Influence of Moisture Content and Reinforcement on LoadCarrying Capacity of Housings in Solid Timber Beams By Rafid Shams Huq B.Sc., Bangladesh University of Engineering and Technology (BUET), Dhaka, Bangladesh, 2016 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE IN ENGINEERING UNIVERSITY OF NORTHERN BRITISH COLUMBIA July 2022 © Rafid Shams Huq, 2022 i Abstract Timber construction is a sustainable solution to address the increasing global demand for housing. Timber framing currently uses mostly mechanical fasteners, which are made from metals with high carbon footprint, and – depending on the type of fasteners– provide poor aesthetics. These challenges can be addressed by using traditional joints such as timber housings. However, there are no design guidelines available that account for the joint geometric parameters, the wood moisture content during fabrication and when loaded as well as the impact of mechanical reinforcements. This thesis extends on previous research to investigate the influence of wood moisture and reinforcement on the load-carrying capacity of solid timber beams with housings. A total of 60 beams with single housings and 30 beams with double housings were prepared and tested while varying the moisture condition (either wet or dry) at the time of cutting the housings and at the time of testing the beams. Along with 32 test data from previous research, statistical analysis was conducted to evaluate the statistical significance of the investigated parameters. In addition, 198 small-scale specimens were tested for shear, tension and compression to evaluate the impact of small clear specimen material strength on the beam load-carrying capacity. The tests confirmed that larger depth below the mortise and the use of self-tapping screws as reinforcement lead to increase of load-carrying capacity; however, the moisture condition during cutting and testing was only shown to have a significant impact on the load-carrying capacity of double housings, but not single housings. The correlation between the material strength tests and load-carrying capacities demonstrated that the strength of small clear specimens from the same beam had no influence on the load-carrying capacity of the housings. ii Table of Contents Abstract .....................................................................................................................................ii List of Figures ........................................................................................................................... v List of Tables ..........................................................................................................................vii Acknowledgements .............................................................................................................. viii 1 2 3 Introduction ...................................................................................................................... 1 1-1 Background ................................................................................................................. 1 1-2 Load Bearing Timber Housing .................................................................................... 1 1-3 Load and Moisture Induced Stresses in Housings ...................................................... 2 1-4 Reinforcements of Connection Details ....................................................................... 2 1-5 Objectives .................................................................................................................... 2 1-6 Scope and Limitations ................................................................................................. 3 Literature Review ............................................................................................................. 4 2-1 Material Properties of Wood ....................................................................................... 4 2-2 Timber Housings ......................................................................................................... 6 2-3 Reinforcing Wood Joints ............................................................................................. 8 2-4 Preceding Work at UNBC ........................................................................................... 9 2-5 Summary of Literature Review ................................................................................. 12 Experimental Investigations .......................................................................................... 13 3-1 Materials ........................................................................................................................ 13 3-2 Specimen Description and Test Matrix.......................................................................... 16 3-3 Test Methods .................................................................................................................. 18 3-4 Statistical Analyses ........................................................................................................ 20 3-5 Results ............................................................................................................................ 21 iii 3-5-1 Compression, Tension and Shear Tests .................................................................. 21 3-5-2 Single Housing Beams ............................................................................................ 26 3-5-3 Double Housing Beams .......................................................................................... 30 3-6 Discussion ...................................................................................................................... 33 3-6-1 Correlation between material strengths and housing load-carrying capacity ......... 33 3-6-2 Single Housing Beams ............................................................................................ 37 3-6-3 Double Housing Beams .......................................................................................... 41 4 Conclusions and Outlook ............................................................................................... 46 References ............................................................................................................................... 48 Appendices .............................................................................................................................. 51 Appendix 1 – Test results from previous research (Heal et al., 2019) ................................. 51 Appendix 2 – Detailed test results of all specimens in S-0.80 series ................................... 52 Appendix 3 – Detailed test results of all specimens in S-0.65 series ................................... 60 Appendix 4 – Detailed test results of all specimens in D-0.80 series .................................. 68 Appendix 5 – Detailed test results of all specimens in D-0.65 series .................................. 76 iv List of Figures Figure 1-1: Load bearing housing (Nehil & Trojniak, 2013) .................................................... 1 Figure 2-1: Effect of moisture content on wood strength properties ......................................... 5 Figure 2-2: a) Typical single and double timber housings (Heal et al., 2019) and b) photo ..... 7 Figure 2-3: Failure Modes (a) Splitting, (b) Bearing, (c) Combined Bearing/Splitting .......... 11 Figure 2-4: Typical load-displacement curves for failure modes (a) Splitting, (b) Bearing, (c) Combined Bearing/Splitting (Heal et al., 2019) ...................................................................... 11 Figure 3-1: Test Assembly for (a) Compression Perpendicular to Grain Test (b) Tension Perpendicular to Grain Test (c) Shear Test .............................................................................. 14 Figure 3-2: Fully threaded STS of 120mm, 180mm and 200mm from left to right ................ 15 Figure 3-3: Schematic of a) single housing and b) double housing ......................................... 16 Figure 3-4: Single & Double Housing Specimens with Reinforcements ................................ 17 Figure 3-5: Specimen naming convention ............................................................................... 17 Figure 3-6: Experimental setup: a) photo, b) detail sensor, and c) schematic ......................... 19 Figure 3-7: Failure modes of Compression, Tension and Shear Tests .................................... 21 Figure 3-8: Failure Modes: a) Splitting, b) Bearing c) Combined Bearing/Splitting .............. 27 Figure 3-9: Typical load-displacement curves for different failure modes ............................. 27 Figure 3-10: Load-displacement tests of typical 0.80 series under bearing load (a) Load vs Beam Displacement (b) Load vs Mean Crack Opening .......................................................... 30 Figure 3-11: Load-displacement tests of typical 0.65 series under bearing load (a) Load vs Beam Displacement (b) Load vs Mean Crack Opening .......................................................... 30 Figure 3-12: Failure Modes Observed (a) Splitting Failure, (b) Bearing Failure, (c) Combined Bearing/Splitting Failure .......................................................................................................... 31 Figure 3-13: Load-displacement of D-0.80 series: a) Load vs Beam Displacement (b) Load vs Crack Opening for Side #1, c) Load vs Crack Opening for Side #2 ....................................... 33 v Figure 3-14: Load-displacement of D-0.65 series: a) Load vs Beam Displacement, b) Load vs Crack Opening for Side #1, c) Load vs Crack Opening for Side #2 ....................................... 33 Figure 3-15: Relation between capacity and material strength ................................................ 34 Figure 3-16: Relation between material strength and density ................................................. 35 Figure 3-17: Capacity vs Mechanical Strength based on Failure Modes ................................ 36 Figure 3-18: Relation between Load-carrying capacity and Mortise Depth (Single Housings) .................................................................................................................................................. 38 Figure 3-19: Relation between load-carrying capacity and reinforcement (Single Housings) 39 Figure 3-20: Relation between load-carrying capacity and moisture condition (single housings) .................................................................................................................................................. 40 Figure 3-21: Relation between load-carrying capacity and mortise depth (double housings) 42 Figure 3-22: Relation between load-carrying capacity and moisture condition (double housings) .................................................................................................................................. 43 Figure 3-23: Relation between load-carrying capacity and reinforcement (double housings) 44 vi List of Tables Table 2-1: Parameter overview and results of tests by Heal et al. (2021) ............................... 10 Table 3-1: Overview test series................................................................................................ 18 Table 3-2: Compression Test Results ...................................................................................... 23 Table 3-3: Tension Test Results .............................................................................................. 24 Table 3-4: Shear Test Results .................................................................................................. 25 Table 3-5: Estimated Bearing Capacity ................................................................................... 26 Table 3-6: Load-carrying capacity, crack opening, and density: single housing specimens ... 28 Table 3-7: Load-carrying capacity, beam displacement and crack opening: double housing . 32 Table 3-8: Proportion of failure modes: single housing .......................................................... 37 Table 3-9: Three Factor ANOVA results: single housing ....................................................... 41 Table 3-10: Proportion of failure modes: non-reinforced double housing .............................. 42 Table 3-11: Two factor ANOVA (Analysis 1) results: double housing .................................. 45 Table 3-12: Two factor ANOVA (Analysis 2) results: double housing .................................. 45 vii Acknowledgements I am grateful to Allah for all His blessings. I thank my parents and wife for their unwavering support throughout the entire journey of graduate studies. It has been an honor and great privilege working with Dr. Thomas Tannert. His prompt and detailed feedback kept me going in the right direction. His continuous guidance and support made it possible to get this thesis to the present state. I also appreciate the valuable suggestions provided by the committee members Dr. Jianhui Zhou and Maik Gehloff. I want to take this opportunity to thank Maik Gehloff for ensuring that I had the beams with required dimensions on time for my tests. Finally, I thank the entire team at Wood Innovation Research Laboratory –Michael Billups, James Andal and Ryan Stern– for preparing the test samples, setting up the equipment/test assembly and ensuring smooth running of everything required for completing all my tests related to my thesis. viii 1 Introduction 1-1 Background The increasing global population requires extensive levels of construction. The construction sector contributed 39% of energy and process-related carbon dioxide emissions in 2018, out of which one third was due to manufacturing building materials and products such as cement and steel (IEA, 2019). In comparison, timber requires little energy for processing, about 60 to 80% less than an equivalent structure made of concrete (Börjesson & Gustavsson, 2000). Increased use of wood and engineered wood products in construction along with sustainable forestry is required. Hence, increased use of traditional joints such as housings can be a potential solution. 