Banca de DEFESA: PRINCE AZSEMBERGH NOGUEIRA DE CARVALHO

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : PRINCE AZSEMBERGH NOGUEIRA DE CARVALHO
DATE: 13/12/2019
TIME: 08:30
LOCAL: Sala 414 do CTEC - UFRN
TITLE:

MULTIOBJECTIVE OPTIMIZATION OF AERODYNAMIC PROFILES USING GENETIC ALGORITHM

KEY WORDS:

Multiobjective optimization, aerodynamics profiles, genetic algorithm.

PAGES: 73
BIG AREA: Engenharias
AREA: Engenharia Mecânica
SUMMARY:

In aviation, the pursuit of more efficient aircraft has grown in the face of concerns
by governments and organizations about the environment, forcing new aircraft
projects to emit less and less CO2 into the atmosphere. Aircraft manufacturers then,
to remedy the various problems, focus primarily on wing design, empennage,
fuselage and engines, which are the most significant aerodynamic components. All
of them have in common a vital part in the projects, which is the profile. In the
present work we sought to find an optimal profile with a multiobjective approach. In
Matlab, a genetic algorithm will optimize the profiles through an iteration with XFOIL,
where the aerodynamic characteristics of the profiles, represented by their
coefficients (Cl, Cd, Cm and Cl / Cd), are obtained. In order to obtain a good
population diversification and thus finding the ideal profile, combining the most
diverse geometric characteristics, 300 different profiles were available as source for
the initial population generation, which were available in the Airfoil Data Site
database. For XFOIL simulation, the air density was adopted as 1.225 kg / m³, a
viscosity of 1.79e-5 Pa.s, reynolds no. 5.00e5, mach no. 0.05, ncrit 9, no. maximum
convergence iterations of 100 and an angle of attack range from 0º to 18º. For the
genetic algorithm, a 90% crossover probability, 5% for mutation, a search
exploration outside the 30% range and a maximum number of 100 generations were
adopted as stopping criteria for the population. The optimized profiles outperformed
other studies in relation to the aerodynamic coefficients such as efficiency and Cl,
but not so much in relation to Cm. The algorithm proved to be satisfactory in finding
profiles that aimed to improve Cl, Cd, and consequently efficiency, finding higher
profiles by 32% Cl for flying wings, 67% Cl for empennage and better efficiencies
by about 70% for wind turbines. The algorithm was able to achieve better and better
profiles compared to others commonly used in the mentioned applications, besides
obtaining characteristics such as softer stall and wind turbine noise reduction. For
future work, it is suggested to investigate a larger number range of Reynolds, and
other aerodynamic solvers, in addition to implementing the inverse method
optimization, aiming to obtain an optimal pressure distribution.

BANKING MEMBERS:
Presidente - 1338331 - RAIMUNDO CARLOS SILVERIO FREIRE JUNIOR
Interno - 1445637 - WALLACE MOREIRA BESSA
Externo ao Programa - 1647050 - SANDI ITAMAR SCHAFER DE SOUZA
Externo à Instituição - RÔMULO PIERRE BATISTA DOS REIS - UFERSA