Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Self-tapping screw assemblies under monotonic and reverse cyclic load Closen, Max 2012

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

Item Metadata

Download

Media
24-ubc_2012_fall_closen_max.pdf [ 11.85MB ]
Metadata
JSON: 24-1.0072901.json
JSON-LD: 24-1.0072901-ld.json
RDF/XML (Pretty): 24-1.0072901-rdf.xml
RDF/JSON: 24-1.0072901-rdf.json
Turtle: 24-1.0072901-turtle.txt
N-Triples: 24-1.0072901-rdf-ntriples.txt
Original Record: 24-1.0072901-source.json
Full Text
24-1.0072901-fulltext.txt
Citation
24-1.0072901.ris

Full Text

SELF-TAPPING SCREW ASSEMBLIES UNDER MONOTONIC AND REVERSE CYCLIC LOAD by  MAX CLOSEN  Dipl.-Ing. (FH), University of Applied Sciences, Rosenheim, Germany, 2008  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE  in  THE FACULTY OF GRADUATE STUDIES  (Forestry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  July 2012  © Max Closen, 2012  ABSTRACT In the past century old-growth forests in Canada and the USA provided sufficiently large, clear wooden construction material which have been extensively used. Today, the importance of high-quality structural timber and wood products has increased by far. This increase in demand for high-quality timber and wood products can only be satisfied with second-growth wood, some remaining old-growth forests, and of course engineered wood products. The performance of these materials in structures is, however, largely influenced by the capacity of connections. The envelope in timber construction can only be pushed forward if research on mechanical fasteners and connections that are strong, reliable and cost efficient is conducted. Primary focus of research must address the inherent tensile and shear weaknesses of wood perpendicular and parallel to the wood grain. The thesis presented here experimentally investigates the performance of newly evolved structural self-tapping full thread wood screws as a primary fastener in Canadian Douglas-fir glulam and Cross-Laminated-Timber. The screws as primary fasteners were investigated in a commonly used shear connection and a recently developed moment resisting assembly under reverse cyclic load. Both connection systems utilize the high withdrawal resistance and tensile strength of the fastener with inclined (screw-in angles between 30° and 45°) arrangements. The inclined arrangement allows force transfer along the fastener axis and therefore reduces perpendicular to grain splitting and parallel to grain shear failure and provides high connection capacities and stiffness. The results show that structural self-tapping wood screws can effectively be used as primary connector under reverse cyclic loading conditions. In addition to the screw’s superior ii  withdrawal resistance and tensile strength the research showed that self-tapping screws can be applied efficiently with commonly available machinery and tools.  iii  TABLE OF CONTENTS ABSTRACT…… ................................................................................................................ ii TABLE OF CONTENTS ................................................................................................... iv LIST OF TABLES ........................................................................................................... viii LIST OF FIGURES ............................................................................................................ x ACKNOWLEDGEMENTS .............................................................................................. xv 1. Introduction…. ................................................................................................................ 1 1.1.  Background and problem statement......................................................................... 1  1.2.  Significance.............................................................................................................. 3  1.3.  Thesis overview ....................................................................................................... 4  2.  Literature survey and background ............................................................................... 5  2.1.  Introduction to wood as a construction material ...................................................... 5  2.2.  Wood and its properties ........................................................................................... 7  2.3.  Moisture induced stresses in timber joints ............................................................... 8  2.4.  Design of timber joints............................................................................................. 9  2.4.1.  Ductility, stiffness and energy dissipation ......................................................... 11  2.4.2.  The group effect ................................................................................................. 14  2.5. 2.5.1.  Timber moment joints under cyclic load ............................................................... 15 Steel tube reinforced timber moment joints ....................................................... 16  iv  2.6.  Introduction to STS in timber construction ........................................................... 20  2.7.  STS reinforced timber moment joints .................................................................... 23  2.7.1.  Application possibilities for STS ....................................................................... 28  2.7.2.  Alternative connection details with STS ............................................................ 31  2.7.3.  Available design equations for STS assemblies ................................................. 35  2.7.4.  Design equations for tension and compression assemblies with ZD-plates....... 40  2.7.5.  Spacing’s and end distances for STS ................................................................. 42  2.8.  Estimated connection capacities ............................................................................ 45  2.8.1.  Estimated connection capacity – shear connection ............................................ 45  2.8.2.  Estimated connection capacity – moment connection ....................................... 46  2.9.  Loading procedure ................................................................................................. 48  2.9.1.  CUREE loading procedure – shear connection .................................................. 49  2.9.2.  CUREE loading procedure –moment connection .............................................. 53  3.  Experimental procedures and materials ..................................................................... 56  3.1.  Introduction ............................................................................................................ 56  3.2.  Specimen identification ......................................................................................... 56  3.3.  Timber members .................................................................................................... 57  3.3.1.  Shear connection specimen ................................................................................ 58  3.3.2.  Moment connection specimen ............................................................................ 64  3.4.  Required hardware ................................................................................................. 65 v  3.4.1.  Steel plates.......................................................................................................... 66  3.4.2.  Self-tapping screws ............................................................................................ 68  3.4.3.  Tension and compression plates ......................................................................... 69  3.5.  Test setup ............................................................................................................... 71  3.5.1.  Equipment shear connection .............................................................................. 71  3.5.2.  Equipment moment connection .......................................................................... 77  4.  Results and discussion ............................................................................................... 81  4.1.  General overview ................................................................................................... 81  4.2.  Shear connection .................................................................................................... 81  4.2.1.  GG-4X series and GG -8X series ....................................................................... 82  4.2.2.  GCLT-4X series and GCLT -8X series.............................................................. 92  4.2.3.  GG-4X-30 series and GCLT -4X-30 series...................................................... 100  4.3.  Data analysis – shear connection ......................................................................... 104  4.4.  Moment connection ............................................................................................. 105  4.5.  Data analysis – moment connection .................................................................... 118  5.  Conclusions.............................................................................................................. 120  5.1.  General ................................................................................................................. 120  5.2.  Shear connection .................................................................................................. 120  5.3.  Moment connection ............................................................................................. 123  5.4.  Recommendations ................................................................................................ 124 vi  BIBLIOGRAPHY ........................................................................................................... 125 APPENDIX…….. ........................................................................................................... 132  vii  LIST OF TABLES Table 1: Summary of the test results from (Lam, et. al., 2007®), by permission ............. 24 Table 2: Summary of the test results from (Lam et.al., 2010®), by permission ............... 25 Table 3: Comparison among calculated characteristic capacities of conventional joints and 5%-tile values of tested moment mitre joints (Trautz & Koj, 2009®) ................ 28 Table 4: Spacing’s end and edge distance abbreviations used in design standards .......... 43 Table 5: End, edge and spacing requirements according to DIN 1052: 2004-08 ............. 44 Table 6: Estimated shear-connection capacities ............................................................... 46 Table 7: Cycle amplitudes for shear connection tests ....................................................... 52 Table 8: Cycle amplitudes for moment connection tests .................................................. 54 Table 9: Average recorded moisture content (MC) and density of shear member specimen ................................................................................................................................... 58 Table 10: Recorded connection capacities and stiffness for the GG-4X series ................ 86 Table 11: Recorded connection capacities and stiffness for the GG-8X series ................ 88 Table 12:Recorded observations on screws and wood members after specimen dissasembly for the GG-4X and GG-8X series ......................................................... 91 Table 13: Recorded connection capacities for the GCLT-4X series ................................ 95 Table 14: Recorded connection capacities for the GCLT-8X series ................................ 97 Table 15: Recorded observations on screws and wood members after specimen dissasembly for the GCLT-4X and GCLT-8X series ................................................ 99 Table 16: Recorded connection capacities and stiffness for the GG-4X -30series........ 104 Table 17: Statistics summary of recorded data ............................................................... 117 Table 18: Measured moisture contents (MC) and wood densities for GG-4X series ..... 132 Table 19: Measured moisture contents (MC) and wood densities for GG-8X series ..... 132 viii  Table 20: Measured moisture content (MC) and wood densities for GCLT-4X series .. 133 Table 21: Measured moistre content (MC) and wood densities for the GCLT-8X series ................................................................................................................................. 133 Table 22: Measured moisture contents (MC) and wood densities for the GG-4X-30 series ................................................................................................................................. 134 Table 23:Recorded moment connection test data ........................................................... 135 Table 24: Measured moisture contents (MC) and wood densities for moment connection ................................................................................................................................. 136  ix  LIST OF FIGURES Figure 1: Load curve and definitions ................................................................................ 10 Figure 2: Typical examples of joints and respective ductility curves ............................... 14 Figure 3: Conceptual example of the expanded tube type reinforcement ........................ 17 Figure 4: (a) example of a beam-to-column connection type with densified wood in the dowel bearing zone, (b) schematic energy dissipation capacity between unreinforced and reinforced connection types. ............................................................................... 20 Figure 5: Variety of STS head shapes and tapping tip compared to drill bit .................... 22 Figure 6: Connection layout examined by (Lam et.al., 2007®), adapted by permission .. 23 Figure 7: Tested connection layouts from Lam et.al., (2010®), adapted by permission. Left side: bolt layouts, right side: STS layout. All dimensions in mm. ..................... 26 Figure 8: Typical screw arrangements: purlin to rafter connection, beam extension or repair detail and shear connection. ............................................................................ 30 Figure 9: Panel to panel shear connection ........................................................................ 32 Figure 10: Panel to beam connection ................................................................................ 32 Figure 11: Typical post-to-beam connection: transferring shear forces and moments ..... 33 Figure 12: Typical rigid STS assembly: transferring moments ........................................ 34 Figure 13: Typical connection: transferring tension and compression forces .................. 34 Figure 14: Example of reinforcements in CLT ................................................................. 35 Figure 15: Assumed force distribution in ZD-plate connection ....................................... 42 Figure 16: Simplified assumptions for connection capacity estimate .............................. 47 Figure 17: Principle forces of the ZD-plate system. Before load application (a), after screw head engages with lid in compression (b), force perpendicular to shear plane resulting from eccentricity of the steel plate.............................................................. 48 Figure 18: Modified CUREE loading protocol (shear connection) .................................. 51 x  Figure 19: CUREE loading protocol (moment connection) ............................................. 53 Figure 20: Specimen identification ................................................................................... 57 Figure 21: GG-4X and GG-8X series specimen (dimensions in mm) .............................. 60 Figure 22: GG-4X-30 series specimen (dimensions in mm) ............................................ 61 Figure 23: GCLT series specimen (dimensions in mm) ................................................... 63 Figure 24: GCLT-30 series specimen (dimensions in mm) .............................................. 64 Figure 25: Required millwork for specimen assembly ..................................................... 65 Figure 26: Custom made steel-bracket (dimensions in mm) ............................................ 67 Figure 27: Dimension of the employed fasteners from SWG........................................... 69 Figure 28: ZD-plate (Blass, 2010b) .................................................................................. 70 Figure 29: Shear test rig front view .................................................................................. 73 Figure 30: Shear test rig side view .................................................................................... 74 Figure 31: General shear connection test setup and transducer positions ........................ 75 Figure 32: Shear connection actuator load application device ......................................... 76 Figure 33: Clamping detail of steel cross bars .................................................................. 77 Figure 34: Steel bars stretching over horizontal connection member (left), fixture to avoid horizontal movement (right). ..................................................................................... 78 Figure 35: Vertical actuator movement recording (left), rotation recording (right) ......... 79 Figure 36: General test set-up for moment connection tests............................................. 80 Figure 37: Typical tension fracture of screws, pull-in failure and screw bending ........... 83 Figure 38: Typical example of connection failure and specimen separation .................... 84 Figure 39: Typical load displacement curve for GG-4X series ........................................ 85 Figure 40: Typical load displacement curve for GG-8X series ........................................ 87 Figure 41: Specimen separation (top left), typ. screw failures (top right) and splitting ... 89 xi  Figure 42: Withdrawal, push-out failure (top), tension fracture, fastener yielding (middle), specimen separation (bottom) .................................................................... 93 Figure 43: Typical load displacement curve for GCLT-4X series ................................... 94 Figure 44: Typical load displacemnt curve for GCLT-8X series ..................................... 96 Figure 45: Typical splitting of CLT at the bottom of the side member under large loads 98 Figure 46: Typical shear failure parallel to grain and expected point of stress concentration at the screw tip for GCLT specimen ................................................. 101 Figure 47: Typical load displacement curve for GG-4X-30 series ................................. 102 Figure 48: Typical fastener yielding for the GG-4X-30 series and GCLT-4X-30 series 103 Figure 49: Observed wood crushing at the upper end of the beam member .................. 106 Figure 50: Yielding of compression screws and fracture of tension screws (top), screw withdraw on all three ZD-plates (bottom). .............................................................. 108 Figure 51: Bolt push-out and bending ............................................................................ 109 Figure 52: Combined compression and transverse shear failure at column member. .... 110 Figure 53: Wood crushing under hold-down steel pipe (left), developed gap at bottom of column member (right) ............................................................................................ 111 Figure 54: Plastic deformation of steel shoe side plates after testing ............................. 112 Figure 55: Splitting of column member perpendicular to the grain................................ 112 Figure 56: Steel plate yielding and screw withdraw at bottom steel plate...................... 113 Figure 57: Typical moment rotation plot with tensile fracture of screws followed by a steep load drop. ........................................................................................................ 115 Figure 58: Typical moment rotation plot with intense wood crushing at the column member followed by screw withdraw...................................................................... 115 Figure 59: Examples for calculated equivalent viscous damping ratios ......................... 118 Figure 60: Moment rotation response specimen Mc-1 ................................................... 137 xii  Figure 61: Moment rotation response specimen Mc-2 ................................................... 137 Figure 62: Moment rotation response specimen Mc-3 ................................................... 138 Figure 63: Moment rotation response specimen Mc-4 ................................................... 138 Figure 64: Moment rotation response specimen Mc-5 ................................................... 139 Figure 65: Moment rotation response specimen Mc-6 ................................................... 