Xin Sun,Chunying Zhu,Taotao Fu,Youguang Ma,Huai Z. Li
Dynamics for droplet breakup with tunnels and formation of satellite droplets in a symmetric microfluidic T-junction are investigated using a high-speed digital camera. The breakup process of droplet could be divided into four sequential stages: squeezing, transition, pinch-off, and thread rupture stages. Effects of the viscosity ratio of both phases λ , the capillary number of the continuous phase Ca c , and the dimensionless droplet length l 0 / w c on the breakup process of droplet are analyzed. In the squeezing stage controlled by squeezing pressure, the variation of the minimum width of the droplet neck with time could be scaled as a power-law relationship with exponent related to λ and l 0 / w c . The transition stage is dominated by the velocity of the fluid, and the pinch-off stage is mainly driven by the capillary force. For these two stages, the evolution of the minimum width of the droplet neck with the remaining time could be described as other power-law relationship with exponent related to λ . In the thread rupture stage controlled by viscous stresses of both phases and the surface tension, the minimum width of droplet neck decreases linearly with time, and the coefficient of the linear function depends on λ . Additionally, for the formation of satellite droplets, the size of satellite droplet increases with the increase of superficial velocity of the fluid and droplet viscosity. Furthermore, a critical capillary number 0.03 is observed in low viscosity of the droplet, which divides the variation of the size of the main satellite droplet with the capillary number into constant area and growth zone.
Deep learning has recently been leveraged in the microfluidic research for intelligent image processing. However, it usually requires massive human efforts to label sufficient images for obtaining a precise deep learning model. We herein propose an integrated microfluidic image processing method based on transfer deep learning for the rapid deployment of Mask R-CNN, which is capable of reutilizing models trained in previous microfluidic domains to accelerate the analysis of droplets and bubbles in the images of the target microfluidic domain. The transfer deep learning is validated in the microfluidic research with a constrictive microchannel, which is commonly used to break large droplets or bubbles to enhance emulsification. With only 40 labeled images and about 10 min training, the integrated image processing method achieves precise recognition results for both droplets and bubbles in the microfluidic experiments. Applying the trained Mask R-CNN to unlabeled images, the breakup of droplets is found to be highly dependent on the low flow rate, but the bubble breakup is much easier to happen in the same microfluidic equipment. Critical conditions of droplet breakup and size laws of droplet and bubble are precisely determined based on the recognition results of the trained Mask R-CNN.
危险品化学品经营许可证（带存储）