The lockbolt structure is essential in railway wagons, and a scientific lockbolt layout can ensure uniform load distribution, thereby preventing failure. However, current engineering lacks layout optimization methods that address multidimensional failure modes. This paper presents a new lockbolt structure layout optimization method based on submodel, parametric models, and a multi-strategy integrated NSGA-III (MSNSGA-III), adhering to the DVS EFB 3435-2 standard. This method simultaneously optimizes the number and spacing of lockbolts to prevent tensile, bearing, shear, and other static failure modes under specified load conditions. The proposed method was applied during the design phase of a container flatcar. Optimization results indicate that, compared to NSGA-III, this method achieves the best IGD and HV values across multiple complex test functions, demonstrating superior performance in solving complex Pareto front optimization problems. Additionally, the optimized lockbolt structure's safety margins increased by a maximum of 59.81%, passing the full vehicle strength test and significantly enhancing resistance to multidimensional failure modes. These results highlight the method's significant practical application value in addressing the optimization of railway wagon lockbolt structures under complex multidimensional failure modes.
The overall stiffness and modal frequency of the car body of a rapid box car are reduced by the design of the full-side open movable side door structure. The vibration fatigue performance of the welded structure in this car body needs to be verified. The rigid-flexible coupling model of the rapid box wagon was established first, and the model was verified by modal test data. By the application of the virtual iteration method on this model, the displacement excitation loads of this vehicle were acquired. The effectiveness of the load reverse obtaining technology was verified through the comparison between calculated data and the experimental data. Based on the rigid-flexible coupling model and the load obtained by reverse engineering, the fatigue life of the welded structure in the car body was evaluated through the modal structural stress method. The calculated results show that the car body structure obtains obvious modal vibration, which leads to short fatigue life in several weld lines. According to the application requirements of this wagon, the local improvement scheme was proposed, and the effect of the improvement program was evaluated. In this paper, a new fatigue evaluation technology based on the load reverse method of test data was proposed, which provides a theoretical basis for the structural design and program improvement of railway vehicles.
3 May 2026 in Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science
To address the issues of insufficient feature extraction ability and low efficiency of manual parameter optimization in traditional single deep learning models for processing data under variable operating conditions, this study proposes a physics-prior-based time-frequency dual-channel model, CNBiG-AM. The model is designed to handle the different impacts of variable conditions on the manifestation of time-frequency domain fault features by employing a dual-channel time-frequency approach for complementary feature learning. At the same time, a lightweight and efficient model and attention collaboration mechanism are established, utilizing decision-level deep feature fusion for diagnostic decisions. The Sparrow Search Algorithm (SSA) is used as an intelligent optimization strategy to ensure the optimal performance of the complex model under variable operating conditions. Ablation experiments using the CWRU and HNU datasets demonstrate that CNBiG-AM outperforms baseline models in all metrics, with accuracy improvements of 6.5% and 5.5%, respectively, proving the necessity of the BiGRU-attention collaboration mechanism. After intelligent optimization using SSA, the model performance is further enhanced, with accuracy improving by 2.5% on the CWRU dataset and 3.5% on the more generalized HNU dataset. The t-SNE visualization graph shows a significant improvement in the feature space aggregation effect. This study provides an efficient and robust ensemble model for bearing fault diagnosis under variable operating conditions, demonstrating the effectiveness of dual-frequency separation and intelligent optimization algorithms.