Content uploaded by Charles R. Frihart
Author content
All content in this area was uploaded by Charles R. Frihart
Content may be subject to copyright.
Chapter
18
Wood Structure
and
Adhesive
Bond Strength
Charles R. Frihart
Abstract
Much of the literature on the bonding of wood and other lignocellulosic materials has concentrated on
traditional adhesion theories. This has led to misconceptions because wood is a porous material on both
the macroscopic and microscopic levels. A better understanding of wood bonding can be developed by
investigating the theories of adhesion and bond strength, taking into consideration the unusual structure
of wood. Wood is not uniform in the millimeter, micrometer, and nanometer scales. The interaction of
adhesive with wood needs to be considered on these different spatial scales. In addition, emphasis needs
to be placed on the stress concentration and dissipation mechanisms that are active in bonded wood.
Because most adhesives bond wood sufficiently to give wood failure underdry conditions, the emphasis
is on durable bonds, especially those exposed to moisture and/or heat variations. The new hypothesis
emphasizes that for durable bonds, the adhesive needs to give during wood expansion or to restrict wood
expansion to lower stress in the interphase regions. Among the experiments that support this hypothesis,
one study involves the failure mechanism of epoxy wood bonds. Available information indicates that
the fracture occurs near the surface within the epoxy layer. A second study is the bonding of acetylated
wood with epoxy adhesives. Under wet conditions, acetylated wood expands less than does untreated
wood and less stress thus occurs at the interface. In addition, this hypothesis proposes that the primer,
hydroxymethylated resorcinol, is not a coupling agent but stabilizes the wood surface.
Keywords: adhesive, strength, wood. adhesion. failure, epoxy, acetylated, scanning electron microscopy,
swelling, bond, durability
Background
Much of the wood bonding literature has addressed standard adhesion bonding mechanisms. There
has been limited consideration of how these mechanisms need to be modified when wood is the
substrate. Studies have focused on interfacial failure and weak boundary layers. Marra (1980) and
Wellons (1977) addressed many aspects of wood
-
adhesive interactions during the bonding process.
River et al. (1991) studied the preparation of wood surfaces in detail. River (1994) also reviewed
work on the fracture analysis of bonded wood assemblies.
In discussing the processes of bonding and de
-
bonding, it is important to emphasize that the unique
properties of wood need to be examined on several spatial scales. In discussing adhesion and adhesive
241
242
Chapter 18
strength, it is important to separate the process of bond formation from tests of bond perfomance.
To form a bond, the adhesive flows into cell lumens and sometimes into cell walls to form an intimate
contact with the wood surface. If intimate contact does not occur, the bond is poor since all adhesion
requires contact on a molecular level between the adhesive and the substrate.
On the other hand, for an adhesive to have satisfactory strength after solidification, it needs to resist
flow and fracture under a given set of conditions. Adhesive strength is a mechanical property because
it is defined as the ability to hold two materials together under a given set of conditions. However, the
chemical structure determines the mechanical properties of materials. Thus, it doesn’t make sense to
separate whether bond strength is strictly mechanical or chemical given the interdependence of these
factors. When stress is highly concentrated at a location, some bonds will usually fracture there. On
the other hand, if the stress is distributed, then even weaker bonds may not be ruptured.
Adhesive strength is more than adhesion at the interface, although without adhesion there is no
bond. Thus, while the process of bond formation is dependent on thermodynamics and rheology, the
process of bond fracture is mainly a function of viscoelastic dissipation of
energy.
The interphase re-
gions of the adhesive and substrate also play an important role, particularly when the substrate is wood.
One useful method for understanding adhesive strength is the chain link analogy (Marra 1980).
Different areas of the substrate and adhesive are likened to a series of chain links, with the weakest
link being the site of fracture (Figure 18.1). Link 1 is a bulk adhesive layer; this link represents
the properties that are normally measured for an adhesive. At the extremes are links 8 and 9, which
represent the bulk properties of the wood substrate. The smaller links represent smaller layers of the
interphase. Links 4 and 5 are typical interface links, where the adhesive contacts the wood surface.
Fig. 18.1 Chain analogy for bond strength. Link 1, adhesive film; links 2 and 3, intra-adhesive boundary
layers; links 4 and 5, adhesive–adherend interface; links 6 and 7, adherend subsurface; links 8 and 9,
adherend proper. (Figure adapted from Marra 1980. Used by permission.)
