Visit https://ebookultra.

com to download the full version and

explore more ebooks

Structural Analysis with the Finite Element Method

Linear Statics Volume 1 Basis and Solids Lecture
Notes on Numerical Methods in Engineering and
Sciences v 1 1st Edition Eugenio Onate
_____ Click the link below to download _____
https://ebookultra.com/download/structural-analysis-
with-the-finite-element-method-linear-statics-
volume-1-basis-and-solids-lecture-notes-on-numerical-
methods-in-engineering-and-sciences-v-1-1st-edition-
eugenio-onate/

Explore and download more ebooks at ebookultra.com

Structural Analysis with the Finite Element Method
Linear Statics Volume 1 Basis and Solids Lecture Notes
on Numerical Methods in Engineering and Sciences v 1
1st Edition Eugenio Onate Digital Instant Download
Author(s): Eugenio Onate
ISBN(s): 9781402087325, 1402087322
Edition: 1
File Details: PDF, 25.23 MB
Year: 2009
Language: english
Structural Analysis with
the Finite Element Method
Linear Statics
Volume 1. Basis and Solids
 Lecture Notes on Numerical Methods in
Engineering and Sciences

Aims and Scope of the Series

This series publishes text books on topics of general interest in the field of computational
engineering sciences.
The books will focus on subjects in which numerical methods play a fundamental role for
solving problems in engineering and applied sciences. Advances in finite element, finite
volume, finite differences, discrete and particle methods and their applications to classical
single discipline fields and new multidisciplinary domains are examples of the topics covered
by the series.
The main intended audience is the first year graduate student. Some books define the
current state of a field to a highly specialised readership; others are accessible to final year
undergraduates, but essentially the emphasis is on accessibility and clarity.
The books will be also useful for practising engineers and scientists interested in state of the
art information on the theory and application of numerical methods.

Series Editor
Eugenio Oñate
International Center for Numerical Methods in Engineering (CIMNE)
School of Civil Engineering
Universitat Politècnica de Catalunya (UPC), Barcelona, Spain

Editorial Board
Francisco Chinesta, Ecole Nationale Supérieure d'Arts et Métiers, Paris, France
Charbel Farhat, Stanford University, Stanford, USA
Carlos Felippa, University of Colorado at Boulder, Colorado, USA
Antonio Huerta, Universitat Politècnica de Catalunya (UPC), Barcelona, Spain
Thomas J.R. Hughes, The University of Texas at Austin, Austin, USA
Sergio R. Idelsohn, CIMNE-ICREA, Barcelona, Spain
Pierre Ladeveze, ENS de Cachan-LMT-Cachan, France
Wing Kam Liu, Northwestern University, Evanston, USA
Xavier Oliver, Universitat Politècnica de Catalunya (UPC), Barcelona, Spain
Manolis Papadrakakis, National Technical University of Athens, Greece
Jacques Périaux, CIMNE-UPC Barcelona, Spain & Univ. of Jyväskylä, Finland
Bernhard Schrefler, Università degli Studi di Padova, Padova, Italy
Genki Yagawa, Tokyo University, Tokyo, Japan
Mingwu Yuan, Peking University, China

Titles:

1. E. Oñate, Structural Analysis with the Finite Element Method.

Linear Statics. Volume 1. Basis and Solids, 2009
Structural Analysis with
the Finite Element Method
Linear Statics
Volume 1. Basis and Solids

Eugenio Oñate
International Center for Numerical Methods in Engineering (CIMNE)
School of Civil Engineering
Universitat Politècnica de Catalunya (UPC)
Barcelona, Spain
ISBN: 978-1-4020-8732-5 (HB)
ISBN:978-1-4020-8733-2 (e-book)

Depósito legal: B-11715-09

A C.I.P. Catalogue record for this book is available from the Library of Congress

Typesetting: Mª Jesús Samper, CIMNE, Barcelona, Spain

Lecture Notes Series Manager: Adriana Hanganu, CIMNE, Barcelona, Spain

Cover page: Pallí Disseny i Comunicació, www.pallidisseny.com

Printed by: Artes Gráficas Torres S.L.

Morales 17, 08029 Barcelona, España
www.agraficastorres.es

Printed on elemental chlorine-free paper

Structural Analysis with the Finite Element Method. Linear Statics.

Volume 1. Basis and Solids
Eugenio Oñate

First edition, March 2009

¤ International Center for Numerical Methods in Engineering (CIMNE)

Gran Capitán s/n, 08034 Barcelona, Spain
www.cimne.upc.es

No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form
or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise,
without written permission from the Publisher, with the exception of any material supplied
specifically for the purpose of being entered and executed on a computer system, for exclusive
use by the purchaser of the work.

Tomyfamily
Preface

This two-volume book presents an overview of the possibilities of the Finite

Element Method (FEM) for linear static analysis of structures. The text
is a revised extension of the Spanish version of the book published by
the International Center for Numerical Methods in Engineering (CIMNE)
in 1992 and 1995 (2nd edition). The content of the book is based on the
lectures of the course on Finite Element Structural Analysis taught by the
author since 1979 to final year students in the School of Civil Engineering
at the Technical University of Catalonia (UPC) in Barcelona, Spain.
Volume 1 of the book presents the basis of the FEM and its application to
structures that can be modelled as two-dimensional (2D), axisymmetric
and three-dimensional (3D) solids using the assumptions of general linear
elasticity theory.
Volume 2 covers the FEM analysis of beam, plate, folded plate, axisym-
metric shell and arbitrary shape shell structures. Emphasis is put in the
treatment of structures with composite materials.
Each chapter of the book presents the main theoretical concepts on the
particular structural model considered, such as the kinematic description,
the constitutive relationship between stresses and strains and the equili-
brium equations expressed by the Principle of Virtual Work. This is follo-
wed by a detailed derivation of the FEM equations and some applications
to academic and practical examples of structural analysis. Complementary
topics such as error estimation, adaptive mesh refinement, mesh generation
and visualization of FEM results and computer programming of the FEM
are also covered in the last chapters of Volume 1.
The book is particularly addressed to those interested in the analysis and
design of solids and structures, understood here in a broad sense. The FEM
concepts explained in the book are therefore applicable to the analysis
of structures in civil engineering constructions, buildings and historical
constructions, mechanical components and structural parts in automotive,
naval and aerospace engineering, among many other applications.
The background knowledge required for study of the book is the stan-
dard one on mathematics, numerical analysis, elasticity and strength of
materials, matrix structural analysis and computer programming covered
in the first courses of engineering and architecture schools at technical
universities. In any case, the key theoretical concepts of each chapter are
explained in some detail so as to facilitate its study.
Chapter 1 of Volume 1 presents first the concepts of structural and compu-
tational models. Then the basic steps of matrix analysis of bar structures
are summarized. This chapter is important as the FEM follows very closely
the methodology of matrix structural analysis. Understanding clearly the
concept of splitting a structure in different elements, the equilibrium of
the individual elements and the assembly of the global equilibrium equa-
tions of the structure from the contributions of the different elements is
essential in order to follow the rest of the book.
Chapters 2 and 3 introduce the FEM formulation for the analysis of sim-
ple axially loaded bars using one-dimensional (1D) bar elements. The
key ingredients of the FEM, such as discretization, interpolation, shape
functions, numerical integration of the stiffness matrix and the equivalent
nodal force vector for the element are explained in detail, as well as other
general concepts such as the patch test, the conditions for convergence of
the FE solution, the types of errors, etc.
Chapter 4 focuses on the study of structures under the assumption of
2D elasticity. These structures include dams, tunnels, pipes and retain-
ing walls, among many others. The key ideas of 2D elasticity theory are
explained, as well as the formulation of the 3-noded triangular element.
Details of the explicit form of the element stiffness matrix and the equiv-
alent nodal force vector are given.
Chapter 5 explains the derivation of the shape functions for 2D solid ele-
ments of rectangular and triangular shape and different orders of approxi-
mation. The resulting expressions for the shape functions are applicable to
axisymmetric solid elements, as well as for many plate and shell elements
studied in Volume 2.
Chapter 6 focuses on the formulation of 2D solid elements of arbitrary
shape (i.e. irregular quadrilateral and triangular elements with straight or
curved sides) using the isoparametric formulation and numerical integra-
tion. These concepts are essential for the organization of a general FEM
computer program applicable to elements of different shape and approx-
imation order. Examples of application to civil engineering constructions
are presented.
Chapter 7 describes the formulation of axisymmetric solid elements. Use is
made of the concepts explained in the previous two chapters, such as the
derivation of the element shape functions, the isoparametric formulation
and numerical integration. Applications to the analysis of axisymmetric
solids and structures are presented.
Chapter 8 studies 3D solid elements of tetrahedral and hexahedral shapes.
3D solid elements allow the FEM analysis of any structure. Details of the
derivation of the stiffness matrix and the equivalent nodal force vector
are given for the simple 4-noded tetrahedral element. The formulation of
higher order 3D solid elements is explained using the isoparametric formu-
lation and numerical integration. Applications of 3D solid elements to a
wide range of structures such as dams, buildings, historical constructions
and mechanical parts are presented.
Chapter 9 covers miscellaneous topics of general interest for FEM analysis.
These include the treatment of inclined supports, the blending of elements
of different types, the study of structures on elastic foundations, the use
of substructuring techniques, the procedures for applying constraints on
the nodal displacements, the computation of stresses at the nodes and the
key concepts of error estimation and adaptive mesh refinement strategies.
Chapter 10 introduces the basic ideas of mesh generation and visualization
of the FEM results. The advancing front method and the Delaunay method
for generation of unstructured meshes are explained in some detail.
Chapter 11 finally describes the organization of a simple computer pro-
gram for FEM analysis of 2D structures using the 3-noded triangle and
the 4-noded quadrilateral using MATLAB as a programming tool and the
GiD pre-postprocessing system.
The four annexes cover the basic concepts of matrix algebra (Annex A),
the solution of simultaneous linear algebraic equations (Annex B), the
computation of the parameters for adaptive mesh refinement analysis (An-
nex C) and details of the GiD pre-postprocessing system developed at
CIMNE (Annex D).
I want to express my gratitude to Dr. Francisco Zárate who was responsible
for writing the computer program Mat-fem explained in Chapter 11 and
also undertook the task of the writing this chapter.
Many thanks also to my colleagues in the Department of Continuum Me-
chanics and Structural Analysis at the Civil Engineering School of UPC for
their support and cooperation over many years. Special thanks to Profs.
Benjamı́n Suárez, Miguel Cervera and Juan Miquel and Drs. Francisco
Zárate and Daniel di Capua with whom I have shared the teaching of the
course on Finite Element Structural Analysis at UPC.
Many examples included in the book are the result of problems solved
by academics and research students at UPC and CIMNE in cooperation
with companies which are acknowledged in the text. I thank all of them
for their contributions. Special thanks to the GiD team at CIMNE for
providing the text for Annex D and many pictures shown in the book.
Many thanks also to my colleagues and staff at CIMNE for their coo-
peration and support during so many years that has made possible the
publication of this book.
I am particularly grateful to Prof. O.C. Zienkiewicz from University of
Swansea (UK) and Prof. R.L. Taylor from University of California at
Berkeley (USA). Their ideas and suggestions during many visits at CIMNE
and UPC in the period 1987-2007 have been a source of inspiration for
the writing of this book.
Prof. Zienkiewicz, one of the giants in the field of computational mecha-
nics, unfortunately passed away on January 2nd 2009 and has been unable
to see the publication of this book. I express my deep sorrow for such a
big loss and my recognition and gratitude for his support and friendship
throughout my career.
Thanks also to Mrs. Adriana Hanganu from CIMNE for supervising the
joint publication of the book by CIMNE and Springer.
Finally, my special thanks to Mrs. Marı́a Jesús Samper from CIMNE for
her excellent work in the typing and editing of the manuscript.

Eugenio Oñate
Barcelona, January 2009
Foreword

It is just over one-half century since papers on element based approximate

solutions to structural problems first appeared in print. The term Finite
Element Method was introduced in 1960 by Professor R.W. Clough to
define this class of solution methods. In 1967, Professor O.C. Zienkiewicz
published the first book describing applications of the method. Since these
early contributions the finite element method has become indispensable
to engineers and scientists involved in the analysis and design of a very
wide range of practical structural problems: These include concrete dams,
automobiles, aircraft, electronic parts, and medical devices, to name a few.
Professor Eugenio Oñate, the author of Structural Analysis with the Finite
Element Method, is a well recognized educator and research scholar in the
area of computational mechanics. He completed his doctoral studies under
the supervision of Professor O.C. Zienkiewicz at the University of Wales,
Swansea. Professor Oñate is the founder and director of the International
Center for Numerical Methods in Engineering (CIMNE) at the Universitat
Politècnica de Catalunya in Barcelona, Spain. He has more than thirty
years experience in development of finite element methods and related
software.
This two volume book presents the results of the author’s extensive expe-
rience in teaching and research on the finite element method. The content
of the book develops the theory and practical implementation of the fi-
nite element method for application to linear structural problems. In the
first volume, the finite element method is described to solve linear elastic
problems for solids. The second volume extends the method to solve beam,
plate and shell structures.
The style of presentation allows the reader to fully comprehend the fun-
damental steps in a finite element solution process. In the first volume,
the equations of elasticity are developed explicitly and are combined with
the principal of virtual work to describe the matrix problem to be solved.
The book starts with one dimensional problems and builds systemati-
cally through two and three dimensional applications for solids. The first
nine chapters present the theory of finite element analysis in detail – inclu-
ding the required steps to approximate element variables by isoparametric
shape functions, to carry out numerical integration, and to perform assem-
bly of final equations. Numerous examples are completely worked out and
are complemented by color plates of results from analyses of practical
problems. The first volume concludes with a chapter on mesh genera-
tion and visualization and a chapter on programming the finite element
method. Use of the GiD program permits the reader to rapidly generate
a mesh, while the chapter on programming describes how the reader can
combine the computational advantages of MATLAB with the graphical
capabilities of GiD to solve problems and visualize results. The reader can
attain a deeper understanding of the finite element method by studying
these chapters in parallel with the earlier theoretical chapters.
The second volume builds on the first to develop finite element formu-
lations for beam, plate and shell problems. The pattern of development
is identical with the first volume – namely starting with beam theories
and building systematically through the development of various plate and
shell finite element forms.
These two volumes enhance the reader’s ability to master the basic con-
cepts of the finite element method. Moreover, they provide the necessary
background for further study on inelastic material behavior, contact inte-
ractions, and large deformation of solids and shells. Thus, the book is an
extremely valuable contribution toward practical application of the finite
element method in analysis and design of structures.

