Creep: Fatigue Models of Composites and

Nanocomposites 1st Edition Leo Razdolsky install

download

https://ebookmeta.com/product/creep-fatigue-models-of-composites-
and-nanocomposites-1st-edition-leo-razdolsky/

Download more ebook from https://ebookmeta.com

We believe these products will be a great fit for you. Click
the link to download now, or visit ebookmeta.com
to discover even more!

Mathematical Methods and Models in Composites, 2nd

Edition Manti´`

https://ebookmeta.com/product/mathematical-methods-and-models-in-
composites-2nd-edition-manti/

Primary Mathematics 3A Hoerst

https://ebookmeta.com/product/primary-mathematics-3a-hoerst/

Encyclopedia of Materials: Composites: Composites.

Volume 1 1st Edition Dermot Brabazon (Editor)

https://ebookmeta.com/product/encyclopedia-of-materials-
composites-composites-volume-1-1st-edition-dermot-brabazon-
editor/

Nordic Equality and Anti-Discrimination Laws in the

Throes of Change 1st Edition Anne Hellum

https://ebookmeta.com/product/nordic-equality-and-anti-
discrimination-laws-in-the-throes-of-change-1st-edition-anne-
hellum/
Designing and Implementing Effective Professional
Learning 1st Edition John M. Murray

https://ebookmeta.com/product/designing-and-implementing-
effective-professional-learning-1st-edition-john-m-murray/

How Civil Wars Start: And How to Stop Them Barbara F.

Walter

https://ebookmeta.com/product/how-civil-wars-start-and-how-to-
stop-them-barbara-f-walter/

Last Hunt 1st Edition Adam Nicholls

https://ebookmeta.com/product/last-hunt-1st-edition-adam-
nicholls/

China and The Pursuit of Harmony in World Politics

Understanding Chinese International Relations Theory
1st Edition Adam Grydehøj

https://ebookmeta.com/product/china-and-the-pursuit-of-harmony-
in-world-politics-understanding-chinese-international-relations-
theory-1st-edition-adam-grydehoj/

Economics of Maritime Business 1st Edition Shuo Ma

https://ebookmeta.com/product/economics-of-maritime-business-1st-
edition-shuo-ma/
Cambridge International AS A Level Mathematics
Mechanics Question Workbook Willard

https://ebookmeta.com/product/cambridge-international-as-a-level-
mathematics-mechanics-question-workbook-willard/
Creep – Fatigue Models
of
Composites
and
Nanocomposites

LEO RAZDOLSKY
President, LR Structural Engineering, Inc.
Lincolnshire, Illinois, USA

p,
p,
A SCIENCE PUBLISHERS BOOK
A SCIENCE PUBLISHERS BOOK
First edition published 2023
by CRC Press
6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742
and by CRC Press
4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

© 2023 Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, LLC

Reasonable efforts have been made to publish reliable data and information, but the author and
publisher cannot assume responsibility for the validity of all materials or the consequences
of their use. The authors and publishers have attempted to trace the copyright holders of
all material reproduced in this publication and apologize to copyright holders if permission
to publish in this form has not been obtained. If any copyright material has not been
acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted,
reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means,
now known or hereafter invented, including photocopying, microfilming, and recording, or in
any information storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, access
www.copyright.com or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. For works that are not available on CCC please
contact [email protected]
Trademark notice: Product or corporate names may be trademarks or registered trademarks and
are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data (applied for)

ISBN: 978-1-032-21301-9 (hbk)

ISBN: 978-1-032-21302-6 (pbk)
ISBN: 978-1-003-26772-0 (ebk)
DOI: 10.1201/9781003267720

Typeset in Times New Roman

by Radiant Productions
Preface

Difficulties in determining the resources of engineering objects are directly

related to the complexity of the processes that occur in structural materials under
the existing operating conditions. Understanding the laws of these processes
will allow building a reliable mathematical model that will contain specific
parameters of the stress-strain state (SSS) that meet the working conditions of
the object, and which, ultimately, can become the theoretical basis for creating
methods and algorithms for assessing the resources of objects in accordance
with the individual histories of their use. Since the processes of damage
accumulation are closely related to the kinetics of the SSS, the accuracy of the
calculated estimates of the strength and service life of the structure elements
will depend on the degree of certainty with which the ratios determining the
mechanics of defective materials (MDM) represent danger zones of deformation
in structural elements at predetermined operating conditions. Viscoelastic
deformation parameters such as length and type of trajectory, type of stress
state, history of its change, and others significantly affect the rate of damage
accumulation. Thus, the main goal of research in the field of mechanics of a
deformable solid is rather not to clarify the various formulations necessary
for determining macroscopic deformations from a given loading history, but
to seek to understand the basic laws of phenomena that prepare the ultimate
state of the material and structure until it fails. This manuscript represents a
logical extension (in terms of applied methods of mathematics used in this
book) of the author’s previous book “Phenomenological Creep Models of
Composites and Nanocomposites” that was published by CRC (Taylor and
Francis) Company back in 2019. The subject matter of the manuscript offered
here is significantly different, since the phenomenon of fatigue-creep behavior
of composites and nanocomposites at high temperatures is considered. In
recent years, the use of composites and nanocomposites in various aerospace,
automotive, marine, and civil engineering applications has been constantly
increasing. Without exaggeration, it can be said that it is among the most
complex and urgent problems of the mechanics of deformable solids, since
phenomenological models in this case are much more complex, since they
have to take into account an additional number of factors arising during cyclic
iv Creep-Fatigue Models of Composites and Nanocomposites

loading. They are primarily associated with the development of cumulative

damage. Currently, the problem of structural strength under cyclic loading
is considered in a much broader sense. This is due to the development of
new industries with modern technology, such as aircraft, power engineering,
aviation and rocket technologies. The cyclic loading significantly reduces the
creep-fatigue resistance in the entire frequency range, and it became obvious
that such traditional characteristics of strength as endurance limit, static creep
limits and long-term static strength can no longer suffice the design criteria
for reliable performance. As a result, new directions appeared in the areas of
high-temperature strength-cyclical creep and long-term cyclic strength. These
circumstances led to the creation of new methods and means for determining
the resistance of composite and nanocomposite materials and a continuum
of damage development under cyclic loading; to the creation of appropriate
physical models. Particularly relevant is the intensification of creep by high-
frequency cyclic loading in composite materials, which usually occurs at
high temperatures. Most of the known studies in the field of cyclic creep are
experimental, and the direct use of the number of cycles to define the damage
model cannot escape the empirical relation that predicts fatigue life level.
This chapter presents new phenomenological cyclic creep-fatigue models
for describing the fatigue life and behavior of time-dependent composites and
nanocomposites. Certain criteria imposed on selecting the creep functions
have the potential to describe a wide range of material behaviors. Unlike
metals, composite materials are substantially heterogeneous and anisotropic.
This author presents the new phenomenological cyclic creep-fatigue models
describing the fatigue life and behavior of time-dependent composites and
nanocomposites. Damage does not accumulate in a localized form, and failure
does not always occur because of the propagation of a single macroscopic crack.
Due to the high specific stiffness and strength of composite materials, they
are often used for weight-critical structural applications. However, existing
imperfections in the methods for evaluating the strength of these materials often
leads to the fact that large safety factors are required in structural calculations
and designing. Although the technology of composite materials is developing
rapidly, their use in real structures is hampered by the lack of sufficient reliable
experimental fatigue data, which contains the main operational parameters.
The analysis of the stress state under variable high temperature loads presented
below requires taking into account the anisotropy of the averaged elastic
properties of the composite and nanocomposite materials. The book consists
of five chapters. The first chapter considers issues related to the formulation
of the creep-fatigue problem in general on the basis of the classical Volterra
integral equation of the second kind in the case of nonlinear creep. Of course,
this integral equation is supplemented with a periodic function reflecting the
fatigue process of a composite or nanocomposite. In addition, the kinematic
 Preface v

differential equation of the continuous damage function has been added. This
chapter also provides the required parameters that are obtained from simple
tensile (compression) experiments on a composite sample. In order to “test”
the proposed model, an example of a creep-fatigue calculation of a simple
mechanical element of the Kelvin – Voigt model is given.
The second chapter is devoted to the application of the cumulative
damage function to the solution of the problem of creep-fatigue of composites
or nanocomposites under the action of high temperatures. Without loss of
generality of the problem posed, some restrictions are introduced on the
function of accumulative damage and the integrand of the Voltaire equation of
the second kind. In the case of nanocomposites, the functions of cumulative
damage consist of two parts, reflecting the technological process of creation
(nucleation and growth of clusters) under the action of high temperatures.
In the third chapter, various phenomenological models of the maximum
lifetime of a composite or nanocomposite are considered. At the same
time, it is noted that some parameters of the system (such as the volumetric
percentage of any filler or the temperature dependence of the viscosity
coefficient significantly affect the life of the composite. In the fourth chapter
parameters and variables (dependent and independent) in dimensionless form
are used to obtain fatigue curves analytically. It should also be noted that,
in addition to the usual dependence (a0 – Nf), it is also possible to obtain
other dependences of the cumulative damageability (ω - Nf). In this case, the
limiting state of the system is f (ω) = 1 @ ϴ = ϴmax or the usual dependence
of the fatigue curve (a0 – Nf) of composites and nanocomposites. Thus, the
creep-fatigue problem is solved in a deterministic formulation. In structural
engineering practice an empirical so-called safety factor is usually introduced.
Chapter five contains various methods of the applied theory of probability
applied to evaluate this coefficient qualitatively and quantitatively. It appears
inevitable that the structural engineering community, as well as many other
engineering communities that are ultimately responsible for life safety issues,
will eventually incorporate probabilistic analyses methods to some degree.
Probabilistic analyses methods, unlike traditional deterministic methods,
provide the means to quantify the inherent risk of a structural design and
to quantify the sensitivities of the most important parts of the design in the
overall reliability of the structural system as a whole. The degree to which
these methods are successfully applied depends on addressing the issues
and concerns discussed in this book. The importance of a probabilistic
approach in the issues of the cyclic creep-fatigue process and the long-
term behavior with a specified reliability is addressed in NASA reports and
other technical publications. One of the prime issues for these structural
components is assured long-term behavior with a specified fatigue life. The
present book addresses issues pertaining to the probabilistic creep-fatigue
vi Creep-Fatigue Models of Composites and Nanocomposites

life of composites under combined thermal/mechanical cyclic loads through

some typical examples. The focus of ongoing research has been to develop
advanced integrated computational methods and related computer codes to
perform a complete probabilistic assessment of composite structures. These
methods account for uncertainties in all the constituent properties, fabrication
process variables, and loads to predict probabilistic micro, ply, laminate,
and structural responses. The results are supported by step-by-step practical
design examples and should be useful for practicing structural engineers, code
developers as well as researchers and university faculty personnel.
It should also be noted that some restrictions are introduced in this book
(not violating the generality of the issue under consideration), namely:
1. All differential and integral equations, as well as independent, dependent
variables and their parameters are presented in a dimensionless form. At
the same time, in accordance with the theory of dimensionlessness, it is
possible to achieve the minimum number of parameters and independent
and dependent variables.
2. Solutions of all integral and differential equations describing the behavior
of composite and nanocomposite materials under the action of static,
dynamic and high temperature loads are presented in the form of numerical
integration of the corresponding equations. This computer output is
presented in three parts: systems of differential equations, explicit inputs,
and implicit inputs. Due to the fact that these output datasets differ only
in the value of one load parameter, in this case the complete solution (all
three parts) is given for the smallest load value, and in all other cases the
system of differential equations and the desired expression of a particular
function (for example, the dependence of the oscillation amplitude on the
variable temperature) are given.
3. The proof of the mathematical method for solving integra-differential
equations is purely mathematical (Existence and Uniqueness Theorem
for Volterra Integral Equations of second rang) and not included in this
book. However, the simple result of this theorem in this particular case
is as follows: in order to substitute the Volterra Integral Equations with
corresponding Differential Equation the Initial Condition must be added.
Contents

Preface iii
Nomenclature x

1. Introduction and Assumptions 1

1.1 Introduction 1
1.2 Fatigue curve and endurance limit 3
1.3 Creep-fatigue process under cyclically changing strain 6
1.4 The concept of effective stresses 6
1.5 Use of scalar internal variable to quantify damage 7
1.6 The scalar measure of damage 8
1.7 Strength as endurance limit of composites and 9
nanocomposites
1.8 Continuous damage accumulation model 10
1.9 Damage accumulation function for composites 13
1.10 Paris’ Law and Miner’s Rule 14
1.11 The Bergman-Milton theory 16
1.12 Failure criteria 17
1.13 The Bergman-Milton theory 20
1.14 Failure criteria 23
1.15 Principle of stress equivalence 25
1.16 Standard linear model 28
1.16aStandard Linear Model with different 35
viscosity – temperature relationships
1.17 Temperature effect on viscosity 36
1.18 Viscosity of dispersed systems 38

2. Cumulative Damage Model (CDM) of Cyclic Creep-Fatigue 50

Process
2.1 Introduction 50
2.2 The concept of effective stress 53
2.3 Classification of composite materials 55
viii Creep-Fatigue Models of Composites and Nanocomposites

2.4 Objectives of this research 58

2.5 Cyclic loading types 64
2.6 Creep-fatigue constitutive model with cumulative 66
damage law
2.7 Proposed constitutive model of the creep-fatigue process 73
2.8 Analytical formulas for S – N fatigue curves 77

3. Phenomenological Creep-Fatigue Models 97

3.1 Introduction 97
3.2 Effect of temperature – time relationships on 100
creep-fatigue behavior of composites
3.3 Effect of chemical energy on cyclic creep-fatigue 109
process
3.3.1 Chemical kinetic effect on nanocomposites 109
creep-fatigue process
3.4 Nanocomposite material under cyclic creep-fatigue 112
conditions
3.5 Viscosity change effect at high temperature 114
3.6 Analytical expression of crystallization function f3 118
3.7 Effect of increase in frequency ‘p’ 124

4. Peculiarities of Phenomenological Models of Nanocomposites 130

4.1 Introduction 130
4.2 The concept of effective stresses 133
4.3 Creep-fatigue behavior of nanocomposites 134
4.4 Defining the damaged medium mechanics 137

5. Probabilistic Approach to Creep-Fatigue Models 179

5.1 Introduction 179
5.2 Creep-fatigue process under periodic loads 180
5.3 Continuum damage mechanics and durability of composites 180
5.4 Damage function ω and decrease of cross section area 181
5.5 “Forward” and “Reversed” probabilistic problem 183
5.6 Phenomenological models of creep-fatigue of composites 184
5.7 Failure criteria 186
5.7.1 Maximum stress theory 186
5.7.2 Haskin’s failure theory 186
5.8 Specifics of constitutive equation of creep-fatigue of 187
composites
5.9 Effect of high temperature on fatigue curves (S – Nf) 189
5.10 Examples 190
 Contents ix

5.11 Probabilistic approach for creep-fatigue model of 200

composites
5.11.1 Main assumptions 200
5.11.2 Continuum damage and temperature effects on the 202
probabilistic approach
Conclusions 220
Index 223
Nomenclature

A constant
Ai unknown parameter
[a;b] interval
E modulus of elasticity
C; C1 specific creep
D diffusion rate, D is a constant,
Q is an activation energy
J compliance function (often also called the creep function)
K stress memory function
Sd esign load
R structural resistance
t and t1 time t > t1
T temperature
t
retardation time
tr
relaxation time
Ș viscosity
Į material property parameter
İ strain
ı stress
p frequency
Nf number of cycles to failure
Ȧ FRQWLQXXPGDPDJHIXQFWLRQ
ij YROXPHWULFFRQFHQWUDWLRQ
F force
Ea
T [T  T* ]  Dimensionless Temperature
RT*2

RT* E
E1 N m 7* %DVH7HPSHUDWXUH>0 .@
Ea Ef
h2
Time: t W [sec]
a
 Nomenclature xi

RT*2
Temperature: = T è θ+ T
+* T [K], where T* = 600oK is the base line
E *
temperature
k The thermal conductivity that has the dimensions W/m*K or
J/m*s*K
T Temperature
d thickness in the direction of heat flow.
ρ is the air density
cp is its specific heat capacity at constant pressure
K is the number of collisions per second resulting in a reaction At is
the total number of collisions
E activation energy
R is the ideal gas constant
P Heat loss due to thermal radiation
e Emissivity factor
σ Stefan-Boltzmann constant (σ=5.6703(10-8) watt/m2K4);
To ambient temperature
cp average specific heat at constant pressure
t time

v(u; v; w) velocity vector
D Diffusion coefficient [m2/sec]
p is the pressure
ν kinematic viscosity; ν = μ/ρ
θ dimensionless temperature
“τ” dimensionless time
“h” height of the specimen [m]
“a” thermal diffusivity [m2/sec]
h2
Time: t = τ [sec]
a RT*2
Temperature:
= T θ + T* [K], where T* = 600oK is the base line
temperature E

Coordinates: x = x / h and z = z / h - “x” and “z” – dimensionless coordinates.

