WO1996023837A2 - Electrically conductive adhesive bondlines - Google Patents

Abstract

An improved method and composition is provided for the preparation of electrically conductive bondlines for composite structures while retaining the desirable physical strength properties for an effective composite bondline. The bondlines are formed through use of an expanded metal foil filled with a non-conductive composite adhesive, the latter selected from a group consisting of thermosetting and thermoplastic adhesive polymers. The improved bondlines provide electrical resistivity less than 0.05 Φ.cm and bond strength in excess of 3000 psi. The expanded metal foils are selected from a group consisting of nickel and stainless steel.

Description

ELECTRICALLY CONDUCTIVE ADHESIVE BONDLINES

Background of the Invention

1. Field of the Invention.

The present invention is broadly concerned with electrically conductive bondline adhesive materials and a process of formation of the materials involving a metallic structure dispersed within a quantity of non- conductive adhesive within a bondline for the purpose of creating a high strength joint having low resistivity.

2. Background of the Invention. A. INTRODUCTION

Carbon fiber reinforced composite materials are being used increasingly as replacements for aluminum or aluminum alloys in aircraft manufacture (Jeanne, 1987). Composite materials are strong, durable, and damage tolerant. They meet many design requirements and offer significant weight advantages. Composite materials have a higher strength-to-weight ratio than aluminum. They can be formed into novel aerodynamic shapes more easily than metals (Constance, 1992). In addition, fiber reinforced composites can be oriented to provide strength along specific directions. Extensive use of composite materials reduce aircraft weight, improve payload, and lessen corrosion problems.

The materials used in aircraft manufacture and the methods used to hold these materials together to form the aircraft structure are the important factors in the protection of a modern aircraft from hazardous natural environ¬ ments. Conventional aluminum airframes of riveted construction have, by virtue of their excellent inherently electrical conductivity, rarely suffered critical damage from lightning strikes. These structures have provided excellent protection for more vulnerable systems.

Carbon fiber reinforced composites currently employed on the new generation of composite aircraft have fairly low inherent electrical resistivity, 6*10^"3 Ω-cm. However, structural adhesives used as part of secondary bonding to fasten composite parts together are not electrically conductive. These adhesives alone are insulators and unable to dissipate the electrical charge caused by a lightning strike, whereas the carbon fiber reinforced composite within the bulk material has this ability. Without continuity of the electrical conductivity within the aircraft structure, the strength of adhesives might be lost if the adhesives are exposed directly to lightning strike. The energy of the strike will be converted to heat and can result in puncture or severe delamination of the composite skin (DeMeis, 1984).

Presently aluminum rivets, bolts and other traditional fasteners are used along with structural adhesives during composite aircraft manufacture, in order to provide sufficient electrical conductivity within the aircraft structure. However, adding these aluminum components not only complicate the aircraft manufacturing, but also increase drag and create point of stress concentration in the aircraft structure (Constance, 1992). Furthermore, the riveting and bolting process is the most costly part of an aircraft's assembly (The Engineer, 1992). The elimination of these processes from the composite aircraft manufacture definitely will benefit both the efficiency and economy of the aircraft industry.

Adhesives not only weigh less than mechanical fasteners, but they also distribute stresses over a larger area. Adhesives thus allow aircraft designers to use thinner materials, reducing both weight and concerns about stress concentration at fastener holes. Performance experience has shown that adhesively bonded structures have dramatically reduced structural repair rates as a consequence of corrosion and fatigue cracking (Reinhart, 1989). Using adhesive technology to replace or support more conventional joining methods is an issue currently at the forefront of the aircraft industry. However, conventional adhesives with good electrical conductivity typically possess poor bonding strength and are, therefore, not practical. B. ELECTRICALLY CONDUCTIVE ADHESIVES As applications for advanced composite materials in the aircraft industry increase, the need to develop techniques to join structural parts of composite materials becomes increasingly urgent. Adhesive bonding is a simple and well-developed technique (Landrock, 1985; Shields, 1988). Automated computer-controlled robotic assembly lines for adhesive bonding have been developed in many industrial areas (Ludbrook, 1990). Because of the specific requirements of lightning protection of the composite aircraft, adhesives or adhesive systems with good electrical conductivity as well as high bond strengths are highly desirable.

Existing electrically conductive adhesives have so far been widely used in the electronic and microelectronic industries (Ogunjimi, 1992). These applications include connecting large-scale integrated circuits or liquid crystal display elements to printed circuit boards, bonding of flexible flat cables to external connecting terminals, and sealing electronic packages. Bonding of battery terminals is another useful application when soldering temperatures may be harmful to the terminal materials. However, these electrically conductive adhesives can not meet strength and volume resistivity criteria

(such as 3000 psi and 6*10^"3 Ω-cm, respectively) simultaneously, these criteria being desirable for aircraft applications.

(1) Electrically Conductive Adhesives

Traditionally, adhesives are made electrically conductive by the addition of either metallic fillers or conductive carbons to a polymer matrix.

Many electrically conductive adhesives have been developed and patented utilizing a variety of matrix and conductive filler systems, (a) Adhesive Matrices

A wide variety of thermosetting as well as thermoplastic polymers have been used to formulate the matrices of electrically conductive adhesives.

Such polymers, including epoxies, acrylics, polyimides, silicones, polyesters, polyethylenes, polystyrenes, polyolefins, and polyurethanes, etc., have been used as adhesive matrices for different applications. Acrylics can be used in low-temperature applications (under 100°C) while epoxies or silicones are more robust and can be used at higher temperatures (up to about 200 °C)

(Pujol, 1989). Polyimides are used in the harshest environments, where temperatures can approach 300°C.

Among the polymers mentioned above, epoxies are the most widely used matrix materials because of their versatility, excellent adhesion, ease of application, and resistance to weathering. Those epoxies which cure quickly are very useful when increased productivity is desired or when it is necessary to minimize time required for holding the bonded parts at curing temperatures. (b) Conductive Filler?

Turning now to conductive fillers, in principle, gold and silver are the best fillers among all conductive fillers, simply because they are chemically inert. Silver is preferable to gold since it is less costly and has a lower resistivity. Usually finely divided silver flake or silver powder is dispersed into the adhesive matrix. The highest silver loading possible is about 85% by weight (Landrock, 1985). Silver loadings lower than the optimum (about 65% by weight) cause conductivities to drop sharply, but offer higher adhesive strengths. Aside from silver and gold, other metallic fillers have included aluminum, chromium, copper, nickel and zinc (see Table 1). Each of these metals presents a common problem. Non-conductive metal oxides form on the surface of the particle resulting in a discontinuity of the conductivity at the contact points of particles. Silver has the advantage of having moderately conductive salts and oxides so that a slight amount of oxidation or tarnishing can be tolerated (Pujol, 1989). Because of this, other metal particles coated with silver are considered to be quite practical in providing relatively inexpen¬ sive conductive particles. The surface properties of these silver coated particles are improved considerably; yet, the conductivities may not be as good as those of pure silver particles.

Non-conductive particles with a thin silver coating also have been reported as conductive fillers in the literature (Lyons, 1991). The base material is either plastic or glass. Plastic particles coated with silver can be made to deform between the opposing contact surfaces and thus provide a large contact area. Silver coated glass particles, however, lead to controlled bondline thickness.

Sometimes carbon (graphite) filler is used as an alternative to metals, yielding fairly low electrical conductivity. Since the recent development of high T_c (transition temperature) ceramic superconductors, Y₁Ba₂Cu₃O₇, has been reported as a conductive filler (Kusakawa, 1989).

(2) Metal-Filled Thermoset Polymers

Metal-filled thermoset polymers were first patented as electrically conductive adhesives in the 1950s (Wolfson, 1956; Matz, 1958; Peck, 1958). In the ensuing years, development of such adhesive systems and their subsequent use in adhesive bonding, particularly for adhesive bonding of miniature devices, e.g. light emitting diodes and integrated circuits, has resulted in partial replacement of traditional weld bonding and thermal compression techniques. However, fundamental understanding of the scientific principles and physico-chemical processes governing the onset of electrical conduction in these materials has lagged far behind their commercial utilization.

It has been generally recognized that percolation theory, first introduced by Broadbent and Hammersley in 1957, can be used to describe the mechanism of electrical conduction by metallic particles within an insulating polymeric matrix (Broadbent, 1957). The random addition of metal particles into an insulating polymer matrix changes the electrical properties of the filled polymer system in a discontinuous way. As the volume fraction of filler increases, no significant change occurs until a critical volume fraction, p_c, is reached. This point, where the electrical resistance decreases dramatically, is called the percolation threshold and has been attributed to the formation of a network of conductive particles that span the polymer (Kirkpatrick, 1973).

For volume loadings smaller than the percolation threshold, i.e. p < p_c, the probability is zero that a conductive passage will span the polymer matrix (Davis, 1975). If enough metal particles are added to form a network within the polymer matrix, electrons can flow across the particle contact points, making the mixture electronically conductive. Resistances are introduced at the particle contact points by layers of adsorbed organic molecules and surface oxides. It is this surface oxide layer that rules out the use of most metals as conductive particles since this oxide film electrically insulates the particles from each other at the point of contact. Only metals such as silver and gold, which form thin, relatively conductive oxides, can be used to provide stable volume resistivities less than about 0.001 Ω-cm (Bolger, 1990).

