Eventos Anais de eventos
COBEM 2021
26th International Congress of Mechanical Engineering
REDUCED ORDER MODELING BASED ON COMPUTATIONAL FLUID DYNAMICS: A STUDY OF CASE FOR A STIRRED TANK
Submission Author:
ALEXANDRE COSTA , PR , Brazil
Co-Authors:
ALEXANDRE COSTA, Marcelo José Alba
Presenter: Marcelo José Alba
doi://10.26678/ABCM.COBEM2021.COB2021-0301
Abstract
Progresses in hardware computational power and numerical algorithms had expanded the class of industrial flows that can be approached by Computational Fluid Dynamics methods. In recent years, Reduced Order Modeling (ROM) had been applied extensively in engineering problems tackled by Computational Fluid Dynamics. Reduced Order Methods derives on splitting a higher order problem in two steps: a computational expensive (CFD runs, ROM generation and evaluation) and less expensive ROM consumption. The ROM consumption step can also be perfomed in a simpler hardware platform or integrated in a full multiscale model of an industrial plant (digital twin) In this work is applied the ROM workflow for a laboratory scale mixer CFD model. The practical implementation of the ROM to estimate the relationship between rotation velocity, fluid properties and mixing time is explained in detail. Computational fluid dynamics (CFD) verification runs are used to assess the reliability of the reduced-order model. Findings indicate that the ROM is capable of predicting the mixer performance for a wide range of input conditions
Keywords
reduced order model, CFD, Digital Twin
