Géza Pattantyús-Ábrahám Doctoral School of Mechanical Engineering 

BME Faculty of Mechanical Engineering, Department of Fluid Mechanics

Supervisor: Dr. Vad János

Diagnostic of Axial Flow Fans with the Involvement of the Phased Array Microphone Technique

Introduction of the research topic

The axial flow fans are the parts of our everyday life (ventilation, aviation technique, etc.). These machines may be small sized, but their efficient operation is necessary because of their large number. Besides, if humans or animals are near the machine, the quiet operation is also required. The aim of this research is to better understand the aeroacoustic noise generation mechanism of axial turbomachinery blading, and the development of concerted aerodynamics-aeroacoustics diagnostic methods for axial flow fans.

 

Brief introduction of the research place

The Department of Fluid Mechanics (Faculty of Mechanical Engineering) is one the few aerodynamic research institute in Hungary. The wind tunnel laboratory of the Department is unique in Central-Eastern Europe. Since the activity of Dr. József Gruber, one of the most important research area of the Department has been the aerodynamic and acoustic investigation of fans, where phased array microphone (PAM) technique has gained an increasingly important role over the past years. Nowadays, the leading figure of the research is Dr. János Vad, the Head of the Department and my supervisor.

 

The history and context of research

 

The energy losses and the emitted noise of an axial fan can frequently be related to the unfavourable flow phenomena in the fan blade channels and  often have the same common cause (for example the thickening of the boundary layer on the blades). Therefore, he combined reduction of the aerodynamic loss and noise is possible as early as during the design phase. However, the electric input power and the emitted noise of an axial fan built in an industrial environment may exceed the catalogue data, because the operating conditions and installation environment can differ from the design considerations. In this case, the cause of the problem can be discovered with an aerodynamic simulation or on-site acoustic and aerodynamic measurements. The simulation requires a lot of time and expertise, and the acoustic measurements should be performed at low background noise, therefore it may interfere with the production process. The latter can be solved with the PAM technique, which means synchronized sampling with multiple microphones. From the outcoming signal the position and strength of the noise sources in the investigated area can be determined with the help of beamforming algorithms. An example is shown in Fig 1. Another advantage of this technique is that it can be used in noisy environments. 

 

Figure 1: Noise filtration investigation of a window [1]

 

The aim of the research

 

PAM technique was used in various cases in the acoustic tests of turbomachinery, but its combination with aerodynamic measurements is very rare. Therefore, the main goal of the research is the development of aerodynamic-acoustic diagnostics methods of turbomachinery, which combine PAM technique with spatially resolved aerodynamic measurements and satisfy the requirements of on-site investigations. The elaborated methods aimed at understanding the following phenomena:

 

Boundary layer noise:

In the near vicinity of solid walls (called boundary layer) the velocity of the fluid is equal to the velocity of the wall, and moving away from this surface it gradually reaches the velocity of the main flow. With increasing pressure in flow direction (like in the blade channels of a fan), the boundary layer is thickening and it can separate from the wall, i.e. the flow will not follow the shape of the surface. [2] With the thickening of the boundary layer on the blade surface, emitted noise and aerodynamic losses increase as well, and at the point of separation, a large, sudden increase can be observed. [3][4] One of the main goals of the research is the development of a diagnostic method, that can estimate the emitted noise and the aerodynamic loss associated with the blade’s boundary layer, and enable to propose a method for their reduction, for example by influencing the velocity profile in the inlet cross section of the blading.

 

Leakage flow noise

There is flow above the fan blade tips from the pressure side to the suction side due the pressure difference. This flow can become more chaotic if the fan is installed in a duct, due the shearing effect between the standing and moving surfaces. [3][4] Although it is known that this leakage flow constitutes a loss and noise source, but from the noise source maps, made during the investigation, it became apparent that the source of noise could not be unambiguously connected either to the leakage flow or the turbulent ingestion of the leading edge of the blades. The other goal of the research is to eliminate this uncertainty.

 

Methods

 

Measurement technique

The aerodynamic measurements required a robust but portable device capable of providing spatially resolved data. Therefore, a vane anemometer was chosen, which can be used for flow velocity measurements in a wide velocity range.

For the acoustic measurements the above mentioned PAM technique was used. For noise source maps generation, the ROSI beamforming algorithm [5] was used, which is capable of generating noise source maps in a co-rotating reference frame, which enables better investigation of the noise sources associated with the fan blading.

