Download the full version of the textbook now at textbookfull.

com

Fundamentals of RCS Prediction Methodology

using Parallelized Numerical Electromagnetics
Code NEC and Finite Element Pre processor
Vineetha Joy
https://textbookfull.com/product/fundamentals-of-
rcs-prediction-methodology-using-parallelized-
numerical-electromagnetics-code-nec-and-finite-
element-pre-processor-vineetha-joy/

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

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

Electromagnetics through the Finite Element Method: A

Simplified Approach Using Maxwell's Equations 1st Edition
José Roberto Cardoso
https://textbookfull.com/product/electromagnetics-through-the-finite-
element-method-a-simplified-approach-using-maxwells-equations-1st-
edition-jose-roberto-cardoso/
textbookfull.com

Numerical Solution of Differential Equations: Introduction

to Finite Difference and Finite Element Methods 1st
Edition Zhilin Li
https://textbookfull.com/product/numerical-solution-of-differential-
equations-introduction-to-finite-difference-and-finite-element-
methods-1st-edition-zhilin-li/
textbookfull.com

Harmonic Balance Finite Element Method: Applications in

Nonlinear Electromagnetics and Power Systems 1st Edition
Junwei Lu
https://textbookfull.com/product/harmonic-balance-finite-element-
method-applications-in-nonlinear-electromagnetics-and-power-
systems-1st-edition-junwei-lu/
textbookfull.com

Reimagining Philosophy and Technology Reinventing Ihde

Glen Miller

https://textbookfull.com/product/reimagining-philosophy-and-
technology-reinventing-ihde-glen-miller/

textbookfull.com
The Empire of Disgust: Prejudice, Discrimination, and
Policy in India and the US 1st Edition Zoya Hasan

https://textbookfull.com/product/the-empire-of-disgust-prejudice-
discrimination-and-policy-in-india-and-the-us-1st-edition-zoya-hasan/

textbookfull.com

The Coconut Palm Cocos nucifera L Research and Development

Perspectives V. Krishnakumar

https://textbookfull.com/product/the-coconut-palm-cocos-nucifera-l-
research-and-development-perspectives-v-krishnakumar/

textbookfull.com

Handbook of Sustainability in Additive Manufacturing

Volume 1 1st Edition Subramanian Senthilkannan Muthu

https://textbookfull.com/product/handbook-of-sustainability-in-
additive-manufacturing-volume-1-1st-edition-subramanian-senthilkannan-
muthu/
textbookfull.com

Neurorheumatology A Comprehenisve Guide to Immune Mediated

Disorders of the Nervous System Tracey A. Cho

https://textbookfull.com/product/neurorheumatology-a-comprehenisve-
guide-to-immune-mediated-disorders-of-the-nervous-system-tracey-a-cho/

textbookfull.com

The Making of Islamic Heritage: Muslim Pasts and Heritage

Presents 1st Edition Trinidad Rico (Eds.)

https://textbookfull.com/product/the-making-of-islamic-heritage-
muslim-pasts-and-heritage-presents-1st-edition-trinidad-rico-eds/

textbookfull.com
Puppy Love (Her Secret Menagerie #1) 1st Edition Katelyn
Beckett

https://textbookfull.com/product/puppy-love-her-secret-
menagerie-1-1st-edition-katelyn-beckett/

textbookfull.com
SPRINGER BRIEFS IN ELEC TRIC AL AND COMPUTER
ENGINEERING  COMPUTATIONAL ELECTROMAGNETICS

Vineetha Joy
G. L. Rajeshwari
Hema Singh
Raveendranath U. Nair

Fundamentals
of RCS Prediction
Methodology using
Parallelized Numerical
Electromagnetics
Code (NEC) and Finite
Element Pre-processor

123
SpringerBriefs in Electrical and Computer
Engineering

SpringerBriefs in Computational Electromagnetics

Series Editors
K. J. Vinoy, Indian Institute of Science, Professor, Electrical Communication
Engineering, Indian Institute of Science, Bangalore, Karnataka, India
Rakesh Mohan Jha (Late), Centre for Electromagnetics, CSIR-National Aerospace
Laboratories, Bangalore, Karnataka, India
The phenomenon of electric and magnetic field vector (wave) propagation through the
free-space, or any other medium is considered within the ambit of electromagnetics. The
media themselves, in general, could be of diverse type, such as linear/non-linear,
isotropic/non-isotropic, homogeneous/inhomogeneous, reciprocal/non-reciprocal, etc. Such
electromagnetic wave propagation problems are formulated with the set of Maxwell’s
equations. Computational Electromagnetics endeavors to provide the solution to the
Maxwell’s equations for a given formulation. It is often difficult to find closed forms
solutions to the Maxwell’s equation formulations. The advent of computers, and particularly
the initial developments of efficient coding for numerical analysis, encouraged the
development of numerical electromagnetics. A second motivation came from the interaction
of the electromagnetic wave with the matter. This could be visualized as scattering bodies,
which required incorporation of the phenomena of reflection, refraction, diffraction and
polarization. The finite/large nature of the scatterer required that problem of electromagnetics
is considered with respect to the operation wavelength leading to the classification of
low-frequency, high-frequency and resonance region problems. This also inspired various
asymptotic and grid-based finite-method techniques, for solving specific electromagnetic
problems. Surface modeling and ray tracing are also considered for such electromagnetic
problems. Further, design optimization towards hardware realization have led to the recourse
to various soft computing algorithms. Computational Electromagnetics is deemed to
encompass the numerical electromagnetics along with all other above developments. With
the wide availability of massively parallel high performance parallel computing platforms,
new possibilities have emerged for reducing the computation time and developing macro
models that can even be employed for several practical multi-physics scenarios. Both volume
and surface discretization methods have been given a new boost, and several acceleration
techniques including GPU based computation, learning based approaches, and model order
reduction have been attempted. Limitations of generating meshes and modifying these for
parametric estimation have been addressed by statistical approaches and smart solvers. Many
nature-inspired algorithms and other soft computing approaches have been employed for
electromagnetic synthesis problems. One of the recent additions is Game Theoretic
optimization.
Finally, the emergence of Computational Electromagnetics has been motivated by myriad
applications. These diverse application include but are not restricted to those in Electronics
and Communication, Wireless Propagation, Computer Hardware, Aerospace Engineering,
Biomedical Engineering, Radio-astronomy, Terahertz Technology, Photonics, etc for
modelling of devices, components, systems and even large structures. Some of the
well-known applications are in Analysis and design of radio frequency (RF) circuit, antennas
and systems, Analysis of antenna on structures, radar imaging, radar cross section
(RCS) computation and reduction, and analysis of electromagnetic wave-matter interactions
at discrete, random and periodic geometries including metamaterials. Authors are encouraged
to submit original research work in the area of Computational Electromagnetics. The content
could be either theoretical development, or specific to particular applications. This Series also
encourages state-of-the-art reviews and easy to comprehend tutorials.

More information about this subseries at http://www.springer.com/series/13885

Vineetha Joy G. L. Rajeshwari
• •

Hema Singh Raveendranath U. Nair

•

Fundamentals of RCS
Prediction Methodology
using Parallelized Numerical
Electromagnetics Code
(NEC) and Finite Element
Pre-processor

123
Vineetha Joy G. L. Rajeshwari
Centre for Electromagnetics (CEM) Centre for Electromagnetics (CEM)
CSIR-National Aerospace Laboratories CSIR-National Aerospace Laboratories
Bengaluru, Karnataka, India Bengaluru, Karnataka, India

Hema Singh Raveendranath U. Nair

Centre for Electromagnetics (CEM) Centre for Electromagnetics (CEM)
CSIR-National Aerospace Laboratories CSIR-National Aerospace Laboratories
Bengaluru, Karnataka, India Bengaluru, Karnataka, India

ISSN 2191-8112 ISSN 2191-8120 (electronic)

SpringerBriefs in Electrical and Computer Engineering
ISSN 2365-6239 ISSN 2365-6247 (electronic)
SpringerBriefs in Computational Electromagnetics
ISBN 978-981-15-7163-3 ISBN 978-981-15-7164-0 (eBook)
https://doi.org/10.1007/978-981-15-7164-0
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021
This work is subject to copyright. All rights are solely and exclusively licensed 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, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
To
Our Families
Preface

With the advent of stealth technology, precise computation of radar cross section
(RCS) has become an inevitable component in the efficient design and development
of military vehicles. However, the scattering characterization of stealth platforms is
undoubtedly the most demanding problem in modern applied electromagnetics.
When it comes to the computation of scattered fields from conducting structures,
the accuracy and efficacy of a method of moments (MoM)-based integral equation
formulation are difficult to be surpassed. Among MoM-based solvers, Numerical
Electromagnetics Code (NEC) is a versatile open-source computer program used
for the electromagnetic analysis of metallic structures in the presence of sources or
incident fields. Since its development, NEC has continued to be one of the most
widely used electromagnetic simulation codes even in the presence of commercial
MoM-based solvers. This can be attributed to the availability of the well-
documented computational engine. However, a serious drawback of NEC is the
absence of an appropriate module for the wire-grid-based meshing of geometries
and the consequent generation of segmentation data in an NEC compatible format.
This is one of the most intricate and time-consuming steps in the computation of
RCS using NEC. In this regard, this brief presents a detailed methodology for the
computation of RCS of metallic structures using a parallelized version of NEC in
conjunction with a finite element pre-processor, which has been strategically
incorporated for geometry modelling catering to NEC guidelines. It includes a
thorough overview of the theoretical background of NEC including all relevant
aspects of formulation and modelling. This book will serve as a valuable resource
for students, researchers, scientists, and engineers working in the field of RCS
predictions and measurements.

