In this study we present the Computational Fluid Dynamics (CFD) modeling of the combustion process using detailed chemistry in a spark-ignited (SI) optical access engine operated at part load using gasoline and ethanol as fuels. Simulation results are compared against experimental optical and indicating data. The engine is installed at the Department of Engineering of the University of Perugia, and it features a four-valve head, a transparent flat piston and a port-fuel-injection (PFI) system. Full open cycle simulations have been performed using the commercial code CONVERGE. The combustion process has been simulated using detailed chemistry and adaptive mesh refinement (AMR) to resolve in detail and track the reaction zone, in a Reynolds Averaged Navier-Stokes (RANS) modeling framework. In-cylinder pressure, heat release, and flame morphology have been compared with experimental indicating and imaging data. Tests and simulations span different air-fuel ratios in lean and rich conditions (relative air-fuel ratio λranges from 0.9 to 1.1). Results indicate that simulations are able to predict experimental data with high accuracy. Variations due to changing fuel type and air-fuel ratio are well captured. The computational cost to achieve grid-independent results has been evaluated and it is also not excessively high. Taking into account that the engine speed was quite low, i.e., 900rpm, we conclude that, in this condition, detailed chemistry coupled with RANS works satisfactorily without turbulence chemistry interaction sub-models, and therefore without any tunings

The present work is the continuation of the research activity developed by the same authors in last years about the use of recent technologies (Artificial Neural Networks) for the set up of “software redundancy” modules to be implemented On Board for the use in Diagnostic Systems. In the present work, a system based on Artificial Neural Networks models for automotive engines Fault Diagnosis and Isolation purposes is set-up and analysed. Four sensors/actuators (throttle valve, rotational speed, torque and intake manifold pressure) are considered, and the respective acquired data are used to train and test four ANN modules correlating the different quantities. An FDI scheme is presented which generates fault codes sequences by suitably treating the primary residuals, obtained by comparing experimental data with the calculated ones by the ANN modules. The robust fault isolation capabilities of the proposed FDI system are presented and discussed. The proposed model architecture appears as suitable to solve either OBD problems or quality controls at factory engine production lines.