ABSTRACTS

The following is a list of the abstracts for papers which will be presented in THIRD 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.)

Structural Integrity of Adhesively-bonded Joints in GRE Pipework Systems

Filament wound GRE epoxy pipes are increasingly being used in critical applications. Recent applications include firewater systems and process piping on offshore oil and gas platforms, and cooling circuits in nuclear power plant. In these applications, structural integrity of the connections is paramount. Frequently the connections are achieved by adhesively-bonded pipe-to-socket joints, in many cases completed on site.

The paper presents the results of a number of studies, both experimental and numerical, of the structural integrity of such adhesively-bonded connections. Aspects covered include basic defect tolerance, creep performance at elevated temperatures and fatigue resistance. Factors which influence the quality of the completed joints are also discussed and outline design guidance for such joints is presented.

Experimental Verification of a Stress Singularity Model in Predicting the Effect of Bondline Thickness on Joint Strength in Bonded Joints

A stress singularity model based on the evaluation of stress intensity factors at the free edge interface corner of an uncracked single-lap joint is used to predict failure in single-lap joints with varying bondline thickness. The theoretical results are compared with experimental results for verification purposes. Essentially the experimental analysis is split into two parts. The first part determining the Young's Modulus and Poisson's Ratio of the adhesive selected for this verification study and the second part measuring the failure loads for a series of single-lap joints with varying bondline thickness. Based on the experimental data of the adhesive properties, predictions are made of the joint strength (provided failure is in the adhesive) for varying bondline thickness. The predicted failure loads, after an initial increase in the low bondline thickness range, is found to decrease with increasing bondline thickness. This agrees well with the trends predicted by the experiments. The stress singularity theory sets the basis of a method for analyzing and designing adhesively bonded joints.

Nondestructive Evaluation of the Interface between a Die Attach Adhesive and a Copper Leadframe in IC Packaging

The silicon die chip and the copper leadframe in IC packaging are bonded by a die attach adhesive, and the quality of the interface is a critical issue in the reliability of IC packaging as well as during the manufacturing process. The common defects such as cracks and delamination can be detected using C-mode scanning acoustic microscopy. However, a weak interface due to poor adhesion has often gone undetected and may become a potential defective area at a later stage. This paper describes the work done to evaluate the quality of the weak interface between a die attach adhesive and a copper leadframe. An interface spring model is used to predict the ultrasonic reflection coefficients. Normal incidence reflection coefficients are measured from a two-layer specimen bonded with a die attach adhesive. The quality of the interface in the samples and its degradation are effected by copper oxidization, and by applying shear stress loading. It is shown that the reflection coefficient depends strongly on both the interface quality and stress loading, indicating that a nondestructive characterization of the interface is possible and the reflection coefficient can be used as a criterion.

Investigation of Optimal Surface Treatments for Carbon/epoxy Composite Adhesive Joints

There are two kinds of joints for composite structures, i.e. mechanical and adhesive joints. The mechanical joint is created by fastening the substrates with bolts or rivets, but the adhesive joint uses an adhesive interlayer between the adherends. An adhesive joint can distribute the load over a larger area than the mechanical joint, requires no holes, adds very little weight to the structure and has superior fatigue resistance. However, an adhesive joint requires careful surface preparation of the adherends, is affected by service environments and is difficult to disassemble for inspection and repair.

In this paper, the suitable conditions of surface treatments such as plasma surface treatment, abrasion by sandpaper treatment and sandblast treatment were investigated to enhance the load capabilities of carbon/epoxy composite adhesive joints. A capacitively coupled radio frequency plasma system was designed for the plasma surface treatment of the carbon/epoxy composites. The suitable plasma surface treatment conditions were experimentally investigated with respect to gas flow rate, vacuum pressure, power intensity, and surface treatment time by measuring the surface free energy of the specimen and the strength of single lap composite adhesive joint. The optimal sandpaper treatment conditions were investigated with respect to the mesh number of sandpaper. The optimal sandblast conditions were investigated with respect to sandblast pressure and media size through the measurement of geometric shape change of adherend by sandblast media. Also the failure mode of the composite adhesive joint was investigated with respect to surface treatment.

* To whom correspondence should be sent.

