Affiliations: [a] College of Science, Xi'an University of Posts and Telecommunications, Xi'an, Shaanxi, China | [b] State Key Laboratory for Strength and Vibration of Mechanical Structures, Xi'an Jiaotong University, Xi'an, Shaanxi, China
Correspondence:
[*]
Corresponding author: Li Wang, College of Science, Xi'an University of Posts and Telecommunications, Xi'an, 710121, Shaanxi, China. E-mail:wl_3691@163.comorwangli_06@stu.xjtu.edu.cn
Abstract: Quantitative evaluation of crack profiles using Eddy
Current Testing (ECT) signals often underestimates depth of a Stress
Corrosion Crack (SCC) because of the crack complicated property, especially
the conductivity property. In this paper, to improve reconstruction
precision of crack depth, an SCC numerical model considering crack
conductivity distribution is adopted. To solve the high-dimension
optimization problem with multiple conductivity parameters, a hybrid
inversion strategy combined Particle Swarm Optimization (PSO) algorithm with
Powell direction acceleration method is proposed. Owing to overcoming local
optimum and slow search speed in a high-dimension solution space of PSO
algorithm, and solving the difficulty of initial value selection for Powell
direction acceleration method, the hybrid inversion strategy can improve SCC
reconstruction accuracy and numerical stability. Results from simulated and
measured ECT signals of several artificial SCCs demonstrate that the
proposed strategy is effective to improve reconstruction precision of crack
depth.
Keywords: Eddy current testing, stress corrosion cracks, reconstruction, particle swarm optimization, high-dimension optimization, Powell direction acceleration method
