Adhesive bonding in steel construction challenge and innovation

Adhesive Bonding in Steel Construction Challenge and Innovation Procedia Engineering 172 ( 2017 ) 186 – 193 1877 7058 © 2017 Published by Elsevier Ltd This is an open access article under the CC BY NC[.]

Available online at www.sciencedirect.com ScienceDirect Procedia Engineering 172 (2017) 186 – 193 Modern Building Materials, Structures and Techniques, MBMST 2016 Adhesive bonding in steel construction - Challenge and innovation Yvonne Ciupacka*, Hartmut Pasternaka, Christoph Metteb, Elisabeth Stammenb, Klaus Dilgerb a Chair of Steel and Timber Structures, BTU, Konrad-Wachmann-Allee 2, 03046 Cottbus, Germany Institute of Joining and Welding, Technische Universität Braunschweig, Langer Kamp 8, 38106 Braunschweig, Germany b Abstract Despite much advancement of typical joining techniques in steel construction, fundamental problems, like residual stresses for welds as well as weakening of the cross section for bolts and screws, still remain The application of bonded joints could improve the situation The automotive industry shows the potential of this certain joining technique since years Even in civil engineering and especially in steel construction, researchers are continuously establishing the bonding technology as a structural element On the one hand bonded joints mean a challenge, but on the other hand an innovation, what is shown in this paper © 2017 Published by Elsevier Ltd This is an open access article under the CC BY-NC-ND license © 2016 The Authors Published by Elsevier Ltd (http://creativecommons.org/licenses/by-nc-nd/4.0/) Peer-review underresponsibility responsibility of organizing the organizing committee of MBMST Peer-review under of the committee of MBMST 2016 2016 Keywords: Adhesive; joining technology; steel structure Challenge 1.1 Design of bondline geometry A great number of adhesives with different carrying and deformation behavior, as well as varying handling properties are available for the design of bonded structures Due to a general lack of experience the selection of suitable adhesive systems is a difficult task for a civil engineer General requirements for an adhesive are characterized by a specific viscosity for the manufacturing and a sufficient carrying and deformation capacity of the bondline Despite these parameters the properties of the adherent surface play an important role for the compound * Corresponding author Tel.: +49-355-69-2255; fax: +49-355-69-2144 E-mail address: yvonne.ciupack@b-tu.de 1877-7058 © 2017 Published by Elsevier Ltd This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Peer-review under responsibility of the organizing committee of MBMST 2016 doi:10.1016/j.proeng.2017.02.048 Yvonne Ciupack et al / Procedia Engineering 172 (2017) 186 – 193 Thus, cleaning and degreasing of steel surfaces are essential during the bonding process In addition, mechanical pre-treatments such as blasting can improve the adhesion properties Uniform load Shear: "suitable" Pressure: "suitable" Tension: "less suitable" peel: "bad" cleavage: "bad" Uneven and combinated load Shear and peel: "bad" Fig Influence of geometric design and loading of a bondline Moreover, the kind of loading has a big influence on the carrying capacity of a bonded steel construction Uneven and combined loads as well as peel stresses should be avoided, as shown in Fig But uniform distributed tension, pressure and shear impacts are suitable By slopping the adherent ends stress peaks can be reduced, which increases the carrying capacity of the bonded connection 1.2 Carrying and deformation behavior of bonded constructions The carrying and deformation behavior of bondlines is nonlinear and time dependent According to the kind of adhesive basis distinct creep and relaxation effects can be observed The long-term mechanical behavior is influenced by environmental effects, e.g temperature, UV-radiation and humidity These impacts can lead to irreversible degradation of the mechanical bondline properties Particular attention is paid to the temperature dependency, because adhesives are characterized by a so-called glass transition temperature If a bondline is exposed to high thermal impacts, higher than the glass transition, it shows increased deformations and a decreased ultimate stress Low temperatures lead to a high stiffness and an increased notch sensitivity of the bondline These relations can be utilized in application, e.g elastic adhesives and sealing materials are used for temperatures higher than the glass transition to avoid thermal residual stresses and compensate thermal strains Because of the complexity of the carrying and deformation behavior, an appropriate description of the stress distribution in a bondline means complex calculations, e.g numeric simulations Since simple and standardized concepts are necessary for the design in civil engineering, the development of suitable analytical models poses a special challenge a b Fig (a) Butt joint test; (b) Lap