Eventos Anais de eventos
DINAME2019
DINAME2019
Kriging/FORM-Based Reliability Analysis of Rotor-Bearing Systems
Submission Author:
Mateus Barbosa , SP
Co-Authors:
Domingos Rade, William Alves
Presenter: Mateus Barbosa
doi://10.26678/ABCM.DINAME2019.DIN2019-0202
Abstract
In this work, it is proposed a Kriging/FORM-based reliability assessment of rotor-bearing systems. The motivation is to signi cantly reduce computation effort in limit-state function evaluation by using surrogate models for rotordynamics. Usually, the analysis of rotor-bearing systems dynamic features relies on the Finite Element (FE) method, which may lead to complex and time consuming simulation procedures. Approximations of the input/output relationships created by the FE simulations can be obtained using the Kriging interpolating method. The derived models work as fast-running surrogates for the underlying computer model. Comparison of the results from Kriging surrogates generated using different training samples shows that these low-dimension mathematical relationships can accurately predict rotor-bearing system outputs with considerably low training effort. Subsequently, these surrogates are used to approximate and replace implicit limit-state functions for a Kriging-based reliability assessment by means of the fi rst-order reliability method (FORM). The proposed methodology shows impressive gains in reducing the computation cost with minimal loss in terms of reliability prediction when compared with the classical Monte Carlo simulation approach.
Keywords
Rotor-bearing system, Surrogate model, Kriging, Structural Reliability, FORM