Bengaluru, India Vineetha Joy

G. L. Rajeshwari
Hema Singh
Raveendranath U. Nair

vii
Acknowledgements

First and foremost, we would like to thank the God Almighty for being the beacon
of hope during the course of this research work.
Further, we would like to thank Shri. Jitendra J. Jadhav, Director, CSIR-National
Aerospace Laboratories, Bengaluru, for the permission to write this SpringerBrief.
We would also like to acknowledge the valuable suggestions from our col-
leagues at the Centre for Electromagnetics during the course of writing this book.
But for the concerted support and encouragement of Ms. Swati Meherishi,
Editorial Director, Applied Sciences and Engineering, and Ms. Muskan Jaiswal of
Springer Nature India Private Limited, it would not have been possible to bring out
this book within such a short span of time.
We are also forever beholden to our family members for their incessant inspi-
ration which supported us to stay at rough tracks. Vineetha Joy would like to
specially thank her husband Johnu George for his constant support and encour-
agment during the course of this work.

ix
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 Theoretical Background for the Computation of Radar
Cross-Section (RCS) Using NEC . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1 Formulation of Integral Equations (IE) . . . . . . . . . . . . . . . . . . . . . 5
2.1.1 Electric-Field Integral Equation (EF-IE) . . . . . . . . . . . . . . 5
2.1.2 Magnetic-Field Integral Equation (MF-IE) . . . . . . . . . . . . . 6
2.1.3 EF-IE/MF-IE Combined Equation . . . . . . . . . . . . . . . . . . . 7
2.2 Computation of Scattered Fields . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Numerical Solution of Integral Equations . . . . . . . . . . . . . . . . . . . 9
2.3.1 Selection of Basis Functions and Weight
Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3.2 Solution of Matrix Equation . . . . . . . . . . . . . . . . . . . . . . . 11
2.4 Strategy for Parallelization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3 Modelling Guidelines and Input/Output Formats in NEC . . . . . . . . 17
3.1 Modelling of Wires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2 Modelling of Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3 Format of NEC Input File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.4 Format of NEC Output File . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4 Methodology for the Computation of RCS using Parallelized NEC
and Finite Element Pre-processor . . . . . . . . . . . . . . . . . . . . . . . . . .. 23
4.1 Overview of Finite Element Pre-processor . . . . . . . . . . . . . . . . .. 23
4.2 Revised Methodology for the Computation of RCS using
Parallelized NEC and Finite Element Pre-processor . . . . . . . . . .. 25
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27

xi
xii Contents

5 Studies on Pre-selected Structures . . . . . . . . . . . . . . . . . . . . . . . . . . 29

5.1 Monostatic RCS of a PEC Dihedral . . . . . . . . . . . . . . . . . . . . . . . 29
5.2 Monostatic RCS of NASA Almond . . . . . . . . . . . . . . . . . . . . . . . 30
5.3 Monostatic RCS of Metallic Single Ogive . . . . . . . . . . . . . . . . . . 33
5.4 Monostatic RCS of Rocket-Shaped Model . . . . . . . . . . . . . . . . . . 35
5.5 Monostatic RCS of Micro Aerial Vehicle (MAV) . . . . . . . . . . . . . 37
5.6 Bi-Static RCS of Unmanned Aerial Vehicle (UAV) . . . . . . . . . . . 39
5.7 Monostatic RCS of Automobile . . . . . . . . . . . . . . . . . . . . . . . . . . 40
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Appendix A: Libraries used for Parallel Computations . . . . . . . . . . . . . . 55

Appendix B: Format of Command Lines in NEC . . . . . . . . . . . . . . . . . . 61
Appendix C: Format of Various Output Files . . . . . . . . . . . . . . . . . . . . . 67
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
About the Authors

Mrs. Vineetha Joy is working as Scientist at Centre for Electromagnetics of

CSIR-National Aerospace Laboratories, Bangalore, India since March, 2016. She
obtained B. Tech in Electronics and Communication Engineering with Second
Rank from University of Calicut in 2011 and received M. Tech degree with First
Rank in RF & Microwave Engineering from Indian Institute of Technology (IIT),
Kharagpur in 2014. She is actively involved in various R&D programs spanning
domains like Computational Electromagnetics, Electromagnetic Design and
Performance Analysis of Radomes, Design and Development of broadband radar
absorbing structures (RAS), Hybrid Numerical Techniques for Scattering Analysis,
and Electromagnetic Characterization of Potential Materials for Airborne
Structures. She has authored/co-authored several research publications including
peer reviewed journal papers, symposium papers, technical documents and test
reports.

Ms. G. L. Rajeshwari obtained M. Tech. in Digital Electronics and Communi-

cation in 2019 from Department of Electronics & Communication Engineering,
Bangalore Institute of Technology, Bengaluru, India and B.E. in Electronics &
Communication Engineering, in 2017 from Nitte Meenakshi Institute of Tech-
nology, Bengaluru. Her research interest includes Computational Electromagnetics,
Hybrid Numerical Techniques for Scattering Analysis etc.

Dr. Hema Singh is working as Senior Principal Scientist in Centre for Electro-
magnetics, National Aerospace Laboratories (CSIR-NAL), Bangalore, India. She
has received Ph.D. degree in Electronics Engineering from IIT-BHU, Varanasi
India in Feb. 2000. For the period 1999-2001, she was Lecturer in Physics at P.G.
College, Kashipur, Uttaranchal, India. She was a Lecturer in EEE of Birla Institute
of Technology & Science (BITS), Pilani, Rajasthan, India, for the period
2001-2004. She joined CSIR-NAL as Scientist in January 2005. Her active areas of
research are Computational Electromagnetics for Aerospace Applications, EM
analysis of propagation in an indoor environment, Phased Arrays, Conformal
Antennas, Radar Cross Section (RCS) Studies including Active RCS Reduction.

xiii
xiv About the Authors

She has contributed in the projects not only sponsored by DRDO on low RCS
phased array, Active RCS reduction, RAS development, but also in the project
sponsored by Boeing USA on EM analysis of RF field build-up within Boeing 787
Dreamliner, She received Best Woman Scientist Award in CSIR-NAL, Bangalore
for period of 2007-2008 for her contribution in the area of active RCS reduction.
Dr. Singh co-authored 14 books, 2 book chapters, 7 software copyrights, 370
scientific research papers and technical reports.

Dr. Raveendranath U. Nair is working as Senior Principal Scientist & Head in

Centre for Electromagnetics, National Aerospace Laboratories (CSIR-NAL),
Bangalore, India. He holds a Ph.D. degree in Microwave Electronics from MG
University, India. He has over 20 years’ experience in the field of electromagnetic
design and analysis of radomes, design and development of FSS based structures
for airborne platforms, radar cross section (RCS) studies, design and development
of artificially engineered materials etc. He has contributed significantly to the
National Radome Programs including Doppler Weather Radar (DWR) radome
installed at SHAAR Sreeharikota, Fire Control Radar (FCR) radome for Jaguar
aircraft, Astra missile ceramic radome, nosecone radome for Saras aircraft, multi-
band radomes for TU-142M aircraft etc. He has authored/co-authored over 200
research publications including peer reviewed journal papers, symposium papers,
technical reports, and two books.
List of Figures

Fig. 2.1 a 1D column distribution, b 2D block-cyclic distribution . . . .. 14

Fig. 2.2 Variation in memory requirements of NEC at a particular node
with respect to number of segments . . . . . . . . . . . . . . . . . . . . .. 15
Fig. 3.1 Geometry of a wire segment . . . . . . . . . . . . . . . . . . . . . . . . . .. 18
Fig. 3.2 Surface patch a general geometry b rectangular patch
c triangular patch d quadrilateral patch . . . . . . . . . . . . . . . . . . .. 20
Fig. 4.1 Graphical user interface (GUI) of HyperMesh . . . . . . . . . . . . .. 24
Fig. 4.2 Workflow for the computation of RCS using NEC
in conjunction with HyperMesh . . . . . . . . . . . . . . . . . . . . . . . .. 25
Fig. 4.3 Procedure for computation of RCS using NEC
and HyperMesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27
Fig. 5.1 Wire grid model of PEC dihedral in HyperMesh . . . . . . . . . . .. 30
Fig. 5.2 A portion of the NEC compatible segmentation data,
corresponding to PEC dihedral, extracted from Nastran file . . .. 31
Fig. 5.3 NEC input file for the computation of monostatic RCS
of PEC dihedral at 600 MHz . . . . . . . . . . . . . . . . . . . . . . . . . .. 31
Fig. 5.4 Monostatic RCS of PEC dihedral . . . . . . . . . . . . . . . . . . . . . . .. 32
Fig. 5.5 Wire grid model of NASA almond in HyperMesh . . . . . . . . . .. 33
Fig. 5.6 A portion of the NEC compatible segmentation data,
corresponding to NASA almond, extracted from Nastran file . .. 34
Fig. 5.7 NEC input file for the computation of monostatic RCS
of NASA almond (horizontal polarization) . . . . . . . . . . . . . . . .. 34
Fig. 5.8 NEC input file for the computation of monostatic RCS
of NASA almond (vertical polarization) . . . . . . . . . . . . . . . . . .. 35
Fig. 5.9 Monostatic RCS of metallic NASA almond. a Horizontal
polarization b vertical polarization . . . . . . . . . . . . . . . . . . . . . .. 36
Fig. 5.10 Wire grid model of metallic single ogive in HyperMesh . . . . .. 37
Fig. 5.11 A portion of the NEC compatible segmentation data,
corresponding to metallic single ogive, extracted
from Nastran file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38