shear test Another important challenge is the determination of characteristic material parameters for the design of bondlines Whereas the temperature dependency of material properties for steel adherents need only to be considered for fire design, it is not sufficient for adhesives The detection of time dependent material parameters, by consideration of damaging environment impacts, is imperative But currently, only simple test methods are available, which exclute the implementation of suchlike influence factors On the one hand the shear-stress to shearstrain relation can be described with shear lap tests according to DIN EN 14869-2 [1] On the other hand evidence about the carrying behavior of bondlines under normal stress states can be obtained from butt joint tests according to DIN EN 15870 [2] The specimens and loading situation for these experimental investigations are shown in Fig 187 188 Yvonne Ciupack et al / Procedia Engineering 172 (2017) 186 – 193 With these destructive tests, the bondline properties can be characterized, but initially without consideration of time and environment dependent effects Interaction relations can be deduced from multi-axial tests, such as arcan-tests The offhanded transfer of knowledge from these small scale specimen tests to typical steel construction dimension is not possible, so specimen component tests still remain necessary Innovation 2.1 Application of bonding technology for faỗade construction The Chair of Steel and Timber Structures of the Brandenburg University of Technology deals with two application examples of bonding technology in different research activities A bonded connection of a trapezoidal faỗade and a bonded reinforcement of a hollow section, for a typical transom-mullion-faỗade, are investigated Faỗades with trapezoidal sheets are deployed predominantly for light weight halls and industry constructions After the erection of a steel skeleton as the primary structure, a substructure of light weight steel profiles is installed, facilitates the connection of the trapezoidal sheets by screws or cartridge-fired pins This joining method with fasteners weakens the cross section and can lead to inadvertent dents, scratches or holes in the faỗade view In addition, the self cleaning effect is dysfunctional in the area of salient fastener heads Due to these problems of typical joints, the possibilities of indirect bonded faỗade connection have been developed, which are shown in Fig 3a The connection concepts are designed so that dead loads are carried by special load application points at the top of the faỗade Thus, creep effects of the bondline are avoided Specially formed connection profiles in addition to a bolt connection, realized with an elongated hole, allow the deformation of the outer shell without introducing constrains into the bondline cover b a seal L-profile T-profile 3-profile outside isolation glass beam screw self-drilling inside reinforcement bondline Fig (a) Structures for bonded faỗade connections with different shape of connection profile; (b) Bonded faỗade reinforcement Principal requirements of architects and contractors for representative buildings are high side view transparency, structured and light filled faỗades In a lot of cases the transom-mullion-construction is realized, so in order to achieve the desired impression, it is necessary to put the primary structure in the background That requires a minimization of the outer dimensions by increasing the stiffness of the structured member at the same time An innovative solution is offered by the adhesive bonding technology With a steel reinforcement, bonded to the inside of a typical hollow section, a new type of compound cross section with an increased stiffness and carrying capacity is created (Fig 3b) The advantage of a reinforced transom-profile compared to a hollow section with higher thickness is the lesser slenderness of the webs, thus avoiding the buckling of webs under pressure The principle of faỗade hollow profiles with an inner reinforcement is similar to glued laminated timber, but in contrast to timber structure, a bondline design and evidence is necessary, due to the higher stiffness of the steel adherents The presented constructions are developed so that all bonded connections can be realized in a laboratory under controlled conditions Non-reproducible bondlines and bonding on the side are avoided and the pre-prepared elements can be mounted on site The joining with adhesives is introduced as an additional process step, without changing the established principles and methods of faỗade construction 189 Yvonne Ciupack et al / Procedia Engineering 172 (2017) 186 – 193 2.2 Experimental investigations of bonded steel constructions under static loading 2500 b a Force [N] 2000 1500 1000 500 0 10 15 20 Deformation [mm] 25 Fig (a) Test setup; (b) Test results The application examples from Fig were experimentally investigated, in order to determine the load and deformation behavior, as well as