Continuum Mechanics 1st Edition Andrus Koppel PDF Download
Continuum Mechanics 1st Edition Andrus Koppel PDF Download
pdf download
https://ebookgate.com/product/continuum-mechanics-1st-edition-
andrus-koppel/
Get the full ebook with Bonus Features for a Better Reading Experience on ebookgate.com
Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...
https://ebookgate.com/product/continuum-and-solid-mechanics-concepts-
and-applications-victor-quinn/
ebookgate.com
https://ebookgate.com/product/continuum-mechanics-modeling-of-
material-behavior-1st-edition-martin-h-sadd/
ebookgate.com
https://ebookgate.com/product/elements-of-continuum-mechanics-and-
thermodynamics-1st-edition-joanne-l-wegner/
ebookgate.com
https://ebookgate.com/product/introduction-to-continuum-mechanics-
solutions-manual-4th-edition-w-michael-lai/
ebookgate.com
Geometric Continuum Mechanics Advances in Mechanics and
Mathematics 42 Band 42 1 ed. 2020 Edition Reuven Segev
(Editor)
https://ebookgate.com/product/geometric-continuum-mechanics-advances-
in-mechanics-and-mathematics-42-band-42-1-ed-2020-edition-reuven-
segev-editor/
ebookgate.com
https://ebookgate.com/product/principles-of-continuum-mechanics-a-
study-of-conservation-principles-with-applications-1st-edition-j-n-
reddy/
ebookgate.com
https://ebookgate.com/product/the-web-writer-s-guide-1st-edition-
darlene-maciuba-koppel/
ebookgate.com
https://ebookgate.com/product/the-continuum-companion-to-ethics-1st-
edition-christian-miller-ed/
ebookgate.com
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
CONTINUUM MECHANICS
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or
by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no
expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of information
contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in
rendering legal, medical or any other professional services.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
CONTINUUM MECHANICS
ANDRUS KOPPEL
AND
JAAK OJA
EDITORS
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Copyright © 2010 by Nova Science Publishers, Inc.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or
transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical
photocopying, recording or otherwise without the written permission of the Publisher.
For permission to use material from this book please contact us:
Telephone 631-231-7269; Fax 631-231-8175
Web Site: http://www.novapublishers.com
Independent verification should be sought for any data, advice or recommendations contained in
this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage
to persons or property arising from any methods, products, instructions, ideas or otherwise
contained in this publication.
This publication is designed to provide accurate and authoritative information with regard to the
subject matter covered herein. It is sold with the clear understanding that the Publisher is not
engaged in rendering legal or any other professional services. If legal or any other expert
assistance is required, the services of a competent person should be sought. FROM A
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
CONTENTS
Preface vii
Chapter 1 Manifolds on Continuum Mechanics 1
Jari Mäkinen
Chapter 2 Analysis of Shell Structures Applying Triangular Finite Elements 53
C. W. S. To
Chapter 3 Transport Control of Fluid and Solutes in Microchannels Using AC
Field and Semiconductor Diodes 85
Dimiter N. Petsev and Orlin D. Velev
Chapter 4 Continuum Description of Flow-Like Landslide Dynamics 105
M. Pirulli
Chapter 5 Examination of Cracks Based on Continuum-Mechanics 147
Agnes Horvath
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
vi Contents
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
PREFACE
Continuum mechanics is a branch of mechanics that deals with the analysis of the
kinematics and mechanical behavior of materials modeled as a continuum (e.g., solids and
fluids, liquids and gases). A continuum concept assumes that the substance of the body is
distributed throughout — and completely fills — the space it occupies. Differential equations
can be employed in solving problems in continuum mechanics. Some of these differential
equations are specific to the materials being investigated and are called constitutive equations,
while others capture fundamental physical laws, such as the conservation of mass (the
continuity equation), the conservation of momentum (the equations of motion and
equilibrium), and the conservation of energy (the first law of thermodynamics). This new and
important book gathers the latest research from around the globe in this field.
Chapter 1 - The authors study differential geometry very elementarily, but hopefully in a
practical way. The authors divide vector spaces into material and spatial spaces since these
spaces behave differently in the observer transformation and with objective derivatives (Lie-
derivatives). All the vector spaces, which they consider, have a metric tensor thus they are
metric vector spaces, and all the finite dimensional manifolds are Riemannian manifolds that
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
are embedded in a Euclidean space. Additionally, the authors may identify a dual vector space
by its primary vector space. In classical tensor analysis, this identification is applied, but here
they make distinction between primary and dual spaces in the formulation, and the
identification can be accomplished later. If the identification of dual and primary vector
spaces is done a priori, then push-forward and pull-back operations are not uniquely defined.
As an example, the authors will study the rotation manifold and its underlying geometric
structure in the terms of differential geometry.
In continuum mechanics, the authors have different manifolds. The placement field of
continuum medium takes values in a Hilbert space, where a chart parametrization maps
vector-valued functions into vector-valued functions. The placement field needs an infinite
number of basis functions in order to present an arbitrary placement field on continuum,
yielding infinite-dimensional manifolds.
In Lagrangian mechanics, forces are divided differently into constraint forces and applied
forces. In addition, rich mathematical methods like variational calculus and other mechanical
principles are included. In terms of differential geometry, Lagrangian mechanics describes a
motion on an event manifold with a Lagrangian functional on the tangent bundle of the event
manifold. An event manifold is a time-placement manifold which is also a constraint
manifold, i.e. the time and placement variables satisfy all the constraints.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
viii Andrus Koppel and Jaak Oja
The authors think that there is a need for another type of mechanics between Newtonian
and Lagrangian mechanics, as it is sometimes done. This mechanics could be named
d‘Alembertian mechanics where the principle of virtual work is its cornerstone. Here the
authors include inertial forces also in the virtual work form. The virtual work may be viewed
as a linear form on the tangent field-bundle. This field-bundle is also a tangent bundle of the
placement manifold at fixed time. The authors give definitions for the virtual work in the
finite-dimensional and infinite-dimensional cases. In addition they give definition for the
variation, Lie derivative and Lie variation. The concept of push-forward and pull-back
operators is essential for understanding Lie derivatives and variations.
Finally, the authors consider constraint point-manifolds that arise from point-wise
holonomic constraint equations. The usual geometric joints of a multibody system like
spherical, revolute, cylindrical, universal, helical, prismatic, and sliding joints can be
presented via holonomic constraint equations that only depend on displacement at
corresponding geometric points. All these constraints generate a smooth point-manifold that
can be parametrized. Also, the principle of virtual work and its geometric structure are
naturally related with the parametrization of the constraint manifold
Chapter 2 - Over the years many shell finite elements has been developed for applications
in aerospace, automotive and shipbuilding industries. Misuse or abuse of some of these shell
finite elements in the relatively mature shell finite element technology is not uncommon.
Some of these shell finite elements were based on the principles of classical shell theory in
which the simplest one is the theory of Love. Others were based on intuitive or heuristic
arguments. For reasons of economy, mathematical simplicity and accuracy, lower order flat
triangular shell finite elements are popular among designers. This article is concerned with
the review, development and application of triangular shell finite elements. Emphasis is on
mixed formulation based lower order flat triangular shell finite elements. Finite element
representation of shell structures is introduced. Linear analysis of static and dynamic, and
nonlinear analysis of static and dynamic problems, are included in this article.
Chapter 3 - This chapter presents an overview of recent results on the electric field
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
control and manipulation of fluids in microfluidic devices. The newer approaches are based
on using alternating or a combination of alternating and direct current fields. The alternating
field can be locally converted to direct by semiconductor diodes that may be placed at key
locations where an electroosmotic force has to be applied to the fluid. Such techniques allow
to design and fabricate small micrometer sized pumps and mixers. The latter are important
because of the inherent low Reynolds characteristics of the flow in microchannels. The diode
mixers are simple to fabricate and can be turned on and off depending on the operational
requirements. Combining alternate and direct current fields and diode pumps makes possible
the decoupling of the electroosmotic fluid flow from the electrophoretic particle or
macromolecular mass flux. This can be exploited for precise analyte focusing,
preconcentration and separation.
Chapter 4 - Landslide run-out is a complex phenomenon, much more difficult to simulate
by models than flow of fluids. The main complicating aspects concern that landslide material
is often heterogeneous and its characteristics may change during the landslide movement due
to drainage, hydraulic interaction between fluid and grains, comminution of grains or mixing
with surface water or partly or fully liquefied superficial material entrained from the path.
The continuum mechanical theory, treating the heterogeneous and multiphase moving
mass as a continuum, has emerged in the last years as a useful tool for describing the evolving
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Preface ix
geometry and the velocity distribution of a mass flowing down a surface. A hypothetical
material, ―equivalent fluid‖, whose rheology is controlled by a small number of parameters is,
in fact, introduced to represent the bulk behaviour of a landslide.
After a brief introduction on landslide characteristics and dynamics, new advances in the
continuum mechanical description of flow-like landslides are discussed in dedicated sections.
Each section deals with one of the main aspects that characterize the physical behaviour of a
landslide and presents the simplifying, but nevertheless realistic, assumptions made to
streamline their mathematical formulation.
The mathematical formulation is then implemented in a numerical code (RASH3D) to
test the capability of each mathematical assumption in allowing the modelling of real
phenomenon dynamics. Results of numerical simulations of laboratory tests and real events
are discussed in this chapter to this aim.
Chapter 5 - The phenomenon of failure by catastrophic crack propagation in structural
materials poses problems of design and analysis in many fields of engineering. Cracks are
present to some degree in all structures. They may exist as basic defects in the constituent
materials or they may be induced in construction or during service life. The continuum-
mechanics can be applied for macro cracks.
Over the past decades the finite element technique has become firmly established as a
useful tool for numerical solution of engineering problems. In order to be able to apply the
finite element method to the efficient solution of fracture problems, adaptations or further
developments must be made. Using the finite element method, a lot of papers deal with the
calculation of stress intensity factors for two- and three-dimensional geometries containing
cracks of different shapes under various loadings to elastic bodies. In order to increase the
accuracy of the results, special singular and transition elements have been used. They are
described together with methods for calculating the stress intensity factors from the computed
results. These include the displacement substitution method, J-integral and the virtual crack
extension technique. Despite of the large number of published finite element stress intensity
factor calculations there are not so many papers published on J-integral to elastic-plastic
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
bodies.
At the vicinity of a crack tip the strains are not always small, but they may be large ones,
too. In this case the J-integral can also be applied to characterise the cracks in elastic or
elastic-plastic bodies.
This chapter describes the computation of the two dimensional J-integral in the case of
small and large strains to elastic and elastic-plastic bodies and represents some numerical
examples, too.
Chapter 6 - Stiffness degradation for laminated composites such as carbon fiber/epoxy
composites is an important physical response to the damage and failure evolution under
continuous or cyclic loads. The ability to predict the initial and subsequent evolution process
of such damage phenomenon is essential to explore the mechanical properties of laminated
composites. This chapter gives a general review on the popular methodologies which deal
with the damage initiation, stiffness degradation and final failure strength of composite
laminates. These methodologies include the linear/nonlinear stress calculations, the failure
criteria for initial microcracking, the stiffness degradation models and solution algorithms in
the progressive failure analysis. It should be pointed out that the assumption of constant
damage variable which is introduced into the constitutive equations of laminated composites
to simulate the stiffness degradation properties is less effective and practical than that of
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
x Andrus Koppel and Jaak Oja
changed damage variable with loads in the framework of continuum damage mechanics
(CDM). Also, different damage evolution laws using CDM should be assumed to describe
three failure modes: fiber breakage, matrix cracking and interfacial debonding, respectively.
Chapter 7 - The continuum mechanics in terms of the differential forms is proposed. The
authors introduce the dual material space-time which consists of the strain space-time and the
stress space-time. In this case, there is a one-to-one correspondence between the kinds of the
basic equations in the continuum mechanics and the kinds of the basic operators in the
differential forms. That is, the kinematic and constitutive equations can be derived by the
exterior differential operator and the Hodge star operator, respectively. Other compound
equations such as the Navier equation, Laplace (wave) equation and the incompatibility
equation can be derived by the combination of the basic operators. This systematic approach
allows us to find (i) the anti-exact solution of the Navier equation and (ii) the J-integral in
fracture mechanics. The result (ii) means that the continuum mechanics in terms of the
differential forms describes a partial aspect of the fracture mechanics. Moreover, the
differential form approach allows us to link the deformation field with the non-deformation
field such as the electromagnetic field. As an example, the authors take up the piezoelectric
and Villari effects and derive the constitutive equations for these effects. These constitutive
equations can be interpreted geometrically as the interaction among the geometrical objets of
the space-time.
Chapter 8 - In this chapter the authors develop a general framework for the modelling of
morphogenesis by introducing a growth process in the structural elements of the cell, which
in turn depends on the stress state of the tissue. Some experimental observations suggest this
feedback mechanism during embryo development, and only very recently this behaviour has
started to be simulated.
The authors here derive the necessary equilibrium equations of a stress controlled growth
mechanism in the context of continuum mechanics. In these derivations the authors assume a
free energy source which is responsible for the active forces during the elongation process,
and a passive hyperelastic response of the material. In addition, they write the necessary
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
conditions that the active elongation law must satisfy in order to be thermodynamically
consistent. The authors particularise these equations and conditions for the relevant elements
of the cytoskeleton, namely, microfilaments and microtubules. The authors apply the model
to simulate the shape changes observed during embryo morphogenesis in truss element. As a
salient result, the model reveals that by imposing boundary stress conditions, unbounded
elongation would be obtained. Therfore, either prescribed displacements or cross-links
between fibres are necessary to reach a homeostatic state.
Chapter 9 - In the nearest-nodes finite element method (NN-FEM), finite elements are
mainly used for numerical integration; for each quadrature point, shape functions are
constructed from a set of nodes that are the nearest to the quadrature point, nodes from
neighbour elements may be involved in the construction. Based on this strategy, there are
several techniques available for constructing shape functions. In this paper, the moving local
polynomial interpolation method is adopted. Benefiting from the above strategy, NN-FEM
has several attractive features. High-order shape functions can be constructed from simplex
finite element meshes; Analysis accuracy of NN-FEM is not influenced by element distortion;
NN-FEM can deal with extremely large deformation, etc. Furthermore, NN-FEM provides a
favourable environment for implementing an adaptive algorithm.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Preface xi
Chapter 10 - In this chapter, an adaptive finite element method is formulated based on the
newly developed nearest-nodes finite element method (NN-FEM). In the adaptive NN-FEM,
mesh modification is guided by the gradient of strain energy density, i.e. a larger gradient
requires a denser mesh and vice versa. A finite element mesh is iteratively modified by a set
of operators, including mesh refinement, mesh coarsening and mesh smoothing, to make its
density conform with the gradient of strain energy density. The selection of a proper operator
for a specific mesh region is determined by a set of criteria that are based on mesh intensity.
The iteration loop of mesh modification is stopped when the relative error in the total
potential energy is less than a prescribed accuracy. Numerical examples are presented to
demonstrate the performance of the proposed adaptive NN-FEM.
Chapter 11 - The natural finite element approach introduced by John Argyris in the early
sixties is characterized by the distinction between rigid body motion and deformation, on the
one hand, and by the description of the latter in compliance with the element purpose and
geometry, on the other hand. For triangular and tetrahedral elements the concept suggests
strain and stress measures defined along the sides or the edges respectively as homogeneous
normal quantities, free of shear. In the mechanics of continua the corresponding infinitesimal
elements represent minimum configurations to define local deformation in two- and three
dimensions.
This treatise concerns utilization of the natural approach on the continuum level within a
consistent theoretical framework. It is proposed to begin with a reference system of
supernumerary coordinates associated with the elementary tetrahedron in the space or with
the triangle in the plane. Vectorial quantities are defined, the operations of gradient and
divergence are interpreted in this system. The natural deformation rate is deduced from the
velocity field, the stress is introduced as work conjugate measure. The condition for local
equilibrium is presented in natural quantities as well as the stress definition in association
with the resultant forces. The set up of material constitutive relations is exemplified for the
elastic solid and for viscous media. Beyond the description of the momentary kinematics as
from the velocity field, the appearance of finite deformation is considered basing on
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
displacements. Illustration of the methodology for a plane elastic case terminates the part
regarding the mechanics of solids. Extension to fluid motion and to thermal phenomena is
appended.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
In: Continuum Mechanics ISBN: 978-1-60741-585-5
Editors: Andrus Koppel and Jaak Oja, pp.1-52 ©2010 Nova Science Publishers, Inc.
Chapter 1
Jari Mäkinen*
Tampere University of Technology, Department of Mechanics and Design, P.O. Box 589,
FIN-33101 Tampere, Finland
ABSTRACT
We study differential geometry very elementarily, but hopefully in a practical way.
