This thesis presents an investigation of the influence of turbocharging on the performance and combustion behaviour of a dual fuelled, spark-ignition engine fuelled with natural gas and gasoline. The investigation was carried out using a combination of experimental and analytical methods. The experimental data was obtained from an instrumented, four cylinder, Toyota engine mounted in a test cell. An electrically driven Roots blower was used to provide compressed air to the engine, and a restriction was placed in the exhaust pipe to simulate the effects of an exhaust-driven turbine. Cylinder pressure data were recorded and analysed using a computer routine in order to provide information on mass burning rates and burning velocities. Computer routines were also developed to simulate the compression, combustion and expansion processes in the engine. It was found that the laminar burning velocity of natural gas is 50% to 60% lower than gasoline, under engine-like conditions of temperature and pressure. Mass-burning rate analyses of measured cylinder pressure data showed that the lower burning velocity of natural gas has its greatest influence during the ignition delay period (up to 1% mass burned) and that it can cause increases in ignition delay of between 50% and 100% relative to gasoline. It was observed that the low burning velocity of natural gas also affects the main combustion period, but to a much lesser extent, increasing it by up to 10% relative to gasoline. It was concluded that the main combustion period is dominated by turbulence effects and that it is relatively unaffected by variations in fuel type, air/fuel ratio or boost pressure. Results from the engine tests and simulation program indicated that it is possible to recover the power loss experienced by an engine running on natural gas by boosting the intake pressure to 3 psig (20 kPa) above that provided when the engine is running on gasoline. This increase in boost pressure does not significantly reduce the efficiency or raise the specific fuel consumption. It was found, however, that the peak cylinder pressures attained can be as much as 20% higher on natural gas than on gasoline at the same power level. / Applied Science, Faculty of / Mechanical Engineering, Department of / Graduate

Orientador: Luiz Carlos Sandoval Góes / Dissertação (mestrado profissional) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecânica / Made available in DSpace on 2018-08-19T02:42:23Z (GMT). No. of bitstreams: 1 Silva_RobertoCarlosdeCastro_M.pdf: 3523448 bytes, checksum: 056edbb15555679526f8d88b82ca288c (MD5) Previous issue date: 2011 / Resumo: Neste trabalho construiu-se e verificou-se um modelo matemático capaz de realizar o cálculo da rotação do eixo do turbocompressor aplicado a um motor Diesel, para futura implementação de uma função de software no programa da unidade de comando do motor, de forma a permitir a criação de estratégias de proteção contra sobre rotação do eixo do turbocompressor. O modelo foi construído com utilização de modelagem de fenômenos físicos que ocorrem em algumas partes do motor, junto com tabelas obtidas através de testes empíricos do motor. O modelo foi criado de modo a aproveitar os sensores atualmente disponíveis no motor de série, utilizados pelo sistema de injeção eletrônica. O parâmetro de vazão de ar admitido, calculado pela unidade de controle do sistema de injeção também foi aproveitado. Após a criação do modelo, este foi verificado inicialmente simulando-se variações nas entradas (quantidade injetada de combustível e acionamento da válvula EGR) e analisando-se a reação ocorrida na saída do modelo (rotação do eixo do turbocompressor). Foram realizadas medições em um banco de testes para o levantamento do mapa de temperatura do coletor de escape em condição estacionária. Para a verificação do modelo em um caso real, foram coletados dados de diversos parâmetros de funcionamento de um motor diesel instalado em um veículo de testes da MWM International. Os parâmetros de entrada coletados foram inseridos no modelo e o valor de rotação medido confrontado com o valor calculado. Ajustes nos submodelos foram necessários para a estimação dos parâmetros do modelo através da minimização do erro encontrado. O erro encontrado diminuiu sensivelmente, porém ainda se manteve alto para ser considerado como uma informação confiável para a implementação no software da unidade de comando do motor / Abstract: In this work, it was intended to create a model capable to calculate the turbocharger shaft speed of a Diesel engine, for future implementation of an algorithm in the engine's Electronic Control Unit software, as a protection strategy against turbocharger shaft overspeed. The model was built using physical phenomenon modeling from some parts of the engine, among tables obtained by empiric tests of the engine. The model was created in order to use current sensors available in the series production engine, which is used by electronic fuel injection system. The intake air flow parameter, already calculated by ECU, was also used. After the model creation, it was verified simulating input variation (fuel injected quantity and EGR valve command) and analyzing the model output (turbocharger shaft speed). Measurements in bench test were performed in order to create an exhaust manifold temperature map in steady condition. In order to check the model in a real case, several operating parameter data were collected from a Diesel engine installed in a test vehicle from MWM International. The acquired data was inserted in the model and the turbocharger shaft speed was crosschecked against the calculated value. Adjustments in submodels were necessary to parameter estimation through minimizing of error found. The error decreased significantly, however it remained too high to be considered a reliable information for implementation in the controller's software / Mestrado / Eletrônica / Mestre em Engenharia Automobilistica