

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

(c) Damage to the top plate; and (d) Plateau reaction of the wall specimen with load increments (Fig. 10). The peak load from the test is addressed as the compressive capacity of each 1.2 m (4 ft) wall system. Two types of loading are tested for each wall system: concentric and eccentric, as shown in Fig. 9(b).

423

3.1. I-Joist wall System at 406.4-mm (16-in.) spacing TJI wall studs (a product line of Weyerhaeuser NR Company) are occasionally used instead of lumber studs in regions where energy-efficiency requirements necessitate such measures. These regions also tend to use thicker walls for higher thermal resistance. Although TJI products are meant to be used in flooring or roofing, they may also be used as an alternative when a thick wall system

Hourly Average Outdoor Temperatures (February-June, 2013) WS1-Stud

40

WS1-Cavity

Temperature (°C)

30

WS2-Stud WS2-Cavity

20

WS3-Stud 10

WS3-Cavity WS4-Stud

0

WS4-Cavity -10

WS5-Stud WS5-Cavity

-20

T_outside

-30

Fig. 5. Hourly average outdoor temperatures of selected wall systems during winter, spring, and summer seasons.

Relative Humidity (February-June 2013) 100 90

Relative Humidity (%)

80 70 WS1

60

WS2 50

WS3

40

WS4

30

WS5

20

Ambient

10 0

Fig. 6. Hourly average relative humidity of selected wall systems during winter, spring, and summer seasons.

Thermal resistance convergence factors at stud and cavity of selected walls systems over time (n=12 hours)

Convergence factor

0.5

WS1-Stud

0.4

WS1-Cavity

0.3

WS2-Stud WS2-Cavity

0.2

WS3-Stud 0.1

WS3-Cavity

0 1 -0.1

121 241 361 481 601 721 841 961 1081 1201 1321 1441 1561 1681 1801 1921

WS4-Stud WS4-Cavity

Time (hour)

Fig. 7. Thermal resistance convergence factors at stud and cavity of the selected wall systems at n = 12 h.

424

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

Fig. 8. Assembly RSI values of the selected wall systems.

Fig. 9. (a) Wall specimen set-up shows the positions of vertical, in-plane, and out-of-plane deflectometers; (b) loading set-up for both concentric (left) and eccentric (right) tests.

-8 -10

-200

-12

-250

Time (s)

-4

-14

-21

-38

-69

-114

-156

0

-6

-150

-189

2

-2 -4

-100

-212

-223 0

-14

MTS Force MTS Displacement

(a) MTS 6000 Force (KN) and Displacement (mm) over Time (s) for Concentrically-Loaded I-Joist Wall System

-2 -4 -6 -8 -10 Transverse Load (KN)

-12

West Vertical East Vertical

(b) Vertical (East and West) Deformations (mm) od Concentrically-Loaded I-Joist Wall System

Fig. 10. A sample of load-deformation curves for concentrically loaded I-joist wall system at 406.4 mm (16-in.) spacing.

Deformation (mm)

-50

mid-rise construction, the I-joist wall system is the first choice for investigation. The main reason for the failure of the concentrically loaded wall specimens is the crushing of top plates, causing the entire wall

Displacement (mm)

Loading (KN)

0

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00

with a high R-value and minimal thermal bridging at framing members is desired [22]. In this section, both concentric and eccentric load-bearing tests are addressed in detail. Since this study aims at evaluating the thermal resistance and structural performance of

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

425

Fig. 11. Details of concentrically-loaded 38 mm × 140 mm (2 in. × 6 in.) staggered wall system deformation.

