ANALYSIS OF THE PENETRATION OF ADHESIVES AT FINGER-JOINTS IN BEECH WOOD

Transcription

1 ANALYSIS OF THE PENETRATION OF ADHESIVES AT FINGER-JOINTS IN BEECH WOOD Thomas Volkmer 1, Bettina Franke 2, Anna Schusser 3 ABSTRACT: The penetration of different adhesives on finger joints in hard wood for the production of glued laminated timber as a structural material has been investigated. Finger joints are the key element for longitudinal assembling of the lamellas as known for softwood glulam. For reliable strengths of finger joints, the penetration and the distribution of the adhesive is an important parameter, because the anatomical structure of hard wood is different to the ones of soft wood. First results show that the production parameters, like the application method of the adhesives, the open and closed time, the pressure as well as pressing time influence the penetration of the adhesives and the achievable strength of the finger joints. KEYWORDS: Hard wood, Adhesives, Penetration, Finger joints 1 INTRODUCTION 1 Current research projects in Europe concentrate on the use of hard wood as a construction material for timber structures. Ongoing research projects deal with the production of glued laminated timber of ash, beech or oak, [1], [2], [3]. Parallel to the necessary surface gluing, finger joints are an important element to produce the lamellas known from soft wood products. It enables the improvement of wood and the continuous assembly of lamellas. The finger joints have also an influence on the final strength of the structural timber elements, because the mechanical properties of hard wood lead to higher strength classes than solid wood and the finger joints are more stressed. Consequently finger joints are a key point in the production of structural timber and the quality must be ensured. 1 Thomas Volkmer, Bern University of Applied Sciences, 2500 Biel, Switzerland. thomas.volkmer@bfh.ch 2 Bettina Franke, Bern University of Applied Sciences, 2500 Biel, Switzerland. bettina.franke@bfh.ch 3 Anna Schusser, Bern University of Applied Sciences, 2500 Biel, Switzerland. In Switzerland the hardwood stock has increased since Hardwood covers about 31% of the entire wood stock, in which the biggest part with 18 % is beech wood [5]. In addition, the first realized timber constructions from hardwood show the potential of hardwood compared to conventional construction material, [4]. Therefore the following research results concentrate on finger joints from beech. Different adhesives and there penetration according to the various production characteristics were investigated. Hard wood has got another anatomical structure than soft wood. Due to differences in the cell type and cell dimensions, the penetration of the adhesive will be different and leads to a different composite of the adhesives with the wood, which again results in different achievable strength classes. 2 MATERIAL AND METHODS 2.1 MATERIAL BEECH For the experimental test series, beech wood (Fagus sylvatica) from Switzerland, the Jura area, was used. The average moisture content after conditioning was about 10%. The mean density was about 700 kg/m³. The test program comprises finger joint lengths of 15 mm and 20 mm and the samples were glued together with the following adhesives: Polyurethane (PUR), Melamine- Urea-Formaldehyde (MUF) and Emulsion Polymer Isocyanate (EPI). The production was done under both laboratory and manufactory conditions according to the

