2012 IEEE Symposium on Business, Engineering and Industrial Applications Experimental Performance of Mortice and Tenon Joint Strengthened with Glass Fibre Reinforced Polymer under Tensile Load

Rohana Hassan, Azmi Ibrahim and Zakiah Ahmad Institute for Infrastructure Engineering and Sustainable Management (IIESM) Faculty of Civil Engineering, Universiti Teknologi MARA, Shah Alam, MALAYSIA Email address: [email protected]

Abstract— This paper reported on the experimental beam vertically restrained by holding down with roller performance of mortise and tenon joint for beam-to-column supports to a strong floor at 47.24 in. centres. The tenon was structural member made of Kempas species. The joints were loaded through a 0.98 in. steel dowel, through articulation to a dowelled with glass fibre reinforced polymer (GFRP) and load cell and hydraulic jack operated from a hand pump. All compared to the joints strengthened with steel dowel and wood joints were found failed in tension by withdrawal of the tenon dowel. The failure modes performance analyses were observed following failure of the dowel. As stated by Schmidt [12], the based on European Yield Model (EYM) theory according to National Design Specification (NDS) 2008. Results show that the tension strength of a mortise and tenon joint connected with experimental performance of the mortise and tenon joints one or more wood dowels can be predicted using the yield dowelled with GFRP under tensile load may provide comparable model approached in the NDS. strength to the joints dowelled with steel. The sequence of having the higher capacity to hold the mortise and tenon joints is by II. GFRP- IN TIMBER APPLICATIONS strengthening it using GFRP, followed by steel and wood dowel. Fibre reinforced polymer (FRP) is composed either from The failure behavior of both joints dowelled with steel or GFRP carbon (CFRP) or glass polymer (GFRP). The fiberglass cloth were almost similar. Nevertheless, different behavior was shown was used to strengthen the single-bolted double shear wood in wood dowelled joints, the flexibility of wood dowels extended the displacement longer made the yield load and load at rapture connections by Soltis et al. [13]. They found out that the extended compared to the other types of joints. reinforced joints has increased its strength by 33 % in comparison to the non-reinforced parallel to grain and more than twice the strength for perpendicular to grain stresses. Keywords-Structural joint, Kempas, steel dowel, wood dowel Their findings also proved that the catastrophic failure associated with tension-perpendicular to grain stresses was I. INTRODUCTION changed to a ductile mode of failure when the joints were bonded with the fibre-glass cloth. In other related research, Tension capacity is also known as pull-out capacity of the timber beams dowelled with GFRP dowels has been observed mortise and tenon joint. According to Brungraber [1], when according to its development and applications by Gentile et al. the tenoned member is in tension, the members will never bear [14]. They studied the flexural behavior of creosote-treated solidly on the sides of the mortise. In facts, the wood dowel sawn Douglas fir timber beams dowelled with GFRP dowels. will simply continue to deform in shear and crush until the The results have shown that using the proposed experimental tenon pulls out of the mortise. Most of the mortise and tenon technique changed the failure mode from brittle tension to joint were studied more in tension [1] ,[2], [3], [4], [5] and [6] compression failure, and flexural strength increased by 18 to rather than bending [1], [6], [7], [8] and [9] and shear capacity 46%. Research findings indicate the use of near-surface GFRP [1], [6] and [10]. Comparison between mortise and tenon dowels overcomes the effect of local defects in the timber and dowelled with steel and wood dowel has also been reported by enhances the bending strength of the members. Hassan et al., [11]. Sandberg et al. [3], has tested the mortise and tenon in tension using the double shear or sandwich Other studies using the cloth-fibre made of Carbon Fibre concept. This method was developed to represent the common Reinforced Polymer (CFRP) is the investigation of the flexural mortise and tenon joints by arranging the middle pieces loaded strength of timber structures by Schober and Rautenstrauch, in parallel to the grain and the side members loaded [15] from Germany. The tests showed, the arrangement of the perpendicular to the grain. Shanks [6] reported that the tension reinforcement, the bond surface quality and the stiffness of the tests for mortise and tenon for green Oak has the average of load transmitting materials were of decisive influence for the 1686 lbs. as the ultimate load. The test was done by having the overall strength of the specimen.

