MECHANICAL PROPERTIES OF BEECH CLT Steffen Franke1 ABSTRACT: The use of relatively new constructions products like Cross laminated timber (CLT) is increasing significantly. It is planned to extend the production of CLT by producing them out of beech or of beech and spruce in combination as hybrid product. The objective is to provide high performing materials which compensate weak points in soft wood products. In order to use and implement the product, the mechanical behaviour of a CLT plate of beech were investigated. The potential of beech is shown in terms of known strength values. Experimental tests for the evaluation of the strength and stiffness values for beech CLT for different situations as well as delamination tests were performed. Failure cases of the mechanical tests are presented and discussed where the rolling shear failure was in major focus for the discussion. KEYWORDS: Hardwood, CLT, mechanical properties
1 INTRODUCTION1 Hardwood shows a higher natural strength potential than soft wood. Due to their high adaptability, hardwood species increase their population. In the long term, the hardwood population is growing faster than they are needed and currently used in the construction sector. Furthermore, hardwood, with its good mechanical properties, can be the key in realizing long spanned and high stressed timber constructions. On the other hand, the available cross sections and lengths are less in dimensions. To compensate this, wood products such as glulam or Cross laminated timber (CLT) are one main opportunity. CLT is getting popular for softwoods, but only less research was done using hardwood, . The paper presents first results on the investigation of the mechanical behaviour with evaluation of several strength and stiffness values of CLT made of beech wood.
softwood products due to higher handling costs but mainly due to non-standardization, non-effectiveness and individual production. Compared to softwood, the mechanical properties of hardwood have higher values. A comparison of mechanical properties of spruce and different hardwood species is shown Figure 2. Especially the tension strength perpendicular to the grain for hardwood can reach up to
2 POTENTIAL OF HARDWOOD 2.1 General Figure 1: Distribution of wood species in the Swiss forest, 
Beech (690 kg/m³)
52 50 60
Ash (690 kg/m³)
Oak (670 kg/m³)
2.7 4.0 7.0 7.0
5.8 11.0 11.0 10.0
7.5 11.5 13.0 10.0
Spruce (470 kg/m³)
95 105 120
Failure strenght [N/mm2]
The high potential of hardwood is caused by various aspects like their high adaptability to climate changes. Compared to conifers, deciduous trees species are comparably young and have a more developed and complex cell structures that enable rapid adaptation to changing living conditions. Especially maple and beech population show a steady increase in Swiss forests , Figure 1. This potential is rarely used by the building industry even so hardwood logs are up to 30% cheaper than softwood logs [Gross 2010]. However, so far, the prices for the products made of hardwood are higher compared to
Steffen Franke, Bern University of Applied Sciences, Institute for Timber Construction, Structures and Architecture, Solothurnstrasse 102, 2504 Biel, Switzerland. Email: [email protected]
Tension Compression Compresion Shear ft,90
Figure 2: Comparison of mechanical properties of hard-
wood and softwood species
260% of the softwood strength values. In bending and compression parallel to the grain, the strength values are up to 175 % respectively 150 % higher. Therefore, the use of hardwood allows larger spans and smaller cross sections. The higher natural strengths are based on the relation of the density to the mechanical material behaviour. The densities of hardwood are obviously higher with e.g. around 610 kg/m3 for maple wood or around 690 kg/m3 for beech wood in average. Furthermore, due to its higher density, beech CLT performs better in terms of sound insulation and can reduce the use of concrete, . At the moment, the natural higher strength potential is mainly used for partial reinforcements in timber structures, e.g. for strengthening the lateral compression capacity at supports or loading plates or the tension capacity perpendicular to grain at notches and holes or in tapered and curved beams . 