The longitudinal flow rate of air through bar type specimens of fifteen species including domestic and imported woods was measured. From the results, the permeability characteristics for each species, and the effects of specimen length (l) and moisture content (u) on the permeability were discussed. (1) In the longitudinal direction of wood, the permeability value calculated based on Darcy's law (kg) was a function of l with some exceptions, and a linear relationship was found between l and logarithm of kg in the range of 5-45cm sample length, as expressed by kg=K0exp(-bl), where kg is the superficial gas permeability. (2) The longitudinal permeability of wood was extremely different from one species to another. In the case of air dried specimens of broad leaved trees, the values of K0 ranged from 0.02 to 180 darcy, and constant b was 0.03-0.56. In the case of those of coniferous trees, K0 was from 0.04 to 1.2 darcy and b was 0-0.03. Apitong (Dipterocarpus grandiflorus BLCO.) was the most permeable wood among the species tested and Balsa (Ochroma lagopus SW.) was also very permeable. Witch hazel (Distylium racemosum SIEB. & ZUCC.) had the least value of permeability. The heartwood of Cryptomeria (Cryptomeria Japonica D. DON) and Formosan cypress (Chamaecyparis obtusa ENDL. var formosana HAYATA) were less permeable than thier sapwood. (3) The value of kg showed a low value for green specimens, but the value increased rapidly with drying at least below 25% of u with an exception of the heartwood of Siebold's beech (Fagus crenata BL.) and b became more or less smaller with decreasing u. This variation in permeability with u seemed to be reversible unless the specimens were dried to the moisture content below 5% with severe drying conditions.
As a part of the investigation on the structure of vessel walls in hardwoods, the structure of the primary wall was studied using ultra-thin sectioning and replica techniques. Microfibrils of the primary wall extend straight and are arranged parallel one another. The primary wall is composed of three layers, PO, PM and PI, each showing the different microfibrillar orientation. Microfibrils of the outer layer PO are oriented transversely with respect to the vessel axis and microfibrils of the middle layer PM are oriented at random. The inner layer PI consists of the crossed polylamellate structure. On the other hand, microfibrils of the primary wall pass through the pit membrane maintaining their orientation. The crossed polylamellate structure is also present in the pit membrane. From these facts, it is considered that the structure of the pit membrane is essentially same as that of the primary wall.
Loading often occurs on coated abrasive paper and fuzzy grain appears on a surface of wood sanded with coated abrasive paper, while no loading occurs on equisetum (scouring rush; horse-tail) which gives wood great gloss. The purpose of this study was to reveal these characteristics of polishing wood with equisetum, a natural abrasive meterial, compared with sanding wood with coated abrasive paper. A work specimen was quarter-sawn Japanese cypress (Chamaecyparis obtusa Endl.) and abrasive tools were equisetum (Equisetum hiemale L. var. japonicum Milde.) and coated abrasive paper (silicon carbide 2C, 240 and 600 grit). A quantitative analysis by means of TAPPI standard method T 15 m-58 indicated that an entire stem of equisetum contains about 20% of ash. An observation through a scanning electron microscope showed that lots of removed chips were stored in grooves running on both sides of the warty processes while chips firmly overlapped one another among grits on coated abrasive paper and loading occurred there and that the chips removed with equisetum were thinner than those removed with coated abrasive paper. Since the heights of tips of grits on coated abrasive paper are approximately represented by a normal distribution, deep scratches appear on the surface of wood sanded, which may be responsible for fuzzy grain. On the other hand, no scratch is left on the surface of wood polished with equisetum probably because the heights of grits of warty processes are about equal and their tops approximately semi-spherical. The measurements with simulated single cutting edges (semi-spherical tips with radii of curvature of 0.4 and 1.2mm for equisetum and a quadrangular pyramidal tip with an apical angle of 80 degrees for grit) indicated that the ratio of vertical cutting force to horizontal one was significantly greater for the model of equisetum than for that of grit; equisetum required more vertical force than coated abrasive paper to abrade at an equal indentation. These experimental evidences suggest that the warty processes of equisetum might act on a surface of wood like a calender in the paper-making process. It is concluded that a unique structure of equisetum gives a surface of wood great gloss while coated abrasive paper causes rather fuzzy grain on wood.