to deduce a Eurocode-based design concept [3] A strip coated trapezoidal profile with a thickness of mm was used for the faỗade connection, and a simplified L-profile with a thickness of mm was chosen to create the connection profile With a two-component copolymer based adhesive a bondline thickness of mm was set The surface was pretreated by blasting with rounded cut wire to improve the adhesion properties of the steel adherents The experimental set up, shown in Fig accurately represents the conditions in the mounted state of the faỗade connection, because the general impact results from wind, acting perpendicular to the bondline During the conduction of the displacement-controlled test, the ends of the trapezoidal element were braced against the machine table All joints failed with a cohesive failure and behaved strongly non-linear (Fig 4b) The bondline allows large deformations, which is positive in regard to advance notice of failure Specific experimental investigations were carried out, in order to evaluate the structural behavior of the faỗade reinforcement A typical hollow section with a width of 60 mm, height of 181.5 mm and a wall thickness of 2.5 mm was chosen Due to the interesting bondline failure, which should be investigated in the tests, the profile length was set to m, based on pilot tests and analytical analysis With a special pneumatic method, which was developed for this application example, a sheet metal steel (width: 50 mm; height: 20 mm) was bonded to the inner side of the hollow profile The pre-treatment of the specimens was defined by blasting the reinforcement with rounded cut wire, as well as cleaning and refining all surfaces with acetone To accurately reflect the install conditions, the load was transferred to the specimen by two double shear connections at every load application point and at the supports These points were designed to allow tension-free horizontal deformations and rotations F w A w w A F/2 90 90 F/2 136.5 136.5 273 136.5 136.5 273 1000 136.5 136.5 273 90 90 181.5 w 218 14 250 200 50 40 w b 710 20 w 44 F w Cross section A-A 10 66 F/2 Force [kN] F/2 120 a 150 100 50 0,0 2,0 4,0 6,0 8,0 Deformation in axis [mm] Fig (a) Four point bending test; (b) Test results 190 Yvonne Ciupack et al / Procedia Engineering 172 (2017) 186 – 193 The evaluation of the experiment shows that the bondline of all specimens failed with a cohesive failure During the tests a distinct leveling of the ultimate load was observed, which develops with the bondline failure Compared to a reference specimen without reinforcement (dashed line in Fig 5b), an improvement of stiffness and carrying capacity can be deduced 2.3 Design of bonded joints in steel structures In a previous research project [3] Eurocode-based design rules for the presented application examples were developed This was realized by a juxtaposition of test results with solutions of analytical models, which are described in [4] The main idea of this procedure is to separate the bonded joint into theoretical components, which can be treated separately Based on the weakest-link-approach, the carrying capacity of the connection or compound can be expressed by the smallest value of calculated resistances of the bondline and the steel adherents Herein the bondline is modeled by spring elements, which can be characterized by small scale specimen tests (see chapter 1.2), representing the specific loading situation of the application example The principle of the component model is shown in Fig Fig Component spring model Meinz [4] developed two analytical models to describe the carrying behavior of the presented applications, based on the theory of elastic supported slabs He assumed the properties of a bondline to be expressed by characteristic values of tension strength (Vk), shear strength (Wk), Young’s modulus (Ek) and shear modulus (Gk) These parameters were determined by butt joint and lap shear tests (see Fig 2) according to the Eurocode requirements and are summarized in table Table Characteristic material properties of different adhesives [N/mm²] Adhesive basis Ek Vk Gk Wk Copolymer 5.64 1.32 0.54 1.32 Epoxy 3063 38.7 308.9 29.3 In accordance to the regulation of the Eurocode concept [5], semi-probabilistic studies for both models were carried out The main objective was to determine partial safety factors JM for the resistance models and to introduce so-called conversion factors to take into account temperature dependent (Kt) and manufacturing dependent effects (Km) Regarding the cohesive bondline failure mode, the design value of the resistance can be expressed with Eq (1) Rd Rk JM ˜Kt ˜Km ˜Ki (1) The calibration process resulted in a partial safety factor for the bonded faỗade connection of 2.16 and for the faỗade reinforcement of 1.56 These values are based on defined reference conditions during the test, which are assumed to be 20°C and mm bondline thickness for the elastic copolymer or 0.2 mm for the epoxy resin based adhesive To consider influences