ABSTRACTS

The following is a list of the abstracts for papers which will be presented in SECOND INTERNATIONAL SYMPOSIUM ON ADHESIVE JOINTS: FORMATION CHARACTERISTICS AND TESTING. The listing is alphabetical by presenting author. This list is updated continually to add abstracts as they become available and make appropriate corrections. This list may be conveniently searched by using the editor provided with most popular browsers (e.g. Microsoft Explorer, Netscape, ... etc.)

The Behaviour of Single Lap Joints under Bending Loading

Although there has been a lot of work on the tensile loading of single lap joints (SLJ) there exists only a small amount of work on joints loaded in bending. A structural bonded joint will usually face a combination of tensile, compressive and bending loads. The importance of the bending moment applied at the edges of the overlap of a SLJ has long been addressed by Goland and Reissner⁽¹⁾, Hart-Smith⁽²⁾, and others. However, all these analyses concerned the bending moment induced due to the eccentricity of the load path in the SLJ configuration.

The behaviour of bonded SLJ under bending loading was investigated in the current work. A single part toughened epoxy adhesive and hardened steel adherends were used for the manufacture of the joints. The effect of adherend thickness, adhesive thickness and overlap length was investigated both experimentally and theoretically, using the finite element method and the commercial code ABAQUS. Non-linear large displacement analyses were carried out and the adhesive was modelled using a yield criterion that is dependent on the hydrostatic stress component of stress.

Experimentally it was found that there is a significant effect on the strength of the joints from the adherend thickness while there is no effect in the strength from the overlap length or the adhesive thickness. The joint strength increases almost linearly as the adherend thickness increases while it remains almost constant for joints with different overlap lengths or different adhesive thicknesses.

(1) Goland, M., Reissner, E., 'The stresses in cemented joints', Journal of Applied Mechanics, March 1944.

(2) Hart-Smith, L.J., 'Adhesive bonded single lap joints', NASA Report CR-112236, CR-112237, Langley Research Center, 1973.

Cure Shrinkage in Epoxy Adhesives

The build-up of shrinkage over time during curing and on cooling of an epoxy adhesive has been investigated. A new device, designed and built in-house, was used to measure the shrinkage of the epoxy during cure at room and elevated temperatures. In this investigation, the cure of several epoxy resin systems have been successfully monitored using this device and through the way of measurement of shrinkage change with time. By careful monitoring of the shrinkage, we have shown that the cure contraction at room temperature is of the order of 3.75% by volume and 1.24% expansion, 3.45% contraction at 60ºC for the two-part adhesive. It is also shown how the overall volume change caused by expansion and shrinkage is distributed during the cure process.

The shrinkage has been correlated with the variation of the dynamic shear modulus of the epoxy with time during cure, measured by using a novel torsion pendulum. It was found that the shear modulus increased rapidly during the gel stage, when strong contraction was still occurring. It is therefore likely that significant internal stresses can be built up by cure contractions, even for room temperature cure.

Fatigue and Fracture Behavior of Adhesively Bonded Composite Repairs

(Abstract not yet available)

Effect of Geometrical Non-linearity on the Stress and Deformation States of Adhesive Joints

When the adhesively bonded joints may be subjected to reasonably high loads a non-linearity arises because significant changes in the geometry of the bonded joint although the adhesive and adherends are still elastic. This is called geometrical non-linearity. In this case the linear elastic analysis may be misleading in predicting the stress and deformation states of the adhesive and adherends. Since the displacements and rotations may become large equilibrium equations must be written for the deformed configuration rather than the original configuration.

In this study the small strain-large displacement theory was recalled for the analysis of the adhesively bonded joints and its application to the incremental finite element method was reviewed. Therefore, the material non-linearity in the adhesive and adherends was not considered. This non-linear finite element method was used to analyse the two dimensional stress distributions of an adhesively bonded tee joint with single support plus angled reinforcement.

The tee joint configurations bonded to a rigid base and a flexible base were considered. For each case the linear and non-linear stress analyses of the adhesive tee joint were compared for the different plate end conditions causing the substantial changes (large displacements and rotations) in the joint geometry. The geometrical non-linear analysis showed that the large displacements have considerable effects on the stress and deformation states of the adhesive and adherends. In addition, the effects of the support length and the angled reinforcement length on the peak stress were investigated, increasing these dimensions resulted in an evident decrease in the peak stresses until a specific support or angled reinforcement length/joint length ratio.

1) Dept. of Pure and Applied Chemistry, Thomas Graham Building,

University of Strathclyde, Glasgow, G1 1XL, UK.

2) Dept. of Mechanical Engineering, James Weir Building,

University of Strathclyde, Glasgow, G1 1XJ, UK.

