Beech Glulam - Investigations of surface preparation and gluing

Beech Glulam - Investigations of surface preparation and gluing Bettina Franke 1, Christian Lehringer 2, Andreas Müller 3, Flavien Sauser 4 ABSTRACT: In Central Europe, beech wood is the main available hardwood species and its use is also broadly advertised for engineered wood products. Used as engineering material, beech wood offer a great potential for high performance timber structures. However, the physical material behaviour of beech wood and its differences from the material properties from commonly used softwoods require specially adapted production processes for engineered wood products and therefore among other reasons delay the broad introduction into the market. The presented results reveal some obstacles and chances for the production of glued laminated timber. The test programme comprises the investigation of different surface preparation techniques as well as the use of different certified adhesives. The characterisation of the different arrangements is based on the requirement of the production of glued laminated timber according to EN 14080:2013. The results show that it is generally possible to produce high quality beech glued laminated timber with commercially available adhesives. KEYWORDS: Glued laminated timber, Cross laminated timber, Beech, Adhesives, Shear strength, Delamination 1 INTRODUCTION 123 Glued laminated timber (Glulam) or cross laminated timber (CLT) are well known and established products in wood engineering. The mainly used products in the timber construction sector are based on softwood materials. Currently they are no standard products of Glulam or CLT in hardwood. Thus, individual solutions are needed, making hardwood constructions complicated, expensive and less attractive for use. Market researches from Bern University of Applied Sciences show a high potential of beech wood products, [1]. So far there are, in general, no established parameters like price, potential, production amount, delivery time and/or connections available for hardwood products. However, the potential of beech Glulam on the market can be approximated for ten years after industrial production start to 3.900 m 3 up to a maximum of 25.400 m 3 per year depending on the scenarios, [1]. Regarding the opinion of experts, the maximum amount can be achieved because beech Glulam is then available as a 1 Bettina Franke, Bern University of Applied Sciences, Institute for Timber Construction, Structures and Architecture, Solothurnstrasse 102, 2504 Biel, Switzerland. Email: bettina.franke@bfh.ch 2 Christian Lehringer, Henkel & Cie. AG, Industriestrasse 17a, 6203 Sempach Station, Switzerland. Email: christian.lehringer@henkel.com 3 Andreas Müller, Bern University of Applied Sciences 4 Flavien Sauser, Bern University of Applied Sciences high performance and high-tech product which, moreover, can show the full strength for many engineering areas in combination with a good marketing. The field of application of beech Glulam ranges from replacement of steel or concrete columns or members in multi-story buildings, realisation of large span halls, but also in smaller halls to reach higher effective useable volume due to reduced cross sections, and in family homes for column-free floor plans. Compared to softwoods, the mechanical properties of beech show higher values. For example, the strength values in bending and compression parallel to the grain are up to 175% respectively 150% higher of beech wood compared to softwood. The higher natural strength is based on the correlation of the density and the mechanical material behaviour. With 690 kg/m 3, the average raw density of beech wood is significantly higher than e.g. Norway spruce (450 kg/m 3 ). At the moment, the natural higher strength potential of beech wood is mainly used for partial reinforcements in timber structures, e.g. for strengthening the lateral compression strength at supports or loading plates or the tension strength perpendicular to grain at notches and holes or in tapered and curved beams [2]. Currently, there grows more wood than been used in the Swiss forest. Especially the hardwood stock has been increasing since 1995. 31% of the entire wood stock is hardwood, where the biggest part of 18 % counts for beech wood [3]. In Switzerland, hardwood is mainly sawn to unedged boards 38%, [4]. Caused by the