Establishment of a conductive network requires uninterrupted particle-to-particle contacts of the metallic filler. It was shown theoretically that the initiation of conductivity will take place when the average number of contacts per particle, M, becomes greater than one (Aharoni, 1972). At this point, for every particle with zero contacts, there exists another particle with two contacts; therefore, the probability of conductive chain initiation becomes non¬ zero and the resistivity of the composite begins to decrease. This will continue until M = 2, the point at which all particles have a sufficient statistical probability of participating in infinite chains. Further increase in M beyond 2 will not cause any appreciable decrease in resistivity, since two contacts per particle suffice to assure full conductivity.

A recent study has demonstrated that a short-range percolation coherence length, e, exists when the volume fraction of the metal particles is below the percolation limit, p_c (Etemad, 1986; Quan, 1987). Thus, even if the metal-filled polymer exhibits no bulk conductivity, conduction can occur within domains that are smaller than e. As the volume fraction of metal particles approaches p_c, e approaches infinity and the polymer becomes isotropically conductive. In this theory the actual mechanism of conduction is assumed to be metallic, i.e. the electrons will travel along continuous chains formed by metallic particles in contact rather than by hopping across insulating gaps. The value of p_c has been shown to depend on several factors including processing technique (Bigg, 1986), particle settling (assuming that a random distribution has been achieved) (Nicodemo, 1978), and the size of metal particles relative to any other structures present in the polymer matrix (Malliaris, 1971). It was found that the nature of the matrix does not play a significant part in the value of the conductivity (Pujol, 1989).

(3) An Electrically Conductive Adhesive Database Currently, no adhesives are known which satisfy the adhesive strength and conductivity criteria of 6x10^'3 Ω-cm and 3000 psi when used as structural adhesives. Citations appearing in Table 1 disclose electrically conductive adhesives; almost all, however, are related to electronic and semiconductor applications. Further, while Table 1 discloses a number of adhesives that could meet the resistivity criterion of 6x10^"3 Ω-cm, none could satisfy the adhesive strength criterion of 3000 psi.

Table 1. Citations from Chemical Abstracts Meeting the Volume Resistivity Criterion (6x10³ Ω-cm).

C. ADHESION AND ADHESIVES

Adhesives have been used as a means of joining materials for many centuries. However, it is only in the last fifty years or so that the science and technology of adhesion and adhesives have progressed significantly. The main reason for this is that the adhesives currently applied in nearly all technically demanding applications are based upon synthetic polymer matrices. These materials possess properties that enable them to adhere readily to other materials and to have adequate strength to transmit the applied forces from one substrate to the other. In essence, adhesion is an interfacial phenomenon (Huntsberger,

1970). Physical and chemical driving forces are always operative when materials come together to form an interface. Intimate molecular contact at the adhesive-substrate interface is required for developing strong and stable joints. To achieve the intimate contact, it is necessary to generate intrinsic adhesion forces across the interface. These forces range in magnitude from strong covalent or ionic chemical bonds to weaker physical adsorption, e.g. H- bonding, dipole-dipole, and Van der Waals interactions (Good, 1967).

The mechanism of adhesion is usually considered in two stages: wetting and curing. However, the entire process of adhesion must include the behavior of the joint because the joint performance is used to quantatively evaluate adhesion. Another fact that should be included is the application environment, which often influences the resistance of the adhesive. For example, adhesive joints may appear to be satisfactory at first, but they may fail drastically under end-use conditions, such as a moist environment or mechanical impact. The bonding between the adhesive and substrate depends on the types of interaction prevailing at the interface, whether the species contributing to adhesion are sensitive to moisture and whether voids due to improper wetting are present.

(1) Wetting Phenomena The concept of wetting, originally defined by reference to a liquid drop resting on a solid surface (Figure 1), has been used in adhesive technology. A small contact angle, θ, indicates that the adhesive is wetting the substrate surface effectively; large contact angles show that wetting is poor. The adhesive must spread over the substrate, establishing intimate contact between the adhesive and the substrate. This molecular-level contact, which allows the adhesive forces to develop, removes entrapped air from the interface and maximizes the area over which bonding can develop. Without wetting, strong adhesive bonding is impossible.

In any liquid, surface tension develops as a result of the attractive force exerted by the bulk molecules upon those at the surface layer. This attraction tends to reduce the number of molecules in the surface region, resulting in an increase in intermolecular distance. It was found that the cosine of the liquid contact angle against a solid surface (cosθ) increased linearly toward unity with decreasing values of the liquid surface tension, y, for a homogeneous series of liquids (Zisman, 1963; Zisman, 1977). By extrapolating the curve to zero contact angle (cosθ = 1.0), a critical surface tension, γ_c, a characteristic of that solid surface, can be obtained. Table 2 lists critical surface tensions, γ_c, of certain polymers and metals, and Table 3 lists the surface tensions, y, of common liquids, including epoxy resins typical of those used in standard adhesive formulations.

To ensure adequate wetting, the adhesive should have a surface tension, y^, less than the critical surface tension, γ_c, of the substrate (Zisman, 1962), i.e.,

Liquids having surface tensions below the critical surface tension will have contact angles of zero and will wet the surface completely. Adhesives with surface tensions above the critical surface tension will have finite contact angles. Good wetting results when the adhesive flows into the valleys and crevices on the substrate surface. Poor wetting occurs when the adhesive bridges over the valley and results in a reduction of the actual contact area between the adhesive and substrate, leading to a lower overall joint strength (Petrie, 1975).

(2) Curing Process

After an adhesive wets the surface of a substrate, it is necessary to convert it to a hardened state so that the joint will be capable of bearing loads. This is accomplished by either chemical reaction, cooling from a molten liquid or by drying through solvent evaporation (Petrie, 1989). These processes cause the adhesive to shrink, resulting in internal stresses within the adhesive bondline and possible formation of cracks and voids. Internal stress within the adhesive can significantly reduce the maximum joint strength. In order to minimize internal stresses within a joint, it is desirable to minimize the volume changes by the adhesive during the cure. The adhesive should be less rigid than the substrate; otherwise stress concentrations also can develop at the interface. In addition, it is beneficial for the thermal expansion coefficients of the adhesive and substrates to be similar, especially important when the cured adhesive has a high modulus of elasticity.

Adhesives formulated from neat polymeric resins shrink relatively little during curing. Solvent-based adhesives experience the most shrinkage during cure. The fact that epoxy resins shrink only about 3% upon curing is one reason for their good performance (Tanaka, 1988). Another advantage of epoxy resins compared to many other condensation polymers is that no small molecules are evolved during cure, e.g. water; these can interfere with bonding and form voids within the bondline. Polyurethane reactions are also favorable in this regard. (3) Joint Design

Referring to Fig. 2, there are four types of stress which are commonly studied when considering adhesive bonded joints 20 formed by adherends 28 and adhesive 30. These stresses include (a) normal stress, (b) shear stress, (c) cleavage stress, and (d) peel stress (Kinloch, 1987) (arrows indicate the direction of the applied stresses).

The total stress on any arbitrary plane can be resolved into two components: normal and shear stresses. Normal stress develops when forces act perpendicular to the plane of the joint. The stress is distributed uniformly over the entire area of the bond. Normal stress may be compressive or tensile. When loaded in pure compression, a joint is less likely to fail than when loaded in any other manner. However, compression loaded joints have limited application. Tensile stress is evenly distributed over the joint area, but it is not always possible to be sure that other stresses are not present. If the applied load is offset to any degree, the advantage of an evenly distributed stress is lost and the joint is more likely to undergo failure.

Shear stress occurs when applied forces are parallel to the plane of the joint. This type of loading imposes an even stress across the whole bonded area, utilizing the area of the joint to best advantage and providing an economical joint that is most resistant to failure. Adhesive joints designed for shear stresses are preferable since adhesives generally show considerable strength under this type of stress. Stress should be transmitted through adhesively bonded joints as a shear stress whenever possible.

Cleavage stress arises as the result of an offset tensile force or bending moment. It occurs when forces at one end of a rigid bonded assembly act to split the substrates apart. The stress is not evenly distributed but is concentrated on one side of the joint. A sufficiently large area is needed to accommodate this type of stress in practical applications.

Peel stress arises if one or both of the substrates are flexible. In this type of loading, a very high stress is applied to the boundary line of the joint. Unless the joint is wide or the load is small, failure of the bond will occur.

This type of loading should be avoided if at all possible. Since adhesives generally have poor resistance to cleavage and peel stresses, joints designed to load the adhesive in tension should have physical restraints to ensure axial loading (Shields, 1984). Joints for adhesive bonding must be designed for use with typical adhesives. The following are general principles which will result in maximum effectiveness (Snogren, 1970):

(i) The maximum bonded area should be used within allowable geometry and weight constraints. (ii) The bond should be stressed in its strongest direction, i.e. in shear or tension, (iii) Stress in the weakest bond direction should be minimized, i.e. in cleavage and peel, (iv) Residual stresses due to differential thermal coefficients of expansion should be included in joint design.

