Skip navigation

Air slide basic modelling

Air slide basic modelling

Valciu, Serena Carmen (2015) Air slide basic modelling. PhD thesis, University of Greenwich.

[img]
Preview
PDF
Serena Carmen Valciu 2015.pdf - Published Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (14MB) | Preview

Abstract

This thesis addresses the evaluation, modification and modelling of a mass flow feeding silo, stand pipe and standard air slide system for use in handling and transporting alumina in the aluminium smelting industry. Stability and repeatability of gravitational flow rates from feeding silos and stand pipes (addressed as interfaces) is crucial from operational point of view and needs to be achieved before one can further estimate and model the flow of alumina and the capacity of an air slide. Although techniques for design of storage vessels (based on the flow characteristics of particulates) have been in the public domain since the 1960’s, they have been slow to gain acceptance in the industry (due to a lack of awareness of their existence amongst engineers). The use of the stand pipe concept to increase flow rates during gravity discharge and to achieve stability of flow has been examined experimentally in a small rig at Tel-Tek department of POSTEC. The stand pipe should be integrated into the mechanical design of an air slide, because if the material cannot be delivered (discharged) properly from the feeding unit onto the air slide, there will be no stable conveying conditions. It has been found by Gu, Arnold and McLean (1993) that the stand pipe is efficient only if it is kept full of powder - the feeding silo acting more as a buffer for the stand pipe. Their concept was implemented in practice during the tests conducted at POSTEC by monitoring the weight of material in the stand pipe and the flow rates of alumina in real time by using LabView software and weighing cells. What makes the research work reported in this document unique is the integration of different mechanical and modelling concepts from different engineering fields (continuum and soil mechanics, fluid dynamics, rheology, cybernetics and powder technology).

Previous work undertaken by Haugland (1998) for Hydro focused the flow of alumina as a Newtonian fluid and was itself based on the modelling work of Woodcock (1978). Yet, besides their work, no accurate mathematical model has been developed which could be used to predict the velocity and mass flow rates of alumina flow and capacity of an air slide for a range of alumina qualities. Such a model would be extremely useful when designing new air slide systems, as the capacity of air slides could be predicted from the rheology of powder and the knowledge of bulk particulate characteristics. Previous research, measurement campaigns and extensive literature studies, all point out that in terms of controlling discharge from gravity flow equipment (and after studying the transport behaviour of an air slide), it is often the gravity discharge approach that generates the greatest degree of variability, in terms of both consistency and repeatability. The root of the problem of inconsistent discharge rates has been identified to lie with the flow channel development within the feeding silo and/or the stand pipe (discharge head). The potential causes of flow irregularities are even more critical to be aware of, in the cases where mass flow principles are intended to be applied - especially in the case of highly segregable materials like alumina blends.

This thesis is aimed at developing a reliable means for predicting the average velocity of alumina flow and capacity and also to understand the mechanisms that govern the flow behaviour. To enable online estimation and further modelling of full-scale air slide capacity, a small-scale rig with adjustable length was built at POSTEC in 2012 and further modified in 2013. Air slide capacity (alumina flow rates) for different lengths of 3 m, 7 m and 15 m and inclination angles from 0 to 3.1 degrees of air slide were measured by using pressurized air in the range of 3 to 6.5 barg. Empirical models have been developed from test results as a first step. It was found based on empirical models that the velocity of alumina bed has a power law behaviour. However, these models would require the use of expensive and time-consuming air slide trials to determine the values of the power law coefficients for each alumina quality.

The degree of segregation when handling and transporting alumina has proven to be quite considerable, in terms of undesirable effects on the production process. This thesis was prompted by several measurement programmes of work undertaken for Hydro by the author with the aim of mapping the performance of volumetric feeders and the degree of segregation in a feeding air slide rig at The Reference Centre in Årdal. Sampling from feeders and chemical analysis of the samples have clearly shown the degree of segregation, especially when fluidizing and transporting binary mixtures of alumina and aluminium fluoride. Pressure measurement results showed pressure drops along the air supply tube, which also have been found to have an impact on segregation. Fluidization and shearing trials have been conducted both on alumina and binary mixtures in order to establish a benchmark procedure for powder characterisation that could be used as an operational support tool. Currently the minimum fluidization velocity for all Hydro Aluminium plants is set to 2.0 cm/s, which is much higher than what it is actually needed to achieve an optimal fluidization.

There is a strong analogy with the flow of Newtonian or non – Newtonian fluids in an open channel that can be applied to the flow of fluidized alumina on an air slide. In this thesis, the hydrological non- Newtonian model developed from the general Saint-Venant model of open channel in a rectangular channel has been applied as a second step. Such theoretical mechanistic models, based on the mass and momentum balance with bottom friction along the powder bed, are numerically challenging to solve. The model requires use of rheometrical benchmark tests to determine the flow index coefficients for a specific quality of alumina for a given initial inlet bed depth and steady state estimated capacity. The use of an Ordinary Differential Equations (ODE) solver in MATLAB to predict the height and the velocity of alumina bed in an open channel seemed promising and showed similar trends when compared to the alumina bed results obtained from the air slide tests conducted at POSTEC. Results obtained so far indicate that a more detailed analysis needs to be conducted in order to find out how to ‘fine tune’ the model parameters to further improve the model fit to measurement data. Although more experimental data is needed, e.g. shear stress measurements and flow coefficients at different fluidization velocities and for different alumina qualities, the correlation between measurements and the model obtained so far confirmed that further investigation would be justified.

Thus, the effect of interfaces, the feeding silo and the stand pipe should be considered and included into further design approaches for mechanical equipment, by balancing headroom availability in pot rooms versus increase in transport capacity of bulk solids to optimize production.

Item Type: Thesis (PhD)
Uncontrolled Keywords: powder flow behaviour; air slide test rig; Empirical modelling; mathematical modelling;
Subjects: Q Science > QA Mathematics
T Technology > TA Engineering (General). Civil engineering (General)
Faculty / School / Research Centre / Research Group: Faculty of Engineering & Science
Faculty of Engineering & Science > School of Engineering (ENG)
Last Modified: 22 Nov 2017 12:14
URI: http://gala.gre.ac.uk/id/eprint/18121

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics