摘要:High-temperature mechanical strength is a key performance determinant for the long term, stable operation of advanced energy systems and components in high-temperature service and has been a disciplinary branch in the mechanical strength theory. Its research and development have accompanied major industrial technological advances. The research paradigm has shifted from early empirical formulas and single damage model to a structural integrity assessment framework characterized by mechanistic interpretability, prediction orientation, and evidential reproducibility. Building on the historical trajectory of the field together with bibliometric analysis and keyword clustering, the phase specific migration of research hotspots and the evolving knowledge structure were delineated. Recent progress was synthesized along three complementary themes, namely multiscale modeling, multiple damage coupling, and multidisciplinary integration. The synthesis covered material deformation and damage mechanism, damage evaluation and life assessment, and in-service monitoring and reliability assessment, thereby establishing a traceable mapping from microstructural mechanisms to engineering applications. Looking ahead, advances are expected to deepen in multiphysics coupling, intelligent decision-making algorithms, and standards system development. Critical challenges include bridging high-fidelity models and real-time prediction, establishing robust mappings from microstructure to service life, and translating theoretical modeling into engineering codes.
摘要:In order to study the evolution laws of microstructure and properties during the hot ring rolling, carburizing heat treatment, and deep cryogenic treatment of high-speed railway bearings, the quantitative relationships among the forming manufacturing conditions, microstructure states, and mechanical properties of high-speed railway bearings were established. The optimal process conditions for high-performance forming manufacturing of high-speed railway bearings were determined. The electron back-scatter diffraction (EBSD), scanning electron microscope (SEM), X-ray diffraction (XRD) microstructural testing technologies and tensile, friction and wear, rolling contact fatigue performance testing technologies were used to reveal the evolution laws of the microstructure and mechanical properties of high-speed railway bearing rings during the forming and manufacturing process, and a forming and manufacturing process method for high-performance high-speed railway bearing rings was proposed. The research shows that ring rolling can refine grains, promote the refinement of carbides and increase the dislocation density after carburizing once quenching and tempering, reduce the grain size and carbides, and improve the volume fraction of carbides after secondary quenching and tempering. The deep cryogenic treatment process promotes the decomposition of retained austenite and the precipitation of carbides, reduces the content of retained austenite, enhances the stability of retained austenite, decreases the average size of carbides, and increases the volume fraction of carbides. The wear resistance of high-speed railway bearings is improved by 82.7%, and the contact fatigue performance is improved by 322.1% by applying the optimal hot ring rolling and carburizing heat treatment processes. The research can provide a scientific basis and technical method for the high-performance forming manufacturing of high-speed railway bearings.
关键词:High-speed railway bearing rings;Rolling and carburizing heat treatment;Secondary quenching;Deep cryogenic treatment;Microstructure and property evolution;High-performance manufacturing
摘要:Exponential distribution is widely applied to describe product life in reliability engineering. Correspondingly, product failure rate is a constant (not changing with the service time of the product). Nevertheless, only the products with special property or under particular load condition have exponentially distributed life and constant failure rate. Unrealistic hypothesis of exponentially distributed life will lead to serious error in reliability and failure rate analysis results. In the situations that component life follows exponential distribution, to assume component failures being independent of each other will mislead system reliability evaluation. The the property of product failure rate was analyzed and inferred from the aspects of product strength performance and load environment.The conditions for product failure rate to follow exponential distribution were revealed, and product failure dependency issues in condition of life following exponential distribution were explained.
摘要:To address the issue of characterizing static and dynamic mechanical behaviors of the surface-modified layer (SML) in 18CrNiMo7-6 alloy steel, a layered inversion method for the Johnson-Cook (J-C) constitutive model of SML was proposed. The SML was subjected to layered processing, and dynamic compression tests were conducted on cylindrical specimens with different SML thicknesses. Through progressive parameter inversion, the strain rate sensitivity coefficient C at each depth of the SML was determined. Combined with quasi-static thin plate tensile tests at different temperatures for each depth of the SML, the corresponding yield strength A, strain hardening coefficient B,strain hardening index n,and thermal softening exponent m were determined. Test results show that the SML of 18CrNiMo7-6 alloy steel exhibits significant strain hardening, strain rate strengthening, and temperature softening effects. Additionally, a correlation model between J-C constitutive parameters and dimensionless depth h/hb (distance to SML surface/SML effective depth) was established, providing support for subsequent composite strengthening simulations.