xv
xvi List of Figures

Fig. 5.12 NEC input file for the computation of monostatic

RCS of metallic single ogive (horizontal polarization) . . . . . . . . 38
Fig. 5.13 NEC input file for the computation of monostatic
RCS of metallic single ogive (vertical polarization) . . . . . . . . . . 39
Fig. 5.14 Monostatic RCS of metallic single ogive. a Horizontal
polarization b vertical polarization . . . . . . . . . . . . . . . . . . . . . . . 40
Fig. 5.15 Cross-sectional view of the metallic rocket-shaped model . . . . . 41
Fig. 5.16 Wire grid model of metallic rocket shaped geometry
in HyperMesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Fig. 5.17 A portion of the NEC compatible segmentation data,
corresponding to metallic rocket shaped model, extracted
from Nastran file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Fig. 5.18 NEC input file for the computation of monostatic
RCS of rocket-shaped model . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Fig. 5.19 Monostatic RCS of metallic rocket shaped model. . . . . . . . . . . . 43
Fig. 5.20 CAD model of metallic MAV . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Fig. 5.21 Wire grid model of Metallic MAV in HyperMesh . . . . . . . . . . . 44
Fig. 5.22 A portion of the NEC compatible segmentation data,
corresponding to metallic MAV, extracted from Nastran file . . . 45
Fig. 5.23 NEC input file for the computation of monostatic
RCS of metallic MAV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Fig. 5.24 Monostatic RCS of metallic MAV . . . . . . . . . . . . . . . . . . . . . . . 46
Fig. 5.25 CAD model of metallic UAV . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Fig. 5.26 Wire grid model of metallic UAV in HyperMesh . . . . . . . . . . . 47
Fig. 5.27 A portion of the NEC compatible segmentation data,
corresponding to metallic UAV, extracted from Nastran file . . . 48
Fig. 5.28 NEC input file for the computation of bi-static
RCS of metallic UAV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Fig. 5.29 Bi-static RCS of metallic UAV model . . . . . . . . . . . . . . . . . . . . 49
Fig. 5.30 CAD model of metallic car . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Fig. 5.31 Wire grid model of metallic car in HyperMesh . . . . . . . . . . . . . 50
Fig. 5.32 A portion of the NEC compatible segmentation data,
corresponding to metallic car, extracted from Nastran file . . . . . 51
Fig. 5.33 NEC input file for the computation of monostatic
RCS of metallic car . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Fig. 5.34 Monostatic RCS of metallic car model . . . . . . . . . . . . . . . . . . . . 52
Fig. A.1 Structure of a general MPI program . . . . . . . . . . . . . . . . . . . . . . 58
Fig. A.2 Format of a typical PBS script . . . . . . . . . . . . . . . . . . . . . . . . . . 59
List of Tables

Table 3.1 List of geometry commands . . . . . . . . . . . . . . . . . . . . . . . . . .. 21

Table 3.2 List of program control commands . . . . . . . . . . . . . . . . . . . . .. 22
Table 5.1 Simulation and meshing parameters to be used for the
computation of monostatic RCS of PEC dihedral . . . . . . . . . .. 30
Table 5.2 Simulation and meshing parameters to be used for the
computation of monostatic RCS of NASA almond . . . . . . . . .. 33
Table 5.3 Simulation and meshing parameters to be used for the
computation of monostatic RCS of metallic single ogive . . . .. 37
Table 5.4 Simulation and meshing parameters to be used for the
computation of monostatic RCS of metallic rocket-shaped
geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 42
Table 5.5 Simulation and meshing parameters to be used for the
computation of monostatic RCS of metallic MAV . . . . . . . . .. 45
Table 5.6 Simulation and meshing parameters to be used for the
computation of bi-static RCS of metallic UAV . . . . . . . . . . . .. 48
Table 5.7 Simulation and meshing parameters to be used for the
computation of monostatic RCS of metallic car . . . . . . . . . . . . 50
Table B.1 Options available for Int-1 (EX command) . . . . . . . . . . . . . . . . 63
Table B.2 Options available for Int-1 (RP command) . . . . . . . . . . . . . . . . 64
Table B.3 Options available for P (RP command) . . . . . . . . . . . . . . . . . . 64
Table B.4 Options available for N (RP command) . . . . . . . . . . . . . . . . . . 65
Table B.5 Options available for G (RP command) . . . . . . . . . . . . . . . . . . 65
Table B.6 Options available for A (RP command) . . . . . . . . . . . . . . . . . . 65

xvii
Chapter 1
Introduction

The burgeoning research and developmental activities happening in various domains

like antenna design, wireless communication, stealth technologies, electromagnetic
interference/compatibility (EMI/EMC), etc., have made systems much more complex
than ever before. Considering the huge expenses related to fabrication, it is not
practically possible to modify a device if its performance is not within the stipulated
limits. Therefore, techniques capable of precise characterization of complex systems
are the need of the hour. Closed-form analytical solutions are available for only
particular problems, which can hardly be directly applied in real-world systems.
This gap has led to extensive research on computational electromagnetics (CEM)
where numerical methods are used for solving electromagnetic problems. Among
all the numerical techniques, method of moments (MoM) is highly accurate and
efficient especially for the analysis of scattering from conducting structures. In this
regard, Numerical Electromagnetics Code (NEC) is a versatile MoM-based open-
source code (written in Fortran) used for the electromagnetic analysis of metallic
structures in the presence of sources or incident fields (Burke and Poggio 1981;
Burke et al. 2004). The development of NEC has been spearheaded by Lawrence
Livermore Laboratory, USA, with the support of the Naval Ocean Systems Center
and the Air Force Weapons Laboratory.
Since its development, NEC has continued to be one of the most widely used
electromagnetic simulation codes even in the presence of commercial MoM solvers.
This attributes to the availability of the well-documented powerful computational
platform, which can be used as a foundation for further developments in the area of
electromagnetic analysis. NEC does not use many of the simplifying assumptions
employed by other high-frequency asymptotic techniques and thus provides a highly
accurate and resourceful tool for electromagnetic studies.
The numerical solution of integral equations, for the induced currents on metallic
structures by sources or impinging electromagnetic waves, is the essence of NEC.
The excitations can be defined in terms of voltage sources or plane waves of
linear/elliptical polarization. Various electromagnetic entities like non-radiating

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 1
V. Joy et al., Fundamentals of RCS Prediction Methodology using Parallelized Numerical
Electromagnetics Code (NEC) and Finite Element Pre-processor,
SpringerBriefs in Computational Electromagnetics,
https://doi.org/10.1007/978-981-15-7164-0_1
2 1 Introduction

networks/transmission lines, perfect/imperfect conductors, perfectly or imperfectly

conducting ground planes, etc., can be easily included in NEC. The output of NEC can
be tailored for specific requirements like induced current and charge densities, near
magnetic or electric fields, impedance or admittance, radiated fields, etc. Further-
more, many of the commonly encountered performance parameters such as gain and
directivity are also available. Therefore, the program is well suited for both antenna
analysis and radar cross section (RCS) predictions (Burke and Poggio 1981).
Although numerical methods like MoM can be used in a wide range of radiation
and scattering problems, the computational complexity increases with increase in the
electrical size of the structure under consideration. The memory requirements have
been found to exhibit a quadratic growth with increase in the number of segments. A
parallelized version of NEC has been introduced (Rubinstein et al. 2003, Rubinstein
2004) to tackle this problem. The two most time-consuming steps in evaluating the
response of a structure, viz. (i) computation of the interaction matrix elements and
(ii) solution of the resultant matrix equation, have been efficiently parallelized in
Rubinstein et al. (2003).
In addition to the memory constraints, a serious issue in the computation of RCS
using NEC is the wire-grid-based meshing of geometries and the generation of NEC
input file with the segmentation data included in accordance with NEC guidelines.
This is an extremely difficult and error-prone process, especially for complex struc-
tures (Hubing et al. 1994; Toit and Davidson 1995; Ross et al. 1999), and further-
more, the code will often run in the presence of such errors and produce results that
are plausible but incorrect. Therefore, an appropriate pre-processor is essential for
the accurate prediction of RCS using NEC.
The aim of this book is to provide a strong foundation for the computation of
radar cross section (RCS) of aerospace structures using a parallelized version of
NEC in conjunction with a finite element pre-processor. This book has five chapters.
It begins with a thorough overview of the theoretical background of NEC including
all relevant aspects of formulation of integral equation and its numerical solution.
The strategy used for parallelization of the code is also included in Chap. 2. The
geometrical modelling of structures and the input/output formats are described in
Chap. 3. The revised methodology for the computation of RCS incorporating all the
required steps and details is discussed elaborately in Chap. 4 followed by case studies
and validations in Chap. 5.