Initial design considerations require a knowledge of the chemical and physical properties of both adhesive and substrates. An adhesive joint is expected to perform satisfactorily under the expected service conditions for the planned lifetime of the structure. Thus the changes in the properties of adhesive or substrates involved as a function of the effect of environment, fatigue, temperature, loading rate, and service age must be known or predicted (Reinhart, 1973).

(4) Conventional Test Methods

Mechanical test occupies an important position in adhesive technology. It is essential for the development, qualification, processing, and use of structural adhesives. The most commonly used standard test methods for assessing performance of the adhesive joints are those issued by the American Society for Testing and Materials (ASTM series). In some areas, the U.S. Military and Federal Adhesive Specifications are used as standard. The test methods most commonly used for adhesive tensile butt joint and single lap shear joint are ASTM D 2095, "Standard Test Method for Tensile Strength of Adhesives by Means of Bar and Rod Specimens", and ASTM D 1002, "Standard Test Method for Strength Properties of Adhesives in Shear by Tension Loading (Metal-to-Metal)". Although these tests provide information useful for comparison purposes, they give little indication of how the adhesive will perform in an actual bonding application. They remain popular because they are relatively quick and simple to perform and enable unsuitable candidates to be screened out. Other methods, involving durability and fatigue testing, are used to predict the long-term performance of adhesively bonded structures.

(5) Adhesive Selection

In most cases, the primary criterion for the selection of an adhesive is its ability to support the design loads under the service conditions required for the planned life of the structure; thus, the mechanical properties, durability, and environmental resistance of the bonded structure are obviously important. Of equal concern are the nature of the substrates, the application technique, the cure conditions, the handling requirements and the cost. The performance may vary considerably within a given class of adhesives; it is essential to consider the properties of each particular adhesive system for use in a given adhesively bonded joint.

Table 2. Critical Surface Tension of Some Polymers and Metals (Petrie, 1989; Kinloch, 1987)

Table 3. Surface Tension of Some Adhesives and Liquids (Petrie, 1989; CRC, 1993)

Accordingly, the requirements for the preparation of electrically conductive bondlines for securing composite structures together are exceed¬ ingly stringent, with the most troublesome difficulties being the preparation of such bondline that is both strong and electrically conductive to enhance electrical ground planes throughout a structure.

Summary of the Invention

The present invention overcomes the problems outlined above, and provides an improved method and composition for the preparation of electrically conductive bondlines for composite structures while retaining the physical strength properties required for effective adhesive bondlines. The invention is predicated upon the discovery that an improved bondline can be formed through use of a continuous metallic structure in the form of an expanded metal foil filled with a non-conductive composite adhesive, the latter selected from a group consisting of thermosetting and thermoplastic adhesive polymers. The improved bondlines provide electrical resistivities less than 0.05

Ω-cm and bond strength in excess of 3000 psi. The expanded metal foils are selected from a group consisting of nickel and stainless steel.

In more detail, the metal foil is configured with an array of openings ranging between 200-2500/in², each such opening having dimensions of about 0.05-0.125 in. (LWD), bondline thicknesses of between 0.005-0.012 inches and foil volume fractions of between 0.08-0.22%.

Brief Description of the Drawings Figure 1 shows a liquid drop resting on a solid surface.

Fig. 2 shows various types of stresses: (a) normal stress; (b) shear stress; (c) cleavage stress; (d) peel stress.

Fig. 3 shows the joint geometry used to measure tensile strength and bondline conductivity for a tensile butt joint and a lap shear joint. Fig. 4 shows a DSC curve of 3M Scotch-Weld (TM) EC-3448

Structural Adhesive (1XA) used to evaluate ΔH_P and ΔH.

Fig. 5 shows the estimated degree of curve of 3M Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) as a function of time, at different temperatures. Fig. 6 shows the estimated degree of curve of 3M Scotch-Weld

(TM) EC-3448 Structural Adhesive (1XA) as a function of temperature.

Fig. 7 shows the curing time of 3M Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) as a function of temperature for different conver¬ sions. Fig. 8 is a schematic of a portion of a metal foil defining various characteristic dimensions.

Fig. 9 shows a schematic diagram of the volume resistivity measurement system for a butt joint adhesive bondline.

Fig. 10 is a plot of volume resistivity of the adhesive bondline as a function of the number of openings for the copper foil systems.

Fig. 11 is a plot of volume resistivity of the adhesive bondline as a function of the long way of diamond (LWD) for the copper foil systems.

Fig. 12 is a plot of volume resistivity of the adhesive bondline as a function of the bondline thickness for the copper foil systems. Fig. 13 is a plot of volume resistivity of the adhesive bondline as a function of the volume fraction of the metal for the copper foil systems.

Fig. 14 is a plot of volume resistivity of the adhesive bondline as a function of the number of openings for the nickel foil systems.

Fig. 15 is a plot of volume resistivity of the adhesive bondline as a function of the long way of diamond (LWD) for the nickel foil systems. Fig. 16 is a plot of volume resistivity of the adhesive bondline as a function of the bondline thickness for the nickel foil systems.

Fig. 17 is a plot of volume resistivity of the adhesive bondline as a function of the volume fraction of the metal for the nickel foil systems. Fig. 18 is a plot of volume resistivity of the adhesive bondline as a function of the number of openings for the stainless steel foil systems.

Fig. 19 is a plot of volume resistivity of the adhesive bondline as a function of the long way of diamond (LWD) for the stainless steel foil systems.

Fig. 20 is a plot of volume resistivity of the adhesive bondline as a function of the bondline thickness for the stainless steel foil systems.

Fig. 21 is a plot of volume resistivity of the adhesive bondline as a function of the volume fraction of the metal for the stainless steel foil systems.

Fig. 22 is a plot of the average resistances of the composite adherends for adhesive lap shear joint specimens. Fig. 23 is a plot of the average resistances of the adhesive bondline incorporating different metal foils for lap shear joint specimens.

Fig. 24 is a plot of the tensile strength of lap shear joints having adhesive bondlines incorporating different metal foils.

Description of the Preferred Embodiments

EXAMPLE 1 A. ADHESIVE CHARACTERIZATION

(1) Adhesives

Ten commercially available structural adhesives were investi- gated in the past (Liaw, 1990). It was found that Scotch-Weld (TM) EC-3448

Structural Adhesive (1XA), manufactured by the Aerospace Materials Department of 3M, was the most appropriate adhesive among all investigated for the purposes of the examples. Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) is an adhesive which currently is employed extensively in composite aircraft manufacturing.

(2) Adhesive Characterization

Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) consists of about 60 - 85% epoxy resin, with 10 - 20% resorcinol, 5 - 15% non-volatile amide, and other minor components. A Perkin Elmer DSC 7 Differential Scanning Calorimeter was used to characterize the cure reaction of Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA). Approximately 30 mg of adhesive was transferred to a standard aluminum sample pan (Perkin Elmer Part No. 0219-0041) and the net adhesive weight measured on a Mettler PE 360 electronic scale. An aluminum sample pan cover then was placed on the sample pan and the adhesive encapsulated by a standard Perkin Elmer crimper press (Perkin Elmer Part No. 0219-0048). The sealed sample was loaded into the sample holder of the DSC 7 and an empty sample pan with cover was placed in the reference compart- ment. The temperature was scanned from 40 to 210 ^CC at a constant rate of

2°C/sec. Nitrogen gas was used as a purge gas at an inlet rate of 40 psi. The DSC data were obtained following the DSC 7 operating instructions (Perkin Elmer, 1987).

(3) Data Analysis Data analysis was performed on the Perkin Elmer 7700

Professional Computer by DSC 7 Kinetics software package (Perkin Elmer).

This program uses a multilinear regression to fit DSC data obtained at a constant heating rate assuming Borchardt-Daniels kinetics (Borchardt, 1957).

The curing processes of the adhesive can be written as a chemical reaction

k A > p

where A is the uncured adhesive, P is the cured adhesive, and k is the degree of rate constant. An nth order reaction is assumed with the reaction rate expressed in terms of the reaction conversion, x (Prime, 1973),

— k ( l x ) ⁿ (A . l ) dt

where t is the time in seconds. It is assumed that the reaction rate constant can be written in an Arrhenius form

E RT

Z e (A . 2 ) where Z is the pre-exponential constant, E_a is the activation energy of the reaction in J/mol, R is the universal gas constant 8.314 J/K-mol, and T is the absolute temperature in K. Substituting Eq. A.2 into Eq. A.1,

-- Z(l x)- e ^B- ^M (A.3) dt

In differential scanning calorimetry, the experiment can be designed such that the temperature of a reference material varies linearly with time; thus,

T T₀ β (A.4)

and dT

(A.5) dt

where T₀ is the initial temperature, and β is the scanning rate. Combining Eqs. A.3 and A.4,

f_— Z(l x)ⁿ e ^{E* RT} (A.6) dT

The reaction conversion, x, is obtained by

ΔH

(A.7)

ΔH

where ΔH_P and ΔH are partial heat at any time t and total heat for the reaction, respectively. The values of ΔH_P and ΔH are evaluated from the area under the DSC curve, as shown in Fig.4.

