Explore the full ebook collection and download it now at textbookfull.

com

Extended finite element and meshfree methods Timon

Rabczuk

https://textbookfull.com/product/extended-finite-element-
and-meshfree-methods-timon-rabczuk/

OR CLICK HERE

DOWLOAD EBOOK

Browse and Get More Ebook Downloads Instantly at https://textbookfull.com

Click here to visit textbookfull.com and download textbook now
 Your digital treasures (PDF, ePub, MOBI) await
Download instantly and pick your perfect format...

Read anywhere, anytime, on any device!

Computational Methods for Fracture in Porous Media:

Isogeometric and Extended Finite Element Methods 1st
Edition Renã© De Borst
https://textbookfull.com/product/computational-methods-for-fracture-
in-porous-media-isogeometric-and-extended-finite-element-methods-1st-
edition-rena-de-borst/
textbookfull.com

Extended Finite Element Method Tsinghua University Press

Computational Mechanics Series 1st Edition Zhuo Zhuang

https://textbookfull.com/product/extended-finite-element-method-
tsinghua-university-press-computational-mechanics-series-1st-edition-
zhuo-zhuang/
textbookfull.com

Computational Finite Element Methods in Nanotechnology

Sarhan M. Musa(Editor)

https://textbookfull.com/product/computational-finite-element-methods-
in-nanotechnology-sarhan-m-musaeditor/

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
Mathematical Theory of Subdivision Finite Element and
Wavelet Methods 1st Edition Sandeep Kumar (Author)

https://textbookfull.com/product/mathematical-theory-of-subdivision-
finite-element-and-wavelet-methods-1st-edition-sandeep-kumar-author/

textbookfull.com

Finite Element Analysis A Primer 2nd Edition Musa

https://textbookfull.com/product/finite-element-analysis-a-primer-2nd-
edition-musa/

textbookfull.com

Novel Finite Element Technologies for Solids and

Structures Jörg Schröder

https://textbookfull.com/product/novel-finite-element-technologies-
for-solids-and-structures-jorg-schroder/

textbookfull.com

Finite Element Analysis A Primer Second Edition Sarhan M

Musa

https://textbookfull.com/product/finite-element-analysis-a-primer-
second-edition-sarhan-m-musa/

textbookfull.com

Finite Element Analysis for Design Engineers Second

Edition Paul M. Kurowski

https://textbookfull.com/product/finite-element-analysis-for-design-
engineers-second-edition-paul-m-kurowski/

textbookfull.com
EXTENDED FINITE
ELEMENT AND
MESHFREE METHODS
EXTENDED FINITE
ELEMENT AND
MESHFREE METHODS

TIMON RABCZUK
Bauhaus Universität Weimar
Weimar, Germany

JEONG-HOON SONG
University of Colorado at Boulder
Boulder, CO, United States of America

XIAOYING ZHUANG
Tongji University
Shanghai, China
Leibniz Universität Hannover
Hannover, Germany

COSMIN ANITESCU
Bauhaus Universität Weimar
Weimar, Germany
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1650, San Diego, CA 92101, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2020 Elsevier Inc. All rights reserved.
MATLAB® is a trademark of The MathWorks, Inc. and is used with permission.
The MathWorks does not warrant the accuracy of the text or exercises in this book.
This book’s use or discussion of MATLAB® software or related products does not constitute
endorsement or sponsorship by The MathWorks of a particular pedagogical approach or particular
use of the MATLAB® software.
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, further information about the Publisher’s permissions policies and our arrangements
with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency,
can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described herein. In
using such information or methods they should be mindful of their own safety and the safety of
others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of any methods, products,
instructions, or ideas contained in the material herein.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

ISBN: 978-0-12-814106-9

For information on all Academic Press publications

visit our website at https://www.elsevier.com/books-and-journals

Publisher: Matthew Deans

Acquisition Editor: Brian Guerin
Editorial Project Manager: Isabella C. Silva
Production Project Manager: Surya Narayanan
Jayachandran
Designer: Mark Rogers
Typeset by VTeX
Contents

Preface xiii
Nomenclature xix

1. Introduction 1
1.1. Partition of unity methods 1
1.2. Moving boundary problems 6
1.3. Fracture mechanics 8
1.4. Level set methods 10
1.4.1. Implicit interface and signed distance functions 11
1.4.2. Discretization of the level set 12
1.4.3. Capturing motion interface 12
1.4.4. Level sets for 3D fracture modeling 14
References 15

2. Weak forms and governing equations 19

2.1. Strong form for pure mechanical problems 19
2.1.1. One dimensional model problem 19
2.1.2. Model problem in higher dimensions 20
2.1.3. Total Lagrangian formulation 21
2.1.4. Updated Lagrangian formulation 22
2.2. From the strong form to the weak form 24
2.2.1. Weak form for the one-dimensional model problem 24
2.2.2. Weak form for the total Lagrangian formulation 26
2.3. Variational formulation 27

3. Extended finite element method 29

3.1. Formulation and concepts 29
3.1.1. Standard XFEM 29
3.1.2. Hansbo-Hansbo XFEM 34
3.2. Blending, integration and solvers 36
3.2.1. Blending 36
3.2.2. Isoparametric 2D quadrilateral XFEM element for linear elasticity 40
3.2.3. Shape functions 41
3.2.4. The B-operator 42
3.2.5. The element stiffness matrix 44
3.2.6. Integration 46
3.3. XFEM for static/quasi-static fracture modeling in 2D and 3D 50
3.3.1. XFEM approximation for cracks 50
3.3.2. Discrete equations 54

v
vi Contents

3.3.3. Crack branching and crack junction 57

3.3.4. Crack opening and crack closure 59
3.4. XFEM for dynamic fracture modeling in 2D and 3D 60
3.4.1. Diagonalized mass matrix 60
3.4.2. Limitations 64
3.5. Smoothed extended finite element method 65
3.5.1. Introduction to SFEM 67
3.5.2. Enrichment in SXFEM and selection of enriched nodes 70
3.5.3. Displacement-, strain field approximation and discrete equations 72
3.5.4. Numerical integration 75
3.6. XFEM for coupled problems 77
3.6.1. Hydro-mechanical problems 77
3.6.2. Thermo-mechanical problems 89
3.6.3. Piezoelectric materials 92
3.6.4. Flexoelectricity 100
3.7. XFEM for inverse analysis and topology optimization 105
3.7.1. Inverse problem 105
3.7.2. Optimization problems 115
3.7.3. Mathematical form of a structural optimization problem 116
3.7.4. Solid isotropic material with penalization (SIMP) 117
3.7.5. Level set based optimization 118
3.7.6. Nanoelasticity 118
3.7.7. Nanopiezoelectricity 130
3.8. Conditioning and solution of ill-conditioned systems 146
References 147

4. Phantom node method 153

4.1. Formulation and concepts 153
4.2. A crack tip element for the phantom node methods 154
4.2.1. Three-node triangular element 154
4.2.2. Four-node quadrilateral element 157
4.3. Multiple crack modeling 158
References 159

5. Extended meshfree methods 161

5.1. Introduction to meshfree methods 161
5.1.1. Basic approximation 161
5.1.2. Completeness and conservation 162
5.1.3. Consistency, stability and convergence 164
5.1.4. Continuity 165
5.1.5. Partition of unity 165
5.1.6. Kernel functions 167
Contents vii

5.2. Some specific methods 171

5.2.1. Approximation of the displacement field 171
5.2.2. Spatial integration 177
5.2.3. Essential boundary conditions 185
5.2.4. Comparison of different methods 186
5.3. Numerical instabilities 190
5.3.1. Instability due to rank deficiency 192
5.3.2. Tensile instability 193
5.3.3. Attempts to remove instabilities 193
5.3.4. Material instability in meshfree methods 194
5.4. Fracture modeling in meshfree methods 209
5.4.1. The visibility method 209
5.4.2. The diffraction method 212
5.4.3. The transparency method 215
5.4.4. The “see through” and “continuous line” method 217
5.5. The concept of enrichment 217
5.5.1. Intrinsic enrichment 219
5.5.2. Extrinsic enrichment 222
5.6. (Extrinsically) enriched local PU meshfree methods 225
5.6.1. Enriched methods with crack tip enrichment 226
5.6.2. Enriched methods without crack tip enrichment 230
5.6.3. Crack branching and crack junction 236
5.7. Extended local maximum entropy (XLME) 238
5.7.1. Local Maximum Entropy (LME) approximants 239
5.7.2. Numerical integration 243
5.7.3. Condition number 245
5.8. Cracking particle methods 245
5.8.1. The enriched cracking particles method 246
5.8.2. Applications to large deformations 250
5.8.3. The cracking particles method without enrichment 250
5.8.4. Cracking rules for cracking particle methods 251
5.9. Comparison of different methods 253
5.9.1. The mode I crack problem 253
5.9.2. The mixed mode problem 260
5.10. Extensions to mode II kinematics 263
5.10.1. Enriching in the shear band plane 263
5.10.2. Enforcing mode II-kinematics with the penalty method 265
5.11. Discrete system of equations for pure mechanical problems 265
5.11.1. Methods without enrichment 265
5.11.2. Enriched methods 267
5.11.3. Extension to dynamics 270
5.12. Spatial integration 283
viii Contents

5.13. Time integration 286

5.13.1. Explicit-implicit time integration 286
5.13.2. Explicit time integration, critical time step and mass lumping 287
5.13.3. Crack propagation in time 304
References 306

6. Extended isogeometric analysis 315

6.1. Formulation and concepts 315
6.1.1. B-splines and NURBS 315
6.1.2. Bézier extraction 317
6.2. Hierarchical refinement with PHT-splines 320
6.2.1. PHT-spline space 321
6.2.2. Computing the control points 323
6.3. Analysis using splines 324
6.3.1. Galerkin method 325
6.3.2. Linear elasticity 327
6.4. Numerical examples 329
6.4.1. Infinite plate with circular hole 329
6.4.2. Open spanner 330
6.4.3. Pinched cylinder 331
6.4.4. Hollow sphere 332
6.5. Adaptive analysis 333
6.5.1. Determining the superconvergent point locations 333
6.5.2. Superconvergent patch recovery 337
6.5.3. Marking algorithm 340
6.6. Multi-patch formulations for complex geometry 341
6.7. XIGA for interface problems 341
6.7.1. Governing and weak form equations 342
6.7.2. Enriched basis functions selection 345
6.7.3. Enrichment functions 347
6.7.4. Greville Abscissae 348
6.7.5. Repeating middle neighbor knots 349
6.7.6. Inverse mapping 350
6.7.7. Curve fitting 351
6.7.8. Intersection points 353
6.7.9. Triangular integration 354
References 355

7. Fracture in plates and shells 359

7.1. Fractures in shell and plates using XFEM 359
7.1.1. Weak form 359
7.1.2. Implementation based on the Q4 element 361
7.1.3. Shear locking 362
Contents ix

7.1.4. Curvature strain smoothing 363

7.1.5. Extended finite element method for shear deformable plates 365
7.1.6. Smoothed extended finite element method 367
7.1.7. Integration 368
7.2. Fractures in shell and plates using the phantom node method 370
7.2.1. Phantom node method for the Belytschko-Tsay shell element 370
7.2.2. Phantom node method based on the three-node isotropic
triangular MITC shell element 378
7.3. Extended meshfree methods for fracture in shells 392
7.3.1. Shell model 393
7.3.2. Continuum constitutive models 396
7.3.3. Crack model 397
7.4. An immersed particle method for fluid-structure interaction 402
7.5. XIGA models for plates and shells 408
7.5.1. Kinematics of the shell 408
7.5.2. Weak form 410
7.5.3. Discretization of the displacement field and enrichment 412
7.5.4. Discrete system of equations 419
7.5.5. Edge cracked plates under tension or shear 422
7.5.6. Pressurized cylinder with an axial crack 428
References 432

8. Fracture criteria and crack tracking procedures 437

8.1. Fracture criteria 437
8.2. Cracking criteria 437
8.2.1. Criteria in LEFM 437
8.2.2. Global energy criteria 440
8.2.3. Rankine criterion 440
8.2.4. Loss of material stability condition 441
8.2.5. Rank-one-stability criterion 443
8.2.6. Determining the crack orientation 444
8.2.7. Computation of the crack length 444
8.3. Crack surface representation and tracking the crack path 445
8.3.1. The level set method to trace the crack path 447
8.3.2. Tracking the crack path in 3D 451
8.3.3. Adaptive crack propagation technique 462
8.3.4. Comments 464
References 466

9. Multiscale methods for fracture 471

9.1. Extended Bridging Domain Method 472
9.1.1. Concurrent coupling of two models at different length scales 474
9.1.2. Consistency of material properties 479
x Contents

9.2. Extended bridging scale method 479

9.2.1. Consistency of material properties 481
9.2.2. Upscaling and downscaling 483
9.3. Multiscale aggregating discontinuity (MAD) method 491
9.3.1. Overview of the method 491
9.3.2. Coarse graining method 494
9.3.3. Micro-macro linkage 500
9.4. Crack opening in unit cells with the hourglass mode 503
9.5. Stability of the macromaterial 504
9.6. Implementation 507
9.7. Numerical examples 508
9.7.1. 3D modeling of cracks in a nanocomposite 508
9.7.2. Hierarchical multiscale example 508
9.7.3. Semi-concurrent FE-FE coupling example 510
9.7.4. Concurrent FE-XFEM coupling example 512
9.7.5. MD-XFEM coupling example 513
References 516

