MOGA-FE composite laminate stacking sequence optimization for ultralight pressure vessels

RODRIGUEZ EXEQUIEL; MORAN J

Miami

Congreso; American Advanced Materials Congress 2016; 2016

American Advanced Materials Congress 2016

The weight of each component is a major concern in the aerospace industry. For this reason, the design stage comes to a crucial decision of two competing objectives: minimum weight and maximum stiffness. This is the case of the high-pressure 100% composite ultralight tanks (called linerless or type V). These tanks are used to contain pressurizing gases in launch vehicles. For a particular material, (i.e. 60% carbon fiber reinforced epoxy), the ply stacking sequence needs to be optimized in order to maximize the mechanical properties of the component while minimizing the number of layers. The search space becomes practically infinite if one takes into account that the main variable (orientation) is practically continuous for the particular manufacturing technique used (Filament Winding). The number of layers is usually located in a few tens, so the number of combinations rises exponentially (≈90N), considering a symmetrical and balanced laminate. The design using finite element programs (fixed ply-stack sequences) becomes impractical, as it consumes too much time and computational resources [1]. This is the reason why a combined algorithm based on evolutionary techniques (Multi Objective Genetic Algorithms - MOGA) and finite elements (FE) is proposed to solve the problem of multi-objective optimization [2]. The MOGA algorithm is applied to a simplified model following the Classical Laminate Theory [3]. The various design and manufacturing constraints [4] are incorporated in the evaluation stage of the GA Fitness function as penalty-based constraints. Then a subset of the optimum solutions (Pareto front) is evaluated in more detail by the finite element method (FE), verifying the ability of the designed composite material to withstand workloads within specified ranges and therefore selecting the best option as the definite stacking sequence for the structural application.