We divide vector spaces into material and spatial spaces since these spaces behave
differently in the observer transformation and with objective derivatives (Lie-
derivatives). All the vector spaces, which we consider, have a metric tensor thus they are
metric vector spaces, and all the finite dimensional manifolds are Riemannian manifolds
that are embedded in a Euclidean space. Additionally, we may identify a dual vector
space by its primary vector space. In classical tensor analysis, this identification is
applied, but here we make distinction between primary and dual spaces in the
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
formulation, and the identification can be accomplished later. If the identification of dual
and primary vector spaces is done a priori, then push-forward and pull-back operations
are not uniquely defined. As an example, we will study the rotation manifold and its
underlying geometric structure in the terms of differential geometry.
In continuum mechanics, we have different manifolds. The placement field of
continuum medium takes values in a Hilbert space, where a chart parametrization maps
vector-valued functions into vector-valued functions. The placement field needs an
infinite number of basis functions in order to present an arbitrary placement field on
continuum, yielding infinite-dimensional manifolds.
In Lagrangian mechanics, forces are divided differently into constraint forces and
applied forces. In addition, rich mathematical methods like variational calculus and other
mechanical principles are included. In terms of differential geometry, Lagrangian
mechanics describes a motion on an event manifold with a Lagrangian functional on the
tangent bundle of the event manifold. An event manifold is a time-placement manifold
which is also a constraint manifold, i.e. the time and placement variables satisfy all the
constraints.
We think that there is a need for another type of mechanics between Newtonian and
Lagrangian mechanics, as it is sometimes done. This mechanics could be named
*
Academy Research Fellow, E-mail: [email protected], Fax: +358 3 3115 2107, Phone: +358 50 5366632
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
2 Jari Mäkinen
d‘Alembertian mechanics where the principle of virtual work is its cornerstone. Here we
include inertial forces also in the virtual work form. The virtual work may be viewed as a
linear form on the tangent field-bundle. This field-bundle is also a tangent bundle of the
placement manifold at fixed time. We give definitions for the virtual work in the finite-
dimensional and infinite-dimensional cases. In addition we give definition for the
variation, Lie derivative and Lie variation. The concept of push-forward and pull-back
operators is essential for understanding Lie derivatives and variations.
Finally, we consider constraint point-manifolds that arise from point-wise holonomic
constraint equations. The usual geometric joints of a multibody system like spherical,
revolute, cylindrical, universal, helical, prismatic, and sliding joints can be presented via
holonomic constraint equations that only depend on displacement at corresponding
geometric points. All these constraints generate a smooth point-manifold that can be
parametrized. Also, the principle of virtual work and its geometric structure are naturally
related with the parametrization of the constraint manifold
primary vector spaces is done a priori, then push-forward and pull-back operations are not
uniquely defined. We also itemize terms a vector space and a linear space where the linear
space is considered as a trivial manifold (or a linear manifold, or a flat manifold). Vector
spaces usually appear from the tangent spaces of the manifold which are distinct at different
points of a nontrivial manifold
In this section, we give the definitions for vector and tensor algebra on topological vector
spaces1, definitions for manifolds, and tensor algebra on manifolds. We recommend
consulting, especially, the paper [Stumpf & Hoppe 1997], and the textbooks [Wang &
Truesdell 1973] or [Marsden & Hughes 1983] for tensors on manifolds, and textbooks
[Arnold 1978] or [Abraham et al. 1983] for differentiable manifolds. A reader is assumed to
1
We consider the topological vector space as a general vector space without explicit knowledge of a metric.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 3
be familiar with classical tensor algebra on Euclidean spaces2, text books like [Ogden 1984]
or [Truesdell 1977] or [Bonet & Wood 1997].
:V V R, bf , ag f a R ,
which have two properties: bilinearity, i.e. it is linear with respect to each of its two members,
and definite, i.e. if f V is fixed and f a 0 a V , then a 0 . Conversely, if a V is
fixed and f a 0, f V , then f 0 . If f a 0 , the vector a is said to be orthogonal to the
covector f, and vice versa. Note that a covector space is also a vector space satisfying the
vector space properties. Because the vector space and its co-covector space are canonically
isomorphic3, i.e. V V , we have the symmetry property of the dot product: f a a f .
Let F L (V ,W ) be a linear operator from V W . The adjoint operator
F L (W ,V ) is defined with the aid of the dot product as
Fw a w Fa R a V , w W ,
where the first dot product is on the vector space V , and the latter on the vector space W ,
see Figure 1.
On notation: we omit -symbol when there is no source of confusion. Then the terms Fa
and F a are identical. The brackets are used for purpose of dependency, e.g. F(x) a denotes
the linear operator F(x) acts (linearly) on a where the operator depends on x. In this case, the
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
dot symbol may not be omitted. In addition, in the composite mapping of operators, like FG ,
the dot symbol is omitted.
F 1
V W
F
bf v g
V
R bt w g W
R
F
V W
F
Figure 1. The diagram of domains and ranges for the operator F Liso (V ,W ) and its derivatives.
2
The Euclidean space is a real, finite-dimensional, linear, inner-product space with an Euclidean metric.
Topological vector spaces are isomorphic, denoted by , if there exists a (continuous) linear bijection, called
3
isomorphism, between these spaces. Two vector spaces are isomorphic iff they have the same dimensions.
Vector spaces are canonical isomorphic, denoted by , if there exists a natural (‗almost trivial‘) isomorphism.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
4 Jari Mäkinen
A covector space is commonly called a dual vector space, and the elements of the
covector space are called covectors, or dual vectors, or linear forms (which never make any
sense). Additionally, the dot product is also called duality pairing, and an adjoint operator is
called a dual operator. We do not make any notational difference between the elements in the
vector and covector spaces since we desire to use the notation similar to classical tensor
algebra. For example, force quantities like moment and force vectors, and Lagrange
multiplies are the elements of covector spaces. We will use the byte ‗co-‘ instead of the word
‗dual‘ because of its simplicity and compactness.
(a f ) w (f w) a V , w W ,
where the tensor a f belongs to the tensor space produced by V and W , i.e.
a f V W L (W ,V ) . The tensor product is a linear mapping for each member
separately, i.e. a bilinear operator, because of the bilinearity of the dot product. The tensor is
called a two-point tensor if it is defined on two different vector spaces. The general two-point
tensor space T can be denoted by
T : V V V V W W W W
r s t u
that is the space of r-fold on the vector space V , s-fold on the covector space V , t-fold on
the vector space W , and u-fold on the covector space W . This can be shortly denoted by the
tensor space T (r, s; t, u) with the order of r s t u .
4
This vector space V could be a covector space, or more generally, a tensor space
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 5
Note that any other permutation of vector spaces is possible, thus e.g. the notation
, ;0,1) could mean the tensor spaces V W or W V . In the case of one-point tensor
(10
spaces, defined on the same vector or covector space, we use a simplified notation: e.g. the
tensor space T (1,1) for the tensor spaces V V or V V defined on V , or
correspondingly for the tensor spaces W W or W W defined on W .
There are two possible points of view to comprehend a tensor: operational or quantitative.
The operational aspect informs ‗how it works‘, and quantitative responds to ‗how much is it‘.
Mathematicians represent the operational point of view and engineers the quantitative point of
view. Although we will define the tensor by quantitative, we shall keep in mind its
operational aspect: a tensor is a multilinear operator.
T:V V V V W W W W R
r s t u .
The two-point tensor T is an element of two-point tensor space such that it assigns a
tensor for its two-point domain.
The tensor space is a vector space itself by satisfying all vector space properties. Then we
may state that the tensors are vectors and the vectors are tensors. However, we consider the
first-order tensors as vectors, and the higher-order tensors as tensors. Sometimes the tensors
are characterized by their component transformation laws under the change of the basis: the
object is a tensor if its components change like tensor components under a coordinate
transformation. For example, Christoffel symbols are not tensors. Conversely, the vectors are
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
characterized by direction, magnitude, and, especially, by the parallelogram law: the vector
can be added to another vector by the parallelogram law. For example, it is often incorrectly
claimed that the finite rotation does not satisfy the parallelogram law, whereupon the finite
rotation vector is not a vector quantity. We keep these characterizations rather old-fashioned
and they can lead to serious misunderstandings. The vectors and tensors may be characterized
by studying if they are elements of corresponding vector and tensor spaces, respectively.
