Experimental investigation of laterally loaded double-shear-nail connections used in midply wood shear walls

Construction and Building Materials 101 (2015) 761–771

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Experimental investigation of laterally loaded double-shear-nail connections used in midply wood shear walls Wei Zheng a, Weidong Lu a,⇑, Weiqing Liu a, Lu Wang a, Zhibin Ling b a b

College of Civil Engineering, Nanjing Tech University, Nanjing 211816, PR China School of Civil Engineering, Southeast University, Nanjing 210096, PR China

h i g h l i g h t s
Laterally loaded double-shear-nail (DSN) connections with different variables.
The failure modes are affected by the sheathing thickness and the nail edge-distance.
Influence of different variables on the load-displacement behavior of DSN connections.
The analytical model can be used to describe the response of DSN connections.
The results can be used to develop baseline data for DSN connection model.

a r t i c l e

i n f o

Article history: Received 28 May 2015 Received in revised form 8 September 2015 Accepted 16 October 2015

Keywords: Nail connection Double-shear Load–displacement behavior Failure mode Wood shear wall

a b s t r a c t Midply wood shear wall is regarded as a preferable lateral resistance system for timber frame structures. One of the most important reasons is attributed to the good lateral performance of double-shear-nail (DSN) connections. In this study, eight groups of specimens were tested under monotonic lateral loads to investigate the influence of different variables on the lateral performance of the DSN connections. Representative variables derived from practical midply wood shear walls were considered, including sheathing thickness, nail edge-distance and loading direction to the grain of framing. Test results indicate that the failure modes of the DSN connections depend on sheathing thickness and nail edge-distance. The increase of nail edge-distance and sheathing thickness can significantly improve the ultimate strength and the ductility, while it had little influence on the initial stiffness. Moreover, loading direction to the grain of spruce–pine–fir (SPF) framing had significant influences on the ultimate strength and initial stiffness of DSN connections. Finally, the testing data was idealized to translate the load–displacement behavior of the DSN connections through a typical exponential function model. This study can be used to develop baseline data for the finite-element DSN connection model to predict the performance of midply wood shear walls. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Platform type timber frame structure is a kind of comfortable, economical and environmental friendly constructions, which is commonly used in residential houses and multi-storey buildings in North America and Europe. For this construction, timber shear walls, working as vertical components, play an important role in resisting lateral loads, e.g. earthquake loads and wind loads. A few years ago, FPInnovations [1] proposed an advanced shear wall concept named midply wood shear wall system, which was

⇑ Corresponding author at: 30# Puzhu South Street, Nanjing 211816, PR China. E-mail address: [email protected] (W. Lu). http://dx.doi.org/10.1016/j.conbuildmat.2015.10.100 0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.

regarded as a preferable structural system that can provide superior lateral resistance. In this system, the sheathing is placed in the center of the wall and sandwiched by a serious of studs and plates on both sides of the wall sheathing. The studs and plates are placed at a 90° rotated position relative to those in standard shear walls, as shown in Fig. 1. Due to this innovative arrangement of components, midply wood shear wall system is able to provide higher lateral resistance compared to standard shear wall system. Previous studies [2–4] revealed that the load-carrying capacity and stiffness of midply wood shear walls were nearly three times those of standard shear walls. Moreover, the common failure modes of the sheathing connections [5–8], which were directly related to the destruction of standard shear walls, can virtually be eliminated

762

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

Sheathing

Standard shear wall 406mm

Midply shear wall 610 or 406mm

Stud 406mm

Nail

406mm

Stud

Sheathing

Nail

610 or 406mm

Fig. 1. Top view section of a standard wood shear wall and a midply wood shear wall.

in midply wood shear walls. Therefore, the double-shear-nail (DSN) connections which were distinct from the single-shear nail connections in standard wood shear walls (see Fig. 2) was regarded as one of the most important reasons for the superior performance of midply shear walls. It is commonly accepted that the performance of a standard timber shear wall is primarily determined by the response of the nail connection rather than the properties of the timber members themselves [9–15]. Hence, many researchers developed various sophisticated finite-element models to predict the lateral performance of shear walls with nail connections in the past decades. Dolan and Foschi [16] employed a finite-element model to simulate standard shear walls. Comparison between the simulated results and the experimental results of full-size standard shear walls indicated that the model was effective to predict the performance of standard shear walls. Judd [17,18] developed a new analytical model for nail connections to predict the performance of standard shear walls. In this model, a nonlinear spring pair oriented along the initial displacement trajectory of nail connections was adopted to simulate the actual behavior of nail connections. The results indicated that this model provided a closer fit to test data compared to previous nail connection model. Likewise in midply wood shear wall system, it can be considered that the properties of the DSN connections among the sheathing and framings have a significant effect on the performance of midply wood shear walls. In the meantime, it is possible to predict the performance of midply wood shear walls by using similar finite-element methods. Despite the studies on the response of DSN connections are few, extensive studies on single nail connections in standard timber shear walls can provide some references. Dolan and Madsen [19] investigated the lateral monotonic and cyclic behavior of singleshear nail connections with the intent of finite-element modeling. The connections fabricated with different sheathing types were loaded parallel and perpendicular to the grain of sheathing and framing. The results indicated that the influence of loading direction on the connection response was small, whereas the sheathing type had an effect on the connections load–displacement curves near the ultimate load capacity of the connection. Girhammar