3) Non-destructive Evaluation Branch, Materials Directorate,

Air Force Research Lab., MLLP Building 655,

Wright Patterson Air Force Base, Ohio, 45433-7817, USA

Monitoring the Solvent Uptake in Adhesive Bonded Joints Using a Novel Non-destructive Technique - High Frequency Dielectrics

This paper will present the application of High Frequency Dielectric analysis (frequency and time domain) for the examination of adhesive bonded joints. Adhesive bonding of both primary and secondary structures within the aerospace and automotive industry is becoming evermore popular, necessitating further research into adhesive environmental aging processes and non-destructive evaluation (NDE). In this study, aluminum (chromic acid anodized) to aluminum and carbon fiber reinforced plastic (CFRP) to CFRP epoxy bonded specimens will be monitored, with respect to their aging environment. The aluminum joints were subjected to water under three temperature regimes, with the bonded composite structures being exposed at 65 C to a variety of solvents typically found around aircraft in service, including butanone, hydraulic fluid, ethylene glycol, aviation fuel and simulated sea water. The function of high frequency dielectric measurements as an NDE technique will be introduced, with respect to solvent uptake within the system and the subsequent detection of voids and debonds caused by such solvent ingress. Destructive mechanical shear and cleavage test results will also be commented on. These illustrate the excellent correlation to mechanical strength that can be achieved by this new NDE approach, allowing the user to probe and asses deterioration caused by such aging, within the adhesive and at its interface with the respective substrate layer.

Experimental Evaluation of Methods for Rapid Heating and Curing of

Polyurethane Adhesives

(Abstract not yet available)

Ageing Effects of Environmentally-Friendly Cleaners on Adhesive Bond Integrity

Because of the 1990 Clean Air Act Amendment many chlorinated solvents are being phase out of use in manufacturing industries. Replacement of the ODC (ozone-depleting chemicals) with less volatile, non-ozone depleting cleaners has been extensively studied over the past nine years at Thiokol Propulsion.

Many of the non-ODC cleaners contain compounds that can potentially degrade over time under conditions of high temperature, humidity and exposure to light. The chemical composition of environmentally conditioned cleaners and the subsequent effect on aluminum/Tiga 321epoxy bond integrity as measured by Tapered Double Cantilever Beam were evaluated.

From this study it is observed that moisture content increases for those cleaners containing polar compounds. Non-volatile residue content increases as stabilizers are depleted and the chemical compound limonene is oxidized. A change in aluminum/Tiga 321 bond fracture toughness is observed as some of these cleaners age.

1) Department of Chemical Engineering and Polymer Interface Center, Lehigh University, Bethlehem, PA 18015

2) Boeing Commercial Airplane Company, Seattle, WA 98124

An Arrhenius Method to Study the Hydrolytic Stability of Polymer-metal Adhesion

A theory of rate-dependent bond fracture is used to study the environmental crack propagation of a polymer-metal interface. The model system consists of polystyrene, chemically bonded to an aluminum surface using a silane coupling agent. The aluminum was prepared by phosphoric acid anodization, chromic-sulfuric acid etching and sonicating in a solution of alkali detergent. A styryl-functional silane coupling agent, H₂C=CHC₆H₄CH₂NH(CH₂)₂NH(CH₂)₃Si(OCH₃)₃, was added to the styrene monomer during the polymerization in order to improve adhesion at the polymer-metal interface. The hydrolytic stability of the polystyrene-aluminum interface was investigated using a hydrothermal peel test at 90 deg. configuration. An analysis of the crack propagation rate using an Arrhenius-activated rate theory provided the apparent activation energy of crack propagation. The fracture energy of the interface was found to be velocity dependent. Theoretical model, which based on the Erying chemical activation rate theory, suggested that the fracture of the aluminum-polystyrene interface is influenced by a number of interfacial processes including stress-induced hydrolysis of the bonds at the interface, scission of polystyrene chains, as well as plastic deformation in the polymer. Good agreement between the numerical simulation and the experimental results is shown in this paper.

Thickness Effect in Adhesive Joints

(Abstract not yet available)

York Y01 5DF, UNITED KINGDOM

Easy Removal of Pressure Sensitive Adhesives for Skin Applications

There are two essential requirements of medical pressure-sensitive adhesives: that they should stick firmly to a difficult substrate (skin) and that they should be easily and cleanly removed from that substrate when desired. These requirements would seem to be in conflict: ability to stick firmly is usually characterised by peel testing, while removal of the dressings is by peeling, where the force must be low to minimise the trauma to the skin.

A number of ways have been considered to resolve this conflict and these will be discussed further in the paper. These may be divided into two broad categories: those that make the best of existing pressure sensitive adhesives technology, broadly taking a physical approach, and those that introduce novel chemistry into the process. Physical approaches consider such details as the dependence of peel force on peel angle, peel rate, backing materials and the deformation of the skin during peeling. Careful selection of peel angle and backing material can influence the peel force, though peel rate tends not to have a very significant effect. In addition, the use of crosslinked adhesives, or adhesive microspheres, can increase the ease of peeling of the dressing on removal, but also reduce the ability of the dressing to adhere in the first place.

