Use of small thinning logs in a round-wood trussed bridge.
The purpose of this study was to determine the flexural properties of a truss system fabricated with thinnings from a Taiwania plantation and their application to pedestrian bridge construction. The maximum bending capacity at failure of a 6-m truss system fabricated with round-wood members 150 mm in diameter is 1.7 times that fabricated with 120 mm diameter members, and 2.4 times that of a 4-m truss system with 120-mm members. The yield bearing strength of single and double bolted joints for 120 and 150 mm round wood members ranges from 73 percent to 85 percent of maximum dowel bearing strength. The maximum load capacity of bolted joints for 120 and 150 mm round wood members was 5.5 times and 5.4 times the allowable design value. The flexural deflection of a trussed pedestrian bridge subjected to dead load and live load of 3924 N/[m.sup.2] was L/416, which met the code criteria of L/300. The creep behavior of a 6-m pedestrian bridge was in steady stage after 2 months of live loads, and the flexural deflection reached 90 percent of that measured after 6 months. The MC of chord members after the trussed bridge had been completed for 6 months dropped to between 23.6 to 29.2% from the initial green state.
The utilization of immature thinning logs from rapidly grown forests presents challenges during processing, such as warp, split, knots, low mechanical properties, low lumber recovery, and finally low quality of products. The more process steps put on these juvenile materials, the less profit can be obtained from low end products. Wang (1985) reported that the juvenile wood is located within the 18th to 20th annual rings or 70 to 120 mm around the pith for China fir, and the 1lth to 19th annual ring or 50 to 80 mm around the pith for Japanese cedar. Consequently, many small logs generated from thinning operations are simply dumped in the forests without any further commercial usage. Relatively low strength values of bending, compression, and tension of Finnish and UK Scots pine, Finnish spruce, Dutch larch, Finnish Douglas-fir, and UK Sitka spruce small-diameter (80 to 130 mm) round wood were reported in recent European research (Stern 2002). In general, the quality of raw materials required for structural application is much lower than that for furniture products, and the processes are also less complex, which may have cost advantages or make products more satisfactory. Furthermore, Wolfe (2001) suggested that small-diameter logs have a greater load capacity than the largest prismatic timber sawn from the same logs, and strength variability for round logs is about one-half to two-thirds that of lumber.
The objective of this research was to determine the properties of trusses made with small-diameter plantation wood and evaluate their performance in pedestrian bridge construction.
Materials and methods
Wood from a fast grown Taiwania plantation (Taiwania cryptomerioides Hay.) was used in round shape as structural members, which enclosed most natural defects inside of log shanks. These round-wood members were constructed into different configurations of efficient truss systems with bolted joints at each node. The bearing strength of bolts in juvenile wood was evaluated with a tensile test, and the flexural properties of trussed round-wood beams were determined with a static test. A pedestrian bridge was then constructed with trusses and loaded for the evaluation of long-term performance in service.
Taiwania logs of 200 mm in diameter or less were selected from a 20-year-old mixed forest plantation during a thinning operation. All the logs were peeled and shaped into round wood members of 120 mm (designated as D120) or 150 mm (designated as D150) in diameter. A groove 3 mm wide with a depth equal to one-third of diameter was cut along the whole length of each member before air drying. This is proven effective to control splitting of round wood during the air-drying stage. The grooves at both ends of round wood were further cut through to accommodate plates for bolt connection. The MC of air-dried round wood was 15.0 [+ or -] 0.4 percent, and the average growth ring width was 5.8 [+ or -] 0.1 mm. The SG of round wood was 0.42 [+ or -] 0.07 and the compressive strength parallel to the grain and modulus of elasticity were 29.4 MPa and 9.32 GPa, respectively.
Bolt joint test
A tensile test was used to evaluate bolted joint strength of round-wood members to simulate the structural performance at critical node of the wood truss. Single-bolt joints were fabricated with round-wood members of both 120 and 150 mm in diameter and 500 in length, while double-bolt joint was fabricated with 700 mm long members (Fig. 1). Each testing condition had six replications. The testing speed was 2 mm/min and the dowel bearing strength ([f.sub.e]) can be estimated as:
[f.sub.e] = P/Ld 
where P: load, L: bearing length of dowel, d: diameter of dowel.
[FIGURE 1 OMITTED]
Warren trusses 6 m in length were assembled with both D120 and D150 members, while trusses 4 m in length were assembled with D120 members only (Fig. 2 and Figure 3). Two sets of structures were assembled for each truss configuration. Each end of round wood members was connected with the required number of bolts, 15.88 mm in diameter, and single stainless steel plate, 5 mm in thickness, inserted in the middle of the cross section (Fig. 4). The required number and placement of bolts for each node of the truss structures were estimated based on the specifications of the building code (MOI 2003) and design loads for the pedestrian bridge (MOTC 2001).
