Lateral shear performance of sheathed post-and-beam wooden structures with small panels

Abstract

Sheathed post-and-beam wooden structures are distinct from light-wood structures. They allow for using sheathing panels that are smaller (0.91 m × 1.82 m) than standard-sized panels (1.22 m × 2.44 m or 2.44 m × 2.44 m). Evidence indicates that nail spacing and panel thickness determine the lateral capacity of the wood frame shear walls. To verify the lateral shear performance of wood frame shear walls with smaller panels, we subjected 13 shear walls, measuring 0.91 m in width and 2.925 m in height, to a low-cycle cyclic loading test with three kinds of nail spacing and three panel thicknesses. A nonlinear numerical simulation analysis of the wall was conducted using ABAQUS finite element (FE) software, where a custom nonlinear spring element was used to simulate the sheathing-frame connection. The results indicate that the hysteretic performance of the walls was mainly determined by the hysteretic performance of the sheathing-frame connection. When same nail specifications were adopted, the stiffness and bearing capacity of the walls were inversely related to the nail spacing and directly related to the panel thickness. The shear wall remained in the elastic stage when the drift was 1/250 rad and ductility coefficients were all greater than 2.5, which satisfied the deformation requirements of residential structures. Based on the test and FE analysis results, the shear strength of the post-and-beam wooden structures with sheathed walls was determined.

Graphical abstract

Keywords

post-and-beam wooden structures with sheathed walls / low reversed cyclic loading / bearing capacity / stiffness / numerical simulation

1 Introduction

The system of post-and-beam wooden structures with sheathed walls is a type of modern timber construction applied to three-story and under three-story houses [1]. It is a new type of timber construction method that is different from light wood frame construction (Fig.1) approach. In this construction method, the vertical load is supported by the beam-post system, whereas the membrane effect of the oriented strand board (OSB) formed by the panel, nail, and wooden frame supports the lateral load. Although sheathing wood frame shear walls have been extensively investigated, these studies primarily focused on light-framed shear walls. The end studs of light-framed shear walls are made of dimensional lumber with small cross-sections, and the end stud deformations cannot be neglected during the wall lateral resistance [2]. Unlike light-framed shear walls, wood-frame shear walls in post-and-beam wooden structures have larger post cross-sections; accordingly, the deformation can be neglected during the wall lateral resistance [1], which helps improve the overall lateral stiffness of the wall. This type of wall was first proposed by Japanese scholars, with limited research conducted in other countries. It is a new type of wall construction method using an innovative structural system with research significance and value.

van de Lindt [3] summarized the research results for light wood structures between 1983 and 2001. Ugalde et al. [4] reviewed studies focused on the supplemental damping in light-framed shear walls between 1990 and 2004. The walls considered in these studies typically assumed the frame shown in Fig.2(b) with panels measuring 1.22 m × 2.44 m or 2.44 m × 2.44 m. Some enterprises in China are currently adopting 0.91 m × 1.82 m panels instead of standard-sized panels. Such smaller panels are easier to transport than the standard-sized panels. However, more research is required to verify whether the lateral resistance performance of walls constructed with smaller panels is identical to that of walls with standard-sized panels.

The cross-section of the post is smaller than that of the traditional column is ≥ 105 mm × 105 mm; the post spacing is usually short in this new type of structure. The stud width is ≥ 30 mm, typically between 30 and 60 mm. Construction can be done using various wood floor forms. This new wooden construction methods combines the benefits of traditional beam−column wood construction and modern light wood frame construction approaches. The differences between them are presented in Tab.1. Fig.2 compares the differences between the three construction methods.

Some enterprises in China have built several post-and-beam wooden structures with sheathed wall construction based on Japanese codes. However, these constructions have not been included in the Chinese code, Standard for Design of Timber Structures (GB 50005-2017) [5]. The current design procedure of this construction method in China can only be regarded as a square timber structure and refers to the relevant clauses in Subsection 7.3 of GB 50005-2017. The design approach provided in GB 50005-2017 is derived from Japanese codes and primarily considers the structural plywood specified in the Japanese Agriculture Standard rather than the OSB [5]. It is necessary to study the lateral resistance of post-and-beam wooden structures with sheathed walls to provide a more accurate design basis for such construction.

