Search

With the development of nanotechnology, the preparation method of surface-enhanced Raman scattering (SERS) substrate has made significant progress, which promotes the application of SERS in food safety, environmental protection, and medical testing. We reported on the fabrication of silver-coated magnetic Fe3O4@phenolic resin (Fe3O4@RF) core–shell particles by a facile solvothermal method, followed by deposition of high-density Ag nanoparticles onto the phenolic resin surfaces through an in-situ reduction process. Analysis of Raman spectroscopy was performed by laser Raman spectrometer. The as-synthesized Ag-coated Fe3O4@RF particles as sensitive SERS substrates showed outstanding SERS performances toward probe molecules of 4-aminothiophenolas with a detection limit of 1 × 10–9 mol L–1. Moreover, the enhancement factor of the substrate was calculated to be about 2.86 × 105. Importantly, profiting from the excellent magnetism, the signal attenuated slightly only after three cycles measurements of SERS based on such composite substrates, indicating that Fe3O4@RF@Ag magnetic composite microspheres possessed the outstanding repeatability. On the basis of these results, it is believed that such Fe3O4@RF@Ag substrates are promising for the reusable detection of toxic molecules in practical applications.

Ultrasonic method is promising in oil film thickness measurement, while the commonly used commercial transducers are bulky and then sometimes hinder the application of this technology in field. Using the piezoelectric element is an alternative method and has the advantage of being convenient to be fixed. In this work, the key points of the use of the piezoelectric element are summarized. A calibration rig is set up to test the performance of oil film thickness measurement using piezoelectric element. The results show that using the piezoelectric element can measure the oil film thickness effectively and accurately.

Fukushima nuclear power plant accident caused huge losses and pollution and it showed that the reactor coolant pump is very important in a nuclear power plant. Therefore, to keep the safety and reliability, the condition of the coolant pump needs to be online condition monitored and fault analyzed. In this paper, condition monitoring and analysis based on support vector machine (SVM) is proposed. This method is just to aim at the small sample studies such as reactor coolant pump. Both experiment data and field data are analyzed. In order to eliminate the noise and useless frequency, these data are disposed through a multi-band FIR filter. After that, a fault feature selection method based on principal component analysis is proposed. The related variable quantity is changed into unrelated variable quantity, and the dimension is descended. Then the SVM method is used to separate different fault characteristics. Firstly, this method is used as a two-kind classifier to separate each two different running conditions. Then the SVM is used as a multiple classifier to separate all of the different condition types. The SVM could separate these conditions successfully. After that, software based on SVM was designed for reactor coolant pump condition analysis. This software is installed on the reactor plant control system of Qinshan nuclear power plant in China. It could monitor the online data and find the pump mechanical fault automatically.