Bonding of aluminum 1 Sapa Technology (ST) is a research and development center within Sapa. Sapa is an international industrial group. It develops, manufactures and markets aluminum and plastic based products that have a high degree of added value. Aluminum and plastic are both lightweight materials. ST has been assigned the role of supporting and stimulating the technical development of the products and processes of Sapa companies. We are a high level technical resource for the individual companies and, consequently, their customers. ST’s specialists supply expertise on how aluminum’s properties can be influenced by controlling manufacturing conditions and the mix of alloying elements. Our laboratories have advanced measuring equipment and equipment for the technical investigation of materials. We often work closely with universities, university colleges and research institutes. An important part of our activities is the dissemination of knowledge about aluminum’s properties and the use of aluminum. Cover photo: The ¾ tonne “Profilen” (“Profile”), the world’s first sailing boat to be made using bonded 2 aluminum profiles. Developed and built by Sapa Profiler, Vetlanda. Introduction Contents Many designers know that adhesives are used to bond load-bearing structures in aircraft. Double-sided sticky tape is also used to join certain metal components in aircraft. Between these two extremes, there are very many different adhesives. Introduction 3 The bonding of structures – an overview 4 Preconditions for bonding 5 Bonding as a jointing method offers wide possibilities in a very broad area. However, the seemingly simple jointing offered by bonding can be illusory if the joint is not designed with a knowledge of the basic factors that affect the joint’s strength and service life. The effect of the service environment 6 The effect of temperature 7 The effect of joint design 7 We hope that this guide helps to increase knowledge of the possibilities and limitations inherent in the bonding of aluminum. The text is based on studies of relevant literature and on the testing we carried out when producing the first edition, published in July 1984. The effect of mechanical loads 8 The effect of adhesive type 9 The effect of alloying elements and condition 9 This, the third edition, has been updated as regards, amongst other things, contents and adhesive designations. The effect of pretreatment 9 How the bond is made 11 Bonding to materials other than aluminum 11 Choice of adhesive 12 Epoxy adhesives 13 Polyurethane adhesives 14 Hot-melt adhesives 16 Polysulphide rubbers 17 Anaerobic adhesives 18 Test methods 18 References 19 Tests of a few selected adhesives 20 Examples of bonded aluminum products 21 3 The bonding of structures – an overview Perhaps the main difference between bonding, on the one hand, and soldering, brazing and welding, on the other, is that the filler material in bonding is a plastic rather than a metal. For bonding to take place, the adhesive and the material to be bonded must come closer to each other than 0.5 nm (1 nm = one millionth of a millimetre). This is only possible if the adhesive not only wets the appropriate surface(s), but also spreads and “penetrates” so that that the gaps and dips in the surface(s) are filled out. Wetting can only occur if the adhesive has a lower surface tension than the surface that has to be wetted. So that there is sufficient contact between adhesive and material, the adhesive must have good liquid properties. However, for the transmission of loads to be possible, it must set into a load-transmitting material. Broadly speaking, setting can be divided into three types – drying, cooling and polymerisation. Different types of adhesive set in different ways: Adhesive type: Setting via: Solutions of plastics/elastomers Drying (evaporation) Adhesives that, at room temperature or above, can appear elastic or soft can, at lower temperatures, become hard and, perhaps, brittle. Below, an attempt is made to roughly rank adhesives as regards the durability of the finished bonds at various temperatures. Bond durability at various temperatures: A rough ranking of adhesive types, from most to least temperature sensitive: Dispersions of plastics/elastomers Drying (evaporation) Hot-melt adhesives Cooling Prepolymers Polymerisation (curing) Where bonds between metals and reinforced plastics are subject to high loads, only the last of the above types of adhesive (i.e. curing) is suitable. Curing (polymerisation) can be initiated by: • Mixing of, or contact between, two components • Heating (heat-curing) • Illumination with UV or blue light • Environmental changes such as: - presence of moisture - altered pH value - absence of oxygen + metal ion contact When bonded joints are subject to loads, there is always so much energy in the bonds between the adhesive and the bonded matter that the bonds are stronger than the weaker (weakest) of the materials involved. Thus, rather than the bond itself coming apart, any failure would be in the form of a break in the adhesive or the bonded material. The foregoing is conditional on the adhesive and the surface(s) having come into truly good contact with each other. This is not always the case when fast-setting adhesives are used. Furthermore, not all types of bonds are so rich in energy that they can withstand the effect of another medium, e.g. water. Adhesion can then reduce and even become negative. This very much depends on the combination of adhesive, material and surface treatment. All plastics and, therefore, all adhesives are viscoelastic. Thus, from the point of view of loads, they are more open to the influences of temperature and time than are, for example, metals. Consequently, for an adhesive, it is not possible to set fixed values for a large number of strength parameters (e.g. modulus of elasticity, yield point and creep strength). These values differ even with relatively small temperature differences. They also depend on the rate of deformation. 4 • Thermoplastic hot-melt adhesives • Double-sided sticky tapes • Thermoplastic adhesives, drying • Elastomers (rubber, contact adhesives) • Curing elastomers • Curing hot-melt adhesives • “Environment-curing” adhesives • Two-component adhesives that cure at room temperature • The same 2-component adhesives, but heat-curing • One and two-component adhesives that require heat curing • Curing adhesive films, heat-curing, 125 – 175°C The spread within the various groups is wide. It is highly recommended that meticulous attention should be paid to each adhesive’s data sheets. However, these do not always provide the answers and own, supplementary testing may be necessary. The adhesives that cure without any heat input are seldom of practical use at temperatures above 100°C. Silicon adhesives, which can be used up to around 250°C, are an exception. There are extremely few adhesives that are of practical use at temperatures above 300°C. Unfortunately, most heat-resistant adhesives are often relatively hard (silicon once again being an exception). As a result, their ability to spread the stresses presented by peeling and cleaving forces is limited. Bonded joints should be designed so that, as far as possible, they are exposed to pure shearing forces. This is partly because adhesives are viscoelastic (yield/creep characteristics of soft adhesives) and partly because of the notch sensitivity of bonds in hard, heat-resistant adhesives. The durability of a bonded joint very much depends on how well the adhesive fills out the pores and unevennesses in the surfaces of whatever is to be bonded. This is particularly important when bonding metals that are to be used in corrosive environments. Poorly filled surfaces provide space for water, which can cause boundary layer corrosion. Moisture getting into incompletely filled surfaces can, when a bond is subjected to minus temperatures, also lead to frost erosion in the boundary layer. Thus, when using high-viscosity adhesives and curing adhesive films, it may be necessary to first saturate the surface with a low-viscosity primer. The high-viscosity adhesive can then be applied. This can give the bond higher green strength and improved durability. In this connection, it should be pointed out that the results of metal bonding are highly dependent on which metal oxides are on the metal surfaces at bonding and on how securely these oxides sit on the surface. See “The effect of alloying elements and condition”. Preconditions for bonding The term “adhesion” is frequently used in a bonding context. “Adhesive forces” are the attractive forces arising in the interface between two surfaces that are in contact with each other. The attractive forces between an adhesive’s molecules and the surface(s) that are to be bonded have a maximum range of 0.5 mm. To achieve such closeness, the adhesive must have a lower surface tension than the material that is to be bonded. This is so that spontaneous wetting can occur (fig. 1). If the surface tension relationship is inverted, the adhesive tries to pull itself together into a droplet (fig. 2). The strength of the adhesive (“the plastic”) Figure 1: Spontaneous wetting. Adhesive forces between adhesive and aluminum oxide (degree of wetting) Strength of the aluminum alloy The adhesion between oxides and the aluminum substrate The strength of the oxide-hydroxide layer Figure 2: No wetting = No adhesion Figure 3: Possible weakest links in a bonded joint It can also be said that the surface must have sufficient free surface energy to prevent the adhesive pulling itself together into a droplet. When bonding aluminum to another metal, the surface tension relationship is not a problem. This is because all adhesives (plastics) have a lower surface tension than all metals. However, when bonding aluminum to a plastic, this phenomenon should be borne in mind. Nonetheless, in practice, knowing about the surface tensions of the parent materials is not enough. This is because the properties of a surface are often different from those of the parent material. Production methods, surface treatments, handling and storage are just some of the factors that can affect the chemical and structural properties of the surface layer. What we regard as aluminum is, in reality, often aluminum coated with an oxide. The surface of untreated aluminum is often made up of magnesium oxides. In its turn, this surface may be coated with grease, dirt, adsorbed molecules from gases and liquids or the products of chemical reactions between the material and its surroundings. The strength of the bonded joint depends on the weakest link in the chain. This is illustrated schematically in fig. 3. “Relative strengths” immediately after bonding are not stable. The adhesive can be affected by, for example, water absorption, UV irradiation, heat-attributable softening or creep under load. Similarly, the oxide layer on a metal surface can reform in such a way that strength is reduced and/or the volume is changed. This can set up internal stresses in the bonded joint. Achieving a bonded joint that has good long-term strength requires a good knowledge of how the factors set out below affect the joint’s quality. Factors that affect the quality of bonded joints • Environment • Temperature • Mechanical loads • Pretreatment • Alloying elements and condition • Adhesive type • Joint design • How the bond is made 5 A surface that may appear smooth and even (perhaps even polished) is very uneven when viewed at high magnification (fig. 4). A viscid adhesive needs a very long time to take over the space occupied by the air in such a surface. Complete filling never take place. The thicker and more fast-setting an adhesive, the greater the need to first apply a runny primer that saturates the surface before the adhesive is applied. Figure 4 a: An aluminum plate at 500x magnification Figure 4 b: The same plate at 25,000x magnification The effect of the service environment Figure 5: Variously pretreated, unloaded joints kept outdoors in an industrial environment. Alloy 6061-T6. Ref. 3. Figure 6: Unloaded joints kept outdoors in a coastal environment. Alloy 6061-T6. Ref. 4. 