Equation A.6 is the rate equation used to obtain kinetic parame- ters from the DSC data and is programmed into the DSC 7 Kinetics software package. Eq. A.6 can be linearized by taking the natural logarithm of both sides. rix ln(β— ) ln(Z) E /RT nln(l x) (A.8) dT ^a A multilinear regression is performed using ln(βdx/dt), 2-1/RT, and ln(1-x) as variables, from which Z, E_a and n can be obtained. The following kinetic parameters for the cure reaction of Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) were obtained. Z = 6.95*10¹⁶ (1/sec)

E_a = 146.54 (kJ/mole) n = 1.55 Isothermal reaction kinetics can be simulated based on the kinetic data above. Integration of Eq. A.3 with an initial condition of x = 0 at t = 0 and for a reaction order, n, not equal to 1 yields

E /RT

[ 1 Z ( n 1 ) e t ] 1 n (A . 9 )

Substituting the values of Z, E_a, and n into the equation above, one obtains x 1 ( 1 3 . 82 * 10¹⁶ e ^{17626 τ} t ) ¹ - ⁸¹⁸ (A . 10 )

A plot of conversion, x, verses time, t, at different temperatures according to Eq. A.10 is shown on Fig. 5. Based on these simulations, conversions in excess of 99% cure at temperature between 120 and 130°C for 120 minutes easily can be achieved. The same conclusion can be drawn from Fig. 6 in which conversion is plotted against temperature. In addition, Figure 6 also shows that no conversion is anticipated rf the cure temperature is less than 60 °C. This is a very important information because heating the adhesive to 60 °C will reduce the viscosity of the adhesive. This allows the adhesive to be preheated during sample preparation without concurrent reaction. Rearranging Eq. A.10, one obtains the following equation.

( 1 x ) ° - ⁵⁵ 1

(A . 11 ) 3 . 82 ^χ l 0¹⁶ e ^17626/τ

Fig. 7 generated from Eq. A.11 gives another perspective on control of the adhesive cure reactions under discussion. A conversion of 99% can be reached by holding the sample at 130°C for 120 minutes. Therefore, the following conditions were chosen for curing the Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) in the experimental work described herein. Temperature: 130°C Time period: 120 minutes

B. EXPERIMENTAL - BUTT JOINT AND LAP SHEAR JOINT TESTS As is common of an adhesive bonding procedure used in the composite aircraft manufacture, a nylon scrim is embedded in the bondline in order to control the bondline thickness. In Examples 2 and 3, however, an expanded metal foil is included in the adhesive bondline instead of nylon to improve the electrical conductivity of the bondline. The change affects the adhesive bondline properties such as the strength and the electrical conductiv¬ ity of the bondline. In Examples 2 and 3, a series of adhesive joint experiments were conducted to examine these properties, based on the different expanded metal foils incorporated.

(1) Materials for Butt Joint and Lap Shear Joint Tests

In the examples, a continuous metallic structure is provided in the form of MicroGrid precision-expanded foils. The expanded foils were obtained from Delker Corporation, and are generally similar to the foil shown in Figure 8. In the practice of the invention, any net or grid of conducting material may be used so long as it is capable of also functioning as a bondline spacer.

An explanation of foil product codes is given for the code "4Cu4.6-125" by making reference to Fig. 8. The first number represents nominal, original base material thickness 38 (4 = 0.004"). The letters are chemical symbol for material (Cu = Copper). The number immediately following letters represents strand width 40 (4.6 = 0.0046"). The last number indicates long way of diamond (LWD) 42 (125 = 0.125"). The copper, nickel and stainless steel foils with the properties listed in Table 4 were utilized in Examples 2 and 3. The number of openings 41 is based on one square inch of expanded foil. It is related to the long way of diamond (LWD) 42 of the foil, as shown in Fig. 8, according to Delker's specification. Table 4. Specifications for the MicroGrid Precision-Expanded Foils.

Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) was used as the adhesive for these examples. Brass plugs were employed as adherends for butt joint adhesive bonding. The adhesive is epoxy-based, having a reported surface tension of 47 mN/m (Table 3). The critical surface tension of 1 ,000 mN/m, a value for copper (Table 2), was assumed for the brass alloy used herein. Therefore, this adhesive-substrate selection meets the require¬ ment of good wetting; i.e. the surface tension of the adhesive should be less than the critical surface tension of the solid surface, as discussed above.

EXAMPLE 2 (2) Adhesive Butt Joint Tests

Axially loaded butt joint tensile strength measurements of Scotch- Weld (TM) EC-3448 Structural Adhesive (1XA) into which expanded metal foils were incorporated were conducted according to standard test method ASTM D 2095 (ASTM D 2095, 1993). The joint geometry used in this example is shown in Fig. 3 and involves adherends 32 placed in abutting relationship to define a butt joint 34 having therein an adhesive bondline 36. The tensile strength of adhesive bondline 36 is measured by axially loading in the direction of the arrows as shown in Fig. 3. Prior to the tensile strength measurement, the resistivity of bondline had been determined. (a) Sample Preparation

Butt joint adhesion specimens were prepared as follows: (I) Polish the surface of each brass plug on which the adhesive is to be applied by a wet wheel method with abrasive paper disks having 320 and 600 grid (Carbimet Paper Discs, Buehler).

(ii) Clean the surface with water. Dry the surface. Rinse with methyl ethyl ketone to remove any grease and possible copper oxidants remaining on the surface.

(iii) Cut the expanded metal foil to the same size of the brass plug surface, and weigh the foil on an electronic scale (Mettler PE 360). Rinse the expanded foil with methyl ethyl ketone.

(iv) Apply the adhesive onto the surface of both brass plugs with a flexible plastic stick.

(v) Place the cleaned metal foil on the surface of one brass plug, close the bondline with another brass plug, and set the brass plug joint specimen in the assembly jig.

(vi) Put the assembly jig in the Isotemp vacuum oven (Fisher Scientific, Model 281 A) at 50°C for 10 minutes to soften the adhesive.

(vii) Take the brass plug joint specimen, together with assembly jig, from the oven.

(viii) Press the top of the bonded brass plug specimen by hand as hard as possible to ensure good contact between the metal foil and the surfaces of brass plug in the bondline. Tighten the screws on the assembly jig to maintain the contact. (ix) Wipe off any excess adhesive squeezed from the joint with methyl ethyl ketone wetted Kimwipes.

(x) Place the brass plug joint specimen together with assembly jig in the oven. Cure the adhesive at 130°C for 2 hours.

(xi) Take the brass plug joint specimen out of oven and cool it to room temperature.

(b) Butt Joint Resistivity Measurements

Resistance measurements of the composite adhesive bondline were based on standard test method ASTM D 2739 (ASTM D 2739, 1993). Volume resistivity was calculated from the measured resistance through the cross-sectional area of the bondline. The cured butt joint adhesion specimen was connected to the circuit as shown in Fig. 9. A constant power supply 50 (E-C Apparatus Corporation, Model EC-400) was used to provide constant DC voltage over a range from 0 to 1000 volt. The constant DC current output option was chosen in the experiments. The connection was made to the outer pair of terminals 52 of the butt joint adhesion brass plugs 54 and 56. A Keithley 195A Digital Multimeter (Keithley Instrument Inc.) was employed as the voltmeter. The potential drop was measured across butt joint bondline 60 through the inner pair of the terminals. A Fluke 8050A Digital Multimeter 62 (John Fluke Manufacture Co., Inc.) was utilized as the ammeter. Standard and miniature banana plugs were applied to connect the electrical apparatus with the tensile butt joint adhesion plugs. The junction between the brass plugs and the wire banana plugs was covered with a transparent, half-spherical plastic cover to reduce the impact of ambient conditions, such as air conditioning or body heat, on the measuring system. Detailed discussions including precautions on low voltage measurement can be found elsewhere (Liaw, 1990).

The procedures for resistance measurement of the butt joint adhesive bondline 60 are given below:

(I) Turn on the ammeter, voltmeter, and DC power supply. (ii) Set the meters to the proper measurement range.

(iii) Warm up the entire system for 2 hours to allow equilibrium, (iv) Ensure that the readings of ammeter and voltmeter stabilize for at least 5 minutes before taking data.

(v) Record the current and voltage drop. (vi) Calculate the resistance by Ohm's law and volume resistivity as follows:

The electrical resistance, R, is calculated from Ohm's law using measured current and electrical potential data

^R -j

where

R = electrical resistance, Ω V = electrical potential, V I = current, A The volume resistivity can be calculated from the electrical resistance πd ²R

P

4 1

where p = volume resistivity, Ω-cm d = diameter of brass plug, cm

I = thickness of the bondline 60, cm.

(c) Butt Joint Tensile Strength Measurements

After having determined the volume resistivity of the adhesive- metal foil bondline, the butt joint tensile strength of this composite adhesive system was measured on a Tinius Olsen Super "L" Type Hydraulic Universal Testing Machine (Tinius Olsen Testing Machine Co., Inc.) Threaded attach- ment fixtures were used to connect butt joint adhesion brass plugs 54 and 56 with the Tinius Olsen standard specimen holders, one mounted on the fixed top crosshead and the other on the movable lower crosshead of the test machine. The procedure below was followed.

