Design and Implementation of Nondestructive Evaluation Instrument for Acoustics-based Prediction of Wood Modulus of Elasticity and a Method to Assess the Recorded Longitudinal Stress Wave

Abstract

Modulus of elasticity (MOE) of a wood is a value that can predict the mechanical characteristics of a wood. It can predict the impact of a wood crack and internal damage to the overall strength. However, it is still rare to find a high graded wood in Indonesia due to the complexity of the evaluation using a conventional method based on Destructive Evaluation (DT) where the sample tested are purposely destroyed or static Nondestructive Evaluation (NDE) where heavy tools are required. This research examined an NDE method to obtain MOE values and proposed a handheld instrument to conduct dynamic test based on longitudinal stress wave method (LSWM), a popular method among several NDE to examine the MOE values. Earlier researchers show that for various local wood i.e. Sonokeling, Teak, Sukun, Acacia, Munggur, and Mahogany LSWM has a good performance. It has an average correlation of dynamic MOE (MOEd) to the MOE values obtained from the static method (MOEs) or R-squared of 0.898. In this research an instrument was developed based on a proposed enhanced version of the LSWM. In this enhanced version, aside from utilizing the fundamental natural frequency, f0, as the non-destructive parameter, a method, and parameter to assess how much the recorded wave representing the actual longitudinal stress wave is introduced. In this method, the recorded wave assessment is done by performing band-pass filter with center frequency on f0 and extracting the envelope of an impulse response of the filtered longitudinal stress wave and perform fitting to the extracted envelope. The instrument uses a Raspberry Pi 2 microcomputer as the signal processor, an LCD touchscreen, a USB soundcard and a dynamic microphone with a frequency range of 0.1-5 kHz and -64 ± 3dB. It provides semi-automatic NDE calculation and a handheld mobile instrument given the small dimension of 12x8x4 cm and powered with a 1200 mAh USB power supply. The design in overall is less in weight, practical, easy to use, with an internal storage for data transfer capability and the most benefit is the NDE capability to predict modulus of elasticity of woods. A new parameter named R2envl. It is the coefficient to determine the envelope fitting in the damping ratio calculation. After the recorded signal is obtained for the MOEd measurement, we calculate the damping ratio. If a low R2envl is obtained, it is concluded that there is a high probability for the data of not being able to represent the longitudinal stress wave signal and measurement needed to be redone.