Optimization of Automobile Air Cleaner for Effective Filter Area Utilization Using Acusolve CFD P. Balasankar Assistant Manager - R & D Mahindra Two Wheelers Ltd. D1 Block, Plot No. 18/2 (part), Chinchwad MIDC, Pune 411019. INDIA. Anish Gokhale Deputy Manager - R & D Mahindra Two Wheelers Ltd. D1 Block, Plot No. 18/2 (part), Chinchwad MIDC, Pune 411019. INDIA. Keywords : Air cleaner, porous media. Abstract The primary function of Air cleaner is to provide clean air to the engine with minimum flow restriction. The performance and the life of the engine are dependent on the quality and quantity of the air inducted. This paper focuses on the optimization of air cleaner for filter area utilization using AcuSolve CFD code. The air cleaner is analyzed to understand the flow behaviour. The filter element is modelled using porous media definition which modifies the momentum equation to add the flow resistance. The values for the resistance are obtained from the test data. Post-processing is done using HyperView to identify the zones of flow recirculation and restriction. The air cleaner geometry is optimized to reduce the flow recirculation to achieve lesser pressure drop. Guiding vanes are introduced into the air cleaner to guide the flow over the filter element for better filter area utilization. The pressure drop predicted through the CFD analysis correlates well with the test results. Introduction Air cleaner is an important element of the automobile intake system. The clean air required by the engine is supplied by the air cleaner. The filter element in the air cleaner filters the dust particles, thus the utilization of filter area is important for improving the service life of the filter element. The shape of air filter casing is critical in distributing the flow evenly over the filter element and to reduce the pressure loss due to flow separation and flow recirculation. Computational fluid dynamics play a very important role in understanding the flow behavior inside the air cleaner for improving the pressure loss and filter area utilization. The paper describes the optimization of the air cleaner geometry using AcuSolve CFD code. Model Geometry: The air cleaner geometry is shown in fig (1). The dusty air enters the air cleaner through the entry duct. The air then passes over the foam element where it gets filtered. The clean air then passes through the outlet tube to the engine. Fig. (2) shows the modified air cleaner assembly with guiding vanes to guide the flow over the filter element. The shape of the casing has been modified to reduce the recirculation zone and to improve the filter area utilization. The life of filter element can be significantly improved by utilizing the filter surface to the maximum extent.
Inlet Casing Outlet Filter element Figure 1 : Base air cleaner Guide vanes Figure 2 : Modified air cleaner
Process Methodology Mesh Setup: The fluid domain of the air cleaner was captured well with the tetrahedral mesh. Figure 3 and 4 shows the tetrahedral mesh of air cleaner and around the guide vanes. Boundary layers are inserted such that the surface Y+ is in the range of 30 to 100. Figure 3 : CFD volume mesh Figure 4 : CFD volume mesh around guide vanes Porous Media Setup: The filter element is modeled as a porous media. The material property for the filter element volume is assigned as porous using the volume manager. The porosity model [1] in AcuSolve modifies the momentum equation as in equation (1).... (1) where, ρ - density; φ - porosity; u = [ u1 u2, u3 ] T - is the velocity vector; f = [ f1 f2, f3 ] T - is the porous media contribution, R - The rotation tensor which rotates 'f' to the global coordinate axes; p - Pressure;
= [ ij ] - the viscous stress tensor; b - Specific body-force. The porous media forces are calculated from equation (2). The pressure drop vs. velocity data for the filter element is taken from the test. Curve fitting the data will give the 'A' and 'B' values and further the Darcy and Forchheimer coefficients can be found.... (2)... (3) Where, µ - the viscosity, C Darcy and C Forch - Darcy and Forchheimer coefficients k x - Permeability in the principal direction 'x' Boundary condition: S. No Boundary Type Boundary condition 1 Inlet Inflow - Stagnation pressure 2 Outlet Inflow - Mass flow rate (negative) 3 Casing Wall - no slip 4 Porous surface Internal interface Results & Discussions Figure 5, shows the velocity vectors along the cross section of the filter element, which is modeled as a porous media. The flow vectors can be seen entering the porous media radially. Figure 6 shows the pressure plot generated by AcuProbe for the inner and outer surface of the filter element (porous). Figure 7 and 8, shows the velocity vector plots for the base air cleaner geometry and the modified air cleaner geometry. In the base air cleaner, the air flow is biased towards the bottom, thereby utilizing only the bottom portion of the filter element. Majority of the filter area remains unutilized. In the modified air cleaner, the flow distribution is improved; zones of flow separation were eliminated by modifying the shape of the air cleaner as shown in the figure 8, guide vanes are provided to improve the flow over the filter element utilizing filter area to greater extent. Figure 9 and figure 10, shows the velocity contours on the surface of the filter element. The velocity distribution on the filter element for the modified air cleaner is uniform compared to the base design. The zones of low flow velocity on the filter element are significantly reduced as compared to the base.
Figure 5. Velocity vectors - Filter element Figure 6. Pressure plot - Filter inner and outer surface Figure 7. Base design - Velocity vectors Figure 8. Modified design - Velocity vectors Figure 9. Base design Velocity contours on filter surface Figure 10. Modified design - Velocity contours on filter surface
Benefits Summary The geometry optimization and prototyping time is reduced almost by 70%. The flow field visualization and understanding is possible using CFD simulation which is an additional advantage from the conventional method of geometry optimization of air filter. Conclusions The reduction in pressure drop with the modified air cleaner is 15 % compared to the base design. The filter area utilization is significantly improved. The prediction of pressure drop across the air cleaner through CFD conforms well to the test results and difference is within 10%. This is mainly because of the turbulence at the inlet boundary. ACKNOWLEDGEMENTS We would like to thank Mahindra Two Wheelers R & D employees, who are involved directly or indirectly in this work. Also we would like to thank Altair AcuSolve team for their continued support. [1] AcuSolve user manual [2] "Computational fluid dynamics", Chiang,Hoffmann REFERENCES