Corrosion Mechanism And Prediction Model Of Hot-rolled Steel In Dynamic Atmospheric Environment

The atmospheric corrosion behavior of metallic materials is affected by their service status.The current research methods for atmospheric corrosion are mainly static exposure tests.Although this method can reflect the actual working conditions of stationary metal materials,it cannot well verify the corrosion laws under actual dynamic service conditions.Therefore,this thesis starts from the dynamic and static corrosion behavior of hot-rolled steel SPHC（Steel Plate Heat Commercial）,and analyzes the corrosion mechanism of SPHC steel in different environments by designing dynamic and static atmospheric exposure tests.Finite element numerical simulation and computer learning algorithms are used to predict the static atmospheric corrosion depth and dynamic atmospheric corrosion rate respectively.First,for the study of the corrosion behavior of SPHC steel in a static environment,a static atmospheric exposure test device was designed in accordance with the standard,and a 30-month static atmospheric exposure test was carried out.Corrosion kinetics test,SEM（scanning electron microscope）,XRD（X-ray diffraction）,electrochemical test and other methods are used to analyze the corrosion behavior.At the same time,numerical simulation of static atmospheric corrosion depth is carried out with the help of finite element software.The results show that in the static environment of Tianjin,the corrosion rate of SPHC steel increases first and then decreases.The corrosion products are mainlyγ-Fe OOH and gradually transform toα-Fe OOH and Fe₂O₃,which improves the compactness and stability of the rust layer and delays the corrosion process.The numerical simulation results show that the change trend of corrosion depth is the same as the change trend of SPHC steel weight loss under the actual situationSecondly,for the research on the corrosion behavior of SPHC steel in a dynamic environment,a dynamic atmospheric exposure test platform was built,and dynamic atmospheric exposure tests were carried out in Dongli District of Tianjin and Daoli District of Harbin.The Si O₂ and Cl^-deposited on the surface promoted the corrosion reaction under dynamic exposure in Tianjin.The content of Si O₂ deposited on the surface of the dynamically exposed samples in Harbin is relatively high.The corrosion products generated and Si O₂accumulate on the surface of the samples to form a dense rust layer,which makes the rust layer high in resistance and hinders the occurrence of corrosion reactions.Finally,for the prediction of corrosion rate in dynamic atmospheric environment,fuzzy clustering is used to extract the main related factors of vehicle driving characteristics,and GA-SVR,GA-BPNN,SVR and BPNN algorithms are used to predict the dynamic atmospheric corrosion rate.The results show that the dynamic-to-static ratioand the average speed(1(1can be used as the main correlative factors of vehicle driving characteristics.The average errors of the four algorithms are 6.84%,7.90%,7.90%,and 11.36%,respectively,and the fitting accuracy of the predicted and actual values are 0.9771,0.9610,0.9664,and 0.9333,respectively.In summary,the GA-SVR algorithm is more effective in predicting the dynamic atmospheric corrosion rate of small samples.