10. A short overview of alternatives for fracture 521

10.1. Numerical manifold method (finite cover method) 521
10.1.1. The cover approximation 522
10.1.2. The least square-based physical cover functions 523
10.1.3. The imposition of boundary conditions 524
10.1.4. Fracture modeling 524
10.1.5. Geometric and material nonlinear analysis 527
10.2. Peridynamics and dual-horizon peridynamics 528
10.2.1. Dual-horizon peridynamics 531
10.2.2. The dual property of dual-horizon 539
10.2.3. Wave propagation in 1D homogeneous bar 543
10.2.4. Numerical examples 544
10.3. Phase field models 562
10.3.1. Concepts 563
10.3.2. Governing equations 568
10.3.3. Discretization 569
10.3.4. Solution schemes 572
10.3.5. Implementations 573
References 575

11. Implementation details 581

11.1. Computer implementation of enriched methods 581
11.1.1. Pre-processing 582
11.1.2. Processing 585
11.1.3. Post-processing 589
Contents xi

11.2. Numerical examples 590

11.2.1. Crack propagation angle 591
11.2.2. Hydro-mechanical model with center cracks 591
11.2.3. Hydro-mechanical model with edge crack 592
References 597

Part 1. Appendices

A. Derivation of shape derivative for the nanoelasticity problem 601

B. Derivation of the adjoint problem for the nanopiezoelectricity

problem 603

Index 607
Preface

The objective of this book is to provide an overview and the theoreti-

cal/computational background of partition-of-unity based computational
methods, their implementation and applications. The focus is on extended
finite element and meshfree methods and their application with focus on
modeling material failure. It is assumed that the readers are already familiar
with finite element methods or similar computational approaches including
their implementation. The content of this book is written from an engi-
neering point of view. It explains concepts and formulations and provides
details on the implementation through simple Matlab® codes. We pro-
vide classical benchmark problems for which state-of-the-art computational
methods are tested at and present some interesting numerical examples to
demonstrate the power and performance of the outlined methods. The
book however does not contain mathematical proofs concerning for in-
stance the convergence of the above methods. Convergence plots are just
shown numerically for specific examples. Though some of the methods are
implemented in commercial software such as ABAQUS, the book does not
provide a description on the use of these functions within such commercial
codes.
The book is aimed for students and researchers who are interested in
learning and implementing partition-of-unity method, especially extended
finite element and meshfree methods. It is well suited for students and post-
doctoral fellows to start research in this direction and who are interested in
method development or the application of described methods to challeng-
ing problems in engineering and materials science. It is also of interest to
readers who are interested in state-of-the-art computational methods for
linear and nonlinear fracture and choosing an adequate method for their
problem of interest. The content of this book is too extensive to be cov-
ered in a single course though parts of it could be the basis for a 1-semester
course on meshfree methods or extended finite elements.
Chapter 1 provides an introduction to computational challenges which
occur in problems with moving boundaries such as fracture, fluid mechan-
ics, fluid-structure interaction, inverse problems or optimization. It also
presents the level set method which is commonly combined with partition-
of-unity methods for those problems. Chapter 2 summarizes the governing
equations for purely mechanical problems for applications in statics and

xiii
xiv Preface

dynamics as most of the methods are presented in such a setting. They

are provided in strong and weak form including a Total Lagrangian and
updated Lagrangian description of motion. Chapter 3 is focused on the ex-
tended finite element method (XFEM) and variations or improvements of
it. Within this chapter, challenges related to so-called blending, numerical
integration, enrichment and solution procedures are discussed and potential
solutions are derived. The implementation of XFEM for static and dynamic
fracture problems is described in detail and the representation of complex
features such as crack nucleation, crack branching and crack coalescence is
discussed. We also propose a variation of the classical XFEM for fracture, i.e.
the smoothed extended finite element method (SXFEM). SXFEM avoids
the integration of the singularity in case of asymptotic crack tip enrich-
ments. It facilitates the subtriangulation commonly employed in cracked
elements, and inherits certain superior properties of the smoothed finite
element method including less sensitivity to mesh distortion and its high
accuracy for triangular elements. We subsequently present the extended fi-
nite element formulations for coupled fracture problems including different
enrichment strategies, implementation details and potential challenges re-
lated for instance to the ill-conditioning of the final system of equations to
be solved. Thermo-mechanical, hydro-mechanical and electro-mechanical
are described; the latter ones include piezo- as well as flexoelectric mate-
rials. Finally, two other important applications of XFEM are given: Inverse
analysis and (topology) optimization. For those problems, XFEM allows an
exact – implicit – representation of the topology through level set functions
and hence employing always the same mesh during the iterations while
maintaining optimal convergence rates. The performance of the method
is demonstrated for several challenging problems in the associated section
for selected problem. The phantom node method, another “variation” of
XFEM, in Chapter 4 is not based on enrichment functions but overlap-
ping elements. It has the advantage of being easily implementable into an
existing finite element code but can be applied only to fracture problems.
Concepts for how to incorporate multiple cracks and specific crack tip el-
ements are devised in this chapter.
The topic of Chapter 5 is extended meshfree methods. Firstly, the basic
concept of meshfree methods is explained. Subsequently, we present several
popular meshfree approximations including the Smoothed Particle Hydro-
dynamics (SPH) and improvements of it such as the Reproducing Kernel
Particle Method (RKPM). The mostly rational shape functions and lack
of the so-called Kronecker-delta property impose additional challenges on
Preface xv

meshfree methods compared to FEM. In this context, we provide different

approaches of spatial numerical integration, imposition of essential bound-
ary conditions as well as solutions to avoid instabilities of different sources
which occur in many meshfree methods such as SPH. We subsequently
present classical methods on how to incorporate strong discontinuities
and finally model discrete fracture within meshfree methods including the
visibility, diffraction and transparency method. Different extended mesh-
free methods are presented which are based either on an intrinsic or an
extrinsic enrichment. In an intrinsic enrichment, the enrichment func-
tions are included in the polynomial basis used to construct the meshfree
shape functions while extrinsic enrichment schemes are mainly focused on
partition-of-unity enrichments. Two classes of discrete fracture methods
are presented. The first class ensures a continuous crack path and requires
special approaches to represent the crack’s topology and crack tracking
algorithms while the second class, the so-called cracking particle meth-
ods, represent the crack as set of crack segments and avoid crack tracking
algorithms and methods to represent the crack surface entirely. The per-
formance of various extended meshfree methods are compared for several
classical benchmark problems mostly in linear elastic fracture mechanics.
Chapter 6 presents formulations based on extended Isogeometric Anal-
ysis (XIGA). We first describe popular IGA basis functions including
B-splines, Non Uniform Rational B-splines (NURBS) and PHT-splines,
which are useful for h-adaptive refinement procedures within IGA. Imple-
mentation details of IGA for problems in linear elastostatics are provided
before different XIGA approaches for weak discontinuities are discussed
in detail. We dedicated an entire chapter to extended finite element and
meshfree methods for modeling fracture in plates and shells. Formulations
based on Mindlin-Reissner as well as Kirchhoff Love shells are devised. In
the latter case, the higher-order continuity of the associated meshfree or
IGA method has been exploited, so that no additional rotational degrees
of freedom are needed, which drastically facilitates the enrichment strategy
requiring fulfillment of a constraint condition. We also present a method
that can efficiently deal with fluid-driven fracture due to fluid-structure
interaction.
Each of the above mentioned methods are capable of dealing with
discrete fracture efficiently. However, they all require a fracture criterion
which determine the orientation and “length” of the crack. Therefore,
Chapter 8 is related to state-of-the-art fracture criteria and crack track-
ing algorithms for methods requiring a continuous crack surface. Fracture
xvi Preface

criteria for problems in linear elastic fracture mechanics (LEFM) as well

as nonlinear continua are described. While the fracture criterion provides
the orientation of the crack surface, criteria for propagating the crack are
needed as well. Different approaches for how to represent the crack surface
based on triangular facets and level sets are explained and related to differ-
ent computational methods. In this context, we describe in detail efficient
crack tracking algorithms in three dimensions and discuss challenges and
limitations.
Chapter 9 presents different multiscale methods for fracture which are
useful for applications such as computational materials design. The focus
will be on so-called concurrent multiscale methods for fracture where the
geometry of a fine-scale model is directly integrated into the geometry of
the coarse-scale model. All those methods are based on extended finite ele-
ment methods to represent fracture either on one or two length scales. Two
approaches are described: In the first approach, the fine-scale domain and
the coarse-scale domain is coupled at a discrete interface. This approach
seems promising for static applications while the second approach is better
for dynamic fracture as artificial wave reflections are minimized through
a handshake coupling which contains both the fine-scale and coarse-scale
domain. Efficient strategies to coarse grain cracks are presented which are
required in adaptive multiscale methods to guarantee computational effi-
ciency. The multiscale methods for fracture are described for coupling two
continuum models as well as coupling atomistic and continuum models.
Chapter 10 briefly gives a short introduction to competitive and popu-
lar alternative methods for fracture, i.e. the numerical manifold method
(NMM), peridynamics (PD) as well as phase field models for fracture.
The NMM shares some features of the phantom node method though
it has been proposed much earlier, even before the extended finite element
method. PD is a very efficient method for dynamic fracture as – similarly
to the cracking particles method – it does not require crack tracking proce-
dures. The crack path in PD is a natural outcome of the simulation. Phase
field models are somehow related to gradient damage models and smear
the crack over a certain width. The beauty of the phase field model lies in
the thermodynamic consistent framework, which allows a straightforward
implementation into a finite element framework. Though they belong to
the class of continuous approaches for fracture, we included them in our
book due to their growing popularity.
The last chapter of the book is dedicated to the implementation details
of the presented approaches in this book, i.e. XFEM, extended meshfree
Preface xvii

methods and XIGA. The focus is on the development of CAD-compatible

formulations in the framework of (X)IGA is briefly explained. Some nu-
merical examples are included and a link to an open-source repository
containing corresponding Matlab code with additional documentation and
explanations is provided.
We would like to thank our collaborators and current and former stu-
dents, whose research contributed to this book, among whom are: Fatemeh
Amiri, Stéphane Bordas, P.R. Budarapu, C.L. Chan, Thanh Chau-Dinh,
Lei Chen, G.R. Liu, Mohammed Msekh, S.S. Nanthakumar, Vinh Phu
Nguyen, Nhon Nguyen-Thanh, Hung Nguyen-Xuan, Harold S. Park,
Huilong Ren, Mohammad Silani, Hossein Talebi, Navid Valizadeh, Nam
Vu-Bac and Goangseup Zi. We would also like to express our gratitude to
the Elsevier editors and technical staff, in particular Brian Guerin, Sabrina
Webber, Thomas van der Ploeg, and Isabella C. Silva for their support in
the realization of this project.

Weimar, Germany T.R. and C.A.

Boulder, Colorado J.H.S.
Hannover, Germany X.Z.
September, 2019
Nomenclature

The following symbols are standard in engineering literature and are also used in this book.
However, the notation might vary somewhat among the different chapters and sections.

Greek symbols
 increment
δ variation or Kronecker delta or Dirac delta function
,  strain
 boundary
λ Lagrange multiplier or eigenvalue or Lamé’s first parameter
 domain
φ level set
potential
ψ enrichment function
ρ density
σ, σ Cauchy stress

Latin symbols
B B-operator
b body force
C (tangent) material matrix
d displacement vector
F deformation gradient
f force
I identity tensor
K stiffness matrix
M mass matrix
n normal
A area
a, a accelerations
E Young’s modulus
G shear modulus
H Heaviside function
h mesh size parameter
J Jacobian
KI , KII stress intensity factor
N shape function
p polynomial degree of basis
P, P first Piola Kirchhoff stress
R support size
t, t traction
u, U displacement
V volume

xix
xx Nomenclature

v, v velocities
w quadrature weight
X, X material coordinates
x, x spatial coordinates
 CHAPTER ONE

Introduction
1.1. Partition of unity methods
The Finite Element Method (FEM) was developed in the 1950s and
1960s as a convenient way to solve partial differential equations arising
from various scientific and engineering applications. Rigorous mathemat-
ical analysis of the method started in 1970s and since then thousands of
papers have been published on this topic. The finite element method is a
Galerkin method that approximates the solution of the partial differential
equation (PDE), posed in a variational form. The method involves parti-
tioning the domain into a finite number of “elements”, and an approximate
solution is sought in a finite dimensional space of piecewise polynomials,
defined relative to the elements. Naturally, the quality of the approxima-
tion depends on the fineness (or coarseness) of the discretization, i.e. on the
size of the elements, the degree of the underlying polynomials in the finite
element space and on the regularity of the exact solution.
During the 1970s and 1980s, as computational resources became less
expensive, the popularity of the finite element method grew rapidly and a
large number of both mathematical results and computer programs were de-
veloped [25]. Nevertheless, even with increasing computational capabilities,
certain problems (e.g. crack propagation problems, multi-scale problems,
problems with complex boundaries, etc.) were, and still remain too expen-
sive (due to size or complexity) to be solved satisfactorily by the classical
finite elements. Therefore interest grew in taking a more general approach,
with essentially two features. These features involve: (a) either not using
a mesh at all or using a very simple mesh to discretize the domain, and
(b) suitably choosing approximation spaces that are not based on polyno-
mials. The associated methods are broadly denoted as meshless or meshfree
methods.
These methods can also be grouped into two classes – the classical parti-
cle methods [52,53,55,56] and the methods based on the idea of data fitting
techniques [11,18]. The classical particle methods were developed for time
dependent problems or conservation laws. They involve a discrete set of
points called particles to discretize the underlying domain, and the solutions
of a system of time-dependent PDEs for all the particles are sought. On the
other hand, the meshless methods based on data-fitting techniques, which