1-2 Load Bearing Timber Housing Load bearing housings are quite common; these consist of a joist (secondary beam) with a tenon attached to a main beam (primary beam) resting on a mortise, as shown in Figure 1-1. Mortise Tenon Figure 1-1: Load bearing housing (Nehil & Trojniak, 2013) Such joints are also known as beam pockets, butt cogs, or partial-width notches (Sobon, 2014). Previous research on load bearing timber housings is limited, but provides valuable understanding of its performance. Amongst other findings, a linear relationship was observed between the load-carrying capacity and depth below the notch (Nehil & Trojniak, 2013), (Heal 1 et al., 2019). However, most previous tests were done under constant service conditions, and the effect of changes in moisture content (MC) due to seasoning has not been evaluated. 1-3 Load and Moisture Induced Stresses in Housings Wood is a natural and anisotropic material with low strength in the perpendicular to grain direction. Hence, it is preferable to avoid or reduce stresses perpendicular to the grain. However, such stresses inevitably develop in housings, caused by loads and increased due to changes in moisture content. Since wood is also hygroscopic in nature, surrounding environmental conditions such as relative humidity and temperature impact its moisture content, which in turn influences mechanical properties of wood. Drying of wood not only leads to shrinkage, but can also lead to internal tensile stresses and internal cracks (United States Department of Agriculture, 2010), (Dietsch, 2017). 1-4 Reinforcements of Connection Details To counteract the effect of moisture-induced stresses, and to increase a connection detail’s load-carrying capacity, several reinforcement methods exist. Self-tapping screws (STS) are quite popular in modern construction as means of reinforcement (Tannert & Lam, 2009). Previous work on load bearing housings established that STS increased the load-carrying capacity and the deformation capacity (Heal et al., 2019). 1-5 Objectives The objective of this study is to investigate the structural performance of timber housings. The specific objectives are to: 1) Investigate the effect of moisture changes on housings by testing dry specimens, which were fabricated under both wet (wet-dry) and dry (dry-dry) conditions and compare these to previous results obtained by testing wet specimens (wet-wet). 2 2) Investigate the relationship between the wood mechanical properties (such as shear parallel to grain, tension and compression perpendicular to grain strength properties) and the housing load-carrying capacities. 3) Investigate the influence of reinforcement with self-tapping screws (STS) on load-carrying capacities and the failure modes of housings. 1-6 Scope and Limitations This research focused on testing solid timber beams under wet and dry conditions, but not any range of moisture contents in between. Therefore, all conclusions are limited to these two extreme conditions. The development of design provisions for housings was beyond the scope of this thesis. In addition, all tests were conducted under quasi-static monotonic short-term loading at room temperature. Hence, long-term creep effects, influence of fire and durability on load-carrying capacity are outside the scope of this research. Moreover, the present study is limited to the use of Douglas fir grade # 1 and fully threaded 8 mm diameter STS. Hence, the performance of housings in different species or different reinforcement methods is not known. 3 2 Literature Review 2-1 Material Properties of Wood Wood is the most widely used renewable construction material. It has high strength-to-weight ratio and good thermal insulation properties. As a naturally grown material, its properties are highly variable and vary not only between species, but also between trees of one species and even within an individual tree. The latter may be due to difference in density and presence of knots, spiral grain, slope of grain, etc. Hence, timbers are graded to satisfy the requirements of diverse usage and to ensure reliability of product quality (Harte, 2009). Wood is an organic composite of cellulose fibers and lignin in which the orientation of microfibers in cell wall has great influence on strength and shrinkage properties of wood. It is an orthotropic material having higher strength and stiffness in parallel-to-grain direction while being weaker in the perpendicular-to-grain direction (Harte, 2009). Therefore, tension stresses perpendicular to grain should be minimized. Barrett et al. (1975) used Weibull’s theory of brittle fracture to derive allowable stresses while taking into account the size effect, proving that the allowable stress is related to the structure’s volume. Wood is a hygroscopic material; it exchanges moisture with the surrounding environment to attain an equilibrium state (Sjödin & Johansson , 2007), which equilibrium is called Equilibrium Moisture Content (EMC). The EMC depends on the surrounding environment, i.e. relative humidity and temperature (Harte, 2009). The MC during construction varies based on construction material and surrounding conditions, from as low as 6% to up to 20%. However, unseasoned green wood has MCs higher than 30%, and kiln-drying (expensive) or air drying (time consuming) are used to reduce MC. Moisture exchange with the surrounding on the wood surface is much faster compared to inner regions due to the distance of diffusion. Hence, moisture content depends on both time and 4 location inside the cross section. It should be noted that both physical and mechanical properties of wood change mostly below the Fiber Saturation Point (FSP); the FSP ranges from 28% to 30% depending on species. Widehammar (2004) showed that wood exhibits higher strength in dry condition compared to wet condition. Note: Wood is called dry if MC is less than 19%, wet if the MC is higher but still below the (MC of approx. 28%), and green if the MC is higher than the FSP. The change of different strength properties (A: Tension parallel to grain, B: Bending, C: Compression parallel to grain, D: Compression perpendicular to grain, E: Tension perpendicular to grain) as a function of moisture content is shown in Figure 2-1, (United States Department of Agriculture, 2010). Figure 2-1: Effect of moisture content on wood strength properties Reduction in moisture content also leads to shrinkage of wood and the shrinkage varies based on direction. Shrinkage is lowest in the longitudinal to fiber orientation (0.1 to 0.2%) and maximum in the tangential orientation (6 to 13%) (Green, 2001). The differential shrinkage or swelling based on orientation is reflection of the anisotropic nature of wood and leads to strains of different degrees across the cross-section. During drying, the inner wood restrains the wood near the surface which results in tensile stresses perpendicular to the grain near the surface. 5 Whereas, there are compressive stresses inside the cross-section to counterbalance the tensile stresses in order to maintain stress equilibrium in the section. Although such stresses reduce over time due to relaxation processes, these stresses can lead to cracks if the local tensile strength perpendicular to grain of timber is exceeded. Hence, the load-carrying capacity of structural member will be reduced (Dietsch, 2017). Häglund (2007) investigated moisture penetration and moisture gradients at various climatic location over 23 years to conclude that rapid daily changes do not affect inner part of the timber. However, slow annual variations have a significant effect. Chaplain & Valentin (2010) demonstrated that cyclic change in air humidity produced rapid change of the crack velocity and this was more prominent during drying change. Drying of wood not only leads to shrinkage, but can also lead to internal tensile stresses and which in turn can lead to cracks (Franke & Franke, 2014). It occurs when the strength of bending, shear or tension of the timber is exceeded, for example in beams with notches. Rammer & Winistorfer (2001) proved the effect of moisture content on the stiffness and the load-carrying capacity of connections. Although, temperature treatment is a viable option to improve the hygroscopic properties of wood (Roszyk et al., 2020), moisture content may vary due to change in ambient conditions inducing stress perpendicular to the grain. 2-2 Timber Housings Timber-framing is a load-bearing wooden structure usually held together with mortise and tenon (or similar) joinery. Currently, there are no design guidelines for housings and use of such connections are limited based on experience of the engineers. Limited research focused on carpentry-type joints. Tannert (2008) tested rounded dovetail connections of different materials, geometry, loading conditions, moisture contents, manufacturing tolerances and reinforcement with STS and concluded that capacity and failure mode is impacted by the load 6 transfer mechanism with size effect playing a major role in capacity. A housing is a mortise cut on the side of the primary beam to support a secondary beam, enabling load to be transferred. Housings can be either single or double housings i.e. mortise cut on only one side or both sides of a beam respectively. Schematics and a photo for typical timber housings are shown in Figure 2-2. a) b) Figure 2-2: a) Typical single and double timber housings (Heal et al., 2019) and b) photo Nehil & Trojniak (2013) tested load-bearing housings in green timbers with different depth of material remaining below the mortise. They concluded that the capacity is linearly proportional to the depth of material below notch and strength of the specimens themselves. The weaker softwoods failed in sudden rupture (tension perpendicular to grain), whereas the stronger hardwoods showed ductile-like splitting with increased load-carrying capacity after the split developed. Although not related to the observed failure mechanisms, results were consistent with the National Design Specification (NDS) Equation 3.4-7 (Wood Design Standards Committee, 2018). Similar consistency with the NDS Equation was observed by (Iuliano, 2013), based on 36 tests on load-bearing housings in 8" × 8" timber beams with varying housing heights. Garbin et al. (2006) tested four different configurations of dovetail connections in shear, with two configurations designed to fail the mortises and the other two designed to fail the tenons. They concluded that the location of defects and pith (if present) had a substantial impact and 7 they proposed an empirical model based on the slip of the joint for the tested configurations. However, the results may be skewed as only short lengths of joist and beams were tested. Thus, the supports may have provided additional resistance against splitting which would be absent in reality. More recently, Heal et al. (2019) investigated housings in timber beams to analyze the influence of variations in housing depth, housing width, use of reinforcement, shape of housing, etc. to provide insightful findings on the stated parameters. 2-3 Reinforcing Wood Joints Timber strength perpendicular to the wood grain is significantly lower than parallel to the grain, hence timber structures and joints should be detailed minimizing stresses perpendicular to the grain (Blass & Bejtka, 2004), (Dietsch & Brandner 2015). Where it is inevitable, the timber should be reinforced to compensate for the low strength properties. Reinforcement is possible using various means, such as glued-in rods, bolts, STS, etc. STS are considered the best choices for many applications as they are fast to install (Dietsch & Brandner 2015). Angst & Malo (2012) studied the effectiveness of STS as reinforcement under variable climatic conditions and concluded that during wetting season in comparison to unreinforced glulam, tensile stresses were reduced significantly by 70% to 30%. However, slight increase in tensile stress was observed during the drying season due to the additional restraint of reinforcing STS. It was shown by Jockwer (2014) and later confirmed by Danzer et al. (2016) that installation of screws at an angle to the grain is an efficient way to increase the capacities both for tension perpendicular to the grain and in shear for the notched timber members. Dietsch (2017) investigated the effect of reinforcement on shrinkage stresses in timbers. He concluded that restraining effect of reinforcement was more critical for timber members in buildings with a constant dry climate. A reduction of moisture content of 3% around reinforcement could lead to critical tensile stresses perpendicular to the grain, even after 8 considering the influence of relaxation processes. In fact, even a 1% reduction in moisture was shown to nullify the comparable stress transfer by the reinforcement in the un-cracked member. When the reinforcements were placed at an angle of 45°, the moisture induced stresses were reduced to half whereas the stressed volume was reduced to about 15%. Hence, for areas of high stresses, e.g. near notches, he recommended placing reinforcements at an angle to allowing release and proportionate transfer of shear stresses. Zhang et al. (2019) compared between the effectiveness of fully threaded screws (100% thread length) and partially threaded screws (33% thread length) in improving the mechanical properties of single-dowel timber connections with artificial cracks of three different widths which represented splitting of the wood due to variation in moisture. They concluded that partial threaded STS were sufficiently suitable to reinforce cracks up to 4.5 mm, but not effective for higher cracks of 6 mm width. 