21,0 21,0 17,5 17,5 14,0 14,0 10,5 10,5 7,0 7,0 3,5 3,5 Strength in N/mm2 mon and most difficult environmental “stresses” to which an adhesive bond involving a metal can be exposed. The effect on the boundary layer is even more negative if the water contains salts. Normally, water gets into the adhesive-aluminum boundary layer via penetration into the incompletely filled unevennesses in the aluminum surface. This once again highlights the value of pretreating with a runny primer. For aluminum, the long-term strength of a bonded joint exposed to moisture is directly dependent on: how well the adhesive fills out surface unevennesses; and, the strength Strength in N/mm2 Adhesives can, of course, be affected by the service environment. As there are clear indications of the durability of various adhesives in various environments, choosing an appropriate adhesive is no great problem in this particular respect. When bonding aluminum, low-strength bonded joints are often a boundary layer problem, i.e. the consequence of phenomena that have undesired effects on the boundary layer between adhesive and aluminum oxide. Water, in either its liquid or vapour phase, is one of the most com- 12 24 Exposure in months 36 48 60 72 84 96 12 24 36 48 60 72 84 Exposure in months Key to figs. 5 and 6: Degreased in trichloroethylene vapour Blasting with silicon sand + vapour degreasing Chromating Anodizing in sulphuric acid followed by sealing Pickled with chromic/sulphuric acid paste Pickling in phosphoric acid/ethanol Pickling in chromic/sulphuric acid 6 96 and durability of the aluminum oxides that are being bonded to. Through experience, the aircraft industry has long been aware of this. However, it is only in recent decades that scientific experiments have been carried out to provide explanations of the observed phenomena (ref. 8). Bonding to a normally formed aluminum oxide does not give the best long-term strength. If there is water or high air humidity in the service environment, a surface conversion must be carried out to increase the durability of the bonded joint. See also “The effect of pretreatment”. The graphs in figs. 5 and 6 give an idea of the effect that environment and pretreatment have on the strength of a bonded joint. Testing was carried out on single lap joints (as per ASTM D 1002-72) with a 2-component, roomtemperature (RT) curing epoxy adhesive. The effect of temperature The properties of an adhesive are strongly temperature dependent. Adhesives soften as temperature rises and harden as it falls. The most critical changes are: reduced creep resistance at high temperatures when subject to loads; and, increased sensitivity to stress concentrations and shock loads at low temperatures. Different adhesives, even when within the same group, are affected to different degrees by temperature. The graph on the right shows the temperature sensitivity of various epoxy adhesives. It should be borne in mind that these values result from short-term testing, normally as per ASTM D 1002-99, where a single lap joint in 1.55 mm aluminum is subjected to a load at a constant tensile speed of 1.33 mm per minute. (As a rule, equivalent test methods are also used for the strength values given in the data sheets for adhesives.) If it is known that bonded joints are to be exposed to long-term loads in a single direction at elevated temperatures, steps should be taken to ensure that the adhesive’s creep strength is sufficient. This often demands own/inhouse testing. N/mm 2 45 40 1-comp., viscous, heat cured 35 2-comp., viscous, RT cured 30 2-comp., hard, heat cured 25 20 1-comp., hard, heat cured 15 2-comp., hard, RT cured 10 5 2-comp., fast, RT cured -50 0 50 100 150 200 250 °C Figure 7: Strength of several different groups of epoxy adhesives at various temperatures. The blue curves are for the same adhesive. The effect of joint design A bonded joint does not necessarily have to be seen as a film between the joint’s surfaces. Instead, it can be regarded as a large number of tightly packed elements that join the surfaces. P P MV P P Imagining these elements as glass when the adhesive is hard, and as rubber when it is soft, gives an idea of the differences in load distribution. The glass elements (hard adhesive) do not stretch much under load. This means that the outermost row of bars takes up most of the load. The bars continue to do this until they fracture or come away from the substrate. It is only then that the underlying bars take up the load – until they too break, etc. Using the rubber-element model (elastic adhesive), the result is that the outer row stretches under a moderate load without transmitting much of this to the anchorages (the joint surfaces). Rows two, three, etc. also transmit a certain part of the load. This is part of the philosophy behind the advantages of using a more elastic adhesive. However, the increased elasticity is often paid for in poorer heat resistance and reduced resistance to creep. It is also important to bear in mind that an adhesive’s hardness is temperature dependent. An adhesive that can behave like rubber bars at room temperature can become like glass bars in cooler conditions. Figure 8: Deformation of adhesive and aluminum plate when a single lap joint is loaded 7 sive, torsional or tensile, such joints experience them as shear stresses. However, any differences in coefficients of thermal expansion should be taken into account here. A joint’s edges are always exposed to the greatest stress. Thus, especially when using a hard adhesive, it is important that the design spreads any load evenly throughout the bonded joint. Fig. 10 shows various ways of reducing edge effects. 2 1 3 4 5 Figure 10: Lap joints modified to reduce stress at a joint’s edges Figure 9: Stress distribution in a single lap joint – viewed using tension spectrometry The joint designs labelled 1 – 4 make the material more yielding at its edges. The idea is that the material should deform before it transmits large loads to the bonded joint. Joint number 5 offers a further possible solution, i.e. making the bonded joint considerably thicker at its edges. This gives a bonded joint with a larger “deformation zone”. In bonding, joint modification is of greatest benefit with rigid adhesives. Especially when using hard adhesives, the design of a bonded joint is of great significance. The basic rule is that a bonded joint must be designed so that the loads it is exposed to are transmitted as shear forces. Cylindrical bonded joints (pins in holes, pipes in pipes, etc.) come the closest possible to the ideal. Whether loads are compres- It is advisable to try and form a picture of the force flows throughout a bonded joint and, as rule of thumb, not to use a harder adhesive than necessary. The effect of mechanical loads 8 36 Unloaded 30 Residual strength in N/mm2 Bonded joints are normally regarded as rather insensitive to vibration and fatigue at high frequencies. They are often used as vibration dampers and crack barriers. Nonetheless, mechanical loads can exacerbate boundary layer problems. The simultaneous effects (synergy) of temperature, environment and mechanical load result in significantly faster reduction in strength than would occur if these three stresses operated individually and had their outcomes added together. The stress concentrations that can arise when a structure is loaded manifest themselves at the edges of bonded joints (especially if the joint has not been designed to minimise such concentrations), where environmental impact is also greatest. This can result in more rapid aging of the bonded joint than would otherwise have been the case. Fig. 11 compares the reduction in the residual strength of loaded and unloaded joints in 100% humidity at 60°C. Pretreatment was chromic/sulphuric acid pickling. The choice of pretreatment has a large effect on results. See also “The effect of pretreatment”. 24 18 6,2 N/mm 2 8,3 N/mm 2 12 6 10 20 30 40 50 Exposure in days Figure 11: Differences in residual strength of differently loaded joints in moist heat (ref. 1) The effect of adhesive type To correctly understand the effect that different adhesives have on a bonded joint, the bonded joint should be viewed as an independent structural element in a composite structure. Naturally enough, this structural element can have various mechanical properties and be affected differently by temperature, environment and chemicals. Through the electrolytes that can form in the presence of water, different adhesives can also differently affect the aluminum surface in the boundary layer. “Choice of adhesive” and “Tests of a few selected adhesives” (table 5, page 18) present a selection of adhesives that are used for bonding aluminum. The effect of alloying elements and condition In certain conditions, alloys that contain magnesium or copper are more difficult to bond (to). High concentrations of these elements in the adhesive/aluminum oxide boundary layer form oxide types that are of lower strength and are more sensitive than aluminum oxide. However, the boundary layer’s composition is highly dependent on pretreatment and can be widely affected in a number of ways. Heat treatment of magnesium-containing alloys increases both the oxide layer’s thickness and its magnesium content. By contrast, anodizing in phosphoric acid (for example) gives a thin and magnesium-poor boundary layer with very good durability. It has also been shown that, as with magnesium, the presence of copper in the boundary layer can impair the durability of the bonded joint. In the chromic/sulphuric acid pickling of copper-containing alloys, the copper can, if pickling is taken too far, be enriched in the oxide layer (ref. 2). Table 1 gives an idea of how various surface treatment methods effect magnesium content in the oxide and the durability of the bonded joint. See also “The effect of pretreatment”. Surface treatment Al:Mg ratio Joint’s long-term strength Degreasing 6:1 Very poor Sand blasting 15:1 Moderate Chromic acid pickling 57:1 Good Phosphoric acid anodized 110:1 Very Good * The alloy’s Al:Mg ratio was 53:1. Table 1: Magnesium content in the surface in relationship to longterm strength (ref. 1) The effect of pretreatment Pretreatment can have several purposes: • To achieve a clean surface that can be bonded to. • To create a surface that has better durability in corrosive environments. • To provide a material with a strong oxide that can take high loads. • To give a surface a decorative appearance. (These four purposes can sometimes conflict with each other.) As previously intimated, aluminum in its as-delivered condition provides a comparatively poor surface to bond to. The composition of the oxide layer on the surface varies and the oxide layer has relatively poor adhesion to the parent metal. If loads are moderate and the joint is not exposed to moisture, a pretreatment that gives a clean surface free from dirt may be sufficient. Examples of such pretreatments are washing with water or, previously, trichloroethylene vapour. Bonded joints for harsh service environments require other pretreatments. Some methods that are used with various degrees of success are given below. Strong, alkaline washing solutions give a surface that is undoubtedly clean, but the oxide layer of which has been converted to brittle aluminum hydroxide. Mechanical pretreatments such as blasting or lapping with a compliant abrasive give a strong surface. However it has no corrosion protection. Dipping for around 10 minutes in water at 95 – 100°C after pickling in, for example, caustic soda gives improved corrosion resistance in the adhesive/aluminum boundary layer. Yellow or green chromating is a common treatment before painting. It gives good corrosion protection, but load bearing properties are relatively poor. When bonding, a chromate layer should be thin. This method is thus best for moderate loads and elastic adhesives. Pickling in chromic/sulphuric acid gives a very good surface for bonding. This method has long been used by aircraft manufacturers in the USA. Anodizing has proven to be the best pretreatment if durability of loaded joints in corrosive environments is required. Thick anodic oxide layers give excellent corrosion resistance, but are brittle and thus have poor load-bearing capabilities. Anodizing in sulphuric acid is common and gives the thickest oxide layers (see table 2). Anodizing in chromic acid without subsequent sealing gives a thin oxide layer that has good load-bearing capabilities. This method has long been used by the European aircraft industry. From the strength point of view, the method is excellent. However, the process requires precise control. 