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.

Aiming at the problem of the weld fatigue test of the new energy vehicle subframe, a program load spectrum compilation method based on the failure dominant load was proposed. Firstly, the finite element model of the rear subframe was established, and the stress distribution under unit load was coupled with the load of each connection point. The structural stress method was used to evaluate the fatigue life of the subframe welds, and six dangerous points that were easy to fail were selected. Secondly, by comparing the load damage of each connection point, the failure dominant connection point corresponding to the weld dangerous unit was determined. Then, the failure dominant load was determined by the principal stress analysis, time domain correlation and uniaxial damage contribution at the weld, so as to reduce the dimension of the multi-axial load and reduce the difficulty of the bench test loading. Finally, a pseudo-damage matrix was output based on the failure dominant load, the characteristic working conditions and their proportions were selected to obtain the load spectrum of the fatigue accelerated test program, and the minimum number of cycles was determined according to the principle of the damage equivalence. The numerical simulation results show that the program load spectrum can reproduce the damage of dangerous points and has a high acceleration coefficient, which verifies the effectiveness of the accelerated test spectrum.

Due to the needs of transportation, installation, grid connection, and maintenance, the construction of offshore wind farms in inshore areas often cannot be far from busy surrounding waterways, which significantly increases the probability of offshore wind turbines being impacted by ships. To analyze the performance and damage of different hollow ratios of protective devices when offshore wind turbines are hit by ships，the collision process of a 5000-ton bow-downward ship with an offshore wind turbine at a speed of 2.0 m/s was simulated using Ansys/Ls-Dyna. The influence of the hollow ratio on the anti-collision performance of Ogden rubber，Mooney-Rivlin rubber and Aluminum foam aluminum constitutive protective devices was studied and compared. The results show that with the increase of hollow ratio, the impact duration of aluminum foam protective devices increases, and the contact force decreases accordingly, while rubber materials show the opposite trend. As the hollow ratio decreases, the protective device is more similar to a solid tube, with a relatively smaller maximum indentation depth. Under the influence of hyperelastic properties, the indentation depth of rubber materials is smaller than that of aluminum foam after the collision is completed, and the material damage of the protective device is smaller. However, the proportion of internal energy in the support area gradually increases, so the influence of hollow ratio on the leg support needs to be considered in the design and research of protective devices.

The main types of carbides in M50 bearing steel are MC, M₂C and M₂₃C₆. Under the scanning electron microscopy (SEM), they exhibit significant differences in the shape, size, and distribution. Some carbides have larger sizes and uneven distribution. They become areas of stress concentration under loading，which has a negative impact on the bearing fatigue performance. So an improved mask region-based convolutional neural network (Mask R-CNN) model was proposed which can batch identify the types of three kinds of carbides in SEM pictures, the diameters of carbides were measured, and the distribution of carbides was showed. The output images and histogram results show that the size of M₂C carbide in M50 bearing steel is large and unevenly distributed, but the distribution of MC carbide with the largest size and M₂₃C₆ with the smallest size is reasonably uniform.

In order to improve the high temperature wear resistance and extend its service life of 304 stainless steel, the high temperature wear-resistant NiCrAlY/Co coating was prepared on the surface of 304 stainless steel by the laser cladding.The morphology, phase composition and microhardness of the coating were analyzed.The tribological properties of 304 stainless steel and NiCrAlY/Co coating at different temperatures（the room temperature to 800 ℃) were studied, and the wear mechanism was analyzed. The results show that the coating is metallurgically bonded to the 304 stainless steel substrate; the coating is mainly composed of γ-Co,(Cr, Ni) and AlNi₃ phases; the average microhardness of the coating（303 HV) is about 1.6 times that of the substrate（194 HV); Compared with the substrate, the coating has a smaller friction coefficient at 200-600 ℃,the friction coefficient is comparable at 800 ℃, and the lowest friction coefficient of the coating is 0.5 at 600 ℃.The wear rate of the coating from the room temperature to 800 ℃ is lower than that of the substrate，and the lowest wear rate is 1.91×10^-5 mm³/(N·m) at 400 ℃, which is about 1/3 of the substrate, indicating that the NiCrAlY/Co coating improves the high temperature wear resistance of 304 stainless steel. At medium and low temperatures，the wear mechanism of the substrate is mainly abrasive wear and adhesive wear, and the wear mechanism of the NiCrAlY/Co coating is mainly abrasive wear and gradually slight adhesive wear.At 800 ℃, the wear mechanism of the substrate is plastic deformation, and the wear mechanism of the coating is oxidation wear.

