| As the development of pneumatic conveying, especially the wide application of dense-phase conveying, understanding of the underlying mechanisms becomes more and more concerned by the researchers. By reviewing of current research status of pneumatic conveying, it is found that in order to achieve the goal of flow regime identification, pressure drop prediction and mass flow rates measurement, particle motion analysis and force analysis should be involved in the research studies. What’s more, considering the fact that pipe diameters are usually small, hydrodynamics of gas-solid two-phase flow will be strongly influenced by the particle-wall interactions. However, most of the investigations on particle-wall interactions were focused on the measurement of Janssen coefficients which were used for the prediction of pressure drop of plug flow. There is a serious lack of studies on the influence of particle-wall interactions on the hydrodynamics and flow stability of the gas-solid two-phase flow. It is of vital importance to investigate the role of particle-wall interactions in the field of pneumatic conveying.In this work, acoustic emission (AE) technique, which is very sensitive to particle motions, was combined with wavelet analysis and power spectrum density analysis to establish the method of online measurement of particle-wall interactions. By using this method, with the aid of pressure fluctuation, camera and electrostatic detection techniques, the influence of particle-wall interactions on pressure drop, flow patterns and flow stability were systematically investigated in lab-scale experiments. Then we applied this method to a high pressure dense-phase pneumatic conveying system of pulverized coal in the field. Wavelet transform and rescale analysis tools have been used for multi-scale resolution analysis of the AE fluctuation signals to investigate the flow behavior. The principal of classification of different scales is established first, followed by the discussion of the physical meanings of each scale of AE signals. The influence of particle-wall interactions on flow stability was also involved. At last, a novel measuring technique of mass flow rate based on the multi-scale analysis of AE signals was established. Research work has been carried out on the following aspects:1. The concepts of particle-wall collision (normal direction) and friction (tangential direction) were introduced into the analysis of AE signals for the first time. Acoustic analysis of particle-wall interactions detection method was proposed in this work. The influence of particle size, particle velocity and normal stress on the main frequency and energy of particle-wall collision and friction signals were studied via experiment. Based on the experimental results, models for the prediction of the main frequency and energy of AE signals were established. These models can be used for the detection of mass flow rates and particle-wall impact angle in dilute phase pneumatic conveying in the vertical pipe.(1) The method of acoustic analysis of particle-wall interactions including the following steps: first, the main frequency and characteristics of energy fraction distributions of collision and friction signals was studied by the experiment. The results showed that the main frequency of collision signals were much higher than that of friction signals for particles belong to different Geldart types. Then, wavelet reconstruct method was used to quantitatively extract the particle-wall collision part and friction part in the original AE signals.(2) Combining Hertz contact theory and piecewise plasticity model, we proposed a model for the prediction of the main frequency of particle-wall collision AE signals as follows The predicted main frequency fit with the experimental value very well. By introducing the contact time number, a model for the prediction of the main frequency of particle-wall friction signals was proposed as followsThe experiment results showed that main frequency of collision and friction signals decrease as the increase of particle size, main frequency of friction signals decrease as the increase of particle velocity and normal stress. These results can be well interpreted by the developed models.(3) It was found that suspended particle and cluster exist simultaneously in dilute-phase pneumatic conveying in vertical pipe, and the way they interacted with pipe wall were different. According to this fact, we established the model for the prediction of AE energy of particle-wall collision and friction. Based on the model, we developed the relationship between energy fraction of collision with solids loading ratio, Solid mass flow rates can be calculated by this equation, and the mean relative error is 3.78%. Also, we developed the relationship between energy fraction ratio with impact angle2. Multiple detection methods including acoustic emission, pressure drop and electrostatic were used to reveal the influence law of particle-wall interactions on flow regimes transition near the minimum transport velocity. Two new flow regimes transition diagrams were developed.