Similar to light wood frame constructions, the performance of nail connections determines the lateral resistance of post-and-beam wooden structures with sheathed walls. Several experimental studies have been conducted on nail connections worldwide [6–9]. Evidence suggests that the material type and nail diameter significantly influence the performance of sheathing-frame connections. These connection studies have been fundamental to the study of the lateral behavior in post-and-beam wooden structures with sheathed walls.

This study investigates the lateral resistance performance of post-and-beam wooden structures with sheathed walls. The failure mode, bearing capacity, stiffness, ductility, and ultimate load degradation law of the shear walls with different nail spacings were obtained through a low-cycle repeated load test and numerical simulation of 13 full-scale walls.

2 Materials and methods

2.1 Test specimens

The wall specimens were designed according to the Standard for Design of Timber Structures (GB 50005-2017 [5]). All specimens were one-sided panel shear walls with a width of 910 mm and height of 2925 mm. Tab.2 presents the dimensions of the frames and panels.

The posts and beams were connected by a set of steel pins (detail 1, as shown in Fig.3). The stud and beams were joined using two pairs of 120 mm long inclined nails (detail 2, as shown in Fig.3). The panels and wooden frame (including the column, stud, top beam, and bottom beam) were connected using nails. The layout of the nails on the upper panel was in the shape of the letter “m”, while the layout of the nails on the lower panel was in the shape of the letter “w”. The nail-panel edging distance was 10 mm. The nail-post and nail-beam edging distances were both 35 mm. Thirteen specimens were divided into seven groups according to nail spacing and panel thickness. Fig.3 shows the specimen structures. Tab.3 lists the specimen numbers and specific parameters.

In this study, the wooden beam, post, and stud were made of hard pine (Pinus sylvestris). As reported in Tab.4, the strength indices for hard pine were provided by the manufacturer, Dalian Shuanghua Wooden Structure Construction Engineering Co., Ltd. The OSB panels used in this study were imported from Germany under the brand name EGGER. The design value of the shear modulus for this brand of OSB panel was 388 MPa, which was measured by a preliminary shear test. All connections between the panel and skeleton comprised ordinary galvanized iron nails with a diameter of 3.1 mm and length of 90 mm.

2.2 Test setup and loading procedure

A 100 mm square steel tube with a thickness of 10 mm was set as the loading beam on the upper part of the shear wall. A hydraulic servo actuator provided the horizontal push−pull force. One end of the actuator was connected to the loading beam and the other end was supported on a rigid frame. The bottom beam of the wall was fixed with two M18 bolts to the foundation steel beam and the top beam was fixed with two M18 bolts to the loading steel beam. Two HPB300 hold-down bars with a diameter of 18 mm were used on both sides of the wall to connect the top and foundation steel beams. A fixture was installed at both ends of the rod to prevent the specimen from floating. A steel angle with a thickness of 10 mm and width of 90 mm was installed at both ends of the wall bottom beam to prevent horizontal sliding between the wall and base. A pair of lateral braces with pulleys was set at the center of the wall to prevent out-of-plane deflection of the wall. A servo-loading controller and microcomputer controlled the entire test process. The test setup and layout are shown in Fig.4(a). Four linear variable displacement transducers (LVDTs) were arranged in four wall corners to measure the deformation, as shown in Fig.4(b). LVDTs 1 and 2 were installed horizontally at the top and bottom of the wall specimen, respectively, to measure the horizontal displacement of the upper and lower ends of the test wall. LVDTs 3 and 4 were vertically installed on the post feet on both sides of the test wall to measure the vertical displacement of the posts.

The test procedure adopted Test Method B described in ASTM E2126 [10]. This procedure is also specified in ISO 16670 [11]. The same loading protocol was used for nail connection tests.

