Experience Seamless Full Ebook Downloads for Every Genre at textbookfull.

com

Dynamics and Control of Advanced Structures and

Machines 1st Edition Hans Irschik

https://textbookfull.com/product/dynamics-and-control-of-
advanced-structures-and-machines-1st-edition-hans-irschik/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Advanced Flight Dynamics with Elements of Flight Control

1st Edition Nandan K. Sinha

https://textbookfull.com/product/advanced-flight-dynamics-with-
elements-of-flight-control-1st-edition-nandan-k-sinha/

textboxfull.com

Dynamics and Control of Industrial Cranes Keum-Shik Hong

https://textbookfull.com/product/dynamics-and-control-of-industrial-
cranes-keum-shik-hong/

textboxfull.com

Mathematical Structures of Ergodicity and Chaos in

Population Dynamics 1st Edition Pawe■ J. Mitkowski

https://textbookfull.com/product/mathematical-structures-of-
ergodicity-and-chaos-in-population-dynamics-1st-edition-pawel-j-
mitkowski/
textboxfull.com

Advanced Data Structures Theory and Applications 1st

Edition Suman Saha

https://textbookfull.com/product/advanced-data-structures-theory-and-
applications-1st-edition-suman-saha/

textboxfull.com
Control and dynamics in power systems and microgrids 1st
Edition Fan

https://textbookfull.com/product/control-and-dynamics-in-power-
systems-and-microgrids-1st-edition-fan/

textboxfull.com

Control Oriented Modelling of AC Electric Machines

SpringerBriefs in Electrical and Computer Engineering
Masmoudi
https://textbookfull.com/product/control-oriented-modelling-of-ac-
electric-machines-springerbriefs-in-electrical-and-computer-
engineering-masmoudi/
textboxfull.com

Shell-like Structures: Advanced Theories and Applications

1st Edition Holm Altenbach

https://textbookfull.com/product/shell-like-structures-advanced-
theories-and-applications-1st-edition-holm-altenbach/

textboxfull.com

Induction Machines Handbook: Transients, Control

Principles, Design and Testing 3rd Edition Ion Boldea

https://textbookfull.com/product/induction-machines-handbook-
transients-control-principles-design-and-testing-3rd-edition-ion-
boldea/
textboxfull.com

Dynamics of Coupled Structures Volume 4 Proceedings of the

34th IMAC A Conference and Exposition on Structural
Dynamics 2016 1st Edition Matt Allen
https://textbookfull.com/product/dynamics-of-coupled-structures-
volume-4-proceedings-of-the-34th-imac-a-conference-and-exposition-on-
structural-dynamics-2016-1st-edition-matt-allen/
textboxfull.com
Hans Irschik · Alexander Belyaev
Michael Krommer Editors

Dynamics
and Control
of Advanced
Structures and
Machines
Dynamics and Control of Advanced Structures
and Machines
Hans Irschik • Alexander Belyaev •
Michael Krommer
Editors

Dynamics and Control

of Advanced Structures
and Machines

123
Editors
Hans Irschik Alexander Belyaev
Institute of Technical Mechanics Institute of Problems in Mechanical
Johannes Kepler University Linz Engineering
Linz, Austria Russian Academy of Sciences
St. Petersburg, Russia
Michael Krommer
Institute of Mechanics and Mechatronics
Vienna University of Technology
Vienna, Austria

ISBN 978-3-319-43079-9 ISBN 978-3-319-43080-5 (eBook)

DOI 10.1007/978-3-319-43080-5

Library of Congress Control Number: 2016955567

© Springer International Publishing Switzerland 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
Preface

This book presents a collection of 25 contributions presented during the 2nd

International Workshop on Advanced Dynamics and Model Based Control of
Structures and Machines, which was held at TU Wien in September 2015. It
contains 9 full-length papers of presentations from Austria, 7 from Russia, 4 from
Japan, 3 from Italy and 2 from Taiwan.
The general goal of the workshop was to present and discuss the frontiers in
the mechanics of controlled machines and structures. The workshop continued a
series of international workshops, the Japan-Austria Joint Workshop on Mechanics
and Model Based Control of Smart Materials and Structures, the Russia-Austria
Joint Workshop on Advanced Dynamics and Model Based Control of Structures
and Machines and the 1st International Workshop on Advanced Dynamics and
Model Based Control of Structures and Machines. The first two workshops took
place in Linz, Austria, in September 2008 and April 2010, and the third one
in St. Petersburg, Russia, in July 2012. The key objective of the workshop was
to further strengthen the long-standing cooperation between research teams from
Austria, Japan and Russia and to initiate new collaborations with other participating
renowned scientists from Europe and Taiwan.
We dedicate the book to Franz Ziegler who passed away on January 4th, 2016.
Franz Ziegler delivered his last scientific talk on “Free and forced vibrations of fuzzy
structures” at the workshop. This contribution, which is included in the present
volume, brings his long list of numerous papers published in the most prestigious
peer-reviewed journals, books, and conference proceedings to an end.
We, the undersigned Editors, together with all his other friends and colleagues
all over the world, will hold Franz Ziegler in grateful memory.

St. Petersburg, Russia Alexander Belyaev

Linz, Austria Hans Irschik
Vienna, Austria Michael Krommer
June 2016

v
Acknowledgements

Support of the 2nd International Workshop on Advanced Dynamics and Model

Based Control of Structures and Machines from the K2 area of the Linz Center
of Mechatronics GmbH is gratefully acknowledged. This area is promoted as a
K2 project with the project name Austrian Competence Center of Mechatronics
(ACCM) in the context of Competence Centers for Excellent Technologies (COMET)
by BMVIT, BMWFJ and by the country Upper Austria.
The editors also wish to thank the Institute of Mechanics and Mechatronics at
TU Wien for serving as host of the workshop and Mrs. Silvia Schilgerius from
SpringerWienNewYork for her support during the preparation of this book.

vii
Contents

Free and Forced Vibrations of Fuzzy Structures . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 1

Franz Ziegler
Resonance of a Series of Train Cars Traveling Over Multi-Span
Continuous Beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 11
J.D. Yau and Y.B. Yang
Stability and Supercritical Deformation of a Circular Ring
with Intrinsic Curvature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 23
Yury Vetyukov
Estimation of Mechanical Properties of Micro-Lattice Panel
with Irregular Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 33
Kuniharu Ushijima and Dai-Heng Chen
Overview Reading and Comparing the Seismic Proof
Capability of Displacement Dependent Semi-Active Hydraulic
Damper and Accumulated Semi-Active Hydraulic Damper . . . . . . . . . . . . . . . . 43
M.-H. Shih and W.-P. Sung
On Necessary and Sufficient Conditions for Eigenstrain-Type
Control of Stresses in the Dynamics of Force-Loaded Elastic Bodies . . . . . . 53
Juergen Schoeftner and Hans Irschik
Variational Principles for Different Representations of
Lagrangian and Hamiltonian Systems . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 65
Markus Schöberl and Kurt Schlacher
Hardware-in-the-Loop Testing of High-Speed Pantographs
Using Real-Time Catenary Emulation . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 75
Alexander Schirrer, Guilherme Aschauer, and Stefan Jakubek
Swelling-Induced Bending of Hydrogel Bistrips . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 85
Takuya Morimoto, Fumihiro Ashida, and Yu Hayashi

ix
x Contents

Determination of Parameters of the External Electric Circuits

Providing Maximum Damping of Vibrations of Electroelastic Bodies . . . . . 93
V.P. Matveenko, N.V. Sevodina, N.A. Yurlova, D.A. Oshmarin,
M.A. Yurlov, and A.S. Ivanov
Forming of Woven-Reinforced Thermoplastic-Matrix
Composites: Characterization, Modelling, and Validation .. . . . . . . . . . . . . . . . . 101
Martín Machado and Zoltan Major
Model Predictive Temperature Control of a Distribution
System for Chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 109
S. Koch, M. Ponikvar, M. Steinberger, and M. Horn
Hidden Oscillations in Electromechanical Systems . . . . . .. . . . . . . . . . . . . . . . . . . . 119
Maria Kiseleva, Natalya Kondratyeva, Nikolay Kuznetsov,
and Gennady Leonov
Effect of Material Layers in a Compound Circular Receiver
Model Design for Concentrating Solar Power . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 125
Ryuusuke Kawamura, Yoshinori Nagase, and Shigeki Tomomatsu
On Multiple Support Excitation Analysis of Bridges . . . .. . . . . . . . . . . . . . . . . . . . 137
R. Heuer and D. Watzl
Control of Friction by Surface Microgeometry Variation .. . . . . . . . . . . . . . . . . . 145
Irina Goryacheva
Peculiarities of the Magnetic Behavior of Pipe Steels with
Different Initial Stress–Strain States Under Elastic Deformation . . . . . . . . . . 153
E.S. Gorkunov, A.M. Povolotskaya, S.M. Zadvorkin, and
Yu.V. Subachev
Cable-Stayed Bridges: A Monitoring Challenge.. . . . . . . . .. . . . . . . . . . . . . . . . . . . . 161
L. Faravelli
Dynamics and Control of Motion for Systems Containing
Internal Moving Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 169
F.L. Chernousko
Elaborations from the TKB Monitoring Database . . . . . . .. . . . . . . . . . . . . . . . . . . . 179
S. Casciati and L. Elia
Reduced Order Models and Localized Nonlinearity:
An Approach to the Design of Meta-Structures . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 187
F. Casciati
Contact of Flexible Elastic Belt with Two Pulleys . . . . . . . .. . . . . . . . . . . . . . . . . . . . 195
A.K. Belyaev, V.V. Eliseev, H. Irschik, and E.A. Oborin
Control over Internet of Oscillations for Group of Pendulums . . . . . . . . . . . . . 205
Mikhail S. Ananyevskiy and Alexander L. Fradkov
Contents xi

Effect of the Load Modelling Strategy on the Dynamic

Response Prediction of Bridges Subjected to High-Speed Trains . . . . . . . . . . 215
Christoph Adam and Patrick Salcher
Mechanical Properties of Epoxy Resins Filled with Nano-Silica
Particles . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 225
Tadaharu Adachi, Markus Karamoy Umboh, Tadamasa Nemoto,
Masahiro Higuchi, and Zoltan Major
Free and Forced Vibrations of Fuzzy Structures

Franz Ziegler

Abstract Stochastic free and forced vibrations of layered beams are analyzed that
result from a single bounded random stiffness parameter whose probability density
function is considered to be zero outside of a given interval, i.e., it is a member of
a fuzzy set with bounded uncertainty. The relevant properties of natural vibrations
of an ensemble of sandwich beams with three perfectly bonded layers under hard
hinged support conditions are worked out in detail when a bounded random shear
stiffness of the core material is assigned by employing interval mathematics. The
main structure of a compound single-span railway bridge, effectively modeled as
a two-layer beam, is subjected to a single moving load as well as to a series
of repetitive moving loads traveling with constant speed. It serves as a complex
example for the resulting forced random vibrations and resonances under the severe
condition of an elastic interface slip of bounded random stiffness. In both cases
exact homogenization yields a stochastic sixth-order partial differential equation of
motion of the layered beam. Light modal damping is considered. The analysis of
the illustrative problems is based on the interval representation with a triangular
membership function of the stiffness modulus assigned. A short comment provides
information on the limits of such triangular membership functions. Membership
functions in the form of envelopes of the random natural frequencies, the dynamic
magnification factors, and the phase angles in free vibrations are determined.
Both, fuzzy peak deflection and acceleration are derived for the forced single-
span compound railway bridge subjected to the moving loads. Approximating
superposition of modal maxima is considered by standard routines of reliability
analysis.

