Archives of Civil and Mechanical Engineering最新文献

Carbon fiber reinforced concrete (CFRC) often suffers from weak bonding between its cement matrix and carbon fibers (CF) because the fibers have an inert surface.To overcome this limitation, we introduce phenylpropyl emulsion (SAE) as an interfacial modifier and examine its strengthening mechanism through macroscopic and microscopic experiments and multiscale simulations using discrete element and molecular dynamics methods. Macroscopic experiments show that both CF and SAE significantly improve the mechanical properties of CFRC, although the effectiveness of SAE depends on its dosage. Specifically, a 4% SAE dosage increases the compressive and shear strengths of CFRC (with 1.5% CF) by 16.5% and 12.1%, respectively, while a 2% dosage boosts early flexural strength by 24.4%. At the microscopic level, SAE enhances the bond at the interface by providing physical anchorage through benzene rings and by forming chemical interactions via COO⁻-Ca²⁺complexes. Discrete element simulations confirm that the modified material has higher ultimate strength and toughness, and molecular dynamics simulations show that SAE lowers the interfacial energy and strengthens the bond between CF and the matrix through ionic and hydrogen bonding. This work offers a solid theoretical basis for improving CFRC interfaces and informs the design of innovative polymer modifiers.

Graphical abstract

The safety and durability of reinforced concrete (RC) structures face significant impacts from the combined action of corrosion conditions and fatigue stress. To determine the corrosion-fatigue life of RC structures under combined corrosion environment and fatigue loading, a corrosion-fatigue life assessment model for RC structures under the combined action of carbonation environment, chloride-contaminated environment, and fatigue loading, aiming to analyze their corrosion-fatigue life, was developed. Taking a 6 m-span RC crane beam in an industrial plant as an example, the proposed model was illustrated, and its corrosion-fatigue life was evaluated. Effects of crane operating frequency, lifting capacity, carbonation environment grade, chloride-contaminated environment grade, ambient temperature, and relative humidity on corrosion-fatigue life were explored. Results indicate that increased loading frequency and stress range shorten the corrosion-fatigue life of RC structures. The combined action of corrosion and fatigue notably reduces corrosion-fatigue life. Compared with single carbonation or chloride-contaminated environments, the combined carbonation and chloride-contaminated environment has a stronger effect on the corrosion-fatigue performance of RC structures. Additionally, ambient temperature and relative humidity influence the corrosion-fatigue life of RC structures. This study can establish a basic framework for assessing the corrosion-fatigue life of RC structures under combined multi-corrosion environments and fatigue loading.

To enhance the real-time sensing and explosion location recognition capabilities in underground engineering, this study investigates structural vibration responses under different detonation source locations through model testing and numerical simulations. By integrating the simplex method with an improved particle swarm optimization algorithm, a novel localization algorithm is proposed, leveraging the moment differences of peak acceleration. The algorithm’s accuracy is validated through model tests and numerical results. The findings reveal that when the detonation source is located in the middle and far field, significant dispersion exists between the distance from the measurement point to the detonation source and the peak acceleration. While the general trend shows an increase in distance correlating with a decrease in peak acceleration, a negative correlation is observed only at very short distances. Numerical simulations confirm that the time of peak vibration acceleration and detonation source distance exhibit a strong linear positive correlation, with the coefficient ranging between 0.00218 and 0.00336. Furthermore, the larger the distance difference, the greater the time difference in peak acceleration. Peak vibration acceleration time as a principle, combined with the simplex method and improved particle swarm optimization algorithm can be inverse performance and the real detonation source coordinates similar to the distance of the detonation source coordinates. And the improved particle swarm algorithm of detonation source location calculation results are more accurate, based on the numerical simulation data positioning accuracy analysis results show that its accuracy is stable at more than 99%.

