Paper presented to the 10th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics, Florida, 14-16 July 2014.
To improve the efficiency of diesel engines several actions have been performed. Beside the reduction of internal friction by applying new and more effective materials, the thermodynamics offer different opportunities to influence the engine performance and emissions. Especially the interaction of fuel and air flow within the combustion chamber should be investigated. Therefore an optimized in-cylinder flow can enhance air-fuel mixing and lead to lower exhaust emissions and less fuel consumption. In addition the turbulent flow within the cylinder exhibits large- and small-scale cyclic variations. The turbulence characteristics and cycle to cycle fluctuations of the in-cylinder flow can also have a pronounced influence on the combustion process and evoke the need to be thoroughly investigated.
Nowadays CFD simulations are widely used to predict and optimise the air flow within internal combustion engines, but such numerical calculations require a comparison with experimental data. Particle Image Velocimetry (PIV) was applied on an optically accessible single cylinder diesel engine to receive data sets for validation purpose. This engine provides access through a glass ring with 30 mm height and a modified piston bowl with an integrated glass bottom. In order to detect the local dependencies of the air flow, the velocity fields were quantified in two horizontal and three vertical measurement planes. A high resolution double shutter camera and a high energy double pulse laser were applied to measure the velocity fields during the intake and compression phases. In order to study the cyclic fluctuations, instantaneous snapshot pairs from 100 successive cycles were taken. As a result of the strong turbulences inside the internal combustion chamber strong cyclic fluctuations were observed at all investigated measurement planes.
The measured averaged in-cylinder velocities and indicated pressures were compared with a k-ε turbulence model simulation, based on an Unsteady Reynolds Averaged NavierStockes (URANS) approach. In addition, the swirl and tumble characteristics were calculated and checked against the measured ones. Despite a sufficient accordance between the experimentally determined and k-ε model calculated in-cylinder velocities, discrepancies in swirl and the tumble flow could be observed. Therefore the air flow was also simulated using a Scale-Adaptive Simulation (SAS) turbulence model in order to resolve the small scale turbulences inside the combustion chamber. The achieved solution was compared with the velocities fields, averaged over 100 cycles, as well as the single cycle velocities to avoid the elimination of small scale turbulences due to cyclic variations.