Short and long-term mechanical performance of adhesively bonded timber-concrete structural elements

Hybridizing timber and concrete in timber-concrete composites (TCC) can improve load capacity, bending stiffness, sound insulation, vibration response, and fire performance compared to timber alone with the same geometry. These benefits depend on a strong timber-concrete connection at their interface, as weak connections reduce composite efficiency. Traditional methods like mechanical fasteners (mainly relying on metallic connectors) have limitations, including partial rigidity and wood damage from cutting or drilling. Adhesive bonding has emerged as a cost-effective alternative, offering better interface rigidity. It can be applied using wet techniques (bonding fresh concrete to timber while the adhesive is still wet) or dry techniques (bonding prefabricated concrete to timber). However, the shear and bending behaviors of adhesively bonded timber-concrete composites, especially their long-term behaviour, using these methods remain under-researched. Chapter 1 introduced the TCC concept, emphasizing its applications in enhancing load-bearing capacity, bending stiffness, thermal performance, and sound insulation. The chapter highlighted the research gap in using adhesives as connectors and the absence of established guidelines for their application, particularly for wet and dry manufacturing techniques. It also outlined the research objectives and dissertation structure. Chapter 2 provided a comprehensive literature review on TCC systems, categorizing studies into short- and long-term investigations involving mechanical fasteners, notches, metal plates, and adhesive bonding. Chapter 3 detailed the materials, geometry, and manufacturing processes for timber-concrete joints and panels, along with the experimental setups for shear and bending tests. Chapter 4 investigated a short-term study examining the shear bond strength of wet and dry spruce wood-concrete joints bonded with ductile polyurethane (PUR) and brittle epoxy adhesives. The ductile PUR adhesive exhibits a relatively low tensile strength (6.1 MPa) and high strain at break (19%), in contrast to the brittle epoxy adhesive, which has a high tensile strength (38 MPa) but a low strain at break (1.7%). Key experimental parameters included adhesive amount for both wet and dry joints and waiting time for wet joints, defined as the interval between adhesive polymerization and fresh concrete casting. The study also conducted a comprehensive failure mode analysis guided by a well-defined criterion. In Chapter 5, a short-term study investigated the bending behavior of glulam-concrete panels bonded with ductile PUR and brittle epoxy adhesives. Key parameters included ultimate bending load, effective bending stiffness, mid-span deflection, mid-span strain distribution, and maximum principal strain between the support and 300 mm toward the loading point. Two panel designs were studied: Design (a) with a depth ratio of 50/85 mm/mm and a 2-m span, and Design (b) with a depth ratio of 50/80 mm/mm and a 4-m span. For Design (a), the study examined adhesive type (epoxy vs. PUR) and mesh steel reinforcement in wet-bonded panels. For Design (b), the focus was on manufacturing technique (wet vs. dry) and adhesive type (epoxy vs. PUR). Adhesive shear stress along the bond line was monitored using fiber optic sensors. A numerical simulation explored the effects of adhesive thickness, concrete-wood depth ratio, concrete strength class, wood type, and span length on load-mid-span deflection curve and shear and peel stress distributions within the bond lines. Chapter 6 presents a long-term study on wet and dry spruce wood-concrete joints bonded with brittle epoxy and ductile PUR. These joints were conditioned for nearly a year under three scenarios: outdoor, outdoor with mechanical shear load, and indoor with mechanical shear load. Destructive shear tests were conducted at 2, 4, 6, and 12 months for outdoor-only joints to evaluate changes in shear stiffness and strength over time. Joints exposed to mechanical shear loads were monitored continuously for shear deformation throughout 12 months and subsequently tested destructively to assess both shear stiffness and strength. Chapter 7 evaluated glulam-concrete panels with Design (b) under combined outdoor exposure and sustained bending loads for 86 weeks. Key parameters included adhesive type (brittle epoxy vs. ductile PUR), manufacturing method (wet vs. dry), and steel reinforcement in wet-bonded panels. The panels' creep mid-span deflection was monitored throughout, and after 86 weeks, they underwent destructive bending tests. Results were compared with reference panels discussed in Chapter 5. A predictive model incorporating viscoelastic behavior, concrete shrinkage, and environmental strain effects was employed to estimate the mid-span creep deflection of the panels over a 55-year service life. During long-term exposure and post-destructive bending tests, shear stress distribution