9 Thickness, approx. Chromic/sulphuric acid pickling 40 nm Phosphoric acid anodizing 400 nm Chromic acid anodizing 1,500 nm Sulphuric acid anodizing 15,000 nm Table 2: Layer thickness with various surface treatments for aluminum ≈ 10 nm Fig. 13: Service lives of joints with various surface treatments and loads (alloy 6061-T6, 1-component nitrile modified epoxy, cured 15 minutes at 204°C) 16,8 14,0 11,2 8,4 5,6 2,8 Applied tensile stress in N/mm2 Surface treatment (7 dagar) Days) (7 100 nm 40 nm 7 nm Oxide Figure 12 b: Isometric sketch of the oxide layer of a phosphoric acid anodized surface This surface is easily destroyed by mechanical contact and must be directly coated with a primer or a runny adhesive. Fig. 13 gives an idea of long-term strength when loaded in moist heat. (69 Days) 4 3 (694 Days) 10 10 10 10 6 10 7 Time (in minutes) until fracture Key: Degreased in trichloroethylene vapour Anodized in sulphuric acid, sealed Anodized in sulphuric acid, not sealed Pickled in chromic/ sulphuric acid Anodized in chromic acid Anodized in phosphoric acid (ref. 7). 10 100 nm 2 5 Table 3 gives some useful guidelines for selecting a pretreatment (in relation to what is required of the bonded joint). The properties of a pretreated surface are rapidly impaired if it is exposed to moist air or greasy, airborne contaminants. Similarly, it is sensitive to, for example, fingerprints during handling. It is often appropriate to apply an adhesive primer directly after pretreatment. The surface can then be stored a long time without impairment. The use of primer often brings increased green strength and better corrosion protection/long-term strength. Adhesive primers frequently have corrosion-inhibiting additives that further improve the bond’s chances of providing long-term strength in difficult environments (ref 5). “Wash-primers” are general primers that give good results. They contain polyvinyl butyral with a phosphoric acid hardener (curing agent). This type of primer can, for example, be advantageously applied before bonding with a wide range of polyurethane adhesives. There are special primers for many special adhesives, e.g. those used in the aircraft industry (ref. 8). Table 3: Useful guidelines for selecting a pretreatment (in relation to what is required of the bonded joint) Surface quality Unloaded or mildly Loaded Joint loaded joint Environment Environment Dry Water Brine Dry Water Brine Untreated (as delivered) (+) Vapour degreased + (+) + Mechanically processed +++ + +++ + Dipped in hot water +++ ++ ++ ++ Primed with polyvinyl butyral ++ ++ + ++ (+)+ + Chromic acid pickled +++ +++ + +++ +++ + Chromated ++ + ++ + + Sulphuric acid anodized no subsequent sealing ++(+) ++(+) ++(?) ++ ++ ++ (+) Sulphuric acid anodized ++(+) ++(+) ++(+) + subsequent sealing + + Chromic acid anodized +++ +++ +++ +++ +++ (+)++ Phosphoric acid anodized +++ +++ +++ +++ +++ +++ The number of “+”s indicates suitability. (+) is a doubtful + (?) is an unverified assumption 10 (+) How the bond is made The strength of an adhesive bond is very much determined by the way in which it is made. At a first glance, it may seem that this is largely a question of applying adhesive to the parts and holding them together until the bond sets. “Can there really be many other ways of doing it?” If the process of making a bond is divided into subprocesses, it can be seen that there are rather a lot of points where each subprocess can have an effect on the end results. Such a division could be as set out below. All these The parts that are to be bon- Intermediate storage ded The adhesive: • Conditioning • Receiving inspection – • Dimension checking storage of adhesives The surfaces that are to be • Stirring of the components • De-aeration of the bonded components • Cleaning • Dosing of adhesive and • Drying hardener • Surface conversion • Mixing • Priming • Drying/curing of the primer • De-aeration of the mixture subprocesses could be involved in a single bonded joint! To ensure consistently good bonding results, it is clearly important that descriptions are drawn up detailing how a bonded joint is to be made. Each subprocess can be further divided into smaller steps. Carrying out the subprocesses does, of course, require knowledge. Each subprocess also presents opportunities for going wrong. Carrying out series bonding without detailed work descriptions can hardly be appropriate. Making the bond: • Controlling joint thickness • Application • Waiting time before putting the components together • Time during which the adhesive mixture can be used • Assembly • Fixing – pressing • Curing • Cooling after heat curing • Removal from jig/press Subsequent storage/curing Testing – Checks – Logging Further processing Packing Delivery Bonding to aluminum of materials other than aluminum When bonding metals, the long-term strength of the bond is often regarded as a metal problem. It could thus be assumed that the bonding of other materials would have fewer problems associated with it. Tests have shown that this is indeed the case as regards long-term strength in moist environments. Difficulties in bonding plastics are most often associated with finding an adhesive with a surface tension low enough for it to wet the plastic that is to be bonded. Many plastics can be treated to increase their surface tension and thus improve bonding possibilities. Below, there are some comments on the bonding of various plastics. • As a rule, PVC, polycarbonates or thermosetting resins present no bonding problems. • Polyamides (nylon), acrylic (Plexiglas) and ABS plastics limit the number of adhesives that can be used. • Because they have very low surface tension, olefinic plastics (polyethene and polypropene) and acetal plastics (e.g. Delrin) are very difficult to bond. Surface tension can be increased by plasma, flame or arc (e.g. corona) treatments. Using a cyanoacrylate adhesive in combination with a primer for plastics that have a low surface tension can result in successful bonds. There is now also a 2-component acrylic adhesive that provides good adhesion to polyolefines. • Polytetrafluorethylene plastic (e.g. Teflon) has the lowest surface tension. To “alloy” (“graft”) the Teflon surface with sodium, thereby increasing surface tension and improving bondability, the plastic can be treated in a liquid that contains metallic sodium. • From the point of view of bonding, painted aluminum surfaces are to be regarded as plastic surfaces. • Glass, ceramics and wood are regarded as easily bonded. The same applies to many other metals than aluminum. The presentation of various adhesive types in “Choice of adhesive” takes up, amongst other things, each adhesive’s suitability for various materials. Solids Surface tension (10-1Pa) Liquids 5 000 Diamonds Glass Platinum Steel 1 000 Aluminum Tin Lead Sodium 100 90 80 Ice Water 70 Glycerol Wood 60 Polyester (unsaturated) 50 Nylon Epoxy PVC Polystyrene Polyester (linear) Acetal (Delrin) Polypropylene Polyethene 40 Epoxy adhesives Polyurethane adhesives Cyanoacrylate adhesives 30 Aromatic hydrocarbons Petrol 20 Ethyl alcohols Teflon 10 Ethyl ethers Figure 14: Surface tension of several materials – for wetting to occur, the adhesive must be below (in the chart) the material that is to be bonded (ref. 7) 11 Choice of adhesive The choice of adhesive is principally determined by three different considerations: • The adhesive must wet the materials that are to be bonded. • In its final state, the adhesive must have material properties that, in the intended service environment, are sufficient to transmit the loads in question. • It must be possible to use the adhesive in production of the product. There is no great difficulty in finding some 1,000 different adhesives on the market. The difficulty is largely in obtaining sufficient knowledge about the adhesive (especially in its final, set state). One way of gaining an overview of the area may be to consider the ways in which adhesives set. After all, adhesives do have to be liquid at some point in the bonding process. Essentially, there are only three ways of setting. However, combinations of these do occur. Setting through drying The solvent or water evaporates. Only 20 to 50% of the original adhesive remains (= shrinkage). Most of the drying must take place through the material. Consequently, this adhesive type is not suitable for bonding aluminum to aluminum. The binder is based on thermoplasts or elastomers (rubber). One way of using a drying adhesive is to employ it as a contact adhesive. Most of the solvent must here be allowed to disappear before the parts are brought together. However, significant quantities are left in the joint after such compression and, in principle, the problem remains. Drying adhesives are not excluded from use in the bonding of thin, compliant and porous materials (mats, etc.) to aluminum. However, they are not discussed any further here. Setting through cooling Some of the drying adhesives can be heat activated. The adhesive is applied to one or both joint surfaces and dried completely. At jointing, the adhesive is activated (melted) on one of the parts and quickly joined with the other. This adhesive type facilitates rapid production, but does not give filled joints. It is only suitable if one of the materials is readily deformable (thin, soft). This bonding method is also discussed no further in this text (however, see example 7 on page 24). Hot-melt adhesives are applied hot, usually to roomtemperature surfaces. They give filled joints and rapid fixing. This type of adhesive is discussed more fully on page 14. Setting through polymerisation (curing) Curing can take place through: • Mixing of two components (adhesives 1, 2, 4, 5 and 8 in tables 4 and 5 – table 5 is on page 18). Table 4: Overview of various curing adhesives Adhesive type (tested adhesive no.) Strength of single lap joint (N/m2) Peel strength (N)m* Handling/ curing Suitable for large surfaces Suitable for thermoplasts Suitable for thermosetting resins Epoxy 16 - 29 4 - 6 2-comp. (1, 2) Dosing + + + – – – + + Mixing 20 - 150°C Epoxy 1-comp. (3) approx. 35 approx. 9 Cures in heat >90°C Polyurethane 6 - 20 2 - 6 Dosing 2-comp. (4) (T-peel test) Mixing 20°C Polyurethane approx. 2 approx. 7 elastomer, 1-comp (5) Polysulphide approx. 2 approx. 5 rubber (7) + + – – –– + + + + – – + + +++ Cured by + – + + + + + atmospheric moisture Hot melt approx. 7 approx. 10 Cooling - curing (6) cured + + If reactivated Dosing + + + Mixing – – +++ (–) + + + + ”Benign” smell Misc. Many different adhesives Other curing methods exist +++ Slow curing Acrylic 20 - 25 2,5 - 3,5 Adhesive + + adhesive (T-peel test) & hardener (SGA) (8) on own surface –– + + + + + Strong smell Short fixing time Aerobic approx. 20 2,5 - 3,5 Adhesive + + SGA (T-peel test) & hardener adhesive (9) on own surface –– + + + Strong smell Short fixing time Anaerobic 12.5 - 20 adhesive (10) –­– 0,5 - 1,5 Cures by (T-peel test) itself + * Roller drum peel tests give higher values than T-peel tests. Strength values are taken from the manufacturers’ data sheets. 