3(1): 1-7, 2018 Design Optimization of Axial Flow Fan Azeem Mustafa, Lu Yiping and Feng Mingpeng Abstract Axial flow fan occupies an important role in many industrial applications. In this paper, the computational fluid dynamics (CFD) modeling of the axial flow fan of the 7500 kw air-cooled motor is presented. The numerical simulations are performed to analyze the effect of installation angle, pressure variations and the number of blades on the performance of axial flow fan using ANSYS Fluent 1. Based on finite volume method, threedimensional turbulent flow equations are numerically solved. The results show that the volumetric flow rate and efficiency of the axial flow fan are higher when the installation angle is 30 and blades of the fan are 19. Furthermore, volumetric flow rate decreases with the increase in outlet pressure and vice versa. This paper could provide an insightful understanding for the design optimization of axial flow fan and be helpful in designing a fan to improve the overall cooling performance of the systems. Keywords Axial flow fan; Computational fluid dynamics; Design optimization; Installation angle; Mesh generation. I. Introduction The primary objective of axial flow fan is to provide a large volume of airflow parallel to the axis of rotation for heat and mass transfer operations and it is widely used in various industrial equipment and processes. All the major industries such as textile mills, fertilizer industries, chemical and pharmaceuticals, power generation etc. use different capacity axial flow fan for different purposes. However, the benefit of using axial flow fan for the purpose of augmenting heat transfer is particularly evident in those industries where large capacity motors are extensively used. These motors require large amounts of air for cooling due to the generation of large heat losses within them. Due to the wide adaptability of axial flow fan in many industries, numerical investigations have been performed in recent years to quantify its performance and characteristics. Li et al. [1] presented a modification of conventional axial fan blades with the numerical and experimental investigation to discuss the influence of adverse flow conditions at the fan hub and improve fan aerodynamic performance. B. Bizjan et al. [2] studied the effect of blade tip geometry on axial fan operation in four different operating points along with the underlying flow energy conversion and dissipation mechanisms. The effects of blade thickness on the performances of an axial-flow fan using two fans that differ only in the thickness of their blades were studied by Sarraf et al. [3]. Ye et al. evaluated the effects of an abnormal stagger angle, the pressure pulsation and transient flow field under the normal and abnormal regulation of the stagger angle using unsteady 3D modeling [4]. The Panigrahi et al. [5] presented the Computational Fluid Dynamics (CFD) simulations of drag and lift coefficients of six different airfoils using the ANSYS Fluent software to improve the fan efficiency. Hurault et al. discussed the wall pressure fluctuations in axial flow fans using CFD coupled with semi-empirical aero-acoustic models [6]. Visualization of flow around blades of axial fans for large capacity open-type motors was studied by Nakahama et al. [7]. Li et al. [8] demonstrated the efficiency improvement and pressure ratio of the forward-skewed blade fan. Based on Navier-Stokes equations and standard k-ω turbulence model, the numerical simulation of the internal three-dimensional turbulent flow field in the axial-flow fire-fighting fan under the design operating conditions was made by Shuang-lei [9] et al. The use of modern CAD and CFD techniques in the conception and simulation of industrial products has huge applications in the mechanical, automotive and aerospace industries [10] and the development of reduced forms of Navier-Stokes equations for Newtonian fluid dynamics is an active area for the designers of various mechanical and electrical components. With the development of fast and validated numerical procures, and the continuous increase in computer speed, larger and larger problems are being solved using CFD methods [11]. CFD Numerical analysis based on N-S equations has also become practical for the design and simulation of mechanical components [12-16]. Energy sector plays a long and important role in the economies of developing countries and the development of any country directly links with its energy resources. With the industrial development, per capita consumption of Technical and Scientific Publication
