Modelling the solids transport phenomena within flighted rotary dryers

This thesis presents the development and testing of a solids transport model for\ud flighted rotary dryers based on the physical and geometric properties of the system.\ud Particular emphasis was placed on understanding the internal flows and phenomena.\ud An introduction to flighted rotary dryers is given in Chapter 1, where the context and\ud relevance of this research is outlines. Chapter 2 gives a review of literature pertaining\ud to the modelling and analysis of solids transport in flighted rotary dryers.\ud \ud Chapter 3 discusses the development of the solids transport model based on the\ud physical behaviour of a flighted rotary dryer. The solids transport model was\ud developed based on numerical methods, dividing the dryer into a number of discrete\ud slices, and each slice was further separated into two discrete phases. One phase\ud selected to represent the material contain in the flights and in the bottom of the drum,\ud whilst the other phase was selected to represent the solids falling through the moving\ud gas stream. The flow of solids between phases was based on the physical movement\ud of solids that occurs within an actual dryer. The magnitude of these flows was\ud described using solid residence times and partitioning coefficients.\ud \ud The solids transport phenomena occurring in the two phases were described using the\ud geometry of the dryer and the physical properties of the solids. Chapter 4 presents a\ud model for the unloading profile of a generic unserrated, straight, two-section flight,\ud which was developed based on geometric analysis of the holdup within a flight. This\ud unloading profile was then used to calculate the average fall path of a solid particle\ud within a dryer, and thus the time spent within each phase of the dryer. Using\ud measurements from CSR Invicta Mill’s raw sugar dryer number 2, the average fall\ud time of a particle was found to be in the order of 0.9 seconds, and the average time a\ud particle spent in the flights to be in the order of 9 seconds. These residence times were\ud then used to govern the flow of solids within the overall solids transport model, and\ud the methodology describes a generic approach to modelling flighted rotary dryers.\ud \ud A flight unloading apparatus was used to validate the geometric flight unloading\ud model, the methods and results of which are presented in Chapter 5. The apparatus\ud consisted of a 1m length of a full scale industrial dryer flight, which was rotated at a\ud controlled rate, and the rate at which material was discharged recorded. Tests were\ud conducted using three different flight geometries and three different solid materials at\ud rotational speeds between 1 and 8 rpm. It was found that the geometric unloading\ud model accurately represents the experimental unloading profiles across the full range\ud of conditions tested. High-speed photography was used to observe the solid material\ud during unloading, and to measure factors such as the cascading curtain thickness,\ud surface particle velocity and dynamic angle of repose. It was found that the dynamic\ud angle of repose of the solids was dependent on the rotational velocity of the apparatus\ud and showed significant variability. Surface velocities were found to be in the order of\ud 1 m/s and surface thickness was found to be closely linked to the unloading rate of the\ud flight. The data from these experiments was used in simulating interactions between \ud gas and solids in the falling curtain in the following chapter.\ud \ud Study of the high-speed photographs and unloading profiles revealed that the\ud unloading of the flight was discontinuous, even though the materials used were\ud generally regarded as free flowing. Observation of both the unloading profile and the\ud high-speed images showed the flight unloading in pulses, with periods of high flow,\ud and periods where less material was unloaded. This resulted in a varying material\ud surface within the flight, which contributed to the high variability in measurements of\ud surface properties. The effects of flight serrations and the methodology of modelling\ud air drag in particulate curtains was also described.\ud \ud Chapter 6 presents a study on using computational fluid dynamics (CFD) to simulate\ud the gas-solids interactions within a falling curtain of solids. Experimental results from\ud wind tunnel experiments conducted at Monash University, Melbourne, were used to\ud verify and validate the CFD model. The simulated results showed good agreement\ud with the experimental data for solids displacement and velocity. Data from the flight\ud unloading experiments were then used to simulate the behaviour of the falling curtains\ud of solids that occur within a rotary dryer. Simulations with a single curtain showed\ud that gas-solids interactions were minimal for the conditions studied, with solids only\ud being displaced in the first 10-20cm of the falling curtain. Simulation using multiple\ud curtains in close proximity (50-80mm apart) showed that channelling of the gas flow\ud between the curtains was significant (increases in gas velocity of up to 25% were \ud observed), and resulted in greater displacement of the falling curtains. However,\ud limits on computational requirements prevented further study of this phenomena.\ud \ud Chapter 7 presents the validation of the solids transport model using experimental\ud data from an industrial flighted rotary dryer. The geometric flight unloading model\ud was integrated into the overall solids transport model for a rotary dryer, and\ud experimental data from Invicta Sugar Mill’s dryer number 2 was used to statistically\ud determine the remaining model parameters to validate the model. For a model using\ud 33 slices, the kilning phase residence time was estimated to be 7.7 seconds, with 54%\ud of the falling solids undergoing backmixing. It was found that the solids transport\ud model provided a good fit to the experimental data, however it was unable to match\ud the extended tail of the experimental curve. It was found that the number of slices\ud used in the model to represent the dryer had a minimal influence on the quality of the\ud statistical fit to the experimental data. Due to the inability of the solids transport\ud model to match the extended tail of the experimental RTD, alternative model\ud structures were considered and studied. However, the alternative model structures\ud considered showed similar or poorer fits to the experimental data, and techniques to\ud enhance the fit are described.\ud \ud Further study of the model predictions revealed an improbable amount of kilning\ud material within the dryer, with less than 5% of the solids within the dryer present in\ud the falling curtains of solids. This in turn resulted in large amount of solids\ud undergoing kilning, resulting in kilning being the dominant mode of solids transport\ud within the model. This is believed to be unrealistic, and emphasised the need to\ud undertake further experimental research into kilning and holdup in flighted rotary\ud dryers. Based on the observations made in this thesis, a number of recommendations\ud are made for the further development of solids transport models for rotary dryers, and\ud these are presented in Chapter 8