12 Work environment aspect – • Contact between two components, each of which has been spread on its particular surface (adhesives 9 and 10 in tables 4 and 5). • Temperature increase to at least 100°C. • Environmental change for the adhesive, e.g. - moisture contact (adhesive 6 in tables 4 and 5). - absence of oxygen + metal ion contact (adhesive 10 in tables 4 and 5). - UV irradiation. From the large number of available curing adhesives, we selected 10 for testing. These are detailed hereafter. Three of them were epoxy adhesives. The remaining seven were each typical of their group. All the adhesives were fairly common and readily available. Table 4 gives an overview of the various adhesives. Table 5 on page 18 gives the results of the aging tests that we carried out. Clearly enough, adhesives other than those discussed here may be available and may be the best choice for a certain design. In addition, the tested adhesives have also been further developed and, in some cases, replaced by others. Thus, the original test results are here given with any replacement products stated. regards the relationship between strength and test method, see “Test methods”. Various adhesive types The presentation of various adhesive types takes up, amongst other things, each adhesive’s suitability for various materials. When bonding dissimilar materials to each other, it should be borne in mind that movement caused by moisture or temperature will not be the same for each material. Bondable materials: Aluminum and other metals and minerals. Thermosetting resins such as Bakelite and polyester laminates can also be bonded with good results. Nylon gives fair results. The adhesive is not otherwise especially suitable for thermoplasts. This is because the latter normally have a lower surface tension than does the adhesive. EPOXY ADHESIVES Araldit AV 106/HV 953 U, “Araldit Standard” (from CibaGeigy), is probably the most widely used metal adhesive there is. It is a comparatively old adhesive. We also included a newer all-round epoxy adhesive (AV 144/HV 997 from Ciba-Geigy) in the tests. Both these adhesives have relatively low heat resistance. Two-component epoxy adhesives that cure at room temperature can rarely be loaded at temperatures above 80°C. Epoxy adhesives with considerably higher heat resistance are available. However, these require heat curing. Unfortunately, the price paid for higher heat resistance is usually lower peel strength. As an alternative to 2-component epoxy, we chose a heat-curing 1-component adhesive, EC 2214 from 3M. This is an old, well known adhesive. The disadvantages of heat curing must be weighed against the advantages of not having to “dose” and mix adhesives. Furthermore, as regards resistance to heat and environmental stresses, the resultant bond is also stronger. Heat curing gives a fully cured joint, something that can never be achieved with a 2-component epoxy adhesive that has to cure at room temperature. One-component epoxy adhesives are available as pastes and, in various thicknesses, films. The main difference between film and paste adhesives is the ways in which they can be applied. Using epoxy adhesives, it is possible to fill joints that have large gaps. Information about the various adhesives has largely been taken from the manufacturers’ data sheets. As Tested adhesive 1 EPOXY ADHESIVE ARALDIT AV 106, HARDENER HV 953 U Components: 2 Mixing ratios: 100:80 (weight), 100:100 (volume). The mixing ratios can vary within comparatively wide limits. Adding more hardener (curing agent) gives a softer adhesive. However, it should be borne in mind that, besides serving as a softener, the surplus hardener will remain in a liquid “unreacted” form in the adhesive. The adhesive’s comparatively poor heat resistance then becomes even poorer. The components should be mixed by weight. Function: The reaction starts when the components are mixed. Curing rate is not affected by the quantity of hardener. Owing to the heat generated during the reaction, large mixes cure more quickly. Curing times: 25°C – 18 hours, 70°C – 50 minutes, 150°C – 5 minutes Suitable areas of use: In the engineering industry, the adhesive is regarded as an all-rounder. It gives filled joints and, where strength requirements at elevated temperatures are small or moderate, can be used for bonds to the above cited materials. Previously, the adhesive was used for sandwich structures. It has now, practically speaking, been completely replaced by polyurethane adhesives. Strength at room temperature Single lap joint, tensile speed 10 mm/min: After curing at 25°C for 18 hours – 16 N/mm2 After curing at 70°C for 50 minutes – 22 N/mm2 After curing at 150°C for 5 minutes – 29 N/mm2 Peel strength: Roller drum peel test: 4 – 6 N/mm2 Temperature range -60° to +60°C. At 60°C, the adhesive has lost around 50% of its strength in respect of short-term loads. Creep strength at this temperature is low. Durability/resistance: Good against petrol and mineral oil (SAE HD 30); otherwise moderate to poor. (Refer also to the data sheets.) Work environment: Must be handled in such a way that the adhesive does not come into contact with the skin. Ventilation must be provided where handling is continuous. In Sweden, the use of epoxy adhesives is regulated by AFS 1996:4, Härdplaster (ordinance 1996:4 of the Swedish Work Environment Authority, Thermosetting resins). Alternative suppliers See under tested adhesive 2. 13 Tested adhesive 2 EPOXY ADHESIVE ARALDIT AV 144, HARDENER HV 997 Components: 2 Mixing ratios 100:40 by weight. The mixing ratio should not be varied. This adhesive is now available in packs that supply readymixed adhesive direct from tubes or cartridges. Function Curing starts when the components are mixed. Pot life of a 100 gram batch at 23°C is 50 – 70 minutes. Curing times 10°C – 24 hours, 20°C – 8 hours, 60°C – 45 minutes, 120°C – 5 minutes Remark: Curing times are slightly shorter than for AV 106. Maintaining the same curing time, the curing temperature can be kept lower. Bondable materials The same as for AV 106 Suitable areas of use The adhesive should be regarded as a development of AV 106. All its properties are better than those of AV 106. The prime area of use is all-round bonding with metal as one or both of the materials. The adhesive is grey. Strength at room temperature Single lap joint, AlMgSi alloy, ground surface: After curing at 20°C for 24 hours – 23 to 24 N/mm2 After curing at 60°C for 30 minutes – 21 to 24 N/mm2 After curing at 120°C for 5 minutes – 26 to 28 N/mm2 Peel strength Roller drum peel test: 6.0 – 6.5 N/mm2 Temperature range -60° to +80°C, with approx. 50% of short-term strength maintained at 80°C Durability/resistance Considerably better than AV 106 throughout (refer also to the data sheets). Work environment The same as for AV 106 Leverantörer av 2-komponent epoxilim: Tested adhesive 3 EPOXY ADHESIVE EC 2214, 3M SWEDEN Components: 1 Mixing ratios The adhesive is always ready for use. This adhesive type has a more restricted storage time than do 2-component adhesives. At no more than 4°C, the storage time is 8 months. 14 Function Curing accelerates with temperature increases (it takes place slowly even at room temperature). The lowest curing temperature is 95°C. Curing times 95°C – 120 minutes, 120°C – 40 minutes (= best results), 204°C – 2 minutes Bondable materials The same as other epoxy adhesives, i.e. metals, minerals and thermosetting resins. It should also be borne in mind that the materials must tolerate the curing temperature. When bonding materials with different thermal expansion coefficients, internal stresses may arise or, on cooling, warping may occur. Suitable areas of use Primarily for bonding metals and other materials that have equal thermal expansion coefficients. Uses include the bonding of body plates in the automotive industry and bonding cemented carbide tools in steel holders. Strength Single lap joint, Al/Al, optimal curing: -40°C – 21 N/mm2 24°C – 35 N/mm2 82°C – 35 N/mm2 121°C – 14 N/mm2 177°C – 3.5 N/mm2 The values are short-term at a tensile speed of 10 mm/ min. Peel strength T-peel test with 0.9 mm steel sheet: approx. 9.5 N/mm2 at 24°C Temperature range -55° to +95°C Durability/resistance The same as that of the better of the 2-component epoxy adhesives Work environment As there is no dosing and mixing of components with 1-component epoxies (the constituents of which also present a lower vapour pressure), the health hazards are considered to be less. However, it must be ensured that there is no contact with the skin. At curing, the air in the curing zone must be evacuated, i.e. there should be underpressure in the curing oven. Temperature rises increase the vapour pressure exerted by the adhesive’s constituents. This also increases the health risks. See also AV 106. POLYURETHANE ADHESIVES Initially largely for work environment reasons, polyurethane adhesives were regarded as possible replacements for epoxy adhesives. However, the risks of isocyanates (the hardeners for polyurethane adhesives) have meant that they cannot be seen as a safer option than epoxies. When using epoxy, skin allergies are the greatest work environment risk. For polyurethane adhesives, breathing difficulties and symptoms of asthma dominate the risks. In the solvent-free 2-component adhesives now on the market, the hardener is, as a rule, based on methylene diisocyanate (MDI). This has a considerably lower vapour pressure than toluene diisocyanate (TDI) types, and is thus less hazardous to work with. The occupational exposure limit for MDI and TDI is set as low as 0.005 ppm. Tested adhesive 4 POLYURETHANE ADHESIVE CASCOBOND 1852, HARDENER 1853, CASCO NOBEL In practice, this adhesive has been replaced by 1849 PUR 2K and hardener 1821. This combination gives longer assembly times and shorter press times. Components: 2 In Sweden, the use of polyurethane adhesives is regulated by AFS 1996:4, Härdplaster (ordinance 1996:4 of the Swedish Work Environment Authority, Thermosetting resins). Polyurethane adhesives’ hardeners (isocyanates) react readily with water. Carbonic acid is formed in this reaction. This means that 2-component polyurethane adhesives always become fully cured if moisture is present. Surplus hardener that reacts with moisture makes the end product harder. However, this final curing takes a long time. When bonding diffusion-resistant materials and large surfaces (1 dm2 and above), it is not certain that this final curing will take place. Thus, even for polyurethane adhesives, it is import to dose and mix the components carefully. The hardener’s moisture sensitivity means that many polyurethane adhesives have a slight foaming tendency when mixed by hand. As there is normally some moisture in the air in/on metal surfaces, such foaming can even occur in the boundary layer. In more recent formulations of 2-component polyurethane adhesives, this foaming tendency has been eliminated. The possibility of reacting with moisture has enabled the formulation of 1-component polyurethane adhesives (e.g. adhesive 6) that cure in moist environments. Here, the moisture is the curing agent (hardener). Curing temperature has less effect on the curing times of polyurethane adhesives than it does on those of epoxy adhesives. Curing temperatures as high as those for epoxy adhesives cannot be used – 70°C is a practical upper limit. There are polyurethane adhesives that have very short curing times even at room temperature. However, once mixed, the pot life of such an adhesive is also very short. Polyurethane adhesives have their widest use in the bonding of sandwich panels. As a rule, polyurethane adhesives are considerably cheaper than epoxy adhesives. As regards use with plastics, polyurethane adhesives are, broadly speaking, better than epoxy adhesives. This is because of the former’s lower surface tension. They are often used in various combinations that include plastics. Using polyurethane adhesives, it is possible to fill joints that have large gaps. Before bonding with a 2-component polyurethane adhesive, the application of a wash-primer (e.g. polyvinyl butyral + a phosphoric acid hardener) gives better filled surfaces and, consequently, improved long-term strength in moist environments. Mixing ratios 100:20 by weight (1849/1821, 100:22). The mixing ratio must not be varied. Function Curing starts when the components are mixed. Mixing is not spontaneous and great care must be taken when dosing the hardener into the adhesive. This adhesive has constituents that have to take care of moisture before the hardener reacts with any moisture. Normally, foaming as a result of carbonic acid formation does not have to be feared. During storage, the adhesive component can sediment in its packaging. Thus, for fully satisfactory results, stirring is very important. The pot life of a 500 gram batch at 20°C is around 1 hour (approx. 25 minutes for 1849/1821). Curing times At 20°C, fixing times of approx. 12 – 16 hours are usually sufficient (around 3 – 4 hours for 1849/1821). It should be noted that, at room temperature, full curing usually takes several weeks and requires the presence of moisture. Bondable materials Hard PVC, polycarbonates and acrylic plastics are amongst the thermoplasts that can be successfully bonded using polyurethane adhesives. In sandwich structures, bonding is between different cellular plastic cores. This adhesive adheres well to metals (aluminum included therein). Applies also to 1849/1821. Suitable areas of use In various sandwich structures and where a filled joint is desired, often where plastics are involved. Polyurethane adhesives can also be used for bonding corner joints in frameworks made from profiles. They are also often used when aluminum is to be bonded to other materials, e.g. plastics. Strength Single lap joint at 20°C, tensile speed 10 mm/min: Approx. 14 N/mm2 when bonding to aluminum Peel strength T-peel test: Approx. 