1) A monotonic loading test was conducted to determine the ultimate displacement ∆_m of the wall under monotonic loading. According to the test results obtained by Dean and Shenton [12], Guíñez et al. [13], and Uang and Gatto [14], ∆_m was taken as 150 mm.

2) The ultimate displacement ∆ of the wall under reciprocating loading was equal to ∆_m. The formal procedure involved two displacement patterns.

3) The first pattern included five single fully reversed cycles of 1.25%∆, 2.5%∆, 5%∆, 7.5%∆, and 10%∆.

4) The second pattern included phases at displacements of 20%∆, 40%∆, 60%∆, 80%∆, 100%∆, and 120%∆. Three equal-amplitude cycles were applied for each phase.

5) If the specimen was not significantly damaged, the test continued at an increased displacement amplitude of 20%∆ per phase until the specimen was substantially damaged or ultimate load decreased to 80% of the peak load under the corresponding primary cycle.

3 Results

3.1 Experimental phenomena

Significant damage was mainly distributed in the corners and edges of the panels, where nail connection failure was accompanied by nail fractures and panel damage. In contrast, the degree of damage to the shear wall in the middle area was relatively minor. Fig.5 shows the failure phenomena of the specimens.

A wall with a 12-mm-thick panel and 150-mm nail spacing was gradually damaged during the loading process.

1) The wall deformation remained in the elastic region when the drift was within 1/250 rad.

2) When the drift reached approximately 1/100 rad, corner tear failure (Fig.5(a)) occurred at the gap between the two panels owing to the inclination of the nail.

3) When the drift reached approximately 1/50 rad, corner tear failure (Fig.5(a)) occurred at the four corners of the upper and lower panels. The feet of posts were pulled up slightly (Fig.5(b)). The nails at the edge of the panel were pulled out of the panel (Fig.5(c)).

4) When the drift reached approximately 1/33 rad, some nail connections failed at the edge owing to panel tear failure (Fig.5(d)). The nails in the middle nail row near the panel gap sank into the panel (Fig.5(e)). Slippage of the panels was apparent (Fig.5(f)).

5) Final failure occurred when the drift reached approximately 1/26–1/20 rad. At this stage, almost all nail connections at the edge of the panel were pulled out or tension-fractured (Fig.5(g)). Some nails were fractured in the middle nail row (Fig.5(h)). The remaining unbroken nails in the middle row maintained the connection between the panels and wooden frame; both sides of the panels were entirely separated from the frame.

With the same panel thickness, the failure stage was approximately the same with a denser nail row, while the degree of damage to the sheathing-frame connections increased. The wall failure process with a panel thickness of 18 mm is approximately the same as that of the wall with a thickness of 12 mm. When the panel thickness increased to 24 mm, the failure form exhibited the following differences.

1) When the drift reached 1/50 rad, corner tear failure occurred only at the gap between panels; the remaining panel corners did not exhibit a significant tear but the nail sank into the panel (Fig.5(i)).

2) When the drift reached 1/33 rad, the panel warped visibly but there was no significant tear failure, as observed in the thinner panel (Fig.5(j)). In the final stage, apparent uplift occurred at the stud foot (Fig.5(k)).

3.2 Hysteretic curve and skeleton curve

The force–drift hysteretic and skeleton curves of each specimen are shown in Fig.6.

Fig.6 indicates that the force−drift hysteresis curve of post-and-beam wooden structures with sheathed walls contained the following characteristics.

1) The hysteresis curves of all phases were accompanied by stiffness degradation. In the small-displacement stage (the drift was less than 1/200 rad), the stiffness degradation of the hysteretic curve was not apparent and the curve was plump. As the load and displacement increased, the slip of the nail connections increased. Subsequently, the curve was pinched with a noticeable smooth segment.

2) Three equal-amplitude cycles were loaded into the large-displacement stage for each phase. In each phase, the ultimate strength of the secondary cycle gradually decreased. Three prominent skeleton curves were formed with noticeable stiffness and strength degradation.

