Experimental Study On Uv—visible Sepctra Of Engine Oil

Lubricating oil plays an important role in the operation of equipment. The used oil carries a certain amount of pollutants and metal wear particles.Evaluating the lubricating oil reasonably can effectively reduce mechanical wear and prolong the service life of the machine. Therefore it is necessary to monitorthe lubricating oil.Spectrum analysis technology is an important research direction in oil monitoring.This article comprehensively studies the existing research results. In view of the problems and the insufficiencies existing in the spectrum monitoring of oil, the paper established model both between the pollutants concentration and the reflectance spectrum and the transmission spectrum in ultraviolet visible light to analysis the degree of oil pollution. It lays the foundation for using the optical method to measure the oil pollution and has a positive reference value. The main contents are as follows:This article focuses on the physical and chemical properties of lubricating oil and designs the experimental simulating the process of continuous change of the pollutant concentration in the using lubricating oil and measures its reflectance spectrum and transmission spectrum.The paper selects the best pretreatment method to analyze the transmission spectra of lubricating oil.Then extracting the characteristic wavelengths by successive projections algorithm.Ultimately the experimental results set up the relational mode between pollutant concentration and light transmittance of lubricating oil combining with partial least squares. This model can well forecast validation sets. The study provides a new method for lubricating oil pollution degree analysisAccording to the principle of reflection spectrum of the material,this article uses correlation coefficient and PLS projection analysis combining with threshold setting method respectively to extract the sensitive wavelengths of different pollution degree of lubricating oil. Finally this paper established the partial least squares models between reflectivity and the concentration. The models was used to verify samples and the result show that the models has good prediction ability. The results of the study provides experimental basis for using reflection spectrum to detect lubricating oil pollution.