Shot peening process is widely used in the manufacturing process of gears and other basic components, and its own limitations limit the enhancement of the surface integrity of the workpiece. In order to further improve the surface integrity of the workplece, A combination of numerical simulation and experimental methods was utilized to study the effect of two surface composite strengthening processes, such as double shot peening and shot peening-ultrasonic rolling, on the surface integrity of 18CrNiMo7-6 carburization gear steel samples, and mainly analyzed the effect of the two composite strengthening processes on the improvement of surface integrity of the shot peened samples. The results show that the maximum value of the residual compressive stress of the double shot peening sample was 1 359.56 MPa, locates at the depth of 0.08 mm, and the maximum value of the residual compressive stress of the shot peening-ultrasonic rolling peening sample was 1 329.05 MPa,locates at the depth of 0.25 mm. Compare with the single shot peening sample, the surface roughness of the double shot peening sample and the shot peening-ultrasonic rolling sample was 29.42% and 29.42% lower than that of the single shot peening sample. Compare with the single shot peening samples，the surface roughness of the double shot peening samples and shot peening-ultrasonic tumbling peening samples decreased by 29.42% and 62.76%，respectively, the surface microhardness increased by 8.70% and 17.60%, and the standard deviation of the surface node compressive residual stress value decreased by 23.36% and 89.50%. The shot peening-ultrasonic rolling process is more effective in enhancing the surface hardness,thickness of the residual stress layer and uniformity of the residual compressive stress, as well as reducing the surface roughness of the specimens，and can effectively improve the surface integrity of the shot peened samples.

Minimal surfaces are characterized by spatially continuous smoothness, which can effectively avoid the problem of the stress concentration, and have become a significant focus in the research of mechanical metamaterials. A Scherk single-periodic minimal surface (SPMS) was studied, and the spatially continuous smooth geometrical model was generated using a voxel reconstruction technique with multi-software association method. The mechanical behavior and energy absorption characteristics of the single-period Scherk surface structure were investigated using finite element simulation. The impact of the mathematical parameters of the minimal surface on the geometric configuration was examined.Five minimal surfaces with different parameters were established, and the deformation patterns and stress distributions of the five structures were explored under positive and lateral compression conditions. The Scherk surface structures were fabricated by metal printing technology, and quasi-static compression tests were conducted. The results show that the Scherk single-period surface has an obvious negative Poisson ratio effect, and the SPMS structure exhibits X-and V-shaped deformation modes, which can well withstand the external loading effects. The compression tests show the hump phenomenon of the reaction force and displacement curves of the SPMS structure, indicating that the structure has a negative Poisson ratio and negative stiffness property. This property provides a wide space for its application research in large deformation damage and structural energy absorption.

In pressurized-water reactor nuclear power plants, the vessel nozzles of large carbon steel equipment such as reactor pressure vessel (RPV), steam generator (SG), and main pumps are connected to austenitic steel pipes through dissimilar metal welds (DMWs). The thick-walled DMW has material inhomogeneity and complex weld residual stress (WRS), which easily leads to the generation of fatigue or stress corrosion cracks.Firstly, the WRS of DMW in nuclear power plants obtained through international measurements and numerical analyses was investigated. Then, based on a rapid WRS simulation method for volume uniform heating of unit cells, the WRS of DMW in the hot leg of the primary loop (the connecting pipe section from RPV outlet to SG inlet, which is the pipe section with the highest operating parameters in the primary loop pressure boundary) was obtained.The numerical simulation results are consistent with the trend of the fitting envelope curve recommended by the United States, and the overall results can be enveloped by the fitting curve recommended by the United States, indicating that the described rapid WRS simulation method is feasible. The WRS of thick-walled DMW is relatively high, and the stress values at the inner and outer surfaces of the pipe are more conservative than the recommended values of the United States, suggesting that more safety margins can be obtained in actual structural analyses.

To study the response characteristics of the hollow extruded profile of the vehicle after the collision, the generalized incremental stress state-damage model (GISSMO) was introduced and the finite element simulation was carried out. Firstly, based on the test results of 6082-T6 aluminum alloy, the dynamic and static mechanical properties and fracture behavior under different stress states were characterized by the modified Johnson-Cook (MJC) model and DF2016 model respectively. Secondly, the parameter calibration of GISSMO was carried out based on the combination of LS-OPT soft ware and manual optimization. Then, according to the mesh size effect, the mesh size dependence correction was carried out, and the effectiveness of the model and correction were verified by comparison between the experiment and simulation. Finally, the impact simulation analysis of a hollow extruded profile on the side wall of a vehicle body was carried out, and the impact of material damage and fracture on the simulation results was compared. The results show that GISSMO can more accurately reflect the response of profiles under longitudinal impact than without considering the damage and fracture of materials.

In practical operation, the loads borne by electric excavators often exhibit significant non-stationary random characteristics, leading to complex vibration phenomena of the battery pack, which directly affects the safe and reliable operation of the battery pack. To address this issue, the characteristics of road excitation, plunger pump pressure pulsation excitation, and impact excitation on electric excavators under complex working conditions were investigated. The vibration transmission paths under various excitations were analyzed, a dynamic model of the battery pack was established, the vibration characteristics of the battery pack under non-stationary random excitation were revealed, and case studies were conducted for analysis and verification. The research shows that reconstructing road excitation signals based on wavelet transform and Grey Wolf Optimization-Variational Mode Decomposition (GWO-VMD) signal analysis algorithm can effectively reflect the characteristics of road excitation. The road excitation borne by electric excavators under driving conditions exhibits significant non-stationary characteristics. Under non-stationary random excitations such as road excitation, the battery pack of electric excavators produces complex and changeable vibrations, whose power spectral density of dynamic response changes significantly with time, showing obvious non-stationary random characteristics.This study provides a reference for the safe and reliable operation of battery packs in electric excavators.