Currently, different countries adopt different methods for determining the mechanical property parameters of a wall according to the skeleton of the hysteretic curve. Because this construction method was mainly derived in Japan, this study considered the skeleton of the hysteresis curve according to the Japanese wooden structure design standards [15] to obtain the stiffness and strength characteristics of the wall. Fig.7 presents the curve-processing method used in the Japanese standard. After obtaining the target values on the curve, we determined the strength value F₀ according based on Eq. (1). The stiffness value K corresponds to the straight-line slope connecting the origin (0,0) and yield point.

(1)F_0=min\left\{\begin{array}{c}{F}_{\mathrm{y}}\\ F_{\text{μ}}=F_{\text{μ}}\times 0.2\sqrt{2\mu -1}\\ {\displaystyle \frac{2}{3}}{F}_{max}\\ F_{D=1/150\mathrm{r}\mathrm{a}\mathrm{d}}\end{array}\right\},

(2)\mu =\frac{D_{\mathrm{y}}}{D_{\mathrm{v}}},

where F_y is the actual yield force, F_u is the ultimate bearing capacity determined by the elastic−plastic theory, F_max is the maximum value of the skeleton curve, F_{D = 1/150 rad} is the corresponding strength value when the drift is 1/150 rad, D_y is the yield drift value, D_v is the drift corresponding to the ultimate bearing capacity, D_u is the ultimate drift value, and μ is the wall ductility. These values were obtained using the following steps.

1) Straight Line 1 is obtained by connecting the points whose ordinates are 0.1F_max and 0.4F_max on the rising stage of the skeleton curve (if the curve has no descent section, then the ordinate of the point on the curve whose abscissa is 1/15 rad is regarded as F_max).

2) We connect the point whose abscissa is 0.4F_max and point whose abscissa is 0.9F_max on the skeleton curve to obtain Line 2.

3) Line 2 is translated toward the positive direction of the Y-axis until the line is tangential to the skeleton curve to obtain Line 3.

4) The ordinate of the intersection between Lines 1 and 3 is F_y.

5) The abscissa of the point whose ordinate is F_y on the rising stage of the skeleton curve is D_v. D_u takes a smaller value regarding the abscissa of the point whose ordinate is 0.8F_max in the descending section of the skeleton curve and 1/15 rad.

6) Connect the origin (0,0) and (D_v,F_y) as Line 4, where the slope of Line 4 denotes the stiffness K.

7) Determine the area S surrounded by the skeleton curve, X-axis and line X = D_u.

8) Create Line 5 parallel to the X-axis so that the trapezoidal area is surrounded by Line 5, X = D_u, X-axis, and Line 4 equaling S. The ordinate of the intersection for Lines 4 and 5 is F_u, and the abscissa is D_v.

9) µ is the extension rate calculated according to Eq. (2).

Tab.5 reports the target values of the skeleton curve and average stiffness and strength values for different specimen types. The average strength F_0,unit and stiffness K_unit of the unit-length wall are stated in Eqs. (3) and (4); the results are presented in Tab.6.

(3)F_{0,\text{unit}}=F_{0,\text{AVG}}/L,

(4)K_{\text{unit}}=K_{\text{AVG}}/L,

where L is the length of the wall (m) set to 0.91 m. F_0,AVG and K_AVG are the average strength and stiffness values for the same set of specimens, which are listed in Tab.5.

4 Numerical simulation

4.1 General description

We developed a set of two-dimensional finite element (FE) models using ABAQUS. The software does not have an appropriate element for modeling nail connections. Therefore, the connections adopted user-defined elements through the subroutine VUEL in ABAQUS. This user-defined element used the hysteretic model proposed by Wu [16]. The model configuration and hysteretic roles are illustrated in Fig.8. This hysteretic model improves the model proposed by Folz and Filiatrault [17]. These two models share the same expression for the skeleton curve, as stated in Eq. (5). In the equation, f_peak is the peak load of the skeleton curve, δ_peak is the corresponding deformation when f_peak is reached, K₁ is the initial stiffness of the skeleton curve, slope of the tangent line of the rising section of the skeleton curve at point (δ_peak, f_peak) is K₂ and intercept is f_I, and K₃ is the slope of the falling section of the skeleton curve. The hysteresis rules of the two models are different: f_I and K_p change with δ_max (the maximum displacement reached in the current loading direction) in Wu’s model, whereas both f_I and K_p are constant in Folz’s model.

