CHAPTER 3 – FRACTURE TEST METHOD
In many wood composites, the bondline can be stressed beyond its capacity to resist crack formation. These cracks greatly affect the mechanical properties and durability of the composite material. There are many factors during the manufacturing of a wood-adhesive interface that can affect the integrity of the bondline. In order to characterize the robustness of this interaction between wood and resin, scientists started looking at the fracture energy, or the energy it takes to pull apart the two wood faces of the bondline. Fracture toughness is a material property and describes the ability of a material to resist crack propagation.
There are three modes of fracture (Figure 3.1). Mode I opening does not create a net moment because the forces are co-linear. A net moment can only be created by two forces that have some distance of separation, as in the case of Mode II or Mode III. Mode II is longitudinal shear, and mode III is transverse shear. Mixed mode fracture consists of two or more of these modes. Mode I fracture generally occurs at the lowest energy, and is considered the critical fracture mode. Modes II and III fracture have also been investigated (Yoshihara 2000; Ehart 1999).
In mode I fracture, the stress intensity factor (K) is related to fracture energy (GIc) by the Poisson’s ratio (ν ) and the modulus of elasticity (E): GIc=K 2 (1−ν 2 )
The stress intensity factor characterizes the stresses and displacements near the crack tip and is related to the fracture energy, as shown above (Anderson, 1995). The total energy of the system (G) is assumed to remain constant and can be described by: dG = dD + dγ = 0
where D is the potential energy of deformation, and γ is the surface energy of the new surfaces.
where Pc is the load at which crack propagation occurs, C is the compliance of the beams, a is the crack length, and b is the width of the specimen. As the crack length increases along the bondline in a DCB specimen, the stiffness of the beams decreases and the compliance C of the beams increases. A linear relationship exists between the cubic root of compliance, 3 C , and the crack length a: 3 C = ma + b
where m is the slope of the line, and b is the y-intercept (Williams 1989). The correction factor is the distance from the origin to the x-intercept of the line, and can be found using the slope and the y-intercept: χ = b
Figure 3.2 illustrates the correction factor and the correlation between cube root of compliance and the crack length for the DCB fracture specimen.
The DCB fracture test method used in this study follows the form of Equation [3.13].
Contoured dual-cantilever beams (CDCB) were used in the form of a pure epoxy as well as aluminum beams with an epoxy adhesion line. Fracture surfaces were examined to determine fracture ductility and resistance at several temperatures. A method of finding the optimum bond thickness was established, and the effects of bond thickness and temperature on fracture energy were also described (Hunston, et al., 1980).
Ebewele, River, and Koutsky (1979) showed that the adhesive thickness and grain orientation affect the fracture energy. They also found that increases in cure time increased the fracture toughness considerably. The test specimen configuration they used evolved over time and is currently quite simple to manufacture. First, a very precisely machined curvature was used on the two bonded wooden beams in order to have constant stiffness as the crack length increased.
Since then, others such as River and Okkonen (1993) used an aluminum tapered backing reinforcement to keep the stiffness of the beams constant for all crack lengths. This required additional adherents between the wood and the aluminum, which may have caused additional experimental variation. Blackman, et al. (1991) developed four linear elastic fracture mechanics methods for analyzing the data from load-displacement measurements from testing machines. This allowed them to use double rectangular wooden members as the beams, which are bonded together and subjected to cantilever pulling forces. The dual cantilever beam (DCB) method uses a setup similar to Figure 3.3. This methodology has been widely accepted and is now generally used for testing of fracture strengths (Scott, et al. 1992; Lim, et al. 1994).
The DCB method for analyzing fracture mechanics data depend on the specimen behaving in a linear elastic manner. The test gives a load – displacement history as well as crack lengths at arrest. The goal in fracture testing is to keep the fracture within the adhesive bondline. If the wood fails, we do not learn anything about the bondline except that the adhesive is stronger than the wood. But if the adhesive fails, then we can analyze and compare the effects of different environmental exposures, different adhesives, and production parameters on fracture energy.
Studies at Virginia Tech have investigated the fracture energy of DCB specimens before and after a two-hour boil test (Gagliano and Frazier, 2001). Fracture testing is an energy-based test and makes possible direct comparisons of the relative quality of bonding. Differences in performance of types of resins, different press times, or different treatments before or after pressing may be subtle and difficult to see from strength-based tests. The fracture test method has shown significant differences in adhesives that were not shown using the IB test (Gagliano, Frazier 2001).
Fracture testing is valuable for durability studies of composite materials. The fracture characteristics of Parallam PSL have been compared to solid wood and particleboard. PSL had a higher fracture toughness in both weak directions as well as the principle stress direction, using a wedge fracture test (Ehart, et al. 1998). Fracture energy was shown to be significantly reduced when a specimen was loaded for long time periods.
Although common tests such as the compression shear block test and IB test can give an average stress over the area of the bondline, we can learn more about adhesion using an energy based test. For a compression shear block test, 100% wood failure occurs when the adhesive bond is stronger than the wood in tension perpendicular to the grain. Since the adhesive is stronger than the wood in shear parallel to grain, the fracture occurs completely in the bulk wood material and we can not learn much about the adhesion in the bondline. The DCB fracture test method focuses the failure in the bondline, and the analysis is more indicative of surface variations and interactions between the resin and the wood material.
