Shear Capacity of Dowelled Mortise and Tenon in Tropical Timber

Size: px

Start display at page:

Download "Shear Capacity of Dowelled Mortise and Tenon in Tropical Timber"

Alannah Hopkins
5 years ago
Views:

IOP Conference Series: Materials Science and Engineering Shear Capacity of Dowelled Mortise and Tenon in Tropical Timber To cite this article: Rohana Hassan et al 2011 IOP Conf. Ser.: Mater. Sci. Eng. 17 012012 View the article online for updates and enhancements.

1 IOP Conference Series: Materials Science and Engineering Shear Capacity of Dowelled Mortise and Tenon in Tropical Timber To cite this article: Rohana Hassan et al 2011 IOP Conf. Ser.: Mater. Sci. Eng View the article online for updates and enhancements. Related content - Externally Prestressed Monolithic and Segmental Concrete Beams under Torsion: a Comparative Finite Element Study M A Al-Gorafi, A A A Ali, I Othman et al. - Moisture Absorption Characteristics of Epoxy Based Adhesive Reinforced with CTBN and Ceramic Particles for Bondedin Timber Connection: Fickian or Non- Fickian Behaviour Z Ahmad, M P Ansell and D Smedley - Test Simulation using Finite Element Method M B Ali, S Abdullah, M Z Nuawi et al. This content was downloaded from IP address on 10/11/2018 at 16:23

2 Shear Capacity of Dowelled Mortise and Tenon in Tropical Timber Rohana Hassan*, Azmi Ibrahim, Zakiah Ahmad Faculty of Civil Engineering, UiTM, Shah Alam, Selangor Darul Ehsan. * address: Abstract. The European Yield Models (EYM) concept has been applied to observe an experimental study of single-dowelled mortise and tenon. Specimens consisting of structural size mortise and tenon dowelled with 20.6 mm diameter were axially loaded in shear. Test variables are the material for the dowels namely steel; glass fibre reinforced plastic (GFRP) and wood. Connections were tested until failure to observe the ultimate capacities and possible modes of failure as the variables were changed. Results show that the shear strength capacity of the mortise and tenon connection single-dowelled with steel, GFRP and wood is not directly reflected the capacity of the dowel. The proportional limit, 5% diameter offset and maximum strength of the connection are in the order of steel, wood and GFRP but the dowel yield modes are in the order of steel, GFRP and wood. Failure modes of GFRP were found as stiff as steel (mode Im) compared to wood (mode IIIs). 1. Introduction The load carrying behaviour of timber joints made with Malaysian tropical timber with regard to rigidity, load capacity and ductility are not well documented. The design of timber joint using steel connectors is given in MS 544: Part5:2001, however the design code for traditional structural connection such as mortise and tenon is not available. On the other hand, the current European Yield Model (EYM) is also developed according to the double shear strength mechanical behaviour which is not the true mortise and tenon joint. Thus the true mechanical behaviour of mortise and tenon joints is largely yet to cover. For that reason, this study is made as an attempt to observe on the capacity and the failure behaviour of the steel dowelled mortise and tenon. Along with the connections were also dowelled with GFRP and wood dowel. However, the observations for GFRP and wood dowels were made on the same principle as those developed for the design of steel dowel joints as no code is currently applicable for the design using GFRP or wood as dowel material. The use of GFRP as a substitute s to dowel material is still new. Various studies concerning GFRP as a composite to civil engineering material has been done, however none has replaced it as the structural dowel in mortise and tenon connections. GFRP is well known with its benefits such as having a high local force transfer, very stiff connections, particularly when the rods are loaded in the axial direction. It is also a low cost material and production, improved aesthetically as the connections is completely hidden in the timber structure as well as light weight connectors. Uses of GFRP as a construction material is widening due to the increasing cost of steel. The new innovation approach of this Published under licence by Ltd 1

