Proceedings. W056 - Special Track 18th CIB World Building Congress May 2010 Salford, United Kingdom

Transcription

1 Proceedings W056 - Special Track 18th CIB World Building Congress May 010 Salford, United Kingdom CIB W056 - Sandwich Panels CIB Publication 34

2 CIBWORKINGCOMMISSION W056 SANDWICHPANELS PAPERSANDPOSTGRADUATEPAPERSFROMTHESPECIALTRACK HELDATTHECIBWORLDBUILDINGCONGRESS010,10 13MAY010 THELOWRY,SALFORDQUAYS,UNITEDKINGDOM SelectedpapersfromtheProceedingsofthe18 th CIBWorldBuildingCongress. Proceedingseditedby:ProfessorPeterBarrett,ProfessorDilanthiAmaratunga,Dr.Richard Haigh,Dr.KaushalKeraminiyageandDr.ChamindaPathirage W056SpecialTrackPapers(excludingPostgraduatePapers)reviewedby:ProfessorJose AmorimFaria,ProfessorJukkaAalto,ProfessorJörgLange,Dr.LarsHeseliusandProfessorJ. MichaelDavies CIBPublication34

3 W056 SANDWICHPANELS PAPERSANDPOSTGRADUATEPAPERSFROMTHESPECIALTRACK The objectives of this Commission are to exchange information and coordinate research programmesinalltechnicalaspectsoflight weightconstructionsusingtraditionalandnew buildingmaterialsandtopreparesynthesisedreportsonmattersofparticularinterest.

4 CONTENTS Papers StabilisationofBeamsbySandwichPanels NewRegulationsandRecent 1 ResearchResults Misiek,T.Kapplein,S.Durr,M.Saal,H. AUnifiedApproachfortheLocalBucklingofSandwichPanelsandTrapezoidalSheeting 14 Misiek,T.Hassinen,P. ECCS/CIBJointCommitteeonSandwichConstructions,RecentEuropean 7 RecommendationsonDesignandTesting Davis,J.M.Hassinen,P.Heselius,L.Misiek,T. MethodstoMeasuretheDurabilityofStructuralSandwichPanels 38 Hassinen,P.Pfeiffer,L. ReliableCompositeRoofing:LearningfromExperience 51 Roberts,K. PostgraduatePapers OpeningsinSandwichElements 61 Warmuth,F.Lange,J. OptimizationofGeometryandCoreMaterialsofSandwichPanelswithMetallicFaces 73 Kurpiela,A.Lange,J.Berner,K. CIBBrochure 85 Disclaimer 87

5 Stabilisation of Beams by Sandwich Panels New Regulations and Recent Research Results Misiek, T. Karlsruhe Institute of Technology, Karlsruhe, Germany ( Käpplein, S. Karlsruhe Institute of Technology, Karlsruhe, Germany ( Dürr, M. Montana Bausysteme AG, Villmergen, Switzerland ( Saal, H. Karlsruhe Institute of Technology, Karlsruhe, Germany ( Abstract Sandwich panels are modern pre-fabricated construction components used as cladding elements for different types of buildings. Sandwich panels consist of an insulating core material covered by two faces which are typically made of thin metal sheets. In standard applications, the panels are mounted and fixed on a load-bearing substructure of beams or purlins. Sandwich panels can reduce the problem of lateral torsional buckling of this substructure of beams or purlins by providing stabilization either by shear stiffness or by torsional restraint. The new edition of the German standard for the design of steel structures DIN gives formulae for the calculation of the stiffness of the torsional spring for restraint of the substructure under vertical downward loading. These new regulations are based on experimental investigations and parametric finite element analyses. These formulae only apply for sandwich panels with steel facings and polyurethane and mineral wool as core material. The paper explains the load-bearing mechanisms of the stabilisation effect by the sandwich panel. A mechanical model is developed for extending the range of application of the design formulae and to include the effects of creep and elevated ambient temperature. It presents the new regulations of DIN and explains the tests on which these regulations are based.the spectrum of applications not yet examined is investigated by tests and accompanying numerical calculations within the framework of the EASIE project. As a result of these investigations the torsional restraint of panels with facings made of aluminium and glass fibre reinforced plastics (GFRP) and with cores made of EPS are dealt with. The load case wind suction is discussed in addition. The increase of the torsional restraint obtained by fixing roof panels at the upper flange (which is mainly with saddle washers) is also explained and quantified by these investigations. Keywords: sandwich panels, lateral-torsional buckling, torsional restraint 1

6 1. Stabilising effects on beams Sandwich panels increase the resistance of substructures (beams, purlins) against lateral torsional buckling by restraining the lateral displacements and rotations. The high in-plane shear stiffness of sandwich panels can be used for stabilizing the lateral displacement of the substructure and thus preventing lateral torsional buckling of the substructure. This type of stabilization requires the exact knowledge of the in-plane shear stiffness. Special considerations are necessary for the design of the fastenings because the flexibility of the connection to the substructure and that of neighbouring panels to each other reduces the effective shear stiffness significantly. The torsional restraint is governed by the stiffness of the connection of the sandwich panel to the substructure. Recent research carried out by Dürr (008) showed that this stiffness significantly depends on the load transferred by the sandwich panel to the substructure. Dürr (009) gives formulae for calculating the parameters of this moment-rotation-relation for sandwich panels under deadweight loading and with two different core materials. So far only connections through the lower flange of the outer face with two fasteners per element have been investigated. Other types of connections (e.g. connection through upper flange of the outer face with saddle washers) and different core materials are important yet unknown parameters of the moment-rotation-relation. The focus of the present paper is on the stabilisation of beams by torsional restraint.. General description of the effects.1 The spring stiffnesses The torsional restraint by sandwich panels can be calculated by using the mechanical model of a torsion spring with the spring stiffness c. This spring stiffness is a combination of the bending stiffness of the attached panel c C, the stiffness of the connection c A and the distortional stiffness c B of the beam to be stabilised. The stiffnesses c C and c B depend on the geometry of the sandwich panels and type of beams used. They can be easily calculated. We will focus on the stiffness c A of the connection between the sandwich panel and the subjacent beam because this is the weak link dominating the value of the combined stiffness c of the chain of springs. In the following text, the stiffness c A will be simply denoted as c to ease reading and to reduce the number of subscripts.

7 c Figure 1: Stabilisation: torsional restraint. Sandwich panels with deadweight loading Figure shows a generalised moment-rotation-relation for the spring stiffness of the connection of a sandwich panel under deadweight loading. m c m K c 1 c 1 c 0 m K Figure : Generalised moment-rotation-relation for deadweight loading In this generalised relation we assume that all of the fasteners are mounted on one side of the web as shown in Figure. Here, the positive direction of rotation is defined as an anticlockwise rotation. We can differentiate three parts of the moment-rotation-relation. For small rotations, there is the value c 1. The load q acting on the panel is always transferred by contact from the inner face of the panel to the upper flange of the beam. The rotational stiffness only depends on the width of the flange and the indentation stiffness (Figure 3). This indentation stiffness is dominated by the compression stiffness E Cc of the core material. The rotational stiffness does not depend on the position of the fasteners because the fasteners are not activated in this situation. 3

8 Figure 3: Mechanical model for c 1 The area of contact decreases, with increasing rotation until it is reduced to the final contact line with the outer edge of the flange. At this stage, the restoring moment is the contact moment m K q b When the deflecting moment to be stabilized exceeds the contact moment m K the value c applies. At this stage, tensile forces in the fasteners are activated. These tensile forces F t cause an indentation u w of the fasteners heads and washers into the outer face of the sandwich panel. This additional deformation decreases the stiffness significantly: The value c is significantly smaller than the value c 1. b K u w Figure 4: Mechanical model for c positive rotations The value c depends on the indentation stiffness k w of the fastener and the indentation stiffness k f at the line of contact at the outer edge of the flange. This stiffness depends on the direction of rotation as defined in Figure with regard to the position of the fastener and the distance b K of the fastener from the contact line as defined in Figure 4. For positive rotations we still have a distinct value of c, while for negative rotations c is comparatively small because of the small distance b K and the small corresponding contribution to the restoring moment. With an alternating fixing pattern the values c are the same for both directions of rotation, provided b K is the same for both directions of rotation. However, due to the aforementioned influence of the indentation stiffness of the fasteners head, the total number of fasteners has to be doubled. If not, the value c reduces to half of the value. 4

9 .3 Sandwich panels with uplift loading Figure 5 shows a generalised moment-rotation-relation for the spring stiffness of the connection of a sandwich panel subjected to uplift loading. The same assumptions regarding fasteners position and direction of rotation apply as for deadweight loading. m c 1 0 c 1 0 K c Figure 5: Generalised moment-rotation-relation for uplift loading For small values of the rotation, there is no torsional restraint of the beam by the sandwich panel. This is due to the gap between the upper flange of the beam and the adjacent face of the panel (Figure 6). This gap is caused by the indentation of the fastener s head and washer when the panel is subjected to uplift loading and the fasteners are therefore loaded with tensile forces. The tensile load induces a torsional moment and (negative) rotation in the beam, giving a preferential direction for lateral torsional buckling and enforcing the tendency for lateral-torsional buckling of the beam. b K u w Figure 6: Mechanical model for c 1 0 After reaching a rotation of = K, the stiffness increases. At this stage, the gap between the outer edge of the flange of the beam and the adjacent face is closed. Therefore for this load case the governing parameter is the contact rotation K K u b w K k w F t b K 5

10 with F t being the tensile forces in the fasteners caused by the uplift loading and k w being the indentation stiffness of the fastener. For negative rotations, the rotational stiffness remains approximately zero because contact occurs only after very large rotations due to the small lever arm b K. b K u w Figure 7: Mechanical model for c 0 A significant increase in stiffness can only be found for large positive rotations. Due to the direction induced by the tensile load in the fasteners this is a rather theoretical case. The mechanical model is the same as for deadweight loading (Figure 4), but with a different direction of rotation. In this case, the values c are the same for uplift and deadweight loading but can not be taken into account. In practice, there is no torsional restraint with uplift loading for applications with all of the fasteners mounted on one side of the web. The use of an alternating fixing pattern is possible, too. With an alternating fixing pattern the values c are the same for both directions of rotation. However, due to the aforementioned influence of the indentation stiffness of the fasteners head, with the equal partitioning of the fasteners to both sides of the web c reduces to half of its value. With the alternating fixing pattern the following disadvantage is avoided which occurs with the arrangement of the fasteners on one side of the web: The tensile forces in the fasteners resulting from the uplift loading induce a torsional moment enforcing the tendency for lateral-torsional buckling of the beam. 3. Regulations and standards: The new German design code Most recently, the German design code for steel structures, DIN , was updated, now including the possibility to use sandwich panels for the stabilisation of beams against lateral-torsional buckling. These regulations are based on the investigations described in Dürr (008) which received financial support by the IFBS. The basic construction of the formulae (derivation of the influencing parameters) for c 1 and c was derived from an FE-analysis whereas the parameters c 1 and c were derived from a statistical evaluation of test results. These regulations are only given for downward loading and for the core materials PUR and mineral wool. Also, the parameter range for E C and t K is restricted. A secant value of 6

11 c can be taken into account, using the simplified moment-rotation relation shown in Figure 8. The necessary values and parameters are given in the following tables. mk m K Figure 8: Moment-rotation-relation Table 1: Values c 1 and c c 1 c m K Double-symmetric beams with 60 mm 100 mm b 8 b Z- or C-section with 60 mm b 80 mm c1 E c E C 1 C c E t b 8 C K 0 b q d q d b Table : Parameters c 1, c parameter according to Table 3.0 N/mm² E C 6.0 N/mm² Young s modulus of the core material b [mm] width of the flange of the beam 0.4 mm t K 0.67 mm sheet thickness of the outer face layer q d parameter depending on the pattern of fixings = 1.0 alternating application of fixings = 1.5 one-sided application of fixings = 0.0 hidden fixings design value of the downward load to be transferred from the panel to the beam 7

12 Table 3: Parameters c 1 and c Core material Geometry of outer face (at the head of the fasteners) c 1 c PUR/EPS Mineral wool profiled mm² mm lightly profiled/flat mm² mm profiled mm² mm lightly profiled/flat mm² mm alternating application of fixings one-sided application of fixings Figure 9: Fixing patterns 4. Further developments in the EASIE project 4.1 Mechanical model Two simplified mechanical models and an idealised moment-rotation-relation were used to model the compliance between the beam and the sandwich panel. Figure 10 shows the models for the values c 1 and c.the model for c 1 consists of a single spring for taking account of the indentation at the edge of the flange of the beam. Using the contact moment we obtain c 1 k 1 b 4 8

13 The spring stiffness k 1 has the unit force per square of length because k 1 is referring to the length of the flange. k 1 is predominantly depending on the Young s modulus E C of the core. For the value c a model with two springs is used. The second spring is the indentation of the screws and the washers into the outer face. We obtain c b 1 k K 1 k 1 n b 1 K K 1 k 1 The indentation of the fasteners is the dominating effect and therefore we can simplify this equation to c n K b K While the spring stiffness k also has the unit force per square of length, K has the unit force per length. k is dominantly depending on the Young s modulus E C of the core and the number n of fasteners per length. value c 1 value c b k = f (n, E C ) b k 1 = f (E C ) w 1 = q k 1 w k 1 = f (E C ) w 1 b b b K Figure 10: Mechanical models for the values c 1 and c 4. Numerical investigations Numerical investigations were performed to verify the mechanical model and to study the influence of different parameters. These parameters were the essential dimensions and the Young s modulus of the core and the face material. We obtained the following results. 9

14 - For panels with two flat or lightly profiled faces (wall panels) both c 1 and c depend on the thickness D of the panels. c 1 and c increase with the thickness D. They converge to the value of the panel with a strongly profiled outer face (usually roof panels) with similar arrangement of the fasteners. - Both values c 1 and c increase with the depth of profiling of the outer face due to the support by this profiling. Beyond 10 mm of depth of this profiling no further increase of this is possible such that c 1 and c there attain their limit values for a panel with a profiled outer face. This applies for the investigated geometry with a distance of ribs of 333 mm. - Both c 1 and c increase with Young s modulus E C of the core material with the power of 0.9. The approximation by a linear function is justified. - c 1 increases with the bending stiffness (EI) F of the inner face with the power of 0.1. The influence of this stiffness can therefore be neglected for the common parameter range (faces made of steel with thickness 0.38 mm t K 0.,71 mm, faces made of aluminium with 0.50 mm t 0.65 mm). For faces made of GFRP a reduction factor c F is required. As expected there is no influence of (EI) F1 of the outer face on c 1 such that (EI) F1 can also be disregarded. c increases with the bending stiffness (EI) F1 of the outer face with the power less than 0.1 so the same applies as for c 1. There is no significant increase of c with increasing bending stiffness (EI) F of the inner face. This justifies the mechanical model introduced above that c only depends on the core material and the type of profiling of the outer face. - c 1 does not increase with the square of b but with the power of 1.3 (thin wall panels) to 1.7 (thick wall panels): The actual lever arm is smaller than b because of the indentation and because of bending of the panel. c increases with the square of b. 4.3 Evaluation of tests results Based on the aforementioned considerations we used an approach of the form c 1 c1 EC b to determine the value c 1 and an approach of the form c c n E C b K to determine c. The number n of fasteners per meter length and the distance b K depend on the fixing pattern and the direction of rotation. c should be set to zero unless b K 0.5 b. For double-symmetric beams with one sided fixing at one fourth of the flange b K = 0.75 b or b K = 0 applies, depending on the direction of rotation. For one-sided application of fixings this means that for one direction of rotation c is always zero. In principle the same is true for alternating fixing patterns: b K is always 10

15 the longer lever arm and for n only the number of fasteners per length corresponding to this lever arm can be taken into account. Attention has to be paid to the units: c 1 is non-dimensional value whereas c has the dimension meter because n as defined above has the dimension m -1. E C and b K are input with their units. Creep tests were performed and evaluated to consider the effect of duration of loading on Young s modulus E C of the core material. The extrapolation of the results of these tests according to EN to 000 hours (representing snow loading) and hours (representing self-weight loading) resulted in values C,t much higher than the value C, = 1.0 given in Dürr (008). This increase to almost twice the value is mainly due to the statistical evaluation and the scatter of the test results. Finally the values C,t are provided for use with the formula E C, t 1 E C C, t The effect of temperature on Young s modulus E C of the core material can be taken into account by using the reduction factor k 1 according to EN 14509: E C, E C, k 3 1 E C, E E Ct, 80 C Ct, 0 C From the tests with sandwich panels with faces made of GFRP additional reduction factors c F = 0.38 for c 1 and c F = 0.41 for c were obtained. We recommend to use c F = 0.38 both for c 1 and c. The final result is summarized in Table 4. The values listed for panels with a profiled outer face can be used for panels fixed in the lower flange of the outer face or in the upper flange and also for fixing with or without saddle washers as well because the differences of values found both in numerical and experimental investigations were to small to be seriously quantified. Table 4: Values c 1 and c c 1 c Double-symmetric beams Z- or C-section c 1 cf EC, t, b c1 cf EC, t, b c c n E b 0 F C, t, K E C,t, E C, t, 1 E C C, t k E C C, t E E Ct, 80 C Ct, 0 C m K b q d q d b Table 5: Parameters 11