Bone regeneration remains a persistent clinical challenge, with optimal therapeutic strategies yet to be established. Magnesium alloys reinforced with bioactive particles represent an emerging biomaterial class demonstrating favorable tribological properties and inherent osteogenic potential. The study presents a comparative investigation of the dry reciprocating wear behavior of AZ91D magnesium alloy and its friction stir processed surface composites, each reinforced with 15 wt% of bioactive nanoparticles: HA, ZrO₂, and Y₂O₃. Wear tests were conducted using a ball-on-disc tribometer under a 5 N load, 5 Hz frequency, and 5 mm stroke length for 900 s against ZrO₂, Si₂N₄, and SS440 counterface balls. Among the composites, the AZ91D matrix reinforced with ZrO₂ exhibited the lowest wear volume, showing reductions of 24.2% and 38.7% compared to the Y₂O₃ and HA-reinforcement, respectively, along with a relatively lower coefficient of friction. This synergistic enhancement is attributed to the uniform distribution of fine reinforcement, which suppressed adhesive and delamination wear while increasing load-bearing capacity during repeated sliding. Microstructural and surface characterizations using SEM, EDS, and XRD confirmed a transition in the dominant wear mechanism from adhesive and abrasive modes in the monolithic alloy to more abrasive and erosive mechanisms in the composites. The presence of reinforcement and in situ formation of protective magnesium-based oxides played a key role in this shift. The novelty of this work lies in revealing the tailored tribological behavior of AZ91D through bioactive reinforcement via friction stir processing for improved wear resistance in magnesium-based composites.

Hot-extruded aluminum alloy profiles with complex cross-sectional characteristics are widely used in new energy vehicles, rail transportation and aerospace industries. By optimizing the structure of die, a uniform flow rate at the profile die outlet is obtained, thereby reducing the possibility of defects. Subsequently, the extrusion transient simulation is performed by setting the billet skin. The two flow modes of the billet skin during the extrusion process are explained through transient simulation, and the front- end defect of the profile and the rear-end defect of the billet are simulated and determined. An extrusion experiment is carried out based on the process parameters in the simulation, and EBSD tests are performed on the as-cast billet, the as-cast billet after homogenization treatment, the rear-end defect, the front-end defect and the normal profile. By analyzing the ODF of the as-cast billet, it is found that the billet cast by low-frequency electromagnetic casting formed a <110> // ND texture and <110> // TD texture due to characteristics similar to directional solidification. Homogenization treatment can effectively alleviate the segregation and slightly increase the grain size. The flow pattern of the billet skin in the simulation is verified by analyzing the grain size and fiber texture of the back-end defect. By analyzing the grain size of the profile and combining it with simulation, the extrusion coarse-grained ring effect is revealed. By comparing the front-end defect and the texture of the profile, the reason why the front-end defect is difficult to reach the normal profile usage standards is explained from a microscopic perspective.

In this study, a novel hybrid manufacturing process (HMP) is proposed to improve the global yield strength and hardness of 316L stainless steel (316L SS) by integrating interlayer laser directed energy deposition (LDED) additive, milling subtractive, and ultrasonic rolling (UR) equivalent manufacturing processes. The results showed that the porosity of 316L SS subjected to HMP decreased by 96.4% to 0.02%, grain size refined by 58.9% to 20.2 μm, dislocation density increased by 145.5% to 2.7 × 10¹⁵ m^-2, hardness increased by 25.5% to 269.1 HV_0.1, yield strength improved by 41.4% to 670.3 MPa, and ultimate tensile strength improved by 26.1% to 746.7 MPa than those obtained by the single LDED process. The dislocation strengthening and grain refinement strengthening were responsible for the enhanced global yield strength and hardness of 316L SS. The HMP can be applied to realize the defect-free and high-performance manufacturing of metals by eliminating internal defects, reconstructing gradient structures of grain sizes, dislocation densities, grain boundary angles, texture intensities, and twins.

High-strength engineered cementitious composites (ECC) with polyethylene fiber exhibits superior strength characteristics; however, its crack width is significantly larger, which compromises both its transmission performance and self-healing capability compared to conventional ECC with polyvinyl alcohol. To this end, this study prepared a lightweight high-strength ECC by synergistically utilizing fly ash cenospheres (FAC) and expanded vermiculite (EV) as lightweight fillers. The tensile and compressive properties of ECC with different contents of FAC and EV were examined experimentally, and its self-healing behavior was verified through wet-dry cycling tests. Additionally, the microstructure of the matrix was analyzed using scanning electron microscopy. The results indicated that, compared to reference ECC with only FAC (FE0), the incorporation of EV led to a maximum increase in density of 14%, while the compressive strength decreased by up to 9%. Despite this, the maximum density reached 1777 kg/m³, and the minimum compressive strength was 65.8 MPa, still placing it within the category of lightweight, high-strength ECC. The number of cracks increased from 21 in FE0 to 123 in FE75 (a combination of EV and FAC), while the crack width decreased from 75 μm to 20 μm. These narrow cracks contributed to the near-complete self-healing of cracks within the tensile specimens, except for those that led to ultimate failure.

