22 July 2020 in IEEE Transactions on Instrumentation and Measurement
Silicone rubber insulators (SRIs) play an important role in high-voltage power transmission lines. Given that SRIs critically threatens the safe operation of the power grid, quantitatively assessing the aging status of SRIs in situ is crucial. In this paper, a portable unilateral magnetic resonance (UMR) sensor with U-shaped magnet structure and its measuring circuit were designed for the nondestructive detection of the aging status of SRIs in situ. First, the optimization efficiency of the magnet structure was improved through a uniform design method. Then, the planar radio frequency (RF) coil was designed by the finite element method. Furthermore, the demagnetization effect of eddy currents induced in metal materials on the RF magnetic field was analyzed. We measured the SRI sheds at different service times and obtained the corresponding 1H transverse relaxation decay curves by importing Carr-Purcell-Meiboom-Gill (CPMG) sequence. Finally, the transverse relaxation decay curves were post-processed using inverse Laplace transformation, and the 1-D spectrum distributions of transverse relaxation time (T2) were obtained. Results reveal that the T2 of the surface of the SRI shed decreases as the service time of the SRI increases. Additionally, the difference in T2 between the surface and interior of the SRI shed increases. The comparison experiments demonstrate that the U-shaped sensor designed in this paper is superior to the curved sensor previously designed by our team. The effect of temperature variation on SRIs aging assessment was taken into account at last.
10 June 2022 in IEEE Transactions on Instrumentation and Measurement
We report a low-field (160 mT) and lightweight (2.45 kg) Halbach nuclear magnetic resonance (NMR) sensor for non-destructive measurement of soybean moisture content (MC) on site. The sensor was composed of Halbach magnets and a solenoid radio frequency (RF) coil. A uniform design (UD) method was applied to enhance the efficiency of optimizing structural parameters of magnets by selecting a reasonable combination of parameters for simulation. An RF coil structure was then optimized with the goal of meeting the requirement of maximizing signal-to-noise ratio (SNR). Third, the portable sensor prototype with the dimensions of $\Phi 11.5$ cm $\times $ 15 cm was utilized to measure soybeans with different MCs using Carr–Purcell–Meiboom–Gill (CPMG) sequence to obtain $^{1}\text{H}$ transverse relaxation decay curves. The 1-D $T_{2}$ spectrum distributions of soybeans with different MCs were then postprocessed by applying inverse Laplace transform. The experimental results illustrate that $T_{2}$ of soybeans increases with the increase in soybean MC. The sensitivity of Halbach sensor is 1.2241 normalized peak area/%, which is increased by 28.85%. The maximum nonlinearity and resolution are 6.388% and 1% in the moisture ranging from 5% to 60%. Besides, the weight of Halbach sensor is 63.97% lighter than the previous mobile sensor.
28 April 2022 in IEEE Transactions on Instrumentation and Measurement
The sensor determines the basic performance of the nuclear magnetic resonance (NMR) logging-while-drilling (LWD) system. However, the electromagnetic calculation and structure optimization of the sensor are complicated and difficult. In this study, we proposed an effective design method for inside-out NMR LWD sensor from the perspective of electromagnetic field. First, a region of interest (ROI) was set first, the static magnetic field strength range in ROI corresponding to the ideal operating frequency bandwidth was defined as sensitive region, and the magnet system was optimized efficiently by introducing a uniform design method to maximize the sensitive region. Different simplified finite element models were established to calculate different electromagnetic parameters of the RF coil system. The inside-out NMR LWD sensor was processed. Static magnetic field measurement result shows that the proposed static magnetic field optimization method can effectively obtain the sensitive region with ideal radial position and strength and is superior to the traditional optimization method. The NMR signal detection ability was tested by importing the Carr–Purcell–Meiboom–Gill (CPMG) sequence and phase-alternate pulse sequence technology. In particular, a porous and radially layered measurement device was designed, the signal sensitive region was quantitatively measured based on changes in sample porosity, and the width and height of signal sensitive region are 50 and 90 mm, respectively. The differences between static magnetic field sensitive region and signal sensitive region were analyzed, and suggestions for improving the sensor optimization method were provided at last.
20 April 2023 in IEEE Transactions on Geoscience and Remote Sensing
Nuclear magnetic resonance (NMR) well logging is an important method of oil measurement, and the radiofrequency (RF) coil is an important component for obtaining accurate parameters of rock formation. The use of Litz wire as RF coil can help minimize the resistance and increase the signal-to-noise ratio (SNR) of NMR signals. However, the electromagnetic field computing and performance evaluation of Litz wire are complicated and difficult. In this study, we proposed an effective model for analyzing the Litz wire in inside-out NMR logging sensor. Considering the electromagnetic parameters of the coil itself, the frequency response of excitation and tuning and matching circuit, we proposed a new and accurate model for SNR analysis. We then introduced a homogenization-based finite element method to calculate the Litz wire around a ferrite core, and the electromagnetic parameters of Litz wire with different structures were obtained accurately and quickly. We compared the Litz wire and enameled wire through our proposed model. Results show that the SNR of Litz wire is remarkably higher than that of enameled wire, and the model analysis results are consistent with the measured results. Suggestions for further reducing the AC resistance of RF coil composed of Litz wire are provided.
25 June 2024 in IEEE Sensors Journal
Nuclear magnetic resonance (NMR) well logging plays an important role in oil measurement. The magnet system is a crucial component of a unilateral NMR logging sensor as it generates a high-strength static magnetic field with a strong gradient in the remote region of interest (ROI). However, current works on magnet system design are based on experiential structures, and lack a theoretical exploration of the optimal magnet system structure. What is more, during magnet system design, current works often ignore the impact of ferrite core on the radio frequency (RF) coil efficiency, which is crucial for NMR signal strength. In this study, we propose a topology optimization-based design approach for the magnet system in a unilateral NMR logging sensor, taking into account the RF coil efficiency. Firstly, we conduct sizing optimization using the genetic algorithm to obtain the regular magnet system structure under a given RF coil structure, both with and without considering coil efficiency. Subsequently, we further optimize the ferrite core structure using topology optimization and finite element method. We compare and analyze the optimization results for different coil structures and ROI ranges. We manufacture a prototype of the magnet system based on optimization results. The results demonstrate that a desired magnet field distribution is obtained, and the measured results align with the calculated results. Furthermore, we test the signal quality of the designed unilateral NMR logging sensor after tuning and matching the RF coil, and the analysis is presented at last.
19 June 2024 in IEEE Transactions on Magnetics
In continuous element density-based topology optimization (TO), global fine mesh is usually employed to obtain clear and smooth boundaries of electromagnetic devices, which is time-consuming because a substantial number of element densities need to be updated based on derivative information. To reduce the number of elements and expedite TO, this paper introduces a novel strategy via adaptive mesh refinement and mesh deformation. Initially, the rough boundaries of materials are rapidly determined by employing a global coarse mesh. Subsequently, the mesh surrounding the boundaries with intermediate densities undergoes refinement, while the mesh in other regions is coarsened automatically. In addition, the mesh deformation technique is proposed to reduce the thickness of the gray zone surrounding the boundaries by considering the variations in node densities. Consequently, clearer and thinner boundaries are attained with much fewer elements in comparison with traditional TO, and fewer iteration steps and iteration time are required. The optimization results of a unilateral nuclear magnetic resonance logging sensor verify the effectiveness of our proposed method.