摘要:Different parts of high-speed train bogies are usually designed with aluminum alloy materials of varying strengths, and welding is adopted to connect these different parts. When high-speed trains operate under complex road conditions, the bogies will be subjected to tensile overload, which will produce a coupled superposition effect with the strength difference of welded joints. Therefore, tensile overload tests were carried out on the welded structural components of bogies to study the fatigue crack growth behavior and intrinsic mechanism of aluminum alloy welded joints with different strengths under the action of tensile overload. The compliance method was used to measure the crack growth rate under tensile overload; the digital image correlation (DIC) technology was applied to analyze the change in the size of the plastic zone at the crack tip before and after the application of tensile overload; the scanning electron microscope (SEM) was employed to observe the fracture morphology characteristics of different aluminum alloys in the region affected by tensile overload. The crack growth behavior and intrinsic mechanism under tensile overload were explained based on the change in the size of the plastic zone at the crack tip and the corresponding fracture morphology characteristics. The results show that a single tensile overload can reduce the fatigue crack growth rate and extend the fatigue life. Further analysis indicates that during the tensile overload process, the plastic zone at the crack tip expands and the crack tip is blunted, which together lead to the reduction of the fatigue crack growth rate. The lower the material strength, the more severe the deformation at the crack tip and the more obvious the hysteresis effect under the same tensile overload. The test results of welded joints under tensile overload are consistent with those of the base metal, suggesting that the strength of the hysteresis effect depends only on the inherent strength of the material itself.
摘要:To investigate the influence of structural parameters on the yield strength and deformation behavior of truss lattice structures, face-centered cubic (FCC) porous lattices were fabricated and the mechanical responses were systematically studied. A finite element model was developed to evaluate the yield strength and failure modes of structures with varying geometric parameters. The analysis revealed a correlation between the member slenderness ratio and the transition from progressive collapse to global yielding under different loading orientations. Structures with lower slenderness ratios tend to exhibit global yielding, while those with higher slenderness ratios are prone to layer-by-layer compression failure. Furthermore, lattices supported along the face diagonals demonstrate more uniform global deformation, whereas those supported along the body diagonals are characterized by localized deformation.
关键词:Finite element analysis;Porous lattice structure;Gibson-Ashby model;Progressive collapse;Failure mode
摘要:Under the background of automotive lightweighting, the use of resistance spot welding to achieve effective connection of aluminum to steel structures is an unremitting pursuit. However, in actual welding production, fluctuation of working conditions occur frequently, seriously affecting the quality of weld points. Firstly, the resistance spot welding process was adopted to connect aluminum alloy and low-carbon steel plates. The influences of different inclination angles, plate gaps, and fluctuations in cooling water flow conditions on the resistance spot welding of aluminum-steel were investigated. Then, the quality of the weld points was evaluated by comparing the diameter and thickness of the nugget, the thickness of the intermetallic compound, the coach peel performance, as well as the fracture mode. The research results show that an increase in the inclination angle and the gap between the plates within a certain range, as well as a decrease in the cooling water flow within a certain range, will both reduce the quality of the weld points. Therefore, they should be avoided as much as possible in actual production. The results of this study provide a theoretical basis and practical guidance for optimizing the resistance spot welding process of aluminum to steel.