4.3 N/mm2 Temperature range -60°C to approx + 80°C Durability/resistance Good against water, oil and certain solvents (refer also to supplier’s data sheets). 15 Tested adhesive 5 POLYURETHANE ELASTOMER SIKAFLEX 221 SIKA SVERIGE Components: 1 Function The adhesive cures on contact with moisture. Curing times At 20°C and 65 % RH, 3 mm/24 hours Bondable materials Aluminum, epoxies, polyester, acrylics, polyamides, hard PVC, minerals, etc. Suitable areas of use This is a very soft and elastic product. Less soft variants of the product are available. For joints with little surface expansion and moderate requirements as regards load transmission, products of this type can be a good choice, especially where material movements are significant and/ or there are high requirements in respect of impact and vibration resistance. Because they exclude moisture, large diffusion-resistant surfaces cannot be bonded. Strength Tensile strength is 1.4 N/mm2 Temperature range -30° to +70°C Durability/resistance Good against water; temporary resistance to oils and grease. Work environment In Sweden, use is regulated by AFS 1996:4, Härdplaster (ordinance 1996:4 of the Swedish Work Environment Authority, Thermosetting resins). Remark Sika has products that give similar final properties and can be heat cured (lowest at about 70°C). There are also products that can be cured both with heat and moisture as well as 2-component products that cure after mixing. HOT-MELT ADHESIVES For the most part, hot-melt adhesives (“hot melts”) are various mixes of thermoplasts. Characteristically, hot melts must be made runny at application. Many hot melts can be kept molten for several hours at temperatures of between 150 and 250°C. These adhesive are often ethyl-vinyl-acetate mixes. Despite the rather high application temperatures, the heat resistance of these hot melts is low. It is unrealistic to count on such adhesives being able to transmit loads, even for moderate periods, if the temperature rises above 50°C. There are also hot melts that are based on polyamides or thermoplastic polyester. These thermoplasts have higher service temperatures. However, strength when subjected to long-term loads is already very low at 70 – 80°C. 16 The application temperature for this adhesive type is high (approx. 250°C). At this temperature, the adhesive breaks down if it comes into contact with oxygen. Consequently, equipment for these adhesives is often provided with a shielding gas. The great advantage of hot-melt adhesives is the speed with which they form bonds. However, this speed has a drawback. When the hot adhesive meets a surface at room temperature, setting is often so rapid that the adhesive does not wet the surface (cf. “dry joints”). This disadvantage increases: the greater the difference between the temperature of the surface and that of the adhesive; and, the greater the thermal conductivity of the material that is to be bonded. For these reasons, metals are often heated before bonding. Curing hot-melt adhesives are now available. These adhesives are based on polyurethanes that cure on contact with moisture. The adhesive has a solid form even before curing. The melt and, consequently, the application temperature is considerably lower than for the thermoplastic hot-melt adhesives. There are also curing hot melts that, after application and cooling, have a tape-like character for around one hour. Thus, if a relatively high (momentary) press force can be applied, parts at room temperature can be assembled. After setting, the adhesive has moderate strength. On contact with moisture, curing is to a polyurethane with comparatively good strength properties. For full curing to occur, the diffusion paths for moisture must not be too long. Tested adhesive 6 CURING HOT-MELT ADHESIVE SUPERGRIP 9802, BOSTIK Components: 1 Function The adhesive works like a hot melt, i.e. it is applied in its molten state and sets on cooling. The main difference compared with other hot melts is that, after setting, it also cures. The application temperature is relatively low, 100°C (as opposed to 200 – 250°C for ordinary hot melts). As a result, assembly times (i.e. the time before the adhesive sets) are longer. Curing occurs on contact with moisture in the air. Curing times The purely physical setting (cooling) depends on the thermal conductivity of the bonded materials. On aluminum, the setting time is about 10 seconds. To prolong the assembly time when bonding aluminum to materials that have lower thermal conductivity, the adhesive should be applied to the latter. Curing depends on the presence of moisture. If the relative air humidity is not too low and the diffusion paths not too long, curing usually takes place within 12 hours. As with other moisture-curing adhesives, curing times are long when bonding large diffusion-resistant surfaces. Bondable materials Aluminum and other metals, glass, thermosetting resins, PVC, polycarbonates, acrylics, woods, many rubbers, soft PVC, etc. Suitable areas of use Those where, for technical production reasons, hot melts were the first choice but, because of their low heat resistance, could not previously be used. The adhesive requires equipment that prevents it coming into contact with moisture. For test bonds, samples in metal tubes that can easily be heated in an oven are often available. Strength Single lap joint at 20°C: Al/Al, 7.4 N/mm2; PVC/PVC, 6 – 8 N/mm2; polycarbonates, 9.4 N/mm2 Where a soft, rubber-like joint is required. As it does not depend on moisture for curing, this product enables the bonding of large surfaces. Strength Tensile strength, approx. 0.9 N/mm2; elongation at rupture, approx. 100 %; hardness, approx. 55” Shore A Temperature range -30° to +70°C Durability/resistance Excellent against water, moderate against oils Work environment Skin contact must be avoided. SGA (SECOND GENERATION ACRYLIC) ADHESIVES Peel strength Ground rubber, 6 N/mm2 Halogenated rubber, 13 N/mm2 Soft PVC, 10 N/mm2 Temperature range At 70°C, about 50% of the strength at room temperature. At 100°C, approx. 2 N/mm2. For structures that are constantly loaded at elevated temperatures, the creep strength should be investigated. Durability/resistance When bonding anodized aluminum, water resistance is good. Resistance to other media – contact Bostik. Work environment The adhesive contains isocyanates. In Sweden, work is regulated by AFS 1996:4, Härdplaster (ordinance 1996:4 of the Swedish Work Environment Authority, Thermosetting resins). POLYSULPHIDE RUBBERS Besides polyurethane, polysulphide rubber is one of the few rubber materials that can be made to cure at room temperature and, at the same time, adhere to other materials. Its widest use is in the manufacture of insulating glazing where bonding is to aluminum profiles and as a “joint compound” in cladding elements (e.g. aluminum sections in buildings). The material has low strength, but large elongation at rupture. Tested adhesive 7 POLYSULPHIDE RUBBER NAFTOTHERM M82 METALLGESELLSCHAFT/YTTEKNIK Components: 2 Mixing ratios 100:10 by weight. Pot life is approx. 1 – 2 hours. Function Curing starts when the components are mixed. Curing times Around 8 – 12 hours at room temperature Bondable materials Aluminum and other metals, minerals, thermosetting resins This adhesive type is characterised by the hardener being spread on one surface and the adhesive on the other. Curing to a handleable state occurs within one or several minutes. This adhesive type is also very tolerant of oil on surfaces. Adhesive and hardener must be spread on their respective surfaces in such a way that they almost completely overlap each other when brought together. Uncured adhesive will otherwise be left at the joint’s edges. The adhesive is best suited for small to moderately large surfaces. This adhesive type is also available in 2-component packs with a mixer tube. The adhesive emerges mixed and air-free directly from the packs. Curing times are then usually set to be a little longer. As regards work environment, this adhesive type requires contactless handling and good ventilation. The components have a strong smell. Tested adhesive 8 SGA ADHESIVE MULTIBOND 330, LOCTITE Components Two, which are spread separately, i.e. adhesive and hardener on their own surfaces. Function There is a very fast reaction when the adhesive and hardener come into contact. The press force needs to be maintained between 1 and 3 minutes. Curing times Fifty percent strength is reached in around 20 – 30 minutes. Full strength is achieved after 3 – 6 hours. Bondable materials Aluminum and other metals, glass, PVC, polystyrenes, polycarbonates, acrylics, thermosetting resins Suitable areas of use Small to moderately large bonding surfaces where short fixing times and high impact and peel strengths are required. On lightly oiled surfaces that cannot be cleaned. Strength Single lap joint: Al/Al (AlCuMg alloys), 20 – 25 N/mm2; PVC/ PVC, 8.5 – 14 N/mm2 Peel strength Suitable areas of use 17 T-peel test, Al/Al: 2.5 – 3.5 N/mm2 Temperature range -40° to +100°C. At 75°C, about 70% of the strength at room temperature remains. Durability/resistance Good against water and oils (contact the supplier for further details). Work environment Requires good ventilation and handling with no skin contact. In Sweden, work with acrylic adhesives is regulated by AFS 1996:4, Härdplaster (ordinance 1996:4 of the Swedish Work Environment Authority, Thermosetting resins). Tested adhesive 9 SGA ADHESIVE GZM, GARCO Components Two, which are spread separately, i.e. adhesive on one surface and hardener on the other. Function There is a very fast reaction when the adhesive and hardener come into contact. Giving less vapour emission, shorter curing times, better gap filling capabilities and improved resistance to heat and solvents, this product is regarded as a development of SGA adhesives. Curing times Fixing, 15 – 30 seconds. Final curing, 1 – 2 hours. Bondable materials Aluminum and other metals, glass, thermosetting resins, PVC, glass-filled nylon (poorer with other thermoplasts) Suitable areas of use As for other SGA adhesives, where a short fixing time is the goal. Strength Single lap joint: Al/Al, 21 N/mm2; epoxy/fibreglass laminate, 11.2 N/mm2 ANAEROBIC ADHESIVES Because the first adhesives with an anaerobic function were used to lock screw joints, they are usually referred to as locking fluids. Broadly speaking, anaerobic means “without air”. For curing to take place, these products require the absence of air. The majority also require metal ions for curing. Development of these “locking fluids” has been rapid and the group now has products that act as adhesives. The original curing system (air free + metal ions) has been complemented by products that cure using ultraviolet light, heat or activator. Anaerobic adhesives are suitable for small to moderately large surfaces. They are widely used in the electronics industry. Anaerobic adhesives are not so suitable for plastics. Tested adhesive 10 ANAEROBIC ADHESIVE LOCTITE 326, G A LINDBERG Components: 1 The adhesive can be used with or without activator. Activator is used when bonding surfaces that do not have active metal ions and/or to accelerate curing. Function Cures in thin gaps and in contact with metal ions. Curing on anodized aluminum requires activator. Curing times Curing times very much depend on the metal that is to be bonded. Fixing times of around 30 minutes are usual. To cure in a reasonable time, aluminum alloys with a copper content under 1% or a lot of magnesium require heat input or an activator. Bondable materials Aluminum and other metals, ferrite magnets, glass Suitable areas of use Assembly in the engineering and electronics industry (small bonded surfaces) Strength Single lap joint Al/Al: 12.5 – 20 N/mm2 Peel strength T-peel, 0.5 – 1.5 N/mm2 Temperature range -50° to +120°C. At 90°C, strength is about 50% of that at room temperature. Work environment Skin contact must be avoided. Work with anaerobic adhesives (which are classed as acrylic adhesives). Test Methods The most common test method is the single lap joint. This most often provides the basis for adhesive suppliers’ data sheets. The test is illustrated in fig. 15. 