A Concrete Core Void Imaging Approach and Parameter Analysis of Concrete-Filled Steel Tube Members Using Travel Time Tomography: Multi-Physics Simulations and Experimental Studies

Wenting Zheng, Bin Xu, Zongjun Xia, Jiang Wang, Jingliang Liu, Yudi Yao, Yifei Wang2024 · DOI

TL;DR

This paper introduces a novel approach using curved ray theory-based travel time tomography (TTT) to visualize internal concrete core voids in concrete-filled steel tube (CFST) members. By optimizing inversion parameters and validating with simulations and experiments, the method offers a promising tool for assessing the structural integrity of these widely used construction elements.

What Problem Does This Paper Address?

Concrete-filled steel tube (CFST) members are increasingly used in large-scale civil engineering projects due to their superior mechanical properties. However, internal defects like concrete core voids and interface debonding can significantly compromise their load-carrying capacity and stiffness, posing safety risks. Visualizing these internal defects is challenging due to the composite nature of CFSTs. While existing non-destructive testing (NDT) methods like stress wave and vibration analysis can detect defects, they often struggle with precise localization and quantification of internal voids. This paper aims to bridge this gap by developing and validating a TTT-based imaging technique specifically for visualizing concrete core voids in CFST members.

Methods

The core of this research is the application of curved ray theory-based travel time tomography (TTT). Unlike traditional straight-ray methods, curved ray theory accounts for the complex wave propagation paths within the heterogeneous CFST structure, leading to more accurate imaging. The TTT method works by measuring the travel times of stress waves emitted by piezoelectric lead zirconate titanate (PZT) actuators and received by PZT sensors mounted on the exterior of the CFST member.

The process involves:

  • Wave Measurement: Generating stress waves using PZT actuators and recording their arrival times at PZT sensors.
  • Travel Time Extraction: Identifying the first arrival time of the stress wave response signals.
  • Curved Ray Tracing: Calculating the theoretical travel paths of stress waves, considering the material properties of steel and concrete.
  • Inversion Algorithm: Employing a least square iterative linear inversion algorithm to reconstruct the velocity (or slowness) distribution within the CFST cross-section. This algorithm iteratively refines the estimated slowness field to minimize the difference between measured and calculated travel times.
  • Parameter Optimization: Investigating the influence of key parameters on imaging quality, including:
  • Model Weighting Matrix (Wm): Explored different operators (unit, first-order, second-order difference) to constrain the inversion.
  • Weighting Factor (λ): Adjusted to balance data fitting and regularization terms.
  • Grid Size (d): Determined the resolution of the reconstructed image.
  • Multi-Physics FEM Simulation: A finite element model (FEM) was developed to simulate the stress wave propagation and generate synthetic travel time data for numerical studies.

Datasets

The paper utilizes simulated datasets generated from multi-physics finite element models (FEMs) of CFST members with various void configurations. Additionally, experimental data was collected from six fabricated CFST specimens with intentionally created concrete core voids. These experimental results serve as the ground truth for validating the proposed imaging approach.

Key Insights and Ideas

  • Curved Rays are Crucial for CFSTs: The inherent heterogeneity of CFSTs (steel tube and concrete core) necessitates the use of curved ray tracing for accurate wave path prediction, significantly improving imaging compared to straight-ray methods.
  • Second-Order Difference Operator is Optimal for Wm: The "second-order difference operator" for the model weighting matrix (Wm) yielded the best imaging results, striking a balance between capturing defect details and suppressing noise.
  • Optimal Parameters are Key: The choice of weighting factor (λ) and grid size (d) significantly impacts imaging quality. The study identified specific optimal values (λ=5, d=1 cm) for the tested numerical model, demonstrating that these parameters are critical for accurate defect visualization.
  • TTT is Robust to Void Size and Location: The proposed TTT approach showed good generalizability, effectively imaging voids of varying sizes and at different positions within the CFST cross-section, although imaging quality slightly degraded for voids near the corners due to reduced ray coverage.
  • Experimental Validation Confirms Efficacy: The experimental results closely matched the numerical findings, confirming the practical applicability and accuracy of the TTT method for real-world CFST defect detection.

Key Results

  • Numerical Simulations: The optimized TTT method successfully visualized concrete core void defects in simulated CFST models, accurately identifying their positions and sizes. Parameter analysis revealed that the second-order difference operator for Wm, a weighting factor (λ) of 5, and a grid size (d) of 1 cm provided the best imaging results.
  • Experimental Verification: Six CFST specimens with mimicked concrete core voids were tested. The imaging results from the experimental data demonstrated that the proposed approach could effectively visualize the sizes and locations of these voids with acceptable accuracy, with the outer steel tubes also being accurately imaged. The size of the imaged voids was generally slightly larger than their actual counterparts.

Limitations and Future Work

The authors acknowledge several limitations:

  • Ray Coverage: The imaging quality is heavily dependent on the coverage rate of the curved rays across the CFST cross-section. Areas with sparse ray coverage may lead to less accurate imaging.
  • Travel Time Extraction Accuracy: The accuracy of the proposed method relies on the precise extraction of travel times from the stress wave response signals.
  • Preliminary Study: The current study focuses on rectangular CFST members and a simplified mesoscale heterogeneous model for the concrete core.

Future work suggested includes:

  • Combining with Vibration-Based Methods: Integrating TTT with vibration-based inspection technologies could enhance void defect detection capabilities.
  • Extending to Other Geometries: Investigating the feasibility of the approach for two-dimensional circular CFST members and three-dimensional circular or rectangular CFST members.
  • Advanced Material Modeling: Exploring the defect visualization of CFST members where the concrete core is modeled as a mesoscale heterogeneous material.

Summary updated

BibTeX
@misc{zheng2024concrete,
  author = {Wenting Zheng and Bin Xu and Zongjun Xia and Jiang Wang and Jingliang Liu and Yudi Yao and Yifei Wang},
  title = {{A Concrete Core Void Imaging Approach and Parameter Analysis of Concrete-Filled Steel Tube Members Using Travel Time Tomography: Multi-Physics Simulations and Experimental Studies}},
  year = {2024},
  doi = {10.3390/s24082503},
}
Figure