摘要:At present, the deterministic design and evaluation of high-temperature structural strength have been fully developed and formed a relatively complete system framework, laying the theoretical foundation for the design and manufacturing of numerous mechanical equipment under harsh service conditions. Considering the randomness of high-temperature structural failure and the small-sample characteristics of failure data, safety factors are usually adopted for conservative design in engineering. However, this often leads to structural redundancy and cost waste, so there is an urgent need to conduct research on design methods from determinism to uncertainty. Nevertheless, no universal and mature theoretical methods for high-temperature structural reliability or national/industrial standards have been established so far, making it difficult to effectively predict and guarantee the reliability of high-end equipment such as China's aero-engines during operation. Based on this, firstly, uncertainty analysis was elaborated. The damage-threshold interference criterion was introduced in detail, and its differences from and connections with the stress-strength interference criterion was explained. Finally, taking a certain steam turbine rotor as an example, it illustrates the engineering application of the damage-threshold interference criterion in the reliability analysis of high-temperature structures.
摘要:To enhance the assembly connection performance of adhesive structures in heavy machinery and aerospace equipment, a numerical analysis model based on the cohesive force element was developed to investigate the failure behavior of adhesive joints. The evolution of shear stress distribution in the adhesive layer during the tensile-shear failure process under different loading stages was analyzed. The variations in ultimate failure load and structural stiffness with different adhesive joint parameters were systematically studied, and tensile-shear failure tests were conducted. The results indicate that the shear stress distribution in the adhesive layer transitions from an initial U-shaped profile to an M-shaped and finally evolves into an approximately inverted U-shaped pattern as the load increases. Increasing the length or width of the adhesive layer significantly improves both the ultimate failure load and overall structural stiffness. However, increasing the adhesive layer thickness or substrate thickness exhibits a minor effect on the ultimate failure load. Notably, the structural stiffness decreases with increasing adhesive thickness but increases with higher substrate thickness.
摘要:In response to the problems of reduced surface quality and severe twist drill wear in HR-2 hydrogen resistant steel deep small holes drilling, integrated thermal coupling finite element simulation with experimental investigations of deep small holes drilling to analyze the variations in tool wear, drilling temperature, and the quality of the machined surface during the process of machining deep small holes. The improvement effect of introducing ultrasonic vibration on poor surface quality and severe tool drilling wear was analyzed by comparing ultrasonic vibration assisted drilling (UVAD) and conventional drilling (CD). The results show that as the drilling depth increases, the heat accumulation of the tool's transverse and cutting edges significantly increases, gradually leading to problems such as coating peeling, edge passivation and chipping. Concurrently, the accumulation of cutting heat on the surface of the machined hole causes temperature rise, resulting in material coating, debris adhesion, oblique scratching and other problems on the machined surface, resulting in a deterioration of surface quality and an elevation in surface roughness. Compared to CD, the UVAD effectively reduces drilling temperature, helps to reduce tool wear and maintain cutting edge integrity, while suppressing the increase in surface roughness during machining, ultimately improving surface quality.
关键词:Hydrogen resistant steel;Deep small holes;Ultrasonic vibration assisted drilling;Finite element simulation;Tool wear;Surface quality
摘要:To address the problems of unclear internal load evolution law of the electromechanical composite transmission system for high-speed tracked vehicles and the lack of fatigue life prediction method for planetary gear bearings, a dynamic model of planetary gear bearings in the electromechanical composite transmission system was established to obtain the distribution law of contact loads on planetary gear bearings. A fatigue life prediction method based on the dynamic load characteristics of planetary gear bearings was proposed, which provides certain guiding significance for the optimization and design of the planetary mechanism in the electromechanical composite transmission system. Considering the effects of multi-row coupling effect of the electromechanical composite transmission system, gear time-varying meshing stiffness excitation, and nonlinear support stiffness excitation of planetary gear bearings, a dynamic model of the electromechanical composite transmission system was established using Simpack to obtain the load-bearing conditions of planetary gear bearings. Furthermore, a dynamic model of planetary gear bearings was established using the lumped mass method to analyze the evolution law of contact loads on planetary gear bearings. Then, a fatigue life analysis and prediction model for planetary gear bearings was established using the L-P formula to analyze the variation law of fatigue life of cylindrical roller bearings for planetary gears under different working conditions. The results show that at high rotational speeds, the centrifugal force of the planetary gear bearings during revolution has a significant impact on the contact loads and fatigue life of the planetary gear bearings. At lower rotational speeds, the service life of the planetary gear bearings in the reduction gear increases with the rotational speed. At higher rotational speeds, the service life of the planetary gear bearings in the reduction gear decreases with the increase in rotational speed. At high rotational speeds, measures such as profile modification of the rolling elements of planetary gear bearings can be taken to extend the service life according to the life requirements of the electromechanical composite transmission system.
