The pursuit of higher rail speeds and increasing traffic density have intensified collision risks, making effective crashworthiness design a critical safeguard for mitigating accident consequences. Modern multi-car trains present a particular challenge, as their safety during collisions is governed by complex inter-vehicle interactions. While finite element (FE) methods offer high accuracy in simulating these scenarios, their prohibitive computational cost has limited their utility in rapid safety design and optimization. To address this issue, four machine learning-based framework model for rapid prediction of structural dynamic response was proposed. The surrogate model is coupled with multi-body dynamics to enable efficient multi-objective optimization of train crashworthiness. The AE-GRU model demonstrated excellent convergence, with final losses below 0.0009. More importantly, when applied to large-scale optimization involving 10,000 iterations, the computational cost was reduced to under 25% of that required by conventional FE analysis. This approach provides a robust and efficient data-driven foundation for enhancing collision safety in modern train systems.

This study innovatively combines Fourier curves with circular tube design and proposes a novel corrugated multi-cellular tube structure (NCMTS), aiming to improve crashworthiness. The accuracy of the finite element model was validated through axial quasi-static compression tests. The influence mechanism of different structural parameter combinations on the energy absorption capacity of NCMTS was further investigated. The results reveal that the EA of NCMTS increases monotonically with rising of parameter ru, with a notable acceleration in growth rate as parameter Rc is elevated. Notably, the combination of larger values of ru and Rc, and tout and tc can significantly improve the crashworthiness performance. Moreover, appropriately increasing C1 and C can increase the EA of NCMTS by approximately 50%. Through multi-objective optimization analysis utilizing the global response surface method, the optimal structural parameters for NCMTS were identified: at Rc = 0.68, C = -0.57, C1 = 0.16, tout = 2.0 mm, tc = 0.68, the EA, SEA and IPCF reached 7.02 kJ, 15.33 kJ/kg and 400 kN respectively. This study offers valuable guidance for the design and optimization of high-performance collision energy-absorbing structures.

This paper examines the thermal and structural characteristics of high-speed train axle -mounted brake discs. Initially, investigate the mechanical and thermophysical properties of C/C-SiC composites. Then, dynamic data including temperature and friction coefficients of brake discs under various pressures and speeds were obtained using a full-scale train brake dynamometer. Based on the above test data, a finite element model of the C/C-SiC brake disc under emergency braking conditions at 350 km/h is established using ABAQUS software. The accuracy of the finite element model was validated through experiments, followed by a thermal -structural coupling analysis of the brake discs. Simulation results indicate the highest temperature of the brake disc occurred at 68.80 s, reaching 869.90 degrees C; the maximum thermal stress reached 165.44 MPa at 80.54 s; the maximum axial deformation measured 97.56 mu m at 91.03 s. Although the times at which the highest temperature, maximum thermal stress and maximum axial deformation occurred were not synchronised, the overall trend was consistent. Based on the verification of experimental and simulation results, it has been confirmed that the brake disc meets the requirements of a 350 km/h high-speed train. This provides a solid theoretical foundation for the structural design and iterative upgrades of high-speed train brake discs.

Carbon fibre reinforced silicon carbide (C/C-SiC) composite materials have attracted increasing attention in brake system of high-speed trains, due to their low density, high specific strength, thermal stability, and frictional wear performance. This study aims to explore the mechanical properties of the 2.5D C/C-SiC composites, in order to characterise its potential application to the friction braking system of high-speed trains. To comprehensively evaluate the mechanical performance of the 2.5D C/C-SiC composites, the chemical vapor infiltration (CVI) method was adopted to prepare novel samples with high densification and low content of residual Si. The mechanical properties and failure mechanisms of the composites were investigated under various loading conditions, including tension, compression, bending, and shear tests. The experimental results indicated that the 2.5D C/C-SiC composites exhibited superior mechanical properties, with average in-plane tensile strength, compressive strength, bending strength, and interlaminar shear strength reaching 127.67 MPa, 326.92 MPa, 355.74 MPa, and 9.77 MPa, respectively, which are 2-8 times higher than the mechanical properties of C/C-SiC composites in existing publications. Under different loading conditions, the composites demonstrated characteristics of ductile fracture, pseudo-plastic compression fracture, pseudo-plastic bending fracture, and brittle shear fracture. Both high fibre content and the formation of SiC structure were advantageous for enhancing the load-bearing performance of C/C-SiC composites. The outstanding mechanical properties of the 2.5D C/C-SiC composite render the brake disc highly resistant to deformation and cracking under high stress, thereby enhancing its durability and establishing it as the ideal material for the next generation of high-speed train brake discs.

