Modern Diesel engines have become ever more complex systems with many degrees of freedom. Simultaneously, with increasing computational power, simulations of engines have become more popular, and can be used to find the optimum set up of engine operation parameters which result in the desired point in the emission-efficiency trade off. With increasing number of engine operation parameter combinations, the number of calculations increase exponentially. Therefore, adequate models for combustion and emissions with limited calculation costs are required. For obvious reasons, the accuracy of the ignition timing is a key point for the following combustion and emission model quality. Furthermore, the combination of mixing and chemical processes during the ignition delay is very challenging to model in a fast way for a wide range of operation conditions. This work focuses on the description of a physics-rich diesel engine spray ignition model, which offers the possibility of real-time calculation.
The spray ignition model uses a simplified 3-Arrhenius approach for the modelling of chemical reactions, with time-variant conditions of temperature and mixture fraction replicating the history of the igniting fuel portion. To identify the mechanisms for fuel mixing and ignition conceptually, 3D computational fluid dynamics simulations were performed using a conditional moment closure (CMC) combustion model. A Lagrangian tracking method which is capable of capturing spray pattern and fuel trajectory in liquid and vapour phases was applied to characterise the mixing history of the fuel. The result of the tracking is the information of local equivalence ratio and temperature along the trajectory of the igniting fuel portion of the spray, from injection to ignition. A variation in operating conditions (changing nozzle diameter, injection pressure, ambient temperature and fuel temperature) including reactive and non-reactive calculations provided a broad understanding of the role of mixing and low temperature reactions in the trajectory of the igniting fuel. For the spray ignition model, this trajectory has been deduced in a conceptual way, using distinct phases that could be identified from the CFD simulation. The resulting equivalence ratio and temperature evolution has been used to complete an ignition integral using a simplified 3-Arrhenius approach. The final model contains 6 parameters, which need to be adapted for operation in a Diesel engine.
The model has been calibrated on a medium-speed diesel engine and validated for a broad variation of engine operating conditions (load, boost pressure, start of injection, EGR, intake temperature, intake valve timing, etc.) in two different engines. The accuracy of the modelled versus the measured ignition delay is very good. The dominant effects of the spray ignition process have been identified and successfully transformed into a simplified description, which can be used for real-time calculations.