Sheathing panel

et al. [20] also conducted an experimental study on single-shear nail connections with different sheathing materials, considering the influence of loading direction to grain. The results showed that both loading direction to grain and nail edge-distance had obvious effects on the failure modes and the maximum loading capacity of the connections. Buitelaar [21] and Karacabeyli et al. [2] tested nine types of DSN connections with different nails and sheathing materials under monotonic and cyclic loads. The results indicated that the maximum load of DSN connections is about 80% greater than that of single-shear nail connections, and the initial stiffness of the former is about three times that of the latter. However, these conclusions were drawn just basing on the test results of DSN connections in a specific configuration without regard to other important variables, such as the nail edge-distance and the loading direction to the grain of framing and sheathing. Therefore, their conclusions were not representative and comprehensive. To evaluate the influences of different variables on the response of DSN connections, eight groups of single fastener DSN connections were tested under lateral loads in this study. The experimental variables consisting of nail edge-distance, sheathing thickness and loading direction to the grain of framing were derived from practical midply wood shear walls. It was also intended to develop baseline data for a finite-element DSN connection model so as to predict the performance of midply wood shear walls as was similarly done by Dolan and Madsen [19]. Up to now, minimal studies have been conducted on DSN connections. Improving the understanding of DSN connections will subsequently aid in improving the understanding of midply wood shear walls and their performance under lateral loads. 2. Materials and methods 2.1. Connection fabrication A typical DSN connection was fabricated with a sheathing panel sandwiched between two exterior framing members. In this test, Oriented strand boards (OSB/3) [22] in two different thicknesses (12.5 and 15.5 mm), with an average moisture content of 12%, were used as sandwich sheathing. No. 2 spruce–pine–fir (SPF) 38  89 mm lumbers imported from Canada, with an average moisture content of 12%, were used as exterior framing members. Galvanized grooved nails with 3.6 mm diameter and 82 mm length were adopted to connect sandwich sheathing and exterior framings. The nails were manually driven by a hammer. Table 1 shows the design parameters of specimens.

2.2. Description of test specimens and variables Eight groups of DSN connection specimens were tested under monotonic loads, with ten replicates for each group. Four types of typical specimen configurations derived from practical midply wood shear walls [23] were selected for testing, as shown in Fig. 3. The first three configurations were used to simulate the sheathing displacement condition that was perpendicular to the grain of framing in a midply wood shear wall, consisting of (1) with a nail edge-distance of 10 mm for Specimens DSN12.5-PE-10 and DSN15.5-PE-10 (see Fig. 3(a)), corresponding to the nail connections in the upper and bottom edges of the sheathing panels; and (2) with a nail edge-distance of 22 mm for Specimens DSN12.5-PE-22 and DSN15.5-PE-22 (see

Nail Grain orientation

38mm

89mm

38mm

Nail

Grain orientation

Stud or plate

(a)

Stud or plate

Sheathing panel

(b)

Fig. 2. Contrast between (a) a single-shear nail connection and (b) a double-shear-nail connection.

763

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771 Table 1 Test matrix of the DSN connection specimens. Test group

t (mm)

Loading direction

l (mm)

Replicates

DSN12.5-PE-10 DSN12.5-PE-22 DSN12.5-PE-50 DSN12.5-PA-50 DSN15.5-PE-10 DSN15.5-PE-22 DSN15.5-PE-50 DSN15.5-PA-50

12.5 12.5 12.5 12.5 15.5 15.5 15.5 15.5

Perpendicular Perpendicular Perpendicular Parallel Perpendicular Perpendicular Perpendicular Parallel

10 22 50 50 10 22 50 50

10 10 10 10 10 10 10 10

Note: t = sheathing thickness; l = nail edge-distance (which means the distance from the nail to the edge of sheathing panel); perpendicular = loaded perpendicular to the grain of framing; parallel = loaded parallel to the grain of framing.