139 Figure 66: Moment rotation response specimen Mc-7 ................................................... 140 Figure 67: Moment rotation response specimen Mc-8 ................................................... 140 Figure 68: Load deformation response specimen GG1-4X ............................................ 141 Figure 69: Load deformation response specimen GG2-4X ............................................ 141 Figure 70: Load deformation response specimen GG3-4X ............................................ 142 Figure 71: Load deformation response specimen GG4-4X ............................................ 142 Figure 72: Load deformation response specimen GG5-4X ............................................ 143 Figure 73: Load deformation response specimen GG1-8X ............................................ 144 Figure 74: Load deformation response specimen GG2-8X ............................................ 145 Figure 75: Load deformation response specimen GG3-8X ............................................ 146 Figure 76: Load deformation response specimen GG4-8X ............................................ 147 Figure 77: Load deformation response specimen GG5-8X ............................................ 148 Figure 78: Load deformation response specimen GG1-4X-30 ....................................... 149 Figure 79: Load deformation response specimen GG2-4X-30 ....................................... 150 Figure 80: Load deformation response specimen GG3-4X-30 ....................................... 151 Figure 81: Load deformation response specimen GG4-4X-30 ....................................... 152 Figure 82: Load deformation response specimen GG5-4X-30 ....................................... 153 Figure 83: Load deformation response specimen GCLT1-4X ....................................... 154 Figure 84: Load deformation response specimen GCLT2-4X ....................................... 155 xiii  Figure 85: Load deformation response specimen GCLT3-4X ....................................... 156 Figure 86: Load deformation response specimen GCLT4-4X ....................................... 157 Figure 87: Load deformation response specimen GCLT5-4X ....................................... 158 Figure 88: Load deformation response specimen GCLT1-8X ....................................... 159 Figure 89: Load deformation response specimen GCLT2-8X ....................................... 160 Figure 90: Load deformation response specimen GCLT3-8X ....................................... 161 Figure 91: Load deformation response specimen GCLT4-8X ....................................... 162 Figure 92: Load deformation response specimen GCLT5-8X ....................................... 163 Figure 93: Load deformation response specimen GCLT1-4X-30 .................................. 164 Figure 94: Load deformation response specimen GCLT2-4X-30 .................................. 165 Figure 95: Load deformation response specimen GCLT3-4X-30 .................................. 166 Figure 96: Load deformation response specimen GCLT4-4X-30 .................................. 167 Figure 97: Load deformation response specimen GCLT5-4X-30 .................................. 168  xiv  ACKNOWLEDGEMENTS I would like to thank Professor Frank Lam for the guidance provided throughout my graduate studies and this project. Also I want to thank George Lee for the help and advice provided during testing. My gratitude’s are also expressed to the MITACS Accelerate program for the financial support throughout this research project. The corporation and funding received from Schraubenwerk Gaisbach GmbH (SWG) (www.swg-produktion.de), funding received from ChiuHippmann Consulting Structural Engineers. I also wish to thank my family and in particular my girlfriend Simone for their lovely support during this mission.  xv  1. Introduction 1.1. Background and problem statement Self-tapping wood screws (STS) have been introduced as cost effective and simple solutions to the timber construction market. Many innovative engineers and architects view them as relatively straightforward and simple to detail while builders view them as functional, understandable, and economically effective to install. Over the past decade self-tapping screw type fastener have evolved into a diverse spectrum of applications that include use as reinforcement, fastener in heavy timber and light-weight post-beam connections, and recently, fastener in moment resisting timber connections. The design community in Europe has accepted this fastener as a popular choice in many different applications due to its outstanding performance under design dead and live loads (Blass, et. al., 2006, Blass, 2010). The major part of the research has been focused on understanding the behavior of STS assemblies and their performance under monotonic loading (Blass et. al., 2006, Blass, 2010, Lam et. al., 2010). In general, monotonic loading simulates a sustained static load at stresses that do not cause permanent deformation (Anderson, 2001). Generally, the design also accounts for a live load by an estimation of the said load coupled with a load factor that accounts for rate, intensity and duration of the load. Available published research yields adequate design rules for STS assemblies considering dead and live loads. From research in the field of earthquake engineering information on dynamic load events is available to investigate the performance of STS assemblies under such loads. The 1  nature of an earthquake causes lateral acceleration of the structural system inducing inertial components applied in a cyclic manner in addition to gravitational components. Earthquake events are prevalent throughout western Canada and the western United States of America. As an example, the west coast of Canada near the “Ring of Fire” is prone to major earthquakes. Given the tectonic setting with tectonic plates sliding along each other and subducting beneath one another, shallow crustal earthquakes (magnitudes of 5.5 to 7.5) and deep subduction zone earthquakes (magnitudes of 4 to 9.5) are known to have occurred in the distant past (National Resources Canada). Canada’s Queen Charlotte fault for example caused a magnitude 8.1 earthquake on August 22nd 1949 (NRC,2012). According to Earthquakes Canada (at www.nrcan.gc.ca), more than 100 earthquakes of magnitude 5 or greater have occurred in the offshore region west of Vancouver Island during the past 70 years. Stronger earthquakes with magnitudes ranging from 5.5 to 9.5 have the ability to damage and destroy large and highly populated areas in a very short period of time. In Canada, the National Building Code of Canada (NBCC) and in the USA the Uniform Building Code (UBC) consider structural damage to be inevitable and complete destruction or collapse as unacceptable. Recent earthquakes in urbanized areas such as the Haiti Earthquake on January 12th 2010 and the Chile Earthquake on March 11th 2010 show the frequency of dynamic loading events. Typically, static monotonic test methods have been used to evaluate the capacity of newly developed timber connections (Lam et.al., 2007). Results from static monotonic 2  tests sufficiently describe the static connection behaviour, however, these results have limited use in dynamic analyses with cyclic loading. Most timber joints will degrade in strength and stiffness when subjected to dynamic loading. Therefore, concerns are raised regarding the adequacy and reliability of design equations for STS assemblies and their performance under cyclic loading. Without sufficient knowledge about the performance and expected amount of damage, the market for self-tapping screw assemblies could be adversely affected. Successful understanding of the performance of a moment resisting STS assembly under cyclic load requires quantifying and recording the parameters that govern strength and stiffness in laboratory dynamic tests. Also, it is necessary to determine if current design methods and design equations yield acceptable results on capacity loads, yielding mechanisms, failure mechanisms and group effects for this connection type subjected to dynamic loading. This paper summarizes the obtained engineering parameters.  1.2. Significance The introduced construction material Cross-Laminated-Timber (CLT) and respective panel-to-panel and panel-to-beam connections were investigated under static loads. However very little research was published on these connections when subjected to dynamic load. With the presented work of this thesis a fundamental database for the performance of STS under dynamic loading conditions in typical CLT connections is presented.  3  Accurate characterization of the moment resisting STS assembly in terms of momentrotation response is an important prerequisite to reliably predicting the response of the structure to seismic events. Development of an advanced database for self-tapping screw assemblies and in particular moment resisting self-tapping screw assemblies under cyclic load currently exists neither in Europe nor in North America. The goal to achieve is to provide moment-resisting frames designed to allow columns and beams to deform elastically and connections to perform ductile with energy dissipation potential.  1.3. Thesis overview The literature survey in chapter 2 briefly reviews published research and current design recommendations for self-tapping screw assembled connections, reinforcements, testing procedures and calculation models. The following chapter 3 describes the experimental set up and also provides information about specimen geometry, materials used, and experimental procedures. The results and discussion section highlight observations made during testing and presents recorded data in detail. The performances under cyclic loading for the investigated STS assemblies are discussed. The last chapter summarizes the work conducted and conclusions are drawn.  4  2. Literature survey and background 2.1. Introduction to wood as a construction material Timber was generally available across the entire planet wherever humans settled and has been used to build shelters by most societies. The use of timber to build structures, machinery and boats has a long history and reaches back to the point where mankind first learned to manufacture tools. The wide use of this sustainable, natural material during the past millenniums has resulted in the development of great expertise and knowledge for the use of wood as a construction material (Kuklik, 2008). Wood as a natural, organic cellular solid is aesthetically pleasing with an excellent ratio between strength and density. Therefore, it has been considered as suitable building material in modern architecture and design. However, timber generally possesses properties which can cause difficulties if not properly considered in the design. Because timber is anisotropic with lower stiffness and strength perpendicular to grain than parallel to grain, larger shrinkage and swelling perpendicular to grain than parallel to grain and subjected to cracking, combustible and degradable by insects and fungi the public and many engineers are hesitant in using wood in large scale structures (Augustin, 2008). Most housing and commercial structures constructed in the 20 century in North America used wood as the major building material. This was because of large and easily accessible old-growth forests. Today, North American timber construction is still present in many single family dwellings, commercial and industrial buildings. Because of newly available wood products and construction techniques there has been increased interest in using 5  timber in commercial structures, including multi-story housing (mgb et. al., 2012). Typically, connections represent a major part of the overall strength and stiffness of such structures and have the ability to withstand short and long term loads. Long and short term loads directly affect the safety, reliability and durability of the structure especially when subjected to short term loads generated by seismic events. Past research conducted by Rainer & Karacabeyli, (1999) addressed the performance of typical wood-frame construction in a number of earthquakes. It was concluded that the majority of typically stick-frame houses subjected to peak ground accelerations of 0.6g and greater resisted the shaking without significant structural damage or collapse. This structural behaviour satisfies the life-safety criterion implemented in current building codes. In order to ensure satisfactory performance of new fastener materials and connection types in large and small timber structures and to further enhance the safety of timber structures, extensive research on new fastening systems has to be conducted. Continuing research and many centuries of experience in timber construction will help to overcome the remaining scepticism and build up further knowledge to design safe, aesthetical, functional and economical earthquake resistant timber structures of all types and sizes. Timber structures whether on their own or in combination with other suitable construction materials may become more and more important to sustain our planet as wood is a renewable building material.  6  2.2. Wood and its properties The part of primary interest to the structural engineer is the trunk of a tree since most structural elements are obtained by milling of the trunk. Generally, structural wood is obtained from either hardwoods (angiosperms) or softwoods (gymnosperms). In order to design timber structures economically some basic understanding of the wood from the trunk is a necessity for the structural engineer. Wood consists out of a chemical complex of cellulose, hemicelluloses, lignin and extractives. Because of the elongated shapes of wood cells, the oriented structure of the cell walls and also from variation of cell sizes throughout a growth season, wood is highly anisotropic. Also, it is of importance to know that wood is 20 to 40 times stiffer in its longitudinal direction than in its transverse direction. The macrostructure, i.e. growth irregularities, of wood however is responsible for the large variation in tensile strength. Many wood specific factors keep wood in the forefront of raw materials but its availability in many species, dimensions and aesthetically pleasing grain patterns and colour suit almost every demand. In addition to its durability and outstanding structural performance with a high strength to weight ratio dry wood has good insulating properties against heat, cold, sound and electricity. Wood can easily be shaped and fastened with nails, screws, bolts and glue. It is also of great advantage that wood generally resists oxidation, acid, salt water and that damaged wood can easily be repaired and remodelled. Furthermore, wood can be treated with preservatives and combined with many other materials for aesthetical purposes (Miller, 1999).  7  2.3. Moisture induced stresses in timber joints Because wood constantly exchanges moisture with its environment and variations in moisture within the timber can be significant during construction and service live, recent research (Sjoedin & Johansson, 2007), (Sjoedin et.al., 2004) addressed the influence of moisture induced stresses in steel to timber joints. This effect is important because the mechanical properties of wood are strongly affected by the moisture content. Their work considers multiple steel-to-timber dowel joints with exposure to controlled climate changes. Joints were assembled and then stored in the climate chamber for the first part of the study. The second part of this study considered specimen subjected to a controlled climate change before assembling. All specimens were tested to failure and the effects on load carrying capacity after an initial decrease in moisture are summarized. The research concluded that typical timber joints such as bolted joints with slotted-in steel plates can restrain swelling or shrinkage deformations of the timber when subjected to high humidity or dry climate conditions, respectively. Restraining occurs when the position of certain parts of the timber is fixed. Thus, internal stresses or unexpected movement of the timber may cause cracks in the timber and therefore, serious damage to the structure can occur. The effect of moisture induced stresses can be even worse if stress distribution caused by moisture variations interacts with stress distributions caused by mechanical loads. They concluded that moisture induced stresses can in fact decrease the load bearing capacity of dowel type joints where swelling and shrinkage is restrained by dowels and steel plates. Furthermore, a decrease in capacity was noticed when specimens were 8  exposed to an initial decrease in moisture level with no bolts, dowels or steel plates fitted to them during the drying process. From the results of this research and future research on this topic the issue of moisture induced stresses is conceivable to be considered in design codes.  2.4. Design of timber joints Given by natural restrictions, i.e. limited dimensions in length and cross section, the field of timber engineering is strongly affected by efficient, reliable jointing techniques so that the individual structural members can fulfill their intended function in resisting heavy loads. For instance, connections with dowel type fasteners commonly only provide 40% to 60% of the load carrying capacity of the members they connect (Werner, 1995). Thus, in many cases, the dimension of the structural element is governed by the required cross section determined by the design of the joint. Characteristic quantities to describe the behaviour of timber joints are, beside the ultimate capacity, the static ductility and the stiffness of the joint. These are typical parameters to describe the mechanical behaviour of joints and fasteners and their respective load-slipcurves (Figure 1).  9  Figure 1: Load curve and definitions  The static ductility Ds is defined as the ratio of displacement at failure load uu to displacement at yield load uy. If a joint behaves ductile, large deformations can occur before the ultimate load is reached. Brittle joints however, only have a small capacity for deformation before the ultimate load is reached and failure occurs. Generally, in earthquake prone regions, ductile behaviour of joints should be used to sustain load at large displacements without sudden collapse or deformity. Beside the ductility a further design parameter must be considered. This parameter is stiffness and defined as the flexibility of joints due to mechanical actions. Flexibility of joints influences deformations and therefore the serviceability of buildings. To describe stiffness, Figure 1 illustrates two important parameters, the slip modulus Kser and the slip modulus in the ultimate limit state Ku. Here, Kser is the slope of the load-slip-curve in the linear-elastic range and Ku is the secant to the ultimate load of the joint. The initial slip (Figure 1) which can occur in timber joints is the initial deformation independent of the applied load. Initial slip especially occurs in multiple fastener connections mostly because of manufacturing tolerances. Because production tolerances largely influence the 10  stiffness and ductility of joints, required tolerances should be as low as possible. Low production tolerances however will automatically increase the production cost and therefore a compromise has to be considered in the design process.  