Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Lateral resistance of traditional Japanese post-and-beam frames under monotonic and cyclic loading conditions Stefanescu, Maria 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-0151.pdf [ 3.49MB ]
Metadata
JSON: 831-1.0089420.json
JSON-LD: 831-1.0089420-ld.json
RDF/XML (Pretty): 831-1.0089420-rdf.xml
RDF/JSON: 831-1.0089420-rdf.json
Turtle: 831-1.0089420-turtle.txt
N-Triples: 831-1.0089420-rdf-ntriples.txt
Original Record: 831-1.0089420-source.json
Full Text
831-1.0089420-fulltext.txt
Citation
831-1.0089420.ris

Full Text

LATERAL RESISTANCE OF TRADITIONAL JAPANESE POST-AND-BEAM FRAMES UNDER MONOTONIC AND CYCLIC LOADING CONDITIONS by MARIA STEFANESCU B. Eng." Transilvania " University, Brasov, Romania, 1992 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in FACULTY OF GRADUATE STUDIES Department of Wood Science We accept this thesis as conforming to the required standards THE UNIVERSITY OF BRITISH COLUMBIA March, 2000 ® Maria Stefanescu, 2000 UBC Special Collections - Thesis Authorisation Form http://www.library.ubc.ca/spcoll/thesauth.html In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements f o r ah advanced degree at the Un i v e r s i t y of B r i t i s h Columbia, I agree that the Li b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I further agree that permission f o r extensive copying of t h i s thesis f o r s c h o l a r l y purposes may be granted by the head of my department or by his or her representatives. It i s understood that copying or p u b l i c a t i o n of t h i s thesis f o r f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada Date 1 of 1 3/27/00 8:59 AM ABSTRACT Full-scale tests were conducted on three types of traditional Japanese post-and-beam wall frames (2-Brace, 4-Brace and OSB sheathed frames) to determine the lateral loading resistance under monotonic and cyclic loading procedures. Twelve tests were conducted on 2.62 x 2.70 m frames, constructed with British Columbia Hem-fir timber and oriented strand board panels as sheathing (JIS - Japanese grade) provided by Ainsworth Lumber Ltd. Five specimens were tested monotonically using a loading rate of 0.13 mm/sec and seven specimens were tested cyclically using various loading protocols (UBC, UBC - modified and MOC). The ultimate loads measured in the monotonic tests were close to those measured in the cyclic tests but the corresponding displacements were much smaller for the cyclic tests in comparison with the monotonic tests. The experimental results showed the influence of the various cyclic loading procedures on the structural performance of the post-and-beam frames. The MOC-protocol induced a slightly lower capacity in comparison with the UBC-protocol. The 2-Brace and 4-Brace frames experienced higher initial stiffness and higher loads under lateral loading but they were relatively brittle systems. The sill failed in tension perpendicular to grain due to the fact that the nails from the metal plates used for the sill-post connection created a zone of concentrated tension perpendicular to grain stresses in the sill. In comparison with the 4-Brace frames the OSB frames experienced lower peak load but a substantially higher ductility. The results of this project suggest that the connectors used for these types of frames can be improved to obtain higher capacity and higher ductility. i i TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iii LIST OF TABLES v LIST OF FIGURES vi ACKNOWLEDGEMENT x CHAPTER I - INTRODUCTION 1 1.1 O V E R V I E W l 1.2 O B J E C T I V E S 6 CHAPTER I I - BACKGROUND 8 2.1 G E N E R A L 8 2.2 T E S T I N G P R O C E D U R E S 9 2.3 H I S T O R I C A L R E V I E W 16 2.4 P R O P E R T Y DEFINITIONS - E A R T H Q U A K E P E R F O R M A N C E INDICATORS 34 CHAPTER I I I - SPECIMEN CONFIGURATIONS 37 3.1 P O S T A N D B E A M F R A M E C O N F I G U R A T I O N 37 3 .2" T " J O I N T S P E C I M E N C O N F I G U R A T I O N 48 CHAPTER I V -EXPERIMENTAL SETTINGS AND TESTING PROCEDURES 49 4.1 P O S T - A N D - B E A M F R A M E S 49 4.2 " T " C O N N E C T O R 55 CHAPTER V - MONOTONIC EXPERIMENTAL RESULTS FOR POST-AND-BEAM FRAMES.58 5.1 L O A D - D I S P L A C E M E N T C U R V E S 58 i i i 5.2 FAILURE MODES 65 5.3 OVERALL FRAME RESPONSE 72 CHAPTER V I - CYCLIC EXPERIMENTAL RESULTS FOR POST-AND-BEAM FRAMES 76 6.1 LOAD - DISPLACEMENT CURVES 76 6.2 FAILURE MODES 81 6.3 OVERALL FRAME RESPONSE 86 CHAPTER V I I - MONOTONIC EXPERIMENTAL TEST RESULTS FOR " T " CONNECTION.91 CHAPTER V I I I -CONCLUSIONS AND RECOMMENDATIONS 99 8.1 CONCLUSIONS 99 8.2 RECOMMENDATIONS 102 BIBLIOGRAPHY 103 i v LIST OF TABLES TABLE 4.1 TESTING P R O C E D U R E S 53 TABLE 5.1 MONOTONIC TEST RESULTS 72 TABLE 5.2 VERTICAL D ISPLACEMENTS OF THE END STUDS AT ULTIMATE CAPACITY....74 TABLE 6.1 CYCL IC TEST RESULTS 87 TABLE 6.2 VERTICAL D ISPLACEMENTS OF THE END STUDS 89 TABLE 7.1 EXPER IMENTAL TESTS RESULTS 92 TABLE 7.2 VERTICAL D ISPLACEMENTS AND TILTING A N G L E S 97 V LIST OF FIGURES F I G U R E 1.1 J A P A N E S E C A R P E N T E R S U S I N G W O O D 1 F I G U R E 1.2 J A P A N E S E A R C H I T E C T U R E 2 F I G U R E 1.3 T R A D I T I O N A L J A P A N E S E C O N S T R U C T I O N 2 F I G U R E 1.4 P R E F A B R I C A T E D H O U S E S 3 F I G U R E 1.5 F O R C E S O N A W A L L E L E M E N T 4 F I G U R E 1.6 P O S T - A N D - B E A M F R A M E C O N F I G U R A T I O N S 6 F I G U R E 1.7 C O N N E C T O R S 7 F I G U R E 2.1 S P D - L O A D I N G S E Q U E N C E S 10 F I G U R E 2.2 Y I E L D D I S P L A C E M E N T M E T H O D F O R S P D P R O T O C O L 10 F I G U R E 2.3 C E N - L O A D I N G P R O C E D U R E S 11 F I G U R E 2.4 Y I E L D D I S P L A C E M E N T M E T H O D F O R C E N P R O T O C O L 12 F I G U R E 2.5 M O C - L O A D I N G P R O C E D U R E 13 F I G U R E 2.6 Y I E L D D I S P L A C E M E N T M E T H O D F O R M O C P R O T O C O L 14 F I G U R E 2.7 U B C L O A D I N G M E T H O D 15 F I G U R E 2.8 T E S T F R A M E 16 F I G U R E 2.9 B R A C I N G M E T H O D S 18 F I G U R E 2.10 T E S T I N G S Y S T E M - " T " - S H A P E D M O R T I S E A N D T E N O N J O I N T 19 F I G U R E 2.11 F R A M E G E O M E T R Y 22 F I G U R E 2.12 B E A M - C O L U M N J O I N T S 22 F I G U R E 2.13 O S B P A N E L S 24 F I G U R E 2.14 F O S C H I ' S M O D E L 25 F I G U R E 2.15 S H E A T H I N G - T O - F R A M I N G C O N N E C T O R E L E M E N T 26 F I G U R E 2.16 S H E A T H I N G - T O - F R A M I N G C O N N E C T O R L O A D - D E F L E C T I O N C U R V E 27 F I G U R E 2.17 F R A M E - F R A M E C O N N E C T I O N 29 F I G U R E 2.18 H Y S T E R E T I C C U R V E 30 F I G U R E 2.19 M O D E L W I T H C O N T A C T E L E M E N T S 32 vi F I G U R E 2.20 L O A D A N D B O U N D A R Y C O N D I T I O N S 33 F I G U R E 3.1 2 - B R A C E F R A M E 37 F I G U R E 3.2 P O S T - G I R D E R J O I N T ( D E T A I L A) 38 F I G U R E 3.3 P O S T - S I L L J O I N T ( D E T A I L B) 39 F I G U R E 3.4 P O S T - S I L L - B R A C E J O I N T ( D E T A I L C ) 40 F I G U R E 3.5 P O S T - G I R D E R - B R A C E C O N N E C T I O N ( D E T A I L D) 41 F I G U R E 3.6 M A B A S H I R A - G I R D E R J O I N T ( D E T A I L E ) 42 F I G U R E 3.7 M A B A S HIRA-SI L L J O I N T ( D E T A I L F) 42 F I G U R E 3.8 M A B A S H I R A - B R A C E J O I N T ( D E T A I L G ) 43 F I G U R E 3.9 4 - B R A C E F R A M E 44 F I G U R E 3.10 M A B A S H I R A - B R A C E J O I N T ( D E T A I L H) 45 F I G U R E 3.11 O S B S H E A T H E D F R A M E 46 F I G U R E 3.12 S P E C I F I C A T I O N O F T H E J O I N T S - O S B S H E A T H E D F R A M E 47 F I G U R E 3.13 " T " J O I N T C O N F I G U R A T I O N 48 F I G U R E 4.1 F R A M E T E S T S E T - U P 49 F I G U R E 4.2 T R A N S D U C E R P O S I T I O N 52 F I G U R E 4.3 L O A D I N G P R O C E D U R E - U B C - M O D I F I E D P R O T O C O L 54 F I G U R E 4.4 " T " J O I N T - T E S T S E T - U P 55 F I G U R E 4.5 T R A N S D U C E R P O S I T I O N S 56 F I G U R E 5.1 L O A D - D I S P L A C E M E N T C U R V E S 58 F I G U R E 5.2 L O A D - D I S P L A C E M E N T C U R V E S - O S B F R A M E 59 F I G U R E 5.3 Y I E L D D I S P L A C E M E N T - O S B F R A M E - M O C P R O C E D U R E 60 F I G U R E 5.4 R A C K I N G T E S T S R E S U L T S - O S B F R A M E 60 F I G U R E 5.5 L O A D - D I S P L A C E M E N T C U R V E S - 4 - B R A C E F R A M E 61 F I G U R E 5.6 Y I E L D D I S P L A C E M E N T - 4 - B R A C E F R A M E - M O C P R O C E D U R E 62 F I G U R E 5.7 R A C K I N G T E S T R E S U L T S - 2 - B R A C E F R A M E S 63 F I G U R E 5.8 Y I E L D D I S P L A C E M E N T - 2 - B R A C E F R A M E - M O C P R O C E D U R E 64 vii FIGURE 5.9 TYPICAL DEFORMATION CONFIGURATION FOR SHEATHED FRAMES 65 FIGURE 5.10 NAILS PULLED OUT 66 F IGURE 5.11 NAILS PULLED T H R O U G H 66 FIGURE 5.12 NAILS PULLED OUT 67 F IGURE 5.13 MABASHIRA-SILL FAILURE 67 FIGURE 5.14 SILL FAILURE 67 FIGURE 5.15 STUD - SILL FAILURE 68 FIGURE 5.16 MABASHIRA - SILL FAILURE 69 FIGURE 5.17 MIDDLE STUD - SILL FAILURE 69 FIGURE 5.18 STUD-SILL FAILURE 70 FIGURE 5.19 MABASHIRA-SILL FAILURE 70 FIGURE 5.20 MIDDLE STUD - SILL FAILURE 71 FIGURE 6.1 RACKING AND CYCLIC LOAD - D ISPLACEMENT C U R V E S FOR OSB FRAMES -MOC PROTOCOL 76 FIGURE 6.2 RACKING AND CYCLIC LOAD - D ISPLACEMENT C U R V E S FOR OSB FRAMES -UBC - MODIFIED PROTOCOL 77 FIGURE 6.3 RACKING AND CYCL IC LOAD - D ISPLACEMENT C U R V E S FOR 2-BRACE FRAMES - MOC PROTOCOL 78 FIGURE 6.4 RACKING AND CYCL IC LOAD - D ISPLACEMENT C U R V E S FOR 2-BRACE FRAMES - UBC - MODIFIED PROTOCOL 78 FIGURE 6.5 RACKING AND CYCL IC LOAD - D ISPLACEMENT C U R V E S FOR 4-BRACE FRAMES - MOC PROTOCOL 79 FIGURE 6.6 RACKING AND CYCL IC LOAD - D ISPLACEMENT C U R V E S FOR 4-BRACE FRAMES - UBC - MODIFIED PROTOCOL 80 F IGURE 6.7 RACK ING AND CYCL IC LOAD - D ISPLACEMENT C U R V E S FOR 4-BRACE FRAMES - MOC PROTOCOL 80 FIGURE 6.8 NAILS PULLED THROUGH 81 viii FIGURE 6.9 NAILS PULLED OUT AT " C P - T " CONNECTOR 82 FIGURE 6.10 SHEATHED FRAME DISTORTION 82 FIGURE 6.11 SILL FAILURE - TENSION PERPEND ICULAR TO GRAIN - END C O R N E R S 83 FIGURE 6.12 SILL FAILURE - TENSION PERPEND ICULAR TO GRAIN - MIDDLE STUD 83 FIGURE 6.13 SILL FAILURE - TENSION PERPEND ICULAR TO GRAIN 84 FIGURE 6.14 MABASHIRA - SILL FAILURE 84 FIGURE 6.15 MIDDLE STUD - SILL FAILURE 85 FIGURE 7.1 LOAD - D ISPLACEMENT C U R V E S OF THE " T " CONNECTION 91 FIGURE 7.2 MAXIMUM CAPACITY VS. SPECIFIC GRAVITY - "T" CONNECTION 93 FIGURE 7.3 ST IFFNESS VS. SPECIFIC GRAVITY - "T" CONNECTION 94 FIGURE 7.4 FAILURE MODE - SPLITTING OF THE SILL 95 FIGURE 7.5 FAILURE MODE - NAILS PULLED OUT 96 FIGURE 7.6 FAILURE MODE - METAL PLATE FAILURE IN SHEAR 96 FIGURE 7.7 TILTING A N G L E 98 ix ACKNOWLEDGEMENT I wish to express my deep appreciation to Drs. Frank Lam, David Barrett and Helmut Prion for their assistance and valuable suggestions throughout the project. My thanks are extended to Forest Renewal of British Columbia for funding support provided through my project and Ainsworth Lumber Co. for providing the OSB panels. I also express my thanks to Mr. Leo Oesterle, my colleagues Bingning Zhao and Dr. Liping Cai who have helped me to build the post-and-beam frames. x CHAPTER I - Introduction 1.1 Overview Ancient Japanese architecture is known for its use of wood in construction of houses, temples, or castles (Figure 1.1). Such wooden structures typically feature an unique proportion which is demonstrated through columns and beams and combinations of gentle curves. The beauty of these structures (Figure 1.2) is generally different than that of masonry structures, which features decorations on the surfaces of stones or bricks. Japanese people are fascinated by the warmth, lightweight and the life force of wood, considering that the wooden structure's components can live as parts of the building even after being cut from trees. Figure 1.1 Japanese carpenters using wood 1 Originally, the basis of traditional Japanese woodworking was the intricate system of joinery constructed without using nails or glue. Figure 1.3 shows a post-and-beam system held together by these types of joinery. This construction method is an honored technique for Japanese homebuilders. 