The timber used for the test is pine (Pinus banksiana) from Canada, with straight grain and some knots, an average moisture content of about 13%, and an end face dimension of 38 mm × 89 mm. The glulam is made of Spruce-Pine-Fir (SPF) from Canada, with an average moisture content of about 13% and an end face dimension of 120 mm × 120 mm. The detailed dimensions are shown in Figs. 2-5.
The test walls are consisted of frames and wall panels. Studs of the WPC, pine, glulam and lightweight steel are used to assemble the frames. Wall panels include exterior wall hanging plates and interior panels.
The WPC studs are made of hollow structural polyethylene (PE) WPC. The dimension of end face is 40 mm × 90 mm and the wall thickness of the WPC studs is 7 mm.
The exterior wall hanging plates are made of PE co-extruded wood plastic. There are two kinds of exterior wall panels. One is rib profiles with end face dimensions of 20 mm × 200 mm, the other is hollow sheets with end face dimensions of 50 mm × 200 mm, and both of them have wall thickness of 7mm. The interior panel is made of PVC WPC with 1220 mm × 2440 mm exterior dimensions, 8 mm wall thickness and 930 kg/m3 density. The WPCPE means PE co-extruded wood plastic, and WPCPVC means PVC co-extruded wood plastic. Fig. 6 shows the test materials of wall stud and wall panels.
The snap-fit PVC plastic exterior wall hanging plates have a dimension of 230 mm × 4000 mm and a wall thickness of 1.5 mm. The lightweight steel for the walls is cold-formed thin-walled C steel, with a cross-sectional height of 73 mm, a cross-sectional width of 33 mm, a seam width of 26 mm and a thickness of 1 mm. The glass fiber cotton filled between the studs is composite aluminum foil glass wool, with a thickness of 50 mm, a dry density of 11 kg/m3, and a fire resistance grade of A non-combustible. The specification of fireproof paper plasterboard for wall cladding is 1220 mm × 2440 mm × 9.5 mm, the surface density is 9.5 kg/m2, and the fire resistance grade is A non-combustible, environmental protection grade E0 (formaldehyde emission ≤ 0.5 mg/L) (GB 8624-2912: Classification for Burning Behavior of Building Materials and Products). Breathing paper for waterproof and moisture proof, the appearance was layered and concave, a thickness of 0.5 mm, and a fracture elongation of 88%. Waterproof and moisture-proof respiratory paper has a laminar concave and convex appearance, the thickness is 0.5 mm, the elongation at break is 88%, and the impermeability is 2 m water column 24 h impermeability.
Figs. 2–5 show the typical structure of the wall. Referring to GB 50005-2017 (Standard for Design of Timber Structures) and GB 50361-2018 (Technical Standard for Infills or Partitions with Timber Framework), the test wall was constructed in the form of a frame shear structure, i.e., framed in glued laminated timber and shear-resistant by nailing oriented structural particle board (OSB) to the frame and to the studs in the frame, with nail spacing of 150 mm.
The test walls are divided into single-layer frame walls and double-layer frame walls. The single-layer frame wall means that the outer surface of the wall using a hanging board carried by the wall, and the double-layer frame wall refers to a kind of wall which arranges a protective wall on the outside of the frame shear wall carried by the foundation and supported by frame shear wall. The wall studs are of lightweight steel, WPCPE and pine.
The inner wall cladding is covered with 9.5 mm thick fireproof paper plasterboard and the outer wall panels are fixed to the outside of the framed walls by hanging plates, in the form of 20 mm thick WPCPE, 50 mm thick WPCPE and PVC hanging plate. The wall frame cavity is filled with glass fiber cotton. Fig. 7 shows the structure details of wall (W4).
Designs of 11 groups of wall samples are shown in Table 1. In order to evaluate the influence of the wall studs on the insulation performance of wall, lightweight steel, WPCPE, pines are respectively adopted as the wall studs. The external wall hanging plate was suspended from the outside of the wall to achieve the functions of insulation, decoration, and protection. The structural form, material and sealing performance of the external wall hanging panel will affect the heat insulation capacity. In order to evaluate the influence of the external wall hanging plate on the wall insulation performance, 15 mm thick PVC board, 20 mm thick WPCPE board, and 50 mm thick WPCPE board are used as the external wall hanging plates. The walls were assembled and tested in the laboratory of Nanjing Forestry University.