As an alternative to simply making the best of the physics of the peeling process, various workers have devised chemical systems for making the adhesive less strongly adhering at the time of removal. These systems usually consist of introducing a 'switch' mechanism into a strongly-adhering adhesive so that its adherence may be reduced significantly at the time of removal by operation of the 'switch'. Means of activating the 'switch' include: heat (warming or cooling), application of water via an absorbent backing and exposure to visible light. These may produce physical or chemical changes in the adhesive.

While these approaches bring benefits to patients, consideration of the science behind them is leading to an enhanced understanding of the peeling process.

Fracture of Polymer Interfaces, Interfacial Structure and Crack Tip

Plasticity

When the fracture toughness of an interface is measured via a fracture mechanics tests, the resulting value of G_c is due to dissipative mechanisms at the crack tip. Sometimes these mechanisms are very localized and sometimes they are much more diffuse. We investigated how this crack tip plasticity depends on the molecular structure at the interface, the microstructure near the interface, in particular crystallinity and second-phase, and the degree of mode mixity of the applied stress field. A combination of surface analysis, microscopic observations and fracture mechanics tools will be used to interpret the results which will then be discussed in view of recent theoretical advances on the molecular origin of crack tip plasticity.

Adhesion Behavior of Epoxy-Siloxane to Contaminated Surfaces

During manufacturing, contamination of bond surfaces is always a concern. Selection of an adhesive that is contamination insensitive can help to reduce processing costs by potentially relaxing inspection and cleaning procedures. One of the most dangerous contaminates is silicone because of it propensity to easily wet most surfaces.

A family of functionalized siloxane structural resins has been found to exhibit significant ability form strong interactions with cured silicone rubber. In addition, average lap shear strength data on silicone and hydrocarbon pre-contaminated 304 stainless steel these materials are statistically equivalent to cleaned substrates up to 1.023 mg/cm² of surface contamination. In general, commercially available structural adhesive materials showed high bond-surface contamination sensitivity. Of the three commercially available structural adhesives evaluated silicone surface concentrations higher than 4.65x10^-2 mg/cm² produced bondlines that could not be removed from the bonding fixture without falling apart. These results have shown that these materials offer significant strength advantages over typical UV or two part adhesive materials when bonded to heavily contaminated surfaces.

50 S CENTRAL CAMPUS DR, SALT LAKE CITY UT 84112 9208

Interpreting Adhesive Joint Tests

A myriad of tests are available for determination of the strength of adhesive joints. Organizations such as The American Society for Testing Materials (ASTM) have standardized a variety of these including describing the manner/form of reporting test results. The exact meaning and nature of the test results is not as clear. While the tests often appear rather simple and the results obvious there are usually subtleties that make interpretation very difficult. How the results might be used to predict the strength of other practical joints or in design of practical means of joining parts is even more in question. Even the definition of the strength of adhesive is subject to different interpretations. Test results are typically reported as load at failure divided by the bond area or, for peel tests, force per unit bond width. Such "average stress" results are often misleading, since stresses in joints are typically very non-uniform with the maximum stress differing markedly from this average. Furthermore, tensile stresses normally dominate, even in what are termed "lap shear joints". Increasing the bond area does not result in a concomitant increase in joint strength. This paper discusses research (analytical, numerical and experimental) on tensile, lap, and split-cantilever test configurations. It emphasizes the complex states of stress in joints and the use of test results to predict failure in other joint geometries.

Supported by NSF-CMS-9522743

Rational Joint Strength Analysis by Fracture Mechanics

The central issue in this presentation is rational analysis of adhesive joints by fracture mechanics. Applied calculations and tests relate to various kinds of wood lap joints.

The models discussed are linear fracture mechanics, a quasi non-linear fracture mechanics model and a non-linear model (Wernersson, 1994) taking into account gradual damage and strain softening in the bond line. For the linear and quasi-linear models analytical methods of calculation are dealt with, while results of finite element analysis are shown for the nonlinear fracture model. The bond line properties are for the linear model characterized by fracture energy (critical energy release rate), for the quasi nonlinear model by fracture energy and bond strength and for the fully nonlinear model by a stress-deformation curve which shows the bond line response when the deformation is increased from zero to complete separation of the bond. The nonlinear model is at 2 D and 3D stress analysis defined by two curves and a damage criterion in the normal shear deformation space.

Experimental determination of the complete stress deformation curve of a bond line, including the post peak descending part of the curve, requires use of a stiff testing machine and a stiff test set up or else sudden unstable failure will occur at peak stress. Test results for a few different adhesives are shown for loading in pure shear, pure tension perpendicular to the bond, mixed loading and for pull-out or an axi-symmetric rod.