different energy resources continues to increase. Therefore, the scientific community is paying much attention to explore and develop such methods and techniques that not only can generate more power to meet the consumption but can optimize [17-19] the current design, methods etc. to save or minimize the use of natural resources. Combined effects of radial clearance, number and geometry of blades, installation angle, speed and other factors determine the performance of the axial flow fan. A thorough understanding of the positive and negative impact of these factors, overall layout, working principle and basic design is important to achieve a good performance index. The performance of axial flow fan is directly related to the efficiency of air-cooled mechanical components such as large motors and generators, so it requires designing it carefully through cost-effective means of simulation by the numerical solution of governing equations. Considering the scope of latest researches and profound importance of design optimization in energy sector, this paper, including all the steps from modeling of geometry up to analysis of simulation results, presents the design optimization of axial flow fan that is associated with 7500kW motor by studying the effect of pressure variation, different blade angles, and the number of blades on the volumetric flow rate and the efficiency. II. Physical Model And Mesh Generation The computational domain of three-dimensional geometric model of axial flow fan has been established using the mainstream CAD modeling software Solidworks and the structure has been made no simplification to ensure the accuracy and reliability of calculation results. The computational domain has casing and hub diameters of 1146 mm and 980 mm respectively. The axial length from hub to the casing is 93 mm, shown in Fig.1. The fan has 19 blades with an installation angle of 30, however, the number of blades and installation angle has been changed to (17 and 15) and (26 and 28 ), respectively to analyze the performance of present fan model. Like other axial flow fans, the air stream moves along the axis of the fan and the axis of rotation of the fan is coincident with the z-axis, and the positive direction of the z-axis is towards down. A fine mesh of the computational domain is generated to numerically solve the partial differential mass and momentum equations (Fig.2). The complete body of the fan has been divided into small volumes and then each volume is meshed using different mesh schemes and due to the complex geometry of the model and involvement of angular surfaces, the unstructured mesh is also used. Spin wall Inlet Outlet Blades Casing Hub Fig. 1 Physical Model of Axial Flow Fan Fan body Blades 2
Fig.2. Distribution of Mesh over Different Parts (e.g. Blades, Inlet and Hub) of the Axial Flow Fan Inspection of the mesh indicates that the total number of cells is 2786740 (approximately same for different blade angles) and skewness ratios for all volumes are less than 0.97 that ensures high-quality mesh. The mesh is also refined into those regions where large velocity and the pressure gradient are expected. After employing different interval sizes and meshing schemes, grid independence (solution) is obtained. III. Computational Method And Boundary Conditions Commercial CFD software ANSYS Fluent 1 is used for the calculations. The air that enters through the inlet of axial flow fan is in the state of turbulence (Re>5000) and incompressible (Ma<0.3). Three-dimensional flow and heat transfer coupling equations including mass and momentum conservation equation and the relation of absolute velocity vector u (the velocity viewed from the stationary frame) and relative velocity vector u r (the velocity viewed from the rotating frame) can be written as follows: ( u ) 0 (1) ( u u ) (2 u r) p F (2) r r r u ur r (3) In the fixed reference frame, mass, momentum, and energy conservation equations can be replaced by a general control equation which is given in equation (4). grad r u S (4) Where, ɸ is universal variable, which can be replaced by unknown variables u, v, w, T, etc. Γ is the generalized diffusion coefficient and S is the generalized source term. The SST k-ω model is adopted in the present work for the calculation of numerical results because it accounts for