Extended Finite Element and Meshfree Methods Copyright © 2020 Elsevier Inc. 1
https://doi.org/10.1016/B978-0-12-814106-9.00007-3 All rights reserved.
2 Extended Finite Element and Meshfree Methods

were initially developed for stationary problems, also discretize the domain
by particles. Here, each particle is associated with a “patch” (an open set),
such that the union of the patches covers the underlying domain. Suit-
able finite dimensional spaces (which may not be based on polynomials and
could be obtained by data-fitting techniques) are defined on each of these
patches. Finally, the associated “shape functions” are used in a Galerkin or
collocation method to obtain a matrix equation. The solution of this linear
system is then used to obtain the solution of the meshless method.
Several meshless methods, based on the idea of data fitting techniques
described above, have been developed primarily by engineers and they dif-
fer essentially in the choice of the finite dimensional spaces on the patches.
The first of these methods is known as Shepard’s method [61]. This idea was
further generalized into a method called Smoothed Particle Hydrodynam-
ics (SPH), [27,28,43,49,50,71] and the Method of Clouds [16,17]. Method
of Clouds used moving least-squares (MLS) ideas based on polynomials to
construct the finite dimensional space. Many methods, with slight varia-
tions of these ideas, were also developed; for example the Diffuse Element
Method [54] and Method of Spheres [15]. Later, a class of methods were
introduced where the finite dimensional spaces were constructed using the
ideas of reproducing kernels [38,40–42]. Also, Radial Basis Functions were
used to construct meshless methods, and the approach was developed by
both the mathematicians and engineers [23,24,35,36,73,74]. We mention
that though the methods described above (other than those based on RBFs)
were not mathematically investigated in detail, the generic ideas and the as-
sociated mathematical analysis were given in 70’s in [3,4,64]. For a modern
mathematical treatment, we refer to [8].
The Partition of Unity Finite Element Method (PUFEM) could also be
viewed as a meshless method in a broad sense. On the other hand, it is not
a specific method, but is a general flexible framework. In fact, the classical
FEM and many meshless methods, discussed before, could be cast into the
PUFEM framework in certain situations.
The idea of PUFEM was partially introduced in [6], where piecewise
L-splines [72] instead of piecewise polynomials were used to approximate
locally the solution of a boundary value problem with “rough” coefficients.
However, this method could not be directly generalized to higher dimen-
sions. In 1994, Babuška, Caloz and Osborn [7] introduced the so-called
“Special Finite Element Method” where they used certain non-polynomial
functions for local approximation on “finite element stars” (which served
as patches). These local approximations were “pasted together” using the
Introduction 3

finite element hat functions (which form a partition of unity) to obtain

a global approximation. This idea was used to approximate the solution
of a PDE, modeling a particular unidimensional composite. It was further
refined in [5,44] and a general abstract framework was developed. It was
shown that the local approximations (in the finite dimensional spaces de-
fined locally on a patch) could be pasted together using any partition of
unity. A basic approximation result was obtained in [44], which indicated
that “accurate” local approximations, pasted together using a partition of
unity, yields an accurate global approximation of the unknown solution of
the PDE. These ideas were used to approximate solutions of Laplace and
Helmholtz equations. The basic approximation result for PUFEM was fur-
ther elaborated in [9]. The framework presented in [5,44] was successfully
implemented [65–67] by Strouboulis et al. on engineering applications in-
volving composites made of fibers and domains involving multiple voids
and cracks. The name of Generalized Finite Element Method (GFEM) was
first used in these papers. Several issues related to the implementation of
GFEM, e.g. hierarchical construction of the patches (open cover), numer-
ical integration, and solving the linear system, were addressed in a series of
papers [30–33] by Griebel and Schweitzer. We note however, that the local
approximation in these papers were based on polynomials. Ideas similar to
GFEM and PUFEM have also been developed, to some extent indepen-
dently and in parallel, under the name of Extended Finite Element Method
(XFEM), [62], [68]. This method is also an extension of the standard
FEM, where certain “enrichment” functions (based on the available in-
formation on the solution) are used for approximation, locally in specially
chosen finite element stars. These enrichment functions are then pasted
together using standard finite element hat functions. XFEM has been suc-
cessfully implemented for various crack propagation problems [1,22,46–48,
62,69,70]. We mention that XFEM could also be cast in the framework of
PUFEM, where standard finite elements are used as the partition of unity.
The PUFEM framework can be described by the following discretiza-
tion steps:
• We consider an overlapping finite open cover {ωi } of the underlying do-
main . The sets ωi , called patches, could be the interiors of the finite
element stars, with respect to a simple finite element mesh that “trian-
gulates” a region containing . We note that precisely these patches are
used in XFEM and were also used by Strouboulis et al. in [65,66] and
[67]. They could also be spherical, as used in the Method of Spheres
and Method of Clouds.
4 Extended Finite Element and Meshfree Methods

• Relative to the open cover {ωi }, the PUFEM uses a partition of unity {φi }.
Standard finite element hat functions, with respect to a finite element
mesh, could serve as the partition of unity, as are used in XFEM and
[65,66] and [67]. On the other hand, the “reproducing kernel particle
(RKP) shape functions” [38,40–42], can also be used as a partition of
unity.
• Special, problem dependent, finite dimensional spaces Vi , defined on
ωi ’s, are used for local approximation in PUFEM. These finite dimen-
sional spaces may contain constant functions together with singular
functions, harmonic polynomials or other special functions based on
the available information [9,44]. In fact, a one-dimensional Vi contain-
ing only constants, together with reproducing kernel particle functions
as a partition of unity, define many meshless methods. The functions
in Vi could also be constructed using data-fitting ideas, e.g. MLS func-
tions, as it was done in the Method of Clouds.
We finally mention that PUFEM is a Galerkin method, where the trial
space is constructed as
 
SPUFEM = φj Vj = {v ∈ H 1 () : v = φj vj , vj ∈ Vj }.
j j

If {νji }ni=(j1) is a basis for Vj , then

SPUFEM = span{φj νji };

the functions φj νji will be referred to as shape functions of PUFEM.

With appropriate choice of φj and Vj , the space SPUFEM yields precisely
the trial spaces used in various meshless methods and XFEM. See [8,9].
It is apparent that PUFEM is essentially an extension of the standard fi-
nite element method, but some added flexibility is obtained by the selection
of the partition of unity functions φj and local approximation spaces Vj . In
fact it has been shown that the PUFEM/GFEM using the hat-function par-
tition of unity and Vj = Pk (ωi ) is equivalent to the finite element method
using Lagrange shape functions in certain situations [9].
One of the main advantages of PUFEM is the ability to use a discretiza-
tion scheme that is easy to construct compared to the “meshing” step of the
standard FEM. While tremendous advances have been made in the area of
mesh generation, generating a satisfactory mesh for problems with compli-
cated boundaries or discontinuities, especially in three dimensions, remains
a time-consuming process which often requires human interaction. This is
Introduction 5

particularly inconvenient for problems where the meshing needs to change

at each time-step, such as crack propagation or certain fluid flow prob-
lems. The difficulties of mesh-generation have led to the development of
meshless or “mesh-free” methods, which were briefly discussed above.
Another attractive feature of PUFEM is that it allows the use of local
approximation spaces which can be adapted to the problem at hand. In
general, the better the functions in Vj approximate locally the exact solu-
tion of the PDE, the better will be the global approximation quality of the
PUFEM solution. Therefore it is possible (and desirable) to tailor the ap-
proximation spaces Vj to the information available about the exact solution.
For example, if it is known that the exact solution is harmonic, one could
choose Vj to be the space of harmonic polynomials rather than all poly-
nomials. Also, if the exact solution has singularities due to the boundary
geometry, one could select a Vj whose functions accurately approximate
the singular behavior of the exact solution (as in XFEM, see [9,62,68]).
Conversely, if the exact solution is known to be smooth, one can
select a smooth partition of unity and local approximation spaces that con-
tain smooth functions (e.g. polynomials). Then the resulting global shape
functions of PUFEM are also smooth and will approximate well the un-
known exact solution. This is an improvement over the standard finite
elements, where it is more difficult to obtain an approximation which is
C 1 -continuous or smoother.
The added flexibility of PUFEM is not without some costs. The main
drawback, which is shared with most of the other mesh-free methods, is
the added complexity of the numerical integration. In PUFEM, instead of
integrating over shape-regular elements, one has to consider the support
of each φj which can have a more complicated structure. For example, if
the φj have radial supports, then their intersection is a lens-shaped region
which is difficult to handle by numerical integration routines. Moreover,
PUFEM and XFEM allow the use of approximating functions that may
not be piecewise polynomials of low degree, therefore accurate numerical
integration will require the use of a large number of quadrature points. In
general, numerical integration required in the assembly of the linear system
becomes the most time-consuming part of the solution process. On the
other hand, this step is more amenable to parallelization and the recent
advent of parallel CPU architectures is likely to make computational costs
manageable in a majority of cases [33].
Another potential pitfall is that certain choices of partition of unity
functions and local approximation spaces can give rise to (almost) lin-
6 Extended Finite Element and Meshfree Methods

early dependent PUFEM shape functions. If that is the case, the resulting
linear system is consistent but the stiffness matrix will be singular and spe-
cial algorithms will be needed to deal with these situations. Furthermore,
multi-grid solvers and other efficient methods available for the linear sys-
tems arising from the standard finite elements may need to be modified for
use with PUFEM [32].
We also mention that solving a Dirichlet boundary value problem by
PUFEM will require that the functions in the trial space SPUFEM satisfy the
essential boundary conditions. This in turn requires that the functions in
Vj , for ω̄j ∩ ∂ = ∅, satisfy the essential boundary conditions, which may
be very difficult to accomplish. Several different approaches have been pro-
posed to overcome this problem, such as coupling to mesh-based methods
close to the boundary [37], penalty or perturbation methods [2,9,39], the
Lagrange multiplier method [10,39,51], and a method due to Nitsche [34,
59]. Most of these approaches either add an extra layer of complexity to the
problem, or lose some of the advantages of PUFEM, or result in a loss of
the optimal rate of convergence. The most promising idea in the literature
seems to be Nitsche’s method which works by modifying the variational
formulation of the problem to account for the boundary data. This ap-
proach retains the optimal convergence rate and the mesh-free character of
the method and was implemented in [59].

1.2. Moving boundary problems

Moving boundary problems are of major importance in engineer-
ing and particularly challenging for computational methods. They in-
clude multi-phase flow, fluid-structure interaction, biofilm/tumor growth,
shape/topology optimization, inverse analysis or image registration, to
name some of the applications. Fracture is another challenging moving
boundary problem which will be the focus in the next section. There are
two subtle differences to the aforementioned problems: (1) A crack is not a
closed boundary but an open boundary and (2) boundary does not propa-
gate orthogonal to its interface but from the crack front.
There are two ways to model the boundary: (1) as smeared inter-
face and (2) as sharp interface. The first category commonly simplifies
the implementation of computational methods on cost of accuracy and/or
computational efficiency. A good example is the 88 line code for topology
optimization based on Solid Isotropic Material with Penalization (SIMP).
In SIMP, the material properties are interpolated and therefore continuous.
Introduction 7

Hence, the SIMP method cannot capture jumps in the strain/stress field
as they occur at material interfaces. Though SIMP and its improved ver-
sions have been successfully applied to numerous challenging problems, it
has its limitations. For instance, lack of robustness and mesh dependency
has been reported for multi-material and nonlinear optimization problems.
Methods based on level sets for instance capture the sharp interface but the
implementation effort is higher; the 129 line implementation of the level
set method from [14] is probably among the shortest available ‘level set’
codes. Note that also the solution procedure of the forward problem might
be more complex. In FEM for instance, the interface can be represented
‘explicitly’ or ‘implicitly’. In other words: When the interface is aligned to
the discretization, we call it an ‘explicit representation’; otherwise an ‘im-
plicit’ one. The first scenario requires the generation of new meshes when
the topology changes during the iterations which in turn deteriorates com-
putational efficiency when the entire stiffness matrix has to be reassembled.
When the interface is to be captured within an element by the level set
function, enrichment techniques as described in this book are needed. It
is of course possible to avoid enrichment schemes by employing a sim-
ple sub-triangulation techniques for integration and the outcome of the
topology optimization might not differ much from enrichment schemes in
many cases. However, the jump in the stress/strain field cannot be captured
and there might be applications where this is important, for instance when
interface phenomena needs to be accounted for as in topology optimiza-
tion of nano-structures. It is imperative to say that without enrichment, the
convergence rates will be sub-optimal when material interfaces are inside
an element.
Another issue is how to deal with moving interfaces. This apparently
depends on the application and method employed. In level set based topol-
ogy optimization where the interface is represented ‘implicitly’ by the level
set, the new interface (i.e. the new level set) is obtained through the solu-
tion of the (stabilized) Hamilton-Jacobi equation where commonly shape
derivatives are used as velocity normal to the interface. The procedure is
quite similar in fluid mechanics problems when level sets are used to implic-
itly represent the interface between two fluids. However, in this case, the
velocity normal to the interface is the velocity of the fluid which is inter-
polated from the nodes. In problems involving fluid-structure-interaction,
one can distinguish between interphase capturing methods and interface track-
ing methods. The most classical interface capturing method is probably ALE
(Arbitrary Lagrangian Eulerian) which adapts the Eulerian (fluid) mesh to
8 Extended Finite Element and Meshfree Methods

the deforming Lagrangian mesh of the solid. Classical interface tracking

methods are the marker and cell method, volume of fluid method and the
level set method.

