Racking performance of poplar laminated veneer lumber frames and frame-shear hybrid walls

Abstract

This study examined the racking performance of poplar laminated veneer lumber (LVL) frames using bolted steel filling plates to connect beam-column joints, poplar LVL frames using the embedment bars to connect beam-column joints, and frame-shear hybrid walls made of poplar LVL studs and oriented strand board (OSB) sheathing panels. A new design load spreader beam was used on the side of the top of a specimen to apply monotonic and cyclic loadings. It was found that the lateral force resistance, stiffness, and ultimate loads of poplar LVL pure frames with bolted steel filling plate connections and closed rod connections were much lower than those of the poplar LVL frame-shear wall hybrid structure. The highest initial stiffness of the poplar LVL hybrid frame-shear wall was 1.77 kN/mm, which was 24% and 22% lower than that of the conventional shear wall made with spruce-pine-fir studs and OSB or plywood sheathing panels, respectively. The poplar LVL frame-shear wall hybrid structure showed lower degradation in stiffness than the conventional shear wall. The hybrid frame-shear wall structures made of poplar LVL could meet the requirements of Chinese standard; however, diagonal braces were required in use of poplar LVL pure frames.

Full Article

Racking Performance of Poplar Laminated Veneer Lumber Frames and Frame-shear Hybrid Walls

Yan Liu,^a Zizhen Gao,^b Hong-wei Ma,^a,* Meng Gong,^b and Honghe Wang ^a

This study examined the racking performance of poplar laminated veneer lumber (LVL) frames using bolted steel filling plates to connect beam-column joints, poplar LVL frames using the embedment bars to connect beam-column joints, and frame-shear hybrid walls made of poplar LVL studs and oriented strand board (OSB) sheathing panels. A new design load spreader beam was used on the side of the top of a specimen to apply monotonic and cyclic loadings. It was found that the lateral force resistance, stiffness, and ultimate loads of poplar LVL pure frames with bolted steel filling plate connections and closed rod connections were much lower than those of the poplar LVL frame-shear wall hybrid structure. The highest initial stiffness of the poplar LVL hybrid frame-shear wall was 1.77 kN/mm, which was 24% and 22% lower than that of the conventional shear wall made with spruce-pine-fir studs and OSB or plywood sheathing panels, respectively. The poplar LVL frame-shear wall hybrid structure showed lower degradation in stiffness than the conventional shear wall. The hybrid frame-shear wall structures made of poplar LVL could meet the requirements of Chinese standard; however, diagonal braces were required in use of poplar LVL pure frames.

Keywords: Poplar LVL; Frame; Frame-shear wall; Load spreader beam; Racking performance

Contact information: a: College of Civil Science and Engineering, Yangzhou University, Yangzhou, Jiangsu 225127, China; b: Wood Science and Technology Center, University of New Brunswick, Fredericton, New Brunswick, E3C 2G6, Canada; *Corresponding author: hwma@yzu.edu.cn

INTRODUCTION

Timber frame structures are common construction systems for low-rise residential houses, commercial buildings, and industrial structures in North America and Northern Europe. Modern frame timber buildings have good performance when subjected to horizontal actions, such as wind or seismic loads, as the lateral forces are transferred from the horizontal diaphragms to the vertical bracing systems and then to the foundation (Gattesco and Boem 2016). Wood has been used for thousands of years in China. However, due to the decline in China’s forest resources, few modern, heavy-wood-frame buildings have been built in the last several years. Therefore, it is essential to find new materials and products made in China to construct modern frame timber buildings.

In the 1960s and 1970s, China successfully introduced poplar (Populus euramericana cv. I-214) from Italy, and it has become popular in China. After decades of development, the domestic poplar resources are quite rich, and many wood products, such as laminated veneer lumber (LVL), are manufactured from domestic poplar and widely used as packing materials and furniture components in China. However, they are seldom used in building construction. Limited research has been conducted on the racking behavior of LVL frames, despite their extensive use in traditional Chinese wood construction. Racking behavior is the structural performance such as displacement, stiffness, and energy consumption of a specimen when subjected to lateral (racking) loading, in particular under the low-cycle fatigue loading. This study was undertaken to address the gap.