anatomy and the crooked trunks only 40-50% of the log can currently be used. For the production of engineered wood products such as Glulam or CLT, it must be considered that hardwood differs from softwood in terms of growth habit and growth cycle as well microscopic structure. Therefore, technologies and production processes of softwood cannot be used one-to-one and need adaptations and further inventions. Hence, the production of Glulam and CLT is almost limited to softwood and only some guidelines are available for beech wood, e. g. [5], [6]. However, no standards for the industrial process are available and the industry still tends to avoid the use of beech wood as a material resource. The existing technologies for the production of structural timber should be optimized for an economic fabrication of beech wood and standards have to be created or adopted to eliminate the constraints of the application of hardwoods in general. 2 MATERIALS AND METHODS 2.1 GENERAL The paper presents results that were obtained from investigations concerning the production of beech Glulam. The results were partly generated within a publicly funded research project at the Bern University of Applied Sciences in the year 2013, investigating the face gluing performance of several adhesives on beech wood. One of the project partners was the Henkel & Cie. AG/Purbond, who contributed experience and know-how for the application of 1-Component Polyurethane adhesives (PUR). And who independently investigated and developed an industrially applicable and certified solution for the face gluing of beech wood with PUR. The face gluing of the lamellas is a key point for the production of beech Glulam. The durability and strength of the glue line has to be ensured over the life span of the product. Therefore, the anatomic differences of hardwood and softwood have to be considered. The penetration behaviour is different, [7] and the possible red heartwood of beech wood influences the penetration as well, [8]. There is also an influence of the lamella thickness on the quality of the beech Glulam, Frühwald (2010). Thicknesses of 18 mm, 24 mm and 30 mm beech lamellas were tested using the adhesives Melamine Urea Formaldehyde (MUF), Polyurethane (PUR), and Phenol Resorcinol Formaldehyde (PRF). The quality of the glue line increases with thinner lamellas. The final Glulam member of beech wood can reach strength classes of GL 48 and higher, [5], [9], [10]. However, the high natural potential of beech wood which can lead to strength classes up to GL 60 or more is still not been used. Therefore, the research work is still ongoing. 2.2 QUALITY ASSESSMENT 2.2.1 General The verification of the production of Glulam is according to EN 14080:2013. The bondline integrity has to be proved by delamination or shear tests. The requirements in the standards for both test setups are established for softwood and not for hardwood. However, these regulations were applied for the assessment of beech Glulam, [9], [12], [13]. A harmonized European standard for Glulam made of selected hardwoods is in preparation, which however will not be available on the short run. At the moment, Glulam can be produced according to EN 14080:2013 with some hardwood species, however out of the range for a CE-conformity. Hence, the only way for a CEconform production of engineered wood products is via a national approval or a European Technical Assessment (ETA). 2.2.2 Delamination test The delamination test according to EN 14080:2013, Annex C is used to evaluate the bondline integrity against the climate exposure during the life time during factory production control. The test begins with a fully adsorption of water of the specimen by applying controlled vacuum and compression cycles while the specimen is under water. Finally the specimen is dried using an oven with air circulation. The wet and dry cycle results in tension stress transvers to the glue line in the specimen. For the assessment, the sum of the openings along the glue line at the end grain developed during the tests has to be taken in relation to the total glue line length at the end grain of the specimen. To respect different service classes according to EN 1995-1-1:2004, the delamination test standard differs between three methods for the wet/dry cycles. The specimen size used during the presented investigations was b/h/l = 100/160/75 mm, as shown in Figure 1. R L T h b Figure 1: Specimen for delamination test according to EN 14080:2013, Annex C l Lamella Glue line

b Lamella Glue line Loading plate l L T R Figure 2: Test setup for shear test according to EN 14080:2013, Annex D 2.2.3 Shear test The second required test for the bonding quality of Glulam during factory production control is a shear test of the glue line loaded in longitudinal direction according to EN 14080:2013, Annex D. The standard provides two specimen shapes, on the one hand the block or bar with a cross section of 50 mm by 50 mm including all glue lines of the member depth (as shown in Figure 2) and on the other hand a drill core including only one glue line. Parallel to the shear strength, the percentage of wood failure (PWF) at the failure plane after testing has to be visually examined. 