1 Japanese Architecture (http.7Avww.kippo.or.jp/culture/build/history_e.htm). 1 Figure 1.2 Japanese architecture2 Figure 1.3 Traditional Japanese construction 3 One big problem for the traditional Japanese post and beam housing industry is the shortage of skilled carpenters and as a result the builder consumers have started searching for less expensive and simpler building techniques. Since Japan is situated in a high seismic zone the structural performance of the construction is also very important. The wood frame houses must be properly attached to the foundation and tied together structurally to resist at these loads and reduce the likelihood of damage. 2 Japanese Architecture (http://www.kippo.or jp./culture/build/history_e.htm). 3 Traditional Art of Japanese Woodworking (http://www.sonic.net/~kiarts/art.html). 2 Wood's natural flexibility is an advantage when seismic and wind loads have to be carried by the structure. However, diagonal bracing or structural panels and metal connectors is also required to minimize the risk during earthquakes. Prefabricated house technology (Figure 1.4) was first introduced into Japan in the early 1970s. The houses are built with a variety of methods and materials, including wooden post-and-beam frames, panelized 2x4 (shear walls systems), heavy steel frame and Ferro-concrete construction. Post-and-beam structure 2x4 Shear wall structure Figure 1.4 Prefabricated houses 4 Since its introduction this method of construction has made steady gain in term of market share in Japanese residential construction. The good behaviour of 2x4 shear walls structures under earthquake conditions was proven during the Kobe quake in 1995. In contrast a large number of collapses occurred in traditional post-and-beam structures. The cause of the poor performance of the traditional post-and-beam construction lies in both the design of the structures and deterioration due to rot or insect damage. The major elements, that determine the response of a post-and-beam frame during an earthquake, are the connections between sill, studs and braces. Once a connection fails the structure cannot carry any more load. 4 Big Time "Frame-Up": U. S. Wood Exports to Japan (http:/Avww.fas.usda.gov/info/agexporter.html). The major elements, that determine the response of a post-and-beam frame during an earthquake, are the connections between sill, studs and braces. Once a connection fails the structure cannot carry any more load. For a shear wall the nail-frame-sheathing joints mainly determine the structural capacity. A shear wall is more redundant than a post-and-beam frame, due to the fact that even when a frame connection fails the load can be redistributed to the other nail-frame-sheathing joints, creating a ductile system. Wood frame walls are the primary lateral resisting elements in wood frame structures. A post and beam frame or a shear wall must provide the necessary lateral strength to resist horizontal earthquake forces and transfer these forces to the next element in the load path below them, such as other frames or shear walls, floors or foundation walls. It also has to provide lateral stiffness to prevent excessive side-sway under service loads (Fig. 1.5). Side-sway Anchorage Anchorage Connection Connection Figure 1.5 Forces on a wall element 4 In North America, up to the present, research on the performance of traditional Japanese post-and-beam frames has been limited, whereas in Japan a large volume of prior research on this topic exists. Examples of some of the research published in Japan can be found in publications by Hirashima and Kamiya (1981), Imayaha and Nakamura (1990) and Komatsu (1990). After the collapse of a large number of traditional Japanese post-and-beam frames during the Kobe earthquake, confidence associated with the seismic performance of these types of frames has been significantly reduced. Consequently, there is a strong need for more test data on the full-scale seismic response of these wood frame structures to improve their structural performance. 5 1.2 Objectives The objective of this study was to evaluate the performance of three types of post-and-beam frames (Fig. 1.6) with metal "CP.T", "VP" and "BP" connectors (Fig 1.7), using British Columbia Hem-fir timber and built with the traditional mortise and tenon Japanese carpentry techniques. a) 2-Brace frame b) 4-Brace frame c) OSB Sheathed frame Figure 1.6 Post-and-beam frame configurations Figure 1.7 Connectors This objective was achieved step by step as follows: • Build a test facility to perform the static and cyclic tests • Experimentally investigate the performance of a "CP.T" connector, which joins the sill and the stud • Experimentally investigate the structural performance of the three types of post-and-beam frame (a) Frame with 2-braces, b) Frame with 4-braces and c) Frame with OSB-sheathing (Fig. 1.6) (under static and various cyclic protocols) • Compare the results from the static and cyclic tests 7 CHAPTER II - Background 2.1 General Post-and-beam structures often span large open spaces, presenting the beauty of exposed connections and the natural texture of wooden' posts and beams creating warmth and charm. A traditional Japanese post-and-beam frame is composed of vertical posts, spaced at regular intervals (455 mm), horizontal bottom and top beams and braces. Japanese houses are based entirely on traditional fixed modules of 910 mm (2X455mm). Metal connectors are commonly used to fasten the components together. Increasingly, sheathing panels may replace the diagonal braces in modern post and beam construction. The sizes of sheathing panels are mostly 915 by 1830 mm (3 by 6 feet) or 915 by 2745 mm (3 by 9 feet) in Japan versus 1220 by 2440 mm (4 by 8 feet) in the U.S. and Canada. The fasteners are nails of different lengths and diameters, depending on the elements to be connected. Like shear walls, which are vertical elements in the lateral force resisting system and made of wood frames sheathed with OSB or plywood panels using nails, the post-and-beam frames transmit lateral forces from the diaphragm above to the diaphragm below or to the foundation. A vast amount of literature is available on the structural performance of wood frames and connectors but their behaviour during earthquakes and high wind is still not completely understood. Experimental studies are still basis for the understanding their behaviour under loading. This data is often used for the development of analytical models and the design code. This chapter summarizes important findings on wood frame and connection behavior and a discussion of different testing procedures. 8 2.2 Testing procedures For post and beam frames, a specific testing method to evaluate their performance under lateral loading is not available. Due to the similarity in geometry and function between a post and beam frame with or without sheathing and a shear wall, it is believed that these types of post and beam frames can be tested and analyzed using the same methods as for shear walls. A brief description of defferent test protocols is given below. 2.2.1 ASTM-Protocol The most common testing procedures to evaluate the racking performance of shear walls are found in standard ASTM E-72 and ASTM E-564. The shear strength and stiffness of a framed wall are determined by racking a wall specimen from a rectangle into a parallelogram shape. This is realized by anchoring the bottom edge of the wall to the floor and applying an in-plane force laterally parallel to the floor at the top of the wall. No vertical loads are applied. Griffiths (1984) and Dolan (1991) made a comparison between these test procedures. They found that the ASTM E-72 test overestimated the test specimen stiffness and strength due to the addition of two tie-down roads at the ends of the specimen. They concluded that the ASTM E-564 test provided too optimistic results on the behaviour of timber shear walls. 2.2.2 SPD-Protocol (1993) The SPD-Protocol, also known as the "Sequential Phased Displacement Procedure", consists of triangular reversed cyclic loading at incrementally increasing and decreasing displacement levels with a maximum frequency of 1.0 Hz. The incremental displacement is related to the first major event (FME) determined from the monotonic test. The FME is defined as the displacement at which the structure starts to deform inelastically (yield displacement). The process of loading is repeated until failure (Fig 2.1). 9 ~ 400% + UJ FME -300%+ -400%+ Number of cycle groups until failure Time Figure 2.1 SPD-loading sequences Each type of protocol has its own method for determining the yield displacement. According to this protocol the yield displacement is given by the intersection between the line that joins the origin and the point on the monotonic curve at 0.4Pmax and a horizontal line at a load level equal or greater than 0.8Pmax. The ultimate displacement is found by equaling the triangular areas Aland A2 (Fig 2.2). P * 0.4P, Pyield max max Ayield Al Figure 2.2 Yield displacement method for SPD protocol 10 2.2.3 CEN Protocol (1995) The CEN Protocol is a European standard originally developed for testing mechanical fasteners. It has two procedures: a short procedure (Fig. 2.3 (A)) which consists of three cycles with an amplitude equal to the product of the ductility and yield slip from the monotonic test followed by a pushover test until failure, and a long procedure (Fig. 2.3 (B)), which consists of two single cycles followed by groups of three cycles with an amplitude that represent 25%, 50%, 75%, 100%, 200%... of the yield slip until failure. A) CEN - Short Procedure Time B) CEN - Long Procedure Figure 2.3 CEN - Loading procedures 11 The yield deformation is defined by the intersection of two lines (Fig. 2.4). The first line joins the points corresponding to O.IPmax and 0.4Pmax on the load deformation curve. The second line is the tangent to the load-slip curve having an inclination of 1/6 of the first line. P + Figure 2.4 Yield displacement method for CEN protocol 12 2.2.4 MOC Protocol The MOC (Ministry of Construction) Protocol is a Japanese standard for testing wood frame constructions. It consists of two single reversed triangular cycles with the amplitudes equal to % yield slip and Vz yield slip, respectively, followed by groups of three cycles with amplitudes equal to: 3A, 2, 4, 6, 8, 12, ...of the yield displacement until failure (Fig 2.5). Each cycle should take more than one second. • Time Figure 2.5 MOC - loading procedure The method for determining the yield displacement from the monotonic test is described below and presented in Fig. 2.6. Procedure: - Draw a line through the points on the load-deformation curve, which correspond to the load level of O.IPmax and 0.4Pmax (Line 1). - Draw a line through the points on the load-deformation curve, which correspond to the load level of 0.4Pmax and 0.9Pmax (Line 2). - Draw a tangent line to the load-deformation curve that is parallel to Line 2. 