Robert L. Taylor
University of California, Berkeley, USA
December 2008
Contents

1 INTRODUCTION TO THE FINITE ELEMENT

METHOD FOR STRUCTURAL ANALYSIS . . . . . . . . . . 1
1.1 WHAT IS THE FINITE ELEMENT METHOD? . . . . . . . . . 1
1.2 ANALYTICAL AND NUMERICAL METHODS . . . . . . . . . 1
1.3 WHAT IS A FINITE ELEMENT? . . . . . . . . . . . . . . . . . . . . . 3
1.4 STRUCTURAL MODELLING AND FEM ANALYSIS . . . 3
1.4.1 Classification of the problem . . . . . . . . . . . . . . . . . . . . . 3
1.4.2 Conceptual, structural and computational models . . . 3
1.4.3 Structural analysis by the FEM . . . . . . . . . . . . . . . . . . 8
1.4.4 Verification and validation of FEM results . . . . . . . . . 8
1.5 DISCRETE SYSTEMS. BAR STRUCTURES . . . . . . . . . . . 14
1.5.1 Basic concepts of matrix analysis of bar structures . . 14
1.5.2 Analogy with the matrix analysis of other discrete
systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.5.3 Basic steps for matrix analysis of discrete systems . . 20
1.6 DIRECT ASSEMBLY OF THE GLOBAL STIFFNESS
MATRIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.7 DERIVATION OF THE MATRIX EQUILIBRIUM
EQUATIONS FOR THE BAR USING THE PRINCIPLE
OF VIRTUAL WORK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.8 DERIVATION OF THE BAR EQUILIBRIUM
EQUATIONS VIA THE MINIMUM TOTAL
POTENTIAL ENERGY PRINCIPLE . . . . . . . . . . . . . . . . . . 25
1.9 PLANE FRAMEWORKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.9.1 Plane pin-jointed frameworks . . . . . . . . . . . . . . . . . . . . . 27
1.9.2 Plane rigid jointed frames . . . . . . . . . . . . . . . . . . . . . . . 29
XIV Contents

1.10 TREATMENT OF PRESCRIBED DISPLACEMENTS

AND COMPUTATION OF REACTIONS . . . . . . . . . . . . . . . 34
1.11 INTRODUCTION TO THE FINITE ELEMENT
METHOD FOR STRUCTURAL ANALYSIS . . . . . . . . . . . . 36
1.12 THE VALUE OF FINITE ELEMENT COMPUTATIONS
FOR STRUCTURAL DESIGN AND VERIFICATION . . . 42
1.13 CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

2 1D FINITE ELEMENTS FOR AXIALLY LOADED

RODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.2 AXIALLY LOADED ROD . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.3 AXIALLY LOADED ROD OF CONSTANT CROSS
SECTION. DISCRETIZATION IN ONE LINEAR ROD
ELEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.3.1 Approximation of the displacement field . . . . . . . . . . . 46
2.3.2 Derivation of equilibrium equations for the elements 48
2.3.3 Assembly of the global equilibrium equations . . . . . . . 51
2.3.4 Computation of the reactions . . . . . . . . . . . . . . . . . . . . 51
2.3.5 Computation of the axial strain and the axial force . 52
2.4 DERIVATION OF THE DISCRETIZED EQUATIONS
FROM THE GLOBAL DISPLACEMENT
INTERPOLATION FIELD . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.5 AXIALLY LOADED ROD OF CONSTANT CROSS
SECTION. DISCRETIZATION IN TWO LINEAR ROD
ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.5.1 Solution using the element shape functions . . . . . . . . . 57
2.5.2 Solution using the global shape functions . . . . . . . . . . 60
2.6 GENERALIZATION OF THE SOLUTION WITH N
LINEAR ROD ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.7 EXTRAPOLATION OF THE SOLUTION FROM TWO
DIFFERENT MESHES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.8 MATRIX FORMULATION OF THE ELEMENT
EQUATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.8.1 Shape function matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.8.2 Strain matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.8.3 Constitutive matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.8.4 Principle of Virtual Work . . . . . . . . . . . . . . . . . . . . . . . . 71
2.8.5 Stiffness matrix and equivalent nodal force vector . . . 72
 Contents XV

2.9 SUMMARY OF THE STEPS FOR THE ANALYSIS OF

A STRUCTURE USING THE FEM . . . . . . . . . . . . . . . . . . . . 75

3 ADVANCED ROD ELEMENTS AND

REQUIREMENTS FOR THE NUMERICAL
SOLUTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.2 ONE DIMENSIONAL C 0 ELEMENTS. LAGRANGE
ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.3 ISOPARAMETRIC FORMULATION AND
NUMERICAL INTEGRATION . . . . . . . . . . . . . . . . . . . . . . . . 82
3.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.3.2 The concept of parametric interpolation . . . . . . . . . . . 82
3.3.3 Isoparametric formulation of the two-noded rod
element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.3.4 Isoparametric formulation of the 3-noded quadratic
rod element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.4 NUMERICAL INTEGRATION . . . . . . . . . . . . . . . . . . . . . . . . 89
3.5 STEPS FOR THE COMPUTATION OF MATRICES
AND VECTORS FOR AN ISOPARAMETRIC ROD
ELEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.5.1 Interpolation of the axial displacement . . . . . . . . . . . . 92
3.5.2 Geometry interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . 93
3.5.3 Interpolation of the axial strain . . . . . . . . . . . . . . . . . . . 93
3.5.4 Computation of the axial force . . . . . . . . . . . . . . . . . . . 94
3.5.5 Element stiffness matrix . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.5.6 Equivalent nodal force vector . . . . . . . . . . . . . . . . . . . . . 95
3.6 BASIC ORGANIZATION OF A FINITE ELEMENT
PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.7 SELECTION OF ELEMENT TYPE . . . . . . . . . . . . . . . . . . . 97
3.8 REQUIREMENTS FOR CONVERGENCE OF THE
SOLUTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.8.1 Continuity condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.8.2 Derivativity condition . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.8.3 Integrability condition . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.8.4 Rigid body condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.8.5 Constant strain condition . . . . . . . . . . . . . . . . . . . . . . . . 103
3.9 ASSESSMENT OF CONVERGENCE REQUIREMENTS.
THE PATCH TEST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
XVI Contents

3.10 OTHER REQUIREMENTS FOR THE FINITE

ELEMENT APPROXIMATION . . . . . . . . . . . . . . . . . . . . . . . 105
3.10.1 Compatibility condition . . . . . . . . . . . . . . . . . . . . . . . . . 105
3.10.2 Condition of complete polynomial . . . . . . . . . . . . . . . . 106
3.10.3 Stability condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
3.10.4 Geometric –invariance condition . . . . . . . . . . . . . . . . . . 108
3.11 SOME REMARKS ON THE COMPATIBILITY AND
EQUILIBRIUM OF THE SOLUTION . . . . . . . . . . . . . . . . . . 108
3.12 CONVERGENCE REQUIREMENTS FOR
ISOPARAMETRIC ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . 110
3.13 ERROR TYPES IN THE FINITE ELEMENT SOLUTION 111
3.13.1 Discretization error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
3.13.2 Error in the geometry approximation . . . . . . . . . . . . . . 113
3.13.3 Error in the computation of the element integrals . . . 113
3.13.4 Errors in the solution of the global equation system . 114
3.13.5 Errors associated with the constitutive equation . . . . 116

4 2D SOLIDS. LINEAR TRIANGULAR AND

RECTANGULAR ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . 117
4.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.2 TWO DIMENSIONAL ELASTICITY THEORY . . . . . . . . . 119
4.2.1 Displacement field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4.2.2 Strain field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4.2.3 Stress field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.2.4 Stress-strain relationship . . . . . . . . . . . . . . . . . . . . . . . . 121
4.2.5 Principal stresses and failure criteria . . . . . . . . . . . . . . 125
4.2.6 Virtual work expression . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.3 FINITE ELEMENT FORMULATION. THREE-NODED
TRIANGULAR ELEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
4.3.1 Discretization of the displacement field . . . . . . . . . . . . 132
4.3.2 Discretization of the strain field . . . . . . . . . . . . . . . . . . 135
4.3.3 Discretization of the stress field . . . . . . . . . . . . . . . . . . . 137
4.3.4 Discretized equilibrium equations . . . . . . . . . . . . . . . . . 137
4.3.5 Stiffness matrix and equivalent nodal force vectors
for the 3-noded triangular element . . . . . . . . . . . . . . . . 142
4.4 THE FOUR NODED RECTANGULAR ELEMENT . . . . . . 146
4.4.1 Basic formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
4.4.2 Some remarks on the behaviour of the 4-noded
rectangle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
 Contents XVII

4.4.2.1 Reduced integration of the shear stiffness

terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
4.4.2.2 Addition of internal modes . . . . . . . . . . . . . . . . 152
4.4.2.3 Addition of incompatible modes . . . . . . . . . . . 153
4.4.2.4 Use of an assumed strain field . . . . . . . . . . . . . 153
4.5 PERFORMANCE OF THE 3-NODED TRIANGLE AND
THE 4-NODED RECTANGLE . . . . . . . . . . . . . . . . . . . . . . . . 154
4.6 CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

5 HIGHER ORDER 2D SOLID ELEMENTS. SHAPE

FUNCTIONS AND ANALYTICAL COMPUTATION
OF INTEGRALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
5.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
5.2 DERIVATION OF THE SHAPE FUNCTIONS FOR Co
TWO DIMENSIONAL ELEMENTS . . . . . . . . . . . . . . . . . . . . 158
5.2.1 Complete polynomials in two dimensions. Pascal
triangle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
5.2.2 Shape functions of C o rectangular elements. Natural
coordinates in two dimensions . . . . . . . . . . . . . . . . . . . . 159
5.3 LAGRANGE RECTANGULAR ELEMENTS . . . . . . . . . . . . 161
5.3.1 Four-noded Lagrange rectangle . . . . . . . . . . . . . . . . . . . 161
5.3.2 Nine-noded quadratic Lagrange rectangle . . . . . . . . . . 164
5.3.3 Sixteen-noded cubic Lagrange rectangle . . . . . . . . . . . 166
5.3.4 Other Lagrange rectangular elements . . . . . . . . . . . . . . 167
5.4 SERENDIPITY RECTANGULAR ELEMENTS . . . . . . . . . 167
5.4.1 Eigth-noded quadratic Serendipity rectangle . . . . . . . 169
5.4.2 Twelve-noded cubic Serendipity rectangle . . . . . . . . . . 171
5.4.3 Seventeen-noded quartic Serendipity rectangle . . . . . . 171
5.5 SHAPE FUNCTIONS FOR C 0 CONTINUOUS
TRIANGULAR ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.5.1 Area coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.5.2 Derivation of the shape functions for C 0 continuous
triangles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5.5.3 Shape functions for the 3-noded linear triangle . . . . . 175
5.5.4 Shape functions for the six-noded quadratic triangle 176
5.5.5 Shape functions for the ten-noded cubic triangle . . . . 177
5.5.6 Natural coordinates for triangles . . . . . . . . . . . . . . . . . . 178
5.6 ANALYTIC COMPUTATION OF INTEGRALS OVER
RECTANGLES AND STRAIGHT-SIDED TRIANGLES . . 178
XVIII Contents

5.7 GENERAL PERFORMANCE OF TRIANGULAR AND

RECTANGULAR ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . 183
5.8 ENHANCEMENT OF 2D ELASTICITY ELEMENTS
USING DRILLING ROTATIONS . . . . . . . . . . . . . . . . . . . . . . 185
5.9 CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

6 ISOPARAMETRIC 2D SOLID ELEMENTS.

NUMERICAL INTEGRATION AND APPLICATIONS 187
6.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
6.2 ISOPARAMETRIC QUADRILATERAL ELEMENTS . . . . 187
6.2.1 Stiffness matrix and load vector for the
isoparametric quadrilateral . . . . . . . . . . . . . . . . . . . . . . . 190
6.2.2 A comparison between the 8- and 9-noded
isoparametric quadrilaterals . . . . . . . . . . . . . . . . . . . . . . 193
6.3 ISOPARAMETRIC TRIANGULAR ELEMENTS . . . . . . . . 194
6.4 NUMERICAL INTEGRATION IN TWO DIMENSIONS . . 197
6.4.1 Numerical integration in quadrilateral domains . . . . . 198
6.4.2 Numerical integration over triangles . . . . . . . . . . . . . . . 199
6.5 NUMERICAL INTEGRATION OF THE ELEMENT
MATRICES AND VECTORS . . . . . . . . . . . . . . . . . . . . . . . . . 200
6.5.1 Numerical integration of the stiffness matrix . . . . . . . 200
6.5.2 Numerical integration of the equivalent nodal force
vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
6.6 COMPUTER PROGRAMMING OF K(e) AND f (e) . . . . . . 203
6.7 OPTIMAL POINTS FOR COMPUTING STRAINS AND
STRESSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
6.8 SELECTION OF THE QUADRATURE ORDER . . . . . . . . 209
6.9 PERFORMANCE OF 2D ISOPARAMETRIC SOLID
ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
6.10 THE PATCH TEST FOR SOLID ELEMENTS . . . . . . . . . . 214
6.11 APPLICATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
6.11.1 Analysis of concrete dams . . . . . . . . . . . . . . . . . . . . . . . 219
6.11.2 Analysis of an earth dam . . . . . . . . . . . . . . . . . . . . . . . . 221
6.11.3 Analysis of an underground tunnel . . . . . . . . . . . . . . . . 223
6.12 CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . 224

7 AXISYMMETRIC SOLIDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

7.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
7.2 BASIC FORMULATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
7.2.1 Displacement field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
 Contents XIX

7.2.2 Strain field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227

7.2.3 Stress field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
7.2.4 Constitutive equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
7.2.5 Principle of virtual work . . . . . . . . . . . . . . . . . . . . . . . . . 230
7.3 FINITE ELEMENT FORMULATION. THREE-NODED
AXISYMMETRIC TRIANGLE . . . . . . . . . . . . . . . . . . . . . . . . 231
7.3.1 Discretization of the displacement field . . . . . . . . . . . . 231
7.3.2 Discretization of the strain and stress fields . . . . . . . . 232
7.3.3 Equilibrium equations . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
7.3.4 The stiffness matrix for the 3-noded triangle . . . . . . . 235
7.3.5 Equivalent nodal force vectors for the 3-noded triangle238
7.4 OTHER RECTANGULAR OR STRAIGHT-SIDED
TRIANGULAR AXISYMMETRIC SOLID ELEMENTS . . 240
7.5 ISOPARAMETRIC AXISYMMETRIC SOLID ELEMENTS243
7.6 ANALOGIES BETWEEN THE FINITE ELEMENT
FORMULATIONS FOR PLANE ELASTICITY AND
AXISYMMETRIC SOLIDS . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
7.7 EXAMPLES OF APPLICATION . . . . . . . . . . . . . . . . . . . . . . 245
7.7.1 Infinitely long cylinder under external pressure . . . . . 245
7.7.2 Cylindrical tank with spherical dome under internal
pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
7.7.3 Semi-infinite elastic space under point load . . . . . . . . 249
7.8 CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . 249

