Prediction of mechanical properties of welded steel X70 pipeline using neural network modelling. (September 2020)

• Record Type: Journal Article
• Authors: Saoudi, Adel; Fellah, Mamoun; Hezil, Naouel; Lerari, Djahida; Khamouli, Farida; Atoui, L'hadi; Bachari, Khaldoun; Morozova, Julia; Obrosov, Aleksei; Abdul Samad, Mohammed
• Is Part Of: International journal of pressure vessels and piping. Volume 186(2020)
• Publication Date: 2020-09
• Subjects: API X70 -- Artificial neural network (ANN) -- Submerged arc welding -- Modeling -- Mechanical properties -- Chemical composition; Pressure vessels -- Periodicals; Pipe -- Periodicals; Récipients sous pression -- Périodiques; Tuyaux -- Périodiques; Pipe; Pressure vessels; Periodicals; 681.76041
• DOI: 10.1016/j.ijpvp.2020.104153
• Languages: English
• ISSNs: 0308-0161

Abstract

Abstract: An artificial neural network (ANN) model was developed to predict tensile and impact properties of a submerged arc helical welded (SAHW) pipeline steel API X70 based upon its chemical composition. Weight percent of the elements was considered as the input, while the tensile and Charpy impact properties were considered as the outputs. Scatter diagrams and two statistical parameters (absolute fraction of variance and relative error) were used to evaluate the prediction performance of the developed artificial neural network model. The predicted values were found to be in excellent agreement with the experimental data and the current model has a good learning precision and generalization (for training, validation and testing data sets). The results revealed that the developed model is very accurate and has a strong potential for capturing the interaction between the mechanical properties and chemical composition of welded high strength low alloy (HSLA) steels. Highlights: A neural network model was developed to predict multiple tensile and impact properties of welded steel API X70. Chemical composition of the test material was used as a key input to the developed model. Scatter diagrams, absolute fraction of variance R² and relative error (MRE) were used to evaluate the model performance. Results showed high levels of agreement between experimental and predicted data. The obtained model can reduce the need for time-consuming and expensive experimental testing of low alloy steels.