188宝金博页面版

Online available since 2005/Nov/15Pattern Classification of Acoustic Emission Signals During Wood Drying by Principal Component Analysis and Artificial Neural Network Ki Bok Kim1 ,a, Ho Yang Kang2,b, Dong Jin Yoon1 ,c and Man Yong Choi1 ,d 1Center for Environment and Safety Measurement, Korea Research Institute of Standards and Science, Daejeon, Korea 2Department of Forest Products Engineering, Chungnam National University, Daejeon, Korea akimkibok@kriss.re.kr, bhykang@cnu.ac.kr Keywords: Acoustic emission, wood drying, principal component analysis, artificial neural network Abstract. This study was performed to classify the acoustic emission (AE) signal due to surface check and water movement of the flat-sawn boards of oak (Quercus Variablilis) during drying using the principle component analysis (PCA) and artificial neural network (ANN). To reduce the multicollinearity among AE parameters such as peak amplitude, ring-down count, event duration, ring-down count divided by event duration, energy, rise time, and peak amplitude divided by rise time and to extract the significant AE parameters, correlation analysis was performed. Over 96 % of the variance of AE parameters could be accounted for by the first and second principal components. An ANN was successfully used to classify the AE signals into two patterns. The ANN classifier based on PCA appeared to be a promising tool to classify the AE signals from wood drying. Introduction Acoustic emission (AE) is widely used to nondestructively monitor structural integrity and characterize the behavior of materials when they undergo deformation, fracture, or both. The AE technique has been used to minimize defects during wood drying. During wood drying differential shrinkage between the rapidly dried surface of a board and the saturated internal regions causes surface tensile stresses and balancing internal compressive stresses. These stresses generate the AEs. Researchers have been investigated the characteristics of acoustic emissions for using them as parameters in a reactive control system for the wood drying process [1-6]. Breese et al. [7] reported that AE and steaming treatments could decrease checking during klin drying without prolonged klin residence time. It was revealed that AEs are dependent on wood density, fiber length, moisture content and temperature concerning their frequency, amplitude and energy [8, 9]. During wood drying AE activity increases once the surface instantaneous strain attained the proportional limit [8], and cumulative AE energy increases rapidly soon after the first small checks could be observed [5]. Considerable AE activity was monitored in both soaked and dried wood in the radial plane due to the weak bonds between the cells of wood rays and axial tracheids [10, 11]. It is important to identify the sources of acoustic emissions to control the wood drying process. Booker [12] proposed that AEs were due to slips in the crystalline portion of cellulose, and that checking occurred when the rate of such slips exceeded a critical value. The studies of AE in mechanically stressed wood have proved that the sources of AEs are the slip lines (slow AE), brittle micro cracks in the cell walls, and the processes of delamination (rapid AE) [13, 14]. With previous experience it could be assumed that the major AE sources are the surface tensile stress related to moisture movement below the proportional limit, the process of cracking and the thermal stress related to the abrupt change of wood temperature. Key Engineering Materials Vols. 297-300 (2005) pp 1962-1967online at http://www.scientific.net© (2005) Trans Tech Publications, SwitzerlandAll rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of thepublisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 130.203.133.33-14/04/08,11:07:55)