关键词:Electromechanical composite transmission system;Planetary gear bearing;Dynamic modeling;Contact load;Fatigue life
摘要:Residual stress is the main factor affecting the machining performance and service life of brazed diamond tools. At present, in simulations, diamond is often simplified as a spherical shape, which leads to changes in the degree of structural constraint at the joint, resulting in significant differences between the calculated results and the actual situation. The evolution law and distribution characteristics of residual stress in vacuum-brazed diamonds, as well as the influence of residual stress on wear resistance, were investigated. Firstly, the geometric model of the diamond coating was optimized based on macroscopic morphology. Then, a finite element model of the stress field of vacuum brazed diamond coating was established using thermal elastoplastic mechanics, and the stress field distribution law of the diamond coating was obtained. Subsequently, residual stress measurement experiments were conducted on diamond and nickel based coatings to verify the reliability of the model. Finally, the influence of residual stress on the wear resistance of diamond coatings was explored through wear-resistant weight loss experiments. The main form of wear failure of diamond tools is diamond detachment caused by insufficient grip of the coating on diamond abrasive particles. The diamond wrapped by the brazing material layer is mainly affected by residual compressive stress, which increases the coating’s grip on the diamond. The higher residual stress inside the nickel based coating can also effectively suppress the wear-resistant failure mode of coating peeling.
摘要:Aiming at the limitations of current monitoring methods in the accuracy of early fault diagnosis for rolling mill bearings, a structural design method for intelligent rolling mill bearings based on embedded multi-source microsensors was proposed.A multi-source microsensor module integrating temperature and acceleration signals was developed, and an optimized layout structure of axial sensing leads in the bearing housing was designed, breaking through the bottleneck of sensor integration under the space constraints of traditional bearings. A mechanical performance evaluation system for the slotted structure was established, and the reliability of the intelligent structure was verified through strength check and service life calculation.The results showed that when the slotted area was 10 mm×5 mm, the maximum equivalent stress was 99.71 MPa, which had sufficient safety margin compared with the material yield limit; the maximum overall deformation of the structure was only 0.24 mm, and the local deformation was less than 0.02 mm, with the theoretical service life consistent with that of conventional bearings. The optimized intelligent bearing ensured monitoring functionality while meeting industrial application requirements in terms of structural strength and service life.The research results not only provide a high-precision monitoring method for early fault diagnosis of rolling mill bearings under extreme working conditions but also achieve an integrated "monitoring-structure" process through embedded design. Its strength check standards and service life evaluation methods can directly guide the transformation and upgrading of intelligent bearings in industrial sites, holding significant engineering value for improving the operation and maintenance efficiency of rolling production lines.
关键词:Intelligent bearings for rolling mill;Structural design;Embedded sensing;Finite element calculation;Mechanical property
摘要:As a typical self-excited vibration phenomenon in metal cutting processes, chatter leads to deteriorated machining surface quality, manifested by texture fluctuations, increased dimensional errors, and compromised surface integrity. Effective detection and suppression of chatter is crucial for ensuring machining efficiency and enhancing component performance. Current research has established a multi-dimensional technical framework encompassing physics-model-based offline prediction methods, multi-sensor signal-dependent experimental detection schemes, and intelligent algorithm-integrated online monitoring frameworks. However, existing review literature lacks in-depth dissection of this domain. Addressing this gap, this study conducts a systematic technical review and analysis focusing on chatter detection and suppression technologies. For chatter detection, an analytical-experimental dual methodological framework is established, emphasizing the dissection of applicability scenarios and performance boundaries of various techniques. In terms of chatter suppression, a triple control strategy classification system integrating active-passive-parameter adjustment is constructed, comparing implementation costs and vibration attenuation effects of different solutions. Based on multi-dimensional technical comparisons and cross-disciplinary method integration, existing challenges and potential solutions in this field are explored, providing comprehensive theoretical support and technical references for subsequent research.
