To enhance the crashworthiness of subway vehicles, this study presents a novel hybrid anti-climb energy-absorbing device. It integrates inner and outer double-layer thin-walled square tubes with honeycomb structures for synergistic energy absorption. To analyze the energy absorption characteristics of the anti-climb energy-absorbing device, collision simulations were conducted by mounting the device on the front end of a test trolley. Based on Optimal Latin hypercube sampling (OLHS), a Kriging surrogate model characterizing energy absorption (EA) and peak crushing force (PCF) was developed for the parameters of aluminum honeycomb average crushing stress (c), partition thickness (t1), thin-walled outer tube thickness (t2), and thin-walled inner tube thickness (t3). The main effects of parameters on EA and PCF were analyzed, and the results indicated that t2 exerted the most significant influence on the energy absorption characteristics. Multi-objective optimization via the Non-dominated Sorting Genetic Algorithm II (NSGA-II algorithm) was implemented for different vertical offset conditions (h = 0 mm and 40 mm), simultaneously maximizing EA while minimizing PCF. The optimal solution is identified at c = 3.9 MPa, t1 = 4 mm, t2 = 2.9 mm, and t3 = 2.4 mm, with the surrogate model's prediction errors relative to finite element simulation results below 10%. Collision simulations of 6-car subway vehicles demonstrated that key collision indicators all met the requirements of the EN15227 standard, verifying that the device exhibits excellent energy absorption and anti-climb performance, and providing critical references for its optimized design and application in subway vehicles.

To address the fatigue aging of 6005A-T6 aluminum alloy-widely used in rail transit structures-under longterm service, this study investigates its crack growth behavior and remaining useful life (RUL) prediction under different fatigue aging conditions. The simulation covered 4 fatigue aging states, achieved by applying different numbers of pre-fatigue cycles. Compact-tension-shear specimens were tested under mixed-mode I + II fatigue crack growth at 4 loading angles (0 degrees, 30 degrees, 45 degrees, and 60 degrees). Digital image correlation was employed to capture crack tip strain fields for analyzing crack growth behavior. Experimental results show that fatigue aging significantly reduces the material's resistance to crack growth. While increasing the loading angle suppresses crack growth rate, this suppressive effect is weakened under severe fatigue aging conditions. The antagonistic interplay between fatigue aging and increased loading angle in determining RUL is investigated for the first time. Fractographic analysis reveals that the reduction in fatigue striations and the increase in microcrack formation are the key microstructural mechanisms responsible for the fatigue aging-induced decline in crack resistance. Furthermore, an extended finite element model based on an energy release rate attenuation mechanism was developed. The simulation results show high agreement with experimental data, with a maximum standard deviation of 1.3887 and a maximum life prediction error within 7.5 %. These findings provide theoretical support and technical guidance for service life prediction and failure assessment of aluminum alloy structures.

ObjectiveTopology optimization is a widely used optimization method that optimizes material construction, finds the optimal shape of structural design under given design space and constraints, achieves lightweighting, and satisfies constraints. Based on the idea of variable density topology optimization method, taking the fixture structure of a certain test bench as the analysis object, and the establishment of a mathematical model and solution method for topology optimization were elaborated in detail, in order to provide reference for structural design based on advanced design methods.MethodsFirstly, a finite element model of the tooling was established in the HyperMesh simulation software using the OptiStruct module. With the minimum mass and maximum first-order frequency as the optimization objectives, topology optimization and size optimization were conducted on the tooling structure under a single working condition. The material distribution was reasonably arranged and the thickness of the plates was optimized to improve the quality of tooling design.Secondly, based on the optimization results and the practicality of manufacturing and processing, a new tooling structure was designed. Finally, in the Nastran module, the strength check of the tooling model was carried out in accordance with the GB/T 21563—2018 standard. After the actual processing of the tooling structure was completed, the tooling was placed on a vibration table, and longitudinal, transverse and vertical sweep frequency tests were conducted respectively. Subsequently, the test results were compared with the simulation results.ResultsThe results show that the first-order modal frequency has increased from 766 Hz to 1 009 Hz, representing a rise of 31.72%. Meanwhile, the fixture mass has decreased from 51 kg to 44.26 kg, a reduction of approximately 13.24%, its strength meets the design standards and the overall objective is successfully achiered.