Theoretical analysis suggested that increased wood strength should be obtained if using a weak adhesive so that failure would take place in the bond layer, not in the wood along the bond line. This finding was verified by tests of single lap joints. It was moreover verified by tests that significantly increased strength of a timber structural element is possible by sawing the element into two pieces and then glue the pieces together again by a ductile adhesive, giving a bond with a high fracture energy. This finding has been developed into a bond layer design with a thin rubber mat glued in between the two wooden adherends. With such a design it is possible to achieve glued timber joints with extremely high load bearing capacity.

Applied fracture mechanics joint strength analyses with comparison to test results are indicated also for heel joints in glued timber trusses, for finger joints and for timber joints with glued in steel rods.

Surface Engineering of Polymer Interphase/Interface System for Enhanced Adhesion of Adhesives and Polymeric Coatings

It has been widely discussed in the subject literature that the creation of covalent bonds across the substrate/adhesive or reinforcement/matrix interface is sufficient for creating viable adhesion in composites, adhesively boned assemblies or in other adhesion-related applications.

The analysis of interatomic and intermolecular interactions reveals that the introduction of the maximum attainable number of chemical bonds between adhering surfaces leads to a theoretical fracture energy G_1c = U_c/A (where U_c: energy required to break a covalent bond; A: the area occupied by molecules bonded through covalent bonds) equal to approximately G_1c = 1J/m². This value, although relatively high in comparison with van der Waals interactions, is, however, frequently unacceptably low for practical, long term durable adhesion.

In contrast with the above, the experimental values of fracture energy for high molecular weight entangled polymers is in the order of G_1c = 100 to 1000 J/m². These values are 2 to 3 orders of magnitude higher than the theoretical fracture energy of covalent bonds, and thus must be investigated and treated as a benchmark for designing polymer interface-interphase systems for acceptable adhesion.

This paper reviews the principles of macromolecular design of polymer interface/interphase systems for obtaining maximum adhesion. Practical and technologically feasible solutions of the theoretical concepts are discussed in detail in the paper. It is shown that the use of chemically attached graft chemicals or controlled spatial geometry and chemical functionality enables significant increase of the strength and fracture energy of bonds to the point of cohesive fracture of the substrate. This occurs even upon prolonged exposure of bonded or painted materials to adverse environments such as hot water, thermal shock, UV radiation and other hostile environments.

Testing and Evaluating the Adhesion of Wood Dispersions by a Semi-quantative Dolly Pull-off Test Following Iso 4624

The adhesion of wood coatings especially under wet conditions is critical for modern waterborne acrylic dispersions. Scientific and technical results received from the "quantitative" tensile strength values derived from the pull-off test prescribed in ISO 4624 were generally poor in repeatability, reproducibility and general meaning for wood coatings. Originally the method and procedure described in ISO 4624 was designed for testing and evaluating coatings on metals and plastics. Hence, a modified pull-off test taking advantage of the weakening of the coating/wood bonds by the influence of water was developed, which results in highly interpretable semi-quantitative data. The adhesion of typical water-borne wood dispersions on pine sapwood samples was measured following this concept. Liquid water has been deposited with a lab injection device for 3 hours into 0,2 mm deep circular grooves around the metal dolly before the pull-off test. The tensile strength values indicated a small standard deviation and the mode of failure was over 90% caused by losses either in the adhesion between wood and coating or the cohesion within the primer. This method has been further used to determine differences in wood/ coating adhesion after wetting for two tropical wood species Red Meranti and Sipo Mahogany and the European softwood pine species (heartwood and sapwood). The findings show a significant dependance of the adhesion means (N/mm²) from the wood species.

Organic Matrix Composites: Factors that Affect Fiber-Matrix Interface Performance

Although adhesion at the fiber-matrix interface in composites has a profound affect on a composite's performance and failure, assessment of this critical parameter is often based on the use of simplistic assumptions that preclude a full understanding of how to fully utilize this information in the design of composite structures. Recent investigations in the NIST laboratory have shown that the determined adhesive strength of the fiber-matrix interface is dependent on testing rate. This result has been linked to the non-linear viscoelastic behavior of the matrix material and its impact on stress concentrations that arise at the fiber-matrix interface. Since "sizing" agents are coated on the reinforcing fibers to promoted durability of the fiber-matrix interface, preliminary results have been obtained on how the "sizing" influences interface failure modes. Research results suggests that the chemical composition, morphology, and thickness of the "sizing" agent layer have a profound affect on the failure modes exhibited by the interface when fiber fracture occurs. The details of this research will be presented and the potential impact of these results on full composite performance will be discussed.

