Identification of defects of wood by using wave analysis of tap tone and application to machine stress rating of lumber

研究了利用敲击声波形分析识别木材缺陷及其在木材机械应力等级评定中的应用.通过波形分析识别缺陷，采用传递矩阵法计算了试样缺陷引起的光谱峰位移。得到了估算缺陷引起的谱峰位移的数值方程。通过使用具有圆孔的试样的实验显示出与传递矩阵法的数值模拟相同的趋势。具有多个孔的样品的峰位移通过单个孔的影响的线性组合来评估。提出了一种新的缺陷位置识别原理。确定了位置，使实验值与估算值之间的吻合度达到最大.通过波形分析确定振动参数敲击音不仅包括基本振动中的有用信息，还包括更高阶振动波中的有用信息。利用傅里叶变换的滤波过程，分离出单个模态的波。然后采用希尔伯特变换得到时间包络，并确定对数衰减量。此外，通过希尔伯特变换得到的瞬时频率显示了对踢踏音进行准时频同时分析的潜力.应用波动信息对木材机械应力进行评定。利用杨氏模量、对数衰减率和高阶振动频率的不规则性进行的多元回归分析比利用杨氏模量进行的简单回归给出了更高的相关性。

Identification of defects in lumber by a wave analysis of tap tone and application to a machine stress rating of lumber were studied.1. Identification of defects by a wave analysis.A transfer matrix method was used to evaluate the spectrum peak shift due to defects of a specimen. A numerical equation which estimates the spectrum peak shift by defects was obtained. Experiments by using specimens having round holes showed the same tendency as the numerical simulation of the transfer matrix method. The peak shifts of the specimen having multiple holes were evaluated by a linear combination of the effect of individual hole. A new principle to identify the defects positions was proposed. The position was identified so that the coincidence between experimental and estimated values becomes maximum.2. Determination of vibration parameters by a wave analysisA tap tone includes useful information not only in a fundamental vibration but also in higher order vibration waves. Isolation of the wave of an individual mode was done by using the filtering procedure of Fourier transformation. Then Hilbert transformation was adopted to obtain a time envelope and logarithmic decrement was determined. Moreover, an instantaneous frequency given by Hilbert transformation showed a potential of a quasi time-frequency simultaneous analysis of a tap tone.3. Application of wave information to machine stress rating of lumberSmall and full size wood specimens were used. A multiple regression analysis using Young's modulus, logarithmic decrement and irregularity in frequencies of higher ordered vibrations gave higher correlation than the case of a simple regression which used Young's moduli