Impact Localization in Complex Cylindrical Shell Structures Based on the Time-Reversal Virtual Focusing Triangulation Method

Xiufeng Huang, Rongwu Xu, Wenjing Yu, Shiji Wu2024 · DOI

TL;DR

This paper introduces a novel method for pinpointing the location of impacts on large, complex cylindrical structures like ship hulls. By combining time-reversal focusing with a triangulation approach, it achieves accurate localization without needing prior knowledge of the structure's wave properties, offering a practical solution for structural health monitoring.

What Problem Does This Paper Address?

Accurately detecting and locating impacts on large, complex engineering structures is crucial for structural health monitoring (SHM). This is particularly challenging for anisotropic, stiffened cylindrical shell structures, such as ship hulls, where impacts can lead to significant damage. Existing methods often struggle with the complexity of these structures, require detailed prior knowledge of material properties (like wave velocity), or are computationally intensive. The authors aim to develop a robust and practical impact localization method that overcomes these limitations, enabling precise identification of impact points even in noisy environments.

Methods

The proposed method, "Time-Reversal Virtual Focusing Triangulation," combines several techniques:

  • Energy Power Sensor Selection: Sensors are ranked based on the energy of their vibration signals, with the top-ranked sensors being prioritized for data input. This is a heuristic approach to identify sensors likely closer to the impact.
  • Wavelet Packet Decomposition: This technique is used to extract narrowband Lamb wave signals from the complex impact response. It allows for a more refined analysis of the signal's time-frequency characteristics, helping to isolate relevant frequencies.
  • Time-Reversal Virtual Focusing: This core component enhances the signal-to-noise ratio and signal recognition. Instead of physical re-emission, it uses digital signal processing to virtually "re-emit" the time-reversed signals, causing them to focus at the impact location.
  • Average Wave Velocity Calculation: For anisotropic structures where wave velocity varies, an average wave speed is estimated using the time-reversal focusing results and sensor distances. This estimated speed is then used in the triangulation step.
  • Triangulation Localization: The final impact location is determined using a geometric triangulation method based on the arrival times of the signals at the selected sensors and the calculated average wave velocity.

What's novel is the combination of these techniques, particularly the "virtual focusing" aspect and the integration of an estimated average wave velocity for complex structures, which avoids the need for pre-measured wave speeds.

Datasets

The study was validated using experimental data collected from a large-scale Q235 material compartment model (Φ3 m × 8.8 m × 0.01 m). Impacts were applied at 37 distinct locations (labeled Q1 to Q37) under two conditions: with and without a motor operating (simulating background noise). The paper does not describe the creation of a new benchmark dataset but rather uses experimental results from this specific test structure.

Key Insights and Ideas

  • Sensor Selection Matters: Prioritizing sensors with higher energy power significantly improves the likelihood of capturing the most intense response, likely indicating proximity to the impact source. The study found that using the top three sensors in terms of energy power captured the closest measurement point to the excitation source in over 94% of cases.
  • Virtual Focusing Enhances Signals: The time-reversal virtual focusing technique effectively amplifies the impact signal, making it easier to identify key features and improving the signal-to-noise ratio, which is crucial in noisy environments.
  • Average Wave Velocity is a Practical Solution: For complex, anisotropic structures, estimating an average wave velocity is a viable workaround for methods that typically require precise wave speed information, making the approach more broadly applicable.
  • Noise Impacts Accuracy, but Robustness Remains: While background noise (simulated by motor operation) does increase localization error, the proposed method still maintains a high level of accuracy and a low error ratio compared to traditional triangulation methods.

Key Results

  • Non-Motorized Conditions: Average localization error of 0.89 m, average localization error ratio of 6.96%, and 100% localization accuracy (error ratio < 20%).
  • Motorized Conditions (with noise): Average localization error of 1.12 m, average localization error ratio of 8.76%, and 97.30% localization accuracy.
  • Comparison to Traditional Triangulation: The proposed method consistently demonstrated lower localization error and superior performance across various impact points and operating conditions. For instance, at point Q1, the error was 0.60 m compared to 6.29 m for traditional triangulation (motor disengaged).

Limitations and Future Work

  • Limitations:
    • The presence of background noise (motor operation) does increase localization error, although the method remains robust.
  • Specific impact points close to structural boundaries (e.g., upper edge of the compartment) or near ribs showed slightly higher errors, potentially due to complex wave reflections and deformations.
  • Future Work: The paper suggests that the method is applicable to large-scale anisotropic ribbed cylindrical shell structures and demonstrates promising application prospects. Further research could explore optimizing sensor placement strategies and investigating the method's performance in even more complex and dynamic environments.

Summary updated

BibTeX
@misc{huang2024impact,
  author = {Xiufeng Huang and Rongwu Xu and Wenjing Yu and Shiji Wu},
  title = {{Impact Localization in Complex Cylindrical Shell Structures Based on the Time-Reversal Virtual Focusing Triangulation Method}},
  year = {2024},
  doi = {10.3390/s24165185},
}
Figure