8 THREE DIMENSIONAL SOLIDS . . . . . . . . . . . . . . . . . . . . . 250

8.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
8.2 BASIC THEORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
8.2.1 Displacement field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
8.2.2 Strain field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
8.2.3 Stress field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
8.2.4 Stress-strain relationship . . . . . . . . . . . . . . . . . . . . . . . . 253
8.2.5 Principal stresses, stress invariants and failure criteria 254
8.2.6 Virtual work principle . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
8.3 FINITE ELEMENT FORMULATION. THE FOUR-
NODED TETRAHEDRON . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
8.3.1 Discretization of the displacement field . . . . . . . . . . . . 257
8.3.2 Strain matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
8.3.3 Equilibrium equations . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
8.3.4 Stiffness matrix for the element . . . . . . . . . . . . . . . . . . 262
8.3.5 Equivalent nodal force vector for the element . . . . . . . 263
XX Contents

8.3.6 The performance of the 4-noded tetrahedron . . . . . . . 265

8.4 OTHER 3D SOLID ELEMENTS . . . . . . . . . . . . . . . . . . . . . . 265
8.5 RIGHT PRISMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
8.5.1 Right prisms of the Lagrange family . . . . . . . . . . . . . . 267
8.5.1.1 Linear right prism of the Lagrange family . . . 267
8.5.1.2 Quadratic right prism of the Lagrange family 270
8.5.1.3 Other hexahedral elements of the Lagrange
family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
8.5.2 Serendipity prisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
8.5.2.1 20-noded quadratic Serendipity prism . . . . . . 271
8.5.2.2 32-noded cubic Serendipity prism . . . . . . . . . . 274
8.6 STRAIGHT-EDGED TETRAHEDRA . . . . . . . . . . . . . . . . . . 277
8.6.1 Shape functions for the 10-noded quadratic
tetrahedron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
8.6.2 Shape functions for the 20-noded quadratic
tetrahedron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
8.7 COMPUTATION OF ELEMENT INTEGRALS . . . . . . . . . 284
8.7.1 Analytical computation of element integrals . . . . . . . . 284
8.8 3D ISOPARAMETRIC ELEMENTS . . . . . . . . . . . . . . . . . . . 286
8.9 NUMERICAL INTEGRATION . . . . . . . . . . . . . . . . . . . . . . . 290
8.9.1 Hexahedral elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
8.9.2 Tetrahedral elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
8.10 NUMERICAL INTEGRATION OF ELEMENT MATRICES292
8.10.1 Isoparametric hexahedral elements . . . . . . . . . . . . . . . . 292
8.10.2 Isoparametric tetrahedral elements . . . . . . . . . . . . . . . . 295
8.10.3 Selection of the quadrature order . . . . . . . . . . . . . . . . . 295
8.11 PERFORMANCE OF 3D SOLID ELEMENTS . . . . . . . . . . 296
8.12 EXAMPLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
8.12.1 Analysis of a gravity dam . . . . . . . . . . . . . . . . . . . . . . . . 297
8.12.2 Analysis of a double curvature arch dam . . . . . . . . . . . 298
8.12.3 Analysis of arch dams . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
8.12.4 Analysis of a flat in a building . . . . . . . . . . . . . . . . . . . . 301
8.12.5 Analysis of prismatic cellular caissons for harbour
piers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
8.12.6 Analysis of a nuclear containment building . . . . . . . . . 302
8.12.7 Analysis of historical constructions . . . . . . . . . . . . . . . . 302
8.12.8 Analysis of mechanical parts . . . . . . . . . . . . . . . . . . . . . 306
8.13 FINAL REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
 Contents XXI

9 MISCELLANEOUS: INCLINED SUPPORTS,

DISPLACEMENT CONSTRAINS, ERROR
ESTIMATION, MESH ADAPTIVITY ETC. . . . . . . . . . . . 309
9.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
9.2 BOUNDARY CONDITIONS IN INCLINED SUPPORTS . 309
9.3 JOINING DISSIMILAR ELEMENTS . . . . . . . . . . . . . . . . . . . 311
9.4 DISPLACEMENT CONSTRAINTS . . . . . . . . . . . . . . . . . . . . 314
9.4.1 General procedure to eliminate constrained DOFs . . 314
9.4.2 Use of Lagrange multipliers . . . . . . . . . . . . . . . . . . . . . 318
9.4.3 Penalty method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
9.5 NODAL CONDENSATION AND SUBSTRUCTURES . . . . 320
9.5.1 Nodal condensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
9.5.2 Substructuring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
9.6 STRUCTURAL SYMMETRY . . . . . . . . . . . . . . . . . . . . . . . . . 323
9.6.1 Symmetric solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
9.6.2 Cyclic symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
9.7 STRUCTURES ON ELASTIC FOUNDATION . . . . . . . . . . 327
9.8 COMPUTATION OF NODAL STRESSES . . . . . . . . . . . . . . 329
9.8.1 Global smoothing of stresses . . . . . . . . . . . . . . . . . . . . . 330
9.8.2 Direct local extrapolation . . . . . . . . . . . . . . . . . . . . . . . . 332
9.8.3 Superconvergent patch recovery techniques . . . . . . . . . 334
9.8.4 Iterative enhancement of the solution . . . . . . . . . . . . . 336
9.9 ERROR ESTIMATION AND MESH ADAPTIVITY . . . . . 338
9.9.1 Basic concepts of error estimation . . . . . . . . . . . . . . . . 338
9.9.2 Error measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
9.9.3 Error estimation techniques . . . . . . . . . . . . . . . . . . . . . . 339
9.9.4 Mesh adaptation strategies . . . . . . . . . . . . . . . . . . . . . . . 341
9.9.4.1 Mesh optimality criterion based on the
equal distribution of the global energy error . 343
9.9.4.2 Mesh optimality criterion based on the
global distribution of the density of the
energy error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
9.9.4.3 Mesh refinement strategy based on the
point-wise error in stresses . . . . . . . . . . . . . . . . 346
9.9.5 Construction of an adapted mesh . . . . . . . . . . . . . . . . . 348
9.9.6 Examples of mesh adaptivity . . . . . . . . . . . . . . . . . . . . . 348
9.9.6.1 Thick circular cylinder under internal pressure349
9.9.6.2 2D hollow dam under water pressure and
self-weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
XXII Contents

9.9.7 Conclusions from the examples . . . . . . . . . . . . . . . . . . . 354

10 GENERATION OF ANALYSIS DATA AND

VISUALIZATION OF NUMERICAL RESULTS . . . . . . . 355
10.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
10.2 THE IMPORTANCE OF A GOOD PRE AND POST
PROCESSING SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
10.3 THE GEOMETRICAL REPRESENTATION OF
STRUCTURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
10.4 MESH GENERATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
10.4.1 The advancing front method . . . . . . . . . . . . . . . . . . . . . 363
10.4.2 The paving method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
10.4.3 The Delaunay method . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
10.4.4 3D mesh generation with the Delaunay method . . . . . 375
10.5 VISUALIZATION OF NUMERICAL RESULTS . . . . . . . . . 378
10.6 CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . 381

11 LEARNING TO PROGRAM THE FEM WITH

MATLAB AND GID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
11.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
11.2 MAT-fem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
11.3 DATA FILES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
11.4 START . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
11.5 STIFFNESS MATRIX AND EQUIVALENT NODAL
FORCE VECTOR FOR SELF-WEIGHT . . . . . . . . . . . . . . . 390
11.5.1 Generalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
11.5.2 Computation and assembly of K(e) and f (e)
(self-weight) for 3-noded triangles and 4-noded
quadrilaterals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
11.6 EXTERNAL LOADS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
11.7 PRESCRIBED DISPLACEMENTS . . . . . . . . . . . . . . . . . . . . 396
11.8 SOLUTION OF THE EQUATIONS SYSTEM . . . . . . . . . . . 396
11.9 NODAL REACTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
11.10 STRESSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
11.10.1Generalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
11.10.2Computation of the stresses at the nodes . . . . . . . . . . 398
11.11 POSTPROCESSING STEP . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
11.12 GRAPHICAL USER INTERFACE . . . . . . . . . . . . . . . . . . . . 403
11.12.1Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
11.12.2Program execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
 Contents XXIII

11.12.3Postprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
11.13 EXAMPLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408

A MATRIX ALGEBRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414

A.1 DEFINITION OF MATRIX . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
A.1.1 Transpose of a matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
A.1.2 Square matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
A.1.3 Symmetric and antisymmetric matrix . . . . . . . . . . . . . 415
A.1.4 Null matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
A.1.5 Diagonal matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
A.1.6 Identity matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
A.1.7 Triangular matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
A.2 OPERATION WITH MATRICES . . . . . . . . . . . . . . . . . . . . . . 417
A.2.1 Multiplication of matrices . . . . . . . . . . . . . . . . . . . . . . . 417
A.2.2 Associative rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
A.2.3 Distributive rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
A.2.4 Product of a matrix by a vector and a scalar . . . . . . . 418
A.2.5 Sum and subtraction of matrices . . . . . . . . . . . . . . . . . . 419
A.2.6 Partition of a matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
A.2.7 Determinant of a matrix . . . . . . . . . . . . . . . . . . . . . . . . . 420
A.2.8 Inverse of a matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
A.3 EIGENVALUES AND EIGENVECTORS OF A MATRIX 422

B SOLUTION OF SIMULTANEOUS LINEAR

ALGEBRAIC EQUATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
B.1 DIRECT SOLUTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
B.2 ITERATIVE SOLUTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426

C COMPUTATION OF THE ELEMENT

REFINEMENT PARAMETER FOR AN
EQUAL DISTRIBUTION OF THE ERROR
AND ACCOUNTING FOR THE CHANGE IN THE
NUMBER OF ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428

D APPENDIX D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
D.1 THE GID PRE/POSTPROCESSING SYSTEM . . . . . . . . . 431
D.1.1 General features of GiD . . . . . . . . . . . . . . . . . . . . . . . . . 433
D.1.2 More features of GiD . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
D.1.3 The handling of geometrical data with GiD . . . . . . . . 437
XXIV Contents

D.1.4 Generation of analysis data and interfacing with

computer simulation codes via GiD . . . . . . . . . . . . . . . 438
D.1.5 Visualization of numerical results with GiD . . . . . . . . 441
D.1.6 Who can benefit from using GiD? . . . . . . . . . . . . . . . . 441
D.1.7 Which are the advantages of using GiD? . . . . . . . . . . . 444
D.1.8 Why is it worth using GiD? . . . . . . . . . . . . . . . . . . . . . . 444
D.1.9 How can one learn to use GiD? . . . . . . . . . . . . . . . . . . . 444
D.1.10How can one access GiD? . . . . . . . . . . . . . . . . . . . . . . . . 444
D.1.11The GiD team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447

Author index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463

Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467

1
INTRODUCTION TO THE FINITE
ELEMENT METHOD FOR
STRUCTURAL ANALYSIS

1.1 WHAT IS THE FINITE ELEMENT METHOD?

The Finite Element Method (FEM) is a procedure for the numerical solu-
tion of the equations that govern the problems found in nature. Usually the
behaviour of nature can be described by equations expressed in differential
or integral form. For this reason the FEM is understood in mathematical
circles as a numerical technique for solving partial differential or integral
equations. Generally, the FEM allows users to obtain the evolution in
space and/or time of one or more variables representing the behaviour of
a physical system.
When referred to the analysis of structures the FEM is a powerful
method for computing the displacements, stresses and strains in a struc-
ture under a set of loads. This is precisely what we aim to study in this
book.

1.2 ANALYTICAL AND NUMERICAL METHODS

The conceptual difference between analytical and numerical methods is
that the former search for the universal mathematical expressions repre-
senting the general and “exact” solution of a problem governed typically by
mathematical equations. Unfortunately exact solutions are only possible
for a few particular cases which frequently represent coarse simplifications
of reality.
On the other hand, numerical methods such as the FEM aim to pro-
viding a solution, in the form of a set of numbers, to the mathematical
equations governing a problem. The strategy followed by most numerical
2 Introduction to the finite element method for structural analysis

methods is to transform the mathematical expressions into a set of alge-

braic equations which depend on a finite set of parameters. For practical
problems these equations involve many thousands (or even millions) of un-
knowns and therefore the final system of algebraic equations can only be
solved with the help of computers. This explains why even though many
numerical methods were known since the XVIII century, their develop-
ment and popularity has occurred in tandem to the progress of modern
computers in the XX century. The term numerical method is synonymous
of computational method in this text.
Numerical methods represent, in fact, the return of numbers as the true
protagonists in the solution of a problem. The loop initiated by Pythagoras
some 25 centuries ago has been closed in the last few decades with the
evidence that, with the help of numerical methods, we can find precise
answers to any problem in science and engineering.
We should keep in mind that numerical methods for structural engi-
neering are inseparable from mathematics, material modelling and com-
puter science. Nowadays it is unthinkable to attempt the development of
a numerical method for structural analysis without referring to those dis-
ciplines. As an example, any method for solving a large scale structural
problem has to take into account the hardware environment where it will
be implemented (most frequently using parallel computing facilities). Also
a modern computer program for structural analysis should be able to in-
corporate the continuous advances in the modelling of new materials.
The concept which perhaps best synthesizes the immediate future of
numerical methods is “multidisciplinary computations”. The solution of
problems will not be attempted from the perspective of a single discipline
and it will involve all the couplings which characterize the complexity of
reality. For instance, the design of a structural component for a vehicle (an
automobile, an aeroplane, etc.) will take into account the manufacturing
process and the function which the component will play throughout its
life time. Structures in civil engineering will be studied considering the
surrounding environment (soil, water, air). Similar examples are found in
mechanical, naval and aeronautical engineering and indeed in practically
all branches of engineering science. Accounting for the non-deterministic
character of data will be essential for estimating the probability that the
new products and processes conceived by men behave as planned. The
huge computational needs resulting from a stochastic multidisciplinary
viewpoint will demand better numerical methods, new material models
and, indeed, faster computers.
 Structural modelling and FEM analysis 3

It is only through the integration of a deep knowledge of the physi-

cal and mathematical basis of a problem and of numerical methods and
informatics, that effective solutions will be found for the large-scale multi-
disciplinary problems in structural engineering of the twenty-first century.

1.3 WHAT IS A FINITE ELEMENT?

A finite element can be visualized as a small portion of a continuum (in this
book a solid or a structure). The word “finite” distinguishes such a portion
from the “infinitesimal” elements of differential calculus. The geometry of
the continuum is considered to be formed by the assembly of a collection
of non-overlapping domains with simple geometry termed finite elements.
Triangles and quadrilaterals in two dimensions (2D) or tetrahedra and
hexahedra in three dimensions (3D) are typically chosen to represent the
“elements”. It is usually said that a “mesh” of finite elements “discretizes”
the continuum (Figure 1.1). The space variation of the problem parameters
(i.e. the displacements in a structure) is expressed within each element by
means of a polynomial expansion. Since the “exact” analytical variation of
such parameters is more complex and generally unknown, the FEM only
provides an approximation to the exact solution.

1.4 STRUCTURAL MODELLING AND FEM ANALYSIS

1.4.1 Classification of the problem
The first step in the solution of a problem is the identification of the pro-
blem itself. Hence, before we can analyze a structure we must ask ourselves
the following questions: Which are the more relevant physical phenomena
influencing the structure? Is the problem of static or dynamic nature?
Are the kinematics or the material properties linear or non-linear? Which
are the key results requested? What is the level of accuracy sought? The
answers to these questions are essential for selecting a structural model
and the adequate computational method.

