The vibration signal of mechanical equipment in operating environments is the key to describing fault characteristics, but due to thez influence of equipment density and environmental interference, the accuracy of fault diagnosis is often affected by noise. In this paper, a fault diagnosis method based on a 1D Multi-Channel Improved Convolutional Neural Network (1DMCICNN) is proposed. By introducing BiLSTM, an attention mechanism and a local sparse structure of a two-channel Convolutional Neural Network, the feature information of the noisy timing signal is fully extracted at different scales while reducing the computational parameters. The model is verified through experiments under different signal-to-noise ratios and loads. The results show that the accuracy of 1DMCICNN is 98.67%, 99.71%, 99.04%, and 99.71% on different load and speed datasets. Meanwhile, compared with the unoptimized two-channel Convolutional Neural Network, the training parameters are reduced by 55.58%.

The contact fatigue performance of carburized gear steels is critical for transmission durability, yet the mechanisms linking alloy-specific microstructure to failure modes remain complex. This study systematically compares the contact fatigue behaviors of 20MnCr5 and 20CrMoH gears using step-loading tests and multi-scale characterization. The results demonstrate a significantly higher contact fatigue limit for 20MnCr5 of 1709 ± 12 MPa compared to 1652 ± 40 MPa for 20CrMoH, despite the latter exhibiting higher initial surface hardness. This hardness–toughness paradox is mechanistically elucidated by the distinct roles of alloying elements: while Molybdenum in 20CrMoH refines the grain size for high static strength, it limits retained austenite stability, resulting in a brittle hard-shell and soft-core structure prone to interface decohesion at martensite lath boundaries. Conversely, Manganese in 20MnCr5 promotes a gentler hardness gradient via favorable diffusion kinetics and stabilizes abundant film-like retained austenite. This microstructure activates a Stress Compensation Mechanism, where strain-induced martensitic transformation generates compressive volume expansion to counteract cyclic stress relaxation. Consequently, 20MnCr5 exhibits mild plastic micropitting driven by transformation toughening, whereas 20CrMoH undergoes severe brittle spalling driven by the Eggshell Effect. These findings confirm that balancing matrix toughness with hardness is more critical than maximizing surface hardness alone for contact fatigue resistance.