(i) Turn on the instrument power. (ii) Select an appropriate load range; 6000 lb was the upper limit in most of the tests.

(iii) Reset the load gauge to erase the previous record, (iv) Zero the load gauge.

(v) Start the motor that drives the hydraulic pump. (vi) Adjust the movable lower crosshead by a pinion shaft handle until the brass plugs are tightly assembled between two threaded attachment fixtures with the load value pointer still reading zero.

(vii) Apply load to specimen at an approximate rate of 40 lb/sec according to ASTM D 2095. (viii) Stop the motor immediately upon the specimen failure and record the ultimate load applied.

(ix) Calculate the tensile strength as follows: Tensile strength of the butt joint is calculated from the breaking strength or maximum load of the butt joint tensile measurement divided by the cross-sectional area of the adhesive bondline

4 F σ πd '

where σ = tensile strength, lb/in² F = breaking load, lb d = diameter of brass plug, in

(x) Carefully tear off the adhesive-metal foil layer from the brass plug with a knife and measure the average thickness of the bondline with a micrometer (Mitutoyo). (xi) Calculate the volume fraction of metal in the adhesive bondline as follows:

The volume fraction of the metal foil in the adhesive bondline is calculated from

4m πd ²lp

where f = volume fraction of metal foil in the bondline m = metal foil mass, g p = density of the metal, 8.941 g/cm³, 8.889 g/cm³, and 8.166 g/cm³ for copper, nickel, and stainless steel, respectively, d = diameter of brass plug, cm I = thickness of the bondline, cm

EXAMPLE 3 (3) Adhesive Lap Shear Joint Tests

In this example, single lap shear joints (shown in Fig. 3), having composite adherends 64 defining a lap shear joint 66 having therein a bondline 68, were examined. Meanwhile, accelerated aging simulation of these adhesive joints were also investigated. Tensile tests were performed in accordance with the requirements of the standard method ASTM D 1002 except that composite adherend was used instead of metal (ASTM D 1002, 1993). (a) Lap Shear Joint Test Materials

Scotch-Weld (TM) EC-3448 Structural Adhesive (1XA) was the adhesive used. Expanded metal foils prepared from three different materials, 4Cu4.6-125 for copper, 3Ni5-077(AN) for nickel, and 4SS(316L)6.5-100(AN) for stainless steel, were incorporated into single lap shear joint adhesive bondlines. Three groups of specimens, twenty-one each representing each type of metal foil, were prepared. Carbon fiber composite lap shear adherends were used for the joint. The critical surface tension of the composite material was assumed to be 900 mN/m, a value reported for graphite (Table 2). It is greater than the surface tension of the epoxy-based adhesive (47 mN/m, Table 3), thereby meeting the requirement for adequate wetting by the adhesive.

(b) Lap Shear Joint Sample Preparation All the lap shear joint specimens were fabricated by Beech Aircraft using Beech Specification 13201 D-1 panels. Metal foil was cleaned with methyl ethyl ketone and bonded with the adhesive in the temperature range of 120-150°C for 60-90 minutes. 20 psi pressure was applied for adhesive bonding. The resistance measurements were performed on the adherends alone before the bonding and on the joints after the curing of the adhesive.

Three groups of single lap shear joint specimens were placed in two Blue M Vapor-Temp Controlled Temperature and Humidity Chambers for accelerated aging. The conditions in the humidity chambers were controlled at 70% relative humidity for the first three weeks through setting the dry bulb temperature at 35°C and the wet bulb temperature at 30°C. Then the condition was adjusted to 85% relative humidity by resetting the wet bulb to 33 °C. The relative humidity could be converted with the dry and wet bulb readings from the Relative Humidity Chart which came with the humidity chambers. The moisture pick-up of each specimen was monitored by weighing the specimen on an electronic scale (Mettler PE 360) periodically until no weight gain was observed. The specimens remained in the chambers for the next twenty-nine consecutive weeks before taking the tensile strength test. (c) Lap Shear Joint Tensile Strength

Measurements

Lap shear tensile strength tests were carried out on the Tinius

Olsen Super "L" Type Hydraulic Universal Testing Machine. A pair of rack & pinion-type flat, "V" wedge grips were installed on both the top and bottom crossheads in order to hold the lap shear specimen. The test procedure below was followed.

(I) Turn on the power to the testing machine, (ii) Select 6000 lb as the load range. (iii) Reset the load gauge to erase the previous record.

(iv) Zero the load gauge.

(v) Start the motor that drives the hydraulic pump, (vi) Insert the specimen between grips and adjust the movable lower crosshead with a pinion shaft handle to secure specimen. (vii) Ensure that the load value on the pointer still reads zero.

(viii) Apply load to specimen at an approximate rate of 20 lb/sec according to ASTM D 1002.

(ix) Stop the motor immediately upon the specimen failure and record the ultimate load applied. (x) Measure the adhesive bondline area with a ruler.

(xi) Calculate the tensile strength as follows: Tensile strength of the lap shear joint is calculated from the breaking strength or maximum load of the lap shear joint tensile measurement divided by the cross section area of the adhesive bondline

F ab

where σ = tensile strength, lb/in² F = breaking load, lb a = length of the bondline, in b = width of the bondline, in

C. RESULTS AND DISCUSSION The volume resistivity and tensile strength of the bondlines formed from

Scotch-Weld (TM) EC-3448 structural adhesive with embedded expanded metal foils were measured in Examples 2 and 3 to determine whether such adhesive systems are suitable for bonding composite materials in aircraft manufacture. This application requires good adhesive bond strength as well as sufficient electrical conductivity within the bondline to avoid damage from possible lightning strike to composite aircraft.

(1) Results

In Examples 2 and 3, both tensile butt joints and single lap shear joints were fabricated from three different expanded metal foils - copper, nickel and stainless steel. Tensile butt joint specimens were fabricated with four to six mesh sizes for each expanded foil. At least four replicates were made for each measurement. The lap shear joint specimens were made from a single mesh size for copper, nickel and stainless steel. Twenty-one replicates were obtained for the single lap shear measurements,

(a) Adhesive Butt Joint Tests

Six different copper foils were incorporated into tensile butt joints. A summary of the volume resistivity and tensile strength results for this series are shown in Table 5. A plot of the volume resistivity of adhesive-expanded copper foil bondline against the number of openings of copper foil is shown in

Fig. 10. As the number of openings increases, the volume resistivity increases. A plot of the volume resistivity versus the long way of diamond (LWD) of copper foils is shown in Figure 11. As the LWD increases from 0.031 inch to 0.100 inch, the volume resistivity decreases; it stays constant for LWD's ≥ 0.125 inch. Figs. 10 and 11 indicate the same trend since the number of openings is proportional to the reciprocal of the LWD (see Table 4). Fig. 12 illustrates the volume resistivity of the adhesive-copper foil bondline as a function of bondline thickness. The volume resistivity decreases with increasing bondline thickness, but remains constant when the bondline thickness is greater than 0.009 inch. A plot of the volume resistivity against volume fraction occupied by the copper foil in the adhesive bondline is shown in Figure 13. The resistivity does not correlate with the volume fraction of copper. The lowest volume resistivity achieved from the adhesive-copper foil systems is approximately 0.04 Ω-cm. Table 5. Butt Joint Results of Adhesive-Copper Foil Bondlines

10

I

Average (Standard Deviation) I

15

Table 5 also shows the butt joint tensile strengths of these adhesive specimens. Most of the adhesive bondlines incorporating copper foils meet the strength criterion of 3000 psi. Based on all of the data, copper foil with product code 5Cu14-189(AN) has the best properties as a conductive foil filler among the copper foils tested.

Only four types of expanded nickel foils were available for incorporation into adhesive butt joints. The volume resistivity and tensile strength results are given in Table 6. Fig. 14 is a plot of the volume resistivity of adhesive-nickel foil bondlines as a function of the number of openings of the nickel foil. There is an increase in the volume resistivity as the number of openings increases from 275 to a maximum at 635. Fig. 15 illustrates the volume resistivity of the adhesive bondline as a function of the LWD of nickel foil. There is an increase of the volume resistivity as the LWD increases from 0.050 inch to 0.077 inch, but the volume resistivity decreases as the LWD increases further. Fig. 16 shows the volume resistivity of the adhesive-nickel foil bondline as a function of the bondline thickness. The decrease in the volume resistivity is nearly linear with increasing bondline thickness. A plot of the volume resistivity against volume fraction occupied by nickel foil in the adhesive bondline is shown in Fig. 17. The volume resistivity of the adhesive bondline increases with increasing volume fraction. The lowest bondline volume resistivity obtained from nickel foils was approximately 0.03 Ω-cm.

The butt joint tensile strength data for the nickel foils are given in Table 6. Only one adhesive bondline strength exceeds 3000 psi. Among the four nickel foils that were tested, the foil with product code 3NM0-125 performed best as the conductive foil filler.