The trace of the second order tensor is usually defined by the contraction of its
components. This is a contradiction with the component independence of the tensor, although,
the trace is component-independent. We follow the definition of the trace given in [Truesdell
1977; App. II].
b g
tr f a : f a R
.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
6 Jari Mäkinen
Also the trace operation for the tensor on V V can be applied by noting V V , but it
is not defined for two-point tensors. The double-dot product for the tensors f t V W
and v w V W is defined via the ordinary dot product
bf tg:bv wg : bf vg bt wg
V W
R
,
where the subscripts indicate the vector space of the corresponding dot product. Therefore,
the double-dot product is a mapping L (V W V W , R) that is a four-linear operator.
All tensors, which we have considered, have been presented by the tensor product of the
vectors, e.g. the tensor f a . However, a general tensor can not be expressed directly in that
way. We may present a common tensor with basis vectors of tensor space. Let {Gi } , with the
index i 12, ,3 , be an ordered basis for the vector space V and let {g i } ( i 12, ,3 ) be an
ordered basis for the vector space W, then we may present a general second-order two-point
tensor T V W by the linear combination of the basis vectors, namely (with the
conventional summation)
T TijGi g j (1)
where Gi g j V W corresponds the basis vector of the tensor with the coefficient Tij R .
The coefficient matrix [Tij ] R33 is called the component matrix of the tensor T with respect
to the bases {Gi } and {g i } 5. Higher order tensors are represented a similar way. In order to
represent tensors on covector spaces, we have to define the bases for the covector spaces.
bases (dual bases) {Gi } and {g i } on the covector spaces V and W are defined by formulas
Gi G j ij , gi g j ij ,
where ij is the Kronecker‘s delta symbol. Then, for example, the tensor T V W may be
represented by T TijGi gj .
We have defined a tensor algebra on a topological vector space. These vector spaces are
often induced by a manifold, yielding a tensor algebra on the manifold that we define next.
5
The component matrix is an isomorphism between the tensor space V W and the Cartesian space R33 .
6
a mapping is a bijection if it is injective and surjective, i.e. one-to-one and onto mapping
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 7
under a mapping, see Figure 2. A pair (Ui , i ) is called a chart or a parametrization chart, and
the mapping i is called a chart mapping or a parametrization mapping
(t ) (0)
x lim , where (0) x, (t ) M
t 0 t .
The tangent vector x belongs to a tangent space of the manifold, namely x Tx M , see
Figure 3. The tangent (vector) space Tx M is a set of tangent vectors at x M .
7
a diffeomorphism is a bijection with continuously differentiable mapping and its inverse mapping
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
8 Jari Mäkinen
Tx M
x x
( t )
Figure 3. The tangent vector x and its tangent space Tx M on the manifold M at the point x .
b
TM := x , Tx M
x M
g
.
The dimension of the tangent bundle is twice the dimension of the manifold M .
Especially, the pair of state vectors, the placement x(t ) and velocity vectors v(t ) , belongs to
the tangent bundle, (x, v)(t ) TM .
For a two-point tensor, its domain of points is divided into two separate but not
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
independent regions which are defined in the vector spaces V and W . It is convenient to
choose a material body B, containing all material points of body, for one region of the domain
and another region which is obtained via a mapping of the material body B. The material
body B is a set of points and its elements are denoted X , Y, Z,B.
The initial reference placement B0 is defined as the special case of the current placement
B by setting t 0 , giving
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 9
Figure 4. The material body B with the body-point X , the material placement B0 with the material
place vector X and the material vector A , and the spatial placement B with the spatial place vector
x and the spatial vector a .
8
A base point is a point of the manifold where a tangent space is induced.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
10 Jari Mäkinen
Now we could set V =TXB0 , W =TxB giving, for example, the type of (1,1;1,1) two-
point tensor at a body-point X B with mappings X 0 ( X ) and x ( X )
T:TXB0 TXB0 TxB TxB R , where TXB0 and TxB are the covector spaces for the
vector spaces TXB0 and Tx B , respectively. The two-point tensor T is an element of
multilinear operators, denoted as T L (TXB0 TXB0 TxB TxB , R ) . For the sake of
simplicity, we omit body points and mappings when expressing tensors and vectors, and we
call the place vectors X and x as the material base point and the spatial base point,
respectively.
So far we have studied vectors and tensors in vector spaces without knowledge about its
metric. A metric of the vector space is a symmetric positive-definite bilinear operator9 , called
a metric tensor. Let pairs (V , G) and (W , g) indicate metric vector spaces in the material and
spatial representation, with the (material) metric tensor G L (V ,V ) and the (spatial) metric
tensor g L (W ,W ) . Metric tensors are used for measuring distances and deformation,
which is impossible without introducing metric. Since manifolds are embedded in the
Euclidean space E3 , we could choose metric tensors as the identity elements. This can be
achieved by identifying the metric vector spaces (V , G) and (W , g) with the Euclidean vector
space E3 . However, this identifying is not accomplished at this moment since it is
informative to comprehend the existence of the metric tensor in different operators like
deformation and strain tensors.
, :V V R, (a, b) a, b G : Ga b ( a b)
,
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
where the dot product is defined in Def. 1. For simplicity, the covector Ga is often denoted by
b g
a . The tensor F L (V , G),(W , g ) , its transpose operator F T is defined via the inner
product
FTw, v w, Fv g
w W , v V
G .
9
a metric is a scalar valued function that induces a linear bijection (isomorphism), called a metric tensor
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 11
b
r0 e p 0 e ,g (2)
r1 r0 cos e p 0 sin p 0 , e, r0 , r1 E 3 , R ,
where denotes the cross product on E3 . Now the rotated vector p1 can be expressed with
the aid of (2)
p 1 p 0 r0 r1
b g b g
p 0 1 cos e e p 0 e p 0 sin p 1 , p 0 , e, r0 , r1 E 3 , R ,
(3)
r1
r0
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
E3 p1 r1 e p 0 sin
e p0
r0
Figure5. A rotational motion about e -axis where p 0 is the original vector and p1 is the rotated vector.
(Note that e r0 e p0 ).
: e e 1, e E3 , R (4)
where the unit rotation axis vector e and the non-negative rotation angle are oriented such
that they form a right-handed screw, see Figure5.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
12 Jari Mäkinen
Note that the length of the rotation vector is equal to the rotation angle, i.e. |||| . Here
we do not restrict the angle of rotation; it may have any non-negative values. The rotation
vector lives in a three-dimensional vector space that is isomorphic to the Euclidean space
E3 . This issue will be realized later.
Now Eqn (3) can be written in terms of the rotation vector, which yields the expression
of the rotation operator
sin 1 cos
p1 p 0
p0
2
b
p0 g
F sin ~ 1 cos ~ IJ p
GI
(5)
H
2
K 2 0 ,
~
where the skew-symmetric tensor called the rotation tensor, is defined by formula
~ 3 , or more formally ~
a a, a E : .
sin ~ 1 cos ~ 2
R: I ,
2
.
Then in Figure5, the rotation operator R transforms the vector p 0 into the vector p1 , i.e.
p1 Rp0 .
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
2n1 1 b g,
b g 2n 1 b g 2
b g
~ n1 2 n1 ~ ~ n1 2 n1 ~
(6)
~ 1 ~ 1 ~ ~
R I 2 3 :exp .
2! 3!
e j (7)
10
There exists no effective criterion for the term canonical.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 13
This is a significant property of the rotation operator and offers the shortest relationship
between the rotation vector and the rotation operator. We also note that the transpose of the
rotation operator is equal to the reverse rotational motion
sin
e~ j 1 cos e~ j b g
2
RT I R , (8)
2
~T ~
due to the skew-symmetry of the rotation tensor, .
This yields the proper orthogonal features of the rotation operator
R T R RR T I,
b g
det R 1 ,
(9)
where I is the identity element. It is evident since the inverse of the rotation operator is the
reverse rotation operator. If an operator satisfies Equation (9a) solely, there exist two possible
values for its determinant, namely det(R) {1,1} , where the first value (+1) produces the
preservation of the orientation.
A rotation operator can be written also with the aid of a rotation axis e , yielding
R I sin ~
e (1 cos )~
e2, , (10)
This relation makes it comprehensible that the rotation operator does not depend on the
multiples of the rotation revolution counts, i.e. R() R( 2ie), i N .
Def. 14 gives a canonical parametrization of the rotation manifold SO(3). The
parametrization can represent a rotation operator only locally, and there exists no
parametrization that is global as well as non-singular. Note that a parametrization is a
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
mapping from an open set of Euclidean space into some open set of the manifold. The
rotation vector parametrization is singular at the rotation angle equal to 2 and its multiples.
