Identification Of Boiling Flow Patterns In Triangular Pit Microchannels And Their Heat Transfer Characteristics

In recent years,flow boiling in microchannels has received a lot of attention because of its high heat dissipation rate and ability to keep system temperatures within safe operating ranges.The artificial cavity can promote the nucleation of bubbles,improve the uneven temperature distribution of the wall and reduce the flow boiling instability at low wall superheat.At the same time,the cavity will have an effect on the bubble behavior in the microchannel,and then affect the overall flow boiling heat transfer performance of the microchannel.Further exploration of the evolution characteristics of the bubble behavior in the cavity microchannel will help to better understand and understand the mechanism of flow boiling heat transfer in the cavity microchannel.Aiming at the phenomenon of flow boiling in triangular cavity microchannels,this paper uses the VOF（volume of fluid）model and user-defined functions to numerically simulate the behavior of bubbles and flow boiling heat transfer characteristics in triangular cavity microchannels.The dynamic evolution process of bubble nucleation and growth,fusion and detachment in the cavity microchannel,and the reasons for the enhanced boiling heat transfer in the triangular cavity microchannel are analyzed.The research results show that compared with the straight microchannel,the triangular cavity structure increases the vaporization core,which is beneficial to the nucleation of the vapor bubble and reduces the boiling temperature.At the same time,the triangular cavity structure can shorten the bubble detachment cycle,reduce the bubble detachment diameter,weaken the local drying phenomenon,and help improve the stability of flow boiling heat transfer.Comparing the depths of different triangular cavity structures（the width of the cavity remains unchanged,the depth of the cavity is increased）,the effect of the structure form on the boiling flow pattern and heat transfer performance is studied.The results show that increasing the depth of the triangular cavity can improve the heat transfer performance to a certain extent,but it also increases the pressure loss.A neural network was used to identify the vapor-liquid two-phase flow pattern in the triangular cavity microchannel.According to the flow pattern obtained by numerical calculation,the relevant feature data is extracted and processed,the corresponding neural network model is constructed,the training samples are input into the network for training,and the overall performance of the network is tested.The results showed that the recognition success rates of BP,RBF,GA-BP,and GA-RBF neural networks on the test set flow patterns were 92%,88%,96%,and 92%,respectively.The flow pattern recognition of the four kinds of neural networks has good accuracy,which provides a new idea for the flow pattern recognition in the cavity microchannel.