In Wu’s model, the first loading reaches A(δ_max, f_max). Then, the loading follows the A-B-C′-C-D-E′-E-F-G path shown in Fig.8(b). The slopes of AB and DE correspond to the unloading stiffness K_u, calculated using Eq. (6). The slopes of EF and BC correspond to the pinch stiffness K_p, calculated using Eq. (7). The vertical coordinates of the intersection points C′ and E′ on the coordinate axes are f_I, calculated using Eq. (8). The coordinates of point G are (δ_max,α_d·f_max), and the strength reduction coefficient α_d is calculated using Eq. (9). The slope of segment FG corresponds the reloading stiffness K_r, calculated using Eq. (10). h₁–h₅ are the hysteresis coefficients stated in Eqs. (6)–(10).

(5)

(6)K_{\text{u}} =h_5{K}_1,

(7)K_{\text{p}} =h_1(1-{\delta }_{max}/{\delta }_{\text{u}}{)}^2,

(8)f_{\mathrm{I}}=h_2{f}_{\mathrm{u}}\sqrt{1+{\delta }_{max}/{\delta }_{\text{u}}},

(9){\alpha }_{\mathrm{d}}=1.0+\sqrt{({\delta }_{max}/{\delta }_{\text{u}})}\times (h_4-1.0),

(10)K_{\mathrm{r}}=K_1{\left(\frac{{\delta }_{\text{u}}}{{\delta }_{\text{u}}+{\delta }_{max}}\right)}^{h_3}.

4.2 Previous nail connection tests

Prior to the wall tests, a series of low-cycle repeated nail connection loading tests was conducted following the American standards ASTM D1761 [18] and ASTM E2126 [10] which consider under the same loading regime for the wall tests.

The nail tests were conducted with 30 specimens. The specimens were divided into two types: nail connection perpendicular to the frame and nail connection parallel to the frame. Each type contained 12-, 18-, and 24-mm plate thicknesses. When the wall deforms under a horizontal force, the deformation direction of the nail connection is always multi-angular. It cannot maintain a state perpendicular or parallel to the skeleton. Therefore, the characteristic value of the nail connection bearing capacity for each panel thickness value was determined by the mean value of the parallel and vertical groups with a given thickness. The 11 parameters used in Wu’s model, obtained experimentally, are listed in Tab.7. The load-displacement curves of the nail connections were used for the numerical simulation tests.

4.3 Model verification

Seven FE models were established and simulated for seven wall specimens SJ1-1, SJ2-1, SJ3-1, SJ4-1, SJ5-1, SJ6-1, and SJ7. The B21 beam element was used in the top beam, bottom beam, and posts, and the CPS4R stress element was used in the panels. The beam–post and beam–stud joints were set as pin joints in the FE model. The sheathing-frame connection used the user-defined elements mentioned previously. The bearing condition of the bottom beam elements was set as pinned. The element lengths of the panels, beams, and posts were determined based on the nail spacing. For instance, when nail spacing was 150 mm, the element length was defined as 150 mm; and when the nail spacing was 75 mm, the element length was defined as 75 mm.

The simulation results were consistent with the experimental results. For instance, the deformation of FE model 3 with a 12-mm-thick panel and 75-mm nail spacing under the ultimate displacement is shown in Fig.9. A noticeable dislocation between the upper and lower panels is observed. The maximum dislocation displacement at the corner of the panel was 34 mm (detail 1 in Fig.9(a)). The deformation of the experimental specimen was 35 mm (Fig.9(b)), and the error between numerical simulation and experimental results was only 2.9% of the experimental results.