Fig. 3(b)), corresponding to the nail connections in the left and right edges of the sheathing panels; and (3) with a nail edge-distance of 50 mm for Specimens DSN12.5-PE-50 and DSN15.5-PE-50 (see Fig. 3(c)), corresponding to the inner nail connections in the sheathing panels. The fourth configuration, (4) with a nail edge-distance of 50 mm for Specimens DSN12.5-PA-50 and DSN15.5-PA-50 (see Fig. 3(d)), was used to simulate all the sheathing displacement condition that was parallel to the grain of framing in midply wood shear walls. These four configurations contained all the possible sheathing displacement conditions in a typical midply shear wall. It should be noted that 50 mm nail edge-distance is assigned to the inner nail connections loaded perpendicular to the grain of framing, as well as the nail connections loaded parallel to the grain of framing. The reason why to choose 50 mm nail edge-distance is that it is large enough to ensure realistic sheathing displacement condition for DSN connections.

Fig. 4. Nail bending test setup. 2.3. Nail bending test A preliminary study was conducted to evaluate the bending performance of the nails which were used in the DSN connection specimens. A total of 10 nails were tested following ASTM F1575-03 [24] to obtain the bending yield strength of nails. The test nails were centered across a 45-mm-span with a concentrated load applied at mid-span with a loading rate of 6 mm/min (see Fig. 4). The load and deflection were recorded at a rate of 2 Hz. The bending yield strength Fyb can be calculated as:

Loading Direction A

Loading Direction A

OSB

150 250

38

50

50

38

150 250

Section A-A

A

A

300 38

Section A-A (b)

Loading Direction A

Loading Direction A

A

38

38

Section A-A (c)

250 50

50

SPF 50

Nail Grain Orientation SPF

50

300

89

Nail

50

211 50 39

Grain Orientation

OSB

150

OSB

150 250

38

50

(a)

50

SPF Nail

89

22 67

SPF Nail

89

12 10 67

Grain Orientation

50

Grain Orientation

300

223

211

OSB

30.5

38

89 30.5 150

38

Section A-A

A (d)

Fig. 3. Test specimen configurations for DSN connections: (a) DSN12.5-PE-10 and DSN15.5-PE-10; (b) DSN12.5-PE-22 and DSN15.5-PE-22; (c) DSN12.5-PE-50 and DSN15.5PE-50; (d) DSN12.5-PA-50 and DSN15.5-PA-50.

764

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

Fig. 5. Typical load–deflection diagram from nail bending tests.

Table 2 Test results of nails. Test number

D (mm)

L (mm)

ke (N/ mm)

Pmax (N)

Pyield (N)

Fyb (MPa)

1 2 3 4 5 6 7 8 9 10 Mean COV (%)

3.57 3.65 3.71 3.58 3.59 3.68 3.56 3.63 3.74 3.47 3.62 2.8

80.79 82.2 83.68 83.90 83.15 82.3 81.54 83.50 80.42 80.72 82.22 1.5

485.6 497.9 522.0 520.2 496.0 533.4 457.0 532.4 470.6 521.7 503.7 5.0

592.9 491.7 591.7 577.5 597.4 539.3 594.8 580.8 587.2 594.8 574.8 5.6

501.9 412.8 514.7 468.2 520.8 466.0 500.0 501.2 495.5 503.8 488.5 6.2

726.1 597.2 744.7 677.3 753.5 674.2 723.4 725.2 716.9 728.8 706.7 6.2

Fig. 6. Electromechanical test machine.

Note: L = length of nail; ke = initial stiffness of nail in bending test; Pmax = maximum load; COV = coefficient of variation.

F yb ¼ M y =S

ð1Þ

where My and S are the mid-span moment and effective plastic section modulus of the nail, respectively, which can be calculated by

M y ¼ Pyield sbp =4

ð2Þ

S ¼ D3 =6

ð3Þ

where Pyield is the yield load calculated from the load–deformation curves by the 5% offset method in which yield load is defined as the intersection of load–deformation curve and a line parallel to the initial linear portion of the curve offset by 0.05 times the diameter of the nail; sbp is the spacing of bearing points which is equal to 45 mm; D is the nail diameter calculated using an average of three measured root diameters. A typical load–deformation curve is shown in Fig. 5, in which the 5% offset method is illustrated as well. The test results of nails are summarized in Table 2. The calculated bending yield strength of nails was 706.7 MPa on average, which was consistent with the requirement in NDS standard [25].

2.4. Test set-up DSN connection lateral tests were conducted by a universal electromechanical testing machine (see Fig. 6) in accordance with ASTM D1761 [26]. The connection specimens were loaded monotonically until the applied load diminished to 10% of the peak load, with a displacement rate of 3 mm/min. As shown in Fig. 7, specimens were mounted to the base of a rigid steel frame. Two steel plates were placed on the top of the two SPF framings as hold downs, and secured down to the base of the steel test frame with four 14-mm-diameter threaded rods. Loading was applied on the middle panel that was clamped tightly by the upper-jig of the machine without slippage. The load and displacement were recorded synchronously at a rate of 50 Hz.