1.3. Fracture mechanics

Material failure, or more precisely the nucleation and propagation
of cracks, is another challenging moving boundary problem. We devoted
a separate section on this topic as it is (1) one of the key applications in
this book and (2) different in several ways to above mentioned moving
boundary/interface problems. As already mentioned in the previous sec-
tion, cracks are open surfaces. Furthermore, the crack consists actually of
two crack surface (interfaces) which propagate from their crack front. Cri-
teria for crack propagation (and crack nucleation) is another topic itself and
will be addressed in Chapter 8 of this book.
Computational methods for fracture can be classified into continuous
approaches to fracture and discrete approaches. Popular continuous ap-
proaches are gradient models, nonlocal models, viscous models or phase
field approaches. They all smear the crack over a certain width and avoid
representing the crack topology and crack tracking algorithms. If the global
response is of interest, such approaches often work as well as discrete crack
approaches. However, the computational cost is commonly higher as a finer
discretization is needed in order to resolve the crack width. In certain ap-
plications such as fracture in heterogeneous structures, the implementation
effort and the computational cost might raise drastically. For instance, the
interphase in polymer-matrix-composites is of the order of several nanome-
ters and the length scale parameter of a phase field model should be less
than approximately 10% of the characteristic length of the composite; in
this case the interphase thickness. It requires at least one element to resolve
the characteristic length. On the other hand, similar results can be achieved
by discrete crack approaches where it might be sufficient to resolve the in-
terphase with a single element. Also, the application to dynamic fracture is
challenging for continuous approaches for fracture. The micro-branch in-
stability problem from [19–21] which could be modeled by discrete fracture
approaches [58] is a good example which has never been solved success-
fully by a continuous crack approach. Fluid-driven fracture problems such
as fluid flow through fracturing thin structures is another example where
continuous approaches for fracture are prone to fail. Though there are
continuous approaches for fluid-driven fracture, they cannot model the
Introduction 9

fluid flow through the propagating cracks, i.e. through the solution of the
Navier-Stokes equation which can be easily done by discrete crack ap-
proaches.
The discrete approach for fracture requires two key ingredients:
• a method to capture the crack kinematics (the discontinuous displace-
ment field) and
• a fracture criterion for crack propagation/nucleation.
Classical representatives of computational methods for discrete fracture are
element deletion, cohesive elements, remeshing techniques, boundary ele-
ment methods, embedded elements, extended finite element and meshfree
methods, cracking particles methods and peridynamics. Cohesive elements,
boundary elements and remeshing techniques align the crack to the dis-
cretization while embedded elements and extended finite element and
meshfree methods allow for arbitrary crack propagation through the dis-
cretization by exploiting partition of unity enrichment. Usually, they model
the crack as continuous surface which requires (1) a crack surface represen-
tation (e.g. by triangles in 3D) and (2) crack tracking algorithms. Note that
ensuring a continuous crack surface in 3D is nontrivial. This issue will be
discussed in Chapter 8.
Fracture can be categorized as brittle fracture, quasi-brittle fracture and ductile
fracture. Brittle fracture can be modeled by linear elastic fracture mechanics
(LEFM). Brittle materials are characterized by linear elastic material behav-
ior in the bulk and a small fracture process zone. However, LEFM cannot
be used when the failure process zone is of the order of the size of the
structure. The relative size of the fracture process zone lpz with respect
to the smallest critical dimension D of the structure is important for the
choice of the fracture model [26]. For D/lpz > 100, LEFM is valid while
for 5 < D/lpz < 100, a quasi-brittle fracture approach is required which
accounts for the energy dissipation at postlocalization. Gdoutos [26] rec-
ommends the use of nonlocal damage models for D/lpz < 5. The length of
the fracture process zone is of the order of the characteristic length [46]:
E Gf E Gf
lch = or lch =   (1.1)
ft 2
1 − μ2 ft2
where E is the Young’s modulus, μ the Poisson’s ratio, Gf the fracture en-
ergy and ft denotes the tensile strength of the material. Carpinteri [13]
G
introduced a non-dimensional brittleness number sE = ft bf where b is a
geometric measure. He tested pre-notched beams under 3-point bending
where b denotes the distance between the notch and the upper boundary of
10 Extended Finite Element and Meshfree Methods

Figure 1.1 Different failure modes.

the beam. He concluded if the process zone is small compared to b, the fail-
ure is brittle and LEFM is applicable. Otherwise, the energy dissipation at
postlocalization cannot be neglected and a quasi-brittle approach is needed.
A common approach to account for the energy dissipation which is well
suited for discrete crack methods are cohesive zone models. When large
plastic deformations occur before the material looses stability, the fracture
is ductile.
There are three failure modes, see Fig. 1.1. Mode-I failure is related to
crack opening, mode-II failure (sliding) is a pure shear failure mode, and
mode-III failure (tearing) can be considered as out-of-plane shearing. In
many applications, materials will fail due to a mixed mode failure.

1.4. Level set methods

In the modeling of moving boundaries problems such as crack prop-
agation, phase-changing, topology optimization and inverse analysis, a ge-
ometry description of interfaces is necessary. Therefore, a numerical scheme
that can be incorporated with the XFEM to track and update the geometry
should be devised. The level set method is perhaps one of the most widely
used method apart from other method such as parametric description. The
level set method (sometimes abbreviated as LSM) is a recently developed
numerical method to capture the motion of interfaces and shapes in an
implicit way. The name “level sets” refers to the sets which collect points
having the same certain level, i.e. value of signed distance. It provides an
implicit way of describing the geometry of a surface by measuring the
 Another Random Document on
Scribd Without Any Related Topics
(Doolittle) agreed, with the exception of one teacher who was a southerner, that there
was never anything but the most natural mingling in the classrooms, about the
building and on the playground. At a school 30 per cent Negro (Drake), the principal
of which stated that the relations between the races were not harmonious, the
investigator observed a free and natural grouping of Negroes and whites of all ages
on the playground. The principal explained that this was "a forced rather than a
natural grouping because of lack of apparatus for all." The white children at a school
20 per cent Negro (Haven) were Italians, Jews, and Greeks, and all the races played
so naturally together that passersby frequently stopped to watch them.
Social contacts.—There are few social organizations and gatherings in the elementary
schools. The principal of a school 93 per cent Negro (Raymond) said that there were
clubs through all the grammar grades and that the friendliness between the two races
was marked, but added:
We have not more than fifty or sixty white children in this particular building. One white
child was elected vice-president, the first white child elected in eight years. It shows the
friendly relationship when a white child could be elected to office with a large
preponderance of colored children. A Jewish boy was elected to a smaller office of clerk.
The white children are not foreign. In their meetings the question of color never arises
at all.

In a few instances principals had found that graduation presented some difficulties, as
white mothers would appear at the school a few days before and request that their
children do not march with Negro children. "About the only time I see a white mother
is near graduation," said the principal of a school 38 per cent Negro (Forrestville).
"They always say they wouldn't care for themselves, but a friend might see and they
would feel ashamed." "White children prefer not to march with colored at graduation,"
said a teacher at Oakland School (26 per cent), "and mothers sometimes come to ask
that it be so arranged that their girls can march with white girls. They usually say that
for themselves they don't mind, but friends might see and wonder why that should
be."
A number of the schools have orchestras or occasional musical programs. The
investigator heard one orchestra of eleven pieces in Doolittle School (85 per cent),
which played remarkably well. All but one of the children were Negroes. A teacher in
Webster School (30 per cent), where there was reported to be constant friction
between Negro and white children, gave an incident of a Negro boy in the school
playing the violin with a white accompanist and being enthusiastically applauded by
the children.
The principal of a 92 per cent Negro school (Colman) reported an unpleasant
experience when pupils from her school were invited to take part in a musical program
at a West Side Park.
A group of sixty went with two white teachers in charge. On the way over a group of
foreign women called out insulting remarks to the teachers, but no one paid any
attention. After the program the group started marching out of the park and were met
 at the gate with a shower of stones. The teacher told the children to run for their lives,
and they all had to scatter and hide in the bushes in the park or run toward home if
they could. A rough set of boys had got together and were waiting for those children,
stones all ready to throw. Since that time we have never accepted an invitation to sing
outside our own neighborhood. Invitations have come from time to time, but the
children all come with excuses. All of them, children and parents throughout the
neighborhood, are afraid but you can't get anyone to come out and say it.

Attitude of parents.—Principals and teachers were questioned about their relations

with the parents of both Negro and white children—whether they received co-
operation from the parents in matters of discipline; what was the attitude of the
parents toward Negro teachers; and whether many requests were received from
Negro or white parents for transfers to schools where there were fewer Negroes.
In general it may be said that the principals who found Negro parents unco-operative,
unambitious, and antagonistic were those who believed in separate schools, found
Negro children difficult to discipline, and would have no Negro teachers in their
schools. Such principals declared that Negro parents were "10 to 1 in the complaints
brought into the office,"[38] and that "they fuss over everything and tell their children
not to take anything from a white child." They also cited cases of insolence and
threats which appeared to be exceptional rather than typical.
Some teachers said the reason they did not receive any co-operation from Negro
mothers was because a large proportion of them were working. Tardiness and
absence were due mainly to this cause, according to one principal, though a teacher
of a room for retarded children in another school said there was little tardiness and
practically no absence in her group. This teacher expressed the conviction, as did
many others, that Negro parents were appreciative of school advantages and eager to
have their children learn. Principals who came in contact with both Negro and foreign
parents found the Negro parents much more interested and ambitious than the
foreigners. Even the principal of a school 30 per cent Negro (Webster), who was
somewhat prejudiced in her attitude toward Negroes in the school, said she had more
Negro than white boys able to go to work whose parents wished them to remain in
school.
Negro teachers were apparently acceptable to Negro parents, only one of the
principals or teachers interviewed reporting objections by Negro parents. One teacher
in a school 30 per cent Negro (Webster) said that Negro parents had their children
transferred there from schools with more Negroes, so that they would have white
teachers. The district superintendent said he had had some difficulty in placing Negro
teachers in Negro schools, which he attributed to the fact that Negro parents felt that
Negro teachers had not had the same opportunity for thorough training as white
teachers. Some Negro parents, however, had indicated that their attitude was not due
to belief that Negro teachers were inadequately trained, but to fear that too general
placing of Negro teachers over Negro pupils was a step toward segregation.
The principal of a school 90 per cent Negro (Keith) thought Negro mothers preferred
Negro teachers because several had said to her that the "colored teachers understand
our children better."
The district superintendent in the area including most of the schools largely attended
by Negroes said that few requests for transfers were made during the year, but he
believed more were made at the request of Negro than of white parents. A number of
these Negro children transferred not to go to a school largely white but to a school 70
per cent Negro, because they said they were afraid to go to the school in their own
district which was across Wentworth Avenue. The race feeling between certain groups
in this district was very intense, according to the superintendent. It was especially
violent between the Negro children and the Italians and between the Jews and the
Bohemians. The principal of a school 93 per cent Negro (Raymond) also testified to
the spirit of antagonism along Wentworth Avenue:
Wentworth Avenue is the gang line. They seem to feel that trespass on either side of
that line is ground for trouble. While they will admit colored members to the school
without any trouble for manual training, they have to be escorted over the line, because
of trouble, not from members of the school, but groups of boys outside the school. To
illustrate: We took a kindergarten group over to the park. One little six-year-old girl was
struck in the face by a man. The condition is a tradition. There does not seem to be any
malice in it. "He is from the east side," or "Hit him, he is from the West Side," are
remarks frequently heard.

Transfers from schools with a predominant Negro membership were reported by one
or two principals and teachers in schools with a Negro minority, who said that the
Negro mothers objected to having their children in schools "where there are so many
common niggers." One of the principals said she had many requests from Negro
mothers for transfers from the branch of the school with 90 per cent Negroes to the
main school with 20 per cent. The Commission did not find in its inquiry among Negro
mothers that such an objection was prevalent, but that most of the transfers
requested were due to the reputation of the school for being overcrowded, poorly
taught, and generally run down.

2. HIGH SCHOOLS

Classroom and building contacts.—In the high schools the ordinary contacts in classes
and about the building become subordinate to the more difficult problems created by
the increased number of social activities—athletics, gymnasium exhibitions, clubs, and
parties.
The dean of Englewood High School, which has only about 6 per cent Negro children,
said that the white and Negro children mingled freely with no sign of trouble or
prejudice but thought that if more Negro children came to the school the spirit would
change. A teacher in this same school who had formerly been at Wendell Phillips,
where the majority are Negro, said that a spirit of friendliness had grown up there
between the two races, and race distinction had disappeared.
 WENDELL PHILLIPS HIGH SCHOOL
Located at Thirty-ninth Street and Prairie Avenue, 52 per cent Negro attendance.