Mechanics of the Sandwich Beam Specimen for Determining Adhesive Shear Properties

There are a number of experiments to determine the shear properties of an adhesive, but most of these tests require very careful measurements to give good results. The loads must be directed precisely to produce a simple shear stress, the specimens must be machined with high accuracy and alignment to get an even adhesive layer, etc. Consequently, a simple test for adhesive shear properties is highly desirable. The sandwich beam test seems to fit this requirement. In this test, two beams are bonded together and the resulting sandwich specimen is subjected to bending. Such loading produces shear in the adhesive so in theory, a measurement of bending modulus for the sandwich can be analyzed to yield a shear modulus for the adhesive. The advantage of this test is that the sandwich geometry is very easy to prepare and most laboratories have the capability to perform simple bending tests. An important limitation is that the sandwich specimen can provide only modulus while some of the other shear tests can generate a complete stress-strain curve. At present, all of the analyses for the sandwich specimen are elastic, and often it is the viscoelastic properties of an adhesive that are of interest. Since most standard viscoelastic characterization equipment can perform the bending experiment (in both transient and dynamic modes), we have been investigating the extension of the sandwich beam test to the characterization of viscoelastic properties for adhesives. In the initial results, the viscoelastic trends in the data looked very promising, but the limiting values in the elastic range did not agree with the predictions from the current available analyses. One possible explanation is that the assumptions made in these analyses are not valid for the test geometry that is required for the viscoelastic equipment. The assumptions are most appropriate when the adhesive layer is either very thick or very thin relative to the adherend thickness. In our experiments, the two thicknesses are similar. To examine this question, we have performed a finite element analysis for the exact geometries used in our tests. In addition, we independently measured the shear properties of the adhesive for comparison with the data from the sandwich specimen. The talk will discuss the results of these experiments.

Synergistic Effects of Environmental Exposure and Cold Temperature Testing on the Fracture Behavior of Bonded Joints

Many aerospace applications of bonded joints require that the joint be durable for many years of exposure to cyclic loads, cyclic temperatures and sustained times at elevated temperatures. For some transport/airliner applications the elevated temperature may be 160(F simulating hot runway conditions. For supersonic aircraft the exposure temperatures may be much higher. All of these aircraft experience cold temperatures (typically -65(F) at a subsonic cruise condition. As such, it is important to determine the strength/toughness of typical bonded structure after long term exposure when subjected to loading at the cold temperature.

This paper presents the results for two bonded systems: Aluminum adherends bonded together with FM-73 adhesive and Graphite/BMI composite adherends bonded together with AF-191.

As suspected, the adhesives are brittle at the cold temperature. However, testing the joints at cold temperature did not result in a large decrease in fracture toughness. The environmentally exposured specimens tested at room temperature also did not show an overly dramatic drop in fracture toughness. BUT when the exposed specimens were tested at cold temperatures the resulting fracture toughness is dramatically decreased. These results clearly indicate the need for such testing to insure the long-term durability of bonded systems. The reason for such behavior will be explained.

Korea Advanced Institute of Science and Technology, ME3221, Kusong-dong, Yusong-gu, Taejon-shi, Korea 305-701

Modeling of Environmental Effects on the Strength of Adhesively Bonded Joints

The strength of adhesively bonded joints degrade when the joints are exposed to high temperature or high humidity environment because of the property degradation of adhesive. In this paper, the strength of adhesively bonded joints was analyzed by finite element method with respect to specific moisture concentration. During finite element analyses, the mechanical properties of adhesive were modeled as a function of specific moisture concentration with the assumption that the fully degraded surface between the adhesive and the adherend could not transfer load due to separation of the adhesive from the adherend.From the investigation, it was found that the strength of adhesives decreased as the absorbed moisture content increased, while the strength of the adhesively bonded joints decreased less compared to that of the adhesives due to moisture swelling effect. Also it was found that the moisture swelling released the thermal residual stress induced during fabrication.

Photocuring Induced Adhesion evolution in marine systems

We report on the use of a proprietary caulk-like photo-polymerizable resin as a marine adhesive and its curing characteristics as a function of water temperature and dose. The photo-polymerizable resin is a two-component mixture of bis-phenol-A-diglycidylether dimethacrylate (bis-GMA) with triethyleneglycol dimethacrylate (TEGDMA) as a viscosity modifier. The photo-sensitive component includes camphorquinone, for which the maximum in absorption is located at 468 nm. This wavelength corresponds closely to the 470 nm wavelength of a home-built lamp system consisting of a grid of 24 LED's. The ingredients are mixed in a dark hood and the resin can be stored for a long time away from light. Measurements of elastic modulus and flexural strength have been done on a photopolymerizable acrylic resin cured underwater at various temperatures and times. These bulk properties show the well-known increasing dependence as a function of the curing time. On the other hand, the water bath temperature does not seem to preclude the cure evolution of the resin. A better control of the sample effective area by the way of a pre-cure degassing and an absence of contact with the bath water may allow us to compensate for the difficulty of its measurement and then to increase the sensitivity of the mechanical characterization.