1.4.2 Conceptual, structural and computational models

Computational methods, such as the FEM, are applied to conceptual mo-
dels of a real problem, and not to the actual problem itself. Even experi-
mental methods in structural laboratories make use of scale reproductions
of the conceptual model chosen (also called physical models) unless the
4 Introduction to the finite element method for structural analysis

Fig. 1.1 Discretization of different solids and structures with finite elements
 Structural modelling and FEM analysis 5

actual structure is tested in real size, which rarely occurs. A conceptual

model can be developed once the physical nature of a problem is clearly
understood. In the derivation of a conceptual model we should aim to
exclude superfluous details and include all the relevant features of the
problem under consideration so that the model can describe reality with
enough accuracy.
A conceptual model for the study of a structure should include all the
data necessary for its representation and analysis. Clearly different persons
will have different perceptions of reality and, consequently, the conceptual
model for the same structure can take a variety of forms.
After selecting a conceptual model of a structure, the next step for the
numerical (and analytical) study is the definition of a structural model
(sometimes called mathematical model).
A structural model must include three fundamental aspects. The geo-
metric description of the structure by means of its geometrical compo-
nents (points, lines, surfaces, volumes), the mathematical expression of
the basic physical laws governing the behaviour of the structure (i.e. the
force-equilibrium equations and the boundary conditions) usually written
in terms of differential and/or integral equations and the specification of
the properties of the materials and of the loads acting on the structure.
Clearly the same conceptual model of a structure can be analyzed using
different structural models depending on the accuracy and/or simplicity
sought in the analysis. As an example, a beam can be modelled using the
general 3D elasticity theory, the 2D plane stress theory or the simpler
beam theory. Each structural model provides a different set out for the
analysis of the actual structure. We should bear in mind that a solution
found by starting from an incorrect conceptual or structural model will be
a wrong solution, far from correct physical values, even if obtained with
the most accurate numerical method.
The next step in the structural analysis sequence is the definition of
a numerical method, such as the FEM. The application of the FEM in-
variably requires its implementation in a computer code. The analysis of
a structure with the FEM implies feeding the code with quantitative in-
formation on the mechanical properties of the materials, the boundary
conditions and the applied loads (the physical parameters) as well as the
features of the discretization chosen (i.e. element type, mesh size, etc).
The outcome of this process is what we call a computational model for the
analysis of a structure (Figure 1.2).
6 Introduction to the finite element method for structural analysis

Fig. 1.2 The path from the real structure to the computational model

In this book we will study the application of the FEM to a number

of structural models covering most structures found in the engineering
practice. The material properties will be considered to be linear elastic.
Furthermore the analysis will be restricted to linear kinematics and to
static loading. The structures are therefore analyzed under linear static
conditions
conditions. Despite their simplicity, these assumptions are applicable to
most of the situations found in the everyday practice of structural analysis
and design.
The structural models considered in this book are classified as solid
models (2D/3D solids and axisymmetric solids), beam and plate models
and shell models (faceted shells, axisymmetric shells and curved shells).
Figure 1.3 shows the general features of a typical member of each struc-
tural model family. The structures that can be analyzed with these models
 Structural modelling and FEM analysis 7

Fig. 1.3 Structural models for some structures

include frames, buildings, slabs, foundations, retaining walls, dams, tun-

nels, bridges, cylindrical tanks, shell roofs, ship hulls, mechanical parts,
airplane fuselages, vehicle components, etc.
Volume 1 of this book studies structures that can be analyzed using
solid finite element models. The finite element analysis of beam, plate and
shell structures is covered in Volume 2 [On].
8 Introduction to the finite element method for structural analysis

1.4.3 Structural analysis by the FEM

The geometry of a structure is discretized when it is split into a mesh of
finite elements of a certain accuracy. Clearly the discretization introduces
another approximation. With respect to reality we have therefore two error
sources from the outset: the modelling error and the discretization error.
The former can be reduced by improving the conceptual and structural
models which describe the actual behaviour of the structure, as previously
explained. The discretization error, on the other hand, can be reduced by
using a finer mesh (i.e. more elements), or else by increasing the accuracy
of the finite elements chosen using higher order polynomial expansions for
approximating the displacement field within each element.
Additionally, the use of computers introduces numerical errors asso-
ciated with their ability to represent data accurately with numbers of
finite precision. The numerical error is usually small, although it can be
large in some problems, such as when some parts of the structure have
very different physical properties. The sum of discretization and numerical
errors contribute to the error of the computational model. Note that even
if we could reduce the computational error to zero, we would not be able
to reproduce accurately the actual behaviour of the structure, unless the
conceptual and structural models were perfect.
Figure 1.4 shows schematically the discretization of some geometrical
models of structures using finite elements. Figure 1.5 shows the actual
image of a car panel, the geometrical definition of the panel surface by
means of NURBS (non-uniform rational B-splines) patches [PT] using
computer-aided design (CAD) tools (see Chapter 10), the discretization
of the surface by a mesh of 3-noded shell triangles and some numerical
results of the FEM analysis. The differences between the real structure
of the panel, the geometrical description and the analysis mesh can be
seen clearly. A similar example of the FEM analysis of an office building
is shown in Figure 1.6.

1.4.4 Verification and validation of FEM results

Developers of structural finite element computer codes, analysts who use
the codes and decision makers who rely on the results of the analysis face
a critical question: How should confidence in modelling and computation
be critically assessed? Validation and verification of FEM results are the
primary methods for building and quantifying this confidence. In essence,
validation is the assessment of the accuracy of the structural and compu-
 Structural modelling and FEM analysis 9

Fig. 1.4 Discretization of structural models into finite elements

tational models by comparison of the numerical results with experimental

data. Experiments are usually performed in laboratory using scale models
10 Introduction to the finite element method for structural analysis

Fig. 1.5 (a) Actual geometry of an automotive panel. (b) CAD geometrical des-
cription by NURBS patches. (c) Finite element mesh of 3-noded shell triangles
discretizing the panel geometry. (d) FEM numerical results of the structural ana-
lysis showing the equivalent strain distribution. Images by courtesy of Quantech
ATZ S.A., www.quantech.es

of a structure, and in special occasions on actual structures. The correct

definition of the experimental tests and the reliability of the experimental
results are crucial issues in the validation process.
Verification, on the other hand, is the process of determining that
a computational model accurately represents the underlying structural
model and its solution. In verification, therefore, the relationship between
the numerical results to the real world is not an issue. The verification of
FEM computations is made by comparing the numerical results for sim-
ple benchmark problems with “exact” solutions obtained analytically, or
using more accurate numerical methods. Figure 1.7 shows an scheme of
the verification and validation steps [ASME,Sch].
A careful examination of the verification process indicates that there
are two fundamental parts of verification: 1) code verification, in order
to establish confidence that the mathematical model and the solution al-
gorithms are working correctly, and 2) calculation verification aiming to
establish confidence that the discrete solution of the mathematical model
is accurate.
Among the code verification techniques, the most popular one is to
compare code outputs with analytical solutions. As the number of such
 Structural modelling and FEM analysis 11

Fig. 1.6 FEM analysis of the Agbar tower (Barcelona). Actual structure and dis-
cretization into shell and 3D beam elements. Deformed mesh (amplified) under
wind load. Images are courtesy of Compass Ingenierı́a y Sistemas SA,
www.compassis.com and Robert Brufau i Associats, S.A. www.robertbrufau.com

solutions is very limited, a code verification procedure with the potential

to greatly expand is the use of manufactured solutions.
The basic concept of a manufactured solution is simple. Given a partial
differential equation (PDE) and a code that provides general solutions of
that PDE, an arbitrary solution to the PDE is manufactured, i.e. made
up, then substituted into the PDE along with associated boundary condi-
tions, also manufactured. The result is a forcing function (right-hand side)
that exactly reproduces the originally selected manufactured solution. The
code is then subjected to this forcing function and the numerical results
12 Introduction to the finite element method for structural analysis

Fig. 1.7 Scheme of the verification and validation processes in the FEM. Flowchart
concept taken from [ASME,Sch] and reprinted by permission of the American
Society of Mechanical Engineering (ASME)

compared with the manufactured solution. If the code has no errors the
two solutions should agree [Sch].
As an illustration of a manufactured solution, let us consider the ordi-
nary differential equations for an Euler-Bernouilli beam of length L with
a constant cross section (Chapter 1 of Volume 2 [On] and [Ti])

d4 w
EI = f (x)
dx4
where w is the beam deflection, E and I are the Young modulus and the
inertia of the beam cross section, respectively and f (x) is a uniformly
 Structural modelling and FEM analysis 13

distributed loading. The following manufactured solution is assumed

αx
w(x) = A sin + Bex/L + C
L
where the four constants A, α, B and C are determined from the boun-
dary conditions. Substitution of the manufactured solution into the beam
equation results in the following expression for the loading term
· ³ ´ ¸
α 4 αx B x/L
f (x) = EI A sin + 4e
L L L
This loading function would be prescribed as input data to the discrete
beam finite element code and the code’s solution for w(x) is then compared
with the selected manufactured solution.
Code verification is only half of the verification effort. The other half
is the calculation verification, or, in other words, estimating the error in
the numerical solution due to discretization. These errors can be appraised
using error estimation techniques (Chapter 9). A more accurate numerical
solution can be found with a finer discretization or by using higher order
elements.
The subsequent validation step (Figure 1.7) has the goal of assessing
the predictive capability of the model. This assessment is made by com-
paring the numerical results with validation experiments performed on
physical models in laboratory or in real structures. If these comparisons
are satisfactory, the model is deemed validated for its intended use. In
summary, the validation exercise provides insight on the capacity of the
overall structural model to reproduce the behaviour of a real structure (or
the physical model chosen) with enough precision. Although both the ac-
curacy of the structural model and the computational method are assessed
in a validation process, a large validation error for an already verified code
typically means that the structural model chosen is not adequate and that
a better structural model should be used.
In conclusion, verification serves to check that we are solving structural
problems accurately, while validation tell us that we are solving the right
problem. Simply put, if the model passes the tests in the verification and
validation plan, then it can be used to make the desired predictions with
confidence. More details on the issue of verification and validation of the
FEM in solid mechanics can be found in [ASME,Ro,Sch].
In the following sections we will revisit the basic concepts of the matrix
analysis of bar structures, considered here as a particular class of the so-
called discrete systems. Then we will summarize the general steps in the
14 Introduction to the finite element method for structural analysis

Fig. 1.8 Some discrete systems. Elements and joint points (nodes)

analysis of “continuous” structures by the FEM. The interest of classical

matrix structural analysis is that it provides a general solution framework
which reassembles very closely that followed in the FEM.

1.5 DISCRETE SYSTEMS. BAR STRUCTURES

The solution of many technical problems requires the analysis of a network
system formed by different “elements”connected by their extremities or
joints, and subjected to a set of “loads” which are usually external to the
system. Examples of such systems, which we will call discrete systems,
are common in structural engineering (pin-jointed bar structures, frames,
grillages, etc.) and in many other different engineering problems, e.g.:
hydraulic piping networks, electric networks, transport planning networks,
production organization systems (PERT, etc) amongst others. Figure 1.8
shows some of these discrete systems.
Discrete systems can be studied using matrix analysis procedures which
have a very close resemblance to the FEM. In Appendix A the basic con-
cepts of matrix algebra are summarized. An outline of matrix analysis
techniques for bar structures and other discrete systems such as electric
and hydraulic networks is presented in the next section.

1.5.1 Basic concepts of matrix analysis of bar structures

Matrix analysis is the most popular technique for the solution of bar struc-
tures [Li,Pr]. Matrix analysis also provides a general methodology for the
application of the FEM to other structural problems. A good knowledge
of matrix analysis is essential for the study of this book.
 Discrete systems. Bar structures 15

Fig. 1.9 Deformation of a bar subjected to axial end forces. Number in brackets
at joints denotes global joint number

The matrix equations for a bar structure are obtained from the equa-
tions expressing the equilibrium of forces for each bar and for the structure
as a whole. Let us consider an isolated bar, e, of length l(e) subjected to
(e) (e)
axial forces Fx1 and Fx2 acting at the beam joints (Figure 1.9). The x
axis has the direction of the bar. Strength of Materials defines the strain
at any point in the bar by the relative elongation [Ti], i.e.
(e) (e)
∆l(e) u −u
ε= (e)
= 2 (e) 1 (1.1)
l l
(e) (e)
where u1 and u2 are the displacements of the joint points 1 and 2 in
the x direction, respectively. In Eq.(1.1) and the following the superindex
e denotes values associated to an individual bar. Generally indexes 1 and
2 are local joint numbers for the bar and correspond to the actual global
(e) (e)
numbers i, j of the joints in the structure. Hence u1 = ui and u2 = uj
(Figure 1.9 and Example 1.1).
The axial stress σ is related to the strain ε by Hooke law [Ti] as
(e) (e)
u −u
σ = Eε = E 2 (e) 1 (1.2)
l
where E is the Young modulus of the material. The axial force N at each
section is obtained by integrating the stress over the cross sectional area.
The axial force N is transmitted to the adjacent bars through the joints.
For homogeneous material we have (Figure 1.9)
(e) (e)
(e) u2 − u1 (e)
N2 = A(e) σ = (EA)(e) = N1 (1.3)
l(e)
The force equilibrium equation for the bar of Figure 1.9 is simply
Fx(e)
1
+ Fx(e)
2
=0 (1.4a)
16 Introduction to the finite element method for structural analysis

(ltx )(e) (e) (e)

Fx(e) + = k(e) (u1 − u2 )
1
2
µ ¶(e)
(ltx )(e) (e) (e) EA
Fx(e) + = k(e) (u2 − u1 ) , k (e) =
2
2 l
Fig. 1.10 Equilibrium equations for a bar subjected to axial joint forces and a
(e)
uniformly distributed axial load tx

with (e) (e)

(e) (e) u2 −u1 (e) (e)
Fx2 = N2 = (EA)(e) l(e)
= k (e) (u2 − u1 )
and (1.4b)
(e) (e) (e) (e) (e)
Fx1 = −Fx2 = k (e) (u1 − u2 ) = −N1
¡ ¢(e)
where k (e) = EA l . Eqs.(1.4b) can be written in matrix form as
( ) · ¸ ( (e) )
(e)
(e) Fx1 (e) 1 −1 u1
q = (e) =k (e) = K(e) a(e) (1.5a)
Fx2 −1 1 u2

where · ¸
(e) (e) 1 −1
K =k (1.5b)
−1 1
is the stiffness matrix of the bar, which depends on the geometry of the
(e) (e)
bar (l(e) , A(e) ) and its mechanical properties (E (e) ) only; a(e) = [u1 , u2 ]T
(e) (e)
and q(e) = [Fx1 , Fx2 ]T are the joint displacement vector and the joint
equilibrating force vector for the bar, respectively.
(e)
A uniformly distributed external axial load of intensity tx can easily
be taken into account by adding one half of the total external load to
each axial force at the bar joints. The equilibrium equations now read
(Figure 1.10)
( ) · ¸ ( (e) ) ½ ¾
(e)
(e) Fx1 (e) 1 −1 u1 (ltx )(e) 1
q = (e) =k (e) − = K(e) a(e) − f (e)
Fx2 −1 1 u2 2 1
(1.6a)
where  
fx(e)  (lt )(e) ½ ¾
(e) 1 x 1
f = = (1.6b)
f (e)  2 1
x2
 Discrete systems. Bar structures 17