Grazing Incidence X-ray Diffraction Studies of Adhesion of Polyurethane Thin Films to an Al Substrate

Three unaged and three aged polyurethane films with various OH numbers were prepared by solution casting on an Al surface. Using grazing incidence X-ray diffraction scans, the structural characteristics of polyurethane films bonded to an Al substrate were measured as a function of X-ray penetration, and the results were related to the bond strength of the films. Substrate-induced ordering of the polymer molecules was found to be significant in non-aged samples; aging led to the improvement of bulk crystallinity in all the samples. The bond strength was found to depend upon the crystallinity, measured from the scattering, which included the polyurethane crystalline interphase. It was also found that the bulk crystallinity in the polyurethane films contributed considerably to the adhesion. A relationship between the bond strength and the ratio of the (021) and (100) peak intensities, obtained from the region near the polymer/Al interface was found. Under the condition that the polymer chains align in parallel to the Al surface, ordering in the direction perpendicular to the (021) planes inversely affects the bond strength. A strain-induced distortion, evaluated from the variation in the interplanar spacing of (021) reflection, also appears to change the bond strength.

Local Modelling of Transverse Normal Stress in Double-lap Joint

Although the double-lap assemblage is intended to reduce the normal adhesive stresses, these may cause premature failure of brittle adhesives. Some epoxy adhesives with low ductility, which are brittle in comparison to metals, are reported by R.D. ADAMS (1989). Nevertheless, it seems difficult to predict this type of brittle behavior except possibly with the finite element method.

The present study concerns obtention of approximate analytical solutions, when the behavior remains elastic only for the normal adhesive stresses. This assumption has been made by L.J. HART-SMITH and is mentioned by J.R. VINSON and R.L. SIERAKOWSKI (1986).

Analytical solution for the peel stress were obtained by M. GOLAND and E. REISSNER only for rigid adherends and flexible joint using the hypotheses of "Strength of Materials". However, T.M. ROBERTS (1989) considers that this simplified theory is necessary also in the general case, to obtain simplified analytical solutions for initial design and parametric studies.

The presented modelling concerns originally the local bending of the outer adherends of double-lap joints, at their free edge, to interpret electrical straingage mesurements in the case of a brittle, linear adhesive. Besides, these specimens have been elaborated and tested with an experimental methodology that has been developed by another way. Equations of "Strength of Materials" can be obtained in different forms. This local modelling may furnish closed-form solutions for the peel stress, possibly similar to those of T.M. ROBERTS (1989) which are similar to the results of the finite element method.

Neglecting the peel stresses in the double-lap assemblage corresponds to a "balanced design" which has been employed particularly by N.S. OTTOSEN and K.-G. OLSSON (1988) for the modelling of the shear softening in case of adhesives in conventional epoxy, "non plastified". Closed-form solution is obtained for the optimized length of the overlap which maximizes the force of total breaking. Some assumptions concerning the nonlinear behavior of the joint allow to explain that the shear softening develops more thoroughly in the case of rigid adherends and relatively flexible adhesive layer.

In Situ Cure Monitoring for Adhesively Bonded Joints by Dielectrometry

Since the properties of thermosetting adhesives for adhesively bonded joints are dependent on the degree of cure of the adhesives, on-line cure monitoring is important. In this work, the dissipation factor which is a function of temperature and viscosity was measured and the relation between the dissipation factor and the degree of cure was investigated using the dielectrometry apparatus by auto balancing Wheatstone bridge and the joint self-sensor which did not act as the defect in the adhesive layer. Also DSC results were compared with the dielectrometry. From the investigation, it was found that the dissipation factor showed a similar trend to the cure rate of the adhesive.

Key words: Dielectrometry; Differential scanning calorimetry; On-line cure monitoring; Adhesively bonded joint; Self sensor.

Department of Mechanical Engineering, ME3261, Korea Advanced Institute of Science and Technology, Guseong-dong, Yuseong-gu, Daejeon, 305-701, REPUBLIC OF KOREA

Dependency of Peel Strength of Foam Core Sandwich Beams on the Resin Impregnated Interfacial Surface

The interfacial adhesion characteristics of the foams are very important for the structural integrity of sandwich structures. The peel strength between the face plate and the foam core is one of the appropriate criteria for the interfacial characteristics of the sandwich structures and its peel energy is also measured for the interfacial characterization. The peel strength is the first peak force per unit width of bond line required to produce progressive separation by the wedge or other crack opening type action of two adherends where one or both undergo significant bending and the peel energy is the surface energy required for debonding per unit width of bond line. In this work, to investigate the dependency of the peel strength of foam core sandwich beams on the resin impregnated interfacial surface, the peel strength as well as peel energy of test specimens were obtained by the cleavage peel test and compared with those of non surface treated polyurethane foam core sandwich structures.