There was only one Negro teacher in the high schools of Chicago at the time of this
investigation, the teacher of manual training at Wendell Phillips. He is a graduate of
the University of Illinois and had substituted around Chicago for several years.
Although they spoke very highly of him, none of the principals of three high schools
with small Negro percentages and in which there were vacancies could use him. The
principal of Wendell Phillips, with a large proportion of Negroes, told, however, of a
different experience when this teacher was at that school. "In answer to complaints
by pupils I told them that this man was a graduate of the University of Illinois, a high-
school graduate in the city, and a cultured man. 'Go in there and forget the color, and
see if you can get the subject matter.' In the majority of cases it worked."
Racial friction about the buildings and grounds was not reported by any of the high-
school principals. "I have not known of a fight between a colored and a white boy in
fifteen years," said the principal of Hyde Park.
Two principals said that the Negro children voluntarily grouped themselves at noon,
either eating at tables by themselves in the lunchroom or bringing their own lunches
and eating in the back part of the assembly hall. The gymnasium instructor at Wendell
Phillips said that she had no difficulty in her work if she let the children arrange
themselves. The gymnasium instructor at a school with a small proportion of Negroes
said that the white girls had objected to going into the swimming-pool with Negro
girls, but that she had gone in with the Negro girls, which had helped to remove the
prejudice.
Athletic teams.—In the field of athletics there seems to be no feeling between the
white and Negro members of a school team, but the Negro members are sometimes
roughly handled when the team plays other schools. "The basketball team is half and
half," said the principal of Wendell Phillips. He reported some friction in previous years
but said that "this year it is not shown at all." "They played a strenuous game with
Englewood last week. A colored boy was roughly treated by the other team. Our white
boys were ready to fight the whole Englewood team."
The principal of Hyde Park High School also said that there was no feeling in his
school against Negro members of athletic teams, and that he did not know of a single
instance in which a Negro boy was kept off an athletic team if he was the best for the
place.
Two Seniors in a high school mainly white (Tilden) thus described the way they
handled the Negro members of a visiting basket-ball team:
On the way over here fellows on the outside bawled them out, but our fellows sure got
them on the way home. There were three black fellows on the team and those three
got just about laid out. Our team wouldn't play them, so there was a great old row.
Then, when they went home some of our boys were waiting for them to come out of
the building to give them a chase. The coons were afraid to come out, so policemen
had to be called to take them to the car line. The white fellows weren't hurt any, but the
coons got some bricks.

Transfers between high schools.—Requests for transfers from Wendell Phillips to

Englewood and Hyde Park schools had been made by both white and Negro children,
according to the principals of the latter schools. The permits of the Negro children had
frequently been revoked after they had been admitted to classes, and the children
returned to Wendell Phillips. A teacher at Wendell Phillips pointed out the injustice of
transferring a child in the middle of a term. After a child has been admitted to classes
he should be permitted to remain through the semester, she believed, for otherwise a
full term's work was lost because the courses in the schools were different. "All this
transferring is nonsense, anyway," she said. "Children should be made to go to school
in the district where they live and that would end the trouble."
This teacher told of an incident at Tilden School when a group of Negro boys
registered for entrance:
About sixty colored boys entered Tilden High School either for the regular high-school
course or prevocational work and were thrown out by the Tilden boys. They made it so
hot for the colored boys that the sixty had to withdraw. Some came back here; others
dropped out of school entirely. It's pretty bad when one set of boys can put out another
set and nothing is done to punish one and call back the other group.

Two boys at Tilden who took part in this affair gave this version of the incident:
About thirty colored boys registered at Tilden last fall, but we cleaned up on them the
first couple of days and they never showed up again. We didn't give them any peace in
the locker room, basement, at noon hours, or between classes—told them to keep out
of our way or we'd see they got out. The fellows who were in school before we didn't
tackle—they know where they belong. There's one colored fellow in our class everybody
likes. He's a smart nice fellow to talk to, and he doesn't stick around when you don't
want him. He didn't say anything when we made the new coons step around, but I
guess he didn't like it very well.
It was this same group of boys who objected to playing a visiting basketball team with
three Negroes on it and "just about laid them out."
Social activities in high schools.—In high schools, with their older pupils, there is an
increased race consciousness, and in the purely social activities such as clubs and
dances, which are part of high-school life, there is none of the general mingling often
found in semi-social activities such as singing and literary societies. Although Negro
pupils do not share in the purely social activities, they do not organize such activities
among themselves.
"The colored never come to social affairs," said the dean of one school. "They are so
much in the minority here that they leave all organizations to the whites." The
principal of this school told of having seen two colored girls at a class party who
danced together for a while and left. "It is the only time I've seen the two races at the
same social gathering."
The dean of Englewood said: "We have colored children in singing clubs, in the
orchestra, in literary societies, in class organizations, and on athletic teams. Always
when there is a class party there will be five or six colored children. They will always
dance together, but they are present and welcomed by the white. Between dances it
is not uncommon to see white and colored talking."
An incident showing lack of feeling against individuals of special achievement was
given by the principal of this last school:
Several years ago we organized a voluntary orchestra which met after school. The
director accepted all applications, among them a number of colored boys. The white
boys balked; it should be white membership or they would leave. As it was near the end
of the year the orchestra was dissolved. The next year I suggested to the teacher that
he fill the orchestra places by a general tryout, so understood, but really with the policy
of excluding the colored. This was done and a white orchestra organized. Shortly, the
father of H. F., a colored boy who had been excluded, protested in my office, saying that
his boy had been excluded because of race prejudice and that he was going to carry his
protest to the Board of Education, for he knew his boy played better than any boy in
school. I admitted that it was a choice in the school of white orchestra or no orchestra,
but that if his boy was the fine musician he said he was I would gladly see what could
be done. Soon after that H. appeared on a school program and played with remarkable
skill and technique. He was applauded enthusiastically and recalled three times.
Straightway the orchestra members asked him to play with them. He became unusually
popular throughout the school. His standing was the highest and he was awarded a
scholarship of $100 allowed by the Board of Education for the best student. He was also
chosen to represent the school on the Northwestern University scholarship, and in his
Freshman year he won another scholarship for the next year. The death of his parents
made it necessary for him to leave college to support his brothers and sisters. At this
time he was stricken with infantile paralysis. The interest on Liberty bonds taken out by
the high school is paid in to H., and when the colored people gave a benefit for him the
pupils sold 500 tickets. He is improving and teaching violin to thirty pupils at present.
His sister is in the school now on a scholarship and is doing remarkably well also.

At Wendell Phillips the situation was quite different, for there were no school or class
social affairs which were general. There were invitational affairs to which the Negroes
were not invited. All the clubs in the school were white, Negroes being excluded. The
principal said he would not insist on mixed clubs until he saw the parents of the
children mixing socially. The glee club was an especially difficult problem because of
its semi-public as well as social character. The Negro children maintained that a glee
club composed entirely of whites was not representative of a school in which the
majority were Negroes. The Negroes had not responded to the suggestion of the
principal that they form a glee club of their own, and as the white children would not
be in a glee club with Negro children, there was constant friction over this club.
Other principals expressed the conviction that the racial problem of school social
affairs could not be solved until the prejudice and antagonism of adults had
disappeared. One principal said he had had to call off an arrangement for a class affair
because the hotel would not accommodate the Negroes. Another principal thought
that the schools would not wait to follow the lead of the parents in forgetting the race
prejudice but would themselves be the greatest factor in destroying it.
Relations with parents.—In most cases the high schools were receiving splendid
support from Negro parents in matters of discipline. "I have never had a case where
the parent did not back up the teacher in the treatment given to a colored child," said
one principal, speaking of cases where children had got into difficulty when they
complained that the teacher had "picked on them" because they were Negroes. The
parents always made the child withdraw the statement and admit that the trouble was
not due to color at all.

3. TECHNICAL HIGH SCHOOLS

Reports were received from three technical high schools, Lane, Tilden, and Lucy H.
Flower. Lane and Tilden had few Negro students, while in Lucy H. Flower the Negroes
were about 20 per cent. The principals of Lane and Tilden said they were not
conscious of any racial difference in their pupils, that no special methods of instruction
were necessary for the Negro children, that there were no quarrels with a racial
background in the schools, and no voluntary or compulsory groupings of white and
Negro. The principal of Lucy H. Flower found racial differences between the Negroes
and whites which she believed created special problems of education and discipline.
The children got along together very well in school, and whatever quarrels there were,
the principal thought were due to personal dislikes rather than to race prejudice. The
colored girls grouped themselves voluntarily at noon and at dismissal time, and the
white girls did the same.

III. RETARDATION

1. RETARDATION IN ELEMENTARY SCHOOLS

With the assistance of the Board of Education a selection was made of three groups of
schools to be studied for comparative retardation. The group comprised six schools
having the largest percentage of Negro children, six attended mainly by whites in
neighborhoods where the family income might be comparable, and twelve attended
mainly by children of immigrants. Table XII gives the number and percentage of
accelerated, normal, and retarded children for each school, for each group, and for
the whole group of twenty-four schools.
This table shows the much greater amount of retardation among schools attended by
Negroes than in schools attended by white Americans or by children of immigrants.
The percentage for the group attended by Negroes is 74, while for the different
schools in the group it varies from 67 to 81. For the two groups of schools attended
by white Americans the percentage of retardation is the same, 49, though there is
greater variation among these schools than among the schools attended by Negroes.
In the group attended by children of immigrants, for instance, only 32 per cent are
retarded in the Jungman (Bohemian) School, while 71 per cent are retarded in the
Holden (Polish) School. A similar discrepancy appears in the group attended by white
Americans, where the figure is 40 per cent for the Armstrong School and 62 per cent
for the Byford School.
TABLE XII
Number and Percentage of Children in Accelerated, Normal, and Retarded Groups in Schools
Attended Mainly by White Americans, by Negroes, and by Children of Immigrants

Retarded
School Accel. Percentage Normal Percentage Retarded Ungrad. Percentage Total
[39]

Attended
mainly by
white
Americans:
Armstrong 202 21 365 39 355 19 40 941
Byford 118 9 361 29 783 — 62 1,262
Harper 291 17 609 35 829 — 48 1,729
Howe 220 17 421 35 577 — 48 1,218
Key 173 25 205 29 314 — 46 692
Morse 169 14 450 37 581 — 49 1,200
Total 1,173 17 2,411 34 3,439 19 49 7,042

Attended
mainly by
Negroes:
Coleman 54 8 124 17 561 2 75 743
Doolittle 267 16 261 16 1,099 24 68 1,651
Douglas 136 9.3 197 13.7 1,126 — 77 1,463
Keith 77 11 93 14 497 — 75 667
Moseley 62 7.5 95 11.5 551 122 81 830
Raymond 112 13 179 20 578 — 67 869
Total 708 11 949 15 4,412 148 74 6,217
 Attended
mainly by
children of
immigrants:
Bohemian
Bryant 385 21 735 37 809 15 42 1,944
Hammond 161 12 503 34 795 — 54 1,459
Jungman 375 35 350 33 357 — 32 1,082
Polish
Chopin 298 17 631 36 818 1 47 1,748
Hibbard 392 29 445 32 535 — 39 1,372
Holden 122 11 208 18 759 — 71 1,089
Italian
Goodrich 157 14 240 22 693 — 64 1,090
Jackson 360 15 731 32 1,174 — 53 2,265
Jenner 176 11 524 33 875 — 56 1,575
Jewish
Herzel 609 25 731 30 1,085 — 45 2,425
Lawson 466 16 944 32 1,407 20 52 2,837
Von 528 22 848 34 1,072 — 44 2,448
Humboldt
Totals 4,029 19 6,890 32 10,379 36 49 21,334
Totals 5,910 17 10,250 30 18,230 203 53 34,593
for
three
groups

The retardation figures for the group of twenty-four schools studied are close to those
for the city at large, 53 per cent retarded in the special group and 51 per cent for the
city at large. In the accelerated group the percentage of accelerated Negro children,
11, is smaller than the percentage of accelerated white children, 17, or the percentage
of accelerated foreign children, 19. This variation is not so striking as that in the
normal group where only 15 per cent of the Negro children appear to make normal
progress as compared with 34 per cent of the white children and 32 per cent of the
foreign children. From this it would appear that there are factors in the lives of many
Negro children which prevent them from making normal progress.
The degree of retardation, as shown in Table XIII is again quite different for the white
and Negro groups.
The largest single groups of backward white American and foreign children are
retarded less than one year (42 per cent of the white American and 39 per cent of the
foreign group), and the numbers decrease rapidly as the degree of retardation
increases. In the case of the Negroes 19 per cent are retarded less than one year. The
decrease as the degree of retardation increases is slower than in the white groups,
and many more children are retarded two, three, four, five years and more. In the
white American group only one child out of 3,439 retarded children is retarded five
and one-half to six years, while there are forty-one in the corresponding Negro group
out of a total of 4,412. One white child is retarded six and one-half to seven years,
while seventeen Negro children are retarded this amount; twelve foreign children out
of 10,379 retarded children are retarded six to ten years, and thirty-seven Negro
children are found in these groups.
Though the main reasons for the high degree of retardation among Negro children are
set forth in the next section under "Causes of Retardation," a partial explanation is to
be found in the fact that Negro parents are frequently more interested in keeping their
over-age children in school than white parents, especially foreign parents, whose
anxiety to have their children leave school as soon as they are old enough to get
work-permits is well known.
Causes of retardation.—It is generally understood of course that comparisons of
Negro with white children are hardly fair, since Negro children have not had the same
opportunities as whites to make normal progress.
A study was made of the reasons why children were retarded in the groups of schools
attended mainly by Negroes, by white Americans, and by children of immigrants.
Records were obtained at the schools for 1,469 Negro children and 1,560 white
children who were listed according to the Board of Education's classification for
retarded children.
Table XIV shows clearly that the predominating cause of retardation among Negroes is
late entrance, which, according to the board's classification, means that they did not
enter school until more than six years of age. This is generally explained by the fact
that the family came from the South, where there was no school near enough for the
child to attend, or the school was overcrowded, or the family was uneducated and
indifferent. In some cases the parents have come North, leaving the child with
grandparents who made no effort to see that it went to school.
TABLE XIII
Number of Children in Table XII Who Are Retarded One-half to One Year, One Year to One and
One-half Years, etc.