25 1,6 12,5 140 Figure 15: Standard shear test, ASTM D 1002-99 18 Adhesive strength is given in N/mm2, which is an expression of average stress. Locally, there are both higher and lower stresses in the joint. The stiffer the adhesive that is used, the higher the stress peaks the test returns at the edges (see “Joint design”). The test is normally performed against aluminum. However, apart from the adhesive itself, it is not always stated which materials make up the bonded joint. Because harder materials normally give higher strength values for the joint, this absence of information raises further questions. The tensile speed when the test load is exerted is of great significance for the measured results. Before 1994, a constant tensile speed of 10 mm per minute was used. This gave short test times, but no idea of creep resistance in the joint. A tensile speed of 1.3 mm per minute is now used. However, this also does not give an idea of the adhesive’s creep resistance. Where the adhesive is to be exposed to a constant load at elevated temperatures, data on the joint’s creep strength must be compiled. As intimated in “The effect of the service environment”, the “composite stress” exerted by temperature, environment and load is the only “realistic” test. In cyclic testing (water soaking-freezing-drying) of an object, the thermal movements and the taking up of water induce stresses in the joint. This induced stress arises more rapidly than with soaking alone. Table 5 on page 18 shows, amongst other things, the results of this test carried out on the adhesives tested by Sapa Technology. Storage in 100% humidity at 60°C can give a quicker idea of any degradation of the joint than can corresponding storage in water. The wedge cleavage test (ASTM D 3762-98) is comparatively easy to carry out and does not require special equipment. This test enables load, environment and temperature stresses to be combined. Bonded test pieces are prepared (see fig. 16). When the bond has cured, a 3 mm thick wedge is pressed into one end of the test piece. The initial crack formed at the wedge tip is marked out. Next, the test pieces are exposed to 100% RH at 60°C for 60 – 75 minutes. Crack propagation is then marked and measured again. This can be repeated throughout the period available for testing. A week of testing usually reveals clear differences between different adhesives and pretreatments. At the conclusion of testing, the bonded joint is taken apart and the types of failure that have occurred are analysed. With heat-curing epoxy adhesives on pretreated aluminum, there is most often a cohesive failure (failure in the adhesive) in the first five millimetres. The longer the crack then grows, the greater the ratio of boundary layer failures that arise. This method has been used by the “Institute for Metallic Construction Materials” at Chalmers University of Technology to investigate the long-term strength of bonded joints to aluminum that has undergone various pretreatments (refs. 2 and 5). Such testing can be comparative or used as a production check (e.g. production parameters – checking of pretreatment baths, etc.). The method can be modified so that it is suitable for use with other materials bonded to aluminum. a a Figure 16: Wedge cleavage test, ASTM D 3762-98 References (1) ”Durability of Structural Adhesives” A.J. Kinloch (editor), Applied Science Publishers London (1983), ISBN 0-85334-214-8. (2) ”Förbehandling för limning mot aluminum” Tore Rönnhult och Bengt Nilssor, Institutionen för Metalliska Konstruktions-material, Chalmers Tekniska Högskola, Göteborg (1982). (3) ”Effect of Surface Preparations on Adhesive Bonding of Aluminum”, J.D. Minford, Adhesives Age, July 1974. (7) ”Handbook of Aluminum Bonding Technology and Data” J. Dean Minford. Marcel Dekker Inc, NY, Basel, Hong Kong, 1993 (744 sidor, 4686 referenser) ISBN 0-8247-8817-6. (8) ”Adhesion in Bonded Alurninium Joints for Aircraft Construction”, W. Brockmann, O-D Henneman, H. Kollek och C. Matz. International Journal of Adhesion and Adhesives,.6(3), July 1986. (4) ”Effect of Surface Preparation on Stressed Aluminum Joints in Corrosive Saltwater Exposure” J.D. Minford, Adhesives Age, October 1980. (5) ”Surface Treatment of Aluminum Alloys for Adhesive Bonding” Laszlo Kozma, Institutionen för Metalliska Konstruktionsmaterial, Chalmers Tekniska Högskola, Göteborg. (1984) Internrapport R 445/84. (6) ”Ytbehandlingens betydelse vid limning av aluminumdetaljer”, Bengt Nilsson och Tore Rönnhult, Institutionen för Metalliska Konstruktionsmaterial, Chalmers Tekniska Högskola, Göteborg (1983), Internrapport R 424/83. 19 Tests of a few selected adhesives Sapa Technology tested a number of different adhesives on aluminum that had either been ground or anodized in sulphuric acid and then sealed. For each of these two pretreatments, ten different adhesives were tested. Some of the adhesives were comparatively common epoxy adhesives, others were chosen to give a more rubber-like joint. These adhesives are presented in more detail in the “Choice of adhesive” chapter. After full curing, the adhesives were subjected to soaking or cyclical variations (moisture, cold, heat) and then tested. The results were compared with reference samples stored in dry air at 20°C. The test results are shown in table 5 on page 18. Test data Tensile speed, 10 mm/min at 20°C Tested adhesives 1. 2-component epoxy, Araldit AV 106/HV 953 U (Standard) 2. 2-component epoxy, Araldit AV 144/HV 997 3. 1-component epoxy, 3M Scotch-Weld EC 2214 (heatcuring, 120°C) 4. 2-component polyurethane, Casco 1852/1853 5. 1-component polyurethane, Sikaflex 221 (flexible) 6. Hot-melt adhesive, moisture-curing polyurethane, Bostik 9802 7. Polysulphide, Metallgesellschaft M 82 8. SGA adhesive, Loctite Multibond (acrylic rubber) 9. SGA/“Aerobic”, Garco GZM 10. Anaerobic, Loctite 326 (on anodized material, used along with activator NF) Test format Single lap joint on 2 mm Al (gives large stress concentrations). Alloy – condition SS 4104-06 Surface treatments • Anodizing in sulphuric acid + subsequent sealing (carried out by Sapa). Washing in warm water to which “Candoeleane Al” had been added, rinsing and drying before bonding. • Grinding with abrasive nylon (Scotch-Brite 7447, 3M Sverige). Joint thickness calibrated to 0.15 mm for all adhesives. Test environment A. Dry air B. 14 days in water at 60°C. Testing 2 – 7 days after drying. C. Cyclic: 1 week in water at 20°C 1 week in cold, -20 to -25°C 1 week in dry air at 20 to 23°C (Testing after 3 cycles and after 1 year of cycles.) Table 5: Strength, N/mm2, average value of 5 tests Various adhesives on alloy SS 4104-06 Single lap joint on 2 mm bars On a ground substrate On an anodized substrate Storage environment before testing Storage environment before testing Air 20°C Water 20°C Cycles 3(1) Cycles 1 year (1) Air 20°C Water 20°C Cycles 3(1) Cycles 1 year (1) 1 2-component epoxy Araldit AV 106/HV 953 U 8.9 G 8.3 G 13.3 G 10.8 G 9.3 G 13.3 G 16.3 G 13.9 G 2 2-component epoxy Araldit AV 144/HV 997 7.0 G/L 14.1 G 13.3 G 13.7 G 9.9 L(G) 16.8 L(G) 16.8 L(G) 15.5 G(L) 3 1-component epoxy 3M Scotch-Weld EC 2214 10.6 G(L) 13.3 G(L) 17.5 G(L) 11.5 G(L) 10.6 L(G) 20.7 L 21.8 L(G) 20.2 L 4 2-component polyurethane 10.3 L 7.5 G/L 16.9 L(G) 15.3 L(G) 8.8 L 13.3 L 15.8 L 14.3 L 5 1-component polyurethane elastomer Sikaflex 221 1.6 L 3.8 G 3.5 G/L 7.4 L(G) 1.5 L 2.9 L 3.1 L 3.4 L Moisture-curing hot-melt adhesive Bostik 9802 3.8 G 3.8 G 7.2 G 2.8 G 0.6 G 5.1 G 6.2 G 7.7 G 7 2-component polysulphide Metallgesellschaft M 82 1.0 L 1.3 L 1.9 L 1.7 L 0.9 L 1.0 L 1.9 L 1.4 L 8 Acrylic rubber adhesive (SGA) 7.9 G 10.1 G 13.7 G 8.4 G 9.0 G 9.9 G 11.8 G 10.9 G Acrylic rubber adhesive (Aerobic) 10.4 G/L 9.6 G 18.6 G 13.9 G(L) 8.7 G 3.9 G 4.2 G 8.4 G Anaerobic-curing adhesive 11.0 G/L 11.2 G 13.5 G 11.9 G(L) 7.3 G 7.8 L/G 13.4 G 14.8 G(L) 6 Casco 1852/1853 2) Loctite Multibond 330 9 10 Penloc GZM (från Garco) Loctite 326 1) One cycle per week 2) Replaced by Casco 1849/1821 PUR 2K 20 Failure type L = Clear break in the adhesive. Adhesive on both aluminum surfaces to 100%.. G = Boundary layer failure. G/L = Approx. 50% of each of G and L. L(G) and G(L) = The letter in brackets indicate a “small proportion”. Results – Discussion As shown by table 5, the strength of several adhesives increases after soaking in water and passing through the cycles. This may be because the relevant adhesives cure more fully in contact with water. In particular, this applies to the polyurethanes and the polysulphides. For the harder adhesives, this may be a question of a certain, non-harmful, taking up of water. The water then serves to soften the adhesive. This gives better stress distribution in the adhesive and, consequently, reduced loading of the boundary layer. Aluminum surfaces that have been pickled in sulphuric acid and then sealed have often been regarded as unsuitable for bonds. However, the test results show that many adhesives adhere just as well to anodized surfaces as they do to ground surfaces – in some cases even better! After one year of cycles, several adhesives show considerably less deterioration in strength on anodized surfaces than they do on ground surfaces. This applies especially to the more highly viscous adhesives. The explanation for this is probably: • Anodized surfaces are more resistant than ground surfaces and do not develop weak oxide layers that can delaminate under load. • Anodized surfaces are more even and are thus more easily filled by the adhesive (especially if it is viscous). This leaves less space for water in the boundary layer. • The tested adhesives were very different. For example, there were great differences in hardness. The three epoxy adhesives were all softened to give better load distribution and counteract stress concentrations. Adhesives 5 and 6 were very soft and elastic. Thus, the adhesives must not primarily be compared with each other, but with themselves in respect of how they respond to the different pretreatments. Examples of bonded aluminum products Example 1 – radiator unit Radiator unit in AA 3005-0 aluminum (clad with AA 4104); the tank (upper part) is in polyamide 6.6 with a 30% fibreglass filling. The aluminum radiator is bonded to the plastic tank. The intended service environment is -40° to +120°C in constant contact with a water/glycol mixture. As the radiators are manufactured in very large numbers, streamlined production is essential. Pretreatment The cladding is rich in magnesium. After vacuum soldering of the radiator unit, magnesium oxides make up most of the surface. Even if it is clean, this is not a suitable surface to bond to if there is moisture/water in the service environment. Radiator unit: Alkaline pickling + chromating (trials with a chromate-free alternative are in progress). Polyamide tank: Manufactured without release agents and not pretreated. Adhesive One-component heat-curing epoxy adhesive, similar to no. 3 of the tested adhesives in table 4. Alternative adhesives None that offer any advantages with maintained safety. Bonding procedure To improve the flow of the adhesive, the components are preheated to 60°C. The adhesive is applied to the plastic component. Curing in an oven at 130°C for 60 minutes. Results Sealed, durable fixing of the stub ends of the radiator unit’s pipes. The adhesive’s thermal expansion coefficient matches that of the plastic. Thus, variations in temperature do not present problems. Disadvantages Comparatively long curing time at 130°C; requires large oven capacity for large series. Remarks When used as a component in vehicle production, any faults have very serious consequences. For this reason, tested methods and components are preferred to others that might, for example, give shorter curing times. Example 1: Bonded radiator unit alongside unbonded components 21 Example 2 – outside planking (boat) “Profilen” is a ¾ tonne sailing boat that is some 10 metres long. It is built of extruded aluminum profiles (alloy AA 6063) developed by Sapa. to the preceding, underlying profile. The adhesive is now available in a double cartridge that delivers ready-mixed and air-free adhesive direct from the pack. To make the boat rigid and watertight, the profiles are bonded to each other. The overlying profile is then fitted within 60 minutes and secured by clamps. Curing takes place at room temperature and lasts at least 16 minutes. As with wooden boats, the skins of boats of this type are built plank by plank. This requires an adhesive that gives long handling times and which is able to fill comparatively large gaps. The adhesive must cure at room temperature. The service environment is the normal