2026-04-02
Chunlong Xu
IntroductionConducting reliability analyses for engineering problems with small failure probabilities and expensive computational models is challenging. Subset simulation (SS) is an excellent method that has been applied in many fields. However, SS still has several limitations that need to be addressed, such as correlated samples, a large coefficient of variation (COV), the risk of deviating from dominant failure regions, and inaccuracies in problems with multiple failure regions.MethodsIn this paper, a subset simulation algorithm integrating importance sampling is developed to address the aforementioned limitations, focusing on complex reliability problems characterized by low-to-moderate dimensionality and small failure probabilities. First, a seed placement strategy on intermediate limit-state curves is developed to reduce the COV and generate independent samples within each subset. Second, interval estimation combined with clustering algorithms is applied to precisely identify seeds. This strategy is designed to handle problems featuring multiple failure regions while mitigating the risk of divergence from dominant failure regions.ResultsThe performance of the proposed algorithm is demonstrated through seven case studies from the literature, including problems with multiple failure regions, nonlinear problems, system reliability problems, SS counterexamples, and structural reliability problems.DiscussionThe results show that the proposed method provides more accurate and robust failure probability estimates than the other tested methods.
DOI: 10.3389/fmech.2026.17136742026-03-31
Ali Malik Saadoon, Nassear R. Hmoad, Suhair G. Hussein, Mohammad Qasim Abdullah
In this research, a detailed finite-element (FE) analysis of the combined influence of the drilled-hole position, the shape of the hole, and the fillet design on the structural and dynamic performance of spur gears is investigated. ANSYS R16.2 was used to create a three-dimensional numerical model that can be used to assess the bending stress distribution and vibration response under realistic loading conditions. A trochoidal fillet and four circular fillet radii (0.5, 1.0, 1.5 and 2.0 mm) were studied to determine their effect on the stress concentration behavior. FE-guided hole-suggestion process was introduced which is an automated process in which low-stress zones to be cut away are identified so as to allow systematic recommendation of optimal locations, orientations and size of holes without any empirical relation. It was found that root stress decreased dramatically as fillet radius was increased, and 2 mm fillet had the minimum bending stress of all circular arrangements. The baseline configuration (Rf = 0.5 mm, without holes) exhibited a maximum bending stress of 69.45 MPa, whereas increasing the fillet radius to 2.0 mm resulted in a stress reduction of approximately 35%. The trochoidal fillet provided less stress gradients and a larger zone of low stress surrounding the tooth root. The holes proposed by FE were further incorporated, which increased structural performance. Hole size out of the chosen geometric parameters was statistically most impactful on bending stress and dynamic response, which ANOVA proved to be accurate (p < 0.001). The holes in the top the most desirable performance were medium-size (≈2.0–2.4 mm) drilled horizontally, which minimized bending stress by about 46%–50% relative to the baseline gear and ensured very low peak dynamic displacement (∼3.4 × 10−5 m at approximately 73 Hz). Structural integrity is well enhanced by optimizing fillet radius and drilled holes sizes, directions, and locations regarding the strength and dynamic stability. The proposed methodology offers a reliable and scientifically grounded framework for gear modification with strong potential for integration into advanced gear design and light weighting applications.
DOI: 10.3389/fmech.2026.18006592026-03-25
Wei Song
IntroductionIn order to solve the problems of low accuracy and multi-objective optimization imbalance in traditional hybrid electric vehicle energy management strategies under dynamic conditions.MethodsA study was conducted to design an improved reinforcement learning energy management strategy based on dual delay deep deterministic strategy gradient (TD3), aiming to improve fuel economy, extend battery life, and enhance strategy robustness. Firstly, a multi energy system dynamics model was constructed, which includes an engine, power battery, and electric motor. Secondly, in order to solve the problems of slow convergence and easy getting stuck in local optima in traditional reinforcement learning for multi-objective optimization, adaptive reward functions and priority experience replay mechanisms are introduced.ResultsThe results indicate that the initial value of the state of charge for all three strategies is 0.5, and the research strategy maintains it at 0.5.DiscussionITD3 can more accurately control the state of charge, making it close to the initial value and reducing excessive energy consumption; Overall, compared with traditional strategies, this research strategy exhibits better battery state of charge retention ability under two typical operating conditions. This strategy can achieve precise energy management, effectively reduce costs, improve energy utilization efficiency, support environmental sustainability, and provide better solutions for energy management of new energy hybrid vehicles.