Resin characteristics affect the behavior of adhesive bondlines during cure and during the useful life of the product. The resistance of the adhesive to moisture and temperature changes has a great affect on the durability of the adhesive bond. This chapter addresses the effect of adhesive properties on the fracture energy of bonded wood laminates.
The objective of this study is to compare the durability of phenol formaldehyde (PF) and polymeric methylene diphenyl diisocyanate (pMDI) adhesives. PF and pMDI are two common structural adhesives used for various composite wood products. Since structural adhesives are used in products that are rated for use in outdoor exposure, it is important to learn as much as possible about how these adhesives perform in wet or humid conditions.
The effect of 2-hour boil cycles and 4-day accelerated aging cycles on the fracture energy of wood laminates bonded with PF and pMDI was investigated, and the performance of PF and pMDI was compared.
In order to compare PF and pMDI adhesives, it is important to understand the differences between them and the similarities they have. The following is a brief background about the basic characteristics, manufacturing process, and properties of PF and pMDI.
Phenol formaldehyde resins are thermosets. Thermosets are resins that cure in the presence of heat by an irreversible cross-linking process. The most common PF starts out as a liquid with small polymers of phenol and formaldehyde suspended in an aqueous solution. PF is also used in the form of a powder or an impregnated paper. As temperature increases in the hot-press, the resin softens and flows into the wood surface contours. Provided the wood surface is sufficiently active, the flow of the adhesive allows intimate contact between wood and adhesive. As the temperature continues to increase, the molecular weight of the PF polymers increases due to cross-linking that occurs, and the resin stops flowing, hardens, and turns into a brittle glassy solid. This solidification is the mechanical development of the adhesive.
There are two types of PF resins; novolaks are catalyzed in acidic conditions, and resols are catalyzed by alkaline conditions. Resol PF resins are normally used for wood composite products. PF resins are normally used for exterior grade structural products such as plywood, OSB, and PSL. The speed of cure can be controlled by the formaldehyde to phenol (F:P) ratio, by the pH of the adhesive mixture, and by temperature.
Most PF resins used in the industry are water based solutions, but too much water from other sources, either in the wood or introduced during hotpressing, may cause washout of the resin. Washout occurs when water affects the concentration and viscosity of the resin enough to weaken the adhesion to the wood fibers. The basic components of the resin, such as sodium hydroxide (NaOH) or potassium hydroxide (KOH) are hydrophilic and absorb water and may cause detrimental thickness swell. PF adhesives are not harmed by temperatures up to 250 C, so the products can be hot-stacked, or placed in a pile without having to be cooled down. PF resins are less expensive than isocyanates, but generally cure slower.
Like PF, isocyanate resin is also a cross-linking thermoset. The isocyanate starts out as a monomeric, low viscosity liquid. It is also nonpolar. The liquid readily wets the wood surface, and the small molecular weight facilitates deep penetration of the adhesive into the wood material. PMDI resins penetrate further than PF resins. The resin cures by reacting with the water in the wood and creating urea linkages, which creates a rigid, polar network. This adhesive network has been shown to create urethane linkages with molecules in the wood. This is an important contributor to the properties of adhesion of isocyanates.
Isocyanate resin is used primarily in LSL and OSB. It is highly stable in the presence of water; conversely, the resin does not cure in the absence of water. This hydrolytic stability allows the use of higher MC wood, without risking washout. Some isocyanate wood composite products with larger cross-sections use steam injection hot-pressing to transfer heat uniformly throughout the cross-section.
Since the synthesis of isocyanate resin involves the use of phosgene, a toxic gas, it is more expensive than PF. Isocyanates are not as thermally stable as PF, but they are faster curing. Despite the added difficulty and expense of handling the toxic isocyanates, these adhesives have the advantage of not producing formaldehyde emissions.
Specimens made with each adhesive (PF and pMDI) were exposed to 2-hour boil cycles ranging from 0 to 4 cycles, and other specimens were exposed to 4-day environmental cycles ranging from 0 to 2 cycles (see Tables 4.1 and 4.2). Details of these aging conditions are given later.
CHAPTER 1 – INTRODUCTION
DEFINING THE PROBLEM
SCOPE AND LIMITATIONS
CHAPTER 2 – UNDERSTANDING THE PROBLEM
CHAPTER 3 – FRACTURE TEST METHOD
CHAPTER 4 – USING THE FRACTURE TEST METHOD TO COMPARE THE DURABILITY OF PF AND PMDI ADHESIVES
RESULTS AND DISCUSSION
CHAPTER 5 – FEASIBILITY OF FRACTURE TEST METHOD WITH WOOD COMPOSITES
CRACK PROPAGATION TEST
COMPARISON OF OSB FRACTURE TEST METHOD AND INTERNAL BOND TEST.
RESULTS AND DISCUSSION
CHAPTER 6 – CONCLUSIONS.
GET THE COMPLETE PROJECT
Characterizing the Durability of PF and pMDI Adhesive Wood Composites Through Fracture Testing