Set code Wall stud type Interior wall panel Exterior wall panel Remarks G1 Light steel WPCPVC - Single-layer wall S1 Pine - - Single-layer wall S2 Pine WPCPVC - Single-layer wall S3 Pine WPCPVC WPCPVC Single-layer wall S4 Pine WPCPVC WPCPE with thickness of 20 mm Single-layer wall S5 Pine WPCPVC WPCPE with thickness of 50 mm Double-layer wall W1 WPCPE - - Single-layer wall W2 WPCPE WPCPVC - Single-layer wall W3 WPCPE WPCPVC WPCPVC Single-layer wall W4 WPCPE WPCPVC WPCPE with thickness of 20 mm Single-layer wall W5 WPCPE WPCPVC WPCPE with thickness of 50 mm Double-layer wall Notes: "-" in the table indicates that there is no such item. WPCPVC, polyvinyl chloride co-extruded wood plastic; WPCPE, polyethylene co-extruded wood plastic.
Table 1. Structural wood wall samples integrated with WPC.
According to GB/T 13475-2008 (Thermal Insulation—Determination of Steady-state Thermal Transmission Properties—Calibrated and Guard Hot Box), the thermal insulation performance of walls was tested by the guard hot box method. The test walls were made according to the design and had the dimension of 1100 mm × 910 mm.
Fig. 8 shows the layout of measuring points for wall's insulation test. Eighteen temperature sensors were equally interval fixed at the specified position on both side and outside surfaces of the wall by tin foil. One measuring point was fixed in the middle of the wall filled with glass fiber cotton, and 8 measuring points were evenly fixed in the thermal bridge part where the surrounding studs were located. The spacing between measuring points was 300 mm. The air temperature measuring point was arranged in a suspended position between the wall and the cold-hot box. The heat flux plates were located in the center and sides of the wall. The temperature of the hot box was set to 50 ℃ and the time interval of measuring point temperature for 30 min. After reaching the steady state, the cold surface air temperature, the temperature of each measuring point on the cold surface, the hot air temperature, the temperature of each measuring point of the hot surface and the heat flux value at each interval were recorded. The heat transfer coefficient value of each wall was calculated according to GB/T 23483-2009 (Test Standard for Overall Heat Transfer Coefficient of Building Envelope and Heat Supply for Space Heating).
2.3. Heat transfer coefficient test method for walls
Fig. 9 shows heat insulation test of materials. In the test, 11 groups of walls show the same regularity. Take W5 wall with 50 mm thickness of WPC exterior wall panel as an example. Fig. 10 shows the test temperature of W5 wall at each point. The indoor environmental temperature (thermostat temperature of the hot box) rised first, then fell, and tended to be stable after 6 h.
After the temperature-controlled hot box was heated, the temperature rised, when the threshold value was exceeded, the heating stopped, and the temperature slowly dropped to the set temperature. The heating system was automatically controlled by the system, and the hot box was always in a dynamic environment with dynamic balance. After reaching steady state, the temperature of the measuring point between the position of the wall studs on the external surface of wall was generally higher than the temperature of the measuring point of the position of the glass fiber cotton. The wall studs were the main part of heat conduction. There was a certain thermal bridge effect leading to the increase of heat flow, heat consumption and temperature. Fig. 11 shows that heat flux value tends to be steady after 12 h. Due to the different materials used in the construction of different parts of the wall, the heat flux value was different. The heat flux value at the wall studs was generally higher than the that at the glass fiber cotton. After reaching steady state, the heat flux fluctuated in the range of 5-10 W/m2. The WPC wall was a frame-shear structure composed of multi-layer materials, in which the WPC wall studs were arranged at a certain interval. The heat transfer rates of the wall studs and the cavity portion were different. The wall studs reduced the thermal resistance of the wall and then promoted heat conduction.
In the figure, q1 and q2 indicate heat flux of the position of wall studs and insulation cotton, respectively.
