Aug 2026· Applied Sciences· 0 citations· 67 references
TL;DR
TRA-DNN employs the DNN model to perform SRA via Crude Monte Carlo Simulation, yielding results that are in high agreement with conventional SFEA, yet with significantly reduced computational time (1,388,250 times faster).
Abstract
The aim of structural reliability assessment (SRA) is to guarantee the safety, durability, and performance of structures; however, traditional methods like stochastic finite element analysis (SFEA) can be computationally prohibitive to use in practical situations. This paper introduces a novel framework for SRA utilizing deep neural networks (DNNs) implemented in an open-source program called TRA-DNN, replacing the resource-intensive finite element (FE) analysis with a DNN model. The DNN is trained using 6874 FE column models, including factors like geometric imperfections, resulting in a training database with 419,314 data records. It accurately predicts axial load-deformation curves for corroded steel CHS columns, enabling the determination of the ultimate capacities for columns with varying properties. The model’s accuracy is confirmed through rigorous quantitative and qualitative validation, including various failure modes. TRA-DNN employs the DNN model to perform SRA via Crude Monte Carlo Simulation (MCS), yielding results that are in high agreement with conventional SFEA, yet with significantly reduced computational time (1,388,250 times faster). In addition, TRA-DNN can be used to estimate the service life of CHS columns considering both corrosion propagation and load increase with time. Future research can utilize TRA-DNN to optimize column design and maintenance to minimize both risk and cost.
Box culverts are critical components of transportation infrastructure, yet their load-carrying capacity is often compromised by reinforcement corrosion and geometric deterioration. This study proposes an integrated finite element (FEM) and machine learning (ML) framework to evaluate and predict the ultimate capacity of corroded reinforced concrete box culverts strengthened with shape memory alloy (SMA). Severe corrosion was simulated by reducing the slab thickness by 20 mm, and SMA was subsequently applied to the corroded culvert in different strengthening configurations to assess its rehabilitation effectiveness. Case-2 showed the best performance in increasing the capacity of the corroded culvert. A total of 314 FEM-based generated data points were used to train and test five regression models: GB, ANN, SVM, RF, and KNN. The performance of the model was evaluated in terms of R2, MAE, RMSE, and MAPE. The SHAP and Partial Dependence Plot (PDP) analysis was conducted for model interpretability and understanding the physical relevance. The analysis suggests that SMA helps in restoring and improving the carrying capacity and resistance to deformation of the corroded culvert. Among the predictive models, GB achieved the highest accuracy, followed by ANN and SVM. Interpretability analyses consistently identified culvert width as the dominant parameter, followed by depth and length, while concrete strength showed a moderate positive influence. The proposed FEM–ML–XAI framework offers a reliable and interpretable tool for assessing and rehabilitating corrosion-damaged culverts using SMA.
K. Majid, F. Dang, Abdulmoez Al Ismaeel et al.· Buildings· 0 citations
Performance estimation of composite structural members like slabs, beams and columns is crucial for structural integrity, durability and economy. However, conventional methods including testing, analytical formulations and finite element analysis are known to be inefficient, computation-intensive and limited in terms of capturing the highly non-linear behaviour of composite structures. Recently, machine learning (ML) has been considered as a fast and effective data-driven technique for modelling structural responses based on experimental, numerical and monitoring data. In this context, the current work presents a comprehensive review of the recent advancements in the application of ML techniques to composite structural members. Various methods used by researchers including shallow learning, deep learning and hybrid methods are summarised, and their capability in predicting the load-bearing capacity, deflection, fatigue response, cracking behaviour and fire resistance is investigated in detail. Contrary to other previous works that have reviewed either specific components or specific ML algorithms, this work provides a comparative evaluation of the existing ML-based methodologies in the context of composite slabs, beams and columns.
