Detection And Evaluation Of Tank Bottom Corrosion Based On Acoustic Emission Technology

With the development of petrochemical industry,China's oil demand and output are increasing,and the demand for oil reserves is also rising.As the main storage container of petroleum,the number and scale of atmospheric pressure metal storage tank are increasing quickly as well.At the same time,the safety of storage tank is getting more and more attention.As an important part of tank safety evaluation,the evaluation of tank bottom corrosion has always been the focus of attention.Compared with the traditional open tank detection method,acoustic emission detection technology,as a dynamic,online and overall detection method,has the advantages of short detection time and low economic loss.In this thesis,with the help of Tianjin Research Institute for Water Transport Engineering,M.O.T,aiming at the application of acoustic emission testing technology in tank safety evaluation,the relationship between acoustic emission parameters and tank corrosion evaluation is studied.Based on the actual situation of the bottom of the in-service metal tank,the metal corrosion experiment is designed and completed.Taking the corrosion rate as the research object,the corrosion behavior of Q235 B steel in 0.1g/l chloride solution,tap water and naphtha ponding was tested and analyzed.In order to study the acoustic emission information in the process of corrosion,an acoustic emission signal acquisition device is built based on the principle of acoustic emission detection,and the acoustic emission signal acquisition experiment is carried out in the laboratory.The continuous acoustic signal acquisition of the simulated tank is carried out under the same experimental conditions as the metal corrosion experiment,the characteristic parameter information of acoustic emission signal is obtained by processing and calculating the collected acoustic signals.Then,Pearson correlation analysis was carried out between the corrosion rate and the acoustic emission signal characteristic parameters of impact,count and energy.The results show that the corrosion rate of metal is highly correlated with the three characteristic parameters of acoustic emission signal.The acoustic emission testing and evaluation method of metal storage tank is analyzed,the acoustic emission characteristic parameters that can represent the corrosion of tank bottom plate are determined,the dangerous chemicals that need to be supervised are analyzed and sorted out,ae well as whose deposit water is highly corrosive.Combined with the information of tank itself,the evaluation method of tank floor corrosion is established.After that,based on RBF neural network and BP neural network,the evaluation model of tank bottom corrosion is established.The network was trained and tested with 40 groups of in-service storage tank information and on-site detection data,including 25 groups of training samples and 15 groups of test samples.The test results show that the accuracy of RBF neural network model is73.33%,the accuracy of BP neural network model is 93.33%.After that,three different in-service tanks were tested by acoustic emission testing,and the more accurate BP network corrosion evaluation model is used to evaluate the tank,then the output of the model is compared with the evaluation results of acoustic emission test experts.The results show that the evaluation model of tank bottom corrosion based on acoustic emission detection constructed by BP neural network has high accuracy,the model is reliable to evaluate the corrosion of tank bottom.