The elastic constants of plywood can be mathematically calculated when those values of veneers and glue layers have previously been known. In the fact, however, that glued veneers in the plywood have plenty of minute cavities filled with adhesives, it is suggested that there must be variation from the single veneer to the glued veneer in their mechanical properties. It is necessary in the calculation of the constants that such variation will be taken into account. So in the first place the variation in the elastic constants of Japanese Oak was experimentally examined (under three different conditions); in solid wood, in the rotary-cut veneer and in the“veneer-laminated wood”, in which all the veneers were laid in parallel with their grain direction. Another investigation was carried out to calculate Young's moduli and the modulus of rigidity of plywood constructed of Red Lauan, Basswood and Japanese Oak using the elastic constants obtained from the veneer-laminated wood. The results of the tests are summarized as follows: (1) Of the longitudinal direction there has been but little difference made to Young's modulus (Figs. 4 and 6) by the lathe checks and gluing process. (2) The veneers 3mm thick showed remarkable drop in Young's modulus tangential to about 15% as hard as solid wood, while with the veneers 1mm and 2mm thick, the drop was only to 50% and 30% as hard respectively (Fig. 5). (3) When glued together the veneers 3mm thick recovered its stiffness to be 85% as hard as solid wood, while thinner veneers showed full recovery, or even showed excess stiffness (Fig. 7). (4) Altogether 11 samples of plywoods were examined to calculate their elastic constants, and there has been good agreement of their calculated values with their observed values (Table VII).
Under the axial tensile force in the fiber direction the distribution of crystal lattice strain in the b-axis direction of the cellulose (I) in the cell wall of tracheids were measured by the X-ray diffraction technique using (040) reflection of cellulose micelles. In this study the transit X-ray method was used and the diffraction profiles were obtained by 40sec fixed time and θ-2θ method. The macroscopic surface strains of the specimens in the fiber direction were measured by the electrical strain meter and resistance type strain gage, pasted on both sides of the specimens. The lattice strain in the fiber direction was a little smaller than the surface strain of the specimens, but the average ratio of the lattice strain to the surface strain was in the range of 0.90∼0.97 below the surface strain level of 3000×10-6. The lattice strains decreased with the increasing inclination from the fiber axis. In the inclined direction from the fiber axis, the lattice strains were a little smaller than the normal strains of the specimen calculated from Eq. (10), of homogeneous orthogonal anisotropic body, but the appearance of distribution of lattice strains in the cell wall was similar to that of the normal strains of the specimens. From those results it is considered that the micelles which were arranged in the cell wall have actually taken over a portion of the applied stress and acted an effective role in the mechanical properties of wood. It is qualitatively observed that the micelles have largely been improved in their orientation by the axial extension in the fiber direction.
The aim of this study is to investigate the viscoelastic anisotropy in the transverse direction of wood in terms of its porous structure. Considering wood as a porous material consisting of the substance and the void, the following relation is given: logE=nlogθ+logES, where E is the apparent modulus, ES is the modulus of wood substance, θ is the volume fraction of wood substance and the numerical value of n is called “form exponent”. The contribution of the porous structures such as geometry and distribution of cells to the modulus of wood can be evaluated by the two factors, θ and n. The dependence of grain angle and angle of annual ring on n is discussed qualitatively as follows; we find that n in any direction is larger than three principal directions, and especially in RT-plane the angle of annual ring which is 45° takes the largest value of about 4 for n. The effect of the thickness of the sample on Young's modulus has been examined and it is found that the discrepancy between the theoretical and the experimental value of form exponent can be explained in terms of the effect of the thickness of the specimen. Finally, on the relaxation modulus at 20°C, 45% R.H., we find that n takes the value of about 1.1 in radial direction and about 1.5 in the tangential direction, and also that they are both independent of time.
When timber and wooden based materials are used as members of the structure, they are often subjected to the bending load and there are varieties of supporting conditions at the edges. There are for example, simply supporting edges and built-in edges besides the elastically built-in edges which are the commonest. The mechanical behavior of plates and of many other materials naturally changes in the bending with the supporting conditions at the edges. In this report, the coefficient α and k which depend on the rigidity of restraint along the edges in the bending of the strip are taken into consideration (cf. Eqs. (5) and (8)) and the quantity α and k have been obtained experimentally in the bending of hardboard strip as to a few supporting conditions at the edges……by Sugi edge and Rubber edge (cf. Fig. 3). The results are shown in Table II. When the edges are supported elastically (by Sugi edge and Rubber edge), the quantity α and k take the values between those of the simply supporting edges and the built-in edges. Then the cofficient k is applied to calculation of the deflection of the square plate with elastically built-in edges (by Sugi edge and Rubber edge). Substituting the coefficient k and other values in Eq. (21), we can calculate P/wo of the square plate. P: Concentrated load acting at the center of the square plate, wo: deflection at the center. The calculated value of P/wo and the experimental value are shown in Table III. They coincide with each other very well when the sides of the square plate are 30 and 35cm respectively in length. Although the problems on the supporting conditions at the edges are difficult and complicated, this is a helpful substitute in calculating the deflection of the plate with elastically built-in edges.