16 c 1, c parameter according to Table 6 c F C,t b [mm] b K [mm] n [m -1 ] q d parameter depending on the pattern of fixing c F = 1.00 c F = 0.38 face materials steel and aluminium face material GFRP parameter depending on the pattern of fixing C,000 = 1.9 core materials PUR and EPS C, = 1.83 core materials PUR and EPS C,000 = 1.35 core material mineral wool C, =.31 core material mineral wool width of the flange of the beam distance between governing line of fixing and contact line number of fasteners per meter length in the governing line of fixing (n = 0.0 for hidden fixings and for b K < 0.5 b) design value of the downward load to be transferred from the panel to the beam Table 6: Parameters c 1 and c Core material PUR/EPS Mineral wool Geometry of outer face (at the head of the fasteners) c 1 c profiled m lightly profiled/flat m profiled m lightly profiled/flat m Table 7: Application range 60 mm b 180 mm for double-symmetric beams 60 mm b 80 mm Z- or C-sections.0 N/mm² E C 8.0 N/mm² Young s modulus of the core material 0.38 mm t K 0.71 mm sheet thickness of the face layers (steel) 0.50 mm t 0.65 mm sheet thickness of the face layers (aluminium) 1.7 mm t.0 mm sheet thickness of the face layers (GFRP) 1 m -1 n 4 m -1 number of fasteners per meter length in the governing line of fixing 1

17 5. Conclusion Sandwich panels increase the resistance of substructures (beams, purlins) against lateral torsional buckling by restraining the lateral displacements and rotations. The torsional restraint by sandwich panels can be calculated by using the mechanical model of a torsion spring with the spring stiffness c. This spring stiffness is a combination of the bending stiffness of the attached panel c C, the stiffness of the connection c A and the distortional stiffness c B of the beam to be stabilised. The new rules given in the German design code DIN for the calculation of the stiffness of the connection c A which are based on investigations by Dürr (008) are presented and their range of application is extended. The investigations for this extension were performed within the framework of the EASIE project. Acknowledgements This research has received financial support from the Industrieverband für Bausysteme im Metallleichtbau (IFBS), the German association of the producers of sandwich panels and thin-walled profiles. This paper also presents results obtained from the research of the EASIE project. The EASIE project has received financial support from the European Community s Seventh Framework Programme FP7/NMP-SE-008 under grant agreement No We express our sincere gratitude for this support. References Dürr M (008) Die Stabilisierung biegedrillknickgefährdeter Träger durch Sandwichelemente und Trapezbleche Berichte der Versuchsanstalt für Stahl, Holz und Steine der Universität Fridericiana in Karlsruhe, 5. Folge Heft 17. Karlsruhe 008. Dürr M and Saal H (009) Die drehbettende Wirkung von Sandwichelementen beim Biegedrillknicknachweis in der Neufassung der DIN Bauingenieur 84: DIN :008-11: Stahlbauten Teil : Stabilitätsfälle Knicken von Stäben und Stabwerken EN 14509:006: Self-supporting double skin metal faced insulating panels Factory made products Specifications 13

18 A Unified Approach for the Local Buckling Of Sandwich Panels and Trapezoidal Sheeting Misiek, T. Karlsruhe Institute of Technology, Karlsruhe, Germany ( Hassinen, P. Faculty of Engineering and Architecture, Helsinki University of Technology, Espoo, Finland and Pontek Consulting Engineers Ltd, Espoo, Finland ( Abstract Failure of thin-walled building components like sandwich panels or trapezoidal sheeting is normally initiated through a local buckling of the plane elements of the cross-section. For trapezoidal sheeting, EN gives an equation for the determination of the effective width of these plane elements and thus for the calculation of the load-bearing capacity of these components. The cross-sectional parts of a lightly or strongly profiled facing of a sandwich panel can be regarded as elastically supported plane elements, whereas the elastic support is provided by the core material. For the determination of the load-bearing resistance of sandwich panels with lightly or strongly profiled facings, a calculation procedure to determine the effective width of the elastically supported plane elements is needed. Some approaches for the calculation of the effective width already exist. The papers published so far are using a modified buckling coefficient for the calculation of the buckling strength, following Winter s approach such as given in EN Because no generally accepted design procedures exist, the load-bearing resistance of sandwich panels is determined experimentally. Based on the basic principles of structural stability, the buckling strength of the elastically supported plane element can be calculated, taking into account the material properties of the core material and the associated buckling wavelength for minimum buckling strength. Then the design procedures of EN for thin plate buckling of trapezoidal sheeting can be used, expanded by the procedures of EN for taking into account the column type buckling behaviour for buckling wavelengths smaller than the total width of the plane element. Comparison of the test results with different arrangements to the calculated values shows a good consistency. For lightly profiled faces, depending on the depth of the profiling, failure will finally take place through a plate buckling of the plane elements or by a column buckling of the stiffening profiles, whereas the latter failure mode is looking similar as the wrinkling failure of a flat facing. Comparison of the experimental results obtained with different test arrangements to the calculated values show the present limits of the applicability of the proposed design procedure. The differences in failure modes are discussed. Keywords: sandwich panels, face layer, buckling, elastic foundation, elastic half-space 14

19 1. Introduction The determination of the load-bearing capacity required for the design of sandwich panels is to a large degree based on test results. In contrast, there is a large number of references for the determination of the load-bearing capacity of the trapezoidal sheeting, documented in national standards such as the Swedish StBK-N5 or the German DIN as well as international standards such as EN This is astonishing because both building components are com-parable in materials, geometry and load-bearing behaviour. Compared to trapezoidal sheeting the foam core provides an additional elastic foundation to the plane thin-walled elements. In the following we will show that the calculation procedures developed for trapezoidal sheeting can be modified to be applied for sandwich panels. This will start with the basic module, of which all cross-sections consist: a simple plane element, like a flange of a strongly profiled face of a sandwich panel.. Theoretical background and standardized design procedures The local buckling of the plane cross-sectional elements shall be taken into account in the determination of the load-bearing capacity of the thin-walled building components. Because of the local buckling of the plates of a medium or high slenderness only an effective width, which is smaller than the total width, can be taken into account. x b a y EI F = f (E F, t, F ) x z c = f (E C, G C, C ) Figure 1: Plate on elastic foundation The calculation of the effective width is based on the critical buckling stress cr. An additional support provided by an elastic foundation increases the critical buckling stress cr. This increase is an 15

20 addition to the local buckling stress of the plate without foundation calculated according to EN The increase of the critical buckling stress leads to a decrease of the relative slenderness of the plate and thus to an increase of effective width and load-bearing capacity. Up to now, most of the present studies capture the effect of the elastic foundation by adjusting the buckling value k as proposed by Davis, Hakmi and Hassinen (1991). This adjustment is based more on statistical evaluations of test results. The determination of the critical buckling stress can be based on the elastic potential, which is shown in Timoshenko and Woinowsky-Krieger (1959) for the plate without foundation. The calculation of the stiffness of the Winkler s foundation out of the material parameters E C and G C of the core material is shown in Stamm and Witte (1974). We obtain the following equation: cr, p 1 E 1 F t F a m m a n b 1 C 1 3 C 4 C E C t a m m a n b with E C EC GC 1 C according to Stamm and Witte (1974) to take into account very approximately the anisotropy of the material properties of core materials such as for example polyurethane and polystyrene foam. The effective width of the plate can be calculated from the critical buckling stress and the relative slenderness. EN gives the following expressions: p f y cr, p 1,0 beff, p 1 0, p b p p p for p p 0,673 0,673 These equations derived by Winter assume pure plate-like behaviour. This can be found for ratios = a/b > 1,0. Because of the elastic foundation we obtain a buckling wave pattern with an aspect ratio a/b which is smaller than 1,0. This has to be taken into account when calculating the critical buckling stress, leading to an increase of the buckling coefficient k. Unfortunately this also results in a minor support of the mid-part of the plate by the supports at the longitudinal edges compared to buckling patterns with higher aspect ratios. Therefore the load-bearing capacity is lower than the one calculated according to Winter s equation. Column-like buckling behaviour of the plate has to be taken into account. This can be done by utilising the procedure given in EN In this case an interpolation between the plate buckling curve according to Winter and the column buckling curve c 16

21 is done. This interpolation is based on the ratio of the elastic critical buckling stress of the plate and the column: c p c c with cr, p 0 1,0 where 1 cr, c and cr, c 1 E 1 F t F m a 1 C 1 3 C 4 C E C t a m c f y cr, c 0,5 1 0,1 c 0, c c 1 c 1,0 3. Comparison with results from numerical calculations and tests with flat plates The presented approach for the determination of the load-bearing resistance of a flat plate on an elastic foundation was verified using finite-element-calculations and comparisons with test results. The plates were modelled using four-node shell elements with a linear-elastic ideal-plastic stressstrain relationship. For the elastic foundation, spring elements with a linear elastic stress-strain relationship were used. The parameters such as sheet thickness t, yield strength f y and stiffness of the foundation c (via E C, G C and C) were varied. Figure shows the results of the comparison between the theoretically derived critical buckling stress and the buckling stress based on FE-calculations. Using this FE-model and the initial imperfection corresponding to the first eigenvalue of the model, the ultimate load has been computed. The initial imperfection was scaled to w 0 = 0,1 t. The comparison of the computed results with the results of the procedure based on the interpolation between the plate buckling curve according to Winter s equation and the column buckling curve c is shown in Figure 3. 17

22 cr,p,fem [N/mm²] cr,p,calc [N/mm²] Figure : Comparison of critical buckling stress based on the finite-element-calculations and on the analytical expressions 1,0 0,8 c,fem [-] 0,6 0,4 0, 0,0 0,0 0, 0,4 0,6 0,8 1,0 c,calc [-] Figure 3: Comparison of the results of finite-element-calculations with results of the calculation procedure: normalised effective width c To verify the calculation procedure, a series of tests was performed. These tests were done with trapezoidal sheeting (five ribs) and with cross-sections cut out from the sheeting (one rib). The trapezoidal sheeting was the upper face of a panel with one strongly profiled face (Figure 4). The 18

23 inner flat face was removed. Thus, the trapezoidal sheeting provided the only load-bearing component of the specimen. Due to the remaining foam core, the plane elements of the face still were supported by an elastic foundation. The normalized effective width c was back-calculated using the calculation procedure for trapezoidal sheeting according to EN , supplemented with the procedures of EN mm 3 mm 114 mm t F1 = 0,47 mm 35 mm 61mm Figure 4: Trapezoidal sheeting with elastic foundation no face According to the calculation procedure, the aspect ratio of the buckling waves should be about = 0,3. Figure 5 shows the buckling waves in the upper flange confirming the calculated value. Figure 5: Local buckling of the upper flange An additional series of tests was done with the upper face layer without any foam material. The geometry and material was identical to the one used in the tests described above. In this case no elastic foundation for the plate is available and we have a cross-section of a standard trapezoidal sheeting. Results of the tests are shown in Figure 6. In addition, the tests results of Davis and Hakmi (1991) and the results of Pokharel and Mahendran (003) were used for comparison (Figure 6). Davis and Hakmi performed bending tests on C-shaped 19

24 members with a foam core. Pokharel and Mahendran glued flat sheets of different thicknesses and steel grades to layers of plastic foam material. 1,0 0,8 c,test [-] 0,6 0,4 0, 0,0 Davis and Hakmi 1991 Pokharel and Mahendran 003 trapezoidal sheeting with elastic foundation trapezoidal sheeting without elastic foundation 0,0 0, 0,4 0,6 0,8 1,0 c,calc [-] Figure 6: Comparison of calculation procedure with test results: normalised effective width c All the test results show a good correlation with the calculated values. The calculated values make the lower boundary to the test results. Interestingly the values with the highest discrepancy between test and calculation being on the unsafe side are the ones for the standard trapezoidal sheeting without any elastic foundation. This is a result of the very high slenderness of the upper flange, the relative slenderness being of the order of 5,5 to 5,6. It is interesting to discuss about the effect of the material properties used in the calculation procedure. The first difference arises in the test method for the determination of the shear modulus G C. Davis and Hakmi performed tests with small cubes of the core material which were loaded with a pure shear load. No information is available about the tests of Pokharel and Mahendran. However, it can be assumed that the tests were done in a similar way based on small cubes. Our tests for the determination of the shear stiffness were performed with small short-spanning beams according to EN The latter test method might lead to higher discrepancies between the measured value and the actual material properties, resulting in a higher discrepancy between the calculated and experimental values c,calc and c,test. The elastic modulus E C of the core material is normally determined as an average value corresponding to the total thickness of the specimen, determined separately in a compression and tensile test. For plastic foam materials made in a continuous process the local values strongly deviate from this mean value, depending on the relative position of the sample in the thickness direction of the panel. For these sandwich panels, investigations on the distribution of the stiffness presented in 0

25 Dürr (008) show that the elastic modulus near to the faces is higher than in the centre plane of the panel. In addition, different values of the modulus can be obtained close to the faces depending on their position (top or bottom) during the production. These effects have to be kept in mind when looking for the results of our tests (which were done with real panels) and presumably also for the tests of Davies and Hakmi. For the tests of Pokharel and Mahendran which were performed with foam material glued to the face this effect can be neglected. In fact, when calculating the load bearing resistance with the values from the tests, calculated values are on the safe side, because the higher stiffness near the faces gives a stiffer elastic foundation than assumed by using the mean values over the total core thickness. 4. Strongly profiled sandwich panels In addition to the buckling tests with trapezoidal sheeting, tests with sandwich panels were also performed. The panels had a similar geometry as the specimens introduced before. In these tests, the inner face was not removed, so real sandwich-type load-bearing behaviour existed. The results of the tests were back-calculated, too. Now, the lower flange of the outer face is also under compression loading. This can be easily seen by comparing the pictures of this flange taken after the failure (Figure 7). Figure 7: Upper and lower flange of the specimens after failure: trapezoidal sheeting with elastic foundation (left) and sandwich panel (right) The effective cross-section of the outer face can be calculated with the introduced procedure. At first, the effective cross-section of the outer face under compression has to be determined by strictly following the calculation procedure introduced in chapter (approach A). In fact, there will be always a difference between the calculated effective width and the test results. These differences in the determination of the effective width will cause further differences in the calculation of the loadbearing capacity of the sandwich cross-section. In our case, we have also the possibility to use the effective width of the upper flange from the tests on tests described in chapter 3 (approach B) to minimize the effect of the former differences. The use of both the both approaches allows us to compare the results. The calculation has to be done iteratively. For the first iteration we assume a 1

26 constant compression stress distribution over the whole cross-section of the outer face. We obtain the following results: 0 mm 3 mm 114 mm 11,1 mm 35 mm 11 mm 11,1 mm Approach A: 15,4 mm Approach B:,8 mm 11 mm Figure 8: Effective cross-section of the outer face after the first iteration Table 1: load-bearing capacity of the sandwich panel after the first iteration Approach A Approach B Test result Ultimate bending moment 5,7 knm 6,74 knm 6,94 knm According to this calculation, the compression stress in the lower flange reaches a value of = 148 N/mm² which is below the yield strength. The effective width in the lower flange is larger than assumed at first. Same is true for the lower part of the webs. Therefore the effective cross-section is larger than calculated in this first step. The accuracy of the calculated load bearing capacity can be improved iteratively. For the lower flange, this can be done by calculating the effective width with = 148 N/mm² instead of f y = 400 N/mm². Special considerations are required for the web. We can use the simple assumption given in EN that the effective width at the lower edge is 1,5-times the one at the upper width were = f y. After the first iteration, the stresses in the lower flange are further decreasing, and the stresses in the inner face increase. Load-bearing capacity is increasing with every iteration. However, the values are converging very soon. Stress distribution based on the fifth iteration can be accepted to represent the ultimate load: 0 mm 3 mm 114 mm 11,1 mm 35 mm 16,7 mm Approach A: 15,4 mm Approach B:,8 mm Approach A: 0,0 mm Approach B: 19,0 mm Approach A: 0,0 mm Approach B: 19,0 mm Figure 9: Effective cross-section of the outer face after the fifth iteration