It is clear that the singularity naturally appears in dynamical analysis with large rotations, and
cannot be omitted. Singularity should be considered as a non-differentiable hole that must not
be omitted by skipping. The singularity is due to fact that the rotation manifold is compact,
and there does not exist a single continuous parametrization from an open set of the Euclidean
space E 3 onto this compact manifold, see details in [Stuelpnagel 1964].
A rotation operator can be presented by higher dimensional, singularity-free
representations where a unit quaternion is a four-parametric example. The coordinates of a
quaternion are not independent, in fact, a quaternion produces a three-dimensional manifold
into a four-dimensional Euclidean space, that is a unit three-sphere S3 (surface) embedded in
E 4 . Hence, we do not speak about parametrization when considering a mapping between
different manifolds, for example in the case of a unit quaternion this mapping is S 3 SO(3) .
The description of a rotation motion has been studied for a long time, so there exists a
large number of different representations of a rotation motion. Three-dimensional
representations are rotation vectors, Euler angles, Bryant angles, Rodrigues parameters
(Gibbs vector), and four-parametric representations are unit quaternions (Euler-Rodrigues
parameters), linear parameters, Euler rotation, and Cayley-Klein parameters, see [Spring
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
14 Jari Mäkinen
1986] and a historical aspect for Euler-Rodrigues parameters in [Cheng & Gupta 1989]. In
addition, there exists a higher dimensional representation of a rotation operator, like a rotation
matrix, that has a dimension equal to nine.
Four-dimensional descriptions are topologically connected with a unit three-sphere S3 and
to the proper unitary group SU(2) that is a group of complex 2-by-2 matrices, and their joined
algebra is an even Clifford subalgebra (quaternion algebra) and an algebra of 2-by-2 skew-
Hermitian traceless matrices (Lie-algebra su(2)), respectively, see [Choquet-Bruhat et al.
1989].
Correspondingly, three-dimensional descriptions are the parametrizations of the rotation
manifold and their algebras are the cross product in the Euclidean space E 3 (Lie-algebra in
E 3 ) and an algebra of skew-symmetric tensors (Lie-algebra so(3)). We consider three-
dimensional descriptions and especially the rotation vector a simple, geometrical significance
representation. The major drawback of the rotation vector parametrization, singularity, can be
passed by introducing another parametrization chart such that the parametrization mappings
cover the rotation manifold globally.
C : 2 ,
.
Then the rotation angle of the complement rotation vector is C 2 and the
rotation axis is e C e .
After substituting the complement rotation vector into Def. 14, we notice that the rotation
vector and its complement represent the same rotation operator, i.e. R(C ) R() . Def. 15
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
is a change of parametrization in the parameter space E3 , see Figure 2 and Figure 6. This
change of parametrization is a continuously differentiable mapping on the open domain
0 2 , giving a smooth construction of the rotation manifold SO(3) at this domain.
Note that the complement of a complement rotation vector is a rotation vector itself, i.e.
(C )C , hence there is no priority over these parametrization charts.
We could represent the rotation manifold globally with these two parametrization charts.
When a rotation angle exceeds straight angle ( ), we accomplish the change of
parametrization according to Def. 15, giving a new rotation angle smaller than straight angle.
Thus, we never get into trouble with singularity at 2 . As it is illustrated in
Figure 6, the change of parametrization maps rotation angle outside of straight angle into
inside of straight angle. Note that there exists no other canonical parametrization with rotation
less than perigon such as those parametrizations given in Def. 15.
The zero rotation vector is an isolated point, the centre of the domain, for the
parametrization change. Using a limit process, we find out that the rotation operator
approaches to the identity element when the rotation angle is decreased. Hence, we could
modify the domain of the parametrization where the rotation angle is less than perigon, i.e.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 15
< 2 including the zero rotation angle. This domain is still an open domain in the Euclidean
space E 3 , indeed, it is an open ball in E 3 with 2-radius.
parametrization
~
parametrization
~ exp( C )
exp( ) mapping
mapping
2 2
change of
parametrization
3
Figure 6. The change of parametrization in the parameter space E for the canonical representation of
the rotation manifold.
manifold, called the rotation manifold, we show that the rotation manifold is a Lie group, too.
This issue is significant especially in composite rotations.
b g b g
[1] the internal operation is associative A BC AB C, A, B, C G ,
[2] there is a unique element I G called identity such that AI IA A, A G ,
[3] for each A G there exists a unique element of G called the inverse of A such that
A 1A AA 1 I .
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
16 Jari Mäkinen
o
SO(3): R: E3 E3 linear R T R RR T I, det( R ) 1 t.
Since SO(3) is a group it has to fulfill all group properties given in Def. 16.
A rotation operator defined in Def. 14 is also an element of the special orthogonal group
as it was shown in Equation (9). The rotation operators form a non-commutative group with
the internal operation, called composite mapping, and an identity element as the identity
operator I . Hence, we may denote R SO(3) .
Bruhat et el 1989; pp. 181-182]. It is based on reality that any two composite rotations can be
represented by Eulerian angles, giving a differentiable change of parametrization mappings
between different sets of Eulerian angles.
a, b, c b, c, a c, a, b 0, a, b, c l
.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 17
Ad R:l l, Ad Rb : RbR1 , R L .
Note that b l is an element of the Lie algebra l and R is an element of the Lie group L.
The Lie algebra adjoint transformation maps an element of the Lie algebra into another Lie
algebra element.
a, b
e b
d R()b R( ) gj
1
d 0.
Lie brackets is a bilinear skew-symmetric form and satisfies Jacobi‘s identity, given in
Def. 19.
Especially, let R() SO(3) be a -parametrized rotation operator, an element of the
~
special orthogonal group, given by formula R() exp() . Differentiating the expression
~
exp() with respect to the parameter at 0 gives the tangent vector space at the
identity I SO(3) , yielding
~
d exp( ) ~
. (11)
d 0
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
~
Thus, the skew-symmetric tensor belongs to the tangent space of the rotation manifold
~
SO(3), denoted by TI SO(3) , where the identity I SO(3) represents a base point of the
~
rotation manifold. The skew-symmetric tensor is also an element of Lie algebra so(3) for
corresponding Lie group SO(3). We could also mark so(3) TI SO(3) , i.e. Lie algebra is
canonical isomorphic to the tangent space of the rotation manifold at the identity. Moreover,
we may denote the Lie algebra so(3) as a set of skew-symmetric operators (tensors)
~
{ ~ ~
so(3) : E3 E3 linear T . } (12)
We obtain the Lie brackets of the Lie algebra so(3) by differentiating the Lie algebra
~
adjoint representation Ad R ( ) with respect to at 0
~ ~ ~ ~
d(Ad R ( 0 ) ) d(exp( ) exp( )) ~ ~ ~~
. (13)
d d 0
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
18 Jari Mäkinen
~ ~ ~ ~ ~~ ~ ~
Hence, the Lie brackets for the Lie algebra so(3) is [, ] , , so(3) .
The vector cross product (): E3 E3 E3 in the Euclidean space E3 is a Lie algebra
with Lie brackets defined by
The vector cross product () is a bilinear, with respect to vector addition and scalar
multiplication, and a skew-symmetric operator over E3 and satisfies Jacobi‘s identity in Def.
19. The Lie algebra so(3) can be identified with the cross product on E3 by formula
~
a a, a E3 , (15)
~
where the vector E3 is the axial vector for the skew-symmetric tensor so(3) .
b g b g bg
[, ]g , l , , g
.
b g~
~~ ~~
, (16)
see proof e.g. in [Marsden & Ratiu 1999; p. 290]. The tilde mapping is also a linear bijection
giving isomorphic correspondence between the elements of the Lie algebras, denoted by
E3 so(3) . For computational purposes, the Lie algebra in the Euclidean space E3 is simpler
than the Lie algebra so(3) and, hence, it will be utilized in following. The Lie algebra in the
Euclidean space E 3 equipped with the cross product as the Lie bracket is a rather unusual Lie
algebra since, by our knowledge, there does not exist a Lie group whose the Lie algebra it is.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 19
~
mat
Left R R inc : RR inc
mat
R exp(R ), mat
R inc , R SO(3)
,
where R inc
mat is a material incremental rotation operator, and
R is a material incremental
rotation vector with respect to the base point R SO(3) . This description is called material
since the incremental rotation operator acts on a material vector space.