Fig. 12 shows the heat transfer coefficient of different locations of W5 wall. The calculation of heat transfer coefficient of different parts of the wall should be based on the requirements of JGJ/T 132-2009 (Standard for Energy Efficiency Test of Residential Buildings). The final calculated value of the thermal resistance of the main part of the retaining structure was not more than 5% compared with the calculated value before 24 h. According to the average heat transfer coefficient of 0.237 W/(m2·K) and 0.188 W/(m2·K) after 12 h in the location of studs and in the location of insulation cotton, the heat transfer coefficient at the studs was 20.7% higher than that at the insulation cotton on average.
In the figure, K1 and K2 indicate heat transfer coefficient of the position of wall studs and insulation cotton, respectively.
Considering the obvious influence of thermal bridge effect in the location of studs, the total heat transfer coefficient of the wall must take the influence of heat transfer in the location of the studs into account. The total heat transfer coefficient of the walls can be calculated by the area-weighted method (Equation (1)):
where U is the total heat transfer coefficient of the test wall (W/(m2·K)); Si is the proportion of the location of insulation cotton (glass fiber cotton) area to the wall area, which is calculated to be 53%; Ki is the heat transfer coefficient in the location of insulation cotton (glass fiber cotton) (W/(m2·K)); Sf is the proportion of the location of WPC wall studs area to the wall area, which is calculated to be 47%; and Kf is the heat transfer coefficient in the location of WPC wall studs (W/(m2·K)).
The total heat transfer coefficients of each groups of walls calculated according to Equation (1) are shown in Table 2. According to GB50189-2015 (Design Standard for Energy Efficiency of Public Buildings), the effective heat transfer coefficients U of G1, S1 and W1 < 0.6 W/(m2·K), which meet the standard of wall thermal level Ⅲt and is suitable for cold regions or hot summer and cold winter regions. The S2, S4, W2, W4 effective heat transfer coefficients U < 0.5 W/(m2·K), meet the standard of wall thermal level Ⅱt, and is suitable for cold regions. The S3, S5, W3, W5 effective heat transfer coefficients U < 0.4 W/(m2·K), which meet the standard of wall thermal level Ⅰt and can be applied to severe cold regions. The walls with the WPC as the wall studs are made of the WPC exterior wall hanging panels with thickness of 50 mm, which have the lowest total heat transfer coefficient and the best thermal insulation performance.
Set code Heat transfer coefficient in location of wall studs Heat transfer coefficient in location of insulation cotton Total heat transfer coefficient G1 0.583 0.510 0.544 S1 0.580 0.530 0.553 S2 0.503 0.434 0.466 S3 0.399 0.365 0.381 S4 0.439 0.382 0.408 S5 0.403 0.354 0.377 W1 0.587 0.466 0.523 W2 0.484 0.386 0.432 W3 0.338 0.258 0.295 W4 0.481 0.354 0.414 W5 0.237 0.188 0.207
Table 2. Measured heat transfer coefficient of structural wood wall integrated with WPC (W/(m2·K)).
The thermal conductivity of the WPC and OSB was measured by thermal characteristic analyzer (Model: ISOMET 2104). The thermal parameters of other materials were obtained according to GB50176-2016 (Code for Thermal Design of Civil Building). Table 3 shows the parameters of wall materials. The thermal conductivity of lightweight steel was the highest, and the relationship between thermal conductivity of materials was as follows: light steel > WPCPE > PVC > WPCPVC > pine. The higher thermal conductivity, the worse the thermal performance. As a result, the thermal performance of pine was better than WPCPVC, WPCPVC was better than PVC, PVC was better than WPCPE, and the thermal performance of WPCPE was better than light steel. In addition, the thermal conductivity of the WPCPE panel is close to that of common building wall covering panel (such as paper-based fireproof gypsum board, OSB, etc.). In terms of thermal performance, the WPCPVC and PVC were close to the requirements of thermal insulation material (thermal conductivity < 0.14 W/(m·K)). Compared with light steel materials, the WPCPE has a great thermal insulation advantage. It can be found that using WPC materials as building external wall panels, interior wall panels and wall studs has obvious advantages in building thermal performance.