To enhance the energy absorption characteristics of the energy-absorbing structure in metal alloy 3D printing, various heat treatment conditions were studied on the 3D printed hourglass multi-cellular energy-absorbing structure (HTMEAS). The results reveal that as the annealing temperature rises, the structural morphology of the alpha-Al+Si phase in AlSi10Mg alloy undergoes alterations, particularly when the annealing temperature remains below 300 degrees C. After heat treatment, the HTMEAS can contract layer by layer along the gradient and undergo controllable and orderly deformation. The maximum energy absorption, displacement, and peak force reached 758 J, 30 mm, and 30 kN, respectively.These values represent a significant enhancement of 146.8 %, 157.8 %, and 50 % compared to those without heat treatment, showcasing commendable mechanical properties. This advancement establishes a solid theoretical framework and assurance for the future development and refinement of 3D printing energy-absorbing structures.

Deep learning is attracting increasing attention due to its excellent predictive power and its ability to be applied in traditional research areas. In this paper, we propose a multisource response prediction network architecture based on a long short-term memory (LSTM)-stacked autoencoder to predict key crashworthiness indicators and complete curve reconstruction. Taking an expansion tube as an example using an equivalent scaling research method, a scaled expandable tubular (SET) finite element model was established and verified using a quasistatic compression test and a full-size coupler and buffer system experiment. A design of experiments (DOE) approach was used to obtain a dataset for training the prediction network. Neural network hyperparameters are critical to network prediction accuracy, and after comparison, the multisource response prediction network architecture showed good computational efficiency and satisfactory prediction accuracy when appropriate hyperparameters were selected. Subsequently, multiobjective constraint optimization was performed using the nondominated sorting genetic algorithm-II (NSGA-II) based on a prediction network architecture, which greatly improved the energy-absorption structure optimization accuracy. The results are expected to provide a research methodology for solving complex engineering problems by establishing a new framework for deep learning algorithms combined with optimization methods.

This study introduce an innovation okra biomimetic corrugated multi-cellular Taper tube (OBCMT) designed for the energy absorption, drawing inspiration from the okra, which comprises a Fourier curve wall and X-shaped ribs. Validation is accomplished through quasi-static crushing experiments, ensuring the accuracy of the finite element simulation model. Numerical simulations investigate the influence of critical interaction structural parameters on the performance of the OBCMT, and a theoretical model based on the super folding theory is deduced to predict the mean crushing force of the OBCMT. The result show that the specific energy absorption ( SEA ) exhibited the heightened sensitivity to the variation of C_height under the same C _ R , approximately reached 8.36 kJ/kg which increased by 43.2 % compared to the minimum value in the identical group. Simultaneously, the structure parameters of height coefficient( C_height ) and radiu coefficient( C_R ) have the most significant impact on the deformation mode of OBCMT. Comparative analysis demonstrates that the OBCMT exhibits the extraordinal crashworthiness compared to the contemporary typical energy absorption structures under identical mass, with the SEA reaching 21.8 kJ/kg. This represents approximately 290.7 %, 205.7 %, 187.9 %, 189.6 %, and 174.4 % relative to the six-cell hexagonal tube (SHT), quadruple-cell circular tube (QCT), multi-cell bicircular tube (MBT), five-cell square tube (FST), and quadruple-cell square tube II (QST_II), respectively. This research significantly advances the development of high-performance bionic energy-absorbing structures for crash applications and enhances our comprehension of the biomimetic engineering.