In order to examine the wear of tungsten carbide saw tips, laminated experimental blocks of particleboard were interruptedly or continuously turned with a single saw tooth tipped with fine, F, medium, G, and coarse, H, grained carbide. The wear of the cutting edge was observed using a light and a scanning electron microscope, and the four wear parameters and the three cutting force components were measured. The cutting edge of carbide tip wore more on the back face than on the front face, irrespectively of the kind of tip and the method of cutting. The back wear land surface showed a smoother appearance than that of the front face, and for tips F and G striations and grooves developed in the cutting direction on it. In both interrupted and continuous cutting, tip H wore and blunted most among three kinds of tips. Tip G had a constant higher wear resistance, particularly in continuous cutting. Tip F showed a higher performance in interrupted cutting, although its back wear land was relatively wide. It was concluded that the impact in interrupted cutting does not accelerate the wear of carbide tip so much, but some unknown factor in continuous cutting affects more strongly its wear and blunting.
Bending fatigue is one of the important factors which must be considered in designing stressed-skin panels. In the present study, bending fatigue tests were carried out on stressed-skin floor panels with glued plywood or particleboard skin under concentrated cyclic load. The function of repeated load was sinusoidal and the minimum load was 5kgf. There were several failure modes both in static and fatigue tests. The order of the change in failure modes with increasing number of load repetitions in the fatigue test was the same as that with increasing load in the static test. The ratio of the fatigue strength at 106 cycles to the static strength (RFS) was 43-54%, and RFS of delamination of glue line between frame and skin was also nearly the same. Although both plywood-skin panels (PL) and particleboard-skin panels (PB) have nearly the same rigidity of flexure, the static and fatigue strengths of PB were merely 50-65% of those of PL. The fatigue strength of wet PB-panels was 17% lower than that of air-dried PB-panels. The increase of deflection under repeated load was also observed. A sudden increase of deflection was caused by sudden propagation of crack in glue line between frame and skin. Therefore, the delamination of glue joint between frame and skin by fatigue is one of the most important factors to be considered for designing of stressed skin panels.
Partly pulsating tension load with the frequency range of 1Hz to 16.7Hz was applied to timber butt-joints with metal plate connectors. The time dependent effect on fatigue strength was discussed with respect to the relation between cycles to failure (N) and loading components, i. e., repetitive load (Pr) and constant load (Pc). The results obtained are as follows; (1) N obtained by partly pulsating tests at 16.7Hz and 10Hz did not depend on Pc but dominantly on Pr, when the sum of Pc and Pr was less than 70% of the static strength (Po) and Pr was grater than the fatigue limit (Pe). (2) The failure mode of the joints in partly pulsating tests was mostly the withdrawal of teeth of the connector. This tendency was observed only in high stress levels of pulsating tests. (3) In lower frequency tests at 5Hz and 1Hz, N decreased to 1/2 (5Hz) and 1/3 (1Hz) of the value of N obtained in 10Hz tests, when Pr/Po was from 0.4 to 0.6 and the sum of Pc and Pr was 750kg.
In this paper, the relation between the anatomical features and the elastic anisotropy in the transverse direction of softwood was investigated experimentally and numerically by using a simple model. The cell arrangement in the cross section of wood was approximated by two kinds of porous models, which were made of an isotropic and homogeneous material. They corresponded to earlywood and latewood. Tensile tests were carried out in various directions between the two principal axes of these models. Then, the numerical values of n, which is an index of anisotropy decided by both the geometrical features of deformable unit of wood and its volume fraction, were determined. The modulus of ray tissue was calculated using the anatomical model of its fine structure. On the basis of these experimental and calculated data, the elastic moduli in various directions of whole wood were simulated by considering wood as a complex laminate with two stages of lamination, i. e. a laminate of alternating layers of transversely anisotropic earlywood and latewood perpendicular to ray tissue. The mixture rule was used in order to compute the elastic modulus of this complex laminate. Young's modulus perpendicular to the grain of sugiwood and hinokiwood at any given angle to the radial direction was obtained by measurements with special u-gages designed to eliminate various errors. From these results, it was found that the elastic anisotropy of wood was caused predominantly by the arrangement of pores and macroscopically laminated structures.
