ISSN 1068-798X, Russian Engineering Research, 2007, Vol. 27, No. 8, pp. 513–518. © Allerton Press, Inc., 2007.

Original Russian Text © Yu.G. Kabaldin, S.N. Murav’yev, 2007, published in Vestnik Mashinostroeniya, 2007, No. 8, pp. 46–50.

Information Models of Self-Organization and Fatigue Damage

of Metallic Materials
Yu. G. Kabaldin and S. N. Murav’yev
Komsomol’sk-on-Amur State Technical University
DOI: 10.3103/S1068798X07080072

There are various views at the present time on the the structural transitions and what controls the struc-
mechanisms of the structural transformations in the tural changes; is the order parameter the information
metallic materials. In several works, the structural carrier?
transformations are examined from the viewpoint of In this connection, it is necessary to note the impor-
synergetics theory as a nonequilibrium phase change tance of the studies that have been performed at the
with the forming of dissipative structures as a conse- Institute of Metallurgy and Machine Science of the
quence of self-organization [1–10]. Thus, in [11] the Russian Academy of Sciences under the guidance of
structural transformations are examined from the view- V.S. Ivanova [1, 4–6], in which the structural informa-
point of the forming and movement of the partial discli- tion is proposed as the measure of the orderliness of the
nations. In [12] the forming of the substructure is asso- system upon disruption of its symmetry. The technique
ciated with the polygonization processes. According to of multifractal parametrization of the materials has
the studies [9, 10], the structural transformations in the been created on the basis of this approach. It is well
defect subsystem are controlled by the scalar density of known that the fractals also have informational proper-
the partial dislocations and the accumulated strain ties. The connection of the fractal dimension with the
energy. It is noted in [3] that the structural changes in mechanical properties of the materials was shown in [1,
the defect subsystem may be determined by the form- 7, and others], and a new interdisciplinary scientific
ing of the collective mode carriers (structons), which direction was created (fractal materials science).
causes the “crystallization” of the material of the sub-
We should also note the extremely fruitful idea of
boundaries upon reaching the critical density of the
V.E. Panin [8] on the role of information in the struc-
chaotic dislocations.
tural transformations in the highly excited state of the
In accordance with the principles of synergetics the- crystals, in accordance with which there is embedded in
ory, self-organization, i.e., the transition from the the electron structure of the metals and alloys a genetic
unstable state to the stable state, is accompanied by code. This code determines the structural transforma-
decrease of the entropy production [2]. In this case the tions in the highly excited state of the crystals and the
open system exchanges both energy and information appearance of the new permitted states and degrees of
with the external medium. The founder of synergetics freedom. In accordance with Haken’s classification
theory, H. Haken [13], analyzed the changes of the [13], this form of information should be termed genetic.
information (the Shannon information entropy Su) in However the open systems interchange information
the nonuniform phase transition. It was shown that in with the environment, and the exchange of the genetic
this case the system acquires the ability to store infor- information that is embedded in the system with the
mation. Therefore Su can characterize the orderliness information of the environment leads to the creation of
(stability) of the system upon disruption of its symme- new information.
try. However, the role of the information has not been The prediction of the failure of materials and of their
studied in the literature on the basis of the principles of remaining service life is possible only on the basis of
synergetics theory in the case of the structural transfor- the development of the promising new techniques for
mations in the defect subsystem of the materials. From evaluating in real time the structural changes in the
the viewpoint of synergetics theory, it is important to materials that precede their failure. One such method is
determine which parameter in the system is the order acoustic emission (AE). It is well known that the AE
parameter that gives rise to the collective behavior of signals carry information on the forming and the move-
the dislocations. Several other questions arise when ment of the dislocations, on the correlated collective
examining the role of the information in the structural effects, and so on [14]. Thanks to the rapid develop-
transformations, specifically: how do the individual ment of computer technology and the new methods for
dislocations transfer the information (signal); what is analyzing the nonstationary signals (wavelet analysis
the mechanism of their coordinated (cooperative) inter- [15–17] and the neural networks), it is possible not only
action; how is the information stored in the course of to analyze the AE signals in real time but also to model

