Reliability-based design and topology optimization of aerospace components and structures

Abstract

his work presents a research on Reliability Analysis (RA) and Reliability-Based Design Optimization (RBDO) with the target of solving practical aerospace structures problems. Two types of problems are considered, each one solved with a different approach. The first one applies Reliability-Based Topology Optimization (RBTO) to industry-like aerospace structures, while the second performs RBDO on aerospace components that require meticulous and computationally expensive simulations in order to predict accurately their behaviour. The RBDO method used in both approaches is the Sequential Optimization and Reliability Assessment (SORA) due to its uncoupled nature, which allows to separate the Deterministic Optimization (DO) and RA phases being easily combinable with external software. On the other hand, the RA method selected is the Hybrid Mean Value (HMV) given its robustness. Both algorithms have been implemented in MATLAB codes working in a High Performance Computing (HPC) environment. In the first approach they are combined with an external optimization software (Altar Optistruct) able to perform Deterministic Topology Optimization (DTO) problems. In the second approach the MATLAB codes are combined with external Finite Element (FE) analysis software (Abaqus), and the RA phase benefits from a Polynomial Chaos Expansion (PCE) based metamodel in order to save computating time. Several application examples have been carried out, including a wing, a rear fuselage and a pylon in the first approach and a stiffened composite panel in the second approach.