27 Table : Load-bearing capacity after five iterations Approach A Approach B Test result Ultimate bending moment 5,94 knm 7,14 knm 6,94 knm This is a considerable improvement compared to the results of the first iteration. However, in the example we have a difference between the values of the pure calculational approach A and the test result. The difference is in the same order as for the simple flat plate element. 5. Resistance of lightly profiled faces A local buckling phenomena can also be observed in tests with panels having a slightly profiled face. Figure 10 shows the face of a lightly profiled sandwich panel (depth of lining approx.,0 to,5 mm) in a bending test before reaching the failure load. Finally, failure will occur by global buckling or wrinkling of the face over the complete width of the panel, depending on the support conditions at the longitudinal edges of the compressed face. Figure 10: Local buckling of the slightly profiled face of a sandwich panel For an extension of our approach to lightly profiled faces on an elastic support, the results of the buckling tests of Pokharel and Mahendran (005) were used. Pokharel and Mahendran (005) varied the thickness of the faces and the width b of the plane elements of the profiling. The total dimensions of the faces were 400 mm and 100 mm in the width and length directions, respectively. All the edges were fully supported. The effective width of the plane elements between the stiffeners can be calculated according to the presented calculation procedure. After incorporating the restrictions in EN requiring some reductions in width also the cross-sectional values of the stiffeners itself can be calculated. The part of the critical buckling stress provided by the stiffeners can now be calculated using 3

28 cr, s 1, 8 E b I s t b 3 e Assuming a global buckling failure of the stiffened plate and recalculating the tests with the equations introduced in section of this paper, the results shown in Figure 11 were obtained. 40,0 3,0 4,5 mm 4,5 mm 13 mm 1,0 mm b N u,test [kn] 4,0 16,0 8,0 b = 78,5 mm b = 8,5 mm 0,0 0,0 8,0 16,0 4,0 3,0 40,0 N u,calc [kn] Figure 11: Comparison of results of calculation procedure with test results It can be shown that for the geometry investigated in Pokharel and Mahendran (005) the effects of the stiffeners can be neglected and the support at the longitudinal edges as well as the elastic foundation of the core material are the dominating stabilising effect. In this case, failure occurred by local buckling of the entire stiffened plate. The overall width dominated against the width of the lining between the stiffeners. Typical flat or lightly profiled faces of sandwich panels do not have supports at the longitudinal edges. The wrinkling stress of these panels is usually calculated using the so-called Plantemaequation. cr 0,8 E F EC GC The coefficient 0,8 is in most cases modified to take into account the effects of imperfections etc. By doing so the transition form the elastic critical buckling tress to the ultimate buckling stress is done. Further modifications of this coefficient, sometimes denominated as a wrinkling factor, can be justified by taking into account the geometry of the faces, leading to a value derived from tests. Further adjustments of this equation are discussed in Pokharel and Mahendran (005). 4

29 This problem concerning the flat or lightly profiled faces of typical sandwich panels without a support on the longitudinal edges could also be calculated within the framework of EN The wrinkling failure can be interpreted as a buckling failure of the stiffeners of the face. After calculating the elastic critical buckling load of the stiffener, a buckling curve for the stiffener available in EN could be used. The relative slenderness and the buckling curve are given by d f y cr, s and d 1,47 1,0 0,73 0,66 d for 0,65 d d 0,65 d 1,38 1,38 d This buckling curve was developed for trapezoidal sheeting using the assumption that the elastic foundation of the stiffener is provided by the support of the neighbouring longitudinal edges of the stiffened element. Further investigations have to be done before applying the design procedures of EN for the design of lightly profiled sandwich panels. At present, the observations can be used to make a differentiation between lightly and strongly profiled faces of sandwich panels. When calculating the critical buckling load of the stiffener, the width b P = s W of the web has to be fully effective, thus c = 1,0. As a second criteria, column-buckling of this stiffener must not be the significant failure mode, so d = 1,0 for column buckling of the stiffener. Then, the face can be regarded as strongly profiled and the load-bearing capacity of the panel can be calculated as presented in the previous chapters. 6. Conclusion The determination of the load-bearing capacity required for the design of sandwich panels is to a large degree based on test results. In contrast, there is a large number of references for the determination of the load-bearing capacity of trapezoidal sheeting. Because both building components are comparable in materials, geometry and load-bearing behaviour, a calculation procedure for sandwich panels which relies on the design principles developed for trapezoidal sheeting seems to be possible, by taking into account the effects of the elastic foundation provided by the core layer. The basic principles of the design model are introduced and compared with Finite-Elementcalculations as well as with test results from different sources, showing the applicability of the procedures. Apparently this calculation procedure can be applied to strongly profiled faces of 5

30 sandwich panels as well as for trapezoidal sheeting with bonded plastic foam insulation. The lastmentioned building components have found increasing dissemination on the market in the last years. Finally, an overview to the design of lightly profiled sandwich panels was made. Comparison with test results showed the present limits of the applicability of the design procedures of EN if applied to sandwich panels. Lightly profiled faces show a complex behaviour with interactions from plate buckling and column buckling failure. This can not be captured on a sufficient safety level at the moment. Acknowledgements ThyssenKrupp Steel Europe AG provided specimens for the experimental investigations. We express our sincere gratitude for this support. References Timoshenko S P Woinowsky-Krieger S (1959) Theory of Plates and Shells, New York, McGraw- Hill. Stamm K and Witte H (1974) Sandwichkonstruktionen: Berechnung, Fertigung, Ausführung, Wien New York, Springer Verlag Davies M J, Hakmi M R and Hassinen P (1991) Face buckling stresses in sandwich panels Contributions 1991 ECCS Nordic Steel Colloquium, p Davies M J and Hakmi M R (1991) Postbuckling behaviour of foam-filled thin-walled steel beams Journal of Constructional Steel Research 0: Pokharel N and Mahendran M (003) Experimental investigations and design of sandwich panels subject to local buckling effects Journal of Constructional steel research 59: Pokharel N and Mahendran M (005) An investigation of lightly profiled sandwich panels subject to local buckling and flexural wrinkling effects Journal of Constructional steel research 61: Dürr M (008) Die Stabilisierung biegedrillknickgefährdeter Träger durch Sandwichelemente und Trapezbleche, Berichte der Versuchsanstalt für Stahl, Holz und Steine der Universität Fridericiana in Karlsruhe, 5. Folge Heft 17. Karlsruhe 008. EN : Eurocode 3: Design of steel structures part 1-3: General rules - Supplementary rules for cold-formed members and sheeting EN : Eurocode 3: Design of steel structures part 1-5: Plated structural elements 6

31 ECCS/CIB Joint Committee on Sandwich Constructions: Recent European Recommendations on Design and Testing Davis, J.M. 83 Park Road Hale, Altrincham Cheshire WA15 9LQ UK ( Hassinen, P. Helsinki University of Technology, Finland ( Heselius, L. LHH Consulting Oy Ab Ltd ( Misiek, T. Karlsruhe Institute of Technology, Karlsruhe, Germany ( Abstract Light-weight sandwich panels, made of two metal faces separated by an insulating core material, are modern pre-fabricated construction components. The design, manufacture and use of these composite structural elements have required continuous development of regulations and standards. The Joint Committee is the combination of two working groups; CIB W056 Lightweight constructions and ECCS TWG 7.9 Sandwich panels and related structures. It therefore provides a powerful forum for the consideration of international research and development of the design, testing and end-use of sandwich panels. This paper introduces the Joint Committee and its work by outlining its history and its most important publication, European Recommendation for sandwich panels. The significant influence of the Committee on the recently published European Product Standard EN is also described. Focus is then directed to the latest publications of the Joint Committee, namely the Preliminary European Recommendations for Testing and Design of fastenings for sandwich panels and the State of the art report for Design of Sandwich Panels with Openings. The latter is still under preparation and will probably be published in 010. Keywords: sandwich panel, light-weight, fastening, opening, CIB, ECCS, recommendations 7

32 1. The joint committee on sandwich constructions 1.1 CIB steering group S56 and working commission W056 The first sign of the existence of the CIB Steering Group S56 was at the International Symposium on Low-Rise Lightweight Constructions held in Budapest in April, A significant product of this era was CIB Publication 59 Recommendations for the Structural Design of Lightweight Sandwich Panels which was published in This document gives principles for design with regard to static out-of-plane and in-plane actions, impact loads and dynamic loads. It also gives recommendations for design for environmental effects and fire. Finally, it discusses the properties of connections and construction systems including the principles of the quality control. The report was based on ISO standard which introduces the principles of the modern limit state design. The report, which was far ahead of its time, was written as a framework for national regulations and it probably provided the first formalised recommendations for the design of sandwich panels. Nowadays, the emphasis in Europe has moved away from national guidelines and towards harmonized standards for the whole continent. However, historically, Recommendations produced by European-wide Committees of experts have led to a similar unified outcome. The name of Steering Group S56 was changed to Working Commission W056 and the membership reconstituted. It held its first meeting in Espoo, Finland in May, Initially, developments in lightweight constructions were introduced and discussed without limitation as to materials or types of construction. Later, the subjects of W056 became focused more towards mineral wool cored sandwich panels because of the need for guidelines for this new sandwich panel product. The coordinators of W056 during this period were Professor J M Davies in and Lars Heselius in Thus, W056 extended the Preliminary European Recommendations for Sandwich Panels written originally by ECCS TWG 7.4 (see 1.) and the outcome was CIB Publication No 148 which was published in 1993 and reprinted in ECCS technical working group TWG 7.4 The European Convention for Constructional Steelwork (ECCS) has a long and distinguished record of pre-standardisation work and its Technical Committee TC7 Cold-formed thin-walled sheet steel in buildings has been one of its most successful working committees and, since its inception in 1974, it has produced numerous significant documents. The workload of TC7 quickly increased to the point where a number of separate Working Groups were required and in 1983, noting the increasing interest of the market in light-weight metal sheet faced sandwich construction, it was decided to form a new technical working group TWG 7.4 Sandwich Panels to produce European Recommendations for design, testing and good practice in the use of sandwich panels. The work of TWG 7.4 started in 1983 under the chairmanship of Dieter Stemman taking full advantage of his experience with German Zulassung documents. 8

33 TWG 7.4 produced Preliminary European Recommendations for Sandwich Panels in two parts. Part I Design was published in 1991 and relates to the analysis, design and testing for the loadbearing capacity of the sandwich panels. This significant document has provided the basis for a number of subsequent Recommendations and Standards. Part II Good Practice was published in 1990 and introduces subjects such as building physics, fire, installation and erection work. In the late nineties needs to update the European Recommendations became obvious and ECCS TC7 founded a new Technical Working Group TWG 7.9 Sandwich Panels and related Subjects. Because of the mutual interest, in 1998, ECCS TWG 7.9 and CIB W056 combined to form the Joint Committee on Sandwich Construction, which is still active today. The first two years of work resulted in the final manuscript of the European Recommendations for Sandwich Panels, Part 1: Design, which was published as CIB Publication No 57 in 000 and ECCS Publication No 115 in 001. The authors pay tribute to the work of Antti Helenius who acted as secretary of W056 and the Joint Committee for more than ten years. Based on the work of the CIB W056 Working Commission, a group of members combined together to write a book covering all aspects of sandwich panel design and construction. Although individual members took responsibility for individual chapters, this book is unique in that the authors took joint responsibility for the whole and each chapter was discussed in detail at a series of meetings. The contributions were edited by Professor J M Davies and the result was published in 001 entitled Lightweight Sandwich Construction. The above Recommendations have formed the basis of the European product standard for selfsupporting double skin metal faced insulating panels, EN 14509, which was published in December 006 and cited in the Journal of CEN in December 008. Drafting of the product standard was in the hands of CEN Committee TC18/SC11, which includes most members of the Joint Committee. The product standard extends the European Recommendations to include all of the Essential Requirements of the Construction Product Directive in accordance with formal mandates. Because it is a European product standard, EN defines the requirements and the methods of verification of the essential properties of the factory made sandwich panels. The standard does not and cannot specify requirements for operations carried out after manufacture such as cutting, fixing and installation. Consequently, in 004, the reactivated Joint Committee identified a number of subjects in which further direction and harmonization is needed. The first items to be considered were the fastening and openings in sandwich panels. 9

34 . The recent European recommendations.1 Testing and design of fastenings of sandwich panels.1.1 Background The product Standard EN does not give any requirements regarding the joints and fastenings of the sandwich panels. In order to remedy this, the Joint Committee has produced preliminary recommendations which discuss the experimental determination of the load-bearing resistance of fastenings, the evaluation of the test results and the principles of the design of fastenings. These preliminary recommendations, now available as ECCS Publication No 17 and CIB publication No 30 in 009, are based on the earlier versions of the European Recommendations for Sandwich Panels ECCS & CIB (000) updated to accord with current knowledge and experience..1. Fastening of sandwich panels Figure 1: Fixings of sandwich panels: typical application and fastener (right hand Figure courtesy of Würth). Because of the inherent rigidity of the sidelap, seam fasteners are not generally used in sandwich construction. Attention is therefore concentrated on the connections to the supporting structure. These may be concealed in the sidelap or may be visible and pass through the complete panel. Whichever solution is adopted, self-tapping or self-drilling screws are generally used. Both of these are able to cut their own threads into the substructure, however self-tapping screws require predrilling. The screws have a formed drilling bit that allows the screwing and drilling in a single 30

35 operation. In order to produce a rainproof joint, sealing washers with a vulcanised EPDM layer are used between the screw head and the face. The washers of the screws fixed to the upper flange of the trapezoidally profiled facing, may take the form of saddle washers to tighten and to support the profile. The fasteners between the sandwich panel and the substructure may be loaded by tensile and shear forces as well as by bending moments (head deflection) due to the thermal movements of the faces..1.3 Tensile and shear resistance of fastenings The load-bearing performance of fastenings is mainly determined experimentally. Therefore, tensile, shear and bending tests have to be performed. Attention is generally concentrated on the strength (capacity) of the connection. However, it is important to appreciate that stiffness and ductility are also important. In particular, sandwich panel assemblies are generally extremely rigid with regard to in-plane displacements so that any such movements (due, for example, to thermal elongation or parasitic stressed skin action) must be accommodated primarily by ductility in the fastening system. e C e 3 F e 1 e direction of span Figure : Test arrangement for pull-through resistance based on small-scale specimens at an end support (ECCS & CIB, 009). e C e 3 F e 4 e 4 direction of span Figure 3: Test arrangements for pull-through resistance with small-scale specimens at an intermediate support (ECCS & CIB, 009). The tensile resistance of a fastening represents the minimum value of the pull-through resistance and the pull-out resistance. The load-bearing capacity of the screw itself does normally not play an important role in the resistance of the fastenings of sandwich panels. The Recommendations ECCS & CIB (009) deal with the pull-through resistance only. For determining the pull-out resistance reference is made to ECCS (008). 31

36 The distance between the end of the panel and the fastener has an influence on the tensile resistance of the fastening. Therefore, two separate test arrangements have been developed to take into account the different nature of the fastening at an end support and at an intermediate support as shown in Figs and 3. Figure 4 shows the principal set-up of a test for the determination of the shear resistance of a screw fastening. Since the influence of the external face decreases with the increasing thickness e C of the core layer, the tests are to be performed with the largest envisaged panel thickness. As an alternative, the sole direct load transmission between the internal face and the substructure may be investigated. t 1 F e C e 1 u Figure 4: Shear test assembly for screws passing through the panel (ECCS & CIB, 009). For the determination of the resistance of a screw to the deflection of the screw head, the shaft of the fastener is subjected to a repeated deflection of u, where u is the maximum lateral displacement of the head. The deflection spectrum is as follows: cycles with a deflection of 4/7 u,.000 cycles with a deflection of 6/7 u and 100 cycles with a deflection of u. This load spectrum is based on the assumption of a service life of 50 years in a location in central Europe. During the test, the screw shall not fail and, after cyclic loading, the screw has to achieve at least 80 % of the mean value of the pull-out resistance without the repeated load.. Openings in sandwich panels..1 Background In the majority of buildings, functional requirements necessitate openings in the wall and roof cladding. The current state of the art with regard to openings within the sandwich panel wall and roof construction generally requires the addition of reinforcement in the form of additional framing to replace the load-bearing capacity which has been removed by the opening (Fig 5). Recent research results indicate that this additional framing is not always needed. The state of the art report of the Joint Committee will introduce the possibility of designing panels with openings without the need for any additional strengthening... Remaining resistance of panels with openings The cross-section of the panel which remains after cutting an opening may be able to carry the applied loads. In this case, the opening is classified as small. The evaluation of the remaining 3