U.S. Government work not protected by U.S. copyright 856 Wood members dowelled with mechanically fastened with TABLE 1: EYM FOR DOUBLE SHEAR TIMBER-TO-TIMBER cloth- FRP has also been reported by Dempsey and Scott [16] EQUATIONS ACCORDING TO NDS, 2005. from Savannah. Their findings supported the Schober and Rautenstrauch [15] by stated that the proposed strengthening technique induced a gradual failure of the composite members Failure Mode Failure and increased ultimate moment, initial stiffness, and ductility Descriptions Mode over that found for the control specimens. Dempsey and Scott [16] increased the fastener spacing and found that it has decreased the member ultimate moment, initial stiffness, and Im ductility ratio. The moisture content of the wood material greatly affected the ductility ratio of the wood members. Triantafillou [17] has studied the mechanical behavior of wood members either reinforced or dowelled with fiber- Is reinforced plastic (FRP) materials in the form of cloth (laminates or fabrics) externally bonded to the shear-critical zones. He as well found the result was quite satisfactory. Similar studies were done by Plevris and Triantafillou [18] with concerns on the tension zone of wood beams and beams columns using epoxy resins. The analysis is extended to IIIs establish a methodology for the optimum selection of the FRP reinforcement to optimize mechanical performance.

III. NATIONAL DESIGN STANDARD (NDS, 2005) The theoretical background of NDS, 2005 [19] is IV based on European Yield Model (EYM). It has been accepted globally as an engineering approach in joint design and it was developed by Johansen [20].The double shear timber-to- timber connections as stipulated in NDS, 2005 marked four

(4) types of failure modes. These failure modes are listed as IV. EXPERIMENTAL SET-UP below and illustrated in Table 1. Eight (8) series of tests from Kempas were composed. The actual structural size of column and beam dowelled with • Yield Mode Im = Wood crushing in the main member. Dowel stiffness is greater than wood mortise and tenon joint inserted with single dowel were tested. strength. Pre-drilling of the timber elements was adopted in all specimens. The rate of loading was applied at 0.07 in/min. for • Yield Mode Is = Wood crushing in the side members. Dowel stiffness is greater than wood strength. all tests. Limiting rotation was defined by the maximum stroke of the hydraulic jack that is at 9.84 in. All joints were tested • Yield Mode IIIs = Dowel yield in bending at one plastic hinge point per shear plane and associated until total failure. Full-size column and beam with mortise and wood crushing of side members. tenon connection specimens were prepared and constructed as • Yield Mode IV = Dowel yield in bending at two shown in Fig. 1. plastic hinge points per shear plane and associated wood crushing.

857 Tension Load tenon dowelled with GFRP has the highest strength value; closely followed by joint dowelled with steel and with wood. U-Shape Bolt (to grip) The mortise and tenon joints of Kempas dowelled with GFRP steel Plate is 4.66% stiffer compared to joints when it was dowelled with steel. Joints strength dowelled with wood was at 26.24% lesser than joints dowelled with steel. It was also found that the LVDT 1, 2 Restrained mortise and tenon joints dowelled with GFRP are the stiffest and 3 (Rear by a roller side of compared to strengthened with other types of dowels. support bolted to strong LVDT When increasing loads, the load-displacement behaviour of floor mortise and tenon in tension passes through four different (Both phases. These four phases are clearly identifiable, but the sides) transitions are different from each type of dowel material. The

LVDT Strong Wall

Tenoned member Dowel load-displacement behaviour of the mortise and tenon joint of Strong Wall different dowel material is illustrated in Fig. 2.

5000 4500 Mortised member 4000

) 3500 3000 lb s.

( Figure 1: Configuration of experimental tensile test set up 2500 2000 Phase V V. RESULTS AND ANALYSIS Load Phase III T-KPS-S The performance and the load carrying capacity of the joints 1500 Phase IV were observed through their strength, stiffness and failure 1000 T-KPS-G Phase II modes behavior. 500 Phase I T-KPS-W A. Strength And Stiffness 0 The mean shear strength and stiffness values of the joints are shown in Table 2 respectively. -0.5 0 0.5 1 1.5