from temperature or changes of the thickness of the adhesive layer, small scale tests were repeated while varying certain test conditions To determine the conversion factors, probabilistic 191 Yvonne Ciupack et al / Procedia Engineering 172 (2017) 186 – 193 conceptions with respect to probability functions are employed The distribution function of resistance must be formulated depending on the variables temperature and bondline thickness This procedure results in conversion factors, which are summarized in table for temperature dependent effects and in table for manufacturing dependent influences Table Conversion factors for temperature dependent effects Copolymer Kt Epoxy -20°C ≤ T ≤ 20°C 20°C < T ≤ 80°C -20°C ≤ T ≤ 20°C 20°C < T ≤ 80°C Shear 1.00 0.31 1.00 0.37 Tension 1.00 0.86 1.00 0.46 Table Conversion factors for manufacturing dependent effects Km Copolymer Epoxy mm ≤ dk ≤ mm 0.2 mm ≤ dk ≤ 0.5 mm 0.5 mm < dk ≤ mm Shear 0.42 0.83 0.45 Tension 0.81 1.00 0.89 As shown within the results, the investigated effects depend on the type of loading Consequently different conversion factors for normal and shear stress conditions are recommended Due to the strong decrease of the material parameters at high temperatures and different bondline thicknesses, the conversion factors are declared for specific ranges 2.4 Adhesive bonded steel structures under cyclic loading As mentioned before, the performance of adhesive bonded joints depends on time and environment influences The carrying and deformation behavior of a bondline is characterized by creep and relaxation effects, affected by temperature and permanently changed by ageing processes The objective of an ongoing research project is to carry out cyclic test procedures for the presented application examples, taking into account these polymer specific properties [6] In a first step, representative load functions for the test sequences are deduced from real measured data Because the studied faỗade connection is mainly stressed by wind and temperature, this investigation focuses on environmental data from different weather stations in Germany The extreme temperatures for a specific period, as well as the variation of thermal impacts over time, are determined by statistical means For the deduction of the faỗade temperature based on measured air temperature, different thermo-physical effects, such as radiation and convection, are considered and shown in Fig 7a The obtained values from this simulation describe a temperature range from -20°C to 80°C, in agreement with Eurocode [7] and the building code for external walls [8] a b Inside Direct and diffuse solar radiation Forced convection Free convection Heat radiation Heat radiation Heat passage Jan Faỗade temperature [C] Outside Feb Mar Apr May June July Aug Sept Oct 90 80 70 60 50 40 30 20 10 -10 -20 Fig (a) Faỗade temperature influencing parameters; (b) Statistically determined temperature load cycle Nov Dec 192 Yvonne Ciupack et al / Procedia Engineering 172 (2017) 186 – 193 A suitable loading cycle (see Fig 7b) was determined based on statistical studies regarding an hourly approach The minimum temperature is assumed to be -20°C (dashed line in Fig 7b) according to the code requirements [7] Thus, for every month a so-called model day is provided, which ranges from the minimum to the maximum temperature The curve in Fig 7b shows the variation of thermal impacts during the monthly model day During the experimental investigations a model day needs to be repeated, according to the number of days a month Moreover, a test procedure for simulation of wind loading is extrapolated from actual measured wind speed data Due to the combination of the typical wind load by a quasi-static mean and a turbulent time-dependent part, the test cycle was developed from measurements with various sampling rates To describe the time steady wind load, the 10 minute mean value of the windiest German weather station in Helgoland (from the archive of “Deutscher Wetterdienst”) is used This is combined with the wind turbulences measured in Karlsruhe [9] with a sampling rate of Hz The composite wind speed function is transferred to a load to number-of-cycles function for a 10 m wide quadratic building, as a model faỗade In the framework of the project, this specific function was analyzed by the rainflow method, which results in an amplitude collective for wind load shown in Fig 1400 Amplitude [N] 1200 1000 800 600 400 200 10 100 1.000 10.000 100.000 1.000.000 10.000.000 Number of cycles, cumulative [-] Fig Amplitude collective for wind loads The aim for the next step, the testing of application examples for cyclic loading, is to simplify the load collective for wind and reduce the testing time by means of endurance strength theory In order to describe the general behavior of bonded joints under cyclic loading, some pre-investigations on small scale specimens and the faỗade connection were carried out under constant amplitudes The stress ratio is chosen with a value