the transport of the turbulence shear stress in the definition of the turbulent viscosity making it more accurate and reliable for wider class of flows e.g. adverse pressure gradient flows, air flows and transonic shock waves than the other turbulence models and also includes two extra transport equations to represent the turbulent properties of the flow. The equations for transported variables, turbulent kinetic energy k and dissipation rate ω are given below [20]: k k kui ( k ) Gk Yk S (5) k t x x j x t i x x ui ) G Y S i j j ( (6) x j In these equations, G k represents the generation of turbulence kinetic energy due to mean velocity gradients. G ω represents the generation of ω. Γ k and Γ ω represent the effective diffusivity of k and ω, respectively. Y k and Y ω represent the dissipation of k and ω due to turbulence. D ω represents the cross-diffusion term. S k and S ω are userdefined source terms. The fan inlet pressure is set to 0 Pa, and the volumetric flow rate is calculated at different outlet pressures; 0, 100, 200-700Pa and the rated rotating speed is 1486 rpm. The values used for density and viscosity are 1.093 kg/m 3 and 1.96 10-5 kg/m s, respectively. In this paper, the Multi-Reference Frame model [20] that is a steady state approximation for multiple zones is adopted because this problem involves multiple moving parts and can t be modeled with the Single Reference Frame approach. So, the model has been broken up into multiple cell zones and zones that contain the moving components have been solved using the moving reference frame equations, whereas stationary zones are solved with the stationary frame equations. The Coupled algorithm has some advantages over the segregated approaches as it obtains a robust and efficient single phase implementation with superior performance and is used in this paper for the coupling of pressure and velocity fields [20]. A second-order upwind scheme is applied for the discretization of turbulent viscosity coefficient. The current solution is assumed converged when the difference in volumetric rate between inlet and outlet is less than 10-5, and the residuals in all directions and SST k-ω are simultaneously less than 10-4. IV. RESULTS AND DISCUSSION In this paper, computational study of axial flow fan is carried out, and the volumetric flow rate (Q v ) and efficiency (η) of axial flow fan with different installation angles (α) i.e. 26, 28 and 30 and number of blades i.e. 19, 17 and 15 is calculated at 1486rpm using the SST k-ω model with inlet pressure of 0 Pa and the outlet pressure is gradually increased from 0 Pa to 700 Pa. A. Pressure and Velocity Contours on Blades Fig.3 shows the static pressure and air velocity in the flow field. It can be seen that the velocity and pressure distributions for all blade angles follow almost similar pattern and changes from one side to another side of the blade. Pressure distributions on blades and pressure range can be seen in figures given below. The pressure intensity is 3
Q v (m 3 /s) more on that face of blades which is directly exposed to the inlet air i.e. on the front side, the intensity of pressure increases from left to right side of the blade while on the back side of the blade is higher but does not change in any direction, as it transports air from inlet to outlet. Pressure ranges from -3490 Pa to 1780 Pa at α=26. However, the velocity contours at α=26 remain the same for both sides of the blade. The magnitude of velocity on both sides is lower at the bottom of the blade and gradually increases towards top of the blade. Its value from bottom to top surface is within 76.8 m/s to 90.3 m/s. In case of α=28 and α=30, the pressure intensity and velocity distribution follows the same pattern as that in α=26 ; pressure intensity increases from left to right on the front side and is higher and does not change on the back side in each blade of two models but the their ranges are different in both models. It is -2980 Pa to 1020 Pa and - 3390 Pa to 1640 Pa for α=28 and α=30, respectively. The value of maximum and minimum velocity is 76.8 m/s and 76.2 m/s to 90.3 m/s in both cases. Fig.3b Velocity distribution over backward side blades at α=26 B. Effect of Installation Angle on Flow Rate and Efficiency Fig. 4 and 5 present the curves of volumetric flow rate and efficiency corresponding to pressure and different blade angles. Volumetric flow rate and efficiency decreases and increases respectively when pressure ranges from 0 to 700 Pa. The work performed by a fan is basically a function of geometry of blades i.e. angle and width of the blade and higher flow rate leading to higher efficiencies could be attained if the blade is tapered [21] and well designed because when the blade is more tapered it can direct more air volume towards the outlet. That s why, due to more degree among all the blades, the values of volumetric flow rate is maximum for 30 and lower for 26, 8.065 m 3 /s and 7.753 m 3 /s, respectively, at the same outlet pressure. 