 

Semi-empirical models

As it exercises a huge impact on the aerodynamic losses and the emitted noise, it is essential to know the thickness of the blade surface boundary layer. However, its measurement is complicated and is only possible under laboratory conditions. In order to solve this problem, semi-empirical models [3][6][7] were used, which were developed for axial flow fans, and the boundary layer thickness on the blade surface can be estimated from easily measurable or available data (blade geometry, inlet velocity profile, speed of revolution, ...).

A semi-empirical model was also developed [8] (further referred to as BPM model) to estimate the noise of a single air foil. Using this model the spectrum of the boundary layer noise can be calculated from the inlet velocity, the angle of attack and the size of the air foil. The fan blade sections can be considered as air foils, therefore the application of this model can be useful in the present case.

 

Flow simulation

The numerical flow simulations have a huge role in the better understanding of flow phenomena. In the present case the finite volume method was used. In this method, the geometrical model of the flow domain is constructed first and divided in to small cells (this is the numerical mesh), and boundary conditions are prescribed on the boundary surfaces. Solving the governing equations (mass conservation, momentum conservation, effect of the turbulent phenomena, …) on the numerical mesh the current flow can be determined. The geometrical model and the boundary conditions of the present case can be seen in Fig. 2.

Figure 2: The geometrical model and the boundary conditions of the simulations

Results

The presented results relate to short ducted axial fans.

 

Suction side boundary layer noise

During the research a method was developed to compare the results from the PAM measurements and from the above mentioned BPM model in case of axial fans. The difficulties of the comparison are the following:

      I.        the BPM model was developed for single, infinite air foils, but in the case of axial fans there is a cascade, and the geometry of the blades can change along the radius.

     II.        in the noise source maps resulted from beamforming, for a single noise source there is no zero value even at locations where there is no noise source, but a spatially distribution of the source strength can be seen instead. This distribution is determined by the point spread function calculated from the position of the source and the geometry of the PAM. In the case of multiple sources the noise source map is the sum of the point spread functions.

With the developed method, the radial distribution of the suction side boundary layer thickness can be estimated from the inlet velocity profile and the blade geometry using semi-empirical models. Using the resulting boundary layer thickness distribution as the inlet parameter of the BPM model the radial distribution of the noise source strength of the boundary layer noise can be calculated. The convolution of the calculated noise source strength distribution and the point spread functions can be compared to the circumferentially averaged noise source strength data from the PAM measurement which was performed from the upstream direction of the fan. (Fig 3.)

 

 

Figure 3: The radial distribution of the source strength levels from the PAM measurements and the BPM model

 

 

Based on the results, the BPM model is in good agreement with the PAM measurements, the uncertainty is +/- 3 dB in a wide parameter range and through the 40 dB source strength range. Therefore, the BPM model can be used for the boundary layer noise estimation in case of axial fans.

 

Leakage flow noise

As it was mentioned above, in the noise source maps of axial fans (Fig. 4) it cannot clearly be decided whether the noise sources near the blade tips are caused by the leakage flow or the turbulent ingestion of the leading edges. In order to eliminate this ambiguity the following process was developed:

      I.        performing a PAM measurement from the upstream direction

     II.        repeating the measurements PAM with a blade elongated with a light, flexible plate (for example cardboard) to reduce the leakage flow of this blade. There should be no contact between the elongated blade and the duct wall.

    III.        comparing the noise source maps from the two measurements: if there is a decrease in the noise source strength in the blade channel after the elongated blade (label 1 in Fig. 4), the cause of the noise is the leakage flow. If the noise source strength in the second blade channel after the elongated blade decreases (label 2 in Fig. 4), the double-leakage flow has a role in the noise generation as well.

 

Figure 4: Noise source map [dB] of an axial fan with uniform tip clearance (left hand side) and with an elongated blade (right hide side) on 4000 Hz. The black arrow shows the elongated blade

 

As a results of the research, it has also been shown, that the noise reducing effect of the bellmouth entry in the case of short ducted axial fans is related to the leakage flow as well. The numerical simulations showed (Fig. 5) that the bellmouth entry eliminates the low-velocity zone in the inlet section of the duct, therefore the axial velocity of the flow is higher near the blade tip. This causes a reduced leakage flow and eliminates the strongest noise sources near the leading edge and the blade tip. (Fig. 6)

 

Figure 5: Streamlines from the blade tip without (left hand side) and with (right hand side) bellmouth entry - black arrow: direction of rotation, white arrow: axial direction

 

Figure 6: Noise source map [dB] of an axial fan without (left hand side) and with bellmouth entry (right hide side) on 5000 Hz

 

 

 

Expected impact, further research

 

Based on the results the aerodynamic-acoustic diagnostic of axial fans built in industrial environment will be easier, the noise source maps of the fans can be better understood. The results ‒ thanks to the research at the Department ‒ can be extended for axial fans installed in long ducts and for aircraft engines.