Wood Structure and Adhesive Bond Strength
243
Links 2 and 3 are the adhesive interphase regions, the adhesive layers next to the wood surfaces.
These layers are often not fully formed because it is difficult to form a full polymer network in
the constrained environment and the curing chemistry can be influenced by the wood. The lack of
mobility near the surface limits the molecular collisions needed for a normal curing process. The pH
of the wood or adsorption of components can alter the formation of the adhesive structure. Generally,
stress is concentrated at the adhesive interphase region when there are differences in expansion and
contraction characteristics between the adhesive and the wood.
Links 6 and 7 are the wood interphase regions. These areas are often weaker than the wood itself,
because the process of preparing the surface often causes fracture of the normal wood matrix. In
addition, as with the adhesive interphase layer, when there are differences in expansion or contraction
characteristics, forces are highly concentrated at the wood interphase layers. Ways of preparing the
wood surface that lead to a weak surface layer have been discussed by River (1994) and others. Stehr
and Johansson (2000) separated the weak interphase layer into two forms: mechanical weakness,
caused by fracture of the wood surface from planing, sawing, or sanding, and chemical weakness,
caused by the movement of extractives to the surface. I will discuss this model again, but it is
important to remember that more than the adhesive
-
wood interface needs to be considered when
thinking about wood adhesion.
Bond
formation
Bond formation involves several steps. The first is the macroscopic wetting and flow of the adhesive
across and into the wood pores. Normally, the main concern is about equilibrium contact angle (q),
which determines the wetting of the surface. Everything is fine as long as the adhesive has sufficient
time to come to equilibrium. However, in most wood bonding applications, the process is not at
equilibrium because some solvent migrates into the wood and polymerization takes place. Both of
these factors raise the viscosity of adhesive and therefore reduce penetration. Given these factors, it
is more important to be concerned about dynamic wetting [x(dx/dt)], which is similar to equilibrium
wetting but also has an inverse of the viscosity term (Pocius 1997). Dynamic wetting is important
for flow across and into the wood surface. For flow into the pore of distance x, there is an additional
term for the pore radius: normal capillary theory would favor flow in smaller capillaries. However,
the viscosity factor predominates; thus, larger pores are more easily filled.
An additional aspect, which is unique to wood, is the significant flow of some adhesives into the
wood polymer wall. This absorption into the cell walls is different from the adsorption on the cell walls
that is usually considered in adhesion theories. Using a variety of techniques, a number of authors
have shown that some adhesives end up in the wall (Nearn 1974; Marcinko et al. 2001; Schmidt and
Frazier 2001). This flow is obviously controlled by factors on the molecular level, such as solubility
parameter, molecular size, and shape. Simple absorption into the cells is probably not sufficient to
stabilize the wood because the uncured adhesive can leach out again. Reaction with the cell wall
components or polymerization in the cell wall is necessary to provide a stable interphase region.
Polymerization can be inhibited by selective adsorption onto the lignin, hemicellulose, and cellu
-
lose components. In a two
-
component system, the cure could be disrupted by selective adsorption,
but this would not affect a single
-
component adhesive. It is not clear exactly whether the adhesive
reacts within the cell wall, but there is substantial evidence, gained by a number of different methods,
that many adhesives flow into the cell walls. The adhesives in the wall should reduce dimensional
changes with variations in moisture and, consequently, should reduce bondline stress. In summary,
244 Chapter 18
adhesive penetration of the cell walls could play some role in the strength of those walls and the
adhesive bond, but that role has not yet been determined.
Most discussions about wood bonding have used general adhesion models; however, these models
should be adjusted to reflect the particular characteristics of wood that make it different from most
other substrates.
The
mechanical aspects of adhesion are divided into mechanical interlock and
diffusion. For wood, mechanical interlocking could be a significant factor. Certainly, given the
porous nature of wood (hollow cells) and the likely fracture of cell walls, mechanical interlock can
occur in many places in wood as compared to other types of substrates. Although diffusion of organic
adhesives into the substrate does not occur with metals and occurs very infrequently with organic
adherends, the wood cell wall is porous enough that some adhesives can penetrate it, as discussed
in
the previous paragraph. Because a number of adhesives penetrate into the wall, the diffusion of low
molecular weight material on a molecular level, followed by polymerization, can generate nanoscale
mechanicalinterlock.
The chemical aspects ofadhesion involve interfacial interactions between the adhesive and wood.