Lattice structures and biomimetic structures are lightweight and have high specific energy absorption and mechanical properties. They have been developed and widely used as energy-absorbing structures. Bionics and lattice structures are integrated in this study to design a biomimetic lotus root lattice structure (BLRLS) with hourglass-shaped cells. The BLRLS unit cells are stacked in the x, y, and z directions utilizing the technology of additive manufacturing. The mechanical characteristics of the BLRLS under uniaxial compression were studied, and the corresponding finite element model was verified by quasi-static experimental results. The deformation and the energy absorption mechanisms of the BLRLS were also analyzed using the verified finite element (FE) model. An empirical formula for the mean crashing force of BLRLS was subsequently derived. The results show that: the crash performance indicators initial peak crushing force (IPCF), energy absorption (EA) and the average of the crushing force (MCF) of the experimental and numerical were found to be 2.68 %, 0.52 % and 3.54 %, respectively. As height coefficient (CH) and wall thickness (t) increase, the crashworthiness (mean force, energy absorption and specific energy absorption) of BLRLS will be improved. The radius coefficient (CR) of the BLRLS has a significant effect on its deformation mode. When CR is less than 0.6, the BLRLS exhibits a petal shaped deformation mode; when CR is greater than 1 (R > r), the cell body of the energy absorbing structure is inward concave, which R and r represent the cell major radius and cell minor radius, respectively. The errors between the theoretical predicted values and the experimental and finite element models are 7.8 % and 11.1 %, respectively. Additionally, the BLRLS exhibits desired deformation mode during compression, resulting in high EA and specific energy absorption (SEA), and great potential in engineering applications.

This paper introduces a new hybrid friction energy-absorbing structures (HFEAS), which is consisting of a bearing bush, an anti-creeper device, a pretension bolt, a friction plate, and friction metal/CFRP tubes have been designed in the paper. The friction coefficient of raw material of CFRP was determined by the MM 3000 friction testing machine and a finite element model of HFEAS was established. Experimental verification of the finite element model was conducted, as well as a study of the influence of structural parameters on crashworthiness. By analyzing the parametric influence, the pre_force ( P_F ) has the largest influence on the HFEAS, followed by diameter ( D ) and stell_thickness ( S_t ), pre_length ( P_L ) and layer_number ( L_n ). In particular, the peak force increased by 8 times and the platform force was 15 times higher than origin, when P_F increases from 1 to 21 kN; Overall, the HFEAS combines the recycling and tribological characteristics of composite materials, offering a new direction for the iterative upgrading and development of energy-absorbing structures.

This study proposed a method to optimise the crashworthiness for the crush zone structure of a subway vehicle based on the beam element model. In the preliminary conceptual design, a simplified beam element model of the crush zone was established to verify the feasibility of the model. Then, a parameterised modelling method for the mechanical characteristic curves of the energy-absorbing structures and the front-end beam was proposed, and a hybrid-weight prediction model of the mechanical curves was built. Based on the mapping relationship, a database of mechanical and geometric parameters of the crush zone was established. Finally, beam element models of the crush zone with different arrangement schemes were obtained through optimisation based on genetic algorithm. Multiple sets of feasible geometric structure parameters were obtained and the best scheme was determined. The research results showed that the calculation speed for the simplified beam element model of the crush zone structure was 461 times faster than that of the finite element model. The prediction model of the mechanical characteristic curve of the crush zone structure built in this study exhibited sufficient accuracy. The mapping relationship between the geometric structure and the mechanical properties enabled the optimised structure by the discrete beam model to be quickly obtained. The mechanical parameters of the optimised crush zone were obtained with energy absorption of 597.51 kJ, mean crushing force of 2398.38 kN, and crushing displacement of 249.72 mm. The proposed method is expected to provide a convenient guideline for fast optimisation and upgrade of energy-absorbing structures.

