In recent years, the sensitivity of the particle detection sensor has been greatly improved. However, the research on multi-particle distinction detection is less. This paper adopts the method of frequency conversion excitation to explore the response relationship between the inductive effect and eddy current effect of abrasive particles and the excitation frequency. Research shows that the inductance signal of iron particles does not change significantly with the excitation frequency, and the smaller the particles, the smaller the influence of the excitation frequency on the output signal. When the excitation frequency is 0.1MHz, the change in inductance generated by the copper particles passing through the sensor is less than the lower detection limit of the instrument, and when the excitation frequency is 2MHz, the signal changes significantly. Therefore, this paper proposes a frequency conversion detection method to distinguish between aggregated metal particles of different materials. The method changes the intensity of the eddy current effect of copper particles by frequency conversion for individual iron particles, individual copper particles, and aggregated metal particles of different materials. The proposed method avoids the impact on the accuracy of the sensor caused by the simultaneous passage of particles of different materials and provides a new reference method for subsequent accurate detection.

Lubricant is a critical component of a ship’s power plant, and analyzing the information on the change of lubricant viscosity and the wear particle it carries is of great significance to realize the condition monitoring of the power plant. In this paper, we propose and design an inductive sensor capable of simultaneously detecting the viscosity, particle diameter, and velocity of lubricants. COMSOL was used to investigate the magnetic field, particle flow, and particle velocity characteristics. The inductance signal amplitude and velocity values of spherical iron particles of eight diameters passing through four levels of viscosity lubricants were collected. Based on this data, we plotted the particle diameter-inductance signal amplitude curves, viscosity-velocity (for different diameters), and diameter-velocity (for different viscosities) curves. According to the experiment, when the two unknown iron particles passed through the standard oil sample of 50 mm ² /s (20°C), the corresponding particle diameter can be obtained from the particle diameter-inductance amplitude curves as 300 μm and 278 μm, respectively. For the calibrated particle with a diameter of 300 μm, the viscosity-velocity (for different diameters) curve yielded the viscosity value of 53.382mm ² /s with an error of 6.76%. The estimated viscosity value of 54.73 mm ² /s was obtained by diameter-velocity (for different viscosity) curve for the non-calibrated particle of 278 μm, with an error of 9.46%. This detection technology provides a new research idea for marine lubricant testing.

The cleanliness of lubricating oil plays a key role in determining the operational health of mechanical systems, serving as a critical metric that delineates the extent of equipment wear. In this study, we present a magnetic-core-type planar coil particle detection sensor. The detection accuracy and detection limit are improved by optimizing the magnetic field inside the sensor. The optimization of the magnetic field is achieved through the finite element simulation analysis of the coil and the magnetic core. First, the finite element simulation software COMSOL 6.0 is used to model the sensor in three dimensions (3D). Then, we study the distribution of the magnetic field under different coil radii, core conductivity levels, and other parameters. We obtain the sensor structure after optimizing the magnetic field. The sensor is made using experimental methods, and the iron particles and copper particles are detected. The results show that the lower limit of detection of iron particles can reach 46 μm, and the lower limit of detection of copper particles can reach 110 μm.