References

Burke, G.J., and A.J. Poggio. 1981 January. Numerical Electromagnetics Code (NEC)—Method
of Moments, Lawrence Livermore National Laboratory, Livermore, California, Technical
Document, Rep. UCID-18834, 719.
Burke, G.J., E.K. Miller, and A.J. Poggio. 2004 June. “The Numerical Electromagnetics Code
(NEC)—A Brief history.” In Proceedings of IEEE Antennas and Propagation Society Symposium,
vol. 3, 2871–2874, Monterey, California.
References 3

Hubing, T.H., C.H.H. Lim, and J. Drewniak. 1994 March.“A Geometry Description Language for
3D Electromagnetic Analysis Codes.” In Proceedings of the 10th Annual Review of Progress in
Applied Computational Electromagnetics, Monterey, CA, 417–422.
Rubinstein, A., F. Rachidi, M. Rubinstein, and B. Reusser. 2003. A Parallel Implementation of NEC
for the Analysis of Large Structures. IEEE Transactions on Electromagnetic Compatibility 45
(2): 177–188.
Ross, J.E., L.L. Nagy, and J. Szostka. 1999. “CAD tools for vehicular antennas.” In Proceedings of
the 1999 IX National Symposium of Radio Science, Poznan, Poland, March 16–17.
Rubinstein, A. 2004 January. Simulation of Electrically Large Structures in EMC Studies: Applica-
tion to Automotive EMC, Ph.D. dissertation, École polytechnique fédérale de Lausanne (EPFL),
https://doi.org/10.5075, 156.
Toit, C.D., and D.B. Davidson. 1995. Wiregrid: A NEC2 Pre-processor. Journal of Applied
Computational Electromagnetics Society 10: 31–39.
Chapter 2
Theoretical Background
for the Computation of Radar
Cross-Section (RCS) Using NEC

The entire theoretical framework of NEC starting from the definition of integral
equations to the solution of matrix equation is briefly introduced in this chapter. The
basis functions used for representing the unknown current densities and the weight
functions used for testing are also described in detail. Further, the strategy used for
parallelization of NEC (Rubinstein et al. 2003) is also included here.

2.1 Formulation of Integral Equations (IE)

NEC employs both an electric-field integral equation (EF-IE) and a magnetic-field

integral equation (MF-IE) for evaluating the electromagnetic behaviour of conducting
bodies under consideration. EF-IEs are mainly used in the case of thin-wire geome-
tries with small conductor volumes. On the other hand, MF-IE, which cannot be
applied for the thin-wire case, is more suitable for voluminous structures with smooth
surfaces. EF-IE and MF-IE are coupled together for a structure containing both wires
and surfaces. The theoretical formulation presented in the following sections is based
on the work reported in (Burke and Poggio 1981).

2.1.1 Electric-Field Integral Equation (EF-IE)

The integral representation for the electric field due to a surface current distribution
−
→
J S on a perfectly conducting body (S) is given by,
     −
−
→ →
 r ) = (− jη/4π k)
E( JS r  · Ḡ¯ r, r  dA (2.1)
S

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 5
V. Joy et al., Fundamentals of RCS Prediction Methodology using Parallelized Numerical
Electromagnetics Code (NEC) and Finite Element Pre-processor,
SpringerBriefs in Computational Electromagnetics,
https://doi.org/10.1007/978-981-15-7164-0_2
6 2 Theoretical Background for the Computation …

where,
     
G r, r = k 2 I + ∇∇ g r, r
 
 
  e− j k r −r 

g r, r = 
 
 
r − r 


√ μ0
k = ω μ0 ε0 ; η = ; I = x̂ x̂ + ŷ ŷ + ẑ ẑ
ε0

r and r represent observation point and source point, respectively.

−
→
The integral equation for the current induced on S by an incident field E Inc is
obtained by imposing the following boundary condition for r ∈ S:
−
→ −
→
r ) × E Sca (
n̂( r ) + E Inc (
r) = 0 (2.2)

−
→
r ) is the unit normal vector to the surface at r and E Sca is the scattered field
where n̂(
−
→ −
→Sca
due to the induced current J s . Substituting (2.1) for E in (2.2) yields the final
EF-IE as,

−
→ →   2
−   
r ) × E Inc (
−n̂( r ) = (− jη/4π k) × n̂(
r) × J s (r ) · k I + ∇∇ g r, r dA
S
(2.3)

Since NEC employs thin cylindrical wires for representing the conducting surface
(S), the vector integral equation in (2.3) will be transformed to a scalar integral
equation. The assumptions applicable for the thin-wire approximation used in most
of the cases are given below:
(i) Currents in the transverse direction can be ignored with respect to the currents
along axial direction of the wire.
(ii) Circumferential variation in the currents along axial direction can be ignored.
(iii) The unknown current may be expressed as a filament along the wire axis.
(iv) The boundary condition pertaining to the electric field is required to be imposed
only in the axial direction.

2.1.2 Magnetic-Field Integral Equation (MF-IE)

The integral representation for the magnetic field due to a surface current distribution
−
→
J s on a perfectly conducting body is given by,
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
2.1 Formulation of Integral Equations (IE) 7
   
−
→ −→ 
H (
r ) = (1/4π ) J s (r ) · ∇  g r, r dA (2.4)
S

where the differentiation is with respect to the integration variable r .

−
→ −
→
If an external incident field H Inc induces a current J s on S, then the total magnetic
field within the perfectly conducting body vanishes. Hence, the boundary condition
for r immediately within the surface S can be expressed as,
−
→Sca −
→
H ( r ) + H Inc (
r) = 0 (2.5)

−
→ −
→
where H Inc is the incident field without the structure and H Sca is the scattered field
−
→
due to the induced current J s given by (2.4).
−
→
Substituting H Sca in (2.5) yields the final MF-IE as,
     
−
→ −
→ 
r0 ) × H Inc (
−n̂( r0 ) = n̂(
r0 ) × (1/4π ) lim J s r × ∇  g r, r dA (2.6)
r→
r0
S

r0 ) is the normal vector directed outwards at the surface point r0 .

where n̂(
Similar to EF-IE, the vector integral Eq. (2.6) can also be resolved into two scalar
expressions in the directions of the orthogonal surface vectors tˆ1 and tˆ2 where

r0 ) × tˆ2 (
tˆ1 ( r0 ) = n̂(
r0 ) (2.7)

2.1.3 EF-IE/MF-IE Combined Equation

The EF-IE/MF-IE combined equation is used for a structure consisting of both wires
and surfaces. In such cases, r in (2.5) is restricted to wires and the integral for
−
→Sca
E ( r ) extends over the complete structure. Similarly, r0 in (2.6) is restricted to
−
→
surfaces and the integrals for H Sca (
r ) extends over the entire structure. The integral
can be simplified by using the thin-wire approximation on wires.
The final coupled integral equations (Burke and Poggio 1981) are,
    2   
−
→  ∂  dS 
−ŝ · E Inc (
r ) = (− jη/4π k) I (s  )
k 2 ŝ · s  − g 
r , r
∂s ∂s 
L
    
→     2 
− ∂ 
− ( jη/4π k) Js r k ŝ − ∇  g r, r dA ;
∂s
S1

for r on surfaces of wire (2.8)

8 2 Theoretical Background for the Computation …

  
−
→  
tˆ2 ( r ) = (−1/4π )tˆ2 (
r ) · H Inc ( r) · I s  sˆ × ∇  g r, r dS 
L

− (1/2)tˆ1 (
r ) · JS (
r ) − (1/4π ) tˆ2 (
r)
S1
−
→  
· J S (r ) × ∇  g r, r dA ;
for r on surfaces excluding wires (2.9)
  
−
→
−tˆ1 ( r ) = (1/4π )tˆ1 (
r ) · H Inc ( r) · I (s  ) sˆ × ∇  g r, r dS 
L

− (1/2)tˆ2 (
r ) · JS (
r ) + (1/4π ) tˆ1 (
r)
S1
 
· JS (r ) × ∇ g r, r dA ;


for r on surfaces excluding wires (2.10)

where s is the parameter corresponding to distance along

 the wire axis at r and ŝ
is the unit vector tangent to the wire axis at r. and represents integration over
L S1
wires and integration over surfaces excluding wires, respectively.

2.2 Computation of Scattered Fields

The scattered fields due to induced currents can be computed using a far-field approx-
imation. This is applicable when the observation points are far apart from the induced
current distribution by distances much larger compared to the wavelength as well
as the dimensions of the current distribution. The far-zone scattered fields from a
conducting body consisting of a wire region with contour L having linear current
−
→ −
→
distribution I (s) and a surface region S with surface current density J s ( r ) are
given by,
⎧
−
→ ⎨   − → 


r0 ) = ( jkη/4π ) × (e− jkro /ro ) ×
E ( (k · I (s))k − I(s) e j k·r
⎩
L
⎫
   ⎬
 −→  −
→ 
+ (k · J s (
r ))k − J s (
r ) e j k·r dA (2.11)
⎭
S
2.2 Computation of Scattered Fields 9


where k = |rroo | , k = 2π
λ
and ro is the position of the observation point.
The contour integral can be computed in closed form over each wire segment
for the cosine, sine, and constant components of the basis functions and this in turn
becomes a summation over the wire segments. Since the surface current on the patch
is represented by a delta function at the centre of the patch, the second integral
transforms to a summation over the patch elements.
−
→
For an incident plane wave with electric field E Inc , the NEC program constants are
defined such that the radar cross-section (RCS) σ/λ2 is displayed under the column
gain in the output file.
− 
→Sca 2
σ E 
= 4π 
→ 2
(2.12)
λ2 −
 E Inc 

−
→
where E Sca is given by (2.11).

