Sai Ma

Piezoelectric drop-on-demand inkjet printing technology is widely applied in the field of additive manufacturing due to its non-contact deposition, high-precision formation, and excellent compatibility with diverse functional inks. As application requirements gradually extend from planar substrates to complex curved and undulating surfaces, increasing the printing distance has become a key factor in enabling conformal printing. However, under long-distance printing conditions, the existing process is prone to droplet landing deviation and fluctuations in formation quality, thereby limiting the attainable stable working distance. In this paper, the superposition characteristics of flow field oscillations inside the piezoelectric printhead system were analyzed through experiments and simulations, and a method for increasing the printing distance was proposed. A multi-waveform superposition strategy was employed to achieve secondary regulation of the liquid-column formation process during the jetting stage. The results demonstrated that, under the newly designed waveform conditions, the printhead used in this study, with a nominal printing distance of less than 2 mm, could achieve controllable and stable single-droplet jetting even at a printing distance of up to 11 mm. This represented an increase in the printing distance by more than fivefold. In addition, various micro-patterns can be successfully fabricated by controlling the spacing between deposited droplets. This work provides a new strategy for achieving stable long-distance inkjet printing.

Diagnostic models based on deep transfer learning hold the potential to apply diagnostic knowledge across relevant machinery. However, existing methods suffer from several drawbacks. First, the random initialization of convolutional kernels lacks interpretability and may lead to suboptimal solutions, affecting diagnostic accuracy and training convergence. Second, treating all channels equally limits the ability of the model to capture crucial information. Third, traditional domain alignment methods do not effectively utilize high-order features for matching global or local feature information. To address these issues, a novel domain adaptation method named multiorder statistics matching sparse wavelet convolutional neural network (MSM-SWCNN) is proposed in this article. In MSM-SWCNN, a multichannel feature extraction module is proposed to utilize sparse wavelet convolutional kernels for comprehensive feature extraction from both time and frequency domain. In addition, a novel generalized distribution matching measure is proposed to align features and distribution between two domains. For eight transfer tasks in two datasets of bearings and gears, the diagnostic accuracy can reach more than 96.6%. The results demonstrate that the proposed method outperforms other methods for rotating machinery fault diagnosis.

Diagnostic models based on deep transfer learning hold the potential to apply diagnostic knowledge across relevant machinery. However, the existing methods suffer from several drawbacks: 1) most methods use wavelet convolutional kernels to extract features, but the extracted features are redundant and may diminish the diagnostic accuracy; 2) traditional domain alignment methods cannot effectively utilize both low- and high-order measure to match global or local feature information; and 3) contrastive learning in transfer learning requires the assignment of labels to the target domain, but target domain pseudo-labels may not be reliable. To address these issues, a novel domain adaptation method named time–frequency adaptive contrastive neural network (TFACNN) is proposed. This method uses wavelet convolutional kernels to extract features from the time–frequency domain. It then uses sparse constraint and kurtosis constraint to extract features corresponding to the fault frequencies. It utilizes a generalized distribution matching measure to align features, minimizing distribution differences between the two domains. Additionally, an adaptive threshold is designed to ensure the quality of target domain pseudo-labels, thereby improving the performance of contrastive learning. The proposed method is validated on two different datasets, which include rotating machinery fault diagnosis dataset and turning process monitoring dataset, demonstrating its superiority over other methods for rotating machinery fault diagnosis.