2.4.1. Ductility, stiffness and energy dissipation Wood shows brittle failure behaviour, with poor energy dissipation when compared to other construction materials such as steel and reinforced concrete due to its weakness in longitudinal shear and tension perpendicular to grain. This weaknesses can be pictured as a bundle of parallel straws representing the fibres, bonded together using a weak glue, with much higher stiffness and strength parallel to the wood grain than across the wood grain (Kuklik, 2008). These typical mechanical challenges must be considered before conceptually designing a joint to fulfill the main requirements such as:   Load carrying capacity    Stiffness    Ductility and energy dissipation  The load carrying capacity is the main criteria to be satisfied in the ultimate limit state design (ULS). The ULS proves that the connection can resist assumed external forces. Therefore, it is of importance to determine the mode of action, the direction of action and the duration of action. For each of those actions considered for the ULS design it is mandatory to take the following mechanical properties which significantly influence the load-carrying capacity of the connected members into account:  11    timber species    timber density    moisture content    possible wood defects located in the joint area    member cross-section (possible size effect)  Beside the consideration of ultimate load-carrying capacity serviceability which is greatly influenced by stiffness has to be taken into account to economically design timber joints. Deflection limitations largely contribute to the well-being of the user of the building. Timber structures are less stiff compared to other construction materials such as steel and reinforced concrete. Building codes however require the same deflection limitations for most construction materials and therefore stiffness mostly governs the design process for timber. Due to the fact that the stiffness of the joints significantly contributes to the structural deflections it is required to provide accurate stiffness information for practicing engineers. This is especially true when new connection and fastening system are introduced to the market. As mentioned before joints generally represent weaknesses in a structure and therefore should be able to tolerate large ductile deformations before failure occurs. Mischler, (1998) assigned significant influence of ductility to the structures overall performance, joint performance and the group performance of fasteners. For most connection types the timbers cross section is weakened to accommodate bolts, steel-plates or slender fasteners in bore holes or slots. Stress concentrations will form at the bore holes or slots because the flow of forces is disturbed and wood typically allows 12  for little stress relaxation only. Hence, a ductile performance of the structure can only be achieved by the connection itself and respective design considerations. In many cases, ductile design causes a decrease in connection capacity however, increases the ultimateload capacity of the structure due to possible force redistributions. Effects of ductility on the load carrying characteristics of joints made with metal-type fastener are most significant when stressed to their plastic limit (yield point). Plastic deformation of metal-type fasteners or connectors however takes place after larger elastic deformation has already occurred. Thus, the design of the connection has to allow for large deformations and ductile failure, for instance, splitting parallel to the wood grain, should not occur if plastic deformation of the connector is intended. Examples of three typically timber joints and their respective load curves are shown in Figure 2. The hatched area under the curves reflects the work required to cause failure of the joint. From Figure 2 it can be concluded that it will be more “work” for an extreme loading event i.e. an earthquake, to cause collapse of a structure consistent of ductile connections.  13  Figure 2: Typical examples of joints and respective ductility curves  2.4.2. The group effect Because of the ductile joint design requirement in timber engineering, fasteners with a low slenderness ratio are generally not favourable. Slenderness is the ratio between the actual thickness of the timber member and the diameter of the fastener. Fasteners with a low slenderness ratio tend to split the wood open perpendicular to grain and therefore, the related failure mode is brittle if no reinforcement is applied. In contradiction, if the slenderness ratio is high, the fastener will deform in bending and therefore, the joint capacity is governed by the bending-moment resistance of the fastener itself. Using many slender fasteners in a connection significantly reduces splitting and a higher load-carrying capacity and ductility can be achieved. Consequently, in order to achieve a ductile behaviour of the connection it is rather suggested to use many small diameter fasteners. 14  Using many small diameter fasteners instead of a few large ones however increases the amount of uncertainty when calculating the load-carrying capacity of the connection because more production tolerances and geometrical tolerances of the members are involved. In addition, density and moisture within the connection area are not constant. This can influence the load-slip behaviour and the ultimate local fastener capacity. In an optimal connection the ultimate load-capacity would equal the sum of the ultimate loads of the single fasteners. If large differences of the single fastener loads at failure of the connection can be found, some of the fasteners are loaded below their actual ultimate capacity. This results in an ultimate connection capacity smaller than the sum of the ultimate loads of the single fasteners. The effect causing this decrease in ultimate connection capacity is called group effect and is considered by design standards in North America such as the CSA, 2005, NDS, 20011 and in Europe the EC5, 2004 with a decreasing factor for the effective number of fasteners.  2.5. Timber moment joints under cyclic load The behaviour of a timber structure under cyclic loading is determined mainly by the behaviour of its connections joining the primary load bearing structural members. In North America, buildings are designed to resist seismic motions, known as “service” earthquake loads. The service earthquake is coupled to a Peak Ground Acceleration (PGA) having an average return period of 2475 years (NBCC, 2005). The requirements to make a building withstand the service earthquake conditions cannot limit its use, nor should serious deformation or significant damage occur as a result of a service earthquake. Building codes also require the structure to be strong enough to resist an 15  “ultimate” earthquake in which serious damage to the main load bearing structural elements will occur without complete collapse. If seismic events occur, the structure increases its own period of oscillation thus dissipating kinetic energy. Because of the low weight of the wooden structure, high specific strength and high potential of energy dissipation in mechanical connections wood structures have shown outstanding performances in seismic events. Typical beam-to-column connections consist of mechanical fasteners, such as dowels and bolts. The capacity of a doweled or bolted connection type is typically governed by   the dowels bearing strength    the diameter and yield strength of the bolt    the location of the fastener within the connection (end and edge distances).  Because of the material anisotropy of wood it is not possible to design dowel-type joints for unintended moments without reinforcement. New reinforcing techniques for doweltype connections strengthen the perpendicular to grain properties and may provide opportunities to design strong moment-resisting-timber connections. In addition, the goal of high ductility may be achieved.  2.5.1. Steel tube reinforced timber moment joints The performance of bolt-type connections is mainly governed by three parameters mentioned in section 2.5. To enhance the performance of the first parameter, the dowel bearing strength, Rodd & Leijten, (2003) investigated the impact of an overlength hollow steel pipe inserted into a slightly oversized dowel hole that is drilled through all members 16  of the assembly. The overlength of the pipe is used to apply a hydraulic jack which compresses the ends of the pipe hence the steel pipe expands in diameter until it fits the oversized dowel hole perfectly (Figure 3). Because of the expansion of the steel pipe the material surrounding the pipe becomes pre-stressed and therefore the initial stiffness can be increased by up to eight times when compared to a non-reinforced connection. To further increase the bearing strength and amplify the impact of the pipe hardwood or engineered wood products (EWP) can be sandwiched between the glulam lamellas.  Figure 3: Conceptual example of the expanded tube type reinforcement  The hollow steel pipe reinforcement with the EWP sandwiched between the glulam lamellas significantly increased the strength and stiffness of the joint. In fact, the reinforcement was so effective that no failure was observed in the joint but in the beam itself. The stress in the timber member reached about 35 MPa which. Compared to the traditional unreinforced joint where failure occurred at 10 MPa a factor of 3.5 can be calculated highlighting the impact of the reinforcement. The researchers also concluded that the variability of performance among the tests was decreased when compared to unreinforced specimens. Lower variability can be directly related to higher reliability of 17  the joint. Leijten et.al., (2006) reports similar observations with strong energy dissipation capabilities in heavy timber tube connections subjected to reverse cyclic load. Research (Heiduschke et.al., 2008) on laminated timber frames with moment resisting beam-to-column connections investigated the behaviour of small-scale and full scale frames under dynamic loading. Unreinforced and reinforced moment frames were subjected to seismic load and their performance was compared. It was found that densified wood and the fibre reinforcement in the connection zone significantly contributed to mitigate story drift while maintaining ductility and increasing energy dissipation of the connections. Haller & Wehsener, (2003) investigated the effects of densified wood in dowel type joints to increase the dowel bearing strength. They also report noticeable increase in stiffness and ductilty due to the increased bearing strength of the densified wood. A different approach of reinforcement was investigated by Heiduschke et. al., (2004) where reinforcing was done using glass-fibre oriented across the wood fibres with additional high density material in the dowel bearing zone. The glass-fibre reinforcement mitigated brittle failure which usually occurs because of perpendicular to grain splitting. The high density material in the dowel bearing zone was found to increase stiffness, energy dissipation and ductility of the connection. Kasal et.al., (2004) investigated the performance of heavy timber laminated frames Figure 4(a) under cyclic load where reinforcement was done with glass-fibre composite material and wood densification. The specimen used were European Spruce (Picea excelsa) with an average density of 0.44 g/cm3. Material densification was performed according to the method described in Haller & Wehsener, (2003). The densified wood is reported to have an average density of 0.9g/cm3. The results of this research show that locally reinforced 18  frames can withstand large seismic forces without significant damage. It was also reported that the frames behaved as a self-correcting system with the capability of large elastic deformation. They also observed significant energy dissipation due to elastoplastic deformations in the reinforced connections (Figure 4 (b)). In conclusion reinforced laminated timber moment frames have shown excellent performance when subjected to seismic loads. The reinforced beam-to-column connections were able to deform elastically, mitigating residual deformation and hence promoting the re-use of the structure after large seismic events. Because of the high embedment strength of the desified wood and the fibre reinforcement to prevent splitting perpendicular to grain it seems possible to apply stocky dowels. At high loads the thick, stocky dowels will form plastic hinges and therefore dissipate large amounts of energy because plastic deformation in the dowel embedment area is prevented. Because splitting of the wood and plastic deformation in the dowel embedment area is prevented yielding of the bolts and therefore ductile connection behaviour can be achieved.  19  Figure 4: (a) example of a beam-to-column connection type with densified wood in the dowel bearing zone, (b) schematic energy dissipation capacity between unreinforced and reinforced connection types.  2.6. Introduction to STS in timber construction Wood screws were developed to transfer forces perpendicular to their axis from one member to the other. In addition, the screws thread allows transferring tension and compression stresses parallel to their axis. In the past, wood screws were only available in small, short dimensions and therefore could not be used to transfer design loads in large dimension structural members. Today, wood screws with technical approvals and self-tapping screw tips are available up to 1400 mm length and 14 mm in diameter. The 20  self-tapping tip reduces splitting and the fully threaded shaft provides large withdrawal resistance. Typically these self-tapping screws are hardened after rolling on the thread to increase their tension strength, yield moment and torsional strength. The hardening process yields tension strength of 1400N/mm2 which in combination with the self-tapping tip results in a highly efficient fastener. Because of the large variety of screws in length and diameter on the market, the high withdrawal and push –in resistance STS have evolved as a multi-use tool in timber engineering. In Figure 5 a variety of self-tapping wood screws and their self-tapping tip is shown.  21  Figure 5: Variety of STS head shapes and tapping tip compared to drill bit  Primarily intended as reinforcement full thread STS are dedicated to be used in large scale timber structures today. The goal is to apply self-tapping screws not just as reinforcement in all kinds of timber joints but to utilize their capacity and potential as an active connector. In the past years STS have been used in timber to timber connections with only a few screws. Mainly, tension and compression forces were transmitted through the screws limiting their application to small timber joints. However recently, a new preengineered STS-assembly developed and patented in Germany (Blass, 2010) possibly allows for hundreds of screws in a connection and therefore these self-tapping screw assemblies may carry several mega newtons of design loads. 22  2.7. STS reinforced timber moment joints In previous chapters two reinforcement methods to avoid brittle connection behaviour have been briefly introduced and the most important findings were summarized. The following will introduce an entirely new approach to enhance the performance of bolted timber connections under earthquake loads using common full thread self-tapping wood screws. Experimental studies conducted (Lam et.al., 2007) examined the contribution of self-tapping screws as perpendicular to grain reinforcement for bolted glulam connections with slotted in steel plates (Figure 6). Experiments were conducted on a commonly used beam-to-column connection with 19 mm bolts subjected to monotonic and reverse cyclic load.  Figure 6: Connection layout examined by (Lam et.al., 2007®), adapted by permission  SWG-ASSY VG plus self-tapping wood screws with continuous threads were employed as reinforcement. The screws were 8 mm in diameter and 300 mm long. According to the technical approval, the yield strength of the screw steel is 330 N/mm2, and the tensile 23  strength is 410 N/mm2. The applied 8 mm screws provide 18.9 KN tensile strength with a characteristic torsional strength of 23 Nm. To compare improvements in performance static and dynamic tests were done on un-reinforced and reinforced specimen. The bolt diameter and therefore end and edge distances were kept constant and followed the requirements of the German timber building code DIN 1052:2004-08, 2004. Data analysis and observations made during the tests clearly indicated distinct improvement of connection behaviour when the reinforcement was applied. Unreinforced specimens always failed brittle and cracks formed at low rotation angles. Due to the early, brittle failure no yielding of the mechanical fasteners took place in un-reinforced specimen. Tests where reinforced specimens were subjected to reverse cyclic loading kept increasing their ultimate load in both, positive and negative cycles until the actuator stroke was fully utilized. A summary of the recorded moment capacities and rotations is shown in Table 1. Table 1: Summary of the test results from (Lam, et. al., 2007®), by permission  Max Moment [kNm] @ Rotation [°] Ductility Ratio [-]  Monotonic unreinforced Mean (STDV)  Monotonic reinforced Mean (STDV)  Cyclic unreinforced Mean (STDV)  Cyclic reinforced Mean (STDV)  31.49 (5.06)  65.88 (2.12)  35.70 (1.63)  62.54 (1.55)  2.97 (0.70) -  16.59 (0.06) >5.97 (0.62)  4.01 (0.17) -  15.90 (0.17) -  Further experiments on moment-resisting bolted timber connections using 25.4 mm bolts for the connection layout shown in Figure 6 and 25.4 mm bolts with reduced edge distances for the connection layout shown in Figure 7 were conducted by Lam et.al., 24  (2010). This research project considered two different connections with various reinforcement and bolt layouts. Lam et.al., (2007) used 19 mm bolts and 94.5 mm edge distance whereas Lam et. al., (2010) used 25.4 mm bolts and a reduced edge distane of 49.5 mm with two different STS arrangements (Figure 7). The thick, stocky dowels in combination with the reduced edge distances promoted splitting hence, increased the impact of the STS reinforcement however not improving the energy dissipation capacity. The ultimate moment resistance of the connection with 25.4 mm bolts and 49.5 mm edge distance was increased by a factor of 2.9 when compared to unreinforced specimens under cyclic load. An increase in stiffness by a factor of 2.4 was also achieved (Table 2). In summary, due to the reinforcement perpendicular to grain using full thread selftapping screws sudden failure at large rotations was prevented. The bolt holes remained intact to counter the bearing of the bolts and the structure would have been able to deform extensively before collapse occurs. Thus, indicating imminent danger to the occupants and leaving enough time for evacuation. Table 2: Summary of the test results from (Lam et.al., 2010®), by permission  Max Moment [kNm] @ Rotation [°] Elastic Stiffness [kNm/°]  Cyclic reinforced  Cyclic reinforced  Cyclic reinforced  (94.5 mm - 19 mm)  (49.5 mm - 25.4 mm)  (49.5 mm -25.4 mm)  Mean (STDV) 76.64 (3.35) 8.86 (1.99)  Mean (STDV) 103.83 (6.55) 5.64 (0.64)  Mean (STDV) 105.90 (3.66) 6.84 (1.13)  17.25 (2.65)  34.43 (3.65)  35.98 (1.38)  25  Figure 7: Tested connection layouts from Lam et.al., (2010®), adapted by permission. Left side: bolt layouts, right side: STS layout. All dimensions in mm.  Recently, Trautz & Koj (2009) investigated a new approach that utilizes the high axial strength and continuous bond of full thread STS to the wood to form a moment connection. The researchers use the screws to systematically reinforce the weaknesses of wood similar to the principles that are commonly used for concrete reinforcement. 26  Screws driven into the timber members form an internal truss system which transfers stresses through the timber and also function as a primary connector. During testing the moment mitre joint was subjected to positive and negative bending moments with three different screw configurations. When comparing the obtained test data to the calculated characteristic capacity of conventional timber joints such as glued finger joints or dowel rings it can be seen that STS moment mitre joints yield significantly higher capacities than conventional joints (Table 3). From Table 3 it can be seen that the moment mitre joint actually reached the bending capacity of the glulam member for negative bending moments. Trautz reports a bending fracture of the vertical connection member thus highlighting the high capacity of this assembly. For connection configurations containing less screws the typical failure mode was tension fracture of the screw steel. Because the performance and capacity of the screw steel is well known and therefore predictable a connection capacity estimate based on the developed strut-and-tie-model can be done. Furthermore, failure of the screw steel resulted in a small statistical spread for the ultimate capacities of this connection.  27  Table 3: Comparison among calculated characteristic capacities of conventional joints and 5%tile values of tested moment mitre joints according to Trautz & Koj, 2009®  Joint  Negative bending moments Force  beam  Moment  Positive bending moments Force  capacity  beam  Moment  capacity  [KN]  [%]  [kNm]  [KN]  [%]  [kNm]  Glued finger joint  Rc  25.59  37.8  27.38  5.12  8.2  5.48  Dowel ring  Rc  17.02  25.2  18.22  17.02  27.1  18.22  Screw joint  R0.05  59.04  87.3  63.18  43.60  69.5  46.65  Glulam member  Rc  67.62  100  72.35  62.71  100  67.10  2.7.1. Application possibilities for STS Because of intensive research and an increasing variety of STS available on the market many application possibilities have been developed in the past decade. To fully utilize the potential of STS in transferring loads from one member to the other it is preferred to load the screws parallel to their axis. This can be achieved with an inclined arrangement of the screws between 90° and 30° to the wood grain. In Figure 8 possible connections using a STS are shown. Typically this connection causes combined loading of the screw in tension and shear. With increasing inclination the tension stresses parallel to the screws axis will increase whereas the shear stresses will decrease. Because the withdrawal capacity of the screw is much larger than the shear capacity, the screw-in angle is of great importance. In addition, the tension force component in the screw will yield compression 28  forces perpendicular to the shear plane. This causes the friction at the interface of the connected members to increase and additional resistance parallel to grain is gained (Figure 8 bottom). Additional connection capacity to transfer forces parallel to the shear plane is gained when screws are applied crosswise. Since the push– in capacity and the withdrawal capacity of STS can be considered as equal, pairs of crosswise arranged screws are able to transfer tension and compression forces. The downside of this connection type is that no advantageous friction forces at the interface between the connected members can be developed. Figure 8 shows several possible connection types with regular and crosswise arranged STS for force transfer parallel to the shear plane.  