2.3 PRELIMINARY TEST PROGRAM Beech lamellas with a thickness of 20 mm from the Jura region, Switzerland, were used. The lamellas were conditioned at 20 C temperature and 65% relative humidity until use. The moisture content of wood was in average 14.1 M%. The lamellas were free from defects with an average density of 702 kg/m 3. For the investigation of the gluing capability, different surface preparations were carried out in combination with different adhesives. The roughness of the surface was detected before the gluing process. The roughness of the surface as surface profile of the lamellas was determined using a touching measuring device. The adhesives used for the test series ranges from 1C- Polyurethane (PUR), Melamine Urea Formaldehyde (MUF) to Resorcinol Formaldehyde (RF). Table 1 gives an overview of the complete test program. The production comprises homogeneous beech Glulam of beech wood with a cross section of 100 mm by 160 mm, as shown in Figure 3. Table 1: Preliminary experimental test program Surface Adhesive Number of specimen Planed MUF, PUR, RF 10, 10, 10 Rough sanded P40 MUF, PUR, RF 10, 10, 10 Fine sanded P100/P150 MUF, PUR, RF 10, 10, 10 Rotoles/Face milling MUF, PUR, RF 10, 10, 10 Figure 3: Cutting plan for the delamination specimens (D) and shear specimens (S) The gluing process itself was done under production conditions in company (20 C and 50% relative humidity). The laminating pressure was 0.7-0.9 MPa and the pressing time 16 hours. 3 RESULTS AND DISCUSSION 3.1 PRELIMINARY TEST PROGRAM 3.1.1 Surface Roughness In average, the roughness of the surface of the lamellas can be divided into two levels. The planed and fine sanded surfaces are 32 μm respectively 45 μm rough in average. This is less compared to rough sanded and the face milled surface with 151 μm and 174 μm in average. 3.1.2 Shear strength The shear strength of all test series is in average 16.5 MPa and the minimum required percentage of wood failure is fulfilled, according to EN14080:2013. The values of the shear strength for the different adhesives and surfaces used are in a small range, see Figure 5. No clear advantages or disadvantages of different surface preparations or adhesives could be determined, except higher COV for planed surfaces. 3.1.3 Delamination The results of the delamination test according the B procedure of EN 14080:2013 differs clearly for the surfaces and adhesives used, see Figure 7 and Figure 8. Roughness [μm] 210 180 150 120 90 60 30 0 0 1 2 3 4 5 Figure 4: Roughness characteristic of surfaces for Beech Glulam

Shear strength [MPa] 24 22 20 18 16 14 12 10 8 6 4 2 0 MUF PUR RF 0 1 2 3 4 5 Delamination [%] 80% 60% 40% 20% 4 % 0% 68.3% 13.3% 15.0% 0.6% 0.9% 1.9% MUF PUR RF 37.9% 0.0% 0.1% 20.2% 3.3% 3.5% Figure 5: Shear strength of glue line for Beech Glulam Figure 8: Overview of the delamination test results for Beech Glulam (Minimum, Maximum and Average value of each configuration) The test series with RF fulfils the requirements for delamination of less than 4% according to EN 14080:2013 for all surfaces. The MUF adhesives can be used for sanded and face milled surfaces. The results for the PUR adhesive were significantly above the requirements of EN 14080:2013. This, however was due to the fact that an adhesive system developed and certified for bonding softwoods was used. 3.2 LATEST EXPERIMENTAL TEST SERIES AND DEVELOPMENTS Figure 6: Samples for determination of the percentage of wood failure, 100% PWF (top), 60 % PWF (bottom) applying the Wiesner Test, (purple color: wood surface, light color: adhesive surface) Henkel & Cie. AG recognized the need for the development of a solution that is able to bond beech wood (and other alternative wood species) under industrial conditions. Target was set to reduce the delamination results under the existing requirements of EN 14080:2013 and EN 15425, where possible. A feasible way was found to apply a primer pre-treatment to the wood surface and combine it with the established and already certified LOCTITE/PURBOND-adhesive systems. Laboratory test results and trials in factory production revealed the significant effect on the delamination behaviour of glued beech wood (Figures 9 and 10). Also with beech Laminated Veneer Lumber (LVL), a significant improvement of the bonding quality towards normative requirements could be achieved (Figure 11). Figure 7: Differences in the delamination tests for Beech Glulam, planed specimen with MUF (left), rough sanded with MUF (right) Figure 9: Beech Glulam, glued with LOCTITE HB S309 PURBOND and partly with Primer LOCTITE PR 3105 PURBOND, Delamination test acc. to EN 14080:2013, D