ν ν
Velocities: u = u [m/sec] and w = w [m/sec] - horizontal and vertical
h h
components of velocity respectively ; ν – kinematic viscosity [m2/sec]; “u”
and “w” – dimensionless velocities.
gh 3
Fr = Froude number
νa
g is gravitational acceleration
Le = a/D = Sc/Pr - The Lewis number
 xii Creep-Fatigue Models of Composites and Nanocomposites

Sc = ν/D - The Schmidt number

RT
âβ = * - Dimensionless parameter
E
cp RT*2
ãγ = - Dimensionless parameter
QE
eσ K v ( β T* )3 h
P= - Thermal radiation dimensionless coefficient
λ
σ = 5.67(10–8) [watt/m2K4] –Stefan-Boltzmann constant

Kv = Ao h/V – Dimensionless opening factor

E E
=δ
ä ( 2
)Qz(exp(− )) - Frank-Kamenetskii’s parameter
RT* RT*
ν
W = W - Vertical component of composite’s velocity
h
ν
U = U - Horizontal component of composite’s velocity
h
b = L/h, “L” and “h”- Length (width) and height of component accordingly
W; U – dimensionless velocities
Coordinates: x = x / h and z = z / h - “x” and “z” – dimensionless coordinates.
Chapter 1
Introduction and
Assumptions

1.1 Introduction
Without exaggeration, one might say that the cyclic creep-fatigue problem is
among the most complex and urgent problems of the mechanics of solid body
deformation. It is primarily associated with the development of a creep-fatigue
damage mechanism in order to assess the structural stability of composite and
nanocomposite structural elements and systems. Today in modern industry,
especially where it is critically important to ensure the strength and stiffness
of structural elements with minimal weight, the designers of airplanes and
airplane engines are using composite and nanocomposite materials [1].
However, existing imperfections in the methods for evaluating the strength
of these materials often lead to the fact that large safety factors are required
in the structural calculation methods. Therefore, to ensure the required safety
of the composite materials’ products, the structural engineer designers are
using larger margins of safety, which, in turns, reduces their efficiency. These
considerations fully apply to methods for evaluating the fatigue strength of
composites and nanocomposites. Under high temperature cyclic loading of
composite materials, the accumulation of damage can manifest itself in a change
in the integral properties of such materials. Various methods of non-destructive
testing can detect changes in the elastic modulus, electrical conductivity and
damping coefficient. Although the technology of new composite materials
is developing rapidly, their use in real structures is being hampered because
of the insufficient fatigue data required for design. The stress analysis under
variable thermal loads requires taking into account the anisotropy of the
so-called averaged properties of the composite. Unlike metals, composite
materials are substantially heterogeneous and anisotropic. Damage does not
accumulate in a localized form, and failure does not always occur because of
2 Creep-Fatigue Models of Composites and Nanocomposites

the propagation of a single macroscopic crack. Micro structural mechanisms

of damage accumulation, including fiber fracture and matrix fracture, fiber-
matrix splitting and delaminating sometimes occur independently. At low
levels of cyclic loading or in the initial part of the fatigue life, most types
of composites accumulate and diffuse the damage. These damages propagate
throughout the tension zone, and gradually reduce the strength and stiffness
of the composite structure. In the later stages of the fatigue life, the amount of
accumulated damage in a certain region of the composite element can be quite
large. This leads to the fact that the residual strength of the composite of that
region drops below the allowable stress level.
Well before a microstructural understanding of fatigue processes was
developed, engineers had developed empirical means of quantifying the
fatigue process and designs against it. Perhaps the most important concept
is the S-N diagram, where at any given constant cyclic stress amplitude, S,
applied to a specimen, the number of loading cycles Nf until the specimen
fails is counted. Millions of cycles might be required to cause failure at
lower loading levels, so the abscissas of such a diagram are usually plotted
logarithmically. Statistical variability is troublesome in fatigue testing,
because it is necessary to measure the lifetimes of perhaps twenty specimens
at each of the ten or so load levels to construct the S−N curve with statistical
confidence. It is generally impossible to apply a cyclic load to a specimen at
more than approximately 10 Hz (inertia in components of the testing machine
and heating of the specimen often become problematic at higher speeds) and
at that speed, it takes 11.6 days to reach 107 cycles of loading. Obtaining a
full S−N curve is obviously a tedious and expensive procedure. For instance,
a very substantial amount of testing is required to obtain a S−N curve for
the simple case of fully reversed loading, and it will usually be impractical
to determine whole families of curves for every combination of mean and
alternating stress (see Fig. 1.1) [2].
There are a number of strata gems for assessing the residual stress of
fatigue, one common one being the Goodman diagram (see Fig. 1.2) [3]. Here
is the graph with mean stress as the abscissa and alternating stress as the
ordinate, and a straight “lifeline” is constructed from σm on the σalt axis to the
ultimate tensile stress σf on the σm axis. Then for any given mean stress, the
fatigue life limit (the value of alternating stress at which fatigue failure never
occurs) can be read directly as the ordinate of the lifeline at that value of σm.
Alternatively, if the design application dictates a given ratio of σm to σalt, a line
is drawn from the origin with a slope equal to that ratio. Its intersection with
the “life line” then gives the effective endurance limit for that combination of
σf and σm.
 Introduction and Assumptions 3

0.45
0.4
0.45 smax
0.350.4 smax
0.35
0.3
0.250.3
1
0.25
0.2
0.150.2 sm
1
0.15 smin sm
0.1
smin
0.050.1
0.05
–0.5
–0.5 0.5 1 1.5 2 2.5 3 3.5 4 4.5
–0.05 0.5 1 1.5 2 2.5 3 3.5 4 4.5
–0.05
Figure 1.1. Simultaneous mean and cyclic loading.

smax Cyclic (alternate) stress

0.4 smax Cyclic (alternate) stress
sm
0.4
.35 sm
.35
0.3
0.3
.25
.25Cyclic (alternate) stress NOT safe operating zone
NOT safe operating zone

1.2
0.2 Cyclic (alternate) stress

1.2
0.2
.15 Safe operating zone
.15 Safe operating zone
0.1
0.1
.05.05
Mean
Meanstress
stress sf
sf
ssmm
0.50.5 11 1.5
1.5 22 2.5
2.5 33 3.5
3.5 4
4
Figure 1.2. The Goodman diagram.

smax
smax
The maximum completely reversing cyclic stress that a material can
withstand for an indefinite (or infinite) number of stress reversals is known as
the fatigue strength or endurance strength (Se).

1.2 Fatigue curve and endurance limit

1.3
1.3
The ability of the material to withstand the effects of variable loads
sr sr NN
characterizes the endurance limit, the value of which is usually determined
0 0 NN0 0
experimentally. The purpose of the test is to determine the number of cycles
at which each sample fails at a given stress. The first sample is loaded with a

s
8s
8
 0.1

.05
Mean stress sf

0.5 1 1.5 2 2.5 3 3.5 4

4 Creep-Fatigue Models of Composites and Nanocomposites

smax

sr N
0 N0
Figure 1.3. Fatigue curve.

symmetric cycle with stress amplitude that brings it to failure. The following
samples are tested with reducing loads. The compiled test results are the main
source of so-called S–N diagram, which is the fatigue (endurance) curve (see
Fig. 1.3) [4]. For many engineering materials, a horizontal asymptote is a
s
characteristic
8 feature of the fatigue curve. This means that at a certain value
of the amplitude of the cycle stress, the sample can withstand a theoretically
7
infinitely large number of loading cycles. This stress value is the endurance
limit.6
For
5
metals, endurance limits are the values at which fatigue failure
does not occur after 107 cycles. This number of cycles is called basic and is
4
designated N0. For non-ferrous metals, hardened steels, and many composites,
the endurance
3 limit is 20 million cycles, on the fatigue curve. Thermal fatigue
is defined as the cracking of a material primarily caused by repeated heating700°C
2
and cooling which induces cyclic internal thermal stresses. In addition, stresses
800°C
caused
1 by external mechanical loads may also contribute to thermal-fatigue
1100°C 1000°C 900°C tc hr
failure. Stress, temperature, and time conditions may be sufficient to cause
both
–0.5 instantaneous
0.5 1 1.5 viscoelastic
2 2.5 3strain
3.5 and
4 time-dependent
4.5 5 5.5 6 creep 6.5 7 during
7.5 each
8 8.5 9 9.5
cycle.
–1 Therefore, a general method for calculating the thermal-fatigue life
must apply to a cycle in which various types of composite material behaviors
are present in arbitrary combinations. It should be noted that the important
factor in determining the life span of a composite is the temperature—time
dependence of the same type as expressed in creep data, rather than a time-
independent effect such as plastic deformation. There are some general
assumptions made in this manuscript and they are as follows:
(1) Thermal fatigue is considered as a primary creep-fatigue phenomenon
rather than a combination of many factors influencing fatigue failure.
(2) Cyclic creep-fatigue is assumed to have a failure mechanism in any type
of loading combination (including “stop – and – go”), rather than being
limited to cycles that contain tensile mean stress only.
 Introduction and Assumptions 5

(3) The proposed analytical method applies directly to the complete service
cycles of load and temperature, rather than being limited to a specific
idealized laboratory test such as the isothermal strain cycling method.
(4) The well-known cumulative damage continuum models for the creep-
fatigue process have some modifications under non-steady loading
conditions.
(5) The proposed phenomenological model predicts in quantitative and
qualitative terms the effect on the thermal creep-fatigue life of various
cycle parameters such as the maximum amplitude of strain (stress);
frequency of accelerations; range of temperatures and rates of heating
and cooling.
This chapter is devoted to the methods of solving creep-fatigue problems
analytically by using the simplest mechanical system (a Standard Linear
Model (SLM) or Vought – Kelvin model). Also, the cumulative damage
model and the mechanism of application of this model is proposed here. The
Vought – Kelvin model was chosen here with the main purpose of confirming
the correctness of the proposed assumptions in the formulation of the fatigue
problem of composites and nanocomposites under cyclic loading, as well as
obtaining a qualitative confirmation of the results. In the subsequent chapters,
each feature of cyclic fatigue-creep of composites and nanocomposites under
high-temperature loading conditions is considered. Analyses of fatigue-creep
of composites and nanocomposites problems are presented with the numerical
solutions (in dimensionless form). A uniaxial phenomenological model
proposed here takes into consideration an inelastic deformation and failure
criteria of composites and nanocomposites under the combined action of
static and cyclic loads. The Gibbs energy [5,6] and adjacent thermodynamic
[7,8] approaches for describing inelastic rheological deformation and failure
of composites and nanocomposites under unsteady loading provides good
results, and the expediency of their use in computational practice. The
purpose of this work is to generalize the approach [9] that describes the fatigue
phenomena that occur in composite materials under the combined action of
static σst and cyclic loads with an amplitude value of the cyclic component σa.
Consideration is also given to the so-called multi-cycle loading at a frequency
f > 10 Hz and an amplitude coefficient a0 = σa/σst not exceeding a certain
nondimensional critical value acr, having an order of 0.95. In this case, cyclic
loading leads to two main effects: (1) acceleration (or even initiation) of the
creep process at a given static stress σ0; (2) a decrease in the accumulated
inelastic deformation at the time of failure compared with a similar value
under pure static loading. As a rule, these phenomena cannot be described
either within the framework of just ordinary classical approaches, or from the
6 Creep-Fatigue Models of Composites and Nanocomposites

standpoint of phenomenological creep only:7 or from the standpoint of fatigue

in a symmetric (or asymmetric) cycle.

1.3 Creep-fatigue process under cyclically changing

strain
Dr. R. Talreja [10] gave the most general definition of the creep-fatigue
process of composites: “A reliable and cost-effective fatigue life prediction
methodology for composite structures requires a physically based modeling
of fatigue damage evolution. An undesirable alternative is an empirical
approach. A major obstacle to developing mechanistic models for composites
is the complexity of the fatigue damage mechanisms, both in their geometry
and the details of the evolution process. Overcoming this obstacle requires
insightful simplification that allows the use of well-developed mechanical
modeling tools without compromising the essential physical nature of the
fatigue process”. Although the fatigue behavior of composite materials is
significantly different from the behavior of metals, many models have been
developed based on the well-known S – N curves. These models make up
the first class of so-called “fatigue strength models”. This approach requires
large experimental studies and does not take into account the real mechanisms
of damage, such as matrix damage and fiber breaks [11]. The second class
includes phenomenological models for multi-cycle fatigue. These models
offer an evolutionary law that describes the gradual degradation of the strength
or stiffness of a composite sample based on macroscopic properties. Recently,
models have been developed based on the concepts of continuum damage
mechanics. The development of damage is determined by evolutionary kinetic
equations that reflect the irreversible nature of damage [12]. Continuum damage
models introduce scalar, vector, or tensor damage parameters that describe the
degradation of the entire composite material or structural components. These
models are based on physical modeling of the main damage mechanisms that
lead to macroscopically noticeable degradation of mechanical properties [13].
The main result of all fatigue models is the prediction of fatigue life, and each
of these three categories uses its own criterion to determine the failure process
and, as a consequence, the fatigue life of the composite material.