29  purlin to rafter  beam extension or repair  shear connection Figure 8: Typical screw arrangements: purlin to rafter connection, beam extension or repair detail and shear connection.  Inclined STS can also be applied to transfer shear forces in post-to-beam connections. The force transfer through the crosswise arranged screws is independent of the bearing 30  situation of the beam (whether it is hinged or fixed). One screw will always transfer tension stresses and one compression stresses. For post-to-beam connections with only one screw the screw is stressed in tension and hence the withdrawal capacity may govern the ultimate capacity. This is similar to the connection shown at the top of Figure 8. Again, for this case, advantageous compression forces at the interface between the connected members occur causing the connection capacity to increase.  2.7.2. Alternative connection details with STS Further connections that can be done efficiently using STS are panel-to-panel shear connections (Figure 9) or panel to beam shear connections (Figure 10). With CLT-panel products being implemented on the North American timber construction market simple and efficient shear connections are needed and STS may be the fastener of choice for many engineers in the near future.  31  Figure 9: Panel to panel shear connection  Figure 10: Panel to beam connection  32  Particularly in post-to-beam connections with crosswise arranged screws (Figure 11) the screw arrangement is of importance. Moments caused by eccentricities shall be avoided as additional stresses on the fastener will occur. It is optimal to arrange the fastener intersection point at the shear plane between the post and beam member.  (a)  (b) optimal arranged STS  not optimal arranged STS  Figure 11: Typical post-to-beam connection: transferring shear forces and moments  A further application possibility using inclined STS is a rigid connection intended to transmit moments. This connection type may be used in a timber moment frame or as a rigid longitudinal beam joint (Figure 12). Typically, forces caused by bending stresses are transmitted by the screw parallel to their axis. Compression forces present in this type of connection may be partially transmitted by the wood and also parallel to the screws axis.  33  Figure 12: Typical rigid STS assembly: transferring moments  Further usage of STS can be made in regular column-to-beam connections and truss type connections where either tension or compression forces are intended to be transmitted.  Figure 13: Typical connection: transferring tension and compression forces  In summary one can say that STS are most efficient when forces are transferred parallel to the screw’s axis since large withdrawal and push-in resistance is provided by the screws thread. Therefore, the screw may always be arranged in force direction. Even considering this restriction almost unlimited use of STS can be ensured since it is easy to arrange this fastener in such a way that the forces mainly act in axial direction. 34  Applying full thread screws as reinforcement under an angle to the wood grain is one known method to strengthen bending stiffness and ultimate capacity of beams (Trautz & Koj, 2009). Truss principles are utilized to find the most efficient location and angle for the screw reinforcement. Figure 14 shows an example of a typical STS reinforced CLTpanel. The same reinforcement principle can be used to strengthen lumber, timber or glulam material.  Figure 14: Example of reinforcements in CLT  2.7.3. Available design equations for STS assemblies Design equations presented in this chapter are generally valid for the screw types listed below and may not be applicable for any other wood screw commonly used in timber construction (Blass et.al., 2006). In addition, reliable calculations using the presented design equations can only be made for limited penetration depths ls as outlined in the following. 35   Schmid Star Drive screw 6 mm x 100 mm with 20 mm ≤ ls ≤ 60 mm  Wuerth AMO III screw 7.5 mm x 182 mm with 40 mm ≤ ls ≤ 120 mm  Wuerth ASSY screw 8 mm x 340 mm with 40 mm ≤ ls ≤ 100 mm  Schmid Star Drive screw 10 mm x 200 mm with 40 mm ≤ ls ≤ 100 mm  ABC Spax-S screw 12mm x 420 mm with 40 mm ≤ ls ≤ 100 mm However, the design equations presented in this chapter may serve as estimates for a large variety of common wood screws in timbers with densities ρc ≤ 500 kg/m3. The design approach for screw-type fastenings used in timber construction differs from the design of timber members such as beams and columns. This is because the resistance of a particular fastener depends on a number of parameters. These parameters typically are:  screw diameter d in mm  wood density ρc in kg/m3  screw in angle α between the screw axis and the wood grain in degrees (°)  screw penetration depths lef in mm  characteristic withdrawal resistance parameter f1,k In the German timber building code (DIN 1052:2004-08, 2004), the design of screw-type connections is in accordance with the design procedure of nails. Therefore it was assumed that the embedment resistance of the wood (for small diameter fastener d ≤ 8 mm) is not affected by the thread of a screw-type fastener. Also it was assumed that the angle between the force direction and the wood grain does not have significant impact on the wood embedment resistance. The design for screw-type fastener with d > 8mm was 36  suggested to be done according to the design procedure of tight fitting bolts in pre-drilled holes. The design equation to determine the embedment strength of screw type fastener is presented in equation 1 ( )  Recent research (Blass, et.al., 2006) found that the embedment resistance of wood, when stressed by a screw thread, significantly differs from test data when estimated according to Equation 1. In particular for higher wood densities Equation 1 does not yield acceptable estimates for the wood embedment resistance when stressed by a screw thread. Therefore Equation 2 was developed from experimental data for more efficient design. [  ( )  ]  with: d = screw diameter in mm ρc = wood density in kg/m3 ε = angle between the screw axis and the wood grain in° Typically, European building codes such as the DIN 1052 and the EC5 consider the pushin and withdrawal resistance Rax,c of wood screws as equal for 45° ˂ α ≤ 90°. Parameters affecting Rax,c according to DIN 1052 are the screw diameter d, the effective penetration depth lef, the screw in angle α between the screw axis and the wood grain, and f1,k being the characteristic withdrawal resistance parameter depending on the wood density. Considering variations in axial push-in and withdrawal resistance of different screw threads available on the market, f1,k is divided into three groups. For thread group one f1,k can be calculated as  ( ) , for thread group two 37  ( ), and for thread group three  ( ). Additionally  head pull through, only applicable for partially threaded screws with head diameter dk, and the tensile strength of the screw steel must be taken into account. For head pull through DIN 1052 suggests f2,k to be divided in three strength groups. In strength group ( ) in strength group two  one  ( ) and  ( ) The characteristic withdrawal  strength group three  resistance Rax,c according to DIN 1052:2004:08 can be calculated.  {  }  ( )  Current research (Blass et.al., 2006) proved that the withdrawal resistance for screws applied at 90° to the wood grain could also be calculated using equation 10. √  (  )  (  )  Typically the withdrawal resistance declines for angles other than 90°. This effect is accounted for in design standards such as the DIN 1052 with the following equation.  From this equation it can be seen that the withdrawal resistance at α = 45° is approximately 90% when compared to a 90° screw-in angle.  38  Blass et.al., (2006) found that the equation provided by DIN 1052 is a conservative withdrawal resistance estimate for inclined screws. As a less conservative estimate the following equation has been proposed. √  (  )  For connections with crosswise arranged screws as shown in Figure 8 the capacity Rx per screw cross can be estimated based on Rax of equation 13. (  )  (  )  Stiffness estimation for STS assemblies with screws arranged at α = 90° using the following equation can be done.  (  )  [  ]  Due to the lack of data research has not yet derived an equation to accurately estimate stiffness’s for angles other than 90°. Blass et.al., (2006) however do point out that the slip-modul is not significantly affected by the screw-in angle and that further testing must be done to verify this statement. To determine the connection deformation at which the load on an axially loaded screw can no more be increased (limit state of diplacement δax,α) the following relation has been established. √  (  )  (  )  39  2.7.4. Design equations for tension and compression assemblies with ZD-plates To broaden the application of STS in timber construction design equations that estimate capacities of STS assemblies with tension and compression plates (ZD-plates) were derived by Blass (2010b). The following briefly presents the derived equations. The nature of the assembly yields an eccentricity (e). Based on this eccentricity a force (Ft,p,ZD) perpendicular to the shear plane can be estimated with the distance sZD being the distance between the outer most ZD-plates and (Fst) being the tension force applied to the connection.  (  )  (  )  The tensile force ( Ft,s) acting on the screws can be estimated with equation 18.  Compression stresses at the upper ZD-plate location will cause high transverse shear stresses (σc,perp.) that can be estimated with equation 19. In equation 19 (b) and (l) indicate the width and length of the ZD-plate. [  ]  (  )  Under the assumption of load sharing among ZD-plates in a connection and ineffective compression screws in the bottom ZD-plate the load carrying contribution for the 40  remaining compression screws can be estimated. Blass (2010b) suggests reducing the load carrying capacity of the compression screws by 30%. When adding the capacities of individual ZD-plates under consideration of the rope effect and a friction coefficient μ = 0.25 stated in the German timber design standard (DIN 1052:2004-08, 2004) a characteristic ZD-plate connection capacity with steel side plates is estimated with the following equation.  ( (  )  )  with: Rax,c =  characteristic withdrawal resistance of the full thread wood screw or the characteristic tensile capacity of the screw steel. The lesser of these values governs.  nZD =  number of ZD-plates in the connection  t=  steel plate thickness in mm  a1 =  ZD-plate spacing parallel to the shear plane in mm  Because of uplift forces which occur at the bottom ZD-plate sufficient fastening that resists a force (Ftp,screw,c) perpendicular to the steel plates axis must be applied.  (  )  (  )  If fastening which resists the force Ftp,screw,c is applied the connection capacity can be estimated with the following equation. 41  (  )  Figure 15: Assumed force distribution in ZD-plate connection  2.7.5. Spacing’s and end distances for STS The design resistance of fasteners provided in design standards is based on minimum spacing’s between fasteners and distances to the ends and edges of respective timber members. These recommendations are made in particular to avoid splitting but shall also reduce the risk of block shear failure and splitting when wood is not pre-drilled. In Table 4 a summary of commonly used abbreviations from European design codes describing spacing, edge and end distance requirements is shown.  42  Table 4: Spacing’s end and edge distance abbreviations used in design standards  Spacing type  Direction  Abbreviation  -Parallel to the wood grain  a1  -Perpendicular to the wood grain  a2  -Distance to the loaded end  a1,t  -Distance to the unloaded end  a1,c  -Distance to the loaded end  a2,t  -Distance to the unloaded end  a2,c  Spacing  End distance  Edge distance  a2  a1  a2,c α  α  a1,t  α  a2,t  a1,c  European design standards treat end and edge distances and also spacing’s between screws in the same way as nails in either pre-drilled or non-pre-drilled holes. Table 5 shows an overview for recommended end, edge and spacing requirements commonly used in European design standards.  43  Table 5: End, edge and spacing requirements according to DIN 1052: 2004-08  End, edge distance or  Non-pre-drilled holes applicable for characteristic densities of ρc ≤ 420 kg/m3  pre-drilled holes  spacing  d < 5 mm  d ≥ 5 mm  a1  (5+5 cos α) d  (5+7 cos α) d  (3+2 cos α) d  a2  5d  5d  3d  a1,t  (7+5 cos α) d  (10 +5 cos α) d  (7+5 cos α) d  a1,c  7d  10d  7d  a2,t  (5+2 sin α) d  (5+5 sin α) d  (3+4 sin α) d  a2,c  5d  5d  3d  Using engineered structural screws in timber connections with equal end, edge and spacing requirements as used for nails in non-pre-drilled or pre- drilled holes may seem inappropriate as screw diameters can be much larger than nail diameters. Larger diameter fasteners may split wood in heavily loaded connections. However, due to the large number of screw manufacturers with each having its unique screw tip and screw head design, general end, edge and spacing recommendations applicable to all available STS cannot be derived. The required amount of tests to gather enough data exceeds the capacity of research facilities around the world. Because of this restriction engineers are forced to use the conservative approach of nail end, edge and spacing requirements even though this approach may cause inefficient timber connection design.  44  2.8. Estimated connection capacities A recent study (Gehloff, 2011) on the withdrawal resistance of STS in major Canadian timber species proved the high withdrawal resistance of such fasteners and suggests that these connectors can be effectively used as reinforcement and primary connector. Gehloff suggests that the high withdrawal resistance of self-tapping screws is most efficiently utilized if installed in such a way that the main stress transfer is parallel to the axis of the fastener. When looking at the current Canadian timber design standard CSA O86-09 it becomes clear that no available design equation for mechanical fasteners can accurately predict the pull-out resistance of self-tapping wood screws. Design equations found provide very conservative capacity estimations and also fail to account for screw–in angles other than 90° to the wood grain. The same is true for the current National Design Standard of the United States (NDS 2005). Proposed equations found in the German timber design Standard DIN 1052 can, according to Gehloff (2011), conservatively but more accurately predict withdrawal capacities of STS. In addition, DIN 1052 incorporates the screw-in angle in design equations and therefore accounts for respective capacity decreases with increasing inclination of the screw.  2.8.1. Estimated connection capacity – shear connection Capacity estimations provided in Table 6 are based on equations presented in section 2.7.3. The density for the GG-series was selected to 490 Kg/m3 and 440 Kg/m3 for the GCLT-series in accordance with Table A10.1 of the Wood Design Manual 2005. To promote withdrawal resistance failure of the screw in the side member the effective  45  penetration depth, leff, was chosen accordingly as indicated in Table 6. The value of Rax was calculated using Equation 12 and Rx was calculated using Equation 13. Table 6: Estimated shear-connection capacities Series  Density [Kg/m3] 490 490 490  # of screw X  GG-4X GG-8X GG-4X-30  leff [mm] 113 113 110  8 16 8  Rax [kN] 15 15 15  Rx [kN] 21 21 26  Rxtot [kN] 174 349 210  GCLT-4X GCLT-8X GCLT-4X-30  141 141 200  440 440 440  8 16 8  17 17 23  24 24 32  195 390 261  Note: Grey shading indicates tensile strength of screw governs GG series – glulam to glulam connection with 4 or 8 screw crosses GCLT series – CLT to glulam connection with 4 or 8 screw crosses Screw-in angles were 45degree or 30 degree  2.8.2. Estimated connection capacity – moment connection When stresses between the ZD-plate and the wood member parallel to the shear plane occur two screws will be in tension and two screws will be under compression. Therefore, forces acting parallel to the screw axis can be determined using truss principles. Because of the design of the ZD-plate a gap of approximately one to two millimetres between the lid surface and the head of the screws exists. Due to this gab the compression screws do not contribute to the connection capacity before a certain deformation occurs. Before the compression screw head bears on the ZD-plate lid force equilibrium is established by a compression force at the interface between the ZD-Plate and the wood. After the ZD-plate is fully assembled in the connection a height of 27 mm is measured. This height causes eccentricity with respect to the force transmitted through 46  the steel plate and the distance to the ZD-Plate – wood shear plane. The moment caused by this eccentricity is assumed to be transmitted by forces perpendicular to the shear plane. The simplified principle force distribution in the assembly is shown in Figure 16. Figure 17 shows the assumed force distribution inside the ZD-plate under load in different  stages. Preliminary connection capacity estimates are based on research results presented by the University of Karlsruhe (Blass H. , 2010) and converted to the respective connection investigated in this paper. For these tests Blass reports an average connection capacity of 270 KN with three ZD-plates spaced at 100 mm in a shear connection. Considering the steel plate thickness (9.52 mm) and the beam member height (304 mm) a lever arm of 9.52/2 + 304/2 = 156.76 mm can be calculated. Based on these assumptions a moment capacity of 85 KNm was estimated.  Figure 16: Simplified assumptions for connection capacity estimate  47  From testing of 10 mm ASSY screws in Canadian Douglas-fir glulam with penetration depth of 160 mm an average tensile capacity of 30.9 KN was obtained from a total of 28 replicates. All of the 28 screws tested failed, reaching their tensile capacity.  Figure 17: Principle forces of the ZD-plate system. Before load application (a), after screw head engages with lid in compression (b), force perpendicular to shear plane resulting from eccentricity of the steel plate.  2.9. Loading procedure The capacity of a connection derived from static monotonic tests represents the upper boundary of the joint capacity. This is because most mechanical timber joints degrade in strength and stiffness and capacity when subjected to dynamic loading. To be able to establish the displacement limit upon which the dynamic loading protocol is based static tests are used. Dynamic tests represent an attempt to evaluate the joint performance by static deformation controlled tests. Mostly, in North America, the quasi-static cyclic testing protocol developed by the Consortium of Universities for Research in Earthquake Engineering (Fonseca et. al., 2002) is used to evaluate the performance of timber elements under cyclic load. In favour of other commonly used test standards such as the  48  European prEN 12512 or the American Society for Testing of Materials (ASTM) a detailed description of the utilized CUREE procedure is provided in the following.  