Stress Measurement of a Pressurized Vessel Using Ultrasonic Subsurface Longitudinal Wave With 1–3 Composite Transducers
Autor: | Daniel Morrow, Taeyang Kim, Xiaoning Jiang, Howuk Kim |
---|---|
Rok vydání: | 2020 |
Předmět: |
Materials science
Acoustics and Ultrasonics Acoustics Transmitter 01 natural sciences Finite element method Stress (mechanics) Transducer 0103 physical sciences Principal component regression Waveform Ultrasonic sensor Electrical and Electronic Engineering 010301 acoustics Instrumentation Longitudinal wave |
Zdroj: | IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control. 67:158-166 |
ISSN: | 1525-8955 0885-3010 |
DOI: | 10.1109/tuffc.2019.2941133 |
Popis: | This article aims to develop a stress-sensing method for a pressurized vessel based on subsurface longitudinal (SSL) waves confined in a specific waveform by using 1–3 composite transducers. Although ultrasonic SSL waves have been commonly utilized for stress sensing, wave excitation under the predefined function using the composite-type transmitter is not well studied. In this article, composite-type transducers having a wide frequency bandwidth (> 60%) and a predominant thickness mode are utilized to enhance the signal intensity of the SSL wave and the accuracy of the sensor by incorporating a specific toneburst waveform. Finite element analysis demonstrates that the signal intensity of the composite-type transducer is up to 45.3% higher than that of a single-phase transducer. Pulse-echo tests reveal that the frequency bandwidth of the developed transducer reaches up to 60.7% and is, therefore, sufficient (> 57.0%) to transmit and receive Hanning-windowed toneburst signals. Results of stress sensing affirm a linear relationship between the time delay of SSL wave and the mechanical stress of a pressurized vessel (0.335 ns/MPa). Accordingly, the regression model is constructed via principal component regression (PCR) under temperature-varying condition. PCR has a less significant degree of error (0.62 MPa) compared to that of a typical least square regression (9.49 MPa). |
Databáze: | OpenAIRE |
Externí odkaz: |