1.4 The concept of effective stresses

In the continuum damage approach the composite material is considered as a
homogeneous anisotropic elastic medium. Assuming that the model considers
the small strains that take place in the structural element or structural system
 Introduction and Assumptions 7

as a whole, the elastic strain energy function is quadratic and there is a linear
relationship between the stress and strain tensors σ and ε. In contrast to the
mechanics of brittle fracture, which considers the process of equilibrium or the
growth of macrocracks, the continuum mechanics of damage uses continuous
internal variables that are associated with the density of microdefects. The
proposed model uses the concept of so called effective stress that integrally
reflects various types of damage at the micro-scale level (such as the formation
and growth of micro cracks of the matrix, fiber breaks, delaminating and other
microscopic defects) [14].

1.5 Use of scalar internal variable to quantify damage

The damage parameter ω is associated with a decrease in the effective area of
any cross section at a given point on the body and determined by the following
formula:
A − ∆A
ω= (1.1)
A
In Equation (1.1) “A” is the nominal, undamaged cross-sectional area,
“∆A” is the total cross-sectional area of all defects on this site. By definition,
the theoretical value of ω should be in the range 0 ≤ ω ≤ 1. The effective stress
tensor in case of isotropic damage is introduced as follows:
σij
 =
σ (1.2)
1−ω
For a more detailed reflection of the mechanism of fatigue damage of
composites, tensor measures of damage are used. This allows one to take into
account the directed, anisotropic nature of the accumulation of fatigue defects.
To identify the parameters of the models of anisotropic damage, a significant
number of experiments are required with complex test programs that make it
possible to identify the directional nature of fatigue damage.
In considering the geometric interpretation of the anisotropic damage
function, the second rank tensor is introduced. For physical reasons, it is
symmetrical. Generalizing the dependence Equation (1.2) to the case of
the damage tensor of the second rank, the symmetric shape of the effective
stresses σij is obtained as follows:
[σij ]n1 [σij ]n 2
=σ ij A1 + A2 (1.3)
[1 − ω]m1 [1 − ω]m2
8 Creep-Fatigue Models of Composites and Nanocomposites

1.6 The scalar measure of damage

The introduction of a scalar measure of damage also determines the choice
of a mathematical model for describing the accumulation of damage. Under
conditions of a high-temperature stress state, the rate of damage accumulation
should depend on the joint variants of stress tensors and tensors that determined
experimentally characterize the mechanical properties of the composite. The
desire to reflect the characteristics of the cycles for each component of the
stress tensor leads to a complication of theoretical models. A realistic approach
to such situations is to introduce the amplitude of the instantaneous stress a0
and frequency p. The scalar measure of damage ω is considered as a function
that depends on the maximum stress value per cycle; the number of loading
cycles N; the cycle parameter R; temperature T; material properties and other
variables affecting composite material fatigue. In the proposed model it has
been assumed that the damage accumulation rate depends on current stress
level and all required material constants are determined from relatively simple
experiments at fixed values of the cycle parameter and temperature. The rate
of damage level ω is presented as:
dω
= f (σ, ω, n i , mi ) i = 1; 2 (1.4)
dθ
In accordance with the concept of continuum damage mechanics, function
f(ω) can be obtained theoretically based on the number of loading cycles or by
analyzing changes in the modulus of elasticity with the temperature rise [15].
However, for the practical use of the theory, it is more preferable to identify
the functional dependence (1.4) based on the S – N Weller fatigue curves’
results [16].
The geometric interpretation of the damage parameter at ω = 1 corresponds
to the case when the cross section of the composite material is completely
filled with macro-cracks. In practice, the material becomes unstable and
collapses when the damage reaches a certain critical value, less than unity.
Due to significant nonlinearity of the dependence of the damage parameter
on the number of cycles, at the stage preceding the failure, the growth rate
increases and tends to infinity. Therefore, the interval of change of the damage
parameter is close to unity, in the range 0.9 < ω < 1 corresponding to an
insignificant change in the ratio of the number of cycles to fatigue life Nf.
The basic experiments to determine the fatigue characteristics of composite
materials are experiments on cyclic loading under uniaxial stress conditions.
Due to the significant scatter of experimental data, it is necessary to test a
large number of samples at various stress levels in order to construct S – N
curves. The proposed deterministic and probabilistic models are presented by
the examples given below. Moreover, the statistical datasets that are required
 Introduction and Assumptions 9

in order to solve the corresponding probabilistic problems are based on the

solutions of the corresponding problems in a deterministic formulation.
The developed model allows predicting fatigue life taking into account
the influence of the orientation of the main reinforcement directions relative
to the planes of elastic symmetry of the material. The models presented here
are using the dimensionless parameters and variables that in turn require the
minimum number of material parameters and the minimum necessary set of
experiments respectfully.
Until the mid-40s of the last century, the solution to problems of strength
of materials and structures under cyclic loading was mainly reduced to the
assessment of fatigue resistance formulated first by V. Woehler [17].

1.7 Strength as endurance limit of composites and

nanocomposites
The problem of strength under cyclic loading is considered more extensively
in present times. This is due to the development of new industries in modern
technology, primarily stationary and transport power engineering. It is a
very well known fact that in most parts of power machines operating at high
temperatures, cyclic loads varying in a wide range of frequencies and amplitudes
are imposed on static loads of various kinds. If the structures operate in a
nonstationary time mode, then an uneven temperature field with large gradients
arises in its elements, causing considerable cyclic temperature stresses. It was
found that cyclic loading significantly reduces the creep resistance in the entire
frequency range, and the impact of low frequency loads (tenths and hundredths
of a hertz) equal to and exceeding the yield strength of the material reduces
fatigue resistance. It became obvious that such traditional characteristics of
strength as endurance limit, static creep limits and long-term static strength
can no longer be considered sufficient criteria for reliable performance.
As a result, new directions appeared in the section of high-temperature
strength—cyclical creep and long-term cyclic strength. The issues of structural
and surface stability of materials, thermal fatigue and stability of structures are
also very important. These circumstances led to the creation of new methods
and means of determining the resistance of structural materials to deformation
and damage development process under cyclic loading, to the creation of
appropriate theories and physical models. The efforts of many scientists have
already achieved significant progress in the field of theoretical interpretations
and the quantitative description of the phenomena of cyclic creep and long-
term cyclic strength, and in the field of engineering applications of theoretical
results.
10 Creep-Fatigue Models of Composites and Nanocomposites

Particularly relevant is the intensification of creep by high-frequency

cyclic loading applied to the composite and nanocomposite materials, which
usually occurs at high temperatures, i.e., in the conditions most characteristic of
the operation of many critical parts of modern power machines. These include
working blades and disks of gas turbines and compressors, gas pipelines,
linings of combustion chambers, fasteners and other parts of structural
elements for which the mode of high-temperature multi-cycle loading is one of
the main ones. It is also indicative that in composites under high temperatures,
cyclic creep-fatigue develops not only at small values of the amplitude of the
cyclic component, but also at values exceeding the static load in some cases.
The result of the intensification of the creep-fatigue process by a high-
frequency load is, as a rule, premature violation of the structural integrity of
the part and its failure is a result of an excessively developed residual stresses
and failure. In particular, the nature of the fracture of the working blades of
gas turbines does not carry purely static or fatigue damage, but indicates
damage progression due to cyclic creep-fatigue process.
The problem of cyclical creep-fatigue and durability under multi-cycle
loading attracted the attention of researchers around the same period when
it became necessary to consider taking cyclic loads into account when
evaluating strength. These tasks were set by the needs of the turbine industry,
and subsequently arose in the nuclear power industry, chemical engineering,
aviation and rocket technology. Most of the known studies in the field of
cyclic creep-fatigue are experimental, and the first of them are related to the
study of the creep of lead, copper, aluminum, and other pure metals. Recently,
however, the more general questions of the cyclic creep-fatigue process that
develop under the joint effects of static and cyclic loads have become the
focus of attention of researchers [18,19,20]. Due to the high specific stiffness
and strength of the fiber-reinforced composite materials, they are used for
weight-critical structural applications. However, existing imperfections
in the analytical methods for evaluating the strength of composites and
nanocomposites often lead to the requirement of large safety factors in the
structural calculation methods. Therefore, products of composite materials are
often designed with an excessive margin of safety, which reduces the efficiency
of their use. These considerations fully apply to methods for evaluating the
fatigue strength of composite materials.

1.8 Continuous damage accumulation model

A rational choice of the damage accumulation model allows one to lead to a
more efficient use of these composite materials [21]. Under cyclic loading
of composite materials, the accumulation of damage can manifest itself in a
 Introduction and Assumptions 11

change in the integral properties that are included in corresponding constitutive

equations. By various non-destructive testing methods the change in the elastic
modulus, electrical conductivity, and damping coefficient can be obtained.
Damage can occur by wave effects and recognized by acoustic emission.
Although the technology of composite materials is developing rapidly, their
use in real structures is hampered by the lack of access fatigue data, which
contains the main operational parameters of structural systems. Unlike metals,
composite materials are substantially heterogeneous and anisotropic. Damage
does not accumulate in a localized form, and failure does not occur because of
the propagation of a single macroscopic crack. Microstructural mechanisms
of damage accumulation, including fiber fracture and matrix destruction,
fiber-matrix splitting and separation, sometimes occur independently. At low
levels of cyclic loading or in the initial condition, most types of composites
accumulate and diffuse damage. These damages gradually reduce the strength
and stiffness of the composite throughout the stress zone. In the later stages of
the fatigue life, the amount of accumulated damage in a certain region of the
composite can be quite large. This leads to the fact that the residual composite
strength of that region drops to the maximum stress level in cyclic loading and
as a result, the failure occurs.
Although the fatigue behavior of fiber-reinforced composite materials is
significantly different from the behavior of metals, many models have been
developed based on the well-known S – N curves. These models make up the
first class of so-called “fatigue strength models”. Such an approach requires
large experimental studies and does not take into account real damage
mechanisms, such as matrix damage and fiber breaks [22]. The second class
includes phenomenological models for multi-cycle fatigue. These models
offer an evolutionary law that describes the gradual degradation of the
strength or stiffness of a composite sample based on macroscopic properties.
Recently, new developing models based on the concepts of continuum fracture
mechanics are used [23]. In models of this type, damage is quantitatively
described by some internal parameters of the material. The development
of damage is determined by evolutionary kinetic equations that reflect the
irreversible nature of damage. Continuous damage models introduce scalar,
vector, or tensor damage parameters that describe the degradation of the
entire composite material or for structural components. These models are
based on physical modeling of the main damage mechanisms that lead to
macroscopically noticeable degradation of mechanical properties [24]. The
main result of all fatigue models is the prediction of fatigue life, and each of
these three categories uses its own criterion to determine the fracture condition
and, as a consequence, the fatigue life of the composite material. In this paper,
we consider a continuum model of the accumulation of fatigue damage, based
12 Creep-Fatigue Models of Composites and Nanocomposites

on the assumption that the growth rate of the damage parameter depends
on the maximum value of the specific energy of elastic deformation of the
composite.
In the continuum approach to the analysis of the stress state and fatigue
of products made of composites, the material is considered as a homogeneous
anisotropic elastic medium [25]. When constructing the model, it is assumed
that small elastic strains take place. The elastic strain energy function is
quadratic and there is a linear relationship between the stress and strain tensors
σ and ε respectively. In contrast to the mechanics of brittle fracture, which
considers the process of an equilibrium state or the growth of macrocracks, the
continuum mechanics of damage uses continuous internal variables that are
associated with the density of microdefects. The proposed model is based on
the concept of effective stress and integrally reflects various types of damage
at the micro-scale level, such as the formation and growth of micro-cracks
of the matrix, fiber breaks, delamination and other microscopic defects [26].
The complexity and variety of mechanisms for the accumulation of fatigue
damage and the degradation of the strength properties of the composite make
it justifiable. Damage parameter ω is associated with a decrease in the effective
area of any cross section near a given point on the body and is determined by
the following formula:
A−A
ω= (1.5)
A
–
where A is the nominal, undamaged cross-sectional area, A is the total cross-
sectional area of all defects on this site. By definition, the theoretical value of
ω should be in the range 0 ≤ ω ≤ 1. The effective stress tensor in the case of
isotropic damage is introduced as follows:
σ
 ij = ij
σ (1.6)
1−ω
For a more detailed reflection of the mechanism of fatigue damage of
composites, tensor measures of damage are used. This allows one to take
into account the directed, anisotropic nature of the accumulation of fatigue
damage.
To identify the parameters of models of anisotropic damage, a significant
number of experiments are required with complex test programs that make it
possible to identify the directional nature of fatigue damage. When considering
the geometric interpretation of anisotropic damage, a damage rank tensor
of the second rank is introduced. For physical reasons, it is symmetrical.
Generalizations of dependence (2) to the case of the damage tensor of the
 Introduction and Assumptions 13

~ can be obtained
second rank, the symmetric shape of the effective stresses σ ij
as follows [27]:
1  δki σik δljσlj 
=σ ij  +  (1.7)
2  1 − ωki 1 − ω jl 

where, δij is the Kroniker delta and ωij is the damage tensor of the second rank.

1.9 Damage accumulation function for composites

The introduction of a scalar measure of damage also determines the choice of
a mathematical model to describe the process of damage accumulation. Under
complex stress conditions, the rate of damage accumulation should depend on
the joint invariants of stress tensors and tensors characterizing the mechanical
properties of the composite. The change in the individual components of
the stress tensor within the cycle can theoretically occur in different time
dependences. The desire to reflect the characteristics of the cycles for
each component of the stress tensor leads to an excessive complication of
theoretical models. A realistic approach to such situations is to introduce a
cycle parameter for the characteristic invariant of the stress state. As such a
parameter, the ratio of the minimum and maximum values of the first principal
stress per cycle is:
σ1min
R= (1.8)
σ1max
The scalar measure of damage ω considered herein depend on the
maximum value of viscoelastic deformation, the number of loading cycles N,
the cycle parameter R, temperature T, material properties and other parameters
affecting fatigue mode:

ω = f (R, E, σ, θ) (1.9)

Establishing the dependence of fatigue strength on the cycle parameter

and temperature is a difficult problem even for homogeneous materials.
Material constants of composites and nanocomposites are usually determined
from relatively simple experiments at fixed values of the cycle parameters and
temperature. In the proposed model, the hypothesis that the rate of damage
accumulation depends on the current stress value, the ratio of the minimum
stress to the maximum R and the current damage level ω is accepted.
dω
= f (R, σ, θ) (1.10)
dt
14 Creep-Fatigue Models of Composites and Nanocomposites

The form of the function f, which determines the rate of damage

accumulation is usually obtained based on the results of experiments on fatigue
strength. In accordance with the concept of continuum damage mechanics,
function g(ω) can theoretically be controlled with temperature—time (the
number of loading cycles) by a change in the elastic modulus. For practical
use of the theory, it is preferable to identify the functional dependence (6)
based on the results of fatigue strength from S – N Weller curves [28].

