It is assumed that owing to its porous nature wood has sound absorption capability, but it is not clear that the structure of wood is reflected enough in the sound absorption characteristics in the measurable spectrum range up to 6400 Hz with a normal measuring device. Therefore the authors constructed new equipment for measuring in a higher range spectrum and attemped an actual survey. Usually the sound absorption coefficient is given at the 1/3 octave band center frequency. However, with our new machine it was necessary to devise a measuring method that gives reasonable sound absorption coefficients. We suggested a method to extract the local minimum of raw sound absorption coefficients. By this method, it was confirmed that a characteristic spectrum was seen in a high range of more than 7700 Hz of the sound absorption coefficient of wood, showing delicate differences between species of wood or wood sections.

View full abstract

Thin spruce veneer was corrugated to make wooden honeycomb elements for the lightweight core of sandwich panels. Wood honeycomb core (VHC) is a honeycomb element in which the fiber direction of the veneer is aligned in the thickness direction of the panel. Corrugated veneer core (CVC) is similar to corrugated cardboard, in which the fiber direction of the veneer is aligned in the direction of the wave. The cores were put between two sheets of plywoods to make sandwich panels and their compressive strength (σ_c) and compressive Young's modulus (E_c) in the thickness direction were measured. The density (ρ_c) of VHC and CVC varied from 10 to 130kg/m³. The specific strength (σ_c/ρ_c) of VHC was 7 to 20 times greater than that of paper honeycomb. The ρ_c and E_c of CVC were almost the same as those of paper honeycomb while its rupture energy was about 3 times greater than that of paper honeycomb. Finer wave resulted in greater ρ_c and σ_c of CVC, and greater σ_c/ρ_c was achieved by smaller curvature with thicker veneer. Plural veneers could be glued during corrugation, but such a plied-CVC showed smaller σ_c/ρ_c than mono-layered CVC probably because of failure in gluing.

View full abstract

The technique of pattern recognition was applied to identify the blue-stained areas of Pinus radiata boards that were discolored by blue stain fungi. After mosaic processing was performed to a digital camera image of the board surface, color information values according to RGB and L^*a^*b^* color systems were calculated from each mosaic block. Pattern vectors constructed of 2 to 6 dimensions, consisting of the color information values, were used to identify the blue-stained and sound areas. Blue-stained and sound areas were determined by visual inspection by 20 persons, and the specimen was used for both construction and evaluation of the algorithm of the pattern recognition function developed using RCE (restricted coulomb energy) network. The function was constructed by training such that the function was revised step by step according to the results of trial recognition of blue-stained or sound areas. The performance of the constructed function was evaluated by the recognition rate of unknown input patterns. The influence of pattern composition on recognition rate was investigated. When the pattern vector was (G, B, L^*, a^*), the recognition function showed good recognition performance for both blue-stained and sound areas. As the numbers of dimension of pattern space increased, the recognition rate of sound area improved, and the rate of unidentified patterns tended to decrease. Even if the numbers of dimension of pattern space increased, the mean of the recognition rate was about the same in the blue-stained areas, and the standard deviation tended to decrease.

View full abstract

This paper deals with the determination of stress and strength of single strapped adhesive joints made of - bamboo and hinoki wood, when subjected to static bending moments. The resulting material properties of the hinoki and bamboo composite were determined by tension tests parallel, perpendicular, and at an angle of 45 degrees to the fiber direction. Stress and strain were calculated using an elastic and orthotropic finite-element method (FEM). In this test, the two hinoki plates were unbounded. The greatest stress was incurred on the hinoki side of the adhesive layer near the ends of the bamboo plate. The result was calculated following the adhesion of two pieces of hinoki wood to a bamboo plate. The effects of variations in thickness and length of the bamboo element relative to the strength of joints were examined. It was found that the strength of the joints increases as the length of the bamboo increases, but the effect of thickness was not very important. In addition, surface strain distributions under static bending moments were measured using strain gauges. Reasonable correlation was found between the numerical and experimental results. Bending strength tests were performed. It was observed that the bending strength of the laminated composite plate increased with an increase in the proportion of bamboo element thickness. The separation process became clear by experiments and through observation of the ruptured surface. It was found that the separation started from the point of greatest stress, which was calculated by FEM. Verification of the calculated result was shown.

View full abstract