YEARS
SCHOOL ½ to 1 1 to 1½ 1½ to 2 2 to 2½ 2½ to 3 3 to 3½ 3½ to 4
Attended mainly by white
Americans:
Armstrong 143 84 62 31 15 13 3
Byford 317 175 105 64 54 20 11
Harper 364 234 106 67 28 12 6
Howe 275 128 93 57 9 9 2
Key 141 106 33 19 20 2 2
Morse 229 160 81 56 27 14 8
Total 1,470 887 480 294 143 70 32
Attended mainly by Negroes:
Colman 109 114 109 66 50 46 27
Doolittle 229 175 146 152 117 86 69
 Douglas 190 198 191 142 126 83 71
Keith 94 66 71 78 54 34 30
Moseley 95 104 96 54 59 31 34
Raymond 135 115 111 69 50 39 27
Total 852 772 724 561 456 319 258
Attended mainly by children of
immigrants:
Bohemian:
Bryant 369 224 107 51 36 15 3
Hammond 306 225 114 69 47 13 10
Jungman 173 87 49 27 12 6 2
Polish:
Chopin 323 216 125 57 43 23 16
Hibbard 252 158 68 28 18 5 4
Holden 216 190 112 91 60 36 18
Italian:
Goodrich 236 185 104 68 42 28 10
Jackson 369 284 202 141 84 53 16
Jenner 281 253 135 86 42 32 20
Jewish:
Herzel 521 294 124 71 38 19 9
Lawson 574 370 233 109 65 24 19
Von Humboldt 498 145 145 76 37 17 6
Totals 4,118 2,710 1,518 874 524 271 133
Totals for three groups 6,440 2,249 2,722 1,729 1,123 660 423

TABLE XIII (continued)

Number of Children in Table XII Who Are Retarded One-half to One Year, One Year to One
and One-half Years, etc.

YEARS
SCHOOL 4 to 4½ 4½ to 5 5 to 5½ 5½ to 6 6 to 6½ 6½ to 7
Attended mainly by white Americans:
Armstrong ... 4 ... ... ... ...
Byford 32 2 2 ... ... 1
Harper 4 3 1 ... 1 ...
Howe ... 4 ... ... ... ...
Key 1 ... ... ... ... ...
Morse 3 1 1 1 ... ...
Total 40 14 4 1 1 1
Attended mainly by Negroes:
Colman 18 7 8 4 2 ...
Doolittle 44 37 16 8 4 4
Douglas 45 26 21 14 10 7
Keith 32 17 8 6 4 3
Moseley 26 20 15 7 4 2
Raymond 12 10 6 2 ... 1
Total 177 117 74 41 24 17
Attended mainly by children of immigrants:
 Bohemian:
Bryant 1 2 1 ... ... ...
Hammond 8 2 1 ... ... ...
Jungman 1 ... ... ... ... ...
Polish:
Chopin 9 2 1 2 ... ...
Hibbard ... 2 ... ... ... ...
Holden 13 9 8 4 2 ...
Italian:
Goodrich 9 2 3 4 2 ...
Jackson 13 4 4 2 ... 1
Jenner 10 3 4 4 2 2
Jewish:
Herzel 4 4 ... 1 ... ...
Lawson 7 4 2 ... ... ...
Von Humboldt 6 2 1 ... ... ...
Totals 81 36 25 17 6 3
Totals for three groups 298 167 103 59 31 21

TABLE XIII (continued)

Number of Children in Table XII Who Are Retarded One-half to One Year, One Year to One and
One-half Years, etc.

YEARS
7 to 7½ to 8 to 8½ to 9 to 9½ to 10 to
SCHOOL
7½ 8 8½ 9 9½ 10 10½
Attended mainly by white Americans:
Armstrong ... ... ... ... ... ... ...
Byford ... ... ... ... ... ... ...
Harper 1 ... ... 1 ... ... ...
Howe ... ... ... ... ... ... ...
Key ... ... ... ... ... ... ...
Morse ... ... ... ... ... ... ...
Total 1 ... ... 1 ... ... ...
Attended mainly by Negroes:
Colman ... 1 ... ... ... ... ...
Doolittle ... 4 4 1 1 2 ...
Douglas ... ... ... ... ... ... 2
Keith ... ... ... ... ... ... ...
Moseley 3 ... 1 ... ... ... ...
Raymond ... 1 ... ... ... ... ...
Total 3 6 5 1 1 2 2
Attended mainly by children of
immigrants:
Bohemian:
Bryant ... ... ... ... ... ... ...
Hammond ... ... ... ... ... ... ...
Jungman ... ... ... ... ... ... ...
Polish:
Chopin ... ... ... ... ... ... ...
 Hibbard ... ... ... ... ... ... ...
Holden ... ... ... ... ... ... 1
Italian:
Goodrich ... ... ... ... ... ... ...
Jackson 1 ... ... ... ... ... ...
Jenner ... ... ... 1 ... ... ...
Jewish:
Herzel ... ... ... ... ... ... ...
Lawson ... ... ... ... ... ... ...
Von Humboldt ... ... ... ... ... ... ...
Totals 1 ... ... 1 ... ... 1
Totals for three groups 5 6 5 3 1 2 3

TABLE XIV
REASONS WHY 1,469 NEGRO CHILDREN AND 1,560 WHITE CHILDREN WERE RETARDED
IN GROUP OF TWENTY-FOUR SCHOOLS

PHYSICAL
SCHOOLS TOTAL LATE ENTERING FOREIGN ILL HEALTH
DEFECT
N. W. N. W. N. W. N. W. N. W.
Attended mainly
by Negroes:
Colman, 92 29 1 20 ... ... ... ... ... 1 ...
per cent
Negro
attendance
Doolittle, 85 603 6 190 ... ... 6 3 ... 68 ...
per cent
Drake, 24 per 72 58 41 11 ... 3 ... ... 4 15
cent
Farren, 92 171 7 49 ... ... ... 1 1 34 1
per cent
Felsenthal, 20 173 69 58 13 3 2 ... ... 6 12
per cent
Forrestville, 93 5 34 2 ... ... 2 ... 4 ...
38 per cent
Haven, 20 71 59 56 9 ... 6 ... ... 2 9
per cent
McCosh, 15 18 31 7 4 ... 2 ... 1 5 8
per cent
Oakland, 26 50 44 18 6 ... ... 1 ... 4 5
per cent
Raymond, 93 133 6 64 2 ... 2 1 ... 7 1
per cent
Webster, 30 52 31 27 7 ... 1 1 ... 5 9
per cent
Attended mainly
by White
Americans:
Fiske 2 88 ... 3 ... 16 ... 2 ... 1
Rowland ... 101 ... 13 ... 7 ... 2 ... 17
 Kenwood ... 25 ... 1 ... 5 ... ... ... 4
Attended mainly
by children of
immigrants:
Farragut ... 107 ... 12 ... 10 ... 2 ... 11
Goodrich ... 92 ... ... ... 28 ... 3 ... 8
Jackson ... 255 ... 32 ... 48 ... 12 ... 38
Jungman ... 21 ... 11 ... 2 ... ... ... ...
Kosciuszke ... 144 ... 23 ... 17 ... 4 ... 9
Lawson 1 155 ... 7 ... 15 ... 1 ... 19
McCormick ... 21 ... ... ... 2 ... 1 ... 7
Seward ... 131 ... 19 ... 29 ... 2 ... 14
Smyth ... 57 ... 3 ... 14 ... 1 ... 9
Swing ... 46 ... 9 ... 2 ... 1 ... 7
Totals by 1,469 1,560 564 187 3 217 9 33 140 204
races
Totals of 3,029 751 220 42 344
both races

TABLE XIV (continued)

REASONS WHY 1,469 NEGRO CHILDREN AND 1,560 WHITE CHILDREN WERE RETARDED
IN GROUP OF TWENTY-FOUR SCHOOLS

FAMILY DEFECTIVE DEFECTIVE VARIANT

SCHOOLS BACKWARD
DIFFICULTIES VISION HEARING MENTALITY
N. W. N. W. N. W. N. W. N. W.
Attended mainly by
Negroes:
Colman, 92 per ... ... ... ... ... ... ... ... 4 ...
cent Negro
attendance
Doolittle, 85 per 141 ... 7 ... ... ... 35 ... 73 ...
cent
Drake, 24 per ... 3 ... 2 ... ... ... 7 7 12
cent
Farren, 92 per 21 ... 3 ... 1 ... 12 2 18 ...
cent
Felsenthal, 20 per 56 15 ... 2 1 ... 7 1 33 10
cent
Forrestville, 38 10 ... 1 ... ... ... 4 ... 30 ...
per cent
Haven, 20 per ... ... ... 4 ... ... ... 3 7 19
cent
McCosh, 15 per ... ... ... 1 ... ... ... ... 1 1
cent
Oakland, 26 per ... 6 1 2 ... 2 3 6 3 5
cent
Raymond, 93 per 18 ... ... ... ... ... 1 ... 11 ...
cent
Webster, 30 per 6 2 ... ... ... ... ... 1 4 2
cent
Attended mainly by
White Americans:
Fiske ... 34 ... 1 ... ... ... ... 1 21
Rowland ... 3 ... 4 ... ... ... 9 ... 28
Kenwood ... 4 ... ... ... ... ... 2 ... 5
Attended mainly by
children of
immigrants:
Farragut ... 6 ... 3 ... ... ... 5 ... 23
Goodrich ... 7 ... 3 ... ... ... 2 ... 23
Jackson ... 38 ... 8 ... 1 ... 3 ... 45
Jungman ... 1 ... ... ... ... ... ... ... ...
Kosciuszke ... 18 ... 2 ... ... ... 12 ... 4
Lawson 1 20 ... 2 ... 1 ... 25 ... 39
McCormick ... ... ... ... ... ... ... ... ... 1
Seward ... 10 ... ... ... 1 ... 2 ... 13
Smyth ... 4 ... 2 ... ... ... ... ... 14
Swing ... 4 ... 2 ... ... ... 1 ... 6
Totals by races 253 175 12 38 2 5 62 81 192 271
Totals of both 428 50 7 143 463
races

TABLE XIV (continued)

REASONS WHY 1,469 NEGRO CHILDREN AND 1,560 WHITE CHILDREN WERE RETARDED
IN GROUP OF TWENTY-FOUR SCHOOLS

LOW FEEBLE- IRREGULAR TEMP. IN DEMOTED FOR

SCHOOLS
MENTALITY MINDED ATTENDANCE GRADE CONDUCT
N. W. N. W. N. W. N. W. N. W.
Attended mainly by
Negroes:
Colman, 92 per ... ... 1 ... 1 1 ... ... 2 ...
cent Negro
attendance
Doolittle, 85 per 10 ... ... ... 68 ... 8 ... ... ...
cent
Drake, 24 per cent ... 1 ... ... 19 5 ... ... ... ...
Farren, 92 per 11 2 ... ... 21 1 ... ... ... ...
cent
Felsenthal, 20 per 3 4 ... ... 13 2 ... 1 ... ...
cent
Forrestville, 38 per 1 ... 1 ... 6 3 ... ... ... ...
cent
Haven, 20 per cent 2 4 ... ... 4 5 ... ... ... ...
McCosh, 15 per 1 6 ... ... 3 9 ... ... ... ...
cent
Oakland, 26 per 4 3 1 ... 12 9 1 2 ... ...
cent
Raymond, 93 per 6 ... 6 ... 19 1 ... ... ... ...
cent
 Webster, 30 per 9 6 ... ... 1 2 ... ... ... ...
cent
Attended mainly by
White Americans:
Fiske 1 2 ... 1 ... 6 ... 1 ... ...
Rowland ... 10 ... 1 ... 7 ... ... ... ...
Kenwood ... 1 ... ... ... 3 ... ... ... ...
Attended mainly by
children of
immigrants:
Farragut ... 32 ... ... ... 3 ... ... ... ...
Goodrich ... 10 ... ... ... 8 ... ... ... ...
Jackson ... 8 ... ... ... 19 ... 2 ... ...
Jungman ... ... ... ... ... 7 ... ... ... ...
Kosciuszke ... 8 ... 1 ... 39 ... 9 ... ...
Lawson ... 19 ... 2 ... 5 ... ... ... ...
McCormick ... 8 ... ... ... 1 ... ... 1 ...
Seward ... 11 ... 5 ... 20 ... 5 ... ...
Smyth 1 5 ... ... ... 5 ... ... ... ...
Swing ... 6 ... 1 ... 2 ... 4 ... ...
Totals by races 49 146 9 12 167 161 9 24 3 ...
Totals of both 195 21 328 33 3
races

