CFD codes are now widely used to investigate the in-cylinder thermofluid-dynamic processes. Despite the availability of complex physical models, the results are strongly influenced by the mesh quality, which should be kept high for the whole simulation even in presence of extreme boundary deformations like valve penetration into the piston bowl or piston approaching the cylinder head at the top dead center. For this reason the whole simulation is generally covered by using a different number of meshes and each one of them is used for a certain crank angle interval. This provides good mesh quality and resolution when efficient algorithms for field mapping and grid point motion are available: interpolation-related errors should be avoided and the number of target meshes should be kept at minimum to reduce the user turnaround time required for mesh generation and motion. In this context, the authors have implemented in the OpenFOAM code moving mesh algorithms for internal combustion engine simulations. The grid points motion is described by a vertex-based unstructured mesh motion solver: the Laplace equation of motion is solved with variable diffusion on the mesh vertices, using a Finite Element method with polyhedral cell support. This guarantees that an initial valid mesh remains valid for arbitrary boundary motion. Different expressions for the motion diffusion can be used to improve the quality of the mesh during motion. A cell based, distance weighted interpolation technique is used to interpolate the geometric fields from the source to the target mesh. The proposed approach has been applied to simulate the in-cylinder flows in different engine geometries. Firstly a flat-head geometry with a centrally located valve was used to evaluate the efficiency of the field mapping and mesh motion algorithms. Then two real engine configurations were modeled: for both of them the simulation of the intake, compression and combustion phases was performed and a comparison with experimental data of in-cylinder pressure is provided.
A Finite-Element Based Mesh Motion Technique for Internal Combustion Engine Simulations / Lucchini T.; D'Errico G.; Brusiani F.; Bianchi G. M.. - ELETTRONICO. - (2008), pp. 1-8. (Intervento presentato al convegno 7th COMODIA International Conference on Modeling and Diagnostics for Advanced Engine Systems tenutosi a HOKKAIDO UNIVERSITY, SAPPORO, JAPAN nel 28-31 Luglio 2008).
A Finite-Element Based Mesh Motion Technique for Internal Combustion Engine Simulations
BRUSIANI, FEDERICO;BIANCHI, GIAN MARCO
2008
Abstract
CFD codes are now widely used to investigate the in-cylinder thermofluid-dynamic processes. Despite the availability of complex physical models, the results are strongly influenced by the mesh quality, which should be kept high for the whole simulation even in presence of extreme boundary deformations like valve penetration into the piston bowl or piston approaching the cylinder head at the top dead center. For this reason the whole simulation is generally covered by using a different number of meshes and each one of them is used for a certain crank angle interval. This provides good mesh quality and resolution when efficient algorithms for field mapping and grid point motion are available: interpolation-related errors should be avoided and the number of target meshes should be kept at minimum to reduce the user turnaround time required for mesh generation and motion. In this context, the authors have implemented in the OpenFOAM code moving mesh algorithms for internal combustion engine simulations. The grid points motion is described by a vertex-based unstructured mesh motion solver: the Laplace equation of motion is solved with variable diffusion on the mesh vertices, using a Finite Element method with polyhedral cell support. This guarantees that an initial valid mesh remains valid for arbitrary boundary motion. Different expressions for the motion diffusion can be used to improve the quality of the mesh during motion. A cell based, distance weighted interpolation technique is used to interpolate the geometric fields from the source to the target mesh. The proposed approach has been applied to simulate the in-cylinder flows in different engine geometries. Firstly a flat-head geometry with a centrally located valve was used to evaluate the efficiency of the field mapping and mesh motion algorithms. Then two real engine configurations were modeled: for both of them the simulation of the intake, compression and combustion phases was performed and a comparison with experimental data of in-cylinder pressure is provided.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