2.3 Numerical Solution of Integral Equations

The integral equations described in the previous sections are solved numerically in
NEC by using the method of moments (MoM). Method of moments starts with a
conventional linear-operator equation as,

L( f ) = u (2.13)

where f is an unknown response, u is a known excitation and L is an integral operator.

The formulation proceeds by expressing the unknown function f as a sum of
known basis functions f j (defined in the domain of the integral operator L) as,


N
f = βj f j (2.14)
j=1

The unknown coefficients (β j ) are obtained by taking the inner product of (2.13)
with a set of weighting functions wi , i.e.

wi , L( f ) = wi , u ; i = 1, 2, . . . , N (2.15)

The inner product in the present case can be defined as,


 p, q = r ) q(
( p( r ))dA
S
10 2 Theoretical Background for the Computation …

where integration is over the surface of the structure.

Substituting (2.14) in (2.15) gives


N
  
β j wi , L f j = wi , u ; i = 1, 2, . . . , N (2.16)
j=1

The above equation can be expressed in matrix form as,

[G] [B] = [U ] (2.17)

  
where G i j = wi , L f j ; B j = β j ; Ui = wi , u .
The solution to (2.17) is

[B] = [G]−1 [U ] (2.18)

2.3.1 Selection of Basis Functions and Weight Functions

The critical factors determining the accuracy and convergence of MoM are the weight
functions and the basis functions. NEC employs different basis and weight functions
unlike Galerkin’s method. The weighing functions (wi ) are chosen as a set of delta
functions in accordance with the collocation method, where the integral equation is
enforced at specific sample points on the geometry, i.e.

wi (
r ) = δ(
r − ri )

where {ri } are the set of points on the conducting surface. In NEC, conducting wires
are sub-divided into short linear segments where a sample point is defined at the
centre of each segment. On the other hand, surfaces are represented by a set of flat
patches with a sample point defined at the patch centre.
The linear current density on wires and surface current density on surfaces are
the unknown quantities in NEC which are to be represented by a set of known basis
functions. NEC uses sub-domain basis functions to simplify the calculation of inner
product integrals and also to ensure that the matrix G is well conditioned.
In the case of wires, the current on each segment is approximated as a sum of
three terms, viz., a sine, a cosine, and a constant. The total current on ith segment
can thus be expressed as,

Ii (s) = Ai sin k(s − si ) + Bi cos k(s − si ) + Ci , |s − si | < i

(2.19)
2
2.3 Numerical Solution of Integral Equations 11

where si is the value of s at the centre of segment i, i is the length of segment i,

and k is the free space wave number.
The additional equations required to evaluate all the unknown coefficients in (2.19)
for each and every segment will be provided by the local boundary conditions. The
continuity of current and charge is one such condition to be enforced at a junction
of two segments with uniform radius. In the case of a junction with more than
two segments of different radii, the continuity of current may be generalized to
Kirchhoff’s current law.
The final solution requires the computation of the electric field at each segment
due to these currents. NEC uses three approximations of the integral equation kernel
for the computation of fields. These include (i) thin-wire approximation for most
of the cases, (ii) extended thin-wire approximation for thick wires, and (iii) current
element approximation for the case where segments are separated far apart by a
considerable distance. In all the three methods used for the evaluation of field at
a particular segment due to the induced current on a different segment, the field
computation is performed on the surface of the observation segment.
NEC models surfaces using small flat patches and the unknown surface current
on each patch are expressed as a summation of a set of pulse functions except where
a wire connection is present. When a wire connects to a surface, a special local
boundary condition has to be enforced as explained in (Burke and Poggio 1981).
The unknown surface current density over M p patches can be written as,

−
→  Mp
 
J s (
r) = (J1i t 1i + J2i t 2i )vi (
r) (2.20)
i=1

   
where t 1i = t 1 (
ri ), t 2i = t 2 (
ri ), ri is the position of the centre of patch i, vi (
r) = 1
for r on patch i and 0 otherwise. Thus, every patch has two unknown quantities and
the equation is enforced for each vector component at the sample point in a given
patch.

2.3.2 Solution of Matrix Equation

The matrix equation obtained by appropriately substituting the basis functions and
weight functions in (2.16) has to be finally solved for the unknown currents. The
equation can be alternately expressed as,

[G] [I ] = [E] (2.21)

where [G] is the interaction matrix, [I ] is the column vector with unknown currents,
and [E] is the excitation vector. NEC employs Gauss elimination method for solving
(2.21). Here, the matrix G is first factorized into the product of upper and lower
Random documents with unrelated
content Scribd suggests to you:
 Motolinia, Torquemada, Ixtlilxochitl, and Gomara agree in placing
the two oratories or shrines on the extreme eastern edge of the
platform, so that there was only just room for a man to pass round
them on the east side. The two oratories were separate one from the
other, each being enclosed within its own walls with a doorway
towards the west. The oratory of Huitzilopochtli was the larger of the
two and stood to the south. The oratory of Tlaloc stood to the north.
No measurements are given of the area covered by these two
oratories, but there is no suggestion that they were large buildings[7]
except in height. The roof and probably the upper stages were made
of wood (Torquemada), and we know that they were burnt during the
siege.
Height:—

Anonimo—“Ten or twelve men’s bodies.”

Torquemada—“Each in three stories, each story of great height.”
Motolinia—“The Great Temples had three stories above the
altars, all terraced and of considerable height.”

Ixtlilxochitl gives the height of the great Teocalli as over twenty-

seven brazas (150′). If this means the height from the ground to the
top of the Oratory of Huitzilopochtli it would very nearly agree with
the height given on the hypothetical section on Plate B.
In the description of the map of the city published in 1524 [see
‘Conquest of New Spain,’ vol. iii. (Hakluyt Society)] I called attention
to the “full human face probably representing the Sun” between the
Oratories of the Teocalli of Huitzilopochtli. The map is, I believe, in
error in placing the Teocalli on the west side of the Temple
Enclosure, but that the full human face is intended to represent the
sun is confirmed by the following passage from Motolinia[8]:—
“Tlacaxipenalistli.—This festival takes place when the sun
stood in the middle of Huichilobos, which was at the Equinox,
and because it was a little out of the straight[9] Montezuma
wished to pull it down and set it right.”
 The map of 1524 was probably drawn from a description given by
one of the Conquistadores, and if we turn to the pages of Gomara, an
author who was never in Mexico and who wrote only from hearsay, it
is easy to see how such a mistake in orientation arose.
Gomara, Historia General de las Indias—Conquista de Mejico. (El
Templo de Mejico.)
“This temple occupies a square, from corner to corner the length of
a crossbow shot. The stone wall has four gateways corresponding to
the four principal streets.... In the middle of this space is an edifice of
earth and massive stone four square like the court, and of the
breadth of fifty fathoms from corner to corner. On the west side there
are no terraces but 113 or 114 steps leading up to the top. All the
people of the city[10] look and pray towards the sunrise and on this
account they build their large temples in this manner.... In addition
to this tower with its chapels placed on the top of the pyramid, there
were forty or more other towers great and small on other smaller
Teocallis standing in the same enclosure (circuito) as this great one,
and although they were of the same form, they did not look to the
east but to other parts of the heaven, to differentiate them from the
Great Temple. Some were larger than others, and each one
(dedicated) to a different god.”
The confusion of thought between a temple that faced the east and
a temple where the worshippers faced the east is evident.
There can be little doubt that the steps of the Great Teocalli were
on the west side, that the Oratories of Huitzilopochtli and Taloc were
on the east side of the summit platform, and that their doorways
faced the west. The priest and worshippers faced the east to watch
the sunrise at the equinox in the narrow space between the two
oratories, and because the alignment was not quite correct
Montezuma wished to pull down the oratories and rebuild them.
Following from this, it appears to me that Duran was probably not
far from correct in placing the great green sacrificial stone “fronting
the two chambers,” but that Ixtlilxochitl was still more accurate in
placing it towards (hacia) the doorway of the sanctuary of
Huitzilopochtli. The heart of the human victim would be torn out and
held up to the rising sun from the spot where the priest stood to
observe the sunrise.
It will at once be urged against this solution of the difficulties
attending the orientation of the Great Teocalli that the plan and
tracings locate the Teocalli eight degrees from the east and west line,
and that, therefore, my explanation fails. To this I can only reply that
I plotted the measurements, taking the east and west line of the Calle
de Tacuba from the modern map as a datum, and this may vary
slightly from the ancient line of the street. Then I have observed in
Maya temples that sometimes the shrines stand slightly askew from
the base: this is clearly noticeable at Chichén Itzá. If the error of 8°
were divided between the lines of the Temple enclosure, the base of
the Teocalli, and the sides of the oratories, the difference would not
easily be perceptible.
Moreover, we cannot now ascertain the exact spot from which the
observation was made nor the distance between the two sanctuaries.
If, as Ixtlilxochitl states, it was towards the doorway of the sanctuary
of Huitzilopochtli and not between the two sanctuaries as is stated by
Duran, then the error would be reduced.

Recent Excavations.