Since 2014, the adhesive system has been integrated into two general building approvals for Germany (Z-9.1-679, Z-9.1-765) for face gluing of beech wood, [16], [17]. Furthermore, the adhesive system was applied for the production of CLT which was used as construction material for the recently opened Empa NEST Vision- Wood [18]. Figure 10: Beech-spruce hybrid Glulam, glued with LOCTITE HB S309 PURBOND and partly with Primer LOCTITE PR 3105 PURBOND, Delamination test acc. to EN 14080:2013, D 4 CONCLUSIONS AND VIEW The research results presented show clearly the successful development of the production of glued laminated timber of beech wood. The preliminary test series show already a good material behaviour for the shear strength. Though, the delamination tests results, which represent the climate exposure during life cycle, were not satisfied for all configurations at this stage. The face gluing of beech wood with some PUR adhesives under industrial conditions is possible within existing standard requirements (for softwoods), when a treatment with a special primer is conducted prior to adhesive application. Figure 11: Beech Laminated Veneer Lumber, glued with LOCTITE HB S309 PURBOND and partly with Primer LOCTITE PR 3105 PURBOND, Splitting test acc. to EN 14374, B In a later work, Lüdtke et. al [15], investigated the effect of the this primer on beech, oak and ash and could confirm the significant improvement of delamination results for all those wood species (Figure 12). Since 2014, the adhesive system has been integrated into two general building approvals for Germany (Z-9.1-679, Z-9.1-765) for face gluing of beech wood, [16], [17]. Figure 12: Illustrative summary of results from Lüdtke et. al [15]. The graph displays the overall effect of the Primer pre-treatment on delamination behaviour for the wood species beech, oak and ash and the surface mashing methods periphal planing and face milling. Hence, an interesting alternative can be offered to the market, which in addition comprises the advantages of safe and clean application (1-Component), shortest curing times (down to 10 Minutes) and with no emissions of VOC or formaldehyde in the wood member. Latter aspect is a strong argument for the production of CLT with such PUR-adhesive systems. ACKNOWLEDGEMENT The research work was partly done within a Bachelor Thesis of Flavien Sauser and proudly supported by Stuber and Cie AG, Corbat A+C and Henkel & Cie. AG. REFERENCES [1] T. Näher, B. Neubauer-Letsch, Grundlagen zum Marktpotential für verklebte Laubholzprodukte mit Fokus Buchen-Brettschichtholz, Research Report, Bern University of Applied Sciences, Switzerland, 2015. [2] Franke S.: Hoch belastet optimieren mit Laubholz. In: Holzbautag Biel. 2013 [3] Eid. Forschungsanstalt für Wald Schnee u. Landschaft, Schweizerisches Landesforstinventar. Ergebnisse der dritten Erhebung 2004 2006. Birmensdorf, 2010. [4] T. Lüthi, Analyse der Schweizer Laubholz- Sägewerke. Teilbericht 2 zur Studie Entscheidungsgrundlagen zur Förderung von Laubholzverarbeitung und absatz, 2009. [5] H.J. Blass, J. Denzler, M. Frese, P. Glos, P. Linsemann, Biegefestigkeit von Brettschichtholz in Buche, Publisher University of Karlsruhe, Germany, 2005.

[6] Information on http://www.grupo-gamiz.com/en/ 0202.html, 13/02/2013. [7] F.K. Wittel, P. Niemz, H.J. Hermann, P. Hass, M. Mendoza Jiminez, Verklebung von Laubhölzern, Holzforschung Schweiz, pp. 1-4, 2010. [8] D. Ohnesorge, Verklebungseigenschaften von Brettschichtholz aus Buche, Doctoral Thesis, University Freiburg, Germany, 2009. [9] A. Frühwald, J. Ressel, A. Bernasconi, Hochwertiges Brettschichtholz aus Buchenholz, Institut für Holzphysik und mech.technologie, Hamburg, 2003. [10] T. Strahm, Laubholzkonstruktionen Realisierte Bauwerke in der Schweiz, 20. Internationales Holzbau-Forum IHF, Germany, 2014. [11] EN14080:2013Timber structures - Glued laminated timber and glued solid timber - Requirement, European Committee for Standardization CEN, Bruxelles, Belgium. [12] S. Aicher, H.W. Reinhardt, Delaminierungseigenschaften und Scherfestigkeiten von verklebten rotkernigen Buchenholzlamellen, Springer-Verlag, Stuttgart, 2006. [13] M. Schmidt, P. Glos, G. Wegener, Verklebung von Buchenholz für tragende Holzbauteile, Springer- Verlag, München, 2009. [14] EN1995-1-1:2004 Eurocode 5: Design of timber structures - Part 1-1: General - Common rules and rules for buildings, European Committee for Standardization CEN, Bruxelles, Belgium. [15] J. Luedtke, C. Amen, A. van Ofen, C. Lehringer, 1C-PUR-bonded hardwoods for engineered wood products: influence of selected processing parameters, European Journal of Wood and Wood Products, 73 (2), pp. 167-178, 2015. [16] DIBt, General technical approval Z-9.1-679. Glulam from beech and glulam from beech-hybrid. [BS- Holz aus Buche und BS-Holz Buche-Hybridträger]. Studiengemeinschaft Holzleimbau e.v, Germany, 2009. [17] DIBt, General technical approval Z-9.1-765. 1C- PUR-adhesives LOCTITE HB S049 to HB S709 PURBOND for bonding of load bearing wood members. [1K-PUR-Klebstoffe LOCTITE HB S049 bis HB S709 PURBOND für die Verklebung tragender Holzbauteile]. Henkel & Cie. AG, Switzerland, 2016. [18] Empa, NEST Vision-Wood, www.empa.ch/web/nest, link 12.05.2015.