Displacement (in.) TABLE 2: TENSION STRENGTH AND STIFFNESS CAPACITY OF MORTISE AND TENON JOINTS Figure 2 : Typical tension load-displacement curve of mortise and tenon joint dowelled with GFRP, steel or wood dowel. Type of Stiffness 5% Dowel at 5% Offset Ultimate Offset Load Load From Fig. 2 it was noticed that at the phase I, load- (lbs./in.) (lbs.) (lbs.) displacement curves linear due to elastic behaviour of the Steel materials and joints. At the second phase, sudden drop after Mean 24315.80 3763.05 4332.82 yield can be seen in both joinst strengthend with steel or Std.dev. 8904.31 1194.02 1316.35 GFRP dowels whilst an extended of yielding noticed for joints CoV(%). 37 32 30 strenghened with wood dowels. It was noticed that the sudden GFRP drop of both curves for joints with steel or GFRP dowels was Mean 28325.71 3938.37 4540.30 not due to the failure of the dowels, instead it was the failure Std.dev. 4329.94 1096.76 638.81 of the joint member. The steel and GFRP dowels were found CoV(%). 15 28 14 remain in its straight from after failure. A loud snap sound Wood can be heard simultaneously with the drop of curves. Phase III Mean 11611.44 2775.55 3717.22 shown an increment in load and displacement after the sudden Std.dev. 4198.02 912.07 1067.93 drop or the stagnant up to the ultimate load. Phase IV depicted CoV(%). 36 33 29 load-displacement curves shows non-linearity which due to mortise side or tenon-end failure for steel and GFRP dowelled It depicted the results of 5% offset load for the joint joints, or dowel failure for wood dowelled joints. Loud snaps dowelled with GFRP, steel or wood dowel. The mortise and sound were heard during the joints failure caused of either the cracking and spliting of mortise side or tenon end or both.

858 Phase V shown a continuous snap sound and bearing of the dowel at the tenon-end. Full pull-out of tenon-end occurs not long after the wood dowel failed. Nevertheless, the failure behavior of both joints dowelled with steel or GFRP were almost similar. Different behavior was shown in wood dowelled joints, the flexibility of wood dowels extended the Original Wood displacement longer made the yield load and load at rapture Angle Dowel extended compared to the other types of joints. of Dowel B. Failure Mode Behavior

The failure mode behaviour of the joints were summarised according to the NDS, 2008. It was found that, the joints dowelled with GFRP or steel corresponds to a similar mode of failure, which is mode Im. Whilst all joints Figure 5: The Draw-through of wood dowel. dowelled with wood failed in mode IV (Table 3). When a mortise and tenon is loaded by a tension force on TABLE 3: FAILURE MODE BEHAVIOUR the tenoned-member, the load is mainly transferred by bearing Types of Dowel Nos. Test Failure Modes stresses in the contact zone of the mortised-member at the (EYM, NDS 2005) edge side, the tenon part and the dowel. This means that the GFRP 7 I main portion of the load is transferred by bearing stress m perpendicular to the mortised-fibre direction and parallel to Steel 8 I m the tenoned-fibre direction. The bearing stress in this study Wood 8 IV agreed well with the analytical bearing stress model as

predicted by Brungraber [1]. The bearing stress of joint Steel or GFRP dowel occurred without the yielding of the dowelled with wood under tensile load is similar to the ‘late dowel due to the stiffness of the materials (Fig. 3and Fig. 4). ultimate’ in his study. It was found that the localised stressed

at the middle of wood dowel (at the tenon-end) has caused the

dowel to splits or cracks at the centre of the dowel or within the tenon area.

C. Conclusion It can be concluded that the experimental performance of the mortise and tenon joints dowelled with GFRP under tensile load may provide comparable strength to the joints dowelled with steel. The sequence of having the higher capacity to hold

the mortise and tenon joints is by strengthening it using GFRP, (a) (b) followed by steel and wood dowel. The failure behavior of

Figure 3: Mortise and tenon dowelled with GFRP loaded in tensile, after both joints dowelled with GFRP or steel were failed in mode failure (a) mortise-side and tenon-end and (b) dowel-mode Im. Im. Nevertheless, different failure mode shown in wood dowelled joints, the flexibility of wood dowels extended the displacement longer made the yield load and load at rapture extended and failed in mode IV respectively.

ACKNOWLEDGEMENT

The authors would like to thank the Ministry of Science, Technology and Environment, Malaysia for funding the

project through FRGS, UiTM.