摘要:Reaction bonded silicon carbide (RB-SiC) is widely used for manufacturing core components in nuclear energy and optics due to its excellent thermal stability, radiation resistance and chemical inertness. However, RB-SiC is highly hard and brittle, making it difficult to ensure processing quality and efficiency with traditional methods. Ultrasonic diamond wire saw cutting technology, as an efficient machining method for hard brittle materials, has been successfully applied in the processing of single crystal Si, single crystal SiC, and other hard brittle materials. However, the cutting test and process research of this technology in RB-SiC materials need to be carried out. To this end, the ultrasonic sawing test of RB-SiC was carried out for the first time. The ultrasonic wire saw cutting RB-SiC platform was built. The surface quality of parallel and vertical ultrasonic vibration directions was compared and analyzed. The effects of ultrasonic amplitude, line speed and feed speed parameters on surface roughness and surface micro-morphology were studied. Results demonstrate ultrasonic sawing has significant advantages in improving surface quality. Increasing the amplitude from 3 μm to 7 μm reduces surface roughness value by 17.4% and decreases the number of surface scratches and pits. Compared to the vertical feed direction, ultrasonic vibration sawing in the parallel feed direction is more effective in improving the surface quality of RB-SiC materials. This study can provide guidance for the research of ultrasonic sawing process of RB-SiC materials.
摘要:Blind bolted rivets with large flange are widely used as standard fasteners for single-side connection in the aerospace industry, and their minimum tensile load is one of the clearly specified mechanical properties. However, the current method for calculating the tensile strength of blind bolted rivets with large flanges is not yet fully developed. In order to improve the forward design process of blind bolted rivets with large flanges and predict their tensile strength, the failure modes during the tensile process were investigated. First, mechanical analysis reveals three failure modes due to stress concentration: the breakage of the forming sleeve, breakage of the head, and indentation of the nut sleeve. Then, a finite element simulation was used to propose a prediction method for the tensile strength of blind rivet nuts with large flanges, which helps obtain the failure modes and predict the tensile strength. Finally, a hydraulic testing system was used to conduct tensile tests on a specific model of blind bolted rivets with a large flange, the specific failure modes and force-displacement curves are obtained. The accuracy of the proposed prediction method is verified by the test result. This study provides a reference for improving the connection strength of blind bolted rivets with large flanges and enables the prediction of tensile strength in the forward design process.
关键词:Blind bolted rivets with large flange;Failure modes;Tensile strength;Prediction method;Test validation
摘要:In order to improve the comprehensive mechanical properties of 960 MPa high strength steel weld metal, the optimum content of Ti element in 960 MPa high strength steel weld metal was revealed. Firstly, four kinds of weld metals with different Ti contents (0.01%-0.08%) were designed and welded. The effects of Ti content on the microstructure and mechanical properties of welds were systematically studied by scanning electron microscopy, energy dispersive spectroscopy, tensile and impact tests. The effect of Ti content on the initiation energy and propagation energy was evaluated by fracture observation and fracture morphology. The results show that when the Ti content is less than 0.06%, the microstructure of the weld metal changes from granular bainite to granular bainite + acicular ferrite. With the increase of Ti content, the content of acicular ferrite increases significantly. When the Ti content reaches 0.06%, the tensile strength reaches 939 MPa, the elongation reaches 23.5%, the elongation increases by 27% compared with Ti0.01, and the impact absorption energy at -40 ℃ reaches 104 J; when the Ti content increases to 0.08%, the formation of coarse lath bainite and the precipitation of TiN lead to a sharp decrease in plasticity and toughness, the elongation decreases to 18.2%, and the impact energy at -40 ℃ is only 25 J. Ti promotes the nucleation of acicular ferrite and improves the comprehensive mechanical properties by forming TiO2 inclusions. However, excessive Ti will induce the precipitation of brittle phase and the formation of coarse lath bainite, which significantly deteriorates the plasticity and toughness.