Fig. 7. Test setup of specimens.

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

765

Fig. 8. Failure modes: (a) edge-tear of sheathing; (b) withdrawal of nail.

3. Results and discussion 3.1. Failure modes Fig. 8 shows the failure modes of the DSN connection specimens between OSB sheathing and SPF framings. The failure modes can be categorized as two primary types: (1) sheathing edge-tear (see Fig. 8(a)), which occurred in specimens with 12.5-mm-thick sheathing (DSN12.5-PE-10, DSN12.5-PE-22, DSN12.5-PE-50 and DSN12.5-PA-50) and partial specimens with 15.5-mm-thick sheathing (DSN15.5-PE-10 and DSN15.5-PE-22); and (2) nail withdrawal (see Fig. 8(b)), which occurred in the rest of the specimens. The phenomena which resulted in the corresponding failure modes of specimens can be described, respectively, as follows: (1) the nail crushed the sheathing and then tore through the edge of the sheathing; (2) the nail yields and withdraws from the framing member before the load reached the dowel bearing strength of OSB sheathing. It can be concluded that the sheathing thickness and nail edge-distance determined the failure mode of DSN connection specimens. Sheathing edge-tear was the main damage pattern in the DSN connections with thin sheathing (t = 12.5 mm) and small nail edge-distance (l = 10 and 22 mm), by contrast, nail withdrawal was commonly seen in DSN connections with thicker sheathing (t = 15.5 mm) and larger nail edge-distance (l = 50 mm). It is noteworthy that Specimen DSN15.5-PE-22 exhibited both the two failure modes, consisting of eight failures by sheathing edge-tear and two failures by nail withdrawal. This phenomenon was attributed to the variability of dowel bearing strength of OSB sheathing in testing.

3.2. Influence of sheathing thickness The load–displacement curves of specimens are shown in Fig. 9. The test results are presented along with a discussion on the influence of various variables on the lateral resistance performance of DSN connections. The initial stiffness (ke), yield load (Py), yield displacement (Dy), peak load (Ppeak), corresponding displacement of peak load (Dpeak), ultimate displacement (Du) and ductility factor (g) are calculated from the average load–displacement curve of each test group and summarized in Table 3. As illustrated in

Fig. 10, ke is defined as the slope of the secant between 10% and 40% of the peak load; Ppeak and the corresponding displacement Dpeak are obtained from the peak load point of average load–displacement curve; Du is defined as the displacement corresponding to 80% of post-peak strength. The yield load Py and the corresponding displacement Dy is calculated by the 5% offset method as was similarly done in previous nail bending test. The displacement ductility factor g is defined as the ratio of the displacement at peak load Dpeak to the yield displacement Dy [27], thus g = Dpeak/Dy. Fig. 11 shows the comparison of the average load–displacement curves of specimens with two sheathing thicknesses (t = 12.5 and 15.5 mm). Despite the variation in the edge-distance (l = 10, 22 and 50 mm for the results shown in Fig. 11(a)–(c), respectively) and the loading direction (parallel loading for the results shown in Fig. 11(d)), the influence of sheathing thickness on the ultimate strength can be evaluated. Compared to Specimens DSN12.5-PE-10 and DSN12.5-PE-22, the ultimate strength of Specimens DSN15.5-PE-10 and DSN15.5-PE-22 with 15.5-mm-thick sheathing increased by 52.4% and 22.4%, respectively. The thicker sheathing can improve the dowel bearing strength of sheathing and the edge-tear failure load can be increased, thus the ultimate strength of specimens can be enhanced. However, the ultimate strength of Specimens DSN15.5-PE-50 and DSN15.5-PA-50 were only slightly greater than those of Specimens DSN12.5-PE-50 and DSN12.5PA-50, respectively. Therefore, it can be concluded that the effects of sheathing thickness on the ultimate strength became slighter with the increase in nail edge-distance. The calculated initial stiffness of specimens are listed in Table 3. The initial stiffness of specimens with 12.5-mm-thick sheathing was observed to be closed to that of specimens with 15.5-mmthick sheathing under the same nail edge-distance and loading direction. Moreover, as shown in Fig. 11, specimens with 12.5mm-thick and 15.5-mm-thick sheathing exhibited similar initial ascending portions. It can be concluded that the sheathing thickness had little effect on the initial stiffness of DSN connections. The displacement ductility factor g is introduced to evaluate the ductility performance of specimens. As presented in Table 3, the values of g range from 1.54 to 12.84, and the displacement ductility factors of specimens with 15.5-mm-thick sheathing were generally greater than those of specimens with 12.5-mm-thick sheathing. However, Specimen DSN15.5-PE-50 showed a lower

766

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

Fig. 9. The load–displacement curves of specimens: (a) DSN12.5-PE-10; (b) DSN12.5-PE-22; (c) DSN12.5-PE-50; (d) DSN12.5-PA-50; (e) DSN15.5-PE-10; (f) DSN15.5-PE-22; (g) DSN15.5-PE-50; (h) DSN15.5-PA-50.