Material Density (kg/m3) Thermal conductivity (W/(m·K)) Light steel 7830 58.200 PVC 1310 0.170 Pine 430 0.137 WPCPVC 730 0.156 WPCPE 1310 0.338 Paper fireproof gypsum 9070 0.330 OSB 650 0.340 Glass fiber cotton 70 0.035 Notes: PVC, polyvinyl chloride; OSB, oriented structural particle board.
Table 3. Thermal parameters of wall materials.
Fig. 13 shows that the G1 wall had the highest heat transfer coefficient while W2 wall had the lowest. The heat transfer coefficient of the S2 wall was 0.078 W/(m2·K) lower than that of the G1 wall with a drop of 14.3%. The insulation performance of the walls using the pine as the wall studs was better than that of the wall with wall studs made of lightweight steel. Lightweight steel was a good thermal conductor, which was not conducive to heat insulation. The obvious thermal bridge effect in the wall was an important reason for the poor thermal insulation of WPC buildings based lightweight steel structures. The heat transfer coefficient of the W2 wall was 0.034 W/(m2·K) lower than that of the S2 wall, and the drop was 7.3%. The insulation performance of the wall with the WPC material as the wall studs was better than that of pine.
Although the thermal conductivity of the WPC was higher than that of pine, the hollow structure of the pine studs and the addition of an air layer with good thermal insulation to the wood-plastic material resulted in better thermal performance than that of the pine for the wall studs, and the total heat transfer coefficient of the wall was further reduced.
Fig. 14 shows that the total heat transfer coefficient of the wall with the WPC materials as the wall studs were W5 < W3 < W4 < W2 < W1. The W5 had the lowest heat transfer coefficient and the highest heat transfer resistance. The W5 with 50 mm thick WPC external wall panels had the best thermal performance, which was determined by the hollow section structure of external wall panels and the excellent thermal performance of WPC materials. The W3 < W4 indicated that the thermal performance of the wall using PVC panels was better than that of WPCPE exterior wall panels with thickness of 20 mm. The thermal conductivity of the PVC was lower than that of WPCPE, and the thermal insulation performance of PVC was better. Due to the thin thickness of the PVC panels, the WPC exterior wall panels could obtain better thermal performance by increasing the thickness and adopting the hollow structure. The reason for W2 < W1 was the setting of the inner wall panel, which increased the structural thickness, and WPCPVC was an excellent thermal insulation material. As an interior wall facing slab, WPCPVC played a role in internal decoration while also providing good thermal insulation. The heat transfer coefficient of exterior WPC panels of the W4 wall with thickness of 20 mm and the exterior WPC panels of W5 wall with thickness of 50 mm were reduced by 20.8% and 60.4% respectively compared with the W1 wall without the external wall panel. The wall using pine as the wall studs also reflected the same thermal performance rule: S5 < S3 < S4 < S2 < S1. The heat transfer coefficients of the WPC integrated wall with internal wall panels, external wall panels with thickness of 20 mm and 50 mm were 26.2% and 31.8% lower than those without the internal wall panels and external wall panels, respectively. Reasonable setting the wall panels material and wall structure is an important measure to improve the thermal performance of the wall.
The earliest WPC appeared nearly a century ago where wood flour was combined with phenol-formaldehyde resin to create a composite material used as an automobile gearshift knob (Gardner et al., 2015). At present, the WPC is a kind of high-performance composite material, characterized for its excellent mechanical properties, insulation performance, and decorative performance. The frame-shear structure wall in the building has a good seismic performance and flexible space layout. In the past few decades, there have been fewer researchers studying wood-plastic frame-shear walls. The study of the WPC wall is still an area to be developed. Some researchers have studied the seismic performance of recycled plastic wood wall. The results showed that the recycled plastic wood wall had good seismic performance under cyclic loading (Yang et al., 2019). In addition, relevant researchers studied the sound insulation performance of the WPC wall, and the wall stud and the exterior wall panel material of the WPC wall had a significant influence on the sound insulation effect of the wall (Herrera et al., 2018). However, there has been little research on the thermal insulation properties of the WPC walls. The frame-shear wall has a good resistance to horizontal loads. This study is focused on the thermal insulation properties of the WPC frame-shear walls, exploring the factors affecting the integration of the WPC wall and promoting the development of the WPC buildings.