1.10 Paris’ Law and Miner’s Rule

The effective stress range using a rain flow cycle counting algorithm and
Miner’s linear cumulative damage law is as follows [29,30]:
n
∑ Ni =1 (1.11)
i
where ni is the number of cycles at stress range σi and Ni is the constant
amplitude fatigue life at the stress level σi. These values can be rewritten as:

Ni= Aσ−ri n (1.12)

and ni = γi T, where T is the total number of cycles and γ γi is the fraction of the
total cycles occurring at stress level σi. Substituting these values into Miner’s
rule yields:
n γT T A
∑=i
∑ Aσi= ∑ γi σnri or:=
T (1.13)
N −n
i ri A ∑ γi σnr
i

Equation 5 has the same form as Equation 4, the equation for the constant
amplitude fatigue life. The effective stress range, defined as the constant
amplitude stress range that will produce the same fatigue life as the variable
amplitude stress history, is:

(∑ γ σ )
1/ n
n
σ=
re i ri
(1.14)

This relationship was used to relate the typical traffic vehicle spectrum
of fatigue stresses into a constant level stress range that produces the same
fatigue damage.
As introduced in the abovementioned, fatigue in structures is the process
of growth of cracks under the action of repetitive tensile loads. The fatigue
life of a component is dependent upon the applied stress range, the initial
discontinuity introduced during fabrication, and local stress increase due to
 Introduction and Assumptions 15

the joint geometry. Below is an example for the application of Miner’s rule to
a variable spectrum of amplitudes (see Table 1.1 below).
This example applies the concepts of Miner’s rule to determine the
effective dimensionless stress range, 1.645, and provides an assessment of
which cycles contribute most and least to the total fatigue damage. Of note is
that the stress ranges from 0.5 to 1.5 constitute 80 percent of the total number
of cycles but cause only 16 percent of the total damage. The highest three
stress ranges, 3.5 to 4.5, only constitute 3 percent of the total number of cycles
but cause a much larger amount of damage (41 percent)! This is consistent
with life, or damage related to the cube of the stress amplitude ranges, with a
small percentage of high stress ranges using up a large portion of the structural
fatigue life details available .
In 1945, M. A. Miner popularized a rule that had first been proposed by
A. Palmgren in 1924. The rule is variously called Miner’s rule or the Palmgren-
Miner linear damage hypothesis
If θmax = 10.25 then Nf = (10.25)(104.05)/(2π) = 183045 cycles

=E 0 1[GPa];
= α 10−4 [10 K];
= σf 10.25(20)(103 )(10
= −4
) 20.5[MPa]
= 2.97[ksi]
T* =6000 K; Tmax =
βT*θ + T* =
0.0333(600)10.25 + 600 = 8050 K
20(10.25) + 600 =
σmax 0.175[GPa]
= = 175[MPa] 20.5[MPa]

The formulation, based on thermodynamic principles, leads to a system of

kinetic equations for the evolution of damage. An effective viscosity inversely
proportional to the rate of damage increase is introduced to account for

Table 1.1. Application of Miner’s rule to a variable spectrum of amplitudes.

Stress Percentage Amplitude Amplitude Relative damage Percent

interval of amplitudes σ0i (σ0i)3 γi(σ0i)3 Damage
γi (%)
1 40 0.5 0.125 0.05 1
2 25 1.0 1.0 0.25 5
3 15 1.5 3.375 0.50625 10
4 10 2.0 8.0 0.8 16
5 5 2.5 15.625 0.78125 16
6 2 3.0 27 0.54 11
7 1 3.5 42.875 0.42875 9
8 1 4.0 64.0 0.64 13
9 1 4.5 91.125 0.91125 19
Total 100% Total: 4.9075 100%
Effective σeff. = (4.45185)1/3 = 1.645
stress
16 Creep-Fatigue Models of Composites and Nanocomposites

Table 1.2. Palmgren-Miner linear damage.

frequencies Percentage log (log pi)3 Relative damage Percent

@ stress of frequencies pi γi(log pi)3 Damage
level γi (%) per fatigue
life span L = 100%
p = 1 @ σ0i = 0.5 1 0 0.0 0.0 0
p = 10 @ σ0i = 1.0 15 1.0 1.0 0.15 5
p = 10 @ σ0i = 1.5
2
25 2. 8 2.0 10
p = 103 @ σ0i = 2.0 20 3.0 27 5.4 16
p = 10 @ σ0i = 2.5
4
15 4.0 64 9.6 15
p = 105 @ σ0i = 3.0 10 5.0 125 12.5 11
p = 106 @ σ0i = 3.5 10 6.0 216 21.6 8
p = 10 @ σ0i = 4.0
7
3 7.0 343 10.29 13
p = 108 @ σ0i = 4.5 1 8.0 512 5.12 18
Total 100% Total: 66.66 100%
Effective frequency log peff.= (66.66)1/3 = 4.05
p = 104.05

gradual accumulation of irreversible deformation due to dissipative processes.

A power-law relation between the damage variable and elastic modulus leads
to a non-linear coupling between the rate of damage evolution and the damage
variable itself.

1.11 The Bergman-Milton theory

If the loading is a sequence of constant stresses with hold times and unloading
or if it is a cyclic loading, the phenomenon of fatigue occurs as a superimposed
effect. It is called creep-fatigue interaction because the earliest models were a
combination of two terms: one for creep, and one for fatigue with or without
couplings to obtain a model of nonlinear interaction.
The simple Taira rule of linear interaction (1962) applies to isothermal
uniaxial cyclic loading with 0 ≤ σ ≤ σmax and with a hold time of ∆t. If NR is the
number of cycles to rupture corresponding to a time tR ≈ ∆t·NR; if NRF is the
number of cycles to rupture in pure fatigue (∆t ≈ 0) for the same stress range;
and if tRc is the time-to-rupture in pure creep for the same maximum stress
σ = σmax for all t, then the Taira rule reads [31,32]:
tR N
+ R = 1 (1.15)
t RF N RF
 Introduction and Assumptions 17

The time-to-rupture tR at constant stress and constant temperature is the

 = f (σ, ω) with the initial condition
solution of this differential equation for ω
t=0 → ω=0
r
 σ 
ω =   (1.16)
 A ω (1 − ω) 
where, Aω and “r” are material parameters
A composite is an inhomogeneous structure. However, it can be
characterized as homogeneous by using effective parameters, obtained
through averaging, or homogenization, if the sizes of inhomogeneities are
much smaller than the characteristic length of a bulk material. Any existing
homogenization theory tries to ascribe effective parameters to a mixture
of different phases, providing mixing rules. A mixing rule is an analytical
formulation that describes an effective parameter as a function of the size of
inclusions and concentration. There is currently a multitude of mixing rules
in the literature based on different homogenization theories and approaches.
They are applicable to different types of mixtures, rates of generalization, and
depend on which parameters are homogenized, and what their limits are.

1.12 Failure criteria

In most studies, two-component mixtures are considered, where identical
inclusions are embedded in a homogeneous matrix. Effective properties of
such a composite depend on the intrinsic properties of the inclusions and
the matrix, as well as on the morphology of the composite. The morphology
is a characterization of the manner, in which inclusions are distributed in
the composite, including their concentration, shape, and correlations in
the location. Therefore, the morphology determines how inclusions are
shaped and distributed, whether they are mutually aligned/misaligned in the
composite, and what concentrations of inclusion phases and a matrix material
are. A conventional approach to describe the properties of composites employs
mixing rules, i.e., equations that relate the intrinsic properties of inclusions
and the matrix with the effective properties of the composite based on a simple
idealized model considering an ellipsoidal-shaped inclusion. Typically, the
characterization of the concentration and the shape of inclusions are included
explicitly in the mixing rules, and the account for other morphological
characteristics is attempted by a proper selection of the mathematical form
of mixing rules. A number of mixing rules are found in the literature. The
basic mixing rules are the Maxwell Garnet equation (MG) [33], Bruggeman’s
Effective Medium Theory (EMT) [34], and the Landau-Lifshitz-Looyenga
mixing rule (LLL) [35].
 18 Creep-Fatigue Models of Composites and Nanocomposites

Effective medium approximations (EMA) pertain to analytical modeling

that describes the macroscopic properties of composite materials.
The Maxwell Garnett equation is as follows:
ψ eff − ψ m  ψ − ψm 
= ϕ  incl  (1.17)
ψ eff + 2ψ m  ψincl + 2ψ m 
where ρeff is the effective constant of the medium, ρinc of the inclusions, and ρm
of the matrix; φ is the volume fraction of the inclusions ρinc.
Bridgeman’s model
Without any loss of generality, we shall consider the study of the effective for
a system made up of spherical inclusions with different arbitrary values. Then
the Bridgeman formula takes the form:
ψ m − ψ eff  ψ − ψ eff 
(1 − ϕ) + ϕ  incl  =0 (1.18)
2ψ eff + ψ m  ψincl + 2ψ eff 

The Landau−Lifshitz−Looyenga mixing rule (LLL) is written as

(ψ eff + 1)1/ 3 − 1 = ( )
ϕ (ψincl + 1)1/ 3 − 1

ψ eff ψincl
= ϕ
1 + nψ eff 1 + nψincl
(1.19)
1 1 + n(1 − ϕψincl )
= ψincl
nψ eff nϕ
ϕψincl
ψ eff =
[1 + n(1 − ϕ)ψincl ]

The LLL expression was extensively used to describe the properties of

dispersive systems composed of powders or exhibiting porous structures.
It was even shown that the LLL formula was more reliable when mixtures
contained strongly dissipative particles and was compared to others like
Maxwell Garnett (MG) and Bruggeman,

deff = (1+φ((θ+1)^0.333-1))^3 – 1 (1.20)

 Introduction and Assumptions 19

Equation (1.20) is written for the generalized viscosities of inclusions,

ψincl, and the effective viscosity, ψeff, both normalized to the viscosity of the
matrix. The LLL mixing rule is built up by an iterative procedure starting
from a homogeneous material of inclusions and replacing a small amount of
this material by the material of the matrix. After that, the resulting “effective”
material is regarded as the homogeneous component for the succeeding
substitution step, and so on, which results in Equation (3). The mixing rule
obtained by the same iterative procedure starting with the homogeneous matrix
is referred to as the asymmetric Bruggeman approximation. The result of the
LLL mixing rule is independent of the form factor of inclusions. The LLL
mixing rule is known to be an accurate result for the case when the material
parameter of inclusions differs slightly from that of the matrix. Agreement of
both the MG and EMT mixing rules with the LLL mixing rule in the case of
the susceptibility of inclusions slightly differing from zero is attained only
if n = 1/3. When the volume fraction of inclusions is small, p << pc, and
the interaction between the inclusions is negligible, all three theories are
reduced to:
ϕψ
ψ eff = icl (1.21)
1 + nψicl
Strictly speaking, Equations (1.18), (1.19), and (1.20) are valid for the case
of perfectly spherical shaped inclusions, which have a shape factor n equal to
1/3. For non-spherical particles, the inclusions must be averaged over all three
principal axes of the inclusion (1.21). Two particular cases of non-spherical
inclusions are of practical interest—nearly spherical inclusions and highly
elongated inclusions (long fibers or platelets). For nearly spherical inclusions,
the composites are conventionally described by Equations (1.18), (1.19) and
(1.20) involving an averaged form factor , which is found empirically and may
differ from 1/3. For elongated inclusions, the form factor along the shorter
axis (in the platelet case), or the sum of two form factors along the shorter
axes (in the fiber case) is close to unity, and the orientation of inclusion in
other directions can be neglected. In this case, the above equations are valid
again, with a randomization factor, κ, included in the right-hand part of the
equations to account for an alignment of non-spherical inclusions. For a fiber-
filled composite, κ = 1/3, when the fibers are randomly oriented in space, and
κ = 1/2, when the fibers are randomly oriented in a plane. For composites
filled with platelet-shaped inclusions, κ = 2/3 for the 3D isotropic orientation.
 20 Creep-Fatigue Models of Composites and Nanocomposites

1.13 The Bergman-Milton theory

A generalization of mixing rules may be made with the use of the Bergman-
Milton spectral theory (BM). The theory expresses the effective material
parameter of a composite as
1
b(n)
fi ∫
χeff = dn
0
1 + nχicl
(1.22)
1 1
1 − fi
∫ b(n)dn 1=
= and ∫ nb(n)dn
D
0 0

where the spectral function, b(n), is introduced as a quantitative characterization

of the composite’s morphology. One can see from Equation (1.22),
the BM theory accounts for a distribution of effective form factors of
inclusions in a composite. This distribution may be associated with the
following statistical parameters and processes within the composite: a spread
in the shapes of individual inclusions comprising the composite; possible
agglomeration of inclusions to clusters; and the effects of multiple scattering
and inhomogeneous fields affected by neighboring inclusions. Again, the
spectral function is the same for all susceptibilities of a particular composite.
The sum rules relate the spectral function b(n) to the volume fraction of
inclusions fi for a macroscopically isotropic composite in D dimensions. The
practically important cases are D=3 (an isotropic 3D composite with non-
aligned randomly distributed inclusions, the shape of which is arbitrary in the
general case) and D=2 (an assembly of infinitely long cylinders). The sum
rules provide an agreement of the spectral theory with the LLL mixing rule
at χincl→0.
A well known example of such a combination is the Lichtenecker mixing
rule [36], which is written as:
k k
ψ eff = ϕψincl + (1 − ϕ)ψ kmatrix (1.23)

In Equation (1.23), k has a physical meaning of a critical exponent, which

is conventionally treated as a fitting parameter to obtain an agreement with
measurements. Equation (1.23) may be considered as an empirical combination
of the LLL mixing rule and the introduction to long range correlations.
This kind of test may be helpful in the validation of the unified damage
law for creep-fatigue damage if the strain to rupture εpR and the strain at the
inflexion point corresponding to the threshold εpω are recorded. The integration
of the differential law for σ=const is particularly easy.
 sr N
0 N0

Introduction and Assumptions 21

s
8

1.4
4

2 700°C
800°C
1
1100°C 1000°C 900°C tc hr

–0.5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5
–1
Figure 1.4. Isochronous curves of super alloy IN 100 at different temperatures.