(1) (2)
Px2 − Fx(1)
2
− Fx(2)
1
=0 , or Fx2 + Fx1 = Px2

(1) (2)
Fig. 1.11 Equilibrium of axial forces Fx2 and Fx1 and external force Px2 at joint 2
connecting bars 1 and 2. Number in brackets at joint denotes global joint number

is the vector of forces at the beam joints due to the distributed loading.
The equilibrium equations for the whole structure are obtained by im-
posing the equilibrium of axial and external forces at each of the N joints.
This condition can be written as [Li,Pr]
ne
X
Fx(e)
i
= Pxj , j = 1, N (1.7)
e=1

The sum on the left hand side (l.h.s.) of Eq.(1.7) extends over all bars
ne sharing the joint with global number j and Pxj represents the external
(e)
point load acting at that joint (Figure 1.11). The joint forces Fxi for each
bar are expressed in terms of the joint displacements using Eq.(1.6). This
process leads to the system of global equilibrium equations. In matrix form
    
K11 K12 · · · · · · K1N  u1   f1 
   f2 
 K21 K22 · · · · · · K2N  u2
 ..  .. = ..
.  . 
   .  
KN 1 KN 2 · · · · · · KN N uN fN
or
Ka = f (1.8a)
where K is the global stiffness matrix of the structure and a and f are
the global joint displacement vector and the global joint force vector,
respectively. The derivation of Eq.(1.8a) is termed the assembly process.
Solution of Eq.(1.8a) yields the displacements at all joint points from
which the value of the axial force in each bar can be computed as
(e) (e)
u2 − u1
N (e) = (EA)(e) (1.8b)
l(e)
The axial forces at the joints can be computed from Eqs.(1.4b) and
(1.6a) as ( )
(e)
−N1
q(e) = (e) = K(e) a(e) − f (e) (1.9)
N2
18 Introduction to the finite element method for structural analysis

(e) (e)
Note that N2 = −N1 = N (e) .
The components of q(e) can therefore be interpreted as the joint equi-
librating forces for each bar necessary for imposing global equilibrium of
forces at the joints (Eq.(1.6a)), or as the axial forces at the bar joints
(Eq.(1.9)) which are useful for design purposes. This coincidence will be
exploited later in the book for computing the resultant stresses at each
node for bar and beam finite elements by expressions similar to Eq.(1.9).
The assembled expression for vector q(e) yields the reactions at the
nodes with constrained displacements. The vector of nodal reactions can
be computed from the global stiffness equations as

r = q = Ka − f ext (1.10a)

where r contains the reactions at the constrained nodes and f ext contains
global joint forces due to external loads only. Clearly the sum of the re-
actions and the external joint forces gives the global joint force vector f,
i.e.
f = f ext + r (1.10b)

1.5.2 Analogy with the matrix analysis of other discrete systems

The steps between Eqs.(1.1) and (1.8) are very similar for many discrete
systems. For instance, the study of a single resistance element 1-2 in an
electric network (Figure 1.12a) yields the following relationship between
the currents entering the resistance element and the voltages at the end
points of the resistance (Ohm law)

(e) (e) 1 (e) (e) (e) (e)

I1 = −I2 = (e)
(V1 − V2 ) = k (e) (V1 − V2 ) (1.11a)
R

This equation is identical to Eq.(1.4) for the bar element if the current
intensities and the voltages are replaced by the joint forces and the joint
¡ ¢(e)
displacements, respectively, and 1/R(e) by EA l . Indeed, if uniformly
(e)
distributed external currents tx are supplied along the length of the ele-
ment, the force term f (e) of Eq.(1.6a) is found. The “assembly rule” is the
well known Kirchhoff law stating that the sum of all the current intensities
arriving at a joint must be equal to zero, i.e.
ne
X (e)
Ii = Ij , j = 1, N (1.11b)
e=1
 Discrete systems. Bar structures 19

Fig. 1.12 a) Electrical resistance, b) Fluid carrying pipe. Equations of equilibrium

where Ii is the external current intensity entering joint i and N is the

total number of joints. Note the analogy between Eqs.(1.11b) and (1.7).
The same analogy can be found for fluid carrying pipe networks. The
equilibrium equation relating fluid flow q and hydraulic head h at the ends
of a single pipe element can be written as (Figure 1.12b)
(e) (e) (e) (e)
q1 = −q2 = k (e) (h1 − h2 ) (1.12a)

where k (e) is a parameter which is a function of the pipe roughness and

the hydraulic head. This implies that the terms of the stiffness matrix
(e)
K(e) for a pipe element are known functions of the joint heads hi . The
equilibrium equation for each pipe element is written as in Eq.(1.6) where
(e) (e) (e) (e) (e)
ui and Fxi are replaced by hi and qi , respectively and tx represents
the input of a uniformly distributed flow source along the pipe length.
The assembly rule simply states that at each of the N pipe joints the
sum of the flow contributed by the adjacent pipe elements should equal
the external flow source, i.e.
ne
X (e)
qi = qj , j = 1, N (1.12b)
e=1

The global equilibrium equations are assembled similarly as for the bar
element yielding the system of Eqs.(1.8a). In the general problem matrix K
will be a function of the nodal hydraulic head via the k (e) parameter. Ite-
rative techniques for solving the resulting non-linear system of equations
are needed in this case.
20 Introduction to the finite element method for structural analysis

1.5.3 Basic steps for matrix analysis of discrete systems

What we have seen this far leads us to conclude that the analysis of a
discrete system (i.e. a bar structure) involves the following steps:
a) Definition of a network of discrete elements (bars) connected among
themselves by joints adequately numbered. Each element e has known
geometrical and mechanical properties. All these characteristics con-
stitute the problem data and should be defined in the simplest possible
way (preprocessing step).
b) Computation of the stiffness matrix K(e) and the joint force vector
f (e) for each element of the system.
c) Assembly and solution of the resulting global matrix equilibrium equa-
tion (Ka = f ) to compute the unknown parameters at each joint, i.e.
the displacements for the bar system.
d) Computation of other relevant parameters for each element, i.e. the
axial strain and the axial force, in terms of the joint parameters.
The results of the analysis should be presented in graphical form to fa-
cilitate the assessment of the system’s performance (postprocessing step).

Example 1.1: Compute the displacements and axial forces in the three-bar
structure of Figure 1.13 subjected to an horizontal force P acting at its
right hand end.

Fig. 1.13 Analysis of a simple three-bar structure under an axial load

- Solution
The equilibrium equations for each joint are (see Eq.(1.5a))
(
(1)
) · ¸ ( (1) )
Fx1 (1) 1 −1 u1
Bar 1 =k
(1)
Fx2 −1 1 (1)
u2
 Discrete systems. Bar structures 21

( ) · ¸ ( (2) )
(2)
Fx1 (2) 1 −1 u1
Bar 2 =k
(2)
Fx2 −1 1 u
(2)
2
( ) · ¸ ( (3) )
(3)
Fx1 (3) 1 −1 u1
Bar 3 =k
(3)
Fx2 −1 1 (3)
u2

with k (1) = k (2) = EA

l and k
(3)
= 2EA
l .
The compatibility equations between local and global displacements are
(1) (1) (2)
u1 = u1 ; u 2 = u3 ; u 1 = u2
(2) (3) (3)
u2 = u3 ; u 1 = u3 ; u 2 = u4
Applying the assembly equation (1.7) to each of the four joints we have

3
X 3
X
joint 1: Fx(1)
i
= −R1 , joint 2: Fx(1)
i
= −R2
e=1 e=1
3
X 3
X
joint 3: Fx(1)
i
=0 , joint 4: Fx(1)
i
=P
e=1 e=1
(e)
Substituting the values of Fxi from the bar equilibrium equations gives
(1) (1) (2) (2)
joint 1 : k (1) (u1 − u2 ) = −R1 , joint 2 : k (2) (u1 − u2 ) = −R2
(1) (1) (2) (2) (3) (3)
joint 3 : k (1) (−u1 + u2 ) + k (2) (−u1 + u2 ) + k (3) (u1 + u2 ) = 0
(3) (1)
joint 4 : k (3) (−u1 + u2 ) = P

Above equations can be written in matrix form using the displacement com-
patibility conditions as

1 2 3 4
 (1)
 (1)    
1 k 0 −k 0  u −R1 
 1   
2 0 k (2) −k (2) 0   u2 −R2
=
3 −k −k (k + k (2) + k (3) ) −k (3) 
(1) (2) (1)
u3 
    0  
4 0 0 −k (3) k (3) u4 P

Note that an external point load acting at node j can be placed directly in
the jth position of the global joint force vector f .
Substituting the values of k (e) for each bar and imposing the boundary con-
ditions u1 = u2 = 0, the previous system can be solved to give

Pl Pl P
u3 = ; u4 = ; R 1 = R2 =
2EA EA 2
Discovering Diverse Content Through
Random Scribd Documents
 CHAPTER IV
BORROW AND BOWRING

We now have Borrow a youth of twenty-two. His life has been full of
weird adventure, but to all appearances quite unprofitable in any
worldly sense. His future is nebulous. Dreams are dreamed; visions
are vanished. He seems to be farther from fame and fortune than
when he set off in the coach for London, with the green box in the
boot carrying his Danish ballads and his “Ab Gwilym.” His castles in
the clouds have come crashing to earth in irremediable ruin.
Borrow was indignant with a scurvy world which had treated him
harshly. The plain truth was that the world had no feeling about him
at all, one way or the other. He had nothing to sell that anybody
wanted to buy, and no means of making a living. He had a long
road to travel before he found himself. In 1825 he went home to
Norwich a failure, with the sense of defeat very strong upon him.
The mother who was at once his best adviser and sincerest
worshipper was not likely to chide his folly as the father had done.
She was ready to receive him with demonstrations of love, and to
share her little with him. This was part of the ignominy which he
hated—that he was obliged to impose himself upon the household in
Willow Lane. In a world out of joint, the cursed spite was that he
could do nothing to set it right.
Long time he struggled hard to lift himself out of this rut. He
continued to fail. When at last he did succeed, these years became
to him a horrible nightmare. He would not speak of them; he tried
not to think of them. He resolutely refused to permit the public a
glimpse into the sordid secrets they contained. From 1825 to 1832
he lived a life of which he wished nobody to know anything. Out of
some correspondence between him and Richard Ford arose the
phrase, “the Veiled Period.” Ford implored him to lift the veil a little
and allow his admirers to know what he was doing. There were
many reasons why he declined to do so. He endeavoured to puzzle
the public about it, and perhaps succeeded partly in mystifying
himself. He suggested a kind of vague romance of wanderings in
remote parts of Europe. Some of the suggestions were founded on
a slight basis of fact; that is all that can be said for them.
As to the facts: there is no doubt that he did buy a horse with
money lent to him by Ambrose Smith, and sell it at a profit. As in
the case of Isopel, it may not be unwise to allow some discount off
the published accounts of the transaction. Very possibly the horse
was not such a fine horse as that noble animal with whose
assistance Lavengro electrified the jockeys at Horncastle Fair;
perhaps the profit on the sale was not so great as it was made to
appear in “The Romany Rye.” But there was such a transaction.
Ambrose Smith reminded him of it, long years afterwards, when he
visited the great author at Oulton.
Soon after his return to Norwich, he was busy again about his
literary schemes. He tried to sell copies of his translation of Klinger,
which he took from the publisher in lieu of payment for the work.
While with Phillips in London, he had projected a volume of poetical
translations of Danish ballads. The plan then came to naught. Now
he printed the book in Norwich by subscription, after a
correspondence with Allan Cunningham about it. Cunningham was
full of admiration for the old songs drawn from the “Kjaempe Viser.”
“Swayne Vonved” was his favourite, and it remained Borrow’s own
pet throughout life. Five hundred copies of the “Romantic Ballads”
were printed, of which 200 were subscribed for. These, at ten and
sixpence a copy, paid all the expenses of the issue. There was an
arrangement under which the London publisher, John Taylor, took
the rest and placed his imprint on the title-page. Cunningham gave
the young poet a great deal of good advice about promoting the
interests of the book. He neglected it, with characteristic self-
sufficiency. He had published ballads, and if the great public did not
share Mopsa’s affection for ballads in print, the nineteenth-century
Autolycus could not help it, and would be content with what he
could get out of the local subscribers in Norwich.
In 1826 he was in London, and in correspondence with Benjamin
Haydon about sitting for a figure in one of his pictures—possibly the
“Mock Election.” In the course of the correspondence Borrow speaks
of proceeding presently to the South of France. This is the first hint
of those brief travels on the Continent which became magnified by
the pervading haze into world-wide wanderings. “Were you ever at
Kiachta?” Bowring asked him in a letter some years later. He was
never within some thousands of miles of Kiachta. In 1826 he
probably did go tramping through part of Europe, but he did not
reach the East, as some confused references in the books suggest.
The tale of Murtagh in “The Romany Rye” may incorporate some of
his adventures. At any rate, that alluring narrative was certainly not
given to Borrow in the year 1825 at Horncastle Fair. There is clear
evidence of that in the fact that a portion of it was picked up nearly
thirty years later in very different circumstances.
The real itinerary of the tour of 1826 is probably by way of Paris on
foot to Bayonne; across the Pyrenees into Spain; Pamplona, the
Riviera, Italy, Genoa, and thence home by ship. Slight traces can be
found of such a journey. There is the lightly-touched meeting with
Vidocq in Paris. That delectable rascal’s career always had a strong
fascination for Borrow, whose appetite for picturesque blackguards
was greedy. Vidocq at this time was fifty years of age. A quarter of
a century of adventure as a showman, a soldier, a galley-slave, and
a highwayman had terminated in 1812 with his appointment to the
head of a detective office in Paris, on the principle of setting a thief
to catch a thief. By the year 1825 the authorities were persuaded
that the principle was unworkable, and dismissal ended Vidocq’s
career of corruption and swindling. If Borrow met him in Paris the
next year, therefore, he found his hero a free lance. The Mémoires
of M. Vidocq, which appeared in 1828, and are probably at least as
trustworthy as Baron Munchausen, were among Borrow’s favourite
reading; his relish for literature, embloomed with the flowers of
crime and perfumed by the breath of criminals, had been cultivated
by the compilation of the “Celebrated Trials,” and it never left him.
Vidocq and Peyrecourt loom large in passages of his works; whether
they made so great a figure in his actual experiences in France is
another question. He appears to have met Baron Taylor at Bayonne,
and naturally found in the “picturesque and romantic” voyager a
congenial companion. From these lofty associations the descent on
the other side of the Pyrenees to Quesada [72] and his “Army of
Faith,” the gang of frontiersmen who were helping themselves freely
in the name of the Church, was sudden and severe. But Borrow
seems to have fallen even further, for there is a dim suggestion of
his imprisonment at Pamplona, of his emergence from gaol in a state
of beggary, and his succour at the hands of a party of gypsies whose
patteran he followed in the mountains. He tramped eastwards,
ultimately brought up at Genoa, penniless, and was assisted by
some person or persons unknown to get ship for England.
This is as far as Dr. Knapp has been able to trace the elusory course
of the Wandering Jew of Literature. The theory that he acted as the
travelling commissioner of a London newspaper finds no support. By
1827 he was back in Norwich, keeping his mother’s small household
accounts, visiting the Tombland Fair to inspect “Marshland Shales,”
the glorious chieftain of all the equine race, grubbing for booksellers,
writing articles for newspapers. It was a mean and anxious way of
life, abominable to Borrow, who hated poverty and was ashamed of
it. Therein may be sought the real reason why he “veiled” these
years of his life. His next appearance in the literary arena is in the
distinguished company of Dr. John Bowring.
The Bowring episode in Borrow’s life is one of its most remarkable
and least explicable features. Bowring seems to have been a good
friend to Borrow for many years, to have engaged with him in
literary collaboration, and to have exerted himself in various
directions on his behalf. His reward, so far as Borrow’s works go, is
a scurrilous sketch of himself in “Lavengro,” a long denunciation in
the Appendix to “The Romany Rye,” and the bitter hatred of a man
who knew how to hate as fiercely as he could love intensely. The
whole story of their severance is obscure, but there can be little
doubt that Borrow was entirely in the wrong, that the charges he
made against Bowring of treachery and falsehood were baseless,
and that of many people pilloried in Borrow’s books Bowring was
among the least deserving such scurvy treatment. We have
observed already the circumstances of the first meeting between
Borrow and Bowring at Taylor’s house in Norwich. We shall see that
Bowring came to his rescue when he was in the sorest straits, and
was, in fact, doing much to help him during part of the “veiled
period.”
It has been the writer’s fortune to secure [73] a series of letters from
Borrow to Bowring, which throw much light upon his schemes and
modes of life in the last three of those mysterious years between his
return from the Continent and his engagement by the Bible Society.
He did not remain long in Norwich. In 1829 he was in London,
residing at No. 17, Great Russell Street, Bloomsbury, and deeply
employed about certain translations of Scandinavian poetry which
were to form the basis of a new book on more elaborate lines than
those of the “Romantic Ballads.” Bowring and Borrow had a plan for
issuing in collaboration a collection of English versions, with
interpretations, of those Northern poets whom a purblind public, not
yet obsessed by the Scandinavian spirit in poetry and music,
resolutely disregarded and despised. This was the “literary project”
of which the world heard so much in the Appendix. The
arrangements went so far that a prospectus of the work was put
out. The title proposed was “The Songs of Scandinavia,” and the
collection was to be published in two volumes octavo. The project
remained a project, and the niche left by expectant librarians for the
two octavo volumes was never filled. But in connection with the
negotiations and arrangements between Borrow and Bowring a
correspondence occurred which is full of interest and contains one or
two characteristic bits of Borrovian humour. Incidentally, the letters,
if taken in sequence, and read together with another one of the year
1842, show that, up to a time not far ante-dating the publication of
“The Romany Rye,” with its gross attack on Bowring, the two men
were on the best possible terms. Indeed, in 1842 Borrow speaks of
his old collaborator as “my oldest, I may say my only, friend.” [75]
It were greatly to be wished that the sordid dispute with Bowring
might be numbered among the delenda of Borrow’s history, but
some mention of it will be necessary. Unhappily, no satisfactory
explanation can be given which is at all flattering to Borrow. For
these letters prove conclusively that he introduced into “Lavengro”
and its sequel opinions about Bowring which he certainly did not
hold at the time of which he was writing.
In 1829 their Scandinavian scheme was in the tideway. They had
written and they had met for the discussion of their plans; Borrow
had done a great deal of translation. He was exceedingly anxious
that at any rate the first volume should appear at once; for, as he
said in a letter written on the last day of the year, he was “terribly
afraid of being forestalled in the Kiampe Viser by some of those
Scotch blackguards, who affect to translate from all languages, of
which they are fully as ignorant as Lockhart is of Spanish.” The
italicised passage is underlined in Borrow’s letter; it is a curious
foretaste of some of the choicer invective which he afterwards
bestowed on Scott and the Scots, and of his disagreement with
Lockhart. The preparations were hurried on with a view to the
appearance of the first part of the book in February. The drafting of
the prospectus was left to Borrow, and on January 8th (1830) he
sent a copy to Bowring for his inspection, inviting “the correction of
your master-hand.” He had, he said, “endeavoured to frame a
Danish style,” but was not sure whether he had succeeded. “Alter, I
pray you,” he exclaimed, “whatever false logic has crept into it, find
a remedy for its incoherencies, and render it fit for its intended
purpose.” There follows a delightful touch of egotism. He has, he
explains, had a rising headache for two days, which has “almost”
prevented him from doing anything. But, he adds with fine
nonchalance, “I sat down this morning and translated a hundred
lines of the ‘May Day’”—as though a hundred lines of English verse
were a trifle which he threw off without effort, malgré his “rising
headache.”
Bowring examined the prospectus, made what revisions he thought
necessary, and sent it back.
“I approve of the prospectus in every respect,” wrote Borrow
(January 14th). “It is businesslike, and there is nothing flashy in it.
I do not wish to suggest one alteration.” He goes on to describe the
energy with which he is working, and speaks of having rendered
four hundred lines in one day! The last paragraph of this letter
displays Borrow in a different attitude towards reviews and reviewing
from that which he adopted in after years. “When you see the
foreign editor,” he tells Bowring,