Table 6. Butt Joint Results of Adhesive-Nickel Foil Bondlines

10

Four stainless steel foils were investigated in butt joint adhesion experiments. The results of the volume resistivities and tensile strengths obtained are summarized in Table 10. A plot of the volume resistivity of the adhesive-stainless steel foil bondline against the number of openings of stainless steel foil is given in Fig. 18. As the number of openings increases, the volume resistivity also increases. The volume resistivity decreases as the LWD of stainless steel foil increases (Fig. 19). These data are consistent with those of copper and nickel foils. Fig. 20 illustrates the volume resistivity of adhesive- stainless steel foil bondline as a function of the bondline thickness. As the bondline thickness increases (up to 0.012 inch), the volume resistivity decreases steadily. This behavior is similar to that of copper and nickel foils. The thicker bondline results from thicker stainless steel foils which correspond to larger LWD's of the foils. This plot confirms that adhesive bondlines incorporating stainless steel foils with larger LWD's will give lower bondline volume resistivities. A plot of the volume resistivity of the adhesive-stainless steel foil bondline against volume fraction occupied by the stainless steel foil in the adhesive bondline is shown in Fig. 21. The volume resistivity remains constant for bondline thicknesses from about 0.11 to 0.14, but the volume resistivity increases as the volume fraction increases beyond 0.14. The lowest volume resistivity achieved from stainless steel foils was approximately 2 Ω-cm.

The volume resistivities measured remain essentially constant when the bondlines are fabricated using stainless steel expanded foils with less than 600 openings/in², with the LWD greater than 0.075 in., and a bondline thickness greater than 0.008 in. The butt joint tensile strengths of the adhesive-stainless steel foil systems fabricated are shown in Table 10. Three out of four tensile strengths are over 3000 psi. This indicates that the adhesive bondline tensile strength will be in the neighborhood of 3000 psi when stainless steel foil is incorporated into the bondline. Table 7. Butt Joint Results of Adhesive-Stainless Steel Foil Bondlines

10 Average (Standard Deviation)

(b) Adhesive Single Lap Shear Joint Tests

Examples of copper, nickel and stainless steel expanded foils with different LWD's (product codes 4Cu4.6-125, 3Ni5-077(AN) and 4SS(316L)6.5-100(AN), respectively) were selected for lap single shear joint tests. Selection of these particular materials was made before the adhesive butt joint experiment results were known. Therefore, the data may not be indicative of optimal performance. These three metal foils were incorporated in the adhesive bondline of single lap shear joints having composite adherends. The resistance measurements, including the composite resistance and the resistance of adhesive-metal foil bondline, were made. The results are summarized in Table 7.

Each group of adhesive-metal foil lap shear joints contain twenty-one replicate measurements. After thirty-two weeks of accelerated aging (85% relative humidity at 35°C), the lap shear joint tensile strength measurements were performed. Because of omission of some adhesive bondline resistance readings in the original data (see Table 8), their corresponding composite resistance were not included in the average and confidence interval calcula¬ tions. Some of the experimental observations appear to be inconsistent; therefore, a statistical test following ASTM E 178 standard method was conducted to reject these outliers (ASTM E 178, 1993). The results listed in Tables 9, 10, and 11 are the individual data, averages, and 95% confidence intervals for adhesive lap shear joints with copper, nickel, and stainless steel expanded foils, respectively. The omissions and rejected outliers have been excluded.

Table 8. Resistance (in Ω) of Composite Single Lap Shear Joints with MicroGrid Precision-Expanded Foil in the Bondline

Specimen Group I Group II Group III

Number 4Cu4.6-125 3Ni5-077AN 4SS(316L)6.5-100AN Composite Bondline Composite Bondline Composite Bondline

Note : NR - - No reading

Table 9. Adhesive Lap Shear Joint Results with 4Cu4.6-125 Foil Incorporated in Bondlines

Data Reported as Average ± 95% Confidence Interval

Fig. 22 shows the relationship between the composite resistance and the three materials used to form single lap shear joint specimens. Fig. 23 illustrates the resistance of adhesive-metal foil bondline as a function of the incorporated material. It is seen that copper foil provides the lowest bondline resistance while nickel foil has the highest. The bondline resistance for stainless steel foil system stays in between though it has the lowest resistance of composite adherends which are made of the lap shear joints. Fig. 24 is a plot of the average single lap shear tensile strength of the adhesive-metal foil bondline system against metal used in the bondline. The specific values measured will have a strong dependence on the specific geometry of the metal foils incorporated into the bondlines. All the single lap shear strengths remain above 3000 psi. (4) Discussion

(a) Adhesive Butt Joint Tests

Essentially, the volume resistivity of the adhesive-metal foil bondline increases as the number of openings of the expanded metal foil in the adhesive bondline increases (as in Figs. 10, 14 and 18). This is consistent with the plot of the volume resistivity against the LWD of the foil (Figs. 11 , 15 and 19).

These plots may imply that the larger LWD's provide better adhesive penetra¬ tion through the expanded foil network; extra adhesive applied in the bondline can be squeezed out easily. Dense openings of the foil inhibit flow of the adhesive into the foil network simply because the adhesive has a high viscosity. Preheating adhesive during bondline preparation presumably reduces the adhesive viscosity, if the degree of cure can be limited. Larger LWD's may provide more intimate contact between the expanded metal foil and the substrates (brass plugs in this instance); this is crucial for the enhance¬ ment of electrical conductivity in the adhesive bondline. Thinner adhesive-metal foil bondlines usually result from foils which have a smaller LWD and a thinner original base material. Figs. 12, 16 and 20 illustrate that the volume resistivity of the bondline decreases as the bondline thickness increases. The volume resistivity behavior observed indicates that a thicker bondline does not necessarily mean decreasing contact between the expanded metal foil and the substrates. In contrast, foils with larger LWD's provide better contact and lead to lower resistivity. This agrees with the data discussed above. In addition, experience from aircraft manufacturers suggest that a bondline thickness less than 0.010 inch will lead to weaker adhesive bonding due to a lack of adhesive in the bondline. Therefore, a relatively thicker adhesive bondline is preferred. Plots of the volume resistivity of adhesive-metal foil bondlines against the volume fraction occupied by the expanded foil are shown in Figs. 13, 17 and 21. For copper-based systems, there seems to be no strong dependence of the volume resistivity on the volume fraction of the expanded copper foil in the bondline. This suggests that volume fractions of copper foil as low as 7% in the adhesive bondline exceeds the copper content required to reduce the volume resistivity of the bondline. It is much less than the critical volume fraction predicted by percolation theory. If this is the case, then much less copper is needed in the adhesive-copper foil system to make the adhesive bondline conductive than the traditional method having particles as the conductive filler. The possible reason is that the expanded copper foil has a preexisting network which guarantees contact in the two dimensions parallel to the surface of the substrates. However, the traditional particle filler has to reach two or more contacts per particle (M ≥ 2) to get good conductivity. Therefore, the weight fraction of silver particles required for currently available electrically conductive adhesives typically exceeds 65%. Volume fractions of the conductive particles observed in practice often exceed 60%. Overfilling conductive particles in the adhesive may occur to guarantee the conductivity. This is one of the advantages of metal foil-based conductive adhesive over the conventional type of conductive adhesive in some applications since the latter requires at least 20% by volume of the conductive filler (Gurland, 1966; Aharoni, 1972; Miller, 1966). Less metal filler in the adhesive dramatically reduces the weight of composite adhesive system. This is a crucial factor in the application of conductive adhesive systems in aircraft fabrication. The volume resistivity tends to increase as the volume fraction of expanded foil for nickel and stainless steel systems increases. This is not what would be usually expected, because more conductive foil within the bondline should reduce the resistivity of the bondline. However, other effects may dominate here. One possible reason may be related to the configuration of the individual nickel and stainless steel foils. The surfaces of the expanded foil

(whether nickel, copper, or stainless steel) which contact the two substrates typically are not flat initially. Each foil has its own geometry. This geometry may play a more important role in the outcome of the volume resistivity. Intimate contact between the nickel or stainless steel foil and the substrates could rely more on the geometry of the foil rather than the volume fraction of the nickel foil in the adhesive bondline. One point worth noting is that both nickel and stainless steel foils are much tougher than copper foils. The formers' geometry is not as easily altered when a limited amount of pressure is applied in adhesive bondline preparation. Copper foil is more ductile. The amount of pressure applied during processing will flatten the surface of the foil.

Noticeable changes have been observed in the shape of copper foils during its use in the examples. Flat surfaces of the foil can increase contact area between the substrates and the foil. This may be the reason that the volume resistivities of copper foil incorporated adhesive bondlines behave differently as compared with nickel and stainless steel foils.

Among the three groups of expanded metal foils studied, incorporation of copper and nickel foils into the adhesive bondlines resulted in volume resistivities as low as 0.03 - 0.04 Ω-cm; stainless steel-based systems reached volume resistivities only on the order of 2 Ω-cm. Therefore, copper and nickel with larger LWD's are more appropriate for low resistivity applications.

Both copper and stainless steel foils incorporated into adhesive bondlines have tensile strengths exceeding the strength criterion of 3,000 psi. However, only one of four nickel foil incorporated adhesive bondline systems exceeded this value. It is suspected that the adherends of the butt joint specimens may have been misaligned.