Van der Waals interactions certainly occur, because for an adhesive to bond, it is necessary to have
molecular
-
level contact. Any chemical that is near another on a molecular level exerts a van der
Waals force. The adhesive strength of van der Waals forces has been shown to be significant by the
ability of geckos to walk vertically or upside down on almost any solid surface (Autumn et al. 2002).
Because most wood adhesives and the wood itself are polar, dipolar interactions should take place
between the adhesive and the wood. In addition, because of the polar nature of most adhesives and
wood, polar bonds, such as hydrogen bonds, should also take place as well as acid
-
baseinteractions
in some cases.
The interfacialmechanism that is still open to considerable controversy is the presence of covalent
bonds between the adhesive and the wood. A number of people have speculated that covalent bonds
occur, but the proof is weak. The other chemical model is electrostatic attraction, which generally
does not occur during bond formation but could occur during bond fracture. Thus, electrostatic forces
are not really a general aspect of adhesion but are more an aspect of separation. In summary, many
adhesion concepts need to be evaluated in consideration of the difference between wood and other
substrates, such as metals and plastics.
Analysis of wood bonds
Wood is an unusual material in that it bas many different spatial levels of structure. Therefore,
adhesive bonding and de-bonding need to be evaluated at different spatial levels (Frazier 2002).
For the purposes of this chapter, I have divided the spatial levels of examination into three tiers:
macroscopic, micrometer, and nanometer. The macroscopic tier is at the millimeter or larger level
that can be seen with the naked eye. In the bonding process, wood is examined for defects and damage
to the surface during preparation, such as the crushing of cells. At this level, wetting is measured
and percentage of wood failure is determined after bond fracture (ASTM 1999). On the micrometer
level as observed with the scanning electron microscope, the concern is wetting of the lumen walls
and failure within the adhesive and wood interphase regions. On the nanoscale level, one area of
study is the penetration of adhesives into the cell wall.
On the macroscale level, areas of concern in the de-bonding process are the applied stress, typical
fracture analysis, delamination, and visual failure (River and Minutti 1975). The applied stresses are
normally macroscopic, but internal stresses are visible on a smaller level. The micrometer scale is
Wood Structure and Adhesive Bond Strength
245
where the fracture is influenced by the cell structure. The nanoscale is the level at which fracture
actually begins and is propagated.
One tier that has not been discussed as much as it should for wood bonding and de
-
bonding is the
cellular level. Because wood structure is really based upon cells joined together, it is appropriate that
adhesion andadhesive strength are considered at the cellular level. Changes that occurduring surface
preparation, such as fracture of surface cells and crushing, are important and have been examined
in relation to weak mechanical layers (River et al. 1991). The examination of wood has involved
transverse wood sections and not the actual bonding surface of the radial or tangential planes. At this
level, adhesive penetration into cells near and far from the bondline has been measured. However, it
is important for failure analysis to investigate this stress distribution and concentration on the basis
of what is happening at the cellular level, as will be discussed later. In addition, the failure within
the cell walls after fracture has been examined
on
only a limited basis (Saki 1984).
The nanoscale level is generally the most difficult to examine. Certainly any adhesive bonding
occurs at
this
level, whether it is van der Waals forces, dipolar interaction. hydrogen bond, or actual
chemical bonds. It also should be noted that in the construction of cell walls, the diameter of the
cellulose fibers, hemicellulose domains, and lignin domains are generally on the scale of tens of
nanometers. This is the level at which fracture propagation and bond breaking actually take place.
To elaborate further on the importance of examination at the micrometer level, a schematic of
the structure ofwood surface as viewed in the transverse plane is shown in Figure 18.2. This cross-
sectional view of surface cells shows that the cells are joined by the middle lamella and are composed
of the primary cell wall and the S1 , S2 , and
S
3 layers; some surface cells also have a warty layer
next to the S3 layer. The bonding surface of wood can be generated by fracturing the middle lamella,
fracturing any of the cell wall layers, or fracturing across the cell walls to open the walls of the
lumen. In fact, because the earlywood cells of some woods have very thin walls that are easily split
in a longitudinal trans-wall mode, the lumen walls can constitute up to 80% of the bonding area.
Thus, the chemical structure of the warty layer is very important for understanding bonding.
The literature indicates that the warts consist of highly cross-lied lignin (Baird et al. 1974), but
in some cases the S3 wall is also exposed, which is rich
in
cellulose. A highly lignin-rich surface
should provide less hydrogen bonding than does a cellulose or hemicellulose layer. However, if the
layer
is
exposed, then this surface is likely to be highly cellulosic. It is not clear what fractions
S3
are exposed in split cell walls. Splitting in the middle lamella exposes a lignin
-
rich layer. Areas
Fig. 18.2 Schematic of wood bonding surface from transverse view.