2.2 AVAILABLE PRODUCTS The mainly used products in the timber construction sector are based on softwood materials. Currently they are no standard products in hardwood. Thus individual solutions are needed, making hardwood constructions complicated, expensive and less attractive to use. In Switzerland, hardwood is mainly sawn to unedged boards 38 %, see Figure 3. Caused by the anatomy and the crooked trunks only 40-50% of the log can be used currently. For the production of engineered wood like glulam or CLT it needs to be considered that hardwood differs from softwood in terms of growth habit, growth cycle as well microscopic structure. Therefore technologies and production processes of softwood cannot be used one-to-one. A practical application of hardwood in timber structures is for example the roofing of a car park garage in Arosa (Switzerland), where the roof is made of softwood glulam with local reinforcement in ash wood. The span reaches up to 19.7 m, . Further the agriculturally structure in Lauenen is an example for a hall construction. It is built in beech glulam wood of GL48, . Hardwood is also used in several structural components in the House of Natural Resources at ETH Zurich. Here e.g. an innovative post-tensioned timber structures is designed. The joints are strengthened with ash hardwood, so that no further steel elements are required in
the moment-resisting timber joint; only a single straight tendon is placed in the middle of the beam, . In addition timber-concrete composite slabs using beech wood plates are installed. 2.3 POTENTIAL OF BEECH CLT Market researches from the Bern University of Applied Sciences show a high potential of beech CLT, . Numerous applications of beech CLT are conceivable as the use of beech CLT for highly stressed load bearing that opens the possibility to the use this material in multi storey timber structures. Besides this, a promising application is the use in floor constructions, especially for biaxial spanned floor slaps compared to spruce CLT. Using Beech CLT for load carrying floor constructions, the modulus of elasticity is of high importance. As 95% of the stiffness properties of CLT are defined by the outer lamellas, hybrid cross sections are possible, with high strength and stiffness demands on the outer lamellas. To reduce the price and weight, the inner lamellas can have lower strength and stress values, . The use of beech CLT can open up new fields for timber construction. It can play an important role building staircases and can replace concrete structures.
3 LOAD BEARING BEHAVIOUR OF CLT CLT can be loaded in two principals; vertical to the plane (out of plane) and in plane. This causes panel stresses and shell stresses. Figure 4 shows typical loading situation of CLT panels. The load carrying behavior of CLT is affected by the orthogonal order of the single layers and their anisotropic material characteristic. The cross wise orientation of the layers influence the deformation and stress distribution and has to be considered. 3.1 LOAD OUT OF PLANE Due to the low shear stiffness of the cross layers, out of plate bending load leads to a different behaviour for the stress distribution and deformation than known for solid wood, as shown in Figure 5 and Figure 6. The ductile composite of each layer has to be considered like a semirigid connection. The load carrying capacity is governed
Loading out of plane - parallel (∥) to top layer
Loading in plane - parallel (∥) to top layer Figure 3: Sawing of hardwood, 
Loading out of plane - vertical ( ) to top layer
Loading in plane - vertical ( ) to top layer
Figure 4: Possible loading types for CLT-panels
by the rolling shear strength of the cross layers. The tension and compression stress in the center of each layer decreases while the bending stress and, therefore, the maximum bending stress increases. Tension stresses in the cross layers develop only for edge bonded cross layers. 3.2 LOAD IN PLANE CLT elements can be used as wall systems to carry vertical and horizontal loads. A description of the load carrying effect can be found in . An optimal vertical load transfer can be achieved arranging the outer layer vertically. According to , the semi rigid connection of the vertical layer, caused by the low shear stiffness of the horizontal layer, need to be taken into account for buckling and deformation proof. For complex and highly stressed structures, CLT elements need to be calculated according to shear analogy. For residential, school and industrial buildings, it is sufficient to calculate with an effective stiffness. This method is based on the shear analogy but the bending and shear stiffness are summarized to an effective overall stiffness .