13 - The displacement that corresponds to the point of intersection of Line 1 and Line 3 represent the yield deformation. - The gradient (K) of Line 4, which connects the origin and the yield point is the stiffness of the elasto-plactic model - The deformation corresponding to 0.8Pmax, after the peak has been reached, is the ultimate displacement. Figure 2.6 Yield displacement method for MOC protocol 14 2.2.5 UBC Protocol The UBC protocol was developed by He et al. (1999), who found that the commonly used long protocols caused low cyclic fatigue fractures in the metal connectors (nails), which resulted in a premature failure of the shear walls tested. Since such failure seldom occurred under real earthquake conditions, where typically only a few large amplitude cycles are applied to a structure, a shorter protocol with large amplitude cycles was deemed more appropriate. The loading procedure consists of two groups of three reversed triangular cycles with the amplitudes equal to the displacements at 50% and 80% of the maximum load followed by a single reversed triangular cycle with an amplitude equal to the displacement at 50% of the peak load and finally a push over test (as a racking test) until failure (Fig. 2.7). 8 at 0.5Pmax Time 5 at 0.5Pmax 8 at 0.8Pmax Figure 2.7 UBC loading method 15 2.3 Historical review 2.3.1 Experimental studies In 1981, Hirashima, Kanaya, Hatayama and Kamiya conducted racking tests for seventeen-different types of frames. The geometry of the frames is shown in Fig. 2.8. 3020 ?-Nqnr a Figure 2.8 Test frame The braces were connected to the sill and girder with different connection methods as shown in Figure 2.9. 16 17 Figure 2.9 Bracing methods After testing the frames they found out that the maximum load recorded was between 7.28 kN and 14.10 kN. The lateral deformation corresponding to the maximum load was between 47.65mm and 189.94mm. The tensile brace joined with only two or three toenails (Fig. 2.9 - a), b), c) and d)) carried little or no load when the horizontal load was maximum. Comparing the frames with halving joints and reinforced braced joints (Fig. 2.9 -e), f), g), h), i), j), k), I), m), n) and o)) the allowable shear load and the maximum load were between 1.3 and 1.6 times greater than those without reinforcements. 18 In 1990 Imayaha, Ikehada, Sugiura and Nakamura investigated the effect of the tenon length (I) on the bending strength of a T-shaped mortise and tenon joint (Fig. 2.10). The tenon depth (h) was held constant. The length/depth ratio (l/h) varied between 0.3 and 2.1. The essence of the results was that within the interval 0.3<l/h<1.0 the bending strength of the mortise and tenon joint increased linearly as l/h increased. Within the range of 1.1<l/h<2.1 the bending strength of the joint gradually approached a constant value; this constant value was exactly the same as the modulus of rupture in bending of the material. The bending moment (M) was calculated using the following relation: M = (l/6)*k*ab»b*h*l Where k is the ratio of the maximum bending strength (o-max) to the modulus of rupture (o-b). In this case it is possible to estimate the tenon dimensions according to the expected bending moment acting on the joint. Figure 2.10 Testing system - "T"-shaped mortise and tenon joint 19 Nails are used in all types of wood construction. In fact, research and practice have shown that nails are excellent connectors for wood structures with light to moderate loads. They are used in locations where a ductile connection is required. It is well known that the behaviour of metal fasteners determines the performance of the wood frames subjected to monotonic or cyclic loading. The modes of failure for a nailed joint have been modeled by Johansen and modified by Larsen (1977). The capacity of a nail connection is theoretically based on the yield moment of the nail and the embedding strength of the wood. The withdrawal resistance of a nail is a function of the length of penetration into the main member, the diameter of the nail and the specific gravity of the wood. Komatsu (1990) summarized the results of single shear tests of nailed timber joints with steel side plates. He pointed out that there was no clear linear portion in the load-slip relationship. The correlation between the initial slip modulus K and the timber density tends to be greater as the nail diameter becomes larger and there was a positive correlation between the maximum nail load and the timber density. Many factors influence wood joint performance, such as wood density and mechanical properties, grain angle, connector orientation and its features. Melo et al. (1995) analyzed the influence of wood density on metal-plate connector mechanical behaviour under cyclic loading. Static and cyclic load tensile tests were carried out on joints assembled at a 90-degree angle. The joints were connected with metal plates for two categories of density ranging from 0.34 to 0.38 and from 0.45 to 0.56 (12% moisture content) of Norway spruce. They found out that high-density wood joints were not much more rigid but could bear an average of 30% greater load than low-density joints. 20 Seo et al. (1999) investigated the behaviour of wooden frames with tenon joints under lateral load tests that constitute ancient Korean commoners' houses. Two types of full-scale wooden frames (Fig. 2.11) were fabricated using pine lumber. The joints between the column and top beam, the cross-beam and column, the high column and beams are presented in Fig. 2.12. s e a m c Column C r o s s beams / Dooj^^Po^si^ 7 2 0 2 4 0 0 D LT) CO LO CU o CD in A) Ordinary frame - Type 1 21 High Colunn ' / B e a n Co lunn 5 L J: C D CO 0> C r o s s beans o r -CO CD C D UO 1200 2400 B) High column frame - Type 2 Figure 2.11 Frame geometry 22 Static and cyclic lateral load tests were performed. The test results show a significant nonlinearity and inelasticity in the load-displacement curve. The ultimate load of the Type 1 frame was 1,184 N at a displacement of 200 mm. The ultimate load capacity of the Type 2 frame was 4,160 N at a displacement of 200 mm. The failure modes of joints were either shear or bending failure at the mortise branches of the tenon. Normally the strength of wood frames sheathed with wood structural panels (plywood or oriented strands board) comes from the sheathing fasteners. This phenomenon can be explained by the fact that under lateral loading the wood frame will deform as a parallelogram and due to the rigidity of the sheathing panel the fasteners will deform and carry the load. This behaviour is achieved only if the nails will not pull through the thickness of the sheathing panel or tear any of the edges. Therefore the two main functions of a structural panel are in-plane strength, rigidity and the nail holding characteristics. Oriented Strand Board (OSB) panels are commonly used for sheathing timber frames and they have been replacing plywood in many applications. In 1996 Canadian plywood accounted for 31 per cent of total structural panel demand in Canada, compared to 69 per cent for OSB. In 2012, on information provided by Resource Information System Inc. (RISI) the ratios will be 12 per cent for plywood and 88 per cent for OSB. This big gain for OSB is related to the lower cost and raw material availability. OSB (Fig. 2.13) is a structural panel product made from wood strands bonded with adhesives such as phenol-formaldehyde or isocyanate, under intense heat and pressure. The OSB industry uses aspen and poplar in the northern part of North America and southern yellow pine in the south. The strength of the board comes from the size of the strands, overlapping area and strand orientation. 23 Figure 2.13 OSB Panels 5 Panels are normally cut to a size of 1220 x 2440 mm ( 4 x 8 feet), while the thickness varies from 6mm to 38mm. They are also manufactured in other sizes for export markets, e.g. 915 x 1830 mm (3x6 feet) or 915 x 2745 mm (3x9 feet) OSB panels with a basic thickness of 9.375mm (3/8 inch) that are certified by the Japanese Agricultural Standard (JAS). 5 Oriented Strand Board (Glenn Bailey) (http://www.anu.edu.au./Forestry/wood/osb/Product.html). 24 2.3.2 Analytical Modeling Vast amounts of information exist with regard to modeling and design of framed walls under racking and cyclic displacements. This section will present a review of some of the techniques for wood frame analysis. 2.3.2.1 Foschi's model Foschi (1977) developed a finite element analysis program for the monotonic analysis of wood diaphragms. The model considers the combination of the sheathing, the framing members, the connection between frame members, and the sheathing-frame connections. Load-deformation relationship of the connector was assumed non-linear and is represented by an exponential function (Fig. 2.14). P - (mo + mA)(l - exp(/cA/mo)) Figure 2.14 Foschi's model where: A = connector displacement, mo = Y-intercept of the mi line, k = slope of the curve at the origin, and mi = slope of the curve at the maximum displacement 25 It was considered that the non-linearity is due to the inelastic deformation behaviour of the nails between the sheathing and the frame. All elements are considered as linear elastic. 2.3.2.2 SHWALL Program Dolan (1989) extended Foschi's program (1977) by including the bearing effect between adjacent sheathing panels, the out-of-plane behaviour of sheathing panels, and the prediction of the ultimate load capacity of the shear wall. The framing element is a typical finite element used in many plane frame analysis programs with two nodes. Each node is assigned three degrees of freedom, which represent the displacements in the X- and Y-directions, and a rotation. The sheathing element is a four-noded plate element and the sheathing material is assumed to be elastic and orthotropic. The sheathing to framing connector is modeled using three independent nonlinear springs. A frame element is connected to a sheathing element through the nail connection. An example of a nail connection is shown in Fig. 2.15. The connector is located at point P on the sheathing element, and P' on the framing element. Initially, P and P' are coincident, then, when a load is applied, point P moves relatively to point P'. The relative in-plane movement represents the slip of the connection. 4 3 F Y X a) Original geometry b) Relative deflection Figure 2.15 Sheathing-to-Framing Connector Element 2 6 The load-deflection curve used in the analysis program is shown in Fig.2.16. Figure 2.16 Sheathing-to-Framing Connector Load-Deflection Curve The relationship between in-plane components of the connector force, F, and the deflection can be expressed as: |F| = (Po + Ki\ A|)[l - exp(-^i| A| / Po)] for | A| < A max \F\ = (Po + ^2Amax)[l - exp(-.Ki A max/Po)] - Ki(\A\ - Amax) for |A| )>- Amax The relationship between the out-of-plane component of the connector force and the deflection can be expressed as: \Fz\ = (Poz + AT2z|A|)[l - exp(-Ku\Aw\/Poz)] for 0 < Aw < Aw max \Fz\ = (Poz + KlzA max)[l - eXp(-AJzA max/ P02)] ~ K3z(\ A| - A max) fOr AW y Aw max where: Aw = connector deflection in Z-direction 27 The sheathing-to-sheathing bearing element was modeled as a bilinear spring with a very low stiffness in tension and a high stiffness in compression. The connector only prevents the sheathing elements from overlapping. Static uni-directional shear wall tests, performed at University of British Columbia (Dolan 1989) were used to verify the accuracy of the program. The results showed a good prediction of both the ultimate load capacity and the load-deformation characteristics when compared to shear wall test results. Dolan and Madsen (1991) investigated the behaviour of timber shear walls subjected to racking and slow cyclic test by experimental and analytical study. The results showed that the nail connections and the mechanical properties are the predominant factors governing the wall load-displacement characteristics and it was confirmed that the hysteresis from cyclic test is contained within the envelope defined by the monotonic load-displacement curve. 