246 Chapter 18
that are worth further exploration are the chemical nature of the wood surface and a more detailed
understanding of the morphology at the wood surface. This knowledge is important to be able to
characterize how the adhesive interacts with the surface.
Another area that is not clearly understand
is
how much adhesive bond strength is derived from
true interface contact and how much is derived from penetration of the adhesive into the walls
and bonding to the wood components. Penetration of adhesive into the cell wall could be either
mechanical interlocking at the molecular level or reinforcement of the surface cell wall structure
by an intelpenetrating polymer network. Studies have concentrated on interfacial interactions and
the role of extractives. Extractives may not play a large role since some adhesives can solubilize
extractives, exposing the cell wall for adhesive bonding. Solubilization is less likely in the case of
overheating that changes the cell wall itself(Christiansen 1990, 1991).
Another issue is how well the cell wall layers bond to each other compared to the adhesive bond
to the wood. If the adhesive bonds to the warty layer, how well is the warty layer attached to the S3
layer or the S3 layer bonded to the S2 layer? It has often been assumed that the adhesive bonds to
cellulose. However, isn’t it more likely to bond to the lignin or hemicellulose portion since much of
the cellulose
is
tied up in crystalline domains? To understand why some adhesives are better than
others, we need to understand interactions at the molecular level.
Real
bonding
surfaces
The normal appearance of a wood surface is not that of a carefully microtomed scanning electron
microscopy (SEM) specimen, but it is important to understand what the real bonding surface looks
like (Wellons 1980). Figure 18.3 shows a freshly planed tangential softwood surface. The surface of
softwoods has many cells that are split open for good access of the adhesive, but it also has much
debris that could serve as failure points. It is not surprising that in normal planing, surface cell walls
tend to fracture unevenly and generate debris that can result in a weak bonding area. In the hardwood
Fig. 18.3 SEM of southern yellow pine tangential surface
Wood Structure and Adhesive Bond Strength
247
Fig. 18.4 SEM of hard maple tangential surface
surface shown in Figure 18.4, the surface is less open for bonding since the ray cells are sealed and
the fiber cells are small. Thus, longitudinal surfaces are often not open to adhesive penetration. In
fact, most open areas in hardwoods are vessels. Thus, there are areas of poor penetration and good
penetration, making stress distribution very uneven.
Next, it is important to consider the sites where failure can occur. True adhesive failure, that is,
interfacial failure, occurs along a very contorted path that involves the lumen walls and the edges of
the cells. Normally, the percentage ofwood failure is determined visually using limitedmagnification
(ASTM 1999). However, cracks in a layer ofadhesive near the surface, where stress is concentrated,
can appear to be bondline failure. Cell wall failure can also occur by separation of the warty layer
from the S3 layer or the S3 layer from the S2 layer, or by failure within the middle lamella. All of
these, in a visual observation, would appear to be 0% wood failure, but they are interphase failures,
not true adhesion failure. Thus, a number of modes for failure need to
be
understood. The changes
in
an adhesive formulation to solve an adhesion failure would
be
quite different from those needed
to solve an interphase failure.
Examination of epoxy failure modes
One test that has been used extensively for determining the durability of wood bonds is ASTM D
2559 (ASTM 2004). This test involves repetitive cycles of vacuum pressure soaking followed by
rapid heating at 65°C. These cycles cause great stress within the bondlines. First, the soaking cycle
can cause greater expansion of the wood than that of the adhesive. Second, the rapid drying does not
allow the stress relaxation of the wood, resulting in extensive cracking within the wood. In addition,
warping of the wood can cause high normal (Type I) forces on the bondline.
Epoxies do not give highly durable wood bonds (AITC 1992). This is quite surprising given
their ability to give durable bonds to metals and plastics. The literature suggests that the failure of
epoxy
-
woodbonds is interfacial, given the low percentage for wood failure values, but we have
found otherwise (Frihart 2003). To determine where failure may occur on a finer scale, we have
248 Chapter 18
Fig. 18.5 SEM of ‘“wood” side of epoxy bondline failure with southern yellow pine.
exposed delamination sections of ASTM D 2559 (ASTM 2004) specimens by cutting laminated
pieces vertically to expose the failed surface. Visual examination of the failed bondlines revealed
two apparent types of surfaces, one containing a film of the adhesive (“adhesive” surface) and the
other containing many characteristics ofwood (“wood” surface).