2.9.1. CUREE loading procedure – shear connection CUREE developed a quasi-static cyclic testing procedure based on ground motion recordings from real earthquakes in California. This displacement controlled test method was developed based on the statistical analysis of seismic demands on light-frame constructed California buildings. The loading procedure starts with six equal initiation cycles at small amplitudes that serve to check the testing apparatus and the load displacement response at small amplitudes. Cycles following the initiation phase consist of primary cycles and trailing cycles. For calibration of the loading sequences a fraction of the monotonic connection capacity ∆ultimate is taken as the reference deformation (∆). Typically in cyclic tests loss of strength from cumulative damage can be expected. Therefore, it is suggested to take ∆=0.6·∆ultimate to account for these potential differences in load displacement behavior between monotonic and cyclic tests. Each new displacement phase starts with a primary cycle followed by a trailing cycle with 75% the amplitude of the respective primary cycle. Initiation cycles with amplitudes of 0.05·∆ followed by the first primary cycle with an amplitude of 0.075·∆ represent the initial loading. Following this, six trailing cycles with amplitudes of 0.75· (0.075·∆) occur. Subsequent primary cycles following the 1.0·∆ increase their amplitude by 0.5·∆ with six trailing cycles following. For trailing cycles 49  following 0.2·∆ a slight modification to the original CUREE loading procedure was applied to subject the hardened screws to higher stresses. Instead of three trailing cycles six trailing cycles followed each primary cycle until connection failure occurred. A detailed listing of applied amplitudes in trailing and primary cycles is shown in Table 7. No loading suggestions are given in the CUREE loading procedure but can be obtained from ISO standards. A loading rate of 14 mm per minute for the GG-4X series, CLTG4X, GG-8X series and CLTG-8X series was selected to achieve specimen failure after approximately 20 minutes. Therefore, several dummy tests prior to specimen testing were made to adjust the loading procedure and achieve failure within a reasonable time frame. The reference deformation value ∆ for shear connection tests was calculated from the measured average deformation ∆ultimate of the three dummy tests. From the dummy tests an average ∆ultimate of 11.2 mm was obtained hence a reference deformation ∆ = 0.6 * ∆ultimate = 6 mm was selected (Figure 18).  50  Figure 18: Modified CUREE loading protocol (shear connection)  51  Table 7: Cycle amplitudes for shear connection tests  Cycle Description  Amplitude  6 initiation cycles  0.05∆  Amplitude in [mm] 0.3  1st primary cycle  0.075∆  0.45  6 trailing cycles  (0.075∆)·0.75  0.33  2nd primary cycle  0.1∆  0.6  6 trailing cycles  (0.1∆)·0.75  0.45  3rd primary cycle  0.2∆  1.2  6 trailing cycles  (0.2∆)·0.75  0.9  4th primary cycle  0.3∆  1.8  6 trailing cycles  (0.3∆)·0.75  1.35  5th primary cycle  0.4∆  2.4  6 trailing cycles  (0.4∆)·0.75  1.8  6th primary cycle  0.7∆  4.2  6 trailing cycles  (0.7∆)·0.75  3.15  7th primary cycle  1.0∆  6  6 trailing cycles  (1.0∆)·0.75  4.5  8th primary cycle  1.5∆  9  6 trailing cycles  (1.5∆)·0.75  6.75  9th primary cycle  2.0∆  12  6 trailing cycles  (2.0∆)·0.75  9  10th primary cycle  2.5∆  15  52  2.9.2. CUREE loading procedure –moment connection For the moment connection assembly the original CUREE loading procedure (Figure 19) was applied. Table 8 highlights details of the applied loading procedure. To achieve connection failure after approximately 40 minutes a loading rate of 45 mm per minute was selected. The reference deformation ∆ultimate for the moment connection was calculated from two dummy tests with an average displacement of the actuator of ∆ultimate = 86 mm. Because movement of the test rig and specimen fixture was expected during cyclic loading a reference deformation of ∆ultimate = 100 mm was assumed to appropriate. The reference displacement ∆ was therefore selected to be 0.6 * 100 mm = 60 mm.  Figure 19: CUREE loading protocol (moment connection)  53  Table 8: Cycle amplitudes for moment connection tests  Cycle Description  Amplitude  6 initiation cycles  0.05∆  Amplitude in [mm] 2.7  1st primary cycle  0.075∆  4.05  6 trailing cycles  (0.075∆)0.75  3.037  2nd primary cycle  0.1∆  5.4  3 trailing cycles  (0.1∆)0.75  4.05  3rd primary cycle  0.2∆  10.8  3 trailing cycles  (0.2∆)0.75  8.1  4th primary cycle  0.3∆  16.2  6 trailing cycles  (0.3∆)0.75  12.15  5th primary cycle  0.4∆  21.6  2 trailing cycles  (0.4∆)0.75  16.2  6th primary cycle  0.7∆  37.8  2 trailing cycles  (0.7∆)0.75  28.35  7th primary cycle  1.0∆  54  2 trailing cycles  (1.0∆)0.75  40.5  8th primary cycle  1.5∆  81  2 trailing cycles  (1.5∆)0.75  60.75  9th primary cycle  2.0∆  108  2 trailing cycles  (2.0∆)0.75  81  10th primary cycle  2.5∆  135  54  Past research has not addressed the issue of dynamic loads on connections where STS are used as primary fastener. Currently designers must use engineering judgement to predict the performance of screw connections under dynamic loads. The experimental work presented in the following chapters addresses this issue and provides a fundamental database for STS connections under dynamic loads.  55  3. Experimental procedures and materials 3.1. Introduction The first part of this research project was to investigate if self-tapping wood screws can perform well (high connection capacities) as primary connector under dynamic load. The connection layout tested used standard timber materials that are available on the market. The second part was to examine the performance of a moment resisting self-tapping screw assembly under cyclic load. Rigid moment resisting timber joints are crucial in large scale structures but not often investigated under cyclic loads. The assembly tested employs components which are readily available on the market. Steel plates in the dimension used can be obtained from many steel yards and further processed with standard machinery. The loading procedure in either of the test series followed the CUREE reverse cyclic protocol described previously. The following chapter describes the varying parameters that this study has taken into account to determine the connection performances under dynamic load.  3.2. Specimen identification For shear connection specimen identification during testing and data analysis the following nomenclature was developed. The nomenclature consists of three parts, for instance, “GCLT1-4X-30”. The first section (GCLT1) refers to the timber members and actual specimen number of this series. The first letter (G) indicates material of the centre member, here Glulam. The following letters (CLT) refer to the site plate material, here cross-laminated-timber. At the end of the first section (1) indicates the specimen number 56  of this test series. Each test series consisted of 5 specimens. Section two (4X) of the nomenclature refers to the number of screw crosses on each side. In nomenclature section three (30) the screw-in angles other than 45° are indicated. Glulam CLT  4 pairs of STS at 45°  Figure 20: Specimen identification  The nomenclature used for the moment resisting assembly for instance “Mc-1” is as follows. The letters “Mc” refer to moment connection and the respective number “1” to the actual specimen number.  3.3. Timber members The timber materials CLT and Douglas-fir glulam used for this study are locally available and of commercial importance in the near future.  57  3.3.1. Shear connection specimen The shear connection consisted of one centre piece and two side plates. The selected side plate material was either Canadian Douglas-fir glulam or Canadian made pine-crosslaminated-timber. The G-side members were cut from Douglas-fir 24f-E grade glulam beams that were available in the UBC Department of Wood Science lab. To ensure consistent specimen properties, only the lower quality D-graded lamellas from the centre of the beam were used. Typically the middle layers of glulam beams are intended to resist high shear stresses. Particularly high shear stresses parallel to grain were expected in the centre piece of the GG-4X-30 test series. Therefore, cutting needed specimens from only D-graded lamellas was thought to be appropriate. The dimensions of each specimen member, its weight and moisture content was recorded shortly before assembly and testing. The recorded average specimen properties obtained from these measurements are summarized in Table 9. Table 9: Average recorded moisture content (MC) and density of shear member specimen Series  GG-4X GG-8X GCLT-4X GCLT-8X GG-4X-30 GCLT-4X-30  Left member MC [%] 11.2 10.9 10.6 10.2 10.6 9.9  Density [Kg/m3] 525 550 447 505 583 433  Centre member MC [%] 10.7 10.9 11.4 11.1 11 11.6  Density [Kg/m3] 546 566 548 556 551 559  Right member MC [%] 11.3 11 10 10.4 10.8 10  Density [Kg/m3] 554 553 457 475 568 463  The GG-4X series specimen and GG-8X series specimen were assembled from glulam material with the dimensions shown in Figure 21. Two screw crosses and respectively  58  four screw crosses consisting of 8 mm x 260 mm ASSY VG screws were driven in prior to testing at an angle of 45º to the wood grain.  59  Figure 21: GG-4X and GG-8X series specimen (dimensions in mm)  60  The GG-4 X-30 series specimens were assembled from glulam material with the dimensions shown in Figure 22. Two screw crosses consisting of 8 mm x 300 mm ASSY VG screws were driven in prior to testing at an angle of 30º to the wood grain. To allow for similar embedment lengths of the screw inside the side member among GG-4X specimen and GG-4X-30 specimen the side member thickness was reduced from 80 mm to 57 mm.  Figure 22: GG-4X-30 series specimen (dimensions in mm)  For the GCLT-4X series, GCLT-4X-30 series and GCLT-8X series side member and centre member dimensions slightly varied from specimen dimensions used for the GGseries. The side member thickness could not be altered or adjusted to allow for equal embedment depth of the screw among the 45º and 30º screw-in angle due to the predetermined thickness of the available 3-ply cross-laminated-panel. ASSY VG fasteners 61  with a dimension of 8 mm x 300 mm were used for the GCLT-4X and GCLT-8X specimen whereas the GCLT-4X-30 specimen were assembled with an ASSY VG 8 mm x 340 mm. Typically full thread screws do not draw connection members tight together and a gap is likely to develop at the interface between the side and centre member during assembly. To ensure a gap free assembly two 8 mm x 200 mm washer head screws were used to draw the connection members tight prior to the application of the full thread screws. After the full thread screws were driven-in on both sides the washer head screws were removed. Figure 23 and Figure 24 provide detailed information about tested GCLT specimen.  62  Figure 23: GCLT series specimen (dimensions in mm)  63  Figure 24: GCLT-30 series specimen (dimensions in mm)  3.3.2. Moment connection specimen Glulam beams of grade 24f-E in dimensions of 130 mm x 912 mm were cut to the needed dimensions in a way that only D-grade lamellas were obtained. A common cross-section for glulam material in Canada is 304 mm x 130 mm therefore, the glulam material was cut to this dimension. For the horizontal member of the assembly an average density of 540 Kg/m3 with 11.1% moisture content was recorded prior to testing. The vertical member averaged a density of 532 Kg/m3 with 11.2% moisture content. The moisture content of each connection member was recorded using a Delmhorst 2-Pin moisture meter with 25 mm long pins. To fit the required steel bracket and ZD-plates appropriately onto a specimen, extensive milling work had to be done to the vertical and horizontal members. The available Hundegger K2 fully automated joinery machine in the Centre of Advanced Wood 64  Processing (CAWP) was used to cut the specimen with the required accuracy. Also the Hundegger K2 was used to cut specimens to the required length of 830 mm for the vertical member and 950 mm for the horizontal member.  Figure 25: Required millwork for specimen assembly  3.4. Required hardware For this research project several custom made steel-brackets/steel-shoes were needed to allow testing of the shear and moment connection. All steel plates were ordered from a local steel shop and custom cut, drilled and welded at the timber engineering lab.  65  3.4.1. Steel plates Steel plate stock material was ordered in two different dimensions to suit the needs for the custom made steel brackets. The outer, vertical parts of the steel bracket were made from 101.6 mm x 6096 mm x 9.525 mm A-36 Hot Rolled Mild Carbon Steel. Horizontal steel parts of the bracket were made from the same material but in raw material dimension of 203.2 mm x 6096 mm x 9.525 mm. All required holes were drilled using a 17 mm diameter drill bit. Further specification of the steel bracket is shown in Figure 26.  66  Figure 26: Custom made steel-bracket (dimensions in mm)  67  3.4.2. Self-tapping screws Full threaded ASSY plus VG screws with self-tapping tips in various lengths and diameters were employed to assemble all connection layouts. The screws were kindly donated and shipped to Canada by the Schraubenwerk Gaisbach Gmbh (SWG) in Germany. ASSY plus VG wood screws are available in various dimension ranging from 6 mm x 80 mm up to 14 mm x 1400 mm. The manufacturing standards for structural wood screws require the steel wire material to be of high grade and strength. ASSY screws are typically manufactured using a steel wire material with 330 N/mm2 yield strength (Rp,0.2) and must provide an ultimate tensile strength (Rm) of 410 N/mm2. For 8 mm ASSY screws the characteristic tensile strength (Rt,u,k) is 18.9 kN with a characteristic torsional strength of 23 Nm and a characteristic yield moment (My,k) of 16.7 Nm. ASSY screws with a diameter of 10 mm yield a characteristic tensile strength of 24 kN with 45 Nm characteristic torsional strength and a characteristic yield moment (My,k) of 35.8 Nm. Further characteristic strength values can be obtained from the technical approval Z-9.1-614 (DIBt, 2006) if required. For assembly of the glulam to glulam (GG-series) shear connection, only 8 mm x 260 mm screws were used, whereas the CLT to glulam (GCLT-series) shear connection employed 8 mm x 300 mm and 8 mm x 340 mm screws for 45° and 30° screw-in angles, respectively. Different screw lengths were used to allow for screw penetration depth sufficient to reach the tensile capacity of the screw. A definition for length and diameter of the screws is shown in Figure 27.  68  Figure 27: Dimension of the employed fasteners from SWG  3.4.3. Tension and compression plates To establish a connection between the steel shoe and the vertical timber member through self-tapping wood screws the tension and compression plate (ZD-plate) developed by the German screw manufacturer Schraubenwerk Gaisbach Gmbh (SWG) was employed. This ZD-plate is made from two 86 mm x 55 mm steel parts with yield strength ≥ 550 N/mm2. One part (“base”) is pre- installed to the vertical member using four 10 mm by 240 mm ASSY VG plus full thread screws arranged at an angle of ± 30° to the shear plane. Two screws are intended to transfer stresses in tension and two are intended to transfer stresses in compression. The second part (“lid”) is pre-installed using two 4 mm x 10 mm pins that are driven into holes that penetrate the lid and partially penetrate into the base. These pins temporarily fix the lid to the base until a permanent connection between the steel shoe and the ZD-plates is established using a M16 - 10.9 metric bolt on site. The M16 bolt must have a minimum length of 35 mm. The lid ensures transmission of the compression stresses from the screws back into the ZD-plate. The dimensions of the tested ZD-plates are shown in Figure 28.  69  Figure 28: ZD-plate (Blass, 2010b)  70  3.5. Test setup All experiments of this research project were conducted at the University of British Columbia Timber Engineering and Applied Mechanics test facility.  3.5.1. Equipment shear connection For required load application to the shear connection specimen two different actuators were used. One actuator was used for the GG-4X, GG-4X-30, GCLT-4X and GCLT-4X30 series and a second stronger actuator was used for GG-8X and GCLT-8X series to provide sufficient push and pull capacities. Also, to allow for appropriate spacing’s and therefore avoiding unintended fastener interaction different specimen heights and therefore different actuator mounting heights were needed. Because of limited pull capacities of the strong actuator and lack of sufficiently long screws no experiments with 8 screw crosses per side arranged at 30 degree were conducted. Additionally, parallel to grain shear failure of the glulam specimen was already expected for the GG-4X-30 specimen making it unnecessary to further investigate GG-8X series or GCLT-8X series without reinforcement perpendicular to the grain. Multi-purpose Test Software (MTS) actuators were used for loading of the shear connection layouts. The front base of the actuator was mounted to a solid 5cm thick steel plate and attached to 2-U380 horizontal steel beams of the test frame. The actuator displacement was controlled by MTS FlexTest GT Software Version 3.2C-970. A total of seven transducers with at least 25.4 mm stroke were used to record specimen movements during testing. Four transducers were mounted at the front side of the specimen to record expected separation of the side members from the centre member. Two additional transducers were mounted at the bottom of the centre 71  piece to record the up and down motion of the centre piece. In addition, the actual actuator-head displacement was recorded. Due to the high loads that were expected one transducer was also mounted on top of the horizontal steel beam fixture to record occurring deflections in the negative cycle. During data analysis these deflections were subtracted from the measurement obtained from the two bottom transducers to eliminate errors in stiffness calculations in the negative cycle.  72  transducer locations  Figure 29: Shear test rig front view  73  Cross bar deflection measurement  Figure 30: Shear test rig side view  74  transducer Figure 31: General shear connection test setup and transducer positions  To attach the actuator head to the top of the shear connection specimen the custom built fixture shown in Figure 32 was used. This device consisted of two 19 mm thick steel plates with four 16 mm threaded rods tying them together. By placing one steel plate at the bottom and one on the top of the specimens centre piece push and pull forces could be transmitted. To allow for free rotation of the specimen around the horizontal and vertical axis two hinges were built into the fixture. The first hinge that allowed rotation around the vertical axis was formed at the connection point of the fixture to the actuator head (Figure 32). The design of the thread and the respective counterpart allowed for specimen rotations of up to 180° around the vertical specimen axis. Expected rotations were however much lower. A second hinge that allowed for rotations around the horizontal axis was formed by the 38 mm pin and the respective counterpart (Figure 32). 75  Figure 32: Shear connection actuator load application device  Because the specimen was subjected to cyclic load, specimen uplift was expected to occur in the negative cycle. To avoid any uplift of the specimen from the concrete floor a steel frame was anchored to the concrete utilizing the 31.75 mm threaded rods (Figure 33). These threaded rods were fitted through pre drilled holes in doubled-up 76.2 mm x 50.8 mm x 6.35 mm hollow steel bars that clamped down the 101.6 mm x 101.6 mm x 9.52 mm hollow cross bars. Before each test the specimen was leveled and the cross bars 76  were lowered down to the surface of the specimen’s side members to provide uplift resistance.  31.75 mm threaded rod  Figure 33: Clamping detail of steel cross bars  3.5.2. Equipment moment connection A MTS actuator with 240 kN pull and 265 kN push capacity was used to subject the STSmoment assembly to cyclic load. A general test set-up is shown in Figure 36. The horizontal actuator was mounted to a vertical steel column that was rigidly fixed to the strong concrete floor and braced diagonally for additional stiffness. To allow for free rotation about every axis a ball joint fixture was used at the rear end of the actuatorcolumn connection point. The horizontal member of the assembly was tied to the strong concrete floor using 31.75 mm threaded rods and two metal bars with dimensions of 76.2 mm x 50.8 mm x 6.35 mm (Figure 34). To avoid horizontal movement of the specimen parallel to the actuator axis an adjustable steel fixture mounted on a steel beam that was fixed to the concrete floor using the existing 31.75 mm floor- holes (Figure 34) was used.  77  Figure 34: Steel bars stretching over horizontal connection member (left), fixture to avoid horizontal movement (right).  A total of five displacement recordings (see Figure 36) using cable extension position transducers and linear position transducers were done during each test. The displacement measurements were used to calculate horizontal force and displacement vectors that eventually return values of the vertical member’s rotation. One transducer was mounted vertically below the actuators front end to record the up and down movement and allow horizontal force and displacement corrections. A total of two measurements were recorded on the right side of the vertical member to return values of the movement of the top and bottom of the vertical member. These measurements also allowed for calculation of the vertical members rotation that was needed to derive moment rotation plots.  