3 connection is also expected to promote the new engineering design concept for the recent industrial building system (IBS). Therefore, the specific objectives of this study are to experimentally obtain the load-carrying capacities of the connections when dowelled using steel, GFRP and wood by axially loaded in shear. These values are the nominal limit state values namely proportional limit, 5% diameter offset and ultimate load capacity. It is also to identify the mode of failure of the dowels and the behaviour of mortise and tenon failures. 2. Mortise and Tenon Mortise and tenon is the most common type of connection for jointing the beam and column member in timber frame structures. This connection is also one of the typical connections applied in the recent building construction system, such as IBS. The basic mortise and tenon comprises of a tongue (tenon) cut into one rail and a mortise cut all the way through the other rail. A neater version is the stub tenon, since the mortise does not go all the way through the timber, so the end of the tenon is concealed (blind mortise and tenon). The joint can be strengthened by securing with dowels or wedges. The load-carrying behavior of timber connection is primarily controlled by the dowel bearing strength of wood, slenderness, and the yield strength of the dowel and the arrangement of the dowels. Dowel to timber interaction has been investigated in depth in previous research to aid the design of nailed, screws, bolted and dowelled connection. However, current research into traditional mortise and tenon is very limited. Concerns of structural timber-frame studies started prior to the publication of a dissertation by Brungraber [1] where the mortise and tenon using wood dowel as an individual joint, together with full-scale frame testing, finite element of analysis of joint behaviour and a computer model that incorporated connection behaviour. Erikson [2] commented that in continuation to Brungraber s work, the researchers from Germany, Kessel and Augustin (1995) reported that the oak dowel has sufficient strength for use in modern wood construction. Investigation of structural performance of timber pegged mortise and tenon joints and timber frame structures to develop methods for analyzing their strength and stiffness were then explored by few researchers such as Bulliet et al.. [3], Sanberg et al. [4], Erikson [2], Miller [5], Shank and Walker [6] and Walker et al, [7]. There are also studies on the related parameters of mortise and tenon are such as the dowel bearing strength [8,9], joint forms and dimension [10], widthdrawal resistance of the steel dowel [11,12] and effects on the dowel grain direction [13]. 3. European Yield Model (EYM) EYM describes the mechanism by which the components of a timber connection are deformed beyond the elastic region. The equations based on this theory predict the load-carrying capacity of a single dowel, per shear plane, loaded perpendicular to the axis. It is depending on the material properties of the dowels and on the geometry of the connections. Figure 1 shows the failure assumed by EYM for double shear, dowel type, timber-to-timber connections as defined by AFPA [14]. 2

4 Fig. 1: EYM double shear dowel type, timber-to-timber failure modes Where; Yield Mode I Wood crushing in either the main member or side members. Dowel stiffness is greater than wood strength. Yield Mode II Localized wood crushing near the faces of wood members based on the pivoting of rigid dowel about the shear plane. Yield Mode III Dowel yield in bending at one plastic hinge point per shear plane and associated wood crushing. Yield Mode IV Dowel yield in bending at two plastic hinge points per shear plane and associated wood crushing. The 2001 NDS requires the calculation of lateral connection design values for all applicable yield modes, the smallest value being the design value corresponding to the predicted yielding mechanism. 4. Experimental Ten (10) series of tests composed from one of the commonly construction species, namely Kempas (koompassia malaccensis). The actual structural size of column and beam connected with mortise and tenon joint inserted with single dowel were tested and observed. Pre-drilling of the timber elements, in exact size as possible to the dowel diameter was adopted in all specimens. All dowel diameters were as closed as possible to the available on shelve 20.6 mm diameter of GFRP. A gap within 1 mm was allowed within the dowel and the pre-drill hole seeing that according to the common construction practice. Nominal moisture content of the specimens was within 8% at the time of test. Results were studied based on load to displacement relations loaded in shear. Failure modes of the dowels and the mortise and tenon members were observed according to the double shear standard stated in National Design Specification for Wood Construction (AFPA)