 

 

Publications, references, links

Own publications

[S1] Benedek, T., 2012. Mikrofontömbös akusztikai mérési módszerek a Budapesti Műszaki és Gazdaságtudományi Egyetem Áramlástan Tanszékén. Kolozsvár, Romania, OGÉT 2012 - Nemzetközi Gépészeti Találkozó .

[S2] Benedek, T., 2014. Axiális átömlésű ventilátor mikrofontömbös diagnosztikája a zajcsökkentés és a hatásfoknövelés érdekében. Energiagazdálkodás, 55(3), pp. 2-5.

[S3] Benedek, T. & Tóth, P., 2013. Beamforming Measurements of an Axial Flow Fan in an Industrial Environment. Periodica Polytechnica, Mechanical Engineering, 57(2), pp. 37-46.

[S4] Benedek, T. & Vad, J., 2014. Concerted Aerodynamic and Acoustic Diagnostics of an Axial Flow Industrial Fan, Involving the Phased Array Microphone Technique. Düsseldorf, Germany, ASME Turbo Expo, 2014.

[S5] Benedek, T. & Vad, J., 2015a. Spatially Resolved Acoustic And Aerodynamic Studies Upstream And Downstream Of An Axial Flow Fan. Madrid, Spain, ETC11 - European Turbomachinery Conference.

[S6] Benedek, T. & Vad, J., 2015b. Case-specific Empirical Guidelines for Simultaneous Reduction of Loss and Noise in Axial Flow Fans. Budapest, Hungary, CMFF '15 - Conference on Modelling Fluid Flow.

[S7] Benedek, T. & Vad, J., 2016a. An industrial on-site methodology for combined acoustic-aerodynamic diagnostics of axial fans, involving the Phased Array Microphone technique. International Journal of Aeroacoustics.

[S8] Benedek, T. & Vad, J., 2016b. Study on the Effect of Inlet Geometry on the Noise of an Axial Fan, with Involvement of the Phased Array Microphone Technique. Seoul, South Korea, ASME Turbo Expo 2016.

[S9] Kalmár-Nagy, T., Bak, B. D., Benedek, T. & Vad, J., 2015. Vibration and Noise of an Axial Flow Fan. Periodica Polytechnica - Mechanical Engineering, 59(3), pp. 109-113.

[S10] Vad, J., Halász, G. & Benedek, T., 2015. Efficiency gain of low-speed axial flow rotors due the forward sweep. Proceedings IMechE, Part A - Journal of Power and Energy, 229(1), pp. 16-23.

 

List of references

[1] Raman, G. at al., 2014. Remote detection of building air infiltration using a compact microphone array and advanced beamforming methods. Berlin, Germany, Berlin Beamforming Conference 2014.

[2] Lajos, T., 2015. Az Áramlástan Alapjai. 5th. ed. Budapest, Hungary: Dr. Lajos Tamás

[3] Carolus, T., 2003. Ventilatoren. 3rd ed. Germany: Teubner Verlag.

[4] Gruber, J., 1978. Ventilátorok. Budapest, Germany: Műszaki Könyvkiadó.

[5] Sijtsma, P., Oerlemens, S. & Holthusen, H., 2001. Location of Rotating Sources by Phased Array Measurements. Maastricht, Netherlands, 7th AIAA/CEAS Aeroacoustics Conference.

[6] Lieblein, L., 1965. Experimental Flow in Two-Dimensional Cascades. In: Design of Axial-Flow Compressors, Chapter VI. Washington D. C.: NASA SP-36.

[7] Howell, A. R., 1942. The present basis of axial flow compressor design: Part I, Cascade theory and performance. ARC R and M, 2095

[8] Brooks, T. F., Pope, D. S. & Marcolini, M. A., 1989. Air foil Self-Noise and Prediction, NASA Langley Research Center: NASA Reference Publication 1218.