A comparison between the hysteresis curves of the shear wall test and the numerical simulation is shown in Fig.10. Tab.8 presents a comparison between the experimental and numerical simulation results. The average absolute errors of unit stiffness and unit strength were 7.19% and 4.94%, respectively. The maximum differences were 10% and 14%, respectively. Therefore, the errors between the experimental and numerical simulation results were within an acceptable range, indicating that the FE analysis results agreed well with the test results. Thus, we conclude that the established shear wall FE model accurately and effectively simulates the lateral resistance of the wall.

Fig.11 shows a comparison between the cumulative energy dissipation obtained in the experimental and numerical simulation tests. Equation (11) was used to evaluate the goodness-of-fit R² of the two results, where e_i denotes the cumulative energy dissipation of the FE model, G_i denotes the cumulative energy dissipation of the test specimen, and G denotes the average value of the cumulative energy dissipation in the test at each displacement. The closer R² is to 1.00, the better the FE model results fit the test specimen results. The R² values for each specimen type are shown in Fig.11. As shown in Fig.11(a)–Fig.11(g), the overall R² was greater than 0.93, and most individual R² values were greater than 0.97. Consequently, we conclude that the numerical simulation model can simulate the energy dissipation of the experimental test relatively accurately, and the finite model can be applied to other subsequent parameter studies.

(11)R^2=1- \frac{{\displaystyle \sum _{i=1}^n{(g_i-G_i)}^2}}{{\displaystyle \sum _{i=1}^n{(G_i-\overline{G})}^2}}.

5 Discussion

5.1 Lateral stiffness

When comparing the K_0,AVG values of specimens listed in Tab.5 for panel walls with the same thickness, larger nail spacings weakened the connection between panels and wooden frame. Consequently, the wall stiffness decreased. When the panel thickness was 12 mm, compared to the stiffness of the wall with 75 mm nail spacing, the stiffness of the wall with 100 mm nail spacing decreased by 9.66% and that of the wall with 150 mm nail spacing decreased by 35.71% (Fig.12(a)). When the panel thickness was 24 mm, the wall stiffness design values with 100 mm and 150 mm nail spacing decreased by 8.76% and 11.28%, respectively (Fig.12(b)). When the panel thickness was increased, the stiffness decreased more slowly with increasing nail spacing. When the same nail spacing was 150 mm, the stiffness of the wall with 12 mm thick OSB panels was 5.32% lower than that of the wall with 18-mm-thick OSB panels, and 11.04% lower than that of the wall with 24-mm-thick OSB panels (Fig.12(c)). Therefore, the panel thickness has a small positive effect on wall stiffness.

5.2 Ultimate load

Tab.9 compares the tertiary and main cyclic ultimate loads. The tertiary cyclic ultimate load decreased by 15%–22% compared to the main cyclic ultimate load. This rate is approximately the same as the result (19%) of the light-wood frame shear wall tests conducted by Cheng et al. [19], indicating that the strength degradation of the post-and-beam wooden structures with sheathed walls under low reversed cyclic loading is independent of the wooden frame structure but is determined by the hysteresis characteristics of the nail connections.

5.3 Shear strength

Fig.13(a) and Fig.13(b) summarize the average shear strength, F_0,AVG, of wall specimens with the same thickness and different nail spacing. When the panel thickness was 12 mm, the shear strength of specimen SJ3 was 5.93 kN. The shear strengths of specimens SJ2 and SJ1 were 9.4% and 35.2% lower than that of specimen SJ3. When the panel thickness was 24 mm, the shear strength of specimen SJ4 was 7.33 kN. The shear strengths of specimens SJ5 and SJ6 were 17.05% and 36.56% lower than that of specimen SJ4, indicating that the shear strength is inversely proportional to the nail spacing. The smaller the nail spacing was, the higher the shear strength became. When the nail spacing was 150 mm, the shear strengths of the wall with 18-mm-thick OSB panel and 24-mm-thick OSB panel were not significantly different. The shear strength of the wall with 12-mm-thick OSB panels was 11.04% and 13.50% lower than that of the first two walls, respectively (Fig.13(c)). The thickness of the panels had a specific influence on the shear strength of the specimen.