37 resistance is evidently the primary design task and this includes verification of the resistance of the net section to the design bending moments and shear forces at the critical points in the vicinity of the opening taking account of any stress concentrations. A section A - A framework A replacement with crossbeams, e.g. C-sections Figure 5: Openings requiring additional supporting members (ECCS & CIB, 010). remaining cross-section gross cross-section Figure 6: Remaining cross-section in the vicinity of an opening (ECCS & CIB, 010). The ECCS & CIB report (010) gives design formulas to calculate the remaining resistance of a panel with openings (Fig 6)...3 Activation of the load-bearing resistance of neighbouring panels If the remaining cross-section of an individual panel is not able to carry the applied loads, the loadbearing resistance of the neighbouring panels can be activated. Due to the difference of the stiffness between the panels with and without openings, the whole or a part of the loads applied directly to the panel with openings will be transferred via the longitudinal joints to the adjacent panels (opening A in Figure 7). The most severe case is the load transfer from a completely cut sandwich panel 33

38 (opening B in Figure 7). While in case A the panel might have sufficient capacity to withstand the load, case B relies entirely on the load transfer. In both cases, the neighbouring panels will receive additional loads which can be calculated on the basis of the compatibility of the deflections in the longitudinal joints (Figure 8). These loads are transferred by these joints and therefore, the assessment of the strength and stiffness of the joints is essential. Furthermore, the load transfer will result in an eccentric line load applied to the adjacent panel without an opening thus activating its torsional rigidity and causing additional shear stresses due to the torsional moment. A B Figure 7: Small and full-width openings in sandwich panel walls (ECCS & CIB, 010). Figure 8: Line loads on the longitudinal joint between neighbouring sandwich panels with different bending and shear stiffness (ECCS & CIB, 010). The intensity of the load, which can be transferred through the longitudinal joints to adjacent panels, depends on the bending, shear and torsional rigidity of the complete panels and, in addition, on the shear rigidity and shear resistance of the longitudinal joint. Simple framework-software can be used to compute the internal forces and deflections. ECCS & CIB (010) presents the basic principles of the design procedure. A test set-up for the determination of the stiffness and resistance of the longitudinal joints is presented as well as formulas to determine the required bending and torsional stiffnesses of the panels. 34

39 ..4 Large-sized openings In some cases load transfer from the panel with openings is not possible due to the limited shear resistance of the longitudinal joint or the limited load-bearing resistance of the neighbouring sandwich panels. An additional frame has then to be designed to carry the whole load of the sandwich panel with openings. This frame may be placed within the longitudinal joints of the panel (Figure 9). longitudinal beam F c F c F1,c F1,c F c cross-beam F c F t F t longitudinal beam Figure 9: Frame-construction within the panel using special aluminium profiles with web parts made of plastic sections (ECCS & CIB, 010). 3. The EASIE project The Joint Committee is the birthplace of the European research Project Ensuring Advancement in Sandwich Construction through Innovation and Exploitation (EASIE). The EASIE project commenced in October 008 and is funded to approximately 4 million Euro with financial support from the European Community's Seventh Framework Programme FP7/ NMP-SE-008 under grant agreement No EASIE provides support to small and medium sized enterprises to develop the design, manufacturing and use of products related to sandwich constructions. The Joint Committee first defined subjects and tasks of the project reflecting the needs and interests of the industries were collected. In addition, EASIE will also study subjects relevant to the updating of the product standard EN It includes a number of innovative subjects such as the performance of sandwich panels subject to inplane shear and axial resistance. The second goal of EASIE is to disseminate existing and new information for use in practice. This will be carried out through seminars and the production of practical guidelines and e-learning modules. The subject areas include the principles of the design and use of sandwich panels, fastening and fixing, properties in extreme situations and the information given by the research results of EASIE. Practical guidelines and seminars will help and broaden the correct and safe use of sandwich panels in Europe and ICPC. 35

40 4. Future tasks of the Joint Committee The earlier work of the Joint Committee has produced independent technical harmonised guidelines and relevant background documents for the European product standard. The technology of sandwich panels is still evolving and the Committee seems still to have a useful role as author of technical guidelines based on new and reliable information without pressure from commercial or other interests. Because of this continuous development, keeping the available documents up to date is an important task of the Committee. Future tasks are likely to include the preparation of new guidelines on the basis of the technical information emerging from projects such as EASIE. Finally, questions arising from environmental issues in buildings and building products may open new technical tasks requiring a measure of European collaboration. References CIB (1978) Recommendations for the structural design of lightweight sandwich panels. Rotterdam: CIB - International Council for Research and Innovation in Building Construction, S56 Lightweight Constructions. International Organization for Standardization (1973) General principles for the verification of the safety of structures. ISO 394. Davies J M et al. (001) Lightweight sandwich construction, Oxford, Blackwell Science. ECCS & CIB (000) European Recommendations for Sandwich Panels Part 1: Design. Brussels/Rotterdam: ECCS - European Convention for constructional steelwork, TWG 7.9 Sandwich panels and related structures & CIB - International Council for Research and Innovation in Building Construction, W056 Lightweight Constructions. ECCS Publication No 115 and CIB Publication No. 57. ECCS (008) The Testing of Connections with Mechanical Fasteners in Steel Sheeting and Sections. Brussels: ECCS - European Convention for constructional steelwork, TWG 7.10 Connections in Cold-formed Steel Structures. ECCS Publication No 14. ECCS & CIB (009) Preliminary European Recommendations for Testing and Design of Fastenings of Sandwich Panels. Brussels/Rotterdam: ECCS - European Convention for constructional steelwork, TWG 7.9 Sandwich panels and related structures & CIB - International Council for Research and Innovation in Building Construction, W056 Lightweight Constructions. ECCS Publication No 17 and CIB Publication No. 30. ECCS & CIB (010) State of the art report for Design of Sandwich Panels with Openings. Brussels/Rotterdam: ECCS - European Convention for constructional steelwork, TWG 7.9 Sandwich panels and related structures & CIB - International Council for Research and Innovation in Building Construction, W056 Lightweight Constructions. (will be published in 010?) 36

41 EN (006). Self-supporting double skin metal faced insulating panels - Factory made products - Specification. Brussels: CEN - Comité Européen de Normalisation. Ensuring Advancement in Sandwich Construction through Innovation and Exploitation: 37

42 Methods to Measure the Durability of Structural Sandwich Panels Hassinen, P. Aalto University Department of Structural Engineering and Building Technology, Finland ( Pfeiffer, L. ThyssenKrupp Steel Europe AG Profit Center Color/Construction, Germany ( Abstract Determination of the intended length of the service life is an increasingly important question and may be soon a crucial parameter in the development and design of civil engineering structures. Length of the service life may be limited by a reduction of the resistance or by an unsatisfactory function at the serviceability state. To determine the service life, knowledge about the deterioration mechanism and their influence on the structural behaviour is needed. In some cases, information about the development of the mechanical and environmental loads during the service life is important, too. European standard EN introduces experimental methods to measure the changes of the selfsupporting sandwich panels and further, to classify the panels to be fit for the use in external walls and roofs. The methods are based on an indirect parameter by measuring the changes of the cross panel tensile strength and not the properties, which are directly used in the analysis and in design. Ageing of the cross panel tensile test specimens is based on accelerated ageing histories using a high temperature and moisture. The method is practical and relatively easy to use, however, it does neither tell about the changes caused by a real natural environment nor the real changes of the ultimate resistance or the function at the serviceability limit state. The contribution introduces the parameters which are essential in structural design and then, studies the possibilities to measure the changes of the parameters in long-term use. The essential parameters are the wrinkling strength of a face under compressive stresses and the shear strength of the core and bond. These are followed by a number of other failure modes, such as the resistance at the supports to pressure and suction loads. The information is based on tests with small-scale specimens, the results of which will be compared to a limited number of results on full-scale specimens having been in service and exposed to natural ageing conditions. The contribution concludes the promising possibilities and draws outlines for development of methods for the use in practice. This publication was prepared within the scope of the EASIE research project. The project is sponsored in EU 7th framework programme. Keywords: sandwich panel, resistance, long-term, durability, ageing, test 38

43 1. Introduction 1.1 Definition of a sandwich panel Sandwich panels in the context of this report are composite structures, consisting of a minimum of two deck layers and a core. The term panel is derived from the Latin, French and Dutch word paneel and means a flat piece of construction (Brockhaus, 1991, a). This describes the general shape of a sandwich panel. The word sandwich describes generally a combination of different layers (Brockhaus, 1991, b). A typical sandwich layer in construction consists of three layers. A rather thick rigid core material is laminated between two thin faces. There are many possible combinations of facing and core materials, depending on the intended use of the sandwich panel. This work, however, deals primarily with the most common form of a sandwich panel used currently in the building construction: factory made engineered elements, consisting of thin metallic facings of a thickness between 0.4 and 0.75 mm and rigid foam or inorganic wool core. 1. Load bearing behaviour of a sandwich panel The individual layers of sandwich panels by themselves have almost no load-bearing capacity. In a sandwich panel the layers no longer act separately but are connected in a process of adhesion. Only this rigid connection makes the structure a sandwich panel and increases the load-bearing capacity substantially. While the metal sheet facing alone is almost without bending capacity as is the core by itself, the connection of the three layers creates a completely new composite structure with widely enhanced capacities. The composite sandwich panel possesses substantially higher load-bearing capacities with regard to bending, shear and torsion impact than the sum of its individual components. Compared to the metal faces, the core in a sandwich panel generally possesses very limited tension or compression stiffness. The bending moment is therefore distributed to the two faces. For a single span, three-layer sandwich panel loaded by a distributed load, this leads to compression and tension in the two faces that are kept apart at a fixed distance by the core. Based on the properties of the layers and the distribution of the internal stresses, the core carries through almost all of the shear force. Additionally, the core provides foundations to the faces of the sandwich panel. Because of that, the properties of the core have a significant influence on the over all performance of the sandwich panel and need to be carefully evaluated in sandwich panel design. Losses in the stiffness and strength of the core have immediate influences on the performance of the whole panel. Such losses can, for example, be caused by durability related degradation. 39

44 1.3 Durability of construction products A survey by Sarja (00) on the typical share of civil engineering products for European countries shows that: Civil engineering products represent about 70 to 80% of the national assets. The energy use for these products during production and maintenance is about 40% of the total national energy consumption. Civil engineering products produce about 35% of the total waste. These figures emphasise the enormous economical importance of civil engineering products for the societies. Due to the importance and costs of these products, it becomes obvious that there are high demands to the service ability during the life span of the investment. Economical effects are dominated by the fact that buildings generally require large investments. The expectations in the investment can only be fulfilled if the building keeps its functionality over a certain time period. The estimated time period depends on the type of building. For an industrial building like a warehouse, for example, it is generally assumed that the building is designed for a service period of 5 to 30 years. After this period a substantial refurbishment of the building is generally required. The most important technological influence for a civil engineering product is the guarantee of the safe use over a time. Building materials need to be designed in a way, which guarantees an adequate structural performance. If a structural building material cannot resist against the received load impacts, the building is damaged or may even collapse. In a worst case scenario, this may lead to a loss of lives and properties. Therefore, the main task in the regulation is to ensure an adequate structural performance of the building materials throughout the whole life cycle of the structure. Legally binding design procedures, which are very much harmonized across Europe, are the base to guarantee a safe structural performance. For sandwich structures, such procedures are defined in the European standard on sandwich panels EN 14509, where also requirements towards the durability of the sandwich panels together with experimental evaluation procedures are described. In practice, requirements to the fitness for use and to the withstanding the extreme actions during the use are set down to a structure. The previous requirement defines the serviceability limit state and the latter one the ultimate limit state. The structure shall maintain the properties in its environment during its design working life, which defines the requirements on the durability of the structure. The durability requirements may be achieved by choosing correct materials and manufacturing techniques. As a result, the deterioration of the structure will stay small enough or the function and resistance will be guaranteed by regular inspections and refurbishment. In constructions, the first principle is used in the major cases of structures. The available durability tests give information about the changes of the strength on a macroscopic level without explaining the physical and chemical mechanisms affecting the deterioration of the material. It is believed, that the durability test is able to classify if the product is to fit for the purpose 40

45 or not. The durability test methods and the loading histories are therefore strongly dependent on the application.. Durability of sandwich panels The European standard, EN , makes in chapter 5..3 a statement concerning durability; Durability and other long term effects. Detailed information on testing scenarios are written down in the normative annex B, Durability testing method for sandwich panels. Currently, the durability performance is only evaluated qualitatively. This means that the loss in cross panel tensile strength under an artificial, durability accelerating climate must stay within certain boundaries over a time. The cross panel tensile strength is seen as an indicator for over all panel performance. Depending on the type of degradation pattern (Fig 1), a core material is either fit for the application or not. The boundaries chosen for an acceptable loss of the strength are, however, lacking the scientific foundation. Even a loss of 60% of the initial tensile strength can be acceptable in accordance with the standard. In particular, such a loss in the tensile strength is allowed without any impact on other relevant design parameters. Because it is not possible to determine the remaining wrinkling stress after the artificial ageing of the tensile strength, a new small-scale test to study the development of the wrinkling stress is developed. Figure 1. Possible degradation patterns for tensile strength (Davies et al.). 3. Wrinkling stress of sandwich panels 3.1 Wrinkling failure modes The wrinkling of the face under compression is a typical and very important failure mode for sandwich panels with membrane-like faces (Fig. ). As stated before, in a loaded sandwich panel, one 41

46 face is in tension while the other one is loaded by compressive stresses. A membrane under compression has, on its own, no stability, but it fails immediately through buckling. In a sandwich panel, the core provides a lateral support for the thin face and prevents the early buckling failure. The lateral support is activated when the face deforms in a wave-like pattern and induces stresses in the core material. The loss of the tensile strength of the core or the bond results in an immediate wrinkling failure of the face. Wrinkling is one of the most critical failure modes in a sandwich panel construction and it very often determines the ultimate limit state. Figure. a) Wrinkling failure in a span and b) on an intermediate support. 3. Determination of wrinkling stress The wrinkling stress is defined as the ultimate stress, a face can take in compression before it fails in buckling. The compression in the panel face is, in practice, induced by a bending moment that can, be caused by wind loads or temperature loads in multi span panels. Determination of the wrinkling stress of a sandwich panel generally requires full scale testing. The full scale test may be carried out by subjecting a simply supported beam to four line loads. In a test setup according to figure 3 the ultimate bending moment (M u ) and the wrinkling stress ( w ) for a lightly profiled panels are given by: M u FBu L 8 M u w and ec A1 (1), () where F u is ultimate load including self weight of panel and loading equipment L span of the test specimen e C distance between the centroids of the faces cross-section area of the face in compression A 1 In order to be able to determine wrinkling stress after an artificial ageing, it is necessary to develop a small scale wrinkling test set-up which allows the determination of the wrinkling stress or more precisely the time dependent change of the wrinkling stress. The small sample can then fit into a climate chamber and tests on the aged samples can be performed. 3.3 Small-scale wrinkling test The mechanical parameters necessary for the design may be obtained on the basis of the small scale testing. An exception is the wrinkling strength. For a durability research this poses the problem that it is not possible to determine the wrinkling strength after an artificial ageing. Therefore, a small scale test setup determining wrinkling strength has to be developed. When developing a small scale 4