A novel cold-formed steel (CFS) built-up closed L-section column (CSBC-L) is proposed, which consists of two irregular sections connected with screw connections and can be used for mid-rise CFS buildings. This paper investigates the nonlinear behavior of the CSBC-L under axial compression. A numerical model is developed considering initial geometric imperfections, material nonlinearity, and the interaction performance between screws and components. The numerical model accurately predicts the ultimate capacity, buckling mode, and stiffness degradation phenomena well. The maximum errors in the ultimate capacity between the finite element simulation and experimental test results were no more than 11.8%. Furthermore, the effects of various parameters, such as the slenderness ratio, height-thickness ratio of the web, and screw spacing, on the axial performance of the CSBC-L specimens are investigated through parametric analyses. The ultimate capacity is significantly affected by the slenderness ratio and web height-thickness ratio. With varying screw spacing from 150 to 750 mm, most specimens demonstrate adequate jointing effectiveness, with most specimens showing a reduction of up to 3% with increasing screw spacing. The experimental and numerical results are compared with the predicted ultimate capacity calculated by the current codes. The direct strength method (DSM) specified in the current AISI standards is overly conservative for predicting the ultimate capacity of built-up closed L-section columns. Finally, correction factors are proposed to modify the current calculation. The maximum error for the modified calculation is changed from 25% to no more than 16%, and it can be used for the compressive design of the CSBC-L.

Based on hot compression experiments and grain growth experiments of 6063 aluminum alloy, this study established a microstructure model for 6063 aluminum alloy. The accuracy of the developed model was verified by embedding it into hot compression numerical simulations through subroutine secondary development. Subsequently, the microstructure model was applied to the porthole die extrusion of complex cross-sectional profiles to investigate microstructure evolution during profile formation and analyze the influence of process parameters on microstructure. The results demonstrate that the established microstructure model can accurately predict the microstructural evolution of 6063 aluminum alloy during hot deformation. The results indicate that the established microstructure model can accurately predict the microstructure of the 6063 aluminum alloy. During the porthole die extrusion of complex cross-sections, the dynamic recrystallization (DRX) volume fractions from the surface to the interior of the profile are 92.44%, 73.03%, and 66.11%, respectively, with grain sizes of 28.57 μm, 42.80 μm, and 47.92 μm. At the welds, the DRX volume fraction is 97.67%, and the grain size is 29.49 μm. As the billet temperature increases from 490 to 540 °C, the average DRX volume fraction of the profile cross-section increases from 81% to 83.9%, and the average grain size increases from 37.39 to 38.67 μm. When the ram speed decreases from 4 to 0.4 mm/s, the average DRX volume fraction increases from 81% to 82.57%, and the average grain size increases from 37.39 to 41.88 μm. As the extrusion ratio decreases from 61.66 to 44.57, the average DRX volume fraction decreases from 81% to 79.09%, and the average grain size increases from 37.39 to 38.45 μm. This research provides theoretical guidance at the microstructural level for the production of aluminum alloy profiles.

Within the scope of linear-elastic fracture mechanics (LEFM), several dimensionless parameters, known as geometry factors, exist. Calculating geometry factors is a time-consuming and computationally resource-intensive process. This study evaluates the predictive abilities of various machine learning models for dimensionless fracture parameters (Y_I, Y_II, T*). In this regard, three fracture specimens were considered, including single-edge notched bend (SENB), semi-circular bend (SCB), and edge-notched disc bend (ENDB). The evaluations were conducted under a full range of Mode I/II brittle fracture conditions. As the first step, the datasets were statistically described, including assessment of the distributions, correlation of parameters, and examination of the dataset’s sensitivity to input variables. Using a diverse dataset, the study tests linear, Lasso, Ridge, partial least squares, decision tree, artificial neural networks, random forest, gradient boosting machine (GBM), boosted tree, support vector regression, and Gaussian process regression models. Different train/test ratios were assumed to evaluate the dependency of models on the train ratio.