TY  - THES
AU  - Perini, Federico
T1  - Optimally reduced reaction mechanisms for Internal Combustion Engines running on biofuels
Y1  - 2011/04
PB  - University of Modena and Reggio Emilia

UR  - http://www.himech-phdschool.unimore.it/site/home/download/tesi-di-dottorato/documento92016812.html

KW  - Biofuels
KW  - Chemical Kinetics
KW  - Genetic Algorithms
KW  - Internal Combustion Engines
KW  - Mechanism reduction
KW  - Optimization



N1  - Owner: Federico
N1  - Added to JabRef: 2012.01.17
N2  - Research for new combustion concepts for internal combustion engines has been made possible in recent years by the adoption of CFD codes which are capable of computing complete reaction mechanisms, as current pollutant regulations are leading to an increasing need for accurate predictions of the spatial distribution and time evolution of species within the combustion chamber. However, the adoption of full or detailed reaction mechanisms in multi-dimensional studies is still too computationally demanding, and the development of accurate reduced mechanisms is of fundamental importance for mantaining the predictive capabilities of the simulations. In this context, research is urged by the need for finding adequate substitutes to petroleum-based fuels for the transportation sector of the near future. Biofuels, such as bioethanol, biobutanol, biodiesel currently seem to be suitable for gradually being introduced as additives to traditional hydrocarbon fuels, and then to fully substitute them, to contribute to a viable and environmentally sustainable alternative to oil. The present work introduces a novel approach for the automatic development of reduced reaction mechanisms for biofuels' combustion, with the aim of generating reaction mechanisms which can be efficiently adopted for engine simulations with detailed chemistry. \p>First of all, efforts have been devoted to the development of a new computer code for the computationally efficient solution of detailed chemistry in the simulation of combustion systems. The software has been coded as a package for the estimation of finite-rate chemical kinetics in zero dimensional batch reactors and internal combustion engines. All the routines have been developed in a fully vectorized fashion, for maximum computational efficiency. The performance of the code has been compared adopting different ODE solvers, showing robustness and computational efficiency. Finally, the code has been coupled with KIVA-4, well-known CFD tool for the simulation of internal combustion engines, and parallelised adopting a shared memory paradigm. \p>Performance of the solution of the chemistry ODE system has then been improved through the study of a novel explicit ODE solver, exploiting time scale separation. The solution of the reacting environment is carried out according to the assumption that, if the time integration interval is small enough, a linear analysis can yield accurate predictions of the characteristic times of each species, and thus the evolution of each of them can be stopped after its estimated expiration time. This method thus significantly reduces the stiffness of the ODE system, and allows a significant reduction in total computational times, if compared to the more commonly used implicit integration approach, which cannot exploit its potential if bounded by strict integration timesteps (as happening during CFD simulations, where the fluid-dynamic timestep usually rules over the chemistry source terms).\p>Once that all the numerical tools have been validated and tested, the mechanism reduction procedure has been set up as an iterative algorithm, with the aim of gradually reducing the number of species involved in the mechanism, while still mantaining its predictiveness in terms of not only ignition delay times, but also the time evolution of important species. In particular, a global error function is defined taking into account a set of ignition delay calculations at different, engine-relevant, initial mixture compositions, temperatures and pressures. The choice of the species to be deleted is performed exploiting the element flux analysis method; when a global error function of the reduced mechanism exceeds the required accuracy, the collision frequencies and activation energies of selected reactions are corrected by means of a GA-based code. The results show that significantly smaller, accurate reaction mechanisms can be automatically generated and applicated to internal combustion engine simulations. Further research needs to be devoted to the development of skeletal, multi-component reaction mechanisms for allowing new combustion concepts such as multi-fuel combustion and reactivity-controlled combustion to be explored.
ER  -