4 PRODUCT DEVELOPMENT 4.1 MATERIAL AND EXPERIMENTAL TEST SERIES Because of their good mechanical properties and increasing population, CLT board out of beech wood, was built together with industry partners. It consists out of layers of different thickness crosswise stacked to a total thickness of 120 mm. The cross section is displayed in Figure 7. Due to the cross layers, the direction and position of Bending stress own I part
normal stress Steiner part
the lamellas and the loading directions need special consideration (Figure 4). This is considered in the experimental test setups and evaluation of the results. For the investigation of the mechanical behaviour, experimental test series were carried out. Besides the bending tension and compression strength, the shear, rolling shear and the delamination of the CLT plate was determined according the SN EN 408:2012, SN EN 14080:2013, FprEN 16351:2013 and SN 302-2:2013. The test set ups are displayed in Figure 8 to Figure 10. The four-point-bending test, Figure 8, was performed using specimens with horizontal and vertical lamellas to determine the bending stiffness and bending strength for both cases. The edge bending stresses (horizontal lamellas) is determined by formula (1). For the bottom layer formula (2) was used. The bending stress (vertical lamellas) was determined by equation (4). The test samples have not been preconditioned, because of the sizes. They were stored at room temperature after delivery. The rolling shear was tested according to Figure 10. The samples were not preconditioned and loaded to its maximum load. The rolling strength was determined on the basis of the maximum load using equation (5). The compression strength was determined in three different grain orientations (see Figure 9). Parallel to the grain, the specimens were tested up to ultimate load, while perpendicular to the grain, the maximum load was determined at 7 mm deformation. The compression strength was determined du to equation (6) and (7). The specimens were not preconditioned. The shear test of the glue line was determined according to SN EN 14080:2013. To avoid failure due to rolling shear, the specimens were loaded in a 45° angle (see Figure 11). The shear strength was determined using equation (8). The specimens were stored under standard climate for 5 days (20 °C/65 % relative humidity).
∗ own I part
∗ Figure 5: Bending stress and shear stress of CLT elements, 
Figure 6: Shear deformation of glued CLT element, 
∗ 1.5 ∗
2∗ ∗ (6)
4.2 Additional Rolling shear tests As the rolling shear plays an important role for the load carrying capacity of CLT boards, additional tests with changing layer thicknesses were performed (see Figure 12). Based on these test results, the influence of the layer thickness on the rolling shear was determined. Figure 7: Cross section of a CLT out of beech
The top and bottom layer of asymmetrical compositions of CLT elements are orientated in the opposite directions. Therefore, four specimens with a cross orientated layer at the top and four specimens with a cross orientated layer at the bottom were tested (Figure 13). The load was applied on the top surface,  The asymmetrical composition of the elements needs to be considered for the evaluation of the results. Due to the different layer thicknesses, the effective depth changes which influences the stress distribution and maximum stress. Figure 14 shows the shear distribution for two cross sections of 33 and 10 mm layers respectively. The shear stress belongs to 100 kN loading and is calculated be the elastic composite theory.
Figure 8: Test setup for bending
It is obvious that the shear stresses are reduced with increasing effective depth and decreasing layer thickness. The shear stresses can distribute more evenly and lead to a smaller maximum. Only specimen with rolling shear failure were used for the evaluation of the results. Some specimen showed failure in bending of the bottom layer due to gluing problems, as shown in Figure 15 and Figure 16.
Figure 9: Test setup for compression
5 RESULTS AND DISCUSSION 5.1 CLT PLATE The CLT plate was tested due to its stiffness as well as bending, shear, compression and tension strength. Figure 17 and 18 show typical bending failure and Figure 19 a rolling shear failure. The characteristic values determined are displayed in Table 1.
Layer thickness 2.4 mm
Layer thickness 10 mm
Layer thickness 20 mm
Layer thickness 33 mm
Figure 12: Variation of the layer thickness to determine the rolling shear of CLT elements Figure 10: Test setup for rolling shear
Cross layer at top
Cross layer at bottom Figure 11: Shear test specimen
Figure 13: Cross layer at the bottom and top, 
Table 1: Mechanical properties of Beech-CLT
Description Bending strength Rolling shear In plane compression Compression perp. to the plane Shear strength Modulus of Elasticity (global)
160 140 120
60 40 20 0 0.0
Shear stress x [MPa] 200
fm,05 fR,v,k,05 f c,0,05 f c,90,05 f v,05 Emean
In general, the achieved values are relatively high compared to CLT out of spruce. Especially the rolling shear, bending and compression strength perpendicular to the grain are significantly higher than comparable CLT plates out of spruce, compare Figure 20. The presented test results were determined using test samples which were cut from a single plate. The test results achieved has to be confirmed in further test series to get representative material parameters. But nevertheless, it shows the great potential of the developed product and the use of hardwood in the construction sector.