28 2.3.2.3 DAP Program Foschi (1993) modified the finite element program developed in 1977 into a new program called DAP (Shear Wall/Diaphragm Analysis Program) to analyze the behaviour of wood based shear walls. The sheathing panel was modeled as a 2D twelve-node cubic isoparametric element. The frame consists of one-dimensional beam elements with three degrees of freedom at each node and two types of nail connectors were modeled as nonlinear spring elements. The sheathing-frame connector was modeled as in Fig. 2.17. Y Figure 2.17 Frame-Frame Connection This model can reproduce the nonlinear behaviour observed in the experimental tests. 29 2.3.2.4 WALSEIZ Program White and Dolan (1995) developed the Walseiz program, which is a modification of the program developed by Dolan (1989). The modifications include: a reduced number of degree of freedom in the plate, sheathing-bearing connector, and sheathing-to-framing connectors' elements to reduce the analysis time and include the capability of calculating forces and stresses. Assumptions in the model include: hinged connections between the top and bottom plates and the studs; framing elements are homogenous and isotropic; the out-of-plane buckling is excluded and the sheathing elements are orthotropic. The framing element is a two-dimensional, linearly elastic, six-degree-of-freedom beam element. The sheathing element is an eight-degree-of-freedom, orthotropic, linear-elastic, rectangular, plane stress element with translations in the global X- and Y-directions at each of the four corner nodes. The sheathing-to-framing connector element consists of two independent nonlinear springs. The curve for monotonic loading is governed by the equations developed by Foschi (1977) and Dolan (1989). The hysteretic curve, used for cyclic loads, is broken into four sections (Fig.2.18) \ \ Deflection Pt2=(u2,F2) Hysteresis Figure 2.18 Hysteretic curve 30 The equations used for each section are: Fiu = - P i + K*A + [exp(aiA) -1] ax = ln(Fi + Pi - Kwi +1) ui Fiu = - P i + K*A - [exp(a2| A|) -1] ai = ln(-P\-Fi + K4U2 + l) \U2\ Fiu = Pi + ^ 4 A - [exp(a3| A|) -1] l n ( P i - F 2 + i«:4M2 + l) Fiu = Pi + KAA + [exp(a4A) -1], l n ( P i - P i - ^ 4 M i + l) ui where: A = connector displacement; P1 = load at zero displacement; K4 = slope at zero deflection; u1 and u2 = maximum and minimum deflections, respectively; F1 and F2 = maximum and minimum loads, respectively. The sheathing-bearing element is a bilinear spring with a high modulus of elasticity in compression and a low modulus in tension. The high modulus in compression accounts for the effect of adjacent sheathing elements bearing on each other and the low modulus in tension allows for free movement of the sheathing panels when they are not bearing on one another. The program predicted the maximum strength of a wall subjected to monotonic loading to within 2%, and correlated with dynamic tests results with correlation coefficients of 0.835 and 0.846 for displacements. 31 2.3.2.5 Mihailescu and Nicholls Finite Element Model Mihailescu and Nicholls (1999) presented a finite element solution for the analysis of the performance of the square and round mortise and tenon joint. The model was developed in terms of parametric 3D modeling, using the features provided by ANSYS™ Parametric Design Language and includes: geometrical parameters, such as the dimensions of the wood joints; physical parameters, such as mechanical parameters of the wood; boundary conditions and external loads; and internal parameters such as the types and numbers of finite elements. Orthotropic material properties of wood were modeled by using SOLID73 a 3D finite element. The interaction between mortise and tenon was represented by a 3D contact element (Contact 49 in ANSYS™). This type of element has five nodes with three degree of freedom at each node: translation in the x, y and z direction (Fig 2.19). This element can be used to represent the contact and sliding between two surfaces in three dimensions. (Contact 49) Figure 2.19 Model with contact elements 32 The stresses in the connections were studied using the finite element model of the mortise and tenon joints with boundary conditions shown in Fig. 2.20. Figure 2.20 Load and boundary conditions The model was tested experimentally. The experimental results and the solutions from the finite element model were compared in terms of the resultant deflection under certain loads and boundary conditions. The errors were below 10%. 33 2.4 Property Definitions - Earthquake performance indicators The forces that a structure sustains during an earthquake result directly from the distortions induced in the structure by the motion of the ground. The earthquake loads are inertia forces related to the mass, characteristics of the structures, stiffness and energy-absorption (e.g., hysteretic damping and friction). It is uneconomical or impractical to design buildings to resist forces resulting from very severe earthquakes within the elastic range of stress. Generally buildings are designed to resist lower levels of forces by using a measure of ductility and the energy-absorbing capability of the structure to limit the life threatening damage. Deformation at failure shows the ability of the system to deform and resist loads but the yield deformation depends on the elastic stiffness and capacity of the structure. The ductility is a parameter, which should be considered with other performance indicators because a system with low elastic stiffness, high capacity and large failure displacement may have a low ductility due to the large yield displacement. 2.4.1 Energy dissipation - damping For cyclic tests the energy dissipation of a wooden wall incorporates yielding of the nails, internal friction and non-recoverable deformations. The amount of energy dissipated is represented by evaluating the area within the hysteresis loop per cycle. 2.4.2 Ductility (D) For practical design purposes, ductility is defined as the capacity of the building materials, systems, or structures to absorb energy by deforming in the inelastic range. The ductility ratio is the displacement of failure divided by the yield displacement. ^failure Ayield 34 2.4.3 Redundancy Redundancy is another important characteristic for earthquake resistant design. When the primary element or system yields or fails, the lateral forces can be redistributed to the secondary elements or systems to prevent progressive failure. 2.4.4 Capacity This property can be determined from the load-displacement curve for the monotonic tests and for the structures tested cyclically by analyzing the hysteresis loops. The capacity represents the highest load, F p e a k, resisted by the structure. 2.4.5 Failure The failure is considered to occur when the post a load reached 80% of the peak load. 2.4.6 Elastic stiffness The elastic stiffness, K, is the gradient of the line that connects the origin and the yield point on the load displacement curve. 2.4.7 Ultimate shear strength (KN/m) The ultimate shear strength represents the ratio between the maximum carried load and the wall length. Pmax where: P m a x = Maximum load L = Length of the frame measured parallel to the loading direction 35 2.4.8 Shear stiffness (MN/m) Pmax»i7 G = 2 Ayield • L where: H = Height of the frame Ayieid = Yield displacement 36 CHAPTER III - Specimen configurations In total of twelve frames, four of Type A (2-Brace frame), four of Type B (4-Brace frame) and four of Type C (OSB Sheathed frame) were manufactured using Canadian Hemlock lumber with the moisture content between 10% and 15%. The lumber was graded to meet the requirements of the JAS Type A No. 1 and No. 2 grade rule. The configurations of the specimens are presented below. All measurements are in millimeters 3.1 Post and beam frame configuration a) Type A - 2-Brace frame (Fig. 3.1) Figure 3.1 2-Brace frame 37 The post - girder connection (Detail A) and the post - sill connection (Detail B) are realized through a mortise and tenon joint and a " CP-T " metal plate using ZN65 nails (L = 65mm, D = 3.33mm, JIS) as shown in Figures 3.2 and 3.3. Figure 3.2 Post - Girder joint (Detail A) 38 Figure 3.3 Post - Sill joint (Detail B) 39 The post - sill - brace assembly (Fig.3.4 - Detail C) is realized through a mortise and tenon joint, a " VP " metal plate fastened to the wood elements by using ZN90 nails (L = 90mm, D = 4.11mm, JIS) and a " BP " metal plate using ZN65 nails and a M12 bolt. Nail: Type ZN90(JIS) Figure 3.4 Post - sill - brace joint (Detail C) 40 Detail D (Fig. 3.5) is similar to detail C and it represents the assembly between the middle post, the girder and the two braces. Figure 3.5 Post - girder - brace connection (Detail D) 41 Detail E, Detail F and Detail G are shown in Figures 3.6, 3.7 and 3.8, respectively. Girder 105x135 ,30, Tenon Stud 105x30 (Mabashira). 30 Figure 3.6 Mabashira-Girder joint (Detail E) Stud 105x30 (Mabashira). Nail: Type ZN75 (JIS)_ Sill 105x105 ft ft # II Figure 3.7 Mabashira-Sill joint (Detail F) 42 ure 3.8 Mabashira-Brace joint (Detail G) 43 b) Type B - 4-Brace frame (Fig. 3.9) j < M 1 T I D i s s Figure 3.9 4-Brace frame The connections in the 4-Brace frame are similar to those in the 2-Brace frame; the main difference is in the joint between Mabashira and the two braces (Detail H-Fig. 3.10). 44 Figure 3.10 Mabashira-Brace joint (Detail H) 45 c) Type C - OSB Sheathed frame (OSB frames) (Fig. 3.11) 400 Girder 105x 135 ^ Nail: Type N50 (JIS A5508) 400 Post 105x105 OSB Thickness 9.0mm Mabashira 105x30-Sill 105 x 105 150 -455-f -455--1820-150 -455 1 455 135 O o 0^ o r-O J oj 105 -105 Figure 3.11 OSB Sheathed frame Oriented Strand Board panels provided by AINSWORTH LUMBER CO. LTD, BC, Canada, JAS - grade, 910 x 2580mm, 9mm in thickness were used for sheathing the frames. All the sheathing panels were oriented with the longitudinal axis of the panel parallel to the height of the frame. The Post - Sill and Post - Girder connections are " CP -T " joints (Fig. 3.2 - Detail A) and the specification of the other connections are presented in Fig. 3.12. 46 400 Girder 105 x 135 Post 1 0 5 x 1 0 5 -Mabashira 1 0 5 x 3 0 -Sill 105x105 Ax 400 -455 } 455 455 +• 455--1820-135 105 O o \D o OJ OJ Figure 3.12 Specification of the joints - OSB Sheathed frame Detail E and detail F are similar to those from Figures 3.6 and 3.7. The sheathing panels were connected to the frame elements using N50 (L=50mm, JIS A5508). The nail spacing was 150mm for both the perimeter and the interior attachment. 47 3.2 " T " Joint Specimen Configuration Due to the fact that the weakest point of the wooden frames was expected to be the bottom post-to-sill connection, a separate series of tests was conducted to determine its behaviour under tension. Eight" T " connections (Fig. 3.13), with the configuration similar to the " CP-T " joint from the 2-Brace and OSB frames, were manufactured and monotonically tested. This type of connection has been chosen because it is the simplest one, without implicating braces. Canadian Hemlock lumber (A1 - grade) with a moisture content of 12%-15% was used for the horizontal and vertical elements which were connected together by mortise and tenon joints," CP-T " metal plates and ZN65 nails. —105— | p ! 