The most important tasks here are related to determining the conditions for
the intensive development of cyclic creep, evaluating the interaction of fatigue
and creep, as well as obtaining quantitative relationships between deformation,
cyclic creep rate, temperature and stress of the cycle. Problems of analytic
estimation of the limiting state of materials under conditions of developed
cyclic creep are of considerable interest. To describe the experimentally
obtained high-temperature curves of creep and fracture defining equations
like Bailey-Norton (the time-hardening formulation of power law creep) [37]
and Rabotnov-Kachanov [38] are used:
σn σn
=ε B = ; 
ω D
(1 − ω)m (1 − ω)m (1.24)
ω(0) = 0; ω(t*) =ω2

It is assumed that for a certain period of time Δt after the application the load
matrix as a result of creep gains has a deformation increment Δεm = f (σm, T) Δt.
From the condition of compatibility deformation the same increment gets
the filler Δεf = Δεm. An increase in fillers’ strains leads to an increment in of
tensile stresses Δσf = Ef Δεf in them and, in turn, reduction of stresses in the
matrix according to the additivity rule.
σ −σ V
σm =c f f (1.25)
1 − Vf
Since the total load on the composite is Pc = σc Fc = const, where Fc is the
cross section of a composite (assumed unchanged in the process of creep of the
composition). The use of the additivity equation is based on the assumption
of uniform distribution of deformations in the matrix. By virtue of small
22 Creep-Fatigue Models of Composites and Nanocomposites

volume fractions of fibers (Vf ≈ 5%) in the studied material this assumption is
justified. However, in general, the principal of taking into account the uneven
distribution of deformations and stresses in components is also important.
In the next time step Δt, the creep of the matrix will occur at lower stresses
σt and thus is modeled as damped creep of the matrix material characterized
by a decrease in stresses in it and elastic reloading from the reinforcement.
Reinforcement of composites involves scattering of strength properties
and this results in the progression of final failure of the material . The
accumulation of cracks in the composite leads to a decrease in their carrying
capacity. Assuming that the accumulation of discontinuities occurs statistically
uniformly in the region of fragmentation of fibers, we introduce the damage
accumulation function W (σf), which characterizes the ratio of the number of
failed fibers in a layer of critical length equal to the fragment length to the
total number of fibers. As the ruptures accumulate, the carrying capacity of
the composite decreases and the stresses are redistributed from the destroyed
fibers to the matrix.
Exponential Model of viscosity
An exponential model for the temperature-dependence of shear viscosity (μ)
was first proposed by Reynolds in 1886 [39]: μ(T) = μ0 exp (− bT), where T is
the temperature and μ0 and b are coefficients. This is an empirical model that
usually works for a limited range of temperatures.
Arrhenius model
This model is based on the assumption that the fluid flow obeys the Arrhenius
equation for molecular kinetics [40]: μ (T) = μ0 exp ( — E ), where T is the
RT
temperature, μ0 is a coefficient, E is the activation energy and R is the universal
gas constant. A first-order fluid is another name for a power-law fluid with an
exponential dependence of viscosity on the temperature.
Williams-Landel-Ferry model
The Williams-Landel-Ferry model, or WLF for short, is usually used for
polymer melts or other fluids that have a glass transition temperature. The
 C1 (T − Tg ) 
model is: µ(T) = µ0 exp   , where T-temperature, C1,C2,Tg and μ0
 C2 + T − Tg 
 
are empirical parameters (only three of them are independent from each other).
If one selects Tg—the glass transition temperature, then the parameters
C1,C2 become very similar for the wide class of polymers. Typically, we get
C1 = 17.44 K and C2 = 51.6 K.
 Introduction and Assumptions 23

1.14 Failure criteria

Progressive damage models often employ a failure criterion along with
degradation models. Different failure criteria have been proposed in the
literature such as the maximum stress criterion, Haskin’s criterion [41] and
Puck’s criterion [42].
Maximum stress theory
The maximum stress theory predicts failure when the stresses in the principal
material axes exceed the corresponding material strength. In order to avoid
failure, it has to be ensured that the stress limits are not exceeded,
−R c < σ < R t
−R c⇔ < σ⇔ < R ⇔
t
(1.26)
| τ |↵ < R ↵
As soon as one of the inequalities above is violated the material fails by a
failure mode that is associated with the allowable stress. This failure criterion
does not take any interaction of stress components into account and therefore
certain loading conditions such as superposition of tensile and shear stresses,
lead to non-conservative results.
Haskin’s failure theory
One of the first failure criteria applied to fatigue distinguishing fiber-failure
and matrix-failure modes was proposed by Hashin and Rotem. They derived
the fatigue failure criterion from their formulation under static loading, which
is stated in fiber mode as:
=σ1 R t for σ1 ≥ 0
| σ1 | R c for σ1 < 0
=
and for the 2D inter − fibre failure : (1.27)
2
 σ   τ 2
 2t  +  12  =
1 for σ2 ≥ 0
 R   R ↵ 
 
2
 σ   τ 2
 2c  +  12  = 1 for σ2 < 0
 R   R ↵ 
 
In case of inter-fiber failure Hashin and Rotem proposed an elliptic
equation, which depends on the transverse stress σ2 and on the in-plane
stress τ12.
24 Creep-Fatigue Models of Composites and Nanocomposites

Phenomenological models include a description of the damage in

composites during fatigue loading where gradual deterioration of macro-
and meso-mechanical properties is described through residual stiffness or
strength models. A phenomenological model based on residual strength uses
experimental observations to describe the strength loss of composites and is
subdivided into two models: sudden death model and wear-out model [43].
The residual strength in the sudden death model is kept constant over a certain
number of cycles and is then suddenly degraded drastically when it reaches
the critical number of cycles to failure Nf.
This model is a suitable technique to describe this behavior for high-
strength unidirectional composites subjected to a high level state of stress
[44]. The residual strength in the wear-out model is contrarily continually
decreasing over the number of cycles following a certain predetermined
equation. These models can be utilized at lower level states of stress. In the
wear-out model according to Halpin et al [45], it is assumed that the residual
strength R(n) is a monotonically decreasing function of the number of cycles
n, and that the change of the residual strength can be approximated by a
power-law growth equation:

R m (n) − σ m n
m m
= 1− (1.28)
R (0) − σ Nf
as reformulated by Shokrieh and Lessard [46] with m as a constant depending
on the material. This model has been used by many authors in probabilistic
and mechanistic models.
The basis of the concepts of “catastrophic” failure considered below is based
on the physical model of damage accumulation in the process of creep deformation.
In no way rejecting directions in the study of micro-stresses and micro-
deformations, it should nevertheless be noted that this kind of theory at the
level of micro-inhomogeneous media are quite complicated even in the case
of a uniaxial stress state and therefore are not well suited for solving, for
example, boundary problems of continuum mechanics. For these purposes,
more preferred are the usual phenomenological theories of creep at the macro
level.
Previous phenomenological creep-fatigue models devoted to life
prediction and damage assessment were mainly developed on the basis of the
Manson-Coffin law [47]. The effects of all these mechanisms are summarized
under the label creep damage in phenomenological theory. The term “damages“
designate the general destructive effect of the various changes occurring in
the material, which eventually limit its usefulness. Thus, the visible damage
may consist of an accumulation of permanent deformation from creep
(viscous flow), plastic flow, initiation of corrosion pits, or development of
 Introduction and Assumptions 25

macrocracks. These visible effects are the result of submicroscopic changes

in the crystalline lattice or in the intercrystalline boundary material. The
processes designated by the terms slip, diffusion, recrystallization, rotation
of grains, phase change, and relaxation involve varying degrees of atomic
rearrangement that take place within the material; any of these processes may
constitute the initial stages of “damage”.
A very successful suggestion for the description of damage is due to
Kachanov [48]. His leading idea is that only a part of the cross-section ‘A’ of
the creep sample has to bear the force, namely the effective or reduced cross-
section Ae, reduced by cracks, pores and more. This leads to effective stresses
F F
σef = > (1.29)
Ae A
If the effective stresses enter the creep law, this causes a growth in the
creep rate. For this aim one introduces a damage parameter ω as the ratio,
A − Ae
ω= (1.30)
A
0 < ω <1

1.15 Principle of stress equivalence

The damaged material behaves like the undamaged one if the stress in the
creep law is substituted by the effective stress [49].
The damage parameter ω is considered as an internal variable, for which
one needs an evolution equation. By analogy to the Burgers Standard model,
one may write the equation as follows

ω =−ασ ef + βσef + γω (1.31)

The rate of the effective stress is then,

F  F ′ F′ F (1.32)
σ ef =  = − 2 A′e
Ae  Ae  Ae Ae
In a monotonous creep test, the applied load F is constant, so that F’=0.
The growth of the effective cross-section area is fed by different sources after
substituting Equation 1.32 and Equation 1.33 into Equation 1.31.
F
σ ef =− A′e
A e2 (1.33)
F F
= α
ω A′e + β + γω
A e2 Ae
 26 Creep-Fatigue Models of Composites and Nanocomposites

In this manuscript, we introduced the function ω, which is the area of

cracks per unit cross-sectional area at a given time. In the case of this model,
the crack rate of area growth is equal to the crack growth rate v, that is,

∂ω/∂t = ωβ v(σ) (1.34)

If β =1, the first derivative dω/dt = ωv (σ). The average stress in the cross
section at normal stress is σ = σ0/(1 – ω). The latter formula expresses the
effect of increasing the stress level in the damaged material, since 0 < ω < 1.
Introducing the expression σ = σ0/(1 – ω) into equation (1.1), we can obtain a
differential equation for function ω (t), where σ0 is a given function of time.
At the initial time t = 0, the function ω is zero, ω = 0, at the failure time
t = T→ ω = 1.
The equation can be easily integrated in the case when the cumulative
damage growth rate is a power function of the stress v = Aσα. In this case,
the variables are separated and are as follows (1 – ω)α ω-βd ω =Ασdt. The
time fraction rule can be used to predict creep-fatigue life with the help of the
following relation,
m  n  m 
∑ th

∑ N  + ∑  
≤1 (1.35)
=i  f i j 1  Tr
1= j
where, n is the number of cycles to failure in creep-fatigue, Nf is the number
of cycles to failure in pure fatigue for the given strain range, th is the hold
time in each cycle, Tr is the time to rupture in pure creep for a given stress
and temperature; and ω is the allowable combined damage fraction. In this
study Nf was found from a fatigue strain-life curve and Tr was calculated
from a Larson-Miller plot. The first term in Equation (1.35) is a cycle
fraction representing fatigue damage and the second term is a time fraction
representing creep damage.
Phenomenological uniaxial energy type models are proposed to describe
the inelastic creep deformation and fracture of composites as the joint action
of the static and cyclic loads. The energy and adjacent thermodynamic
approaches for describing inelastic rheological deformation and failures
modes of composites in conditions of unsteady thermal loading are very
important, and the expediency of their application in the structural calculation
practice is beyond doubt. The aim of this book is to generalize the author’s
approach [50] to describe the class of phenomena occurring in composite and
nanocomposite material under combined action of static σ0 and cyclic loads
with the amplitude of the cyclic component being σa. It is confined here to the
 Introduction and Assumptions 27

consideration of the so-called low-cycle loading at a frequency f > 10Hz and

the amplitude coefficient A = σ0/σa, which does not exceed a certain critical
value Acr. In the case under consideration, cyclic loading leads to two main
effects [51]: (1) acceleration (or even initiation) of the creep process for a given
static stress σ0; (2) reduction of the accumulated inelastic deformation at the
moment of failure in comparison with the analogous value under purely static
loading. This process is called cyclic creep or cyclic—creep-fatigue. These
phenomena cannot be described either within the framework of conventional
classical approaches, nor from the point of view of phenomenological
creep, nor from fatigue positions in the asymmetric cycle [52].
From the analysis of works on cyclic creep-fatigue it follows that at the
phenomenological level, the following approaches can be conventionally
distinguished.
1. The introduction of a reduced stress equal to a static stress at which the
durability in the static creep mode coincides with the durability in the
cyclic creep mode. In this approach, the similarity of the curves of static
and cyclic creep is postulated, which is one of its drawbacks. In addition,
under non-stationary loading regimes such theories give large errors in
both the static and cyclic components. Numerous attempts to create a
universal principle of nonlinear summation of damage to date have not
led to success.
2. The basic model for the solution of the problem posed in the present
manuscript is similar to the model proposed for a quasistatic loading.
3. Below is a procedure for determining the parameters of the proposed
model.
With the energy approach, the following experimental datasets are used
as the base:
- diagram of material stretching at a constant sufficiently high rate of
deformation;
- series of creep curves from the beginning of loading to the moment of
failure at σ0 = const, which are called stationary creep curves; series of
creep curves from the beginning of loading to the moment of failure at
σ0 = const, σa = const, called the stationary cyclic creep curves.
With the introduction of effective stress, isotropic damage and using the
principle of deformation equivalence [53–55], the uniaxial differential type
constitutive creep-fatigue equation has a form.
 28 Creep-Fatigue Models of Composites and Nanocomposites

1.16 Standard linear model

Differential equation in this case is as follows:
Enε + Hε = nσ + σ

E1E 2 η (1.36)
=E E=
1; H = ;n
E1 + E 2 E1 + E 2
If, for example, E = E1 = E2 and ε = ε0 [sin (pt)], then the dimensionless
form of Equation 1.36 is as follows:
E η
E = E1 ; H = ;n = ;ε = ε0 sin pt
2 2E
dσ 1 E
= − m1σ + m1 + E
dθ n 2n
E=E 0 [(0.625 - 0.375*tanh(c* (θ - θg )))]

Ea RT* a
=θ [T − T*=
]; β ; τ= 2 t (1.37)
RT*2 Ea L

a -thermal diffusivity of composite materials

L - linear dimension ; T* - Base Temperature [0 K]

θg − glass trunsional temperature (dimensionless)

For purposes of comparison with all subsequent cases of fatigue creep

in composites we shall first consider the example of pure creep under high
temperature influence.
Example 1.1a—pure creep failure (w/o harmonic elongations and Damage
function 𝝎)
If E ≠ const. [E = (0.625 – 0.375*tanh(c* (θ – θg)))]; σmax/σ0 ≈ 75 and n = const.
then Equation (1.2) has numerical solution (using POLYMATH software [56]):
Data: E0 = 0.583(104) MPa; ε0 = 0.3(10–4); c = 5; θg = 4
 Introduction and Assumptions 29

Calculated values of DEQ variables

Initial Minimal Maximal

Variable Final value
value value value
1 E 1. 0.25 1. 0.25
2 m 0 –0.1895272 0.0705855 –0.1895272
3 m1 0.0405 –0.1400365 0.0405 –0.1400365
4 n 0.1 0.1 0.1 0.1
5 t 0 0 13.5 13.5
6 Y 0.175 0.175 12.70131 12.70131

Differential equations
1 d(Y)/d(t) = –(1/n)*Y*m1 + 0.175*E +0.175* 0.5*E*(1/n)*m1
Explicit equations
1 m = (0.0405*t – 0.01126*t^2 + 0.001462*t^3 – 0.00006868*t^4)
2 m1 = (0.0405 – 0.02252*t^1 + 0.004386*t^2 – 0.0002747*t^3)
3 E = (0.625 – 0.375*tanh(5* (t – 4)))
4 n = 0.1
Model: Y = a1*t + a2*t^2 + a3*t^3 + a4*t^4 + a5*t^5
Variable Value
a1 1.110151
a2 –0.7023468
a3 0.1754299
a4 –0.0181235
a5 0.0006566

Pure creep failure

1.27E+
1.14E+1 –Y
1.02E+1
8.94E+0
7.69E+0
6.44E+0
5.19E+0
3.93E+0
1.5
2.68E+0
1.43E+0
1.75E–1
0.00E+0 1.35E+0 2.70E+0 4.05E+0 5.40E+0 6.75E+0 8.10E+0 9.45E+0 1.08E+1 1.22E+1 1.35E+1
t

Figure 1.5. Pure creep failure.

