Beam Warming Second-Order Upwind Method

Similar documents
Lecture 12 APPROXIMATION OF FIRST ORDER DERIVATIVES

Derivation of 3-Point Block Method Formula for Solving First Order Stiff Ordinary Differential Equations

Numerical Analysis Formulae Booklet

ANALYSIS ON THE NATURE OF THE BASIC EQUATIONS IN SYNERGETIC INTER-REPRESENTATION NETWORK

Analysis of Lagrange Interpolation Formula

u(x, t) = u 0 (x ct). This Riemann invariant u is constant along characteristics λ with x = x 0 +ct (u(x, t) = u 0 (x 0 )):

u(x, t) = u 0 (x ct). This Riemann invariant u is constant along characteristics λ with x = x 0 +ct (u(x, t) = u 0 (x 0 )):

5 Short Proofs of Simplified Stirling s Approximation

PGE 310: Formulation and Solution in Geosystems Engineering. Dr. Balhoff. Interpolation

Arithmetic Mean and Geometric Mean

Chapter 5 Properties of a Random Sample

Cubic Nonpolynomial Spline Approach to the Solution of a Second Order Two-Point Boundary Value Problem

Ordinary Least Squares Regression. Simple Regression. Algebra and Assumptions.

Q-analogue of a Linear Transformation Preserving Log-concavity

DIFFERENTIAL GEOMETRIC APPROACH TO HAMILTONIAN MECHANICS

The Mathematical Appendix

BERNSTEIN COLLOCATION METHOD FOR SOLVING NONLINEAR DIFFERENTIAL EQUATIONS. Aysegul Akyuz Dascioglu and Nese Isler

2006 Jamie Trahan, Autar Kaw, Kevin Martin University of South Florida United States of America

L5 Polynomial / Spline Curves

On the convergence of derivatives of Bernstein approximation

Numerical Simulations of the Complex Modied Korteweg-de Vries Equation. Thiab R. Taha. The University of Georgia. Abstract

Journal of Mathematical Analysis and Applications

F. Inequalities. HKAL Pure Mathematics. 進佳數學團隊 Dr. Herbert Lam 林康榮博士. [Solution] Example Basic properties

Fourth Order Four-Stage Diagonally Implicit Runge-Kutta Method for Linear Ordinary Differential Equations ABSTRACT INTRODUCTION

Solving Constrained Flow-Shop Scheduling. Problems with Three Machines

Bounds on the expected entropy and KL-divergence of sampled multinomial distributions. Brandon C. Roy

1 0, x? x x. 1 Root finding. 1.1 Introduction. Solve[x^2-1 0,x] {{x -1},{x 1}} Plot[x^2-1,{x,-2,2}] 3

18.413: Error Correcting Codes Lab March 2, Lecture 8

Comparing Different Estimators of three Parameters for Transmuted Weibull Distribution

MATH 247/Winter Notes on the adjoint and on normal operators.

Estimation of Stress- Strength Reliability model using finite mixture of exponential distributions

Lecture 3 Probability review (cont d)

ECE606: Solid State Devices Lecture 13 Solutions of the Continuity Eqs. Analytical & Numerical

FREQUENCY ANALYSIS OF A DOUBLE-WALLED NANOTUBES SYSTEM

C-1: Aerodynamics of Airfoils 1 C-2: Aerodynamics of Airfoils 2 C-3: Panel Methods C-4: Thin Airfoil Theory

X ε ) = 0, or equivalently, lim

02/15/04 INTERESTING FINITE AND INFINITE PRODUCTS FROM SIMPLE ALGEBRAIC IDENTITIES

A FINITE DIFFERENCE SCHEME FOR A FLUID DYNAMIC TRAFFIC FLOW MODEL APPENDED WITH TWO-POINT BOUNDARY CONDITION

0/1 INTEGER PROGRAMMING AND SEMIDEFINTE PROGRAMMING

AN EULER-MC LAURIN FORMULA FOR INFINITE DIMENSIONAL SPACES

VARIABLE-RATE VQ (AKA VQ WITH ENTROPY CODING)

Stability For a stable numerical scheme, the errors in the initial condition will not grow unboundedly with time.