767

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771 Table 3 Test results. Test group

ke (kN/mm)

Py (kN)

Dy (mm)

Ppeak (kN)

DPeak (mm)

Du (mm)

g

DSN12.5-PE-10 DSN12.5-PE-22 DSN12.5-PE-50 DSN12.5-PA-50 DSN15.5-PE-10 DSN15.5-PE-22 DSN15.5-PE-50 DSN15.5-PA-50

1.16 1.05 0.84 1.87 1.02 0.81 0.75 1.72

1.69 1.85 2.01 1.89 2.05 2.10 2.19 1.94

1.6 1.9 2.6 1.2 2.2 3.4 3.1 1.3

2.02 3.08 4.00 3.40 3.08 3.77 4.14 3.45

2.6 10.8 25.8 18.7 7.4 15.3 20.5 23.6

5.7 14.9 38.4 28.6 9.8 24.7 29.7 42.6

1.54 5.79 12.84 9.89 3.60 7.29 9.35 12.16

Note: the results of the specimens were obtained from the average load–displacement curve of each test group.

displacement ductility factor than Specimen DSN12.5-PE-50. The reason is that a larger edge distance (50 mm) in Specimen DSN12.5-PE-50 can significantly improve the displacement ductility.

(Ppeak, Δ peak )

Load P

Pu=0.8Ppeak 3.3. Influence of nail edge-distance

0.4Ppeak

tan(α) = ke 0.1Ppeak

α Δ peak

Δu

Displacement (Δ) Fig. 10. Typical load–displacement curve of specimen.

As shown in Fig. 12(a) and (b), it can be found that both the ultimate strength and ductility of specimens were enhanced significantly with the increase in nail edge-distance. When the nail edge-distance was 50 mm, the ultimate strength of Specimens DSN12.5-PE-50 and DSN15.5-PE-50 were equal to 4.00 and 4.14 kN, respectively, which were 98.0% and 34.4% greater than those of Specimens DSN12.5-PE-10 and DSN15.5-PE-10 (l = 10 mm), respectively, and 29.9% and 9.8% greater than those of Specimens DSN12.5-PE-22 and DSN15.5-PE-22 (l = 22 mm), respectively. Similarly, the displacement ductility factors of Specimens DSN12.5-PE-50 and DSN15.5-PE-50 (l = 50 mm) were 12.84

Fig. 11. The influence of sheathing thickness: (a) l = 10 mm, perpendicular; (b) l = 22 mm, perpendicular; (c) l = 50 mm, perpendicular; (d) l = 50 mm, parallel.

768

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

Fig. 12. The influence of nail edge-distance: (a) t = 12.5 mm, perpendicular; (b) t = 15.5 mm, perpendicular.

Fig. 13. The influence of loading direction to the grain of framing: (a) t = 12.5 mm, l = 50 mm; (b) t = 15.5 mm, l = 50 mm. Table 4 Fitted results for the parameters in the analytical model. Test group

K0 (kN/mm)

r1

r2

P0 (kN)

Dult (mm)

Dfail (mm)

DSN12.5-PE-10 DSN12.5-PE-22 DSN12.5-PE-50 DSN12.5-PA-50 DSN15.5-PE-10 DSN15.5-PE-22 DSN15.5-PE-50 DSN15.5-PA-50

1.85 1.63 1.12 2.82 1.58 1.11 1.15 2.29

0.168 0.037 0.039 0.018 0.034 0.075 0.055 0.012

0.050 0.096 0.037 0.018 0.154 0.050 0.051 0.011

1.28 2.43 2.86 2.44 2.73 2.50 2.84 2.79

2.6 10.8 25.8 18.7 7.4 15.3 20.5 23.6

5.7 14.9 38.4 28.6 9.8 24.7 29.7 42.6

K0

Load ( P )

and 9.35, respectively, which were 8.33 and 2.6 times those of Specimens DSN12.5-PE-10 and DSN15.5-PE-10 (l = 10 mm), respectively, and 2.21 and 1.28 times those of Specimens DSN12.5-PE-22 and DSN15.5-PE-22 (l = 22 mm), respectively. However, increasing the nail edge-distance led to little change on the initial stiffness of specimens. Three load–displacement curves exhibited similar initial ascending portions as shown in Fig. 12(a) and (b), respectively. The initial stiffness of Specimens DSN12.5-PE-10, DSN12.5-PE-22 and DSN12.5-PE-50 (t = 12.5 mm, perpendicular) were close, which were equal to 1.16, 1.05 and 0.84 kN/mm, respectively. Likewise, Specimens DSN15.5-PE-10, DSN15.5-PE-22 and DSN15.5-PE-50 (t = 15.5 mm, perpendicular) behaved similar initial stiffness, which were equal to 1.02, 0.81 and 0.75 kN/mm, respectively. Hence, it can be concluded that nail edge-distance exhibited little influence on the initial stiffness of the DSN connections with the same sheathing thickness and loading direction.

