Otimização numérica do contorno de tubeira de motor-foguete para empuxo máximo
Date
2021Author
Vicentin, Izabel Cecilia Ferreira de Souza, 1985-
Metadata
Show full item recordSubject
Motores de foguetesVeiculos espaciais - Sistema de propulsão
Engenharia Mecânica
Dinamica dos fluidos
xmlui.dri2xhtml.METS-1.0.item-type
Tese DigitalAbstract
Resumo: Bocais convergente-divergente, também chamados bocais de Laval, são dispositivos utilizados em motores-foguete e motores a jato. A função de tais dispositivos é acelerar o escoamento de velocidades subsônicas para supersônicas. Em motores-foguete, o propelente é armazenado na câmara de combustão, e constitui de 80% a 90% da massa total do equipamento. Portanto, a tarefa de maximizar a eficiência ou o impulso específico de tal dispositivo é de grande importância. Uma das maneiras de fazê-lo é através da otimização da geometria de bocais do tipo convergente-divergente. A literatura apresenta inúmeros trabalhos na área de otimização da região divergente, porém são poucos os trabalhos que estudam o efeito da geometria da região convergente no desempenho de motores-foguete. Alguns trabalhos afirmam, ainda, que o impacto no desempenho é pequeno e que apenas um formato suave nessa região é suficiente. Trabalhos recentes mostram que essa região impacta significativamente os parâmetros de propulsão. Com base nisso, o presente trabalho propõe a maximização do empuxo através do Método de Otimização de Forma, empregando 1, 2, 4 e 8 variáveis na região do convergente de uma tubeira operando no vácuo. As variáveis constituem-se em ordenadas de pontos sobre o perfil a ser otimizado, interpolados por spline cúbica. As variáveis citadas são obtidas iterativamente pelo algoritmo de otimização denominado SLSQP (Sequential Least SQuare Programming). O escoamento foi modelado matematicamente pela equação de Euler com o código computacional Mach2D e resolvido numericamente pelo Método dos Volumes Finitos. Com o objetivo de aumentar a acurácia das soluções, a técnica de Multiextrapolação de Richardson (MER) foi utilizada no processo de otimização. Para todos os casos o perfil com 8 variáveis apresentou o maior valor para o empuxo. Resultados mostram que é possível aumentar o coeficiente de empuxo em 1,6% em relação à tubeira BMG45-15 da National Aerospace and Space Administration (NASA). Com relação às tubeiras Proto-33 e Proto-36 os aumentos no coeficiente de empuxo foram de 0,5 e 0,6% respectivamente. Após obter o perfil de convergente otimizado, o método de Rao foi empregado para obter a geometria otimizada do divergente. O aumento final no empuxo, considerando toda a tubeira otimizada, foi de 3% com relação à tubeira BMG45-15. Abstract: Convergent-divergent nozzles, also called Laval nozzles, are devices used in rocket engines and jet engines. The function of such devices is to accelerate the flow from subsonic to supersonic velocities in order to produce thrust. In rocket engines, the propellant is stored in the combustion chamber and constitutes 80% to 90% of the total mass of the equipment. Therefore, the task of maximizing the efficiency or specific impulse of such a device is of great importance. One of the ways to do this is by optimizing the geometry of convergent-divergent nozzles. The literature presents numerous works concerned with optimization of the divergent region, however, few works study the effect of the geometry of the convergent region on the performance of rocket engines. Many papers claim that only a smooth format in this region is sufficient, since the impact on performance is small. Recent work shows that this region significantly impacts propulsion parameters. Based on this, the present work proposes the maximization of thrust through the Shape Optimization Method, employing 1, 2, 4 and 8 variables in the convergent region of a nozzle operating in vacuum. The variables are ordinates of points on the profile to be optimized, interpolated by cubic spline. The variables are obtained iteratively by the optimization algorithm called SLSQP (Sequential Least SQuare Programming). The flow was mathematically modeled by the Euler equation with the Mach2D computational code and numerically solved by the Finite Volume Method. In order to increase the accuracy of the solutions, the Richardson Multiextrapolation (MER) technique was used in the optimization process. For all cases, the profile with 8 variables presented the highest value for thrust. Results show that it is possible to increase the thrust coefficient by 1.6% compared to the National Aerospace and Space Administration (NASA) BMG45-15 nozzle. Regarding the Proto-33 and Proto-36 nozzles, the increases in the thrust coefficient were 0,5 and 0,6% respectively. After obtaining the optimized convergent profile, Rao’s method was used to obtain the optimized geometry of the divergence. The final increase in thrust, considering the entire optimized nozzle, was 3% over the BMG45-15 nozzle.
Collections
- Teses [79]