- Title
- A methodology for DEM modelling of a liquid fluidised bed using a random fluid fluctuating velocity
- Creator
- Abbasfard, Hamed
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2019
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- Liquid solid fluidised beds are broadly used in various industries as they can provide excellent mixing and large surface area to exchange energy, momentum and mass between interacting phases. As such, they have been extensively investigated to optimise their performance through an increase in the fundamental understanding of the key parameters affecting particle fluidisation. Despite the relatively practical simplicity of operating fluidised beds, modelling and simulation of such systems is not an easy task due to the complex interplay of interactions required to achieve a desirable level of fluidisation. There has been always a challenge to select a modelling approach that can provide reasonably accurate results without being computationally expensive. The literature shows that CFD-DEM models are becoming the most popular simulation tools to predict various aspects of both continuous and discrete phase in a fluidised bed. However, a DEM-only approach could save up to 50 % of the total computational cost in some cases. In order to develop a DEM-only model and examine the feasibility of this approach, the first step was to analyse the random motion of particles confined in a box. Then, the influence of applying a random fluid velocity, acting through a drag force model, on the settling and fluidisation of single/multiple particle systems was analysed. The random liquid velocity consisted of a mean and a fluctuating part. The mean velocity was actually the interstitial fluid velocity while the fluctuating part followed a Gaussian probability density function with a standard deviation which was a fraction of the superficial fluid velocity. The simulation results showed that the induced liquid fluctuating velocity was able to produce a realistic random motion of the particles like that found in settling and fluidised beds. In the final stage of the model development, a standard procedure to find out the suitable level of fluid velocity fluctuations was developed. In this new approach, the fluid fluctuating velocity was estimated using an empirical correlation of the literature that relates this parameter to the particle fluctuating velocities. The applied fluctuating velocities were updated at certain intervals. As the updating interval revealed a profound effect on particle velocity, a theoretical approach was introduced to provide the correct particle velocity compared to the literature. Bed expansion and concentration predicted by this new approach showed good agreement with experimental data and the Richardson-Zaki equation. In addition, the effects of column diameter, initial bed height and fluid velocity profile were analysed and discussed. In order to investigate the utilization of the model, the mixing of particles in fluidised beds was studied. Two innovative methods to quantify particle mixing ratio were introduced: position-based and path-based. The simulation results for the mixing ratio revealed good agreement with the experimental using both mentioned methods. It was also shown for the first time that the mean free path was longer along the vertical direction in the fluidised state. In addition, the dispersion coefficient as defined by Kinetic Theory of Granular Flow was modified to include particle and fluid densities. Then, the dispersion coefficient was correlated with the mixing time for different superficial velocities. Furthermore, a theoretical approach to calculate granular pressure based on wall collision frequency was developed. Granular pressure showed a peak in a range of porosity like that of theoretical and experimental published works. Finally, the segregation behaviour of a binary system of 5- and 8-mm particle having the same density was also simulated using this new DEM approach. In summary, the simulation results presented in this thesis revealed that using DEM only along with other numerical methodologies proposed here suffices to predict the behaviour of fluidised beds with reasonable accuracy. However, there exists a considerable space to further extend this model and its applications in the simulation of multi-phase flow systems such as gas bubbling fluidised beds.
- Subject
- modelling; fluidised bed; DEM; fluctuating velocity
- Identifier
- http://hdl.handle.net/1959.13/1403487
- Identifier
- uon:35177
- Rights
- Copyright 2019 Hamed Abbasfard
- Language
- eng
- Full Text
- Hits: 6707
- Visitors: 7523
- Downloads: 1097
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Thesis | 3 MB | Adobe Acrobat PDF | View Details Download | ||
| View Details Download | ATTACHMENT02 | Abstract | 516 KB | Adobe Acrobat PDF | View Details Download |