5.4 Ductility

The ratio of failure displacement to yield displacement of the wall specimen was defined as the wall ductility µ . As shown in Tab.5, the displacement ductility coefficient of each wall specimen was > 2.5. According to the American standard [20], the behavior of walls under horizontal loading is deformation-controlled.

5.5 Energy dissipation

This study analyzed and compared the energy dissipation capacities of each specimen through cumulative energy dissipation. The cumulative energy dissipation for each specimen is shown in Fig.14. When the horizontal displacement at the top was small (displacement < 30 mm and drift < 0.01 rad), the cumulative energy consumption of each specimen was essentially the same. With an increase in the top horizontal displacement, the cumulative energy consumption increased with a decrease in nail spacing (Fig.14(a) and Fig.14(b)) and an increase in plate thickness (Fig.14(c)).

For walls with 12-mm-thickness panels (Fig.14(a)), because the panel edges of SJ1-1 and SJ1-2 completely left the frame and stopped the loading process when the horizontal displacement reached 120 mm (drift = 0.04 rad), the cumulative energy dissipation at this time was considered for quantitative comparison. When the horizontal displacement was 120 mm (drift = 0.04 rad), the cumulative energy dissipations of walls with 100 mm nail spacing (SJ2-1 and SJ2-2) and 75 mm nail spacing (SJ3-1 and SJ3-2) were 35% and 63% higher than that of walls with 150 nail spacings (SJ1-1 and SJ1-2), respectively. For walls with 24-mm-thickness panels (Fig.14(b)), when the horizontal displacement was 150 mm (drift = 0.05 rad), walls with 100 mm nail spacing (SJ5-1 and SJ5-2) and 75 mm nail spacing (SJ4-1 and SJ4-2) consumed 36% and 70% more energy, respectively, than walls with 150 mm nail spacing (SJ6-1 and SJ6-2). For walls with 150 mm nail spacing (Fig.14(c)), when the horizontal displacement reached 120 mm (drift = 0.04 rad), the cumulative energy dissipations of walls with 18-mm-thickness panels (SJ7) and 24-mm-thickness panels (SJ6-1 and SJ6-2) were 24% and 44% higher than those of walls with 12-mm-thickness panels (SJ1-1 and SJ1-2), respectively.

Overall, these results indicated that reducing the nail spacing and increasing the thickness of the OSB panel can significantly improve the energy-dissipation capacity. Meanwhile, the bearing capacity of walls with thin panels and large nail spacing was significantly lower than that of the other walls, which were destroyed earlier with low energy dissipation. Engineers should prefer thicker panels and denser nail spacing in actual projects to ensure the energy dissipation capacity of walls.

6 Conclusions

1) Under low reversed cyclic loading, the post-and-beam wooden structures with sheathed walls mainly failed at the corners and edges of the panels. The primary failure mode was failure of the nail connections. The failure mode of the nail connections was mostly shear failure.

2) The force−drift hysteresis curve of the post-and-beam wooden structures with sheathed walls was pinched, exhibiting an apparent S shape. The ratio of the tertiary cyclic ultimate load to main cyclic ultimate load was stable.

3) The experimental and numerical simulation results showed that under the same nail specifications and diameters, an increase in panel thickness could slightly improve the wall stiffness. The strength of the walls with the same nail diameter increased significantly when the slab thickness was increased from 12 to 18 mm. Conversely, there was essentially no change in the wall strength when the slab thickness increased from 18 to 24 mm. Therefore, a wall with panels > 18 mm should use nails larger than 3.1 mm in diameter, which may result in a higher bearing capacity.

4) The FE analysis results obtained using the subroutine VUEL custom spring connection element agreed well with the test results. This custom unit can be used to accurately and efficiently simulate the lateral resistance of a wall.

5) Based on our test results, we recommended that the unit length bearing capacities of post-and-beam wooden structures with sheathed walls and 12-mm-thick OSB panels were 4.2 kN/m (150 mm nail spacing), 5.9 kN/m (100 mm nail spacing), and 6.5 kN/m (75 mm nail spacing), respectively. The unit length bearing capacities of post-and-beam wooden structures with sheathed walls and 24 mm thick OSB panel were 5.1 kN/m (150 mm nail spacing), 6.6 kN/m (100 mm nail spacing), and 8.0 kN/m (75 mm nail spacing), respectively.