“I should feel much obliged if you would speak to him about my

reviewing Tegnér, and inquire whether a good article on Welsh
poetry would be received. I have the advantage of not being a
Welshman. I would speak the truth, and would give translations
from some of the best Welsh poetry; and I really believe that
my translations would not be the worst that have been made
from the Welsh tongue.”

But this condition of things, in which the romantic ferment caused by

Steffens and Oehlenschläger in Denmark was to be reproduced in
England by Borrow’s translations, did not last long. Difficulties arose
in connection with the publication of the proposed book, and the
enthusiasm paled as the year progressed. The two volumes receded
from view; the twin mountain in labour finally brought forth a review
article of some forty pages. This was despatched in the summer to
the Foreign Quarterly Review, was held back for twelve months, and
appeared at last in the number for June, 1831. In this Bowring
wrote in lively style on Danish and Norwegian literature, and Borrow
supplied sixteen specimens of verse.
In the meantime, Bowring was doing what he could to assist his
protégé to some profitable employment. He sent him an ancient
manuscript which Grundtvig, the Danish poet, wanted to have
transcribed. Borrow said (June 7th) the task would not be overpaid
at £49, but as he was “doing nothing particular” at the time, and
might learn something from it, he would do it for £20. Bowring also
exerted his influence to get him work in the magazines. During the
summer of 1830, Borrow flitted from Great Russell Street to No. 7,
Museum Street, and in the autumn, went to Norwich for a holiday.
In the letter (September 14th) in which he tells Bowring of his
proposal to leave London for Norwich, we get the first hint of a
project which now and then flashed through his mind for a year or
two—that of entering the military service: “I have thought of
attempting to get into the French service, as I should like
prodigiously to serve under Clausel in the next Bedouin campaign.”
This remained a thought, though, as we shall see, other plans of the
same character went a little further. In the same letter he
complained that he was very unwell, but traced his malady to ennui
and unsettled prospects, and hoped that cold bathing in October and
November would prove of some service to him. There is no
reference in this correspondence to one task which he himself
asserts he achieved in 1830. That was the translation of Elis Wyn.
At the instance of “a little bookseller of my acquaintance” in
Smithfield, he rendered from the Welsh Wyn’s, “Visions of the
Sleeping Bard.” This was the nearest approach he made to the
promise of literary success; but even here his malign fate dogged
him. When the little bookseller saw the translation, he begged off
the bargain on the plea that “the terrible descriptions of vice and
torment would frighten the genteel part of the English public out of
their wits. . . . Myn Diawl! I had no idea till I read him in English
that Elis Wyn had been such a terrible fellow!” The sly dig at the
“genteel” public may be reasonably attributed to the bookmaker
rather than to the bookseller.
Before he departed from London, Borrow, returning some books to
Bowring, utters (September 17th) one of those ejaculations on
public affairs which he subsequently inserted as tags to many of his
letters: “More Revolutions, I see. The King of Saxony has run away,
and the Kent peasantry are burning stacks and houses. Where will
all this end?”
A dozen plans for carving a way to undying fame and modest
fortune, all equally futile, were built up and fell down about this
time. Apparently Borrow could not rid himself of the delusion that a
hungry world was waiting to devour the beauties of the Gaelic Bards,
if only they were served up in a suitable form for general
consumption. He launched at the devoted heads of the Highland
Society of London a scheme under which the Society was to employ
(and pay) him for two years in translating the Gaelic Bards into
English verse. The scheme left the Highland Society as cold as the
Bards would have left the reading world. He turned his artillery
upon the British Museum. The Codex Exoniensis was to be copied;
he applied for the work, but without success. It was done in 1831
by one of the regular officials of the Museum. Discouraged but not
dismayed, he sought other employment in Bloomsbury, and asked
Bowring to put in a word for him. The Doctor pointed out that in his
position it was necessary to go about such a matter with discretion.
It would not do for him to originate an application, but if the
authorities of the Museum could be induced to seek his opinion, he
would give Borrow such a character as would “take you to the top of
Hecla itself. You have claims, strong ones, and I should rejoice to
see you niched in the British Museum.” But this design failed like the
rest. In a letter to Bowring he described himself, with melancholy
eloquence, as “drifting upon the sea of the world, and likely to be
so.” To Borrow there was “no fiercer hell than failure”; but the
inferno was of his own creation. His greatest failure was the failure
to realise that there was no sort of demand for the work he insisted
on doing, and that its intrinsic value was far below the standard at
which he placed it.
Compelled thus to abandon his literary ambitions for the present, he
turned his efforts in another direction. He began the pursuit of a
shimmering phantom over which, in the course of his life, he
contrived to waste a great deal of valuable time. Upon what he
based the idea does not appear, but Borrow seems to have imagined
that he had some claim to official employment abroad. It did not
much matter whether the work was made for him by the British
Government or by a foreign State, so long as he should be given the
opportunity of displaying his philological prowess in foreign parts.
After the appearance of the joint article in the Foreign Quarterly, as
Bowring seemed to be able to do nothing for him at the British
Museum, Borrow asked him to see what he could do towards getting
him a post under the Belgian Government. Bowring made the
application, but without success; the Belgians were not at the
moment in need of any English assistance, however talented.
Borrow keenly recognised his friend’s diligence in the matter, and
turned his heaviest artillery on the Ministry at Brussels, who were so
obstinately blind to the advantages of having Mr. George Borrow in
their service. They did not seem, he said in a letter to Bowring
written from Willow Lane, Norwich, and dated September 11th,
1831, either to know or to care for the opinion of the great Cyrus,
whose advice to his captains he quoted from Xenophon: “Take no
heed from what countries ye fill up your ranks, but seek recruits as
ye do horses, not those particularly who are of your own country,
but those of merit.” Belgium, having failed to appreciate the worth
of George Borrow, at once became the most contemptible nation on
earth:

“The Belgians will only have such recruits as are born in

Belgium, and when we consider the heroic manner in which the
native Belgian army defended the person of their new sovereign
in the last conflict with the Dutch, can we blame them for their
determination? It is rather singular, however, that, resolved as
they are to be served only by themselves, they should have sent
for 50,000 Frenchmen to clear their country of a handful of
Hollanders, who have generally been considered the most
 unwarlike people in Europe, and who, if they had had fair play
given them, would long ere this time have replanted the Orange
flag on the towers of Brussels and made the Belgians what they
deserved to be—hewers of wood and drawers of water.”

This sardonic outburst is one of the earliest samples of the polemical

style which Borrow was to develop so strongly in later years.
As he could neither go to fight Bedouins under Clausel nor enter the
Belgian service in Europe, it appears to have occurred to his friend
Bowring that he might care to follow in his father’s footsteps, and
that the British service might suit him at a pinch. If Borrow would
like to purchase a commission, Bowring offered to introduce his
name to the War Secretary. Borrow replied that his name had been
down for several years for the purchase of a commission, but he had
never had sufficient interest to procure an appointment. He would
not now mind serving in the militia if they were to be embodied for
service in Ireland (“that unhappy country”), but he wished to leave
the question open for a few months in order to see whether
something more promising turned up. If he had not secured
employment within two or three months, he would then ask Bowring
to redeem his promise in the matter of the War Secretary, and to
recommend him to a corps in one of the Eastern colonies on the
plea that he was “well grounded in Arabic” and had some talent for
languages:

“I flatter myself that I could do a great deal in the East,

provided I could once get there, either in a civil or military
capacity. There is much talk at present about translating
European books in the two great languages, the Arabic and
Persian. Now, I believe that with my enthusiasm for these
tongues I could, if resident in the East, become in a year or two
better acquainted with them than any European has been yet,
and more capable of executing such a task. . . .”
This letter concluded with a postscript in which he requested that his
best remembrances might be presented to Mrs. Bowring and to
Edgar, their son; and, he added, “tell them they will both be starved.

“There is now a report in the street that twelve corn-stacks are

blazing within twenty miles of this place. I have lately been
wandering about Norfolk, and I am sorry to say that the minds
of the peasantry are in a horrible state of excitement. I have
repeatedly heard men and women in the harvest-field swear
that not a grain of the corn they were cutting should be eaten,
and that they would as lieve be hanged as live. I am afraid all
this will end in a famine and a rustic war.”