The next most important cause of retardation among the Negroes is family difficulties.
The fathers are often kept away from home weeks at a time by their work. A large
number of the mothers are working, and the parents' lack of education is frequently
the cause of a home life that is below standard, physically and morally.
Among the whites, late entrance, inability to speak the language, ill health,
backwardness, and low mentality are the main causes of retardation. While it is often
maintained that the Negro is the mental inferior of the white, these figures do not
bear out that contention. Also the retardation figures do not show the home life of the
Negroes to be productive of as much ill health as is the case with the whites.
Approximately the same number of Negro and white children were retarded because
of irregular attendance.
In addition there were forty-two Negro children and 155 white children who were
classified under two, three, or four different causes for retardation. Children who were
late entering also had some physical difficulty, or children who were retarded because
of family difficulties were also of poor mental endowment. In some cases such double
classification represented a realization by the teacher that retardation is a complicated
and delicate thing which cannot be explained by one hard-and-fast reason. Others,
finding it difficult to decide whether children were backward, of low mentality, or
feeble-minded, classified them under all three causes. In two instances Negroes were
found to be retarded because they were late entering and "foreign"—that is, they
were handicapped by an "initial lack of the English language."
Intensive study of 116 retarded Negro children.—The presence of retarded Negro
children in the Chicago public schools within recent years has been regarded by many
teachers and principals as a problem of Negro education. Some assume that this
retardation is due to an inherent incapacity for normal grade work. Inquiries of the
Commission early disclosed the fact that although the retardation rate of Negro
children was higher than that of white, the great majority of the retarded Negroes
were from southern states, and that Negro children born in the North had, as a rule,
no higher rate of retardation than the whites. In the belief that the causes of
retardation among Negro children could be found in the same factors of social
background and environment which operates to retard white children, an intensive
study was made of 116 Negro children taken at random from among all the retarded
Negro children in several schools to learn what elements in their former life and
present home environment might explain their retardation.
Out of the 116 children 101 had been in school before coming to Chicago. Of these
eighty-six had lived in the South and attended southern schools. Since this group was
chosen at random, the large proportion from the South tends to bear out the
statements of school principals and teachers that Negro children from the South
constitute the bulk of retarded children. Previous school records were obtained for
eighty-four of these eighty-six southern children, and in sixty-four cases the children
were retarded when they came to Chicago. Many of them were retarded two and
three years, and some three, four, five, and even six years. Forty-seven of the sixty-
four were retarded more than one year. In a number of cases children who were in
the normal grade for their age in the South were put back one or two grades when
they entered Chicago schools because they were not equipped to do the work of this
grade in the North.
The states from which these children came are Alabama, Arkansas, Florida, Georgia,
Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas,
and Virginia. Twenty-three of the eighty-six children who had lived in the South were
from Mississippi—the largest group from any one state—and of these three were up
with their normal grade, eleven were retarded three or four years, one was retarded
six years, and one who was in the normal grade in the South was demoted two years.
One reason for the poor record of these Mississippi children is undoubtedly to be
found in that state's inadequate compulsory-education law which provides a school
term of eighty days in districts which do not reject the law. Eight of the Mississippi
children lived on plantations which were so far from school that regular attendance
was impossible.
Information gathered concerning the parents of these 116 retarded children showed
that in eighty-six cases the father was living with his family. In six cases the father
was dead, in one case he was insane, in fifteen cases he had separated from or
deserted the mother, and in eight cases there was no report on the father.
The mother was found to be living with her family in 112 cases. In two cases the
mother was dead, and in two cases she had deserted father and child.
All of the eighty-six fathers who lived at home were working, though one was
reported as working irregularly, and two as having deserted their wives occasionally
for periods of several weeks. In two of the cases where the father had separated from
the mother he was reported as contributing to the support of the child.
In forty out of the eighty-six cases where the father was living at home and working,
the mother was also working, and in the fifteen separation cases where the mother
was supporting the child, she was working. The fact that a total of fifty-five out of 112
mothers, or 49 per cent, were working is undoubtedly a large factor in the retardation
of the children. The statement was frequently made by teachers that 40 or 50 per
cent of the Negro mothers worked, and that the child was therefore neglected, and
the teacher could get no co-operation from the mother, as she was never free to come
to school to talk over matters affecting the child.
Some teachers felt that many mothers worked where there was no economic
necessity, as the father was earning enough to support the family. It should be noted
in this connection that at the time this material was gathered there were more
opportunities for work than there were men to fill them. Under ordinary conditions
there would doubtless be a certain amount of unemployment in these Negro families
which would cause more mothers to work from economic necessity. Many of the
families investigated, where both parents were working, were reported as getting on
very well, though there were some cases of real poverty. In a number of instances the
families could not seem to make ends meet on a good income because they were
ignorant and did not know how to spend their money, or because they had not been
able to adjust themselves to city life.
Of the eighty-six fathers who were working, few were in skilled occupations which
would command a substantial wage. Most of the mothers were engaged in work that
took them away from home. A few did sewing, hairdressing, and laundry work in their
homes, but the large majority went out to work. Work carried on in the home
frequently has as bad an effect on the child's school attendance as the mother's
absence, for the child is sometimes kept at home to help and often finds the work
more interesting than school.
The following occupations of mothers of retarded children were noted:
Day work 22
Stock Yards 12
Hairdresser 4
Laundry 4
Maid 4
Barrel factory 3
Seamstress 3
Domestic service 2
Box factory 1
Car cleaner 1
Cleaning (hospital) 1
 Dishwasher 1
Elevator 1
Foundry 1
Housekeeper 1
Lamp-shade factory 1
Waist factory 1

Education of parents.—Of the eighty-six fathers, thirty-one were illiterate, and forty-
eight had gone to elementary school but had completed only the second, fourth, or
sixth grade. Five of the fathers had gone to high school, and two were college
graduates.
The figures are slightly better for the mothers. Out of 112, twenty-one were totally
illiterate, seventy-six had gone to elementary school, ten had been in high school or
college, and five were not reported on. Eighty-eight per cent of the mothers,
therefore, and 91 per cent of the fathers had less than a high-school education.
Though there were many illiterate or poorly educated parents who were eager for
their children to have advantages which they never had themselves, others, as in any
illiterate group, no matter what the color, failed to appreciate the importance of
school.
Home discipline.—A number of teachers reported that they were unable to discipline
the children in school because they were undisciplined at home. In seventy-three of
the 116 homes there was found to be discipline, in twenty-two a lack of discipline, and
twenty were not reported on. Discipline seemed to be the responsibility of the mother
in the large majority of cases, and many of the twenty-two undisciplined children were
boys who were beyond the control of the mother. In every case but four where there
was no discipline the mother was working, so that the child did not receive much care
during the daytime and the mother was too tired to bother about discipline at night.
Lack of discipline can also be traced to the fact that the child has not always lived with
the parents but with relatives who have been lax in the matter of discipline.
Home care.—The physical condition of the home, the preparation and substance of
the meals, may be expected to affect a child's health and therefore his attendance at
school. The homes of eighty-four children were reported to be clean and twenty-five
not clean, while seven were not reported on. In twenty-one cases out of the twenty-
five reported not clean, the mother was working. In forty-seven cases out of the
eighty-four reported clean the mother was working. In many of the forty-seven cases
there was an aunt or grandmother who took care of the house.
In many homes the ignorance of the parents was obviously responsible for failure to
provide the kind of food adapted to the needs of the children. A great deal of fresh
meat, usually pork and bacon, potatoes, rice, and coffee were the staples, while green
vegetable, fruits, cereals, and milk were noticeably lacking. Also, when the mother is
away all day the food is hastily prepared, which usually means that it is fried. The girl
who gets home from school before her mother has finished her day's work usually
starts the dinner, or brings something from the delicatessen. Many children are given
twenty-five cents with which to buy lunch, and in three extreme cases the children
were given money to buy all their meals, with no supervision over what they ate.
Difficulty of adjustment.—When all the causes contributing to retardation were taken
into consideration in the histories of the 116 retarded children studied, it was still
obvious that the greatest stumbling-block to normal progress was previous residence
in the South. The retardation of children from the South is explained in a variety of
ways.
Some of the children from the South did not get along well because they had not
been able to adjust themselves to city life. They had been accustomed to the freedom
and outdoor life of the farm and did not like the confined life of the city. They felt
timid and shy in the midst of so many people, as they did not come much in contact
with people when they lived on southern farms four or five miles from the nearest
town. Most of these children had never gone to school for more than a few months at
a time, either because the school term was short or they lived too far from the school
to attend regularly. Consequently some of them found the nine months' term irksome.
Demotion.—A number of children were found to be over-aged for their grades
because they had been demoted one or two years when they came to Chicago. Some
of these had gone to school regularly in the South and were of normal age for their
grades, but the school term was so short that it was impossible for them to complete
the same amount of work in the same number of years as children in northern
schools. Children who were in the fifth or fourth grade in the South had been put back
to the third or second grade on entering Chicago schools. This sometimes discouraged
them so much that they dropped out of school on reaching fourteen, the age limit of
the compulsory-education law.
Inadequate schools.—Overcrowded and poorly taught schools also are responsible for
the retardation of southern Negro children. One girl attended a school which was in
session only three months a year and where there were 100 to 125 children under
one teacher. Consequently this girl was retarded four years. A boy who, when he
came to Chicago, was fifteen years old and six years behind his grade had always
lived in small country towns in the South. In one of these his teacher was the iceman.
"He didn't come to school until he was through totin' ice around," said the boy. "Then
if anyone wanted ice they comed after him. He wasn't learning me anything so I quit."
This boy was found to be ambitious and was attending school regularly in Chicago in
spite of the fact that he was conspicuously over-age for his grade.
Other causes of retardation.—Some over-age children are extremely sensitive about
their size and are irregular at school on this account. A fifteen-year-old boy who was 5
feet 8 inches tall was in the fifth grade. He refused to go to school because he was
larger than anyone in his class. At one time he was so ashamed of being seen in the
room with smaller children that he would go out of the classroom every time a girl
passed the door.
As in many white families where the importance of regular school attendance is not
fully understood, work at home or work after school hours is sometimes permitted to
interfere materially with school attendance. Older children are kept at home to look
after young children while the parents are away at work and sometimes when the
mother is home. A fourteen-year-old girl who was three years retarded had always
been kept out of school to do housework. The five younger children were all in the
normal grades for their ages but the fourteen-year-old girl had been out of school so
much she had lost interest. Other children were working after school hours selling
papers and delivering packages and wanted to leave school as soon as possible so
that they could work all the time.
The attitude of the teacher seemed in a few instances to be responsible for the child's
lack of interest. In one case the teacher threw a paper at a boy instead of handing it
to him, and the boy had refused to recite to her ever since. He went to school but
recited to his mother at home. Another boy had been kept back in school by a
misunderstanding between his mother and the principal. The principal took the boy
home with her to do some work around her house and kept him until nine o'clock. The
mother became so worried she had the police out looking for him. When she found
out the cause of his lateness coming home, she went to the school and threatened
the principal. The principal afterward refused either to promote the boy or transfer
him to another school.
Recreation.—A study of the favorite forms of recreation among 116 children, aside
from the few who reported that they had no time to play, showed the movies to be in
the lead. Children economized on lunch, buying potato salad and pickles, in order to
have enough left from their lunch money to go to the movies. One boy who worked
outside of school hours made $3 to $5 a week and spent most of it on the movies; he
went three or four times a day if he had the money. A few children played truant in
order to go to the movies.
TABLE XV
Favorite Recreation of 116 Retarded
Negro Children

Movies 85
Baseball 32
Reading 31
Marbles 29
Skating 20
Jumping rope 11
Music 6
Jacks 6
Vaudeville 5
Running games 4
Singing games 4
Sewing 3
Basket-ball 2
 Target practice 1
Pool 1
Mechanical toys 1
Drawing 1
Dolls 1
Bicycle 1
Typewriting 1
Swinging 1
Rolling hoop 1
Card games, checkers, etc. 1
Total 248

Most of these children had two and even three forms of recreation, and the second
was usually some form of outdoor recreation—baseball, marbles, or jumping rope.
Most of the younger ones went to the playgrounds, except those who had housework
to do or the few who did not care to associate with other children.
A reference to the section on "Recreation" will show that Negro children are limited in
their recreational activities by lack of recreation centers where they are welcome.
There are playgrounds for the younger children in the areas of Negro residence, but
no recreation centers with their varied indoor facilities for the older children.

2. OPINIONS ON SCHOLARSHIP OF NEGRO CHILDREN

Progress of the southern Negro.—The retarded Negro child, usually from the South,
who is conspicuous in the elementary schools, has been referred to in the section on
"Retardation in Elementary Schools." In some schools such children are put in the
regular grades, where they receive no special attention and can progress only one
year at a time, though most teachers agree that retardation is due to lack of
educational opportunity rather than to inability to learn. In other schools there are
special rooms for these children where they are advanced through several grades as
rapidly as possible.
Doolittle School (85 per cent) had six first-grade rooms for such children. In one of
these rooms there were about twenty-five children from twelve to seventeen years of
age doing all the lower-grade work up to the sixth. The teacher said that many of
these children who were unable to read or write when they came from the South
showed remarkable progress in a few months, and in less than a year were able to do
fourth-, fifth-, and sixth-grade work.
"One big girl of thirteen, when she arrived from the South," this teacher said,
"pretended to read with her book upside down, but in a little more than a year she
was doing sixth-grade work. One twelve-year-old boy from the South, unable to read
the primer or write his name, after about nine months of applied work just ate up
everything I gave him and during the following year read sixty library books."
A thirteen-year-old girl, just five days in the school, had come from Alabama, where
she had never attended school. "There wasn't room for me," she explained. She read
for the investigator on the tenth page of the primer, haltingly but with understanding.
The teacher was confident that she could put her through several grades next year.
She said further:
 These children who have been deprived in the South of their rights
educationally are very eager. At first they are timid, but they learn
very quickly. They're as smart as whips if they'd just get down to
business. Without question this is the kind of attention all the colored
children from the South need when they enter school in the North.
The plan has been successful and should be adopted throughout the
school system. One appreciates by comparison the injustice of putting
the fifteen-year-old newcomer from the South into second grade,
requiring of him only second-grade work over the nine months'
period.