Iyappan G.R., S. P., Elango D· Proceedings of the Instituti...· 0 citations
Rectangular concrete-filled steel tube (RCFST) columns are widely adopted as primary load-bearing components in engineering structures, and making reliable estimation of their axial compressive capacity crucial to structural design and safety assessment. However, existing theoretical and design equations generally rely on a series of simplifying assumptions and empirical modification coefficients, which limit their prediction accuracy and applicability. In this context, a data-driven strategy is proposed in which capacity prediction and model interpretation are integrated within a unified procedure. The proposed approach establishes the nonlinear mapping between high-dimensional input features, including geometric and material properties, and axial compressive capacity, while employing the SHapley Additive exPlanations (SHAP) method to quantify the contributions of critical input parameters to the model predictions. An experimental database comprising 745 axial compression tests on RCFST columns is established through an extensive literature survey. Prior to training, correlations within the candidate inputs are examined using Pearson coefficients. An SMA-LSSVM–ANN hybrid prediction model is then developed, in which SMA iteratively searches for the optimal combination of the LSSVM-ANN hyperparameters using mean squared error as the fitness function. Comparisons with several benchmark machine learning (ML) models and existing design-code formulations validate the predictive performance of the proposed model, which achieves an R2 of 0.9571 and an RMSE of 302.84 kN on the testing dataset. Furthermore, SHAP analysis is employed to interpret the proposed model from global and feature-dependency perspectives. Global SHAP analysis identifies concrete compressive strength and cross-sectional dimensions as the dominant features, while feature-dependence analysis reveals generally positive effects of cross-sectional dimensions and material strengths and a negative effect of column length on the predicted capacity.
Jun-Wei Xing, Ya-Nan Zhang, B. Qiu et al.· Buildings· 0 citations
Results indicate that support vector regression (SVR) provides the most consistent overall performance across all regimes and offers a strong balance between accuracy and computational efficiency, and a Bayesian neural network (BNN) achieves competitive predictive performance while additionally enabling uncertainty estimation.
Muhammad Bilal Jan, Zengchao Wu, Mengyu Chai· Metals· 0 citations
Type IV composite hydrogen storage vessels must ensure structural safety under high internal pressure, and determining the optimal laminate configuration typically requires repetitive finite element analysis (FEA), leading to significant computational cost during the early design stage. To address this limitation, this study proposes a machine learning (ML)-based design-assistance framework for predicting the minimum laminate configuration required to satisfy structural safety based on the Tsai–Wu failure criterion. Three geometric design variables—liner radius, liner length, and polar hole radius—were used as inputs, and a dataset was generated using ANSYS Workbench-based FEA. Five ML models—SVR, GPR, RF, XGBoost, and MLP—were applied and evaluated using leave-one-out cross-validation (LOOCV) and an independent test set. All models achieved high predictive accuracy, with LOOCV R2 values ranging from 0.9801 to 0.9907 and test-set R2 values from 0.9866 to 0.9949, while maintaining MAPE below 6%. The consistent performance between LOOCV and the independent test set indicates stable predictive behavior. Learning curve analysis demonstrated stable convergence and a small performance gap between training and validation, suggesting stable predictive behavior within the considered design space. Feature importance analysis using the RF model identified the liner radius as the dominant parameter, while the influence of liner length was relatively minor. These results demonstrate that the proposed ML-based surrogate model can serve as an efficient tool for rapid design decision-making, reducing reliance on repetitive FEA in the early design stage.
Jisoo An, Hyeongmin Yoo· Applied Sciences· 0 citations
This paper proposes a data-driven strategy leveraging the XGBoost (Extreme Gradient Boosting) algorithm to accurately predict the required reinforcement and the deflection of reinforced concrete (RC) slabs in real-world high-rise buildings. To achieve the objective, two distinct predictive models with tailored input parameters are developed. For estimating the required reinforcement ratio, the model incorporates the span direction, location, span length, slab thickness, beam height, and the design strengths of both concrete and steel. Meanwhile, for assessing the maximum deflection of the slab, the input parameters are streamlined to the span direction, span length, concrete strength grade, slab thickness, and the reinforcement area. The model evaluation results demonstrate robust predictive performance as a fast, finite element analysis-based surrogate tool, achieving a coefficient of determination (R2) of 0.971 for reinforcement ratio estimation and 0.974 for deflection prediction when validated against code-compliant finite element analysis data from a real-world high-rise building. Furthermore, the model’s cross-project transferability is successfully verified on an independent structural project within the same design-code domain. Overall, this paper demonstrates that the proposed dual-architecture XGBoost framework effectively approximates time-consuming finite element calculations, significantly accelerating preliminary design iterations while contributing to the advancement of data-driven civil engineering workflows.
Chan-Huy Mai, N. Nguyễn, Khanh-Hoang Vu et al.· CivilEng· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