Effect of Metal Surface Morphology on Adhesive Bond Strength

(Abstract not yet available)

X-ray Studies on the Adhesion and Environmental Effects on the Composite Interface

Residual stress in the structural adhesion and environmental effect on the composite are the most important factors to control the composite properties. We would like to propose a new method for " X-ray diffraction method" to evaluate the residual stress at the interface of Al/epoxy resin system and the environmental effects(surface treatment,heat cycle,water etc.) on the composite interface,Al/epoxy resin, carbon fiber and poly(p-phenylene benzobisoxazole)(PBO) /epoxy resin.

Mitsukazu Ochi, Hiroshige Takahashi; Faculty of Engineering, Kansai University, Suita-shi, Osaka 564-8680, Japan

Adhesive Bonding Properties of Liquid Crystalline Type Epoxy Resin

iquid crystalline type epoxy resins containing some mesogenic groups were synthesized and their bonding properties were characterized, comparing with that of bisphenol-A type epoxy resin. Bonding strength of the former resin system was considerably higher than that of the latter resin system. In the liquid crystalline type epoxy resin system, a large deformation of adhesive layer was observed during the lap shear test. In this resin system, the load onto the adhesive joints should be distributed over the large area of the bonding interface. Namely, high bonding strength of the liquid crystalline epoxy system is due to the large deformation ability with the orientation of network chains. Bonding strength of the cured resin had a maximum peak in the initial curing stage, because of the occurrence of the internal stress in the curing process. Bonding strength of the liquid crystalline epoxy system is remarkably increased with a decrease in the internal stress of the cured system.

David A. Dillard; Engineering Science and Mechanics Dept., Virginia Tech, Blacksburg, VA 24060-0219

Effect of Mode Mixity on the Fracture Toughness of Titanium/Polyimide Adhesive Joints

Using several different fracture tests, the fracture toughness of a chromic acid anodized titanium (Ti-6Al-4V)/polyimide (FM-5) adhesive system was evaluated. Mode I, mode II, and mixed mode (I and II) tests were conducted using double cantilever beam (DCB), end notch flexure (ENF), and mixed mode flexure (MMF) geometries. Interfacial failure types were observed in the ENF and MMF specimens as a result of the mode II loading inherent in these tests. Pure mode I loading, as is the case with symmetric DCB specimens, resulted in cohesive failures with a fracture energy around 2500 J/m^2 on as-received specimens. The assymetric DCB specimens had fracture energy values around 2000 J/m^2, the MMF specimens close to 1970 J/m^2, and ENF specimens around 1300 J/m^2. All the above measurements were made on as-bonded (unaged) specimens. Titanium/FM-5 bonds supplied by The Boeing Company were then aged in one of three different environments for 2 and 6 months respectively. The environments included: 177 C in air and 2 psia, and 204 C in air. Following the aging, DCB, ENF, and MMF tests were conducted on the specimens. The results showed that aging in all three environments resulted in decreases in fracture energy for the above specimen testing configurations. The largest drop (20 percent) in fracture toughness was noted in specimens aged for 6 months in air at 204 C. An unusual finding from this study, in contrast to what other researchers have seen on other systems, was that increasing ,mode II loading resulted in significant reductions in toughness. Crack path selection and interaction with the woven glass scrim within the bonded specimen may be responsible for the lower mode II fracture energies. From the tests conducted, failure envelopes were developed to predict failure energy and type for use in designing structural joints.

Damage Zones in Epoxy-Aluminum Joints

Double cantilever beam tests were performed on a number of toughened epoxies bonded between to aluminum adherents. The epoxies toughened with soft, CTBN rubber particles formed a damage zone consisting of cavitated rubber particles and matrix shear bands. Epoxies toughened with rigid glass beads formed a damage zone consisting of debonded spheres and microcracks. In both cases, the extent of the damage was larger than that found in bulk specimens. A discussion on the role of chemical adhesion versus mechanical adhesion will follow.

Thermoplastic Adhesive system: Experiments and Computational Simulation

(Abstract not yet available)

Bonding of Thermoplastic Composite to Wood: Interface Formation

(Abstract not yet available)

Characterization of Environmental Degradation of Adhesively Bonded Joints

(Abstract not yet available)

1) EUROTECH Ltd., Washington, DC USA

2) AMSIL Ltd., Migdal HaEmek, Israel

3) AMSIL Ltd., Migdal Emek, Israel

Modeling of Silica-Silica Adhesive Bond Formation Using the Statistical Polymer Method

Modeling of formation of adhesive bonds systems with branching/crosslinking like "silica-silica" is one of the most complicated in physical chemistry of interface. Most of conventional methods of description of branched cross-linked structures are numerical and do not allow solution in the general case. Their application needs complicated and long computing and the physical sense of the obtained results is not always clear. In some cases, researchers try application of classical chain models to branched systems, although such approximation is incorrect and may cause serious errors.