78  Figure 35: Vertical actuator movement recording (left), rotation recording (right)  79  point of displacement recording  Figure 36: General test set-up for moment connection tests  80  80  4. Results and discussion 4.1. General overview Within this research project, a total of 30 shear connection tests with inclined self-tapping wood screws (STS) in Canadian Douglas-fir glulam and Canadian CLT were evaluated. In addition, a total of 10 moment resisting STS assemblies were tested under dynamic loads. The following sub chapters present observations made during testing and evaluate the recorded data in terms of connection performance, stiffness, ductility and energy dissipation. Furthermore, the analysis also includes information about measured densities of each connection member and moisture content recordings. During testing of the moment connection a slight horizontal movement of the specimen in positive cycles and negative cycles was observed. Because this movement occurred suddenly some moment rotation plots show small load drops followed by subsequent load increases. It is assumed that these slight horizontal movements are due to the oversized holes that were used to bolt the steel fixtures to the concrete floor. Holes on this fixture were oversized by about 2 mm to allow for easy assembly and alignment of the holes.  4.2. Shear connection During the negative cycle of the shear connection tests the horizontal cross bar deflected due to the high loads. Even though this deflection was measured and results were 81  corrected forces recorded in the negative cycle are generally lower as forces recorded in the positive cycle. This may be caused by slight movements of the threaded rods that tied the steel cross bars to the concrete floor. Furthermore, the deflection of the heavy steel reaction beams was neglected and may have caused additional measurement errors that contributed to the lower ultimate loads and stiffnesses in the negative cycle. Furthermore, the positive cycle which always occurred before the negative cycle may have caused some embedment damage or fastener fracture which ultimately leads to a decreased load in negative cycles. Because of a generally lower load in the negative cycle all shear connection specimens ultimately failed during the positive cycle. For better visualization of the plots shown in this chapter and the appendices axis scales may vary and may therefore not be directly comparable.  4.2.1. GG-4X series and GG -8X series A total of five shear connection specimens of each layout were tested under dynamic loading using the CUREE loading procedure. The maximum connection capacities in positive and negative direction along with respective maximum displacements are summarized in Table 10. The most apparent failure mode observed was a sudden tension fracture of screws followed by a steep load drop in the positive and negative cycle. In addition, screw withdrawal and/or head pull –in occured at the side members of the assembly and forming of plastic hinges on the fastener was observed after the specimens were dissassembled (Figure 37).  82  Figure 37: Typical tension fracture of screws, pull-in failure and screw bending  In all five tests failure occured first during the positive cycle. In the following negative cycle a slightly smaller capacity was reached followed. Except for specimen GG1-4X all specimens failed in cycle # 49. Looking at Figure 38 the cause of the sudden load drops due to tension fracture of the screws in positive and negative cycles can be seen. In progressive positive cycles a gab opened at the interface between the centre member and the side member (Figure 38). During negative cycles this gab always closed up entirely.  83  side member  loading plates  transducers  centre member  developed gap  Figure 38: Typical example of connection failure and specimen separation  In Figure 39 the steep load drop in the positive and negative cycle can be seen. A steep load drop was expected due to tension failure of a number of screws on either one or both sides of the specimen. In none of the tests did fatal failure of screws occur on both sides. After the specimens were disassembled it was common to observe damaged screws only on one side of the specimen. In most cases the undamaged screws or slightly bent screws could be removed. It is assumed that these undamaged screws were still able to transfer loads. Furthermore, the high stiffness of the assembly and the mostly sudden failure mode did not allow for determination of a connection yield point because the curve did not flatten enough to match the second tangent line to it.  84  GG3-4X  -6  -4  -2  180 Load in [kN] 160 140 120 100 80 60 40 20 0 2 -20 0 -40 -60 -80 -100 -120 -140 -160 -180  4  6  Displacement in [mm]  Figure 39: Typical load displacement curve for GG-4X series  When comparing the recorded data with previous research (Blass, et.al, 2006), very similar oservations can be made even though static loading according to DIN EN 26891 was applied. The research project conducted in Germany uses similar connection layouts and fastener arrangements with an average timber density of 457 Kg/m3 and averages capacities and stiffnesses per screw cross and shear plane of 21.9 KN and 16 KN/mm respectively. Comparing these average values to the data recordings of this connection under dynamic load with cpacity and stiffness per screw cross and shear plane of 19.32 KN and 7.30 KN/mm one can see the impact of the dynamic load. A capacity decrease of about 13% and a related stiffness decrease of 55% under dynamic loading conditions is  85  observed. In Table 10 recorded connection capacities are shown and measured moisture contents along with wood densities are presented in Table 22. Table 10: Recorded connection capacities and stiffness for the GG-4X series Positive cycle  Specimen  GG1-4X GG2-4X GG3-4X GG4-4X GG5-4X Average  Negative cycle  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  [KN]  [KN/mm]  [KN]  [KN/mm]  18.8 17.0 19.8 19.5 21.3 19.3  8.1 5.7 7.5 7.6 7.3 7.3  -16.5 -15.8 -15.8 -17.0 -16.7 -16.3  4.7 9.6 4.8 5.7 9.7 6.9  The load displacement curve in Figure 40 for the GG-8X series shows the difference between the data recordings in the positive cycle and data recordings in the negative cycle. It appears that the capacity and stiffness of the connection during the negative cycle decreases more when compared to the plot shown in Figure 39. This may be due to generally higher loads and therefore larger deflections of the test rig and loading equipment at progressive cycles. Because of that capacities and stiffnesses per screw cross of the negative cycle may not be appropriate for comparison. When comparing the typical plot of the GG-4X series (Figure 39) to the GG-8X series (Figure 40) similar performance in the positive cycle can be seen. Both connection layouts show significant load drops caused by tension fracture failures of single or  86  multiple screws. In subsequent cycles the connection can again transmit large loads as the stresses are redistributed among remaining fasteners.  Load in [kN]  260 230  Load drop after screw tension failure  200 170  Load level after failure  140  GG1-8X  110 80 50 20 -8  -6  -4  -2  -10 -40  0  2  4  6  8  Displacement in [mm]  -70 -100 -130 -160 -190 -220 -250  Figure 40: Typical load displacement curve for GG-8X series  For the GG-8X series an average connection capacity per screw cross and shear plane of 15.32 KN in the positive cycle and -14.24 KN in the negative cycle was recorded. A stiffness per screw cross of 17.17 KN/mm was recorded in the positive cycle. Table 11 highlights the previously mentioned large variation in terms of stiffness in the positive and negative cycle. Blass, et.al, (2006) reports a small capacity and stiffness decrease for a similar connection layout with multiple screw crosses under static load. Comparing this observation to Table 11 these observations can not be confirmed under dynamic loading conditions. Blass reports an average capacity of 21.3 KN per screw cross and shear plane 87  with a respective stiffness of 20.8 KN/mm. Under dynamic loading conditions a decrease in capacity of approximately 30% and respectively 20% in stiffness is observed for positive cycles. In terms of failure mode comparison among the two studies more tensile fracture failures of the fasteners under dynamic loading conditions are observed. This may be due to the generally higher density of the Canadian timbers and the resulting higher withdrawal resistance at equal penetration depths. Table 11: Recorded connection capacities and stiffness for the GG-8X series Positive cycle  Specimen  GG1-8X GG2-8X GG3-8X GG4-8X GG5-8X Average  Negative cycle  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  [KN]  [KN/mm]  [KN]  [KN/mm]  15.6 14.0 15.4 16.0 15.4 15.3  15.3 15.3 17.2 14.9 23.0 17.1  -13.5 -12.9 -14.8 -14.8 -15.1 -14.2  1.8 3.4 2.1 2.2 2.5 2.4  Because of the high repetitive loads applied in the positive cycle, a further failure mode at the bottom of the side members was observed. This failure was caused by splitting perpendicular to grain or parrallel to grain toward the the end of the test protocoll (Figure 41). During testing of the GG-8X specimen the same gap as already described for the GG-4X series formed between the side and main member. In some occasions, the gap only formed at the bottom or top resulting in a slight inclination of the side member. This inclination caused a force component perpendicular to the wood grain that resulted in 88  splitting of the wood. After specimen dissasembly typical screw failures such as tension fracture failures on either both or only one specimen side, forming of plastic hinges or withdrawal resistance and/or push out failures of the fastener were observed (Figure 41).  splitting  Figure 41: Specimen separation (top left), typ. screw failures (top right) and splitting  After each specimen was tested and dissassembled with hand machinery the failure mode of each fastener was recorded. Table 12 summarizes observed failures of the fasteners. In all of the test specimen tension fracture failure of the screws occured whereas head push 89  out was only observed in two specimen. Screw bending and withdrawal type failures were seen in half of the specimen tested. Shear failure parallel to the wood grain in the centre member was not expected in this test series and was not observed during testing.  90  Table 12:Recorded observations on screws and wood members after specimen dissasembly for the GG-4X and GG-8X series Left side member  Specimen  Screw tension failure  Screw withdrawal  Screw bending  Centre member Screw tension failure  Screw push out  GG1-4X  Right side member  -  4  GG2-4X  4  -  4  GG3-4X  2  -  4  GG4-4X  2  GG5-4X  1  GG1-8X  4  GG2-8X  4  2  -  3  4  GG5-8X  4  4  -  1  3  -  4  -  8  -  8  -  3  3  1 4  Screw push out  2 2  Note: The numbers 1, 2 ..., refer to number of observations made on each failure mode  91  91  Screw bending  -  GG3-8X GG4-8X  Screw withdrawal  3  3  4.2.2. GCLT-4X series and GCLT -8X series A total of 10 glulam to CLT-shear connection specimens were tested under dynamic load using the CUREE loading procedure. The maximum recorded connection capacities per screw cross for the GCLT-4X series and GCLT-8X series and respective recorded displacements are listed in Table 13. The large displacement differences between specimens may have been caused by varying densities of the CLT side members and therefore smaller withdrawal resistances. A further reason for reduced withdrawal resistances in CLT may be caused by gaps inside the panel material which were penetrated by the screws. At 90% of the specimen a screw tension fracture failure occurred among screw withdrawal, head push out and fastener yielding. The tension fracture failure of the screws mostly resulted in a sudden load drop with little remaining connection capacity. Because of the generally lower density of the CLT side members in combination with the smaller screw penetration depth more withdrawal and head push out failures were expected when compared to the side members of the GG-4X and GG8X series (Figure 42). As it can be seen in Figure 42 noticeable fastener yielding along with intense wood crushing occurred. The intense fastener yielding occurred because of additional bending stresses caused by the eccentricity from the shear plane to the location of screw intersection (Figure 24). After approximately one third of the test protocol a gap formed between the side and centre member. In general the gap formed between the members of the GG series seemed to be smaller when compared to the GCLT series. This may be due to the occurrence of more significant screw withdrawal resistance failures in the side member of the GCLT series (Figure 42). 92  Figure 42: Withdrawal, push-out failure (top), tension fracture, fastener yielding (middle), specimen separation (bottom)  When looking at the load displacement curve shown in Figure 43 the performance similarities among the GG-series and GCLT series can be compared. It can be seen that a steep load drop related to the tension fracture failure of the screws occured in the positive and negative cycle. The main difference between positive and negative cycles can be found in the secondary load resistance after the first failure occured in positive cycles. 93  Typically the secondary load after the first failure occured reached approximately 45% of the maximum load. Secondary load resistances indicate a load redistribution to remaining fasteners.  Load in [kN]  180 150 120  GCLT3-4X  90 60 30 0  -8  -6  -4  -2  0 -30  2  4  6  8  Displacement in [mm]  -60 -90 -120 -150 -180  Figure 43: Typical load displacement curve for GCLT-4X series  Comparing the results obtained from glulam to glulam connection testing (Table 10 and Table 18) to results obtained from CLT to glulam connection testing (Table 13 and Table 20) similar performances in terms of connection capacity per screw cross and shear plane and stiffness per screw cross and shear plane can be seen. The decrease in stiffenss may have resulted from the previously mentioned bending stresses that were caused by the 94  slight ecentricity of the intersection point of the screws to the shear plane. One can assume that for the same reason smaller capacities and stiffnesses were recorded during the negative cycles. In addition, once a screw is withdrawn from a piece of wood a significant resistance reduction occurs. Because many withdrawal failures occured in the CLT members a certain stiffness reduction can be assigned toward this failure mode. Furthermore, the generally lower density wood with lower embedment strength also contributed to a stiffness reduction. During progressive cycles yielding of the fastener crushes wood fibres and a cavity is formed. Inside this cavity the fastener is no longer supported by the wood fibres. Hence a stiffness reduction occurs as resistance is only provided by the fastener itself. Table 13: Recorded connection capacities for the GCLT-4X series Positive cycle  Specimen  GCLT1-4X GCLT2-4X GCLT3-4X GCLT4-4X GCLT5-4X Average  Negative cycle  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  [KN]  [KN/mm]  [KN]  [KN/mm]  17.4 17.4 17.3 18.8 18.7 17.9  5.0 5.0 7.2 8.6 7.7 6.7  -16.8 -16.6 -16.1 -16.3 -16.7 -16.5  5.0 2.4 1.7 4.4 5.0 3.7  A typical load displacement plot of the GCLT-8X series is shown in Figure 44. Examining this plot yields performance details such as steep load drops caused by tension 95  fracture failures of the screws and very little energy dissipation. Also, the secondary load resistance after failure occured up to approximately 50% of the ultimate capacity.  Load in [KN]  300 250 200 150 100  GCLT3-8X  50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250 -300  Figure 44: Typical load displacemnt curve for GCLT-8X series  Comparing the average capacities per screw cross and shear plane from CLT to glulam connection with 4 screw crosses (Table 13) to capacities recorded for CLT to glulam connections with 8 screw crosses (Table 14) a slight decrease in capacity can be seen. This observation is consistent when compared to the results of the glulam to glulam connection tests. In Blass, et.al, (2006) a capacity reduction factor of 14% for 7 screw crosses arranged in a row is suggested. Taking the presented values into account a 96  capacity decrease of approximately 20% for glulam to glulam connections was calculated. A capacity decrease of approximately 8% was calculated for the CLT to glulam connections. Table 14: Recorded connection capacities for the GCLT-8X series Positive cycle  Specimen  GCLT1-8X GCLT2-8X GCLT3-8X GCLT4-8X GCLT5-8X Average  Negative cycle  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  [KN]  [KN/mm]  [KN]  [KN/mm]  16.1 17.1 16.3 16.8 17.3 16.7  4.9 37.4 16.8 12.5 14.5 17.2  -11.5 -17.6 -17.3 -17.1 -17.1 -16.1  7.0 3.9 4.0 3.6 4.7  After testing each specimen was disassembled and visually inspected for failures. Table 15 presents a summary of recorded failures such as tension fracture of the screws, fastener yielding and withdrawal resistance failures. As it can be seen no withdrawal resistance failure occured at the centre member which consisted of the higher density material along with larger penetration depth. Two out of the five tested specimen in the GCLT-8X series showed parallel to grain shear failure at high loads. Total separation was however prevented by the heavy steel plate on the top and bottom of the centre member. Furthermore, splitting failure of the side member perpendicular to the wood grain as described for the GG series occured. Also, noticeable shear failure at the glue line in the CLT side member was observed (Figure 45). 97  Figure 45: Typical splitting of CLT at the bottom of the side member under large loads  98  Table 15: Recorded observations on screws and wood members after specimen dissasembly for the GCLT-4X and GCLT-8X series Left side member  Specimen  Screw tension failure  GCLT1-4X  2  GCLT2-4X  2  GCLT3-4X  2  Screw withdrawal  2 2  GCLT5-4X  2 4  GCLT2-8X  1 1  4  3  -  4  -  2  -  1  2  -  2  -  4  Parallel to grain shear failure  2  Screw withdrawal  Screw bending  Screw push out  2 2  2  1  2  2 2  2  -  GCLT3-8X  8  -  GCLT4-8X  8  -  GCLT5-8X  8  Parallel to grain shear failure  Note: numbers 1, 2 ..., refers to numbr of observations made on each failure mode  99  99  Right side member Screw tension failure  Screw push out  2  GCLT4-4X  GCLT1-8X  Screw bending  Centre member  2 1  1  4.2.3. GG-4X-30 series and GCLT -4X-30 series To evaluate the influence of different screw-in angles on the performance of a shear connection under dynamic loading a total of ten tests were performed. Five tests with glulam side members and five tests with CLT side membes where screws are arranged on an angle of 30º to the wood grain are evaluated in the following. Generally higher connection capacities and also connection stiffnesses were expected. Due to the large loads the glulam centre member of the assembly was however prone to shear failures parallel to the grain. Reinforcement with full thread wood screws to avoid this failure mode was not applied. As it can be seen in Figure 46 shear failure parallel to grain always occured right at the screw tip. This indicates that stresses occuring inside the wood are effectively transferred along the screw‘s axis back into the wood. This can yield to high shear stress concentrations in one plane and potentially cause brittle failures parallel to grain. Recorded capacities and displacements per screw cross for glulam to glulam connections with screws arranged at 30° are presented in Table 16  100  Figure 46: Typical shear failure parallel to grain and expected point of stress concentration at the screw tip for GCLT specimen  101  Load in [kN]  250 200 150  GG3-4X-30  100 50 0  -6  -4  -2  -50  0  2  4  6  Displacement in [mm]  -100 -150 -200 -250  Figure 47: Typical load displacement curve for GG-4X-30 series  Figure 47 shows similar connection performances as observed for the tests with 45º screw arrangements. Again a secondary resistance after an initial steep load drop is recorded. Failure modes found after specimen disassembly are similar to which were found in previous test series. Among tension fracture failures of the screws, fastener yielding and withdrawal resistance or push-out failure and parallel to grain shear failure was observed at the centre piece. Because the centre member was clamped tight between the steel plates required for load application the actual shear failure may not be reflected in the load displacement curves and derived strength values. Comparing the average capacities per screw cross and shear plane between the GG-4X series with 19.32 KN and the GG-4X-30 series with 25.10 KN and stiffness’s of 7.30 KN/mm and 10.88 KN/mm in the positive cycle the influence of the 30º screw arranged 102  can be seen. As expected, an increase in capacity and stiffness was observed due to a decrease of the force component perpendicular to the fastener axis. However, yielding of the fastener was still observed in approximately half of the removed screws (Figure 48).  