关键词:960 MPa high strength steel;Impact toughness;Ti content;Acicular ferrite;Weld metal
摘要:Addressing the rotational vibration issues encountered during the operation of seawater pumps, and integrating measured vibration characteristics, a quasi-zero stiffness ring meta-structure isolator was designed for vibration control based on the principles of quasi-zero stiffness isolation and the bandgap features of meta-structure. First of all, taking a typical seawater pump as the research object and based on an integrated quasi-zero stiffness structure, a structural design scheme of the ring meta-structure isolator was proposed. Then, models of quasi-zero stiffness unit cell, one-dimensional quasi-zero stiffness meta-structure and quasi-zero structure ring meta-structure isolator were established using finite element method and theoretical method, respectively. Their static and dynamic mechanical characteristics were calculated and analyzed, and their isolation effect on the output vibration of seawater pumps was evaluated. The calculation and test results show that the quasi-zero stiffness ring meta-structure isolator provides multiple bandgaps at low frequencies, and exhibits significant vibration suppression effects on typical frequencies of seawater pumps.
摘要:To compensate for the intermittency of renewable energy generation, coal-fired power units are required to operate under low-load conditions over a long time, which causes the last stage blades of steam turbine low-pressure cylinders under small flow conditions persistently, leading to increased dynamic stresses and fatigue damage in the blades. To evaluate the safety of these last-stage blades, the fluid-structure interaction analysis of the last two-stage flow path and last stage rotating blades were conducted for a 660 MW air-cooled steam turbine under typical operating conditions. The results demonstrate that as the load decreases, both the maximum equivalent stress and deformation of the last stage rotating blades gradually diminish first, then slightly increasing. The maximum equivalent stress of the last stage rotating blades remains consistently below the material’s yield strength, indicating that the blades are in the stage of elastic deformation and no plastic deformation has occurred. Using the Goodman curve analysis method for the high-cycle fatigue life assessment of the last stage rotating blades, the dynamic stress levels of the last stage rotating blades fall within the safe zone of the Goodman curve, indicating that there is no risk of fatigue damage to the blades.
摘要:To address the insufficient stability of high-altitude line inspection robots under wind loads, this study proposes optimization strategies involving a novel elastic pressing mechanism and an improved wheel groove, which can effectively enhance their walking stability. A power transmission and distribution line inspection robot with dual-mode switching (flight and walking) capabilities was developed. Firstly, a dynamic model of the robot under wind loads was established, and the relationship between the swing decay time and clamping force, contact area, and friction coefficient was derived. Secondly, dynamic simulations were conducted to verify the performance advantages of the two optimization strategies in suppressing swings. Finally, outdoor wind swing tests were performed to validate the effect of structural improvements. The results show that the elastic pressing mechanism can effectively increase the contact area between the pressing wheel and the line, and the improved wheel groove can enhance the friction coefficient of the walking wheel; both significantly shorten the robot's swing decay time and improve its inspection stability in wind load disturbance environments. The effective technical support and engineering practice basis for the stable operation of high-altitude line inspection robots in complex environments were provided.