[FIGURES 2-4 OMITTED]
Static test of truss
The static bending test for round-wood trusses was performed using two-point loading as specified by ASTM D 198-94 (1996). The loading heads were applied on the nodes of the upper chord members, and four sets of lateral supports were against the specimen during the load application. The stiffness (EI) can be estimated as:
E1 = ([DELTA]P x a)/(48 x [DELTA]y) x (3[L.sup.2] - 4[a.sup.2]) 
where [DELTA]P: difference between total loads at two levels within the proportional limit; a: distance from end support to adjacent load point; [DELTA]y: flexural displacement difference corresponding to loads for [DELTA]P; L: supported span of truss specimen measured 75 mm from both ends of truss; and the equivalent distributed loads (W) can be estimated as:
W = (8P x a)/([L.sup.2] x [W.sub.b]) 
where [W.sub.b]: width of pedestrian bridge; P: maximum loads from two-point loading test.
The force distribution of truss beams was also checked with a Space Truss Analysis software (DYNACOMP, Inc. 1994).
Long-term performance of trussed bridge
Two trusses consisting of D150 round-wood members were used to construct a pedestrian bridge. The span of the 6-m wood bridge was 5.2 m long with a 1.1-m wide deck. Trusses were preservative treated with chromated copper arsenate (CCA) after being assembled with bolts. The superstructure of the bridge was then fastened to CCA-treated southern pine sills, which were fixed on the concrete footings with anchor bolts. Cabot water-based clear paint was applied after the construction was completed. Truss members were in green state during construction. The MC of the lower chord was measured every month using an electric resistance type of moisture meter (Hydromette H35 Germany). A total of 216 pieces of cyclindrical concrete blocks with the size of 150 mm (diameter) by 300 mm (height) each, which equal to 3924 N/[m.sup.2] of design distributed live loads, were placed on the wood deck evenly along the bridge span. A dial gage (Mitutoyo 3058F) was installed at the midspan under the truss bridge, and the flexural deflection was recorded every day for the first week, every 3 days for the following 3 weeks, and every week for the rest of the time until unloaded at the end of 6 months. The decay, termite attack, mold, blue stain, and paint on the trussed round-wood members were investigated after long-term exposure to the weather.
Results and discussion
Dowel bearing strength
The relationship between bearing stress and displacement of bolted joints for round wood in the tensile test is shown in Figure 5, which showed a large portion of nonlinear deformation before failure occurred. According to the yield theory, the bearing strength of wood and the yield moment of the dowel type fastener are governing properties for determining the shear strength of bolted joints (Sawata and Yasumura 2003). The resulting strength of round-wood bolted joints is shown in Table 1. There is a relationship between bearing stress of a bolted joint and the length/diameter ratio (L/D). As indicated from the Wood Handbook (USDA 1987), the bearing stress becomes larger as the L/D ratio increases. The values of L/D were 9.4 and 7.6 for round-wood members of 150 and 120 mm in diameter, respectively, in this study. The bearing strength of bolted 150 mm round-wood members was 88 percent and 81 percent of that for 120 mm round-wood members fastened with single and double bolts, respectively, which showed a similar tendency. The procedures outlined for values of parallel-to-grain bolt values in the Wood Handbook are based on the 5 percent exclusion value, duration of load, and the effect of L/D ratio. Consequently, the derived values of 7.32 and 5.79 MPa can be obtained for 120- and 150-mm round-wood members, respectively, which were 34.9 percent and 31.7 percent of test results. On the other hand, when the bolt joint is designed based on the Code procedure, the maximum load capacities of bolted joints for 120- and 150-mm Taiwania round-wood members are 5.5 times and 5.4 times the allowable design values, respectively (Lin 2004).
The 5 percent off-set method was employed to evaluate the yield bearing strength. A line that goes through the points on the curves corresponding to 10 percent and 40 percent of the maximum loads up to 5 mm displacement was moved about 5 percent of the dowel diameter horizontally (Fig. 5). The yielding bearing strength is defined as the intersection of this line and the load-displacement curve. The results showed that the estimated yield bearing strength of both single and double bolted 120-mm round-wood members was 85 percent of maximum bearing strength, while it was 73 percent and 85 percent for 150-mm round-wood members bolted with single and double bolts, respectively. Sawata and Yasumura (2002) also found the yield bearing strength for spruce and fir laminae connected with 8, 12, 16, and 20 mm bolts was 69 percent of maximum values measured at failure. This indicated that a larger range of linear behavior from the load-displacement relation was found for Taiwania thinning wood.