Reform staved off the “rustic war,” and other things intervened to
prevent Borrow from carrying out his half-formed intention of
becoming a military man.
 CHAPTER V
IN FOREIGN PARTS

“Romance brought up” the year 1832. It was a year full of events
with an important bearing on the course of Borrow’s life. In the first
place, he became acquainted with the Skeppers, of Oulton Hall, near
Lowestoft. The introduction to this family issued in a friendship with
Mr. Skepper’s sister, the widow of a young naval officer named
Clarke. In Mrs. Clarke, a woman somewhat older than himself—she
was thirty-six and he was twenty-nine—he met the woman who was
to bring into his life its fairest influence and its rarest happiness. But
the story of this romance must be postponed for a few pages in
order to the relation of a sequence of affairs without which it cannot
be understood. They resulted from sundry conversations about
Borrow—between the Skeppers and the Rev. Francis Cunningham,
Rector of Pakefield, and in turn between Cunningham and Joseph
Gurney, his brother-in-law, from whose meadows at Earlham George
had fished in boyhood.
Both Cunningham and Gurney were interested in the work of the
Bible Society, and between them the idea was hatched of employing
Borrow’s philological learning in its behalf. The Society happened at
the moment to be looking for a man to superintend the printing of
the New Testament in Manchu. There were many negotiations, and
ultimately the engagement was consummated which made Borrow’s
modest fortune.
To go to St. Petersburg on this business of the Bible Society’s was an
adventure after Borrow’s own heart. He had passed through some
exceedingly stormy waters, and in this employment he found a
secure and congenial harbour. He could well afford to regard lightly
the critical attitude of certain people in Norwich, who did not forget
to recall the episode of “godless Billy Taylor.” Their temper was
reflected in the letter of Harriet Martineau referring to Borrow as a
“polyglot gentleman,” and remarking that his appearance as “a
devout agent of the Bible Society” evoked “one shout of laughter
from all who remembered the old Norwich days.” Borrow did not
like their laughter, and he did not forgive their contempt. But for the
time he was too busy with the actualities of his new situation to
trouble about them, and too elated with his suddenly brightened
prospects to be cast down by the jeers of the scornful.
He was going a journey into a far country, and he was going on a
more or less philological errand. His task was to undertake the
production in the Russian capital of the Manchu version of the
Sacred Books made by Lipotsof. Invited to London to see the
officials of the Society, he set off in high spirits—and on foot. The
long road stretched for a hundred and twelve miles between
Norwich and London—that road which some ten years before he had
travelled by coach with the little green box of poetical translations.
He now tramped it in 27½ hours, and his expenses en route
amounted to fivepence halfpenny! This feat was one of his favourite
boasts. It was, in its way, a remarkable achievement. Few big,
healthy young men would care to undertake such long-sustained
exertion on a pint of ale, half a pint of milk, a roll of bread, and two
apples. But such is Borrow’s tale of his commissariat arrangements
on this expedition.
The Society desired him to learn the Manchu language before he set
out for Russia. They gave him six months for the purpose. Even for
a meteoric philologist like Borrow, who swallowed a language by
memorising its dictionary, six months meant short commons. He
could not possibly acquire more than a nodding acquaintance with
that most difficult of the tongues of Babel. However, he set about
his task with zeal.
There is one amusing passage in the correspondence between him
and the Secretary of the Bible Society. Observe the true Borrovian
spirit asserting itself in the letter where he expresses pleasure at the
prospect of “becoming useful to the Deity, to man, and to myself.”
Observe the solemn admonition of the good secretary, when he
perceived that a sense of human frailty was not one of Borrow’s
most striking characteristics: “Doubtless you mean the prospect of
glorifying God.” Thereafter, the Borrovian spirit was subdued (in
correspondence) to the proper standard of orthodoxy.
At the end of June, 1833, he set sail for St. Petersburg, by way of
Hamburg, and was highly delighted with the Russian capital. He
made his way into the acquaintanceship of a number of literary
people, in whose society he found congenial entertainment. Among
them he speedily established for himself quite a reputation. It was
here that he began his long friendship with Hasfeldt, which produced
a prolific correspondence. Hasfeldt was a Dane attached to the
Russian Government, and a linguist of attainments, who added to his
income by the teaching of European languages. He conceived a
remarkable fondness for “tall George,” as he called him; the affection
was returned as fully as Borrow could return a friendship, and that
was in much higher measure than many estimates of him suggest.
He met Russian scholars, and found many opportunities for
extending his philological studies in the direction of the Oriental
languages.
His work on the Chinese version was hard and long. He had to use
German printers, who did not always feel for the task the
enthusiasm which Borrow expected everybody to throw into
anything in which he himself was concerned. They had to be bribed
with vodka, and other things, in order that progress might be
secured. The Bible Society presumably swallowed the vodka in their
delight at the energy Borrow displayed, and they passed a resolution
to pay him any expenses to which he might be put in the execution
of the commission. He had to furbish up an old fount of type in the
Chinese character, that had been lying rusting in a cellar for many
years, and to get everything in order himself, because, of course, it
was impossible to obtain compositors who knew anything of the
Manchu. He even turned printer. So keen was the zest with which
he entered into the work that he submitted a proposal to the Society
to undertake the distribution of the books when they were printed,
going overland to China, and looking in upon the Tartars on the
way! Without doubt he would have done it but for the fact that the
Russian Government refused to grant him a passport for the
purpose. It is characteristic of Borrow that years afterwards he said,
and doubtless thought, that he had been overland to China.
The work of printing done, he paid a hurried visit to Moscow,
gathering impressions for the description of the Kremlin to be found
in “The Bible in Spain,” and on September 9th, 1835, he left St.
Petersburg for England, having spent the previous night in a solemn
leave-taking of Hasfeldt. While in St. Petersburg hard at work, and
feeling run down, he had “the Horrors” several times, but affected to
have found a cure for it in the shape of strong port wine. It was
during his stay in Russia that the news arrived of the death of his
brother John in Mexico. He had discovered other activities to occupy
him besides the translation of the Testament into Chinese. He
turned homilies of the Church of England into Russian and Manchu,
and did translations of some of the sacred Buddhist books from
Manchu into English. He conceived at the moment no high opinion
of the Buddhist philosophy. “You will be surprised,” he writes to the
Rev. F. Cunningham, “that Satan by such inconsistent trash should
have been able to ensnare the souls of millions!” If that had been
read in the Martineau household there might have been another
“burst of laughter.” It was while he was in St. Petersburg, too, that
he published his “Targum,” a collection of poetic translations from
thirty different languages and dialects. When Pushkin, the poet,
after Borrow’s departure, received a presentation copy of this book,
he expressed his great regret that he had not met the author.
Borrow reached London on the 18th September, and went down to
Norfolk, feeling anxious again about his future, and hoping that the
Bible Society would be able to find some further employment for
him. He was not disappointed. The Society had not yet given up
hope that they might find a way to send him to China, but in the
meantime they resolved to commission him to Portugal. On
November 2nd they passed a resolution that he should be asked to
go to Lisbon and Oporto to inquire about “means and channels for
promoting the circulation of the Holy Scriptures in Portugal.” [91]
Here is the origin of two of his books, of which one was “The Bible in
Spain.” On November 6th he sailed from London, touching at
Falmouth on the 8th, and was at Lisbon on the 13th. He was to
confer with one Wilby about the work; but, Wilby being away,
Borrow consoled himself with the company of Captain Heyland, of
the 35th Foot, whose acquaintance he had made on the voyage.
With him he made several trips, upon one of which he met the
bohémienne landlady of Cintra. During this first expedition to the
Peninsula, he set up relations with the gypsies of Spain, which
provided the germ of the first of his books that attracted anything
like general attention. At Badajoz he encountered a gypsy tribe, by
whom he was detained ten days. In that time he had translated the
Gospel of Saint Luke into the Câlo, or Spanish gypsy language, and
the version was subsequently printed by the Bible Society. One of
the Romany chals, Antonio Lopez, accompanied him most of the way
to Madrid, delaying three days at Merida in a gypsy house. Antonio
finally went off with a gitana. Borrow bought a donkey from the girl,
and rode on the animal’s back as far as Talavera, where he sold it to
a Toledo Jew whom he met on the road. The rest of the journey to
Madrid he did by the diligence, like a common Christian.
By the time of his arrival there, he had formed a definite project of
printing the New Testament in Spanish and in Spain, without
comment or note of any sort. The law would prohibit the circulation
of such a book if it were printed outside and brought into the
country. It was decided to use the current Catholic version, in order
not to excite any more prejudices than could be helped, and to sell
cheaply, and thus to spread the book among people who had never
seen it before. This was a time in Spain of constant political
excitement, chronic Ministerial change, and periodical revolution;
and Borrow had much trouble in getting official recognition for the
enterprise, without which he might as well have left it alone. But
the way was smoothed for him by Sir George Villiers, the British
Minister, and at the end of twelve months he returned to England
with an active campaign mapped out in his mind, for which he soon
obtained the approval of the Society. In a letter to his mother about
this, he remarked that his “ordination” would be put off till his
return. This is the first and the last that we hear of any proposal to
enter the Church.
On his way out to Spain the second time, he happened across Santa
Coloma, the Carlist, who is frequently met with hereafter in his
Spanish adventures. “The Bible in Spain” relates very closely the
events of the next two years—his wanderings and escapes, his
enterprise in Madrid, where he set up a bookselling shop, his
imprisonment for insulting the Government and the Catholic Church
—an offence of which he was quite innocent, for such was not his
method at the time. The trouble was brought on him by an
evangelical firebrand, named Lieutenant Graydon, who led Borrow
into one of his scrapes with the Peninsular powers by claiming to be
associated with him in the work of the Bible Society. Borrow’s
imprisonment resulted in a declaration by him in the Spanish Press,
directed against Graydon. He said that neither himself nor the Bible
Society was actuated by any enmity against either the Government
or the Catholic clergy of Spain, and concluded by avowing himself
the sole agent of the Society in the Peninsula. Out of this grew an
estrangement between Borrow and the Society. It happened that
Graydon was one of the pets of Mr. Brandram, joint secretary of the
Society, and was actually regarded as one of their agents, though he
received no pay, being the holder of a Government pension. He was
an enthusiastic evangelist, who seems to have lacked nothing save
discretion, but manifested this defect by fierce attacks upon the
Catholic faith in its stronghold, instead of contenting himself with
prosecuting the primary work of the Society, which was the
distribution of the unadulterated Scriptures. In the event, Graydon
was withdrawn from Spain, but it was expressly stated that this step
was taken only in the interests of his own safety, and that the
Society would pass no judgment on the merits of the dispute
between him and Borrow until Graydon had returned to England and
had an opportunity of vindicating himself. Borrow at the same time
was ordered to cease issuing his advertisement. It is difficult to
judge a man like Graydon. His good faith in all he did can hardly be
doubted, but there is no question that the result of his ill-timed
action was to put an end to the work of the Society and the
circulation of the Bible in Spain for many years.
The relations between Earl Street and Borrow grew more strained,
and very soon he had practically a command to come to London. He
packed up and returned, but such was the force of his character that
he fascinated Earl Street into sending him to Spain a third time. He
was only home a month or two, and got back to the Peninsula on
the last day of 1838. But the mission was not of much further use,
for there had been another change of Ministry in the meanwhile, and
Borrow and the Society were again out of official favour.
He proceeded to Seville, settling there for a purpose, as we shall
presently see. In the sunlit southern city he was encountered by an
English traveller, who has left a most entertaining account of him.
This was Lieutenant-Colonel Elers Napier, in whose “Excursions along
the Shores of the Mediterranean” appears the remarkable figure of a
Man of Mystery, who is easily identified as Don Jorge—though
apparently Napier never learned who he was. Borrow, six feet three,
with piercing black eyes, snowy head, and swarthy, hairless face,
made a profound impression on his new friend—and we may be sure
that he omitted nothing that would deepen it. He showed off all his
best points and maintained a rigid silence upon the question of his
identity, so that in Napier’s recollections he assumes almost
supernatural proportions, and is described throughout as “The
Unknown.” He revealed all his miscellaneous acquaintance with
languages, Occidental and Oriental. He conversed with the Colonel
in Spanish, in Latin, in French (“the purest Parisian accent”), in
Italian. He spoke English perfectly, but did not appear to be an
Englishman. He was even as conversant with Hindu as the Anglo-
Indian himself; he seemed, Napier says, to know everything and
everybody, but was apparently known to nobody himself. His almost
magic power over the gypsies, his familiarity with their patois and
their customs, the way in which they almost worshipped him when
he took Napier by night for a visit to one of their weird
encampments, added to the marvel.
But the real significance of the visit to Seville is not to be sought in
the archives of the Bible Society or in the jottings of Colonel Napier.
Borrow’s friendship with Mrs. Clarke, of Oulton, arose in the fashion
already mentioned. His long absences from England did not impair
it, and in 1838 it developed in peculiar circumstances, which were
the subject from time to time of scandal utterly unfounded, and of
gossip more or less impertinent and irrelevant. Whether Borrow,
during the years from 1832 to 1838 nurtured dreams of any relation
closer than friendship it is hardly possible to determine. He was not
“a marrying man,” and probably the sober little romance that ended
in their wedding was a thing of sudden growth. That theory is
encouraged by a passage in his correspondence as late as 1838,
when he told his friend Usóz that it was better to suffer the halter
than the yoke, and expressed his conviction that bachelordom was
the better kingdom for him. But at the end of the same year, during
his stay in England, he visited his friends at Oulton, and found a
state of affairs that doubtless altered his judgment.
The business of Mrs. Clarke, who was the principal heiress of the
Oulton Hall estate, was in a highly complicated condition. She had
none but professional advisers, save Borrow, and leant with obvious
relief upon his friendship to guide her through a puzzling maze of
family disputes. It would be wearisome to attempt to follow the
controversies about the disposition of the property. They finally
involved Chancery proceedings, and Dr. Knapp asserts that Mrs.
Clarke’s solicitors advised her that it would be well for her to
disappear for a time. The reason for this counsel is obscure, but the
fact that it was followed is important. Mrs. Clarke consulted Borrow
about it, with the result that her evanishment took the form of a
journey to Spain, accompanied by her daughter Henrietta. The fact
created an amazing quantity of idle speculation and not too
generous suggestion. The plan was arranged in March, 1839.
Borrow was then in Madrid, and immediately posted off to Seville to
prepare a house for the reception of the two ladies, having given
them some useful hints, drawn from his long experience of Spain, as
to the household gods they ought to bring with them. They arrived
in June, and were installed at No. 7, Plazuela de la Pila Seca, which
Borrow had modestly furnished and was himself occupying.
The little wind of scandal that played about this arrangement will not
disturb the equanimity of those who know their Borrow. The
ménage was unquestionably a little difficult to explain to the
Spaniards to whom explanation was necessary, and to this difficulty
Dr. Knapp attributes Borrow’s expedition to Tangier at the end of
August. This was the trip with which “The Bible in Spain” suddenly
closed down in the approved Borrovian style. The scandal was of
short duration and small effect. But in after years other suggestions
were made, including the highly improbable and offensive one that
Mrs. Clarke was at this time pursuing Borrow with the object of
matrimony, and “travelled over half Europe in search of him.”
Another friendly theory advanced was that Borrow’s proceedings
were governed by mercenary motives, and that he married Mrs.
Clarke because she had an income of three or four hundred a year.
Meanwhile, the quarrel with the Bible Society was dragging its slow
length along. The correspondence is confused and in general
uninteresting, except that it shows how Borrow’s attitude towards
Earl Street had altered since the time when he climbed down before
the protests of the good secretary in the first days of their
association. He was on his feet now.
He felt surer of his ground than when he was at his wits’ end for
employment and subsistence. Consequently his native impatience of
restraint came out. The Bible Society never gauged their man. In
one despatch to Earl Street, Borrow had said of a certain enterprise
that “his usual good fortune accompanied them.” “This,” replied Mr.
Brandram, “is a mode of speaking to which we are not well
accustomed; it savours, some of our friends would say, a little of the
profane. . . . Pious expressions may be thrust into letters ad
nauseam, and it is not for that I plead; but is there not a via
media?” The breach grew wider and severance was ordained; it was
consummated very shortly after Borrow’s return to England at the
beginning of the next year.
The visit to Tangier occupied some five or six weeks. Borrow
returned to Seville at the end of September, and set to work
compiling notes and making transcripts for his book on the Gypsies
of Spain. The enterprise was assisted by diligent friends, such as
Bailly, [99a] Usóz, [99b] and Gayangos. [99c] The fruits of their curious
researches among dusty and neglected bookshelves may be seen in
the long translations from archaic Spanish authors in “The Zincali.”
It was a Spaniard who invented the epigram on the virtues of old
wood to burn, old wine to drink, old friends to trust, and old books
to read. But we may be excused for excluding from the category of
books which have the bouquet of old crusted port the discourses of
Dr. Sancho de Moncada and others to which Borrow has treated us
so liberally.
He spared time from these labours and from the task of settling up
with the Bible Society to pay considerable attention to Mrs. Clarke
and “Hen”—the affectionate diminutive given to her daughter
Henrietta. The widow had found Seville, as Borrow promised her it
should be, “a most agreeable retreat,” where “the growls of her
enemies could scarcely reach her.” The ladies enjoyed to the full the
startling change from the life of the English fens to that of the sunny
and many-hued Spanish city. They realised his prophecy that it
would be a delicious existence where, “during the summer and
autumn, the people reside in their courtyards, over which an awning
is hung. A very delicious existence it is—a species of dream of
sunshine and shade, of falling water and flowers.” And, incidentally,
of course, a very fit setting for such love-making as came to be
done: the weather is always fine when people are courting, as a
modern sage has remarked. Not much more than a month after his
return from Morocco, Borrow had proposed marriage to Mrs. Clarke,
and had been accepted. The arrangement was to a certain extent a
“convenient” one for both parties. With little prospect of further
employment by the Bible Society, and only a precarious hold on any
profitable literary work, Borrow had no glowing future before him.
Mrs. Clarke felt the need of a man to manage affairs for her at
Oulton. Still, there is ample evidence that this was a fortuitous
concourse of circumstances, and that it had little to do with the
marriage. The warm English friendship had become more intimate
as the years passed, and there was nothing more natural than this
sequel when they were thrown together in the “delightful existence”
in which she hid from her “enemies” at Seville.
Having decided to cross the Rubicon, Borrow determined that the
sooner it was done the better. There was to be no “sweet, reluctant,
amorous delay.” He began at once to make preparations for the
return to England in order that they might be married in their own
country. One of the first steps to be taken to this end was to
procure his passport from the Alcalde. Why this official disapproved
of Borrow cannot be affirmed. As a son of the True Church he may
have conceived a prejudice against the Protestant colporteur; he
may have been infected by the “spy” mania; he may have been
merely anxious to display his own importance. At any rate, he
resolved to give the Ingles rubio as much trouble as possible to
remove himself and his party out of Spain. He raised questions
about the validity of Borrow’s papers, refused the passport, and
would not be pacified by the offer of fees, “lawful or unlawful,” to
quote Borrow, who sent to him apparently under the impression that
authority, though a stubborn bear, might be led by the nose with
gold, as the clown said to Autolycus. When Don Jorge himself went
to the office to inquire into the matter, he was told to go away.
Instead he continued to investigate the motives of the Alcalde, who
thereupon threatened to carry him to prison. Borrow dared him to
do so—and he did it. This was his third acquaintance with the inside
of a Spanish gaol. He sent a reassuring note to Mrs. Clarke, and had
a message taken to the British Consul. Colonel Napier had noticed
earlier in the year that the police kept sharp eyes on Borrow, and
attributed it to the suspicion that he was (of all things in the world!)
a Russian spy. There was clearly something in the suggestion that
he was under espionage, for while he was in prison his house was
searched for papers. Nothing “compromising” being found, he was
released the next night.
His indignation at this outrage reached white heat, and did not die
down for months. His insistence upon redress detained Borrow in
the country much longer than he had proposed to stop. Once
having got his knife into Spanish officialdom, he twisted it round till
he had gouged out his pound of flesh. And even then, after he had
returned to England, and the knife was no longer available, Spanish
officialdom received very severe treatment from that even more
terrible weapon, his pen. From Seville he set working all the
diplomatic machinery that an injured Briton could influence; he went
to Madrid on the business; he wrote incessantly and exhaustively
about it. His return to England and his marriage had to wait until he
had settled accounts with the impertinent Alcalde de Barrio, who had
laid sacrilegious hands upon a subject of her Britannic Majesty—and
that subject George Borrow. While ambassadors and consuls and
State secretaries were busily employed in official correspondence on
his behalf, he proceeded with the work on the “Gypsies,” and did not
get away from Spain till April, 1840.
The embarkation of the colporteur and his party upon the Royal
Adelaide steamer at Cadiz was an impressive ceremony. Borrow was
taking a long farewell of Spain, and he was not going home without
souvenirs of his residence there. In the previous year he had
purchased the Arab horse celebrated in his books as “Sidi Habismilk”
(being interpreted, “My Lord Mustard”). The retinue at Cadiz
included not only Mrs. Clarke and Henrietta, but also Sidi Habismilk
and Hayim ben Attar, “the Jew of Fez,” Borrow’s servant. [103] They
touched at Lisbon, where General Cordova came on board—not on
business of State, but in search of a consignment of cigars that had
been sent to him in the care of the captain. Borrow wrote an
amusing sketch of the General and two Secretaries of Legation
stowing Havana cigars in their pockets “with all the eagerness of
contrabandista.” [104] The vessel arrived in the port of London on
April 16th, and the party put up at the Spread Eagle, in Gracechurch
Street. As soon as the licence could be obtained, the marriage of
“George Henry Borrow, bachelor,” with “Mary Clarke, widow,” was
celebrated at St. Peter’s Church, Cornhill, and witnessed by John
Pilgrim, of Norwich (the bride’s solicitor) and by her daughter
Henrietta. The wedding day was April 23rd.
There remained a very little business to do in London. He had an
interview with the General Purposes Committee of the Bible Society,
received a letter from Mr. Brandram, saying that there was no sphere
open “to which your services in connection with our Society can be
transferred,” and quickly terminated his relations with Earl Street. In
spite of the little differences that had arisen, there was a generous
reference to Borrow in the Report of the Society for 1840. He was
said to have succeeded “by almost incredible pains, and at no small
cost and hazard,” in his last mission to Spain, and to have assisted in
circulating during five years nearly fourteen thousand copies of the
Scriptures. Thus the Bible Society and Don Jorge said good-bye.
At the beginning of May, Mr. and Mrs. Borrow and Miss Clarke went
down to Oulton. The Hall having been let to a farmer, they took up
their residence in a little house on the margin of the Broad, known
as Oulton Cottage.
 CHAPTER VI
THE SUMMER HOUSE AT OULTON