0 5 - | 6 0 0 25 - H •600-o - 3 0 Figure 3.13 T" Joint Configuration 48 CHAPTER I V -Experimental settings and testing procedures 4.1 Post-and-Beam frames 4.1.1 Test set - up The test set-up is depicted in Fig. 4.1. Figure 4.1 Frame Test Set-up 4 9 The tests were performed in the Structures Laboratory, Civil Engineering Department, at the University of British Columbia. A special steel reaction frame (1) (Fig. 4.1) was used for testing the specimens. The frames were positioned between the top steel beam (3) and anchored to the bottom channel (2) with four 15.9mm (cf>=5/8-inch) bolts. Steel plate washers, 53mm x 53mm square, 5mm thick distributed the load from the bolt head and prevented crushing of the wood. The top steel beam (3) was connected to a hydraulic actuator (4) with a displacement range of ± 304.8mm (±12 inches) and capacity of 222KN. The actuator (4), through the top steel beam, induced the movement of the frame (7). In order to prevent the out of plane buckling of the wood frame during the racking and cyclic tests, two side steel beams (5) with two pairs of rollers were used. During the experimental tests the frame ran between the two pairs of rollers. Two hydraulic jacks were mounted on the bottom of the reaction frame and attached through the top steel beam (3) by 12.7mm § steel tie rods (6). The jacks provided a vertical load of 2.275 KN/m which represents the gravity load over an area of 1.82 m x 1.82 m (where 1.82 m is the length of the frame from an end stud to the other end stud). This load is the equivalent of the weight of the second floor of two-storey building supported by four walls of 1.82 m in length. A MTS Micro-Controller (458.10) and Material Testing Function Generator were used for developing the required lateral loading. The hydraulic actuator contained a transducer, which measured the horizontal displacement, and a load cell that recorded the load resisted by the wood frame. A 386/25 computer and LabTech Notebook data acquisition system were used to store the data. Figure 4.2, A), B), and C), depict the location of another three transducers (TR1, TR2 and TR3) that were attached to the frame to record the compression and uplift displacements of the end posts and middle post relative to the wood frame sill. 50 A) T R 1 - Position B) T R 2 - Position 5 1 52 4.1.2 Testing procedures Twelve frames were tested. Each type was tested monotonically with one or two replications per test and cyclically as depicted in Table 4.1. TABLE 4.1 TESTING PROCEDURES Testing procedures Cyclic tests Frame type Monotonic test MOC Protocol UBC-modified Protocol UBC Protocol 2-Brace frame 2 1 1 -4-Brace frame 1 1 1 1 OSB frame 2 1 1 -The loading rate for the monotonic test was 0.13 mm/sec, according to the ASTM Standard E564-73. To see the influence of the loading procedure on the structural performance of a post and beam frame, three types of cyclic protocols were used (Table 4.1). According to He et al. (1999) high amplitude short sequence cyclic tests might be a more realistic representation of earthquake loading in determining the resistance of wooden walls as appears to long sequence cyclic test protocols that often cause slow cyclic fatigue in metal fasteners. As this phenomenon could be expected in the post and beam frames the UBC protocol was modified by using a more severe loading procedure as follows: 53 UBC - modified protocol The loading procedure consists in two groups of three reversed triangular cycles with the amplitudes equal to 50% and 80% of the displacement corresponding to the peak load from the racking test, followed by a single reversed triangular cycle with an amplitude equal to the amplitude from the first group of cycles, and finally a push over test until failure (Fig. 4.3). 0.88pmax Figure 4.3 Loading procedure - UBC - modified protocol The loading procedures for MOC protocol and UBC protocol are explained in detail in Chapter II, section 2.2.4 and 2.2.5. The frequency was 0.03 Hz for MOC and UBC protocols and 0.05 Hz for UBC -modified protocol. 54 4.2 " T " Connector 4.2.1 Test set-up and testing method Eight " T " Joints were tested using a MTS 800-20 Universal Testing Machine. The setup is shown in Figure 4.4. .105. ~i———r .105. 600 400 420 300 =4^ 105. 280 105 10 105 200 105. 10 Figure 4.4 " T " Joint - TEST SET-UP The " T " connection was fixed on the machine table using four wood blocks, two angle steel plates and six 15.9mm (<j>= 5/8- inch) steel bolts (Fig. 4.4 and Fig 4.5) The upper element (post) was connected to the MTS loading unit through a cross steel pin with a diameter equal to 30 mm. The distance between the two bolts that connected the bottom element (sill) to the angle plates was equal to 400 mm. 55 The specimen was loaded in tension under displacement control with a loading rate of 0.13mm/sec. A 486 Computer and Data Acquisition System were used for recording the data. Locations of the six transducers, Tr1, Tr2, Tr3, Tr4 and Tr6, used to measure the displacement of the specimen during testing are shown in Figure 4.5. II 1 Figure 4.5 Transducer positions Tr1 and Tr2 measured the uplift (left and right) of the upper element (post) relative to the bottom element (sill) of the connection. Tr3 and Tr4 measured the uplift (front and back) of the upper element (post) relative to the machine table. Tr5 and Tr6 measured the uplift (front and back) of the bottom element (sill) relative to the machine table. 5 6 Two sensors attached to the hydraulic actuator that controlled the movement of the machine table measured the displacement of the load cell and the load applied to the specimen. Data from the 6 external transducers and two internal hydraulic actuator sensors were collected at a rate of 10 Hz. 57 CHAPTER V - Monotonic experimental results for Post-and-Beam frames 5.1 Load - displacement curves The load-displacement curves for the monotonic racking tests are shown in Figure 5.1. Each type of frame was tested at least once. Due to the fact that some of the data from the first tests was corrupted by excessive electronic noise and considering that the yield displacement from the monotonic test was necessary for establishing the cyclic schedules, a second test was required for some types of frames. 18 16 14 12 10 Z JC Ti 8 _ i 6 4 2 ri : x > o ( ^ ] n 0 20 40 60 80 100 120 140 Displacement (mm) Figure 5.1 Load - displacement curves In order to determine the yield deformation according to the MOC procedures (Chapter I I , section 2.2.4), the data for the first monotonic tests was smoothed using Foschi's exponential function (Chapter I I , section 2.3.2.1). 58 P = (mo + m-iA)(1 - exp(-kA/mo)) The three parameters, mo, mi and k were determined such that a good fit could be obtained between the experimental and model static load displacement curve. 5.1.1 OSB sheathed frames The load-displacement expression is presented below: P = (9.5 + 0.01 A)(1 - exp(-0.9A/9.5)) where: m 0 = 9.5 KN, mi = 0.015 KN/mm and k = 0.9 KN/mm z \ Exponential model fitted to data 0 10 20 30 40 50 60 Displacement (mm) Figure 5.2 Load - displacement curves - OSB frame The yield displacement was found to be equal to 8.5 mm based on the MOC method (Fig 5.3). 5 9 Figure 5.3 Yield displacement - OSB frame - MOC procedure For checking the accuracy of the results, a second monotonic test was also performed. In this way the reliability of the data was confirmed (Figure 5 .4). First test Second test Figure 5.4 Racking tests results - OSB Frame 60 5.1.2 4-Brace frame The values of the initial elastic stiffness (k) and the other two parameters m 0 and mi, which represent the best fit of the data to Foschi's exponential function, are presented below. P = (14J + 0.033A)(1 - exp(-0.975A/14.7)) where: m 0 = 14.7 KN, k = 0.97 KN/mm and mi = 0.033 KN/mm Experimental data Figure 5.5 Load-displacement curves - 4-Brace frame According to the MOC procedure the yield displacement is equal to 12 mm (Fig. 5.6). 61 Displacement (mm) Figure 5.6 Yield displacement - 4-Brace frame - MOC procedure 6 2 5.1.3 2-Brace frame The load-deformation curves from the monotonic tests are shown in Figure 5.7. Displacement (mm) Figure 5.7 Racking test results - 2-Brace frames Foschi's exponential function, which includes two slopes for the load displacement curve (k and m i ) , provided good fits for the load-displacement trends for OSB frames and 4-Brace frame. The model did not fit the data from the first 2-Brace frame test. The reason can be attributed to the fact that the load-displacement curve for the first data of 2-Brace frame presents more than two slopes and a 4-Brace frame is more symmetric and at the same time closer to a shear wall configuration than a 2-Brace frame. A second racking test was required to evaluate the yield displacement. Considering the curve from the second test and using the MOC procedure the yield displacement was found to be 24 mm (Fig. 5.8) 63 Displacement (mm) Figure 5.8 Yield displacement - 2-Brace frame - MOC procedure 64 5.2 Failure modes 5.2.1 OSB sheathed frames Several failure patterns emerged from the monotonic tests. The typical deflection configuration for a sheathed frame under monotonic loading is the frame distorting into a parallelogram while the sheathing panels rotating and remaining rigid (Fig. 5.9). For both tests, the sheathing tended to pull away from the frame, pulling the nails along with it. More of the edge nails were pulled out of the studs (Fig. 5.10) and some of the edge nails were pulled through the sheathing only from the sill and girder (Fig. 5.11). For the frame bottom corner that was in tension, the upper nails from the "CP-T" connector were pulled out (Fig. 5.12). The frame was able to carry the racking load by the next Mabashira -Sill connection (Fig. 5.13) and the middle Stud-Sill connection until a tension perpendicular to grain failure of the sill occurred along the wood grain (Fig. 5.14). Finally, the load carried by the frame dropped to a value below 0.8 Pmax and the experiment was stopped. Figure 5.9 Typical deformation configuration for sheathed frames 65 OSB Panel Figure 5.10 Nails pulled out of the frame Stud OSB Panel Nail pulled through Figure 5.11 Nails pulled through the sheathing 66 Figure 5.12 Nails pull out 5.2.2 4-Brace frame The 4-Brace frame tests confirmed the hypothesis that the connectors play a critical role in the load-deflection behaviour of the wood frame and in resisting the external-racking load. From monotonically testing a 4-Brace frame and analyzing the failure mode (Fig. 5.15) the weakness of the " VP " and " BP " metal plate connections was shown to be the weak link for resisting this kind of load. This was partially caused by the fact that the bottom nails from both connectors were lying on the same horizontal line. Under loading, very high-tension perpendicular to grain stresses were created in the sill, which led to a splitting failure along the bottom nail line. At the beginning the frame failed in tension perpendicular to grain at the bottom corner. Then the external racking load was transferred to the next Mabashira-Sill connection (2 toe-nails - Fig. 5.16) and finally by the Middle Stud-Sill connection (Fig. 5.17). The second tension perpendicular to grain failure took place in the middle of the sill. At this time the load decreased to a value below 0.8 Pmax. Figure 5.15 Stud - Sill failure 68 Figure 5.16 Mabashira - Sill failure Figure 5.17 Middle Stud - Sill failure 69 5.2.3 2-Brace frame Failure occurred in the following sequence. First, a tension perpendicular to grain failure occurred in the sill along the horizontal lines of the bottom nails from the " VP " and " BP " connectors (Fig. 5.18) at the uplift corner. At this point the lateral load was redistributed to the next Mabashira-Sill joint that was able to still carry additional load until the toe-nails were pulled out (Fig. 5.19). Finally, the bottom " CP-T" nails pulled out from the middle Stud-Sill connection (Fig. 5.20). Figure 5.18 Stud-Sill failure Figure 5.19 Mabashira-Sill failure 70 Figure 5.20 Middle Stud - Sill failure 71 5.3 Overall frame response Table 5.1 summarizes the results of the five monotonic tests including some of the frame performance indicators such capacity, maximum displacement, ultimate displacement, yield displacement, ultimate shear strength, shear stiffness, ductility and specific gravity (S.G.) of the sill at oven dry (O.D) conditions. T A B L E 5.1 M O N O T O N I C T E S T R E S U L T S Sample Pmax (kN) ^failure (mm) (mm) Ayield (mm) Su (KN/mm) G' (MN/m) D (mm/ mm) Sill S.G. O.D OSB framel 10.50 115.5 50.8 8.5 5.77 0.916 13.58 0.48 OSB frame2 10.59 112.9 78.8 8.5 5.82 0.924 13.28 0.43 2-Brace framel 11.92 78.8 52.8 N/A 6.55 N/A N/A 0.51 2-Brace frame2 12.25 90.2 62.1 24 6.73 0.379 3.76 0.41 4-Brace frame 16.45 70.4 52.9 12 9.04 1.017 5.87 0.47 Analyzing the load-displacement curves (Fig. 5.1) and the performance values from Table 5.1, it can be seen that the 4-Brace frame has the highest capacity (56% increase compared with the OSB frames and 38%-34% with the 2-Brace frames), the highest stiffness (168% increase from 2-Brace frame and 11%-10% from OSB frames), a lower ductility compared with the OSB frame (56% to 57% reduction) and a high ductility compared with the 2-Brace frame (56% increase). 