~ ~
inc RR inc R ,
R spat R R R R T , and R R R ,
mat T
(17)
where the first relation is called an inner automorphism that is an isomorphism onto itself, the
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
~ ~
second relation is a Lie algebra adjoint transformation AdR R RR RT , see Def. 20, and
the last relation is another Lie algebra adjoint transformation on the Euclidean space with the
vector cross product as the Lie algebra ( E3 , ) .
T SO ( 3)
spat I
~
~
exp b g
T SO ( 3)
d i
~
exp
~
mat I I
T SO ( 3)
left I
T SO ( 3)
spat R
I ~ R
R
T SO(3)
mat R
R SO( 3)
~
R leftTI SO ( 3)
SO( 3)
T SO ( 3)
left I
Figure 7. A geometric representation of the material (on the left) and spatial tangent spaces (on the
right) on the rotation manifold.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
20 Jari Mäkinen
~
{ ~ ~ ~
T SO(3): R : (R , ) with R; R SO(3), so(3)
mat R },
~
where an element of the material tangent space R matTR SO(3) and is a skew-symmetric
~ ~
tensor, i.e. R so(3) . The notation (R, ) , the pair of the rotation operator R and the skew-
~
symmetric tensor , represents the material skew-symmetric tensor at the base point
~
R SO(3) , see Figure 7. Hence, we may express that R is a skew-symmetric tensor, or a
tangent tensor, at the point R in the manifold SO(3) . For simplicity, we could omit the base
~
point R by denoting matTR SO(3) if there is no danger of confusion.
~
{ ~ ~ ~
T SO(3): R : (R , ) with R; R SO(3), so(3)
spat R },
~
where an element of the spatial tangent space R spatTR SO(3) and is a skew-symmetric
~ ~
tensor, i.e. R so(3) . The notation (R, ) , the pair of the rotation operator R and the skew-
~
symmetric tensor , represents a spatial skew-symmetric tensor at the base point R , see
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
~
Figure 7. Again, we could omit the base point R , i.e. spatTR SO(3) if there is no danger of
confusion.
Rotation operators, the elements of the Lie group SO(3), are defined as linear operators
R L ( E3 , E3 ) . Eqns (17b,c) give another interpretation to a rotation operator, it is an adjoint
transformation between material and spatial tangent spaces. Additionally, a rotational motion
induces the rotation operator, since the rotation operator maps the material place vector
X B0 into the spatial place vector x B by the equation x(t ) R(t )X , i.e. R L (B0 , B ) .
More generally, a rotation operator transforms material vectors into spatial vectors, that is
R L (TXB0 , TxB ) .
11
Usually this space is called left-invariant vector field.
12
Usually is called as right-invariant vector field
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 21
~ : E 3 so(3) that is the tilde mapping. The spatial and material tangent spaces are
isomorphic where the isomorphism is an adjoint transformation
Ad R:mat TR SO(3) spatTR SO(3) , given in (17b). Additionally, the Lie algebra so(3) is
isomorphic in the material tensor space with an isomorphism so(3) matTR SO(3) by
~ ~
R . Then we may express
E3 so(3) T SO(3)
mat R T SO(3) .
spat R (18)
Isomorphism states that for any element from a vector space we can take an element from
an isomorphic vector space with a linear one-to-one correspondence. Therefore, the
isomorphic spaces have the same structure and we may associate the elements of the
isomorphic spaces.
e
~ ~
j ~
e j e
~
exp exp exp R , j (19)
~
where we are finding an incremental rotation tensor, the virtual rotation tensor , such that
~ ~ ~
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
it belongs to the same tangent space as the rotation tensor , i.e. such that , matTI SO(3)
~
with the identity as a base point omitted for simplicity. Note that R = exp() , and
~ ~ ~
R matTR SO(3) . We point out that the skew-symmetric tensors and R do not belong
~ ~ ~ ~
to the same tangent space of rotation as it can be verified that exp()exp() exp( ) ,
~
generally. The associated rotation vector for the skew-symmetric tensor is called the
total material rotation vector whose base point is the identity. Taking the derivatives of
(19) with respect to the parameter at 0 gives after the aid of isomorphism
(18), see e.g. [Ibrahimbegović et. al. 1995]
R T ,
sin 1 cos ~ sin
T I , (20)
2 3
, R exp( ), lim T( ) I,
0
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
22 Jari Mäkinen
where the material tangential transformation T T() is a linear mapping between the
virtual material tangent spaces matTI SO(3) matTR SO(3) . Now, we could make another
verification that the virtual rotation vector R and the virtual total rotation vector
belong to different vector spaces on the manifold. This is because the tangential
transformation T is equal to the identity only at 0 . Note that the transformation T has
an effect on the base points, changing the base point I into R.
By examining the tangential transformation T in Eqn (20), we found that the
transformation is non-singular when the rotation angle is less than perigon, i.e. 2 . It is
worth noting that the tangential transformation T() , the corresponding rotation operator
~
R() and the skew-symmetric rotation tensor have the same eigenvectors. Hence, T() ,
~
R() , and are commutative, see [Ibrahimbegović et. al. 1995].
{ b g ~
e j
T : R : , R exp SO(3), E 3
mat R },
where an element of the material vector space is R matTR , which is an affine space with the
rotation vector as a base point and the incremental rotation vector as a tangent vector.
Hence, the tangential transformation T is a mapping T: matTI matTR . Note that the
elements of this material vector space can be added by the parallelogram law only if they
occupy the same affine space, i.e. if their associated skew-symmetric tensors belong to the
same tangent space of the rotation manifold. Def. 28 for a material vector space matTR should
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
be considered as a useful and simple notation with an equivalence relation with a material
tangent space matTR SO(3) , defined in Def. 26.
Respectively, we could determine the spatial tangential transformation, yielding
~ ),
R TT , T T( ), R exp( ( ) , (21)
spat R o b g bg
~ SO(3), E 3 .
T : R : , R exp t (22)
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 23
the rotation operator as given in the Eqn (17c),. From Eqn (17c), it follows with the base point
I SO(3) (note that matTI ).
I II I I , (23)
where ‗ ‘ denotes the canonical isomorphism between the spatial and material vector spaces.
The identity I maps between the vector fields matTI spatTI . Now, the relation between the
spatial and material vectors can be given as (, ) (I, R) where and represent the
base points in the spatial and material vector spaces, respectively. This relation can be written
more compactly as R RR , called a push-forward, where the rotation operator should be
considered as a mapping between the material and the spatial vector spaces of rotation,
R: matTR spatTR , see Figure 8. A push-forward operator maps a material vector space into a
spatial vector space (one-to-one and onto). It makes sense since the rotation operator is a two-
point tensor. We note that the push-forward operator R has no influence on the base point of
the rotation. Another push-forward operator for rotation tensors is given in (17b) where
~ ~
R R R R T is a mapping between the material and spatial tangent spaces of rotation
R R
R R matTR T
spat R
T T
R R
T TT T TT
I I
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
I I T
mat I
T
spat I
Figure 8. A commutative diagram of virtual material and spatial rotation vectors on the rotation
manifold (on the left), and their corresponding vector spaces (on the right).
~
R : RT R
where the dot represents to the time derivative. See justification in [Marsden & Ratiu 1999;
Ch. 8.6 & 15.2].
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
24 Jari Mäkinen
~ ~ .
R R R R (24)
This states that the angular velocity vector is the time derivative of the incremental
rotation vector R ; moreover, (if the base point is omitted) , , mat TR , which is the
material rotation vector space on the rotation manifold. The result in Eqn (24) is often given
as definition for the angular velocity vector in the elementary text books.
Similar expression and derivation can be accomplished for the spatial angular velocity
tensor and vector, yielding
T,
~ : RR
R
~ (25)
~
R R R R ,
where the spatial incremental rotation vector R , its time derivative vector R and the spatial
angular vector R belong to the same spatial vector space on the manifold , , spat TR , the
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
~ ~ ~
R : R , R matTR SO(3) ,
: , R matTR ,
R R
~ ,
~ : ~ T SO(3) ,
R R R spat R
R : R , R spatTR ,
where A R and R are the material and spatial angular acceleration vectors at the base R .