47 wrinkling test for sandwich panels, it is the general idea to no longer induce compressive stresses in the face through a bending moment but to cut out a small panel area and expose it to a direct axial load in the direction of the plane. Baehre (1988) and Pfeiffer (000) have previously tried to establish small scale wrinkling tests for the design of sandwich panels. However, it was not possible to determine the exact wrinkling capacity of a sandwich panel through small scale testing. The main problem in the earlier work was in finding a way to introduce the in-plane loads without damaging the thin face locally, a problem that does not occur in a full scale test. Special load application devices help to overcome the problem. During the test series it has been found that the full-scale bending test loads the sandwich panel twofold. In the upper face, the in-plane bending stress meets the deformation caused by the line loads, which evoke the actual bending moment and compressive stresses in the face. For a direct comparison between six point bending test and a small scale wrinkling test, this load combination must not be neglected. The results gathered in this project suggest that the research conducted by Baehre (1988) would have had a positive outcome if the local deformations were taken into consideration. With the results on hand, it was not possible to find a small scale test defining the wrinkling strength in the same way as a full scale test does. 3.4 New test set-up Figure 3. Arrangements of a full-scale and a small-scale wrinkling test with descriptions of the distribution of the bending moment and compressive stress. The new test combines the two previously used test methods. Instead of causing compression in the upper face through a bending moment, the small scale test loads directly the face in a vertical, inplane direction (Fig. 3). While the authors in their earlier work used a wooden T-shaped load application device, which was glued to a rectangular specimen with completely flat faces causing compression in one face, Baehre (1988) used a micro-profiled specimen with T-shaped load application devices made of steel. The load application device was not glued to the specimen but rather clipped onto it. In both cases the T-shaped device was chosen to gradually introduce the load to the specimen without causing local failures at the point of the loading (Fig. 4). Baehre failed to find a 43

48 small-scale test, giving similar results as obtained in a full-scale bending test. The reasons for this were the local deformations in the compressed face under the loading points in the full scale test. Such deformations weaken the observed wrinkling strength as they cause an additional local moment and additional stresses in the face. For a small scale test comparable to a full scale test this effect must be accounted for. Figure 4. Test set-up of a small-scale wrinkling test from the left to right; Baehre (1988), Pfeiffer (000) and Pfeiffer (004). Figure 5. Introduction of the load in a small-scale wrinkling test. For illustration purposes, the specimen is separated from the load introduction device. The total load (F) is transferred to the specimen directly (F direct ) and through the shear force (F shear ). For this work, a combination of the two previously described small-scale tests, determining the wrinkling stress, was chosen. The profiled specimen from Baehre was enlarged to a bone-shaped sample, again using T-shaped load application devices at both ends of the specimen. This time the application devices were both glued and clamped to the samples. Only the combination of gluing and 44

49 clamping together with the bone shape hinders the specimen from failing early through local crushing. Gluing the load application device to the samples allows the load to be introduced to the panel faces over a larger surface area, reducing concentrated load effects. The load is introduced through a shear force in the glue area and directly to the top and bottom ends, where the specimen touches the T-shaped device (Fig 5). To study the effect of the local deformation at the load introduction areas in the full scale test, samples with and without local deformations were studied. A three step approach was taken. First, samples with completely undisturbed faces were tested, giving the maximum wrinkling stress that can be obtained. In a second step, the faces were dented prior to executing the small scale test. The depth of the imprinted deformations was of the same dimensions as observed in the full scale test. In a third step, the effect was simulated through a bi-axial loading in the small scale test. An additional load device was clamped to the specimen loading it in a perpendicular direction. All obtained results were compared to the results of the full scale bending test. The last procedure resulted in the most accurate results. 3.5 Test set-up based on rectangular specimen Preparing of the bone-shaped specimens requires careful machining of the two metal sheets and the core. To simplify the preparation work, a test set-up has been developed, which base on a rectangular specimen. To avoid the failure at the end and to create a wrinkling failure in the mid-part of the specimen, the specimen is loaded by an additional transversal load (Fig. 6). a) b) c) d) Figure 6. a) Rectangular small-scale test specimen, b) initial displacement caused by the transversal load (initial deformation equal to the face thickness, w 0 t F ), c) displaced mesh and d) displacement caused by the transversal and axial load (w u 3 t F ). 45

50 Transversal load causes an imperfection in the face and compressive stresses in the core layer. The level of the transversal load has to be high enough to cause an initial imperfection to the wrinkling failure and on the other hand, lower than the elastic limit of the core layer (Fig. 7). Experiments on polyurethane-foam and mineral wool cored specimens have shown the test set-up to give acceptable modes of wrinkling failures. 1 axial stress in relation to wrinkling stress 0,8 0,6 0,4 0, 0 Fdef.1 Fdef. Fdef.3 Fdef.4 0 0,0005 0,001 0,0015 0,00 0,005 0,003 axial compressive deformation Figure 7. Influence of the transversal load on the stress-deformation behaviour of the face in the test described in Fig. 6. Transversal loads F def.i cause the transverse stress of 83, 63, 4 and 1 kn/m in the core. The axial stress of the face is compared to an analytically determined ideal wrinkling stress of 15 N/mm in the case. Small-scale specimens will be exposed to an accelerated influence of the environmental actions. Because of the small size of the specimen, it is possible to create a large quantity of reliable information about the changes of the wrinkling stress in function of time and environmental loading history. The data will make the basis to the development of the design models in order to take into account the possible deterioration of the strength already in the design phase. 3.6 Comparing small-scale and full-scale tests In order to calibrate the test setup, a test series, comparing the large-scale and small-scale test arrangements of the wrinkling stress was undertaken. All specimens in this series were taken from the same production batch. The full-scale panels were first tested as simply supported beams as described previously. The test panels had a PUR core between the two steel faces. The wrinkling failure obtained in the full scale tests resulted in an average wrinkling stress of N/mm² (individual test results between min N/mm² and max N/mm²). The wrinkling failure, obtained from the small scale test without the additional effect of the transversal load, averaged to 8.73 N/mm² (individual test results between min N/mm² and max N/mm²). 46

51 Figure 8. Local deformation in the loading point causing an early failure in the six point bending test. The discontinuity (d) occurs under all loading points but interacts with maximum bending moment at the indicated point. Comparing the results given by the two methods, a difference of 6.4 % can be found. This difference is not acceptable if the determination of the accurate wrinkling stress is required. The reason for the observed difference lies in the test setup of the full scale bending test. In a six point bending test the load is transferred to the panel in four line loads (Fig. 8). At the point of the load introduction, this results in deformations of the face causing a failure at the inner point of the loading beam, where maximum local deformations and the maximum bending moment interact. In order to achieve the same results from the small scale testing as determined through full scale testing, this deformation effect needs to be taken into account. An additional loading device generating a transverse load, was introduced in the test. This additional device loads the panel perpendicular to the faces, causing the same local deformation as obtained in the full scale test. The load is applied through two steel profiles, having the same dimensions as the ones used in the full scale test. The loading is controlled with the help of two screws and measured in a load cell. At an average of N/mm² (individual test results between min N/mm² and max N/mm²) the results obtained from the setup show good correlation with the results of the full scale bending test. When comparing the average results, 98.8 % of the strength determined in the full scale bending tests could be reached. The results indicate clearly that it is possible to use the small scale testing to determine the accurate wrinkling strength of a sandwich panel. It is, however, necessary to simulate the full-scale test accurately and to take into account even the small details, like the local deformations at the load introduction points. When lacking knowledge of the maximum load obtained in the full scale test (F max ), which is the case when no full scale test has been carried out on a particular panel, it is necessary to gradually increase 47

52 the deformation load relative to the load parallel to the face. The load needed on the deformation device can be calculated on the basis that the stress parallel to the face shall be equal in the both tests. This means, that a certain load in the small scale test, corresponds to a certain load in the full scale test, both causing the same stresses parallel to the surface. At the same time this load determines the load on the additional load device and can be calculated as described in the following: The membrane stress in the flat face in the six point bending test can be written by f 6P F6 P L 8e A C f 1 6 P where σ f6p is the stress in face in 6 point bending test F 6P total load on panel in 6 point bending test A f16p : cross sectional area of the face in compression in 6 point bending test (3) The corresponding stress in the small scale test is given by fsc F A SC f 1SC where σ fsc is the stress in the face in the small scale wrinkling test F SC total load parallel to the face in the small scale wrinkling test cross sectional area of the compressed face in a small-scale test A f1sc (4) A simulation of the full scale test through the small scale test requires an equal stress parallel to the face (wrinkling stress). Equating the equation 3 and 4 gives the load F 6P 8F e b SC C 6P F6 P L bsc which can also be written as F6 P FSC ec b6 P 4 L b SC (5) in which t f1 net thickness of the face b 6P width of the panel in the 6 point bending test width of the sample in the small-scale wrinkling test b SC Local stresses in perpendicular direction to the face in the load introduction point are def.6p F 4b 6P 6P B full-scale test and (6) Fdef def. SC small-scale test (7) b B SC where B is the width of load application device in the set-ups 48

53 For comparison of the two tests, the local transverse stresses need to be equal, which results in F def F6 P b 4b SC 6P (8) Combining the equations 8 and 5 leads to F def FSC e L C b b 6P SC b b SC 6P FSC e L C (9) The load in the additional deformation device can now be calculated, depending on the axial loading of the sample, the distance between the centroids of the two panel faces, and the span length of the static system in the six point bending test. 4. Conclusions The results presented in the paper indicate that the small-scale wrinkling test can be used in the evaluation of the wrinkling stress. The test requires extensive preparation work such as bone shape or rectangular cutting and gluing and fixing of the load application devices. Laboratories dealing with sandwich panel technology on a regular base can adopt the test with some additional work. For further work on the durability related degradation of the wrinkling stress, it is sufficient to consider only the durability related changes and thus the relative values. In future research work the test arrangements without the additional deformation device will be studied. Also a simplified rectangular wrinkling test set-up will be developed further. The shear and compression strength of the core is determined in most cases on the basis of the smallscale tests. Thus, the durability of the shear and compressive strength can be studied by exposing the specimens to accelerated deterioration processes in a climate chamber. These studies will contribute further information to the evaluation of the durability of the structural sandwich panels. References Sarja A (00) Neue Anforderung an die Dauerhaftigkeit von Konstruktionen (New requirements towards durability of constructions), Der Prüfingenieur 0: EN Self-supporting double skin metal faced insulating sandwich panels Factory made products Specification Davies JM et al. (001) Lightweight Sandwich Construction Blackwell Science, Oxford Baehre R (1988) Experimentelle Ermittlung der aufnehmbaren Knitterspannung von ebenen und leicht profilierten Stahldeckschichten, Research Report, University of Karlsruhe 49

54 Pfeiffer L (000) Der Einfluss der Zugfestigkeit auf die Knitterspannung bei Sandwichelementen, Final Diploma Thesis, Fachhochschule Mainz (D) and Häme Polytechnic (FIN) Pfeiffer L (005) Durability Assessment of Sandwich Panel Construction, Doctoral Thesis, University of Surrey (GB) and Fachhochschule Mainz (D) EASIE Ensuring Advancement in Sandwich construction through Innovation and Exploitation, EU 7 th FP, 008 to 011, 50

55 Reliable Composite Roofing: Learning From Experience Roberts, K. Roberts Consulting, UK ( Abstract In reviewing the findings of roof inspections there are often common issues that recur. When the problem is repeated for the same roof system on different projects there is an opportunity to learn from the experience and to change practice. The paper asks the question how should this information be shared for the common good? Investigations are often commercially sensitive and occasionally the subject of litigation. If information is to be shared through a public forum it is essential that it is fair, balanced and objective. The paper examines previous means in which common construction faults have been recognised and publicised in an anonymous and constructive way. The paper refers to the CIB / RILEM Roofing Materials and Systems Task Group that was established in 007 to improve our understanding of the reliability of roofing systems and specifically to identify recommendations that can improve reliability, such as by learning from experience. The findings are described from two investigations into the weathertightness of composite panel roof systems laid on shallow slopes. Sliding movement has been observed at end laps, highlighting the need for the fixing and sealant details to be designed to accommodate such movements. By developing appropriate means to share our experience in a constructive way we can improve the reliability of the roofs we design and build. Keywords: roof reliability, composite panel roofing 51

56 1. Introduction Building owners, their professional advisers and main contractors are becoming increasingly concerned at the number of recently completed buildings where there are recurring rainwater leaks through the roofs several years after completion. The problem appears to be widespread in the UK and occurs on both pitched and flat roofed buildings. For building owners and facilities managers the reliability of their roofs is an important topic. The author is a civil and structural engineer who specialises in roofing and cladding, working throughout the UK and Ireland. Over the past nine years the author has inspected more than 10 different roofs of different types in all parts of the UK and Ireland. The reason for undertaking 80% of the surveys was because of reports of water ingress through the roofs. The surveys were generally commissioned after the roofing contractor and design team had carried out some remedial works in an attempt to cure the leakage, indicating that the roofs were not readily repairable. Not all roofs of new buildings leak although there is clearly the potential for improvement. This paper offers constructive feedback.. Reliability studies Over the past three decades there has been a growth in reliability engineering studies, particularly in the aerospace, vehicle production and electronics industries, improving the in-service performance of the assemblies. Consumers are acutely aware of the problem of less than perfect reliability in domestic products such as televisions and cars, and have now come to expect these products to work first time and continue until they become obsolete. These studies have been highly developed in Japan where quality and reliability were adopted as national priorities. Building owners expect the same levels of service from their roofs, in particular meeting the basic requirement of providing a dry internal space. Reliability has been defined by O Connor (00) as the probability that an item will perform a required function without failure under stated conditions for a stated period of time. A crude measure of the reliability of a roof is the number of times the roofing contractor needs to be called back to site to resolve a problem. There are many reasons why a roof might leak and result in a need for remedial works. Knowing as far as is practicable the potential causes of failures is fundamental to preventing them, although it is rarely possible to anticipate all of the causes and a level of uncertainty needs to be taken into account. The CIB W83 / RILEM Joint Committee on Roofing Materials and Systems comprises of roofing specialists drawn from more than 15 countries. In 007 a Task Group was established to develop our understanding of the reliability of roofing systems, and specifically to identify and prioritise practical actions that can deliver improvements. 5

57 Introducing element redundancy is encouraged at the design stage such that if one layer is not perfect and does not perform there is a secondary back up. The double layer roof, with a primary weathering outer face and a secondary underlayer that drains freely out of the building, was described in a paper presented by Roberts (007). Some recommendations are no more than common sense, such as engaging competent tradesmen with the appropriate training and experience to carry out the works, supported by active field inspections and testing to check the quality of materials and work. A recurring theme in the Task Group discussions is the retirement of experienced personnel, losing the knowledge built up over years. This will create foreseeable problems as construction output increases when the economy grows again. Learning from experience is another recommendation that can lead to the avoidance of problems. Constructive feedback on the completion of a project can result in product developments. In a paper written by Dr Mann (008) the importance of learning from failures at the interfaces within a construction project was highlighted. The paper reviewed a wide range of problems and summarised the findings of subsequent enquiries. These studies identified poor communications between the parties as a principal failing, recognising that the human interface within any engineering project is perhaps the riskiest of all. The paper concluded that there are many lessons to be learned but corporate memories are weak and it is incumbent on every engineer in each generation to study failures and gain wisdom from them. 3. Weathertightness of composite panel laps The choice of composite panels for low slope roofing applications has become popular in the UK and Ireland over the past decade. The sandwich panels comprising of an insulated core and thin metal faces, have been successful because they are light in weight, energy efficient and can be easily handled and erected. Members of CIB Commission W56 have contributed to a book edited by Professor Davies (001). This offers an authoritative manual on many aspects of composite panel construction, including thermal, acoustic and fire performance, structural behaviour and mechanical testing. The European Convention for Constructional Steelwork has also published European recommendations for sandwich panels. Based on this work a European Standard was approved in June 006 for the specification of self supporting double skin metal faced insulating panels, BS EN The author regularly carries out independent inspections of roofs and is generally commissioned by building owners, their professional advisors and general contractors. Over the past four years there have been twenty buildings inspected with composite panel roofs laid at shallow slopes. The common complaint has been intermittent rainwater leakage into the buildings. At the time of the inspections the age of the roofs varied between 1 months and 10 years. The sizes of the roofs were between 500 m and 40,000 m. The panels had been manufactured by three different companies and installed by different roofing contractors. 53