In order to examine the effect of the structural inhomogeneity of wood on its Young's modulus, an analytical model was proposed. The model was constructed by considering the structural arrangement of earlywood and latewood cells in the cross section of softwoods, and also the fine structure of the cell wall (Fig. 1). The Young's moduli of earlywood and latewood in the transverse direction were evaluated on the basis of Finite Element Solution for the repeating units of the model. Data on the orientation of the microfibrils and area percentage of the layers in the cell wall presented by Saeki for Japanese cypress (Hinokiwood) were adopted, and the elastic constants of the cell wall layers were determined by means of three methods shown in Table I. The results were compared with the experimental values used by Boutelje for the samples consisting of isolated tissue components made of Swedish pine-wood (Table II, III). It was indicated that the agreement between the calculated and experimental values was fairly good in magnitude, and the elastic anisotropy in the radial and tangential directions was affected by the geometry such as shape and size of the cell rather than the fine structure of the cell wall.
In order to elucidate the effects of moisture content on the strain energy to the proportional limit Up and that to the maximum load Um, the mechanical characteristics (σp, εp: stress and strain at the proportional limit, E: modulus of elasticity in bending, σm, εm: stress and strain at the maximum load) in bending were determined by the central loading on beams of Sitka spruce (Picea sitchensis CARR.). The measurements were carried out at 20, 30, 40 and 50°C under the nominal moisture content of 0, 5, 10 and 15%. The strain energy to the proportional limit Up and that to the maximum load Um were calculated by finding the respective areas under the load-deflection curve and dividing them by the volume of the span portion of beam. The results obtained are as follows: (1) The values of σp, σm, εp and E decreased almost linearly, and the value of εm increased, with the increase of moisture content. (2) The strain energies of Up and Um decreased almost linearly with the increase of moisture content. (3) The change of Up and Um with moisture content was dependent on the relative relation between the percentage change of stress and strain corresponding to the 1% change in moisture content. In this experiment, the average percentage change in stress (σp, σm) with the 1% change in moisture content was more than that in strain (εp, εm).
The dynamic Young's modulus and the internal friction of many specimens made of a trunk of Akaezomatsu and that of Kuroezomatsu were measured using a flexural vibration method, and the comparison in these properties between Akaezomatsu and Kuroezomatsu was made. In order to find out the difference in dynamic properties of the wood classified by the experience of piano industry, these specimens were classified into the high grade wood used for grand piano soundboards, the medium grade wood used for upright piano soundboards and the low grade wood which is not used for soundboards, and their properties were investigated statistically. The main results obtained are as follows: (1) It was observed that in acoustical properties Kuroezomatsu was not inferior to Akaezomatsu when used for the soundboards of musical instruments, and that the percentage of the low grade wood of the former was about twice that of the latter. It was also found that such a high percentage of the low grade wood was mainly caused by the colour of Kuroezomatsu. Consequently, it seems that the reason why Kuroezomatsu is not used for the soundboards is not because of its acoustical properties, but because of its low yield. (2) The results of the statistical investigation on the properties of the classified wood agreed with the general opinion on the physical properties which are desirable for the wood for soundboards. Consequently, it was found that the selection method of the wood for soundboards used in piano industry, which was learned through experience, was marvelous.