one for many boats – water (often salt) in the summer and dry and cold in the winter. Pretreatment The profiles are anodized in sulphuric acid and then sealed. A Sapa facility anodises the profiles in full lengths. Transport, cutting and bending must then be carried out in a way that causes the least possible fouling. This is because efficient cleaning of these long profiles is difficult to achieve simply. (Wiping with a rag and solvent does not remove all contaminants.) Adhesive Two-component, rubber-modified, slow curing epoxy adhesive (adhesive no. 2 in table 5). Bonding procedure The adhesive used to be mixed in a suitable quantity for the profile that was to be bonded. It was then applied as bead Results The first bonded sailing boat has competed successfully in races and, during the period it was checked (around 10 years), remained perfectly watertight. Disadvantages The comparatively long curing time means that no more than one plank per side can be bonded per 24 hour period. Alternative adhesives The thick, hard, post-sealed, sulphuric acid anodizing cannot, without breaking, transmit high loads for a long time in corrosive environments. Softer adhesives could, therefore, be interesting. Tests using polyurethane type rubber adhesives (adhesive no. 5) and polysulphide (adhesive no. 7) were carried out in parallel with the epoxy adhesive actually used. The tests were conducted on loaded joints in a brine environment. Remarks Heat curing is not possible. Adhesives that cure more quickly at room temperature reduce the handling tolerances and give poorer results. Example 2: “Profilen”, a ¾ tonne sailing boat built from bonded, extruded aluminum profiles 22 Example 3 – hatch for pleasure boats Hatches (vents) for pleasure boats are made from aluminum profiles and 10 mm thick, smoked, acrylic plastic sheet. better adhesion to plastic than does, for example, an acetic acid curing system. Before assembly, the aluminum profiles are bent, welded and anodized. The adhesive is very elastic with an elongation at rupture of around 500% and a breaking stress as low as 1 N/mm2. The acrylic sheet is bonded to the aluminum profiles. Bonding procedure The window is fixed in the aluminum frame using a butyl rubber strip (sticky) that also forms the bottom seal for the subsequent silicon joint. In the remaining gap, which is 4 – 5 mm wide, the silicon is applied from a 300 ml cartridge. The moisture in the air cures the silicon. Series size is moderate. For this reason, manufacture has been set up as “streamlined craftsmanship”. The hatches have to withstand both salt water and UV light in a temperature range that can be -30°C to + 60°C. The adhesive bond is exposed to its greatest loads at low temperatures. Here, because of the different coefficients of thermal expansion, the acrylic glass window shrinks in relation to the aluminum frame. The forces that arise are on a par with each material’s yield strength. After curing, a scalpel is used to cut away excess silicon. To make the stress as small as possible, an elastic joint is desirable. The bonded joint is not loaded appreciably apart from the temperature-induced stresses, but these are large. Disadvantages Relatively slow curing (24 hours), especially in the winter when the relative air humidity indoors can be low. Pretreatment The aluminum profiles are washed in a mild, alkaline solution. They are then rinsed in running water and air dried. The edges of the acrylic plastic are polished with a gentle abrasive (abrasive nylon, Scotch-Brite 3M, 7447). Adhesive One-component, moisture-curing silicon, “alcoxy system”. Silicon with an alcoxy curing system generally provides Results A sealed, elastic joint with good adhesion. The joint is elastic even in the cold. Alternative adhesives One-component, moisture-curing polyurethane rubbers. However, to ensure adhesion throughout long-term exposure to UV light, many of these adhesives require a black primer on the acrylic window (there are now some products that withstand UV irradiation – they are used for, amongst other things, filling boat decks). Remarks Curing can be accelerated using a moisture tent and moist heat. Great precision and a certain amount of skill are required to achieve a fully filled joint that is free from bubbles. Example 3: Aluminum frame and “acrylic glass” for a boat hatch (before bonding) 23 Example 4 – sandwich panels Sandwich panels with an aluminum covering floor plate and a PVC cellular plastic core. The panels are used as floors in refrigerated vehicles. They can measure up to 2.4 x 15 metres and are made individually for each vehicle. Panel and bonded joint must withstand temperatures of from -40°C to +80°C. The high temperatures may arise on internal washing. Loads are exerted by cargoes and various fastening devices. They occur as vibrations, torsional stresses, impacts during driving and static creep stresses while the vehicle is stationary. Not least, thermal movements are important. It is required that the adhesive must, in all positions, be stronger than the core material and that it does not “release contact”. Making panels of this size demands that the adhesive has a reasonably long open time. The adhesive should also have good filling properties. Pretreatment At manufacture, the core material is calibrated in a “cutter” and protected throughout transport and storage. No other pretreatment is undertaken. Before bonding, the aluminum is machined using ScotchBrite discs. These give a mildly abraded surface. Where the joint has to satisfy severe corrosion resistance requirements, the aluminum surface is then primed using a wash-primer of the polyvinyl butyral type with a phosphoric acid hardener (curing agent). Example 4: Floor panel for a refrigerated vehicle 24 Adhesive Two-component polyurethane with an open time after application of at least 45 minutes. The adhesive employed resembles no. 4 in table 4. Bonding procedure The adhesive used to be spread on both the core material and the covering plate. This ensured, amongst other things, that the open time before application of press force was considerably extended. The adhesive is applied using a “hand applicator” (roller with reservoir). Nowadays, the adhesive is spread on only one of the materials. Spreading is in tight beads, normally using automatic application equipment. A vacuum is used to apply pressing force. The panels are built up on a flat surface and are covered by a rubber sheet, from beneath which the air is evacuated. Per m2, the press force is around 8 tonnes. This was usually held for 12 – 16 hours. With currently used adhesives of this type, the press force need only be applied for 3 – 4 hours. The open time is at least as long and may even be longer. Results A rigid panel that meets the imposed strength requirements. Disadvantages The long press time ties up production surfaces. New adhesives in this group lessen the problem. Naturally enough, even shorter press times are desirable. Alternative adhesives Two-component epoxy adhesives were previously used. However, all bonds are now made using polyurethane adhesives. Example 5 – Honeycomb panels Honeycomb panels as air guides in a wind tunnel with a diameter of 5 metres. Honeycomb materials are used as, amongst other things, cores in the making of sandwich panels for aircraft applications. The honeycomb is made of thin (0.2 mm) aluminum sheets. These are bonded using offset beads of adhesive so that, on drawing/stretching, they form hexagonal holes. Heat-curing adhesives (rubber modified epoxy and phenolic adhesives) diluted with a solvent are used for this. The adhesive is applied to the aluminum sheet in thin beads and the solvent is allowed to evaporate. Sheets are then stacked to form a bundle that is compressed and activated/cured in a hot press at 125 – 175°C for 30 – 60 minutes. The homogenous bundle is then machined to the desired dimensions and drawn out to its final shape. To enable honeycombs to be joined to honeycombs, cores that are 100 mm thick must be given a vertical metal edge (140 mm). Depending on the structure of the panels, the contact surfaces at bonding are very small. In wind tunnels, vibrations arise that may exert a splitting force on bonded joints. Hence, an adhesive that fills and envelops the thin sheet is required. The adhesive must not be brittle and must have high peel strength. It must also take up the correct position and not move away before it has set. Pretreatment To remove contaminants and weak oxide layers, the honeycomb material and the surface of the aluminum sheet are brushed with “Alonyl brushes” (circular brushes of polyamide fibres that contain around 30% aluminum oxide – from Sinjet, Huskvarna). So that the adhesive is evenly distributed across the entire width of the sheet/honeycomb, the fixture is then stood on edge. Before removal from the fixture, curing takes place for 16 hours at room temperature. Results A joint that has good peel strength and transmits the loads that can arise between the honeycomb and aluminum sheet. The wind tunnel has been in use for some ten years with wind speeds of up to twice the speed of sound. No fatigue fractures have been observed. Disadvantages Getting the adhesive out of its container is messy. (This adhesive is now also available in double cartridges that deliver ready-dosed, air-free, mixed adhesive direct from the pack.) Alternative adhesives SGA adhesives (adhesive no. 8 in table 4) where the hardener is spread in the form of a varnish on the aluminum sheet and the adhesive is spread on the honeycomb material. The use of 2-component, mixed acrylic adhesives is also imaginable. These normally have shorter open times and shorter curing times. Remarks For mounting to the wind tunnel structure, 100 mm long aluminum rods were bonded and placed in a ring around a cell hole. Sikaflex 221 (adhesive 5 in table 4) was used for this. Adhesive Two-component, rubber modified epoxy adhesive that is moderately runny before setting at room temperature (similar to adhesive no. 2 in table 4). Bonding procedure Using a toothed spreader, an approximately 1 mm thick layer of adhesive is applied to 100 mm of the aluminum sheet’s width. The honeycomb material is placed in a simple wooden fixture. Hexagonal pencils and rubber bands are used to fix the adhesive-coated sheets to the honeycomb side. Example 5: Wind tunnel with panels mounted Example 5: Honeycomb panel with bonded joint plate and mounting reinforcements 25 Example 6 – body panels for buses Bus/coach bodies comprise a steel skeleton to which a shell of aluminum cladding panels is added. Previously, the panels were fitted using blind rivets. To distribute loads more evenly and minimise the number of rivets, adhesive bonding is now used. Moisture is the main environmental stress and the structure must withstand temperatures of from -40° to +70°C. The loads are moderate. However, the panels must remain in place in all positions. A design that uses sheet metal panels in this way requires an adhesive that has good peel strength. Pretreatment - The steel structure is degreased and primed. The aluminum panels are either primed or chromated. Alternative adhesives Double-sided sticky tape (double-sided PSA tape) approx. 1 mm thick and built up entirely of an adhesive compound is also used for this purpose. 2-component, acrylic rubber adhesives – mixed or with a separately spread hardener. Remarks There are also 1-component polyurethane adhesives that give similar final properties. They cure in heat (min. 70°C) or are applied hot (approx. 