Franz Ziegler was deceased at the time of publication.

© Springer International Publishing Switzerland 2017 1

H. Irschik et al. (eds.), Dynamics and Control of Advanced Structures
and Machines, DOI 10.1007/978-3-319-43080-5_1
2 F. Ziegler

1 Introduction

In general, the application of the fuzzy finite element method, see, e.g., [3, 9], is
required when the matrix in composite structures exhibits fuzzy randomness of the
material parameters. However, when extending the work on two- and symmetric
three-layer beam-, plate-, and shell structures based on an exactly homogenized
model [1, 4, 5, 8] to include either fuzzy interface slip or fuzzy core stiffness,
we can avoid numerical analyses and analytically work out the effects on the
random dynamic properties of these fuzzy structures. Exemplarily, random free
vibrations of a symmetric three-layer single-span beam with hard hinged supports,
exhibiting fuzzy shear stiffness of the core material, are considered first, with
details documented in [6, 7]. The practically very important case of a single-span
compound bridge consisting in its main structure of two steel girders connected
(elastically) to the concrete deck, when exactly homogenized, refers to the model
of an asymmetric two-layer elastic beam exhibiting the practically most important
defect of elastic interlayer slip with the major uncertainty of its elastic modulus.
Life loads of a train are here simply modeled as the passage of concentrated forces,
which pass the bridge with constant speed. The presentation of the resulting forced
vibrations in the form of non-dimensional response quantities is adopted from a
recent comprehensive study on the dynamic effects of high-speed trains on such a
simple fuzzy bridge structure [2]. The fundamental parameters are taken from [12].
For a more sophisticated deterministic analysis, see [13]. A full stochastic analysis
of such a bridge with temperature effects on stiffness, etc., included is presented
in [10].
The analysis of both illustrative examples is based on the interval representation
(interval of confidence at a given level of presumption, i.e., ˛-cut) with a triangular
fuzzy membership function of the relevant random stiffness prescribed. Fuzzy
membership functions of the response are defined using fuzzy set theory [11, 14],
however, avoiding artificial uncertainties. Where possible, envelope functions are
defined representing the bounds of the random response. Consequently, such a ran-
dom stiffness modulus is considered to be an interval number or a member of a fuzzy
set which contains the sure design value. Such an interval representation transforms
the deterministic parameter to inclusive set values with bounded uncertainty, see
Fig. 1a. Another interpretation of such an interval number includes intervals of
confidence for ˛-cuts of fuzzy sets. Such a parametric extension as shown in Fig. 1b
may even include a worst-case scenario, i.e., min k2 .˛ D 0/ D 0.
Free and Forced Vibrations of Fuzzy Structures 3

Fig. 1 Single bounded random structural parameter, e.g., stiffness. (a) Defined interval contains
sure design value k0 . Outside of the interval, probability is zero. (b) Fuzzy set: interval of
confidence at a given level of presumption ˛-cut. Non-dimensional stiffness k2 .˛/ is referred to
its design value. Isosceles distribution assigned

2 Effects on Free Vibrations: Example: Fuzzy Sandwich

Beam

For linear elastic, symmetrically arranged three-layer beams, Fig. 2, the following
assumptions are made: (1) the faces are rigid in shear with their individual flexural
stiffness B1 D B3 considered. (2) The shear stiffness of the elastic core is a bounded
random variable while its bending stiffness is neglected. Following [5], an exact
homogenization renders the homogeneous partial differential equation (PDE) of
sixth order in terms of the common deflection and for the case of free vibrations,

3
X
 
w;xxxxxx  2 w;xxxx C wR ;xx  2 wR D 0; D %i hi ; (1)
B1 B1 iD1

B0 D B1 C B3 ;
h i
B1 D B0 C .D1 C D3 /d2 ; B0 =B1 D 1= 1 C 3 .1 C h2 =h1 /2  1=4;

see again Fig. 2, with the bounded random parameter

  related to the core shear
modulus 22 G2 .˛/ with its assigned design value 22 G2 0 at ˛ D 1 within the level
of presumption 0  ˛  1,

2b B1  2     
2 D  G2 0 k2 .˛/; k2 .˛/ D 22 G2 .˛/ = 22 G2 0 : (2)
h 2 D1 B 0 2

For the single-span beam, length l, under hard hinged support conditions, the
ortho-normalized mode shapes result with a stochastic normalization factor,
  1=2
'n .x/ D sin ˇ1n x; ˇ1n D n=l; An D .l=2B1/ 2 C ˇ1n
2
B1 =B0 :
(3)
4 F. Ziegler

Fig. 2 Dimensions of the sandwich beam. Common cross-sectional rotation after homogenization
, as shown, is crucially eliminated. Fuzzy core shear stiffness in layer 2 is considered. Deflection
w is referred to the elastic centroid O

The random natural frequencies, referred to the sure values at rigid core shear
2 4
stiffness, !n1 D ˇ1n B1 =, are

.B0 =B1 / C 2;n k2 .˛/  

Œ!n .˛/=!n1 2 D ; 2;n D 22 G2 0 2b=ˇ1n
2
D1 h 2 : (4)
1 C 2;n k2 .˛/

With light viscous modal damping n D   1 understood, the random dynamic

magnification factor n .˛/ and the stochastic phase angle n .˛/ are expressed by
textbook formulas; see, e.g., [15],
n   o1=2
n D 1  2 1  2 2 .!=!n .˛//2 C .!=!n .˛//4 ;
"s  2 s #1
! .B0 =B1 / C 2;n k2 .˛/ ! 1 C 2;n k2 .˛/
tan n D2  :
!n1 1 C 2;n k2 .˛/ !n1 .B0 =B1 / C 2;n k2 .˛/
(5)

Numerical results for bounds and even for envelope functions are thus straight-
forwardly derived, putting, e.g., the design values B0 =B1 D 0:1 < 1=4 and
2;n D 0:25 in Eqs. (4) and (5). Exemplarily, results are shown in Figs. 3, 4, and 5,
for light modal damping n D  D 0:04. More details are presented in [6, 7].
Free and Forced Vibrations of Fuzzy Structures 5

Fig. 3 Bounds of the first five natural frequencies when referred to the sure fundamental
frequency: worst-case scenario. Note the increased fuzziness of the higher modes [7]

Fig. 4 Envelopes of the first three random DMFs [7]. ˛ D 0-cut; worst-case scenario ˛ D 0

Fig. 5 Envelopes of the first three random phase angles [7]. Conditions likewise to Fig. 4
6 F. Ziegler

2.1 Effects of Non-symmetric Uncertainty

In the core shear stiffness, Fig. 2, we introduce a non-symmetry factor > 1, such
that min k2 .˛/ D ˛ remains unchanged and max k2 .˛/ D 1 C .1  ˛/. When
considering the bounds of the natural frequencies just in the worst-case scenario,
the possibility of max !n D min !nC1 becomes true for n D 2 at the minimum
of D 12:16, i.e., overlapping natural frequency intervals occur in an ensemble
of fuzzy sandwich beams. Further, for the material parameters B0 =B1 D 1=16
and B0 =B1 D 0:198 < 1=4, the fundamental frequency and the next higher one,
respectively, exhibit singularities, i.e., the non-symmetric triangular distribution of
uncertainty becomes obsolete. However, no singularity is possible for higher modes,
n  3. For details of analysis, see again [7].

3 Effects on Forced Vibrations: Example: Fuzzy Compound

Railway Bridge

The practically very important case of a single-span compound bridge consisting,

e.g., of two steel girders connected (elastically) to the concrete deck, when smeared,
refers to the model of an asymmetric two-layer elastic beam exhibiting the main
defect of elastic interlayer slip with the major uncertainty of its elastic modulus,
Fig. 6. This structural uncertainty of the physical interface can be attributed to
imperfections, modeling inaccuracies, aging effects, and its design complexity.
Consequently, the slip modulus is considered to be a random variable whose

Fig. 6 Model of the main structure of the fuzzy compound bridge [2]: effective elastic two-layer
beam with fuzzy elastic interlayer slip. Layer one refers to the RC-concrete deck. Layer two
represents the rigidity of two steel girders. Bounded random shear flow T D k.˛/ u. Stiffness
center S
Free and Forced Vibrations of Fuzzy Structures 7

probability density function is not known but is non-zero only in the range of an
interval, i.e., it is considered to be an interval number or a member of a fuzzy set, see
again Fig. 1. Exact homogenization of the two-layer beam renders the PDE of sixth
order, analogously to Eq. (1), see again [1]. Consequently, in the design stage, the
simply supported bridge of length l may be simply traveled by a series of repetitive
single concentrated forces with equal distance ld and with constant speed v to model
the forcing in the inhomogeneous Eq. (6), where the random parameters are adapted
to the fuzzy elastic modulus of the interlayer slip k.˛/ with the sure design value
k0 D k.˛ D 1/. The parameters of the two elastic layers j D 1; 2, are: elastic
modulus Ej , effective cross-sectional area Aj , and flexural rigidity Bj D Dj rj2 with
Dj D E j A j ,

  1
w;xxxxxx  2 w;xxxx C wR ;xx  2 p C p;xx ; 2 .˛/ D k.˛/20 =k0 ;
B0 B1 B0
 
2 1 1 d2
B0 D B1 C B2 ; B1 D B0 C d D1 D2 =.D1 C D2 /; 20 D k0 C C :
D1 D2 B0
(6)

Considering just a single concentrated traveling force Fi in Eq. (6), where

w.x; t/ ) wi .x; t/, the lateral load becomes in standard notations [2],

  si si C l
pi D Fi ı.x  i / H.t  ti0 /  H.t  tiE / ; i D vt  si ; ti0 D ; tiE D :
v v
(7)
A properly truncated modal expansion, Eq. (3), with the definitions given in Eq. (6)
applies as well,

X
N
.n/
wi .x; t/ D Yi .t/'n .x/: (8)
iD1

.n/ .n/
By changing the ortho-normalized modal coordinates to Y y .tI ˛/ D An .˛/Yi .t/
the random load in Eq. (6) is “swallowed” and the modal equations result, crucially
.n/
simplified, with a deterministic load participation factor Li . Only the natural
frequencies are left as the remaining bounded random coefficients in Eq. (9), light
modal damping is also added here [2],

.n/ .n/ .n/ 2Fi .n/

YR i C 2n !n YP i C !n2 Y i D L ;
l i
.n/
Li D ŒH.t  si =v/  H.t  .si C l/=v/ sin n.vt  si /=l: (9)

.n/ .n/
Bounds of the modal coordinate Y i and its rate YP i are determined by proper
.n/
time convolution and, subsequently, the corresponding modal acceleration YR i is
8 F. Ziegler

determined from a rearranged form of Eq. (9). The bounds of the natural frequencies
at ˛-cuts are inserted. Thus, for a number M of repetitive single concentrated forces,
e.g., lateral displacement results by summation

X
M
.n/
wn .x; tI ˛/ D Y n .tI ˛/ n .x/ ) Y n .tI ˛/ D Y i .tI ˛/: (10)
iD1

In reliability analysis it suffices to evaluate the maximum response searched for by

considering either the minimum or maximum bound of the interface stiffness in
an ˛-cut, e.g., just in the worst-case scenario. It can be found by various methods
of approximating superposition of a proper finite number of these random modal
coordinates. The following combination rules, common in structural dynamics are
evaluated, the absolute sum rule, ABSUM, rendering an upper limit and the square
root of the sum of squares, SRSS, the latter was found to be inadequate here, see
again [2], and, best suited, the modal series, since it is exact for N modes,
ˇ N ˇ
ˇX ˇ
ˇ ˇ
wmax .xI ˛/ D max ˇ wn .x; tI ˛/ˇ : (11)
t ˇ ˇ
nD1

Resonances are observed at critical speeds due to the rhythmic repetition in the
.n/
load series [12]: for the deterministic bridge Vj D !n ld =2j, j D 1; 2; 3; : : : A
resonance of second order is due to a single traveling force, V n D !n l=n, it is of
minor importance here, see, e.g., [15, p. 626]. For a detailed analysis, see again [2].

