A carregar...
Publicação
Dinâmica de Veículos Elétricos e Veículos a Combustão: Análise, Comparação e Validação de Resultados
Orientador(es)
Resumo(s)
O presente trabalho tem como objetivo analisar e comparar o comportamento dinâmico de
um veículo elétrico e de um veículo com motor de combustão interna, procurando
compreender de que forma a arquitetura do sistema de propulsão influencia a sua dinâmica,
analisando variáveis como aceleração lateral, velocidade angular, forças laterais e cargas
verticais. Para esse efeito, foi desenvolvido um modelo computacional em Python,
inicialmente baseado no modelo de bicicleta e posteriormente evoluído para um modelo de
quatro rodas, integrando a Fórmula Mágica de Pacejka, um algoritmo de controlo Pure
Pursuit e uma representação tridimensional do ensaio. Paralelamente, foi concebido, testado
e adaptado um sistema de aquisição de dados, utilizado na realização de ensaios
experimentais com os veículos em estudo. Adicionalmente, foram analisados os ensaios
dinâmicos normalizados e efetuados na indústria automóvel e selecionada o teste prático a
realizar com os veículos e a implementar no modelo computacional. Por fim, foram feitos
os ensaios experimentais e numéricos dos veículos e analisados os seus resultados.
A comparação numérica e experimental inicial do modelo computacional, realizada com
dados experimentais da equipa de Fórmula Student do Instituto Politécnico de Leira (LART),
demonstrou uma boa aproximação aos resultados reais. No modelo de quatro rodas, o ensaio
de aceleração 0–75 m apresentou um tempo calculado de 4,73 s face a um valor experimental
de cerca de 4,8 s, enquanto no ensaio de Skidpad o desvio face ao valor real foi de
aproximadamente 2,03%. Estes resultados permitiram considerar o modelo adequado para a
análise comparativa proposta.
Os ensaios numéricos e experimentais evidenciaram diferenças entre os dois veículos, mas
de magnitude reduzida. Embora o veículo elétrico tenha apresentado valores ligeiramente
superiores de aceleração lateral e velocidade angular em torno do eixo vertical no ensaio
experimental, os desvios globais observados situaram-se aproximadamente entre 1% e 2,5%,
que seriam 3% considerando uma hipótese de igualdade rigorosa de raio entre ensaios.
Assim, não foi possível concluir que exista uma diferença dinâmica marcadamente
expressiva entre os veículos testados nas condições consideradas. A comparação entre os
resultados numéricos e experimentais mostrou ainda que o modelo desenvolvido reproduziu
de forma coerente a tendência das principais variáveis analisadas, apesar de algumas
diferenças nos valores máximos absolutos, associadas às simplificações do modelo e às
limitações da aquisição de dados.
Conclui-se que os objetivos definidos foram globalmente cumpridos, tendo sido
desenvolvido e validado um modelo computacional capaz de caracterizar o movimento de
um veículo, implementado um sistema funcional de aquisição de dados e realizada uma
comparação coerente entre veículos elétricos e veículos a combustão, tanto em ambiente
numérico como experimental. Os resultados obtidos indicam que o comportamento
dinâmico não depende apenas do tipo de propulsão, mas também de fatores como a massa total e a sua distribuição, as características dos pneus e as condições de ensaio como a aceleração longitudinal do veículo e trajetória descrita.