References

[1] KawaiN. Permissible Stress Method for Wooden Houses Using Post-Beam Construction. 4th ed. Tokyo: Japan Housing and Wood Technology Center, 2008 (in Japanese)

[2] Källsner B, Girhammar U A. Plastic models for analysis of fully anchored light-frame timber shear walls. Engineering Structures, 2009, 31(9): 2171–2181

[3] van de Lindt J W. Evolution of wood shear wall testing, modeling, and reliability analysis: Bibliography. Practice Periodical on Structural Design and Construction, 2004, 9(1): 44–53

[4] Ugalde D, Almazán J L, Santa María H, Guindos P. Seismic protection technologies for timber structures: A review. European Journal of Wood and Wood Products, 2019, 77(2): 173–194

[5] GB50005-2017. Standard for Design of Timber Structures. Beijing: China Architecture & Building Press, 2017 (in Chinese)

[6] Anderson E N, Leichti R J, Sutt E G, Rosowsky D V. Sheathing nail bending-yield stress: Effect on cyclic performance of wood shear walls. Wood and Fiber Science, 2007, 39(4): 536–547

[7] Demir A, Demirkir C, Aydin I. The effects of wood species, nail size, grain direction and layer numbers on lateral nail strength of structural plywood panels. Sigma Journal of Engineering and Natural Sciences, 2020, 11(2): 141–148

[8] Nishiyama N, Ando N. Analysis of load-slip characteristics of nailed wood joints: Application of a two-dimensional geometric nonlinear analysis. Journal of Wood Science, 2003, 49(6): 505–512

[9] Sartori T, Tomasi R. Experimental investigation on sheathing-to-framing connections in wood shear walls. Engineering Structures, 2013, 56: 2197–2205

[10] ASTME2126-09. Standard Test Methods for Cyclic (Reversed) Load Test for Shear Resistance of Vertical Elements of the Lateral Force Resisting Systems for Buildings. West Conshohocken, PA: ASTM, 2009

[11] ISO16670. Timber Structure-Joint Made with Mechanical Fasters-Quasi-Static Reversed Cyclic Test Method. Geneva: ISO, 2003

[12] Dean P K, Shenton H W. Experimental investigation of the effect of vertical load on the capacity of wood shear walls. Journal of Structural Engineering, 2005, 131(7): 1104–1113

[13] Guíñez F, Santa María H, Almazán J L. Monotonic and cyclic behaviour of wood frame shear walls for mid-height timber buildings. Engineering Structures, 2019, 189: 100–110

[14] Uang C M, Gatto K. Effects of finish materials and dynamic loading on the cyclic response of wood frame shear walls. Journal of Structural Engineering, 2003, 129(10): 1394–1402

[15] ArchitecturalInstitute of Japan (AIJ). Standard for Structural Design of Timber Structures. Tokyo: Architectural Institute of Japan, 2006 (in Japanese)

[16] WuG. Light wood frame construction utilizing small-diameter round timber and investigation of the structural behaviour. Dissertation for the Doctoral Degree. Harbin: Harbin Institute of Technology, 2015 (in Chinese)

[17] Folz B, Filiatrault A. Cyclic analysis of wood shear walls. Journal of Structural Engineering, 2001, 127(4): 433–441

[18] ASTMD1761-20. Standard Test Methods for Mechanical Fasteners in Wood. West Conshohocken, PA: ASTM, 2000

[19] ChengHNi CLvX. Performance of perforated wood-frame shear walls with transverse walls and vertical load. China Civil Engineering Journal, 2006, 39(12): 33−47 (in Chinese)

[20] FederalEmergency Management Agency (FEMA). Pre-standard and Commentary for the Seismic Rehabilitation of Buildings, FEMA 356. Washington, D.C.: FEMA, 2000

RIGHTS & PERMISSIONS

Higher Education Press