Figure 48: Typical fastener yielding for the GG-4X-30 series and GCLT-4X-30 series  A generally lower capacity of the GCLT-4X-30 series may be caused by the parallel to grain shear failures that occured in four out of five specimen whereas in the GG-4X-30 series only one parrallel to grain shear failure occured. Comparing the capacities between 45º screw arrangement and the 30º screw arrangement an increase in average maximum capacity of 30% can be calculated. In terms of average stiffnesses no distinct increase can be seen.  103  Table 16: Recorded connection capacities and stiffness for the GG-4X -30series Positive cycle  Specimen  GG1-4X-30 GG2-4X-30 GG3-4X-30 GG4-4X-30 GG5-4X-30 Average GCLT1-4X-30 GCLT2-4X-30 GCLT3-4X-30 GCLT4-4X-30 GCLT5-4X-30 Average  Negative cycle  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  Capacity per screw cross and shear plane  Stiffness per screw cross and shear plane  [KN]  [KN/mm]  [KN]  [KN/mm]  27.2 22.9 24.9 23.8 26.5 25.1 19.4 25.4 25.0 20.2 17.7 21.5  8.5 11.9 11.2 10.1 12.4 10.8 9.4 10.1 10.7 9.1 10.1 9.9  -20.4 -20.5 -23.0 -21.7 -22.2 -21.6 -19.2 -18.8 -16.8 -12.0 -13.1 -16.0  6.1 4.5 7.4 9.5 9.0 7.3 10.2 10.1 6.9 4.9 12.3 8.9  4.3. Data analysis – shear connection The performance of the shear connection assembly in terms of stiffness, ductility and energy dissipation was derived from the recorded load displacement relations. For accurate stiffness estimation of the connection recorded displacements of the actuator head were corrected with the respective cross bar deflection measurements (Figure 30). After these corrections were applied respective load displacement curves were developed. Statistical values such as standard deviations (Stdev) and coefficients of variation (COV) were calculated from the available data of each test. A test was stopped if the load  104  dropped below 80% of the highest recorded maximum of previous cycles or if sensitive and expensive equipment such as transducers were in danger.  4.4. Moment connection A total of ten tests using the tension and compression ZD-plates with STS as primary connector were conducted to evaluate the performance of the moment resisting assembly under reverse cyclic load. The set-up of the connection was selected in a way that failure in the screws was promoted. This was done to create the basic database which then later can be used to improve the connection performance in terms of ductility and energy dissipation. Being able to accurately predict the capacity of the primary connector of this assembly allows scientists and engineers to purposely weaken the steel members accordingly to promote more ductile connection performances. The moment resistance of the STS assembly in the positive cycle ranged between 69.34 KNm and 108.41 KNm with respective rotations of 2.84° and 2.99°. During the negative cycle maximum moment resistances ranged between -80.87 KNm and -101.82 KNm at rotations of -3.24° and -8.46° respectively. Because of little ductility in some of the specimen yield moments and respectively ductility ratios were not not derived. During testing failure modes observed were tension fracture failures of the STS in the beam member, excessive steel plate yielding at the bottom of the assembly and wood crushing along with transverse shear failure at the column member. Beside these major failure modes wood crushing at several locations of the beam and column member at points of load application was observed. Wood crushing on the beam member 105  perpendicular to the grain at the point of load application was apparent in nine of ten tests. An influence on recorded test data due to slight shifting of the load application fixture after wood crushing occured was not found.  Figure 49: Observed wood crushing at the upper end of the beam member  Tension fracture of the screws occured in 5 of the 10 experiments causing brittle failure modes with a sudden load drop. Typically the first tension fracture was observed in the top ZD-plate which always showed signs of high occuring stresses. After failure occured in the top ZD-plate sudden fracture progressed downward toward the bottom of the beam member. In general the lower ZD-plate showed only little signs of high stresses and failure of the fastener possibly only occured because of sudden load redistribution after 106  other fasteners failed. Along with fracture of the tension screws significant yielding of the compression screws was observed. Partially the fasteners were also withdrawn from the wood (Figure 50). In a report about the performance of ZD-plates under static load (Blass, Pruefbericht Nr. 106109, 2010) damage on the ZD-plates in terms of bending deformation is reported. This observation was not confirmed during testing under dynamic load. A possible reason may be the shorter length (86 mm) of the ZD-plates employed compared to the plates used in (Blass, Pruefbericht Nr. 106109, 2010) with 100 mm length. The only damage visually found at the ZD-plate was a slight indentation on the lid. This indentation was caused by the screw head of the compression screw (Figure 50).  107  Figure 50: Yielding of compression screws and fracture of tension screws (top), screw withdraw on all three ZD-plates (bottom).  A further indicator of high stresses at the ZD-plates were the deformed stocky M16 (10.9) machine bolts which were used to connect the steel shoe to the ZD-plates. Along with bending slight damage on the bolts thread was found. In some tests a gap formed 108  between the M16 bolt head and the steel plate (Figure 51). This happened because of little vibrations and slight, sudden motion of the entire specimen during testing. The sudden motion occured because of oversized holes that were required to fit the bolts for the test rig assembly. Once the force exceeded the friction sudden movement of the entire test rig of about 1 mm occured. This behavior may have caused the bolts to loosen up even though they were tightened to a pre set torque of 200 Nm. In a real structure vibrations also occur over time which may loosen bolts as well. Therefore it is suggested to assume extra measures to secure the bolts in place tightly.  Figure 51: Bolt push-out and bending  Due to the layout of the STS assembly moment rotations are counteracted by a tension force on one side of the specimen and a compression force on the other side. In three tests excessive wood crushing and transverse shear failure in the column member regardless of the reinforcement applied perpendicular to the grain occured. Partially this failure was introduced by high compression stresses below the steel shoe on one side and compression stresses under the hold down fixture on the other side. Therefore the column  109  member was continously subjected to high compressive stresses and eventually failed in three tests (Figure 52).  Figure 52: Combined compression and transverse shear failure at column member.  During progressed load cycles wood crushing under the hold down steel pipe and also deflection of the steel pipe caused a gap to develop between the bottom of the column member and the strong concrete floor (Figure 53). This gap measured between 10 mm and 15 mm. Potentially some influence on the rotation of the beam member may be found in slightly larger rotation recordings. During data analysis the influence of these errors was neglected.  110  Figure 53: Wood crushing under hold-down steel pipe (left), developed gap at bottom of column member (right)  Beside the described failures which occured on the wooden members of the assembly failures were also observed on the steel shoe and bottom steel plates. Plastic deformation of the steel plate was observed at two locations. In all tests plastic deformation of the side steel plates occured at the top ZD-plate and near the stiffening bracket at the bottom. Deformations of up to 15 mm were measured at both locations (Figure 54). In addition, screw withdraw from the column member at maximum load was observed along with perpendicular to grain splitting of the column. It is assumed that occuring tensile stresses are transfered along the screws axis to the screw tip. At the screw tip these stresses are now highly concentrated and transferred back into the wood. The poor strength of wood perpendicular to the grain causes splitting failure which eventually leads to a failure of the connection (Figure 55).  111  Figure 54: Plastic deformation of steel shoe side plates after testing  Figure 55: Splitting of column member perpendicular to the grain  112  In one test a single 12.5 mm bolt used to connect the steel shoe to the bottom steel plate failed in tension and eventually stripped the thread of the bolt placed on the opposite side. Both failures now caused significant plastic deformation and yielding of the bottom steel plate along with withdrawn screws (Figure 56).  Figure 56: Steel plate yielding and screw withdraw at bottom steel plate  When examining the moment rotation plots shown in Figure 57 and Figure 58 the two major, previously descibed failure modes can be distinguished among the plots. In Figure 57 the typical tensile fracture of screws is shown. Clearly it can be seen that a steep load drop after the maximum load was reached occurs. In the negative cycle a few minor load drops can be seen. These minor load drops were caused by tensile fracture of single screws in the respective cycle or shifting of the assembly plates. The fracture was clearly noticeable by a loud snapping noise typical for hardened steel. As it can be seen load redistribution among the screws occured allowing the connection to maintain its capacity 113  or even increase the moment resistance. Early fracture of screws may also be initiated by overtightening of the fastener during assembly. As a matter of fact fasteners in steel to wood connections shall always be tightened to a specified torque to avoid high pretentioning stresses and micro-cracks. In Figure 58 a typical moment- rotation plot derived from a specimen with intense wood crushing and transverese shear failure in the column member is shown. Wood crushing is indicated by the slow load decrease in the negative cycle. Because of the high stresses transverse shear failure occured in the column member. The impact of this failure mode was clearly noticeable through loud cracking noises during testing. Sudden cracking of the wood is also reflected in the plot through small load drops in the negative cycle where the curve flattens. The positive cycle in Figure 58 again shows the impact of a tensile fracture of one or more screws at ultimate load. A sudden load drop followed by a continous load decrease can be seen.  114  Figure 57: Typical moment rotation plot with tensile fracture of screws followed by a steep load drop. Yield point Connection failure  at ultimate load  Figure 58: Typical moment rotation plot with intense wood crushing at the column member followed by screw withdraw.  115  In Table 23 a summary of all recorded test values in terms of strength, stiffness and yield point is provided. In connection test Mc-1 an early failure caused by sudden fracture of screws in the positive cycle is found. After specimen dissasembly it was found that the early failure was caused by tensile failure of two screws. Because the load dropped to less than 80% of the previously reached maximum load the recording software reversed the load and stopped loading after reaching the maximum load in the negative cycle. A possible explanation for the early failure may be found in large overtightening stresses which may have occured during assembly. After the first test all screws in the remaining specimen were carefully tightened with a high torque – low rpm drill to avoid overtightening and therefore premature failure. Table 17 reveals an average moment resistance of 93.33 KNm in the positive direction and an average moment resistance of -89.81 KNm in the negative direction. Comparing these results to the specified design capacity of the tested 130 mm x 304 mm Douglas-fir 24f-E glulam beam the high performance of the STS assembly is highighted. During the positive cycle the connection exceeded the factored bending moment capacity of the beam by a factor of 1.7 and by a factor of 1.6 in the negative cycle.  116  Table 17: Statistics summary of recorded data Positive cycle  Negative cycle  Mean  Stdev.  COV  Mean  Stdev.  COV  max M. [KNm]  93.33  12.38  0.13  -89.81  7.08  -0.08  rotation @ max M. [°]  2.92  1.02  0.35  -3.62  4.17  -6.72  failure M. [KNm]  74.66  9.90  0.13  -71.84  5.66  -0.08  rotation @ failure M. [°]  3.56  0.36  0.10  -3.66  0.86  -0.24  40% max M. [KNm]  37.33  4.95  0.13  -35.92  2.83  -0.08  rotation @ max M. [°]  1.30  0.31  0.24  -1.33  0.45  -0.34  yield M. [KNm]  96.62  9.68  0.10  -84.58  9.26  -0.11  rotation @ yield M. [°]  2.64  0.52  0.20  -2.65  0.47  -0.18  stiffness. [KNm/°]  32.34  7.50  0.23  32.46  8.42  0.26  Throughout the entire test series moisture content measurements were taken during specimen assembly shortly before testing. In addition each specimen was measured and weighted to determine its individual density. Due to the large size of the specimen the recorded densities do not reflect possible local variations which may have influenced the withdrawal resistance and finally the connection capacity. The recorded moisture contents and densities shown in Table 24 shall only be considered as an approximation. In Figure 59 a selection of the calculated viscous damping ratios is shown. As it can be seen some recordings yielded significantly higher damping ratios than normal. These extreme values can possibly be explained due to the slight shifting of the specimen in the test rig and slip in the connection. It is assumed that whenever the load was reversed from pushing in the positive cycles to pulling in the negative cycles slight shifting occured. This shifting may have caused the visible high peaks in the equivalent viscous damping ratio plots. Because the layout of the connection was chosen to collect information about the ultimate capacity and performance of the assembly the issue of energy dissipation 117  was not addressed. In general little energy dissipation can be expected due to the high stiffness and little ability of the assembly to yield.  Equivalent viscous damping ratio  0.7  0.6  0.5  0.4  0.3  0.2  0.1  0 0  5  10  15  20  25  30  35  40  45  50  55  60  65  70  75  80  85  90  Half cycles  Figure 59: Examples for calculated equivalent viscous damping ratios  4.5. Data analysis – moment connection The structural performance of moment resisting timber connections is commonly derived from moment rotation relationships. Moment rotation relationships provide information about the connection stiffness, ductility and energy dissipation. To obtain these relationships the recorded load and displacement data had to be converted using geometric measurements from different locations of the specimen. The horizontal force 118  vector was calculated at the point of load application on the vertical member. Using the actual applied horizontal force the moment at the lower end of the vertical member was calculated. Displacement values obtained from the cable extension transducer mounted at the point of load application and displacement values obtained from the linear position transducer mounted at the bottom of the vertical member were used to calculate the column member rotation with respect to the ground. To receive the envelope curve for the cyclic test data an available FORTRAN software tool was used. Descriptive parameters such as yield moments (yield M), maximal moments (max M) and failure moments (failure M) were derived based on the envelope curve and are summarized in Table 23. In addition, the corresponding rotations, elastic stiffnesses and ductility ratios were derived where possible. The potential for energy dissipation and the equivalent viscous damping ratios were calculated using FORTRAN software tools and followed the procedure outlined in (EN12512:2001, 2001). Statistical values such as Stdev and COV’s were calculated from the data of each test.  119  5. Conclusions 5.1. General The primary focus of this investigation was to investigate the performance of self-tapping wood screws in different connection systems subjected to reverse cyclic load. Two representative connections systems were chosen. The shear connection system tested represents a strong alternative to commonly used connections for the sector of panel-tobeam and panel-to-panel connnections. The selected moment resisting assembly provides an alternative to bolted timber connections or post base fixtures. To simulate extreme loading conditions, each connection system was loaded using the CUREE cyclic testing protocol. In both cases monotonic test data was used to calibrate the displacement controlled loading history. From recorded test data design parameters such as strength, stiffness and ductility were derived were applicable. This data can be used to compare the performance of self-tapping screw assemblies under reverse cyclic loading to data bases created under monotonic load.  5.2. Shear connection Results obtained from this study indicate that self-tapping screws can provide outstanding performances under reverse cyclic load and extreme loading conditions. The high pull out resistance provides resistance against high stresses and ultimately leads to high connection capacities with a small number of slender fasteners. To achieve maximum efficiency in a connection the STS shall be installed in a way that the majority of stresses 120  are transferred parallel to the screws axis. Basic information for real connection alternatives in terms of strength, stiffness and efficiency are now available to the engineer. From the results presented for STS shear connections, it can be seen that the capacity and stiffness of self-tapping screw connections decreases when subjected to reverse cyclic loading. In average an ultimate capacity decrease of approximately 20% among the GG4X series and GG-8X series when compared to static loading conditions was found. Similarly a stiffness decrease of approximately 30% was observed under reverse cyclic load. When comparing the connection capacity among the connections with 4 screw crosses (4X series) and 8 screw crosses (8X series) a general capacity decrease of approximately 20% was found. Similar observations were confirmed in STS connections under static loading conditions with a decrease in capacity of approximately 10%. Under dynamic loading conditions generally lower connection capacities for STS shear type connections can be expected and shall be taken into account during design. In addition it seems that the effective number of fasteners per row further decreases under dynamic loading conditions when compared to static loads. Explanations for the capacity and stiffness decrease under dynamic loading conditions may be found in the previously described gap which formed between the main and side member during testing. Due to this gap the fastener was subjected to larger bending stresses which in combination with the cyclic loading and back and forth bending may have weakened the steel. The large amount of cycles which the STS’s were subjected to  121  may have also contributed to the generally lower capacities due to the many up and down bending occurrences which typically weaken hardened steel. Comparing the average capacity and stiffness between the glulam to glulam connection with screws arranged at 45° (GG series) and the glulam to glulam connection with screws arranged at 30° (GG-30 series) an increased capacity of approximately 30% and stiffness of approximately 50% was observed. The larger capacity of the steeply arranged screws can be explained with a smaller shear component acting on the screw. A smaller shear component causes less bending stresses and generally little plastic deformation occurs as stresses are primarily transferred along the screw axis. A generally weaker performance of the connection in the negative cycle is observed. This is because of the different properties of the test rig in positive and negative direction. It can however be assumed that the assembly will perform equally in both directions under more consistent test conditions. Because of the high stiffness and literally no fastener yielding or wood crushing little potential for energy dissipation in the shear connection was expected. Therefore the issue is not addressed for the purpose of this paper. Respective data was however recorded and will be presented in future papers. When designing shear connections with screws arranged at 30 degrees to the wood grain the issue of parallel to grain shear failure must be addressed. Efficient solutions to avoid this failure mode are STS reinforcements. As experienced during testing shear failure parallel to grain governed the ultimate capacity of a few specimens.  