In order to validate the applicability of orthotropic models and the predictions of the simulation models described in chapter 7, two commercially available composite lumber materials were tested for various mechanical parameters:
1. 15-layer yellow-poplar Laminated Veneer Lumber, manufactured from 3.2 mm (1/8 in) thick peeled structural veneer sheets. The LVL beams were produced at TrusJoist, a Weyerhaeuser Business, Buckhannon, WV, using a continuous technology and crushed lap joints to connect the consecutive veneers. Veneer sheets are sorted by stress-wave timing to ensure uniform product properties, but the layup is not optimized for maximum flatwise bending MOE.
2. Parallel Strand Lumber manufactured at the same facility, using a mix of 75%
yellow-poplar and 25% southern yellow pine (Pinus spp.) strands. The lumber is produced as large billets, using a continuous technology, at relatively high levels of densification. The billets are later re-sawn into smaller cross-sections.
Figure 5.9 shows the structure of the composites, and defines three orthogonal axes; x is the longitudinal direction of the composites, y is the orientation of the constituents within the cross-section, and z is perpendicular to both. One can realize that these axes are analogous with the longitudinal, tangential and radial directions in solid wood, respectively, especially in LVL, where the anatomical orientation of the veneer layers concur with the length, width and thickness of the beams.
The assessment of the orthotropic properties of LVL and PSL requires the definition of two angles. By substituting the x, y and z coordinate axes for L, T and R, respectively, on Figures 4.1 and 4.5, it is possible to define load orientation (ϕ’) and strand/layer orientation (θ’). These angles are analogous to grain and ring orientation in solid wood, respectively. Using ϕ’ and θ’, it is possible to examine the compression and shear orthotropy of composite lumber in the same way as that of solid wood.
5.3.2 Bending MOE measurements
Twenty structural size beams of both LVL and PSL were tested in 4-point bending. The approximate beam cross-sections of LVL and PSL were 45 by 95 mm and 75 by 140 mm, respectively, and beam length was 2.5 m. LVL specimens consisted of 4 groups of 5 beams manufactured from the same panel, which allowed the determination of both within-panel and between-panel variation. The exact position of the PSL beams in
y x
z
Figure 5.9 – The definition of orthogonal axes for LVL and PSL
Figure 5.10 – Experimental setup of the composite bending tests 0.75 m
F 2
F 2
F 2
F 2
0.75 m 0.75 m
2.25 m
Bending yoke Linear potentiometer
a.
b.
L L
C
the production flow was uncertain, and the specimens were treated as a representative random sample of the population. A conditioning period of several weeks in an environment of 21°C and 45% RH preceded the bending tests. Measurements were carried out using a high-capacity Baldwin universal hydraulic testing machine, equipped with a 220 kN ± 5 N load cell. Testing procedure included both edgewise and flatwise load application for every beam. The applied load never exceeded the linear elastic limit of the beams.
Figure 5.10 shows the testing setup of the bending test. Beam supports consisted of load bearing rollers, one of which provided free lateral adjustment and longitudinal movement for the beams. The load applicators, mounted on a self-aligning crossbar, had a 650 mm radius of curvature. The vertical displacement of the center point (C) relative to the load application points (L) was measured on the middle plane of the beam, using a linear potentiometer mounted on a bending yoke, which was supported on nails driven into the beam (see Figure 5.10.) The linear potentiometer had a measuring range of 0 ~ 125 mm, and an accuracy of 0.01 mm. A computerized data acquisition system collected load and deflection data with a frequency of 1/s. In the absence of specific testing standards for wood-based composite lumber, the location of the load application points and the testing speed were established in accordance with ASTM D 198 – 94. The evaluation of the collected data followed the specification of this standard, too.
5.3.3 Orthotropy of shear strength
The orthotropic shear strength of LVL and PSL was measured to verify the applicability of the models described in section 4.1.1, for these composites. The orientation of the sheared plane and the direction of load application involved combinations similar to the ones used for solid wood (see section 5.1.3). In this case, however, not all combinations were measured. Shear measurements included all planes for the longitudinal (x) direction (ϕ’ = 0° ; θ’ = 0°, 15°, … 90°), and load applications in every direction in the xy and xz planes. (ϕ’ = 0°, 15°, … 90° for θ’ = 0° and 90°.) Shear strength was also determined at ϕ’ = 90° ; θ’ = 45°, to furnish Equations 4.1, 4.4 and 4.5 with every required constant.
The sample size was 9 specimens for each combination. Instead of the testing setup shown on Figure 5.3, the standard ASTM testing apparatus was used (ASTM D 143 - 94). The target shape and size of the specimens were also those specified in the standard, although beam cross-section sometimes limited the dimensions of the specimens. Other details of the measurements, including specimen conditioning, testing speed and the testing machine, were the same as specified in section 5.1.3.
5.3.4 Orthotropy of compression elasticity
The orthotropic compression properties of the composites were assessed to verify the simulation models that estimate the compression MOE of LVL and PSL in different
Equation 4.10), using a sample size of 10 in each direction. Specimen preparation and testing procedures were the same as described in section 5.1.4, with two notable exceptions:
• Available composite beam thickness did not allow the manufacture of 100 mm long specimens in directions that deviated from the xy plane. In these instances specimen manufacture included re-gluing the composite lumber with cold-setting Polyvinyl Acetate (PVAc) wood adhesive, to create beams with double thickness.
• For LVL, the maximum thickness created using the above practice was 87 mm, which restricted the maximum length of LVL compression in the yz and z directions. For these specimens, the gauge length of the strain measurement decreased to 35.5 mm.