ebook img

A Projection Transformation Method for Nearly Singular Surface Boundary Element Integrals PDF

464 Pages·1992·7.191 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview A Projection Transformation Method for Nearly Singular Surface Boundary Element Integrals

Lecture Notes in Engineering Edited by C. A. Brebbia and S. A. Orszag 73 K. Hayami A Projection Transfonnation Method for Nearly Singular Surface Boundary Element Integrals ..------, Springer-Vertag Bertin Heidelberg New York London Paris Tokyo Hong Kong Barcelona Budapest Series Editors .... .1:>0 Q) C. A. Brebbia . S. A. Orszag s t ) n 0 Consulting Editors 6 oi J. Argyris . K.-J. Bathe' A. QS. Cakmak . J. Connor' R. McCrory P C. S. Desai· K.-P. Holz . F.SP A. Leckie' G. Pinder' A. R. S. Ponof t on J. H. Sein+< feld . P. Silvesteral (' P. Spanos' W. Wun. derlich . S. Yiper ati -- r 'br Autho-r . e mg CDr .& K Ce'_' nI nHfoanynaamtiio n Technodrilatlogy NuInte S _ a d Resea"rch Laboratories qu , 0 q* Gauss 4IMN0K2S-01aE--' 1Bi13y" n-CN: 1a aJ\ -1 m1ACgM~/' .11 )'. /_.-....,. ""0. . 3, Paa.o:1i....... wyeA...,. r.,..... 09it \/ '''./ ", • __ . p\ a,0Na7 oz7K8 q ar/-a9.<J"D. 3a'n\ \li '/' k7-wti5i8 oa4-30ns-- a56k54IsCPijq*dS i02 0-81,& \ \../'. • 04-66'~., 9\ \ "-" Identity 8'spherical' S : -~.~ 4\ 'b." (Polar) = e0.01 ~ d -\ " 'n .. IS·~Telles , B\ ", N, , ---.~ -1\ bo. .. 3........ \. ....... : 9" 7' 8-.~ .................. -3'~"'" -6R,og-L 2 . ':--... 42", -8............... 4'"...... \ 69R,Og-L1 , , 8"..I -4, " '." I, , , I 200500 1000 1500 Is Fig. 10.32 Convergence graph for CPij This work is subject to copyright. All rights are reserved, whether the whole o1r part oof the material jtive bioor s~ rff o ctShaoeidnspcc tpaee1 usrmnbtienblicdgea,,r t9srioep, npe110 r9c-1 oo6ifdr5i cup,a cainltrliyt ois tn tsth h 10 oce-2 enu rr reimrgoehifnct sritso vo fpeilfemr 10 ts-3 rr imaoonnirts .ti elnaad ntai odonn nyp10 ,-4 l eyore ntuhpnnerisdirns etwiirona tng10 hy-5 , ,e f ora eprnu rudoss evsei tso omifor 10 auni-6 lglssuet s o atifrnl awt htdaieo-7ay ntGs10 as eb, b nreaen nocakibn, tsat a.Ct iiDoon10 nupe-8 ,pyd rl iicgahtti oLna w aforom Springer-Verlag. Violations are liable for prosecution under the Gennan Copyright Law. el©r Springer-Verlag Berlin Heidelberg 1992 r RE The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Typesetting: Camera ready by author ACKNOWLEDGEMENTS I wish to express my sincere appreciation and gratitude to my advisor Dr. Carlos A. Brebbia, for his valuable guidance and encouragement throughout the period of this study, which started with my stay at the Wessex Institute of Technology, U. K. I am grateful to Prof. W. S. Hall, Prof. M. Tanaka, Prof. J. C. F. Telles, Dr. B. Adey, Dr. S. M. Niku, Dr. S. Amini, Dr. W. Tang, Dr. M. H. Aliabadi, Prof. T. Honma, Dr. T. Takeda, Dr. L. Gray, Dr. N. Nishimura, Dr. Y. Iso, Prof. M. Mori, Dr. M. Sugihara, Dr. K. Kishimoto and Dr. S. Akaboshi, among many others, for stimulating discussions, valuable remarks and encouragements. I would also like to thank Dr. W. Blain for his general assistance. I am indebted to my company, NEC Corporation, in particular, Mr. Y. Kato, Dr. K. Iinuma, Mr. A. Managaki, Mr. K. Nakamura, Dr. N. Harada, Mr. Y. Nagai, Dr. E. Okamoto, Dr. S. Doi, Mr. K. Tsutaki