关键词:Line inspection robot;Wind load;Elastic pressing mechanism;Improved wheel groove;Oscillation decay time
摘要:Sandwich structures are widely used in aerospace, national defense and other fields due to their lightweight and energy-absorbing properties, and it is of great significance to improve their anti-explosion performance under internal explosion loads.Three structures were designed, including ring sandwich tube (R), polyurethane foam sandwich tube (F), and ring-polyurethane foam hybrid sandwich tube (RF), with the non-filled sandwich tube (A) as the control group. Through internal explosion load tests and finite element simulations, the deformation modes and energy absorption capacities of the four structures under different explosive amounts were compared and analyzed, and the influence of foam filling methods on the mechanical properties of sandwich tubes was explored.The results showed that, compared with the control group, the non-dimensional deflection of structures F, R and RF was reduced to varying degrees under TNT equivalents of 24 g, 36 g and 48 g. At a TNT equivalent of 48 g, the specific energy absorption of RF structure was 5% higher than that of R structure, exhibiting the best anti-explosion performance. In addition, when the TNT equivalent was greater than 37.39 g, the foam-filled ring structure (FR) showed the strongest deformation resistance; when it was less than this value, the foam-filled structure in the gap between the ring and the tube wall (RF) had the optimal anti-explosion performance.
摘要:The challenge in predicting the remaining useful life (RUL) of multi-mode stochastic degradation equipment lies in establishing a class of stochastic degradation models capable of characterizing multiple distinct degradation modes and deriving the remaining useful life distribution of the equipment under such multi-mode stochastic degradation models.First, a generalized stochastic degradation model based on the nonlinear Wiener process was developed, achieving a unified characterization of multi-mode stochastic degradation processes. Second, a maximum likelihood estimation (MLE) method for model parameters was proposed, utilizing historical degradation data from similar equipment. Third, an analytical approximate solution for the probability density function (PDF) of the remaining useful life distribution of multi-mode stochastic degradation equipment was derived under the first hitting time (FHT) framework. Finally, a sequential Bayesian framework for model parameter updating was constructed, enabling online prediction of the remaining useful life of in-service equipment.Numerical simulation analyses and an application case study on bearing remaining useful life prediction demonstrate that the proposed method can effectively model the multi-mode stochastic degradation processes of stochastic degradation equipment and accurately predict the remaining useful life, thereby providing predictive information to support subsequent maintenance decision-making for the system.
关键词:Prognostics and health management;Remaining useful life prediction;Multiple modes degradation modeling;Nonlinear Wiener process;Reliability
摘要:Bearings, as critical rotating components in precision instruments, directly affect the safety and stability of the system. Therefore, accurate prediction of their remaining useful life (RUL) is crucial. Existing RUL prediction methods for bearings can be classified into two types: physical model-based and data-driven approaches. Physical models offer high interpretability and require fewer samples but suffer from low prediction accuracy and cannot be used for online prediction. Data-driven methods, on the other hand, provide higher accuracy and support online prediction but require large amounts of data and have poor generalization ability under varying operating conditions or between different equipment. To address these limitations, a Wiener-ANN hybrid model is proposed for bearing RUL prediction, combining the advantages of both physical models and data-driven approaches. The model optimizes the Wiener process using time-frequency domain features as multi-source input data for the first-stage prediction. Subsequently, a three-layer artificial neural network (ANN) is trained using the first-stage prediction results to optimize the model. The optimized Wiener model is then combined with the ANN to predict the RUL of the test dataset. Comparisons with traditional Wiener models and ANN methods show that the proposed approach significantly outperforms these methods in prediction accuracy and application performance, demonstrating strong potential for engineering applications.
关键词:Bearing;Remaining useful life;Prediction method;Wiener process model;Artificial neural network
摘要:Taking cast iron material of cylinder head as the research object, a series of thermo-mechanical fatigue experiments under different temperature ranges were conducted through bulk sampling. The results show that the fatigue test of cast iron materials exhibits three stages: cyclic softening, cyclic stability and rapid failure. Additionally, the fatigue life of materials under inverse phase loading is significantly shorter than that under positive phase loading. Six typical supervised learning models, including artificial neural networks (ANN) and random forest (RF), were used to predict the fatigue life of the experimental data. However, the results indicate that these models failed to learn the fatigue life distribution trend of the materials. For this problem, the prediction of the thermal mechanical fatigue life of cast iron materials for cylinder heads was achieved by using the self-supervised algorithm based on the generative adversarial network (GAN), and it showed a good prediction effect under the condition of small samples. This research has strong guiding significance and reference value for cylinder head design and fatigue analysis.