122  When designing shear connections with screws arranged on a 30 degree angle it is crucial to check the shear capacity parallel to the grain as this may govern the design. Typically shear failures parallel to grain are considered brittle failure modes and shall be always reinforced using full thread screws in shear connections of that kind.  5.3. Moment connection The high performance of the STS assembly under reverse cyclic load was confirmed. High stiffness and large moment resistance was confirmed throughout the entire test series. Consistent performance of the assembly through an average of approximately 40 consecutive cycles was observed. In general a predictable failure mode and failure capacity was derived from moment rotation plots. When comparing performances among moment resisting bolted moment connections and the assembly tested with this research significant stiffness improvements were observed. The STS assembly exceeds the stiffness of the bolted connection (see section 2.7) by a factor of 2.3. The potential for energy dissipation and ductility was not improved. With the basic database created from this research the ultimate capacity for moment resisting STS assemblies can now be estimated. The aim would now be to purposely weaken steel members, forcing them to yield and therefore optimize potential energy dissipation and ductility. Following this approach cost efficient and safe connections in heavy timber structures assembled with STS can be used.  123  Due to the occurrence of possible high stresses parallel to the grain, sufficient reinforcing full thread screws should be applied to eliminate brittle parallel to grain shear failure in wood.  5.4. Recommendations Future research on moment resisting self-tapping screw assemblies shall address the issue of energy dissipation and ductility for this connection type. It is of great interest to the engineering community to allow for more ductile connection performances under extreme loading conditions. One may purposely weaken the steel shoe to a level below the capacity of the self-tapping screws and tension and compression plates. Forcing the thick steel plate to yield will increase the amount of dissipated energy and therefore contribute to a more ductile connection performance.  124  BIBLIOGRAPHY Anderson, G. T. (2001) Experimental investigation of group action factor for bolted wood connections. Blacksburg: Virginia Polytechnic Institute and State University, USA Augustin, M. (2008) Handbook 1 - Timber Structures (First ed.). Educational Materials for Designing and Testing of Timber Structures - TEMTIS Blass, H. (2010) Expert Report to the Bearing Performance of ZD-Plates as Connectors in Timber Construction. Karlsruher: Research Institute for Steel, Timber and Stone of the Karlsruhe Institue of Technology Blass, H. (2010b) Pruefbericht Nr. 106109 Versuche zur Ermittlung des Trag- und Verformungsverhaten von Biegesteifen Rahmenecken mit ZD-Platten. Karlsruhe: Karlsruher Institut fuer Technology. Blass, H., Bejtka, I., & Uibel, T. (2006) Tragfaehigkeit von Verbindungen mit selbstbohrenden Holzschrauben mit Vollgewinde. Karlsruher Berichte zum Ingenieurholzbau 4, Universitaet Karlsruhe, Germany. CSA (Canadian Stadards Association) Standard (2009). CSA 086-09: Engineering Design in Wood. Canadian Standard Association, Mississauga  125  DIBt (Deutsches Institut fuer Bautechnik) (2006) Allgemeine Bauaufsichtliche Zulassung, Wuerth ASSY VG plus Vollgewindeschrauben als Holzverbindungsmittel. DIBt, Berlin DIN (Deutsches Institut fuer Normung e.V) (2004) DIN 1052:2004-08: Entwurf, Berechnung und Bemessung von Holzbauwerken- Allgemeine Bemessungsregeln und Bemessungsregeln fuer den Hochbau. Beuth Verlag, Berlin DIN (Deutsches Institut fuer Normung e.V) (1991) (ISO 6891:1983) Holzbauwerke, Verbindungen mit mechanischen Verbindungsmitteln, Allgemeine Grundsaetze fuer die Ermittlung der Tragfaehigkeit und des Verformungsverhaltens. Beuth Verlag, Berlin DIN (Deutsches Institut fuer Normung e.V) (2004) DIN 1055-100: Einwirkungen auf Tragwerke - Teil 100: Grundlagen der Tragwerksplanung, Sicherheitskonzept und Bemessungsregeln. Beuth Verlag, Berlin CEN (Comite Europeen de Normalisation) (2004) EN 1995-1-1: 2004 (D). Eurocode 5: Bemessung und Konstruktion von Holzbauten- Teil 1-1: Allgemeines- Allgemeine Regeln und Regeln fuer den Hochbau. CEN- TC 250 Structural Eurocodes EN (European Norm) (2001) EN12512:2001 Holzbauwerke - Pruefverfahren – Zyklische Pruefungen von Anschluessen mit mechanischen Verbindungsmitteln. Schweizerischer Ingenieur - und Architektenverien, Zuerich.  126  Fonseca, F., Rose, S., & Campbell, S. (2002) Nail, Wood Screws, and Staple Fastener Connections. CUREE publication No.W-16, Consortium of Universities for Research in Earthquake Engineering, California, USA. Gehloff, M. (2011) Pull-Out Resistance of Self-Tapping Wood Screws With Continuous Threads. MASc Thesis, The University of British Columbia, Department of Wood Science. Green, D. W., Winandy, J. E., & Kretschmann, D. E. (1999) Mechanical Properties of Wood. In Wood Handbook - Wood as an Engineering Material (pp. 4-34 - 4-36). Forest Products Laboratory USDA Forest Service,Wisconsin. Haller, P., & Wehsener, J. (2003) Entwicklung innovativer Verbindungen aus Pressholz und Glasfaserarmierungen fuer den Ingenieurholzbau. Fraunehofer IRB-Verlag, Stuttgart Heiduschke, A., Haller, P., & Kasal, B. (2004) Zum Tragverhalten textilbewehrtverdichteter Holz-Rahmenecken unter zyklischer Beanspruchung. Bautechnik, 87-103. Heiduschke, A., Kasal, B., & Haller, P. (2008). Shake table tests of small- and full-scale laminated timber frames with moment connections. Bullettin: Earthquake Engineering (pp. 324-338). IBC (International Building Code) (2000) International Code Council.  127  Kasal, B., Heiduschke, A., & Haller, P. (2004) Heavy timber laminated frames with composite material reinforced beam-to-column connections under earthquake loads. Progress in Structural Engineeringing Mater(6), (pp. 84-93) Kopac, J., Sali, S (2002) Wood: an important material in manufacturing technology. Journal of materials processing technology. Volume 133, Issue 1-2 pp. 134-142 Krawinkler, H., Parisi, F., Ibarra, L., Ayoub, A., & Medina, R. (2001) Development of a testing protocol for wood frame structures. CUREE-Caltech Woodframe Project. Report W-02, Stanford, Califonia Kuklik, P. (2008) History of Timber Structures. HANDBOOK 1 - Timber Structures. Educational Materials for Designing and testing of Timber Structures – TEMTIS, (First ed., pp. 1-14) Kuklik, P. (2008) Wood Properties. HANDBOOK 1 - Timber Structures. Educational Materials for Design and Testing of Timber Structures – TEMTIS, (First ed., pp. 15-27) Lam, F., Gehloff, M., & Closen, M. (2010)® Moment -resiting boltet timber connections. Proceedings of the Institution of Civil Engineers, 163(SB4), (pp.267-274) Lam, F., Schulte-Wrede, M., Yao, C., & Gu, J. (2007) ® Moment resistance of bolted timber connections with perpendicular to grain reinforcements. http://www.rjc.ca/sites/default/files/2010_Bolted%20Timber_Yao.pdf (accessed July 12th, 2012) 128  Leijten, AJ.M., Ruxton, S., Prion, H.G.L., Lam, F., (2006) Reversed-cyclic behavior of a novel heavy timber tube connection. Journal of Structural Engineering, ASCE, 132, 2006, pp. 1314-1319 Marynowicz, A. (2008) Durability of Timber Structures. Handbook 1 - Timber Structures. Educational Material for Designing and Testing of Timber Structures – TEMTIS, (First ed., pp. 210-227) Materials and Welding (2012) Behavior of different materials http://materialswelding.blogspot.ca/2007/09/toughness-ability-of-metal-to-deform.html 9accessed May, 14th 2012) mgb ARCHITECTURE + DESIGN, Equilibrium Consulting, LMDG Ltd., BTY Group (2012) Tall Wood “ The case for tall wood buildings” Miller, R. B. (1999) Structure of Wood. In Wood Handbook - Wood as an Engineering Material (pp. 2-1 - 2-4). Forest Products Laboratory USDA Forest Service, Wisconsin Mischler, A. (1998) Bedeutung der Duktilitaet fuer das Tragverhalten von Stahl-HolzBolzenverbindungen. PhD thesis, ETH, Zuerich NBCC (National Building Code of Canada) (2005) National Research Council of Canada, Ottawa  129  NDS (National Design Specifications for Wood Construction) (2005). American Wood Council, Washington NRC (Natural Resources Canada) The M8.1 Haida Gwaii (formerly Queen Charlotte Island) Earthquake of August 22, 1949). http://www.earthquakescanada.nrcan.gc.ca/historic-historique/events/19490822eng.php, 2012 Rainer, H. J., & Karacabeyli, E. (1999) Performance of wood-frame building construction in earthquakes. Forintek Canada Corp., Vancouver Rodd, P., & Leijten, A. (2003) ® High-performance dowel-type joints for timber structures. Progress in Structural Engineering Matter (pp.77-89) Schulte-Wrede, M. (2007) Experimental Investigation of the Effects of Self-Tapping Screws as Reinforcement Perpendicular to Grain in a Moment Resistance Bolted Wood Connection. Unpublished Thesis, Vancouver. Sjoedin, J., & Johansson, C.-J. (2007) Influence of initial moisture induced stresses in multiple steel-to-timber dowel joints. Holz als Roh und Werkstoff (65), (pp.71-77) Sjoedin, J., Johansson, C.-J., & Petersson, H. (2004) Influence of moisture induced stresses in steel-to-timber dowel joints. Proceedings of the 8th World Conference on Timber Engineering, Lahti, Finland  130  Trautz, M., & Koj, C. (2009)® Self-tapping screws as reinforcement for timber structures. RWTH Aachen University, Faculty of Architecture, Chair of Structures and Structural Design, Germany Uibel, T., & Blass, H. (2010) Determining Suitable Spacings and Distances for Selftapping Screws by Experimental and Numerical Studies. World Conference on Timber Engineering, Riva del Garda, Italy, CD-ROM Proceedings Werner, H. (1995) Stiftverbindungen- Spalteffekte und Verstaerkungsmassnahmen. Informationsdienst Holz - Holzbauwerke nach Eurocode 5 Bemessung und Baustoffe, STEP 3, Lecture 9, Reprint for Teaching Purposes, Technische Universitaet Muenschen, Germany  131  APPENDIX Table 18: Measured moisture contents (MC) and wood densities for GG-4X series  GG1-4X GG2-4X GG3-4X GG4-4X GG5-4X Average  Left side member  Centre member  Right side member  MC  Density  MC  Density  MC  Density  [%]  [Kg/m3]  [%]  [Kg/m3]  [%]  [Kg/m3]  10.8 11.6 11.2 11.7 10.8 11.2  547 531 500 523 523 525  10.8 10.4 10.0 11.4 11.3 10.7  523 525 575 544 562 546  12.5 11.0 10.4 11.4 11.2 11.3  539 571 531 587 539 554  Table 19: Measured moisture contents (MC) and wood densities for GG-8X series  GG1-8X GG2-8X GG3-8X GG4-8X GG5-8X Average  Left side member  Centre member  Right side member  MC  Density  MC  Density  MC  Density  [%]  [Kg/m3]  [%]  [Kg/m3]  [%]  [Kg/m3]  12.0 11.0 11.0 9.4 11.3 10.9  603 556 526 520 544 550  11.2 11.4 10.9 10.4 10.8 10.9  564 558 554 564 591 566  11.0 10.2 10.8 11.4 11.7 11.0  574 562 556 538 538 553  132  Table 20: Measured moisture content (MC) and wood densities for GCLT-4X series  GCLT1-4X GCLT2-4X GCLT3-4X GCLT4-4X GCLT5-4X Average  Left side member  Centre member  MC  MC  Density 3  Density 3  Right side member MC  Density 3  [%]  [Kg/m ]  [%]  [Kg/m ]  [%]  [Kg/m ]  9.5 10.8 10.9 12.0 10.0  501 417 428 444 444 447  11.9 10.3 11.9 11.3 11.7  563 531 537 551 558  9.6 12.1 10.0 9.5 9.1  475 438 432 449 493  11.4  548  10.0  457  10.6  Table 21: Measured moistre content (MC) and wood densities for the GCLT-8X series  GCLT1-8X GCLT2-8X GCLT3-8X GCLT4-8X GCLT5-8X Average  Left side member  Centre member  MC  Density  MC  Density  [%]  [Kg/m3]  [%]  9.4 10.7 8.2 11.7 11.3  476 515 515 532 487 505  11.2 11.8 11.7 8.9 12.1 11.1  10.2  Right side member MC  Density  [Kg/m3]  [%]  [Kg/m3]  549 568 541 564 560  8.4 10.7 9.9 11.4 11.5  476 454 471 515 460  556  10.3  475  133  Table 22: Measured moisture contents (MC) and wood densities for the GG-4X-30 series  GG1-4X-30 GG2-4X-30 GG3-4X-30 GG4-4X-30 GG5-4X-30 Average GCLT1-4X-30 GCLT2-4X-30 GCLT3-4X-30 GCLT4-4X-30 GCLT5-4X-30 Average  Left side member  Centre member  MC  MC  Density 3  Density 3  Right side member MC  Density 3  [%]  [Kg/m ]  [%]  [Kg/m ]  [%]  [Kg/m ]  10.0 10.9 10.0 11.3 11.0  590 590 545 579 612 583  11.1 11.3 11.2 11.4 10.4  530 553 555 558 558  11.3 10.3 10.5 11.7 10.2  586 560 554 525 572  11.0 11.0 12.1 11.9 11.3 11.7 11.6  551 550 536 558 572 579 559  10.8 11.0 10.8 10.0 9.5 9.1 10.0  568 481 428 435 452 517 463  10.6 10.9 8.3 11.1 9.6 9.8 9.9  473 469 380 408 433 433  134  Table 23:Recorded moment connection test data Mc-1 pos. neg. cycle cycle  Mc-2 pos. neg. cycle cycle  Mc-3 pos. neg. cycle cycle  Mc-4 pos. neg. cycle cycle  Mc-5 pos. neg. cycle cycle  max M. [KNm] rotation @ max M. [°]  69.34  -102  88.81  -86.0  79.20  -82.9  89.14  -85.5  108.4  -80.8  2.84  -8.46  3.25  -2.67  3.61  -4.21  2.59  -2.51  2.99  -3.24  failure M. [KNm] rotation @ failure [°]  55.47  -81.4  71.05  -68.8  63.36  -66.3  71.32  -68.4  86.73  -64.7  3.30  -4.34  3.90  -2.82  3.82  -5.18  3.46  -3.28  3.00  -3.67  40% max M. [KNm] rotation @ max M. [°]  27.74  -40.7  35.52  -34.4  31.68  -33.1  35.66  -34.2  43.36  -32.3  1.27 -  -2.16 -  1.41  -1.47  1.28  -1.44  0.92  -1.08  0.91  -0.98  -  -  -  -79.6  83.90  -  101.5  -75.6  -  -  -  -  -  -3.46  2.25  -  2.13  -2.29  21.90 -  18.87 -  30.52  33.85  24.75  22.97  38.87  35.05  47.75  33.01  -  -  -  1.17  1.15  -  1.40  1.41  yield M. [KNm] rotation @ yield M. [°] stiffness. [KNm/°] ductility [-]  Mc-6 pos. neg. cycle cycle  Mc-7 pos. neg. cycle cycle  Mc-8 pos. neg. cycle cycle  Mc-9 pos. neg. cycle cycle  Mc-10 pos. neg. cycle cycle  max M. [KNm] rotation @ max M. [°]  107.8  -88.3  99.55  -88.4  95.22  -101  93.26  -93.9  102.5  -89.1  3.60  2.81  3.55  2.19  3.12  2.87  3.46  4.26  3.14  2.76  failure M. [KNm] rotation @ failure [°]  86.28  -70.6  79.64  -70.7  76.17  -80.7  74.61  -75.1  82.02  -71.3  3.55  -3.27  4.24  -2.68  3.33  -3.13  3.65  -4.92  3.36  -3.32  40% max M. [KNm] rotation @ max M. [°]  43.14  -35.3  39.82  -35.3  38.09  -40.3  37.30  -37.5  41.01  -35.6  1.44  -1.14  2.01  -0.90  1.27  -1.05  1.40  -2.08  1.13  -1.05  yield M. [KNm] rotation @ yield M. [°]  106.2  -81.3  -  -  94.83  -99.5  -  -  95.43  -86.8  3.05  -2.35  -  -  3.12  -2.58  -  -  2.71  -2.55  stiffness. [KNm/°] ductility [-]  34.93 1.18  37.50 1.19  27.27 -  47.15 -  29.89 1.00  38.47 1.11  31.18 -  23.79 -  36.36 1.16  33.97 1.08  135  Table 24: Measured moisture contents (MC) and wood densities for moment connection  Test MC-1 MC-2 MC-3 MC-4 MC-5 MC-6 MC-7 MC-8 MC-9 MC-10 Average Stdev COV  Column Member Density in Kg/m³ MC [%] 596 520 497 560 536 531 548 538 518 562 541 28 5.15%  11.3 10.9 10.3 11.6 11.2 11.8 10.1 11.5 11.7 11 11.1 0.58 5.17%  Beam Member Density in Kg/m³ MC [%] 522 525 553 532 525 516 549 543 524 529 532 12 2.32%  11 11.2 11.8 12.3 12.4 10.8 10.3 9.4 11.7 11.3 11.2 0.91 8.14%  136  Moment [kNm]  120 90 60 30  27.74  0 -8  -6  -4  -2  0  1.27  2  4  -30  6 8 Rotation [°]  -60 -90 -120  Figure 60: Moment rotation response specimen Mc-1  Moment [kNm]  120 90 60  35.52  30 0 -6  -4  -2  0  1.41  2  4  -30  6  8 Rotation [°]  -60 -90 -120  Figure 61: Moment rotation response specimen Mc-2  137  Moment [kNm]  120 90 60 30  31.68  0 -8  -6  -4  -2  0  1.28  2  4  6 8 Rotation [°]  -30 -60 -90 -120  Figure 62: Moment rotation response specimen Mc-3  Moment [kNm]  120 90 60  35.66  30 0 -6  -4  -2  0  0.92  2  -30  4  6 Rotation [°]  -60 -90 -120  Figure 63: Moment rotation response specimen Mc-4  138  Moment 120 [kNm] 90 60  43.36  30 0 -6  -4  -2  0  0.91  2  4  6 Rotation [°]  4  6 Rotation [°]  -30 -60 -90 -120  Figure 64: Moment rotation response specimen Mc-5  Moment 120 [kNm] 90 60  43.14  30 0 -6  -4  -2  0  1.44  2  -30 -60 -90 -120  Figure 65: Moment rotation response specimen Mc-6  139  Moment 120 [kNm] 90 60 39.82  30 0 -6  -4  -2  2.01  0  2  4  6 Rotation [°]  4  6 Rotation [°]  -30 -60 -90 -120  Figure 66: Moment rotation response specimen Mc-7  Moment [kNm]  120 90 60  38.09  30 0 -6  -4  -2  0  1.27  2  -30 -60 -90 -120  Figure 67: Moment rotation response specimen Mc-8  140  GG1-4X  -6  -4  -2  Load in [kN] 180 160 140 120 100 80 60 40 20 0 -20 0 2 -40 -60 -80 -100 -120 -140 -160 -180  4  6  Displacement in [mm]  Figure 68: Load deformation response specimen GG1-4X  GG2-4X  -6  -4  -2  Load in [kN] 180 160 140 120 100 80 60 40 20 0 -20 0 2 -40 -60 -80 -100 -120 -140 -160 -180  4  6  Displacement in [mm]  Figure 69: Load deformation response specimen GG2-4X  141  GG3-4X  -6  -4  -2  180 160 140 120 100 80 60 40 20 0 -20 0 -40 -60 -80 -100 -120 -140 -160 -180  Load in [kN]  2  4  6  Displacement in [mm]  Figure 70: Load deformation response specimen GG3-4X  GG4-4X  -6  -4  -2  180 Load in [kN] 160 140 120 100 80 60 40 20 0 -20 0 2 -40 -60 -80 -100 -120 -140 -160 -180  4  6  Displacement in [mm]  Figure 71: Load deformation response specimen GG4-4X  142  GG5-4X  -6  -4  -2  180 160 140 120 100 80 60 40 20 0 -20 0 -40 -60 -80 -100 -120 -140 -160 -180  Load in [kN]  2  4  6  Displacement in [mm]  Figure 72: Load deformation response specimen GG5-4X  143  Load in [kN]  250 200 150 GG1-8X  100 50 0 -8  -6  -4  -2  0 -50  2  4  6 8 Displacement in [mm]  -100  -150 -200 -250  Figure 73: Load deformation response specimen GG1-8X  144  Load in [kN]  250 200 150  GG2-8X  100 50 0 -8  -6  -4  -2  0 -50  2  4  6 8 Displacement in [mm]  -100  -150 -200 -250  Figure 74: Load deformation response specimen GG2-8X  145  Load in [kN]  250 200  150  GG3-8X  100 50 0 -8  -6  -4  -2  0 -50  2  4  6 8 Displacement in [mm]  -100 -150 -200 -250  Figure 75: Load deformation response specimen GG3-8X  146  GG4-8X  -8  -3  270 240 210 180 150 120 90 60 30 0 -30 -60 -90 -120 -150 -180 -210 -240 -270  Load in [kN]  2  7 Displacement in [mm]  Figure 76: Load deformation response specimen GG4-8X  147  -8  -6  270 Load in [kN] 240 210 180 GG5-8X 150 120 90 60 30 0 -4 -2 -30 0 2 -60 -90 -120 -150 -180 -210 -240 -270  4  6 8 Displacement in [mm]  Figure 77: Load deformation response specimen GG5-8X  148  Load in [kN]  250 200  GG1-4X-30 150 100 50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250  Figure 78: Load deformation response specimen GG1-4X-30  149  Load in [kN]  250 200  GG2-4X-30 150 100 50 0 -8  -6  -4  -2  0  2  4  6  8  -50  Displacement in [mm] -100 -150 -200 -250  Figure 79: Load deformation response specimen GG2-4X-30  150  Load in [kN]  250 200  GG3-4X-30 150 100 50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250  Figure 80: Load deformation response specimen GG3-4X-30  151  Load in [kN]  250 200  GG4-4X-30 150 100 50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250  Figure 81: Load deformation response specimen GG4-4X-30  152  Load in [kN]  250 200  GG5-4X-30 150 100 50 0 -8  -6  -4  -2  0  2  4  6  8  -50  Displacement in [mm] -100 -150 -200 -250  Figure 82: Load deformation response specimen GG5-4X-30  153  Load in [kN]  180 150 120  GCLT1-4X  90 60 30 0 -8  -6  -4  -2  0 -30  2  4  6  8  Displacement in [mm]  -60 -90 -120 -150 -180  Figure 83: Load deformation response specimen GCLT1-4X  154  Load in [kN]  180 150 120  GCLT2-4X  90 60 30 0 -8  -6  -4  -2  0 -30  2  4  6  8  Displacement in [mm]  -60 -90 -120 -150 -180  Figure 84: Load deformation response specimen GCLT2-4X  155  Load in [kN]  180 150  GCLT3-4X  120 90 60 30 0  -8  -6  -4  -2  0 -30  2  4  6  8  Displacement in [mm]  -60 -90 -120 -150 -180  Figure 85: Load deformation response specimen GCLT3-4X  156  Load in [kN]  180 150 120  GCLT4-4X  90 60 30 0 -8  -6  -4  -2  0 -30  2  4  6  8  Displacement in [mm]  -60 -90 -120 -150 -180  Figure 86: Load deformation response specimen GCLT4-4X  157  Load in [kN]  180 150 120  GCLT5-4X  90 60 30 0 -8  -6  -4  -2  -30  0  2  4  6  8  Displacement in [mm]  -60 -90 -120 -150 -180  Figure 87: Load deformation response specimen GCLT5-4X  158  Load in [KN]  300 250 200 150 100  GCLT1-8X  50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250 -300  Figure 88: Load deformation response specimen GCLT1-8X  159  Load in [KN]  300 250 200 150 100  GCLT2-8X  50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250 -300  Figure 89: Load deformation response specimen GCLT2-8X  160  Load in [KN]  300 250 200 150 100  GCLT3-8X  50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250 -300  Figure 90: Load deformation response specimen GCLT3-8X  161  Load in [KN]  300 250 200 150 100  GCLT4-8X  50 0 -8  -6  -4  -2  0 -50  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250 -300  Figure 91: Load deformation response specimen GCLT4-8X  162  Load in [KN]  300 250 200 150 100  GCLT5-8X  50 0 -8  -6  -4  -2  -50  0  2  4  6  8  Displacement in [mm]  -100 -150 -200 -250 -300  Figure 92: Load deformation response specimen GCLT5-8X  163  220  Load in [kN]  180 140 100  GCLT1-4X-30  60 20 -8  -6  -4  -2  -20 0 -60  2  4  6  8  Displacement in [mm]  -100 -140 -180 -220  Figure 93: Load deformation response specimen GCLT1-4X-30  164  220  Load in [kN]  180 140  GCLT2-4X-30  100 60 20  -8  -6  -4  -2  -20 0 -60  2  4  6  8  Displacement in [mm]  -100 -140 -180 -220  Figure 94: Load deformation response specimen GCLT2-4X-30  165  220  Load in [kN]  180 140  GCLT3-4X-30 100 60 20 -8  -6  -4  -2  -20 0 -60  2  4  6  8  Displacement in [mm]  -100 -140 -180 -220  Figure 95: Load deformation response specimen GCLT3-4X-30  166  220  Load in [kN]  180 140  GCLT4-4X-30  100 60 20  -8  -6  -4  -2  -20 0 -60  2  4  6  8  Displacement in [mm]  -100 -140 -180 -220  Figure 96: Load deformation response specimen GCLT4-4X-30  167  220  Load in [kN]  180 140 100  GCLT5-4X-30  60 20 -8  -6  -4  -2  -20 0 -60  2  4  6  8  Displacement in [mm]  -100 -140 -180 -220  Figure 97: Load deformation response specimen GCLT5-4X-30  168  

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.24.1-0072901/manifest

Comment

Related Items