Internal Void Defect Identification in Steel Tube Concrete Columns Using Variational Mode Decomposition and Normalized Kurtosis

Literature Overview and Research Background

This paper by Liu Jingliang et al. (2023), published in Journal of Huaqiao University (Natural Science) (Vol. 44, No. 3, pp. 328-335), proposes a novel non-destructive testing (NDT) method for identifying internal void defects in steel tube concrete columns. The research is supported by the National Natural Science Foundation of China (51608122) and the Fujian Provincial Natural Science Foundation (2020J01581). The method combines Variational Mode Decomposition (VMD) with normalized kurtosis analysis to detect and locate internal voids without requiring baseline information from the undamaged state.

From a steel pipe and concrete-filled tube manufacturing perspective, internal void defects are a critical quality concern. During the concrete filling process, voids can form due to improper vibration compaction, air entrapment, or inadequate concrete flowability. These voids compromise the composite action between the steel tube and the concrete core, reducing the structural capacity and durability of the column.

Core Technical Methodology

Signal Processing Framework

The proposed method follows a systematic signal processing pipeline:

1. VMD Decomposition: The measured response signal (typically from impact or vibration testing) is decomposed into multiple intrinsic mode functions (IMFs) using VMD. Unlike traditional Empirical Mode Decomposition (EMD), VMD is a non-recursive variational method that simultaneously decomposes all modes, avoiding mode mixing issues.
2. Effective Component Selection: Weighted kurtosis values are calculated for each IMF. Components with weighted kurtosis exceeding the average value are identified as effective components containing defect-related information.
3. Signal Reconstruction: The selected effective IMFs are reconstructed into a new signal that emphasizes defect-sensitive features.
4. Teager Energy Operator (TEO) Processing: The TEO is applied to the reconstructed signal to enhance transient features associated with impact responses at defect locations.
5. FFT and Normalized Kurtosis: The TEO-processed signal undergoes Fast Fourier Transform (FFT), and normalized kurtosis is computed from the frequency-domain representation to identify defect locations.

Key Advantage: No Baseline Requirement

A particularly significant feature of this method is its independence from baseline (undamaged) reference data. Traditional signal processing methods for damage detection often require comparison with the response of an undamaged structure, which is frequently unavailable in field conditions. The proposed VMD-kurtosis approach eliminates this limitation, making it practical for in-service inspection of existing steel tube concrete columns.

Engineering Practice Implications

Quality Control Applications

For steel pipe concrete column manufacturing and construction, this method offers a practical NDT tool for verifying the quality of concrete filling. During production, columns can be subjected to impact or vibration testing, and the response signals can be analyzed using this methodology to detect voids that may have formed during the filling process.

Welding and Fabrication Relevance

Internal void defects in steel tube concrete columns can be exacerbated by poor welding practices at tube joints. When steel tubes are welded together to form full-height columns, any misalignment or gaps at weld joints can create preferential void formation zones during concrete filling. The proposed NDT method can help identify such fabrication-related defects that might otherwise go undetected by conventional inspection methods.

Comparison with Conventional NDT Methods

Traditional methods for detecting voids in concrete-filled steel tubes include ultrasonic testing (UT), radiographic testing (RT), and acoustic emission (AE). Each has limitations: UT requires access to both sides of the tube and is limited by the steel tube's shielding effect; RT requires specialized equipment and radiation safety measures; AE is sensitive to environmental noise. The VMD-kurtosis method offers a complementary approach that can be applied with relatively simple impact testing equipment.

Study Insights and Critical Assessment

The methodology presented in this paper represents a meaningful advancement in structural health monitoring for composite steel-concrete members. However, several practical considerations merit attention from an engineering standpoint. The method's effectiveness depends on the quality of the measured signal, which can be affected by environmental noise, boundary conditions, and the accessibility of measurement points on completed structures.

The study validates the method through numerical examples and dynamic tests, demonstrating good accuracy in defect location identification. For practical deployment in manufacturing environments, further investigation into the method's sensitivity to different void sizes, shapes, and orientations would be beneficial. Additionally, establishing quantitative relationships between kurtosis values and void severity would enable the method to move from qualitative detection to quantitative assessment.

For steel pipe manufacturers and construction contractors, integrating this signal processing approach into routine quality control procedures could significantly improve the reliability of concrete-filled steel tube columns, particularly for critical applications in high-rise buildings, bridges, and offshore structures where internal void defects can have serious consequences for structural safety and service life.