With cores of the hexagonal honeycomb (HH), staggered corrugated lattice (SCL), bidirectional corrugated lattice (BCL), reinforced hexagonal honeycomb (RHH), and reinforced staggered corrugated lattice (RSCL) made by cutting, bonding, arranging and weaving the trapezoidal corrugation (TC) formed by mold hot pressing, six types of composite sandwich panels of the same macro size were proposed. The failure processes of the six sandwich panels in three-point bending experiments were compared and analyzed, and the effects of core configuration and reinforced stiffener on the bending response were evaluated. A general analytical model was established to predict the bending stiffness and failure mode of six sandwich panels, which was also visually presented by a three-dimensional numerical model. The numerical and experimental results validated the effectiveness of the analytical model. The effects of ply parameters and corrugated sizes on bending responses were studied by a combing application of experiment, analytical and numerical models. The three-dimensional failure mechanism maps were displayed to evaluate the failure mode and failure load. The longitudinal discontinuity of the bonding zone of the core could significantly weaken the bending performance and bonding strength of BCL, SCL, RSCL, and HH. The peak load of SCL and HH was enhanced by 45.97% and 60.29%, respectively, by setting stiffeners. Both RHH and TC could achieve bending stiffness of more than 3.5 kN·mm−1, while the failure load of the latter could be up to 25 kN. The design parameter space of bending stiffness, failure load with respect to ply orientation angle, core wall thickness, corrugated web angle, and horizontal segment length was provided intuitively.
The failure modes, ultimate load, stiffness performance, and their influencing factors of a composite sandwich laminated box beam under three-point bending load are studied by an experiment, finite element model, and analytical method. The three-point bending experiment was carried out on three different core composite sandwich laminated box beams, and the failure modes and bearing capacity were studied. With the use of composite progressive damage analysis and the core elastoplastic constitutive model, the finite element model of the composite sandwich laminated box beam was established, and the three-point bending failure process and failure modes were analyzed. The analytical model was established based on the Timoshenko beam theory. The overall bending stiffness and shear stiffness of the composite sandwich laminated box beam were calculated by the internal force–displacement relationship. The results show that the composite sandwich laminated box beam mainly suffers from local crushing failure, and the errors between the finite element simulation and the experiment result were within 7%. The analytical model of the composite sandwich laminated box beam can approximately predict the overall stiffness parameters, while the maximum error between theoretic results and experimental values was 5.2%. For composite aluminum honeycomb sandwich laminated box beams with a ratio of span to height less than 10, the additional deflection caused by shear deformation has an error of more than 25%. With the ratio of circumferential layers to longitudinal layers increasing, the three-point bending ultimate load of the composite sandwich laminated box beam increases, but the ratio of the overall stiffness to mass reduces. The use of low-density aluminum foam and smaller-wall-thickness cell aluminum honeycombs allows for the more obvious benefits of light weight.
The double-arrow hollow cylinder (DAHC) has excellent mechanical properties under axial compression loads due to its negative Poisson ratio structure. To improve finite element simulation efficiency for parameter optimization analysis, an automatic modeling program for DAHC negative Poisson ratio structures was developed on the ABAQUS platform using Python. The sample was manufactured using selective laser melting additive manufacturing based on the solid model and then simulated using the finite element method. The accuracy of the automatic modeling method was confirmed by comparing load–displacement curves and deformation cloud images with static compression test results. The DAHC automatic modeling program based on Python can reduce workload during parameter optimization analysis and improve finite element analysis efficiency.
Aiming to tackle the challenge of feature transfer in cross-domain fault diagnosis for rolling bearings, an enhanced domain adaptation-based intelligent fault diagnosis method is proposed. This method systematically combines multi-layer multi-core MMD with adversarial domain classification. Specifically, we will extend alignment to multiple network layers, while previous work typically applied MMD to fewer layers or used single core variants. Initially, a one-dimensional convolutional neural network (1D-CNN) is utilized to extract features from both the source and target domains, thereby enhancing the diagnostic model’s cross-domain adaptability through shared feature learning. Subsequently, to address the distribution differences in feature extraction, the multi-layer multi-kernel maximum mean discrepancy (ML-MK MMD) method is employed to quantify the distribution disparity between the source and target domain features, with the objective of extracting domain-invariant features. Moreover, to further mitigate domain shift, a novel loss function is developed by integrating ML-MK MMD with a domain classifier loss, which optimizes the alignment of feature distributions between the two domains. Ultimately, testing on target domain samples demonstrates that the proposed method effectively extracts domain-invariant features, significantly reduces the distribution gap between the source and target domains, and thereby enhances cross-domain diagnostic performance.