Accurate detection of surface defects for steel is essential to improve surface quality and service life. Deep learning (DL) used in steel surface defect detection can solve the problems of low efficiency and poor accuracy of traditional manual detection. The classic YOLOv5 as a DL method is used to accomplish defect detection tasks without attention mechanisms, resulting in a loss of global information. Besides, it is difficult to complete complex network detection tasks with low-configuration hardware, especially for surface defects with complex defect types and variable defect sizes. To solve these issues, this paper introduces an improved global feature reuse and hardware-aware YOLOv5 by using BoTNet, RepGhost, and EfficientRep model (BGE-YOLOv5). The multi-head self-attention layer is used to obtain global information and only part of the convolutional layers is replaced to avoid the excessive computational cost. The RepGhost model is introduced to extract the remaining feature information for feature reuse. EfficientRep is used to replace the original structure to achieve hardware-aware and to balance the detection veracity and efficiency. The distance IOU is replaced by SCYLLA-IOU to accelerate the iteration and improve stability. The results of the framework on the surface defect database (NEU-DET) show that BGE-YOLOv5 achieves a mean average precision of 79.5%, which is 10.3% greater than the baseline. The proposed BGE-YOLOv5 has a better performance in steel surface defect detection.

The predictions of mechanical responses (stress–strain variations) in the machining of Ti-6Al-4V alloy are important to analyze the deformation conditions of machining to optimize the machining parameters and investigate the generation of a machined surface. The selection of a constitutive model is an essential factor that determines the deformation behavior in the machining simulation model. In this paper, two constitutive models of a modified Johnson–Cook (JC) equation and visco-plastic self-consistent (VPSC) model were used to investigate the stress–strain evolutions in the machining process of Ti-6Al-4V. A finite element (FE) machining model was established, considering the influences of grain refinement and deformation twins, based on a modified JC equation. The VPSC model was fitted based on the macro-strain rate sensitivity of the JC equation. The prediction results of the stress–strain curves of two models were compared, and their validities were further proved. The results show that flow stress hardening and inhomogeneities are caused by multi-scale grain refinement during the machining process of Ti-6Al-4V. Five slip deformation modes and one compressive twinning mode were activated in the VPSC model to be consistent with the macro-deformation behavior predicted with the FE model. The validations show the effectiveness of the modified JC equation, considering microstructural changes and the fitted VPSC model, in predicting dynamic behavior in the machining process of Ti-6Al-4V. The results provide two aspects of macro-deformation and polycrystal plasticity to elucidate the stress variations that occur during the machining of Ti-6Al-4V.

The top beam of a hydraulic support, serving as the central component of the hydraulic support assembly, directly withstands the impact loads generated by the subsidence of overlying rock strata. The structural strength and design rationality of the top beam directly determine the support effectiveness of the hydraulic support. In response to this issue, this study focuses on newly developed hydraulic combined supports for tunneling operations. Transient dynamic simulations were conducted at various positions on the top beams of individual supports under different working conditions, yielding stress and displacement changes in various regions of the support canopy when subjected to impact. We obtained the dynamic characteristics of the key parts of the hydraulic unit support during load bearing stress. Future work will explore the hydraulic system model and the interaction between mechanical and hydraulic subsystems.