(b) Adhesive Lap Shear Joint Tests

The average resistances and their 95% confidence intervals of the composite adherends used to fabricate adhesive-copper, adhesive-nickel and adhesive-stainless steel foil lap shear joints are 7.8±0.9 Ω, 8.3±0.9 Ω, and 6.7±0.9 Ω, respectively (see Table 9). It was found that the composite adherends used for adhesive-copper foil and adhesive-nickel systems are similar; but, there is a significant difference in the resistance of the composite adherends for the copper and nickel systems and those in the stainless steel system, based on a statistical test of average resistance difference for Type I error α = 0.05 (Ott, 1988). The average resistance of the composite adherends used for adhesive-stainless steel foil lap shear joints is lower than those used for adhesive-copper and adhesive-nickel systems (Figure 22).

For adhesive-copper foil lap shear joints, the average resistance of the bondline is about 1.7 times as much as the composite material alone. For adhesive-nickel foil and adhesive-stainless steel foil lap shear joints, the average resistances of the bondlines are 4.5 and 3.6 times, respectively, the value of the composite alone. The copper system achieves the lowest bondline resistance, just as observed from the adhesive butt joint results. Interestingly, although the average resistance of the composite adherends used to fabricate adhesive-stainless steel joints appears the lowest, the lowest adhesive bondline resistance does not occur from this group of adherends.

Fig. 24 demonstrates that the tensile strengths of the single lap shear joints are well above 3000 psi. This clearly establishes that incorporation of such expanded foils within a bondline can be used as spacers to fabricate single lap shear samples with acceptable strengths.

The adhesive-metal foil bondlines were examined right after the tensile tests of the lap shear specimens. No corrosion or rust were observed on the metal foils. The metal foils had a glossy appearance as usual. This was a good indication that corrosion within the bondline is minimal after thirty-two weeks of accelerated aging at 35 °C and about 85% relative humidity; but, it is too early to say when corrosion will ultimately not develop within these adhesive systems and become a factor in bondline properties.

D. CONCLUSIONS (1) Generally.

Electrically conductive adhesives with volume resistivities less than or equal to about 6*10^'3 Ω-cm, a value typical of carbon fiber composites used in fabricating composite aircraft, and lap shear strengths exceeding 3000 psi, are highly desirable for applications such as secondary bonding within composite aircraft structures.

It was found that the lowest electrical resistivity achieved from adhesive butt joint measurements was in the range of 0.03 - 0.04 Ω-cm for copper and nickel foils (product codes 4Cu4.6-125 and 3NM0-125). Although this result is somewhat higher than the volume resistivity criterion (6*10^'3 Ω-cm), it dramatically reduced bondline resistivity from that of an insulator to that typical of semiconductors. Independent resistance measurements of single lap shear joints demonstrated that the adhesive bondline resistance with the incorpora¬ tion of copper foil (product code 4Cu4.6-125) is only 1.7 times as great as the composite material alone. The tensile strengths of these lap shear joints were well above 3000 psi, meeting the specified requirements. These newly developed electrically conductive adhesive systems offer many features that the traditional particle filled conductive adhesives do not. They are easy to prepare and their application procedures are simple. They do not have problems typically associated with particle filled conductive adhesives, such as limited lifetime due to precipitation of particles during storage and unstable electrical properties as contact between particles is lost. These adhesive systems do not need to be premixed; simultaneous application is possible. The preexistence of the foil network guarantees the electrical contact and conductivity. (2) Alternatives to Expanded Foils.

All of the examples have been fabricated using the expanded metal foils manufactured by Delker Corporation. The success of these foils in reducing the bondline resistivities indicates the feasibility of the general approach of incorporating an appropriate metal scrim into the bondline. However, successful fabrication of these bondlines in the practice of the invention would not be limited only to the Delker expanded metal foils. One essential concept in the practice of the invention is the incorporation of a net grid of a conducting material within the bondline as a spacer. The Delker MicroGrid expanded foils are merely one such family of materials which has been shown to achieve the properties desired and which is commercially available.

(3) Alternatives to Scotch-Weld EC-3448 Structural Adhesive.

The majority of the examples were based on paste adhesive, Scotch- Weld EC-3448 Structural Adhesive. Other adhesives could be utilized in the practice of the invention. The key issues are the final adhesive strength and impregnation of the adhesive throughout the foil. The final adhesive strength measured will be very dependent upon the choice of the adhesive. Thorough penetration of the adhesive throughout the conductive scrim employed will depend upon developing effective cure schedules for forming the bondline. At a minimum, a two-step heating process will enhance penetration of the adhesive through the metal scrim and subsequent cure of the adhesive. First, the bondline is heated to, and held at, a temperature high enough to reduce the viscosity of the adhesive such that it flows readily throughout the bondline, but sufficiently low to minimize curing of the adhesive during the impregnation process. Second, the temperature of the bondline is increased to complete cure of the adhesive once uniform impregnation of the adhesive has occurred throughout the bondline structure. The details of the cure cycle will depend on the specific adhesive system being employed; and, these conditions would be modified for the particular adhesive to accomplish the goals stated above.

(4) Bonding of Materials Other Than Carbon Fiber-Epoxy Materials. The utility and applicability of this technology extends beyond the bonding of carbon fiber-epoxy composite materials. Even if the materials being adhesively bonded are electrically conductive, the bonded structure will not conduct electricity due to the insulating behavior of the vast majority of current polymer-based adhesives. Therefore, there is a need to consider electrically conductive adhesive bondlines in joining any number of structural materials, even those such as aluminum. The data presented based upon tensile butt joints prepared by adhesively bonding brass plugs show that the bondlines do not conduct without the inclusion of the expanded foil and that inclusion of a metal scrim or expanded foil into the bondline increases the electrical conductivity of the system, even when the adherends themselves are electrically conductive.

(5) Preformed Electrically Conductive Film Adhesives. Fabrication of electrically conductive bondlines require efforts to ensure that the expanded foil is thoroughly impregnated with the adhesive during layup of the structure. Utilizing a film adhesive that has the metal foil or scrim incoφorated into it at the onset addresses this issue. This would greatly simplify the formation of adhesively bonded structures utilizing these materials. Such film adhesives could be formed as below. (I) Pretreat the expanded foil to ensure a clean metal surface. (ii) Impregnate the metal foil or scrim with the adhesive system desired for that application. This could be accomplished via:

(a) solvent impregnation of the adhesive followed by removal of the solvent, or

(b) by heating the metal foil or scrim in the presence of the adhesive at a temperature high enough to lower the viscosity of the resin to the point at which it can impregnate the metal foil or scrim, but low enough to minimize any cure of the adhesive at this point.

(iii) Remove excess adhesive from the surface of the structure to ensure that the metal foil or scrim will contact the adherends during the cure process. (iv) Heat the film adhesive system formed above to partially cure the adhesive system, resulting in a B-stage film. It is undesirable for the resin to flow at ambient conditions and the adhesive film should remain sufficiently flexible to facilitate formation of the bondlines in whatever shape is necessary.

(6) Other Applications of Electrically Conductive Bondlines.

Enhanced electrical conductivity within an adhesive bondline will provide more flexibility in the design of ground planes within the structure, could enhance the ability of the structure to dissipate electric charge, and might improve shielding of electromagnetic interference.

REFERENCES (The teachings disclosed in all references and citations are incoφorated herein by reference.)

Aharoni, S.M. "Electrical Resistivity of a Composite of Conducting Particles in an Insulating Matrix," J. Appl. Phys., 43, 2463 (1972).

ASTM D 1002, Standard Test Method for Strength 1 Properties of Adhesives in Shear by Tension Loading (Metal-to-Metal), ASTM, Philadelphia (1993).

ASTM D 2095, Standard Test Method for Tensile Strength of Adhesives by Means of Bar and Rod Specimens, ASTM, Philadelphia (1993).

ASTM D 2739, Standard Test Method for Volume Resistivity of Conductive Adhesives, ASTM, Philadelphia (1993).

ASTM E 178, Standard Practice for Dealing With Outlying Observations, ASTM, Philadelphia, (1993). Bigg, D.M., Metal-Filled Polymers, Bhattacharya, S.K. Ed., Marcel

Dekker, New York (1986).

Bolger, J.C., "Conductive Adhesive," Handbook of Adhesives, 3rd ed., Skeist, I. ed., Van Nostrand Reinhold, New York (1990).

Borchardt, H.J., Daniels, F., "The Application of Differential Thermal Analysis to the Study of Reaction Kinetics," J. Am. Chem. Soc, 79, 41 (1957).

Broadbent, S.R., Hammersley, J.M., "Percolation Processes: I. Crystals and Mazes," Proc. Cambr. Phil. Soc, 53, 629 (1957).

Constance, J., 'The All-Composite Aircraft," Machine Design, 64(1), 46 (1992). Davis, H.T., Valencourt, L.R., Johnson, C.E., "Transport Process in

Composite Media," J. Amer. Cer. Soc, 58, 446 (1975).