72 Specific gravity of wood is an important physical property. It represents the mass of a body divided by the mass of an equal volume of water. Wood specific gravity within the same species is related to growth rate, vitality of the tree when the wood was produced and location of wood in the tree. Generally, in softwoods the density decreases with height in the tree and increases with distance from the pith. Specific gravity variability is important for estimating the variability in the wood product capacity under loading conditions. Most mechanical properties of wood are closely correlated to the specific gravity. The strength of wood increases with specific gravity. Because in all the frames the failure took place at the sill, small specimens of wood (~ 20 x 40 x 50 mm) were cut from each sill to find its specific gravity. The range of specific gravity varies from 0.41 to 0.51 with a coefficient of variation (COV) of approximately 9.8%. Analyzing the results from Table 5.1, it can be seen that specific gravity variability induces variations in the peak load of the same types of frames and different frames. The vertical displacements of the end studs in relation to the sill for each frame configuration at ultimate capacity (0.8 Pmax) are listed in Table 5.2. For the compression end stud the recorded displacement was due to crushing of the wood sill. Tension end studs vertical displacements were high due to the fact that the dead load induced in the system was small and the studs were not restrained totally against uplifting. For the two and four braces frames the transducers also recorded the gap created between the end stud and the split sill along the grain. The end studs uplift from the OSB frames represented the gap between the studs and sill due to the nail failure from the" CP-T" connector and sheathing. 73 TABLE 5.2 VERTICAL DISPLACEMENTS OF THE END STUDS AT ULTIMATE CAPACITY Frame specimen Vertical displacement - Uplift - Middle Uplift - tension compression corner stud - Sill corner (mm) (mm) (mm) OSB framel -0.57 12.69 37.25 OSB frame2 -0.45 11.36 27.11 2-Brace framel -0.33 12.71 26.22 2-Brace frame2 -0.16 9.78 20.56 4-Brace frame -0.23 11.38 23.53 The following conclusions can be drawn from the monotonic tests presented in this chapter: • For all types of frames tested, the Sill - Stud connection was the critical element of the frames. • For the 2-Brace and 4-Brace frames, the metal plate hardware ("CP-T", "VP" and "BP") induced tension perpendicular to grain stresses in the sill and in this way they experienced brittle failures. • The OSB frames showed a good equilibrium between stiffness and ductility but because the top nails from the "CP-T" metal plate were pulled out the load carrying capacity was low. To achieve a higher load capacity and a higher ductility, the "CP-T" metal plate hardware should be replaced with "Shoe" or "Angle plates" connectors and the number of fasteners from sheathing/frame connection should be increased. • The 4-Brace frame reached high carrying load capacity, high initial stiffness (due to the four braces, the structure was more rigid) compared with 2-Brace and OSB frames. To achieve a more ductile system, the "VP" and "BP" connectors would need improvements. 74 The 2-Brace frame showed relatively poor performance under lateral loading conditions. The stiffness was very low; the yield displacement was high and therefore a very low ductility. The structural design can be improved by using additional braces, sheathing panels and better sill-stud connections. 75 CHAPTER V I - Cyclic experimental results for Post-and-Beam frames 6.1 Load - displacement curves 6.1.1 Frames sheathed with OSB Panels Two frames sheathed with OSB panels (OSB frames) were tested cyclically using the MOC protocol and the UBC - modified protocol as described in Chapter II and Chapter I V . The hysteresis loops from the cyclic tests with the corresponding monotonic curve are shown in Figures 6.1 and 6.2. Figure 6.1 Racking and cyclic load - displacement curves for OSB frames - MOC protocol 76 -18 J Displacement (mm) Figure 6.2 Racking and cyclic load - displacement curves for OSB frames - UBC - modified protocol 6.1.2 2-Brace frames Two 2-Brace frames were tested cyclically using the MOC protocol and the UBC-modified protocol as described in Chapter II and Chapter IV. The hysteresis loops from cyclic tests with the corresponding monotonic envelope curve are shown in Figures 6.3 and 6.4. 77 aj - First failure - Cyclic test g - First failure - Racking test 140 -120 -100 -80 18 60 80 100 120 140 Displacement (mm) Figure 6.3 Racking and cyclic load - displacement curves for 2-Brace frames MOC protocol 18 s - First failure - Cyclic test m - First failure - Racking test 140 -120 -100 -80 -60 -40 -18 140 Displacement (mm) Figure 6 . 4 Racking and cyclic load - displacement curves for 2-Brace frames UBC - modified protocol 78 6.1. 3 4-Brace frames Three 4-Brace frames were tested cyclically using the MOC protocol, UBC - modified protocol and UBC protocol as described in Chapter II and Chapter IV. The hysteresis loops in the cyclic tests with the corresponding monotonic envelope curve are shown in Figures 6.5, 6.6 and 6.7. Displacement (mm) Figure 6.5 Racking and cyclic load - displacement curves for 4-Brace frames -MOC protocol 79 —*t8 1 = Displacement (mm) Figure 6.7 Racking and cyclic load - displacement curves for 4-Brace frames MOC protocol 80 6.2 Failure modes 6.2.1 OSB sheathed frames Most of the damage observed in the OSB frames in the cyclic tests was confined to the sheathing and " CP-T " metal plate fastener connections. The nails were pulled through along the bottom sheathing edge (Figure 6.8) and pulled out (not to the entire nail length) near the top corners and on the interior edges where the sheathing panels met. As the load increased, the damage progressively extended from the lower corners further up and down along the sides of the frame. Figure 6.8 Nails pulled through The main difference in failure mode between the monotonic and cyclic tests was found in the " CP-T " metal plate connectors. For the monotonic tests the upper nails were pulled out; whereas, the bottom nails were pulled out in the cyclic tests (Figure. 6.9). For the frames tested using the UBC - modified protocol, failure initiated in the sill under tension perpendicular to grain at the corner subjected to uplift (tension). After that, the bottom nails from the "CP-T" connectors were pulled out. Nail pulled-through 81 For the cyclic tests, there was also evidence of compression perpendicular to grain damage in the sill at the compression corner, where the end stud crushed the sill (Fig. 6.9). Similar to the monotonic tests, the frame distorted into a parallelogram, while the sheathing rotated and remained relatively rigid under cyclic loading conditions (Fig. 6.10) Figure 6.9 Nails pulled out at" CP-T" connector Figure 6.10 Sheathed frame distortion 82 6.2.2 4-Brace frame Under both cyclic and monotonic load conditions these types of frames failed in the same way. The sill initially failed in tension perpendicular to grain at the connection between end studs and sill for the corner subjected to uplift (tension) (Fig. 6.11), and then the failure progressed to the middle stud sill connection (Fig. 6.12). Figure 6.11 Sill failure - Tension perpendicular to grain - End corners Figure 6.12 Sill failure - Tension perpendicular to grain - Middle stud 83 6.2.3 2-Brace frame The failure mode in the 2-brace frame under cyclic loading was similar to the failure mode in the monotonically tested frames. Due to the fact that the bottom nails of the "' VP " and " BP " connectors created concentrated stress along the wood grain, the sill failed in tension perpendicular to grain at the end stud-sill connection for the bottom corners subjected to uplift (tension) (Fig. 6.13). After this point the damage progressed to the next Mabashira-Sill joint when the toe-nailed connection suffered withdrawal failure (Fig. 6.14). Finally, the bottom nails from the middle " CP-T " connection started to pull out (Fig. 6.15) causing the load to drop below 0.8Pmax (post peak load). Figure 6.13 Sill failure - Tension perpendicular to grain Figure 6.14 Mabashira - Sill failure 84 85 6.3 Overall frame response Table 6.1 shows a summary of the cyclic test results. It lists frame performance indicators such as capacity, maximum displacement, displacement corresponding to Pmax, total dissipated energy and specific gravity (S.G.) of the sill (near the failure location). The frame capacity ranged from 10.358 kN to 15.135 kN, with the 4-Brace frame - UBC protocol having the highest capacity and the OSB frame - MOC protocol having the lowest capacity. Analyzing the results, the influence of the loading procedures on the structural performance of the post-and-beam frames can be seen. The frames that were tested using the MOC-protocol have a slightly lower capacity than the frames that were tested using the UBC - modified protocol. In this way it was shown that frames tested under high amplitude low sequence cyclic test protocol (UBC - modified or UBC) had a higher capacity in comparison with a lower amplitude long sequence cyclic test protocol (MOC). This observation confirmed results of He et al. (1999) on light frame shear walls. The failure displacement (Afaiiure) of the frames (the post peak displacement corresponding to 80% of the maximum load among all cyclic groups) ranged between 55.42 and 101.27 mm, with the OSB frame (MOC protocol) having the highest displacement and the 4-Brace frame (UBC - modified and UBC protocols) having the lowest displacement. The frames tested using the MOC protocol experienced higher failure displacements than the frames tested using UBC - modified and UBC protocols. The specific gravity of the sill ranged from 0.37 to 0.54 with a coefficient of variation (COV) of about 16%. Considering the results from Table 6.1, it is evident that for the 4-Brace and the 2-Brace frames, the higher values for sill specific gravity corresponded to higher frame capacity. This phenomenon is not evident for the OSB frames because the range of specific gravity values was less. 86 TABLE 6.1 CYCL IC TEST RESULTS Frame type Protocol Pmax Afaiiure A Ui Specific (kN) (mm) (mm) (kN mm) gravity (oven -dried basis) UBC- 12.49 57.9 50.4 845.6 0.41 OSB frame modified MOC 10.36 101.3 66.5 4123.2 0.44 UBC- 13.24 94.8 80.8 1746.3 0.54 2-Brace frame modified MOC 11.39 95.2 82.4 1758.5 0.46 UBC- 11.87 55.4 40.8 1427.9 0.37 4-Brace frame modified UBC 15.13 55.4 49.9 947.6 0.53 MOC 14.33 56.9 41.6 2098.2 0.46 The total dissipated energy was calculated as the sum of dissipated energies for all cyclic groups from 0 to post - peak load of 80%Pmax. Analyzing the amount of energy dissipated for each frame, it can be seen that there are differences between frame types and within the same type of frames but different testing methods. The MOC protocol resulted in higher energy dissipation in comparison with the UBC - modified and UBC protocols, due to higher displacements at failure and wider loops. For the OSB frames, the energy dissipated ranged from 845.6 kN mm to 4123.2 kN mm. The lower value of 845.6 kN mm was obtained by the OSB frame during cyclic tests with the UBC - modified protocol. The frame reached its maximum load relatively early during the first cycle from the second group of cycles. 