This work aims to analyse and compare the dynamic behaviour of an electric vehicle and a vehicle with an internal combustion engine, seeking to understand how the architecture of the propulsion system influences vehicle dynamics by analysing variables such as lateral acceleration, yaw rate, lateral forces and vertical loads. To this end, a computational model was developed in Python, initially based on the bicycle model and later extended to a fourwheel model, integrating Pacejka’s Magic Formula, a Pure Pursuit control algorithm and a three-dimensional representation of the test. In parallel, a data acquisition system was designed, tested and adapted, and subsequently used to carry out experimental tests with the vehicles under study. Additionally, standardised dynamic tests used in the automotive industry were analysed, and the practical test to be performed with the vehicles and implemented in the computational model was selected. Finally, experimental and numerical tests were carried out on the vehicles, and their results were analysed. The initial validation of the computational model, carried out using experimental data from the LART team, showed good agreement with the real results. In the four-wheel model, the 0–75 m acceleration test produced a calculated time of 4,73 s compared with an experimental value of approximately 4,8 s, while in the Skidpad test the deviation from the real value was approximately 2,03%. These results made it possible to consider the model suitable for the proposed comparative analysis. The numerical and experimental tests revealed differences between the two vehicles, although of limited magnitude. Although the electric vehicle showed slightly higher values of lateral acceleration and yaw rate in the experimental test, the overall deviations observed were approximately between 1% and 2,5%, or 3% when considering a strict assumption of equal radius between tests. Therefore, it was not possible to conclude that there was a markedly significant dynamic difference between the vehicles tested under the conditions considered. The comparison between numerical and experimental results also showed that the developed model consistently reproduced the trend of the main variables analysed, despite some differences in the absolute maximum values, associated with the simplifications of the model and the limitations of the data acquisition system. It is concluded that the defined objectives were broadly achieved, with the development and validation of a computational model capable of characterising vehicle motion, the implementation of a functional data acquisition system, and a coherent comparison between electric and combustion vehicles, both in numerical and experimental environments. The results obtained indicate that dynamic behaviour does not depend solely on the type of propulsion, but also on factors such as total mass and its distribution, tyre characteristics, the trajectory followed and test conditions, such as the vehicle’s longitudinal acceleration and the trajectory described.
This work aims to analyse and compare the dynamic behaviour of an electric vehicle and a vehicle with an internal combustion engine, seeking to understand how the architecture of the propulsion system influences vehicle dynamics by analysing variables such as lateral acceleration, yaw rate, lateral forces and vertical loads. To this end, a computational model was developed in Python, initially based on the bicycle model and later extended to a fourwheel model, integrating Pacejka’s Magic Formula, a Pure Pursuit control algorithm and a three-dimensional representation of the test. In parallel, a data acquisition system was designed, tested and adapted, and subsequently used to carry out experimental tests with the vehicles under study. Additionally, standardised dynamic tests used in the automotive industry were analysed, and the practical test to be performed with the vehicles and implemented in the computational model was selected. Finally, experimental and numerical tests were carried out on the vehicles, and their results were analysed. The initial validation of the computational model, carried out using experimental data from the LART team, showed good agreement with the real results. In the four-wheel model, the 0–75 m acceleration test produced a calculated time of 4,73 s compared with an experimental value of approximately 4,8 s, while in the Skidpad test the deviation from the real value was approximately 2,03%. These results made it possible to consider the model suitable for the proposed comparative analysis. The numerical and experimental tests revealed differences between the two vehicles, although of limited magnitude. Although the electric vehicle showed slightly higher values of lateral acceleration and yaw rate in the experimental test, the overall deviations observed were approximately between 1% and 2,5%, or 3% when considering a strict assumption of equal radius between tests. Therefore, it was not possible to conclude that there was a markedly significant dynamic difference between the vehicles tested under the conditions considered. The comparison between numerical and experimental results also showed that the developed model consistently reproduced the trend of the main variables analysed, despite some differences in the absolute maximum values, associated with the simplifications of the model and the limitations of the data acquisition system. It is concluded that the defined objectives were broadly achieved, with the development and validation of a computational model capable of characterising vehicle motion, the implementation of a functional data acquisition system, and a coherent comparison between electric and combustion vehicles, both in numerical and experimental environments. The results obtained indicate that dynamic behaviour does not depend solely on the type of propulsion, but also on factors such as total mass and its distribution, tyre characteristics, the trajectory followed and test conditions, such as the vehicle’s longitudinal acceleration and the trajectory described.
Descrição
Palavras-chave
Dinâmica automóvel Veículo elétrico Veículo a combustão Modelo computacional Ensaios experimentais Comparação dinâmica