2-4 Preceding Work at UNBC In previous research by Heal et al. (2019), a total of 110 Douglas-Fir (No.1 or better) simply supported beams with 13 series for single-sided connections and 8 series for double-sided connections were tested. The specimens were all 1,830 mm (6 feet) in length and the crosssections were 89 mm × 285 mm (4″ × 12″), 89 mm × 235 mm (4″ × 10″), or 184 mm × 235 mm (8″ × 10″). The mortise depth and width were mostly 25 mm and 89 mm, respectively. The specimens were also tested with reinforcement using two STS on either side of the mortise. Hence, two STS for single housing series and four STS for double housing series. It should be noted that all these tests were conducted on specimens where the housings were cut under wet conditions (mean moisture content of 29%). The parameters and test results of all test series are summarized in Table 2-1 The test series marked with asterisk (*) will later be referred to as wet-wet (WW) series in the research work presented in this thesis and the detailed results of these wet-wet series from the previous research are provided in Appendix 1. 9 Table 2-1: Parameter overview and results of tests by Heal et al. (2021) Mortice Height [mm] 142 Housing Width [mm] 89 Single/Double Reinforcement Single No Fult [kN] 31.4 S-0.5-89-25-RR 142 89 Single Yes 52.0 S-.65-89-25 100 89 Single No 28.2 S-.65-89-25-Sq 100 89 Single No 19.6 S-.65-89-25-RR 100 89 Single Yes 21.1 S-0.8-89-25 57 89 Single No 14.9 S-0.8-89-25-RR 57 89 Single Yes 18.0 S-0.8-89-25-R 57 89 Single Yes 18.3 S-0.65-50-25 100 50 Single No 19.7 S-0.65-130-25 100 130 Single No 22.6 S-0.65-89-38 100 89 Single No 26.4 S-0.65-WW* 100 89 Single No 20.1 S-0.80-WW* 57 89 Single No 15.0 D-0.5-89-25 118 89 Double No 45.6 D-0.65-89-25* 82 89 Double No 43.5 D-0.65-89-25-RR 82 89 Double Yes 51.0 D-0.8-89-25* 47 89 Double No 31.3 D-0.8-89-25-RR* 47 89 Double Yes 39.8 D-0.8-89-25-R* 47 89 Double Yes 43.7 D-0.65-50-25 82 50 Double No 34.4 D-0.65-130-25 82 130 Double No 54.1 ID S-0.5-89-25 Three different failure modes - bearing, splitting and combined splitting/bearing, shown in Figure 2-3, were identified. Bearing failures were characterized by localized crushing of wood fibers at the mortise base. Sustained load pick-up was observed without reaching a peak load because of wood densification as shown in Figure 2-4. Splitting failures were characterized by the separation of timber into two parts; small cracks generated at small loads, which continued to grow with increasing load. A peak “ultimate” load was reached until a sudden failure led to drop of load carrying capacity. The majority of splitting occurred at the bottom corner of the mortise but this depended on the presence and location of knots. The third failure mode was a 10 combined one of localized splitting of each side of the mortise base together with crushing below the mortise. Splitting usually happened after bearing and this can be verified from the graphs. On rare occasions, some beams did show spitting first and then bearing. However, this did not distinctly show up on graphs and were few in number out of the total number of specimens. Thus, they were not characterized as a separate type of failure mode and considered as combined bearing/splitting failure. a) b) c) 35 35 30 30 30 25 25 20 15 20 15 25 20 15 10 10 10 5 5 5 0 0 a) Load [kN] 35 Load [kN] Load [kN] Figure 2-3: Failure Modes (a) Splitting, (b) Bearing, (c) Combined Bearing/Splitting 10 20 Displacement [mm] 0 0 0 b) 10 20 Displacement [mm] 0 c) 10 20 Displacement [mm] Figure 2-4: Typical load-displacement curves for failure modes (a) Splitting, (b) Bearing, (c) Combined Bearing/Splitting (Heal et al., 2019) As in previous research (Nehil & Trojniak, 2013), mortise height had the highest influence on load-carrying capacity. An overall decrease in load-carrying capacity was observed with 11 increase in the mortise height. On the other hand, load-carrying capacity increased with the increase of the mortise width. An increased bearing area in a mortise with greater width was beneficial. Reinforcement led to higher capacities and greater deformation before failure. It was shown that housings with square bases achieved lower load-carrying capacities than mortise with rounded base; however, this mechanism was not studied further. The specimens which were retested after reinforcement showed higher capacities and greater deformation than when they were tested unreinforced, however there was a reduction in stiffness. 2-5 Summary of Literature Review The hygroscopic nature of wood makes it susceptible to changes in MC, which in turn influences the mechanical properties. Reduction in MC slightly increases the compressive strength perpendicular to grain, and significantly the compressive strength parallel to grain. Wood is inherently weak in perpendicular to the grain direction, and drying can induce stresses perpendicular to grain. The limited previous research on housings showed that the ultimate load-carrying capacity increased with increase in mortise depth, mortise width and height below mortise. In addition, capacity improvements were possible with STS reinforcement. Also, use of reinforcement is more effective in improving capacity when used before testing rather than using them to reinforce an already tested beam i.e. retest with reinforcement. However, there are gaps in the literature regarding the understanding of how moisture conditions during fabrication and loading of housing, influence the load-carrying capacity for both unreinforced and reinforced beams. Also, the relationship between load-carrying capacity and properties of timber such as shear strength, compressive strength perpendicular to grain, tension perpendicular to grain and density are unavailable. Such information would eventually make it possible to have detailed design guidelines for traditional housing connections. 12 3 Experimental Investigations 3-1 Materials All timbers tested were Douglas-fir purchased in 2018, grade of No. 1 or better. A paraffin wax end sealer (AnchorSeal) was applied to the cut ends of all members to limit drying defects. The beams were air-dried by leaving them inside the Wood Innovation Research Laboratory (WIRL) for three years (till 2021). The beams were originally in green condition, were cut to length, and the mortises were machined using a CNC machine, with mortise bases squared off by hand. The beams that were cut wet to be tested dry did shrink resulting in different actual depth below the mortise. In addition, the beams also warped along the length. However, it was possible to minimize these in the dry-dry specimens (housings cut dry and tested dry) by making adjustments when cutting the housings. The wet-dry specimens they were tested as is. Moisture content was measured using a Delmhorst RDM-2 pin-type moisture meter prior to testing, with an average moisture content of 29.2% at 25 mm below the surface when it was wet in 2018. During the research presented herein, the mean moisture content was 7.0% for the single housing specimens, while 6.7% for the double housing specimens. The means are based on one reading taken for each beam specimen at a depth of 25 mm from the surface, individual values varied between 6.2% to 7.4% and 6.1% to 7.5% for single housing specimens and double housing specimens, respectively. While CSA-086 provides specified design values for Douglas Fir-Larch, beam and stringer grades for longitudinal shear (fv = 1.5 MPa), compression perpendicular to the grain (fcp = 7.0 MPa), tension perpendicular to the grain (ft = 2.0 MPa, value from Glulam). These values were adjusted to the 7% moisture content using the Wood Handbook procedure (United States Department of Agriculture, 2010) as 7.5 MPa, 2.4 MPa and 8.6 MPa for compression perpendicular to grain, tension perpendicular to grain, and shear parallel to grain, respectively. 13 Herein, the actual average strengths in shear parallel to grain, compressive and tensile perpendicular to grain were determined on small clear samples cut from the beams with housings after testing. The two beams with the highest and the lowest load-carrying capacity (obtained from the housing tests) in the unreinforced state were identified from each series (resulting in 20 specimens), and samples were cut following the specifications of ASTM D143 (2021). For the compression tests, 20 beams, which exhibited either bearing or combined bearing/splitting failure, were identified along with 6 beams, which exhibited splitting failure. Three samples were tested for each property (compression, tension and shear) from each beam, for a total of 198 small clear specimens. The small clear specimens were obtained at random locations along the beam to assure these were without knots, cracks or other defects. The corresponding test specimens and the test assemblies for each test are shown in Figure 3-1. ASTM D143 (2021) was followed for testing. The compression, tension and shear tests were conducted using rates of loading of 2.0 mm/min, 2.5 mm/min and 0.6 mm/min respectively. For tension and shear tests, strengths were determined at failure, whereas for the compression tests, strength was determined at displacements of 2.5 mm as per ASTM D 143 (2021), and at displacements of 5.0 mm (to get additional perspective). a) b) c) Figure 3-1: Test Assembly for (a) Compression Perpendicular to Grain Test (b) Tension Perpendicular to Grain Test (c) Shear Test 14 For reinforcement, fully-threaded STS (Heco-Topix) with product approval ETA-11/0284 (DIBt, 2019) with 8.0 mm diameter and lengths of 120 mm, 180 mm or 200 mm, as shown in Figure 3-2, were selected. Length of reinforcement was at least double of height below mortise i.e. 120 mm STS for 57 mm or 47 mm depth, 180 mm STS for 82 mm depth and 200 mm STS for 100 mm depth below the mortise. So, the difference in reinforcement length was as required to match the difference in depth below the mortise. Two screws were used for each connection as per the product approval. For the selected diameter, the screws have characteristic tensile strength of 20 kN and withdrawal resistance of 22.7 kN for S-0.65 series and 13.7 kN for S0.80 series (calculated based on ETA specifications). For calculations embedment length was assumed to be the difference between screw length and height below mortise i.e. for S-0.80 it is 63 mm (120 mm of screw – 57 mm of length below mortise) and for S-0.65 it is 100 mm (200 mm of screw – 100 mm below mortise). These values were assumed irrespective of failure modes and location of splitting failure (if applicable) as the exact location of splitting was not recorded. Hence, for computational convenience it was assumed that the splitting was below the mortise corner. The screws were installed at the minimum edge distances (S) away from the mortise base, 3d, but which was taken as 25 mm for convenience. Figure 3-2: Fully threaded STS of 120mm, 180mm and 200mm from left to right 15 3-2 Specimen Description and Test Matrix Similar to the preceding work at UNBC, simply supported beam specimens with single-sided and double-sided housings were tested. For the single housings, 10 test series with 6 replicates each, therefore 60 specimens were manufactured. Whereas, the double housings have 6 test series each with 5 replicates, therefore a total of 30 specimens as shown in Figure 3-3. The single housing specimens had a length of 1,830 mm, width of 100 mm, beam height of 250 mm with mortise cut in to them having depth of 25 mm and width of 89 mm. The mortise height varied among the series, with the depth below mortise either 57 mm for S-0.80 series or 100 mm for S-0.65 series. The double housing specimen length was 1,825 mm, width was 130 mm, height was 225 mm, with the same mortise width and depth as the single housing connections. However, the depths below the mortise were 47 mm for D-0.80 series and 82 mm for D-0.65 series. For S-0.80, 80% mortise height would mean 0.8 × 250 = 200 mm, hence height below mortise should have been 50 mm, but the actual height was 57 mm. Similarly, for S-0.65 it should have been 87.5 mm, but it was actually 100 mm. Similar double housing depths below the mortise should have been different that those used. These 57 mm and 100 mm for the single housings and 47 mm and 82 mm for the double housings were maintained to match the heights used in the previous research by Heal et al. (2019). a) b) Figure 3-3: Schematic of a) single housing and b) double housing 16 Each housing was reinforced with two STS (i.e., two STS for single housing and four STS for double housing series), as shown in Figure 3-4. Reinforcement was