specimen to plateau (i.e., peak loading capacity is already obtained), and afterward the wall specimen starts unloading. TJI wall studs are proven to be strong enough to resist damage, although minor web cracks are observed at the end connections between the top plate and I-joist wall studs. Looking at the eccentric loading, we observe significant east and west out-of-plane deformations occurring in the same direction. This can be attributed to the eccentricity of loading, resulting in the inclination of the entire wall specimen away from the loading head. The tendency of the wall specimen to incline is found to be the primary reason for failure, resulting in the inclination of the entire loading set-up. At this point the test should be stopped immediately to prevent potential damage. 3.2. 38 mm × 140 mm (2 in. × 6 in.) Staggered wall system at 304.8-mm (12-in.) spacing The cause of failure of the concentrically loaded staggered wall specimens is found to be the studs that are not attached to the OSB sheathing. We observe that these studs buckle until they are crushed, causing entire specimen damage (Fig. 11). Similar to the previously mentioned eccentrically-loaded wall specimens, significant east and west out-of-plane deformations occur which can be attributed to the eccentricity of loading, resulting in the inclination of the entire wall specimen away from the loading head. The main reason for specimen failure is the tendency of the wall specimen to incline, resulting in the inclination of the entire loading set-up. 3.3. 38 mm × 184 mm (2 in. × 8 in.) Conventional wall system at 304.8 mm (12-in.) spacing This type of wall system has been proven to have a relatively higher load-bearing capacity than the other wall systems, since all wall studs are attached to the OSB sheathing, and since the load is evenly distributed among the rectangular stud sections. A significant in-plane stud buckling is reported in the three like specimens, and, consequently, all the wall studs are disconnected from the OSB sheathing. (At this point, potential specimen damage was encountered and the test was stopped immediately.) One other reason for specimen failure is the crushing of the top plate in

the last few minutes of the test. For eccentric loading, except for the excessive inclination angle of the wall specimen due to eccentricity, we observe no significant damages. It is concluded that the maximum loading capacity of a concentrically loaded wall system is slightly higher than the corresponding maximum loading capacity while eccentrically-loaded (Fig. 12). The eccentric loading capacity is found to be approximately 15% less than the concentric loading capacity, as shown in Table 4. It is also apparent that the 38 mm × 184 mm (2 in. × 8 in.) conventional wall system at 304.8-mm (12-in.) spacing achieves the highest loadbearing capacity, both concentrically and eccentrically (356 kN and 282 kN, respectively) compared to the I-joist wall system at 406.4mm (16-in.) spacing (210 kN and 185 kN, respectively), and the 38 mm × 140 mm (2 in. × 6 in.) staggered wall system at 304.8-mm (12-in.) spacing (227 kN and 227 kN, respectively). Although it has been hypothesized that the I-Joist wall system should have the highest load-bearing capacity, it is found to achieve the lowest results. However, this can be attributed to the wide stud spacing (406.4 mm (16 in.) instead of 304.8 mm (12 in.)), as well as to the top plate failure, as explained above. Regarding the staggered wall system, the primary reason for failure is that some inner studs are not stapled to the OSB sheathing, leading to excessive in-plane buckling and, ultimately, extensive damage of the specimen. In the tests, the mid-blocks attached to the staggered wall system play an important role in preventing buckling at the weak (vertical) axis; however, buckling occurs at the upper and lower quarters of the free studs (not attached to OSB). The structural test proves that the mid-blocks are not a necessary structural component in the cases of the I-Joist wall system and 38 mm × 184 mm (2 in. × 8 in.) wall system, since all the studs are stapled to the OSB sheathing. For better I-Joist load-bearing capacity, we recommend developing a new end-connection between the engineered-wood I-Joist studs and the lumber top plates. It is also possible to replace lumber top plates with engineered wood plates in order to prevent the uneven pressure applied to the area under the I-joist flanges only.

4. Discussions, conclusions and recommendations This study focuses on the energy efficiency of innovated wall systems for mid-rise wood-frame construction, as well as their

426

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

Fig. 12. Average maximum loading capacities of both concentrically- and eccentrically-loaded wall systems.

Table 4 Summary of the field test and lab test results of selected wall systems. Wall system

Wall system 1

Wall system 2

Wall system 3

Wall system 4

Wall system 5

RSI value (K m2 /W) Concentric loading (kN) Eccentric loading (kN) % Ecc./Conc.

