Year: 2025
Type: article
Abstract: We report a minute-scale, micrometer-resolution workflow based on focal-stack tomography (FST) under incoherent illumination for in situ 3D tracking of electrical trees in epoxy. A synchronized-trigger scheme acquires focal-plane images of the electrical tree under an AC field at 1 slice s−1, while an adaptive algorithm adjusts the z-scan range to fully enclose the tree volume. Accelerated Speeded Up Robust Feature -based rigid registration, pixel-wise Laplacian depth selection, and a 3D k-nearest-neighbor filter yield a high-fidelity point cloud that is distilled into a weighted skeleton graph (WSG). Real-time graph analytics, such as path length, growth rate, tortuosity, and fractal dimension, expose axial branching that 2D projections cannot resolve. Quantitative validation shows that the 2D micrographs underestimate branch length by ≈10%, underreport the true 3D growth rate, and miss nearly 20% of the tree's fractal complexity. The FST–WSG pipeline, thus, furnishes the first non-invasive, graph-based fingerprint of electrical-tree evolution under practical high-voltage stress, providing a powerful method for studying polymer-insulation failure mechanisms.
Source: Journal of Applied Physics
DOI: 10.1063/5.0284906
Institution: Xi'an Jiaotong University
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,500
Award: G1067615827
OpenAlex ID: W4413341257