installed either for specimens that were already tested or – in some double housing series – before testing. All parameters are summarized in Table 3-1. a) b) Figure 3-4: Single & Double Housing Specimens with Reinforcements The naming convention (ID) of each test series is explained in Figure 3-5. The first letter refers to single (S) or double housing (D), the decimal number 0.80 or 0.65 (80% or 65%) refers to housing depth. The third term (WW, WD or DD) refers to the moisture conditions (wet or dry) during fabrication (first letter) and testing (second letter). Exception to this exists in the form of WWD where WW specimens were retested with reinforcement under dry conditions. The last part refers to reinforcement: none (blank), retested with STS (RR) or tested reinforced (R). Figure 3-5: Specimen naming convention 17 Table 3-1: Overview test series # of Tests Mortise Backcut Moisture Condition Reinforcement S-0.65-WWD-RR 6 100 mm, 0.65 Cut Wet, tested Wet Reinforced with Heco-Topix 8.0 Ø × 200mm & retested S-0.8-WWD-RR 6 57 mm, 0.8 Reinforced with Heco-Topix 8.0 Ø × 120mm & retested S-0.65-WD 6 100 mm, 0.65 Cut Wet, tested Wet Cut Wet, tested Dry S-0.65-WD-RR 6 100 mm, 0.65 Cut Wet, tested Dry Reinforced with Heco-Topix 8.0 Ø × 200mm & retested S-0.80-WD 6 57 mm, 0.8 S-0.80-WD-RR 6 57 mm, 0.8 S-0.65-DD 6 100 mm, 0.65 S-0.65-DD-RR 6 S-0.80-DD Series ID Unreinforced Cut Wet, tested Dry Cut Wet, tested Dry Cut Dry, tested Dry Reinforced with Heco-Topix 8.0 Ø × 120mm & retested 100 mm, 0.65 Cut Dry, tested Dry Reinforced with Heco-Topix 8.0 Ø × 200mm & retested 6 57 mm, 0.8 Cut Dry, tested Dry Unreinforced S-0.80-DD-RR 6 57 mm, 0.8 Cut Dry, tested Dry Reinforced with Heco-Topix 8.0 Ø × 120mm & retested D-0.65-DD 5 82 mm, 0.65 Cut Dry, tested Dry Unreinforced D-0.65-DD-RR 5 82 mm, 0.65 Cut Dry, tested Dry Reinforced with Heco-Topix 8.0 Ø × 180mm & retested D-0.65-DD-R 5 82 mm, 0.65 Cut Dry, tested Dry Reinforced prior to Testing with Heco-Topix 8.0 Ø × 180mm D-0.80-DD 5 47 mm, 0.8 Cut Dry, tested Dry Unreinforced D-0.80-DD-RR 5 47 mm, 0.8 Cut Dry, tested Dry Reinforced with Heco-Topix 8.0 Ø × 120mm & retested D-0.80-DD-R 5 47 mm, 0.8 Cut Dry, tested Dry Reinforced prior to Testing with Heco-Topix 8.0 Ø × 120mm Unreinforced Unreinforced 3-3 Test Methods The housing specimens were tested in a Universal Test Machine (UTM) using a custom loading frame, see Figure 3-6a, such that the mortise was loaded in shear only. The force from the secondary beam was simulated by use of loading block inside the mortise, made of 18 Laminated Veneer Lumber (LVL). Restraints were provided on the outside surface to ensure that they did not deflect outwards when loads were applied to the mortise. EN 26 891 (CEN, 1991) was followed for loading where the specimens were initially loaded to approx. 40% of their expected load-carrying capacity to place the loading block properly on the mortise and thus reducing any alignment problems. After holding this 40% load for 30 seconds, the beam was unloaded to 10%, with that load kept constant for another 30 seconds. Afterwards, the beam was loaded to failure, with failure defined as 80% drop in the load-carrying capacity (F ). A displacement controlled loading rate of 5.0 mm/min was used for all the 90 specimens. In case of specimens where no sudden drop of failure was achieved due to nature of failure mode, loading was stopped when capacity continued to drop consistently. a) b) b) Figure 3-6: Experimental setup: a) photo, b) detail sensor, and c) schematic 19 The load-displacement data were obtained from the UTM data acquisition system, while the crack opening data were collected using string pots attached close to the housings as shown in Figure 3-6b. When the housing was loaded, it would cause a crack to develop near the base of the mortise. The relative displacement between was measured. The screw was placed mid-way below the base of mortise for all test series. Two sets of data were obtained from the two sensors with the results being averaged to get mean crack opening for each test specimen. Bearing capacities were estimated by multiplying the compressive strength with the bearing area (89 mm × 25 mm) of the housing. 3-4 Statistical Analyses In this research, analysis of variance (ANOVA) (Montgomery & Runger, 2003) was conducted to determine the statistical significance of mortise depth, moisture condition and reinforcement along with their interactions on . In order to accept or reject a null hypothesis (normally stating that there is no effect of the factor), a p-value is computed. A significance level (α) of 0.05 which is typical in engineering practice (Tannert et al., 2012) was selected to compare the p value against. For the single housing tests, a three-way factorial design was chosen. The first factor, mortise depth had two levels (S-0.80 and S-0.65). The second factor, moisture conditions, had three levels (WW, WD and DD) and the third factor, reinforcement, had again two levels (unreinforced and reinforced). For double housings, two-way ANOVA was performed twice: 1) with mortise depths (two levels: D-0.80 and D-0.65) and reinforcement (three levels: unreinforced, retested with reinforcement and reinforced) as factors. This analysis was only for dry (DD) specimens of all the double housing series; and 2) for the D0.80 series with different moisture conditions (two levels: WW and DD) and reinforcement (three levels: unreinforced, retested with reinforcement and reinforced) as factors. The test results for WW series for both the single and double housing series were taken from previous study in 2018 at UNBC (Heal et al., 2019) and all the rest from this research. 20 3-5 Results 3-5-1 Compression, Tension and Shear Tests The small clear specimen tests resulted in the failure modes shown in Figure 3-7. In the compression perpendicular to the grain tests, as per ASTM D143, the load at 2.5 mm compression was recorded as strength, as shown in Figure 3-7 (a). In the tension perpendicular to grain tests, all specimens showed distinct failure at the middle of the cross-section, as shown Figure 3-7 (b). In the shear tests, all specimens failed by fracture through the shear plane with the majority completely separating into two pieces, as shown in Figure 3-7 (c). a) b) c) Figure 3-7: Failure modes of Compression, Tension and Shear Tests The weights and volumes were measured for the specimens from each of the selected representative beams to calculate the apparent density. These densities and the recorded mean strengths are presented in Table 3-2 for compression perpendicular to grain tests, Table 3-3 for tension perpendicular to grain tests and Table 3-4 for shear tests. The specimens marked with 21 asterisk (*) refer to the specimens taken from beams tested in the previous research by Heal (Heal et al., 2019) and the numbers ‘1-6’ refer to the small clear specimens being collected from the replicate number for that series. These data are later used to correlate the specimen strength to the load-carrying capacity of then housing and the density of timber. The mean compressive strengths at 2.5 mm displacements were 9.8 MPa with a coefficient of variation (CV) of 16% for the S-0.80 series, 10.4 MPa (18%) for the S-0.65 series, 9.5 MPa (24%) for the D-0.80 series and 8.1 MPa (16%) for the D-0.65 series. The corresponding mean densities were determined as 457 kg/m3, 435 kg/m3, 528 kg/m3 and 503 kg/m3. The mean tensile strengths were determined 1.9 MPa with a CV of 20% for the S-0.80 series, 2.2 MPa (17%) for the S-0.65 series, 2.0 MPa (35%) for the D-0.80 series, and 1.9 MPa (27%) for the D-0.65 series. The corresponding mean densities were determined as 465 kg/m 3, 525 kg/m3, 518 kg/m3 and 535 kg/m3. The mean shear strengths were determined as 8.6 MPa (10%) for the S-0.80 series, 9.7 MPa (12%) for the S-0.65 series, 10.0 MPa (9%) for the D-0.80 series and 10.5 MPa (15%) for the D-0.65 series. The corresponding mean densities were 456 kg/m 3, 511 kg/m3, 495 kg/m3 and 524 kg/m3. The compressive strengths of single housing specimens (S-0.65 and S-0.80 series) were used to estimate the bearing capacity of the housing to compare with the load-carrying capacity of the housing as shown in Table 3-5. It was observed that the mean differences between bearing capacities and load-carrying capacities for splitting, bearing and mixed bearing/splitting failure modes were respectively 31%, 10% and 24%. This indicates that compressive strength of wood is certainly crucial for the specimens which failures in bearing failure modes. 22 Table 3-2: Compression Test Results ID Density [kg/m3] Strength at 2.5 mm [MPa] CV [%] Strength at 5 mm [MPa] CV [%] S-0.80-WW* ‘2’ 420 9.5 5 10.1 2 S-0.80-WW* ‘3’ 517 11.2 1 12.8 2 S-0.80-WD ‘2’ 492 11.4 1 12.2 2 S-0.80-WD ‘5’ 500 11.6 2 13.3 4 S-0.80-DD ‘1’ 420 7.9 4 9.0 5 S-0.80-DD ‘2’ 419 8.7 2 9.5 2 S-0.80-DD ‘6’ 428 8.5 2 9.6 4 Average 457 9.8 16 10.9 16 S-0.65-WW* ‘1’ 396 8.8 6 9.4 6 S-0.65-WW* ‘4’ 532 14.3 2 16.7 1 S-0.65-DD ‘5’ 428 8.5 3 9.3 2 S-0.65-WW* ‘3’ 419 8.5 3 9.3 3 S-0.65-WW* ‘6’ 428 11.5 1 13.4 1 S-0.65-WD ‘2’ 425 9.1 2 10.2 1 S-0.65-WD ‘4’ 426 9.6 4 10.8 5 S-0.65-DD ‘2’ 425 11.3 4 12.4 7 S-0.65-DD ‘4’ 418 11.7 1 13.8 4 S-0.65-DD ‘6’ 453 11.0 2 12.1 1 435 10.4 18 11.7 21 Average D-0.80-DD ‘3’ 452 7.0 4 7.6 5 D-0.80-DD ‘5’ 556 9.0 1 9.3 3 D-0.80-DD-R ‘3’ 525 12.5 3 15.4 4 D-0.80-DD-R ‘2’ 577 9.5 7 10.0 3 528 9.5 24 10.6 32 Average D-0.65-DD ‘5’ 459 9.0 3 10.4 3 D-0.65-DD ‘4’ 559 9.7 4 11.4 4 D-0.65-DD-R ‘3’ 430 6.9 2 7.3 2 D-0.65-DD-R ‘2’ 478 6.7 2 7.1 0 D-0.65-DD-R ‘5’ 589 8.3 5 8.5 4 503 8.1 16 8.9 21 Average 23 Table 3-3: Tension Test Results ID Density [kg/m3] Strength [MPa] CV [%] S-0.80-WW* ‘2’ 454 1.7 22 S-0.80-WW* ‘3’ 537 2.5 67 S-0.80-WD ‘2’ 500 1.9 17 S-0.80-WD ‘5’ 422 2.1 3 S-0.80-DD ‘1’ 428 1.4 7 S-0.80-DD ‘2’ 448 1.8 21 465 1.9 20 Average S-0.65-WW* ‘1’ 431 1.6 17 S-0.65-WW* ‘4’ 564 2.1 34 S-0.65-WD ‘5’ 549 2.8 2 S-0.65-WD ‘6’ 578 2.1 9 S-0.65-DD ‘3’ 552 2.4 5 S-0.65-DD ‘5’ 480 2.4 5 525 2.2 17 Average D-0.80-DD ‘3’ 480 1.0 15 D-0.80-DD ‘4’ 500 2.4 12 D-0.80-DD-R ‘1’ 523 2.5 2 D-0.80-DD-R ‘3’ 569 1.9 10 518 2.0 35 Average D-0.65-DD ‘2’ 566 1.7 9 D-0.65-DD ‘5’ 503 1.3 43 D-0.65-DD-R ‘3’ 506 2.5 8 D-0.65-DD-R ‘4’ 566 1.9 23 535 1.9 27 Average 24 Table 3-4: Shear Test Results ID Density [kg/m3] Strength [MPa] CV [%] S-0.80-WW* ‘2’ 431 8.9 4 S-0.80-WW* ‘3’ 531 9.5 2 S-0.80-WD ‘2’ 524 9.5 8 S-0.80-WD ‘5’ 424 7.5 3 S-0.80-DD ‘1’ 397 7.6 3 S-0.80-DD ‘2’ 431 8.8 8 Average 456 8.6 10 S-0.65-WW* ‘1’ 415 7.8 3 S-0.65-WW* ‘4’ 554 9.4 9 S-0.65-WD ‘5’ 537 10.7 7 S-0.65-WD ‘6’ 577 10.9 12 S-0.65-DD ‘3’ 540 10.2 3 S-0.65-DD ‘5’ 445 8.8 3 511 9.7 12 Average D-0.80-DD ‘3’ 475 8.8 2 D-0.80-DD ‘4’ 494 10.2 3 D-0.80-DD-R ‘1’ 484 10.9 2 D-0.80-DD-R ‘3’ 526 10.0 5 Average 495 10.0 9 D-0.65-DD ‘2’ 572 11.1 6 D-0.65-DD ‘5’ 471 8.6 8 D-0.65-DD-R ‘3’ 461 10.0 3 D-0.65-DD-R ‘4’ 593 12.4 12 Average 524 10.5 15 25 Table 3-5: Estimated Bearing Capacity ID Failure Modes Compression Strength at 2.5 mm [MPa] Bearing Capacity [kN] Fult [kN] % Difference S-0.80-WW* 2/6 Splitting 9.5 21.1 10.9 48 S-0.80-WW* 3/6 Splitting 11.2 25.0 17.9 28 S-0.80-WD 2/6 Splitting 11.4 25.3 17.7 30 S-0.80-WD 5/6 Splitting 11.6 25.7 13.9 46 S-0.80-DD 1/6 Splitting 7.9 17.6 13.8 21 S-0.80-DD 2/6 Splitting 8.7 19.4 16.6 14 S-0.65-WW* 1/6 Bearing 8.8 19.5 17.2 12 S-0.65-DD 5/6 Bearing 8.5 19.0 15.5 18 S-0.65-WW* 3/6 Bearing 8.5 19.0 20.8 -10 S-0.65-WD 2/6 Bearing 9.1 20.3 16.8 17 S-0.65-WD 4/6 Bearing 9.6 21.3 21.4 0 S-0.65-DD 2/6 Bearing 11.3 25.2 18.9 25 S-0.65-DD 6/6 Bearing 11.0 24.4 22.4 9 S-0.80-DD 6/6 Bearing/Splitting 8.5 19.0 14.5 24 S-0.65-WW* 4/6 Bearing/Splitting 14.3 31.8 22.6 29 S-0.65-WW* 6/6 Bearing/Splitting 11.5 25.7 19.0 26 S-0.65-DD 4/6 Bearing/Splitting 11.7 26.0 21.0 19 Mean % 3-5-2 Single Housing Beams Three different failure modes: splitting, bearing, and combined bearing/splitting were observed as shown in Figure 3-8; typical load-displacement curves corresponding to these failures are shown in Figure 3-9. The splitting failures were typically characterized by reaching loadcarrying capacity followed by sudden drop in load. Horizontal splits along the length and width of beams were observed, mostly at the mortise base, but in some specimens also at higher locations. The bearing failures were not characterised by a distinct drop after reaching loadcarrying capacity as the load was usually maintained due to continuous crushing and densification of wood fibres at the base of the mortise. Hence, tests were stopped after 6 26 31 10 24 minutes or when localized tear-out occurred, whichever came first. Otherwise, a combined failure, with mostly splitting happing after bearing occurred, where the load continued to increase due to densification of the wood fibres followed by sudden drop at load-carrying capacity similar to pure splitting. However, the splitting observed here were mostly limited to corners of the mortise base. The mean values along with their CVs for density, load-carrying capacity (F ), as well as mid-span displacement and crack opening at F are presented in Table 3-6. All individual results for series S.80 are provided in Appendix 2, and for series S-0.65 in Appendix 3. a) b) c) Figure 3-8: Failure Modes: a) Splitting, b) Bearing c) Combined Bearing/Splitting 25 S-0.80-WD-RR-4 (Bearing-Splitting) Load [kN] 20 S-0.80-WD-RR-6 (Bearing) 15 10 S-0.80-DD-4 (Splitting) 5 0 0 5 10 15 Displacement [mm] 20 25 Figure 3-9: Typical load-displacement curves for different