The “wood” surface is different from bare wood in a number of ways. First. when the “wood”
surface is held under light at different angles, it is glossier than bare wood; thus, there is likely
a thin coating on the wood. Second, the “wood” surface has a more brown character than bare
wood, probably from the interaction of the adhesive. Third, the “wood” surface has weak yellow
fluorescence under long
-
wavelength UV light, as compared with bare wood. The fluorescence is
typical of the epoxy and suggests the presence of a very thin epoxy coating. Light microscopy of the
“wood” surface reveals smallbeads ofadhesive. Staining withpara-dimethylaminocinaminaldehyde
generally shows a reddish characteristic of reaction with amines, while bare wood only acquires a
grayish cast within a couple hours after treatment. SEM shows that many lumens are filled with
adhesive, indicating that the failure is within the epoxy (see Figure 18.5 compared to Figure 18.3).
Although it is hard to clearly identify wood from epoxy components, the “wood” surface apparently
has a thin coating of adhesive. This surface is different from wood that has been exposed to water
soaking anddrying conditions. Many small fragments appear
on
the epoxy side (“adhesive” surface)
of the failure, which are probably fragments from the planed surface.
The top image of Figure 18.6 shows a possible model of the effect of swelling at the cellular level.
Exposure of wood to water soaking is expected to cause the cells to expand tangentially and radially.
This lateral expansion across the tangential plane causes the wood cell walls to tend to separate from
the adhesive, which is not undergoing similar expansion. Thus, high concentrations of stress occur
along the wood cell walls and the wood
-
adhesive interface. This tension can be relieved by fracturing
the adhesive, the wood
-
adhesive interface, or the wood cell walls.
It
is well known that epoxies have
very low tensile elongation; thus, epoxies are likely to fracture from tensile strain.
On the otherhand, ifthe adhesive orprimer stabilizes the surface cell walls and so limits expansion,
less stress concentration occurs (Figure 18.6, bottom image). Consequentially, the stress may not
be great enough to exceed the tensile limit of the adhesive. This raises the question as to how the
Wood Structure and Adhesive Bond Strength 249
Fig. 18.6 Model of adhesive fracture caused by swelling of cell with water. (This figure also is in the
color section.)
adhesive or primer can alter swelling. One idea is that the adhesive or primer penetrates into the cell
wall during application when the wood is in a swollen state from the water in the adhesive or primer.
If the adhesive displaces the water in the cell wall, the wood surface is unable to
shrink.
Since the
wood is fixed in the expanded state, it does not shrink or swell as much under moisture changes.
Another idea is the formation of an interpenetrating polymer network. Cross
-
linking of the adhesive
occurs in the wall to form a grid, which limits cell wall expansion. This dimensional stabilization of
the cell walls would allow some adhesives that cannot withstand much dimensional change to pass
the durability tests that involve swelling in water, compared to a situation in which cell walls have
not been stabilized.
Bonding to acetylated wood
In addition to using the epoxy fracture studies for developing an adhesion model, a second program
that added support to the concept focused on bonding of acetylated wood. Acetylation converts the
exposed hydroxyl groups to acetate groups, which reduce moisture absorption and associated dimen
-
sional changes. Acetate groups will form hydrogen bonds with hydrogen donors, but the bonds tend
to be weaker than those formed by alcohol groups, which are hydrogen donors and acceptors. There
-
fore, the theory is that acetylation should decrease adhesion strength. This hypothesis depends on the
surface attraction (that is, adhesion) dominating bond strength. However, ifthe wood or the adhesive
250 Chapter 18
interphases are the weaker links in the chain, then the alteration of the surface functionality may not
alter bond strength. Examination of the bond strength of acetylated wood provides insight into the
relative importance ofthermodynamic adhesionrelative to viscoelastic dissipation in bonddurability.
These experiments were done under conditions as comparable as possible to those in previous
work (Vick and Rowell 1990). We modified yellow
-
poplar strips by a high level of acetylation
and bonded them using bonded, untreated wood as the control. In addition, we tested both planed
and unplaned acetylated wood. Unmodified groups may he present on the surface of planed wood,
whereas the surface of unplaned wood should he fully acetylated. The specimens were then tested
for compressive shear by ASTM D 905, using both dry and wet tests. From the thermodynamic work
of adhesion, we expected that the lowest bond strength for the epoxy adhesive should for unplaned
acetylated wood. However, the unplaned acetylated wood had higher bond strength and a higher
percentage of wood failure than did the planed untreated and acetylated wood samples (Figure 18.7).