Note that the material incremental rotation vector R , the material angular velocity
vector R and the material angular acceleration vector R (majuscule of alpha-letter) belong
to the same material vector space on the rotation manifold, i.e. R , R , R matTR with the
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 25
~
base point R exp(I ) . At separate moments, these vectors, however, occupy different
vector spaces because the rotation operator depends on time, namely R R(t ) . The base point
is moving in process of time. Vector quantities of this kind may be called spin vectors. Spin
vectors are rather tricky in numerical sense as they always occupy a distinct vector space on a
manifold. Correspondingly, the spatial spin vectors are R , R , R spatTR .
Angular velocity vectors and the time derivative of total rotation vectors are related by,
see (20-21)
R T( I ) , T , for material description,
where R matTR ,
I I I mat I
(26)
R T T ( I ) I where R spatTR , I , I spatTI , for spatial description,
where the tangential transformation depends on the total rotation vector, and the rotation
~ ~ ) . Similar expression for the angular acceleration vector can
operator is R exp(I ) exp( I
+ T
R T where R matTR , I , , T for material description,
I I I I mat I
(27)
R T
T
I + T I where R spatTR , I , I ,
T
I spatTI for spatial description .
~ dR
R : R T R TR
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
ds ,
whose axial vector is called the material curvature vector R (majuscule of kappa-letter). A
spatial curvature tensor ~ is defined, respectively
~ : R RT
R
,
where prime denotes the derivative with respect to the length parameter s.
Material and spatial curvature tensors and corresponding curvature vectors are related by
~ Ad ~
R R R R R R ,
T
R ad R R R R , (28)
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
26 Jari Mäkinen
where the prime denotes the derivative with respect to the length parameter s , and R is the
base point.
h(t, x) 0 ,
where the arguments are time t and the generalized place vector x(t ) En . Constraint
equations are assumed to be an independent set of equations.
A constraint equation, which is impossible to present in a holonomic form, Def. 33, is
nonholonomic, i.e. it is not and cannot be integrated into a holonomic form. It is clear that we
cannot describe nonholonomic constraint equations, they are just kinematic relations, which
are not holonomic. Different kinematic relations are shown in
Figure 9, where geometric constraints include all holonomic and the so called unilateral
constraints, which are given by inequality equations with the function of time and a
13
Hilbert space is a complete inner-product space, and here especially a complete infinite-dimensional inner-
product vector-valued function space, see Hilbert spaces e.g. in [Debnath & Mikusinski 1990]
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 27
generalized place vector only. Unilateral constraints arise when modeling a kinematic relation
between bodies in a contact. Especially in multibody systems, a joint clearance (play) and
collision problems may be modeled by a contact formulation.
Kinematic
Nonholonomic
Holonomic Geometric
h( t , x ) 0 h( t , x ) 0
o b g
M := t x R En h t , x 0 En d t.
The constraint point-manifold M is a d-dimensional smooth manifold with time as 1-
parameter family. The constraint manifold at fixed time t t0 is denoted by M t0 .
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
28 Jari Mäkinen
{ b g
TM t0 := (x, x ) En En x Mt0 , Dx h t0 , x x 0, Dx h is surjection },
where Dx h is a Fréchet partial derivative of holonomic constraints with respect to x at
t t0 . The definition limits the singular (nonregular) points of the constraint manifold out by
demanding the derivative of the constraints is a surjective (onto) mapping.
This surjectivity request yields that dimensionality do not vary in the manifold that has
importance when accomplishing the finite element method, the constraint manifold has a
fixed dimensional independent on constraints. It can be proven, see e.g. [Rheinboldt 1986; p.
44-45], that the null-space of Dx h , denoted ker Dx h (kernel), is equal to the tangent space.
Thus, TM t0 is indeed a tangent bundle.
A vector bundle establishes a tangent vector space for each regular point of the manifold
M at the fixed time t0 . For practical reasons, we need a tangent space that is an element of
the tangent bundle.
{
Tx 0 M := x En ( x 0 , x ) TM t0 }.
Then we may denote for any virtual displacement vector x Tx0 M , where the base
point x 0 is included in the notation as a subscript.
A geometric interpretation of tangent space and its element virtual displacement have
been illustrated with the holonomic and nonholonomic cases in Fig.10..
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
nonholonomic
tangent space Tx 0 M
constraint atx 0
geometric constraint
manifold (with boundary)
x
virtual displacement with
nonholonomic constraint
boundary of constraint
x
virtual displacement without
nonholonomic constraint
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 29
In this Section, we give definitions for Gâteaux and Fréchet differentials, and for a
constraint manifold modeled in infinite-dimensional Hilbert spaces, called field-manifolds14
here. We assume that any operator we consider is Fréchet differentiable, which is stronger
than Gâteaux differentiable. Hence, we use Gâteaux differential for more useful and simple
way to calculate a Fréchet differential or derivative. We note that if an operator is Fréchet
differentiable then its Fréchet and Gâteaux differential are equal, see more details in [Oden &
Reddy 1976; Ch 2].
In following definition, we consider an operator f:X H1 H2 , later called a vector,
from a set X of the Hilbert space H1 into the Hilbert space H 2 . The vector f is a general
vector-valued nonlinear mapping between function spaces. We also assume that the vector
f(x) is Fréchet differentiable, i.e. it has a unique Fréchet derivative.
r(x, u)
0 . Df (x) u is called Fréchet
H2
where the remainder obeys the condition lim
u 0 u H1
differential.
A vector is called Fréchet differentiable if its Fréchet derivative exists. This derivative is
also a linearized form, or better its affine form with together f(x), for a nonlinear vector
f (x u) at x. Def. 37 is rather simple but not so practical way to calculate Fréchet derivative,
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
hence we define Gâteaux differential for a more practical formula to calculate Fréchet
differential and derivative for Fréchet differentiable vector.
D f ( x ) u : lim
f (x u) f (x )
b
d f x u g
0 d 0
,
where the limit is to be interpreted in the norm of H 2 . The later formula is a practical and
simple way to compute the directional derivative that is the term Df (x ) u where u H 1
indicates direction.
14
We use the name (infinite-dimensional) field-manifold contrast to a (finite-dimensional) point-manifold.
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
30 Jari Mäkinen
We have assumed that X is a set, but next we give a more structure. We denote this
structured set as C .
o b g
C : t x R H 1 h t, x 0 H 2 t,
where h(t, x) indicates an independent set of the holonomic constraint equations. A constraint
manifold at a fixed time t t0 is denoted by Ct0 .
Compare the constraint field-manifold with the constraint point-manifold defined in Def.
34. A similar way as in the point-wise case, we could define a tangent field-space which is a
space of vector-valued functions
{ b g
TCt0 := ( x, x ) H1 H1 x Ct0 , Dx h t0 , x x 0, Dx h is surjection },
where Dx h is the Fréchet partial derivative of the holonomic constraints with respect to x at
t t0 . The definition limits the nonsmooth isolated points of the constraint manifold out by
demanding the existence of the Fréchet derivative.
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
{
Tx 0C := x H 1 ( x 0 , x ) TCt0 },
where the tangent field-bundle is defined in Def. 40. We may denote any virtual displacement
field x Tx0C , where the base x 0 is included in the notation as a subscript.
Note that the place field x is a vector-valued function satisfying all the constraints
equations (holonomic), called the constraint field-manifold C .
15
a vector-valued function, more precisely
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Manifolds on Continuum Mechanics 31
n
H 1 ( x, x
TxC := x ) TC s,
where now time is free, not fixed, like in the virtual displacement. Compare with Def. 9, a
tangent vector in a finite-dimensional case. The velocity field that is an element of the
velocity field-space is also denoted by v: x TxC
In this Section, we give definition for the variation, Lie derivative and Lie variation. The
concept of push-forward and pull-back operators is essential for understanding Lie derivatives
and variations.
b g b g b g
h t0 , x, v : Dxh t0 , x, v x Dv h t0 , x, v v
,
where x Ct0 is a place field, x Tx0C is a virtual displacement field, v TxC is a velocity
field, and v: x Tx0C is a virtual velocity field. Moreover, Dx , Dv are Fréchet partial
derivatives with respect to place and velocity, correspondingly.
The variation operator depends linearly on the virtual displacement and the virtual
velocity. Note a minor notational difference between the virtual and variation operators,
and . Calculating the place and velocity variation, after Def. 43, yields
Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.
x x and v v . (30)
This should be interpreted: the variation of place vector x is equal, not the same thing,
as the virtual displacement x . The variation has an operational meaning whereas the virtual
displacement is a geometrical quantity. In generally, the variation of ‗something‘ and the
virtual ‗something‘ are not equal, e.g. a virtual work may exists although there does not exist
a work function at all and neither the work variation.
