In response to the need for multiple complete bearing degradation datasets in traditional deep learning networks to predict the impact on individual bearings, a novel deep learning-based rolling bearing remaining life prediction method is proposed in the absence of fully degraded bearng data. This method involves processing the raw vibration data through Channel-wise Attention Encoder (CAE) from the Encoder-Channel Attention (ECA), extracting features related to mutual correlation and relevance, selecting the desired characteristics, and incorporating the selected features into the constructed Autoformer-based time prediction model to forecast the degradation trend of bearings’ remaining time. The feature extraction method proposed in this approach outperforms CAE and multilayer perceptual-Attention Encoder in terms of feature extraction capabilities, resulting in reductions of 0.0059 and 0.0402 in mean square error, respectively. Additionally, the indirect prediction approach for the degradation trend of the target bearing demonstrates higher accuracy compared to Informer and Transformer models, with mean square error reductions of 0.3352 and 0.1174, respectively. This suggests that the combined deep learning model proposed in this paper for predicting rolling bearing life may be a more effective life prediction method deserving further research and application.
Wire and arc additive manufacturing of stiffened ribs for fabricating large integral panel components holds promise for lightweight construction, but stress and warpage remain intractable problems. This work take stress in central lines as a research focus, theoretically analyzes stress evolution along central lines without a stress distribution hypothesis, then further develops prediction models of warpage, based on general beam theory. The stress and warpage models are then verified by experimental results. The results showed that below 40 layers, the stress in the top layer increased rapidly to the yield strength. Following the continuous deposition of up to 100 layers, stress increased slowly, then gradually trended to a constant level. After 100 layers of deposition, the top-layer-plane stress in the 10th layer decreased from 68 MPa to 5 MPa, and the plane stress in the 90th layer decreased from 296 MPa to 245 MPa. Post-heating treatment effect was important for temperature uniformity. The largest warpage (3.48 mm) happened when stiffener was designed as 30 layers.
High aspect ratio tungsten nanowires have been prepared by selective dissolution of Nickel-aluminum-tungsten (NiAl−W) alloys which were directionally solidified at growth rates varying from 2 to 25 μm/s with a temperature gradient of 300 K·cm−1. Young’s modulus and electrical resistivity of tungsten nanowires were measured by metallic mask template method. The results show that the tungsten nanowires with uniform diameter and high aspect ratio are well aligned. The length of tungsten nanowires increases with prolongation of etching time, and their length reaches 300 μm at 14 h. Young’s modulus of tungsten nanowires is estimated by Hertz and Sneddon models. The Sneddon model is proper for estimating the Young’s modulus, and the value of calculating Young’s modulus are 260–460 GPa which approach the value of bulk tungsten. The resistivity of tungsten nanowires is measured and fitted with Fuchs−Sondheimer (FS) + Mayadas−Shatzkes (MS) model. The fitting results show that the specific resistivity of W nanowires is a litter bigger than the bulk W, and its value decreases with decreasing diameter.
Molten pool uniformity is a prerequisite for wire and arc additive manufacture to achieve a uniform surface appearance. The thermal boundary is a key indicator to evaluate the thermodynamic state of the molten pool. This paper focuses on thermal analysis through finite element simulation and characterizes the thermal evolution of a molten pool during bottom-up deposition. The peak temperature of the substrate plate increases from 375.7 °C to 623.1 °C when peak current increases from 120 A to 180 A. The temperature gradient decreases from 40 °C/mm to 30 °C/mm. Weld speed increases from 0.15 m/min to 0.25 m/min; the temperature gradient is kept at about 30 °C/mm. Dimensionless thermal conductivity, Ge number, is proposed and defined as the criteria to estimate layer size variation. For Ip = 140 A, the average value of the Ge number is 87.7, and it shows the best numerical stability. Dimensionless thermal conductivity, Ge, is proposed and defined as the criteria to estimate the occurrence of layer size variation. Prolonging heating time and increasing cycle frequency, Ge number shows favorable stability during bottom-up deposition. The experimental results are consistent with analytical conclusions, which proves the validity of the Ge number as an indicator of deposition stability.