To improve the energy absorption characteristics of the energy absorbing structure of metal alloy 3D printing, different heat treatment conditions were studied on the the 3D printed hourglass multi-cellular energy-absorbing structure (HTMEAS). The results show that with the increase of annealing temperature, the organization morphology of α-Al+Si phase of AlSi10Mg alloy changes, when the annealing temperature is below 300°C. When the heat treatment at 500 °C, the tensile strength of the AlSi10Mg alloy decreased significantly to only 47% of its original value. The HTMEAS after heat treatment can contract layer by layer along the gradient and undergo controllable and orderly deformation, the maximum energy absorption, displacement and peak force reached 758 J, 30 mm and 30 kN, respectively, which are 146.8%, 157.8% and 50% of those without heat treatment, which present good mechanical properties, which can provide a theoretical framework and guarantee for 3D printing energy-absorbing structures in the future and its iterative improvement.

为提高地铁列车耐撞性,降低碰撞姿态造成的二次损伤,提出一种控制列车碰撞姿态的辅助保护装置(抱轨装置).首先,设计抱轨装置的几何结构并定义抱轨装置的力学特性,采用动力学方法建立8编组地铁列车的碰撞模型;其次,结合实验验证仿真模型的准确性,研究抱轨装置对列车碰撞姿态的影响;再次,提出3种不同的抱轨装置安装方案,对比分析得到经济且有效的方案;最后,基于EN15227:2008标准对方案进行耐撞性评估,基于多目标遗传算法对抱轨装置的设计参数进行优化.研究结果表明:当钩状抱轨装置安装在车体质心正下方时,在满足控制列车碰撞姿态的要求的前提下,可以不占用车体太大的空间,满足EN15227:2008中耐撞性考核指标,是经济且有效的方案;当距离车体质心的纵向相对位移dinstall=10 000 mm,抱轨装置静止阶段距离x1=9.47 mm,抱轨装置线性阶段刚度k1=5 000 N/mm时,列车的车体和轮对的最大垂向抬升量最小,取到最优值,最大轮对位移抬升量dmax=13.87 mm,列车车体最大俯仰角θ =3.12°.

The energy absorption structure of a train is an important part of passive safety protection during train collisions and is the last line of defense to protect both passengers and trains. In the design process of a train energy absorption structure, improved stability and greater energy absorption capacity is required. A cutting anti-climbing energy absorption structure offers good stability and energy absorption in a collision, but it can easily generate considerable heat in the energy absorption process. Therefore, it is important to conduct thermal–solid coupling simulations and crashworthiness optimization for cutting energy absorption structures. To improve the passive safety protection capability of high-speed trains, this paper experimentally and numerically explored the crashworthiness of a cutting-type energy-absorbing structure composed of an anti-creeper device, an energy-absorbing tube, cutting knives and knife-supporting tools. By adopting the Johnson–Cook material model, a finite element model was developed to study its energy absorption characteristics in a coupled heat–solid state. The effects of cutting depth (D), cutting knife front angle (A) and cutting width (W) on energy absorption (EA), cutting platform force (Fmean) and peak cutting force (PCF) were analyzed based on the validated simulation model. The results showed that EA, Fmean and PCF increase with increasing D and W, while EA, Fmean and PCF decrease with increasing A. The GRSM was employed as the optimization algorithm, and a gain matrix–cloud model optimal worst method (G-CBW) multiobjective decision algorithm was proposed to obtain the most satisfactory configurations from the Pareto front solution. The relative errors from the optimal and finite element results of EA, Fmean and PCF were 3.5%, 2.1% and 2.2%, respectively. All the crashworthiness indicators were improved considerably.