87 Beyond this point, the sill split and the bottom row of nails from the "CP-T" connector failed in pulled out mode. After that point the load rapidly decreased to a level lower than 80% of Pmax. The frame tested using the MOC protocol reached a higher displacement of 101.3 mm at failure, which led to a higher amount of dissipated energy. For the 2-Brace frames, the amount of dissipated energy was similar for both the MOC and UBC - modified protocols. The smallest amount of dissipated energy achieved by the 4-Brace frame was obtained during the cyclic test with the UBC protocol (947.6 kN mm), because the load-displacement loops were more pinched than the loops of the frames tested using the UBC - modified (1427.9 kN mm) and the MOC (2098.2 kN mm) protocols. As a general conclusion in terms of energy dissipation, the most ductile system is the OSB frame (96.5% increase compared with 4-Brace frame and 134.47% from 2-Brace frame) for the MOC protocol, and for the UBC - modified protocol the most ductile system is the 2-Brace frame (22.29% increase compared with 4-Brace frame and 106.52% from OSB frame). The vertical displacements of the end stud relative to the sill at ultimate capacity of 80% Pmax (post - peak load) are listed in Table 6.2. All frames failed when the left stud was in compression, the middle stud in tension and the right stud in tension. For the compression end stud the recorded vertical displacement ranged from -0.37 mm (2-Brace frame - MOC protocol) to -1.69 mm (4-Brace frame - UBC - modified protocol). The maximum uplift of the middle stud was 9.86 mm (2-Brace frame - UBC - modified protocol) and the minimum value was 1.49 mm (OSB frame-MOC protocol). For the tension end stud the vertical displacement bounds ranged from 5.28 mm (OSB frame - MOC protocol) to 26.15 mm (2-Brace frame- MOC protocol). 88 TABLE 6.2 VERTICAL DISPLACEMENTS OF THE END STUDS Frame type Protocol Vertical displacement of the left stud (mm) Vertical displacement of the middle stud (mm) Vertical displacement of the right stud (mm) Comp. Tens. Comp. Tens. Comp. Tens. OSB frame UBC-modified -1.34 2.52 14.75 MOC -0.52 1.49 5.28 2-Brace frame UBC-modified -0.89 9.86 16.23 MOC -0.37 7.47 26.15 4-Brace frame UBC-modified -1.69 6.23 17.06 UBC -0.59 5.61 13.71 MOC -1.19 4.46 12.88 The following conclusions can be drawn from the data presented and discussed in this chapter: • There are not major differences in the capacity of the initial cycles of the cyclic test and the monotonic tests for all the frame configurations. Similar to the monotonic test results the 4-Brace frames had the highest capacity followed by the 2-Brace frames and finally the OSB frames. • For the cyclic tests the loading procedures had a significant influence on the structural performance of the post-and-beam frames. The frames that were tested using a high amplitude low sequence cyclic 89 test protocol (UBC - modified or UBC) had a higher capacity in comparison with a lower amplitude long sequence cyclic test protocol (MOC). In terms of energy dissipation the OSB frames were in the first place followed by the 4-Brace and 2-Brace frames. For the cyclic tests the 2-Brace and 4-Brace frames experienced brittle failures in the sill, which is similar to the monotonic test failures. For the OSB frames the main difference in the failure mode between cyclic and monotonic tests was that pull-out failures occurred of the bottom nails rather than the top nails in the "CP-T" connector at the uplift corner. The overall behaviour of the post-and-beam frames was very similar under either monotonic or cyclic loading Capacity and energy dissipation ability of these three types of frames can be greatly improved by replacing or improving the "CP-T", "BP" and "VP" connectors, and increasing the number of nails for the sheathing/frame connection in the OSB frames. 90 CHAPTER VII - Monotonic experimental test results for " T " connection One of the critical elements in wooden structures is the connection. Its strength, ductility, stiffness and the mechanical properties play an important role in determining the structural performance of the whole structure. From a design point of view, "good" joints should be strong enough to support external loads and should not introduce areas of high stress concentration (Curtu et al., 1988). Eight tests were carried out to determine the resistance of a Sill-Stud " T " connection, which consisted of a mortise and tenon and a " CP-T " metal plate. The specimens were subjected to a monotonic ramp tension loading of 0.13mm/sec. The graph in Figure 8.1 shows the load displacement curves for the monotonic tests. Table 7.1 depicts a summary of the experimental results. It lists connection performance indicators such as capacity (Pmax), failure displacement (Dfa„Ure), yield displacement (Dy), yield load (Py), stiffness at yield point (K), sill specific 91 gravity (S.G.) at oven-dry conditions and the failure mode. The joint stiffness influences the rigidity and stability of the entire structure and it is defined as the ratio between the applied external load and the corresponding displacement in the linear elastic range. TABLE 7.1 EXPERIMENTAL TESTS RESULTS Sample Pmax Py Dfailure Dy Dpmax K S. G. Failure No. (KN) (KN) (mm) (mm) (mm) (KN/ mm) Oven-dried mode 1 9.8 4.8 18.5 3 15.7 1.6 0.39 Sill split 2 11.7 5.5 19.2 3.8 18.6 1.45 0.42 Sill split 3 13.9 8.1 31.7 2.7 16.8 2.94 0.41 Nails pulled out 4 11.4 5.6 13.7 2.3 15.8 2.38 0.44 Sill split 5 13.8 7.15 36.3 3.6 15.9 1.98 0.46 Nails pulled out & metal plate failure 6 13.8 6.6 20.5 2.8 16.7 2.36 0.47 Nails pulled out 7 12 6.35 28.7 3.8 15.3 1.67 0.39 Nails pulled out 8 9.6 5.2 36.1 2.9 15.1 1.79 0.42 Sill split Mean 12 6.16 30.8 3.12 16.11 2.02 0.42 Standard 1.73 1.10 9.99 0.54 1.33 0.50 0.03 deviation The yield displacement was calculated according to the MOC testing standard. 9 2 The joint capacity ranged between 9.6 kN and 13.9 kN with a 14% Coefficient of Variation. The yield displacement varied from 2.3 mm to 3.8 mm with a coefficient of variation of 17%. Because wood is a natural material, its variability (between different elements and even for the same element in term of density) is reflected in the variability of the experimental test results. The sill specific gravity range varied from 0.39 to 0.47 with a coefficient of variation of 7.3%. Analyzing the results from Table 8.1, it can be seen that specific gravity variability induced variation in the connection capacity and connection stiffness. Figures 7.2 and 7.3 show the plot of the maximum capacity versus specific gravity and stiffness versus specific gravity, respectively. 8^ . , , 1 , 1 0.37 0.39 0.41 0.43 0.45 0.47 0.49 Specific gravity Figure 7.2 Maximum capacity vs. Specific gravity - "T" Connection 93 35 2 y = 1.3017x + 1.6745 R 2 = 0.0085 • Split Sill • Nails pulled out • - Linear (Split Sill) Linear (Nails pulled out) CO y = 13.843x-3.9745 R 2 = 0.4897 1.5 0.39 0 41 0 43 0.45 0.47 0.49 Specific gravity Figure 7.3 Stiffness vs. Specific gravity - "T" Connection Due to the fact that the number of tested specimens was relatively small and the specific gravity represented a relatively narrow range, it was not possible to identify a strong trend for the plots presented above. The " best " relationship between the maximum capacity and specific gravity seems to be a linear function with R 2 (coefficient of determination) equal to 0.48 and 0.36 for split sill and nails pulled out failures, respectively. In this case 48% for split sill failure and 36% for nails pulled out failure of the total variation in the values of the maximum capacity can be explained by the linear relationship with the values of the specific gravity. For the stiffness vs. specific gravity the " best " trend that could be fitted was expressed by a linear function with R 2 equal to 0.48 and 0.008 for split sill and nails pulled out failures, respectively. Between these two variables 48% of the total variation in the stiffness values can be accounted for the linear relationship with the values of the specific gravity for split sill failure and only 0.8% for nails pulled out failure. 94 The values of R2 are small because another variables (the number of nails, nail type and metal plate type) might contribute to the stiffness and the maximum capacity of the connection. The tests provided useful information regarding the weakest element of the connection and the stress concentrations by analyzing the main failure modes: • Splitting of the sill due to the highly stressed zone along the bottom nail line of the "CP-T" connector (Fig. 7.4) for the specimens with a lower specific gravity (specimen 1,2,4 and 8) • Pulling out of the bottom nails from the sill (Fig. 7.5) for the specimens with a higher specific gravity (specimen 3, 5, 6, 7) and an extreme case (Fig 7.6) when the "CP-T" metal plate failed in shear (specimen 5) Figure 7.4 Failure mode - Splitting of the sill 95 The vertical displacements of the end stud (d1- left displacement, d2 - right displacement) for the failure load in relation to the sill, the uplift of the stud (d3 -front displacement, d4 - back displacement) relative to the machine table, and the uplift of the sill (d5 - front bottom displacement, d6 - back bottom displacement) relative to the machine table are shown in Table 7.2. Due to the fact that the sill and the stud were connected only on ane side through the "CP-T" metal plate and nails, the connection became eccentric during loading and the sill was twisted in the middle (Fig. 7.7). The values of the tilting angle (a°) between the sill and the end stud are presented in Table 7.2. TABLE 7.2 VERTICAL DISPLACEMENTS AND TILTING ANGLES Sample d1 d2 d3 d4 d5 d6 a° (mm) (mm) (mm) (mm) (mm) (mm) 1 13.11 13.85 15.65 19.15 6.16 12.2 3.29 2 14.52 14.9 16.16 20.12 4.06 8.87 2.62 3 26.6 26.74 27.56 32.7 6.34 13.04 3.65 4 11.63 11.77 12.88 16.47 3.19 6.84 1.99 5 32.5 32.62 32.25 37.77 5.85 11.84 3.26 6 14.52 15.5 17.13 21.83 5.51 10.88 2.93 7 24.28 24.03 25.4 30.09 11.25 17.85 3.6 8 33.07 33.37 34.72 36.16 6.11 9.71 1.96 Mean 21.28 21.59 22.71 26.78 6.06 11.40 2.91 Standard 8.90 8.71 8.34 8.35 2.38 3.28 0.67 deviation 97 Figure 7.7 Tilting angle The mean and the standard deviation for the vertical displacements recorded by the transducers are also shown in Table 8.2. Transducer T4 recorded the maximum uplift, followed by T3 (15% decrease), T2 (19.39% decrease), T1 (20.56% decrease), T6 (57.43% decrease) and T5 (77.37% decrease), respectively. The following conclusions can be drawn from the monotonic test data presented and discussed in this chapter: • The "CP-T" metal plate hardware is the critical element of the connection. The bottom nails in the connector induced tension perpendicular to grain stresses in the sill creating a brittle system. • Stiffness and capacity of this " T " connection can be greatly improved by replacing the "CP-T" metal plate with better hardware. 