Thus, the simple act of planing the acetylated surface, which probably freed unmodified hydroxyl
Fig. 18.7 Shear strength and percentage of wood failure for epoxy
-
bonded yellow poplar using ASTM
D 905. CE is planed, untreated wood surface; PE, planed acetylated surface; UE, unplaned acetylated
surface.
Wood Structure and Adhesive Bond Strength 251
groups, created a weaker interphase that led to a great extent of bondline failure. The wet strength
of the control was also low, because this untreated wood expanded and contracted
to
a greater
extent than did the unplaned acetylated wood. These data support the previously proposed model.
Expansion causes the wood cell walls to tend to separate from the adhesive, and the strain exceeds the
tensile elongation of the epoxy. Thus, viscoelastic energy dissipation plays a significant role in bond
durability.
In contrast,
our
experiments with resorcinol formaldehyde (RF) adhesive showedvery little differ
-
ence in the percentage of wood failure whether the wood was wet or dry, acetylated or unacetylated,
or planed or unplaned. All specimens bonded with RF adhesive had a high percentage of wood
failure. Thus, variation in shear strength of bonded assemblies is a measure of the strength of the
wood itself. The RF adhesive stabilized the wood surface so that water soaking did not overstress
the bond. This result could also be due to the fact that the swelling rate of RF adhesive and wood is
similar (Muszynski et al. 2002).
The next area to be considered is the role of hydroxymethylated resorcinol primer (HMR), which
has been used to improve the durability of bonds
in
hard
-
to
-
bond wood, especially with epoxy
adhesives (Vick et al. 1996). The original role ofthe HMR primer was thought to be one ofproviding
a covalent
link
between the wood and the adhesive (Vick et al. 1995); thus, it was referred to as a
coupling agent. On the otherhand, ourstudies have allowed us to develop anothermodel, in which the
HMR serves to stabilize the wood by forming an interpenetrating polymer network or cross
-
linking
of the wood cells (Christiansen 2003). This stabilization of the surface cells reduces their ability to
expand and contract and thus reduces the stress on the adhesive. Given the complexity of wood, both
morphologically and chemically, this model needs further validation.
Conclusions
The bonding of wood is a complicated process. The particular characteristics of wood surfaces need
to be studied in detail and existing models of adhesive bonding and bond failure need to be further
elaborated in light of the characteristics of wood. To do this, we need to consider that the adhesive
bonding area of wood is greater than that of most substrates as a result of the macroscopic porosity of
wood cells. The adhesive can flow into these pores to
develop
large regions of mechanical interlocking.
In addition, fractures in wood surface cells lead to additional penetration of the adhesive and more
regions of mechanical interlock. Some adhesives penetrate the cell walls to provide microscopic
fingers of adhesive that cross the interface. The adhesive in the cell walls can be characterized as
nanoscale mechanical interlocking. Thus, there are many more modes for enhancing the mechanical
bond strength at or near the interface for wood than for most other substrates.
Determination of bond strength is not limited to bond formation. It is important to remember that
adhesives are defined in a mechanical sense. Bond strength is dependent on the concentration or
distribution of internal forces, such as expansion and contraction of the wood in addition to applied
forces. Further work is needed to understand the internal forces at work in tests such as ASTM D
2559 and D 905.
From the examination of epoxy wood bond failure as influenced by water soaking and drying
cycles, we have determined that the failure mode is most likely within the epoxy layer close to the
surface. A proposed model is that expansion of the wood cells causes a tensile strain greater than
the epoxy can withstand and causes fracture of the epoxy. One possible explanation for the ability of
hydroxymethylated resorcinol (HMR) primer to improve adhesive bond strength is that it reduces this
252 Chapter 18
expansion of the wood. Another hypothesis is that HMR bulks up the wood, altering the selectivity
of absorption of the adhesive components and resulting in incomplete cure of the epoxy.
In another set of experiments, we have examined epoxy bonding of acetylated wood. Acetylation
results in more wood failure under water soaking conditions compared to failure of unacetylated
wood. Planing of acetylated wood resulted in bonds with low wood failure when wet. These results
are not easy to understand from a normal surface bonding aspect because acetylation should reduce
hydrogen bonding between the adhesive and the wood. However, the data are more consistent with
the stress concentration model, where the acetylated wood expands less at the surface than it does in
the other two cases.