5 5. Specimens Geometry Selection Mortise and tenon were made of two pieces of timber dimensioned as 200 x 200 x 1200 mm and 100 x 150 x 1000 mm respectively. A square mortise hole dimensioned with 41 x 100 x 150 mm was made at each of the column whilst the tenon is at one end of each beam with 41 x 89 x 150 mm in dimension. Tenon thickness and distance from the face to the tenon was 41 mm for a 20.6 mm peg (twice the peg diameter). A gap, typically mm, was left between the end of the tenon and the mortise base to ensure that as the timber shrinks the tenon does not come to bear on the mortise. The plan view of mortise and tenon configuration is shown in figure 2. Tenon Mortise 89 mm 20.6 mm 100 mm Beam Column 200 mm 200 mm Fig. 2: Plan view of mortise and tenon configuration 6. Shear Test Set Up Four (4) numbers of Linear Voltage Displacement Transducers (LVDT) were mounted at the necessary locations of the members. Loading applied perpendicular to the grain of the beam member, in the axis of the beam, at 225 mm from the face of column member. The beam was also propped at 1000 mm from the centreline of the column member. The rate of loading applied was 5.0 mm/min for all tests. Load head deflection was measured with a digital dial gauge. Testing was continued until the maximum deflection of the machine; 250 mm. Configuration and experimental set up of shear test is as shown in figure 3 and 4 correspondingly. 4

Mortise (Column) Load Fixed LVDT 4 Tenon/Beam 1000 mm Tenon (Beam) LVDT 1,2,3 225 mm 1000 mm Fig. 3: Configuration of Shear Test Pinned STEEL DOWEL Fig. 4: Experimental arrangement of shear test 7.

6 Mortise (Column) Load Fixed LVDT 4 Tenon/Beam 1000 mm Tenon (Beam) LVDT 1,2,3 225 mm 1000 mm Fig. 3: Configuration of Shear Test Pinned STEEL DOWEL Fig. 4: Experimental arrangement of shear test 7. Result and Discussion The nominal limit state values that are the proportional limit, 5% diameter offset and ultimate load is summarized in Table 1. These values resulted from GFRP and wood dowel tests were compared to the value of steel dowel since the design standard using steel is the only available (Table 1). Results show that the average proportional limit of GFRP and wood is lower than the steel by 32.62% and 21.96% respectively. Difference in loading at proportional limit is higher compared to the maximum load capacity. The maximum loads for GFRP and wood is lower than steel at 29.01% and 16% correspondingly. This also means that mortise and tenon dowelled with wood were found 18.3% higher in strength capacity compared to dowelled with GFRP. 5

7 As the nominal load for design is normally based on the 5% diameter offset capacity, comparison was also made between maximum loads and 5% diameter offset. The difference in maximum load to 5% diameter offset for steel dowel is 6.6%, 0.9% for GFRP dowel and 4.6% for wood dowel. This is as predicted, since the dowel stiffness commonly decreases with the decrease strength of the dowel material. This result has also been compared to Shank [15]. It was found that the average maximum load when dowelled with wood is higher for Kempas compared to Green Oak species. The value reported by shank in a similar set up of experimental work using Green oak is within 35 to 71 kn, whereas for Kempas, the values obtained was within 96 to 117 kn. This result was reflected to the modulus of elasticity (MOE) of each species. The MOE for Kempas and Green Oak is in the order of kg/m3 [16] and 680kg/m3 [15] correspondingly. Shank also commented that the tenon member cheeks rolled from the tenon when the wood dowelled connection loaded between 10 to 20 kn. This study noticed the same movement when joint dowelled with GFRP and wood were loaded between 30 to 50 kn. However, it does not appear when the joint dowelled with steel. It can be concluded that the tenon cheeks rolled due to the rougher surfaces friction of GFRP and wood dowel compares to the smooth surfaces of the steel dowel. 8. Failure Mode of Dowels Being stiffer than wood and GFRP, steel dowel acted as an extended rigid loading pin that had transferred the load applied through the dowel and to the tenon. Steel dowel transferring the load to the tenon, reach its maximum load capacity until further cracks of tenon and its stiffness degraded. Figure 5 demonstrates typical load versus displacement for steel, GFRP and wood dowel. The graph indicates GFRP dowel stiffness behaves similar to steel dowel but with a lower value. Fig.5: Typical load versus displacement for steel, GFRP and wood dowel 6

Table 1: Nominal limit states values of steel, GFRP and wood dowel of mortise and tenon Type of Dowel (Nos. of Test) Ave. Proportional limit (kn) Steel (3) 127.28 Different of Prop.