To establish a method for predicting the loosening life of bolts under random vibration conditions, a mechanical model of the bolt-connection structure was analyzed. An equivalent relationship between the bending moment M of the screw under transverse vibration and the transverse displacement D of the connection interface was established. The accuracy of this equivalent relationship was verified through simulation calculations and experiments. A relationship curve (D-NL curve) between transverse displacement and the loosening life of M16 bolts was derived from bolt-loosening tests. In the random vibration tests under transverse excitation conditions, the bending moment-time history of the bolt was measured using a force-measuring bolt, while the displacementtime history was determined using the established equivalent relationship. A bolt-loosening life prediction method for engineering applications was then developed, incorporating the cumulative model of bolt-loosening damage, the rainflow counting method, and the D-NL curve. By applying this prediction method to the displacement-time history, the loosening life of the bolt was calculated. The results were subsequently compared with experimental data on bolt-loosening life, confirming the accuracy of the proposed analysis method under random transverse loads. This method provides valuable insights and has scientific significance for optimizing the design of bolt-connection structures, as well as for monitoring and preventing bolt loosening.

Fatigue fracture often occurs in plug welded joints under shear loads, but an accurate fatigue life evaluation method for this special type of joint has not yet been proposed. Therefore, the fatigue failure behavior of plug welded joints through extensive shear fatigue tests was investigated in this study. To this end, the fatigue failure process and failure modes of plug welded joints under shear loading were obtained, and Fa-N curves were established. The investigation revealed that the overall Fa-N curve exhibited poor correlation and did not satisfy the engineering requirements. Consequently, an equivalent structural stress model that accurately captures the stress state of plug welded joints using the coupled node loads as input conditions was derived. A beam-shell finite element (FE) model that can accurately simulate the stress distribution of the plug welded joint was subsequently established to obtain the load inputs for the model. The forces and moments of the coupled nodes extracted from the FE model were input into the equivalent structural stress model to obtain the equivalent structural stress sigma s of the plug welded joint. Finally, Ss-N and P-Ss-N curves normalized for different Fa-N curves were developed, and a shear fatigue life prediction method for plug welded joints based on the equivalent structural stress method was established. The accuracy of this method was verified through random vibration fatigue test. The findings of this study provide a reference for the structural design and life prediction of plug welded joints.

It is essential to describe a material's yield strength, anisotropic hardening, and plastic flow behavior under various stress states and loading angles to analyze large plastic deformations such as material forming and structural collision. However, few existing yield functions can simultaneously accurately predict these mechanical responses. This paper constructs a new analytical yield function and corresponding plastic potential function in the framework of non-associated plastic flow. The function can analytically describe the yield stress or r-value at different angles for several important stress states and further improve the prediction accuracy and control the curvature of the yield surface or plastic potential surface by introducing additional parameters. Parametric sensitivity analysis and convexity analysis were carried out for the new functions. These were used to predict the mechanical properties of three crystal structures, including high-strength steel, aluminum alloy, commercially pure titanium, and titanium alloy. For comparison, several similar analytical yield functions were selected to predict these materials simultaneously. The results show that the proposed yield and plastic potential functions can more accurately describe the four materials' yield strength, strain hardening, and r-value. The different test data types further verify the new functions' flexibility and robustness. Finally, the stress invariant and stress state parameters are used to expand the yield function, and the simplification and convexity analysis of the function are further discussed.

To provide a unified assessment method for remaining useful life that is conducive to engineering applications and accurately considers the effects of material conditions and loading angles, addressing the current lack of models that consistently and physically incorporate these variables, this study proposes an improved fatigue crack growth normalization model. Through uniaxial and multiaxial fatigue crack growth tests, the crack growth behavior of base metal, as-welded, and post-weld heat treatment specimens under different loading angles was investigated. Using digital image correlation to obtain the local displacement field at the crack tip and combining it with the relative displacement method, an improved crack growth driving parameter, KCJP-eq, was introduced to normalize the description of crack growth behavior. The study shows that the proposed normalization model significantly improves data fitting, with the coefficient of determination increasing from 0.6164 to 0.8516, and exhibits strong adaptability to loading angles and material conditions. Furthermore, the variations in the plastic zone during crack propagation were analyzed, and for the first time, the interaction between the material’s microstructure, welding residual stress, and the crack deflection angle was revealed. Finally, fatigue fracture surface analysis indicates that the crack growth mechanism in the Paris region follows a similar mode I behavior.

