Investigation of the Internal Flow Field of a Diesel Model Injector Using Particle Image Velocimetry and CFD Imperial College London, U.K

Format:

Conference/Event

Author/Creator:

Aleiferis, P.G., author.

Conference Name:

JSAE/SAE International Fuels & Lubricants Meeting (2007-07-23 : Kyoto, Japan)

Language:

English

Physical Description:

1 online resource

Place of Publication:

Warrendale, PA SAE International 2007

Summary:

Controlling the spray characteristics of a Diesel injector means understanding the internal flow field and the way that cavitation is initiated. Measurements of the internal flow field of an injector are rare, although they provide the appropriate information both on the flow pattern and the initiation of cavitation in order to assist the evaluation of computer predictions of flow and cavitation. The purpose of the current work is to report measurements of the internal flow of a Diesel injector and assess the ability of computational fluid dynamics to predict the flow behaviour. Two-Dimensional Particle Imaging Velocimetry (PIV) technique was employed to measure the internal flow field of a Diesel injector. The experiments were conducted by using 20:1 scale transparent models of different sections upstream of the injection nozzle of a commercial Diesel Injector. The same geometries were imported to the Computational Fluid Dynamics code STAR-CD and by using tetrahedral mesh and two different turbulence models (High Reynolds k-ε and k-ε RNG), the flow within these Diesel injector sections has been calculated. The mean and instantaneous flow patterns within sections in planes parallel to the notional axis of the injector were quantified with PIV. Velocity vector maps were quantified by averaging over 1000 pairs of images. Cavitation inside the nozzles was visualised, by means of shadowgraphy, for the cases that occurred. The CFD code predicted the average flow field within the same sections as the experiments and comparison was made. Unsteady Reynolds Averaged Navier Stokes two phase calculations have also been conducted for the cases that experiments showed initiation of cavitation. It was found that the flow field within the measured section was predicted better with k-ε RNG model and in all cases the separation zones were smaller in the calculations than in the experiment, something which is important for the prediction of cavitation. The CFD code predicted initiation of cavitation at higher flow rates than that of the experiments

Notes:

Vendor supplied data

Publisher Number:

2007-01-1897

Access Restriction:

Restricted for use by site license