3.1 Some Numerical Results for a Standard Compound

Bridge: Concrete Deck, Two Steel Girders

The single span has the length l D 40:0 m and a mass density of  D 15;000 kg m1.
The design value of the elastic interface slip modulus is k0 D 60:0  107 N/m2 .
We note the fundamental frequency f1 .˛ D 1/ D 2:81 Hz and, for the ideal rigid
bond, f1;1 D 3:04 Hz [2]. At first, the intervals of uncertainty of the first ten natural
frequencies are evaluated for the bounded random interface stiffness k.˛/ , rendering
qualitatively analogous results as depicted in Fig. 3,

B0 =B1 C 2;n k.˛/=k0 ˇ 4 B1 k0 1 C D2 =D1

!n2 .˛/ D !n1
2 2
; !n1 D 1n ; 2;n D 2 :
1 C 2;n k.˛/=k0  ˇ1;n D2
(12)
In a second step, the randomly fuzzy response of the bridge is calculated when
a single concentrated force passes with constant speed. Maximum deflection and
maximum acceleration, say at mid-span and at the quarter-point, are calculated and
Free and Forced Vibrations of Fuzzy Structures 9

(a)

(b)

Fig. 7 Uncertain peak lateral deflection induced by ten repetitive moving loads, distance ld D
20 m, speed parameter S, limiting cases for two fuzzy intervals of the random interface stiffness,
[2]. (a) at mid-span, (b) at the quarter-point

stored as functions of the speed parameter S D v=˝1 l, where ˝1 D !1 .˛ D 1/.

Maximum deflections induced by ten repetitive moving loads are depicted in Fig. 7,
where w0 .xI ˛/ D wmax .xI ˛/B1 =Fl3 . For details, see again [2].
In Fig. 7 it is observed that the upper limit of a certain fuzzy interval leads to
a larger deviation of the response from that of the deterministic bridge than for the
lower limit. That means, the softer the interlayer-slip-modulus the larger the increase
of the response magnification becomes. An increase of the interlayer stiffness does
not have this pronounced impact on the decrease of the peak deflection.
Finally it is mentioned again that a comprehensive stochastic analysis of a full
model of such a railway bridge is performed in [10].
10 F. Ziegler

4 Conclusions

For fuzzy symmetric three-layer structures (sandwich beams) and for fuzzy two-
layer composites (compound railway bridge) the exact homogenization yields a
stochastic partial differential equation of sixth order. Intervals of confidence of the
dynamic response are determined either for fuzzy core shear stiffness or for fuzzy
elastic interlayer slip. Thus, intervals of confidence of the natural frequencies result.
Bounds of the dynamic magnification factor and the phase angles are presented
in the form of envelope functions. Effects on resonances in repetitive loadings of
the fuzzy railway bridge are studied within reliability measures of the maximum
response; say of flexural displacements, presenting bounds as function of the speed
parameter. Even the worst-case scenarios can be considered. Truncated modal
expansions are transformed such that the modal load participation factors become
deterministic leaving only the natural frequencies as the random parameters.

References

1. Adam C, Heuer R, Jeschko A (1997) Flexural vibrations of elastic composite beams with
interlayer slip. Acta Mech 125:17–30
2. Adam C, Heuer R, Ziegler F (2012) Reliable dynamic analysis of an uncertain compound
bridge under traffic loads. Acta Mech 223:1567–1581
3. Hanss M, Willner K (2000) A fuzzy arithmetical approach to the solution of finite element
problems with uncertain parameters. Mech Res Commun 27:257–272
4. Heuer R (2004) Equivalence of the analyses of sandwich beams with or without interlayer slip.
Mech Adv Mater Struct 11:425–432
5. Heuer R (2014) On equivalences in the dynamic analysis of layered structures. In: Belyaev AK,
Irschik H, Krommer M (eds) Mechanics and model-based control of advanced engineering
systems. Springer, Heidelberg, pp 155–162. ISBN:978-3-7091-1570-1
6. Heuer R, Ziegler F (2011) Vibrations of layered structures with fuzzy core stiffness/fuzzy
interlayer slip. In: Belyaev AK, Langley RS (eds) Proceedings of the IUTAM-symposium on
the vibration analysis of structures with uncertainties. Springer, Dordrecht, pp 29–42
7. Heuer R, Ziegler F (2011) Modal analysis of laminated beams with fuzzy core stiffness/fuzzy
interlayer slip. J Mech Mater Struct (JoMMS) 6(1–4):213–230
8. Irschik H, Heuer R, Ziegler F (2000) Statics and dynamics of simply supported polygonal
Reissner-Mindlin plates by analogy. Arch Appl Mech 70:231–244
9. Mueller B, Beer M (2004) Fuzzy randomness. Springer, Berlin
10. Salcher P, Pradlwarter H, Adam C (2016) Reliability assessment of railway bridges subjected to
high-speed trains considering the effects of seasonal temperature changes. Eng Struct (accepted
for publication). doi:10.1016/j.engstruct.2016.08.017
11. Viertl R, Hareter D (2006) Beschreibung und analyse unscharfer information - statistische
methoden für unscharfe daten (in German). Springer, Berlin
12. Yang YB, Yau JD, Wu YS (2004) Vehicle-bridge interaction dynamics: with applications to
high-speed railways. World Scientific, Singapore
13. Yau JD, Yang YB (2016) Resonance of a train car traveling over multi-unit simple beams. This
paper is part of this Springer book, where also this paper of Ziegler will be published
14. Zadeh LA (1965) Fuzzy sets. Inf Control 8:338–353
15. Ziegler F (1998) Mechanics of solids and fluids. Corr. repr. of the 2nd edn. Springer,
New York/Wien
Resonance of a Series of Train Cars Traveling
Over Multi-Span Continuous Beams

J.D. Yau and Y.B. Yang

Abstract For a train car moving over a multi-span continuous beam of identical
span length L at constant speed v, it may encounter repetitive excitations transmitted
from the sustaining beam of frequency v=L excited by the previous passing cars.
If the exciting frequency v=L coincides with the vehicle frequency fc , namely
v=L D fc , resonance will be developed on the running car. In such a case, when
the train car travels over more and more spans of the beam, the response of the
car will be accumulated and becomes larger and larger, up to the limit imposed
by inherent damping. Using the rigid-vehicle/bridge interaction finite element
developed previously by the authors, each train car is modeled as a two-axle vehicle
and each span of the continuous beam is simulated as a number of beam elements.
Then the resonant response of the train cars running over the multi-span continuous
beam is analyzed. The numerical examples indicate that for a high speed train
composed of a series of cars traveling over a multi-span continuous beam, the
train-induced resonance on the bridge takes place at a rather high speed, but the
bridge-induced resonance on the train cars takes place at a much lower speed.

1 Introduction

To take the advantage of formwork preparation, the railway bridges for carrying high
speed trains are often designed as multi-span continuous beams. For a dynamical
system subjected to a periodic load, resonance takes place at the system when the
exciting frequency coincides with any of the natural frequencies of the system.
A similar resonant phenomenon can be observed from the train-bridge system of
high speed railways as well. Because of the regular arrangement of bogies in a
train, the bridge encounters repetitive excitations caused by a passing train. For

J.D. Yau ()

Department of Architecture, Tamkang University, Taipei 25137, Taiwan
e-mail: [email protected]
Y.B. Yang
School of Civil Engineering, Chongqing University, 400044 Chongqing, China
Department of Civil Engineering, National Taiwan University, Taipei 10617, Taiwan
e-mail: [email protected]

© Springer International Publishing Switzerland 2017 11

H. Irschik et al. (eds.), Dynamics and Control of Advanced Structures
and Machines, DOI 10.1007/978-3-319-43080-5_2
12 J.D. Yau and Y.B. Yang

Fig. 1 R-VBI model

this, numerous researchers and scientists have conducted researches on the train-
induced resonance of a rail bridge, say, by simulating the train action as a sequence
of moving loads with equal intervals [5]. Since a high speed rail bridge must be
designed to provide sufficient structural strength for the traveling trains at high
speeds, the dynamic response of the train, especially with respect to resonance,
has become an issue that dominates the operational safety and riding quality of
the train-bridge system. For this reason, vehicle–bridge interaction (VBI) dynamics
has received the attention of researchers in the past two decades.
Concerning the VBI of high speed rail bridges, many interesting topics were
investigated, such as the train-induced resonance of a bridge, the wind effect on
moving trains, the behavior of moving trains under earthquakes, and train-induced
ground vibrations. However, there exists relatively little information on the resonant
phenomenon of train cars running over continuous beams with multi-spans of
identical length. In this paper, the train car is modeled as a two-axle system with
rigid car body and each span of the continuous beam is discretized as a number of
beam elements. Then the rigid-vehicle/bridge interaction (R-VBI, see Fig. 1) finite
element developed previously by the authors is employed to analyze the dynamic
response of the train cars running over a multi-span continuous beam. The numerical
results indicated that for a high speed train composed of a series of cars traveling
over multi-span continuous beams, the train-induced resonance of the bridge takes
place at a rather high speed, but the resonance of the train cars takes place at a much
lower speed.