(Pult , Δult) r1K 0

r2 K

0

P0

Displacement ( Δ )

Δfail

Fig. 14. Exponential load–displacement curve of sheathing connection.

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

769

Fig. 15. The comparison between the testing curve and the model curve: (a) DSN12.5-PE-10; (b) DSN12.5-PE-22; (c) DSN12.5-PE-50; (d) DSN12.5-PA-50; (e) DSN15.5-PE-10; (f) DSN15.5-PE-22; (g) DSN15.5-PE-50; (h) DSN15.5-PA-50.

770

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

3.4. Influence of loading direction Fig. 13 shows the influence of loading direction to the grain orientation of framing on the ultimate strength and initial stiffness by comparing the average load–displacement curves of specimens under the same nail edge-distance (l = 50 mm). Despite the difference in the sheathing thickness (t = 12.5 and 15.5 mm for the results in Fig. 13(a) and (b), respectively), similar findings can be observed in two figures. It was found that the specimens loaded parallel to the grain of framing behaved lower ultimate strength but higher initial stiffness than those under perpendicular loading. As presented in Table 3, the ultimate strength of Specimens DSN12.5-PE-50 and DSN15.5-PE-50 (perpendicular) were 4.00 and 4.14 kN, respectively, which were 17.6% and 20.0% greater than those of Specimens DSN12.5-PA-50 and DSN15.5-PA-50 (parallel), respectively, whereas, the initial stiffness of the former two specimens were only 44.9% and 43.6% of that of the latter two specimens. Since SPF working as the framing is an orthotropic material, it was expected that the DSN connections would show a dependency on the grain orientation of SPF framing. Concerning the ductility, the influence of the loading direction to the grain orientation of framing on the displacement ductility was not significant. Finally, it is worth noting that similar load–displacement relationship was presented by Buitelaar [21] for DSN connections with 12.5-mm-thick sheathing under parallel loading (DSN12.5-PA-50). The initial stiffness of Specimen DSN12.5-PA-50 was close to the result from Buitelaar, while a 18.1% larger ultimate strength of Specimen DSN12.5-PA-50 can be achieved. The authors attribute this non-significant difference to the usage of different nails. 3.5. Analytical model For standard timber shear walls, the load–displacement curves of sheathing connections (sheathing-framing) present highly nonlinear characteristic under monotonic loading [19]. Essentially, each connection behaves as an elastoplastic pile (steel nail) embedded in a layered nonlinear foundation (sheathing and framing) [28]. Based on this structural analogy, there are several kinds of sophisticated finite-element models developed to characterize the load–displacement behavior of the connections. As one of the most critical model, Foschi model [29] was firstly proposed as a three-parameter exponential curve model and was expanded by Dolan subsequently for the sake of a softening branch until failure afterwards. Folz and Filiatrault [28] made a further modification by terminating the softening branch with failure displacement, Dfail. It has been confirmed that this model can be used to describe the load–displacement relationship of connections exactly [16–19,30,31]. The modified Foschi model proposed by Folz and Filiatrault [28] was employed to translate the load–displacement relationship of the DSN connections in this study, described as follows:

8 K 0 D=P0 ; if D 6 Dult > < ðP0 þ r 1 K 0 DÞ½1  e P ¼ Pult þ r 2 K 0 ðD  Dult Þ; if Dult < D 6 Dfail > : 0; if D > Dfail

ð4Þ

where P is load of the connection; D is the connection displacement; K0 is the initial stiffness of the curve; r1 is the ratio of the second stiffness to the initial stiffness (0 < r1 < 1); r2 is the ratio of the degradation stiffness to the initial stiffness (r2 < 0); P0 is the second stiffness y-axis intercept; Pult is the maximum load obtained directly from the load–displacement data; Dult is the displacement corresponding to Pult; and Dfail is the failure displacement. The six parameters P0, K0, r1, r2, Dult and Dfail are physically identifiable and illustrated in Fig. 14.

