| Stator-rotor interaction and wake are the main unsteady flows of turbo machinery. The interaction noise between stator and rotor and high circle fatigue of rotor blade downstream will be reduced, and the total noise performance of the whole machine set will be improved using trailing edge blowing (TEB). It is very important to study and master the flow characteristic and control method of trailing edge blowing to improve the overall performance of the turbo machine set.Experimental measurement and numerical simulation are the two basic methods to study the internal flow of turbo machinery. With the development of technology, various advanced measurement techniques come forth to serve for the investigation, such as Hotwire Anemometer (HWA), and Particle Image Velocimetry (PIV), which supply experimental study for the new method and device of measurement. Meanwhile, with the rapid development of computer performance and gradual perfection of computational fluid dynamics (CFD), numerical simulation method based on solving 3D viscous Navier-Stokes equation supply a powerful tool for the investigation of internal flow of turbo machinery. Experimental measurement and numerical simulation are combined to study the effect of trailing edge blowing towards a low speed axial fan in this paper.Main contents and conclusions of the investigation are described in the following:1. According to the demand of study, an experimental stator is designed and manufactured. And the system of trailing edge blowing of the stator is designed to supply continuous and stable air flow for the stator. Under different flow rate of trailing edge blowing, four wake statuses can be obtained, that is pure wake (no blowing), weak wake (some blowing), momentumless wake (blowing adjusted to provide a thrust which exactly cancels the blade's drag), and jet (more blowing than necessary to cancel the drag).2. The performance test facility is designed in according with the Chinese National Standard GB1236-2000. The aerodynamic performance curves of stator-rotor system under different trailing edge blowing rate are obtained. The results show that the mass flow rate is too small to affect the aerodynamic performance of stator-rotor system. When trailing edge blowing reaches momentumless wake status, the mass flow rate is about 1.8% of the designed flow rate of the rotor.3. HWA and PIV are used to measure the detailed flow of stator wake, the flow characteristics of four wake status are obtained through axial velocity profiles, wake characteristic length scale, turbulence length scale, etc. The results showed that through trailing edge blowing up to momentumless wake, the velocity deficit of the stator wake can be filled. The shedding vortex of stator was reduced through momentumless wake. The vortex shedding frequency and its harmonics are eliminated by trailing edge blowing up to momentumless wake. The turbulence length scale of pure wake is larger than that of momentumless wake, and it become larger with the increasing of axial position, and for momentumless wake case it become smaller contrarily. The wake characteristic length scales of pure wake and momentumless wake become larger with the increasing of axial positions, but at the same axial position, characteristic length scale of momentumless wake is smaller than that of pure wake. That is to say, trailing edge blowing can reduce the wake width of stator. Through the experimental result, the wake flow of pure wake and momentumless wake takes on similarity characteristic. At last, CFD simulation was carry out to simulation the wake flow of stator to gain more information about wake. SST turbulence model is found to suit for the wake flow simulation. And the results from this turbulence model agree well with the experiments. Through the CFD simulation, some information, such as velocity, pressure and turbulence kinetic energy, and so on, is abstracted to describe the wake flow of stator.4. HWA and PIV are used to measure the interaction flow between stator wake and rotor, the spreading rules of stator wake across the rotor row are obtained. The results of HWA measurement shows trailing edge blowing can eliminate shedding vortex of stator and weaken the periodicity effect of the rotor downstream. The results of PIV tests present the wake flow across the rotor row. Through trailing edge blowing, the inlet of the rotor is more uniform, and the fluctuation of velocity is small compared with non-blowing from the trailing edge of stator. A whole simulation including all stator and rotor blades is carried out to calculate the flow field of stator-rotor system with trailing edge blowing. And the results agree well with the experiments. Through the distribution of vorticity, pressure, turbulence kinetic energy, etc, the two wake flows are compared with each other. The results of PIV experiments and CFD simulation results reveal the flow characteristic of stator wake flowing across the rotor row. Through trailing edge blowing, the inlet of rotor can be more uniform, and the fluctuating of velocity is smaller.5. The sound pressure level and noise spectrum are measured using sound level meter and a dual channel real-time frequency analyzer in a semi anechoic chamber. A hybrid prediction method is adopted to predict the noise radiated from the stator-rotor system, i.e. CFD plus FW-H. For the trailing edge noise, the shedding vortex noise model developed by Fukano is adopted. And for the interaction noise between stator wake and rotor, the point force model for discrete frequency noise developed by Lowson is adopted. The prediction results agree well with the experiment. The noise test and noise prediction results show that trailing edge blowing can reduce the noise level, especially the discrete noise level; the effect to the broadband noise level is small.6. Pulsing blowing is brought forward in this paper. And CFD simulation is used for preliminary study. Meanwhile, the noise prediction is made using above noise prediction models based on CFD results. The noise prediction results show that pulsing blowing can reduce high order discrete noise, and then reduce the total discrete noise level.
|
PDF Full Text Request