Application
The results discussed in this chapter have revealed the need for further study of wood bonding
on
the
macroscopic, microscopic, and nanoscale levels to understand how adhesion takes place and how
fracture occurs. Given the complexity of wood, this requires the development of more sophisticated
techniques
to
investigate the processes of both bonding and de
-
bonding. More knowledge about
wood and adhesive interactions can lead to the systematic design of improved adhesives.
References
AITC. 1992. American national standard for wood products. Structural glued laminated timber.
ANSI/AITC A190.1
-
92. American Institute of Timber Construction, Vancouver, Washington.
ASTM. 1999. Standard practice for estimating the percentage of wood failure in adhesive bond joints.
ASTM D5266-99. ASTM International, West Conshohocken, Pennsylvania.
ASTM. 2004. Standard specification for adhesives for structural laminated wood products for
use under exterior (wet use) exposure conditions. ASTM D2559. ASTM International, West
Conshohocken, Pennsylvania.
Autumn, K., et al. 2002. Evidence for van der Waals adhesion in Gecko setae. Proc. Natl. Acad. Sci.
99: 12252
-
12256.
Baird, W.M., et al. 1974. Development and composition of the warty layer in balsam fir. Wood and
Fiber 6(3):211
-
222.
Christiansen, A.W. 1990. How overdrying wood reduces its bonding to phenol
-
formaldehyde adhe
-
sives: A critical review of the literature. Part I. Physical responses. Wood Fiber Sci. 22(4):441
-
459.
Christiansen, A.W. 1991. How overdrying wood reduces its bonding to phenol
-
formaldehyde adhe
-
sives: A critical review of the literature. Part II. Chemical reactions. Wood Fiber Sci. 23(1):69-84.
Christiansen, A. 2003. The influence of polymer structure on the effectiveness of hydroxymethylated
resorcinol (HMR) primer on epoxy bonding to wood. http:/www.forestprod.org/am03abs.pdf
Frazier, C.E. 2002. The interphase in bio
-
based composites: What is it, what should it he? In Proc.
6th Pacific Rim Bio
-
Based Composites Symp., (1):206
-
212. Portland, Oregon.
Frihart, C.R. 2003. Durable wood bonding with epoxy adhesives. In: Proc. 26
"
Annual Meeting,
Adhesion Society, Inc., pp. 476-478.
Marcinko, J., C. Phanopolus, and P. Teachy. 2001. Why does chewing gum stick to hair and what
does this have to do with lignocellulosic structural composite adhesion? In Proc. Wood Adhesives
2000, pp. 111-121. Forest Products Society, Madison, Wisconsin.
Marra, A.A. 1980. Applications of wood bonding. In R.F. Bloomquist, A.W. Christiansen, R.H.
Gillespie, and G.E. Myers, eds. Adhesive Bonding of Wood and Other Structural Materials,
Wood Structure and Adhesive Bond Strength 253
Chapter IX, pp. 365-418. EMMSE Project, the Pennsylvania State University, University Park,
Pennsylvania.
Muszynski, L., et al. 2002. Short-term creep tests on phenol
-
resorcinol
-
formaldehyde (PRF) resin
undergoing moisture content changes. Wood and Fiber Sci. 34(4):612-624.
Nearn, W.T. 1974. Application of the ultrasound concept in industrial wood products research. Wood
Sci. 6(3):285
-
293.
Pocius, A. 1997. Adhesion and Adhesives Technology: An Introduction. Hanser Publishers,
Munich, Germany.
River, B.H. 1994. Fracture of adhesive-bonded wood joints. In A. Pizzi and K.L. Mittal, eds. Hand-
book of Adhesive Technology, pp. 151
-
177.Marcel Dekker, New York, New York.
River, B.H., and V.P. Minutti. 1975. Surface damage before gluing
-
weakjoints. Wood and Wood
Products, July issue.
River, B.H., C.B. Vick, and R.H. Gillespie. 1991. Wood as an adherend. In J.D. Minford, ed. Treatise
on Adhesion and Adhesives, Vol. 7., pp. 1
-
23.Marcel Dekker, New York, New York.
Saki, H. 1984. The effect of the penetration of adhesives into cell walls and failure of wood bonding.
Mokuzai Gakkaishi 30(1):88-92.
