Eshaafi, Taher (2020) CFD Based Prediction of the Performance and Thermal Analysis of a Centrifugal Fan. Doctoral thesis, University of Huddersfield.
Abstract

One of the most commonly used machines for industrial and ventilation purposes is the centrifugal fan as it can provide high static pressure rise and higher flow rates. However, the aerodynamic performance of large centrifugal fans can be significantly improved. Additionally, centrifugal fans are also used in high temperature applications where the heat transferred from the source can either damage or reduce the performance of the bearing unit and the electric motor. Thus, a special hub containing a cooling disc can be used to prevent the overheating of the bearings and the motor. The usage of a hub with superior thermal performance can also provide a longer useful life for the fan system.

This work, which is focused on improving the performance of specific products, was carried out in collaboration with the University of Huddersfield and Halifax Fan Limited, United Kingdom. A centrifugal fan with backward inclined blades that hold the industrial model named “Mistral” is investigated in this thesis. Furthermore, a special hub designed by the manufacturer for the high temperature fan applications is also explored in this work. The aim of the work is to characterize and improve the aerodynamic performance (i.e. increase the pressure and the efficiency of the fan) by altering the geometrical parameters of the impeller blade angle and shroud angle. Additionally, to achieve a lower temperature at the bearing region, it is important to improve the thermal performance of the hub.

The general trend in the literature for improving the performance of fans is the technique of experimentally altering the geometrical variables. This requires more time and is often limited to global outputs. The necessity of detailed insights on thermal and flow fields’ behaviour dictates the use of sophisticated analytical and numerical tools to model the flow conditions within the centrifugal fan, resulting in an improved understanding of the underlying flow phenomena.

Advanced modelling tools were used for this study to simulate the flow and thermal fields within the fan and the fan hub. Computational Fluid Dynamics (CFD) techniques have been used to model the operation of the fan in order to carry out investigations under various geometric and flow conditions. The fan numerical model employs a sliding mesh approach to model the impeller motion and solves the Reynolds Averaged formulation of Navier-Stokes equations. The two-equation k-ω SST model is used as the closure model in this work. The numerical results are validated by comparing the performance predicted numerically with the experimentally measured values for the pressure rise and efficiency.

The effect of the impeller geometry on the fan performance is quantified by characterising the flow for the design operation, called also Best Efficiency Point (BEP), near surge and near choke operation of the fan at 3000 rpm. The local and global flow features are analysed along with the sensitivity of the fan performance characteristics to the blade and shroud angles. The highest-pressure rise was observed in the case of near surge operation followed by the design point and then the near choke point.

The effect of operational parameters on the thermal and flow fields within the hub and the ambient air are investigated using the numerical model that is validated by corresponding experimental measurements, in which the temperatures measured at specific points are compared with the values predicted by the numerical model. The effect of heat generated from the bearing unit is replicated using the heat fluxes that are used as the boundary conditions in the numerical model. The temperature predicted by the numerical model is in good correlation with the experimental measurements. The analysis of the flow-field shows higher velocities in the vanes of the cooling disc while lower velocities are seen near the shaft. The pressure is diffused in terms of velocities, causing flow in the holes in the hub. Thermal analysis shows the highest temperature near-source followed by a gradual decline up until the core and an increase up until the bearings followed by decline again.

The impact of the core geometry, specifically the different designs of holes and length of the core on the thermal performance, is explored. The design with a higher number of holes demonstrated the best thermal performance among the design modifications investigated. The core design with the best performance is then further used to investigate the impact of the length of the core. A decrease in the core length is seen to deteriorate the thermal performance, an insight which will aid better designs for the hub.

On the basis of this work, the author can make a series of recommendations, ranging from suggested decisions about the numerical modelling of the aerodynamic and thermal characteristics of the fan and the hub; to design considerations for improving the performance of the system. For the investigated fan, the performance of the system is accurately predicted by the computational domain of approximately 2.1 million polyhedral cells with SST k-ω turbulence formulation, wherein the wheel rotation is handled using a sliding mesh approach along with temporal resolution corresponding to 2° rotation per time step.

The aerodynamic performance of the system can be improved by decreasing the impeller shroud angle while keeping the default blade angle. The thermal performance of the hub can be improved by increasing the area of contact for facilitating the convective heat transfer from the hub to the ambient air. This can either be achieved by increasing the diameters of the holes or by adding an extra layer of holes.

Information
Library
Documents
[thumbnail of FINAL THESIS - Eshaafi.pdf]
FINAL THESIS - Eshaafi.pdf - Accepted Version
Restricted to Repository staff only until 18 August 2025.
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (10MB)
Statistics
Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email