2 regulations of the standard SN EN 385:2001, [6], [7], [8], [9]. The main characteristics of the samples are summarized in Table 1. Table 1: Characteristic of the test programme of finger joints in beech Profile Production Adhesive Finger length Angle I-15-Profile Industry PUR+H 2 O 15 mm 5.6º I-15-Profile Laboratory MUF 15 mm 5.6º I-15-Profile Laboratory EPI 15 mm 5.6º I-20-Profile Laboratory MUF 20 mm 6.0º I-20-Profile Laboratory EPI 20 mm 6.0º I-20-Profile Laboratory PUR 20 mm 6.0º I-20-Profile Laboratory PUR+H 2 O 20 mm 6.0º I-20-Profile Laboratory PUR+Primer 20 mm 6.0º 2.2 METHODE The analysis of the penetration behaviour was done with the lab microscope Leica DMLM. A mercury lamp EL 6000 with emission maximas between 350 and 600 nm was used for illumination. Prior to the analysis, the samples were sanded in different steps in order to get an appropriate glossy surface. The images were acquired with a Zeiss camera and saved with the software ImageAccess. Figure 1 shows a sketch of the samples of the finger joints investigated. For each sample the microscopic analysis was done in longitutinal(figure 1, left) and transvers direction (Figure 1, right). cell wall, fibres or vessels. Also the glue line or the direction of the penetration of the glue is visible as shown in Figure 2. Further the wood structure and possible damages or deformations due to the applied pressure during the gluing can be analyzed. Figure 2: Characteristic glue line formation (PUR, wood pre-treated with water) between two beech samples perpendicular to the grain, the following parameters are visible: glue line width, different penetration depth regarding early and late wood, complete or partly filling of the vessels, deformation of the wood rays Because of the heterogeneous structure of wood on the microscopic level, the microscopic approach to analyse the glue line between two wooden parts provides only information about the specific place, which was observed. The summary and conclusions from these investigations can give a tendency and estimation about the properties in the material. Figure 1: Sketch of finger joint with microscopic investigation directions marked, left longitudinal and right transvers direction 3 RESULTS 3.1 GENERAL The microscopic investigation allows a differentiation of the penetration depth and the kind of penetration. It is possible to observe the area of glue accumulations like the 3.2 GLUE PARAMETERS AND PRE- TREATMENT OF THE WOOD SURFACE In the most samples, the glue line was hardly visible and relative thin. Furthermore, the penetration into the two joining wood parts is different. On the one side, the glue penetrated more than 500 µm whereas on the other site almost no penetration could be observed. An accumulation of glue can be seen at the end of the fingers and from there, the penetration takes place into the longitudinal direction of the fibres, as shown in Figure 3. Due to the vessel distribution and the density differences the penetration is deeper in early wood than in late wood, as shown in Figure 4. Further there is no glue distribution across the year ring boundary because the vessel network of beech is only connected within the year ring and no movement from year ring to year ring is possible.

3 The degree of condensation influences the viscosity and that s why the penetration as well, Systems which are more pre-condensated to bigger molecules have a higher viscosity and penetrate fewer into the substrate [8]. Figure 3: Gap with bubbles between the wood pieces at the end of the finger joints, the PUR adhesive is clearly penetrated in the longitudinal direction Figure 5: Penetration of EPI, I-15 profile, laboratory production, the glue line is clearly visible, the penetration depth and distribution is homogenous in both wood parts Figure 4: Penetration of PUR, I-20 profile, laboratory production, the penetration characteristic in early and late wood of beech depends on the interconnecting vessel system and is very different: few penetration in the late wood, deep penetration in the early wood, no resin flow via the year ring border The distribution of the glue in the microstructure shows some specific patterns. Mainly the vessels are filed with the glue. Fibres containing glue can be seen only very close to the glue line. Very often the glue just covers the inner surface of the cell wall, but sometimes the whole lumen is filled. The rays do not support the transport of glue in radial direction. EPI and MUF adhesives are not combined with fluorescent dye, that s why the wood surface needs to be coloured in order to generate a contrast between wood and glue. Both systems (Figures 5 and 6) penetrate less deep into the wood compared with PUR, but the glue distribution is more homogenous. Further the glue line itself is thicker and can cleary differentiated from the wood. Figure 6: Penetration of MUF, I-15 profile, laboratory production, the glue line (also due to the wood structure) is clearly visible, the penetration depth is different for the two joining parts depending on the application process Beside the anatomical parameters, the glue type and the glueing process a pre-treatment of the wood surface can influence significantly the performance of the bonding. In our study the surface was either wetted with pure water (Figure 8) or treated with a primer (Figure 7). The PUR adhesive penetrate comparable less into the wood if the surface is pre-treated with a primer and the lumen of the vessels is completely filled with resin. In samples without any pre-treatment (Figure 10) the glue covers mainly the cell wall surface and the vessels are usually not completely filled, but the penetreation depth is much higher. These findings correspond to the results of a parallel study, where

4 the samples pre-treated with a primer show the lowest bending and tension strength [9]. can be damaged or destroyed (Figure 9). The anatomical composition of hard woods is more complex compared with soft woods. For example in beech wood the vessels are diffuse distributed and within the year ring connected via pits [9]. Because of this vessel with a certain distant from the glue line can be filled with resin as well (Figure 10). But in such cases the maximum penetration depth is not an adequate parameter anymore to describe the penetration behaviour in this substrate, because the contribution of such filled and somehow isolated vessels to the strength of the bond line is hard to quantify [10], [11]. Figure 7: Penetration of PUR in beech wood after pretreatment with a primer, the glue line is not visible, all vessels close to the glue line are completely filled with resin and the penetration depth is very small Figure 9: Penetration of PUR in spruce, the penetration is relative small and the glue line is clearly visible, the tracheid s beside the glue line are partly destroyed or deformed Figure 8: Penetration of PUR in beech wood after pretreatment with pure water, the glue line is faintly visible, most of the vessel close to the glue line are completely filled with resin, the penetration depth is small 3.3 HARD AND SOFT WOOD DIFFERENCES Due to the anatomical specifications the penetration behavior of liquids in soft woods is different to them in hardwoods. The early wood tracheid s in soft woods are connected via numerous bordered pits, that s why the flow of the resin and the distribution is more homogenous. Generally the maximum depth of penetration is smaller than for hardwoods but due to the homogeneity a valuable parameter to describe the penetration behaviour. In the most cases the glue line is clearly visible and depending on the process parameters the tracheid s close to the glue line Figure 10: Penetration of PUR in beech wood (without pre-treatment) the penetration zone is inhomogeneous, the glue line is hard to identify and some isolated cells are still filled with resin 4 CONCLUSIONS The penetration of glue in hardwoods like beech is defined by the material properties, glue type and the gluing

5 process. From the material site, the density, the surface tension and the microstructure dominate the wetting and penetration. The composition and the viscosity of the binder systems define the depth of penetration. The geometry and the precision of the finger joints in combination with the process parameter influence possible accumulation at the end of the finger joints and the macro distribution of the glue in the glue line. There are still unsolved open questions concerning the interaction between glue and wood. Especially, which parameter is the most important one influencing the macroscopic strength and durability of the bonding. So far there is no clear explanation available, how and to which extend the penetration depths influence the macroscopic properties. Reliable finger joints in hardwood lead to high performing wood products. But also new solutions for the quality control are needed in order to ensure the production. The microscopic investigations provide the possibility to describe and evaluate the penetration behaviour and the distribution of the glue. This information are related to the performance of the bond line and will help to ensure the quality of the products. ACKNOWLEDGEMENT The research project is proudly supported by the Federal office for the environment of Switzerland and the industry partners neue Holzbau AG, Purbond, Corbat Holding SA, DeMeth GmbH, and LEUCO Production SAS. Many thanks also to Grecon Dimter Holzoptimierung Nord GmbH & Co. KG (Hannover, Germany) for the use of their production laboratory. [7] A. Schusser, Untersuchung der Keilzinkenverbindung für Brettlamellen in Laubholz, Masterthesis, Berner Fachhochschule Biel, Switzerland, [8] B. Franke, A. Schusser, A. Müller, Laubholz in Tragwerken Der Keilzinkenstoss als Schlüsselelement, Research Report, Bern University of Applied Sciences, Switzerland, [9] B. Franke, A. Schusser, A. Müller, Analysis of finger joints from beech wood. World conference on timber engineering, Quebec, [10] M. P. G. Laborie: Investigation of the wood/phenolformaldehyde adheisive interphase morphology, PhD Dissertation, Virgina Polytechnic Institute and State University, Blacksburg, [11] H. H.Bosshard, L. Kucera: The three-dimensional structure analysis of wood part 1: The network of vessel system in Fagus sylvatica, Holz als Roh- und Werkstoff, 31, , [12] P. Hass, F. K.Wittel, S. A.McDonald, F. Marone, M. Stampanoni, H. J. Herrmann, P. Niemz: Pore space analysis of beech wood: The vessel network, Holzforschung, 64, , [13] P. Hass, F. K.Wittel, M. Mendoza, H. J. Herrmann, P. Niemz: Adhesive penetration in beech wood: experiments, Wood and Fiber Science, 46, REFERENCES [1] H.J. Blass, J. Denzler, M. Frese, P. Glos, P. Linsemann: Biegefestigkeit von Brettschichtholz in Buche, University of Karlsruhe, Germany, [2] A. Fruehwald, J. Ressel, Bernasconi, A.: Hochwertiges Brettschichtholz aus Buchenholz, Research report, Institute für Holzphysik und mechanische Technologie, Hamburg, Germany, [3] Information on 13/02/2013. [4] W. Bogusch, Konstrutkitve Ansätze: Bauen mit Laubholz, CH-Holzbau 3/4, 2013, [5] Eidgenössische Forschungsanstalt für Wald Schnee und Landschaft. Schweizerisches Landesforstinventar. Ergebnisse der dritten Erhebung Birmensdorf, Germany, [6] SN EN 385:2001: Keilzinkenverbindung im Bauholz Leistungsanforderungen und Mindestanforderungen an die Herstellung. Schweizerischer Ingenieur- und Architektenverein, Zürich, 2001.