[FIGURE 5 OMITTED]
Flexural properties of round-wood trusses
The maximum load measured at failure of a 6 m beam assembled with 150-mm round wood (D150-6) was 70.4 KN, which is 72.3 percent higher than that with 120-mm round-wood members (D120-6) as indicated in Figure 6. The major failure occurred at web members of trusses between loading points and supports, i.e., within the range of shear span. One of the failures was a split at the end of the member due to the tensile force, and the rest of the cases were obviously due to bending of bolts. The latter failure mode can be classified as Mode IIIs as specified in National Design Specification (NDS), which represents fastener yield in bending at one plastic hinge point on each shear plane, and bearing-dominated yield of wood member in contact with the fastener (AFPA 1997). This means the flexural properties of trussed round-wood beams were dependent on the strength of both wood members and mechanical fasteners at the joint.
[FIGURE 6 OMITTED]
The maximum bending moment of a truss increased as the diameter of round-wood members and span of the truss increased. The maximum bending moment of D150-6 truss is 1.6 times of that measured from D120-6 beam and 2.4 times of D120-4 truss case (Fig. 7). The failure of beams mainly occurred at the nodes of diagonal web members outside of the load span. These web members were assembled with a single bolt at each end and subjected to calculated tension forces of 29.6, 41.7, and 48.8 KN for D120-4, D120-6, and D150-6 trusses, respectively, based on the maximum bending loads. These calculated tension forces are close or over the tension capacities of single bolt strength as shown on Table 1. Therefore, it is suggested that more bolts may be appropriate at joints where shear forces exist in order to improve the loading capacity of trusses based on the bending stress concern.
[FIGURE 7 OMITTED]
The equivalent EI values were estimated based on the slope of load-displacement curves in Figure 6, and load levels selected were within linear portions of curves, which ranged between 25 to 50 percent of maximum loads depending on each loading case. It is noted that low EI values were due to the high initial stiffness of truss system, which showed nearly no deflection before the applied load reached around 5 kN. If we investigate each truss system, the D 120-6 truss started to deflect around 3.23 kN, which led to less displacement, while D150-6 truss with higher initial stiffness started to deflect around 7.35 kN, which led to larger displacement. Consequently, the actual deflections of wood trusses are higher than analyzed results as a result of initial stiffness. The results indicated that the bending stiffness of a D 120-4 beam was only 22 percent of that of D120-6 beam. This was probably due to larger shear deformation in a D120-4 beam under the small span/depth ratio of 4, which is subjected to 29.5 KN of shear force within the range of shear span, i.e., 47 percent higher than that of a D120-6 beam. Furthermore, the bending stiffness of a D150-6 beam was three times that of a D120-4 beam, showing greater resistance to flexural deflection. However, the bending stiffness of D150-6 is smaller than that of D120-6. This is probably due to a slight lateral deformation occurred during the load application for D120-6 trusses. Under the design live load of bridge, 400 kgf/[m.sup.2], the measured flexural deflection of truss was only 38 percent, 13 percent, and 5 percent of Code criteria of L/300 for D120-4, D120-6, and D150-6 beams, respectively, while the deflection was 49 percent, 20 percent, and 16 percent of L/300 when live load and dead load were combined.
Long-term performance of round wood bridge
The 6-m pedestrian bridge was designed based on live load and actual dead load, and subjected to the exposure of weather. The stresses of critical tension and compression round-wood members of the trussed beam were 39.7 percent and 28.2 percent of allowable design values, respectively. The allowable design value for 15.88 mm bolts in double shear, wood-to-metal connections was 12.3 percent of the tested load capacity as listed in Table 1. Therefore, the structural performance of the constructed bridge is acceptable based on the test results.
The moisture changes of truss members were investigated at the south side of the bridge where there was less exposure to the sun. The MC of members were above fiber saturation point (FSP) during the bridge construction and dropped to between 23.6 and 29.2% after the 6-months investigation. Some members had MC below FSP at the depth of 30 mm in wood, while the MC still remained over FSP in the core of round wood. Several members also showed higher MC during the fifth and sixth month than earlier because of rain showers that lasted for weeks. A small drop in MC, i.e., 3 percent, was also found after 28 months for a stress laminated timber bridge by Ritter and Wacker (1995).
The relationship of deflection-time is shown in Figure 8. The instant elastic deflection at the center of the bridge span right after the distributed load was applied was 5.8 mm, which is equal to L/897. The primary creep occurred mainly during the second day. The deflection of the trussed bridge continued to increase because of a subsequent typhoon event. The creep deflection reached 14.1 mm or the relative creep became 1.43 after 3 months, and then approached leveling off, i.e., the secondary creep stage. The final deflection due to dead load and live load was 14.41 mm or L/416 as compared to the code requirement of L/300. The duration of load factor specified in NDS is 2 months. The data recorded at the end of 2 months was 12.87 mm, 91 percent of total deflection or 1.22 for relative creep, meaning the bridge had already approached the stable creep stage. The instant recovery of the round-wood trussed bridge was 3 mm, half of initial instant deflection, after all the live load was removed. After 1 month of relaxation, the creep deflection recovered 4.33 mm, equal to 50.8 percent of measured 6-months creep deflection. There was no termite or insect attack, no decay or stain observed, and paint remained in good condition after exposure to the weather for 7 months.