Shear stress x [MPa]
Figure 14: Shear stress distributions for different layer thicknesses,  Figure 17: Bending failure (plate test)
Figure 15: Failure of finger joints within the bottom layer, Figure 18: Bending failure (in plane test)
Figure 16: Bending failure of bootom layer due to less glue for the surface and fingerjoints gluing,
Value 43.8 MPa 3.8 MPa 61.5 MPa 9.3 MPa 5.9 MPa 12306 MPa
Figure 19: Rolling shear failure
Spruce CLT 43.8
0 Bending Rolling shear Compression Compression Shear fm,05 fR,v,k,05 f c,0,05 f c,90,05 f v,05 fm,0.5 fR,v,k,0.5 fc,0,05 fc,90,05 fv,05
Figure 20: Comparison of mechanical properties of hardwood and softwood CLT
Rolling shear strength [MPa]
10 9 8 7 6 5 4 3 2 1 0
10 20 Thickness of lamella [mm]
Figure 21: Rolling shear strength depending on the crosssectional structure
5.2 ENHANCEMENT OF ROLLING SHEAR In the design of solid wooden ceiling, the rolling shear strength limits the load bearing capacity. Compared to a CLT out of spruce, the rolling shear strength of a CLT out of beech is more than two times higher. Figure 21 shows the comparison of the rolling shear strength of a CLT board out of spruce and beech. In addition to the material, also the thickness of the lamellas influences the rolling shear strength. The values increase with decreasing lamella thickness up to a thickness of 10 mm. But then shows less strength for thicknesses of 2,4 mm used by veneers.
Figure 22: Rolling shear failure, beech, t ≤ 10 mm ,
Figure 23: Rolling shear failure, beech, t ≥ 20 mm, 
The rolling shear failure also depends on the layer thickness. Whereas thin layers show rolling shear only, Figure 22 and 24, specimen with thicker layers also show tension failure perpendicular to the grain within the layers, Figure 23 and 25. Differences of the behavior and strength values regarding the position of the outer cross layer, top or bottom, could not be seen.
6 CONCLUSIONS CLT elements out of beech show a great potential. Especially the determined mechanical values like the rolling shear and the compression strength perpendicular to the grain open new applications of CLT in timber structures. CLT elements can be used as plate and shell elements loaded out of plane and in plane direction. The rolling shear strength and failure behaviour depends on the layer thicknesses and composition of the elements. However the experimental test series has to be extended to different production lines in order to prove the mechanical properties.
Figure 24: Rolling shear failure, laminated veneer lumber
ACKNOWLEDGEMENT The research work was proudly supported by Swiss Academy of Engineering Sciences, Raurica Wald Holz AG, and by Federal Office for The Environment (FOEN). Figure 25: Rolling shear failure, CLT with 10 mm layers
REFERENCES  Ringhofer, A., Schickhofer, G., Timber in Town – current examples for residual buildings in CLT and task for the future, In: European Conference on Cross laminated timber (CLT), Graz, Austria, 2013.  Brändli, U.-B., Schweizerisches Landesforstinventar. Ergebnisse der dritten Erhebung 2004 – 2006. Bundesamt für Umwelt, BAFU. Bern, 2010.  Franke S., Hoch belastet – optimieren mit Laubholz. In: Holzbautag Biel. 2013  T. Lüthi: Analyse der Schweizer LaubholzSägewerke. Teilbericht 2 zur Studie – Entscheidungsgrundlagen zur Förderung von Laubholzverarbeitung und –absatz.2009.  T. Strahm, Hardwood structures in Switzerland, Proceedings of 20. IHF, Garmisch Partenkirchen, Germany, 2014.  ETH Zurich, House of Natural Resources, http://www.honr.ethz.ch/en/, 17.05.2016.  T. Näher, B. Neubauer-Letsch, Grundlagen zum Marktpotential für verklebte Laubholzprodukte mit Fokus Buchen-Brettschichtholz, Research Report, Bern University of Applied Sciences, Switzerland, 2015.  S. Winter, H. Kreuzinger, P. Mestek: Teilprojekt 15 Flächen aus Brettstapel, Brettsperrholz und Verbundkonstruktionen. TU München. 2009.  R. Blaser: Untersuchung zur Schubbeanspruchung von Furnier- und Brettsperrholz. Bachelorthesis. Bern University of Applied Sciences. Switzerland, 2012.  S. Franke, M. Vogel, M. Lehmann, F. Sauser, Mechanische Eigenschaften von Buchenbrettschichtholz, Research Report, Bern University of Applied Sciences, Switzerland, 2014.