Universität Kaiserslautern, Arbeitsgruppe Werkstoff- und Oberflächentechnik (AWOK), Erwin-Schrödinger-Str. 56/344, D - 67663 Kaiserslautern, GERMANY

Durability of Bonded Joints and Reliable Assessment by Appropriate Short-time Testing Methods

The aim of all short-time testing methods which are used to evaluate adhesive joints is to simulate the effects in laboratory in days, weeks or months which can happen to a joint in reality during its lifetime of ten or more years. The usage of such tests becomes more and more important with shorter production cycles and longer life-time expectations for new products and therefore for adhesive joints being used in these products.

The failure mechanisms which can occur in adhesive joints can be divided up into 4 groups: swelling and degradation of the polymer due to diffusion of water, the degradation of the boundary layer of the adhesive, leading to the weak boundary layer, the hydrolytic degradation of the surfaces in the presence of the adhesive and in metal joints the bond line corrosion.

Swelling and the degradation of the polymer and the bondline corrosion can be simulated in simple ageing tests like the immersion test, in storage at warm and humid conditions and in the salt spray test. An acceleration of the damage mechanisms by elevated temperatures is possible. Typical ageing times at temperatures up to 50°C are in the range between 1000 and 2500 hours.

In contrast, the hydrolytic degradation of the metal oxide and the degradation of the boundary layer are difficult to accelerate with elevated temperatures. To detect these failure mechanisms, ageing times longer than 12 months are necessary.

In the presented work, short-time test methods for adhesively bonded joints of steel, zinc coated steel, stainless steel, aluminum alloys and floatglass are evaluated according to the induced failure mechanisms and their suitability for life time prediction.

Using the Asymmetric Double Cantilever Beam Test for Evaluating Polymer-Polymer and Polymer-Metal Interfaces

(Abstract not yet available)

A Novel Adhesive based on Flexibilized Epoxy Resin

Epoxy resins have been used as adhesives since the 1950s due to their better wetting ability, low cure shrinkage, superior mechanical properties and excellent chemical resistance. However, inherent brittleness of cured epoxy weakens its peeling strength and impact strength and therefore limits their use in many applications where the bonded structure has to encounter mechanical shocks. In the present work, attempt was made to develop a shock resistant adhesive by modifying the epoxy resin cured with an ambient temperature hardener, with a carboxyl-terminated poly (ethylene glycol) adipate (CTPEGA). CTPEGA was incorporated into the epoxy matrix by prereact method. The adhesive formulations containing different concentration of CTPEGA were evaluated for Lap shear strength and T-peel strength using various substrates. Chemical modification showed significant enhancement of adhesive joint strength. This can be attributed to the increase in flexibility of the adhesive formulation as reflected by significant increase in % elongation and decrease in glass transition temperature (Tg) over the unmodified epoxy. The optimum adhesion was obtained at 20 phr (phr stands for parts per hundred gram of resin) of CTPEGA concentration. The results are discussed in terms of tensile properties of the modified networks. Dynamic mechanical analysis (DMA) indicates higher damping ( tan = 0.18 at ambient temperature ) which indicates better shock resistance of the modified epoxy over the unmodified epoxy.

Thermal Peel, Warpage and Interfacial Shear Stresses in Adhesive Joints

An analytical model, which includes both bending and extension of the adherends, and extensional and shear strains in the adhesive is developed for predicting stresses due solely to temperature changes in a simple lap joint with similar adherends. Bond-line peel, shear and axial stresses in the adherend and adhesive can be determined. The analytical solution displays "sinusoidal"deformation, consistent with a warping adherend. Plotted results clearly show such warping. While a modified shear lag model (MSLM), with no adherend bending, leads to peak bond-line shear stresses which occur only at the ends of the overlap, the present bending model shows that such stresses occur not only near the ends, but at interior points of the overlap region. Results using aluminum and titanium adherends with an epoxy adhesive, show how the MSLM underestimates the shear stresses. This model, also, cannot calculate peel deformation or adherend warpage. Physical stresses are calculated using the present bending model for a temperature change of AT = 130°C, and the resulting bond-line shear and axial stresses fall in the range of shear and tensile strengths, respectively, of the adhesive.

1 Department of Mechanical Engineering, Yamanashi University, (4-3-11,Takeda, Kofu, Yamanashi, 400-0016 Japan) 2 Josai High School (Kofu, Yamanashi, 400 Japan)

Three-dimensional Finite Element Stress Response Analysis of Single-lap Adhesive Joints of Dissimilar Adherends Subjected to Impact Tensile Loads

The stress wave propagations and stress distributions in single-lap joints of dissimilar adherends have been analyzed by using an elastic three-dimensional finite-element method (DYNA3D). An impact tensile load was applied to the single-lap adhesive joint by dropping a weight. The end of adherend in the single-lap adhesive joint was fixed and the other adherend to which a bar was connected was impacted by the weight. The effects of Young's modulus of each adherend and the thickness of each adherend on the stress wave propagations and stress distributions at the interfaces have been examined. It was found that the maximum stress occurred near the edge of the interface. The maximum stress increased as Young's modulus of the fixed adherend increased. It was also seen that the maximum stress increased as the fixed adherend thickness decreased. In addition, strain response of single-lap adhesive joints of dissimilar adherends subjected to impact tensile loads was measured using strain gauges. Fairly good agreement is observed between the FEM calculations and the experimental measurements.

