Search

The global shear capacity of steel–concrete composite downstand cellular beams with precast hollow-core units is an important calculation as it affects the span-to-depth ratios and the amount of material used, hence affecting the embodied CO2 calculation when designers are producing floor grids. This paper presents a reliable tool that can be used by designers to alter and optimise grip options during the preliminary design stages, without the need to run onerous calculations. The global shear capacity prediction formula is developed using five machine learning models. First, a finite element model database is developed. The influence of the opening diameter, web opening spacing, tee-section height, concrete topping thickness, interaction degree, and the number of shear studs above the web opening are investigated. Reliability analysis is conducted to assess the design method and propose new partial safety factors. The Catboost regressor algorithm presented better accuracy compared to the other algorithms. An equation to predict the shear capacity of composite cellular beams with hollow-core units is proposed using gene expression programming. In general, the partial safety factor for resistance, according to the reliability analysis, varied between 1.25 and 1.26. [ABSTRACT FROM AUTHOR]

The use of circular hollow sections (CHS) has increased in recent years owing to its excellent mechanical behaviour including axial compression and torsional resistance as well as its aesthetic appearance. They are popular in a wide range of structural members, including beams, columns, trusses and arches. The behaviour of hot‐finished CHS beam‐columns made from normal‐ and high‐strength steels is the main focus of this article. A particular attention is given to predict the ultimate buckling resistance of CHS beam‐columns using the recent advancement of the artificial neural network (ANN). Finite element (FE) models were established and validated to generate an extensive parametric study. The ANN model is trained and validated using a total of 3439 data points collected from the generated FE models and experimental tests available in the literature. A comprehensive comparative analysis with the design rules in Eurocode 3 is conducted to evaluate the performance of the developed ANN model. It is shown that the proposed ANN‐based design formula provides a reliable means for predicting the buckling resistance of the CHS beam‐columns. This formula can be easily implemented in any programming software, providing an excellent basis for engineers and designers to predict the buckling resistance of the CHS beam–columns with a straightforward procedure in an efficient and sustainable manner with least computational time. [ABSTRACT FROM AUTHOR]

This paper aims to propose an Artificial Neural Network (ANN) model that predicts accurately web-post buckling resistance and failure mode of steel beams with elliptically-based openings. A total of 4,344 and 5,400 geometrical models, were developed by finite element method (FEM) and used to train, validate and test the ANN model for the web-post resistance and failure mode classification, respectively. It was concluded that five neurons model were sufficient to predict the web-post buckling resistance and the failure mode with high level of accuracy. The height and the web thickness of the beams had positive impact of the capacity while the web openings height, width and radius of the elliptically-based web opening were the geometric parameters that had negative impact of the capacity. At last, an ANN-based formula was proposed and compared with previous analytical model for web-post buckling resistance of elliptically-based openings, which considered the web-post as a truss model. The ANN-based formula showed high accuracy, since the Regression (R2), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Average (FEM/Predicted), Standard Deviation and Variation were 0.9989, 26.03 kN, 15.0 kN, 1.00, 4% and 0.12%, respectively. Consequently, the ANN-based formula for web-post buckling resistance of steel beams with elliptically-based openings can be safely adopted for design purposes. • The web-post buckling resistance of elliptically-based web openings beams is obtained using ANN. • The failure mode is predicted using ANN. • ANN-based formula provided good predictions of web-post shear buckling and failure mode. • The proposed expression can be utilized in practice for WPB design of beams with elliptically-based web openings. [ABSTRACT FROM AUTHOR]

The present paper aims to develop an Artificial Neural Network (ANN) formula to predict the lateral-torsional buckling (LTB) resistance of slender steel cellular beams. A finite element model is developed and validated through experimental tests followed by a parametric study. 768 models are employed to train the ANN formula. The results are compared with the analytical models as well as the equation predicted by ANN. It was concluded that the ANN model with seven neurons can accurately predict the LTB resistance of cellular beams as well as the LTB combined with web-post buckling or web distortional buckling modes. Hence, the ANN-based formula can be adopted as design tool. • Finite element results are employed to train the ANN. • Neutral and destabilising effects of loading are also modelled. • ANN-based formula for lateral–torsional buckling resistance of cellular beams is developed • ANN predictions are more in-line with FE results that those obtained analytically. [ABSTRACT FROM AUTHOR]

Stainless steel has many advantages when used in structures, however, the initial cost is high. Hence, it is essential to develop reliable and accurate design methods that can optimize the material. As novel, reliable soft computation methods, machine learning provided more accurate predictions than analytical formulae and solved highly complex problems. The present study aims to develop machine learning models to predict the cross-section resistance of circular hollow section stainless steel stub column. A parametric study is conducted by varying the diameter, thickness, length, and mechanical properties of the column. This database is used to train, validate, and test machine learning models, Artificial Neural Network (ANN), Decision Trees for Regression (DTR), Gene Expression Programming (GEP) and Support Vector Machine Regression (SVMR). Thereafter, results are compared with finite element models and Eurocode 3 (EC3) to assess their accuracy. It was concluded that the EC3 models provided conservative predictions with an average Predicted-to-Actual ratio of 0.698 and Root Mean Square Error (RMSE) of 437.3. The machine learning models presented the highest level of accuracy. However, the SVMR model based on RBF kernel presented a better performance than the ANN, GEP and DTR machine learning models, and RMSE value for SVMR, ANN, GEP and DTR is 22.6, 31.6, 152.84 and 29.07, respectively. The GEP leads to the lowest level of accuracy among the other three machine learning models, yet, it is more accurate than EC3. The machine learning models were implemented in a user-friendly tool, which can be used for design purposes. • The paper studies behaviour of stainless steel CHS stub columns. • ANN, SVR, GEP and DTR machine learning algorithms are developed and validated. • Machine learning-based design approach for predicting the capacity is proposed. • A comparison with EC3 is presented to assess the performance of the models. [ABSTRACT FROM AUTHOR]