2 VBI Problem

As was schematically shown in Fig. 1, during the passage of the train over a bridge,
some elements of the bridge will be directly acted upon by the two-axle train car,
while the others are not. In this study, the most commonly used beam element
with 12 degrees of freedom (DOFs) will be adopted to simulate the continuous
beam, of which the axial displacement is interpolated by linear functions and the
transverse displacements by cubic (Hermitian) functions. The number of train cars
directly acting on each beam element changes as the train keeps moving, and so do
Resonance of a Series of Train Cars Traveling Over Multi-Span Continuous Beams 13

the contact points between each bridge element and the train cars. As the contact
points between the running cars and the bridge move from time to time, the system
matrices must be updated and factorized at each time step in the incremental time-
history analysis [1]. To overcome the time-varying nature of the problem, Yang et al.
proposed a method for condensing the DOF of the two-axle car system with rigid
car body into those of the element in contact, after the two-axle car system equations
are discretized by Newmark’s finite difference formulas [4]. The result is an R-VBI
element that possesses the same number of DOFs as the parent element, while the
properties of symmetry and bandedness are preserved. In the following section, a
brief description of the condensation technique for the VBI system considering the
pitching effect of the train car will be briefed.

2.1 Rigid-VBI Element

As shown in Fig. 2, the contact forces acting on the two elements i and j are
composed of four components: (1) the static weights associated with the car body
and wheel assemblies, represented by Mv g and Mw g; (2) the damping forces
resulting from the relative velocity of the rigid car body to the bridge elements,
as indicated by the terms containing cv ; (3) the elastic forces resulting from the
relative displacement of the car body to the bridge elements, as indicated by the
terms involving kv ; and (4) the inertial forces due to the vertical acceleration of the
bridge elements, as indicated by the terms mw uR .
In analyzing the VBI problem, two sets of equations of motion are written, one
for the supporting bridge and the other for each of the moving vehicles. Consider
a typical increment from time t to t C t in time domain. The equations of motion
for the car body can be written for the current time step, with Nt D t C t clearly
inserted as the subscript as:
 " #
Mv 0 yR v 2cv 0 yP v
C 2
0 Iv Rv Nt 0 cv d2 Pv Nt
" #
2kv 0 yv fver
C 2 D ; (1)
0 kv d2 v Nt
frot Nt

where the acting forces are given as

fver cv .Pui C uP j / C kv .ui C uj / C kv .ri C rj /

D : (2)
frot tC t
0:5dŒcv .Pui  uP j / C kv .ui  uj / C kv .ri  rj / tC t
14 J.D. Yau and Y.B. Yang

Fig. 2 VBI system considering the pitching effect: (a) R-VBI element; (b) Free body diagrams

The equations of motion for the bridge elements, i and j, are expressed for the current
time step t C t as

Œmi  fRui gtC t C Œci  fPui gtC t C Œki  fui gtC t D pi;tC t fNci g ; (3)
 ˚  ˚  ˚ ˚
mj uR j tC t C cj uP j tC t C kj uj tC t
D pj;tC t Ncj ;

where the associated contact forces are

" #
cv .Pui  yP v  0:5dPv / C kv .ui C ri  yv  0:5dv /
pi;tC t D ; (4)
C0:5Mv g C mw .g C uR i / tC t
" #
P
cv .Puj  yP v C 0:5dv / C kv .uj C rj  yv C 0:5dv /
pj;tC t D ;
C0:5Mv g C mw .g C uR j / tC t
Resonance of a Series of Train Cars Traveling Over Multi-Span Continuous Beams 15

the shape function vector fNci g at x D xc of the i-the beam element is given as
D        xc 2  xc 3  xc  x2 E T
fNci g D 1  3 xlc 2 C 2 xlc 3 xc 1  xc 2
l
3 l
2 l l
 1 lc ; (5)
˚
and the shape function vector Ncj is given in a similar way.
Following the condensation procedure developed for a rigid-VBI system [4], one
can discretize the two-axle car system in advance using Newmark’s finite difference
formulas and then condensing the corresponding DOFs hyv v i of the two-axle car
system into the beam element in contact. Since the R-VBI elements possess the
same number of DOFs as the parent element, while the properties of symmetry
and bandedness are preserved, this element is particularly suitable for analyzing the
dynamic responses of the VBI problems concerning both the bridge and vehicle
responses. Readers who are interested in derivation of the R-VBI element should
refer to the paper by Yang et al. [4] for further details.

2.2 VBI Analysis Using FEM

Figure 3 shows the response analysis procedure of the train-bridge system using
the condensation technique. Because the VBI element and its parent element are
fully compatible, the conventional element assembly process can be adopted with
no difficulty to form the equations of motion for the entire vehicle–bridge system,
that is

ŒMfUR b g C ŒCVBI fU

P b g C ŒKVBI fUb g D fPb g C fFVBI g; (6)

where ŒM, ŒCVBI , ŒKVBI , respectively, denote the mass, damping, and stiffness
matrices of the entire condensed vehicle–bridge system, fUb g the bridge displace-
ments, fPb g the external loads acting on the bridge, and fFVBI g the condensed
effective vehicular loads acting on the bridge. The preceding equations are typical
second-order differential equations, which can be solved by a number of time-
marching schemes. In this study, the Newmark ˇ method with constant average
acceleration, i.e., with ˇ D 1=4 and  D 1=2 [5], is employed to render the
preceding equations into a set of equivalent stiffness equations, from which the
bridge displacements fUb g can be solved for each time step. Once the bridge
displacements fUb g are made available, the bridge accelerations and velocities can
be computed accordingly. By a backward procedure, the response of the two-axle
car system can be computed as well on the element level, which serves as an
indicator of the riding comfort.
16 J.D. Yau and Y.B. Yang

Fig. 3 Flowchart of VBI analysis using the condensation technique

3 Resonant Speeds

Resonance takes place when the exciting frequency of the external forces coincides
with any of the natural frequencies of a mechanical system. For a train composed
of cars with bogies of equal interval D, when it travels at speed v over a bridge,
an exciting frequency v=D will be generated. When the exciting frequency v=D
matches one of the bridge frequencies fb , resonant response will be developed on
the bridge, for which the resonant speed is vbr D fb D [3]. Such a phenomenon will
be referred to as the train-induced resonance on the bridge.
On the other hand, for a train car moving over a multi-span continuous beam with
identical span length L at speed v, the train car will encounter repetitive excitation
transmitted from the beam with frequency v=L. Once the exciting frequency v=L
coincides with one of the vehicle’s frequencies fc , resonance will be developed
on the car in running, that is, when v=L D fc . In this case, the corresponding
resonant speed is denoted as vvr D fc L [2]. Such a phenomenon is referred to
as the bridge-induced resonance on the train cars. For most high speed railways,
the bridge-induced resonant speed for the train cars takes place at a lower speed
compared with the train-induced resonant speed for the rail bridge.
Resonance of a Series of Train Cars Traveling Over Multi-Span Continuous Beams 17

4 Finite Element Analysis

Figure 4 shows a schematic plot of the planar two-axle car with rigid car body
moving over a continuous beam, assumed to be of 5 spans. The properties of the
5-span continuous beam and planar two-axle system are listed in Tables 1 and 2,
respectively. By letting f1 denote the first frequency of the beam, the resonant speed
induced by the train car is vres D f1 D [1]. In the following examples, each span
of the continuous beam shown in Fig. 4 is modeled by 6 beam elements. Based
on Newmark’s method of direct integration, numerical solutions for the dynamic
response of the bridge due to the moving two-axle car system have been computed
for a time step of 0:0025 s. To focus on the resonance of the continuous beam of
identical spans induced by the train car, the track irregularities will be neglected
in performing the VBI analysis with the condensation technique in the following
examples. Since the vertical acceleration of the moving vehicle has been regarded
as an indicator of the riding comfort or running safety of high speed trains, the
acceleration response of the VBI system is of key concern in this study.

4.1 Resonant Response Analysis

In Table 2, the resonant speeds vr;v and vr;p represent the speeds for the resonance
to occur on the vehicle due to coincidence of the vertical and pitching frequencies,
respectively, of the two-axle car system with the implied frequency v=L of the
continuous beam. For the present purposes, let us consider the cases when the two-
axle car system moves over the continuous beam at each of the two resonant speeds
(vr;v and vr;p ) as listed in Table 2. Figures 5 and 6 show the vertical and pitching
accelerations, respectively, of the midpoint of the two-axle car system computed
by the R-VBI element simulation. Evidently, when the vehicle moves at either the

Fig. 4 Two-axle car system moving over a continuous beam with identical spans

Table 1 Properties of the continuous beam

L/m EJ/Nm2 m/tm1 /% f1 /Hz f2 /Hz f3 /Hz vres /km h1
35 2:82  108 25 1.5 4.3 4.78 5.97 387
18 J.D. Yau and Y.B. Yang

Table 2 Properties of the two-axle car system

D/m D/m My /t Iv /tm2 mw /t cs /kN sm1
25 20 40 2050 4.5 10

ks /kN m1 fv /Hz fp /Hz vr;v /km h1 vr;p /km h1
650 0.91 1.27 115 160

Fig. 5 Resonant response of the vertical acceleration of midpoint of the two-axle car system
moving over the continuous beam

Fig. 6 Resonant response of pitching acceleration of the two-axle car system moving over the
continuous beam

vertical or pitching resonant speed, the acceleration response of the vehicle grows
monotonically as the vehicle passes through more and more spans.
Resonance of a Series of Train Cars Traveling Over Multi-Span Continuous Beams 19

4.2 Maximum Acceleration of Train Cars Versus Train Speed

In this example, a train composed of 16 coaches is considered for simulating its

passage over a continuous beam with identical spans, as shown in Fig. 7. To illustrate
the resonant phenomenon, Figs. 8 and 9 show the maximum acceleration for the
running train and the beam, respectively, against the train speed. For the present
purposes, the maximum vertical acceleration av;max for each of the two-axle cars
shown in Fig. 7 is defined as:
ˇ ˇ!
ˇ DRv ˇˇ
ˇ
av;max D max ˇuR v ˙ ˇ : (7)
ˇ 2 ˇ

The maximum acceleration response av;max of the two-axle system versus the train
speed v was plotted in Fig. 8, which will be referred to as the av;max  v plot for the
train cars.
From Fig. 8, one observes that there exist three resonant peaks at the speeds of
115, 160, and 387 km/h, corresponding to the bridge-induced vertical and pitching
resonant speeds, and the vehicle-induced VBI resonant speed, respectively. The

Fig. 7 A series of train cars moving over a multi-span continuous beam

Fig. 8 av;max  v plot for the two-axle car system moving over the continuous beam
20 J.D. Yau and Y.B. Yang

Fig. 9 Maximum midpoint acceleration at the central span of the continuous beam vs. train speed

pitching resonant speed at vr;p D fp L (= 160 km/h) deserves a special note, since
it creates a peak much higher than that of the vertical resonance at a speed that can
be encountered in practice. Under such a condition, the vertical component ˙Rv D=2
induced by the pitching acceleration Rv dominates the peak acceleration response of
the vehicle, which is a long coach (D D 25 m). Furthermore, larger response is
induced on the train cars as there are more cars passing the series of spans of the
continuous beam, due to the accumulation effect, although it is not shown here.
The speed 230 km/h represents another peak for the vehicle, which is caused by the
beating phenomenon associated with the pitching rotation. The beating phenomenon
is due to the fact that the subresonant excitation of the second frequency of the
bridge, i.e., f2 =2 D 4:78=2 D 2:39 Hz, couples with the pitching frequency of the
vehicles, 1:27 Hz. Their average frequency is fav D .2:39 C 1:27/=2 D 1:83 Hz,
which corresponds to the beating speed of vb D fav L D 1:83 Hz35 m = 64 m/s =
230 km/h. Further investigation is needed for this phenomenon.
In contrast, the vehicle-induced resonance at vres D f1 D D 387 km/h, where f1 is
the frequency of the beam, is caused by the first resonance of the 5-span continuous
beam under the action of a series of train cars with car length D. For reference, the
maximum vertical acceleration of the midpoint of the central-span of the continuous
beam subjected to the same series of train cars constituting the train has been plotted
in Fig. 9.