failure modes 27 Table 3-6: Load-carrying capacity, crack opening, and density: single housing specimens ID Density [kg/m3] Fult [kN] Disp. [mm] Crack Opening [mm] S-0.80-WWD-RR 479 9% 17.3 26% 10.7 23% 2.0 40% S-0.80-WD 455 10% 15.4 9% 7.2 12% 0.3 27% S-0.80-WD-RR 455 10% 20.6 16% 16.6 28% 1.9 39% S-0.80-DD 451 8% 14.9 8% 7.0 22% 0.8 99% S-0.80-DD-RR 451 8% 19.0 17% 10.0 28% 1.5 16% S-0.65-WWD-RR 475 11% 26.5 16% 23.6 17% 2.4 27% S-0.65-WD 479 10% 18.9 14% 11.9 47% 0.5 41% S-0.65-WD-RR 479 10% 26.0 18% 18.8 33% 1.2 37% S-0.65-DD 490 9% 21.9 21% 12.4 45% 0.9 71% S-0.65-DD-RR 490 9% 23.0 30% 14.8 30% 1.0 72% The mean density varied from 451 kg/m3 for S-0.80-DD to 490 kg/m3 for S-0.65-DD with lowest CV (8%) for S-0.80-DD and highest CV (11%) for S-0.65-WWD-RR. The mean F ranged from 14.9 kN for S-0.80-DD to 26.5 kN for S-0.65-WWD-RR, where lowest CV (8%) was obtained for S-0.80-DD and largest CV (30%) was for S-0.65-DD-RR. In case of beam displacement, lowest was 7.0 mm for S-0.80-DD and highest was 23.6 mm for S-0.65-WWDRR. However, lowest CV (12%) was not for S-0.80-DD as in previous two parameters, but for S-0.80-WD while largest CV (47%) was for S-0.65-WD. Mean crack opening varied largely from 0.3 mm for S-0.80-WD to 2.4 mm for S-0.65-WWD-RR. The lowest CV (16%) was for S-0.80-DD-RR and largest CV (99%) was for S-80-DD. High CV indicated that the crack opening varied significantly. Although crack opening is shown for all specimens (to maintain consistency with previous research), practically for beams with bearing failure this represents how much the end of the sensor string displaced vertically. For beams with splitting or bearing/splitting it is assumed that sensor provides the crack opening as string pot is fixed at the beam top while the end of string is placed below the mortise base. 28 Specimens from the S-0.65 series reached higher load-carrying capacity and beam displacement at failure than the specimens from the corresponding S-0.80 series within the same moisture conditions. Re-testing with reinforcement increased the capacity irrespective of the moisture conditions. It is worth noting that the CV (for capacity) is higher when reinforced, suggesting large dispersion of the results around mean. It is also interesting to note that CV of mean crack opening is large for all series and hence obtaining precise measurements is quite difficult for this parameter. Calculating the average CV for S-0.80 series and S-0.65 series shows that for both load-carrying capacity and beam displacement, S-0.65 has larger CV (20% and 34% respectively), than S-0.80 (15% and 23% respectively). Hence, failure mechanism in S-0.65 series is such that it has higher dispersion of obtained results. Representative graphs for each test series are shown in Figure 3-10 and Figure 3-11. All individual load-displacement and load-crack opening graphs are also provided for both series in Appendix 2 and 3. It is seen in Figure 3-10 (a) that specimens with reinforcement (S-0.80WD-RR-1, S-0.80-DD-RR-4 and S-0.80-WWD-RR-2) have higher load-carrying capacity than the unreinforced ones (S-0.80-WD-3 and S-0.80-DD-4), and also fail after significant displacement of the beam. The trend is more prominent when load vs mean crack-opening curves are considered as shown in Figure 3-10 (b). In the reinforced specimens, the beams reach load capacity when the cracks have opened up significantly more than the unreinforced beams. It is observed in the representative curves shown here in Figure 3-11 (which is also applicable to individual curves provided in the appendixes), the load-displacement path for S0.65 series is much more variable than those of the S-0.80. However, similar to S-0.65 series reinforcement increases the load-carrying capacity with possibility of reaching greater displacement at capacity load. Similar to the unreinforced specimens, the reinforced specimens also show wide range of variation in results. 29 S-0.80-WD-3 S-0.80-DD-4 S-0.80-WWD-RR-2 S-0.80-WD-RR-1 S-0.80-DD-RR-4 S-0.80-WD-3 S-0.80-DD-4 S-0.80-WWD-RR-2 30 30 25 25 20 20 Load [kN] Load [kN] S-0.80-WD-RR-1 S-0.80-DD-RR-4 15 10 15 10 5 5 0 0 0 5 10 15 20 Displacement [mm] 25 0 30 1 2 3 Crack Opening [mm] 4 5 a) b) Figure 3-10: Load-displacement tests of typical 0.80 series under bearing load (a) Load vs Beam Displacement (b) Load vs Mean Crack Opening 30 30 25 25 20 20 15 10 S-0.65-WD-RR-3 S-0.65-DD-RR-6 15 10 5 a) S-0.65-WD-3 S-0.65-DD-6 S-0.65-WWD-RR-3 S-0.65-WD-RR-3 S-0.65-DD-RR-6 Load [kN] Load [kN] S-0.65-WD-3 S-0.65-DD-6 S-0.65-WWD-RR-3 5 0 0 5 10 15 20 25 30 Displacement [mm] b) 0 0 5 10 15 20 Crack Opening [mm] 25 Figure 3-11: Load-displacement tests of typical 0.65 series under bearing load (a) Load vs Beam Displacement (b) Load vs Mean Crack Opening 3-5-3 Double Housing Beams Similar to the specimens with singe housing, the specimens with double housings failed in three different modes: splitting, bearing, and combined splitting/bearing, as shown in Figure 3-12. However, the presence of two housings did complicate the observations made in the form of two different sides seemingly failing in two different modes. Hence, conclusion on the global mode of failure for each specimen was to some extent subjective in some cases. 30 a) b) c) Figure 3-12: Failure Modes Observed (a) Splitting Failure, (b) Bearing Failure, (c) Combined Bearing/Splitting Failure The mean values along with their CVs for density, load-carrying capacity (F ), as well as mid-span displacement and crack opening at F are presented in Table 3-7. All individual results are included in Appendix 4 for D-0.80 series and Appendix 5 for D-0.65 series. Similar to the single housing series, specimens in the double housing series with greater height below housing (i.e. in the D-0.65 series) have higher load-carrying capacity and beam displacement at failure than the specimens from the corresponding D-0.80 series. Also, retesting with reinforcement helped in reaching greater capacity and displacement. However, reinforcing the beams first and then testing them lead to even higher increase of capacity, at a lower displacement irrespective of housing depth. It was not possible to directly compare between single and double housing capacities (i.e. whether the capacity in double housing is double of the single housing capacity) because the depths below the mortise were not identical. The density varied within a close range of 503 kg/m3 (CV of 6%) to 515 kg/m3 (CV of 9%) for D-0.80-DD-R and D-0.65-DD-R, respectively. The load-carrying capacity ranged from 22.0 kN (CV of 52%) to 49.6 kN (CV of 14%) for D-0.80-DD and D-0.65-DD-R, respectively. Beam displacement ranged from 8.6 mm (20%) for D-0.80-DD to 20.0 mm (29%) for D-0.65DD-RR. Similar to the single housings, double housings also had large variation in crack openings; values ranged from 0.9 mm (largest CV of 108%) for D-0.65-DD-R to 8.2 mm (CV 31 of 81%) for D-0.80-DD-RR. The mean crack opening results being most dispersed with large CV in general. It is interesting to note that the specimens, which were reinforced and then tested (D-0.80-DD-R and D-0.65-DD-R) had the lowest CV (9% and 14% respectively). Hence, reinforcing the specimens before testing ensured lower variations in capacity compared to testing an unreinforced beam or retesting a beam with reinforcement. Table 3-7: Load-carrying capacity, beam displacement and crack opening: double housing ID Density [kg/m3] Fult [kN] Disp. [mm] Crack Opening [mm] D-0.80-DD 510 6% 22.0 52% 8.6 20% 3.7 86% D-0.80-DD-RR 510 6% 26.1 28% 13.7 41% 8.2 81% D-0.80-DD-R 503 6% 32.7 9% 10.7 12% 1.0 33% D-0.65-DD 506 6% 44.3 19% 15.6 41% 1.2 91% D-0.65-DD-RR 506 6% 45.6 24% 20.0 29% 1.9 29% D-0.65-DD-R 515 9% 49.6 14% 17.5 37% 0.9 108% Representative graphs for each test series are shown in Figure 3-13 and Figure 3-14. Individual graphs are provided in the appendix 4 for D-0.80 series, and appendix 5 for D-0.65 series. The reinforced series (D-0.80-DD-R-4) showed higher capacity than unreinforced (D-0.80-DD-1) or series retested with reinforcement (D-0.80-DD-RR-1). The reinforcing seemed to reduce ductility; failure occurred without the crack opening up significantly. The STS reduced the opening width at failure. Ductility would give warning of failure and thus be preferable over a sudden failure observed in unreinforced housings. A similar trend was observed in Figure 3-14 with reinforced series (D-0.65-DD-R-1) showing higher capacity than the unreinforced (D0.65-DD-4) and retested beam with reinforcement (D-0.65-DD-RR-2). Mean crack opening stays low for the reinforced series (D-0.65-DD-R-1) with the retested beam reinforcement (D0.65-DD-RR-2) showing very large crack openings before failure. 32 Load [kN] D-0.80-DD-1 D-0.80-DD-RR-1 D-0.80-DD-R-4 D-0.80-DD-1 D-0.80-DD-RR-1 D-0.80-DD-R-4 40 35 30 25 20 15 10 5 0 D-0.80-DD-1 D-0.80-DD-RR-1 D-0.80-DD-R-4 40 35 30 25 20 15 10 5 0 0 2 4 6 8 Displacement [mm] 10 40 35 30 25 20 15 10 5 0 0 2 4 6 8 10 0 Crack Opening #1 [mm] 2 4 6 8 10 Crack Opening #2 [mm] a) b) c) Figure 3-13: Load-displacement of D-0.80 series: a) Load vs Beam Displacement (b) Load vs Crack Opening for Side #1, c) Load vs Crack Opening for Side #2 Load [kN] D-0.65-DD-4 D-0.65-DD-RR-2 D-0.65-DD-R-1 70 70 70 60 60 60 50 50 50 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 a) D-0.65-DD-4 D-0.65-DD-RR-2 D-0.65-DD-R-1 D-0.65-DD-4 D-0.65-DD-RR-2 D-0.65-DD-R-1 5 10 15 20 25 30 Displacement [mm] 0 0 b) 5 10 15 20 25 30 Crack Opening #1 [mm] 0 c) 5 10 15 20 25 30 Crack Opening #2 [mm] Figure 3-14: Load-displacement of D-0.65 series: a) Load vs Beam Displacement, b) Load vs Crack Opening for Side #1, c) Load vs Crack Opening for Side #2 3-6 Discussion 3-6-1 Correlation between material strengths and housing load-carrying capacity The Wood Handbook (United States Department of Agriculture, 2010) tabulated mean strengths for compression perpendicular to grain, tension perpendicular to grain and shear parallel to grain at 7% moisture content were 7.5 MPa, 2.4 MPa and 8.6 MPa, respectively. The experimentally determined compressive strength values were on average 35% higher: 9.8 MPa and 10.4 MPa for single housings, S-0.80 and S-0.65 series, respectively. Similarly, the 33 shear strengths (8.6 MPa and 9.7 MPa) were higher. On the contrary, the tensile strengths (1.9 MPa and 2.2 MPa) were on average 15% lower. The results from the small specimen tests (compression, tension and shear) were correlated to the housing load-carrying capacity (Figure 3-15) and densities (Figure 3-16). The loadcarrying capacity values were normalized for variation in depths of mortise (0.65 and 0.80) and number of housings (single or double housing). The R2 value which describes the proportion of the variation in the dependent variable (i.e., load-carrying capacity) which can be attributed to the independent variable (i.e., compressive strength/tensile strength/shear strength) are low for all three material strengths having R2 values of 0.20, 0.24 and 0.25 respectively. Shear Tension Compression Load-carrying capacity [kN] 35 30 R² = 0.2421 R² = 0.2471 25 R² = 0.1985 20 15 10 0 3 6 9 Strength [MPa] 12 15 Figure 3-15: Relation between capacity and material strength Figure 3-16 illustrates that compressive and shear strengths are more strongly correlated to specimen densities with higher R2 values of 0.51 and 0.59 respectively. However, tensile strength values are less dependent on densities with lower R2 value of 0.21, demonstrating the high variability involved of tensile strength of timber perpendicular to the grain direction. 34 16 R² = 0.5142 Strength [MPa] 12 R² = 0.5859 8 Shear Tension 4 Compression R² = 0.2126 0 350 400 450 500 550 600 Density [kg/m3] Figure 3-16: Relation between material strength and density To get a better perspective on the relationship between material strengths and failure modes of the beam from which the small specimens were prepared, data were analysed of failures modes of the unreinforced single housing beams, see Figure 3-17. As before the load-carrying capacity values were normalized for variation in depths of mortise (0.65 and 0.80) and number of housings (single or double housing). For the compressive strength, 21 beam specimens were considered with 6 splitting failures, 10 bearing failures. and 5 mixed failures. For the shear and tensile strengths, 16 beam specimens were considered with 11 splitting failures, 4 bearing failures and 1 combined failure. A limitation that should be noted is that all three tests were conducted on dry specimens whereas the load-carrying of beams included beams tested wet which accounted for 30% of the small specimens tested (for compression) and 25% (for shear and tension). While the compressive strength is expected to be an indicator for possibility of bearing failure modes (if beam have higher depth below the mortise) based on the relatively lower difference between expected bearing capacity and housing capacity as seen in Table 3-5 , the low R2 values in Figure 3-17 confirm that the small clear specimen strengths are not a good predictor of housing load-carrying capacity of timber housings. This is most applicable for tensile strength which has the lowest R2 value of 0.1. All data were normalised as before. 