Search

In order to improve the sealing performance of the high-speed gearbox sealing device， this study analyzes the non-contact labyrinth sealing of the high-speed gear transmission commonly used in petroleum， chemical and other fields. By establishing a numerical calculation model， the influence of factors such as the number of sealing teeth sealing differential pressure， oil-air mixture concentration，and position of oil slinger on the leakage of oil return ports and oil leakage ports of each sealing cavity is analyzed， which has guiding significance for the design of non-contact labyrinth sealing for high-speed gearboxes.

There is a new tooth failure mode for hardened gears called tooth flank fracture(TFF). ISO published a technical specification for calculating the load capacity of tooth flank failure in 2019. This failure mode is a fatigue phenomenon with initial crack initiating in the transition zone between the case hardened layer and the tooth core. Its failure mechanism is different from the traditional pitting and tooth root breakage fatigue failure modes. The main characteristics，research status，calculation methods and main influencing factors of tooth flank fracture failure are summarized. It provides a theoretical basis for understanding tooth flank fracture failure.

The basic concepts and main characteristics of the oil whirl and oil whip are introduced， and the oil whip may cause serious damage to the unit， which must be paid great attention to and dealt with in time. Through the analysis of the frequency spectrogram， the orbit and the operating parameters of the fault unit， the relationship between them is pointed out. At last， the common measures to deal with the problem of oil whip are summarized， which provides a reference to deal with these problems.

Most of existing deep learning-based methods for image splicing forgery detection use convolutional layer for forensics feature extraction.However,convolution kernel conducts the local computation process with the limited reception field.More-over,existing methods mainly apply the location of tampering regions to guide the detection model to train,and it is difficult to learn richer tamper trace features.To overcome above-mentioned limitations,a multitask transformer-based network(MT-Net) is proposed for image splicing detection and localization.The self-attention mechanism of Transformer is leveraged in encoder to learn the pixel correlation,which is able to provide different attention levels for pixels and makes the detection network pay more attention to tampering traces.Meanwhile,MT-Net considers three subtasks simultaneously to guide the detection network expose tampering traces from both local and global information,including tampered edge detection,tampered area detection and the prediction of the tampered area's proportion.Finally,three specific loss functions for their corresponding subtask are designed to better optimize the detection network in the training phase.In experiments,the proposed method(MT-Net) achieves better detection results compared with other state-of-the-art methods on three public available datasets,including CASIA v2.0,Columbia and IDM2020,where F1 scores are 0.808,0.913 and 0.675 respectively.The visualization results also demonstrate that the proposed method has the better capability of localizing the splicing regions.