Timber framed panel walls are load-resisting elements composed of two main components: timber frame elements and sheathing materials such as oriented strand board (OSB) (Thelandersson and Larsen 2003; Premrov and Kuhta 2011). In addition, the connection types are one of the main factors that affect the racking behavior of wood frames. In recent years, many researchers have studied the mechanical properties of wood frames, such as their racking behavior. Komatsu (2004) first evaluated the shear performance of frame structures without any wall elements and then evaluated small prefabricated mud shear walls (PMSWs). The small prefabricated mud components were inserted into the frames to increase shear resistance under static push-pull cyclic lateral shear testing. Five different types of shear wall configurations were prepared and examined. It was found that an increase of stiffness of the wall system was approximately proportional to the number of mud blocks (Komatsu 2004). Shim et al. (2010) studied the lateral load resistance of hybrid structures under cyclic lateral load. They tested five types of wall systems in quasi-static reversed-cyclic load and found that the hybrid wall structures with both column and beam and light frame shear wall could resist higher lateral load than the light frame shear wall. The lateral load resistance of the hybrid wall was an arithmetic summation of the column and beam structure and the light frame shear wall. The hybrid wall with an opening had lower stiffness and lateral load resistance than the hybrid solid wall. The initial stiffness of the hybrid wall with a window opening was small, and as the lateral load increased, the lateral load resistance of the wall increased at the level of the hybrid solid wall (Shim et al. 2010). Suzuki and Maeno (2013) evaluated the seismic performance of traditional wood frames from shaking table tests and static tests using several scale models. They found that the horizontal restoring force of the wooden frame without walls depended mainly on the bending moment resistance from tie beams and the restoring force due to column rocking. When frame deformation was small, a major part of the total restoring force was the restoring force due to column rocking. The bending moments from tie beams became dominant as the deformation increased, and the traditional wooden frame had large flexibility and deformability. Erikson (2003) investigated the effects of lateral load on the stiffness of full-scale timber frames, in which wood columns and beams were connected with wood pegs. One-story, one-bay frames and two-story, two-bay frames made of various wood species were tested in both unsheathed and sheathed conditions and modeled with a structural analysis program. Load-slip characteristics were analyzed for single-fastener SIP (structural insulated panels)-to-timber connections. Excessive displacements of the frames indicated an unacceptable flexibility when subjected to reversible lateral loads, and the knee brace system provided exceptional strength characteristics due to the substantial available compressive action of the joints. Further, the use of SIP sheathing improved frame stiffness to a large degree (Erikson 2003). Lam et al. (2008) investigated the contribution of self-tapping screws as perpendicular-to-grain reinforcements for bolted glue-laminated timber connections with slotted-in steel plates. The test results from the beam-to-column connection specimens showed that the connections reinforced with self-tapping screws had increased capacities by a factor of 2 and 1.7 when compared to un-reinforced connections under monotonic and reversed cyclic loading, respectively (Lam et al. 2008). Schwendner et al. (2018) compared the load-bearing characteristics under cyclic loading of light-frame walls with gypsum fibreboard sheathing and oriented strand board sheathing, respectively. The results indicated that light-frame walls with OSB and gypsum fiberboards as sheathing materials had similar behaviour.

This study aimed to investigate the racking performance of poplar LVL frames and frame-shear walls to verify the feasibility of using poplar LVL to fabricate frame-shear wall hybrid structures and provide a theoretical basis for the design and application of LVL structures.

EXPERIMENTAL

Design of Specimens

The specimens were designed according to the requirements specified by Chinese standards (Editorial Committee of the Wood Structure Design Manual 2005; GB 50005 2017). The LVL pieces made of poplar (Populus euramericana cv. I-214) were purchased from Siyang Jiuhe Wood Industry Co. Ltd., Siyang, Jiangsu, China and used to fabricate three kinds of frames, which included two types of pure beam-column frame specimens (KJ1 and KJ2) and one type of frame-shear wall specimens (KJ3). The properties of the LVL are shown in Table 1 (Liu et al. 2017). According to Table 1, the mechanical properties of poplar LVL were similar to SPF lumbers. The frames of the three types were designed with the same dimensions, namely 2.55 m in height and 2.3 m in width. Each beam had a 75 mm × 150 mm cross-section, and each column had a 150 mm × 150 mm cross-section. The panel in the frame-shear wall (KJ-3) was 9.5-mm-thick, 1.2-m-wide, and 2.4-m-tall OSB board.