8 Table 1: Nominal limit states values of steel, GFRP and wood dowel of mortise and tenon Type of Dowel (Nos. of Test) Ave. Proportional limit (kn) Steel (3) Different of Prop. Limit of GFRP and Wood to Steel (%) - Ave. Max. Load (experiment) (kn) Different of Max. Load of GFRP and Wood to Steel (%) GFRP (4) Ave. Max. Load (Shank, 2002) (kn) - - Ave. 5% Dia Offset (kn) Different of Max. Load to 5% Dia. Offset (%) Failure Mode (EYM) Mode I m Mode I m Wood (3) to Mode III s Despite the fact that though load capacity for wood is consecutively higher then GFRP, its does not directly conveyed that wood is stronger. The observation of the failure mode of the dowels showed that wood dowel failed after the testing compared to the GFRP dowel. Figure 6 showing the conditions of each dowel after the connection failure. Both steel and GFRP remain its original straight form but small cracks and bent occurs in the wood dowel. The mode of failure is in Mode Im, showing the wood crushing in the tenon member. This also shows that the GFRP and steel dowel is stiffer than the tenon member. Connections dowelled with wood is found in Mode IIIs, showing the wood dowel yield in bending at one plastic hinge points per shear plane and associated wood crushing. STEEL GFRP TIMBER Fig. 6: Form of Dowels after Testing 9. Performance of Mortise and Tenon Typical movement undersides of tenon were observed based on the LVDT 2 which positioned exactly underside of the tenon (Figure 7). These displacements were the minimum compared to the rear-side and the inner-side of the tenon displacement due to the distribution of load by the tenon. Behaviour of the dowels during the test was relatively uniform. Typically, wood dowels displaced more than steel and GFRP dowels due to the load transferred into the wood fibres, which caused it to cracks and bent. From the LVDT 2 load displacement curves (Figure 8), indicates the typical initial increase in joint stiffness, the joints exhibited a linear load displacement response until a series of audible fractures occurred in steel and GFRP dowelled joint. It can be highlighted that this flexibility and bending of wood dowel has delayed the displacement in the connection. The wood dowel flexibility behaviour does contribute in avoiding the tenon from any fracture. 7

$Cracks or bent were not appear in both types of dowels. Figure 9 indicates that cracks occur in the tenon with an audible fracture of beam during testing.$

9 Inner- side LVDT LVDT2 Beam member Rear- side LVDT Fig. 7: Position of LVDT 2 underside of tenon Fig. 8: Typical Underside Displacement of Tenon for Different Dowel Material Plotted from LVDT 2 Failure mode of mortise and tenon dowelled by steel and GFRP were found in similar pattern. Cracks or bent were not appear in both types of dowels. Figure 9 indicates that cracks occur in the tenon with an audible fracture of beam during testing. The tenon failure or wall crushing for both steel and GFRP occurred without the yielding of the dowel (Figure 10). These results are similar to studies reported by Erdil et al. [17] that the tenon failure or wall crushing dowelled by steel generally occurred before yielding of the steel dowel. Vice-versa phenomenon happened in wood dowel, where the small cracks and bent occurs in the dowel and no fracture was found in tenon (Figure 11). Even with lowest strength capacity, wood dowelled mortise and tenon connection failure is not as abrupt as steel and GFRP. The gradual failure of wood dowels is due to the friction between the dowels, mortise and tenon surfaces since it is an inhomogeneous material. GFRP Dowel Beam member Fig. 9: Audible fracture of beam member dowelled with GFRP during testing 8

$Tenon Cracks (a) (b) Fig. 10 (a & b): GFRP Dowel; Audible fracture in tenon without yield or cracks in dowel Dowel Bent (a) (b) Fig.$ Conclusions Shear strength capacity of the mortise and tenon connection single-dowelled with steel, GFRP and wood is not directly reflected the capacity of the dowel.

Conclusions Shear strength capacity of the mortise and tenon connection single-dowelled with steel, GFRP and wood is not directly reflected the capacity of the dowel.

Failure modes of GFRP were found as stiff as steel (mode Im) compared to wood (mode IIIs). This result does foresee that GFRP dowel shall be a viable alternative to steel and wood dowels.

Some factors need to be configured and introduced in order to design and applied GFRP as a dowel material.