DeMeis, R., "Lightning Protection for Aircraft Composites," Aerospace America, 22, 62 (1984).

Dunn, J., "Flights of Fancy," The Engineer, 274(7102), 26 (1992).

Etemad, S., Quan, X., Sanders, N.A., "Geometry-Defined Electrical Interconnection by a Homogeneous Medium," Appl. Phys. Lett., 48, 607 (1986). Good, R.J., "Intermolecular and Interatomic Forces," Treatise on Adhesion and Adhesives, vol. 1 , Patrick, R.L. Ed., Marcel Dekker, New York (1967).

Gurland, J., "An Estimate of Contact and Continuity of Dispersions in Opaque Samples," Trans. Met. Soc. AIME, 236, 642 (1966).

Huntsberger, J.R., 'The Nature of Adhesion, " Adhesives Age, 13(11), 43 (1970).

Jeanne, M.A., Composites: Engineered Materials Handbook, Vol.1 , ASM International, Metals Park (1987). Kinloch, A.J., Adhesion and Adhesive, Chapman and Hall, London

(1987).

Kirkpatrick, "Percolation and Conduction," Rev. Mod. Phys., 45, 574 (1973).

Kusakawa, K., Japan Pat. 01 271 474 (1989). Landrock, A.H., Adhesive Technology Handbook, Noyes, New Jersey

(1985).

Liaw, H.K., "Study of Electrically Conductive Adhesives for Composite Systems," M.S. Thesis in Chemical Engineering, Kansas State University, 1990. Lide, D.R., Handbook of Chemistry and Physics, 74th ed., CRC, London

(1993).

Ludbrook, B.D., "High Performance Adhesives Meet Many Industry Needs," Adhesive Age, 33, 25 (1990).

Lyons, A.M. "Electrically Conductive Adhesives: Effect of Particle Composition and size Distribution," Polym. Eng. Sci., 32(6), 445 (1991).

Malliaris, A., Turner, D.T., "Influence of Particle Size on the Electrical Resistivity of Compacted Mixtures of Polymeric and Metallic Powders," J. Appl. Phys., 42, 614 (1971).

Matz, K.R. "Electrically Conductive Cement and Brush Shunt Connection Containing the Same," U.S. Pat. 2 849 631 (1958).

Nicodemo, L., Nicolais, L., Romeo, G., Scafora, E., 'Temperature Effect on the Electrical Resistivity of Metal/Polymer Composites," Polym. Eng. Sci., 18, 293 (1978).

Miller, B., "Polymerization Behavior of Silver-Filled Epoxy Resins by Resistivity Measurement," J. Appl. Polym. Sci., 10, 217 (1966). Ogunjimi, A.O., Boyle, O, Whalley, D.C., Williams, D.J., "A Review of the Impact of Conductive Adhesive Technology on Interconnection," J. of Electronics Manufacturing, 2, 109 (1992).

Ott, L., An Introduction to Statistical Methods and Data Analysis, 3rd. ed., PWS-KENT, Boston (1988).

Peck, D.B. "Printed Electrical Resistor," U.S. Pat. 2 866 057 (1958).

Perkin Elmer, 7 Series DSC 7 Kinetics Software Kit, Norwalk, Connecti¬ cut (1987).

Petrie, E.M., "Plastics and Elastomers as Adhesives," Handbook of Plastics and Elastomers, Harper, C.A. Ed., McGraw-Hill, New York (1975).

Petrie, E.M., "Adhesively Bonding Plastics: Meeting an Industry Challenge", Adhesives Age, 32(5), 6 (1989).

Prime, R.B., "Differential Scanning Calorimetry of the Epoxy Cure Reaction," Polym. Eng. Sci., 13(5), 365 (1973). Pujol, J.M., Prud'homme, C, Quenneson, M.E., Cassat, R., "Electro- conductive Adhesives: Comparison of Three Different Polymer Matrices: Epoxy, Polyimide and Silicon," J. Adhesion, 27, 213 (1989).

Quan, X., "Investigation of the Short-Range Coherence Length in Polymer Composites Below The Conductive Percolation Threshold" J. Polym. Sci. , Polym. Phys. Ed. , 25, 1557 ( 1987).

Reinhart, T.J., "Engineering Properties of Adhesives, " Adhesives Age, 16(7), 35 (1973).

Reinhart, T.J., "Structural Bonding Needs for Aerospace Vehicles," Adhesives Age, 32(1), 26 (1989).

Shields, J., Adhesives Handbook, 3rd ed., Butterworth, London (1988).

Snogren, R.C., "Space-Age Bonding Techniques, Part 2: Adhesive Bonding," Mech. Eng., 92(5), 33 (1970).

Tanaka, Y., Bauer, R.S., "Curing Reactions," Epoxy Resins, 2nd ed., May, C.A. Ed., Marcel Dekker, New York (1988).

Triplett, T.E., "Lightning Strike Protection Concepts for Composite Materials," Proceedings of the 34th Intemational SAMPE Symposium, Society for the Advancement of Material and Process Engineering, Covina, 1656 (1989). Wolfson, H; Elliott, "Electrically Conducting Cements Containing Epoxy Resins and Silver," U.S. Pat. 2 774 747 (1956).

Zisman, W.A., "Constitutional Effects on Adhesion and Cohesion," Adhesion and Abhesion, Weiss, P. Ed., Elservier, New York (1962).

Zisman, W.A., "Influence of Constitution on Adhesion," Ind. Eng. Chem., 55(10), 19 (1963).

Zisman, W.A., "Influence of Constitution on Adhesion," In Handbook of Adhesives, 2nd ed., Skeist, I. Ed., Van Nostrand Reinhold, New York (1977).

Claims

Claims:

1. An electrically conductive bondline adhesive comprising: two or more adherends forming a joint therebetween; a quantity of adhesive selected from the group comprising thermosetting and thermoplastic adhesive polymers; continuous metallic structure within said adhesive forming an adhesive matrix therewith, said matrix forming a bondline in the joint having bondline electrical resistivity to less than 0.05 Ω-cm; and said bondline single lap shear strength in excess of 3000 psi.

2. The combination of Claim 1 , wherein said metallic structure is nickel expanded foil.

3. The combination of Claim 1 , wherein said metallic structure is stainless steel expanded foil.

4. The combination of Claim 1 , wherein said adhesive is non- conductive.

5. The combination of Claim 1 , wherein said metallic structure is nickel expanded foil having openings therein in the range of 600-1400/in².

6. The combination of Claim 1 , wherein said adherends are carbon fiber composite materials.

7. The combination of Claim 1 , wherein said adherends are metallic construction materials.

8. The combination of Claim 7, wherein said metallic construction material is aluminum.

9. An electrically conductive bondline comprising: a quantity of adhesive selected from the group comprising thermosetting and thermoplastic adhesive polymers; continuous metallic structure within said adhesive for decreasing a bondline electrical resistivity to less than 0.05 Ω-cm; and said bondline strength in excess of 3000 psi.

10. The combination of Claim 9, wherein said metallic structure is nickel expanded foil.

11. The combination of Claim 9, wherein said metallic structure is stainless steel expanded foil.

12. The combination of Claim 9, wherein said adhesive is non- conductive.

13. The combination of Claim 9, wherein said metallic structure is nickel expanded foil having openings therein in the range of 600-1400/in².

14. In combination with a pair of adherends forming a joint therebe¬ tween, an adhesive matrix within said joint and forming a bondline therein for securing the adherends together, said adhesive matrix comprising: a quantity of non-conductive adhesive selected from the group consisting of thermosetting and thermoplastic adhesive polymers; and a metal foil disposed within said adhesive for making the bondline electrically conductive, said foil having disposed therein at least one opening with said adhesive disposed therethrough to form said bondline.

15. The combination of Claim 14, said bondline having an electrical resistivity less than 0.05 Ω-cm and a bond strength of at least 3000 psi.

16. A method of fabricating a lap shear joint having therein a bondline, comprising the steps of: providing aircraft panel adherends forming a joint therebetween; providing a metal foil cleaned with a solvent; filling said foil with a non-conductive adhesive adapted for securing said adherends together, said foil being filled with adhesive in a temperature range sufficiently high to reduce the viscosity of the adhesive to permit foil filling and sufficiently low to prevent undesired curing of the adhesive; disposing said adhesive filled foil within said joint to form a bondline therein; and allowing said bondline to cure during a cure period and applying compres¬ sive pressure to said bondline during said cure period, said cured bondline having an electrical resistivity less than about 0.05 Ω-cm and a bond strength greater than 3000 psi.

17. The method of Claim 16, said adhesive being Scotch-Weld EC-

3448 Structural Adhesive (1XA), and the step of curing said bondline further comprising the step of maintaining the temperature in the range of 120-150 °C for a period of 60-90 minutes and applying a compressive pressure of about 20 psi to said bondline.

18. A preformed electrically conductive film adhesive comprising: a quantity of non-conductive adhesive selected from the group consisting of thermosetting and thermoplastic adhesive polymers; and a metal foil disposed within said adhesive for making the bondline electrically conductive, said foil having disposed therein at least one opening with said adhesive disposed therethrough to form said bondline.