1.00E–0
9.00E–1 –K
-- K1
8.00E–1
7.00E–1
6.00E–1
5.00E–1
4.00E–1
1.6
 30 Creep-Fatigue Models of Composites and Nanocomposites

σ = 1.11θ – 0.70θ2 + 0.175θ3 – 0.018θ4 + 0.000657θ5 (1.38)

σmax/σ0 ≈ 75
Example 1.1b—creep failure with harmonic strain: ε = ε0(sin(pt))
If E ≠ const. [E = (0.625 – 0.375*tanh(c* (θ – θg)))]; σmax/σ0 ≈ 75 and n = const.
then Equation (1.3) has numerical solution (using POLYMATH software [55]):
Data: E0 = 0.583(104) MPa; ε0 = 0.3(10-4); c = 5; θg = 4
Note: If E = const. and n = const. then Equation (1.2) has an analytical solution:
dσ 1 E
= − σ+ ε0 (sin pt) + E(ε0 p)(cos pt)
dt n 2n
=σ A sin(pt) + Bcos(pt)

Apn cos(pt) − Bpn sin(pt) + A sin(pt) + B cos(pt) =

ε0 E(0.5sin pt + pn cos pt)
=

Apn + B = ε0 Epn (1.39)


A − Bpn =0.5ε0 E

ε0 E[p 2 n 2 + 0.5] ε0 E[0.5p 2 n 2 + pn]

A = ; B
[p 2 n 2 + 1] [p 2 n 2 + 1]

B [0.5p 2 n 2 + pn]
tan ϕ = = ; σmax = A 2 + B2
A [p 2 n 2 + 0.5]
It should be noted that the phase shift angle between strains and stresses
in this case depends not only on the frequency of steady-state oscillations,
but also on the ratio of the viscous and elastic components of this model.
Thus, for example, the maximum value of the phase shift angle at n = 0.3 is
at a frequency p = 10 Hz, while at n = 0.1 it is at a frequency p = 3 Hz. It is
also interesting to note that when p tends to infinity, the value of the phase
shift angle decreases monotonically and asymptotically approaches φ = 0.5,
regardless of the value of n below).
n = 0.1 p = 10; n = 0.05 p = 20; n = 0.3 p = 3
Maximum stress σmax vs. frequency of steady-state oscillations p is presented
on Fig. 1.6.
 8.94E+0
6.44E+0
5.19E+0
7.69E+0
6.44E+0
1.5
1.5
3.93E+0
5.19E+0
2.68E+0
3.93E+0
1.43E+0
2.68E+0
1.75E–1
0.00E+0 1.43E+0
1.35E+0 2.70E+0 4.05E+0 5.40E+0 6.75E+0 8.10E+0 Introduction
9.45E+0 1.08E+1 and Assumptions
1.22E+1 1.35E+1 31
t
1.75E–1
0.00E+0 1.35E+0 2.70E+0 4.05E+0 5.40E+0 6.75E+0 8.10E+0 9.45E+0 1.08E+1 1.22E+1 1.35E+1
t
1.00E–0
9.00E–1 –K
--1.00E–0
K1
8.00E–1
9.00E–1 –K
7.00E–1 -- K1
8.00E–1
6.00E–1
5.00E–1
7.00E–1
6.00E–1
1.6
1.6
4.00E–1
5.00E–1
3.00E–1
4.00E–1
2.00E–1
3.00E–1
1.00E–1
2.00E–1
0.00E–1
0.00E+0 1.00E–1
1.00E+1 2.00E+1 3.00E+1 4.00E+1 5.00E+1 6.00E+1 7.00E+1 8.00E+1 9.00E+1 1.00E+2
t
0.00E–1
9.42E–1
Figure 1.6.
0.00E+0 Κ
1.00E+1 = 2.00E+1
tan (φ) @ n = 0.3;΄Κ1
3.00E+1 4.00E+1 = tan (φ) @6.00E+1
5.00E+1 n = 0.17.00E+1
t = p. 8.00E+1 9.00E+1 1.00E+2
t
8.60E–1 –N
9.42E–1
7.79E–1
8.60E–1 –N
6.97E–1
7.79E–1

1.7
6.15E–1
6.97E–1
5.34E–1

1.7
6.15E–1
4.52E–1
5.34E–1
3.70E–1
4.52E–1
2.89E–1
3.70E–1
2.07E–1
2.89E–1
1.25E–1
0.00E+0 2.07E–1
1.00E+1 2.00E+1 3.00E+1 4.00E+1 5.00E+1 6.00E+1 7.00E+1 8.00E+1 9.00E+1 1.00E+2
t
1.25E–1
0.00E+0 1.00E+1 2.00E+1 3.00E+1 4.00E+1 5.00E+1 6.00E+1 7.00E+1 8.00E+1 9.00E+1 1.00E+2
t
Figure 1.7. Maximum stress σmax vs. frequency of steady-state oscillations p.
Damage function
1.00E+1
9.00E–1 –X Damage function
Obviously,
8.00E–1
1.00E+1 the actual (dimensional) maximum stress has to be multiplied

by Eε09.00E–1
7.00E–1 . –X

6.00E–1
Example
5.00E–1
8.00E–1
1.8
7.00E–11.2—If E ≠ const. [E = (0.625 – 0.375*tanh(c* (θ – θ )))] and

1.8
g
n = const. then Equation (1.2) has numerical solution (using POLYMATH
4.00E–1
6.00E–1
5.00E–1
software
3.00E–1
[56]):
4.00E–1
2.00E–1
3.00E–1
Data: E = 0.583(104) MPa; ε0 = 0.3(10-4); c =5; θg = 4
1.00E–1
0
2.00E–1
0.00E–0
0.00E+0 1.00E–1
5.00E+1 1.00E+0 1.50E+0 2.00E+1 2.50E+0 3.00E+0 3.50E+0 6.00E+0 4.50E+0 5.00E+0
Differential
0.00E–0
equations t
0.00E+0 5.00E+1 1.00E+0 1.50E+0 2.00E+1 2.50E+0 3.00E+0 3.50E+0 6.00E+0 4.50E+0 5.00E+0
d(Y)/d(t) = –((1/n)*Y*m1 + 0.29*2.01*e0*p*(cos
t (p*t))*E +
1
0.29*2.01*0.5*E*(1/n)*m1*(e0*(sin(p*t))))
Explicit equations
1 m = (0.0405*t – 0.01126*t^2 + 0.001462*t^3 – 0.00006868*t^4)
2 m1 = (0.0405 – 0.02252*t^1 + 0.004386*t^2 – 0.0002747*t^3)
3 E = (0.625 – 0.375*tanh(5* (t – 4)))
4 p = 1 Hz
5 n = 0.1
6 ε0 = 0.3
32 Creep-Fatigue Models of Composites and Nanocomposites

For temperature range 0 < θ < 11 (300ºF < T < 1260ºF) the dynamic stress
oscillations might be considered as periodical with the rapid change around
transitional glass temperature Tg (θg = 4).
The continuum mechanics of damage is based on the fact that the change
in the mechanical properties of materials over time can be phenomenologically
interpreted as the result of the accumulation of damage and various defects.
When damage reaches a dangerous level, failure occurs. Cracking begins at
the earliest stages of deformation and is associated with the growth of existing
and the emergence of new sub- and micro defects. The material always has a
large number of different defects, leading to high local stresses. The decrease
in strength (the properties of the body to resist impacts from the external
environment) of deformable solids can often be explained by the hidden
destruction and micro defect structure of the body. Thus, since damage to a
body significantly affects the nature of its destruction, it becomes obvious that
both the fracture mechanics and the damage mechanics are designed to solve
the main applied problem of estimating the safety margin of a solid [20].
The modeling of damage processes using the damage parameter ω, the
scalar parameter of Kachanov – Rabotnov, which [22] was associated with
the porosity of the material and it was assumed that the damage accumulation
was due to the combined effect of diffusion and viscous pore growth
mechanisms under conditions of high temperature creep. In [23], the intensity
of accumulated creep strain is taken as a measure of material damage. The
model proposed by the author within the framework of the “Standard linear
mechanical system” of Voigt – Kelvin describes the cumulative damage
under creep conditions.. The rate of cumulative damage ω is included in the
dimensionless differential creep-fatigue equation of the Voigt – Kelvin model
that has the following form,
q q
dσ 1 E  1   1 
= − σ+ ε0 (sin pt)   + E(ε0 p)(cos pt)  
dt n 2n 1− X  1− X 
q
dX  σl 
= A   ; A = const.; t = m(T) (1.40)
dt 1− X 
E1E 2 η
E E=
= 1 E2 ; H
= ;=
n
E1 + E 2 E1 + E 2

= 0; X(0)
σ(0) = 0
The parameters ‘q’ and ‘l’ are obtained from the experimental cyclic
fatigue data. Consider first the effect of temperature – time curves (linear and
nonlinear) on the cyclic creep-fatigue process.
Another Random Scribd Document
with Unrelated Content
The Project Gutenberg eBook of Muinais-
tiedustuksia Pohjanperiltä
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: Muinais-tiedustuksia Pohjanperiltä

Author: J. W. Calamnius

Release date: June 23, 2018 [eBook #57377]

Language: Finnish

Credits: Produced by Jari Koivisto

*** START OF THE PROJECT GUTENBERG EBOOK MUINAIS-

TIEDUSTUKSIA POHJANPERILTÄ ***
Produced by Jari Koivisto

MUINAIS-TIEDUSTUKSIA
POHJANPERILTÄ
Kirj.

J. W. Calamnius
 Helsingissä,
Aikakauskirja Suomi, 7 Osa,
Suomalaisen Kirjall. Seuran kirjapainossa,
1868.

Muinais-tiedustuksia Pohjanperiltä[1]

Ala-Tornion pitäjä ynnä Karungin kappeli.

Tornio on nähtävästi saanut nimensä ruotsalaisista sanoista torn,

torni, ja ö, saari, jonka johdosta onkin nimi ennen ollut Torneö,
niinkuin se kirjoitetaankin vanhoissa kirjoissa, ja niinkuin eräs "Suuri-
valtaisimman kuninkaan, itse kuningas Karl XI:nen" tekemä kirjoitus
Tornion kirkossa sen myöskin kirjoittaa. Asema on vanha
kalastuspaikka, jonka asukasten ja kävijäin johdoksi ja tuntomerkiksi
torni oli rakennettu Sventsar-saarelle. Tämä saari oli ennen aikaan
eroitettu sen eteläpuolella olevasta karista, joka, muutoin veden
alainen, ainoastaan kovalla pohjastuulella oli näkyvissä. Taru kertoo
sitten, että, kun Hollantilaiset ja etelä-Ruotsalaiset tänne purjehtivat
kalan pyyntöön ja heidän laivansa kerran joutui tälle karille, he tänne
ensin panivat meriviitan, ja sittemmin rakentivat oikean tornin, josta
kari sai nimen "Tornö", jonka Suomalaiset käänsivät Tornioksi.
Vähitellen muuttui tämä kari saareksi ja yhdistyi viimein Sventsarön
kanssa.[2]
 Meren alinomaa tapahtuva laskeminen ja maan nouseminen on
kaikissa ranta-pitäjissä vaikuttanut suuria muutoksia. Ehk'eivät nämä
lyhyemmässä ajassa niin pääse näkyviin, joutuvat ne kuitenkin ajan
pitkään niin suuriksi, että tuleva vuosisata edellistänsä tuskin
tunteekaan. Näin on Torniossakin käynyt. Vielä 16 vuosisadan
keskipalkoilla mainitaan Torniolaisilla olleen hyötysä kalastuspaikka
"Sventzar-niemellä", vaan maan nousu on tätä jo aikaa sitten
muuttanut. Niinpä oli myös n.k. Kaupunginlahti muinoin niin syvä,
että laivat siellä uivat ja että sen rannoilla laivoja rakennettiin, vaan
nyt se jo on niin matala, että ainoastaan veneet siellä pääsevät
kulkemaan. Näihin asioihin viittaavat myöskin, nimet Laivanjärvi ja
Laivanniemi, joista en kuitenkaan ole mitään likempää kuullut.

Arkeologiallisista muistoista on Tornio köyhä, niinkuin sitä voipikin

arvata maan mataluudesta. Muualla kaikkialla ainakin tunnetaan
Jättejä, Jättiläisiä, ja osoitetaan heidän jättämiä merkkejä, mutta
täälläpä näistä ei saa juuri mitään nähdä eikä kuulla, ja tämä olikin
ainoa paikka, missä minä, näistä Jätteistä kysellessäni, itse sain
vastata siihen vasta-kysymykseen, mitä ne Jätit sitten olivat, "onko
ne niitä munkkilaisia?"

Lappalaisista täällä ei myöskään ole juuri mitään muistoa jälellä.

Ainoat heistä muistuttavaiset nimet tässä seudussa ovat Lapinmäki,
kunnas, missä Haaparannan kirkko nyt seisoo — mikä siis ei enään
kuulukaan meillen — ja Lappikari, mikä Tornion raastuvanoikeuden
pöytäkirjoissa mainitaan vuonna 1643 olleen kalastuskarina "aivan
kaupungin vieressä." Kari on Laivanniemen ja tuon kohta
mainittavan Pirkiön saaren välillä. Kirjoituksessaan "Bidrag till
Finlands kännedom i ethnografiskt hänseende", Suomi 1847,
mainitsee A. Warelius eräästä Torniovaara-nimisestä vuoresta
"pohjoisimmassa Pohjanmaassa", millä vuorella myös olisi raunioita.
Nimestä ehkä luulisi tämän vuoren kuuluvan tähän pitäjään. Niin ei
kuitenkaan ole laita, ja tämä Torniovaara on luultavasti Kemin
Tornivaara, josta edempänä. "Etevimmät vuoret täällä ovat Huitaperi
ja Kiimavaara, mutta uskottavain herrasmiesten ja talonpoikain
sanomain mukaan", — kirjoittaa eräässä kirjeessä mulle rehtori
Thauvon, joka tätä on asian tähden tutkinut ja hyväntahtoisesti
mulle ilmoittanut — "löytyy näillä yhtä vähän kuin muillakaan
vuoriloilla täällä päin selviä Lapinraunioita tahi kivikumpuja.
Kansantarun mukaan olisi kuitenkin Lappalaisia asunut Yli-Tornion
vuorikko-asemoilla, ja Huitaperillä taidetaan vielä muutamista
epäselvistä merkeistä nähdä, missä heillä on ollut tulisijansa."

Kaksi muuta nimeä täällä sen sijaan löytyy, mitkä ovat

paikkakunnan historioitsijain muistoon pantavat. Saari, johon Ala-
Tornion maankirkko on rakennettu, on nimeltänsä Pirkkiö (Birkiö,
Björkö), ja Helsinki on eräs kylännimi. Edellinen säilyttää historialle
Pirkkalaisten, jälkimäinen Helsinkiläisten nimeä — kaksi kansaa,
jotka mainitaan täällä harjoittaneen kauppaliikettä.[3] — Myöskin
löytyy täällä nimi Danskinsaari, saari Torniojoessa Svensar-saaren
ylipuolella, ja Rauma, kylä vähän matkaa itäpuolella Tornion
kaupunkia. Tämä nimi juteltiin tulleen Rauman kaupungin
siirtolaisista, mutta hyvinkin mahdolliselta näyttää, että nimi on
paikalle alkuperäinen, merkiten rämeistä, suoperäistä asemaa — jota
se nimi täällä merkitsee. Muuallakin maassamme löytyy tämä nimi,
niin esm. Lohtajassakin.