5 Concluding Remarks

Using the VBI finite element developed previously by Yang et al. [4], the key
parameters that dominate the vertical and pitching resonance of a two-axle car
system running over a multi-span continuous beam are studied. The analysis has
Random documents with unrelated
content Scribd suggests to you:
 Fig. 23.
They chose the cave-sanctuary of Apollo in which to place, at the
close of their term of office, their votive tablet because it was in this
ancient sanctuary that they had taken their oath of fidelity on their
election. At the official scrutiny[162] of candidates for the archonship
enquiry was made as to the ancestry of the candidate on both
father’s and mother’s side. But it was not enough that he should be
a full citizen, he was also solemnly asked whether he had an Apollo
Patroös and a Zeus Herkeios and where their sanctuaries were. The
Athenians, in so far as they were Ionians, claimed descent through
Ion from Apollo and of course through Apollo from Zeus. The
sanctuary in the cave was therefore to them of supreme importance.
This scrutiny over, the candidates went to a sacred stone near the
Stoa Basileios, and there, standing over the cut pieces of the
sacrificed victim, they took the oath to rule justly and to take no
bribes, and they swore that if any took a bribe he would dedicate at
Delphi[163] a gold statue commensurate in value.
The archons had to prove their relation to Apollo Patroös and to
dedicate a gold statue if they offended the Pythian god under whose
immediate control they stood. Moreover it was not enough that they
should swear at the Stoa Basileios. The oath was doubtless older
than any Stoa Basileios in the later Market Place. After they had
sworn there they had to ‘go up to the Acropolis and there swear the
same oath again[164].’ Then and not till then could they enter office.
And whither on the Acropolis should they go? Whither but to the
cave where a little later they will dedicate their votive tablets, and
where still the foundations of an altar stand, the cave of their
ancestor Apollo Patroös and Pythios?
Whether the second oath, on the Acropolis, was taken actually in
the cave-sanctuary cannot be certainly decided; the votive tablets
make it probable and they make quite certain that the cave-
sanctuary was officially used by the archons. This fact it is necessary
to emphasize. Until these inscriptions were brought to light Apollo’s
cave was thought to be of but little importance, curious and primitive
but practically negligible. Now that it is clear that the archons
selected it as their memorial chapel, such a view is no longer
possible. It was a sanctuary not merely of Apollo Below-the-Heights
but of the ancestral god, the Apollo Patroös of the archons.
Moreover—a fact all important—this Apollo ‘Below-the-Heights’ being
Apollo Patroös was also Apollo Pythios. Demosthenes in the de
Corona[165], calling to witness his country’s gods, says ‘I call on all
the gods and goddesses who hold the land of Attica and on Apollo
the Pythian, who is ancestral (πατρῷος) to the state.’ The sanctuary
in the cave was a Pythion. Apollo coming as he did to Athens from
Pytho was always Pythian whatever additional title he might take,
and every sanctuary of his was a Pythion; his most venerable
sanctuary was not a temple but a hollowed rock.

The Pythion lies before us securely fixed, primitive, convincing.

With the ‘sanctuary of Zeus Olympios’ it is alas! far otherwise. Given
that the Pythion is fixed at the North-West corner of the Acropolis,
and given that, according to Strabo (see p. 69), it was so near the
Olympieion that the place of an altar could be described as ‘between’
them, then it follows that somewhere near to that North-West corner
the sanctuary of Zeus Olympios must have lain. We may further say
that as Thucydides, it will be seen, notes the various sanctuaries and
the city-well in the order from East to West, and begins with the
sanctuary of Zeus Olympios, it lay presumably somewhat to the East
of the Pythion. To the East of the Pythion, near to the supposed site
of the temenos of Aglauros, was found an inscription[166] with a
dedication to Zeus, but, as inscriptions are easily moveable, no great
importance can be attached to this isolated fact. Of definite
monumental evidence for the existence of a sanctuary of Zeus where
we seek it, we must frankly own at the outset there is nothing
certain[167]. It must stand or fall with the Pythion.
Before examining such literary evidence as exists it is necessary to
note clearly that Thucydides mentions not a temple but a sanctuary.
The great temple near the Ilissos, begun by Peisistratos[168], and
not completed till centuries later by Antiochus Epiphanes and
Hadrian, is usually spoken of as a temple (ναός), but we have no
grounds whatever for supposing that on or near the Long Rocks
there was a temple, but only a sanctuary[169], which may very likely
have been merely a precinct with an altar. Such a precinct and altar
might easily disappear and leave no trace. This is of importance for
the understanding of what follows.
When we come to literary evidence one point is clear. Before
Peisistratos began the building of his great temple there existed
another and earlier place for the worship of Zeus, and this is spoken
of as not a temple but a sanctuary. Pausanias[170], when he visited
the great temple, wrote, ‘They say that Deucalion built the old
sanctuary of Zeus Olympios, and, as a proof of the sojourn of
Deucalion at Athens, point to his tomb, which is not far distant from
the present temple.’
It has usually been assumed that this earlier sanctuary was on or
near the site of the later temple, but, as Prof. Dörpfeld[171] has
pointed out, this is no-wise stated by Pausanias. He only says that
there was a tomb of Deucalion, not far from the present temple, and
that the existence of this tomb made people attribute to Deucalion
the building of the early sanctuary. Where the early sanctuary was
he does not say. It should be noted that he is careful to use the
word sanctuary, not temple, in speaking of the foundation of
Deucalion.
From this it follows, I think, that when we hear of a sanctuary of
Zeus Olympios, not a temple, there is a slight presumption in favour
of its being the earlier foundation. In the opening scene of the
Phaedrus[172] an ‘Olympion,’ i.e. a sanctuary of Zeus, is mentioned.
Socrates and Phaedrus meet somewhere, presumably within the city
walls, for Socrates is later taxed with never going for a country walk.
Socrates says, ‘So it seems Lysias was up in town.’ Phaedrus
answers, ‘Yes, he is staying with Epikrates in yonder house, near the
Olympion, the one that used to belong to Morychus.’ The favourite
haunt of Socrates was the agora; a stroll by the Ilissos was to him a
serious and unusual country walk. Our Olympion at the North-West
corner of the Acropolis would fit the scene somewhat better than the
great temple near the Ilissos; but that is all, the passage proves
nothing.

A question more important perhaps than any topographical issue

remains. Do we know anything of the nature of the god worshipped
in the ancient sanctuary, or of the character of his ritual? The
question may seem to some superfluous. Zeus is surely Zeus
everywhere and for all time, his cloud-compelling nature and his
splendid sacrificial feasts familiar from Homer downwards. But then
what of Deucalion? Deucalion is a figure manifestly Oriental, a feeble
copy of the archetypal Noah. Why does he institute the worship of
our immemorial Indo-European Zeus? Are there two Zeuses?
There were, at least at Athens, two festivals of Zeus.
Thucydides[173] himself is witness. He tells us of the trap laid for
Kylon in characteristic fashion by the Delphic oracle. Kylon was to
seize the Acropolis ‘on the greatest festival of Zeus.’ But this
‘greatest festival’ was alas for him! not of the Zeus he, as an
Olympian victor, remembered, but of ‘Zeus Meilichios,’ and—
significant fact for us—it, the familiar Diasia, was celebrated ‘outside
the city.’ This ‘outside the city’ cannot fail, used as the words are by
Thucydides himself, to remind us of our sanctuary, also ‘outside.’
What may be dimly discerned, though certainly no-wise
demonstrated, is this. The name Zeus is one of the few divine titles
as to which philologists agree that it is Indo-European. But the name
Zeus was attached to persons and conceptions many and diverse,
and here in Athens it was attached to a divinity of Oriental nature
and origin. Meilichios[174] is but the Graecized form of Melek, the
‘King’ best known to us as Moloch, a deity who like the Greek
Meilichios loved holocausts, a deity harsh and stern, who could only
by a helpless and hopelessly mistaken etymology be called Meilichios
the Gentle One. His worship prevailed in the Peiraeus, brought
thither probably by Phenician sailors, from his sanctuary there came
the familiar reliefs with the great snake as the impersonation of the
god. It was this Semitic Melek whom Deucalion brought in his ark.
When this Semitic immigration took place it is hard to say. Tradition,
as evidenced by the Parian Chronicle[175], placed it in the reign of
the shadowy Attic king Kranaos, about 1528 b.c.

The sanctuaries of both Zeus and Apollo are alike outside the
ancient city. Zeus had altars on the Acropolis itself; Apollo, great
though he was, never forced an entrance there. The fact is surely
significant. Herodotus[176], it will be remembered, marks the
successive stages of the development of Athens: under Kekrops they
were Kekropidai, under Erechtheus they were Athenians, and last,
‘when Ion, son of Xuthos, became their leader, from him they were
called Ionians.’ Ion was the first Athenian polemarch[177].
One thing is clear, Ion marks the incoming of a new race, a race
with Zeus and Apollo for their gods. From the blend of this new
stock with the old autochthonous inhabitants arose the Ionians. Zeus
and Apollo were called ‘ancestral’ at Athens because they were
ancestral; the new element traced its descent from them, and
presumably the affiliation was arranged by Delphi; but Apollo,
though his sanctuary was on the hill, never got inside.
Ion had for divine father Apollo, but his real human father was
Xuthos. This Xuthos, as immigrant conqueror, marries the king’s
daughter Creousa. Xuthos was really a local hero of the deme
Potamoi[178], near Prasiae. He came of Achaean stock, and therefore
had Zeus for ancestor. Hermes, in the prologue to the Ion[179], is
quite clear. There was war between Athens and Euboea:

And Xuthos strove and helped them with the sword

And had Creousa, guerdon of his aid,
No home-born hero he, but son of Zeus
And Aiolos, Achaean.
And again[180], when Ion questions his unknown mother as to her
husband:

Ion. And what Athenian took thee for his wife?

Cre. No citizen: an alien from another land.
Ion. Who? For a well-born man he needs had been.
Cre. Xuthos, of Zeus and Aiolos the offspring he.

The tomb of Ion, significant fact, was not at Athens but at

Potamoi, and Pausanias[181] saw it there. Well may the sanctuaries
of Zeus and Apollo stand together.

