Reduced order modeling strategies for simulation of unsteady flows

• Grant number: 19/18809-7
• Support Opportunities: Scholarships abroad - Research Internship - Master's degree
• Start date: November 01, 2019
• End date: April 30, 2020
• Field of knowledge: Engineering - Mechanical Engineering - Transport Phenomena
• Principal Investigator: William Roberto Wolf
• Grantee: Victor Zucatti da Silva
• Supervisor: Bergmann Michel
• Host Institution: Faculdade de Engenharia Mecânica (FEM). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Institution abroad: Institut de Mathématiques de Bordeaux (IMB), France
• Associated to the scholarship: 18/11410-9 - On the application of principal component analysis for the construction of reduced order models, BP.MS

Abstract

The current proposal describes the research plan associated to the overseas internship (BEPE) of student Victor Zucatti da Silva. The internship period abroad should be held at the Institut de Mathématiques de Bordeaux, located in Bordeaux, France. During this period, Victor will work under the supervision of Dr. Michel Bergmann who is currently a Chargé de Recherche in the Institut de Mathématiques de Bordeaux. Dr. Bergman has extensive expertise in the fields of numerical simulation, flow modal decomposition and reduced order modeling which are topics related to the present research proposal. This project aims to apply reduced order modeling strategies for simulation of unsteady flows. Currently, high-fidelity simulations provide accurate solutions for design of more efficient engineering devices. However, their high computational costs make these simulations impeditive for preliminary design since long simulation periods need to be computed using small time steps to obtain meaningful converged statistics. Reduced order models (ROMs) are appealing since they can be used in preliminary stages of design due to their inherent reduction in computational costs compared to large-scale simulations. ROMs should be also useful in the context of optimization and flow control studies. In this project, ROMs based on Galerkin and least-squares Petrov Galerkin projections of the Navier-Stokes equations will be applied to investigate unsteady flows obtained by high-fidelity numerical simulations. Here, we will study techniques of error prediction for ROMs. Moreover, methods for interpolation of basis functions computed by proper orthogonal decomposition will also be investigated aiming to generalize the application of ROMs for different flow conditions. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

ZUCATTI, VICTOR; WOLF, WILLIAM; BERGMANN, MICHEL. Calibration of projection-based reduced-order models for unsteady compressible flows. Journal of Computational Physics, v. 433, MAY 15 2021. (19/18809-7, 13/08293-7, 18/11410-9)

ZUCATTI, VICTOR; LUI, HUGO F. S.; PITZ, DIOGO B.; WOLF, WILLIAM R.. Assessment of reduced-order modeling strategies for convective heat transfer. NUMERICAL HEAT TRANSFER PART A-APPLICATIONS, v. 77, n. 7, p. 702-729, APR 2 2020. (18/11410-9, 13/07375-0, 13/08293-7, 19/18809-7)

ZUCATTI, VICTOR; WOLF, WILLIAM. Data-driven closure of projection-based reduced order models for unsteady compressible flows. COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING, v. 386, DEC 1 2021. (19/18809-7, 18/11410-9, 13/08293-7)