The problem of modeling of branched/cross-linked structures was recently solved by statistical polymer method. This method is based on the consideration of averaged structures (statistical polymers) instead o numerous real aggregated forms. In such an approximation, all reactions in equilibrium are considered as reactions between statistical polymers, moreover - between active centers in the statistical polymers able to accept monomers and monomers, with formation of units possessing single bond with the main structure.

The statistical polymer method was tested directly and indirectly. The direct test comprised the exact reproduction of Trommsdorf effect for non-equilibrium polymers, wheras the indirect test allowed the theoretical interpretation of adsorption isotherms for silica and alumina gels. In all cases, the correlation of experimental and theoretical results was very good, sometime excellent.

If the statistical polymer method is applied to modeling of adhesive bond "silica-silica", many of experimental studies are replaced with computer evaluations. That allowed preparation of new adhesive compositions based on quaternary ammonium silicates. Samples prepared with using of theoretical recommendations exhibit excellent technical characteristics.

Ultrasonic Evaluation of Adhesive Joints: An Overview

A number of different ultrasonic nondestructive evaluation techniques have been applied to adhesive joint inspection for decades. Great success has been acquired for delamination detection and cohesive strength determination. Adhesive strength or interfacial weakness, on the other hand, has been difficult to carry out. Some success has been attained for specific material and manufacturing systems.

Presented in this paper will be a description of the literature and current procedures to evaluate adhesive joint integrity. Such topics as modeling aspects, piezoelectric piezocomposites, and electromagnetic acoustic transducer potential, guided waves, focussed beam techniques, "C" scan techniques, high and low frequency aspects, pulse echo and thru transmission testing, oblique incidence, longitudinal and shear wave aspects, pattern recognition and neural net utilitization, "kissing bonds", composite patch bonding, and quadrature phase detection, will be discussed. Future direction will also be outlined including, for example, a comb transducer phase velocity - frequency tuning approach with guided waves.

The Effects of Surface Modification on Wettability and Joint Strength

In order to gain insight into the aspects of surface roughness, surface profile, resin viscosity and surface morphology, experiments are conducted using two different steels (1018 cold rolled and ot rolled weld steel). Four different surface modifications are performed on these samples by grit blasting at 552 kPa from a distance of 25.4mm and chemical etching with a recipe of chromic acid for 2 minutes and 10 minutes respectively. Two different resins are used, namely Epon 815 and epon 830 with viscosities of 5 to 7 poise and 170 to 225 poise respectively, to assess the effect of viscosity in interaction with different surface topographies. Contact angles are measured on all surfaces using the two resins. The metal specimens are bonded under pressure and cured in oven at 121 C for 2 hours. To gain insight into the effect of surface topography, SEM is utilized to observe the modified surfaces at different magnifications. Average and maximum peak-to-valley height values are measured using a profilometer and surface profiles are also obtained for all the samples. The strength and displacement values are measured using single lap specimens at two different cross-head speeds of 1 and 100 mm/min, to assess the interrelation between the failure mechanisms and joint displacements, surface topography and resin viscosity. Results previously obtained using laser ablation of aluminum symmetric single lap specimens are also included for comparison purposes.

Masahiro Yoneno; ( Loctite (Japan) Corp., Fukuura, Yokohama, Kanagawa, Japan )
Hiroshi Kawamura; Loctite Japan -Corp

Yauo Matsunami; Yamanashi University

Stress Analysis and Strength Evaluation of Bonded shrink Fitted Joints

This paper deals with stress analysis and strength evaluation of bonded shrink fitted joints under push-off force and torsion. The stress distributions in the adhesive layer of bonded shrink fitted joints are analyzed by using the axisymmetric theory of elasticity when an external push-off force and torsion are applied to the upper end of shaft. The effect of the outer diameter and the stiffness of rings on the interface stress distributions are clarified by the numerical calculations. Using the interface stress distributions, joint strength is predicted. In addition, joint strength was measured experimentally. It is seen that a rupture of adhesive layer is initiated from the upper edge of the interfaces when a torsion is applied to the upper end of shaft. In addition, it is found that a rupture is initiated from the lower edge of the interface when the push-off force is applied. The numerical results are in fairly good agreement with the experimental results. It is found that the joint strength of bonded shrink fitted joints is greater than that of shrink fitted joints and the availability of the bonded shrink-fitted joints is demonstrated.