35 Load-carrying capacity [kN] 24 R² = 0.2522 22 20 18 R² = 0.1408 16 Splitting Failures 14 Bearing and Mixed Bearing/Splitting Failures 12 6 8 a) 10 12 14 Compressive strength [MPa] 16 35 Load-carrying capacity [kN] 30 R² = 0.0907 25 20 R² = 0.4503 15 Splitting Failures 10 Bearing and Mixed Bearing/Splitting Failures 5 1.0 1.5 b) 2.0 2.5 Tensile strength [MPa] 3.0 Load-Carrying Capacity [kN] 35 30 R² = 0.3118 25 20 R² = 0.0632 15 Splitting Failures 10 Bearing and Mixed Bearing/Splitting Failures 5 7 c) 8 9 10 Shear Strength [MPa] 11 12 Figure 3-17: Capacity vs Mechanical Strength based on Failure Modes 36 3-6-2 Single Housing Beams Analysis of the failure modes, (photos of each failed specimen are included in the appendices) reveal that for S-0.80 series the proportion of splitting, bearing, combined bearing/splitting and failures are respectively 64%, 0% and 36%, see Table 3-8. Whereas, for the S-0.65 series the proportions are respectively 28%, 44% and 28%. It is evident that the S-0.80 series have higher proportion of splitting failures owing to less material below the mortise. The S-0.65 series with thicker depth below the mortise exhibit lower proportion of splitting failures. On closer inspection into the unreinforced specimens (WD and DD), the arguments are even more conclusive with splitting accounting for 89% of failure in S-0.80 series and only 33% in the S0.65 series. However, reinforcing the specimens significantly reduces the splitting failures to 39% and 22% respectively. It should be taken into consideration that the splitting failures are different from the previous splitting failures as the resistance perpendicular to grain now includes the additional withdrawal capacity of the STS. Hence, the use of fully threaded STS in housing connections load them in withdrawal shifting the connection failure modes towards bearing or combined bearing/splitting which are primarily governed by crushing of wood fibers under compressive stresses. Specimens which still failed by splitting after reinforcement, the peak load corresponds to withdrawal failure. Table 3-8: Proportion of failure modes: single housing ID Splitting Bearing Bearing/Splitting S-0.80 (All) 64% 0% 36% S-0.65 (All) 28% 44% 28% S-0.80 (Unreinforced) 89% 11% S-0.80 (Reinforced) 39% 61% S-0.65 (Unreinforced) 33% 67% S-0.65 (Reinforced) 22% 78% 37 Figure 3-18 presents the relation between capacity and mortise depth. The mean load-carrying capacity while ignoring variation in moisture and reinforcement conditions (mean of unreinforced and reinforced WW, WD and DD specimens) are 17.0 kN (S-0.80 series) and 22.7 kN (S-0.65 series). Hence, greater height below the mortise in S-0.65 series resulted in an increase of load-carrying capacity by 34%. Thus, confirming the findings of previous research (Heal et al., 2019) that depth of material below mortise directly influences the load-carrying capacity of housing in beams. Less material below the mortise increases the likelihood of separation of wood fibers induced by tension perpendicular to the grain. Moisture reduction induced internal cracks are widely known in literature as shown by Rammer & Winistorfer (2001), Franke & Franke (2014), Dietsch (2017), etc. The higher capacity in S-0.65 is primarily due to the compressive strength perpendicular to grain governing the failure mechanism. WW WD DD Load-carrying capacity [kN] 25 20 15 10 5 0 S-0.80 Mortise depth S-0.65 Figure 3-18: Relation between Load-carrying capacity and Mortise Depth (Single Housings) Inspection of the influence of reinforcement, illustrated in Figure 3-19, showed that STS increased the load-carrying capacity in all specimens irrespective of moisture conditions (WW, WD or DD) and mortise depth (S-0.80 or S-0.65). However, for both S-0.80 and S-0.65 series, the largest increase was observed in the wet-wet (WW) series (34% and 38% respectively) 38 along with less brittle failure modes in the form of higher displacement achieved. The corresponding dry-dry (DD) series showed increases of 28% and 5% in load-carrying capacity. Load-carrying capacity [kN] 30 S-0.80-WW S-0.80-WD S-0.80-DD S-0.65-WW S-0.65-WD S-0.65-DD 25 20 15 10 5 0 Unreinforced Reinforced Reinforcement condition Figure 3-19: Relation between load-carrying capacity and reinforcement (Single Housings) The relation between load-carrying capacity and moisture conditions is demonstrated in Figure 3-20. Different moisture conditions did not substantially affect the load-carrying capacity in the S-0.80 series. However, in the S-0.65 series, the highest capacity was observed in the drydry specimens which conforms with the existing literature that reduction in moisture content below FSP leads to higher strength. At the same time, reduction in moisture induces internal stresses in wood. The influence of which was clearly observed in the WD specimens (having lowest capacity) which were cut wet, but tested dry. The WW specimens have higher capacity than the WD specimens, suggesting that the influence of shrinkage like shrinkage defects and the induced internal stresses are significantly more than the increase of strength due to reduction in MC. Otherwise, both the WD and the DD specimens would have had higher capacity than the WW specimens with the WD specimens perhaps having slightly lower capacity than the DD specimens. 39 S-0.80 S-0.65 Load-carrying capacity [kN] 25 20 15 10 5 0 WW WD DD Moisture condition Figure 3-20: Relation between load-carrying capacity and moisture condition (single housings) To investigate these findings statistically, ANOVA was conducted to know the significance of the results. The ANOVA results for the single housing tests are shown Table 3-9. It is noted that there was no significant three-way interaction among the factors nor any two-way interactions. Hence, the impact of all factors can be evaluated separately. It is observed that for an alpha value of 0.05, p values for the dependent variables (Housing Depth and Reinforcement) are much less than 0.05. Hence, the differences in load-carrying capacity due to variation in housing depth and presence or absence of reinforcement are statistically significant. However, the influence of environment (moisture condition) on the load-carrying capacity having a p-value of 0.86 is not statistically significant. Hence, previous observation regarding the WD specimens having the lowest capacity and the DD specimens having the highest capacity in the S-0.65 series, could not be confirmed. It can be concluded that the capacity is controlled by depth below mortise, reinforcement, bearing area (mortise depth and width). In addition, usually splitting failures have lower capacity and it is best to avoid such failure due to its abrupt nature. As an optimum mortise 40 depth, 50% of the beam depth should be selected, considering the possible failure in the tenon of the secondary beam (instead of failure of mortise in primary beam). Table 3-9: Three Factor ANOVA results: single housing Parameters SS df MS F p-value A (Housing Depth) 584 1 584.31 41.32 <0.01 B (Reinforcement) 344 1 344.23 24.34 <0.01 C (Environment) 4.14 2 2.07 0.15 0.86 AxB 4.13 1 4.13 0.29 0.59 AxC 20.9 2 10.49 0.74 0.48 BxC 39.3 2 19.66 1.39 0.26 AxBxC 39.6 2 19.82 1.40 0.25 Within 848 60 14.14 Total 1885 71 26.55 3-6-3 Double Housing Beams Investigating the individual failure modes, provided in appendix A-4 and appendix A-5, showed that for both unreinforced D-0.80 series (D-0.80-DD) and unreinforced D-0.65 series (D-0.65-DD) the proportion of splitting failure is 60%, as summarized in Table 3-10. In contrasts to what was observed in the single housing series where only S-0.80 series had the higher proportion of splitting, this suggests that the failure mechanism in unreinforced double housings is significantly influenced by tension perpendicular to grain directions irrespective of the depths of housings. However, similar to single housings, the D-0.80 series fail in splitting (60%) while only 27% of the specimens in the D-0.65 series fail in splitting. Hence, the general trend of data in the double housings is similar to that in single housings where series with lower height below mortise predominantly failed in splitting. 41 Table 3-10: Proportion of failure modes: non-reinforced double housing ID Splitting (%) Bearing (%) Bearing/Splitting (%) D-0.80 (All) 60 13 27 D-0.65 (All) 27 13 60 D-0.80 (Unreinforced) 60 40 D-0.65 (Unreinforced) 60 40 Figure 3-21 compares between the load-carrying capacities under different reinforcement conditions and mortise depth for dry-dry specimens. Calculating the mean F for all specimens from the D-0.80 and D-0.65 series (ignoring difference in reinforcements), showed that D-0.65 series had significantly (73%) higher capacity of 46.5 kN than the D-0.80 series with 26.9 kN. Hence, mortise depth has a significant influence on load capacity. Load-carrying capacity [kN] DD DD-RR DD-R 50 40 30 20 10 0 D-0.80 Mortise depth D-0.65 Figure 3-21: Relation between load-carrying capacity and mortise depth (double housings) This research only considered the dry-dry series for double housings and to evaluate the influence of moisture, results are compared with wet-wet series tested previously (Heal et al., 2019). The relation between the load-carrying capacity and the moisture conditions of D-0.80 series is illustrated in Figure 3-22. All the wet-wet series (D-0.80-WW, D-0.80-WW-RR and D-0.80-WW-R) had higher load carrying capacity than their corresponding dry-dry series (D0.80-DD, D-0.80-DD-RR and D-0.80-DD-R) except the D-0.65-WW series, which showed 42 slightly lower F than the D-0.65-DD series; however this series also had comparatively larger spread of data for the dry-dry series (CV of 19%) than the wet-wet series (CV of 10%). It was also observed that the dry-dry series of D-0.65 had higher percentage of splitting failures, while the wet-wet series of D-0.65 had higher percentage of mixed failure. The lower load-carrying capacity of dry-dry series suggest change in moisture conditions having strong influence, where reduction in moisture content induces early initiation of cracks due to tension perpendicular to the grain. The reduction in moisture content led to greater shrinkage and internal stresses in wood, resulting in lower load-carrying capacity for specimens likely to fail in splitting. The increase of tension perpendicular to grain strength due to reduction of moisture seems to be of lower influence in comparison to the reduction of strength due to development of moisture change induced internal cracks. UR Load-carrying capacity [kN] 50 RR R 40 30 20 10 0 Wet-Wet Moisture conditions Dry-Dry Figure 3-22: Relation between load-carrying capacity and moisture condition (double housings) Figure 3-23 illustrates that reinforcement increased the load-carrying capacity in beams irrespectively of moisture condition or relish depth. Here, only D-0.80 series are considered, as D-0.65 did not have reinforced wet-wet specimens from previous research to compare with. It is also observed that beams, which were reinforced and then tested, always had higher capacity than those beams, which were tested first and then again retested with reinforcement. 43 In other words, reinforcement of uncracked beams leads to higher capacity as the load is carried by both the reinforcement and the tensile strength of wood. However, in cracked beams the load is carried only by the screw in withdrawal leading to lower capacity. Thus, STS always increased the load-carrying capacity, but if the bonding of wood fibers had already been compromised during the initial test, the effectiveness was lower. Hence, having STS from the beginning delays crack propagation in the wood fibers, contributing in higher capacity. Wet-Wet Dry-Dry :Load-carrying capacity [kN] 50 40 30 20 10 0 UR RR Reinforcement conditions R Figure 3-23: Relation between load-carrying capacity and reinforcement (double housings) Two ANOVAs were performed to evaluate the impact of the parameters, housing depth, moisture, reinforcement on for alpha = 0.05. In Analysis 1, presented in Table 3-11, only the dry-dry specimens from series D-0.80 and D-0.65 were considered. ANOVA showed that the interaction between housing depth and reinforcement was not statistically significant; therefore, the factors can be evaluated independently. Housing depth was significant (p < 0.01), but reinforcement (p = 0.13) was shown not to be statistically significant in influencing the capacity. However, the reinforcement would still be statistically significant if a lower confidence level of 85% was considered. In Analysis 2, results presented in Table 3-12, all specimens from the D-0.80 series were considered. The two-factor ANOVA showed that the interaction between environment and 44 reinforcement was not statistically significant; therefore, the factors can be evaluated independently. Both the moisture conditions and reinforcement were statistically significant in influencing the load-carrying capacity with p-values <0.01. While lower moisture content in timber is expected to result in higher strength properties due to better bonding between cellulose, in this research conflicting results were obtained for the S-0.80 or D-0.80 series. Considering the predominant mode of failure (splitting), it seems reasonable to conclude that the internal stresses induced by shrinkage significantly affected the load-carrying capacity. The moisture-induced internal stresses from shrinkage led to lower capacities in D-0.80 specimens and almost unchanged capacities in the S-0.80 specimens. However, the failure mechanism in S-0.65 or D-0.65 were initiated by compression perpendicular to grain which increases slightly at lower moisture contents leading to an