of 0.1, which means a varying but permanent tensile loading for the bondline To prevent heating of the adhesive due to mechanical load, the tests were carried out with a frequency of Hz If 2·106 load cycles are reached and no damage is detected, the value of the load is assigned within the fatigue strength The results of these examinations for a polyurethane adhesive are shown in Fig Maximum load [kN] 101 Adhesive basis: Substrat: Load: Polyurethan Steel Axial, sinusoidal Failure probability: 50 % Small scale specimen: Sk = 1.3 Specimen component: Sk = 1.1 100 R = 0.1 1/Ts = 1.11 Small scale specimen R = 0.1 1/Ts = 1.38 Specimen component Without fracture 10-1 103 104 105 Cycles [-] 106 107 Sk Fatigue strength [kN] Fig Comparison of test results of cyclic tests Yvonne Ciupack et al / Procedia Engineering 172 (2017) 186 – 193 All specimens failed with a cohesive failure of the bondline The notch sensitivity (k) is calculated with a value of 13.74 for the butt joint tests and of 6.48 for the faỗade connection The threshold of the load, which the joint can withstand permanently (fatigue limit Sk), is evaluated with a value of 1.3 kN for the small scale specimen and of 1.1 kN for the specimen component Conclusion and Outlook The bonding technology is more than just an alternative joining method for steel structures, which can be shown by various applications and research projects in different fields Despite many advantages of bonded joints the civil engineering and especially steel construction community is reluctant in embracing this innovative technique Often it is justified by doubts about the durability and a lack of experience Interpreting such doubts as open questions and challenges creates the possibility and the potential for exceptional innovations The general interest in developing standards as a basis for analysis and design of adhesive joints in steel, is expected to rise with a growing number of functional and calculable applications The future work for the presented application examples for faỗade structures focuses on experimental investigations under variable amplitudes and the creation of a scientific based guideline for the design of bonded steel joints For the introduction of alternative innovative bonded steel joints, the effort is limited to the development of engineering-models, planning and construction design General technical approvals or individual approvals can be achieved more easily, thus sustainably increasing the innovative capacity of small and medium-sized enterprises Acknowledgements The IGF research project (IGF-No 18161 BG) of the Research Association for Steel Application (FOSTA), Sohnstraße 65, 40237 Düsseldorf, has been funded by the AiF within the program for sponsorship by Industrial Joint Research (IGF) of the German Federal Ministry of Economic Affairs and Energy, based on an enactment of the German Parliament References [1] DIN EN 14869-2, Structural adhesives - Determination of shear behaviour of structural bonds, –Part 2: Thick adherends shear test (ISO 11003-2:2001, modified), German version EN 14869-2:2004 [2] DIN EN 15870, Adhesives – Determination of tensile strength of butt joints (ISO 6922:1987 modified), German version EN 15870:2009 [3] K Dilger, H Pasternak et al., IGF Project No 16494 BG, Entwicklung eines Eurocode-basierten Bemessungskonzepts für Klebverbindungen im Stahlbau (in Anlehnung an DIN 1990), Germany, 2012 [4] J Meinz, Kleben im Stahlbau – Betrachtungen zum Trag- und Verformungsverhalten und zum Nachweis geklebter Trapezprofilanschlüsse und verstärkter Hohlprofile in Pfosten-Riegel-Fassaden, Dissertation, Brandenburgische Technische Universität Cottbus, Germany, 2010 [5] DIN EN 1990, Eurocode 0: Basis of structural design; German version EN 1990:2002 + A1:2005 + A1:2005/AC:2010, 2010 [6] K Dilger, H Pasternak et al., IGF Project No 18161 BG, Untersuchungen zum Tragverhalten und der Lebensdauer von Klebverbindungen im Stahlbau unter zyklischer Belastung, Germany, 01.04.2014 - 30.09.2016 [7] DIN EN 1991-1-5: Eurocode 1: Actions on structures – Part 1-5: General actions – Thermal actions; German version EN 1991-1-5:2003 + AC:2009, 2010 [8] DIN 18516-1, Cladding for external walls, ventilated at rear –Part 1: Requirements, principles of testing, 2010 [9] KIT, Institut für Meteorologie und Klimaforschung (2015): http://imkbemu.physik.uni-karlsruhe.de/~fzkmast/ 193 ... questions and challenges creates the possibility and the potential for exceptional innovations The general interest in developing standards as a basis for analysis and design of adhesive joints in steel, ... Uneven and combinated load Shear and peel: "bad" Fig Influence of geometric design and loading of a bondline Moreover, the kind of loading has a big influence on the carrying capacity of a bonded steel. .. Procedia Engineering 172 (2017) 186 – 193 Thus, cleaning and degreasing of steel surfaces are essential during the bonding process In addition, mechanical pre-treatments such as blasting can improve