10 Tenon Cracks (a) (b) Fig. 10 (a & b): GFRP Dowel; Audible fracture in tenon without yield or cracks in dowel Dowel Bent (a) (b) Fig.11 (a & b): Wood Dowel; without fracture in tenon but small bent and cracks appears in dowel 10. Conclusions Shear strength capacity of the mortise and tenon connection single-dowelled with steel, GFRP and wood is not directly reflected the capacity of the dowel. The proportional limit, 5% diameter offset and maximum strength of the connection are in the order of steel, wood and GFRP but the dowel yield modes are in the order of steel, GFRP and wood. Failure modes of GFRP were found as stiff as steel (mode Im) compared to wood (mode IIIs). This result does foresee that GFRP dowel shall be a viable alternative to steel and wood dowels. However, it is also found that the nominal 5% diameter offset values is not suitable to be used in designing the mortise and tenon joint with GFRP dowel. Some factors need to be configured and introduced in order to design and applied GFRP as a dowel material. Yield strength and behaviour of mortise and tenon connection dowelled with GFRP, is similar to steel but different when dowelled with wood. Thus, a deeper investigation is priory needed in order to develop more appropriate information on strength and stiffness of steel; GFRP and wood as a strengthening material for mortise and tenon connection. Acknowledgements The authors would like to thank the Ministry of Higher Education, Malaysia for funding the project through FRGS, UiTM. 9

11 References [1] R. Brungraber. Traditional Timber Joinery: A Modern Approach, PhD Thesis, University of Stamford, [2] R.G. Erikson, Behaviour of Traditional Timber Frame Structures Subjected to Lateral Loading. PhD Thesis, University of Wyoming, [3] W. M. Bulleit, L. B. Sandberg, M. W. Drewek &T. L. O Bryant. Behaviour and Modeling of Wood-Pegged Timber Frames, ASCE, Journal of Structural Engineering, Vol. 125, (No. 1. January). P. 3-9 [4] L.B. Sandberg, W.M. Bulleit, and E.H. Reid, Strength and Stiffness of Oak Pegs In Traditional Timber-Frame Joints. ASCE, Journal of Structural Engineering, Vol. 126(No. 6, June): p [5] J.F. Miller, Capacity of Pegged Mortise and Tenon Joinery, Msc. Thesis, University of Wyoming [6] J.D. Shanks, and P. Walker, Experimental Performance of Mortise and Tenon Connections in Green Oak. The Structural Engineer, 2005: p [7] C. R. Walker, F. S. Fonseca, J. P. Judd, & P. R. Thorley, Tensile Capacity of Timber-Frame Mortise and Tenon Connections. 10th World Conference on Timber Engineering (WCTE), Miyazaki, Japan. [8] J. R. Church, & B. W. Tew, Characterization of Bearing Strength Factors in Pegged Timber Connections. Journal of Structural Engineering, (3), [9] R.D. Drake, The Advancement of Structural Connection Techniques for Timber Buildings. PhD Thesis. University of Bath [10] A. N. Tankut, & N. Tankut, The Effects of Joint Forms (Shape) and Dimensions on the Strengths of Mortise and Tenon Joints. Turk. Journal of Agric. (Tubitak) [11] Eckelman, C.A., E. Haviarova, and H. Akcay, Exploratory Study of the Widthdrawal Resistance of Round Mortise and Tenon Joints with Steel Pipe Cross Pins. Forest Product Journal, (11/12): p [12] R.J., Schmidt, Timber Pegs Consideration for Mortise and Tenon Joint Design. Wood Design Focus, (3): p [13] C.A., Eckelman, E. Haviarova, and H. Akcay, Parallel-to-grain End-load Capacity of Round Mortises in Round and Rectangular Timbers. Forest Products Journal, (4): p [14] National Design Specification for Wood Construction (NDS), American Forest and Paper Association (AFPA) Washington D.C [15] J.D. Shanks, Developing Rational Design Guidelines for Traditional Joints in Oak Frame Construction. PhD Thesis. University of Bath [16] Malaysian Timber Industry Board, Ministry of Plantation Industries and Commodities, Malaysia. [17] Y.Z. Erdil, A. Kasal, and C.A. Eckelman, Bending Moment Capacity of Rectangular Mortise and Tenon Furniture Joints. Forest Product Journal, (12): p

Engineering Research into Traditional Timber Joints

Engineering Research into Traditional Timber Joints Richard Harris, Wen- Shao Chang, Peter Walker (BRE CICM, The University of Bath) and Jon Shanks (CSAW Research Fellow, School of Architecture & Design,