Double Cantilever Beam Mode I Testing for Vacuum Infused Repairs of CFRP

(Abstract not yet available)

Studies on the Migration of Surfactances in Acrylic Emulsion Pressure Sensitive Adhesives by XPS Analysis

Four acrylic emulsion PSA¡¯s having almost same formulation were prepared respectively by using two conventional anionic surfactants, sodium dodecyl sulphate (SDS) and ammonium nonylphenol-4-ethoxy-sulfate (Co-436), and two polymerizable anionic surfactants, sodium salt(AMPS-2405) and ammonium salt(AMPS-2411) of 2-acrylamido-2-methylpropanesulfonic acid . These four emulsion PSA¡¯s containing 2.23%(SDS),2.30%(Co-436),2.00%(AMPS-2405)and 2.00%(AMPS-2411) surfactant in the bulk were coated, dried and analyzed immediately atomic sulfur at the tape surfaces by XPS . XPS analysis determined 9.74%, 3.87%, 2.15% and 2.10% surfactants respectively at these surfaces versus 3.26%, 3.36%, 2.15% and 2.10% surfactants determined at the same tape suefaces after the samples were immersed in 40 C water for 24 hours and dried . The results showed that the migration of ammonium salt surfactants from the bulk to the surface occured less than the sodium salt surfactants and any migration did not happened for two polymerizable anionic surfactants during the PSA¡¯s were dried . These results also interpreted why the adhesion properties and water resistance of PSA¡¯s from the ammonium salt surfactants are better than that from the sodium aslt surfactants and the PSA¡¯s from the polymerizable anionic surfactants have better water resistance than that from the conventional anionic surfactances.

1) State Key Laboratory of Mould Technology, Huazhong Univ. of Sci. and Technol., Wuhan 430074,CHINA

2) College of Mech. and Mater. Eng., China Three Gorges Univ., Yichang 443002, CHINA

Analysis on the Model of Stress and Hardness Distribution of Joint Made of Steel Adherends under Cleavage Strength Testing

The defects and inadequacy of the "line forced model" used to illustrate the action of cleavage load in traditional theory is analyzed in this paper. The "face forced model" is established for describing the stress distribution of specimen under cleavage load by the authors and verified with the hardness distribution on the surface of steel adherends. By means of measuring the hardness on the failed bonding surface after removing the adhesive layer by solvent, the results shows that there exist evident fluctuations in hardness on the surface of the adherends. It means that the evident work-hardening might be in the surface of the bonded area of the joints during loading procedure. The results obtained also show that there exists an uneven stress distribution on the bonded zone along the length or width of the adherends in cleavage strength testing of the joints. The authors believe that the so-called "line forced model" is unsuitable to analyze he stress distribution in the adhesive layer of the adhesively bonded cleavage joints. It is suggested that the steel adherends be annealed completely before their reusing to obtain testing results with hoped high testing accuracy.

1. State Key Laboratory of Mould Technology, Huazhong Univ. of Sci. and Technol., Wuhan 430074,CHINA

2. College of Mech. and Mater. Eng., China Three Gorges Univ., Yichang 443002, CHINA

Analysis on the Strength Testing Methods of Adhesively Bonded Joints

The defects or problems existing in the typical strength testing methods including shear, tensile, cleavage and peel are analyzed in this paper. It is clear that the stress concentration and synthetic action may lead to an evident deviation in the strength testing of adhesively bonded joints especially for shear or tensile strength. Some main factors affecting the strength of the joints are investigated such as the thickness of the adhesive layer, the parameters of curing procedure, the dimensions of the tensile specimen and the phenomena of work-hardening on the failed surfaces of the most of specimen even on the surfaces of specimen after peel testing. The authors believe that it had better treat the strength value from cleavage test as the fracture toughness of the joints but for the "line acted force carrying" ability of the adhesively bonded cleavage joints. It is suggested that some improvements should be done to obtain testing results with hoped high testing accuracy such as using the specimen and strength tester in small-scale, investigating the relation between load carrying ability and dimension of the actual structure, viewing the cleavage test as a fracture toughness test, establishing new models based on the viscoelastic mechanics for all the testing methods mentioned above.