The durable fire-retardants based on the colloidal solution of boric acid- and alkali metal containing melamine-formaldehyde condensates (H3BO3·AOH·MFAC) and 10 boron-containing inorganic compounds have been applied to Hinoki (Chamaecyparis obtusa Endl.) sliced veneer, Lauan (Shorea negrosensis Foxw.) plywood, and Sugi (Cryptomeria japonica D. Don.) shake. The condensates were prepared by use of boric acid, hydroxide of alkali metal and melamine-formaldehyde condensation products. These boric acid- and alkali metal-containing melamine-formaldehyde condensates given clear solution in both neutral and alkaline. The borates used were the same as those appeared in a paper of M.A. Kasem and H.R. Richard.5) The treatment of Hinoki sliced veneer was accomplished by soaking samples in an aqueous solution of these condensates and borates, drying at a room temperature and curing at 160°C for 30min. The treatment of Lauan plywood and Sugi shake were done by soaking in the aqueous solution of the fire-retardants and reducing the pressure by using a vacuum pump, drying at a room temperature for 1 month, and then curing at 160°C for 30min. The effects of alkali metal kind and of varying ratio of fire retardant components in treating solutions on the fire-proofing efficiencies of Hinoki sliced veneer, Lauan plywood, and Sugi shake were discussed. Boric acid-containing melamine-formaldehyde condensates combined with alkali metals imparted excellent fire-proofing resistance to wood and plywood. Satisfactory fire-proofing efficiencies can be obtained in Hinoki sliced veneer, Lauan plywood, and Sugi shake at resin content as low as 5-10%. The overall afterglow of the treated samples depended on the competition between the inhibiting effect of boron and the enhancing effect of alkali metals.
Particleboards were manufactured with adhesives containing fire-retardant chemicals. The physical and mechanical properties, and fire resistance properties of the boards were determined and discussed. The fire-retardant chemicals used were boric acid, lithium hydroxide, lithium dihydrogenphosphate and phosphoric acid, and the adhesive was urea-melamine-formaldehyde resin. The fire-retardant adhesive containing boric acid was stabilized by the addition of lithium hydroxide and lithium dihydrogenphosphate. The fire-retardant adhesives imparted excellent fire resistance to the particleboard. The fire-retardant particleboard containing 12-15% resin in oven dry wood particles passed 1st-2nd class extinguishments by JIS A 1322 and the board containing 6-12% resin also passed 2nd-3rd class extinguishments. Increasing the resin content in oven dry wood particles resulted in an effective increase in internal bond strength as well as MOR, a remarkable decrease in thickness swelling in water, and a slight increase in the equilibrium moisture content. Excepting a few results, the internal bond strength and the MOR of the fire retardant particleboards were greater than those of the untreated one, and the thickness swelling of the fire retardant particleboards in water was smaller than that of the untreated one.
Flake type wood particles were oriented through a 5cm deep frame orienter, in which many vertical plates were set in parallel each other at regular intervals. The effects of flake length, plate spacing and free fall distance from the bottom of plates to the top of the particle mat on the degree of alinement, and the effects of face flake length on physical and mechanical properties of 15mm thick three layer boards were investigated. The particles were prepared from thinnings of Sugi (Cryptomeria japonica). All boards had random orientation in the core; half of the boards had random faces, the rest had faces comprised of oriented flakes. The solid content of adhesive (melamine-urea resin) added was 11% for face particles and 7% for core in oven-dry state. Weighing ratio of face to core particles was 1:3:1 in dry state. Press time was 15 minutes at 150°C and the pressure was 35kg/cm2. The average alinement angle increased with increasing free fall distance and plate spacing, and with decreasing flake length. Provided that all other conditions remain unchanged (e.g. slenderness ratio of particle, resin content and board density), the bending strength of boards may increase 1.6 to 2.2 times as much by orienting the face flackes. MOR and MOE values parallel to the orientation of alined boards showed continuous increase up to the particle length-thickness ratio 150, and reached a plateau which is very close to that of solid wood. The average retention of MOR after 2hr hot water (70°C) immersion of oriented boards was 74%. As for the internal bond strength and thickness swelling, there was no difference between the oriented and the random boards.