4.75 – – –

7.62 210 185 88%

5.51 227 200 88%

5.01 – – –

4.62 356 282 79%

structural performance. The objective of the study is to evaluate the thermal resistance (RSI) and load bearing capacity of the selected wall systems for mid-rise (four-to-six storeys) wood-frame construction, where the following wall systems have been analyzed: (a) WS1: Engineering I-joist wall system at 304.8 mm (12 in.) spacing, (b) WS2: Engineering I-joist wall system at 406.4 mm (16 in.) spacing, (c) WS3: 38 mm × 140 mm (2 in. × 6 in.) staggered wall system at 304.8 mm (12 in.) spacing, (d) WS4: 38 mm × 140 mm (2 in.6 in.) staggered wall system at 203.2 mm (8 in.) spacing, and (e) WS5: 38 mm × 184 mm (2 in. × 8 in.) conventional wall system at 304.8 mm (12 in.) spacing. These wall systems have been tested for their thermal resistance performances (R-value) and structural performance through a fullscale testing house and structural lab tests, respectively. 4.1. Discussion Regarding the thermal resistance performance of the selected wall systems through the field test, we observe that all the tested wall systems achieve the minimum assembly RSI (R-value) for wood-frame construction of RSI 3.45 (R-19.6), as recommended by ASHRAE-90.1 [11]. The second wall system (I-Joist wall studs at 406.4-mm (16-in.) spacing) achieves the highest R-value, followed by the third wall system (staggered wall system at 304.8 mm (12in.) spacing) as shown in Table 4. We hypothesized that I-Joist studs reduce significantly the thermal bridge that occurs in conventional wall systems. Contrary to our expectations, the first wall system (I-Joist wall studs at 304.8-mm (12-in.) spacing) has the lowest Rvalue. The estimated R-values from WUFI simulation software show that the first wall system should have the second-highest R-value. It was also shown that R-value is not a constant number for each wall system, where it fluctuates according to the outdoor temperature. In this study, the R-value of each wall system decreased with the increase of outdoor temperature and vice versa. Regarding the structural performance of the TJI joist wall system, the main cause of failure observed is the crushing of the

top plate due to uneven distribution of the applied load over the entire I-joist members, but only at the I-joist flanges. Trus-Joist [22] recommends solid rectangular plates that match the stiffness of the I-joist flanges. In this study, the TJI joist wall system at 406.4mm (16-in.) spacing is tested for both concentric and eccentric load-bearing capacities. However, we recommend that future studies conduct testing of this wall system at narrower spans such as 304.8-mm (12-in.) spacing. This would provide a common platform for the comparison of the three wall systems, assuming that the stud spacing is constant. I-Joist wall systems, as well as staggered wall systems may be used as non-load-bearing wall systems at higher floor levels, considering their high thermal resistance and low load-bearing capacity. We observe that OSB sheathing, as well as mid-blocks play an important role in the structural performance of the wall systems. They support wall studs against significant in-plane and out-of-plane buckling. Consequently, the addition of drywall on the interior side of different wall systems may add significant strength, especially in the case of staggered wall systems, where some wall studs are not attached to the OSB, resulting in significant buckling. The conventional wall system achieves the highest load-bearing capacity, although regarding its thermal resistance performance it achieves the lowest R-values of all the wall systems. 4.2. Conclusions and recommendations Based on the results of this study, I-Joist and staggered wall systems have shown relatively high thermal resistance performance compared to the conventional wall system. On the other hand, the conventional wall system proved to have the highest load-bearing capacity. Further studies should be conducted to optimize wall systems that accommodate both high thermal and structural performances. Although it was not addressed in this point of research, relative humidity was measured within the cavities of the wall systems. Therefore, future work should include data analysis of humidity content within the wall cavities. Regarding the load-bearing capacity of the developed wall systems, we recommend making use of the I-Joist and staggered wall systems as non-load-bearing wall systems, given their high thermal resistance. We also suggest that those wall systems be tested

