Abstract: Pneumatic conveying is widely used in industry, different flow regimes occur when conditions change. In order to minimize the energy losses and reduce pipe wall erosion, a suitable flow regime should be chosen. Thus, it is very important to determine the transition velocity between the flow regimes. Many researchers have investigated the influence of operation conditions and material properties on the transition velocity and minimum pressure drop, such as pipe diameter, gas velocity, gas density, feed pressure, particle density and diameter, and so on. Based on the experimental results, a lot of empirical or analytical approaches have been proposed. These correlations are very useful when a new conveying system is designed. However, due to the complicated interactions between gas phase and solid phase at different scales of time and space, the underlying mechanism of the transition is still not clear and thus limits the application ranges of these correlations. Acoustic emission signals are very sensitive to the particle motions and have been widely used in the detection of in gas-solid two phase flow. In the current work, a comparison study of acoustic emission signals and pressure signals was proposed, followed by the investigation of flow behavior inside a vertical pneumatic conveying pipe during the transition of flow regimes.
Experiments were taken in a vertical section of the pneumatic conveying system. The inner diameter of the pipe is 25 mm and the length of the vertical section is 3.0 m. Polypropylene (PP) particles with an average diameter of 2100 µm was used in the experiment, and the particle density is 900 kg/m3. Compressed air was used as the conveying gas. The mass flow rate of PP particles was determined by a weighing cell. In the experiments, the superficial gas velocity ranged from 5.0 to 13.0 m/s and the PP particle mass flow rate ranged from 0.006 to 0.026 kg/s. The online AE system for collection and analysis was developed by UNILAB Research Center of Chemical Engineering in Zhejiang University. The system consists of an AE sensor, a preamplifier, a main amplifier, A/D conversion module and a computer. The gain of preamplifier is 40 dB. The AE sensor used in this work is a piezoelectric accelerometer (AE 144S, Fuji ceramics corporation). The AE sensor was mounted on the outer surface of the vertical pipe, and the sampling frequency used was 900 kHz, sampling time was 5 s. Pressure drop of the pipe was detected by a couple of pressure probes (CTG121P, China) with a distance of 1.2 m between them. The pressure signal was recorded by the computer, and the sampling frequency used was 400 Hz, sampling time was 30 s. All the experiments were repeated 3 times and the average results were used.
The results showed that for all the mass flow rates conducted, as the gas velocity decreased, both the pressure drop and energy of acoustic signal decreased to a minimum and then increased. Moreover, the transition velocity detected by these two kinds of method agreed with each other very well. The similarity of these two kinds of signals can be interpreted from the sight of energy consumption. To be specific, the pressure drop stands for the energy needed to transport the particles while the energy of acoustic signals reveals the energy loss due to particle-wall collisions and frictions. Additionally, for the gas velocity beyond and below the minimum transport velocity, a nearly linear relationship can be found for the acoustic energy and pressure drop. However, the slopes of these two lines were not the same which indicate that different mechanisms dominated. Wavelet analysis was further applied and the acoustic signals were decomposed into 1-10 of scales detailed signals (d1-d10) and the 10th scale approximated signal (a10) using Daubechies 2nd order wavelet transform. According to our previous work (He et al., Ind. Eng. Chem. Res., 2014), the detailed signals d1-d3, and d4-d5 were recomposed as the particle-wall collision signals and particle-wall friction signals, which represent particle-wall interaction in the normal and tangential direction respectively. The results showed that as the gas velocity decreased, energy fraction (energy of detailed signal divided by the energy of original signal) of particle-wall collision signals decreased to a minimum and then increased while that of particle-wall friction signals varied the contrary. When the gas velocity was larger than the transition velocity, energy fraction of particle-wall collision signals decreased while that of particle-wall friction signals increased as the gas velocity decreased. This was due to the increase of gas phase turbulence which led to the increase of particle velocity fluctuation and thus larger normal velocity was gained. On the other hand, when the gas velocity was lower than the transition velocity, particle-wall collision energy fraction increased while particle-wall friction energy fraction decreased as the gas velocity decreased. Further analysis showed that the increased solid concentration led to the increase of collisions for the reversed particles with those moving upwards. Based on these analysis, it can be concluded that particle-wall interactions is one of the main cause for the change of pressure drop, and the transition of pressure drop was due to the change of the ways particles collide with each other.
Keywords: transition velocity; acoustic emission; flow regimes.