Predictions of loading capacity and stiffness of corroded circular steel tubes based on artificial neural networks

Abstract

Circular steel tubes (CSTs) are commonly used in building structures and mechanical equipment. Due to the deterioration caused by the surrounding environment, corrosion is inevitable for CSTs. Corrosion can decrease the loading capacity (including compression, torsional, and bending capacities) and the corresponding stiffness of CSTs. The change rule of loading capacity and stiffness along with mass loss ratio is initially revealed through stochastic numerical analysis. The artificial neural network is then utilized to predict the residual loading capacity and corresponding stiffness to quantify the influence of corrosion. Corroded thickness and mass loss ratio are regarded as the input variables of the artificial neural network. The output results include six mechanical indices of corroded CSTs. The accuracy of predicted results is compared with the results derived by stochastic numerical analysis. Results indicated that the artificial neural network can accurately predict the mean value of the reduction factor. To capture the random characteristic of the reduction factor, random corroded thickness is utilized as an input variable. Results also indicated that the artificial neural network can be utilized to predict the loading capacity and stiffness of CSTs with high accuracy, and only the corroded thickness and mass loss ratio are needed for the input variable.

Keywords:

• Circular steel tube
• random pitting corrosion
• loading capacity
• loading stiffness
• artificial neural network
• mass loss ratio

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Some or all data, models, or code generated or used during the study are available from the corresponding author by request.

Additional information

Funding

This work was financially supported by the project funded by the China Postdoctoral Science Foundation (Grant No. 2021M690033) and Jiangsu Planned Projects for Postdoctoral Research Funds (2020Z356).