To return to the question of topography. That the cave marked Β

on the plan is sacred to Apollo admits, in the face of the inscribed
votive tablets, of no doubt. But a difficulty yet remains. It was noted
in speaking of the cave above the Klepsydra that it was too shallow
and too exposed to be a natural scene of the story of Creousa. The
same objections, though in a somewhat less degree, apply to the
cave marked Β. The difficulty, however, admits of an easy solution.
The excavators proceeded to clear out cave Γ, and here they
found nothing, no votive tablets, no altar, no inscriptions. But in
carrying on their work further East they came on a fourth cave, of a
character quite different from that of Α, Β, or Γ. The fourth cave, Δ,
has a very narrow entrance; it communicates by a narrow passage
with Δ′ and also with Δ″, but Δ″ has been turned into a small
Christian church, of which the pavement and a portion of a brick wall
yet remain. Here at Δ we have a cave in the full sense of the word,
and here we have in all probability the cave or caves, the ‘seats[182]’
(θακήματα) of Pan.
But, be it remembered, Pan was a late-comer; his worship was
introduced after his services at Marathon. In heroic days, the time of
the story of Creousa, the Long Rocks were shared by the Pythian
god and the daughters of Aglauros. The hollow triple cave marked
Δ′, Δ″, Δ‴ was once the property of Apollo, and it saw the birth of
Ion; later it was handed over to Pan, and is again, as in the
Lysistrata[183], the natural sequestered haunt of lovers. Kinesias, on
the Acropolis, points out to Myrrhine that near at hand is the
sanctuary of Pan for seclusion, and close by the Klepsydra for
purification.
In the countless votive tablets[184] to Pan and the nymphs, the
type varies little. We have a cave, an altar: round the altar three
nymphs are dancing, usually led by Hermes, and, perched on the
side of the cave or looking through a hole, Pan is piping to them.
The three nymphs, three daughters of Kekrops, were then dancing
on the Long Rocks long before Pan came to pipe to them. Concerned
as we are for the present with Apollo and his Pythion, it is only
necessary to note that their shrine, the sanctuary of Aglauros, must
have been near the cave of Pan, somewhere to the East.
Euripides[185] speaks of them as practically one:

O seats of Pan and rock hard by

To where the hollow Long Rocks lie
Where, before Pallas’ temple-bound
Aglauros’ daughters three go round
Upon their grassy dancing-ground
To nimble reedy staves.
Where thou O Pan art piping found
Within thy shepherd caves.

Exactly where that sanctuary of Aglauros was excavations have

not established. At the point where the cavern is closed by the little
modern church, begins a stairway, consisting of seventeen steps (θ-
κ-λ-μ-), cut in the rock. These steps manifestly lead up to the steps
already known, which lead down, twenty-two in number, from the
Erechtheion. This is probably the ‘opening’ (ὄπη) down which the
deserting women in the Lysistrata[186] were caught escaping. Still
further East is a long narrow subterranean passage, a natural cleft in
the rock π-π′, and at the end of this, just above the modern Church
of the Seraphim, is supposed to be the sanctuary of Aglauros. Here
were found a niche in the rock, the basis of a statue, and some
fragments of black-figured vases. Here again there is communication
with the Acropolis, but only by a ladder ascending the cliff for about
twenty feet at a precipitous point. Moreover the upper part of the
stone stairway is of mediaeval date so that it is not likely that the
ascent was an ancient one.

The Sanctuary of Ge.—The site of this sanctuary can, within very

narrow limits be determined.
Pausanias, in describing the South side of the Acropolis, after
passing the Asklepieion, notes the temple of Themis and the
monument of Hippolytus. Apropos of this he mentions and probably
saw a sanctuary of Aphrodite Pandemos (p. 105); he then says
‘there is also a sanctuary of Ge Kourotrophos and Demeter Chloe’;
immediately afterwards he passes through the Propylaea. The
sanctuary of Ge must therefore have been at the South-West corner
or due West of the Acropolis, and presumably somewhere along the
winding road followed by Pausanias (see Plan, p. 38). From the
account of Pausanias[187] we should gather that Ge Kourotrophos,
Earth the Nursing-Mother, and Demeter Chloe, Green Demeter had a
sanctuary together; perhaps they had by the time of Pausanias, but
the considerable number of separate dedications[188] to Demeter
Chloe makes it probable that at least in earlier days these precincts,
though near, were distinct.
The union of Ge Kourotrophos and Demeter Chloe is not the union
of Mother and Maid, it is the union of two Mother-goddesses. Of the
two Demeter belongs locally not to Athens but to Eleusis. Ge
Kourotrophos is obviously the earlier and strictly local figure. But
Demeter of Eleusis, from various causes, political and agricultural,
developed to dimensions almost Olympian, and her figure tended
everywhere to efface that of the local Earth-Mother, hence we need
not be surprised that the number of dedications to Demeter is larger
than that of those to Kourotrophos. Kourotrophos appears among
the early divinities enumerated by the woman herald in the
Thesmophoriazusae[189], and the scholiast, in his comment on the
passage, recognizes her antiquity: ‘either Earth or Hestia; it comes
to the same thing; they sacrifice to her before Zeus.’ Suidas[190]
states that Erichthonios was the first to sacrifice to her on the
Acropolis, and instituted the custom that ‘those who were sacrificing
to any god should first sacrifice to her.’

The Sanctuary of Dionysos-in-the-Marshes.

The name Dionysos at once carries us in imagination to the
famous theatre on the South side of the Acropolis (Fig. 16), and we
remember perhaps with some relief that this theatre is, quite as
much as the Pythion, ‘towards’ the ancient city; it lies right up
against the Acropolis rock. We remember also that Pausanias[191], in
his account of the South slope, says ‘the oldest sanctuary of
Dionysos is beside the theatre.’ He sees within the precinct there two
temples, the foundations of which remain to-day; one of them was
named Eleutherian, the other we think may surely have belonged to
Dionysos-in-the-Marshes. It is true that the ground about the theatre
is anything but marshy now, nor could it ever have been very damp,
as it slopes sharply down to the South-East. Still, from an ancient
name it is never safe to argue[192]; in-the-marshes may have been a
mere popular etymology from a word the meaning of which was
wholly lost.
But a moment’s reflection shows that the identification, though
tempting, will not do. Thucydides himself (p. 66) seems to warn us;
he seems to say, ‘not that precinct which you all know so well and
think so much of, not that theatre where year by year you all go, but
an earlier and more venerable place, and, that there be no mistake,
the place where you go on the 12th day of Anthesterion, and where
your ancestors went before they migrated to colonize Asia Minor.’
It is most fortunate that Thucydides has been thus precise,
because about this festival on the 12th day of Anthesterion we know
from other sources[193] certain important details which may help to
the identification of the sanctuary.

The festival celebrated on the 12th of Anthesterion was the

Festival of the Choes or Pitchers[194]. On this day, we learn from
Athenaeus[195] and others, the people drank new wine, each one by
himself, offered some to the god, and brought to the priestess in the
sanctuary in the Marshes the wreaths they had worn. On this day
took place also a ceremony of great sanctity, the marriage of the
god to the wife of the chief archon—the ‘king’ as he was called. The
actual marriage took place in a building called the Boukoleion, the
exact site of which is not known; but certain preliminary ceremonies
were gone through by the Bride in the sanctuary in-the-Marshes.
The author of the Oration ‘against Neaera[196]’ tells us that there
was a law by which the Bride had to be a full citizen and a virgin
when she married the king, she was bound over to perform the
ceremonies required of her ‘according to ancestral custom,’ to leave
nothing undone, and to introduce no innovations. This law, the
orator tells us, was engraved on a stele and set up alongside of the
altar in the sanctuary of Dionysos in-the-Marshes, and remained to
his day, though the letters were somewhat dim.
But this, though much, is not all. The orator goes on to tell us why
the law was written up in this particular sanctuary. ‘And the reason
why they set it up in the most ancient sanctuary of Dionysos and the
most holy, in the Marshes, is that not many people may read what is
written. For it is opened once only in each year, on the 12th of the
month Anthesterion[197].’ Finally, having sufficiently raised our
curiosity, he bids the clerk read the actual oath administered by this
pure Bride to her attendants, administered before they touch the
sacred things, and taken on the baskets at the altar. The clerk is to
read it that all present may realize how venerable and holy and
ancient the accustomed rite was. The oath of the attendants was as
follows: ‘I fast and am clean and abstinent from all things that make
unclean and from intercourse with man, and I will celebrate the
Theoinia and the Iobakcheia to Dionysos in accordance with
ancestral usage and at the appointed times.’
We shall meet again the precinct, the altar, the stele, the oath; for
the present it is all-important to note that the precinct In-the-
Marshes was open but once a year, and that on the 12th of
Anthesterion. It is impossible, therefore, that this precinct could be
identical with the precinct near the theatre on the South slope[198],
as this must have been open for the Greater Dionysia, celebrated in
the month Elaphebolion (March-April).
The precinct In-the-Marshes has been sought and found; but
before we tell the story of its finding, in order that we may realize
what clue was in the hands of the excavators, it is necessary to say
a word as to the time and place of the festivals of Dionysos at
Athens.

Thucydides himself tells us that the Dionysiac festivals were two,

an earlier and a later. His use of the comparative—‘Dionysos-in-the-
Marshes,’ he says, ‘to whom is celebrated the more ancient Dionysiac
Festival,’—makes it clear that, to his mind, there were two and only
two. The later festival, the Greater Dionysia, was celebrated in the
precinct of Dionysos Eleuthereus; the time, we noted before, was
the month Elaphebolion.
The ‘more ancient Dionysiac Festival’ is of course a purely informal
descriptive title. But it happens that we know the official title of the
two Athenian festivals, the earlier and the later[199].
1. The later festival, that in the present theatre, was called in laws
and official inscriptions ‘the (Dionysia) in the town’ (τὰ ἐν ἄστει), or
‘the town Dionysia’ (ἀστικὰ Διονύσια).
2. The more ancient festival was called either ‘the Dionysia at the
Lenaion’ (τὰ ἐπὶ Ληναίῳ Διονύσια), or ‘the (dramatic) contest at the
Lenaion’ (ὁ ἐπὶ Ληναίῳ ἀγών), or, more simply, ‘the Lenaia’ (τὰ
Λήναια).
We have got two festivals, an earlier and a later, the earlier called
officially ‘Lenaia,’ or ‘the dramatic contest at the Lenaion’; but were
there two theatres also, an earlier and a later? Yes. Pollux[200] tells
us there was a Dionysiac theatre and a ‘Lenaic’ one—just the very
word we wanted. And to clinch the whole argument we find that the
‘Lenaic’ one was the earlier. Hesychius[201], explaining the phrase,
‘the dramatic contest at the Lenaion,’ says, ‘there is in the city the
Lenaion with a large enclosure, and in it a sanctuary of Dionysos
Lenaios. In this (i.e. presumably the enclosure) the dramatic
contests of the Athenians took place, before the theatre was built.’
This ‘theatre,’ where the plays were performed before the theatre
of Eleuthereus was built, was no very grand affair; its seats, it would
seem, were called ‘scaffoldings’ (ἴκρια). Photius[202] in explaining the
word ikria says, ‘the (structure) in the agora from which they
watched the Dionysiac contests before the theatre in the precinct of
Dionysos was built.’
Photius, while explaining the ‘scaffolding,’ gives us incidentally a
priceless piece of information. This early theatre was in the agora.
But then, to raise a time-honoured question, to which we shall later
(p. 132) return, where is the agora? This question for the present we
must not pursue. But the ancient theatre consisted of more than
‘scaffolding’ for seats. It had what was the central, initial, cardinal
feature of every Greek theatre, its dancing place, its orchestra; and
we know approximately where this orchestra was. A
lexicographer[203], explaining the word orchestra, says, ‘a
conspicuous place for a public festival, where are the statues of
Harmodios and Aristogeiton.’
The agora, conducted by successive theorists, has made the
complete tour of the Acropolis, but the statues of the Tyrant-Slayers
cannot break loose from the Areopagus,—beneath which ‘not far’
from the temple of Ares, Pausanias[204] saw them. The statues,
according to Timaeus, were at the site of the ancient orchestra[205],
from the scaffolding of which ‘in the agora’ the more ancient festival
(the Lenaia) was witnessed. Here then, somewhere near the
Areopagus, we must seek the sanctuary of Dionysos-in-the-Marshes.
The Lenaia, though more ancient than the ‘city Dionysia’ was no
obscure festival. Plato[206], in the Protagoras, mentions a comedy
which Pherecrates had brought out at the Lenaia, and it can never
be forgotten that for the Lenaia, in 405 b.c., Aristophanes wrote the
Frogs[207]. The chorus of Frogs[208] assuredly remember that their
home is in the Limnae. There they were wont to croak and chant at
the Anthesteria, on the third day of which festival, the Chytroi or
Pots, came the ‘Pot Contests,’ probably the earliest dramatic
performances that Athens saw.