When Borrow went to Oulton he was thirty-seven. The comforts of

the domesticity to which he settled down were sweet, but its joys
were of a very different quality from those golden matrimonial
projects of which he had dreamed in Mumper’s Dingle. He was
older, sadder, if not much wiser. He had modified the scale of his
ambitions. He was bent upon the acquisition of such fame as he
could attract through the avenue of literature, and not disdainful of
what local celebrity might come his way. But though he was not of
the temperament to apostrophise with Cowper—

“Domestic happiness! Thou only bliss

Of Paradise that has survived the Fall!”

there is everything in favour of the supposition that, in marrying Mrs.

Clarke, Borrow wrought better for himself than a man of his
temperament usually has an actuarial expectation of doing in
matrimony. Moreover, he did infinitely better than a great number of
literary persons who have taken the plunge in similar circumstances.
There was no such tragedy about his marriage as befell his friend
and neighbour Edward FitzGerald; indeed, there was no tragedy at
all. Its absence is due to Mrs. Borrow’s remarkable personality, her
wifely qualities, unfailing devotion to him in all his fads and moods
and whimsies. She was a perfect “helpmeet”; she provided him with
a buffer to absorb some of the shocks of outrageous fortune; she
was a patient amanuensis and an indefatigable secretary.
The picture one constructs of his wife from the materials—slight
enough—that Borrow himself gives, and from the correspondence
extant, is that of the “flower of wifely patience”—a woman in whom
tact has been developed to such a degree as to become a kind of
extra sense. She was married to one of the queerest specimens of
mankind that Nature ever evolved; yet she secured in their union
happiness for both. Her affection for him was true and deep; it was
strong enough even to prevail over idiosyncrasies that might easily
have been fatal to any chance of domestic peace, to say nothing of
marital bliss. She was one of the women to whom “patience hath
such mild composure given” that even Borrow failed to destroy her
equanimity and self-possession. Behind her hero-worship appears
now and then an illuminating gleam of feminine commonsense—just
a shooting ray upon some foible; but whenever it seems likely to
show Borrow in a specifically unfavourable light it is immediately
switched off.
Near the easternmost point of land in England, on the margin of
Oulton Broad, in a spot where the roar of the North Sea could be
heard, was the cottage in which the best of his remaining years
were to be passed. Here he was to prosecute amid the solemn
marshland the eternal search for truth and happiness, and to find
that the pursuit was even more difficult for him than for the majority
of mankind. The house contained few rooms, but sufficient for the
requirements of the little family, and its quietude and isolation were
special recommendations to Borrow in the particular mood in which
he then found himself. The scenery was of a character for which he
had strong affection, and the place itself was linked with one or two
of the powerful emotions of his youth. The Broad stretched away
from the end of his garden, and he overlooked it from the summer-
house he built as a study. Behind the house: and almost
surrounding it, were plantations of pine trees. For the rest, only an
occasional tower or windmill broke the level horizon. The scene is
different, more varied, and much fuller of life at the present day,
when the virtues of the Broads as pleasure waters and of the
country round as a residential district have been discovered and
exploited. But in certain hours and seasons it is easy to imagine
Oulton as George Borrow knew it.
Miss Elizabeth Harvey has left us a picture of Borrow as the friends
of this period recalled him. [109] In his wooden pavilion “on the very
margin of the water,” she tells us, “he had many strange old books in
various languages. I remember he once put one before me, telling
me to read it. ‘Oh, I can’t,’ I replied. He said, ‘You ought: it’s your
own language.’ It was an old Saxon book. He used to spend a great
deal of his time in this room, writing, translating, and at times
singing strange words in a stentorian voice, while passers-by on the
lake would stop to listen with astonishment and curiosity to the
singular sounds.” A note on his personal appearance, by the same
hand, may help to keep his figure in mind: “He was six feet three, a
splendid man, with handsome hands and feet. He wore neither
whiskers, beard, nor moustache. His features were very handsome,
but his eyes were peculiar, being round and rather small, but very
piercing, and now and then fierce. He would sometimes sing one of
his Romany songs, shake his fist at me, and look quite wild. Then
he would ask, ‘Aren’t you afraid of me?’ ‘No, not at all,’ I would say.
Then he would look just as gentle and kind, and say, ‘God bless you,
I would not hurt a hair of your head.’” Here was he, then, when he
set up author in real earnest, and induced “glorious John” to publish
the first book that resulted from his adventures in foreign parts.
This was “The Zincali; or, An Account of the Gypsies of Spain, with
an Original Collection of their Songs and Poetry, and a Copious
Dictionary of their Language.” Most of the compilation—for such it
is, and a desultory compilation at that—had been made during his
five years in Spain. It was written at odd times, “chiefly in ventas
and posadas, whilst wandering through the country in the arduous
and unthankful task of distributing the Gospel among its children.”
In its published form “The Zincali” was an amalgam of several
schemes that had occurred to the author from time to time during
his Spanish wanderings. He had projected a collection of the
rhymes and proverbial sayings of the gypsies of Spain, inspired
thereto by the material he had gathered at Badajoz and Merida, to
which additions were made some years later at Seville with the
assistance of Juan Antonio Bailly, a French courier with a
considerable acquaintance among the Câlé. He had also proposed a
glossary of Câlo and English, which afterwards resolved itself into a
limited vocabulary of words occurring in the songs and sayings that
he and Bailly had collected. Both these schemes were imperfectly
executed. Borrow’s knowledge of the Spanish-gypsy language was
quite empirical, and Bailly’s collections were either written by
illiterate persons, or taken down from the lips of people who spoke a
corrupted jargon. Borrow and Bailly made a large number of
translations from obscure Spanish authors—and this was the
material from which “The Zincali” was constructed. He eked it out
with a quantity of out-of-the-way information and anecdote acquired
during his association with gypsies in England and Russia, and in the
course of much miscellaneous browsing among books. A more
unscientific process of writing “An Account of the Gypsies of Spain,
etc.,” it would be hard to devise. There were half a hundred works
of more or less utility which he might have consulted, and there is
no evidence that he had seen more than a tithe of that number.
But, pari passu, there is certainly no evidence that if he had seen
them all he would have produced a better book. In fact, here, as in
every other case, his work does not depend for its charm and its
value upon any scientific basis whatever, but upon the idiosyncrasies
of Borrow himself, the mordant style, the quaint observation, the
atmosphere with which he contrives to invest his subject. “The
Zincali” was read at first, as it is read now, not so much for the
accuracy of its history or its philology as for its intrinsic interest as
literature.
Having put together at Oulton these notes, memoranda, rhymes,
translations, descriptions, and scraps of a gypsy vocabulary, Borrow
took the compost to John Murray, who agreed to publish an edition
of 750 copies. The book attracted certain minds attuned to the
Borrovian spirit, and it was admitted to display the supreme virtue of
originality. The voice of Murray, above all, was encouraging, and to
Borrow that was the voice of the “Mæcenas of British literature.” In
spite of occasional difficulties, he held Mr. Murray in unfailing honour,
and was proud to have his work sealed with the cachet of Albemarle
Street. The close association of the Murrays with Richard Ford,
whose “Handbook” was long the classic English work on Spain, had
important results for Borrow. Ford was living in retirement at
Heavitree, near Exeter—the haven where, half a century later,
George Gissing found rest in his last days—and to him the
manuscript of “The Zincali” was sent for critical observation. Ford’s
knowledge of Spain was extensive and peculiar, and he immediately
perceived in Borrow a man after his own heart, who preferred
byways to highways, was full of curious learning, and invariably took
the unconventional outlook. [112] His criticism of the book was what
might have been expected. It took the form of a regret that Borrow
had not given his readers more of himself “instead of the extracts
from those blunder-headed old Spaniards, who knew nothing about
gypsies.” But, on the whole, both Murray and Ford were pleased.
So were the reviewers. As to the public, they bought the work very
slowly. It appeared in April, 1841, and by June only three hundred
copies had been sold. Murray explained this genially by declaring
that the state of politics had shed a blight over literature; no book
was selling, and Borrow’s only shared the fate of the rest.
But before this a new enterprise had been designed. It was to be an
account of Borrow’s personal adventures while engaged in the
circulation of the Scriptures in the Peninsula. The scheme appealed
strongly to Ford, and Murray thought well of it. Ford was “delighted”
to know that Borrow meditated such a work. “The more odd
personal adventures the better, and still more so if dramatic; that is,
giving the exact conversations.” “I have given him much advice,”
said Ford in a letter to Addington, “to avoid Spanish historians and
poetry like prussic acid; to stick to himself, his biography, and queer
adventures.” And Borrow wrote to Ford: “I shall attend to all your
advice. The book will consist entirely of my personal adventures,
travels, etc., in that country during five years. I met with a number
of strange characters, all of whom I have introduced; the most
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com