Pirkkiön saarella kerrotaan ensimäiset tänne tulleet Suomalaiset

asuneen, ja tästä sen nimikin olisi tullut, koska nämät Suomalaiset
olivat Pirkkalaisia, mitkä ensin tulivat tänne kalan pyyntöön, mutta
sitten jäivät tänne asumaan. Ne sanotaan omistaneen itsillensä
nimen "Lapin kuninkaat", sillä ne kävivät kolme kertaa vuodessa
kuninkaan nimessä laittomia veroja ryöstämässä Lappalaisilta, jota
pahaa menettelemistä kesti satoja vuosia, siihen asti kuin Ruotsin
hallitus vihdoin laski nämät seudut allensa. Kun Pirkkalaiset ajan
pitkään rupesivat liian ahneiksi, niin sanotaanpa välistä kovia
kahakoitakin syttyneen Pirkkalaisten ja Lappalaisten välillä. Heidän
pukunsa, jolla he koettivat pelättää ja voittaa itsillensä kunnioitusta
Lappalaisilta, sanotaan olleen merkillinen. Heillä mainitaan olleen
korkeat, punasilla nauhoilla ja kulta- sekä hopeareunuksilla
varustetut lakit. Samaten oli myös heidän takkinsa punasilla
nauhoilla ja palteilla, niinkuin myös kullalla ja hopealla, hyvin
koristettu (J.F. Thauvon.).

Kirkkojen ympäri yhdistyy usein pitäjäin historia. Tornion

kaupunginkirkko on rakennettu vuonna 1683 ja näyttää
häkityksillänsä maalauksellansa vähän katoliselta. Kummallakin
puolen alttaria on seinäkoristuksena historiallinen muistomerkki,
toinen se jo mainittu: kuningas Karl XI:nen tekemä ja sittemmin
tauluun maalattu kirjoitus hänen korkeasta käynnistänsä Tornion
kirkontapulissa, kesäkuulla v. 1694, katsomassa kesä-yön aurinkoa,
toinen eräs vaakuna, jonka kirjoitus jo on kulunut. Edellinen löytyy jo
julkaistuna Murman'in kertomuksessa s. 7, jonka tähden se tästä
jääköön. Jälkimäisen mainitsee Brunnius kuuluneen näin: "Kongl.
Maij:ts Troman och Major under Westerbotns infanteri, Wälborne
Herr Alexander Fredrich Sack genannt von Oster, är föd i Liefland på
Oesel d. 10 Novemb. 1655 och blef död i Torne d. 23 Jan. 1694."
Vanha kirkko, joka 1682 paloi, on seisonut vähän eteläisempänä,
missä nyt joukko aittoja on.

Muinaistaruja täällä ei löydy sanottaviksi. Yksi on taru kuitenkin

eräästä pispasta, joka näillä seuduin olisi kastanut joukon
Suomalaisia. Jos on tässä tarussa perää, niin se nähtävästi oli pispa
Hemming, jonka Geijer mainii kastaneen vuonna 1350
"kaksikymmentä Kemin ja Simon Lappalaista ja Karjalaista suuressa
ammeessa Tornion vieressä".[4] Mutta missä paikassa tämä on
tapahtunut, siitä ei ole tietoa; eräs himeä taru viittasi Karunkiin. —
Myös Iso viha ei näy jättäneen varsinaisia muistoja. Tähän aikaan oli
itse kaupunki melkeen kokonansa palanut v. 1714, jonka jälkeen
viholliset hävittivät ja ryöstivät mitä tuli oli säästänyt, jopa 9 kertaa
"ottivat kaupungin valtaansakin", niinkuin raastuvanoikeuden
protokollat oikein todenteolla vakuuttavat. Jos saapi Brunnius'een
luottaa, niin olisivat "meikäläiset vielä vuosina 1715 ja 1716
yksimielisesti yhteen liittyneet, varustettuina pyssyillä, keihäillä ja
miekoilla, ilman kenenkään kehottamatta. Suuresta isänmaan-
rakkaudesta kiihotetuilla olivat he muka nimensä suureksi kunniaksi
puolustaneet isänmaatansa, lyöden viholliset pakoon sekä
ilmitappelussa että myöskin käyttämällä monenlaisia sukkelasti
ajateltuja sota-keinoja. Vihdoin oli kuitenkin vihollinen, näitä
tämmöisiä tappioita pahaksi pannen, vuonna 1717 tänne tullut niin
suuressa joukossa, että hän ei ainoastaan karkoittanut meikäläisiä,
mitkä eivät enään voineetkaan tehdä vastarintaa, vaan myöskin
hävittäen kuleksi näitä seutuja aina Paajalan kylään saakka, puolen
penikulmaa Kengisten tehtaan toisella puolella. Rikkaan saaliin
kanssa olivat he viimein täältä sitten paenneet".[5]

Kemin pitäjä ynnä Tervolan ja Simon kappelit.

Salosten pitäjän ohessa Kemi lienee vanhin asuttu maakunta

pohjoisessa Pohjanmaassa, sitä todistavat sen vanhat muistot.
Nimensä, joka merkitsee jokea, pää-jokea, joen uraa eli ojannesta,
on pitäjä saanut siitä veden-juovasta, joka täällä mahtavana virtana
mereen kulkee. Ennen vanhaan sen nimi myöskin kuului Kymi, Kimi,
Kiemi tahi vaan Kym, Kim. Eräs taru tahtoi johtaa nimen venäläisestä
Kem'istä, josta tänne muka olisi siirtolaisia muuttanut rikkaan
kalansaaliin tähden. Toinen taru taas tiesi kertoa eräästä saman-
nimisestä joesta "Hämeen ja Venäjän välillä", mistä tänne olisi
asukkaita tullut. Mutta verrattomana nimen johdatuksena seisoo
kuitenkin ijäti Ol. Rudbeck'in (vanhemman) selitys, että Kemi tulee
sanasta "Cimmeria", josta Kemi siis on yhtä kuin "Cimmerialaisten
maa."

Merestä aina korkeanlaiseen Kivalo-selänteesen saakka on Kemi

jokseen lakeata tasankomaata, joka helposti johtaa mieleen sitä
aikaa, jolloin kalat ihmisten sijassa tällä alalla liikkuivat ja aallot
lainehtivat Kivalovuoren juurilla. Tähän esi-historialliseen aikaan
kuuluu Jatulinletto, se on: Jättiläisten saari, joka nykyään on Kivalo-
selän korkeimpia kukkuloita, 3 penikulmaa meren rannasta. Näistä
ajoista muistuttaa myös Vammanvuori, joka on korkein
vuorenhuippu Tervolassa, noin 3 penikulman päässä Tervolan
kirkosta, ja jonka päältä on löydetty laivan runko. Tämän ikää on
satu niin korottanut, että jäte viedään aina mailman luomiseen asti
ja myöskin mainitaan itse "Noakin arkiksikin." Esimerkkinä
tapahtuvasta maatumisesta on myös Merikoski Simojoessa, mikä
koski nyt on neljänneksen matkan paikoilla merestä. Ja kun
Walmarin niemellä Kemijoen rannalla, noin neljänneksen matkaa
merestä, kaivo kerran kaivettiin, niin löydettiin maasta kahdeksan
kyynärän syvyydeltä hiiltynyttä puuta ja katajata. Myöskään ei ole
joen laskuhaara sama kuin ennen. Entinen on näet kuivennut ja on
nyt lampena.
 Ehk'ei kyllä nyt enään Kemissä löydy niin paljon muinaismerkkejä
kuin muissa tietävästi nuoremmissa pitäjissä, niin on hyvinkin
todennäköistä, ett'ei asian laita ole näin aina ollut. Sillä jos missään
niin on täällä nuorempi aikakausi hävittänyt vanhemman jälkiä. Tätä
todistaapi paraiten satuisa Kivalo, johon vanhimmat muistotarut
viittaavat. Muinoin sanotaan täällä olleen paljonkin muinaisjätteitä,
vaan nytpä niitä harvassa vaan siellä on.

Jatulinletosta, joka on Jättiläisten viimeinen asuntopaikka näillä

tienoin, on jo mainittu. Huomattava on, että Jättiläiset täällä Kemissä
ja Tervolassa, niinkuin haja-paikoin Simossakin, vaan ei muissa
pitäjissä, kutsutaan Jatuleiksi ja Jotuneiksi. Tälle pitäjälle
erinomaiseksi omituisuudeksi luettavana on myös sen rikkaus
senkaltaisista kivirakennuksista, joita yleisesti kutsutaan "Pietarin
leikeiksi", vaan täällä saavat nimen Jatulin-taraat, se on: Jättiläisten
aitauksia (= tarhoja). Näitä havaitaan hyvin tiheässä, ja suuruutensa
puolesta ne vaihtelevat kahden ja neljän jopa useammankin syllän
välillä. Ne ovat rakennetut päänkokoisista kivistä, jotka ovat yksi-
kertaisesti maan päälle laaditut. Mutta välistä on niitä laitettu
suuremmalla huolella hyvin matalain muurien tapaisiksi, enemmin tai
vähemmin kierteleväisiksi sokkeloiksi eli labyrinteiksi. Kun näitä,
meidän tietääksemme, ainoastaan on huvin ja leikin vuoksi
rakennettu, niin on heidän paljoutensa milt'ei merkillinenkin.
Jatulinletolla on tämmöinen ollut, samaten Ajos saarella,
Hannuksenvaaralla Tervolassa myös y.m. Tästä Hannuksenvaaran
aitauksesta mainitsee Warelius'kin — kirjoituksensa 62 s. — sanoen
tämän olevan kirkonsijan mukaisen. Halu teki siis tätä nähdä, mutta
kun ei aika sitä myöntänyt, toimitin itselleni siitä omin silmin
näkijältä seuraavan kuvaelluksen. Kertoja on Tervolan silloinen
kappalainen M.A. Kolström, joka jo ennenkin oli sen nähnyt, mutta
sitten vielä aivan asian tähden sinne matkusti, ja hyväntahtoisesti
siitä mulle näin kirjoitti: "Hannuksenvaaralla, joka on noin 3/4
penikulman päässä kirkonkylästä, olen käynyt jäännöstä katsomassa.
Se kutsutaan Jatulin eli Jotunin tarhaksi ja on ymmyrkäinen aitaus
suuremmista ja pienemmistä kivistä, noin 10 kyynärää poikki-
mitaten, melkeen kokonaan jo hajallansa. Pesosen isäntä kertoi
mulle tästä Jatulin tarhasta, että se vielä hänen nuorra ollessansa oli
ollut korkeampi sekä toisenlaiselta näyttänyt sisäpuoleltansa, ehkä
kyllä se jo silloin, hänen isänsä sanoman mukaan, olisi ollut suureksi
osaksensa hajonnut. Hän kertoi, että tarhassa oli ollut ovi ja useata
ympyrää ja kierrosta, niin että tottumaton näihin pian eksyi eikä
enään ovelle osannutkaan, vaan täytyi astua yli aitauksen. Kertoja
sanoi myös, että tarhan keskellä oli suurempi kivi, jonka alta he,
nuoruudessansa siellä kaivaen, olivat luupalasia löytäneet." Eräästä
toisesta, Wiianvaaralla olevasta, "Jatulin haudasta" kerrottiin mulle
myöskin Tervolassa. Mutta perille tultuani havaitsin sen olevan, ei
ihmisen, vaan luonnon tekemää. Se oli sammal-pohjainen soikea
hauta eli kuoppa, noin 15 syllän pituinen, ja 5 syllän levyinen —
ilmetty entinen lampi.

Lappalaiset ovat useitakin muistoja pitäjälle jättäneet. Joen

itäisellä puolella, vähän matkaa lauttaus-paikasta, Jakun-ojan
varrella, sanotaan heidän asuneen, mutta ainoastaan muutamia
kuoppia on siellä enään nähtävinä. Nimen ovat he taas antaneet
Lapin-ojalle, Wahtolan alapuolella. Kivalo-harjanteella on eräs
Tornivaara niminen vuorenkumpu, joka nähtävästi on se
Tornionvaara, jota Warelius mainitsee, sanoen siellä löytyvän
muinaisjätteitä "eri laatuisia kuin ennen kuvatut Lapin-rauniot, mutta
sittenkin kansan jutun mukaan Lappalaisten jättämiä." Sen
kuvauksen mukaan, minkä minä näistä sain omin silmin näkijältä —
sopimattoman ajan tähden en itse päässyt käymään Kivalolla — ovat
ne niin kutsutuita "Lapinhautoja", pieniä ympyrkäisiä kivikumpuja.
Ne ovat luvultansa kolme, ovat ennen olleet korkeammat, mutta nyt
on niitä hajoitettu, jott'ei ne enään ole kuin kyynärän korkeisia. Ne
kutsutaan myös torniksi, ja sanotaan antaneen kummulle nimen.
Täällä on myöskin kivistä maahan laadituita syvennyksiä, nämäkin
hajoitettuina, joita kutsutaan Jotulin hautoiksi. Penikat nimisillä
vuoren-kukkuloilla löytyy myöskin siellä täällä enemmin tai
vähemmin selviä Lapinraunioita. _Kokomaalla_kin on näitä n.k.
Jättiläisten hautoja ja syvennyksiä.

Simossa löytyy Lapinraunioita lukuisasti. Melkeen joka harjulla

niitä näkyy. Niin esm. kaikilla seuraavilla: Marostenmäki, Kiimavaara
(selvimpiä), Kirkkovaara, Waresharju, Näätämaa, Isoharju,
Peräharju, Ervastin talon taustalla ja Kirnuvaaralla. Mitä näihin
Kirnuvaaran raunioihin tulee, niin ovat niistä ne, mitkä ovat enemmin
mereen päin, tätä nykyä jo liian epäselviä. Mutta niistä, mitkä harjun
toisessa päässä näkyvät, ovat ainakin kaksi ulommaista varsin selviä.
Ne ovat kivi-ympyröitä, ulottuen 6-9 jalkaa poikkimittauksessa,
joiden seinät, 2-3 jalan korkeiset, ovat kukistuneet, muodostaen
sisässänsä kuopakkeen. Heidän pyörämuotonsa on johtanut kansan
mieltä näkemään niissä kirnuja, josta nimi Kirnuvaara. Seinien
ulkopuolella näkyy syvyyksiä, mitkä kenties ovat siitä tulleet, että
seinä-kivet ovat siitä otetut. Mereen päin vähenevät kuopakkeet yhä
enemmin sekä suuruutensa että selvyytensä puolesta, eivätkä nouse
maanpintaa ylemmäksi.