expected higher capacity in dry conditions. Table 3-11: Two factor ANOVA (Analysis 1) results: double housing Parameters SS df MS F P-value Housing Depth 2876.89 1 2876.89 39.16 <0.01 Reinforcement 331.62 2 165.81 2.26 0.13 Interaction 36.43 2 18.22 0.25 0.78 Within 1763.32 24 73.47 Total 5008.27 29 Table 3-12: Two factor ANOVA (Analysis 2) results: double housing Parameters SS df MS F P-value Environment 1002.25 1 1002.25 21.04 <0.01 Reinforcement 549.47 2 274.73 5.77 <0.01 Interaction 17.03 2 8.52 0.18 0.84 Within 1143.28 24 47.64 Total 2712.04 29 45 4 Conclusions and Outlook In this thesis, the influence of mortise depth, wood moisture and reinforcement on the loadcarrying capacity of solid timber beams with housings was investigated. A total of 60 beams with single housings and 30 beams with double housings were prepared while varying the moisture condition (either wet or dry) at the time of cutting the housings and at the time of testing. In addition, 198 small-scale specimens were tested for compression, tension and shear to evaluate the impact of local material strength on the beam load-carrying capacity. Along with 32 previous tests, statistical analyses were conducted to evaluate the statistical significance of the investigated parameters. Based on the findings in this research, the following conclusions can be drawn: · Housing depths have significant influence on the failure mode. With lower mortise depths, predominantly bearing failure or mixture of bearing and splitting were observed. The greater height below the mortise leads to compression due to bearing forces to govern instead of tension perpendicular to grain. The housing depth consequently also has affected the load-carrying capacity F . The beams failing in pure splitting (beams with higher mortise depth) exhibited significantly lower F . This finding was confirmed to be statistically significant for both single and double housings. · Reinforcement with self-tapping screws increased the load-carrying capacity irrespective of depth. Again, this finding was confirmed to be statistically significant for both single and double housings. However, the increase in capacity is more prominent in the series with higher mortise depth as the withdrawal capacity of STS reduces the crack propagation in the beam due to tensile forces perpendicular to the grain. Also, increase in the loadcarrying capacity is more when un-cracked section is reinforced instead of reinforcing a cracked section. 46 · For the single housing series, the moisture conditions did not prove to be statistically significant as factor influencing the load-carrying capacity . For double housings, however, lower moisture contents were shown to decrease . and the result was statistically significant. · Splitting occurs suddenly and hence not a desirably mode of failure in housing connections. Bearing failure or a combined bearing/splitting failure would generally provide visible warning concerning failure and hence the more preferable mode of failure. Based on the experimental results, to achieve such failure modes, a large depth below the mortise should be provided and fully-threaded STS should be used to reinforce the housing. Although design driven by how much load it should take is expected to lead to higher depth below the mortise and in other words, higher capacity-non splitting failure modes with ample warning. In case load requirements are not as high, still larger depths should be provided below the mortise to achieve bearing or mixed failure. Based on these results and observations, future investigations should address the following: · Only solid Douglas fir beams were tested. Future testing should validate the findings by testing other species and glue-laminated timber (GLT) beams. · Fracture mechanics could be used to evaluate the load-carrying capacity of housings as a function of the mortise parameters. In this context, fracture mechanics parameters need to be determined on samples taken from the timber beams. · Analytical and numerical methods can be used for parametric studies and developing empirical equations to predict the load-carrying capacity of housings. · The fire resistance of housing connections, both un-reinforced and reinforced should be determined to provide guidance for the fire resistance of exposed structures with housings. 47 References Angst, V., & Malo, K. A. (2012). Effect of self-tapping screws on moisture induced stresses in glulam. Engineering Structures, 299-306. ASTM Standard D143. (2021). Standard Test Methods for Small Clear Specimens of Timber. West Conshohocken, PA: ASTM International. Barrett, J. D., Foschi, R. O., & Fox, S. P. (1975). Perpendicular to Grain Strength of Douglas Fir. Canadian Journal of Civil Engineering, 50-57. Blass, H. J., & Bejtka, I. (2004). Reinforcements perpendicular to the grain using self-tapping screws. 8th World Conference on Timber Engnieering, (pp. 233-238). Lahti. Börjesson, P., & Gustavsson, L. (2000). Greenhouse gas balances in building construction: wood versus concrete from life-cycle and forest land-use perspectives. Energy Policy, 575 - 588. CEN. (1991). EN 26891-Timber structures-Joints made with mechanical fasteners general principles for the determination of strength and deformation characteristics. Brussels, Belgium: European Committee for Standardization. Chaplain, M., & Valentin, G. (2010). Effects of relative humidity conditions on crack propagation in timber : Experiments and modelling. World Conference on Timber Engineering, 1-8. Danzer, M., Dietsch, P., & Winter, S. (2016). Reinforcement of round holes in glulam beams arranged eccentrically or in groups. World Conference on Timber Engineering (pp. 2804-2812). Vienna: TU Verlag Wien. DIBt. (2019). ETA 11/0284: HECO-FIX-plus, HECO-TOPIX, HECO-TOPIX-T and HECOTOPIX-CC screws. Schramberg: European Organisation for Technical Assessment (EOTA). Dietsch, P. (2017). Effect of reinforcement on shrinkage stresses in timber members. Construction and Building Materials , 903–915. Dietsch, P., & Brandner, R. (2015). Self-tapping screws and threaded rods as reinforcement for structural timber elements – A state-of-the-art report. Construction and Building Materials, 78-89. 48 Franke, B., & Franke, S. (2014). Causes, assessment and impact of cracks in timber elements. COST Workshop – Highly Performing Timber Structures: Reliability, Assessment, Monitoring and Strengthening (pp. 15-21). Biel: Bern University of Applied Sciences, Switzerland and Institute for Timber Construction, Structures and Architecture. Garbin, E., Valluzzi, M. R., & Modena, C. (2006). Characterization of a dovetail joint for timber roofs. 9th World Conference on Timber Engineering (pp. 571-578). Portland, USA: Oregon State University Conference Services. Green, D. W. (2001). Wood: Strength and Stiffness. Encyclopedia of Materials: Science and Technology, 9732-9736. Häglund, M. (2007). Varying moisture content and eigen-stresses in timber elements. Wood Science and Technology, 477- 490. Harte, A. (2009). Introduction to Timber as an Engineering Material. London: ICE Publishing. Heal, T., Tannert, T., & Zirnhelt, R. (2019). Housings in Timber Construction: Testing and Design Guidance. Timber Framing 133, 6-13. IEA. (2019). Global Status Report for Buildings and Construction. Paris: IEA. Iuliano, G. (2013). Load Bearing Housing Test for Timber Frame Joining (unpublished). Virginia Military Institute. Jockwer, R. (2014). Structural behaviour of glued laminated timber beams with unreinforced and reinforced notches. Zürich: ETH-Zürich. Montgomery, D. C., & Runger, G. C. (2003). Applied statistics and probability for engineers (3rd ed.). New York: Wiley. Nehil, T. E., & Trojniak, B. I. (2013). Capacity of Load-Bearing Housings: Preliminary Results from Load Testing. Alstead: Timber Frame Engineering Council. Rammer , D. R., & Winistorfer , S. G. (2001). Effect of moisture content on dowel-bearing strength. Wood and fiber science, 126-139. Roszyk, E., Stachowska, E., Majka, J., Mania, P., & Broda, M. (2020). Moisture-Dependent Strength Properties of Thermally-Modified Fraxinus Excelsior Wood in Compression. Materials, 1647. 49 Sjödin , J., & Johansson , C. (2007). Influence of initial moisture induced stresses in multiple steel-to-timber dowel joints. Holz Roh Werkst, 71-77. Sobon, J. A. (2014). Historic American Timber Joinery: A Graphic Guide. Becket: Timber Framers Guild. Tannert, T. (2008). Structural Performance of Rounded Dovetail Connections. Vancouver, Canada: University of British Columbia. Tannert, T., & Lam, F. (2009). Progress in Structural Engineering and Materials : Timber Construction Self-tapping Screws as Reinforcement for Rounded Dovetail Connections. Structural Control and Health Monitoring, 16(3), 374-384. Tannert, T., Vallée , T., & Hehl , S. (2012). Experimental and numerical investigations on adhesively bonded hardwood joints. International Journal of Adhesion and Adhesives, 37, 65–69. United States Department of Agriculture, F. S. (2010). Wood Handbook : Wood as an Engineering Material (Centennial Edition ed.). Madison: Dept. of Agriculture, Forest Service, Forest Products Laboratory. Widehammar, S. (2004). Stress-Strain Relationships for Spruce Wood: Influence of Strain Rate, Moisture Content and Loading Direction. Experimental Mechanics, 44-48. Wood Design Standards Committee. (2018). National Design Specification (NDS) for Wood Construction. Leesburg, VA: American Wood Council (AWC). Zhang, C., Guo, H., Jung, K., Harris, R., & Chang, W.-S. (2019). Self-tapping screws as reinforcement on single-dowel connections with artificial cracks. Engineering Structures, 656 - 664. 50 Appendices Appendix 1 – Test results from previous research (Heal et al., 2019) ID Spec # S-0.80-WW 1/6 2/6 3/6 4/6 5/6 6/6 Average S-0.65-WW 1/6 2/6 3/6 4/6 5/6 6/6 Average D-0.80-WW Average D-0.80-WW-RR Average D-0.80-WW-R Average D-0.65-WW Average 1/5 2/5 3/5 4/5 5/5 1/5 2/5 3/5 4/5 5/5 1/5 2/5 3/5 4/5 5/5 1/5 2/5 3/5 4/5 5/5 Wet MC [%] Density [kg/m3] Fult [kN] Ultimate Disp [mm] 30.4 29.6 29.5 30.2 28.6 29.4 29.6 29.1 34.8 35.5 37 34.1 28.3 33.1 428.64 430.11 519.81 519.31 489.71 488.01 479.26 415.13 530.64 428.95 545.86 470.60 456.96 474.69 15.00 10.90 17.90 15.10 15.60 15.60 15.02 17.20 19.60 20.80 22.60 21.30 19.00 20.08 11.90 11.30 20.10 15.70 13.20 17.30 14.92 23.60 18.90 26.30 23.90 26.40 27.80 24.48 28.9 29.2 26.8 26.4 25.4 27.3 28.9 29.2 26.8 26.4 25.4 27.3 28.2 28.4 25.4 27.4 29.5 27.8 29.1 29.7 29.0 30.1 28.9 29.6 565.00 585.00 625.00 658.00 541.00 594.80 565.00 585.00 625.00 658.00 541.00 594.80 652.00 656.00 513.00 518.00 598.00 587.40 583.00 593.00 588.00 536.00 529.00 565.80 29.10 34.90 38.20 36.90 24.50 32.72 39.80 45.50 39.00 43.50 31.10 39.78 48.50 45.00 38.80 36.60 45.90 42.96 44.70 49.80 43.80 40.60 38.70 43.52 9.90 11.90 9.40 8.90 10.00 10.02 17.00 17.80 13.10 16.30 15.80 16.00 15.20 16.40 16.70 20.20 12.90 16.28 17.10 16.20 16.70 22.60 20.80 18.68 51 Failure Mode Splitting Splitting Splitting Splitting Splitting Splitting Bearing Splitting/ Bearing Bearing Bearing/ Splitting Splitting/ Bearing Bearing/ Splitting Splitting Splitting/Bearing Splitting Splitting Splitting/Bearing Bearing/Splitting Splitting Splitting Bearing/Splitting Bearing/Splitting Bearing Bearing/Splitting Bearing Splitting/Bearing Bearing/Splitting Splitting Splitting Bearing/Splitting Bearing/Splitting Appendix 2 – Detailed test results of all specimens in S-0.80 series Table A-2-1: Individual test results: single housing ( S-0.80 series) ID S-0.80-WWD-RR Average S-0.80-WD Average S-0.80-WD-RR Average S-0.80-DD Average S-0.80-DD-RR Average Spec # Wet MC [%] 1/6 2/6 3/6 4/6 5/6 6/6 1/6 2/6 3/6 4/6 5/6 6/6 1/6 2/6 3/6 4/6 5/6 6/6 1/6 2/6 3/6 4/6 5/6 6/6 1/6 2/6 3/6 4/6 5/6 6/6 Dry MC [%] 6.9 7.1 7.3 6.7 7.0 7.1 7.0 29.2 29.2 29.2 29.2 29.2 29.2 29.2 7.2 6.9 7.1 7.1 6.7 7.1 7.0 7.1 7.0 7.0 7.1 7.1 7.3 7.1 7.1 7.0 7.0 7.1 7.1 7.3 7.1 Density [kg/m3] Fult [kN] 429 430 520 519 490 488 479 412 514 421 518 433 434 455 412 514 421 518 433 434 455 431 436 528 431 451 433 451 431 436 528 431 451 433 451 19.4 19.5 16.2 15.5 23.4 10.0 17.3 15.1 17.7 15.4 15.9 13.9 14.2 15.4 21.4 25.2 19.1 22.6 20.4 15.1 20.6 13.8 16.6 13.9 14.5 15.9 14.5 14.9 21.6 23.5 17.2 19.6 14.8 17.1 19.0 52 Ultimate Disp. [mm] 10.0 13.5 11.5 9.5 12.8 6.9 10.7 6.9 8.1 7.8 5.8 7.8 6.9 7.2 14.1 21.9 12.4 19.5 20.6 11.1 16.6 5.4 6.8 5.5 6.9 9.1 8.6 7.0 9.2 9.4 7.0 12.7 7.7 13.9 10.0 Crack Opening [mm] 1.9 3.1 1.7 1.7 3.0 1.0 2.0 0.4 0.4 0.3 0.3 0.2 0.4 0.3 2.4 1.8 2.3 2.9 1.4 0.8 1.9 0.3 0.5 0.3 0.8 2.5 0.7 0.8 1.9 1.5 1.5 1.8 1.3 1.3 1.5 Failure Mode Bearing/Splitting Splitting Splitting Splitting/Bearing Splitting Splitting Splitting Splitting Splitting Splitting Splitting Splitting Bearing/Splitting Bearing/Splitting Bearing/Splitting Bearing/Splitting Bearing/Splitting Bearing/Splitting Splitting Splitting Splitting Splitting Bearing/Splitting Bearing/Splitting Splitting Splitting Bearing/Splitting Bearing/Splitting Splitting Bearing/Splitting S-0.80-WD-2 S-0.80-WD-RR-1 S-0.80-WD-RR-2 S-0.80-WD-3 S-0.80-WD-4 S-0.80-WD-RR-3 S-0.80-WD-RR-4 S-0.80-WD-5 S-0.80-WD-6 S-0.80-WD-RR-5 S-0.80-WD-RR-6 30 30 25 25 20 20 Load [kN] Load [kN] S-0.80-WD-1 15 10 15 10 5 5 0 0 0 5 10 15 20 Displacement [mm] 25 0 30 5 10 15 20 Displacement [mm] 25 30 (b) (a) S-0.80-DD-2 S-0.80-DD-4 S-0.80-DD-6 30 30 25 25 20 20 Load [kN] Load [kN] S-0.80-DD-1 S-0.80-DD-3 S-0.80-DD-5 15 10 5 S-0.80-DD-RR-1 S-0.80-DD-RR-2 S-0.80-DD-RR-3 S-0.80-DD-RR-4 S-0.80-DD-RR-5 