Development of high fidelity reduced order hybrid stick model for aircraft dynamic aeroelasticity analysis

摘要

This paper presents a new high fidelity reduced order modeling methodology based on a hybrid stick model representation approach. Here, the traditional stick model developed by the unitary loading method is augmented by residual mass and stiffness matrices that account for the dynamic imparity between the stick model and the global finite element model, within a frequency range of interest, as well as the degrees of freedom coupling commonly ignored by the simplified stick model. The new method offers the handling flexibilities of the conventional stick model as well as the high dynamic accuracy of matrix based model order reduction methods such as the Guyan and the Craig-Bampton condensation techniques. Retaining the stick model in the proposed hybrid model representation intuitively enables aerospace development engineers to, accurately and efficiently, optimize the airframe mass and stiffness distribution for aircraft loads minimization and performance maximization without the need to engage an expensive global finite element model in such highly iterative analyses. Two hybrid stick models are presented in this paper that are developed based on the Guyan and the Craig-Bampton reduction methods. A case study is presented where the hybrid stick models developed along with their conventional stick model counterpart are employed in the dynamic aeroelasticity loads analyses of a Bombardier aircraft platform. Using monitor points method, the extracted aeroelastic loads using the reduced order models are compared against those generated employing the aircraft global finite element model. The dynamic characteristics of the reduced order models are also assessed based on their modal characteristics using modal assurance criteria along with their loads modal participation factors. Results obtained show that the developed hybrid stick models have superior dynamic characteristics compared to the conventional stick model. (C) 2019 Elsevier Masson SAS. All rights reserved.