[FIGURE 8 OMITTED]
The effective utilization of thinning logs in structural applications is a major concern for management of rapidly grown plantations. It is suggested that the flexural performance of trussed round-wood beams can be improved when using larger sizes of thinning logs and longer spans based on the span/depth ratio consideration. The maximum load capacity of bolt joints is five times higher than the design values, assuring effective joint strength for juvenile wood members. The flexural deflection of a pedestrian bridge subjected to dead load and live load and exposed to the weather for a period of 2 months was at the L/466 level and without any deterioration in wood materials in a further 5-month investigation, showing acceptable structural performance.
AFPA. 1997. National Design Specification for Wood Construction. American Forest and Pap. Assoc. and American Wood Council. ANSI/ AFandPA NDS-1997, P20, 51-74.
American Soc. for Testing and Materials (ASTM). 1996. Standard methods of static tests of timbers in structural sizes. ASTM D198-94. ASTM. West Conshohocken, Pennsylvania.
DYNACOMP, Inc. 1994. Space Truss Analysis with Graphics. New York. 15 pp.
Lin, Y.L. 2004. Application of domestic thinnings on the trussed bridge. Master thesis, National Pingtung Univ. of Sci. and Tech. Taiwan, ROC. 79 pp.
MOI. 2003. Tech. Design and Construction for Wood-framed Buildings. Ministry of Interior, Taiwan, ROC. Chapter 5.
MOTC. 2001. Highway bridge design specification. Traffic Tech. Standard Specification. Highway Engineering Dept. Ministry of Transportation and Communications, Taiwan, ROC. Chapter 2.
Ritter, M.A. and J.P. Wacker. 1995. Design, construction, and evaluation of timber bridge constructed of cottonwood lumber. In: Proc. 4th Inter. Bridge Engineering Conf. Vol. 2. pp.358-370.
Sawata, K. and M. Yasumura. 2002. Determination of embedding strength of wood for dowel-type fasteners. J. of Wood Sci. 48:138-146.
-- and --. 2003. Estimation of yield and ultimate strength of bolted timber joints by nonlinear analysis and yield theory. J. of Wood Sci. 49:383-391.
Stem, E.G. 2002. Construction with small-diameter roundwood. Forest Prod. J. 51(4):71-82.
USDA. 1987. Wood Handbook: Wood as an engineering material. USDA Forest Serv. Forest Products Soc., Madison, Wisconsin. pp.7-13-17.
Wolfe, R. 2001. Res. challenges for structural use of small-diameter round timbers. Forest Prod. J. 50(2):21-29.
Wang, S.Y. 1985. Properties of medium- and small-diameter timber. In: Proc. of Syrup. of the Utilization of Medium and Small Diameter Trees. Taiwan COA Forestry Series No. 1. pp.27-36.
The authors are, respectively, Professor, Research Assistant, Dept. of Wood Sci. and Design, National Ping-tung Univ. of Sci. and Technology, Nei Pu, Ping Tung, Taiwan, 91201 R. O. C. (email@example.com). This project was funded by National Sci. Council. This paper was received for publication in April 2005. Article No. 10046.
Min-chyuan Yeh * Yu-li Lin
*Forest Products Society Member. [C] Forest Products Society 2007. Forest Prod. J. 57(3):34-38.
Table 1.--Tension properties of bolted joints with different diameters of Taiwania round wood from thinnings. No. of Max. embedding Yield embedding bolts Max. tension loads strength strength (KN) (MPa) 120 mm 150 mm 120 mm 150 mm 120 mm 150 mm Single 37.49 41.28 19.70 17.35 16.66 12.74 (5.9%) (a) (13.4%) (6%) (13.6%) (6.5%) (6.9%) Double 79.36 80.53 41.65 33.81 35.48 28.71 (6.7%) (5.6%) (6.8%) (5.5%) (11.3%) (11.9%) (a) Values in parenthesis are coefficients of variation
|Printer friendly Cite/link Email Feedback|
|Author:||Yeh, Min-chyuan; Lin, Yu-li|
|Publication:||Forest Products Journal|
|Date:||Mar 1, 2007|
|Previous Article:||Development of design capacities for residential deck ledger connections.|
|Next Article:||Exploratory study of small timber trusses constructed with through-bolt and cross-pipe heel connectors.|