Mutta vaikka näitä "Lapin-raunioita" näinkin lukuisasti Simossa

löytyy, niin astuu sittenkin Tervola esiimme Lappalais-muistojen
oikeana pesänä. Täällä on Törmävuori Lappalais-kehinensä, täällä on
Wiianvaara Lappalais-jätteinensä, täällä on jo mainittu
Hannuksenvaara, jossa on samanlaisia muistomerkkejä, ja Joonin
vieressä myös näytetään Lappalaisten vanhoja asuntopaikkoja.
Vanhoissa kirjoissa kutsutaan koko Tervolan kappeli Lapinniemeksi,
pappila kutsutaan Lapinniemeksi, ja lavea suo, joka alkaa
puolentoista virstan päässä joesta ja siitä ulottuu Kivalovuoren
seutuihin saakka, kantaa nimen Lapinjänkä. Tässä suossa on
sammalen alta löydetty lautta, josta arvellaan, että Lappalaiset
muinoin olisivat sitä käyttäneet päästäksensä entisen järven toisesta
rannasta toiseen. Tällä jängällä on laskunsa Lapinojan kautta, joka
Jurvan edustalla jokeen purkauu, ja tämä Lapinoja tekee yhdessä
joen kanssa varsinaisen luonnollisen Lapinniemen. Vihdoinpa löytyy
vielä joessa, ihan vastapäätä Lapinniemen pappilata, eräs Seitasaari
niminen luoto, mikä luoto ennen on niemukkeena ollut, niinkuin sitä
vielä voipikin maan laadusta nähdä, ja on Seitaniemeksi kutsuttu.
Tällä saarella, kuten taru tietää mainita, ovat Lappalaiset muinoin
pitäneet käräjiänsä, ja siksi tarpeeksi oli tähän laadittu kiviä
ympyrään, joidenka kivien päällä Lappalaisten päämiehet ja
vanhemmat istuivat tässä tilassa. Keskellä ympyrää oli suurempi kivi,
jonka päällä jumalan-kuva eli seita seisoi. Näitä kiviä on vielä
meidänkin aikana nähty, mutta nyt ovat jo tulvat ne korjanneet ja
ihmiset vieneet tarpeisinsa, esm. kirkon alustukseen ja portaisin.

Kirkollisia muistoja on Kemillä paljon. Nykyinen kirkko on varmaan

neljäs, vaan luultavasti ainakin viides. Muuanna aikana tämän
vuosisadan alkupuolella seisoi täällä kolme kirkkoa yhtä haavaa,
ainoastaan kivenheiton matkaa toisistansa. Vanhin näistä, Kemin
kivinen kirkko, jota kansantarussa mainitaan Jättiläis-tekemäksi, on
rakennettu vuosina 1519-21 katoliseen rakennustapaan, ja näyttää
vallan muinaiselta.[6] Lattian alla säilytetään joukko ruumiita, joiden
seassa myöskin eräs Rungius nimisen miehen, joka täällä oli
pastorina 1610-29, ja jonka ruumis vieläkin on niin hyvästi säilynyt,
että se luullaan balsameeratuksi.[7] Kirkossa, joka ei enään ole
käytettävänä, löytyy seuraavat muinaiskalut: 1. "Ristiin naulitun
Kristuksen veisto-kuva, öljyväreillä maalattu ja kullattu, Kemin
nimismiehen Juhana Matinpoika Wilm'in lahjoittama." Tästä ei ole
jälillä muuta kuin rungon yläpuoli ja osa jalkoja; 2. Kristuksen
veistokuva, jolla vasemmassa kädessä on maan-pallo, oikea käsi
poissa, päässä kruunu. Kuva seisoo eläimen päällä, josta on vaikea
ratkaista, onko se leijona, karhu vai elefantti. Kanto-pylväänä on
paholaisen kuva; 3. Puolentoista kyynärän korkeinen kuva,
luultavasti pyhästä Henrikistä; 4. Neitsy Maarian kuva, jokseenkin
säilynyt pait värejä; 5. Lutherus'en kuva, kömpelö; 6. Vaimo-ihmisen
kuva (arvattavasti Maarian) lapsensa kanssa, vallan kömpelön-
näkyinen.

Ennen tätä kirkkoa seisoi toinen, puusta tehty, Walmarin niemellä,

missä sija vielä näytetään ja missä muutamat kuopat muistuttavat
entisestä hautausmaasta. Messenius'en mukaan olisi tämä kirkko
ollut rakennettu v. 1431. Mutta jos niin on, niin se ei suinkaan
olekaan ollut ensimäinen, koska arvellaan, että täällä kirkkoherra-
kunta jo vuonna 1248 pantiin toimeen. Kirkon mainitaan Venäläiset
polttaneen, minä vuonna, siitä ei tietoa. Luultavasti tapahtui se
"suuren venäläis-sodan" aikana 1473-1510? Samassa tilaisuudessa
sanotaan myöskin kirkonkellojen tulleen upotetuiksi siihen paikkaa
jokea, missä nyt, kun tämä haara on kiini mennyt, pikkuinen lampi
on.

E. Frosterus käsikirjoituksessaan "Breves Observationes ad

antiquitates Ostrobotniae" sanoo Kemissä olevan messinkisen
suitsutus-astian, jota Paavin-aikakautena käytettiin pyhää savua
poltettaessa. Tämä savu-astia on kadonnut, ja kenties samaa tietä
mennyt kuin muutkin vanhat hopeat. Näistä näet kerrotaan, että
niitä kerran sodan aikana kirkkovärti kaivoi vihollisilta piiloon
maahan, johonkuhun Wahtolan viereen, vaan että kirkkovärti kuoli,
eikä löytänytkään kukaan niitä enään. Vanhimmat kalut ovat
nykyään eräs kalkki[8] ja öylätti-lautanen, varustettu
munkkikirjaimilla ja koristuksilla ja vuosiluvulla: 1250. Pörhölän
pappilassa säilytetään myöskin eräs kupari-kalkki, jota ei enään
käytetä; tämän ikä ei kuitenkaan ole korkea, eikä sen muotokaan
muutoin tavaton.

Merkillinen on täällä niinkuin Iissäkin kulkeva taru eräästä

luostarista, joka muka näillä pohjoisilla seuduilla olisi muinoin ollut.
Paikkaa määrätessä, missä tämä luostari olisi seisonut, eivät sadut
kuitenkaan vedä yhtä. Mutta kun viimein saapi luostarin asetetuksi
Walmarin niemelle, niin joutuu asia selväksi: katolinen kirkko ja
luostari sekaantuvat helposti toisiinsa. Montaja on eräs saari Simon
ulkopuolella. Tämä kuuluu sekin paikkakunnan katolisiin muistoihin.
Tähänkin tahtoo näet taru asettaa luostarin. Mutta asian laita on se,
että Montaja vaan on ollut, johonkuhun luostariin kuuluva,
kalastuspaikka. Siihen luuloon, että luostari olisi Montajalla ollut, on
vaikuttimena myöskin ollut Munkkihiedan läheisyys, joka on maalla
aivan vastapäätä saarta. Tämä Munkkihieta on kummallinen luonnon
teos. Tasaiselle meren rannalle on veden ja tuulten tuoma hieta
pystyttäynyt monimutkaiseksi muuriksi, jonka korkeus veden pinnan
yli on noin 36 jalkaa. Merenpuolinen seinä on mereen päin
kallattava, mutta maanpuolinen äkkijyrkkä, 26 jalan korkeinen. Koko
hietarakennus on jonkunmoisen epämukaisen hevosenkengän
muotoinen, jonka leveys toisesta päästä toiseen on 120 jalkaa. Sivut
eli kengän santamuurit ovat 20-25 jalan levyisiä. Kummallinen on
muodostuminen, kummallista myöskin miten näin korkea
hiekkamuuri on kestänyt meren tuulia, ennenkuin siihen on puita
kasvanut, niinkuin nyt on. Tämän oudon hietamuodostelman nimestä
tietää taru kertoa, että se olisi siitä tullut, että munkkilaislaiva kerran
täällä olisi myrskyn valtaan joutunut, kärsinyt haaksirikkoa ja sitten
törmännyt hiekkaan, mihin se munkkinensa aarteinensa haudattiin,
ja haudan päälle sitten tämä hieta-pylväs noussut. Välistä näkyy
vielä muka sydän-yön aikana sininen liekki palavan aarteiden yli
santa-pylvään kukkulalla.[9]

Näistä vanhemmista ajoista on paikkakunnalla toinenkin muisto.

Se on ruotsalainen siirtokunta. Kemijoen länsi-puolella on koko kylä,
jonka nimenä on Liedakkala eli Sihtuna, ja joka on Ruotsin
Sigtunasta asujamensa saanut. Tästä tietääkin vielä kansan muisto
mainita, että tänne on asukkaita siirtynyt "Liedakkalan eli Sihtunan
kaupungista Ruotsin puolelta." Moniaat ruotsalaiset nimet vielä
muistuttavatkin tästä alkuperästänsä, esm. Gunnari. Myöskin
suomalaisia siirtoja on paikkakuntaan tehty, niinkuin sen
kielimurteista voikin päättää. Savo-karjalaisia alkuaineita havaitaan
täällä yhtä selvästi kuin hämäläisiä Torniossa. Varsinkin on ison vihan
jälkeen asutuksia näihin paikkoihin tapahtunut.

Isoon vihaan kuuluvat seuraavat muistot ja tarut: Venäläiset tulivat

kerran tapaansa myöten veneillä alas jokea kuljettaen näissä
ryöstämänsä rahat ja kalliit tavarat, joiden seassa myös kaksi
kirkonkelloa, mitkä he olivat Rovaniemestä ottaneet. He tulivat
Taivalkoskelle, joka on 6 neljännestä Kemin kirkosta. Kun tämä
mahtavan suuri koski heistä näytti kovin hirveältä heidän
laskeaksensa sitä alas omin neuvoin ilman laskijatta, menivät he
maalle hankkiaksensa itsilleen tämmöisen. Mutta kun oli kaikki mies-
puolinen suku lähtenyt pakoon, onnistuivat Venäläiset vaan saamaan
kiini erään ämmän, joka kerskaili laskemis-taidostansa. Hän otettiin
laskijaksi, vene sysättiin vedelle, akka perään. Mutta ämmä oli viekas
ja uskalias. Laskiessa viillätti ämmä kallion sivutse, hyppäsi itse
kalliolle ja jätti Ryssät menemään itsiksensä yhtä vauhtia vaan. Mitäs
muuta, kaikki Ryssät hukkuivat, kaikki tavara painui kosken pohjaan.
Täältä kuuluu vielä joskus kirkonkellot soivan. Mutta kallio, mihin
ämmä itsensä pelasti, kantaa siitä saakka nimen Ämmänpää. — Itse
tapauksen kertomisessa olivat tarut yksimieliset, vaikka eri
tarinoitsija siihen asetti eri seurakunnan kelloja. Niin arveli muuan,
ett'ei kellot olleetkaan Rovaniemen, sillä nämä kuuluvat olevan
Ounasjokeen uponneet, vaan Kemijärven. Toinen taas, että ne oli
Tervolan. Tämä tiesi lisätä, että ainoastaan Tervolan suurempi
kirkonkello olisi Taivalkoskeen painunut, sillä pienemmän, joka oli
niin halpa-arvoinen, ett'eivät Ryssät huolineet sitä kuljettaa, olivat he
Tervolassa upottaneet mutaiseen lähteesen, kosk'eivät saaneet
tahtonsa mukaan sitä rikki, ja tästä onkin lähde saanut nimen
Kellonlähde. — Mikkolan taloon, lähellä Tulkkilaa, liittyy myöskin
näistä ajoista muisto. Venäläiset olivat paikan ottaneet ja tähän
asettuneet. Mutta Suomalaiset, ruotsalaisen päällikön johdon alla,
piirittivät paikan, sytyttivät talon ja tappoivat viholliset. Yksi ainoa
pääsi jäälle pakoon, mutta saavutettiin sitten ja surmattiin hänkin.
Kun sittemmin kiuas hajoitettiin, oli sieltä löydetty rahoja, joita
Ryssät olivat sinne kätkeneet.[10] — Kostaaksensa Ryssäin julmaa
käytöstä olivat Suomalaiset kerran piirittäneet pirtin, missä Ryssiä
nukkui, ja tuleen kaikki polttaneet. — Sotisaari, johon Kemin uusi
kaupunki on määrätty rakennettavaksi, sanotaan saaneen nimensä
siitä, että Suomalaiset täällä — ison vihan aikana, sanoo juttu, vaan
luultavasti vuosien 1580-90 meteleissä — olivat odottaneet ja
sotineet niitä Venäläisiä vastaan, mitkä tulivat jokea alas. — Vallittu-
saari, josta en ole sen tarkempaa tietoa saanut, kantaa nimensä
luultavasti jonkun samankaltaisen tapauksen nojassa. — Kemistä
ryöstettiin muutoin kaksi kirkonkelloa, niin että seurakunta sittemmin
torvella kutsuttiin jumalanpalvelukseen.

Kuopasjärven vieressä Simossa, 10 penikulmaa jokea ylöspäin,

mainitaan löytyvän Ryssän hautoja;[11] nämä niinkuin muutkin
tänlaiset kuuluvat luultavasti 1808-9 vuosien sotaan. — Maassamme
jokseenkin yleinen taru Laurukaisesta sovellutettiin täällä samaan
Kuopasjärveen. — Laurukais-jutun mukaan kerrottiin täällä myöskin
eräästä vaimo-ihmisestä, jonka Ryssät olivat ottaneet ja vieneet
saareen. Täältä olisi hän kuitenkin pelastettu miehensä kautta, joka
tietysti ei Ryssille suonut kaunista ja nuorta vaimoansa. —
Kulkiessansa pitkin joen rantaa olivat Venäläiset myöskin tulleet Yli-
Kärppälään, mistä asukkaat heti lähtivät pakoon, antaen Ryssäin,
ilman estämättä, ryöstää kaikki tavarat. Mutta kun olivat Venäläiset
lähteneet tiehensä, pisti asukkaita heidän oma pelkoisuutensa
vihaksi, ja he päättivät ajaa vihollisia takaa, saadaksensa, jos
mahdollista, omaisuutensa takasin. Nyt tarjouupi sokea ukkokin
seuraan. Häntä ei tahdottu ensin ottaa mukaan, koska mietittiin,
ett'ei sokea juuri mitään hyötyä voi tehdä, pikemmin vaan on
esteeksi. Mutta ukko ei luopunut tarjoumisestaan, ja vihdoinpa hän
otettiinkin. Lähdettiin nyt vihollisia ajamaan, ja tavattiinkin heitä yön
aikana. Nuoriso tahtoi heti tehdä päälle-rynnäkön. Mutta vanha
sokea kysyi, miltä se vihollisen makuupaikka näytti. Vastattiin, että
lukuisia, suuria tulia paloi, ja tulten ääressä ihmisiä liikkui. Vanhus
silloin kielsi rynnäköstä. Ajan perästä hän uudisti kysymyksensä ja
sai vastaukseksi, että tulet olivat pienenneet, liike samaten. Vieläkin
esti hän ryntäämästä. Sitten kysyi hän taas kolmannen kerran, ja
kun hälle nyt vastattiin, että tulet olivat ihan sammumaisillaan eikä
ihmisiä enään näkynytkään liikkuvan, sanoi hän oikean ajan tulleeksi,
sillä nyt nukkuivat viholliset tuhon unta. Suomalaiset karkasivat nyt
vihollisten päälle, tappoivat ne ja ottivat omaisuutensa takaisin.[12]

Tervolassa ei löydy, pait mainittua juttua Kellonlähteestä, ison

vihan aikuisia muistoja. Luultavasti oli paikkakunta niin harvassa
asuttu, ett'ei vihollisilla ollut juuri erinomaista täällä saatavana.
Täälläkin, niinkuin Kemissä ja Simossa, löytyy kuitenkin, varsinkin