Due to the lack of load/displacement sensors in a complex and uncertain crash test/accident of rail vehicles (e.g., vehicle-to-vehicle or train-to-train collision), only structural deformation images can be obtained while the crashworthiness indicators (e.g., force, displacement, energy absorption) cannot be measured directly. This paper aims to propose a transfer learning-based inverse method for extracting the structural parameters and crashworthiness characteristics by the deformed pictures of energy-absorbing structures. A finite element model of an energy-absorbing structure was firstly established and calibrated by experiments. Then, a number of deformation images were captured from the numerical design of experiment (DOE) through coding languages, which were saved as TFRecord format to reduce the computational time during the training of transfer learning models (i.e., VGG16, LetNet, AlexNet and ResNet50). The result showed that the transfer learning model, ResNet50, exhibited the best performance with R2 of 0.736 and 0.981, respectively, for predicting the structural parameters and crashworthiness characteristics. In addition, the number of full connection layers should be reasonably selected on the premise of maintaining accuracy and efficiency. A group of deformation pictures were randomly used as samples to validate the prediction of structural parameters and crashworthiness through the trained transfer learning model, where good consistence was observed. The proposed method is expected to bring the image recognition and big data prediction into the design and test of composite energy-absorbing structures, thus, auxiliary improve the crashworthiness of rail vehicles.

Strain-based structural health monitoring technology has been widely used in the field of transportation. The existing strain damage identification methods have defects such as complex process, lag in state evaluation, and low intelligence. This paper adopts the deep learning method to establish a network model that uses the strain field information to map directly to the damage information, and takes a subway bolster as the engineering background to realise the end-to-end automatic damage identification. Firstly, the problem of damage identification in strain field is described, combined with the idea of fully convolutional network. The basic structure of damage identification network is modularised, and the overall design framework is proposed. Then, the damage simulation method is determined, and the feasibility of using this method to construct a strain field damage dataset is verified. The batch random damage model generation and the random noise signal addition program are coded, and the datasets of the bolster under static/dynamic force are obtained. Finally, the deep learning model is applied to the bolster damage dataset, and a residual module BolRes_Att that integrates spatial attention and channel attention mechanism is proposed. It has better damage identification performance without an increase in model parameters. The average number of faulty elements on the two bolster test sets of the improved damage identification network is 3.02 and 2.92, respectively, accounting for about 0.016% of all elements. The average processing time for a set of data is only 0.014 s. The results show that the deep learning model constructed in this paper can accurately and quickly identify the damage information of elements according to the strain field information.

This paper presents a novel framework for predicting the crashworthiness of a square cone energy-absorbing (SCEA) structure using a machine-learning method. The structure consists of an anti-creep, a thin-walled structure with nonuniform thickness, diaphragms, two types of aluminum honeycombs and a guide rail. The finite element model of SCEA structure was established and validated by full-scale experimental test. Taking the thicknesses of thin walls ( T A and T B ) and diaphragms ( T gb ), strengths of honeycombs ( δ A and δ B ) as parametric variables, the parameters of SCEA structure were changed based on a virtual design of experiments (DOE) to generate training data and test data. To improve the crashworthiness of SCEA structure, the structural parameters were employed as input data, four machine learning models were utilized to predict the energy-absorbing characteristic curve of the SCEA structure, and the prediction accuracy of different models was compared and analyzed. According to the results of comparison, the Gate Recurrent Unit (GRU) model was chosen to predict the structural energy-absorbing characteristics, also employed as the input of optimization. The energy absorption ( EA ) and initial peak crushing force ( PCF ) were adopted as objectives, and the global response surface method (GRSM) was employed as the optimization algorithm. The results showed that the optimal solution was obtained as PCF = 618.41 kN and EA = 297.99 kJ when T A = 2.1 mm, T B = 2.9 mm, T gb = 2.4 mm, δ A = 5.99 MPa and δ B = 4.82 MPa. The machine learning method offers engineers and scientists a potential tool to accelerate the design and optimization of SCEA structures for rail vehicles.