We have now to consider the position of the Great Teocalli in

relation to the excavations made in the Calle de las Escallerillas when
pipes were being laid for the drainage of the city in the year 1900.
These excavations were watched on behalf of the Government by
Señor Don Leopoldo Batres, Inspector General of Archæological
Monuments, who published an account of his researches in 1902,
with a plan showing the position and depth below the surface at
which objects of archæological interest were discovered.
Unfortunately Señor Batres was already fully convinced that the
Great Teocalli faced the south and occupied more or less the position
of the present Cathedral.
At a spot marked a, in Tracing A2, 38 metres from the east end of
the Escalerillas, Señor Batres discovered a stairway of four masonry
steps which he states measured each 29 cm. in the rise and 22 cm. in
the tread, but unfortunately beyond this statement he gives no
information whatever regarding them. However, I presume that the
steps followed the same direction as a stairway of nine steps which
he had previously described and which will be alluded to
immediately. These three steps I have taken to be the central
stairway leading to the forecourt or apetlac of the Great Teocalli.
Señor Batres had already noted a stairway of nine steps, marked b
in Tracing A2, each measuring 22 cm. in rise and 26 cm. in tread.
This stairway was 2 metres wide and faced the west. The stairway
was apparently joined at one or both sides to a sloping wall[11].
Embedded in the débris which covered these steps was found an idol
of green stone measuring 75 cms. in height and 61 cms. in diameter.
The idol is now preserved in the National Museum.
I take the foot of this stairway of nine steps to have been in line
with the great stairway of the Teocalli, and it may have been part of
the great stairway itself; however, a stairway only two metres wide is
not likely to be the beginning of what must have been the principal
approach to the Teocalli, and I can only suggest that it may have
been a stairway leading to a niche which held the idol of green stone
and that the great stairways passed on either side of it. An idol in a
somewhat similar position can be seen on the Hieroglyphic Stairway
at Copan.
The Anonymous Conqueror. A Description written by a
Companion of Hernando Cortés.

XIV. What these Towers are like.

They build a square tower one hundred and fifty paces, or rather
more, in length, and one hundred and fifteen or one hundred in
breadth. The foundation of this building is solid; when it reaches the
height of two men, a passage is left two paces wide on three sides,
and on one of the long sides steps are made until the height of two
more men is reached, and the edifice is throughout solidly built of
masonry. Here, again, on three sides they leave the passage two
paces wide, and on the other side they build the steps, and in this
way it rises to such a height that the steps total one hundred and
twenty or one hundred and thirty.
There is a fair-sized plaza on the top and from the middle [of it]
arise two other towers which reach the height of ten or twelve men’s
bodies and these have windows above. Within these tall towers stand
the Idols in regular order and well adorned, and the whole house
highly decorated. No one but their high priest was allowed to enter
where the principal God was kept, and this god had distinct names in
different provinces; for in the great city of Mexico he was called
Horchilobos (Huitzilopochtli), and in another city named Chuennila
(Cholula) he was called Quecadquaal (Quetzalcoatl), and so on in the
others.
Whenever they celebrated the festivals of their Idols, they
sacrificed many men and women, boys and girls; and when they
suffered some privation, such as drought or excess of rain, or found
themselves hard pressed by their enemies, or suffered any other
calamity, then they made these sacrifices in the following manner....

XXI. About their Temples and Mosques.

 They have in this great city very great mosques or temples in which
they worship and offer sacrifices to their Idols; but the Chief Mosque
was a marvellous thing to behold, it was as large as a city. It was
surrounded by a high masonry wall and had four principal doorways.
Fray Toribio Benavente or Motolinia, Historia de los Indios de
Nueva España, Treatise No. I. Ch. XII.
There have never been seen or heard of before such temples as
those of this land of Anahuac or New Spain, neither for size and
design nor for anything else; and as they rise to a great height they
must needs have strong foundations; and there was an endless
number of such temples and altars in this country, about which a
note is here made so that those who may come to this country from
now onwards may know about them, for the memory of them all has
already almost perished.
These temples are called Teocallis, and throughout the land we
find that in the principal part of the town a great rectangular court is
constructed; in the large towns they measured from corner to corner
the length of a crossbow shot, in the lesser towns the courts are
smaller.
This courtyard they surround with a wall and many of the walls are
embattled; their gateways dominate the principal streets and roads,
for they are all made to converge towards the court; and so as to give
greater honour to their temples they lay out the roads very straight
with rope line for a distance of one or two leagues, and it is a thing
worth seeing from the top of the principal temple, how straight all
the roads come from all the lesser towns and suburbs and converge
towards the Court of the Teocallis.
In the most conspicuous place in this court would stand a great
rectangular block (cepa). So as to write this description I measured
one in a moderate-sized town named Tenanyocan [Tenayoca] and
found that it measured forty fathoms from corner to corner all built
up with a solid wall, on the outside the wall was of stone, and the
inside was filled up with stone only or with clay and adobe; others
were built of earth well tamped.
As the structure rose it contracted towards the centre and at the
height of a fathom and a half or two fathoms there were some ledges
going inwards, for they did not build it in a straight line, and the
thick foundation was always worked towards the centre so as to give
it strength and as the wall rose it got narrower; so that when it got to
the top of the Teocalli it had narrowed and contracted itself seven or
eight fathoms on each side, both by the ledges and the wall leaving
the foundation [mound] on the top thirty-four or thirty-five fathoms.
On the west side were the steps and ascent, and above on the top
they constructed two great altars, placing them towards the east side,
so that there was no more space left behind them than was sufficient
to enable one to walk round them. One altar was to the right and the
other to the left. Each one stood by itself with its own walls and
hood-like roof. In the great Teocallis there were two altars, in the
others only one, and each one of these altars[12] had upper stories; the
great ones had three stories above the altars, all terraced and of
considerable height, and the building (cepa) itself was very lofty, so
that it could be seen from afar off.
One could walk round each of these chapels and each had its
separate walls. In front of these altars a large space was left where
they made their sacrifices, and the building (cepa) itself had the
height of a great tower, without [counting] the stories that covered
the altars.
According to what some people who saw it have told me, the
Teocalli of Mexico had more than a hundred steps; I have seen them
myself and have counted them more than once, but I do not
remember [the number]. The Teocalli of Texcoco had five or six steps
more than that of Mexico. If one were to ascend to the top of the
chapel of San Francisco in Mexico, which has an arched roof and is of
considerable height, and look over Mexico, the temple of the devil
would have a great advantage in height, and it was a wonderful sight
to view from it the whole of Mexico and the towns in the
neighbourhood.
In similar courts in the principal towns there were twelve or fifteen
other Teocallis of considerable size, some larger than others, but far
from as large as the principal Teocalli.
Some of them had their fronts and steps towards others[13], others
to the East, again others to the South, but none of them had more
than one altar with its chapel, and each one had its halls and
apartments where the Tlamacazques or Ministers dwelt, who were
numerous, and those who were employed to bring water and
firewood, for in front of each altar there were braziers which burnt all
night long, and in the halls also there were fires. All these Teocallis
were very white, burnished and clean, and in some of them [the
temple enclosures] were small gardens with trees and flowers.
There was in almost all these large courts another temple, which,
after its square foundation had been raised and the altar built, was
enclosed with a high circular wall and covered with its dome. This
was [the temple] of the God of the Air, who was said to have his
principal seat in Cholula, and in all this province there were many of
them.
This God of the Air they called in their language Quetzalcoatl, and
they said that he was the son of that God of the great statue and a
native of Tollan [Tula], and thence he had gone out to instruct
certain provinces whence he disappeared, and they still hoped that
he would return. When the ships of the Marqués del Valle, Don
Hernando Cortés (who conquered this New Spain), appeared, when
they saw them approaching from afar off under sail, they said that at
last their God was coming, and on account of the tall white sails they
said that he was bringing Teocallis across the sea. However, when
they [the Spaniards] afterwards disembarked, they said it was not
their God, but that they were many Gods.
The Devil was not contented with the Teocallis already described,
but in every town and in each suburb, at a quarter of a league apart,
they had other small courts where there were three or four small
Teocallis, in some of them more, and in others only one, and on
every rock or hillock one or two, and along the roads and among the
maize fields there were many other small ones, and all of them were
covered with plaster and white, so that they showed up and bulked
large, and in the thickly peopled country it appeared as though it was
all full of houses, especially of the Courts of the Devil, which were
wonderful to behold, and there was much to be seen when one
entered into them, and the most important, above all others, were
those of Texcoco and Mexico.
Sahagun, Fr. Bernadino de (Bustmamante Edition), p. 194.
Report of the Mexicans about their God Vitzilopuchtli.
[Huitzilopochtli, Huichilobos.]
 The Mexicans celebrate three festivals to Vitzilopuchtli every year,
the first of them in the month named Panquetzaliztli. During this
festival [dedicated] to him and others, named Tlacavepancuexcotzin,
they ascend to the top of the Cue, and they make life-size images out
of tzoalli: when these are completed, all the youths of Telpuchcalli
carry them on their hands to the top of the Cue. They make a statue
of Vitzilopuchtli in the district [barrio] named Itepeioc[A]. The
statue of Tlacavepancuexcotzin was made in that of Vitznaoao[14].
They first prepare the dough and afterwards pass all the night in
making the statues of it.
After making the images of the dough, they worshipped them as
soon as it was dawn and made offerings to them during the greater
part of the day, and towards evening they began ceremonies and
dances with which they carried them to the Cue, and at sunset they
ascended to the top of it.
After the images were placed in position, they all came down again
at once, except the guardians [named] Yiopuch.
As soon as dawn came the God named Paynal, who was the Vicar
of Vitzilopuchtli, came down from the lofty Cu, and one of the
priests, clad in the rich vestments of Quetzalcoatl, carried this God
(Vitzilopuchtli) in his hands, as in a procession, and the image of
Paynal (which was carved in wood and, as has already been stated,
was richly adorned) was also brought down.
In this latter festival there went in front of [the image] a mace-
bearer, who carried on his shoulder a sceptre in the shape of a huge
snake, covered all over with a mosaic of turquoise.
When the Chieftain arrived with the image at a place named
Teutlachco, which is the game of Ball [that is at the Tlachtli court],
which is inside the Temple courtyard, they killed two slaves in front
of him, who were the images [representatives] of the Gods named
Amapantzitzin, and many other captives. There the procession
started and went direct to Tlaltelulco.
Many Chieftains and people came out to receive it, and they
burned incense to them [the images] and decapitated many quails
before them.
Thence they went directly to a place named Popotla, which is near
to Tacuba, where now the church of S. Esteban stands, and they gave
it another reception like that mentioned above. They carried in front
of the procession all the way a banner made of paper like a fly-whisk,
all full of holes, and in the holes bunches of feathers, in the same way
as a cross is carried in front of a procession. Thence they came direct
to the Cu of Vitzilopuchtli, and with the banner they performed
another ceremony as above stated in this festival.