Schmidt, R.G., and C.E. Frazier. 2001. Applications of
13
C CP/MAS NMR to the study of the wood
-
phenol formaldehyde bondline. In Proc. Wood Adhesives 2000, pp. 341
-
346. Forest Products
Society, Madison, Wisconsin.
Stehr, M., and I. Johansson. 2000. Weak boundary layers on wood surfaces. J. Adhesion Sci. Technol.
14:1211
-
1224.
Vick, C.B., and R. M. Rowell. 1990. Adhesive bonding of acetylated wood. Int. J. Adhesion and
Adhesives 10(4):263
-
272.
Vick, C.B., et. al. 1995. Hydroxymethylated resorcinol coupling agent for enhanced durability of
hisphenol
-
A epoxy bonds to Sitka spruce. Wood Fiber Sci. 27(1):2-12.
Vick, C.B., et al. 1996. Hydroxymethylated resorcinol coupling agent and method for bonding wood.
US Patent 5, 543, 487 assigned to United States of America.
Wellons, J.D. 1977. Adhesion to wood substrates. In I. Goldstein, ed. Wood Technology, Chemical
Aspects, pp. 150-168. American Chemical Society Symposium Series 43. Washington, DC.
Wellons, J.D. 1980. The adherends and their preparation for bonding. In R. F. Bloomquist et al., ed.
Adhesive Bonding of Wood and Other Structural Materials, pp. 87
-
134.The Pennsylvania State
University, University Park, Pennsylvania.
Characterization
of
the Cellulosic Cell Wall
Proceedings of a Workshop Cosponsored by the USDA Forest Service,
Southern Research Station; the Society
of
Wood
Science and Technology;
and Iowa State University, August
25
-
27,2003,
Grand Lake, Colorado
Edited
by
Douglas D.
Stokke
and
Leslie
H.
Groom
Blackwell
Publishing
Douglas D. Stokke is Assistant Professor in the Department
of Natural Resource Ecology and Management, Iowa State
University and is past president of the Society of Wood
Science and Technology.
Leslie H. Groom is Project Leader for the Utilization of
Southern Forest Resources Research Unit of the US
Department of Agriculture, Forest Service, Southern
Research Station.
© 2006 Blackwell Publishing
All rights reserved
Copyright is not claimed for Chapters 12 and 18, which are
in the public domain.
Blackwell Publishing Professional
2121 State Avenue, Ames, Iowa 50014, USA
Orders: 1
-
800
-
862
-
6657
Office: 1
-
515
-
292
-
0140
Fax: 1
-
515
-
292
-
3348
Web site: www.blackwellprofessional.com
Blackwell Publishing Ltd
9600 Garsington Road, Oxford OX4 2DQ, UK
Tel.: +44 (0)1865 776868
Blackwell Publishing Asia
550 Swanston Street, Carlton, Victoria 3053, Australia
Tel.: +61 (0)3 8359 1011
Authorization to photocopy items for internal or personal
use, or the internal or personal use of specific clients, is
granted by Blackwell Publishing, provided that the base fee
of $.10 per copy is paid directly to the Copyright Clearance
Center, 222 Rosewood Drive, Danvers, MA 01923. For
those organizations that have been granted a photocopy
license by CCC, a separate system of payments has been
arranged. The fee codes for users of the Transactional
Reporting Service are ISBN
-
13: 978
-
0
-
8138
-
0439
-
2;
ISBN
-
10: 0
-
8138
-
0439
-
6/2006 $.10.
First edition, 2006
Library of Congress Cataloging
-
in
-
Publication Data
Characterization of the cellulosic cell wall/editors, Douglas
D. Stokke, Leslie H. Groom.
-
1st ed.
p. cm.
“Proceedings of a workshop co
-
sponsored by the USDA
Forest Service
-
Southern Research Station, the Society of
Wood Science and Technology, and Iowa State University,
August 25
-
27,2003, Grand Lake, Colorado USA.”
ISBN
-
13: 978-0
-
8138-0439
-
2 (alk. paper)
ISBN
-
10: 0
-
8138
-
0439
-
6 (alk. paper)
1. Cellulose
-
Congresses. 2. Plant cell walls
-
Congresses.
3. Cellulose
-
Biotechnology
-
Congresses. I. Stokke,
Douglas D. II. Groom, Leslie H.
QK898.C42C53 2006
571.6´82-dc22 2005016442
The last digit is the
print
number: 9 8 7
6
5 4 3 2 1
