Search Results

An automatic hexahedral meshing technique has been combined with CFD to predict the steady flow through practical intake port/valve geometries. Two intake configurations were investigated. An asymmetric design consisting of one helical and one directed port and a symmetric design consisting of two directed ports. The automatic meshing technique dramatically reduces the mesh construction time to less than 15% of that achievable by the conventional route. Detailed LDA, laser sheet, and swirl measurements reveal that for certain flow conditions and port arrangements it is possible to obtain a good match between the measured and predicted in-cylinder flow structure. The best predictions of in-cylinder swirl were within 5% of the measured values. It is also shown that this approach to mesh generation is both economic and robust for practical engine designs, transforming CFD into a practical tool for the design of intake ports.

LDV is used to measure steady flow in a two port loop scavenged model two-stroke engine cylinder. The model cylinder, machined from acrylic for maximum optical access, is geometrically identical to that used in a previous dynamic study of transfer port efflux vectors. The measured flow field is compared with a CFD prediction which employs experimentally measured velocity, mass flow rate, and turbulence intensity as the inlet boundary condition at the transfer port. The finite volume prediction, using the PHOENICS general purpose code recreates the global flow pattern well, but shows some local discrepancies in flow direction and magnitude. Levels of turbulent kinetic energy were poorly recreated using a k-ϵ model of turbulence, especially around impingement of the incoming jets where local errors of up to 60% were seen.

The measurement of transfer port efflux velocities using laser doppler velocimetry (LDV) in a motored model two-stroke engine is described. The single cylinder engine used is of two port loop scavenged design, externally blown to provide scavenge flow into the cylinder during the entire port open period. LDV measurements were taken along a vertical path, central to the transfer duct, at the port exit over a range of crankangles at motoring speeds of 225rpm, 600rpm, and 900rpm. At 225rpm further measurements were taken for a range of delivery ratios from 0.7 to 2.0. Relatively uniform velocity profiles indicate plug like flow issuing from the port under most conditions. The resultant flow direction is seen never to align with the transfer duct walls, but to vary as a function of crankangle. Quantitative analysis of angles defining mean flow direction reveal that dynamic efflux behaviour is essentially similar for all tested speeds and delivery ratios.