Research On The Control Of Corrosion Resistance Of Cr-Containing Low-alloy Steel Based On Corrosion Big Data Technology

A major technical problem affecting the development of low-alloy steel is the understanding of its corrosion resistance mechanism and the evaluation of its corrosion resistance.In the development of traditional low-alloy steel,the evaluation of its corrosion resistance requires a large amount of long-period outdoor corrosion test data to support it.It takes time and effort,sometimes as long as several years to get a corrosion data.In this paper,the latest research results of material corrosion big data theory,combined with corrosion tests and evaluation methods were used.First of all,low alloy steels with different Cr content are used as the object to verify the reliability of the big data technology.Afterwards,the corrosion resistance of Cr-containing low-alloy structural steel was adjusted by the addition of Mo and Sn elements.Then,the microstructure,grain size and cathode phase ratio of the steel was further adjusted;Finally,through artificial neural network,support vector machine,random forest and other data mining methods,the cross-scale micro and macro corrosion data is systematically explained.The results show that the corrosion big data technology is suitable for screening microalloying elements such as Cr,Sn,and Mo on the corrosion resistance of low-alloy structural steel.The corrosion resistance of low-alloy steel after Cr microalloying is significantly improved;in addition,the corrosion resistance of Cr-containing steel modified by 0.1%Mo has been improved to a certain extent,which improves the corrosion resistance of Cr-containing steel;the addition of 0.2%Sn can also significantly promote the corrosion resistance.And the addition of 0.1%Sn has a certain deteriorating effect on its corrosion resistance.The influence of Cr,Sn and Mo on the corrosion resistance of low-alloy structural steel is mainly manifested in the acceleration or inhibition of uniform corrosion or pitting corrosion.The Cr element will be enriched in the inner rust layer,promote the ratio of oxides and hydroxides in the rust layer,and then promote the uniform corrosion resistance of low-alloy steel;at the same time,the hydrolysis reaction of Cr3+ will produce acidification and promote the pitting corrosion behavior of low alloy steel.The role of Mo is manifested in the inhibition of pitting corrosion behavior.In addition,the corrosion product MoO3 is unstable,and the acidification caused by hydrolysis in the rust layer will accelerate the uniform corrosion process;the role of Sn is mainly manifested in the formation of stable SnO2 oxide doped in the rust layer,increasing the stability of the rust layer.The corrosion resistance of low-alloy structural steel has a certain relationship with the grain size of the original austenite and the sub-grain size.The evaluation results of the corrosion big data technology show that as the grain size of the original austenite increases,its corrosion resistance is gradually getting worse;as the bainite laths are gradually refined,the corrosion resistance gradually becomes better.The influence of the original austenite grain size and the thickness of the bainite lath on the corrosion resistance can be attributed to the influence of the cathode and anode ratio in the material on the corrosion resistance.SKPFM proved that the grain boundary in the structure is generally the cathode phase,while the bainitic ferrite matrix is the anode phase.The martensite-austenite component and its ratio in low-alloy structural steel have a certain influence on its corrosion resistance,and its influence can be quickly identified with big data technology.Specifically,the higher the content of martensite-austenite components in the steel,the worse the corrosion resistance of the steel.The martensite-austenite component exists as the cathode phase in steel due to its high potential during the corrosion process,while the bainitic ferrite matrix exists as the anode phase due to its low potential,thus forming a corrosion micro-couple;During the corrosion process,the bainite ferrite matrix will preferentially dissolve under the action of the micro galvanic couple.The more martensite-austenite components,the worse the corrosion resistance.Machine learning methods such as artificial neural network models,support vector machine models,random forest models and deep learning models are suitable for mining macro-corrosion big data such as atmospheric environmental factors and micro-corrosion big data such as material composition and organizational structure factors on corrosion resistance.It can establish cross-scale impact mechanism research based on macro-micro big data technology.At the same time,the deep learning model can be used to dig and analyze the internal corrosion laws of structural factors,and can dynamically predict the corrosion process of low-alloy steel such as changes in the structure and temperature and humidity parameters.