(1) Single detection method can not precisely identify the flow regimes transition near the minimum transport velocity. It was found that pressure drop, AE energy and electrostatic accumulations decrease and then increase as the increase of gas velocity. However, the transition velocities were not the same. Transition velocity obtained by AE energy (7.0 m/s) was a bit less than that of pressure drop (7.5-8.0 m/s), but larger than that of electrostatic accumulations (6.0 m/s). The last transition velocity indicated the transition of flow regimes from suspension flow to plug flow.(2) The minimum transport velocity was influence by gas-wall friction, particle-wall interactions and particle concentration. When gas velocity was higher than minimum transport velocity, the pressure drop caused by gas-wall friction dominated. When gas velocity was lower than flow regimes transition velocity, the pressure drop caused by particle-wall interactions and particle concentration dominated. When gas velocity was between these two velocities, the pressure drop was influence by all the three factors.(3) At the flow regime transition velocity, energy fraction of collision was the minimum, energy fraction of friction was the maximum, and impact angle was the minimum. This was due to the transition of the distribution of particles. When gas velocity was higher than the flow regime transition velocity, single particle had collisions and frictions with pipe wall. As the decrease of gas velocity, particle-wall collision velocity decreased, resulted in the decrease of energy fraction of collision. When gas velocity was lower than flow regime transition velocity, plugs were formed, particles fell down and had collisions with the plugs, this became the dominant way of particle-wall interactions. As the decrease of gas velocity, the relative velocity between plugs and falling particles increased, resulted in the increase of energy fraction of collision.(4) Two new flow regimes transition diagrams were developed based on the relationship between AE energy and pressure drop, as well as electrostatic accumulation and pressure drop, respectively. These new flow regimes transition diagrams can identify both the minimum transport velocity and flow regime transition velocity. Besides, they can also indicate the different ways of pressure loss in different flow regimes.3. The influence of particle-wall interactions on the stability of plug flow was investigated. It was found that the pressure drop would rise for two times when the plug was long or gas velocity was small. The occurrence of this phenomenon was due to the collision between falling particles with the front of the plugs. Thus, this type of collision was involved in the derived AE energy model for plug flow. The results showed that the energy fraction predicted by the model fit well with the experimental values, which indicated that the collision between falling particles with the front of the plugs was the typical characteristic of plug flow.4. Multi-scale resolution analysis of the AE fluctuation signals was developed to investigate the flow behavior in a dense-phase pneumatic conveying system of pulverized coal at high pressure. The principal of classification of different scales was established first, followed by the discussion of the physical meanings of each scale of AE signals. At last, the influence of particle-wall interactions on flow stability was also investigated, and a novel measuring technique of mass flow rate was established.(1) Combining the multi-scale analysis method V statistic analysis and wavelet transform method, the principle for dividing the original AE signal of pneumatic conveying into macro-scale, meso-scale and micro-scale signals had been developed. In the V statistics against sub period length plots, two distinct regions were observed for the detailed signals drd4 (18.75-300 kHz), as well as for the detailed signals dg-d9 (0-2.34 kHz) and the approximated signal a9. However, three distinct regions can be seen for the detailed signals d5-d7 (2.34-18.75 kHz). Based on the differences of complexity, the detailed signals d1-d4, d5-d7, and d8-d10 were assigned to the micro-scale, meso-scale and macro-scale, which represent micro-scale particle-wall interactions, meso-scale interaction between gas phase and solid phase (such as bubbles, plugs, dunes), and macro-scale flow-induced pipe vibration. Further analysis showed that as the mass flow rate of pulverized coal increased, the energy fraction (energy of detailed signal divided by the energy of original signal) of micro-scale signals decreased while that of meso-scale signals increased, which indicated that particles were more likely to move as particle aggregates than individual particles when mass flow rate increases.(2) Based on multi-scale analysis of AE signals, combining with the signals characteristics in time domain and partial leaser square method, we established the detection model for coal mass flow rates and coal concentrations. The mean relative error for the calculated coal mass flow rates and coal concentrations was 4.15% and 4.78% respectively. When the model was used to predict coal mass flow rates about 5800 kg/h, the mean relative error was 10.37%, indicating that the proposed model has the ability of extrapolation. |
PDF Full Text Request