within the bond line of the panels was monitored using fiber optic sensors to evaluate the adhesive bond integrity. In the short-term study of spruce wood-concrete joints (Chapter 4), increasing adhesive amount improved the shear bond strength of wet joints. Longer waiting times enhanced the bond strength of wet joints bonded with 1- and 3-mm PUR equivalent thicknesses but decreased it for epoxy-bonded joints. This behavior was linked to differences in shear storage modulus between PUR (G’=1.2 kPa) and epoxy (G’=229.7 kPa) adhesives at 0-min waiting time. Additionally, wet bonding generally resulted in higher shear strength compared to dry bonding, considering the same adhesive amount and type. In the short-term study of glulam-concrete panels (Chapter 5), the experimental results indicated that factors such as adhesive type (ductile PUR and brittle epoxy), the presence of steel rebars, and wet and dry manufacturing techniques did not significantly alter the bending behavior of the panels, particularly in terms of ultimate bending load and effective bending stiffness. Both ductile polyurethane (PUR) and brittle epoxy adhesives demonstrated sufficient rigidity, with shear stress consistently remaining below 1 MPa in all cases—well under 20% of the bond strength of the adhesives. The bond strength ranged from 4.5 to 6.3 MPa for wet joints and 3.1 to 3.9 MPa for dry joints. Numerical analysis consistently indicated that shear stress prevailed over peel stress in all parametric scenarios (max. τa / max. σa = 7-23). However, with increasing panel span length, peel stress approached shear stress (max. τa / max. σa ~ 3.5). The ultimate load primarily induced concrete compression damage and wood fiber damage. Lower concrete strength (C12/15) and hardwood (Beech and Azobe) favored compression damage, whereas higher concrete-wood depth ratios shifted the failure mode toward concrete tension damage. Adhesive thickness had minimal impact on stress distribution and panel failure modes. In the long-term study on spruce wood-concrete joints (Chapter 6), results indicated a gradual reduction in the shear stiffness and strength of dry joints exposed to outdoor conditions over the 12-month period, primarily due to bond failure at the concrete-adhesive interface. Conversely, wet joints, benefiting from enhanced mechanical interlocking at the adhesive-concrete interface, exhibited no degradation in shear stiffness and strength across those long-term conditions over the same period. The bond failure observed in dry joints was predominantly associated with stresses arising from dimensional changes in the wood. This study did not observe any degradation in the cross-linking density of the adhesive or in the concrete's stiffness and strength. Over the 86-week period, the mid-span creep deflection of 12 glulam-concrete panels remained consistent, regardless of manufacturing method (wet or dry), adhesive type (brittle epoxy or ductile PUR), or the presence of reinforcement in wet-bonded panels. Destructive tests revealed that most reference and long-term panels (7 out of 12) achieved similar ultimate bending loads and stiffness, with maximum deviations of 13.0% and 13.3% from the predicted values. However, 5 out of 12 panels, from both reference and long-term groups, failed prematurely due to defects such as knots, finger joints, or cracks developed during exposure. Adhesive bond integrity was preserved across all panels except the PUR-bonded dry glulam-concrete panel. This research's key contributions include quantifying shear bond strength in spruce wood-concrete joints through short-term studies, focusing on manufacturing techniques (wet vs. dry), adhesive types (ductile PUR and brittle epoxy), adhesive amounts, and waiting times. Long-term studies on the joints evaluated the effects of outdoor conditions and the combined impact of sustained shear loads with both outdoor and indoor conditions on residual bond strength and stiffness, considering manufacturing technique and adhesive type as part of a parametric study. The effects of wet and dry techniques and adhesive types (ductile PUR and brittle epoxy) on short-term bending behavior (e.g., ultimate bending load and stiffness) and mid-span creep deflection under combined outdoor conditions and sustained bending loads (long-term behavior) were investigated. The adhesive bond integrity of the long-term panels was assessed by analyzing the adhesive shear stress distribution monitored using fiber optic sensors. Numerical simulations were conducted to evaluate the influence of adhesive thickness, concrete-to-wood depth ratios, span length, wood species, and concrete strength class on load–mid-span deflection curves and adhesive shear and peel stress distribution within the bond lines of wood-concrete panels. Overall, wet bonding techniques demonstrated superior durability compared to dry bonding techniques under outdoor exposure and sustained loads. Statistical analyses ensured the reliability of the findings, offering practical