DOI: 10.3389/fmech.2026.17696452026-03-23
Luisa Fernanda Mónico Muñoz, Oscar Hernando Venegas Pereira
Environmental concerns have increasingly driven industries worldwide, particularly the automotive sector, to address the challenges posed by pollutant emissions from internal combustion engines. Diesel engines, for instance, offer higher thermal efficiency than gasoline engines but remain major contributors to atmospheric pollution. Their emission characteristics are also strongly influenced by fuel properties. One promising approach to mitigating these emissions is the use of gasoline–diesel fuel blends. Due to their higher volatility and improved vaporization behavior, these blends promote more homogeneous air–fuel mixture formation, making them suitable for compression ignition engines. In addition, modifying key combustion parameters, most notably injection timing, has proven effective in influencing both emissions and combustion dynamics. Alongside injection pressure and intake oxygen concentration, injection timing plays a critical role in determining pollutant formation and the acoustic characteristics of the combustion process. This study examines the impact of a gasoline–diesel blend (G10) on the performance and emission characteristics of a diesel engine, with particular emphasis on the effects of varying injection timing. The aim is to experimentally evaluate how combining this blend with injection timing adjustments influences engine efficiency and emission output. The experimental results show that advancing injection timing improves torque, power output, and thermal efficiency while maintaining relatively low fuel consumption. Conversely, retarding injection timing is more effective in reducing pollutant emissions. The most effective strategy is delaying injection at 80% load and 3,500 rpm, which results in reductions of smoke density, NOX, and CO2 by 77.34%, 34.45%, and 11.34%, respectively. Performance also improves, with torque increasing by 26.25%, power by 14.52%, and specific fuel consumption decreasing by 9.76%. Although a trade-off exists between optimizing performance and minimizing emissions, the findings indicate that strategic calibration of injection parameters can achieve a balanced compromise between both goals. In conclusion, adjusting injection timing emerges as a viable technique for reducing pollutant emissions without significantly compromising—and potentially even enhancing—engine performance.
DOI: 10.3389/fmech.2026.17342702026-03-09
Jing Li, Aoqi Lian, Jiawei Yang, Lihua Liu
IntroductionIn the process of industrial automation, industrial robots are widely used in complex operations such as welding, assembly, and handling. The dynamic response performance under time-varying load conditions directly affects production efficiency and control quality.MethodsTo improve the dynamic response speed and control accuracy of industrial robot Programmable Logic Control (PLC) systems under time-varying loads, an improved Model Reference Adaptive Control (MRAC) strategy that combines Fuzzy Correction Adaptive Law (FCAL) and Particle Swarm Optimization (PSO) algorithm is designed. It combines an Extended Kalman Filter (EKF) load observer with a composite control law to adapt to the discrete characteristics of PLC and optimize multi-task scheduling.ResultsExperiments show that in three scenarios: automotive welding, electronic assembly, and metal cutting, the production efficiency of this system is increased by 20.7%–23.8% compared with traditional PLC methods, and the dynamic response time is shortened from 0.8 s to 0.3 s. The product qualification rate increases from 1.9% to 3.9%, and the positioning error of the assembly robot drops from ±0.1 m to ±0.05 m. The torque fluctuation of the cutting robot motor drops from 1.0 N m to 0.58 N m, the load observation error does not exceed 0.05 N m, and the angular velocity overshoot is less than 1.2%. DiscussionThrough the deep integration of adaptive control strategy and PLC system, the dynamic response speed and control accuracy of industrial robots under time‐varying load conditions are effectively improved, and production efficiency, product qualification rate, and energy consumption indicators are improved. This study provides reliable technical support for the field of flexible manufacturing.
DOI: 10.3389/fmech.2026.1777195