Stratified scavenging in two-stroke engines using OpenFOAM Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 1
Acknowledgements I would like to thank Associate Professor Håkan Nilsson at the Division of Fluid Dynamics at the Department of Applied Mechanics. Erik Lindén and Johan Spång at Husqvarna AB. Tommaso Lucchini at Politecnico di Milano Mikael Bergman at Husqvarna AB. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 2
Outline Aim and purpose. Approach. The two-stroke engine and the stratified charged two-stroke engine. CFD. Mesh operations. Boundary conditions. Results. Future work. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 3
Aim and purpose Evaluate the scavenging performance of the Husqvarna H576X engine. Evaluate the use of OpenFOAM for two-stroke engine simulations. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 4
Approach Determine interesting information. Limit the problem. Method to find the information. Possibilities in OpenFOAM. Set up the case. Extract the results. Validation. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 5
The two-stroke engine The conventional two-stroke engine. The cycle of operation. Advantages and disadvantages. Trapping efficiency. η tr = Mass of delivered air fuel mixture retained Mass of delivered air fuel mixture Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 6
Expansion stroke Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 7
Compression stroke Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 8
The stratified charged two-stroke engine Reduction of emissions and a lowered fuel consumption. The cycle of operation. Trapping efficiency. η tr,fuel = Mass of delivered fuel retained Mass of delivered fuel Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 9
Expansion stroke Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 10
Compression stroke Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 11
The H576X geometry Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 12
The H576X geometry description Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 13
Geometry modifications Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 14
CFD and the Finite Volume Method The Reynolds-Averaged Navier-Stokes equations. Turbulence modelling using the standard k ε turbulence model. Wall functions. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 15
Fluids Approximation with air as the only fluid. No combustion, the combustion is modelled. Air fuel mixture for trapping efficiency Passive scalar transport Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 16
The passive scalar Adding a passive scalar, φ to represent the air fuel mixture. Convection dominated flow. The governing equations independent of the passive scalar. Volume concentration of φ in the cell. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 17
The passive scalar, continued The passive scalar transport equation. (ρφ) t +div(ρuφ) = 0 The cell volume passive scalar concentration. 0 φ 1 A compressible flow requires the mass to be calculated m fuel = ρ φ V cell Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 18
Mesh operations The solver must manage some mesh operations. The cylinder volume is being compressed and expanded, layeradditionremoval. The cylinder and the ports must interact, slidinginterface. The twostrokeengine-library handle these features. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 19
Layer addition Morphing to a certain limit. Layer addition. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 20
Layer removal Compression to a certain limit. Layer removal. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 21
Sliding interface Sliding interfaces. Ports, cylinder and piston pockets. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 22
Sliding interface, continued Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 23
Boundary conditions, inlets and outlets Scavenging channel inlets. Exhaust outlet. Stratifying channel inlets. Flow both into and out of the domain. Adaptive B.C. for temperature, T, velocity,u, turbulent kinetic energy, k, dissipation of turbulent kinetic energy, ε, and the passive scalar, φ. Flow into the domain, Dirichlet B.C. Flow out of the domain, homogenous Neumann B.C. φ x n = 0 Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 24
Boundary conditions, inlets and outlets, continued Temperature set according to results from 1-d simulations. Pressure set according to results from 1-d simulations. Turbulent properties, k and ε are set to small values. Thepassivescalarissettoa100%concentrationinthescavengingchannelinlet. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 25
Boundary conditions, inlets and outlets, continued Isolated walls. Homogenous Neumann for temperature and pressure and the passive scalar No slip for velocity Wall functions used for k and ε Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 26
Combustion modelling No actual combustion. Modelled combustion. Cylinder pressure and temperature set before exhaust port opens. Pressure from 1-d simulations. Temperature from 1-d simulations exhaust channel. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 27
Results, scavenging Trapping efficiency about 93%. Solution is periodic within four revolutions. Trapping efficiency from experiments 92%. 360 CAD 720 CAD 1080 CAD 1440 CAD 99.34 % 93.14 % 93.61 % 93.77 % Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 28
Results, validation Comparison in massflow over scavenging channel inlet with 1-d simulations. 0.02 0.015 1-d data OpenFOAM 0.01 0.005 Massflow [kg/s] 0-0.005-0.01-0.015-0.02-0.025-0.03-0.035 0 100 200 300 400 500 600 700 800 Time [CAD] Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 29
Results, validation, continued Comparison in massflow over stratifying channel inlet with 1-d simulations. 0.008 0.006 1-d data OpenFOAM 0.004 0.002 Massflow [kg/s] 0-0.002-0.004-0.006-0.008-0.01-0.012-0.014 0 100 200 300 400 500 600 700 800 Time [CAD] Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 30
Results, validation, continued Comparison in massflow over exhaust outlet with 1-d simulations. 0.06 0.05 1-d data OpenFOAM 0.04 Massflow [kg/s] 0.03 0.02 0.01 0-0.01 0 100 200 300 400 500 600 700 800 Time [CAD] Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 31
Results, validation, continued Comparison in massflow over scavenging channel inlet with Fluent simulations. 0.03 0.02 OpenFOAM Fluent 0.01 Massflow [kg/s] 0-0.01-0.02-0.03-0.04 0 100 200 300 400 500 600 700 800 Time [CAD] Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 32
Results, validation, continued Comparison in massflow over stratifying inlet with Fluent simulations. 0.008 0.006 OpenFOAM Fluent 0.004 0.002 Massflow [kg/s] 0-0.002-0.004-0.006-0.008-0.01-0.012-0.014 0 100 200 300 400 500 600 700 800 Time [CAD] Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 33
Results, validation, continued Comparison in massflow over exhaust outlet with Fluent simulations. 0.04 0.035 OpenFOAM Fluent 0.03 Massflow [kg/s] 0.025 0.02 0.015 0.01 0.005 0-0.005 0 100 200 300 400 500 600 700 800 Time [CAD] Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 34
Conclusions, OpenFOAM Useful for two-stroke engine simulations. Approximately 24h per revolution with checkmesh on 3 CPU:s. Periodic behavior within four revolutions. Results in correlation with commercial CFD-code and 1-d simulations. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 35
Future work Non-modified geometry. Expand the geometry. The mesh impact on the solution. Add more species and combustion. Heat transfer with walls. Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 36