‘O brood of the mere, the spring,

Gather together and sing
From the depths of your throat
By the side of the boat
Co-äx, as we move in a ring;

As in Limnae we sang the divine

Nyseïan Giver of Wine,
When the people in lots
With their sanctified Pots
Came reeling around my shrine.’

The excavations which have brought to light the ancient sanctuary

of the Limnae were not undertaken solely, or even chiefly, with that
object. Rather the intention was to settle, if possible, other and
wider topographical questions: where lay the ancient road to the
Acropolis, where the ancient agora, and where the city well,
Kallirrhoë. Yet, to some, who awaited with an almost breathless
impatience the result of these excavations, their great hope was that
the precinct of the Limnae might be found; that they might know
where in imagination to picture the ancient rites of the Anthesteria
and the marriage of the Queen and those earliest dramatic contests
from which sprang tragedy and comedy. The wider results of the
excavations will be noted in connection with the Enneakrounos; for
the moment it is the narrower, intenser issue of the Limnae that
alone concerns us.

So far our only topographical clues have been two. (1) Thucydides
has told us that the sanctuary in the Marshes with the other
sanctuaries he mentions was ‘towards’ the ancient city; we have
fixed the Pythion at the North-West corner of the Acropolis, and as
his account seems to be moving westwards, we expect the Dionysiac
sanctuary to be West of that point. (2) We know also (p. 87) that
the ancient orchestra was near the Areopagus. We look for a site for
the Dionysia which shall combine these two directions. If that site is
also a possible Marsh, so much the better; and here indeed, in the
hollow between the Pnyx, Areopagus, and Acropolis, water is caught
and confined; but for artificial drainage, here marsh-land must be.
This, by practical experience, the excavators soon had reason to
know.
 Fig. 24.

A portion of the results of the excavations begun by the German

Archaeological Institute in 1887[209] and lasting for upwards of ten
years is to be seen on the plans in Figs. 24 and 35. The enlarged
plan of a portion of the excavations (Fig. 24) for the moment alone
concerns us. The first substantial discovery that rewarded the
excavators was the finding of the ancient road. It followed, as
Professor Dörpfeld had always predicted it would, the lie of the
modern road. Roads being strictly conditioned by the law of least
resistance do not lightly alter their course. The present carriage road
to the Acropolis is a little less devious in its windings than the
ancient one, that is all (Fig. 35).
 Fig. 25.
Just below where the ancient road passes down from the West
shoulder of the Acropolis, and at a level much higher than that of
the road itself, the excavators came on a building of Roman date and
indifferent masonry, which proved to be a large hall, with two rows
of columns dividing it into a central nave and two aisles. To the East
the hall was furnished with a quadrangular apse. Within this apse
was found an altar[210] decorated with scenes from the worship of
Dionysos, a goat being dragged to the altar, a Satyr, a Maenad, and
the like. This altar would in itself rouse the suspicion that we are in a
sanctuary dedicated to Dionysos, but fortunately we are not left to
evidence so precarious.
Of far greater interest than the altar, and indeed for our purpose
of supreme importance, was another discovery. In the apse, with the
altar mentioned and other altars, was found the drum of a column
(Fig. 25), which had once stood in the great hall; columns just like it
are still standing, so that it belongs without doubt to the building.
On it is an inscription[211], divided into two columns and 167 lines in
length, which from its style may be dated about the third century
a.d. Above the inscription, in a relief in pediment form containing
Dionysiac symbols, two panthers stand heraldically, one to either
side of a cantharus; above is the head of a bull. Inscriptions
arranged in this fashion on columns are not unusual in the third
century a.d.[212]
The inscription contains the statutes of a thiasos, or club of
persons calling themselves Iobakchoi, who met in a place—the hall
where the inscription was set up—called the Bakcheion. This is our
quadrangular building marked Bakcheion on the plan (Fig. 24). The
rules, which are given in great detail, are very interesting, but for
the present one thing only concerns us—the name of the thiasos,
the Iobakchoi. Iobakchos was a title of Dionysos, a title probably
derived from a cry uttered in his worship, and, we remember (p. 85)
with sudden delight, the Gerarae, the attendants of the Queen,
promised in their oath to celebrate, in accordance with ancestral
usage, the Iobakcheia.
But the building, and even the traces of an earlier structure that
preceded it[213], are of late date; we are on the spot, and yet so far
the sanctuary in the Marshes eludes us. But not for long. Digging
deeper down, to the level of the ancient road, the excavators came
on another and an earlier structure, the triangular precinct marked
on the plan, and here at last evidence was found that settled for
ever the site of the sanctuary of Dionysos-in-the-Marshes.

The sanctuary, for such we shall immediately see it was, is of

triangular shape, and lies substantially lower than the roads by
which it is bounded. The sides of the triangle face approximately,
North, East and South-West. The precinct is surrounded by an
ancient polygonal wall, a portion of which from the South end of the
South-West side is shown in Fig. 26. The material is throughout blue
calcareous stone, but the masonry is by no means of uniform
excellence or of the same date. At various periods the wall must
have undergone repairs. The space enclosed is about 560 square
metres. Owing to the fact that the precinct lay deeper than the
surrounding roads, sometimes to the extent of two metres, the wall
is supported in places by buttresses, only one of which is of good
Greek masonry; the rest seem to have been added shortly before
the ancient precinct fell into disuse.
 Fig. 26.

A notable point about this precinct wall is that there is no trace of

any large entrance-gate. We expect a gate at the South-West side,
where the precinct is skirted by the main road. Here the wall is well
preserved, but there is no trace of any possible gate. The only
feasible place is at the South end of the East wall, where there
seems to have been a break, and towards this point, as we shall see,
the small temple is orientated. Here, then, and in all probability here
only, was there access to the precinct.
At the North-West corner the excavators came on a structure so
far unique in the history of discoveries. They found a walled-in floor
4·70 m. by 2·80. This floor is carefully paved with a mixture of
pebbles, stone, and cement, and is inclined to one corner at an
angle of 0·25 m. At this lowest point there is a hole through the wall
enclosing the floor, and outside, let into the pavement, is a large
vessel, 0·50 m. in diameter, quadrangular above, round below. They
had found, beyond all possible doubt, what they had never dared to
hope they might find, an ancient Greek wine-press or lenos, and at
the finding of that wine-press fled the last lingering misgiving. In
Fig. 27 is a view[214] of the wine-press, which shows clearly how it
lies just in the corner of the triangular precinct, with its South-West
wall (in the front of the picture) abutting on the Panathenaic way.
The stucco floor of the wine-press comes out in dead white. In the
background can be seen, to the right, the North aisle of the
rectangular Bakcheion, and, to the left, the foot of the Areopagus
rock.

Fig. 27.

The wine-press, which is shown in section in Fig. 28, had, like the
precinct, had a long history. It had been rebuilt more than once. The
paved floors of two successive structures are clearly visible. The
upper one is smaller than the lower, and, of course, of later date. It
is, however, below the level of the Bakcheion, and must have been
underground when the Bakcheion was built. The lower wine-press is
at the same level as the Lesche, on the opposite side of the road,
which is known to be of the 4th century b.c. Under this 4th century
wine-press is a pavement which must have belonged to a third, yet
earlier structure. It may be noted that these wine-presses are in
every respect exactly similar to those in use among the Greeks to-
day. The wine-press within the precinct is not the only one that
came to light; scattered about near at hand were several others.
Two can be seen on the plan in Fig. 35. It was indeed a place of
wine-presses, a Lenaion.

Fig. 28.
 Fig. 29.

The wine-press in itself would mark the precinct as belonging to

Dionysos, but there was more evidence forthcoming. In the centre of
the precinct is the foundation in poros stone of a large altar, 3·10
metres square (Fig. 29). In this foundation there once were four
holes; three of them remain, and the fourth may be safely supplied.
These holes are evidently intended for the supports on which the
actual altar-table rested. Such altar-tables are familiar in vase-
paintings, and seem to have been in use specially in the cult of
Dionysos; they held the wine-jars offered to the god, and baskets of
fruit such as those on which the attendants of the Queen took their
oath (p. 85). Moreover, the actual altar-slab of just such a table has
been found in Attica, and it bears an inscription to Dionysos
Auloneus[215]. Yet another important point remains. On the West
step of the altar foundation a long groove is sunk in the stone. Its
purpose is obvious. Both on the Acropolis and elsewhere in sacred
precincts such grooves are found, and they served to contain the
bases of stelae, on which decrees, dedications, and the like were
inscribed. Is it not at least possible that we have here not only the
altar on which the Queen took her oath, but the groove in which was
set up the very stele on which it was inscribed, the stele which stood
‘alongside of the altar’ (παρὰ τὸν βωμόν)?
We have, then, a precinct secluded from the main road; within it,
open to the air, a great altar. But inside this precinct not a single
inscription nor any sort of votive offering has come to light. In a
precinct so important this at first sight seems strange. The
explanation lies to hand. Votive offerings are meant to be seen,
meant to show forth the piety of the worshipper as well as the glory
of the god. Was it worth while to dedicate an offering in a precinct
that was open but for one day in the whole year? Apparently not.
This was essentially a ‘mystery’ sanctuary, with no touch of the
museum.

In the sanctuary of Dionysos-in-the-Marshes we expect not only

precinct and altar but an actual temple, the existence of which we
know, not from Thucydides, but from the scholiast[216] on the Frogs
of Aristophanes. Commenting on the word ‘marsh’ he says, ‘a sacred
place of Dionysos, in which there is a dwelling and a temple of the
god.’ Callimachus in the Hekale says,

‘To him, Limnaios, do they keep the feast

With choral dances.’