44 Theorem
Generally, the variation operator and the time derivative operator do not commutate.
Proof: We will prove this theorem by a counter example. Let consider the constraint
equation x t y 0 , where t represents time. Its variation is x t y 0 . On the other hand,
the virtual displacement of the constraint equation is x t y 0 x t y 0 , whose time
derivative is respectively
Continuum Mechanics, edited by Andrus Koppel, and Jaak Oja, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,
Exploring the Variety of Random
Documents with Different Content
Nas naves lateraes, os altares são divididos por pequenos
zimborios. Grande cupula central.
O templo é em fórma de cruz latina.
Commercio e Industria
Foi em 1855 que se effectuou o primeiro censo commercial, em
Buenos-Aires, e o ultimo em 1904. N’aquella data existiam 3:139
estabelecimentos commerciaes, divididos em 9 categorias, e n’esta
época, 17:985, assim especificados:
—Alimentação 9:358
—Construcções 426
—Vestidos e chapéus de senhora 2:465
—Moveis e annexos 289
—Transporte e annexos 903
—Artisticos, adornos e recreio 638
—Tecidos, couros e pelles 56
—Sanidade e chimica 370
—Artes Graphicas e annexas 158
—Depositos e vendas varias 1:169
—Varios negocios 2:153
17:985
Activo
Acções da Empreza do Edificio 49:362
Bancos e Caixa 40:290:72
Credito Argentino Interno 73:186:44
Immovel 756:176:47
Moveis e utensilios 18:734:44
Reparações no edificio 9:593:20
Recibos de subscripção 5:450
Terrenos no Tigre 6:000
Varios 1:210
960:003:27
Passivo
Capital da Empreza do Edificio 770:000
» » Associação 61:162:92
Fundo de Reserva da Empreza 50:525:73
» » » » Associação 16:340:74
» especial para calamidades publicas 8:076:92
Fundo de jubilações de empregados 2:905:34
» » beneficencia 609:07
Dividendos a pagar 2:049
» de 1907, 2.º semestre 42:520
Varios 6:083:55
960:003:27
Jardins e Parques
Parque Tres de Fevereiro—O seu nome popular é Parque de
Palermo. Occupa a superficie de 3:677:487 metros quadrados. A
avenida principal, chamada das Palmeiras, lembra, com saudades, as
suas homonymas do Rio de Janeiro, altas e esbeltas, ao passo que
estas não attingem um terço da altura d’aquellas. A vegetação geral
é rachitica, recordando a densa e frondosa vitalidade e pujança do
Bosque de Bologna, a que os bonaerenses comparam o Parque Tres
de Fevereiro.
As avenidas são amplas, bem tratadas e amenisadas por varios
lagos e cursos de agua. No recinto e na orla d’este amplissimo local
de recreio estão o Jardim Zoologico, o Velodromo Nacional, o
Hippodromo Nacional, a Escola Normal de Tiro, o Club Hippico
Argentino, os restaurantes dos Lagos e de Palermo, o Tiro Federal
Argentino, o Cricket Club e muitas outras installações sportivas.
Decoram tambem este magnifico parque, as estatuas de
Domingos Sarmiento, Carlos Burmeister e de Eduardo Costa.
A 1.ª é do esculptor Rodin, e foi inaugurada em 25 de Maio de
1900. A figura está de pé sobre uma base de marmore branco, que
ostenta artistica allegoria, representando Apollo, o deus da Luz, a
espancar as trevas da ignorancia, symbolisadas pela serpente
Python.
O monumento de Burmeister, foi inaugurado em Outubro de 1900,
e é obra do esculptor allemão Ricardo Aigner. O do doutor Eduardo
Costa, data de 19 de Março de 1902, e não é particularmente
notavel como obra de arte.
No recinto d’este parque teem funccionado varias e
importantissimas exposições da Sociedade Rural Argentina,
patrocinada pelo governo. As terças-feiras e sabbados são os dias da
moda para este parque, frequentado então por milhares de
vehiculos e de peões. A melhor construcção architectonica de todo o
recinto é a do Tiro Federal Argentino, com bellissima frontaria
ladeada por duas altas e elegantes torres.
Nos hippodromos de Buenos-Aires, as apostas ascenderam, em
1906, á somma de 47:218:602 pesos.
Theatros
São 16 os theatros principaes da metropole argentina, em todos
os generos e frequentados, especialmente de Maio a Setembro, por
companhias de primeira ordem, nacionaes e estrangeiras.
Do esplendido theatro Colon, já tratamos no capitulo
Monumentos. Os mais importantes são:
Curiosidades
Palacio do Governo—Visto do porto, este edificio mostra um
aspecto grandioso e imponente, no conjuncto, com os seus quatro
pavimentos, terraços, balcões e um grupo de gigantescas estatuas a
corôal-o, obra do esculptor Bianchi. Observado da praça de Mayo,
para onde olha a frontaria principal, a perspectiva é muito reduzida,
especialmente pela irregularidade da architectura, sem estylo
definido, antes demonstrativa de que pertence a epochas diversas.
No andar nobre estão installados os ministerios das Obras Publicas,
da Justiça e Instrucção Publica, da Fazenda, das Relações Exteriores,
do Interior e a Presidencia da Republica. Todas as installações
constam, pouco mais ou menos, do mesmo numero de peças, umas
doze para cada ministerio, comprehendendo as salas de espera e de
recepção, os escriptorios e o gabinete do ministro. No Ministerio das
Relações Exteriores ha, a maior, o salão de recepção dos diplomatas,
decorado com simplicidade. A Presidencia da Republica, compõe-se
de sala de espera, gabinete do Presidente, com o retrato, a oleo, de
Rivadavia, saleta de conversação e saleta de fumo. Seguem-se as
secretarias dos officiaes militares e civis. O grupo dos tres salões
principaes do edificio é accessivel por duplo lance de escadaria de
marmore branco. O primeiro é o de recepções simples, o do centro é
o salão Branco, ou das grandes recepções officiaes, e o terceiro é o
das reuniões do Conselho do Governo, presidido pelo Presidente da
Republica. Este salão não tem decoração alguma. Apenas se vê, ao
centro, a mesa, forrada de velludo vermelho-escuro e nove cadeiras
da mesma fazenda. Do tecto pende um lustre. O salão Branco
encerra um soberbo e artistico lustre de metal dourado, com 280
lampadas electricas, trabalho argentino de Azaretto. O recinto é
ainda illuminado por 24 candieiros dourados, a 11 lampadas cada
um. Ha uma galeria circular para o publico. Sobre pedestaes de
marmore preto, circumdam o salão os bustos, em marmore branco,
da Republica, do general Roca, do general Sarmiento, do dr.
Quintana, de Rivadavia, de Bartholomeu Mitre, de Urquiza, e de
outros vultos importantes da Republica. Do general Roca ha outro
busto, sobre niveo pedestal, com lettras douradas.
Nos outros pavimentos estão os escriptorios de serviço dos
ministerios e um terraço com habitações de empregados e varias
officinas.
D’ahi gosa-se vistas parciaes da cidade e o panorama geral do
porto. Este palacio mede 125 metros de frente por 81 de fundo.
Das pequenas bibliothecas e dos archivos dos varios ministerios,
avultam os do Ministerio das Relações Exteriores, pela quantidade e
importancia. Em arcas de ferro, estão guardados os tratados
celebrados com as nações estrangeiras, desde 1811, sendo alguns
d’elles verdadeiras obras de arte.
Alimentação Publica
Matadouro de Liniers—No extremo-oeste do municipio de Buenos-
Aires, nos terrenos conhecidos por Nova Chicago, está situado este
novo e colossal matadouro, concluido em 1900. As suas installações
são vastas e numerosas, occupando uma area total de 319:773
metros quadrados.
Comprehende 44 armazens de mercado de compra, venda e peso
de gado e 96 matadouros com espaço para trinta mil animaes.
Todos os pavimentos são de asphalto e os tectos de telha. Ha
tambem uma grande casa de machinas e de bombas; o edificio de
residencia e trabalho dos veterenarios, um pavilhão com forno
crematorio para os animaes condemnados e o grande edificio da
Administração.
Desde 1900 a 1907, a media annual para o consumo de Buenos-
Aires, é de 382:000 bois e vaccas abatidas, 108:000 vitellos,
585:000 carneiros e 55:000 porcos.
Em 1906 os frigorificos forneceram 17 milhões de kilos de carne
congelada, para a alimentação da cidade, que tambem consumiu no
mesmo periodo:
ebookgate.com