Another school, 92 per cent Negro (Farren), has a special room for
children from the South. "Our dull children are almost without
exception those from the South who have never been to school,"
said the teacher. "Those children should not be classed as dull,
either, for they learn remarkably fast and often catch up to grade."
A teacher of the ungraded room in a school 38 per cent Negro
(Forrestville) said:
Practically all of the colored children are from the South, where they
have not been in school. Once they get started they learn very rapidly
and often catch up to the proper grade if they are not too old when
they start school. The older children in this room have good power of
concentration and consequently learn much in a short time. Take, for
example, a boy twelve years old who came here not two months ago
from the South. When he came he had no idea how to write his
name. A few days ago he wrote for me a fourth-grade eight-line
memory passage with but three mistakes in spelling. Now I call that
remarkable. I have taught in this school all my teaching years, and
they have been many, and have never seen any child equal this,
either white or black.

Capacity for advanced work.—Teachers in the seventh and eighth

grades usually found Negro children equal to the work, though in
some cases they felt that these children had been pushed out of the
lower grade because of crowded conditions before they were ready
for the more advanced work. An eighth-grade teacher gave the
following statement:
 When children get this far they have a good foundation and do their
work very well. One of my colored girls is the brightest child in school
—arithmetic is hard for her but she works at it. One of my colored
boys is seventeen years old. He came here from the South last fall to
live with an uncle and to get to a better school. His father wants him
to be a doctor and thought he wasn't getting along as well in the
South as he would in the North. When the boy came to me he said he
had been going to a college[40] in the South. I took him into the
eighth grade but saw he didn't have the fundamentals. On close
questioning he told me he had been in the seventh grade in that
college. Now he is doing excellent work for me. He has much broader
interests than the other children. He reads, reads, reads, all the time
and is well informed.

Other teachers believed that there was nothing to keep the Negro
children from making equal progress with the white, given similar
opportunities. "The progress of the colored children in Drake school
(30 per cent) cannot be compared with that of the white," said an
upper-grade teacher, "because the colored are all from the South
and have had the poorest opportunities. But comparing a Negro
child and a white child who have had the same advantages in school
and equal opportunities for observation and example in the home,
the Negro makes the same progress."
"I say that under the same conditions a Negro child will do as well
every time as a white," said the teacher of an ungraded room in a
school 38 per cent Negro (Forrestville). "Many do as well as the
white and live in very poor neglected homes. I think every person
who is not prejudiced must admit that the colored do fully as well in
school as the white."
An upper-grade teacher in the Felsenthal School (20 per cent) held a
similar point of view: "The colored are making wonderful strides.
They advance just as rapidly as the white, given equal opportunities.
But their background is so slight and so short in years that one
cannot fairly compare them. The southern colored child must be
studied individually to get his point of view in the school or he gets
nowhere in his work."
High-school work.—The principal of Wendell Phillips High School
prepared tables showing the numbers of white and Negro children
dropping out at the end of each school year. They show that the
largest number of Negro children dropped out during the first year,
and the largest number of white children during the first and second
years, the number of drop-outs being the same for both years. Some
children repeat the work so that all of them do not leave school.
One or two teachers in other schools stated concerning Negro
children that a "very limited number go beyond the first year." "They
cannot grasp the subject," said an English teacher; "they do not
understand as the white child does. They lack the mentality."
In the same school the Latin teacher held quite the opposite opinion.
"The colored children are in every way equal to the white children.
They are just as well equipped mentally and make similar progress.
My best student at present is a colored girl. Her choice of English
and her vocabulary and construction are far ahead of that of any
white student."
Several teachers and principals testified to the brilliancy of individual
Negro students who not infrequently had the highest standing in the
school. The principal of an elementary school (Crerar) who had
formerly had experience in a school largely Negro felt that the junior
high school would meet the needs of the Negro children to a large
degree:
More of them than the immigrant enter high school but do not stay to
finish. I suppose the parents insist upon some high-school training,
but it is necessary for the child to go to work before he finishes.
Another reason for the dropping out might be the teachers' lack of
interest in the child. In the high school you don't find the teachers
taking a keen interest in every individual child as you do in the
grades, and just what colored children need is a keen interest in
them. They do better work.

Academic v. other courses.—A preference of Negro children for

academic work was reported by principals and teachers at two high
schools. This may be due in part to the fact, testified to by many
teachers, that Negro children excel in languages and music and find
mathematics and sciences difficult. The usual implication was,
however, that Negro children took academic work because they
thought it gave them better social standing. A principal who said
that "Negroes want to know nothing about industrial training" and
that "the girls don't care for sewing and cooking," said on another
occasion that the majority of children in auto-mechanics, printing,
and household arts were Negroes. He also reported more Negro
than white children in the normal course preparing themselves to be
teachers, though this was the first year that this had been the case.
Comparative scholarship in elementary schools.—Negro children are
reported to be slower than the Jews, less responsive than the
Bohemians, and more ambitious than the Italians. A manual-training
and domestic-arts teacher thought Negroes did as good work as the
Jews, Bohemians, and white Americans whom he taught. A Latin
teacher said that the Negroes were studious and ambitious, and that
in every way she preferred them to the Jews.
Several teachers thought the Negroes were slow and lacked logic
and "sticking qualities." An upper-grade teacher explained the
slowness as partly due to the fact that they had been pushed out of
the crowded lower grades before they were ready for more
advanced work. A physics teacher who was convinced that Negro
children had no ambition said it was his policy to promote a Negro
child if the child had made the effort, because he appreciated that
the child had come "to the limit of his mental ability."
The principal who said that Negroes had no "sticking qualities" gave
a single instance of a boy who wanted to become a mechanical
engineer but gave up the course after five months, because he said
he did not care enough about the course to work at it for several
years. In endeavoring to prove that Negro children are not
successful in completing high-school work, this principal emphasized
the fact that in the 3-B class 20 per cent of the Negroes dropped out
as compared with 6 per cent of the whites. In actual numbers three
Negroes and two whites dropped out. He did not mention that in the
2-A class 12 per cent of the whites (sixteen children) as compared
with 3 per cent of the Negroes (three children) dropped out. In the
4-B grade 21 per cent of the whites (three children) and none of the
Negroes dropped out. The fact that 21 per cent of the whites
dropped out was explained by the principal to be due to the fact that
the white children wished to graduate from a high school wholly
white. However, only three children were involved.
Attendance and failures.—Table XVI shows the record for attendance
and failures in three groups of schools attended mainly by Negroes,
by children of immigrants and by white Americans. It will be noticed
that the best attendance records are found in Douglas and Farren
schools, both mainly attended by Negroes. The other schools,
attended mainly by Negroes, compare favorably with those attended
by whites.
The smallest percentage of failures is at Colman (92 per cent), while
the next to the largest percentage is also at a school attended
mainly by Negroes (Raymond, 93 per cent). This may be explained
to a certain extent by the fact that there is a higher economic class
of Negroes in the neighborhood of the Colman School. In the other
schools the percentage of failures compares very favorably with that
of whites.
TABLE XVI
Enrolment, Average Attendance, and Number of Failures in Twenty Schools

Average Percentage of Number of Percentage

School Enrolment
Attendance Attendance Failures of Failures
Attended mainly by
Negroes:
Colman, 92 per 964 709 73 13 1.8
cent
Doolittle, 85 per 1,784 1,282 72 77 6.0
cent
Douglas, 93 per 1,443 1,341 93 — —-
cent
Farren, 92 per 986 924 93 83 8.9
cent
 Forrestville, 38 1,493 1,085 73 130 12.0
per cent
Haven, 20 per 1,165 700 60 24 3.4
cent
McCosh, 15 per 1,280 1,017 79 — —-
cent
Moseley, 70 per 923 605 66 81 13.3
cent
Raymond, 93 per 1,532 1,299 85 200 15.4
cent
Webster, 30 per 805 654 81 — —-
cent
Attended mainly by
children of
immigrants:
Farragut 1,729 1,502 86 107 7.0
Goodrich 1,305 1,039 78 121 11.6
Kosciusko 1,134 775 68 33 4.2
Lawson 3,069 2,545 83 292 11.5
McCormick 1,432 1,266 88 204 16.1
Seward 1,058 708 67 43 5.9
Smyth 1,106 860 77 69 8.0
Swing 810 629 77 99 15.8
Attended mainly by
white Americans:
Fiske 1,535 1,272 83 45 3.5
Howland 2,161 1,809 84 100 5.0

C. CONTACTS IN RECREATION

In studying contacts between the races at places of recreation a

survey was made of the various recreational facilities maintained by
the Municipal Bureau of Parks, Playgrounds, and Bathing Beaches,
the South Park Commission, the West Chicago Park Commission, and
the Lincoln Park Commission. Recreational facilities maintained by
twelve park boards which control smaller areas in outlying parts of
the city were not included in the survey unless they were in or near
Negro areas. Visits were made by the Commission's investigator to
places in or bordering on the Negro areas at a time of day when the
use of the park would be greatest; the director or one of his
assistants was interviewed and observations were made as to the
relations between Negroes and whites.
The information thus gathered was supplemented by a conference
held by the Commission, at which representatives of the various park
boards discussed policies and experiences with reference to race
relations in the various recreation places under their charge.

I. CLASSIFICATION OF FACILITIES

Although there is no definite city-wide classification, the publicly

maintained recreation facilities of the city may, for the purpose of
this study, be grouped by types and defined as follows:
1. Playground.—A small tract of land, usually adjacent to public
schools, providing space for ball games, gymnastic and play
apparatus, and in most cases a small building used as an office and
storage place for apparatus.
2. Recreation center.—Including outdoor and indoor gymnasiums for
men, women, and children, a swimming-pool, and a little children's
playground out doors, and a field house providing an assembly room
and dance hall, clubrooms, shower baths, and often an infant-welfare
station and branch library.
3. Large park.—A large area with lawns, shrubbery, and general
recreation facilities, such as tennis, golf, baseball, and boating.
4. Bathing-beach.—Intended primarily for swimmers and usually
including no other recreation equipment. A dressing-house, showers,
and towel supply are provided with life guard and attendants on duty.
5. Swimming-pool.—In some instances a swimming-pool or
natatorium is maintained separately from a recreation center.

II. DISTRIBUTION OF FACILITIES IN RELATION TO NEGRO

AREAS
Of a total of 127 public places of recreation excluding the large
parks, thirty-seven are in or near Negro areas. Of the eighty-two
playgrounds, fourteen are in the Negro areas and nine are adjacent.
Of the twenty-nine recreation centers, none is located within the
Negro areas, but seven are adjacent.
Though these figures seem to indicate that the Negro areas are
fairly well supplied with recreation facilities, it should be borne in
mind that their use by the Negroes in their vicinity is by no means
free and undisputed. The reasons for this are shown in the next
section on "Use of Facilities," but the following summary of use will
aid in considering the distribution of recreation facilities in relation to
the Negro areas:
Total for In Negro Near Negro Number Used 10 Per Cent or
City Areas Areas More by Negroes
Playgrounds 82 14 9 13
Recreation 29 None 7 1
centers
Bathing- 8 3 2 1
beaches

The type of recreation facility most commonly found in the Negro

areas is the playground. The lack of recreation centers within the
Negro areas is conspicuous, as is also the fact that six of the seven
recreation centers accessible to Negroes are not used as much as 10
per cent by them. The playground is intended for the use of young
children and has practically nothing to attract older children and
adults, except sometimes a baseball or athletic field. Indoor facilities
are not a part of the equipment of a playground, so that the average
maintenance cost of a playground is not more than $2,000 to $5,000
a year.[41]
 RECREATION FACILITIES

The recreation center is the most unusual and notable feature of

Chicago's recreation system but one from which the Negro gets little
benefit. It is a complete community center, with both indoor and
outdoor facilities. It represents an investment of from $200,000 to
$800,000, according to the amount of ground, the location, and the
extent of its facilities. The yearly expenditure necessary to maintain
such a recreation center where older children and adults can hold
meetings, dances, and entertainments, and where there are
concerts, indoor games, swimming-pools, showers, etc., is shown by
the reports of the park boards to be from $30,000 to $50,000.
Though the argument that wholesome recreation makes for better
citizenship applies to Negroes as well as to whites, no recreation
center has been located within the Negro areas and only seven near
them.[42]
The director of Armour Square, a recreation center which is just
beyond the edge of the main Negro area, but which the Negroes do
not feel free to use for reasons discussed later, was asked what
places of recreation for adult Negroes existed in that neighborhood.
She instanced a social settlement that had been out of existence for
more than six years, an infant-welfare station and a commercial
amusement park known to be in bad repute.
Although in recent years the Negro population has been increasing
in density in the neighborhood directly east of Wentworth Avenue
along which Hardin, Armour, and Fuller recreation centers are
located, this has not increased the use of these centers by Negroes.
It has tended, rather, to increase the antagonism of the whites in the
vicinity to the use of the centers by Negroes. In this neighborhood
the hostility toward Negroes of whites, especially gangs of
hoodlums, is shown by the many attacks upon Negroes in this area
as discussed in the sections on the "Riot of 1919" and "Antecedent
Clashes."
Several representatives of the park boards strongly deprecated the
lack of recreation centers within the Negro area and said that such
facilities should be provided. The South Park representative
recommended the area east of Wentworth Avenue between Thirtieth
and Forty-seventh streets as one needing additional facilities. The
West Park representative said: "A complete all-year-round recreation
center for the colored people should be established at Ashland and
Lake streets. We need greater facilities, or equal facilities, for the
colored people. There isn't any place on the West Side that I know
of, but yet we have many of these complete recreation centers there
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