|Abstract or Summary
- The advent of synthetic adhesives has transformed the structural applications of wood. However, a persistent issue in adhesively-bonded wood products has been moisture durability. When designing wood based composites, moisture durability will depend on both the wood phase and the adhesive phase. A key question, therefore, is how does one rank adhesives for their ability to convey moisture durability in wood composites? Typical wood-based composite tests for moisture durability assessments do not consider the post-peak load regime of the material or are merely qualitative. This research explores the suitability of crack propagation fracture experiments to develop new methodologies for durability assessment of wood composites and adhesives.
Fracture resistance curves (R curves) were measured for solid wood Douglas-fir and laminated veneer lumber (LVL) made with Douglas-fir veneer and a variety wood adhesives, namely, polyvinyl acetate, phenol formaldehyde, emulsion polymer isocyanate and phenol resorcinol formaldehyde. The LVL and solid wood R curves were similar for initiation of fracture, but the LVL toughness rose much higher than solid wood. Because a rising R curve is caused by fiber bridging effects, these differences show that the LVL adhesive has a large effect on the fiber bridging process. This resin effect was exploited to develop a test method for characterizing the ability of a resin to provide wood composites that are durable to moisture exposure. In characterizing toughness changes, it was important to focus on the magnitude and rate of the toughness increase attributed to fiber bridging. A new method was developed for ranking the role of adhesives in the durability of wood-based composites by observing changes in fracture toughness during crack propagation following cyclic exposure to moisture conditions. This new approach was compared to conventional mechanical performance test methods, such as observing strength and stiffness loss after exposure. Comparing changes in fracture toughness as a function of crack length after moisture cycling shows that fracture-mechanics based methods can distinguish different adhesive systems on the basis of their durability, while conventional test methods do not have similar capability. Using steady-state toughness alone, the most and least durable adhesives (phenol formaldehyde and polyvinyl acetate) could be distinguished, but the performance of two other adhesives (emulsion polymer isocyanate and phenol resorcinol formaldehyde) could not. Further analysis of experimental R curves based on kinetics of degradation was able to rank all adhesives confidently and therefore provided the preferred method. The likely cause for the inability of conventional tests to rank adhesives is that they are based on initiation of failure while the fracture tests show that comparisons that can rank adhesives require consideration of fracture properties after a significant amount of crack propagation has occurred.
Additionally, a new method was proposed for determining fiber-bridging cohesive laws in fiber-reinforced composites and in natural fibrous materials. In brief, the method requires direct measurement of the R curve, followed by a new approach to extracting a cohesive law. This new approach was applied to finding fiber bridging tractions in LVL made with four different adhesives. Moreover, observations of changes in the bridging cohesive laws were used to rank the adhesives for their durability. Finally, both analytical and numerical models for fiber bridging materials were developed. The numerical modeling developed using material point method (MPM) simulated crack propagation that included crack tip propagation, fiber bridging zone development, and steady state crack growth. The simulated R curves agreed with experimental results.