Abstract
Predicting atmospheric corrosion rate of metals has important practical significance. The linear model was often used to predict corrosion rate of metals but this provides low precision. This is because the influence of environmental factors to the corrosion rate is not linear. This paper presents the result of building up one artificial neural network using Matlab software, which can predict atmospheric corrosion rate of CT3 carbon steel in Vietnam tropical climate with input and output data of 26 test areas in 15 provinces of the North, Central and South of Vietnam. This neural network has 05 input parameters including the temperature, humidity, time of wetness - TOW, Cl- concentration, number of hours of rain (average of one year). The obtained results of corrosion rate of CT3 steel at 03 locations (Yen Bai, Da Nang, Thai Binh) by using the neural network show much higher accuracy (relative error < 30%) compared with the linear regression models.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