In railway engineering, the instability behavior of couplers under combined compression and bending influences the collision dynamics response of trains, yet the dynamic interaction mechanism between impact velocity and deflection angle remains insufficiently characterized. This study investigates this coupled instability mechanism comprehensively by integrating full scale impact testing, refined finite element simulation, theoretical modeling, and train system dynamics. The results demonstrate that an increase in impact velocity elevates the axial force demand on the expansion collapse tube through the strain rate hardening effect of the material, whereas an increasing deflection angle degrades the available compressive bending load bearing capacity of the drawbar by introducing an eccentric bending moment. When the platform force elevated by velocity exceeds the critical buckling load reduced by the deflection angle before the expansion collapse tube completes its designated crushing stroke, the coupler response transitions from stable crushing to premature drawbar buckling, thereby triggering the valley effect. On this basis, the present study establishes a theoretical framework that integrates a buffer model dependent on velocity, a platform force model sensitive to strain rate, and a dynamic compressive bending interaction buckling criterion. Following rigorous validation, this theoretical model is embedded into the train system dynamics analysis to achieve a bidirectional coupling between local instability and global collision response. The analytical results further reveal that the coupler instability process comprises three sequential phases consisting of buffer compression, expansion collapse tube crushing, and drawbar buckling. The proposed theoretical model restricts the prediction errors for the platform force, critical buckling force, and equivalent bending stiffness to within 15%, 15.2%, and 13.2% respectively, while faithfully reproducing the valley effect. The train system analysis further discloses that under medium and high speed collision scenarios, the transient load redistribution induced by coupler instability significantly intensifies the transient wheelset lift and amplifies the derailment risk. This study provides a solid theoretical foundation for coupler modeling considering instability characteristics and train crashworthiness safety assessment.

[Objective]CFRP(carbon fiber reinforced poly-mer)and GFRP(glass fiber reinforced polymer)have been ap-plied to the structure of rail transit vehicle equipment cabin.Under impact loading,composite materials exhibit mechanical behaviors distinct from those under quasi-static loading.In view of the reduction in structural protection capability of the equipment cabin floor caused by ballast impact during service,it is necessary to conduct an in-depth study on the influence of strain rate effects of CFRP and GFRP on impact response.[Method]A constitutive model of continuous fiber-rein-forced composites considering strain rate effects is established,and a VUMAT(user-defined material subroutine)is developed based on the secondary development function of ABAQUS software.Dynamic tensile tests of CERP and GFRP are carried out,and dynamic enhancement factors are obtained by fitting the test results.Subsequently,dynamic tensile simulations of the two composite materials are conducted to verify the accura-cy of the model and material parameters.On this basis,CFRP and GFRP are considered respectively as panel materials for the equipment cabin floor structure.According to the GB/T 32060-2015 standard,a simulation model of the aluminum projectile impacting the equipment cabin floor at an initial ve-locity of 200.00 km/h is established for calculation.[Result&Conclusion]When CFRP is used as the panel material and the strain rate effect is considered,the residual velocity of the alu-minum projectile increased from-52.16 km/h to-66.33 km/h,and the maximum panel displacement decreased by 9.69%.When GFRP is used as the panel material and strain-rate effect-is considered,the residual velocity of the aluminum projectile increased from-40.25 km/h to-48.87 km/h,and the maxi-mum panel displacement decreased by 14.53%.

Collision safety is a core requirement in the design of rail vehicles. Collision simulation based on explicit finite elements has become a key means for safety assessment and structural optimization. The reliability of its results is highly dependent on the precise characterization of mechanical behavior under complex working conditions by material constitutive and fracture models. This paper systematically reviews the theoretical system, application status, and challenges of metal material constitutive and fracture models suitable for rail vehicle collision simulation, analyzes the theoretical expression, application scope, and parameter calibration methods of rate-independent models, dynamic coupling models, and machine learning-assisted models, and expounds the modeling basis and predictive ability of coupled and uncoupled fracture models. It is pointed out that there are trade-offs among different models in terms of computational efficiency, parameter rationality, and prediction accuracy, and a reasonable selection should be made based on the characteristics of the research object and scenario. In application practice, typical cases of fracture simulation of vehicle body structure, energy-absorbing components, and connection structures are listed, revealing that the selection of models needs to balance accuracy and efficiency. Facing challenges such as material response prediction, computational efficiency, and mesh sensitivity under complex loads, it is necessary to develop multi-mechanism coupling models and intelligent modeling paradigms in the future, promote the application of standardization and digital twin technology, and provide theoretical support and engineering guidance for high-confidence collision simulation.

To calculate the fatigue life of plug-welded structures, three plug-weld specimens with different plate thicknesses and weld core diameters were first fabricated. Load-life (Delta Fpw-N) curves were obtained through fatigue testing. Normalised specimen structure stress-life (Spw-N) curves were derived by combining finite element simulation of the plug weld structure using the CWELD element and the coarse-mesh hybrid structure stress method. Consequently, a cantilever beam specimen was developed for experimental and simulation purposes. The finite element model was validated through comparative analysis of sweep frequency tests and modal simulations. Finally, comparative analysis of the cantilever specimen's fatigue life was conducted using random vibration testing and simulation. The results showed strong consistency between measured, actual, and simulated lifetimes. This comparison not only confirmed the validity of the Spw-N curve but also the effectiveness of the coarse-mesh hybrid structural stress method for calculating equivalent structural stresses at plug welding, as well as the accuracy of fatigue life simulation methods for plug-welded structures. The investigation's findings have significant reference value and guidance implications.

