基于超声多普勒与电导环的油水两相流流速测量

两相流动现象广泛存在于众多工业领域中，其流动过程参数如流速的准确测量对量化体积/质量流量及优化生产工艺和过程设备有重要意义。针对水平油水两相流流速测量问题，提出了一种同侧双晶连续波超声多普勒（CWUD）与电导环传感器相结合的测量方法。非侵入式超声多普勒传感器为双晶超声换能器，由2个倾角相同且中心频率为1 MHz的压电陶瓷晶片组成，两者之间使用隔声材料防止声波干扰，其中发射晶片向流体连续发射超声波，同时接收晶片接收经流体中离散液滴散射的超声波，测量区间覆盖管道横截面的整个径向范围。动态实验在50 mm管径的水平油水两相流装置上完成，通过分析油水两相流多普勒频移响应特性，发现在测量区间内，平均多普勒频移与总表观流速之间随连续相不同而呈现2种线性关系。因此，根据电导环传感器的电学敏感原理，获得无量纲电压参数判断两相流的连续相，继而选取相应流动状态下的测量模型，计算流体总表观流速。实验结果表明:总表观流速估计值均方根误差为0.01 m/s，平均相对误差为3.09%，其中相对误差小于5%的置信概率为70%。 

Oil-water two-phase flow velocity measurement based on ultrasonic Doppler and conductance ring

Two-phase flow exists widely in industrial field. Accurate measurement of the process parameters, such as flow velocity, is of great significance for quantifying flow rate and optimizing production process and equipment. A combination of one-side two-chip continuous wave ultrasonic Doppler (CWUD) sensor and conductance ring sensor is applied to measure the velocity in horizontal oil-water two-phase flow. The non-intrusive CWUD sensor consists of two-chip transducer with resonant frequency of 1 MHz, and there is a sound insulation material between the two chips to prevent sound waves from interfering. One chip continuously transmits sound wave to fluid, the other one receives the echo scattered from droplets. Consequently, its sample volume covers the whole radial range of pipe cross-section. The Doppler shift signal collected by CWUD sensor is directly related to the average velocity of dispersed phase within the sample volume. Dynamic experiments were conducted at a horizontal oil-water two-phase flow loop with a diameter of 50 mm. Based on the analysis of average Doppler shift response characteristic, it was found that there are two linear relationships between average Doppler shift and overall superficial flow velocity for different flow conditions, i.e. water and oil continuous flow condition. Hence, according to the electrical sensing principle of conductance ring sensor, a dimensionless voltage parameter is used to discriminate the continuous phase. And then for each experiment point, the corresponding measurement model is chosen to calculate the overall superficial flow velocity. The results show that the root mean square error of measured overall superficial flow velocity is 0.01 m/s and the average relative error is 3.09%. The confidence probability for relative error less than 5% is 70%.