1) Department of Mechanical Engineering, Yamanashi University

4-3-11, Takeda, Kofu 400-0016, Japan

2) Kofu Jyousai High School

1-9-1, Shimoiida, Kofu 400-0064, Japan

3) Department of Mechanical Engineering, Yamanashi University

Three-dimensional Finite Element Analysis of Stress Response in Adhesive Scarf Joints Subjected to Impact Tensile Loads

The stress wave propagation and the stress distribution in adhesive scarf joints of similar adherends subjected to impact tensile loads are analyzed in an elastic deformation using three-dimensional finite-element method (FEM). An impact load is applied to a joint by dropping a weight. The one end of upper adherend is fixed and the other end of the lower adherend is subjected to an impact load. FEM code employed is DYNA3D. The effects of scarf angles of the adherends, the adhesive thickness and Young's modulus of the adherends on the stress wave propagation at the interfaces are examined. It is found that the maximum value of the maximum principal stress ₁ appears at the interface of the lower adherend subjected to the impact load. In the cases of scarf angle of 45, 52.47 and 60 degree, the maximum value of maximum principal stress decreases and stress distributions at the interface become to be more homogeneous. The effect of adhesive layer thickness was found to be small on the maximum stress in this joint. The maximum stress increases as Young's modulus of adherends increases. In addition, experiments were carried out to measure the strain response of adhesive scarf joints subjected to impact tensile loads using strain gauges. A fairy good agreement is seen between the analytical and the experimental results.

1) GE Corporate Research and Development

2) U.S. Army TACOM-ARDEC

3) Allied Signal/DOE Kansas City

4) Molecular Simulations Inc.

5) Virginia Polytechnic Institute

6) Naval Air Warfare Center

7) U.S. Air Force Research Laboratory

A New Class of Adhesion Promoters for Addition Curable Silicones

Silicones intrinsically have low surface free energy and therefore either primers or additives, which function as adhesion promoters, must be utilized in order for silicones to adhere to surfaces. Alkoxysilanes which are typically added to promote adhesion in condensation curable silicones function by chemically coupling the silicone network to the substrate providing a deformable interfacial layer. We have developed a new class of adhesion promoters for use in addition curable silicones which operate by an entirely different mechanism. These bifunctional materials are only partly miscible in the silicone matrix and bloom to the surface upon curing. The interfacial layer is comprised of an entanglement of the condensed adhesion promoter and the silicone matrix. The adhesion promoter also functions as an inhibitor in these systems. New room temperature adhesion promoters based on these concepts will be described.

Costantino F. Creton and Hamed Lakrout; Ecole Supérieure de Physique et de Chimie Industrielles, Paris, France

Deformation and Failure Modes of Adhesively Bonded Elastic Layers

Adhesively bonded elastic layers with thicknesses that are small relative to their lateral dimensions are used in a wide variety of applications. The mechanical response of the compliant layer when a normal stress is imposed across its thickness is determined by the effects of lateral constraints, which are characterized by the ratio of the lateral dimensions of the layer to its thickness. From this degree of confinement and from the material properties of the compliant layer, we predict four distinct deformation modes: 1) edge crack propagation, 2) internal crack propagation, 3) bulk cavitation, and 4) fingering. The conditions conducive for each mode are presented in the form of a deformation map developed from fracture mechanics and bulk instability criteria. We use experimental data from elastic and viscoelastic materials to verify the predictions of this deformation map. We also discuss the evolution of the deformation to large strains, where nonlinear effects such as fibrillation and yielding dominate the failure process.

1) Center for Quality Engineering and Failure Prevention, Northwestern University, Evanston, IL 60208-3020

2) Physical Acoustics Corp., 195 Clarksville Road, Princeton, NJ 08550

Detection of Non-Linear Ultrasonic Effects for the Determination of Residual Strength of Adhesive Joints

(Abstract not yet available)

Dental Adhesion to Hard Tissues, Ceramics and Alloys

(Abstract not et available)

1) Dept. Mechanical: Engineering & Applied Mechanics

2) Dept. Materials Science & Engineering

University of Michigan, Ann Arbor, MI 48109-2125

Prediction of the Strength of Plastically-Deforming Adhesive Joints

A general approach for analyzing the fracture of plastically-deforming adhesive joints is presented. This approach couples the use of an embedded-process-zone model to represent the behavior of the adhesive layer, and a non-linear finite-element analysis to calculate the elastic-plastic deformation of adherends. The procedures for determining the model parameters required to analyze mode-I and mode-II loading are discussed. it is shown that once these parameters have been determined, they can be combined with a mixed-mode failure criterion to allow generate simulations of the fracture of different geometries of joints. The excellent predictions that can be obtained with this technique are demonstrated for mixed-mode geometries such as the asymmetrical T-peel and the single lap-shear geometry. The numerical calculations allow quantitative predictions of the loads and displacements (and hence energy absorption) during deformation and fracture to be made. They also result in accurate predictions of how the shape of the deformed joint evolves during fracture.

Fracture characterization of Wood Adhesive Joints

(Abstract not yet available)