(a) (b) REFERENCES Figure 4: Mortise and tenon dowelled with steel loaded in tensile, after failure [1] R. Brungraber. ‘Traditional Timber Joinery: A Modern Approach’, PhD (a) mortise-side and tenon-end and (b) dowel-mode Im. Thesis, University of Stamford. 1985. [2] R.J. Schmidt, and E. D. Daniels, Design Considerations for Mortise and Tenon Connections. Report for Timber Framers Guild. Becket. This also indicates that dowelled the mortise and tenon joint MA. 1999. with GFRP or steel performed in a similar manner and well than dowelled with wood. The draw-through of the wood [3] L.B, Sandberg, W.M, Bulleit, and E.H, Reid. Strength and Stiffness of dowel was also noticed during the tests. (Fig. 5). Oak Wood dowels In Traditional Timber-Frame Joints. ASCE,

859 Journal of Structural Engineering. Vol. 126, No. 6, Issue June; p. [15] K.U. Schober, and Rautenstrauch, Experimental Investigations of 21620. 2000. Fleural Strengthening K. of Timber Structures with CFRP. [4] J.F. Miller. Capacity of Wood dowelled Mortise and Tenon Proceedings of the International Symposium on Bond Behaviour of Joints, Master Science Thesis, Department of Civil and FRP in Structures (BBFS 2005). International Institute for FRP in Architectural Engineering, University of Wyoming, Laramie Construction; pp. 457-464. Wyoming.2004 [16] D.D. Dempsey, and D.W. Scott, Wood Members Strengthened [5] C.A. Eckelman, E. Haviarova, and H. Akcay. Parallel-to-grain with Mechanically Fastened FRP Strips, Journal of Composites for End-load Capacity of Round Mortises in Round and Rectangular Construction. 2006. Vol. 10, No. 5; pp. 392-398. Timbers. Forest Products Journal. 2007.Vol. 57, No.4; pp. 66-71. [17] T. C. Triantafillou. Shear Reinforcement of Wood Using FRP [6] J.D. Shanks. Developing Rational Design Guidelines for Materials. Journal of Materials in Civil Engineering.1997.Vol 9, Traditional Joints in Oak Frame Construction.. PhD Thesis. pp 65-69. University of Bath. 2005. [18] N. Plevris, and T. C. Triantafillou, (1992). FRP-Reinforced Wood [7] C.A., Eckelman, E. Haviarova. Load Capacity and Deflection as Structural Material. Journal of Materials in Civil Engineering. Characteristics of Large Wooden Dowels Loaded in Double Shear. Vol 4, pp 300-317. Forest Products Journal . 2007. Vol. 57 No. 5; pp 60-64 [19] NDS, (2005). National Design Specification for Wood Construction [8] R. Hassan, A. Ibrahim and Z. Ahmad. Bending Behaviour of American Forest and Paper Association (AFPA) Washington D.C. Dowelled Mortise and Tenon Joints in Kempas. Scientific [20] K.W. Johansen (1949). Theory of Timber Connection, Research Journal (2008): Vol. 5 (1): ISSN 165-7009: pp. 1-11 International Association of Bridge and Structural Engineering, [9] R. Hassan, A. Ibrahim and Z. Ahmad. Shear and Bending No. 9, pp 249-262. Performance of Mortise and Tenon Connection Fastened with Dowel. Journal of Tropical Forest Science .2010: 22(4): pp. 425- 432 [10] R. Hassan, A. Ibrahim and Z. Ahmad. Shear Capacity of Dowelled Mortise and Tenon Joint in Tropical Timber. Journal Institute of Physics, lop Conferences Series: Material Science and Engineering. 2010: Vol. 17.(1): pp 1-11 [11] R. Hassan, A. Ibrahim and Z. Ahmad. Performance of Mortise and Tenon Connection Fastened with Wood and Steel Dowel. Proceedings of WCTE 2010, Trentino, Italy, 20-24 June 2010. Paper No. 264.

[12] R.J. Schmidt. Timber Wood dowels Consideration for Mortise and Tenon Joint Design. Wood Design Focus. 2006. Vol. 14, No.3; pp. 44-47.

[13] L. A. Soltis, and T. L. Wilkinson. Bolted-Connection Design (General Technical Report No. FPL-GTR-54). United States, Forest Service: Department of Agriculture. 1987. [14] C. Gentile, D. Svecova, and S. H. Rizkalla. Timber Beams Strengthened with GFRP Bars: Development and Applications. Journal of Composites for Construction. 2002.Vol. 6, pp 1-71.

860