Analysing the degree of sensitisation in 5xxx series aluminium alloys using artificial neural networks: A tool for alloy design. (15th April 2019)

• Record Type: Journal Article
• Authors: Zhang, Ruifeng; Li, Jinfeng; Li, Qian; Qi, Yuanshen; Zeng, Zhuoran; Qiu, Yao; Chen, Xiaobo; Kairy, Shravan K.; Thomas, Sebastian; Birbilis, Nick
• Is Part Of: Corrosion science. Volume 150(2019)
• Page Start: 268
• Page End: 278
• Publication Date: 2019-04-15
• Subjects: A. Aluminium -- B. SEM -- B. TEM -- B. Modelling studies -- C. Intergranular corrosion; Corrosion and anti-corrosives -- Periodicals; 620.11223
• DOI: 10.1016/j.corsci.2019.02.003
• Languages: English
• ISSNs: 0010-938X

Abstract

Highlights: Artificial neural network (ANN) model developed to manage complex data set. The constructed ANN model could precisely predict the degree of sensitisation. The weight of input parameters of sensitised 5xxx alloys was determined. Abstract: The 5xxx series aluminium alloys are susceptible to sensitisation during service at elevated temperatures. Sensitisation refers to deleterious grain boundary precipitation resulting in rapid intergranular corrosion in moist environments. A holistic understanding of the variables that can influence the degree of sensitisation in Al-Mg-Mn alloys is presented herein, including the exploration of some custom produced 5xxx series alloys that were prepared to create a significant dataset for which an artificial neural network (ANN) could be applied. An ANN model could reveal complex interactions between various factors that influence sensitisation, with the view to designing sensitisation resistant Al-Mg-Mn alloys.