98 CHAPTER VIII -Conclusions and Recommendations 8.1 Conclusions 8.1.1 Post and beams frames The main objective of this project was to quantify the structural performance of three types of Japanese post-and-beam frames (4-Brace frame, 2-Brace frame and OSB sheathed frame) under lateral loading. Each type of frame was examined under monotonic and cyclic test procedures. The monotonic tests were conducted in accordance with the ASTM Standard, and the cyclic tests were conducted in accordance with the modified UBC and MOC (Ministry of Construction - Japan) loading procedures. Based on the experimental test results, several differences between momotonic and cyclic response could be identified for each type of frame, considering the ultimate load and displacement, maximum capacity, stiffness, energy dissipation, ductility and the mode of failure. For all types of frames the failure loads (80%Pmax - post peak load) measured in cyclic tests were comparable to the monotonic failure loads. However, the corresponding displacements at the failure load were much smaller for the cyclic test compared to the monotonic test. For the cyclic tests the failure displacement was smaller because it depends on the cyclic protocol. All types of frames showed softening after a certain deflection at which the load diminished drastically. The goal is to design wood structures that have the ability to dissipate energy and deform without having brittle or sudden failure. The results indicated that the structural performances of the post and beam frames depend on the loading procedures. Frame ductility was determined by dividing the displacement at failure load by the yield displacement. Therefore, the monotonic test could predict the maximum load carrying capacity of the frame under cyclic loading, but not the ductility. The cyclic test load results were close to the monotonic test load results during the first cycle of each cycle group at a given displacement. Because of the degradation in stiffness at constant amplitude, the load capacity for the second 99 and the third cycle (within the same cyclic group with constant amplitude) was substantially reduced from the first cycle. Although, this behavior was not captured by the monotonic test procedures, research on dynamic response of wood - based shear walls is necessary to determine realistic loads for lateral load resisting systems. For the monotonic tests, in terms of maximum capacity and stiffness, the 4-Brace frames had the highest capacity (56% increase compare to OSB sheathed frames and 38%-34% compared to 2-Brace frames), the highest stiffness (168% increase from 2-Brace frame and 11%-10% from OSB sheathed frames). In term of ductility the 4-Brace frames had a 56%-57% reduction in ductility compared to OSB sheathed frames and 56% increase in ductility compared to the 2-Brace frame. The frames were tested using different loading protocols to demonstrate the dependence of post and beam frame behaviour on the test protocol. The MOC protocol was included so that data can be compared with prior Japanese research. According to the cyclic test results, the frames that were tested using the MOC-protocol had a slightly lower capacity compared to the frames that were tested using the UBC - modified protocol. The most ductile system, considering the amount of energy dissipated, was the OSB sheathed frame (96.5% increase compared with 4-Brace frame and 134.47% compared to 2-Brace frame) for MOC protocol, and for the UBC - modified protocol the most ductile system is the 2-Brace frame (22.29% increase compared with 4-Brace frame and 106.52% from OSB sheathed frames). The damage and the failure modes of the 2-Brace and 4-Brace frames were similar for the monotonic and cyclic tests. These frames experienced brittle failures due to the splitting failure of the sill along the line of the bottom nails of the "VP" and "BP-T" metal plates, at the tension bottom corner. The OSB sheathed frames failed differently for the monotonic and cyclic tests. First, the damage for the monotonic test was characterized by the pulling out 100 failure of the sheathing/frame nail connections and the pull through failure of the nails from the base of the wall. Secondly, the upper nails from the "CP-T" connector were pulled out at the uplift corner and finally the splitting of sill at the Middle stud - sill connection. For the cyclic test the sheathing panel behaviour was similar to that in the monotonic test. The difference in the failure mode resulted from pullout failure of the bottom nails rather than the top nails in the "CP-T" connector at the uplift corners. For further research, the experimental results obtained in this thesis can be used in model development that will provide analytical tools for predicting the general behaviour and the reliability of post and beam wall systems. 8.1.2 " T " Connector The connection behaviour under monotonic tension loading conditions was influenced by: the number and the types of the nails; orientation of the nails related to wood grain; mechanical properties of the "CP-T" metal plate; and the mechanical properties of the wood elements. The experimental results typically showed brittle behaviour of this "T" type of connection (used for 2-Brace and 4-Brace frames). Under loading, the bottom nails from the "CP-T" metal plate induced an area of high-tension perpendicular to grain in the sill. Splitting failure was observed in the specimens with low specific gravity and nail pull out and/or metal plate shear failures were observed in the specimens with high specific gravity. The maximum capacity ranged from 9.6 kN to 13.9 kN and the maximum displacement ranged from 13.7 mm to 36.3 mm. This type of connection could be improved to achieve a more ductile behaviour. 101 8.2 Recommendations The frames with 4 and 2 braces were brittle systems experiencing higher initial stiffness and higher failure loads. The OSB frames on the other hand were more ductile systems experiencing lower peak loads but a substantial increase in ductility. The ductility values taken alone cannot completely define the structural performance. A system with high ductility but low stiffness and strength may experience extensive displacements and costly damage during an earthquake of moderate magnitude. As previously mentioned, ductility is defined by dividing the failure drift by the yield drift. However, the yield displacement is dependent upon stiffness that prevents high deformations during wind and earthquake loading. Therefore, systems with good seismic resistance should have a balance between strength, stiffness and ductility. In the OSB sheathed frames the sheathing-frame nails helped with the dissipation of energy by deforming under cyclic load conditions. To achieve a higher load capacity for these types of frames, the " CP-T " metal plate hardware used in the stud to sill and stud to girder should be improved. Although the 4-Brace frames experienced relatively high carrying load capacity and relative high stiffness, improvement to the hardware used in the stud to sill and stud to girder connections can increase system ductility. "Shoe" or "Angle Plates" connectors can be used to replace the "VP" and "BP" connectors that created high-tension perpendicular to grain stresses in the sill. The 2-Brace frame showed relatively poor performance under lateral loading conditions. The stiffness was very low; the yield displacement was high and therefore resulting in a very low ductility. The structural design can be improved by using additional braces, sheathing panels and replacing the " VP ", " CP-T " and " BP " connectors with better ones. 102 BIBLIOGRAPHY American Society for Testing and Materials (ASTM) (1991). "Standards: Standard Method of Static Load Test for Shear Resistance of Framed Walls for Buildings (ASTM standard E 564-75). "04.07, Philadelphia, USA. American Society for Testing and Materials (ASTM) (1996). "Proposed standard method of cyclic load test for shear resistance of framed walls for buildings." Philadelphia, USA. Canada Mortgage and Housing Corporation (1982). Canadian Wood-Frame House Construction. Canada Canadian Wood Council (1990). Wood Design Manual. Ottawa, Canada. Canadian Wood Council (1991). Wood Reference Handbook. Ottawa, Canada. Dolan, J. D. and Foschi, R. O. (1991). "Structural analysis model for static loads on timber shear walls." J. of Struct. Engrg., ASCE, 117(3), 851-861. Dolan, J. D. and Madsen, B. (1992). "Monotonic and cyclic tests of timber shear walls." Canadian J. of Civil Engrg., 19, 415-422. Dolan, J. D. and White, M.W. (1995). "Nonlinear shear-wall analysis." J. of Struct. Engrg., ASCE, 121(11), 1629-1635. Foschi, R. O. (1974). "Load-slip characteristics of nails." Wood Sci., 7(1), 69-76 Foschi, R. O. (1990). "Analysis of wood diaphragms and trusses, Part one: Diaphragms." Canadian J. of Civil Engrg., 4(3), 345-352. Haygreen, J. and Bowyer, J. (1996). Forest Products and Wood Science: an introduction. Third edition. ISBN 0-8138-2256-4.United States of America He, M., Lam, F. and Prion, H. G. L. (1998). "Influence of cyclic protocol on performance of wood-based shear walls." Canadian J.of Civil Engrg. 25: 539-550. 1 0 3 Hirashima, Y., Kanaya, N., Hatayama, Y. and Kamiya, F. (1981). "The performance of wooden frames with bracing for horizontal shearing force and their structural analysis. Racking tests of wooden frames." Journal of the Japan Wood Research Society, 27(12), 845-854. lamayama, N., Ikehata, T., Sugiura, M. and Nakamura, I. (1990). "Effect of tenon length on bending strength of mortise-and-tenon joints." Journal of the Japan Wood Research Society. 36(1), 85-91. Itani, R. Y. and Obregon, S. A. (1984). "Nonlinear racking analysis of nailed walls." Wood and Fiber Science, 16(3), 454-465. Japanese Government Housing Loan Corporation Building Manual. Translated and distributed with permission of Japanese Housing Loan Corporation by Council of Forest Industries with the support of the Standards Council of Canada. Lam, F., Prion, H.G.L. and He, M. (1996). "Lateral resistance of wood based shear walls with oversize sheathing panels." Int. Council for Building Research Studies and Documentation, Working Commission W18 - Timber Structures, Meeting Twenty - Nine, Bordeaux, France. Mihailescu, T. and Nicholls, T. (1999). "Problems encountered in designing a finite element model of mortise and tenon joints." Conference paper, ICWSF '99, Missenden Abbey, England. National Research Council Canada (1990). National Building Code of Canada. Ottawa, Canada. Seo, J., Choi, I. and Lee, J. (1999)." Static and cyclic behavior of wooden frames with tenon joints under lateral load." J. of Struct. Engrg., 125(3) 104 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089420/manifest

Comment

Related Items