8.0 Q v at =26 =28 =30 Fig.3a. Pressure Distribution over front and backward side of blade at α=26 P out (Pa) Fig.4. Volumetric flow rate at different installation angles 4
65 60 55 50 26 28 30 Q v at 26 19 Blades 17 Blades 15 Blades 45 40 35 30 P outlet (Pa) Fig.5. Fan Efficiency at different installation angles The formula [22] for the calculation of fan total efficiency is given as; PQv (7) Ti Where T is the Torque, ω i is the angular velocity and P is the total pressure. From the above given formula, it can be noted that the efficiency is directly proportional to volumetric flow rate that means efficiency will be higher for higher values of volumetric flow rate and since, the volumetric flow rate for 30 (due to being more tapered) is high, as a result efficiency is also high when the blade angle is maximum (30 ) and lower for minimum blade angle (26 ), 61.13% and 54.26%, respectively at the same outlet pressure. C. Effect of Number of Blades on Flow Rate In the following figures, curves of volumetric flow rates at different installation angles and for 15, 17 and 19 blades show that the flow rate for each installation angle decreases as the outlet pressure increases. At α = 26, for 19 blades, the value of Q v, 7.422 m 3 /s is maximum and 6.943 m 3 /s is minimum for 15 blades at the same outlet pressure. Volumetric Flow rate at α = 28 follows the same trend as it is maximum for 19 blades and minimum for 15 blades and ranges from 7.753 m 3 /s to 7.124 m 3 /s at the same outlet pressure. Q v (m 3 /s) 5.5 5.0 P outlet (Pa) Fig.6. Volumetric Flow rate at α = 26 with different blades Generally, the volumetric flow rate for different numbers of blade is more than the Q v of 26 and 28 for installation angle of 30. At same outlet pressure, Q v for 19 blades is 8.065m 3 /s and for 15 blades is 7.417m 3 /s. It can be concluded that the volumetric flow rate is more when there are more number of blades and it follows the principle that the volumetric flow rate provided by a fan is proportional to its number of blades [22]. In the presence of more blades, more air could be transported towards outlet consequently making it more feasible for the cooling of large capacity motors. Q v (m 3 /s) 5.5 Q v at 28 19 Blades 17 Blades 15 Blades 5.0 P out (Pa) Fig.7. Volumetric Flow rate at α = 28 with different blades 5
Q v (m 3 /s) 8.0 5.5 Q v at 30 19 Blades 17 Blades 15 Blades P outlet (Pa) Fig.8. Volumetric flow rate at α = 30 with different number of blades numbers D. Comparison of Turbulence Models The two turbulence models: SST k-ω and standard k-ε are compared in order to investigate their effects on the numerical simulation results. The results of standard SST k- ω model and standard k-ε model are almost similar; however, the calculation accuracy of SST k-ω model is slightly better than that of standard k-ε model. Overall, for all calculations (for different number of blades and angle), the SST k-ω model achieved the same convergence accuracy with less iterations. V. Conclusion In this paper, the design optimization of axial flow fan associated with 7500kW motor is presented and using CFD theories, the effect of installation angle and number of blades over its volumetric flow rate and efficiency is analyzed. This research concludes; 1. Volumetric flow rate of present research object decreases as the outlet pressure increases and its values are maximum at P out =0Pa. 2. As the blade installation angle increases, volumetric flow rate also increases and its value at 30 is higher than at other angles. 3. The present research shows more blades transfer more flow rate; consequently the volumetric flow rate is higher when fan consists of 19 blades. 