80 – 90°C). After cooling for around 1 hour, they have mastic-tape properties. Moisture is then responsible for final curing. Adhesive - One-component, moisture-curing, polyurethane elastomer (gives a rubber-like joint resembling adhesive 5 in table 4). Bonding procedure The adhesive is applied to the steel structure; the panels are lifted into place and fixed with the necessary number of blind rivets (at least two per panel). Each panel is pressed lightly against the adhesive so that there is contact with no appreciable springback. Thanks to air moisture, the adhesive cures to sufficient strength in around 24 hours. Results - With good peel strength, a rubber-like bonded joint that deforms elastically and does not give rise to stress concentrations. Disadvantages Relatively long curing time if relative air humidity is low. Example 6: Coach with bonded cladding panels and glass windows. Example 7 – thermal foil Thermal foils with pattern-like printed circuits – aluminum foil on thermoplastic polyester foil. Used in, for example, electrically heated door mirrors on cars. This is a case of mass production with various product forms. Manufacture must be streamlined and the bonded joint flexible and comparatively heat resistant. Thermoplastic polyester has low surface tension. This limits the number of adhesives that can wet it (and thus give good adhesion). Pretreatment Aluminum foil: no pretreatment takes place. Polyester foil: corona treated (electrical discharges that increase surface tension) at manufacture. To increase surface cleanness, the aluminum foil can also be treated in this way. Adhesive Polyurethane rubber solution with curing (hardening) additives (isocyanates). The adhesive is “heat activable”, i.e. after it has dried (the solvent has left the adhesive), it can once again be made “runny” by using heat. Bonding procedure The adhesive solution is rolled out onto the polyester film. The solvent (around 80%) is allowed to evaporate and the dry layer of adhesive is heated until it melts. In this condition, the aluminum foil and the adhesive-coated plastic film are roller-pressed together. To obtain the desired pattern, the base laminate is screenprinted with a protective varnish. Unwanted aluminum foil is etched away and the protective varnish is then 26 washed off. The now only partly coated polyester surface is protected by laminating it with a further adhesive-coated polyester foil. This is also done through heat activation and roller-pressing together. Results A flexible thermal foil that has a service temperature of up to 120°C. There are also foils that tolerate considerably higher temperatures. Disadvantages The use of solvents is less desirable. Solvent emissions in workplaces and to the environment must be checked and, through various measures, held at a low level. Remarks “Heat activable” adhesives in the form of water dispersions are now also available. These are also used with an added hardener. Example 7: Thermal foil for a door mirror Example 8 – aircraft Aluminum in aircraft is largely joined using adhesives. Production is meticulously controlled and all adhesive bonds are checked using, amongst other things, ultrasound. The adhesive bonds have to withstand: very low temperatures without becoming brittle; and, at high temperatures, large loads without creeping. They must be of an even and set thickness and must also be free of air bubbles. Remarks Bonded joints delivering the performance and quality required of structural bonds in aircraft cannot be achieved in a simpler way. Pretreatment Automated anodizing in phosphoric acid without subsequent sealing. This is followed by drying in clean air and immediate coating with a primer that is suitable for the adhesive. Long used in Europe, anodizing in chromic acid also occurs. As this leaves a more even surface, priming can then sometimes be omitted. Pure, clean cotton wool gloves must be worn for any contact with the pretreated surfaces. Adhesive Epoxy-based adhesive films are the most common. Even joint thickness is assured by the film format. Modification through the use of rubber (for peel strength) or phenol (for heat resistance) additives occurs. Load conditions determine the choice of adhesive. Bonding procedure All bonds of a structural nature are pressed and cured in an autoclave at a press force of 1 – 3 N/mm2 at a temperature of 125 – 175°C for 30 – 90 minutes (depending on adhesive type). Results Aircraft that do not fall apart! Disadvantages Comprehensive programme of checks for the anodizing bath. Length of curing time and the comparatively high curing temperatures. Alternative adhesives Adhesive films based on phenolic nitrile are also used. These were the first adhesives designed for use in the aircraft industry (used since the 1940’s). Example 8: Saab 340 being assembled Example 8: Parts of an aircraft body going into an autoclave 27 Example 9 – bonded brake linings Brake linings on cast aluminum shoes. The bonding of brake linings requires a joint that has very good heat resistance. Pretreatment An abrasive is used to lightly grind the friction material. The aluminum shoes are blasted with sand or aluminum oxide. Adhesive A phenolic-nitrile solution that is applied to both joint surfaces and is allowed to dry completely. Alternative adhesives Dried but not cured adhesive films are now available. These can be placed dry between the surfaces. They are then melted and cured in the same way. Remarks Adhesive films for this use are a recent innovation and should be regarded as providing an improvement in the work environment. Bonding procedure The dry, adhesive-coated surfaces are put together and subjected to pressure. They are then heated in an oven. As the adhesive melts, it wets the surfaces. With the press force maintained and continued heating, the adhesive cures. Depending on the heat-resistance requirements placed on the bond, curing temperatures and times can vary from a few minutes at 120°C to several hours at 180°C. Results Brake linings that are securely attached, but the heat resistance of which depends on curing time and temperature. Disadvantages As always when handling solvents, workplaces must be well ventilated and the management of extracted vapours must be approved and checked. Example 9: Bonded brake linings Example 10 – power saw handle Power saw handles made from aluminum pipes and bonded to the cast motor block. In cold weather, the handle can be heated by channelling the exhaust gases through it. The joint must give a good seal and withstand temperatures of between -40°C and +70°C. Alternative adhesives One-component, heat-curing, epoxy adhesives. However, these are more viscous and can be difficult to “get into place” in the joint. Heat curing at a minimum of 100°C is also usually seen as a disadvantage. Design The design of the cast motor block provides two cylindrical holes to take the ends of the pipes. Remarks In locked joints of the “pin in hole” type, the thermal expansion coefficients of the materials should be borne in mind if the joint is exposed to temperature fluctuations. This very much also applies to the adhesive. An adhesive’s expansion is often 5 – 10 times greater than that of the metals involved. Amongst the anaerobic adhesives, there are variants that are of a sufficiently low viscosity for them to be applied to assembled joints that have little play. These adhesives are then sucked in by capillary action. Pretreatment The motor block’s holes are reamed to the correct size (H7) for the pipe ends. A fine adhesive belt is used to grind the pipe ends Adhesive Anaerobic adhesive (similar to adhesive no. 2 in table 4). The adhesive is a runny, 1-component adhesive that starts to cure in the closed joint in contact with active metal ions. It requires aluminum with at least 1% copper or an application of activator on the joint surfaces. Bonding procedure The prepared holes in the motor block are moistened with adhesive. The pipe ends are coated more liberally and pushed into place. Curing to handleable takes around 15 – 30 minutes. During this time, the joint should not be disturbed. Excess adhesive outside the joint does not cure and can be easily wiped or washed away. Results - A sealed, strong joint. 28 Disadvantages As the adhesive can be pushed aside at assembly, it is often difficult to achieve fully filled joints where these are of the “pin in hole” type. Removing adhesive outside the bonded joint can give rise to work environment problems. Example 10: Power saw handle Example 11 – corner structures Extruded aluminum profiles are widely used for claddings, doors, windows and surrounds. Using angle pieces in the corners, these profiles are often joined into fames. These angle pieces can be made of punched or cast profiles. They are bonded into the frame profiles. This requires a joint-filling adhesive that does not shrink on setting. The adhesive must withstand a normal outdoor environment. Disadvantages Series are large and, therefore, involve many people working with epoxy adhesives. The occupational safety issues connected with the handling of epoxies are thus felt to present a problem. The long curing times and the need to frequently mix small batches of adhesive are sometimes seen as problems. Pretreatment At delivery to the various production units, the profiles are anodized or painted. Cutting fluid is used when cutting the profiles. Practically, it is not feasible to clean the cutting fluid residues from the inside of the profiles (even though this is exactly where the angle piece is to be seated). Alternative adhesives Sapa Technology has tested a number of alternative adhesives. Joints were made using surfaces that had been moistened with cutting fluid (several different cutting fluids were tested). The tested adhesives were: • 2-component polyurethanes (adhesive 4 in table 4) • 1-component, moisture-curing, polyurethane elastomers (adhesive 5 in table 4) • 1-component, foaming, moisture-curing, polyurethane adhesives • Anaerobic adhesives with activators for corner elements (adhesive 10 in table 4) Adhesive The majority of these joints have long been made using 2-component, slow curing epoxies (adhesives corresponding to no. 1 in table 4). Bonding procedure Adhesive is applied relatively liberally inside the aluminum profile. The angle piece is pushed inside and the corner is fixed by stamping the overlying profile. Frames are then stored, stress-free and with the corner at 90°, for around 16 hours. Results Can be seen in doors, windows and aluminum profile frames in buildings! Testing was carried out on single lap joints and joints comprising square tubes inside square tubes. Results were compared with those for 2-component epoxies. The strength of the cured bonded joints was compared with that of joints aged for 60 days at 60°C in 100% humidity. The test results show a marked deterioration in the strength of all bonded, single lap joints stored in 100% humidity. Apart from one 2-component polyurethane adhesive, the same also applies to “square tube in square tube” joints. The use of a two-component polyurethane adhesive instead of the current epoxy adhesives could be seen as an improvement in quality. Example 11: Bonded corner structure made from extruded aluminum profiles 29 Example 12 – glass windows with aluminum frames Windows for vehicles (buses, trains, loaders, trucks, etc.) are made from extruded aluminum profiles and glass. Aluminum locks and fittings are bonded to the glass. Thermal movements and service environments require an adhesive that does not become too hard and that, in wet environments, does not lose adhesion with the glass. The rather large production volumes mean that short fixing times are desirable. Pretreatment None. The chosen adhesive type is very tolerant of oil. Adhesive SGA adhesive (acrylic rubber adhesives, correspond to adhesive no. 8 in table 4). Bonding procedure The hardener is spread on the glass and the adhesive is spread inside the aluminum profile. Fixing/pressing takes around 5 minutes. Disadvantages The smell and work environment considerations associated with this adhesive type mean that work has to be carried out in very well ventilated areas. For the joint to cure at all points, the adhesive and hardener should be spread so that, on assembly, they always come into complete contact with each other. Alternative adhesives Mixed acrylic rubber adhesives. One and two-component silicon adhesives. However, these adhesives have considerably longer (hours) fixing times and give a softer joint. Remarks Adhesives of this type are available in double cartridges that contain both hardener and adhesive. On being squeezed out, the components pass through a mixing nozzle. This avoids the need to apply to both surfaces. The risk of uncured adhesive also disappears. Results An elastic, strong joint that, both for mechanical loads and for forces arising from thermal movements, gives excellent stress distribution. Example 12: Glass vehicle windows with aluminum frames and handles 30 The information in this text was considered correct at the time of publication. However, Sapa accepts no legal liability for the correctness or completeness of any of said information. 31 © Sapa Profiles, Inc. Sapa Extrusions Address 7933 NE 21st Avenue Portland, Oregon 97211 USA 32 www.sapagroup.com/us/profiles