Table 1. Physical and Mechanical Parameters of Poplar LVL

The KJ1 and KJ2 beam and column joint specimens were fabricated with bolted steel filling plates and rods, respectively. The connections of the beam and column joints for KJ3 were the same as those of KJ1. As shown in Figs. 1 and 2, 8-mm-thick Q235 steel plates and bolts with a nominal diameter of 8 mm and a strength of 8.8 grade were used to make bolted steel filling plate connections. Ten-mm-thick Q235 steel plates and bolts with a nominal diameter of 14 mm and a strength of 8.8 grade, as shown in Figs. 3 and 4, were adopted for the bolted steel filling plate connections at the LVL column foot. The rod connections of the beams and columns of specimen KJ2 are shown in Figs. 5 and 6. A threaded rod of 12 mm in diameter and a strength of 8.8 grade was used. The penetration depth of rods used should be greater than 15 times the diameter of the rods used to avoid brittle pulling-out failure. The penetration depth was 200 mm.

Fig. 1. Elevation of the steel filling plate

Fig. 2. Axonometrical drawing of the steel filling plate

Fig. 3. Profile of the column base joints

Fig. 4. Axonometrical drawing of the column steel filling plate

Fig. 5. Facades of the planting bar joints

Fig. 6. Position site of the planting joints

For the racking resistance test, three types of specimens (KJ1, KJ2, and KJ3) were used. Each group had two specimens, which resulted in a total of six specimens, namely KJ1-a, KJ1-b, KJ2-a, KJ2-b, KJ3-a, and KJ3-b, where “a” denotes the specimen used for the monotonic load test, and “b” denotes the specimen used for the cycle load test. The specimens are illustrated in Figs. 7, 8, and 9.

(a)	(b)
Fig. 7. Steel filling plate bolt connection of LVL frame KJ1: (a) Design; (b) Photo
Fig. 8. Rod joint connection of LVL frame KJ2: (a) Design; (b) Photo
Fig. 9. Frame with frame-shear wall hybrid structure KJ3: (a) Design; (b) Photo

Testing Setup

A novel lateral horizontal loading device developed by Liu et al. (2008) was used, which was available at the Structural Laboratory, Yangzhou University, Yangzhou, China. This device kept the applied load unchanged due to the load head at the lateral edge. The load head shifted along with and to offset the shifting of the specimen caused by the action of lateral force, as shown in Fig. 10. The novel loading beam does not restrict the lateral and vertical movements of a specimen subjected to the lateral loading, and it can also ensure the effective transferring of lateral loads.

Fig. 10. Experimental setup for testing a wall

Arrangement of LVDTs

For the pure LVL frame specimens (KJ1 and KJ2), two linear variable differential transducers (LVDTs) (V₁ and V₂) were mounted at the column foot, and two LVDTs (V₃ and V₄) were installed at the axis points at both ends of the beam. For the timber frame-shear wall hybrid structure specimens (KJ3), in addition to the four LVDTs (V₁, V₂, V₃, and V₄) arranged at the column bottom and beam ends, three additional LVDTs (V₅, V₆, and V₇) were installed on the longitudinal wall bottom beam plate to measure the vertical relative displacement between the stud and steel foundation beam. To optimize the test and analyze the global and local deformation properties of the specimens, a dial indicator was also placed on the frame to measure the displacement of the specimen stud relative to both the top and bottom beam plates and the rotation angle relative to the frame. Figure 11 illustrates the detailed arrangement of the LVDTs (V₁ to V₇).

