Skip to main content
Top

2022 | Book

Physical Modeling and Computational Techniques for Thermal and Fluid-dynamics

Practical Numerical Mathematics

insite
SEARCH

About this book

This book on computational techniques for thermal and fluid-dynamic problems arose from seminars given by the author at the Institute of Nuclear Energy Technology of Tsinghua University in Beijing, China. The book is composed of eight chapters-- some of which are characterized by a scholastic approach, others are devoted to numerical solution of ordinary differential equations of first order, and of partial differential equations of first and second order, respectively. In Chapter IV, basic concepts of consistency, stability and convergence of discretization algorithms are covered in some detail. Other parts of the book follow a less conventional approach, mainly informed by the author’s experience in teaching and development of computer programs. Among these is Chapter III, where the residual method of Orthogonal Collocations is presented in several variants, ranging from the classical Galerkin method to Point and Domain Collocations, applied to numerical solution of partial differential equations of first order. In most cases solutions of fluid dynamic problems are led through the discretization process, to the numerical solutions of large linear systems. Intended to impart a basic understanding of numerical techniques that would enable readers to deal with problems of Computational Fluid Dynamics at research level, the book is ideal as a reference for graduate students, researchers, and practitioners.

Table of Contents

Frontmatter
Chapter 1. Ordinary Differential EquationsOrdinary differential equations
Abstract
This chapter is dedicated to ordinary differential equations of first and higher orders. Emphasis is put on equations which can be written in normal form. Among those a scheme is outlined to choose the optimum strategy for their integration. For equations which can not be directly integrated, the most effective numerical schemes are reviewed. Ordinary differential equations of higher order are reduced to systems of differential equations of first order. Special attention is directed to stiff systems which frequently occur with time scales of different orders of magnitude. Examples are frequent in chemical kinetics and in transport problems with phase changes due to the different orders of magnitudes of densities in the phases. Stiff systems require appropriate numerical treatments for correct integration.
Maurizio Bottoni
Chapter 2. Partial Differential Equations and the Method of Characteristics
Abstract
This chapter is dedicated to partial differential equations of first and second orders, treated with the method of characteristics. Examples are given for the numerical treatment of linear damped waves and quasi-linear undamped waves. Preliminaries are set for the successive analysis of methods of orthogonal collocations developed in Chap. 3. Partial differential equations of second order are discussed with reference to the problem of existence, uniqueness, und continuity of the solutions in the context of the Hadamard conditions of well-posedness. The equations of characteristics are introduced also for the partial differential equations of second order and are elucidated with numerical examples. Numerical solutions of the Laplace equation with three numerical techniques are reviewed, and relative advantages or drawbacks are discussed.
Maurizio Bottoni
Chapter 3. Methods of Orthogonal Collocations (OC)
Abstract
This chapter links to the previous chapter because the theoretical treatment is based on the quasi-linear damped equations introduced in Chap. 2. The implementation of a model equation is explained for the sub-cases of subdomain collocations methods, for the point collocations, and for the classical Galerkin method. Numerical samples are discussed in the frame of a computer program (FRONTI) which solves the model equation with the method explained above and additionally with explicit (Jacobi and Gauss Seidel) and implicit algorithms. The chapter closes with a numerical application of orthogonal collocation methods applied to a one-dimensional equation of sodium vapor flow in the frame of safety analysis of fast breeder reactors. A synopsis of the code used for this purpose (BLOW 3A) is given.
Maurizio Bottoni
Chapter 4. Numerical Methods for the Solution of the Convection-Diffusion EquationConvection-diffusion equation and QUICK Algorithm
Abstract
This chapter describes random walks of Brownian motion leading to convection-diffusion equations which can be used to test random number generators for real randomness. Model diffusion equations, convective equations and diffusion-convection equations, are used to analyze the stability of numerical schemes with a variety of methods. These include matrix methods, leapfrog (Richardson) schemes, Dufort-Frankel schemes, implicit schemes, and upwind and central differences. Stability criteria after von Neumann are used throughout. The problem of numerical diffusion is introduced, and the solution based on the QUICK scheme is explained. The implementation of the QUICK scheme into a two-phase flow homogeneous equilibrium model is worked out in all details. The FRAM correction to the QUICK scheme is mentioned for its capability to improve the numerical results.
Maurizio Bottoni
Chapter 5. Numerical Solution of Large Linear Systems
Abstract
This chapter is dedicated to a review of numerical methods to solve iteratively linear systems which, even with the most powerful computational platform today available, cannot be solved with direct methods. The chapter starts with the review of consistently ordered matrices and the related concept of the so-called “Property A”, which is a necessary condition for large linear systems to guarantee convergence of the related numerical schemes. The Young-Frankel theory of successive over-relaxations (SOR) is then revised in detail including the sub-cases of Gauss-Seidel and Jacobi methods. Several numerical examples are given for simple matrices with and without the Property A. A review of relaxation methods based on minimization of functional is given, and gradient methods are introduced as a subclass of variational methods.
Maurizio Bottoni
Chapter 6. Numerical Solution of Poisson Equation
Abstract
This chapter is dedicated to the numerical solution of a model Poisson equation, defined in a rectangular domain, with a known analytical solution. Five numerical methods are used for computing approximate solutions and comparing them with the known analytical solution. The first two methods, which can be classified as “finite difference” methods, are based on the 5-point formula (with a 5-point stencil) and on the 9-point formula (9-point stencil). The other three methods can be classified as finite element methods: one is based on the classical Ritz method, while the other two belong to the class of residual methods of orthogonal collocations (ROMC) in the Galerkin variant and in the variant of subdomain collocations, respectively. A brief comparison of computer results is given.
Maurizio Bottoni
Chapter 7. Derivation and Numerical Solutions of Poisson-Like Equations
Abstract
This chapter explains the numerical solution of Poisson-like equations, formally identical with a Poisson equation but with space- and time-dependent coefficients. In the frame of safety analysis of fast breeder reactors, two examples of constructing Poisson-like equations are worked out in all details for the pressure distribution in a fluid flow and for the enthalpy distribution in a heated fluid. Numerical solutions of Poisson-like equations are explained with the classical variant of the ADI (Alternating Directions Implicit) and with advanced variants for two- and three-dimensional problems. A novel algorithm applicable to two-phase flow calculations is explained. Synopsis of two computer codes used over the years for nuclear safety analysis in Germany (BACCUS-3D/TP) and in the USA (COMMIX-2) is given.
Maurizio Bottoni
Chapter 8. Numerical Treatment of the Transport Equations of Turbulence
Abstract
This chapter introduces the concepts of turbulence from a didactic viewpoint with particular reference to the analytical expressions of the energy spectrum in different subsections. An overview of the turbulence model is given, from the K-ε model to the Reynolds stress model (RSM) and to large eddy simulation (LES). Historical developments made at ANL are presented. All details necessary to implement turbulence models in a computer program are explained for a class of transport equations including scalar fluxes, variance of temperature fluctuations, the set of equations of the Reynolds stress model (RSM) and the dissipations of turbulent kinetic energy. The numerical treatments of these equations are carried out in all details. The particular case of homogeneous turbulence is explained and compared with experimental results.
Maurizio Bottoni
Backmatter
Metadata
Title
Physical Modeling and Computational Techniques for Thermal and Fluid-dynamics
Author
Maurizio Bottoni
Copyright Year
2022
Electronic ISBN
978-3-030-79717-1
Print ISBN
978-3-030-79716-4
DOI
https://doi.org/10.1007/978-3-030-79717-1

Premium Partners