Optimization of crenellation patterns for fatigue crack retardation via genetic algorithm and the reduction in computational cost. (May 2016)

• Record Type: Journal Article
• Authors: Lu, Jin; Kashaev, Nikolai; Huber, Norbert
• Is Part Of: Engineering failure analysis. Volume 63(2016:May)
• Page Start: 21
• Page End: 30
• Publication Date: 2016-05
• Subjects: Airworthiness -- Damage tolerance -- Design -- Fatigue crack retardation -- Finite element analysis; System failures (Engineering) -- Periodicals; Fracture mechanics -- Periodicals; Reliability (Engineering) -- Periodicals; Pannes -- Périodiques; Rupture, Mécanique de la -- Périodiques; Fiabilité -- Périodiques; Fracture mechanics; Reliability (Engineering); System failures (Engineering); Periodicals; Electronic journals; 620.112
• DOI: 10.1016/j.engfailanal.2016.02.003
• Languages: English
• ISSNs: 1350-6307

Abstract

Abstract: Crenellation is a promising technique to effectively improve the fatigue life of metallic airframe structures. It systematically varies the thickness of the fuselage skin at constant structural weight. In order to optimize the geometric designs of crenellation patterns for the maximum fatigue crack retardation, we apply an approach coupling genetic algorithm with FEM simulations to examine the vast candidate designs. Additionally, a progressively refined searching strategy and an old-individual-filtering technique were also used to minimize the computational cost. It was found that the crenellation patterns with nonlinear monotonous increase and decrease of skin thickness showed the best fatigue performance, which is expected to extend the fatigue life by 10% compared to the existing design based on experience. The applied strategies for reducing computational cost not only shortened the solving time but also improved the solution quality. Highlights: A geometric optimization coupling genetic algorithm with FEM simulations was applied. The optimized designs favored a nonlinear monotonous thickness change of crenellation. They are expected to extend the fatigue life by 10% compared to the existing design. A strategy for the reduction of the computational cost was developed.