The exponential function was fitted to the average experimental load–displacement curve by using a least squares regression method. The fitted values for the parameters were shown in Table 4. As shown in Fig. 15, it can be found that the analytical model curves are in good agreement with the experimental curves. Thus, the fitted model curves can exactly characterize the load– displacement behavior of DSN connections. All the parameters will be used in conjunction with future research on cyclic response of DSN connections to develop baseline date for a simplified finiteelement model of DSN connection. 4. Conclusions This paper presents an experimental study on the DSN connections generally used in midply wood shear wall system. Eight groups of DSN connections with different variables derived from practical midply wood shear walls were tested to failure in monotonic way to investigate the failure modes, the initial stiffness, the ultimate strength and the displacement ductility of the connections. The test results indicated that the responses of DSN connections were affected obviously by nail edge-distance, sheathing thickness and loading direction to the grain of framing. These influences should be considered in the designation and simulation of midply wood shear walls. The detailed conclusions are drawn as follows: (1) Edge-tear of sheathing and withdrawal of nail were two typical ultimate failure modes for DSN connections subjected to monotonic loads. Edge-tear of sheathing, shown as a brittle damage pattern, was the primary failure mechanism for the DSN connections with thinner sheathing and smaller nail edge-distance. By contrast, withdrawal of nails commonly occurred in the DSN connections with thicker sheathing and larger nail edge-distance. (2) The variation of OSB sheathing thickness from 12.5 to 15.5 mm had little effect on the initial stiffness of DSN connections, but the thicker OSB sheathing led to a maximum increase of 57.5% in ultimate strength. This increase became increasingly weaker with the increase in nail edge-distance. Moreover, the ductility of DSN connections increased with the sheathing thickness as a whole. (3) The larger nail edge-distance can result in higher ultimate strength and displacement ductility, but the changes of the initial stiffness of DSN connections were slight. (4) SPF framing is an orthotropic material so that the DSN connections show a dependency on the grain orientation of SPF framing. DSN connections loaded parallel to the grain of framing behaved lower ultimate strength than DSN connections loaded perpendicular to the grain, whereas, the initial stiffness of the former was approximately two times higher than that of the latter. Exponential curves based on the modified Foschi model can be used to evaluate the load–displacement response of DSN connections under monotonic load effectively. Further research will be conducted on DSN connections under cyclic loading. These conclusions can be used to develop baseline data for a finite-element DSN connection model to predict the lateral performance of midply wood shear walls. Acknowledgement The authors would like to acknowledge the General Project of National Nature Science Foundation of China (Grant No. 51378255) to provide funds for the first author to work on this project.

W. Zheng et al. / Construction and Building Materials 101 (2015) 761–771

References [1] E. Varoglu, S.F. Stiemer, Wood wall structure, United State Patent No. 5782054, 1998. [2] E. Karacabeyli, C. Ni, S.F. Stiemer, H. Fraser, D. Lungu, MIDPLY wall system, Final Report, Prepared for Science Council of British Columbia, Forintek Canada Corp, Canada, 2000. [3] E. Varoglu, E. Karacabeyli, S.F. Stiemer, C. Ni, Midply wood shear wall system: concept and performance in static and cyclic testing, J. Struct. Eng. 132 (9) (2006) 1417–1425, http://dx.doi.org/10.1061/(ASCE)0733-9445(2006) 132:9 (1417). [4] E. Varoglu, E. Karacabeyli, S.F. Stiemer, C. Ni, Midply wood shear wall system: performance in dynamic testing, J. Struct. Eng. 133 (7) (2007) 1035–1042, http://dx.doi.org/10.1061/(ASCE)0733-9445(2007) 133:7(1035). [5] J.R. Tissell, Wood structural panel shear walls, Res. Rep. No. 154, American Plywood Association, Tacoma, USA, 1993. [6] E. Karacabeyli, A. Ceccotti, Test results on the lateral resistance of nailed shear walls, in: Proceedings of the International Wood Engineering Conference, vol. 2, New Orleans, 1996, 28–31 October, pp. 179–186. [7] E. Karacabeyli, A. Ceccotti, Nailed wood-frame shear walls for seismic loads: test results and design considerations, in: Structural Engineering World Wide Paper T207-6, Elsevier Science, New York, 1998. [8] D.W. Dinehart, H.W.S. III, Comparison of static and dynamic response of timber shear walls, J. Struct. Eng. 124 (6) (1998) 686–695, http://dx.doi.org/10.1061/ (ASCE)0733-9445(1998) 124:6(686). [9] A. Filiatrault, Static and dynamic analysis of timber shear walls, Can. J. Civ. Eng. 17 (4) (1990) 643–651. [10] F. Lam, H.G.L. Prion, M. He, Lateral resistance of wood shear walls with large sheathing panels, J. Struct. Eng. 123 (12) (1997) 1666–1673, http://dx.doi.org/ 10.1061/(ASCE)0733-9445(1997) 123:12(1666). [11] D.W. Dinehart, H.W.S. III, Model for dynamic analysis of wood frame shear walls, J. Eng. Mech. 126 (9) (2000) 899–908, http://dx.doi.org/10.1061/(ASCE) 0733-9399(2000) 126:9(899). [12] N. Richard, L. Daudeville, H. Prion, F. Lam, Timber shear walls with large openings: experimental and numerical prediction of the structural behavior, Can. J. Civ. Eng. 29 (5) (2002) 713–724. [13] M. Li, R.O. Foschi, F. Lam, Modeling hysteretic behavior of wood shear walls with a protocol-independent nail connection algorithm, J. Struct. Eng. 138 (1) (2011) 99–108, http://dx.doi.org/10.1061/(ASCE)ST.1943-541X. 0000438. [14] Z. Li, Y. Xiao, R. Wang, G. Monti, Studies of nail connectors used in wood frame shear walls with ply-bamboo sheathing panels, J. Mater. Civ. Eng. 04014216 (2014) 1–13, http://dx.doi.org/10.1061/(ASCE)MT.1943-5533. 0001167. [15] Y. Verdret, C. Faye, S.M. Elachachi, L. Le Magorou, P. Garcia, Experimental investigation on stapled and nailed connections in light timber frame walls,

