Study On The Influence Of Surface Characteristics Of Metal Pipeline On LIBS Spectrum And Failure Prediction Model

With the continuous development of modern society,the power demand of the society is also increasing,and high-performance generator sets are put into operation to meet the power demand of the society.The complex structure and system of large generator sets make the operation and maintenance of thermal power plants more and more serious.Boiler is an important part in power plant,of which the superheater tubes,reheater tube,water wall tube economizer tube and high pressure pipeline is the main component inside the boiler under the environment of high temperature and high pressure,metallographic structure and mechanical properties of heat resistant steel pipe will change over time,produce metal material aging failure phenomenon,has direct influence to the safe operation of boiler equipment.The statistical results show that the explosion and leakage of high temperature pressure tube accounts for about 50% of the unplanned shutdown of thermal power plant units.Therefore,it is of great practical significance to accurately predict the failure of high-temperature and high-pressure pipelines in boilers to ensure the safe operation of thermal power plants.Traditional failure detection of heating surface of metal pipeline requires destructive cutting sampling and off-line analysis of the heating surface of the pipeline during off-line maintenance of the power plant.Although the cutting and sampling detection method can obtain the detailed state information of the pipeline,it will cause damage to the inspection section,requiring welding repair.On the one hand,it will affect the service performance of the pipeline,on the other hand,it will delay the operation of the equipment and cause losses.However,the current nondestructive testing method,which is aimed at the existing defects of components,cannot predict and analyze the failure trend.Because of the limitation of time and operation,it is necessary to develop a fast and nondestructive testing technology to effectively test the heating surface of boiler pipelines,and to effectively detect the failure state before the occurrence of pipeline defects.Laser Induced Breakdown Spectroscopy(Laser Induced Breakdown Spectroscopy,LIBS)is a kind of testing technology could meet the above requirements,it is a new type of atomic emission spectrum technology,its principle is pulsed Laser ablation under test object surface plasma,plasma contains the spectral data reflect the composition of the object under test information,by analyzing the spectrum of the object data can be implemented to test the composition and structure information of qualitative analysis and quantitative analysis.In this paper,the LIBS technology applied in the actual operation of coal-fired power plants of T91 tube aging level prediction,to explore the laser effect to the surface of different energy size unprocessed T91 actual pipeline difference spectrum data,research on metal pipeline in thermal power plants under the environment of high temperature and high pressure on the surface of the metal pipe sediment influence on LIBS measurement,and the applications of machine learning algorithms to failure prediction model,the SVM metal aging level prediction model is established.By comparing the surface polishing treatment samples and surface untreated samples of Fe,Cr and other representatives pipeline important properties of matrix and alloy element line spectral line intensity with the change of pulse number,found that the influence of the metal pipe surface features inside the laser pulse is more obvious in the top 100,from 100 to 200 of the laser pulse,grinding and polishing the lines tend to be the same of the sample.Under this phenomenon,the spectral data of 100 to 200 laser pulse segments of surface untreated samples were input into the SVM aging grade model,and a good prediction result similar to that of the polished samples was obtained,which provided a foundation for the application of LIBS technology to the failure detection of heating surface of metal pipelines in thermal power plants.Finally,the paper reviews and summarizes the current research work on the application of LIBS technology to the prediction of the aging grade of actual operating pipelines in thermal power plants,puts forward the plan for further research work in the future.