• Energy Research

Axial fan is an important piece of equipment in the thermal power plant that provides enough air for combustion of coal. This paper focuses on the aerodynamic noise characteristics of an axial fan in the development from stall inception to stall cells. The aerodynamic noise characteristic of monitoring region in time and frequency domains was simulated employing the large-eddy simulation (LES), with the addition of throttle setting and the Ffowcs Williams-Hawkings (FW-H) noise model. The numerical results show that, under the design condition, the acoustic pressure presents regular periodicity along with the time. The noise energy is concentrated with high energy of the fundamental frequency and high order harmonics. During the stall inception stage, the acoustic pressure amplitude starts fluctuating and discrete frequencies are increased significantly in the low frequency; among them, there are three obvious discrete frequencies: 27.66 Hz, 46.10 Hz and 64.55 Hz. On the rotating stall condition, the fluctuation of the acoustic pressure level and amplitude are more serious than that mentioned above. During the whole evolution process, the acoustic pressure peak is difficult to keep stable all the time, and a sudden increase of the peak value at the 34.5th revolution corresponds to the relative velocity’s first sudden increase at the time when the valve coefficient is 0.780.

202303 bcww ; Not applicable ; Others ; Research Institute for Sustainable Urban Development (RISUD) ; Published ; 24 months ; Green (AAM)

202407 bcwh ; Not applicable ; Others ; Hong Kong Polytechnic University ; Published ; 24 months ; Green (AAM)

Abstract Based on Reynolds averaged Navier-Stokes (RANS) equation and Realizable k-e turbulence model, the influence of abnormal deflection of two adjacent moving blades in an axial fan was discussed numerically. The flow field in the rotor, static and dynamic characteristics of normal blades and four abnormal blades combinations under rotating stall are compared. The results show that the abnormal blades have no effect on the number of stall cells. Compared with the normal blades, in the case of non-stall or stall conditions, the positive deflection blades will increase the surface stress of the blade, and the negative deflection blades can reduce it; deformation distribution also has the same law. The deformation distribution of blades is affected by both aerodynamic and centrifugal loading. However, the stress distribution of the blades is mainly affected by centrifugal loading. By modal analysis, the total deformation distribution of the first to fourth-order vibration modes are obtained. The first to third-order vibration modes show the coupling vibration between the blades and hub, and the fourth order is the blade vibration alone.

202303 bcww ; Not applicable ; Others ; Research Institute for Sustainable Urban Development (RISUD); Hong Kong Polytechnic University ; Published ; 24 months ; Green (AAM)

Abstract The development of a more advanced three-dimensional wake model for wind power generation is presented based on a multivariate Gaussian distribution. The newly-presented model is closer to reality as it truly depends on two independent dimensions (namely horizontal and vertical directions) rather than the radius of a circle. For this reason, the general expression of wake expansion rate in each dimension is specifically developed. In addition, by taking into account the inflow wind shear effect, this current model is able to accurately capture the asymmetric distribution of the vertical wake profile. Four cases including experimental data from wind tunnels and field observations as well as high-fidelity numerical simulation are used to validate the present model. Compared with conventional models, this new model is capable of predicting the wake distribution of a single wind turbine reasonably well. The proposed model is highly simple with a low computational cost. Before applying this model, no additional numerical calculation or trial calculation is required. Wake velocity at any given spatial position can be calculated in an accurate and fast manner. Because of its accuracy, universality and low cost, the present three-dimensional wake model is able to make contributions to farm-level applications such as layout optimization and control strategies and therefore benefit the power output of wind farms.