End of Finite Volume Methods Cartesian grids. Solution of the Navier-Stokes Equations. REVIEW Lecture 17: Higher order (interpolation) schemes

TESTS BASED ON MAXIMUM LIKELIHOOD

EVALUATION OF FUNCTIONAL INTEGRALS BY MEANS OF A SERIES AND THE METHOD OF BOREL TRANSFORM

A Remark on the Uniform Convergence of Some Sequences of Functions

MOLECULAR VIBRATIONS

Generalized One-Step Third Derivative Implicit Hybrid Block Method for the Direct Solution of Second Order Ordinary Differential Equation

PROJECTION PROBLEM FOR REGULAR POLYGONS

Part 4b Asymptotic Results for MRR2 using PRESS. Recall that the PRESS statistic is a special type of cross validation procedure (see Allen (1971))

UNIT 2 SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

On Submanifolds of an Almost r-paracontact Riemannian Manifold Endowed with a Quarter Symmetric Metric Connection

DKA method for single variable holomorphic functions

STK4011 and STK9011 Autumn 2016

Chapter 4 (Part 1): Non-Parametric Classification (Sections ) Pattern Classification 4.3) Announcements

1. A real number x is represented approximately by , and we are told that the relative error is 0.1 %. What is x? Note: There are two answers.

Non-uniform Turán-type problems

Investigation of Partially Conditional RP Model with Response Error. Ed Stanek

Analysis of Variance with Weibull Data

Chapter 9 Jordan Block Matrices

Functions of Random Variables

CHAPTER 4 RADICAL EXPRESSIONS

ESS Line Fitting

v 1 -periodic 2-exponents of SU(2 e ) and SU(2 e + 1)

Mu Sequences/Series Solutions National Convention 2014

Point Estimation: definition of estimators

MEASURES OF DISPERSION

D. VQ WITH 1ST-ORDER LOSSLESS CODING

ENGI 4421 Propagation of Error Page 8-01

Discrete Mathematics and Probability Theory Fall 2016 Seshia and Walrand DIS 10b

Summary of the lecture in Biostatistics

1 Lyapunov Stability Theory

EECE 301 Signals & Systems

( ) 2 2. Multi-Layer Refraction Problem Rafael Espericueta, Bakersfield College, November, 2006

The internal structure of natural numbers, one method for the definition of large prime numbers, and a factorization test

Lecture 4 Sep 9, 2015

Multiple Choice Test. Chapter Adequacy of Models for Regression

Construction and Analysis of Multi-Rate Partitioned Runge-Kutta Methods

ρ < 1 be five real numbers. The

A Conventional Approach for the Solution of the Fifth Order Boundary Value Problems Using Sixth Degree Spline Functions

Research Article Gauss-Lobatto Formulae and Extremal Problems

The numerical simulation of compressible flow in a Shubin nozzle using schemes of Bean-Warming and flux vector splitting

NP!= P. By Liu Ran. Table of Contents. The P versus NP problem is a major unsolved problem in computer

LINEAR RECURRENT SEQUENCES AND POWERS OF A SQUARE MATRIX

A Primer on Summation Notation George H Olson, Ph. D. Doctoral Program in Educational Leadership Appalachian State University Spring 2010

ECE 595, Section 10 Numerical Simulations Lecture 19: FEM for Electronic Transport. Prof. Peter Bermel February 22, 2013

2.3. Quantitative Properties of Finite Difference Schemes. Reading: Tannehill et al. Sections and

d dt d d dt dt Also recall that by Taylor series, / 2 (enables use of sin instead of cos-see p.27 of A&F) dsin

Evolution Operators and Boundary Conditions for Propagation and Reflection Methods

Initial-Value Problems for ODEs. numerical errors (round-off and truncation errors) Consider a perturbed system: dz dt

Assignment 5/MATH 247/Winter Due: Friday, February 19 in class (!) (answers will be posted right after class)

Overview of the weighting constants and the points where we evaluate the function for The Gaussian quadrature Project two

Taylor s Series and Interpolation. Interpolation & Curve-fitting. CIS Interpolation. Basic Scenario. Taylor Series interpolates at a specific

Fractional Order Finite Difference Scheme For Soil Moisture Diffusion Equation And Its Applications

Complete Convergence and Some Maximal Inequalities for Weighted Sums of Random Variables

best estimate (mean) for X uncertainty or error in the measurement (systematic, random or statistical) best

The solution of Euler-Bernoulli beams using variational derivative method

A tighter lower bound on the circuit size of the hardest Boolean functions

Strong Convergence of Weighted Averaged Approximants of Asymptotically Nonexpansive Mappings in Banach Spaces without Uniform Convexity

A New Method for Solving Fuzzy Linear. Programming by Solving Linear Programming

Transcription:

Beam Warmg Secod-Order Upwd Method Petr Valeta Jauary 6, 015 Ths documet s a part of the assessmet work for the subject 1DRP Dfferetal Equatos o Computer lectured o FNSPE CTU Prague. Abstract Ths documet cotas the dervato of the Beam Warmg secod-order upwd method ad subsequetly the applcato of ths method s demostrated. It s used for the dscretzato of the lear advecto ad Burgers equatos ad the the order of ths method for both equatos s examed. The stablty codto ad modfed equato were examed oly for the advecto equato accordace wth the requremets. As a part of ths work the scheme was also mplemeted the software package MATLAB R. All source fles ca be foud at http://kfe.fjf.cvut.cz/~valepe7/fles/1drp. Cotets 1 Itroducto 1 Dervato 3 Advecto equato 3 3.1 Dfferece scheme....................................... 3 3. Order of accuracy....................................... 3 3.3 Stablty............................................ 4 3.4 Modfed equato...................................... 4 4 Burgers equato 4 4.1 Dfferece scheme....................................... 5 5 Cocluso 5 1 Itroducto Ths documet cocers a specfc secod-order accurate upwd method proposed by Warmg ad Beam 1976). The Beam Warmg secod-order upwd method exemplfes some of the geeral vrtues ad lmtatos of hgher-order upwd methods [1]. I the frst secto oe ca fd the dervato of ths method, cosequetly ths method s appled to dscretze lear advecto equato ad Burgers equato. The order of accuracy s examed for both equatos as well as the mplemetato of schemes software package MATLAB R. The stablty codto s examed ad the modfed equato s prepared oly for the oe-way wave equato. 1

Dervato I our case, let us have ths goverg equato t + fu) = 0. 1) To beg wth dervato of the Beam Warmg secod-order upwd method, cosder the followg Taylor seres for ux, t + ) ux, t + ) = ux, t) + t + u t + O3 ), ) where the tme dervatves ca be replaced by space dervatves usg the goverg equato 1). Ths has bee doe by so called Cauchy Kowalewsk techque, whch mples t = fu) u ad t = where au) fu). 3) Substtute the precedg expressos for the tme dervatves 1) to the Taylor seres for ux, t + ) ) to obta ux, t + ) = ux, t) fu) x, t) + x, t) + O 3 ). 4) Assume au) > 0. To dscretze the frst space dervatve we use 3-pot secod-order backwarddfferece for costat x, whch s defed as Also let where au) fu) fu) x, t ) = 3fu ) 4fu 1 ) + fu ) + O x ). 5) x, t ) = x 1, t ) + O x) 6) ) x 1, t ) = au 1/ ) fu ) fu 1 )) au 3/ ) fu 1 ) fu )) x +O x ). The resultg method s as follows substtutg / x λ) 7) = u λ 3fu ) 4fu 1) + fu ) ) + λ [ au 1/ ) fu ) fu 1) ) au 3/ ) fu 1) fu ) )], u +1 whch s the Beam Warmg secod-order upwd method for au) > 0. 8) Assume au) < 0. To dscretze the frst space dervatve ow we use 3-pot secod-order forwarddfferece, whch s defed as fu) x, t ) = 3fu ) 4fu +1 ) + fu + ) + O x ). 9) Smlar way oe obtas = u + λ 3fu ) 4fu +1) + fu +) ) + λ [ au +3/ ) fu +) fu +1) ) au +1/ ) fu +1) fu ) )], u +1 whch s the Beam Warmg secod-order upwd method for au) < 0 [1]. 10)

3 Advecto equato The advecto equato s the hyperbolc partal dfferetal equato that govers the moto of a coserved scalar feld as t s advected by a kow velocty vector feld. Let us cosder oly oe space dmeso ad a costat velocty a. The equato takes the followg form t + a = 0. 11) It has clearly be see that the fucto fu), whch occurs the goverg equato 1) satsfes smple equalty fu) = au. The oe ca mmedately get that au) = a. 3.1 Dfferece scheme Now we apply Beam Warmg secod-order upwd method to dscretze the advecto equato 11). It gets two dfferet forms depedg o a sg of a: u +1 u +1 u u + a 3u 4u 1 + u a 3u 4u +1 + u + = a u u 1 + u x for a > 0, 1) = a u u +1 + u + x for a < 0. 13) 3. Order of accuracy I ths secto we wll exame the order of accuracy of the Beam Warmg scheme appled o the advecto equato. Usg Taylor seres expasos drectly o 1) or 13) would have resulted terms of order, x. Oe has to use followg defto []: Defto 1. A scheme P, x u = R, x f that s cosstet wth the dfferetal equato P u = f s accurate of order p tme ad order q space f for ay smooth fucto φt, x) We say that such a scheme s accurate of order p, x q ). Now we llustrate the use of ths defto. From 1) we have ad P, x φ = φ+1 φ P, x φ R, x P φ = O p, x q ). 14) + a 3φ 4φ 1 + φ a φ φ 1 + φ x 15) R, x f = 1 +1 f + f ) a 4 x 3f 4f 1 + f ) 16) We use the Taylor seres o 15) ad 16) evaluated at t, x ) to obta P, x φ = φ t + aφ x + φ tt a φ xx + O, x ) 17) ad for a smooth fucto ft, x) 16) becomes I addto, P φ = φ t + aφ x = f the R, x f = f + f t a f x + O, x ). 18) P, x φ R, x P φ = O, x ). 19) Hece the Beam Warmg scheme s accurate of order, x ). 3