58 Within chapter 3 of the Lightweight Sandwich Construction book there is only a limited treatment of the watertightness issues at joints, with no reference to end laps, fixings and ancillary penetrations. A test method for watertightness is offered which could be of assistance, although it is understood that this is generally applicable to external wall systems with few manufacturers using the test for shallow pitch roofing applications. BS EN offers a specification for the panels themselves and clause 5..6 states that where required, the water permeability (resistance to driving rain) of a complete assembly of sandwich panels shall be assessed. This includes site applied seals, flashings and fixing methods, and recognises that the watertightness of the assembly is a function of its installation. Leaks through roofs have been the cause of disputes, sometimes resulting in litigation. Each party instructs a different surveyor to carry out detail examinations and prepare written reports. It is common in the UK for the experts to be directed to meet, to discuss their findings and to seek agreement on causation. These exercises can produce information that would be useful to the wider industry if shared through an appropriate forum. It is difficult to obtain a true assessment of the national scale of the water ingress problems with shallow pitch composite roof panels, with no independent survey results publically available. Informal discussions with roofing contractors indicate that they also come across such problems, suggesting that the cases inspected by the author are not unique and that further examination would be worthwhile. There is an opportunity to learn from recent experiences. 4. Case study A: Airdrie During heavy rain there were more than 500 active leak positions within the food processing plant, with the extent becoming progressively worse over the five year period since completion. Each roof slope comprised of two long composite panels with an end lap detail at mid slope. More than 95% of the active leaks were directly below the composite panel end laps. Table 1: Basic data for Case A composite panel roof Location Airdrie, Scotland Building use Food processing plant Altitude, site exposure +165 m, severe Roof area 9,700 m Roof slope Up to 7m long at 4 Composite panel size 1m wide, up to 15.5m long External facing 0.5 mm thick plastisol coated steel Insulation 40mm thick polyisocyanurate Internal facing 0.4mm thick coated steel Side laps Overlap End laps Mid slope, 150mm, no stitchers Rooflights In plane, double skin GRP Ancillaries Air extract vents, platform supports 54

59 Figure 1: typical composite panel end lap On closer examination it was found that the end laps were 150mm long with two lines of 9x3mm butyl sealant tape, see Figure 1. There were no tail stitching screws originally installed. The sealant was found to have perished with small pieces crumbling away. On the day of the inspection there were cold weather conditions and it was seen that there was contraction in the composite panels, with the end lap joint opening by up to 5mm, see Figure. Additional end lap stitcher screws had been installed as a remedial measure, although after a season these had worked loose. Figure : sliding end lap movement of 5mm Evidence of movement of the composite panels relative to fixed structural steel penetrations was also observed, with cracks in rigid reinforced coatings. Elsewhere a lack of fit between the in-plane rooflights and the metal faced composite panels contributed to the water ingress. It is noteworthy that 55

60 the building was situated on high ground in a severely exposed location with the design roof slope at a minimum for profiled sheet roof cladding with pierce fixings. The case study observed the sideways movement that can occur at the ends of long lengths of unrestrained composite panels. The recommended position of the primary fixing screws at end laps has subsequently changed to pass through overlapping panels, as shown in the Metal Cladding and Roofing Manufacturers Association Technical Paper 16 (004). This will offer the benefit of restraining the ends of the panels. These fixing screws will be subjected to significant thermally induced shear loads, particularly in cold winter conditions. In addition the recommended sealant size has changed from 9x3mm to 6mm diameter. 5. Case study B: Consett More than 40 active leak positions were reported below the aluminium faced composite panel roofs of an educational building laid to very shallow slopes in the severely exposed location. One surveyor remarked that if ever there was to be a case study on how a composite panel roof should not be designed and built, then this was it. Table : Basic data for Case B composite panel roof Location Building use Altitude, site exposure Consett, County Durham Educational Roof area 1,550m Roof slope Composite panel size External facing Insulation Internal facing Side laps End laps Rooflights Ancillaries +55 m, very severe Up to 17m long at 1 1m wide, up to 17m long 0.7 mm stucco embossed aluminium 40mm thick polyurethane 0.4mm thick coated steel Aluminium snap-on cap None Barrel vault, double skin GRP Large number of gas flues and vents In strong windy conditions the aluminium snap-on cappings became detached resulting in water ingress at side laps. The large number of flue penetrations at the bottom of the shallow roof slopes created long term ponding upslope, with wind driving water up and under the crown fixed rooflights with inadequate kerb heights. 56

61 Figure 3: general view of the composite panel roof The roof became the subject of a legal action which was eventually resolved by mediation. Subsequently all of the aluminium faced composite panel roofs were encapsulated using a polyester reinforced elastomeric liquid coating, with new upstands to support the barrel vault rooflights. It is notable that it took six years for the matter to be resolved after practical completion of the building. The consequential financial costs were significant. It is experiences such as these that are a motivator for building owners and contractors alike to seek improvements to the reliability of the roofs we design and build. Figure 4: large number of flue penetrations at the foot of the roof slope 57

62 6. Sharing information for the common good Investigations are often commercially sensitive and occasionally the subject of litigation. If information is to be shared through a public forum it is essential that it is fair, balanced and objective. The names of parties and products should not be given. Manufacturers' technical information In the first instance when there is a difficulty with a specific product or system it would be reasonable to expect the manufacturer to identify the problem and the means of resolution. This information should then be circulated to the manufacturer s own network of approved installers. This is a procedure that some manufacturers do follow and should be encouraged. With time and further experience it would be appropriate for the manufacturer to update their published technical literature, advising specifiers and designers of the known product limitations. This does not always happen. Building Research Establishment publications At one time the BRE was a publicly funded organisation that undertook much of the government sponsored research into the performance of buildings. Where there were common problems repeated then concise Defect Action Sheets were prepared, circulated and used as effective training material. One of the roles of the BRE is to offer an Advisory Service to the construction and property industry. Over the years an enormous amount of information was gathered about the behaviour of buildings, including the performance of roofing and cladding elements. Much of this knowledge and experience was retained by members of the BRE staff, many of whom have since retired or moved on to alternative employment. The findings from BRE studies on roofs were collated by Harrison (1996). The book was addressed to building surveyors and other professionals, such as architects and builders who maintain, repair, extend and renew the national building stock. Within the preface Harrison recognised that there is no shortage of written guidance on roofing as BRE has consistently said for many years. The problem is that people do not use the guidance that exists. Trade association information sheets One of the roles of a trade association is often to collate and circulate findings from problems reported by members, seeking to promote best practice. The Building Employers Confederation used to provide a technical advisory service to the building industry with technical bulletins issued quarterly between 1974 up until April 000. Sadly this service is no longer offered. Another organisation that has also recently closed is the Flat Roofing Alliance, which has been producing information sheets since

63 Contractor in-house advisory services Many of the large national general contractors used to maintain a central engineering support service that undertook internal investigations and then shared the findings within their company. This proved to be an effective way in which the company could learn from their own experiences, to avoid problems on future projects and thus improve efficiency and reputation. However, this central service came at a financial cost which, unfortunately, has been seen as non-essential by some and has fallen by the wayside as a result of budget cuts. Published journals and conferences Trade magazines publish technical articles, summarising findings from investigations and offering constructive advice. This includes the long running series of technical notes published in the roofing magazine RCI and circulated to 7,500 people within the cladding industry. Conferences such as CIB 10 are also an opportunity for delegates to listen and learn from the experiences of others, although again financial restrictions may limit attendance at such meetings and can be a barrier to this means of knowledge transfer. Internet discussion groups Looking to the future, further growth is seen in the use of the internet as a medium for transferring knowledge. In particular more widespread use of question & answer forums is foreseen which can give outline feedback and reports on problems encountered, often sharing overseas experiences. The challenge will be in moderating the information, attempting to check that the facts are fair, balanced and objective. 7. Conclusions There have been reports of intermittent rainwater leakage through laps and penetrations in composite panel roof systems laid to shallow falls and particularly on exposed sites. Site observations have identified sliding movement at end laps. The fixing and sealant details should be designed to accommodate these movements. The satisfactory performance of standard laps should be demonstrated for shallow pitched roofs using the water permeability tests set out in BS EN Penetrations such as rooflights and vents further increase the risks of unwanted water entry and can be minimised by introducing proper upstands. There is an opportunity to learn from these experiences. In the first instance it should remain the responsibility for the system manufacturer to maintain and keep up to date technical information regarding the installation and performance of their systems. Historically independent feedback on building defects has been provided by the BRE and trade associations, although this work is diminishing. Technical articles in journals and papers presented at conferences can be effective means to transfer knowledge. 59

64 In the future discussion forums on the internet are likely to be used for sharing information, although the need for independent moderation is recognised. By developing appropriate means to share feedback in a constructive way we can learn from experience and improve the reliability of the roofs we design and build. References O Connor PDT (00) Practical reliability engineering, Fourth Edition, Chichester, Wiley. Roberts K (007) Reliable roofing: the double layer roof, Proceedings of the International Conference on the Building Envelope and Systems Technology, March 007, pages , Bath, UK. Mann AP (008) Learning from failures at the interface, Proceedings of Institution of Civil Engineers, November 008, London, UK. Davies JM (001) Lightweight sandwich construction, Abingdon, Blackwell Science. Report 6 (1990) Preliminary European Recommendations for Sandwich Panels: Part II Good Practice, European Convention for Constructional Steelwork, Brussels. Report 115 (001) European Recommendations for Sandwich Panels: Part I Design, European Convention for Constructional Steelwork, Brussels. BS EN (006) Self supporting double skin faced insulating panels factory made products specification, British Standards Institute, London, UK. Technical Paper 16 (004) Guidance for the effective sealing of end lap details in metal roofing constructions Metal Cladding and Roofing Manufacturers Association, Wirral, UK. Harrison HW, Trotman and Saunders (009) Roofs and roofing: performance, diagnosis, maintenance, repair and avoidance of defects, Third Edition, Building Research Establishment Press, Garston, UK. 60

65 Openings in Sandwich Elements Warmuth, F. Institut für Stahlbau und Werkstoffmechanik, Technische Universität Darmstadt ( Lange, J. Institut für Stahlbau und Werkstoffmechanik, Technische Universität Darmstadt ( Abstract Sandwich panels with flat or lightly profiled faces often form the façade of a building. Due to different requirements like windows, doors or ducts, it is necessary to cut openings into the face of a building. At present, openings in sandwich panels require an additional substructure, which transfers the loads to the main structure. The aim of the project is to find a possibility to avoid this kind of substructure. Basically, there are two possibilities: For small openings in a panel, it seems to be obvious to look at the single element only. If the panel is not able to carry the load, it has to be strengthened. Therefore one solution could be a supporting frame, which is integrated in the opening and diverts the loads around it. Also additional profiles integrated in the joints - could be used to strengthen the element stiffness of the panel and to reduce the stress concentration in the corners of the openings. For larger openings it is necessary to include the neighbouring elements in the load transfer. Therefore, detailed information on the rigidity of the longitudinal joint and the torsional rigidity of the elements is needed. A calculation model shall be developed to describe the load transfer. First test results for a single element and an element formation with window openings are presented. These tests are part of the European research project EASIE ( In addition, comparative calculations with existing calculation proposals for openings without window frames are shown. The long-term objective is to generate a calculation model for sandwich elements with different kind of openings. This model should be only dependent on parameters which can be derived from some basic tests. The EASIE project has received financial support from the European Community s Seventh Framework Programme FP7/NMP-SE-008 under grant agreement No Keywords: sandwich, openings, load transfer, substructure, strengthening 61

66 1. Introduction Sandwich panels often form the outer shell of buildings. As façade elements they perform different tasks at the same time. Sandwich panels do not only seal the building, they also transfer the loads to any kind of substructure and form a very good thermal insulation.sandwich elements consist of two thin covering sheets, which enclose the core material. The sheets are usually made of galvanized steel. As core material, different materials with good insulating properties are used. The most common core materials are polyurethane (PUR) foam and mineral wool. Depending on the estimated use, the panels have thicknesses between 40 and 00 mm and a length up to 4 m. Due to different requirements like windows, doors or ducts, it is necessary to cut openings into the face of a building. These openings always result in an attenuation of the sandwich panel. At present, openings in sandwich panels require a substructure, which transfers the loads to the main structure. These replacements lead to an additional effort of erection and are visually not attractive. The aim of the presented project is to find a possibility to avoid this kind of substructure. In this paper, first test results of sandwich elements with windows are presented. Two different kinds of openings have been tested. In the first case, the opening has been cut in one panel and the window frame has been bonded in the opening. In the second case, the width of a complete panel has been replaced by a window. The load had to be transferred to the neighbouring elements.. State of the art / theory Basically, it can be distinguished between different kinds of openings. The most important differentiation is the size of an opening. We can distinguish between little openings like penetrations for supply lines and large openings for windows and doors. A second criterion is the location of the opening. From central openings in one panel to openings across longitudinal joints everything is possible. In recent years, some research projects on openings in sandwich panels were conducted. Especially for openings within one sandwich panel calculation proposals exist. In 1998 Courage/Toma (1998) published a possibility to calculate the stress peaks around openings and the consequences on the bearing capacity of the sandwich panel. Proposal Courage/Toma: If 0,1 0, 8 : In which σ K = wrinkling stress of the panel without opening σ N = maximum stress in the residual face sheet of the cross-section, referring to the complete panel β = b / B b = width of the opening B = width of the panel 6

67 Marc Böttcher (005) proposed a more simple approach, in which he also included openings, which are not arranged at midspan. Proposal Böttcher: If 0,0 < β < 0,4 : σ N (1- β) σ K If 0,4 < β < 0,8 : σ N 0,6 σ K In both publications, the wrinkling stress is reduced depending on the relation between the width of the opening and the width of the panel. Depending on the size of the opening, the results of the two calculation models are more or less similar. For little openings, these calculation models make sense. For larger openings, the formulas given above lead to a very high loss of bearing capacity. Interaction with neighbouring panels For large openings it is not useful, to confine the load transfer only to the weakened element. Especially at openings with the width of one panel (complete cut-out) there has to be an interaction with the neighbouring panels. The loads have to be transferred over the longitudinal joint and the unweakened neighbouring panels can participate in the load transfer. The German association for light metal structures IFBS (006) published a paper, how to calculate these kind of openings. In this calculation model it is assumed, that the load of the opening directly affects a torsional moment in the neighbouring element. For the verification of the system, the bearing capacity of the neighboured element and of the joint has to be proved.in all the publications mentioned above, only openings without any kind of strengthening were discussed. If windows are built into a sandwich façade, there exist always window frames, which almost certainly have a stiffening effect on the whole panel and especially on the covering sheet in the region of the window corners. That probably improves the load transfer. To check this assumption, panels with window frames have been tested. 3. Experimental investigations 3.1 Window-openings within one sandwich panel Test-set-up In a first test series sandwich panels with a window at midspan have been tested in a six point bending test. The width of the panels was 1 m, the window opening in each test 70 x 70 cm. Two different thicknesses of panels were tested. Four panels had a thickness of 60 mm and a total length of 4000 mm (test 1 to 4), two panels had a thickness of 10 mm and a total length of 6000 mm (test 5 and 6). In figure 1 the test set up is shown. Half of the panels were tested in a positive position (equates to wind pressure), the other half in a negative position (equates to wind suction). The important difference between these tests is the geometry of the window frame, which has some influence on the failure criteria. 63

68 Figure 1: Experimental set-up for test series 1 The deflections were measured at 5 different locations. Three displacement transducers were fixed at midspan (in the middle and left and right of the window frame), two at the corner of the window frame. In some of the tests, additional strain gauges have been applied around the window corner to get detailed information about the load transfer around the opening. Displacements 3.1. Results Load-displacement diagrams registered only very little differences of the displacement at all measured points. One of the load-displacement diagrams is exemplary shown in figure. The uniform displacement shows the stiffening effect of the window frame, which avoid different deflections around midspan. The displacement was nearly linear until the failure of the panel. 3,5 load of the hydraulic cylinder in kn 3,5 1,5 1 0,5 0 midspan 1 midspan w indow 1 w indow deflection at midspan ( x = 0 mm) and in the corner of the w indow ( x = 400 mm ) in mm Figure : Load-displacement-diagram of one test at midspan and in the corners of the window 64

69 Comparing the displacements of the test panels with the calculated displacements of an unweakened panel result in the following table: Table 1: Comparison of deflections with and without window frame Load level F = 1 kn x = 0 mm, deflection in mm Test Calculation Deviation without opening in % x = 40 mm, deflection in mm Test Calculation Deviation without opening in % 1 1,36 6,96 77,6 1,04 6,7 79,7 1,67 6,96 8,0 1,43 6,7 85,5 3 1,49 6,96 79,5 1,39 6,7 84,9 4 1,83 6,96 84,3 1,79 6,7 90,9 Load level max. load x = 0 mm, deflection in mm Test Calculation Deviation without opening in % x = 40 mm, deflection in mm Test Calculation Deviation without opening in % 1 35,6 0 78,0 35,1 19,3 81,9 33,65 18,7 79,9 33, ,7 3 37,81 1,1 79, 37,54 0,3 84,9 4 33,8 18,9 78,9 33,5 18, 84, Mode and level of failure: All the panels failed by wrinkling in the corner region of the opening. Depending on the geometry of the window frame the location of wrinkling was a bit different (see figure 3). The very stiff frame leads to wrinkling directly at the frame corner, where the panel is still complete over the whole width. In the case of the less stiff frame the face sheet wrinkles close to the corner of the opening in the panel. For the same panel thicknesses, the load levels by failure are quite similar for each panel. Figure 3: Different locations of wrinkling 65