4. Efficiency of axial flow is in direct relation with volumetric flow rate, and it increases as the volumetric flow rate increase (vice versa), hence is maximum at 30 as well as for 19 blades. References [1] Li, Z., Jin, Y., Huashu, D., & Yuzhen, J. (2013). Numerical and experimental investigation on aerodynamic performance of small axial flow fan with hollow blade root. Journal of Thermal Science, 22(5), 424-432. [2] Bizjan, B., Milavec, M., Širok, B., Trenc, F., & Hočevar, M. (2016). Energy dissipation in the blade tip region of an axial fan. Journal of Sound and Vibration, 382, 63-72. [3] Sarraf, C., Nouri, H., Ravelet, F., & Bakir, F. (2011). Experimental study of blade thickness effects on the overall and local performances of a Controlled Vortex Designed axial-flow fan. Experimental Thermal and Fluid Science, 35(4), 684-693. [4] Ye, X., Ding, X., Zhang, J., & Li, C. (2017). Numerical simulation of pressure pulsation and transient flow field in an axial flow fan. Energy, 129, 185-200. [5] Panigrahi, D. C., & Mishra, D. P. (2014). CFD Simulations for the Selection of an Appropriate Blade Profile for Improving Energy Efficiency in Axial Flow Mine Ventilation Fans. Journal of Sustainable Mining, 13(1), 15-21. [6] Hurault, J., Kouidri, S., & Bakir, F. (2012). Experimental investigations on the wall pressure measurement on the blade of axial flow fans. Experimental Thermal and Fluid Science,40, 29-37. [7] Nakahama, T., & Ishibashi, F. (2004). Analysis and visualization of flow around blades of axial fans for large capacity open-type motors. IEE Proceedings - Electric Power Applications, 151(1), 70. [8] LI, Y., LIU, J., OUYANG, H., & DU, Z. (2008). Internal flow mechanism and experimental research of low pressure axial fan with forward-skewed blades. Journal of Hydrodynamics, Ser. B, 20(3), 299-305. [9] Chu, S. L., Yu, G. S., Qin, R. H., & Cheng, C. (2009, December). The numerical simulation of internal threedimensional flow in the axial-flow fire-fighting fan. In Test and Measurement, 2009. ICTM'09. International Conference on(vol. 2, pp. 351-354). IEEE. [10] Sayma A. (2017). Computational Fluid Dynamics. Abdulnaser Sayma & Ventus Publishing Aps. 6
[11] Menon, K. G., & Patnaikuni, V. S. (2017). CFD simulation of fuel reactor for chemical looping combustion of Indian coal. Fuel, 203, 90-101. [12] Abdel-Fattah, A., Fateen, S. K., Moustafa, T. M., & Fouad, M. M. (2016). Three-dimensional CFD simulation of industrial Claus reactors. Chemical Engineering Research and Design, 112, 78-87. [13] Liu, N., Wang, W., Wang, Y., Wang, Z., Han, J., Wu, C., & Gong, J. (2017). Comparison of turbulent flow characteristics of liquid-liquid dispersed flow between CFD simulations and direct measurements with particle image velocimetry. Applied Thermal Engineering, 125, 1209-1217. [14] Huang, M., Gowdagiri, S., Cesari, X. M., & Oehlschlaeger, M. A. (2016). Diesel engine CFD simulations: Influence of fuel variability on ignition delay. Fuel, 181, 170-177. [15] Lu, Y., Liu, L., & Zhang, D. (2015). Simulation and Analysis of Thermal Fields of Rotor Multislots for Nonsalient-Pole Motor. IEEE Transactions on Industrial Electronics,62(12), 7678-7686. [16] Hassan, A., & Abomoharam, M. (2017). Modeling and design optimization of a robot gripper mechanism. Robotics and Computer-Integrated Manufacturing, 46, 94-103. [17] Jahan, S. A., Wu, T., Zhang, Y., Zhang, J., Tovar, A., & Elmounayri, H. (2017). Thermo-mechanical Design Optimization of Conformal Cooling Channels using Design of Experiments Approach. Procedia Manufacturing, 10, 898-911. [18] Si, B., Tian, Z., Jin, X., Zhou, X., Tang, P., & Shi, X. (2016). Performance indices and evaluation of algorithms in building energy efficient design optimization. Energy, 114, 100-112. Azeem Mustafa Mechanical and Power Engineering Harbin University of Science and Technology Harbin University of Science and Technology, P.R. China 1667802103@qq.com Lu Yiping Mechanical and Power Engineering Harbin University of Science and Technology, P.R. China luyp2010@yahoo.com.cn Feng Mingpeng Mechanical and Power Engineering Harbin University of Science and Technology, P.R. China 992164191@qq.com [19] Schayes, C., Vogt, J., Bouquerel, J., & Palleschi, F. (2015). Rotor Design Optimisation through Low Cycle Fatigue Testing. Procedia Engineering, 133, 233-243. [20] Ansys Fluent 1 Theory Guide (2016). [21] Monroe, C (1997). Maximizing Fan Performance. (1997). Hudson Products Corporation. [22] Yan Xiao-kang, Wang Li-jun, Zhang Jing-song, & Wang Xin-yong. (2010). Numerical and experimental investigation on effect of installation angle of rotor blade on axial flow fan. 2010 International Conference on Mechanical and Electrical Technology. 7