Fig. 11. Test point layout: (a) layout of the pure frame measurement points; (b) Wood frame-shear wall measuring point layout

Loading Scheme

The loading procedure specified by ISO-16670 (2003) and the displacement control loading procedures specified by ASTM E2126 (2009) were adopted for this study. The low-cycle reversed loading tests were conducted on all three groups, as shown in Table 2. The loading rate of the unidirectional loading test was 7.5 mm/min, and that of the reversed loading test was 3 mm/s. The relative deformations of the pure frame specimens (KJ1 and KJ2) became too large when they reached 200 mm with the subsequent termination of loading. Thus, the corresponding load at this displacement was defined as the ultimate load.

Table 2. Load Test Loading Scheme

Calculations

Stiffness degradation

The degradation in stiffness fully reflected the occurrence and propagation of the cracks in members and the development of plastic deformation in members due to the cumulative effects of structural damage under the reversed load. The stiffness of a specimen is usually expressed in the secant stiffness K, which is calculated with Eq. 1,

 (1)

where Q_i indicates the peak load at time i, and Δ_i indicates the peak displacement at time i.

Energy dissipating capacity

The energy dissipating capacity of a specimen can be expressed by the equivalent viscous damping coefficient ξ_eq. The plumpness of the hysteresis curve of a specimen was positively correlated to the equivalent viscous damping coefficient ξ_eq, and they increased together to form a higher energy dissipating capacity. The equivalent viscous damping coefficient (ξ_eq) can be calculated with Eq. 2,

(2)

where E is the energy dissipating coefficient, which can be calculated by using Eq. 3, according to JGT/T 101 (2015),

(3)

where S_{ABC + CDA} is the area enclosed by the hysteresis curve in one cycle (i.e., the energy dissipated), and S_{OBE + ODF} is the area enclosed by the hypothetical elastic line DB and the coordinate axis of displacement. The equivalent viscous damping coefficient is shown in Fig. 12.

Fig. 12. Calculation of the equivalent viscous damping coefficient

RESULTS AND DISCUSSION

Tests Phenomena

During the monotonic loading test, the column bottom bolts and bolted steel filling plate connections continuously generated cracking sounds as the horizontal displacement rate decreased and they were squeezed into the timber. The failure of a specimen was initiated near the column bottom bolt holes; splitting occurred along the height of the poplar LVL column. Further, splitting cracks appeared at the column bottom of the bolted steel filling plate connection frame (KJ1-a) and the penetrated bar connection frame (KJ2-a) when the displacements reached 60 mm and 50 mm, respectively. Such splitting cracks at the column bottom extended but did not pass through the bolt holes as horizontal displacement increased. Figure 13 shows the failure observed in specimen KJ1-a.

Specimens KJ-1b and KJ-2b were still in the stage of elasticity due to the range of the actuator under the low cyclic repeated loading, and there was no failure.

Fig. 13. Failure of KJ1-a: (a) Lateral deformation of KJ1-a; (b) Column foot splits and breakages of KJ1-a

As specimen KJ3-a was monotonically loaded, dislocations occurred at the splicing joints between the panels when the load reached 32 mm. As displacement increased, crushing was found at the wall panel bottom, and the nails were pulled out at the panel splicing joints. Additionally, outward warping was found in the OSB panels, which were separated from the studs to the right of the splicing joints. Throughout the test, stud V₆ at the middle of the timber shear wall was obviously pulled up more than the other studs, whereas the studs at both ends were not pulled up substantially. In this process, no obvious failure was observed at the beam or column ends of the LVL frame. When the LVL frame shear wall failed, the frame remained in good condition with a certain bearing capacity. For KJ3-a specimen, the stiffness was greatly increased since the OSB panels was used. As shown in Fig. 14, the main failure of KJ3-a occurred at the joints between the sheathing and studs, in which the specimens were considered being failed. Since the KJ1-a and KJ2-a specimens did not have sheathing and middle studs, the displacement to lead to failure of the joint between the post and beam were much larger than that of sheathing and studs. According to the standards, the specimens were considered as having failed when the displacement reached 200 mm. In this case, the KJ1-a and KJ2-a specimens during testing were stopped when the displacement reached 200mm although they did not collapse.