航空发动机滑油回油管内的流型识别及含气率预测研究

Flow pattern recognition and void fraction prediction in lubricating oil recirculating pipe of aeroengine

关键词:  润滑系统  回油管  流型识别  压力特征  含气率预测  分布系数修正

Key words:  Lubrication system  Recirculating pipe  Flow pattern recognition  Pressure characteristics  Void fraction prediction  Distribution coefficient correction

摘要:

发动机润滑系统回油管内油气两相介质的流动特性直接影响系统中回油泵及散热器的工作特性。为辨别回油管内油气两相流流型和建立含气率预测关系式，本文搭建了模拟轴承腔回油管流动的实验系统，并依据涡轴发动机回油量及油气比完成了700组不同工况的实验测试。实验空气与滑油流量范围分别为10~200 SL/min与 6~37 L/min。本文对实验范围内的管内流型进行了判别，并修正了含气率的预测模型。首先对两种极限工况的压力信号的时序特征、概率密度函数和傅里叶变换进行分析，判定其流型为分层流和弹状流。将流型结果与Mandhane流型图进行对比，显示 Mandhane流型图并不能准确预测回油管内的油气两相流流型。其次利用含气率经验关系式进行含气率预测，发现Massena预测模型预测的含气率与实验值更接近，预测结果更准确。分层流的含气率在50%以下，而弹状流的含气率在50%以上。最后根据流型对Nicklin经验关系式进行修正，得到分层流和弹状流对应的分布系数分别为0.848和0.919。

Abstract:

The flow characteristics of oil-gas in oil recirculating pipe of engine lubrication system directly affect the working characteristics of oil return pump and radiator. In order to identify the two-phase flow pattern of oil and gas in the return pipeline and establish the relationship between gas content prediction， an experimental system is built to simulate the oil recirculating pipe flow in bearing chamber. According to the oil return amount and oil-gas ratio of the turboshaft engine， 700 sets of different experiments are completed. The experimental air flow and lubricating oil flow ranges were 10~200 SL/min and 6~37 L/min， respectively. The flow patterns inside the recirculating pipe within the experimental range are discriminated. And the prediction model of void fraction is established. Firstly， the time series characteristics， probability density function and Fourier transform of the pressure signals under two extreme conditions are analyzed. And the flow patterns are determined to be stratified flow and slug flow. Comparing the flow pattern results with the Mandhane flow pattern diagram， it shows that the Mandhane flow pattern diagram cannot accurately predict the oil-gas two-phase flow pattern in the recirculating pipe. Secondly， the empirical relationship of void fraction is used to predict the void fraction. It is found that the void fraction predicted by the Massena prediction model is closer to the experimental value and the prediction result is more accurate. The void fraction of stratified flow is less than 50%， the void fraction of slug flow is more than 50%. Finally， the Nicklin empirical correlation is modified according to the flow pattern. And the distribution coefficients of stratified flow and slug flow are 0.848 and 0.919， respectively.