Account of the Buildings of the Great Temple of Mexico.

The court of this Temple was very large, almost two hundred
fathoms square; it was all paved, and had within it many buildings
and towers. Some of these were more lofty than others, and each one
of them was dedicated to a God.
The principal tower of all was in the middle and was higher than
the others, and was dedicated to the God Vitzilopuchtli
Tlacavepancuexcotzin.
This tower was divided in the upper part, so that it looked like two,
and had two chapels or altars on the top, each one covered by its
dome (chapitel) and each one of them had on the summit its
particular badges or devices. In the principal one of them was the
statue of Vitzilopuchtli, also called Ilhuicatlxoxouhqui, and in the
other the image of the God Tlaloc. Before each one of these was a
round stone like a chopping-block, which they call Texcatl, where
they killed those whom they sacrificed in honour of that God, and
from the stone towards the ground below was a pool of blood from
those killed on it; and so it was on all the other towers; these faced
the West, and one ascended by very narrow straight steps to all these
towers.
(Sahagun mentions seventy-eight edifices in connection with the
Great Temple, but it is almost certain that these were not all within
the Temple enclosure.)
Sahagun, Hist. de la Conquista, Book 12, Ch. XXII.
They [the Mexicans] ascended a Cu, the one that was nearest to the
royal houses [i. e. of Axayacatl], and they carried up there two stout
beams so as to hurl them from that place on to the royal houses and
beat them down so as to force an entry. When the Spaniards
observed this they promptly ascended the Cu in regular formation,
carrying their muskets and crossbows, and they began the ascent
very slowly, and shot with their crossbows and muskets at those
above them, a musketeer accompanying each file and then a soldier
with sword and shield, and then a halberdier: in this order they
continued to ascend the Cu, and those above hurled the timbers
down the steps, but they did no damage to the Spaniards, who
reached the summit of the Cu and began to wound and kill those who
were stationed on the top, and many of them flung themselves down
from the Cu: finally, all those [Mexicans] who had ascended the Cu
perished.
Hernando Cortés, 2nd Letter. (The attack on the Great Teocalli.)
We fought from morning until noon, when we returned with the
utmost sadness to our fortress. On account of this they [the enemy]
gained such courage that they came almost up to the doors, and they
took possession of the great Mosque[15], and about five hundred
Indians, who appeared to me to be persons of distinction, ascended
the principal and most lofty tower, and took up there a great store of
bread and water and other things to eat, and nearly all of them had
very long lances with flint heads, broader than ours and no less
sharp.
From that position they did much damage to the people in the fort,
for it was very close to it. The Spaniards attacked this said tower two
or three times and endeavoured to ascend it, but it was very lofty,
and the ascent was steep, for it had more than one hundred steps,
and as those on the top were well supplied with stones and other
arms, and were protected because we were unable to occupy the
other terraces, every time the Spaniards began the ascent they were
rolled back again and many were wounded. When those of the enemy
who held other positions saw this, they were so greatly encouraged
they came after us up to the fort without any fear.
Then I (seeing that if they could continue to hold the tower, in
addition to the great damage they could do us from it, that it would
encourage them to attack us) set out from the fort, although maimed
in my left hand by a wound that was given me on the first day, and
lashing the shield to my arm, I went to the tower accompanied by
some Spaniards and had the base of it surrounded, for this was easily
done, although those surrounding it had no easy time, for on all sides
they were fighting with the enemy who came in great numbers to the
assistance of their comrades. I then began to ascend the stairway of
the said tower with some Spaniards supporting me, and although the
enemy resisted our ascent very stubbornly, so much so that they
flung down three or four Spaniards, with the aid of God and his
Glorious Mother (for whose habitation that tower had been chosen
and her image placed in it), we ascended the said tower and reaching
the summit we fought them so resolutely that they were forced to
jump down to some terraces about a pace in width which ran round
the tower. Of these the said tower had three or four, thrice a man’s
height from one [terrace] to the other.
Some fell down the whole distance [to the ground], and in addition
to the hurt they received from the fall, the Spaniards below who
surrounded the tower put them to death. Those who remained on the
terraces fought thence very stoutly, and it took us more than three
hours to kill them all, so that all died and none escaped ... and I set
fire to the tower and to the others which there were in the Mosque.
Juan de Torquemada, Monarchia Indiana, Vol. II. Book 8, Ch. XI.
p. 144. [Giving a description of the Great Temple.]
This Temple was rebuilt and added to a second time; and was so
large and of such great extent, that it was more than a crossbow-shot
square.
It was all enclosed in masonry of well squared stone.
There were in the square four gateways which opened to the four
principal streets, three of them by which the city was approached
along the causeways from the land, [the fourth] on the east in the
direction of the lake whence the City was entered by water.
In the middle of this enormous square was the Temple which was
like a quadrangular tower (as we have already stated) built of
masonry, large and massive.
This Temple (not counting the square within which it was built)
measured three hundred and sixty feet from corner to corner, and
was pyramidal in form and make, for the higher one ascended the
narrower became the edifice, the contractions being made at
intervals so as to embellish it.
 On the top, where there was a pavement and small plaza rather
more than seventy feet wide, two very large altars had been built, one
apart from the other, set almost at the edge or border of the tower on
the east side, so that there was only just sufficient ground and space
for a man to walk [on the east side] without danger of falling down
from the building.
These altars were five palms in height with their walls inlaid with
stone, all painted with figures according to the whim and taste of him
who ordered the painting to be done. Above the altars were the
chapels roofed with very well dressed and carved wood.
Each of these chapels had three stories one above the other, and
each story or stage was of great height, so that each one of them [of
the chapels] if set on the ground (not on that tower, but on the
ground level whence the edifice sprang) would have made a very lofty
and sumptuous building, and for this reason the whole fabric of the
Temple was so lofty that its height compelled admiration. To behold,
from the summit of this temple, the city and its surroundings, with
the lakes and all the towns and cities that were built in it and on its
banks, was a matter of great pleasure and contentment.
On the West side this building had no stages [contractions], but
steps by which one ascended to the level of the chapels, and the said
steps had a rise of one foot or more. The steps, or stairs, of this
famous temple numbered one hundred and thirteen, and all were of
very well dressed stone.
From the last step at the summit of this Temple to the Altars and
entrance to the Chapels was a considerable space of ground, so that
the priests and ministers of the Idols could carry out their functions
unimpeded and thoroughly.
On each of the two altars stood an Idol of great bulk, each one
representing the greatest God they possessed, which was
Huitzilupuchtli or by his other name Mexitli.
Near and around this Great Temple there were more than forty
lesser ones, each one of them dedicated and erected to a God, and its
tower and shape narrowed up to the floor on which the Chapel and
altar began to arise, and it was not as large as the Great Temple, nor
did it approach it by far in size, and all these lesser Temples and
towers were associated with the Great Temple and tower which there
was in this City.
The difference between the Great Temple and the lesser ones was
not in the form and structure, for all were the same, but they differed
in site and position [orientation], for the Great Temple had its back
to the East, which is the practice the large temples ought to follow, as
we have noticed that the ancients assert, and their steps and
entrance to the West (as we are accustomed to place many of our
Christian Churches), so that they paid reverence in the direction of
the sun as it rose, the smaller temples looked in the other direction
towards the East and to other parts of the heaven [that is to] the
North and South.