In this study, the fatigue behavior of cantilever bolted connections was investigated through numerous bending fatigue tests, and load amplitude-life (Fa-N) curves were established. Cantilever bolted connections are highly sensitive to geometric parameters, resulting in a low correlation of the overall Fa-N curve, which cannot directly predict the fatigue life of complex cantilever bolted connections. Therefore, an equivalent structural stress signal acquisition model that comprehensively considers the influence of geometric and loading parameters and accurately characterize the stress state of these connections was derived. This equivalent structural stress signal overcomes the shortcomings of excessive simplification of nominal stress and the inadequacy of hot-spot stress in characterising stress gradients by converting the three-dimensional local concentrated stress of the complex structure into an equivalent one-dimensional structural stress signal. To obtain the load inputs for the model, a beam-shell equivalent finite element model of the cantilever bolted connection was established. The load parameters of the bolted connection were obtained via simulation and coupled into concentrated forces and moments through the beam element nodes. The equivalent structural stress signal (sigma s) of the connection was obtained by inputting the nodal loads into the equivalent structural stress signal acquisition model. Subsequently, the fatigue characteristic curves (Ss-N and P-Ss-N curves) normalized to Fa-N curves with different geometric and loading parameters were established. Finally, using the proposed fatigue life prediction method, the fatigue life of a cantilever bolted connection for an antenna bracket was predicted. The prediction results were highly consistent with the test data, which fully verified the effectiveness and accuracy of the equivalent structural stress signal method.

In order to study the flexible dynamic vibration characteristics of railway freight cars, a rigid-flexible coupling dynamic model of railway freight cars C80 was established by joint simulation of NASTRAN and SIMPACK. The model was verified by modal test and dynamic calculation, in which the maximum error of mode is 9.70

The derailment behaviour of trains following a collision is a critical factor in driving safety. This paper builds upon the existing train collision dynamics model by integrating a finite-length Euler beam track model with elastic point support and a wheel-rail interaction solver. It employs a modified explicit double-step method for computation. The enhanced model and algorithm are utilized to examine train derailment post-collision. A specialized program for collision calculations is developed, with its accuracy and stability confirmed through comparison with finite element analyses. In determining train derailment, the dynamic model adopts a criterion from finite element simulations. It sets a threshold where the lifting height of at least two wheelsets on the same vehicle must not exceed 50% of the nominal flange height, and these wheelsets must not be on the same bogie. The model further investigates the impact of four factors on derailment behaviour in a collision: initial vertical height difference, lateral displacement, initial pitch angle, and initial yaw angle, across various speeds. This research offers significant insights into enhancing the safety measures against train collision derailments.

The 6005 aluminum alloy, widely used in transportation structures, is prone to dynamic load impacts during service. Therefore, it is necessary to understand its rate-related mechanics and fracture behavior precisely. This paper investigates the dynamic mechanics and fracture behavior of 6005 aluminum alloy by combining experimental data with modeling. Comprehensive tests covering seven stress states and strain rates ranging from 0.001 to 800 s-1 reveal that the material's fracture strain exhibits complex strain rate sensitivity, which is dependent on the stress state. Based on these findings, a new rate-dependent constitutive model was developed by introducing the high-order terms of the stress triaxial degree and Lode parameters into the yield function, which significantly improved the prediction accuracy of yield behavior under complex stress states. In addition, a non-coupled ductile fracture criterion was proposed, which uniquely describes the different influences of strain rate on the three stages of micro-void evolution. The model parameters were finely calibrated and verified through finite element analysis. The results show that the model demonstrates excellent accuracy in predicting fracture strain. The average relative errors at strain rates of 1 s-1 and 800 s-1 are only 4.52% and 2.74% respectively, which are significantly better than the traditional J-C type and parameter strain rate-related type correction models. This study provides a reliable modeling tool for the crashworthiness simulation and safety design of aluminum alloy structures under dynamic load conditions.