3.3 Stablty I ths secto we demostrate the stablty of Beam Warmg scheme for the oe-way wave equato by usg vo Neuma aalyss, the most popular type of lear stablty aalyss. Thus perform the trasformato u +m +k g m e kξ equato 1) ad obta a solate expresso for amplfcato factor g: g = a λ cosξ) sξ) a λ sξ) + a λ cos ξ) + aλ cosξ) sξ) a λ cosξ) aλ sξ) + aλ cos ξ) aλ cosξ) + aλ + 1. 0) The scheme s stable f the followg codto s fulflled: ξ : g < 1. The orm we compute by summg the squares of the real ad magary part. Precedg codto receves the followg form aλaλ + )aλ + 1) cosξ) 1) 0. 1) The equalty 1) mples that the Beam Warmg scheme s stable f the Courat Fredrchs Lewyor CFL) codto s fulflled, x a. ) The CFL umber s, therefor the CFL lmt s larger tha 1. Ths s the major beeft of usg a wder stecl. The whole process of calculatg the amplfcato factor ca be see attached fle stab_bw.mw. 3.4 Modfed equato A useful techque for studyg the behavor of solutos to dfferece equatos s to model the dfferece equato by a dfferetal equato. The dervato of the modfed equato s closely related to the calculato of the local trucato error for a gve method [3]. I order to fd the modfed equato of the dfferece schemes 1), 13) oe has to replace dscrete varables usg the ffth degree Taylor polyomals evaluated at t, x ). If we express the tme dervatves resulted expressos terms of space dervatves we obta the modfed equato, whch s easer to aalyze u t + au x = a 3aλ + a λ ) x u xxx. 3) 6 The modfed equato s dspersve. For hgher order methods ths elmato of tme dervatves terms of space dervatves must be doe carefully ad s complcated by the eed to clude hgher order terms []. The advatage of the tegrated computer algebra systems was mafested o ths very spot. The etre process ca be observed attached fle modf_bw.mw. 4 Burgers equato Burgers equato s a fudametal partal dfferetal equato. It occurs varous areas of appled mathematcs, such as modelg of gas dyamcs ad traffc flow. We wll cosder a vscd type of ths equato oe space dmeso, whch has the followg form [3] t + u = 0. 4) It has clearly be see that the fucto fu), whch occurs the goverg equato 1) satsfes smple equalty fu) = u /. The oe ca mmedately get that au) = u. 4

4.1 Dfferece scheme Now we apply Beam Warmg secod-order upwd method to dscretze the Burgers equato. It gets ad u + 3u ) 4u 1 ) + u ) = 4 x [ 4 x u 1/ u ) u 1) ) u 3/ u 1 ) u ) )] for u > 0 u +1 u 3u ) 4u +1 ) + u + ) = 4 x [ 4 x u +3/ u + ) u +1) ) u +1/ u +1 ) u ) )] for u < 0. u +1 5) 6) 5 Cocluso The mplemetato of the Beam Warmg secod-order upwd method appled o both equatos ca be foud attached fles advecto_equato.m ad burgers_equato.m. Ulke frst-order upwd methods, hgher-order upwd methods may be extremely oscllatory. I fact, hgher-order upwd methods may be eve more oscllatory tha cetered methods, as s the case comparg the Beam-Warmg secod-order upwd method wth the Lax Wedroff method. I ther orgal paper, Warmg ad Beam 1976) recogzed the oscllatory ature of ther secod-order upwd method. They proposed a bledg that used a frst-order upwd method at shocks ad soc pots ad ther secod-order upwd method smooth regos [1]. Refereces [1] Culbert B. Laey, Computatoal Gasdyamcs. Cambrdge Uversty Press, New York, 1998. [] Joh C. Strkwerda, Fte Dfferece Schemes ad Partal Dfferetal Equatos. Socety for Idustral ad Appled Mathematcs, Phladelpha, d edto, 004. [3] Radall J. LeVeque, Numercal Methods for Coservato Laws. Sprger Basel AG, d edto, 199. 5