Simulating three-dimensional acoustic problems with traditional finite element models (FEMs) may result in a large dispersion error that is difficult to manage. To control dispersion error caused by “overly stiff” features of FEM, a face-based smoothed FEM (FS-FEM) model was used to analyze the sound field of a high-speed train. Based on the acoustic Galerkin method’s weak form, smooth regions were reconstructed within the original finite element area, and numerical simulations were performed in MATLAB. Modal analysis of a test example revealed that results obtained using FS-FEM are more accurate than those obtained using the traditional FEM. Finally, it was used to calculate the sound field of China Railway High-speed 380B (CRH380B) high-speed trains in time and frequency domains, achieving higher accuracy than the traditional approach. Results were more akin to the realistic solution, which demonstrated the performance of the FS-FEM.

Traditional structural forms are difficult to meet the lightweight requirements of subsequent spacecraft for load-bearing structures. In the aerospace industry, filling structure with lattices is a popular approach to reduce the weight of a spacecraft. However, this design strategy has deficiencies in the spatial distribution of lattice cells as well as its affection on the mechanical properties. In this study, a two-step topology optimization technique is proposed to solve the spatial distribution problem of nanosatellite. Firstly, an entire nanosatellite box composed of panels which filled with uniform lattices is sent to the vibration test to obtain the frequency data. Then, a finite element (FE) model of the nanosatellite structure which contains the same uniform lattices is built and validated with the obtained frequency data above. For the subsequent calculation of topology optimization. An equivalent model of the verified FE model is established by replacing the lattice cells in the sandwich layer with equivalent fictional elements. Subsequently, a topology optimization problem under the mass constraints is formulated for maximize the nature frequency, and a new light weighted nanosatellite which filled with non-uniform lattices is established by applying the density mapping method and the previous topology configuration result. By separating the design problem of nanosatellite into two steps, the proposed optimization design method achieves the maximum frequency design under the weight constraint. Furthermore, the frequency is also guaranteed to be around the nature frequency. The results reveal that the mass of the new structure with non-uniform lattices is reduced by 50.32% and the frequency is increased by 1.19%. An important technical importance and application value of this proposed technique is that it improves the performance and design efficiency of the load-bearing structures of a nanosatellite, and this method has significant technical significance and application value.

Customizing and optimizing lattice materials poses a challenge to designers. This study proposed a data-driven generative method to customize and optimize lattice material. The method utilizes subdivision modeling to parametrically describe lattice morphologies and skeletons. Next, the homogenization method is employed to analyze elastic moduli for collecting a dataset. Then, a two-tiered machine learning (ML) framework is proposed to predict the elastic modulus for a forward design. The first-tier model employs polynomial regression to estimate relative density, which serves as an additional input feature for the second-tier model. The prediction accuracy of the second-tier model is improved through the additional inputs. The forward and reverse design strategies offer a flexible and accurate means of tailoring lattice properties to meet specific performance requirements. Two case studies demonstrate the practical value of the framework: customizing a lattice material to achieve a desired elastic modulus and optimizing the mechanical performance of lattice materials under relative density constraints. The results show that the prediction accuracy of the elastic modulus using the two-tiered ML model achieved an error of less than 10% compared to finite element analysis, demonstrating the reliability of the proposed approach. Furthermore, the optimization design achieved up to a 25% improvement in mechanical performance compared to conventional lattice configurations under the same relative density constraints. These findings underscore the advantages of combining generative design, machine learning, and genetic algorithms to navigate complex design spaces and achieve enhanced material performance.

In this paper, the authors elaborate how numerical techniques developed from the direct method can be used to predict the load-bearing capacity of the aluminum casting beam structure which is presently used in the Chinese high-speed train CRH5A. The numerical method introduced in this paper is formulated based on the Melan's static theorem; thus, it determines the strength of the cast aluminum beam under both monotonic and cyclic loads without following the entire load profile. In addition to constructing the plastic and shakedown limits of the component, the proposed approach is also employed to study how stiffeners' thickness on the key areas influences the feasible load domains. Based on the intensive sensitivity analysis, an optimal thickness has been determined which gives the highest strength-to-weight ratio. To this end, the paper confirms that the direct method is a viable means for designing structures used in the rolling stock.