Eventos Anais de eventos
COBEM 2023
27th International Congress of Mechanical Engineering
Stochastic Optimization of a Rotating Machine Through Reliability-Based Design Optimization
Submission Author:
Eduardo Henrique de Paula , SP
Co-Authors:
Eduardo Henrique de Paula, Helio Fiori de Castro
Presenter: Eduardo Henrique de Paula
doi://10.26678/ABCM.COBEM2023.COB2023-1136
Abstract
Rotating machines are components of significant importance for engineering, being fundamental for the operation of the most different mechanisms present in society. Given the importance of rotors for the industry, their operating parameters must be studied and optimized, reducing their manufacturing and operational cost. At the same time, their reliability must be ensured, preventing unexpected failures from interrupting their operation or causing accidents with their operators. Most studies involving the optimization of these components consider their parameters as deterministic. However, just like in any real system, their parameters are stochastic, presenting variability in the material properties, loads, and geometry. To consider this variability in the optimization project, it is necessary to employ stochastic or robust optimization. These methods seek an optimal point for the rotor operating parameters in which, even if their properties vary, the parameter to be optimized does not vary considerably. Therefore, the reliability of the rotor's operation and the robustness of the project are guaranteed. In this paper, a detailed analysis of the optimization of a rotating machine will be conducted through the Reliability-Based Design Optimization (RBDO) method. RBDO is an optimization method that focuses on achieving the robustness of the design by seeking an optimal condition for its operating parameters while ensuring that its failure criteria are not exceeded, guaranteeing a predefined level of reliability. To validate this method, the Jeffcott rotor with a non-central disc and flexible bearings will be studied. A detailed application of RBDO methods in formulating and solving the optimization problem will be presented. The convergence of the optimization results and the computational cost of different approaches will be discussed, aiming to identify the most recommended ones. Once this methodology is validated for a simpler model, further research can be developed to expand its application to more complex rotor models.
Keywords
rotating machines, uncertainty quantification, reliability based design optimization, Jeffcott rotor