and Mr. Y. Yanai, among many others, for their support and understanding for the present work. I am grateful to Mr. H. Matsumoto ofNEC Scientific Information System Development Co. Ltd. for preparing tables for the Gauss-Legendre formula. I would also like to thank Ms. M. Nakae, Mrs. K. Ishikura and Ms. T. Kitagawa, for their excellent typing. I would like to thank my sister, Mrs. Y. Mino for helping me with the proof reading. Last, but not the least, I would like to thank my wife Emiko, for her constant support and encouragement. TABLE OF CONTENTS PART I THEORY AND ALGORITHMS CHAPTER 1 INTRODUCTION 3 CHAPTER 2 BOUNDARY ELEMENT FORMULATION OF 3-D POTENTIAL PROBLEMS 2.1 Boundary Integral Equation 8 2.2 Treatment of the Exterior Problem 14 2.3 Discretization into Boundary Elements 18 2.4 Row Sum Elimination Method 24 CHAPTER 3 NATURE OF INTEGRALS IN 3-D POTENTIAL PROBLEMS 29 3.1 Weakly Singular Integrals 30 3.2 Hyper Singular Integrals 40 3.3 Nearly Singular Integrals 53 CHAPTER 4 SURVEY OF QUADRATURE METHODS FOR 3-D BOUNDARY ELEMENT METHOD 72 4.1 Closed Form Integrals 73 4.2 Gaussian Quadrature Formula 74 4.3 Quadrature Methods for Singular Integrals 76 (1) The weighted Gauss method 76 (2) Singularity subtraction and Taylor expansion method 77 (3) Variable transformation methods 78 (4) Coordinate transformation methods 80 VI 4.4 Quadrature Methods for Nearly Singular Integrals 81 (1) Element subdivision 82 (2) Variable transformation methods 84 (3) Polar coordinates 88 CHAPTER 5 THE PROJECTION AND ANGULAR & RADIAL TRANSFORMATION (PART) METHOD 5.1 Introduction 89 5.2 Source Projection 93 5.3 Approximate Projection of the Curved Element 97 5.4 Polar Coordinates in the Projected Element 104 5.5 Radial Variable Transformation 113 (i) Weakly Singular Integrals 113 (ii) Nearly Singular Integrals 114 (1) . Singularity cancelling radial variable transformation 114 (2) Consideration of exact inverse projection and curvature of the element in the transformation 121 (3) Adaptive logarithmic transformation (log-L2) 123 (4) Adaptive logarithmic transformation (log-Ll) 125 (5) Single and double exponential transformations 131 5.6 Angular Variable Transformation 139 (1) Adaptive logarithmic angular variable transformation 140 (2) Single and double exponential transformations 143 5.7 hnplementation of the PART method 144 (1) The use of Gauss-Legendre formula 145 (2) The use of truncated trapezium rule 150 5.8 Variable Transformation in the Parametric Space 153 VII CHAPTER 6 ELEMENTARY ERROR ANALYSIS 6.1 The Use of Error Estimate for Gauss-Legendre Quadrature Formula 155 6.2 Case {3=2 160 (Adaptive Logarithmic Transformation: fog-L2) 6.3 Case {3= 1 Transformation 166 6.4 Case {3=3 Transformation 169 6.5 Case {3 = 4 Transformation 172 6.6 Case {3 = 5 Transformation 185 6.7 Summary of Error Estimates for (3=1-5 188 6.8 Error Analysis for Flux Calculations 190 CHAPTER 7 ERROR ANALYSIS USING COMPLEX FUNCTION THEORY 7.1 Basic Theorem 204 7.2 Asymptotic Expression for the Error Characteristic 207 Function <P n(z) 7.3 Use ofthe Elliptic Contour as the Integral Path 208 7.4 The Saddle Point Method 210 7.5 Integration in the Transformed Radial Variable: R 212 7.6 Error Analysis for the Identity Transformation: R(p) = p 214 c (1) Estimation of the size (J of the ellipse q 217 (2) Estimation of max I j(z) I 218 Z E£q (3) Error estimate En(j) 220 7.7 Error Analysis for the fog-L Transformation 221 2 (1) Case: 0 = odd 223 (2) Case : 0 = even 227 (i) Contribution from the branch line f +, f- 229 c (ii) Contribution from the ellipse q 234 (iii) Contribution from the small circle CE 238 (iv) Summary 239 VIII 7.8 Error Analysis for the eog-Ll Transformation 241 (1) Error analysis using the saddle point method 244 (2) Error analysis using the elliptic contour: Cq 249 (i) Estimation of max I f(z) I 250 z ECq (ii) Estimation of a 252 (iii) Error estimate En(j) 253 7.9 Summary of Theoretical Error Estimates 257 PART II APPLICATIONS AND NUMERICAL RESULTS CHAPTER 8 NUMERICAL EXPERIMENT PROCEDURES AND ELEMENT TYPES 8.1 Notes on Procedures for Numerical Experiments 263 8.2 Geometry of Boundary Elements used for Numerical Experiments 265 (1) Planar rectangle (PLR) 265 (2) 'Spherical' quadrilateral (SPQ) 267 (3) Hyperbolic quadrilateral (HYQ) 270 CHAPTER 9 APPLICATIONS TO WEAKLY SINGULAR INTEGRALS 9.1 Check with Analytial Integration Formula for Constant Planar Elements 273 9.2 Planar Rectangular Element with Interpolation Function rp jj 282 9.3 'Spherical' Quadrilateral Element with Interpolation Function rpij 294 (1) Results for Isrpjju*dS 294 (2) Results for Is rpijq* dS 303 9.4 Hyperbolic Quadrilateral Element with Interpolation Function rpij 313 (1) Results for Is rpij u* dS 313 (2) Results for Is rp jj q* dS 321 IX 9.5 Summary of Numerical Results for Weakly Singular Integrals 329 CHAPTER 10 APPLICATIONS TO NEARLY SINGULAR INTEGRALS 10.1 Analytical Integration Formula for Constant Planar Elements 331 10.2 Singularity Cancelling Radial Variable Transformation for Constant Planar Elements 335 10.3 Application of the Singularity Cancelling Transformation to Elements with Curvature and Interpolation Functions 353 (1) Application to curved elements 353 (2) Application to integrals including interpolation functions 368 10.4 The Derivation of the eog-L2 Radial Variable Transformation 377 (1) Application of radial variable transformations p dp =r 'P dR «(3* a) to integrals Js lira dS over curved elements 377 (2) Difficulty with flux calculation 392 10.5 The eog-Ll Radial Variable Transformation 395 10.6 Comparison of Radial Variable transformations for the Model Radial Integral la.o 397 (1) Transformation based on the Gauss-Legendre rule 408 (i) Identity transformation 408 (ii) eog-L2 transformation 409 (iii) eog-Ll transformation 410 (2) Transformation based on the truncated trapezium rule 411 (i) Single Exponential (SE) transformation 411 (ii) Double Exponential (DE) transformation 412 (3) Summary 412 x 10.7 Comparison of Different Numerical Integration methods on the 'spherical' Element 413 (1) Effect of the source distance d 414 (2) Effect of the position of the source projection Xs 430 10.8 Summary of Numerical Results for Nearly Singular Integrals 438 CHAPTER 11 APPLICATIONS TO HYPERSINGULAR INTEGRALS 441 CHAPTER 12 CONCLUSIONS 445 REFERENCES 451 PART I THEORY AND ALGORITHMS

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.