70 Table : Comparison of failure loads with and without opening Test Thickness of the panel Failure load in kn Calculated failure load of the complete panel in kn Deviation in % ,44 10,81 68, ,5 10,81 69, ,6 7,94 54, ,7 7,94 58, ,80 18,99 64, ,85 13,06 6,83 Allocation of stresses: Strain gauges have been applied in the region around the window corner to get detailed information about the load transfer around the opening. Exemplary one test-set- up is shown in figure 5. Allocation of stress for different load levels (opening 700 mm) 0,00-50,00 Stress in N/mm² -100,00-150,00-00,00 1 kn 3 kn 5 kn 5,8 kn -50,00-300, distance from the longitudinal axis in mm Figure 4: Allocation of stress for different load levels Figure 5: Strain gauge position 66

71 In figure 4 the allocation of stresses along the width of the panel directly in front of the window frame is shown for different load levels. In the diagram, only half of the panel width is presented. Two interesting things can be pointed out. 1) The openings with a size of 70 cm and the window frame of 80 cm lead to explicit stress peaks in the region around the window corner. With increasing loads, the peaks become more definite. ) In the middle of the panel the stress is increasing with higher loads, but has still a very low level. That means, in spite of the high stiffness of the window frame, the main part of the load is transferred through the covering sheet to the outer continuous part of the panel. 3. Window-opening in a panel interconnection system 3..1 Test-set-up In a second test series an interconnection system with 3 sandwich panels and a window in the middle panel has been tested. The width of the panels was 1m, the window opening was a complete panel width (1 m) and 1 m length. Like in test series 1, two different thicknesses of panels were tested in a positive and a negative position. In figure 6 the test set up is shown. Figure 6: Experimental set-up for test series Displacements 3.. Results Whole systems with windows lead to much higher deflections than sandwich panels without openings or windows. For the tests with 60 mm thick panels a comparison of the calculated displacements without window and the measured displacements with window is shown in figure 8. In some tests, 67

72 particularly the negative position, the window frame slipped out of the system. That explains the different deflections between neighbouring points of measuring in figure 7. Deflection in longitudinal direction ,1 0, 0,3 0,4 0,5 Deflection in mm w ithout w indow 60_1 60_ 60_3 60_4 Length / Total length Deflection in lateral direction in x=0 (midspan) 0 Deflection in mm w ithout w indow 60_1 60_ 60_3 60_ Distance from the longitudinal axis in mm Figure 7: Deflections for the load level 4 kn, 60 mm panels and a span of 3900 mm Mode and level of failure: All the panels failed by an overstressing of the longitudinal joint and a following collapse of the complete system including wrinkling of the outer panels. In some cases the window frames slipped out of the joint. The slipping of the frames did not lead to an immediate collapse of the system, but certainly had an influence on the failure load. As the slipping is a question of an exact fabrication of 68

73 the window it can be avoided. Probably according to the inexactnesses of the windows sizes, the failure load levels were different for each system. Figure 7: Collapse of the system Table 3: Comparison of failure loads with and without window frame Test Thickness of the panel Failure load in kn Calculated failure load of the complete panel in kn Deviation in % Mode of failure ,30 15,94 41,66 Complete collapse 60 8,5 15,94 46,68 Complete collapse ,7 1,91 48,10 Slipping out of the window, then collapse ,5 1,91 65,14 Complete collapse without load on the middle panel ,9,59 47,3 Complete collapse ,0 18,05 55,68 Slipping out of the window, then collapse 69

74 Allocation of stresses: Strain gauges have been applied to get detailed information about the load transfer around the window. Exemplary one test-set- up is shown in figure 8. Figure 8: strain gauge position Longitudinal stresses cross the panels Stresses in N/mm² _1 60_ 60_3 60_4-140 Distance from the longitudinal axis in mm Figure 9: Allocation of stress for load level 4 kn In figure 9 the allocation of stresses along the width of the panels directly in front of the window frame is shown for the load level 4 kn. In the diagram, only half of the test set-up width is presented. In each test, the characteristics of the stresses are similar. In the medial panel the stresses are nearly zero, all the load is transferred to the outer panels. There, a concentration of stress can be identified in the region close to the longitudinal joint. 70

75 4. Comparison with existing methods of calculation 4.1 Evaluation of the tests with one panel As described in chapter, different calculation proposals exist for single sandwich panels with openings. In table 4, the failure loads of the tests are compared to calculation results for openings without frames according to calculation proposals of Böttcher (006) and Courage/Toma (1998). Table 4: Comparison of the failure load in kn between test and calculation Failure load in kn Tests Böttcher Deviation in % Courage/Toma Deviation in % 60_1 3,44 1,9 79, _ 3,5 1,9 69, _3 3,60 1, , _4 3,7 1, , _1 6,80 3, , _ 4,85,3 109,87 69 The comparison shows, that window frames have a positive effect on the load bearing capacity. The panels carry up to 155 % more load than the calculation results without window-frames suggest. 4. Evaluation of the tests with three panels In IFBS (006) a proposal exists for calculating a 3 panel system with an opening. According to this proposal, there are three possible failure modes: Shear failure in the outer panel, wrinkling of the outer panel and failure of the longitudinal joint. By recalculating the tests according to the IFBS paper, one asserts that the longitudinal joint is always the critical failure mode. In table 5 the degree of capacity utilization for the different failure modes is shown for a load level of 4 kn per panel. Table 5: Degree of capacity utilization for different failure modes Load level: 4 kn Shear failure Wrinkling Failure of the joint 60 mm panel 0,31 0,38 0,40 10 mm panel 0,18 0,6 0,53 For the bearing strength of the joint, in IFBS (006) only a rough estimate is given. That could be the reason for significant deviations between some of the test results and the recalculation, like shown in table 6. 71

76 Table 6: Comparison between tests and calculations Failure load per panel in kn 60_1 60_ 60_3 60_4 10_1 10_ Test result 9,3 8,5 6,7 4,5 11,9 8 Calculation result 10,0 10,0 7,9 7,9 7,5 6,0 Deviation -7 % -15 % -16 % - 40 % + 59 % + 33 % 5. Conclusions Test results of sandwich elements with windows have been presented in this paper. Two different kinds of openings were tested. The results of test series 1 show, that window frames have a positive effect on the load bearing capacity of single sandwich elements with openings. The failure load can be clearly increased in comparison to panels with frameless openings. Nevertheless, for larger openings the loss of bearing capacity is very high. In the second series, the width of a complete panel has been replaced by a window. The entire load was transferred to the outer panels. The maximum loads in the second series varied within a wide range. The main reason for that is evidently the slipping of the window frames out of the joints. An interesting point is the bearing capacity of the longitudinal joints. Further tests will show, if a more exact knowledge about the bearing capacity of the joint bring better correlations between the tests and the calculations. At the end of the project, there shall exist a calculation model for sandwich elements with different kind of openings. This model should be only dependent on parameters which can be derived from some basic tests. References Böttcher M. (005) Wand-Sandwichelemente mit Öffnungen, Publication of the institute for steel construction and material mechanics, TU Darmstadt Courage, Toma (1998) Structural detailing of opening in sandwich panels, Publication of the European Communities IFBS (006) Berechnungsverfahren für Wand-Sandwichelmente mit Öffnungen, Publication 5.09 of the IFBS, Düsseldorf 7

77 Optimization of Geometry and Core Materials of Sandwich Panels with Metallic Faces Kurpiela, A. Institute of Steel Construction and Material Mechanics, Technische Universität Darmstadt ( Lange, J. Institute of Steel Construction and Material Mechanics, Technische Universität Darmstadt ( Berner, K. Institute for Sandwich Technology, University of Applied Sciences Mainz ( Abstract Due to their numerous advantages, sandwich panels are increasingly used. However, a scientifically sound optimization of sandwich panels has not yet been done. Sandwich producers have made selectively optimization works of their products but mainly with regard to the manufacturing procedures. The aim of the optimization in the current project is an adjustment of the geometry of metal faces together with the properties of the core material in such a way that the load bearing capacity and the costs of sandwich panels are optimal. The production technology should be considered as well. This topic is very extensive mainly due to the numerous parameters which determine the load bearing capacity of sandwich panels. The optimal geometry of the metal sheets and the optimal mechanical properties of the core material have to be developed on the basis of theoretical investigations and checked in following mechanical tests. In the first step the sets of achievable mechanical properties of the used materials are determined. After this, a combination of the mechanical properties and the geometry of the sandwich panels can be created in order to obtain the largest span. In this regard it is important to achieve a high load bearing capacity of the panel and high utilization factors of all mechanical properties at the same time. The conducted investigations demonstrate possibilities for the optimization of sandwich panels. An achievement of larger spans without a significant increase of costs is possible only by changing the core properties. Furthermore, a significant potential for the optimization of the metal sheets was recognised. There is a possibility to diversify both the quality and the geometry of the steel sheets. The numerous parameters which determine the properties of sandwich panels turn the optimization into a complex procedure. Yet, on the other hand, they also provide numerous alternatives in the optimization processes. Keywords: optimization, sandwich panel, core material, metal face, load bearing capacity 73

78 1. Introduction 1.1 Sandwich panel construction Sandwich panels used in civil engineering consist typically of a thick core with low density between two thin high density faces. The materials can be configured in many possible combinations. For example, for the cover layers thin metal faces, timber based plates or glass fibre reinforced plastics can be used. The insulating core layer in most cases is made of structured foams like polyurethane (PUR), polystyrene (PS) or of mineral wool (MW). Sandwich panels are used as light-weight roofs and wall claddings in industrial and commercial buildings. Usually the structures are loaded by permanent loads, like self-weight, snow and wind loads. However, additional loads like temperature differences between external and internal metal faces or creep of the core must be taken in account for statical calculations for sandwich panels. The high load bearing capacity of sandwich panels is the result of a rigid connection between the core material and the cover layers. The bending moment is distributed to the two faces (e.g. for panels with flat faces in the form of axial forces) and the shear loads are borne by the core layer. The optimization of sandwich panels in respect to the load bearing behaviour means to adjust the mechanical properties of the core material and metal faces in such way that the largest possible spans can be reached. Other important structural properties of sandwich panels (e.g. insulating properties) or the manufacturing conditions should be considered at the same time. The current optimization was done on sandwich panels with cover layers made of lightly profiled steel sheets and a PUR core. In this case the thickness of the core can be between 40 and 300 mm. The thicknesses of the steel faces vary between 0.4 and 1.0 mm. 1. Optimization of sandwich panels with metal faces necessity and possibilities Saving money and resources is very important in almost every industry sector. In this sense, application-oriented research has the role to find the optimal exploitation of materials in the manufacturing process. Compared with steel, concrete or timber constructions, sandwich panels are relatively new components. The first applications of sandwich construction were implemented in the seventies of the last century. That is the reason why sandwich constructions were previously not investigated as deeply as the other construction types mentioned above. This is also the reason why no investigations into the optimization of sandwich panels have been done before. In this paper a possible method for the optimization of sandwich panels in respect to their load bearing capacity is described. 74

79 There are two main assumptions concerning the optimization of sandwich panel load bearing. Firstly: all mechanical properties of sandwich panels shall be adjusted in such way that the largest possible spans can be reached. Second: all the mechanical properties shall be maximally exploited. These two assumptions, together with the several statical systems and different possible load cases that possibly occur, turn the optimization of sandwich panels into a complicated process. The several statical systems which can be used, and the different loads that can occur, generate different failure modes as being decisive for the determination of the span widths. According to the sandwich theory, the calculation method for sandwich panels shapes the background for the optimization calculations. In this respect the following main properties of sandwich structures describe the load bearing behaviour. These are the crucial factors in the calculation method: Wrinkling stress of the metal face σ w Compressive strength of the core f Cc Shear strength of the core f Cv Young modulus of the core E C Shear modulus of the core G C Additionally the shear strength and shear modulus at high temperature and the shear strength for long term behaviour can be expressed as a function of the adequate main values, which were measured at room temperature. For a determination of these properties, experimental tests on each panel type have to be done. The exact relations between the several mechanical properties are not known. Furthermore, there are no constitutive equations that describe how the achievable wrinkling stress of the metal face depends on the properties of the used materials and the face geometry. It makes the determination of the load bearing capacity only in an analytical way (without tests) to an impossible process. The achievable wrinkling stress of sandwich panels depends on many parameters. These are not only the used materials with defined properties. A very important point is the bond between the core layer and the cover sheets, which can be influenced by many factors. For example the production procedures or the adhesive that is used in the case of adhesively bonded panels can be crucial for the final properties of each single panel type. To undertake an optimization of sandwich panels by the way of an investigation, a theoretical model for describing the achievable wrinkling stress should be made. In this model the mechanical material properties of the core and also the geometry and the mechanical properties of the metallic faces have to be considered. Because of the existing material and production imperfections the theoretical model can be used only for an estimation of wrinkling stress values. 75

80 Because of numerous different factors, which impact the mechanical behaviour of sandwich panels and additionally because of the lack of constitutive equations describing the behaviour, the existing mathematical optimization methods cannot be rationally used at the moment. As an approach, in the first step a parameter analysis according to the calculation method for sandwich panels shall be done. The aim of the current project is the development of an optimization method that will consider both the core material and the geometry of the faces in only one step. However, due to the mentioned technical hitches, the optimization of sandwich panels will be split into the two steps of an optimization of the core and an optimization of the metallic faces.. Optimization of core materials.1 Theoretical pre-analysis into optimization of core materials In this paper the investigation into optimization of one of the mostly used core materials polyurethane will be shown. The optimization of the core material means to define the mechanical values the strengths and the modulus (like in 1.) which are needed for getting an optimal core material. In the first step it is possible to execute the optimization of core materials for existing panel geometries. In this case any existing panel with a given geometry can be taken and the mechanical properties of the core can be optimized in such a way that the given geometry and the new core material bring an optimal panel regarding to the load bearing capacity. The mechanical properties are included in certain sets. In this case the optimization of core material means to search for combinations of mechanical properties from defined sets. The created combinations should lead to a high load bearing capacity of the panel and maximal utilization factors of all mechanical properties by limitation of the spans. First of all the sets of mechanical properties of core material (PUR) were determined according to known material properties. The background for this procedure is that the material combinations should be set up of material properties which are possible to manufacture. For instance the shear modulus depending on the density of the core material is shown in Figure 1a. After all relevant mechanical properties were collocated in this way the boundaries for achievable mechanical properties of the core were defined (see Figure 1b). Depending on the chosen density, the possible scope of mechanical properties is known and the raster for creation of new combinations can be defined. In this case a raster of five values was chosen. 76

81 Shear modulus G C in N/mm² 8,0 7,0 6,0 5,0 4,0 3,0,0 1,0 0, a) b) Figure 1a: Values of shear modulus of PUR depending on the core density; Figure 1b: Example for taken value raster for creating the combinations of core properties, density of PUR 40 kg/m³; The known scopes for all relevant mechanical values can be determined: PUR, Density 40 kg/m³ Density PUR in kg/m³ Shear modulus G C in N/mm² 8,0 7,0 6,0 5,0 4,0 3,0,0 1,0 0, Density PUR in kg/m³ G C 1,9; 5,5 in N/mm² E C 1,0; 5,0 in N/mm² f Cc 0,060; 0,00 in N/mm² f Cv 0,085; 0,5 in N/mm² For creating the properties combinations for all values scopes a raster of five was chosen Table 1: Creation of combinations of mechanical core properties for PUR, density 40 kg/m³ G C, T=0 C G C,T>0 C f Cv,T=0 C f Cv,T>0 C f Cv,t= f Cc E C,T=0 C in N/mm² in N/mm² in N/mm² in N/mm² in N/mm² in N/mm² in N/mm² 1,9 1,7 0,085 0,077 0,038 0,060 1,0,8,5 0,10 0,108 0,054 0,095,0 3,7 3,3 0,155 0,140 0,070 0,130 3,0 4,6 4,1 0,190 0,171 0,086 0,017 4,0 5,5 4,9 0,5 0,03 0,101 0,00 5,0 The combination of the main values (marked columns in Table 1) gives 65 compositions of mechanical properties. 77