H. Awad et al. / Energy and Buildings 82 (2014) 416–427

at narrower stud spacing, with the objective of achieving higher load-bearing capacity. Also, in order to arrive at an optimal balance between the thermal and structural performance of the studied wall systems, hybrid (combined) wall systems could be developed which yield better results. Acknowledgements This research is conducted by the University of Alberta with financial support from the Natural Sciences and Engineering Research Council of Canada (NSERC) NEWBuildS Strategic Network/Value to Wood program. The authors appreciate the technical assistance of Royal Construction Company, Mr. Ahmad Alrifai, and Mr. Gurjeet Singh in building the experimental set-up. References [1] A. Dodoo, L. Gustavsson, R. Sathre, Building energy-efficiency standards in a life cycle primary energy perspective, Energy and Buildings 43 (2011) 1589–1597. [2] IEA, Towards net zero energy solar buildings, in: ECBCS/SHC Project Factsheet, Energy Conservation in Buildings and Community Systems Programme, United Kingdom, 2011, pp. 1–2. [3] P. Rohdin, A. Molin, B. Moshfegh, Experiences from nine passive houses in Sweden—indoor thermal environment and energy use, Building and Environment (2013). [4] K.B.P. Shanks, S.N.G. Lo, B. Norton, Appropriate energy efficient building envelope technologies for social housing in the Irish climate, Journal of Housing and the Built Environment 21 (2006) 191–202. [5] B. Berggren, M. Wall, Calculation of thermal bridges in (Nordic) building envelopes—risk of performance failure due to inconsistent use of methodology, Energy and Buildings 65 (2013) 331–339. [6] J.O. Luis Pérez-Lombard, Christine Pout, A review on buildings energy consumption information, Energy and Buildings 40 (2008) 394–398. [7] J.R. Qinru Li, Paul Fazio, Development of HAM tool for building envelope analysis, Building and Environment 44 (2009) 1065–1073. [8] Canadian-Wood-Council, Guide for Designing Energy-Efficient Building Enclosures for Wood-Frame Multi-Unit Residential Buildings in Marine to Cold Climate Zones in North America, FPInnovations, 2013.

427

[9] Canadian-Wood-Council, Managing Moisture and Wood, Forintek Canada Corporation, Ottawa, ON, 2004. [10] Krushna Mahapatra, Leif Gustavsson, Kerstin Hemström, Multi-storey woodframe buildings in Germany, Sweden and the UK, Construction Innovation: Information, Process, Management 12 (1) (2012) 62–85. [11] ASHRAE-90.1, Requirements for the Building Enclosure: Understanding the Compliance Paths for Multi-Unit Residential Buildings, ASHRAE, British Columbia, 2007, pp. 8. [12] ASTM-Standard-E72-10, Standard Test Methods of Conducting Strength Tests of Panels for Building Construction, ASTM International, West Conshohocken, PA, 2010. [13] J. Staube, D. Onysko, S. Chris, Methodology and design of field experiments for monitoring the hygrothermal performance of wood frame enclosures, Journal of Building Physics (2002). [14] Mike Blanford. Structural Insulated Panels: An Alternative to Wood Frame Construction, In: Cityscape: A Journal of Polity Development and Research, volume 11, number 2, 2009, pp 113-116. [15] H. Zaman, M. Al-Hussein, R. Mutsuddy, G. Singh, M. Gül, H. Yu, Energy efficient wall systems for mid-rise wood frame buildings, in: Proceedings, 11th International Conference on Sustainable Energy Technologies, Vancouver, BC, Canada, Sep. 2–5, 2012. [16] P.N.P. Zoltan Pasztory, Sandor Molnar, Ilona Peszlen, Modeling the hygrothermal performance of selected North American and comparable European wood-frame house walls, Energy and Buildings. 49 (2012) 142–147. [17] D. Mallav, J. Wiehagen, A. Wood, Winchester/Cemberley Homes New Constructiion Test House Design, Construction, and Short-Term Testing in a Mixed-Humed Climate, NAHB Research Center Industry Partnership, Upper Marlboro, MD, 2012 (p. Medium: ED; Size 45 pp.). [18] ASTM-Standard-C1155-95, Practice for Determining Thermal Resistance of Building Envelope Components from the In-situ Data, ASTM International, West Conshohocken, PA, 2007. [19] ASTM, Standard Practice for Determining Thermal Resistance of Building Envelope Components from the In-situ Data, ASTM International, West Conshohocken, PA, 2007. [20] ANSI-S12.60-2010/Part-1, Acoustical Performance Criteria, Design Requirements, and Guidelines for Schools, American National Standards Institute, Inc., United States of America, 2010. [21] Gül Koc¸lar Oral, Alpin Köknel Yener, Nurgün Tamer Bayazit, Building envelope design with the objective to ensure thermal, visual and acoustic comfort conditions, Building and Environment 39 (2004) 281–287. [22] Trus-Joist, Trus Joist® TJI® Joists as Wall Studs, 2013, pp. 1–4.