[16]

[17]

[18]

[19] [20]

[21]

[22]

[23]

[24] [25] [26] [27]

[28] [29] [30]

[31]

771

Constr. Build. Mater. 91 (2015) 260–273, http://dx.doi.org/10.1016/ j.conbuildmat.2015.05.052. J.D. Dolan, R.O. Foschi, Structural analysis model for static loads on timber shear walls, J. Struct. Eng. 117 (3) (1991) 851–861, http://dx.doi.org/10.1061/ (ASCE)0733-9445(1991) 117:3(851). J.P. Judd, Analytical modeling of wood-frame shear walls and diaphragms (Ph. D. thesis), Department of Civil and Environmental Engineering, Brigham Young University, America, 2005. J.P. Judd, F.S. Fonseca, Analytical model for sheathing-to-framing connections in wood shear walls and diaphragms, J. Struct. Eng. 131 (2) (2005) 345–352, http://dx.doi.org/10.1061/(ASCE)0733-9445(2005) 131:2(345). J.D. Dolan, B. Madsen, Monotonic and cyclic nail connection tests, Can. J. Civ. Eng. 19 (1) (1992) 415–422. U.A. Girhammar, N.I. Bovirn, B. Källsner, Characteristics of sheathing-to-timber joints in wood shear walls, in: Proceedings of the 8th World Conference on Timber Engineering (WCTE), vol. 2, Lahti, 2004, 14–17 July, pp. 165–170. M. Buitelaar, Static and dynamic testing of the midplyTM shear wall system (Ph.D. thesis), Department of Civil Engineering, The University of British Columbia, Canada, 2001. EN 300, Oriented Strand Boards (OSB)—Definitions, Classification and Specifications, European Standards, European Committee for Standardisation, Brussel, Belgium, 1997. C. Ni, M. Popovski, E. Karacabeyli, Midply wood shearwall system: concept, performance and code implementation, in: Proceedings of Meeting 40 of CIBW18, paper 40-15-3, Bled, Slovenia, August 2007. ASTM: F1575-03, Standard test method for determine bending yield moment of nails, ASTM Standards, West Conshohocken, PA, USA, 2007. American Forest and Paper Association (AF&PA), National Design Specification for Wood Construction ASD/LRFD, Washington D.C., USA, 2005. ASTM: D1761-2006, standard test methods for mechanical fasteners in wood, ASTM Standards, West Conshohocken, PA, USA, 2006. R.K.L. Su, L. Wang, Axial strengthening of preloaded rectangular concrete columns by precambered steel plates, Eng. Struct. 38 (2012) 42–52, http://dx. doi.org/10.1016/j.engstruct.2012.01.003. B. Folz, A. Filiatrault, Cyclic analysis of wood shear walls, J. Struct. Eng. 127 (4) (2001) 433–441, http://dx.doi.org/10.1061/(ASCE)0733-9445(2001) 127:4(433). R.O. Foschi, Load-slip characteristics of nails, Wood Sci. 7 (1) (1974) 69–76. Y.L. Shen, J. Schneider, S. Tesfamariam, S.F. Stiemer, Z.G. Mu, Hysteresis behavior of bracket connection in cross-laminated-timber shear walls, Constr. Build. Mater. 2013 (41) (2015) 980–991, http://dx.doi.org/10.1016/j.conbuildmat..05.089. J. Schneider, Y.L. Shen, S.F. Stiemer, S. Tesfamariam, Assessment and comparison of experimental and numerical model studies of cross-laminated timber mechanical connections under cyclic loading, Constr. Build. Mater. 77 (2015) 198–212. 10.1016/j.conbuildmat.2014.12.029.