De Felice, Massimiliano (2024) DEVELOPMENT AND VALIDATION OF NOVEL COMBUSTION AND EMISSION MODELS FOR SUSTAINABLE INTERNAL COMBUSTION ENGINES. [Tesi di dottorato]

Documento PDF PhD_Thesis_De_Felice_Massimiliano.pdf Visibile a [TBR] Amministratori dell'archivio Download (13MB) | Richiedi una copia

Abstract

The internal combustion engine represents one of the most influential inventions of mankind, able to revolutionize and drastically improve people's lives. Nevertheless, it is also one of the main responsible for climate change, being the most utilized engine in the transportation sector. Therefore, a lot of efforts are bent on finding solutions for the environmental impact reduction of the transportation sector, and consequently of the internal combustion engines. Different pathways are available, that can be summarized as follows: improvements on technologies, controls and combustion for the actual internal combustion engines, as valve strategies and low temperature combustions; utilization of e-fuels in the internal combustion engines, able to obtain a net-zero environmental impact; application of carbon capture systems that can remove the CO2 from the exhaust gases; hybridization and electrification of the vehicles. In this situation, it does not appear obvious which of the pathways is the absolute best. It is therefore necessary to study the different possibilities in order to understand the advantages and the drawbacks. In this scenario, simulation tools represent a powerful instrument, since different technical solutions can be studied with reduced computational and economical costs. In light of the above considerations, the topic of the research activity, presented in this Ph.D. thesis, is the development and the validation of a 0D internal combustion engine cylinder model capable to simulate and predict performances and pollutant emissions of a wide variety of combustion concepts with different fuels. The 0D model, integrated into a 1D commercial software to simulate the pipe flows of an internal combustion engine, employs a multi-zone approach to compute the possible species and temperature inhomogeneities inside the combustion chamber, different combustion models to handle various combustion concepts, and emission models to compute the exhaust uHC, CO and NOx concentrations, and consequently the environmental impact of the internal combustion engines. For this goal, several sub-models are integrated into the model framework: an interzonal flow model, consisting of the spray model to calculate the injected fuel distribution (derived from the spray jet model, integrated in the framework in this activity and improved with the out-flow mechanism), and the diffusion model to calculate the species and temperature variation inside the chamber; an evaporation model, available in a simplified version for most of the calculations and in a detailed version for the spray model (derived from the droplet evaporation model and integrated in the framework in this activity); an heat transfer model; an ignition model, to compute the ignition evolution through a tabulated approach (originally developed for conventional SI engines and extended to work on HCCI and RCCI combustion); a flame propagation model, that is based on the fractal model and utilizes the information of the turbulence model (extended to work on RCCI combustion); a CO and NOx model, to estimate the pollutant concentrations derived from the combustion; a uHC model, consisting of different elements, to calculate the uHC exhaust concentration (improved with the temperature calculation for the crevice volumes, the possibility to handle flame propagation into the crevices, an additional contribution to the oil layer for cold engine conditions, a tabulated approach to estimate uHC post-oxidation, and the development of scavenging and short circuit models for exhaust valve composition estimation). In order to validate the 0D cylinder model, different analyses are performed. Firstly, the uHC model is validated with respect to experimental uHC concentrations of different engines, moving from automotive to marine engines and with different fuels. Thereafter, the flame propagation model is validated with respect to two SI engines, one fueled with methanol and the other fueled with gasoline but working in oxy-fuel conditions. Subsequently, the ignition model in the multi-zone framework is validated with respect to 4 different engines running in HCCI conditions and fueled with hydrogen, methane, n-heptane and blends of n-heptane, toluene and ethanol. In the end, the spray model and the complete 0D cylinder model are validated with respect to a marine engine fueled with natural gas and LFO running in RCCI conditions. This Ph.D. activity demonstrates the predictive capabilities of the 0D model in all its sub-models. In particular, the model proves the capability to accurately predict the experimental uHC exhaust concentrations considering different geometrical features, operating conditions, fuel metering, different engines and fuels, thanks to the different uHC contributions (crevices, quenching, oil, bypass, short circuit), the boundary layer and the post-oxidation models, resulting in an uHC average error of 13.13% on a total of 134 operating conditions. In addition, the model manages to predict conventional SI combustion with different engine loads, mixture qualities and other different operating conditions, considering different fuels and oxidants, resulting in an IMEP average error of 0.85% on a total of 64 operating conditions. Further, the model demonstrates an excellent predictive capability with HCCI combustion thanks to the tabulated ignition model and the multi-zone approach. The model is able to obtain an average error 1.0% on the peak pressure and 0.34 CAD on its phasing. To conclude, the model shows an accurate prediction of the RCCI combustion in terms of performances and pollutant emissions thanks to the spray and diffusion models, the combination of flame propagation and ignition models, and the pollutant emission models, resulting in an average error of 1.76 CAD on the CA50 and an error of 18.76% on the uHC concentration.

Downloads

Actions (login required)

Modifica documento