The first discussion of the buckling of plywood appears to have been made by Balabuch.1) March1) has also dealt with the problem both theoretically and experimentally. There have been many writers besides them who have contributed to finding methods in various ways for the buckling of plywood plates. In many cases, the strain energy method has to be used. One of the few cases where an exact and simple solution is known is that of a simply supported 0° or 90° plate under unidirectional compression (P2=0, S=0 in Fig. 1). In this case, equation (4) reduces to the known accurate result (5). On the other hand, Norris has pointed out that the stress pcr at which the buckling first occurs can be written as equation (7) and the buckling stress coefficient kc is given in equation (8), where r is chosen to make pcr the minimun. Previous experimental discussions have been limited to specific construction of plywood of equal veneer thickness. In this report, plates of Lauan Plywood of various construtions as shown in Table I have been tested, and the observed values of kc have been compared with the calculated values. The values of kc is plotted against a/b(E2/E1)1/4. Examples of these curves are shown in Fig. 3 (a)-(g). When a/b(E2/E1)1/4 is greater than 1, the value of kc is approximately independent of a/b, and kc(a/b→∞) which is defined as equation (10) which is arranged to (11) including Q(=EL/ET), A and D, where A and D depend on the construction of plywood. The effect of A and D on kc(a/b→∞) is shown in Table III, when Q is equal to 20. E1 is Young's modulus of plywood parallel to the direction of compression (Parallel to the side a in Fig. 1), E2 is Young's modulus of plywood perpendicular to direction of compression (Parallel to the side b), EL is the longitudinal Young's modulus of the solid wood composing the plywood, ET is the tangential Young's modulus, and A is determined in formula (c), D is determined in (g), where tn are shown in Fig. 4.
It is very difficult to solve precisely the fundamental equations of simply supported orthotropic plywood shallow shells (Eqs. (2)and(3)) when the axes of elastic symmetry do not coincide with those of co-ordinates (γ16≠0, γ26≠0, δ16≠0 or δ26≠0). The authors attempted to solve them by means of the computer FACOM 230-60 (Kyoto Univ.) and by application of the finite difference method, and succeeded in solving them with good approximate accuracy (see Fig. 1). The computed results, especially that on the influence of curvature, are shown in Figs. 2∼4. As shown in Fig. 2, the elliptic paraboloidal shells and cylindrical shells are more rigid than the hyperbolic paraboloidal shells when the four edges are simply supported by means of rollers. The distribution of bending moments and membrane stresses become more complicated as shown in Figs. 3 and 4 when the fiber directions incline to the edges. Examination of these figures will make clear the mechanical characteristics of shells with orthotropic layers. Their experimental analysis was also made using the simply supported shallow cylindrical plywood shells made of beech veneer under uniformly distributed 7×7 points-load (see Fig. 5). The experimental results of the cross laminated shells whose face grains are parallel to the direction of the curvature (Ortho. 0°) and those inclined at 45° (Ortho. 45°) are shown and compared with the computed results in Figs. 6∼10. The comparison shows that good agreement between them has been obtained. But it also shows that there is a little difference between the experimental support condition and the theoretical one (see Figs. 6∼12). This difference is mainly due to the edges of the shells having slipped perpendiculer to the edges on the supports (see Figs. 8, 9 and 11). The transverse shear deflection which is neglected in the theoretical analysis is also considered to be one of the causes of the difference in Fig. 6. The deflection of the parallel laminated shell is about twice as large as that of the cross laminated shells and the strain caused by the membrane stress is about 7 times large (see Fig. 11). And the parallel laminated shell was destroyed along the center line when the total load reached only 126kg (see Fig. 7). The strain distributions of the plywood shells under a concentrated load at the center were also observed and the experimental results are shown with the computed ones in Fig. 12.
Tests were carried out to examine the influence of the construction variables on the structural performance of load-bearing wall panels used in prefabricated wooden houses. Six types of framings were constructed of spruce wood (Fig. 1), and the framing members were cut out for each section, 4.5 by 7cm, except for the bottom members, which were each 7 by 7cm (Fig. 2). The skin materials of 6mm lauan plywood were jointed on both sides or on one side of the framing according to the type of the specimens. In Figs. 4, 5 and 8 the schemata of loading specimens are shown respectively. The bending tests were carried on over the span of 220cm with the load applied at each quarter point. The racking tests were performed without installation of the tie rods as of ASTM standard. The loading was repeated with increased load till a certain specific stage was reached and still continued evenly to the failure of the material (Fig. 7). The compression tests were performed of the specimens, except specimens SR1, with the load applied on their center line of thickness of the specimens, using the knife edges on both their ends as equipment. The specimens SR1 were tested in accordance with ASTM standard. The results of these tests are summarized as follows: In the bending tests the ratio of the observed flexural rigidity to what was calculated is various with the arrangement of framing members, the method of plywood-timber joints and the form of stressed skin construction (Tables I, III and IV). This fact shows that the effective breadth of the skin has considerable effect on the flexural resistance of the panel. In the racking tests on glued and nail-glued specimens without opening it is shown that the racking load is little affected by the difference in the framing construction, so long as the deformation at the upper end is horizontally equal to 1/100 of the height of panel (P1/100kg/m)(Table II). The racking resistance of nailed specimens is weaker than that of glued specimens (Tables II, III and Fig. 7). In the racking tests of specimens with opening the position of the opening makes larger difference to the racking resistance than the opening ratio does (Table III and Fig. 3). The glued specimens (SRC, ERD, SRG) being equipped with powerful plywood skin and reinforced frame around the opening, show but little decrease in the racking resistance. In the compression tests it is shown that the buckling load increases so far as there is increase in the vertical members of framing. The value of the buckling load calculated by Euler's formula for the pin end column, substituting the observed flexural rigidity, is in good agreement with what was observed, though with a few exceptions (Table IV).