82 65 sandwich panels with the same face geometry and different core materials were generated. By determination of the spans for all 65 fictive panels one of them can result as optimal. The optimal panel should give the largest span and the mechanical properties of the core should be exploited maximal. Example: Sandwich panel with given geometry of faces and a PUR core; Core density 40 kg/m³ Thickness of the panel D = 80 mm External face, steel, micro lined profile, t F1 = 0,60 mm; Interior face, steel, lined profile, t F = 0,50 mm; Statical system: Vertical wall panel, one span beam q (wp /ws) in kn/m² L Table : Mechanical properties of the core; basic panel and optimal combination Element G C, T=0 C G C,T>0 C f Cv,T=0 C f Cv,T>0 C f Cv,t= f Cc E C,T=0 C in N/mm² in N/mm² in N/mm² in N/mm² in N/mm² in N/mm² in N/mm² Basic element 3,6 3,6 0,110 0,110 0,050 0,100 3,7 Optimal core material 5,5 5,0 0,189 0,170 0,085 Depend on support width 5,0 Load-span curve 6 6 Load in kn/m fcv=0,189 N/mm² fcv=0,157 N/mm² basic element element with optimal core for defined case 1 fcv=0,09 N/mm² Span length l v in mm 0 Figure : Load-span curve, example for basic panel with given core and a panel with optimal core As a means to simplify the analysis of the calculations results, the achievable spans were shown in dependency on the increasing loads, like in Davies (001), see Figure. 78

83 The results of the investigations show that the increase of the span is possible even only due to an improvement of the core properties. But in the most relevant cases the wrinkling stress or the deflection limitation are decisive for the determination of the span of sandwich panel. This can be observed in the area of low load (between 0,5 and 1,5 kn/m²). The wrinkling stress depends on the core material, but even more on the properties of the metal faces. That is the reason for a deeper investigation into optimization of the geometry of metallic faces in respect to getting higher wrinkling stress. 3. Optimization of lightly profiled metallic faces 3.1 Background The optimization of metallic faces means to adjust the geometry and the quality of the metal face in such a way that the achievable wrinkling stress of the face will be maximal. Many types and forms of metal faces are produced today. There are flat faces and lightly profiled faces for wall elements and strongly profiled faces for roof elements. This paper regards lightly profiled faces only. The optimization of metal faces is shown on example of the ribbed profile. bp1 br bp br tf hr Figure 3: Lightly profiled face of sandwich panel, ribbed profile During the mechanical tests done on numerous types of sandwich panels for their certification it was observed that the achievable wrinkling stress changes depending on different geometries. It can be supposed that deeper ribs (h R in the Figure 3) cause higher wrinkling stress. But, the exact connection between the rib depth and the achievable wrinkling stress is not known. Furthermore, which influence have other geometry factors, e.g. the plate widths (b P ) or the face thickness (t F ) on the achievable wrinkling stress? Beside the influence of the geometry the elastic bedding on the core through its mechanical properties should be considered as well. The influence of the geometry of metal sheets on the achievable wrinkling stress was done both in theoretical investigations and experimental tests. 3. Theoretical pre-analysis into optimization of metal faces At the moment in optional calculation of wrinkling stress, the lightly profiled faces are regarded as flat faces. In this way their explicit higher load capacity is not regarded. To achieve an optimization of the geometry, a new analytical model for an assessment of the wrinkling stress of lightly profiled 79

84 faces depending on the geometry of the profiling was compiled. The model is based on the existing theories by Plantema (1966) and Stamm/Witte (1974). ax x x w0 z x c=k K(EC,GC, C) Figure 4: Stresses and deformation on flat, constantly bedded sandwich panel structure To the differential equation for flat plates a term for consideration of the bedding of the core was added. B F (w '''' w '' w ) Fx w '' Fy w Fxy w ' c w 0 (1) Where B 1t 1 (1 ) 3 F F F stiffness of the plate () t, thickness and Poisson ratio of the face F F c x K rigidity constant of the bedding (3) By consideration of the behaviour conditions, like deformation of flat plate on elastic bedding and the failure mechanism, the differential equation can be reduced to the searched axial force in the plate: where c like in (3), c F B (4) x F x x x a x is the half cycle rate in longitudinal direction (5) and K (1 C) (1 ) (3 4 ) C C E C is the bedding constant (6) where C is the Poisson ratio of the core. 80

85 For consideration of the shear modulus in the elastic bedding due to the core, the relation between shear modulus and young modulus for isotropic material was used EC (1 C) G C (7) and the bedding constant was transformed to (1 ) G E C C C K (6a) (3 4 ) (1 ) C C Inserting the rigidity constant of the bedding in (4) gives K F B (8) x F x x The lowest value for F x is the buckling load of the flat plate F ki : df d ki x 0 (9) 3 3 Fki K BF (10) In the case of a lightly profiled face, the failure mechanism is similar to a flat face. Nevertheless, the ribs should be additionally considered as stiffeners. For this reason the lightly profiled face was divided in the areas of stiffeners (ribs with effective widths of plates 1 ) and flat plates between stiffeners like in Figure 5. The lightly profiled face is considered as a configuration of beams and flat strips on elastic bedding. The buckling load for each beam with the partial cross sections like in Figure 5 can be taken like in (10): 3 3 Fki,i K Bi (11) 1 Effective widths of plates were determined according to DIN ,

86 bp1 br bp br ReH tf hr c=k K(EC,GC, C) beff1/ beff1/ beff/ beff/ A1, Beff,1 A, Beff, A3, Beff,3 A4, Beff,4 Fki EFIi Ai Fki Lv c Figure 5: Lightly profiled face of sandwich panel; fragmentation of ribbed profile into beam on elastic bedding with cross section A i By consideration of stressed cross-section area A i the wrinkling stress of each single part of the face equals: w,i F ki,i A i (1) 3 K B 3 w,i eff,i td (13) and w,i R eh The partial wrinkling stresses were spread in the complete cross-section area and the total wrinkling stress of a lightly profiled ribbed face is: w w,i A i A i (14) 3.3 Testing of panels with lightly profiled faces For investigation of the influence of the depth of ribs by lightly profiled faces on the bending moment capacity full scale tests on simply supported panels were done. The tests were made according to EN (006), A.5. Sandwich panels with steel faces and a core of PUR were tested. The specimens were taken from one production batch. This provided the possibility to keep all parameters constant. Only the depth of the ribs was varied. h R b 1 = b = 50 mm, b 1 b b R b R =,65 mm; t F = 0,60 mm Figure 6: Test specimen; ribbed profile 8

87 Table 3: Test results for specimens with varied depth of the ribs Type of panel Nominal thickness of panel Tested face geometry Depth of the ribs Wrinkling stress Increasing of wrinkling stress by varied depth of the ribs Wall d = 80 mm 1.1. Ribbed profile h R = 0,63 mm 170,4 N/mm² - d = 80 mm 1.. Ribbed profile h R = 0,90 mm 188,5 N/mm² 10,6 % d = 80 mm 1.3. Ribbed profile h R = 1,6 mm 08, N/mm², % According to the test results the achievable wrinkling stresses (average values) of the metal faces were determined. The results show a clear connection between the depth of the ribs by lightly profiled metal faces and the achievable wrinkling stress. There is the possibility to increase the wrinkling stress considerably by insignificantly deepening the profiling. With regard to the optimization processes is it important that the increasing of the wrinkling stress is possible even by using the same mass of steel and core material. The wrinkling stress for the tested panels was calculated according to (14). In Figure 7 the calculated wrinkling stresses and the corresponding test results are presented. There is the varied parameter - depth of the ribs - presented on the abscissa and the achievable wrinkling stress values are shown on the ordinate. 300 Wrinkling strength in N/mm² Calculation Test 0 0,0 1,0,0 3,0 4,0 Depth of the rib h R in mm Figure 7: Wrinkling stress depending on the depth of the profiling by ribbed profile 4. Conclusions A method for optimization of sandwich panels with thin metallic faces in respect to their load bearing behaviour was shown. There is the possibility to optimize both the core material and the geometry of 83

88 the metallic faces. The connection of those two methods will give a possibility for optimizing sandwich panels in only one step. It has been found that very often, for the relevant cases of loading, the wrinkling stress or deflection limit is decisive for limitation of the span widths. These two values can not be significantly improved by only changing the core material. For this reason the improvement of the wrinkling stress and the reduction of the available deflection should be taken as important points for optimizing the cover layers. The changes on the face geometry can lead to a higher wrinkling stress. Additionally the selection of optimal strength and stiffness values of the core material can increase the load bearing capacity. The numerous factors which influence the mechanical behaviour of sandwich panels provide many options for an improvement of the load bearing capacity. As a practical effect, the producers are able to implement selective changes for getting better load bearing of their products. References Davies J M et al (001) Lightweight sandwich construction, Oxford, Blackwell Science Ltd. Plantema F J (1966) Sandwich Construction, London, John Wiley and Sons, Inc. Stamm/Witte (1974) Sandwichkonstruktionen, Berechnung, Fertigung, Ausführung, Vienna, Springer. DIN , Stahltrapezprofile Allgemeine Anforderungen, Ermittlung der Tragfähigkeit durch Berechnung, Juni 1987 EN 14509, Self-supporting double skin metal faced insulating panels - Factory made products - Specifications; November

89 International Council for Research and Innovation in Building and Construction CIB s mission is to serve its members through encouraging and facilitating international cooperation and information exchange in building and construction research and innovation. CIB is engaged in the scientific, technical, economic and social domains related to building and construction, supporting improvements in the building process and the performance of the built environment. CIB Membership offers: international networking between academia, R&D organisations and industry participation in local and international CIB conferences, symposia and seminars CIB special publications and conference proceedings R&D collaboration Membership: CIB currently numbers over 400 members originating in some 70 countries, with very different backgrounds: major public or semi-public organisations, research institutes, universities and technical schools, documentation centres, firms, contractors, etc. CIB members include most of the major national laboratories and leading universities around the world in building and construction. Working Commissions and Task Groups: CIB Members participate in over 50 Working Commissions and Task Groups, undertaking collaborative R&D activities organised around: construction materials and technologies indoor environment design of buildings and of the built environment organisation, management and economics legal and procurement practices Networking: The CIB provides a platform for academia, R&D organisations and industry to network together, as well as a network to decision makers, government institution and other building and construction institutions and organisations. The CIB network is respected for its thought-leadership, information and knowledge. CIB has formal and informal relationships with, amongst others: the United Nations Environmental Programme (UNEP); the European Commission; the European Network of Building Research Institutes (ENBRI); the International Initiative for Sustainable Built Environment (iisbe), the International Organization for Standardization (ISO); the International Labour Organization (ILO), International Energy Agency (IEA); International Associations of Civil Engineering, including ECCS, fib, IABSE, IASS and RILEM. Conferences, Symposia and Seminars: CIB conferences and co-sponsored conferences cover a wide range of areas of interest to its Members, and attract more than 5000 participants worldwide per year. Leading conference series include: International Symposium on Water Supply and Drainage for Buildings (W06) Organisation and Management of Construction (W065) Durability of Building Materials and Components (W080, RILEM & ISO) Quality and Safety on Construction Sites (W099) Construction in Developing Countries (W107) Sustainable Buildings regional and global triennial conference series (CIB, iisbe & UNEP) Revaluing Construction International Construction Client s Forum CIB Commissions (August 010) TG58 Clients and Construction Innovation TG59 People in Construction TG6 Built Environment Complexity TG63 Disasters and the Built Environment TG64 Leadership in Construction TG65 Small Firms in Construction TG66 Energy and the Built Environment TG67 Statutory Adjudication in Construction TG68 Construction Mediation TG69 Green Buildings and the Law TG71 Research and Innovation Transfer TG7 Public Private Partnership TG73 R&D Programs in Construction TG74 New Production and Business Models in Construction TG75 Engineering Studies on Traditional Constructions TG76 Recognising Innovation in Construction TG77 Health and the Built Environment TG78 Informality and Emergence in Construction TG79 Building Regulations and Control in the Face of Climate Change TG80 Legal and Regulatory Aspects of BIM TG81 Global Construction Data W014 Fire W018 Timber Structures W03 Wall Structures W040 Heat and Moisture Transfer in Buildings W051 Acoustics W055 Construction Industry Economics W056 Sandwich Panels W06 Water Supply and Drainage W065 Organisation and Management of Construction W069 Housing Sociology W070 Facilities Management and Maintenance W077 Indoor Climate W078 Information Technology for Construction W080 Prediction of Service Life of Building Materials and Components W083 Roofing Materials and Systems W084 Building Comfortable Environments for All W086 Building Pathology W089 Building Research and Education W09 Procurement Systems W096 Architectural Management W098 Intelligent & Responsive Buildings W099 Safety and Health on Construction Sites W101 Spatial Planning and infrastructure Development W10 Information and Knowledge Management in Building W104 Open Building Implementation W107 Construction in Developing Countries W108 Climate Change and the Built Environment W110 Informal Settlements and Affordable Housing W111 Usability of Workplaces W11 Culture in Construction W113 Law and Dispute Resolution W114 Earthquake Engineering and Buildings W115 Construction Materials Stewardship W116 Smart and Sustainable Built Environments W117 Performance Measurement in Construction PAGE 1

90 International Council for Research and Innovation in Building and Construction Publications: The CIB produces a wide range of special publications, conference proceedings, etc., most of which are available to CIB Members via the CIB home pages. The CIB network also provides access to the publications of its more than 400 Members. CIB Annual Membership Fee Membership will be automatically renewed each calendar year in January, unless cancelled in writing 3 months before the year end Fee Category FM1 Fee level FM Fee level FM3 Fee level AM1 Fee level AM Fee level IM Fee level All amounts in EURO The lowest Fee Category an organisation can be in depends on the organisation s profile: Recent CIB publications include: Guide and Bibliography to Service Life and Durability Research for Buildings and Components (CIB 95) Performance Based Methods for Service Life Prediction (CIB 94) Performance Criteria of Buildings for Health and Comfort (CIB 9) Performance Based Building 1st International State-of-the- Art Report (CIB 91) Proceedings of the CIB-CTBUH Conference on Tall Buildings: Strategies for Performance in the Aftermath of the World Trade Centre (CIB 90) Condition Assessment of Roofs (CIB 89) Proceedings from the 3rd International Postgraduate Research Conference in the Built and Human Environment Proceedings of the 5th International Conference on Performance-Based Codes and Fire Safety Design Methods Proceedings of the 9th International Symposium on Water Supply and Drainage for Buildings Agenda 1 for Sustainable Development in Developing Countries R&D Collaboration: The CIB provides an active platform for international collaborative R&D between academia, R&D organisations and industry. Publications arising from recent collaborative R&D activities include: Agenda 1 for Sustainable Construction Agenda 1 for Sustainable Construction in Developing Countries The Construction Sector System Approach: An International Framework (CIB 93) Red Man, Green Man: A Review of the Use of Performance Indicators for Urban Sustainability (CIB 86a) Benchmarking of Labour-Intensive Construction Activities: Lean Construction and Fundamental Principles of Working Management (CIB 76) Guide and Bibliography to Service Life and Durability Research for Buildings and Components (CIB 95) Performance-Based Building Regulatory Systems (CIB 99) Design for Deconstruction and Materials Reuse (CIB 7) Value Through Design (CIB 80) Themes: The main thrust of CIB activities takes place through a network of around 50 Working Commissions and Task Groups, organised around four CIB Priority Themes: Sustainable Construction Clients and Users Revaluing Construction Integrated Design and Delivery Solutions FM1 FM FM3 AM1 AM IM Full Member Fee Category 1 Multi disciplinary building research institutes of national standing having a broad field of research Full Member Fee Category Medium size research Institutes; Public agencies with major research interest; Companies with major research interest Full Member Fee Category 3 Information centres of national standing; Organisations normally in Category 4 or 5 which prefer to be a Full Member Associate Member Fee Category 4 Sectoral research & documentation institutes; Institutes for standardisation; Companies, consultants, contractors etc.; Professional associations Associate Member Fee Category 5 Departments, faculties, schools or colleges of universities or technical Institutes of higher education (Universities as a whole can not be Member) Individual Member Fee Category 6 Individuals having an interest in the activities of CIB (not representing an organisation) Fee Reduction: A reduction is offered to all fee levels in the magnitude of 50% for Members in countries with a GNIpc less than USD 1000 and a reduction to all fee levels in the magnitude of 5% for Members in countries with a GNIpc between USD , as defined by the Worldbank. (see org/datastatistics/resources/gnipc.pdf) Reward for Prompt Payment: All above indicated fee amounts will be increased by 10%. Members will subsequently be rewarded a 10% reduction in case of actual payment received within 3 months after the invoice date. For more information contact CIB General Secretariat: secretariat@cibworld.nl PO Box 1837, 3000 BV Rotterdam, The Netherlands Phone ; Fax PAGE