S-0.80-DD-RR-6 15 10 5 0 0 5 10 15 20 Displacement [mm] 25 0 30 0 10 15 20 Displacement [mm] 25 30 (d) Load [kN] (c) 5 30 25 20 15 10 5 0 S-0.80-WWD-RR-1 S-0.80-WWD-RR-2 S-0.80-WWD-RR-3 S-0.80-WWD-RR-4 S-0.80-WWD-RR-5 S-0.80-WWD-RR-6 0 5 10 15 20 Displacement [mm] 25 30 (e) Figure A-2-1: Load-displacement of S-0.80 series under bearing load (a) Unreinforced Wet-Dry Specimens, (b) Wet-Dry Specimens after Reinforcement, (c) Unreinforced Dry-Dry Specimens, (d) DryDry Specimens after Reinforcement, (e) Reinforced Wet-Wet Specimens tested Dry 53 S-0.80-WD-2 S-0.80-WD-3 S-0.80-WD-4 S-0.80-WD-5 S-0.80-WD-6 S-0.80-WD-RR-1 S-0.80-WD-RR-2 S-0.80-WD-RR-3 S-0.80-WD-RR-4 30 30 25 25 20 20 Load [kN] Load [kN] S-0.80-WD-1 15 10 5 15 10 5 0 0 -1 0 1 2 3 4 5 6 -1 0 Crack Opening [mm] 5 6 5 6 (b) S-0.80-DD-1 S-0.80-DD-2 S-0.80-DD-3 S-0.80-DD-4 S-0.80-DD-5 S-0.80-DD-6 30 30 25 25 20 20 Load [kN] Load [kN] (a) 1 2 3 4 Crack Opening [mm] 15 10 5 S-0.80-DD-RR-1 S-0.80-DD-RR-2 S-0.80-DD-RR-3 S-0.80-DD-RR-4 S-0.80-DD-RR-5 15 10 5 0 0 -1 0 1 2 3 4 Crack Opening [mm] 5 6 -1 (c) 0 1 2 3 4 Crack Opening [mm] (d) S-0.80-WWD-RR-1 S-0.80-WWD-RR-2 S-0.80-WWD-RR-3 S-0.80-WWD-RR-4 S-0.80-WWD-RR-5 S-0.80-WWD-RR-6 30 Load [kN] 25 20 15 10 5 0 -1 0 1 2 3 4 Crack Opening [mm] 5 6 (e) Figure A-2-2: Load-Mean Crack Opening S-0.80 series: (a) Unreinforced Wet-Dry Specimens, (b) Wet-Dry Specimens after Reinforcement, (c) Unreinforced Dry-Dry Specimens, (d) Dry-Dry Specimens after Reinforcement, (e) Reinforced Wet-Wet Specimens tested Dry 54 S-0.80-WWD-RR-6: Splitting S-0.80-WWD-RR-5: Splitting S-0.80-WWD-RR-4: Bearing/Splitting S-0.80-WWD-RR-3: Bearing S-0.80-WWD-RR-2: Splitting S-0.80-WWD-RR-1: Bearing/Splitting Figure A-2-3: Test Photos of S-0.80 series reinforced wet-wet specimens tested dry 55 S-0.80-WD-1: Splitting S-0.80-WD-2: Splitting S-0.80-WD-3: Splitting S-0.80-WD-4: Splitting S-0.80-WD-5: Splitting S-0.80-WD-6: Splitting Figure A-2-4: Test Photos of S-0.80 series unreinforced wet-dry specimens 56 S-0.80-WD-RR-1: Bearing/Splitting S-0.80-WD-RR-2: Bearing/Splitting S-0.80-WD-RR-3: Splitting S-0.80-WD-RR-4: Bearing/Splitting S-0.80-WD-RR-5: Bearing/Splitting S-0.80-WD-RR-6: Bearing Figure A-2-5: Test Photos of S-0.80 series reinforced retested wet-dry specimens 57 S-0.80-DD-1: Splitting S-0.80-DD-2: Splitting S-0.80-DD-3: Splitting S-0.80-DD-4: Splitting S-0.80-DD-5: Splitting S-0.80-DD-6: Bearing/Splitting Figure A-2-6: Test Photos of S-0.80 series unreinforced dry-dry specimens 58 S-0.80-DD-RR-6: Bearing/Splitting S-0.80-DD-RR-5: Splitting/Bearing S-0.80-DD-RR-4: Splitting/Bearing S-0.80-DD-RR-3: Splitting S-0.80-DD-RR-2: Splitting S-0.80-DD-RR-1: Splitting/Bearing Figure A-2-7: Test Photos of S-0.80 series reinforced retested dry-dry specimens 59 Appendix 3 – Detailed test results of all specimens in S-0.65 series Table A-3-1: Individual test results: single housing (S-0.65 series) ID S-0.65-WWD-RR Spec # Wet MC [%] 1/6 2/6 3/6 4/6 5/6 6/6 Average S-0.65-WD 1/6 2/6 3/6 4/6 5/6 6/6 Average S-0.65-WD-RR 1/6 2/6 3/6 4/6 5/6 6/6 Average S-0.65-DD 1/6 2/6 3/6 4/6 5/6 6/6 Average S-0.65-DD-RR Average 1/6 2/6 3/6 4/6 5/6 6/6 29.2 29.2 29.2 29.2 29.2 29.2 29.2 29.2 29.2 29.2 29.2 29.2 29.2 29.2 Dry MC [%] Density [kg/m3] Fult [kN] Ultimate Disp [mm] 7.1 7.1 7.0 7.3 7.1 7.1 7.1 6.9 6.8 7.0 6.9 7.1 7.2 7.0 6.9 6.8 7.0 6.9 7.1 7.2 7.0 7.1 6.2 7.1 7.4 7.2 7.1 7.0 7.1 6.2 7.1 7.4 7.2 7.1 7.0 415 531 429 546 471 457 475 443 446 451 448 551 535 479 443 446 451 448 551 535 479 531 446 547 458 453 508 490 531 446 547 458 453 508 490 22.6 29.1 28.7 29.7 28.9 19.8 26.5 18.4 16.8 18.3 21.4 15.7 22.7 18.9 29.0 19.4 24.0 23.3 28.3 32.1 26.0 26.1 18.9 27.7 21.0 15.5 22.4 21.9 31.6 18.1 30.8 22.7 14.3 20.4 23.0 21.2 20.8 27.0 21.6 30.1 20.6 23.6 8.5 21.0 8.2 16.6 7.2 10.0 11.9 14.5 29.1 20.2 18.8 11.0 19.5 18.8 10.2 10.5 8.4 23.6 10.4 11.6 12.4 12.5 17.1 10.7 22.5 11.9 14.1 14.8 60 Mean Crack Opening [mm] 1.6 2.4 2.7 2.7 1.7 3.3 2.4 0.8 0.5 0.4 0.6 0.4 0.2 0.5 1.2 1.2 1.0 2.1 0.9 0.9 1.2 1.7 0.7 0.6 1.7 0.4 0.3 0.9 1.3 0.5 1.9 1.8 0.4 0.2 1.0 Failure Mode Bearing Splitting Bearing Splitting Bearing/Splitting Bearing Splitting Bearing Splitting Bearing Splitting Splitting Splitting/Bearing Bearing Bearing/Splitting Bearing Splitting Bearing Splitting Bearing Splitting Bearing/Splitting Bearing Bearing Bearing/Splitting Bearing/Splitting Splitting Bearing Bearing Bearing S-0.65-WD-2 S-0.65-WD-4 S-0.65-WD-6 S-0.65-WD-RR-1 S-0.65-WD-RR-3 S-0.65-WD-RR-5 35 30 25 20 15 10 5 0 Load [kN] Load [kN] S-0.65-WD-1 S-0.65-WD-3 S-0.65-WD-5 0 5 10 15 20 25 Displacement [mm] 30 5 10 15 20 25 Displacement [mm] 30 S-0.65-DD-1 S-0.65-DD-2 S-0.65-DD-RR-1 S-0.65-DD-RR-2 S-0.65-DD-3 S-0.65-DD-4 S-0.65-DD-RR-3 S-0.65-DD-RR-4 S-0.65-DD-5 S-0.65-DD-6 S-0.65-DD-RR-5 S-0.65-DD-RR-6 35 30 25 20 15 10 5 0 Load [kN] Load [kN] 0 35 (b) 0 5 10 15 20 25 Displacement [mm] 30 35 30 25 20 15 10 5 0 35 (c) Load [kN] 35 30 25 20 15 10 5 0 35 (a) S-0.65-WD-RR-2 S-0.65-WD-RR-4 S-0.65-Wd-RR-6 0 5 10 15 20 25 Displacement [mm] 30 35 (d) 35 30 25 20 15 10 5 0 S-0.65-WWD-RR-1 S-0.65-WWD-RR-2 S-0.65-WWD-RR-3 S-0.65-WWD-RR-4 S-0.65-WWD-RR-5 S-0.65-WWD-RR-6 0 5 10 15 20 25 Displacement [mm] 30 35 (e) Figure A-3-1: Load-displacement tests of 0.65 series under bearing load (a) Unreinforced WetDry Specimens, (b) Wet-Dry Specimens after Reinforcement, (c) Unreinforced Dry-Dry Specimens, (d) Dry-Dry Specimens after Reinforcement, (e) Reinforced Wet-Wet Specimens tested Dry 61 S-0.65-WD-2 S-0.65-WD-4 S-0.65-WD-6 S-0.65-WD-RR-1 S-0.65-WD-RR-3 S-0.65-WD-RR-5 35 30 25 20 15 10 5 0 Load [kN] Load [kN] S-0.65-WD-1 S-0.65-WD-3 S-0.65-WD-5 -1 0 1 2 3 4 Crack Opening [mm] 5 S-0.65-DD-2 S-0.65-DD-4 S-0.65-DD-6 35 30 25 20 15 10 5 0 Load [kN] Load [kN] -1 0 1 2 3 4 Crack Opening [mm] 5 6 (b) S-0.65-DD-1 S-0.65-DD-3 S-0.65-DD-5 -1 0 1 2 3 4 Crack Opening [mm] 5 S-0.65-DD-RR-1 S-0.65-DD-RR-2 S-0.65-DD-RR-3 S-0.65-DD-RR-4 S-0.65-DD-RR-5 S-0.65-DD-RR-6 35 30 25 20 15 10 5 0 -1 6 0 1 2 3 4 Crack Opening [mm] 5 6 (d) (c) S-0.65-WWD-RR-1 S-0.65-WWD-RR-3 S-0.65-WWD-RR-5 Load [kN] 35 30 25 20 15 10 5 0 6 (a) S-0.65-WD-RR-2 S-0.65-WD-RR-4 S-0.65-WD-RR-6 S-0.65-WWD-RR-2 S-0.65-WWD-RR-4 S-0.65-WWD-RR-6 35 30 25 20 15 10 5 0 -1 0 1 2 3 4 Crack Opening [mm] 5 6 (e) Figure A-3-2: Load-Mean Crack Opening tests of 0.65 series under bearing load (a) Unreinforced Wet-Dry Specimens, (b) Wet-Dry Specimens after Reinforcement, (c) Unreinforced Dry-Dry Specimens, (d) Dry-Dry Specimens after Reinforcement, (e) Reinforced Wet-Wet Specimens tested Dry 62 S-0.65-WWD-RR-1: Bearing S-0.65-WWD-RR-2: Splitting S-0.65-WWD-RR-3: Bearing S-0.65-WWD-RR-4: Splitting S-0.65-WWD-RR-5: Bearing/Splitting S-0.65-WWD-RR-6: Bearing Figure A-3-3: Test Photos of S-0.65 series reinforced wet-wet specimens tested dry 63 S-0.65-WD-1: Splitting S-0.65-WD-2: Bearing S-0.65-WD-3: Splitting S-0.65-WD-4: Bearing S-0.65-WD-5: Splitting S-0.65-WD-6: Splitting Figure A-3-4: Test Photos of S-0.65 series unreinforced wet-dry specimens 64 S-0.65-WD-RR-1: Splitting/Bearing S-0.65-WD-RR-2: Bearing S-0.65-WD-RR-3: Bearing/Splitting S-0.65-WD-RR-4: Bearing S-0.65-WD-RR-5: Splitting S-0.65-WD-RR-6: Bearing Figure A-3-5: Test Photos of S-0.65 series reinforced retested wet-dry specimens 65 S-0.65-DD-1: Splitting S-0.65-DD-2: Bearing S-0.65-DD-3: Splitting S-0.65-DD-4: Bearing/Splitting S-0.65-DD-5: Bearing S-0.65-DD-6: Bearing Figure A-3-6: Test Photos of S-0.65 series unreinforced dry-dry specimens 66 S-0.80-DD-RR-1: Splitting/Bearing S-0.80-DD-RR-2: Splitting S-0.80-DD-RR-3: Splitting S-0.80-DD-RR-4: Splitting/Bearing S-0.80-DD-RR-5: Splitting/Bearing S-0.80-DD-RR-6: Bearing/Splitting Figure A-3-7: Test Photos of S-0.65 series reinforced retested dry-dry specimens 67 Appendix 4 – Detailed test results of all specimens in D-0.80 series Table A-4-1: Individual test results: double housings (D-0.80 series) ID D-0.80-DD Average D-0.80-DD-RR Average D-0.80-DD-R Average Spec # 1/5 2/5 3/5 4/5 5/5 1/5 2/5 3/5 4/5 5/5 1/5 2/5 3/5 4/5 5/5 Wet MC [%] Dry MC [%] Density [kg/m3] Fult [kN] Ultimate Disp. [mm] 6.9 7.2 7.3 7.5 7.5 7.3 6.9 7.2 7.3 7.5 7.5 7.3 6.3 6.1 6.3 6.2 6.2 6.2 550 492 469 513 528 510 550 492 469 513 528 510 467 527 483 496 541 503 21.0 25.5 11.5 39.6 12.3 22.0 27.1 28.9 15.2 35.4 23.9 26.1 37.3 32.2 30.6 29.5 34.0 32.7 7.4 6.5 8.5 10.0 10.6 8.6 9.5 9.2 12.6 14.1 22.9 13.7 11.3 10.5 12.1 8.8 10.8 10.7 68 Mean Crack Opening [mm] 4.4 0.6 3.6 1.2 8.6 3.7 7.7 1.7 7.5 4.8 19.3 8.2 1.3 0.9 1.4 0.6 1.0 1.0 Failure Mode Splitting Splitting Bearing Splitting Bearing Splitting Splitting Bearing/Splitting Splitting Splitting/Bearing Splitting Bearing/Splitting Bearing/Splitting Splitting Splitting D-0.80-DD-1 D-0.80-DD-2 D-0.80-DD-RR-1 D-0.80-DD-RR-2 D-0.80-DD-3 D-0.80-DD-4 D-0.80-DD-RR-3 D-0.80-DD-RR-4 D-0.80-DD-5 D-0.80-DD-RR-5 Load [kN] Load [kN] 40 35 30 25 20 15 10 5 0 0 5 10 15 20 Displacement [mm] 40 35 30 25 20 15 10 5 0 25 (a) 0 5 10 15 Displacement [mm] 20 25 (b) D-0.80-DD-R-1 D-0.80-DD-R-2 D-0.80-DD-R-3 D-0.80-DD-R-4 Load [kN] D-0.80-DD-R-5 40 35 30 25 20 15 10 5 0 0 5 10 15 20 Displacement [mm] 25 (c) Figure A-4-1: Load-displacement tests of D-0.80 series under bearing load (a) Unreinforced Dry-Dry Specimens, (b) Dry-Dry Specimens retested after Reinforcement, (e) Reinforced DryDry Specimens 69 D-0.80-DD-1: Splitting D-0.80-DD-2: Splitting D-0.80-DD-3: Bearing D-0.80-DD-4: Splitting D-0.80-DD-5: Bearing Figure A-4-2: Test Photos of D-0.80 series (Side 1) unreinforced dry-dry specimens 70 D-0.80-DD-1: Splitting D-0.80-DD-2: Splitting D-0.80-DD-3: Bearing D-0.80-DD-4: Splitting D-0.80-DD-5: Bearing Figure A-4-3: Test Photos of D-0.80 series (Side 2) unreinforced dry-dry specimens 71 D-0.80-DD-RR-1: Splitting D-0.80-DD-RR-2: Splitting D-0.80-DD-RR-3: Bearing/Splitting D-0.80-DD-RR-4: Splitting D-0.80-DD-RR-5: Splitting/Bearing Figure A-4-4: Test Photos of D-0.80 series (Side 1) reinforced retested dry-dry specimens 72 D-0.80-DD-RR-1: Splitting D-0.80-DD-RR-2: Splitting D-0.80-DD-RR-3: Bearing/Splitting D-0.80-DD-RR-4: Splitting D-0.80-DD-RR-5: Splitting/Bearing Figure A-4-5: Test Photos of D-0.80 series (Side 2) reinforced retested dry-dry specimens 73 D-0.80-DD-R-1: Splitting D-0.80-DD-R-2: Bearing/Splitting D-0.80-DD-R-3: Bearing/Splitting D-0.80-DD-R-4: Splitting D-0.80-DD-R-5: Splitting Figure A-4-6: Test Photos of D-0.80 series (Side 1) reinforced dry-dry specimens 74 D-0.80-DD-R-1: Splitting D-0.80-DD-R-2: Bearing/Splitting D-0.80-DD-R-3: Bearing/Splitting D-0.80-DD-R-4: Splitting D-0.80-DD-R-5: Splitting Figure A-4-7: Test Photos of D-0.80 (Side 2) series reinforced dry-dry specimens 75 Appendix 5 – Detailed test results of all specimens in D-0.65 series Table A-5-1: Individual test results: double housings (D-0.65 series) ID D-0.65-DD Average D-0.65-DD-RR Average D-0.65-DD-R Average Spec # 1/5 2/5 3/5 4/5 5/5 1/5 2/5 3/5 4/5 5/5 1/5 2/5 3/5 4/5 5/5 Wet MC [%] Dry MC [%] Density [kg/m3] Fult [kN] Ultimate Disp. [mm] 6.7 6.2 6.2 6.5 6.2 6.4 6.7 6.2 6.2 6.5 6.2 6.4 6.3 6.4 6.4 7.2 7.3 6.7 475 539 536 505 473 506 475 539 536 505 473 506 499 492 459 573 551 515 38.2 53.6 50.1 46.1 33.5 44.3 39.5 47.6 49.1 60.9 31.2 45.6 54.7 44.3 42.8 59.4 46.9 49.6 9.5 10.3 14.1 19.3 24.5 15.5 12.2 19.2 23.4 27.6 17.5 20.0 19.4 10.1 26.6 12.7 18.5 17.5 76 Mean Crack Opening [mm] 0.4 3.1 0.4 0.9 1.3 1.2 1.8 1.4 2.5 2.3 1.3 1.9 0.3 0.3 0.8 0.4 2.5 0.9 Failure Mode Splitting Splitting Splitting Bearing/Splitting Bearing Splitting Bearing/Splitting Splitting/Bearing Bearing/Splitting Bearing Bearing/Splitting Bearing/Splitting Bearing/Splitting Bearing/Splitting Bearing/Splitting D-0.65-DD-2 D-0.65-DD-4 70 60 50 40 30 20 10 0 0 5 10 15 20 Displacement [mm] 25 D-0.65-DD-RR-2 D-0.65-DD-RR-3 D-0.65-DD-RR-4 70 60 50 40 30 20 10 0 30 (a) 0 5 10 15 20 Displacement [mm] 25 30 (b) D-0.65-DD-R-1 D-0.65-DD-R-3 D-0.65-DD-R-5 Load [kN] D-0.65-DD-RR-1 D-0.65-DD-RR-5 Load [kN] Load [[kN] D-0.65-DD-1 D-0.65-DD-3 D-0.65-DD-5 D-0.65-DD-R-2 D-0.65-DD-R-4 70 60 50 40 30 20 10 0 0 5 10 15 20 Displacement [mm] 25 30 (c) Figure A-5-1: Load-displacement tests of D-0.65 series under bearing load (a) Unreinforced Dry-Dry Specimens, (b) Dry-Dry Specimens retested after Reinforcement, (e) Reinforced DryDry Specimens 77 D-0.65-DD-1: Splitting D-0.65-DD-2: Splitting D-0.65-DD-3: Splitting D-0.65-DD-4: Bearing/Splitting D-0.65-DD-5: Bearing Figure A-5-2: Test Photos of D-0.65 series (Side 1) unreinforced dry-dry specimens 78 D-0.65-DD-1: Splitting D-0.65-DD-2: Splitting D-0.65-DD-3: Splitting D-0.65-DD-4: Bearing/Splitting D-0.65-DD-5: Bearing Figure A-5-3: Test Photos of D-0.65 series (Side 2) unreinforced dry-dry specimens 79 D-0.80-DD-RR-1: Splitting D-0.80-DD-RR-1: Bearing/Splitting D-0.80-DD-RR-1: Splitting/Bearing D-0.80-DD-RR-1: Bearing/Splitting D-0.80-DD-RR-1: Bearing Figure A-5-4: Test Photos of D-0.65 series (Side 1) reinforced retested dry-dry specimens 80 D-0.80-DD-RR-1: Splitting D-0.80-DD-RR-1: Bearing/Splitting D-0.80-DD-RR-1: Splitting/Bearing D-0.80-DD-RR-1: Bearing/Splitting D-0.80-DD-RR-1: Bearing Figure A-5-5: Test Photos of D-0.65 series (Side 2) reinforced retested dry-dry specimens 81 D-0.80-DD-R-1: Splitting D-0.80-DD-R-2: Splitting D-0.80-DD-R-3: Bearing D-0.80-DD-R-4: Splitting D-0.80-DD-R-5: Bearing Figure A-5-6: Test Photos of D-0.65 series (Side 1) reinforced dry-dry specimens 82 D-0.80-DD-R-1: Splitting D-0.80-DD-R-2: Splitting D-0.80-DD-R-3: Bearing D-0.80-DD-R-4: Splitting D-0.80-DD-R-5: Bearing Figure A-5-7: Test Photos of D-0.65 series (Side 2) reinforced dry-dry specimens 83