ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Acoustic Emission for CFST Interface Damage Identification: NDT Practice and Signal Interpretation

Overview of AE Testing for CFST

Acoustic emission (AE) testing is a non-destructive method that detects elastic waves generated by damage events such as cracking, debonding, and yielding. For concrete-filled steel tube (CFST) structures, AE testing can identify interface damage between the concrete and the steel tube, which is a common failure mode. The paper presents a method based on principal component analysis (PCA) of AE signal features to identify the damage modes. From a steel pipe perspective, AE testing is a valuable tool for monitoring CFST structures during load tests and in service.

AE Principles and Signal Features

AE testing relies on sensors attached to the steel tube surface. When damage occurs, elastic waves propagate through the tube and the concrete, and the sensors detect the signals. The signal features include amplitude, frequency, energy, rise time, duration, and counts.

Signal Feature Description Damage Mode Indicator
Amplitude Peak voltage of the signal Cracking and debonding
Frequency Dominant frequency of the signal Cracking (high frequency) vs. debonding (low frequency)
Energy Integral of the signal squared Severity of the damage
Rise time Time from onset to peak Cracking (short rise time) vs. debonding (long rise time)
Duration Total time of the signal Cracking (short duration) vs. debonding (long duration)
Counts Number of threshold crossings Damage intensity

The signal features depend on the damage type and the location. Cracking typically generates high-frequency, short-duration signals, while debonding generates low-frequency, long-duration signals. The paper uses PCA to reduce the feature dimensions and classify the damage modes.

Principal Component Analysis for Damage Classification

PCA is a statistical method that transforms the signal features into a reduced set of principal components. The first few principal components capture most of the variance in the data, and they can be used to classify the damage modes.

Principal Component Variance Captured Dominant Feature
PC1 60–70%