The ‘dwelling’ may be some building that contained the wine-press;

the temple happily has been found, and its position in relation to the
precinct is strange and significant.
The foundations of the temple came to light in the South corner of
the precinct. It is of small size (3·96 by 3·40 m.), and consists of a
quadrangular cella and a narrow pronaos. From its small size it
seems unlikely that the pronaos had any columns. The masonry is
very ancient. The walls are polygonal, and the blocks of calcareous
stone of which they are made are on the South-West side unusually
large. In the foundations of the side-walls a few poros blocks occur.
There are no steps serving as foundation to either cella or pronaos.
From this Professor Dörpfeld concludes that in all probability this
temple is earlier than the temple of Dionysos Eleuthereus, close to
the skenè of the theatre. The temple of Eleuthereus belonged to the
time of Peisistratos; it is more carefully built than the one newly
discovered, and it has one step. Early though the newly discovered
building undoubtedly is, it was preceded by a yet earlier structure,
the walls of which, marked on the plan, lie beneath its foundations.
Quite exceptional is the relation of the temple to the precinct. It
does not lie in the middle, and is, moreover, separated from the
inner part of the precinct by a wall and a door that could be closed.
This separating wall is however apparently later than the temple,
which possibly at one time stood free within the precinct. The
separating wall is only explicable on ritual grounds. It made it
possible for the temple to be accessible all the year round, whereas
the precinct, save for one day in the year, was closed.

Are we to give to the ancient sanctuary the name Lenaion? To the

sanctuary itself probably not. The meaning of Lenaion, it would
seem, is not ‘sanctuary of the god Lenaios,’ but rather ‘place of the
wine-press.’ It is noticeable that writers who could themselves have
seen the sanctuary never call it Lenaion. Thucydides[217], the writer
of the oration against Neaera[218], be he Demosthenes or
Apollodorus, and again Phanodemus[219], as quoted by Athenaeus,
all speak of it as the sanctuary of Dionysos-in-the-Marshes.
Isaeus[220] calls it the Dionysion-in-the-Marshes. On the other hand,
when contemporary authors speak of the dramatic contest which
was held not in honour of Dionysos Eleuthereus but at the older
Dionysia, they speak of the contest as at or on the Lenaion, never as
in-the-Marshes. The natural conclusion is that the name Lenaion is
applicable to the place where the contests actually took place,
namely to the ancient Orchestra and perhaps its immediate
neighbourhood. The district of the wine-presses naturally had its
dancing place, and that dancing place was called the Lenaion. To
this day the peasants of Greece use for their festival-dances the
village threshing-floor.

In the theatre of Eleuthereus Dr Dörpfeld[221] has given back to

us the old orchestra. He has shown us deep down below the
successive Graeco-Roman and Roman stages the old circular
orchestra built of polygonal masonry (Fig. 16). On this old orchestra,
with only wooden seats for the spectators, were acted, we now
know, the dramas of Aeschylus, Sophocles, Euripides, nay
tradition[222] even says, and we have no cause to doubt its veracity,
that Thespis was the first (in 586 b.c.) to exhibit a play in the ‘city’
contest (ἐν ἄστει).
But ancient though it was, before it, as we have seen, came the
orchestra in the Limnae. Dr Dörpfeld had hoped that his excavations
would give back this orchestra too; this hope has not been fulfilled.
Traces have been found of a circular structure on the South slope of
the Areopagus and are marked on the plan (Fig. 46), but they are of
uncertain date, and, if they mark the site of any ancient building, it
is probably that of the Odeion of Agrippa. The old orchestra lay at
the North-West corner of the Areopagos.

Tradition records the beginning of the contests ‘in the city,’ i.e. in
the theatre of Eleuthereus, but the beginnings of the other festivals,
the Lenaia and the Chytroi, held in the Limnae, are lost in the mists
before. The two are in all probability but different names for the
same festival, or rather the Chytroi is the whole ceremony of the
third day of the Anthesteria and Lenaia the name given to the
dramatic part of the ceremonies. But though we do not know the
beginning, and though, as will presently be seen, the ‘Pot-Contests’
went back in all probability to a time before the coming of Dionysos,
we have hints as to how the end came, how the splendour and
convenience of the great theatre of Eleuthereus gradually obscured
and absorbed the primitive contests of the orchestra in the Limnae.
It was, we know, the great statesman Lycurgus who, in the 4th
century b.c., built the first permanent stone stage in the theatre and
made the seats for the spectators as we see them now. So pleased
was he, it would seem, with his theatre that he thought it useless
and senseless to have plays acted elsewhere. Accordingly in the
Lives of the Ten Orators[223] we learn that Lycurgus introduced laws,
and among them one about comic writers ‘to hold a performance at
the Chytroi, a competitive one, in the theatre,’ and ‘to record the
victor as a victor in the city,’ which had formerly not been allowed.
He thus revived the performance which had fallen into disuse.
Lycurgus meant well we may be sure, but he was a Butad[224], he
ought to have known better than to pluck up an old festival by the
roots like that and think to foster it by transplantation. The end was
certain; the old precinct, deserted by its festivals, was bit by bit
forgotten, overgrown, and at last in part built over by the new
Iobakchoi.

The precinct had lost prestige by the time of Pausanias[225]. Had

the temple of Dionysos-in-the-Marshes been above ground he would
assuredly not have passed it by. Near to where the precinct once
was he saw a building, a circular or semi-circular one, which may
have been a last Roman reminiscence of the orchestra, and still of
note though it did not occupy the same site; he notes ‘a theatre
which they call the Odeion.’ It is probable that this was the theatre
built by Agrippa and mentioned by Philostratos[226] as ‘the theatre in
the Kerameikos, which goes by the name of the Agrippeion.’

Before leaving the sanctuary in-the-Marshes, a word must be said

as to the Anthesteria or, as Thucydides calls it, ‘the more ancient
Dionysiac Festival.’ I have tried elsewhere[227] to show in detail that
the Dionysiac element in the Anthesteria was only a thin upper layer
beneath which lay a ritual of immemorial antiquity, which had for its
object the promotion of fertility by means of the placation of ghosts
or heroes. On the first day, if I am right, the Pithoigia was an
Opening not only of wine-jars but of grave-jars; the second, the
Choes, was a feast not only of Cups but of Libations (χοαί); the
third, the Chytroi, not only a Pot-feast, but a feast of Holes in the
ground and of the solemn dismissal of Keres back to the lower
world. That the collective name of the whole feast Anthesteria did
not primarily mean the festival of those who ‘did the flowers,’ but
rather of those who ‘revoked the ghosts[228].’
But in trying to distinguish the two strata, the under stratum of
ghosts, the upper of Dionysos, I never doubted that the Pot Contest
on the day of the Chytroi belonged to Dionysos. Dionysos and the
‘origin of the drama’ are canonically connected. It has remained,
therefore, something of a mystery how Dionysos, late-comer as he
was, contrived to possess himself of the ancient ghost-festival and
impose his dramatic contests on a ritual substratum apparently so
uncongenial. Religions are accommodating enough, but some sort of
analogy or possible bridge from one to the other is necessary for
affiliation.
The difficulty disappears at once if we accept Professor
Ridgeway’s[229] recent theory as to the origin of tragedy. The drama
according to him is not ‘Dorian,’ and, save for the one element of the
Satyric play, not Dionysiac. It took its rise in mimetic dances at the
tombs of local heroes. When Dionysos came to Athens with his Satyr
attendants he would find the Pot-Contests as part of the funeral
ritual of the Anthesteria. He added to the festival wine and the
Satyrs. Small wonder that comedy, as in the Frogs, was at home in
the Underworld, and could in all piety parody a funeral[230] on the
stage.
 Thucydides has given us four examples of sanctuaries outside the
polis which are ‘towards that part’ of it, but again, as in the first
clause, he seems to feel that if he has spoken the truth it is not the
whole truth, so he saves himself from misunderstanding by an
additional clause, ‘and other ancient sanctuaries are placed here.’
It would be idle to try and give a complete list of all the
sanctuaries that were situated in this particular region, still more idle
to decide of what particular sanctuaries Thucydides was thinking.
The precinct of Aglauros and the Anakeion on the North side, the
sanctuary of the Semnae and the Amyneion on the West, the
sanctuary of Aphrodite Pandemos and that of Themis on the West
and South-West are all ‘towards’ the approach. Three out of these,
the Amyneion, the sanctuary of the Semnae, and the sanctuary of
Aphrodite Pandemos, are of such interest in themselves and so
essential to the forming of a picture of the sanctities of ancient
Athens that a word must be said of each.

The Amyneion. The Amyneion, or sanctuary of Amynos[231], is

known to us only through monumental evidence, brought to light in
the recent excavations. Its discovery is one of the things that make
us feel suddenly how much of popular faith we, relying as we must
almost wholly on literature, may have utterly lost.
 Fig. 30.

If after leaving the precinct of Dionysos in-the-Marshes we follow

the main road for about 35 metres, we come on a precinct (Fig. 30)
of much smaller size and of quadrangular shape, which abuts on the
road and along the North side of which a narrow foot-path leads up
to the Acropolis. The precinct-walls are of hard blue calcareous stone
from the Acropolis and neighbouring hills, and the masonry is good
polygonal. The entrance-gate (A), which has been rebuilt in Roman
times, is at the North-West corner. A little to the East of the middle
of the precinct, and manifestly of great importance, is a well (B).
The natural supply of this well was reinforced by a conduit-pipe,
which leads direct into it from the great water-course of Peisistratos,
which will later (p. 119) be described. Near the well are remains of a
small hero-chapel, and within this was found the lower part of a
marble sacrificial table (C), decorated with two snakes. The masonry
of the precinct wall, the well, and the shrine all point to a date at the
time of Peisistratos. Even before the limits of this precinct were fairly
made out the excavators came upon a number of fragments of
votive offerings of a familiar type. Such are reliefs representing parts
of the human body, breasts and the like, votive snakes, and reliefs
representing worshippers approaching a god of the usual Asklepios
type. Conspicuous among these was a fine well-preserved relief (Fig.
31), depicting a man holding a huge leg, very clearly marked with a
varicose vein, exactly where, doctors say, a varicose vein should be.
The inscription[232] above the figure is unfortunately so effaced that
no facts emerge save that the dedicator, the man who holds the leg,
was the son of a certain Lysimachos, and was of the deme
Acharnae. The style of the letters and of the sculpture dates the
monument as of about the first half of the 4th century b.c. It was
clear enough that the excavators had come on the precinct of a god
of healing, and a few decades ago the precinct would have been
labelled without more ado as ‘sacred to Asklepios.’ We should then
have been left with the curious problem, Why had Asklepios two
precincts, one on the South, one on the West? We know that
Asklepios made his triumphant entry into the great precinct on the
South slope in 421 b.c.; if he had had a precinct on the West slope
since the days of Peisistratos, why did he leave it?
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!

textbookfull.com