In order that my readers may not think that I speak heedlessly,

and without a limit to my figures, I wish to quote here the words of
Padre Fray Bernadino de Sahagun, a friar of my Order and one of
those who joined very early in the discovery of this New Spain in the
year twenty-nine [1529], who saw this and the other temples.... He
says these words:—“This Temple was enclosed on all sides by stone
walls half as high again as a man, all embattled and whitened. The
ground of this Temple was all paved, with very smooth flag stones
(not dressed but natural) as smooth and slippery as ice. There was
much to be seen in the buildings of this Temple; I made a picture of
it in this City of Mexico, and they took it to Spain for me, as a thing
well worth beholding, and I could not regain possession of it, nor
paint it again, and although in the painting it looks so fine, it was in
reality much more so, and the building was more beautiful. The
principal shrine or chapel which it possessed was dedicated to the
God Huitzilupuchtli, and to another God his companion named
Tlacahuepancuezcotzin, and to another, of less importance than the
two, called Paynalton....”
And he adds more, saying “the square was of such great
circumference that it included and contained within its area all the
ground where the Cathedral, the houses of the Marques del Valle[A],
the Royal houses[16] and the houses of the Archbishop have now been
built, and a great part of what is now the market place,” which seems
incredible, so great is the said area and space of ground.
 I remember to have seen, thirty-five years ago, a part of these
buildings in the Plaza, on the side of the Cathedral, which looked to
me like hills of stone and earth, which were being used up in the
foundations of God’s house and Cathedral which is being built now
with great splendour.
Padre Fray Diego Duran, Historia de los Indias de Nueva
Espana, Vol. II. Ch. LXXX. p. 82.
Having heard what has been said about the decoration of the Idol,
let us hear what there is notable about the beauty of the Temples. I
do not wish to begin by relating the accounts given me by the
Indians, but that obtained by a monk who was among the first of the
Conquerors who entered the country, named Fray Francisco de
Aguilar, a very venerable person and one of great authority in the
order of our Glorious Father Santo Domingo, and from other
conquerors of strict veracity and authority who assured me that on
the day when they entered the City of Mexico and beheld the height
and beauty of the Temples they believed them to be turreted
fortresses for the defence and ornament of the City, or that they were
palaces and royal houses with many towers and galleries, such was
their beauty and height which could be seen from afar off.
It should be known that of the eight or nine temples which there
were in the City all stood close to one another within a great
enclosure, inside of which enclosure each one adjoined the other, but
each had its own steps and separate patio[17], as well as living rooms
and sleeping places for the Ministers of the temples, all of which
occupied considerable amount of space and ground. It was indeed a
most beautiful sight, for some were more lofty than the others, and
some more ornamental than others, some with an entrance to the
East others to the West, others to the North and others to the South,
all plastered and sculptured, and turreted with various kinds of
battlements, painted with animals and figures and fortified with
huge and wide buttresses of stone, and it beautified the city so
greatly and gave it such an appearance of splendour that one could
do nothing but stare at it.
However, as regards the Temple, especially [dedicated to] the Idol
[Huitzilopochtli] with which we are dealing, as it was that of the
principal God, it was the most sumptuous magnificent of them all.
 It had a very large wall round its special court, all built of great
stones carved to look like snakes, one holding on to the other, and
anyone who wishes to see these stones, let him go to the principal
Church of Mexico and there he will see them used as pedestals and
bases of the pillars. These stones which are now used there as
pedestals formed the wall of the Temple of Huitzilopochtli, and they
called this wall Coatepantli, which means wall of snakes. There was
on the top of the halls or oratories where the Idol stood a very
elegant breastwork covered with small black stones like jet, arranged
with much order and regularity, all the groundwork being of white
and red plaster, which shone wonderfully [when looked at] from
below—on the top of this breastwork were some very ornamental
merlons carved in the shape of shells.
At the end of the abutments, which arose like steps a fathom high,
there were two seated Indians, in stone, with two torch-holders in
their hands, from which torch-holders emerged things like the arms
of a cross, ending in rich green and yellow feathers and long borders
of the same.
Inside this [the] court there were many chambers and lodgings for
the monks and nuns, as well as others on the summit for the priests
and ministers who performed the service of the Idol.
This Court was so large that on the occasion of a festival eight or
ten thousand men assembled in it; and to show that this is not
impossible, I wish to relate an event that is true, related by one who
with his own hands killed many Indians within it....
This Court had four doors or entrances, one towards the East,
another towards the West, another towards the South, and one on
the North side. From these commenced four Causeways, one towards
Tlacopan, which we now call the street of Tacuba, another towards
Guadelupe, another towards Coyoacan, and the other led to the lake
and the landing place of the canoes.
The four principal Temples also have their portals towards the said
four directions, and the four Gods which stand in them also have
their fronts turned in the same directions.

Opposite the principal gateway of this Temple of Huitzilopochtli

there were thirty long steps thirty fathoms long; a street separated
them from the wall of the patio[18]. On the top of them [the steps] was
a terrace, 30 feet wide and as long as the steps, which was all coated
with plaster, and the steps very well made.
Lengthwise along the middle of this broad and long platform was a
very well made palisade as high as a tall tree, all planted in a straight
line, so that the poles were a fathom apart. These thick poles were all
pierced with small holes, and these holes were so close together that
there was not half a yard between them, and these holes were
continued to the top of the thick and high poles. From pole to pole
through the holes came some slender cross-bars on which many
human skulls were strung through the forehead. Each cross-bar held
thirty heads, and these rows of skulls reached to the top of the
timbers and were full from end to end ... all were skulls of the
persons who had been sacrificed.

After describing a procession in which the God was carried to

Chapultapec and thence to Coyoacan, the author continues:—
When they arrived at the foot of the steps of the Temple they
placed the litter [on which the image of the God was carried] there,
and promptly taking some thick ropes they tied them to the handles
of the litter, and, with great circumspection and reverence, some
making efforts from above and others helping from below, they
raised the litter with the Idol to the top of the Temple, with much
sounding of trumpets and flutes, and clamour of conch-shells and
drums; they raised it up in this manner because the steps of the
Temple were very steep and narrow [in the tread] and the stairway
was long and they could not ascend with it on their shoulders
without falling, and so they took that means to raise it up.
Hernando Alvarado Tezozomoc, Cronica Mexicana, Ch. XXX, p.
319, writing of the Temple of Huitzilipochtli, says:—
It could be ascended on three sides and would have as many steps
as there are days in the year, for at that time the year consisted of
eighteen months, and each month contained twenty days, which
amounts to three hundred and sixty days, five days less than our
Catholic religion counts. Others count thirteen months to the year.
At all events the steps were arranged on three sides of the ascent.
 The principal ascent faced the south, the second the east, and the
third the west, and on the north side were three walls like a chamber
open to the south. It had a great court and Mexican plaza all
surrounded by a stone wall, massive and strong, [of which] the
foundations were more than a fathom and the height [of the wall]
was that of four men’s stature. It had three gateways, two of them
small, one facing the east and the other the west; the gateway in the
middle was larger, and that one faced the south, and in that direction
was the great market place and Tianguiz[19], so that it stood in front
of the great palace of Montezuma and the Great Cu. The height of it
[the Great Cu or Temple] was so great that, from below, persons [on
its summit], however tall they might be, appeared to be of the size of
children eight years old or less.
Ixtlilxochitl (‘Codice Goupil’).
The Temple and principal Cu of this City, indeed of all New Spain,
was built in the middle of the city, four square and massive as a
mound (terrapleno) of stone and clay, merely and only the surface
[built] of masonry. Each side was eighty fathoms long (445 Eng. ft.)
and the height was over twenty-seven fathoms (150 Eng. ft.). On the
side by which it was ascended were one hundred and sixty steps
which faced the west. The edifice was of such a shape that from its
foundation it diminished in size and became narrower as it rose in
the shape of a pyramid, and at certain distances as it rose it had
landing places like benches all around it. In the middle of the steps
from the ground and foundation there rose a wall up to the summit
and top of the steps, which was like a division that went between the
two ascents as far as the patio which was on the top, where there
were two great chambers, one larger than the other—the larger one
to the south, and there stood the Idol Huitzilopochtli; the other,
which was smaller, was to the north and contained the Idol Tlaloc,
which (Idol) and Huitzilopochtli and the chambers looked to the
west. These chambers were built at the eastern edge and border of
the said patio, and thus in front of them the patio extended to the
north and south with a [floor of] cement three palms and more in
thickness, highly polished, and so capacious that it would hold five
hundred men, and at one side of it towards the door of the larger
chamber of Huitzilopochtli was a stone rising a yard in height, of the
shape and design of an arched coffer, which was called Techcatl
(Texcatl) where the Indians were sacrificed. Each of these chambers
had upper stories, which were reached from within, the one from the
other by movable wooden ladders, and were full of stores of every
sort of arms, especially macanas, shields, bows, arrows, lances, slings
and pebbles, and every sort of clothing and bows for war. The face
and front of the larger chamber was ornamented with stone in the
shape and form of death’s heads whitened with lime, which were
placed all over the front, and above, for merlons, there were carved
stones in the shape of great shells, which and the other with the rest
of the Cu is painted on the following page. * * * * [see Plate D].

PLATE A.
Part of the City of Mexico from a modern Map.

TRACING A1.
After J. García Icazbalceta.

TRACING A2.
Suggested site of the Great Teocalli and enclosure.

PLATE B.
Suggested plan and section of the Great Teocalli.

PLATE C.
Plan by Padre Sahagun, after Dr. E. Seler.

PLATE D.
View of the Great Teocalli, after Ixtlilxochitl (‘Codice Goupil’).

PLATE E.
View of the Great Teocalli and enclosure, from ‘The Chronicle of
Mexico,’ 1576. (Manuscript in British Museum, No. 31219.
Additional.)
PRINTED BY TAYLOR AND FRANCIS,
RED LION COURT, FLEET STREET.

Tracing A1

after J. GARCÍA ICAZBALCETA

PRINTED BY TAYLOR AND FRANCIS,

RED LION COURT, FLEET STREET.