To investigate the effect of adhesive thickness on the mechanical properties of carbon fiber reinforced polymer (CFRP) and aluminum alloy single-lap adhesive-rivet hybrid joints, single-rivet and double-rivet joint specimens with adhesive thicknesses of 0.2 mm, 0.3 mm, and 0.4 mm were fabricated. Through quasi-static tensile tests combined with digital image correlation (DIC) technology, the load-bearing characteristics, failure displacement, and failure modes of the joints were systematically analyzed. The results indicate that the ultimate load of the double-rivet joints (10 kN) is twice that of the single-rivet joints (5 kN). The increase in adhesive thickness has a nonlinear effect on the failure load of the adhesive layer. When the thickness increases from 0.2 mm to 0.3 mm, the adhesive failure loads of the single- and double-rivet joints are improved by 12% and 14.6%, respectively, but when the thickness increases to 0.4 mm, the load tends to be saturated due to the increased eccentric bending moment or deteriorated stress distribution. In terms of failure modes, the single-rivet joints exhibit progressive failure dominated by CFRP delamination, showing a slow process and large displacement. In contrast, the double-rivet joints show sudden failure due to the instantaneous brittle fracture of CFRP induced by stress concentration. The increase in adhesive thickness significantly improves the failure displacement of the adhesive layer (both single- and double-rivet joints are improved by more than 40%). The effect of adhesive thickness on the mechanical properties and failure modes of different types of adhesive-rivet hybrid joints is revealed, providing guidance for the design of lightweight hybrid connections in fields such as rail transportation.

Non-Gaussian signals are prevalent in engineering applications, yet their frequency-domain equivalence and structural damage effects remain understudied. To address the frequency-domain equivalence of non-Gaussian excitation, this study integrates the Winterstein-Hermite model, Gaussian mixture model, Dirlik method, and fatigue damage spectrum theory to propose an equivalence method for non-Gaussian excitation power spectral density. The method's accuracy is validated through comparative analyses of time-domain signals with varying kurtosis and their equivalent PSDs. The finite element model was validated through experimental and simulation comparisons, yielding structural damage calculations under various kurtosis conditions. Results indicate that structural damage monotonically increases with excitation kurtosis. Multimodal structures can also be evaluated using the Gaussian mixture equivalent PSD under non-Gaussian excitation conditions. These conclusions provide theoretical and engineering references for frequency-domain equivalence analysis of non-Gaussian signals and structural fatigue assessment.

Accurate characterization of notch stress fields under anti-plane shear (Mode III) loading is essential for the structural integrity assessment of tubular joints, yet existing analytical solutions are often limited to the immediate vicinity of the notch tip and rely heavily on stress intensity factors. In this study, a traction-based analytical framework is developed to describe the notch stress fields of V-shaped notches with an end hole (V-O notches) in tubular joints subjected to torsional loading. By integrating traction-based structural stress concept with the eigenfunction expansion method, a unified and physically consistent formulation is established. A two-term eigenfunction expansion is introduced to improve the accuracy of stress representation and extend the validity of the solution beyond the near-tip region. The unknown coefficients are determined from membrane and bending components of traction-based structural stresses, ensuring equilibrium consistency and direct compatibility with engineering evaluation procedures. The proposed analytical solution is systematically validated against finite element analysis (FEA) for various notch opening angles, showing excellent agreement in stress contours and shear stress distributions. The formulation can be readily reduced to the case of sharp V-notches, demonstrating its generality. The present method provides an efficient and practical tool for fatigue assessment and structural integrity evaluation of welded tubular joints without requiring highly refined numerical meshes.

Engineering components are susceptible to numerous fatigue fracture issues in the context of long-term service. The failure of a large number of components is often accompanied by the propagation process of fatigue cracks. The elastic-plastic finite element simulation analysis method was employed to deeply investigate the crack propagation mechanism of aluminum alloy materials under fatigue loading in this paper. First, a finite element model of the CT specimen was constructed based on the constitutive relationship of elastic-plastic materials. Additionally, the crack propagation rule was defined using the extended finite element method (XFEM). Subsequently, the validity and accuracy of the simulation model were verified through fatigue crack propagation experiments using a 6005A aluminum alloy CT specimen. Finally, the simulation model was further utilized to investigate the effects of different stress ratios and specimen thicknesses on the crack propagation behavior. The research findings demonstrated that the crack propagation simulation model established by the elastic-plastic material constitutive and the XFEM is capable of accurately simulating the crack propagation behavior of aluminum alloys under fatigue loading. In the validation CT model, the crack of the simulation model expanded from 13mm to 30mm after 160,000 cycles, and the expansion rate ranged from 2.5 x 10(-5 )to 3 x 10(-3). The height and width of the plastic zone at a crack length of 16 mm were 3.1mm and 2.0mm, respectively, which are very close to the experimental results. Furthermore, the simulation model also reveals the significant role of plastic flow at the crack tip in the fatigue crack propagation process.