513
514 KABALDIN, MURAV’YEV

special setup that makes it possible to conduct bending

tests of the flat specimens with a clamped end. The free
end of the specimen is subjected to loading in alternate
bending. The loading of the specimen is accomplished
with the aid of a crank–rod mechanism that is driven by
a dc motor. The system for creating the bending oscil-
lations by the crank-rod mechanism is located directly
on the motor shaft. It is possible to regulate the ampli-
tude of the oscillations by varying the position of the
crank roller. The maximal amplitude of the oscillations
is 20 mm. The use of the dc motor makes it possible to
vary the oscillation frequency up to 30 Hz.
To eliminate the influence of mechanical noise on
(‡) the acoustic recording system, a portion of the setup is
isolated by rubber damping pads. The specimen is also
attached to the setup through an isolating pad, which
prevents transmission of the high-frequency noise that
arises as a consequence of the friction of the specimen
with the mobile clamp. The test specimens are plates of
thickness 2 mm, with the working section length
70 mm. The specimens were fabricated by contour
milling. The width of the widest part of the specimens
is 8 mm. Specimens of steel 20 and the OT-4 titanium
alloy were used in the studies.
The experimental technique was as follows. The
desired amplitude of the oscillations was established on
the basis of the maximal design stress in the specimen.
The selected oscillation frequency was constant for
(b) each series of experiments. A piezoelectric transducer
was installed on the specimen in order to record the
acoustic signals in the course of the fatigue tests of the
materials. The pickup was mounted directly on the
specimen. The signal from the pickup was fed to a
preamplifier and then to the analog-digital converter of
the computer. The acoustic signals arising in the course
of the fatigue damage of the material of the specimens
were also recorded on the computer. Then the AE sig-
nals were processed with the aid of specially developed
software that made it possible to analyze the fatigue
damage of the materials on the basis of fractal analysis
and wavelet analysis, and also made it possible to cal-
culate the Shannon information entropy Su. In the frac-
tal analysis of the AE signals, we used as the criterion
(c) of the stability of the structural state of the material the
fractal dimension DF. Wavelet analysis made it possible
Fig. 1. Electron photomicrographs illustrating the deforma- to evaluate this criterion in real time.
tions with cyclic loading of steel 20 specimens: (a) domi- A model JEM-7A electron microscope was used for
nant deformation in the near-boundary volumes (×450); the electron-microscope studies of the thin foils. We
(b) stable slip bands (×450); (c) microcracks (×800).
studied the steel 20 specimens after their fatigue load-
ing. The damages of the surface layers in the form of
them, examining them as time series, and also to slip bands and microcracks were studied on a model
develop the new structural and informational criteria JSM-3U scanning electron microscope with the various
that are adequately associated with the processes of the loading cycles.
structural transformations and the failure of the materi- Figure 1 presents the photomicrographs of the struc-
als. The present article is devoted to the resolution of tural changes in the steel 20 specimen on the macro-
these important problems. and microlevels. We see in Fig. 1a that in the initial
For the study of the fatigue damage of the materials stages of the cyclic hardening of the materials there
with the application of the AE method, we have used a forms a specific relief in the form of depressions and

RUSSIAN ENGINEERING RESEARCH Vol. 27 No. 8 2007

 INFORMATION MODELS OF SELF-ORGANIZATION 515

(‡) (b)

(c) (d)

Fig. 2. Electron photomicrographs illustrating the changes in the deformed volumes of steel 20 specimens: (a) dislocation structure
at the ferrite grain boundary in the microyielding stage (×14000); (b) chaotic dislocation structure in the ferrite grains (×14000);
(c) cellular dislocation structure in the strain hardening stage (×14000); (d) fragmented dislocation structure (×14000).

protrusions, indicating predominant deformation in the changes in the specimens of the OT-4 alloy that were
near-surface volume of the material. In the course of reconstructed from the AE signals with the various
increase of the number of loading cycles, there form in numbers of loading cycles. Figure 4b presents the
the favorably oriented ferrite grains slip lines, and then dependence of the Shannon information entropy Su on
stable slip bands (Fig. 1b) and microcracks (Fig. 1c). It the number of loading cycles. Similar relations were
was established by the electron-microscope study that obtained in the cyclic loading of the specimens of steel
in the course of the cyclic loading process the forming 20 [18].
of the dislocations is concentrated near the boundaries
of the ferrite grains (Fig. 2a). Their formation takes These studies made it possible to identify our pro-
place in the microyielding stage. With increase of the posed criteria (the fractal dimension and the informa-
number of the loading cycles, there forms in the ferrite tion entropy) for evaluating the structural state of the
grains in the strain hardening stage a chaotic (Fig. 2b) deformed materials in real time and compare this with
and then cellular dislocation structure (Fig. 2c), which the results of the electron-microscope studies of the
with increase of the number of the loading cycles tran- structural transformations, and the micro- and mac-
sitions to the fragmented (Fig. 2d) and then banded rodamages of the materials. There was established a
structure. correlation between the changes of the forms of the
AE signals in time and the multilevel nature of the
Figure 3 presents the photomicrographs of the struc- structural changes that are caused by the collective
tural changes in the grains of the OT-4 titanium alloy. motion of the defects with the forming of the various
Here there also form the slip lines (Fig. 3a), transitioning types of substructure (see Fig. 2), and the forming of
to the stable slip bands (Fig. 3b) with the forming of the stable microdamage and macrodamage slip bands
microcracks in the grains and along the grain boundaries. (see Figs. 1b and 3b).
Figure 4a presents the dependence of the fractal We investigated on the basis of the electron-micro-
dimension DF of the AE signals in the cyclic loading of scope studies, the fractal analysis of the AE signals, and
the OT-4 titanium alloy, and also shows the attractors the reconstruction on this basis of the attractors (see
(indicated by the numerals 1 to 6) of the structural Fig. 4a) the mechanisms of the structural changes in the

RUSSIAN ENGINEERING RESEARCH Vol. 27 No. 8 2007

516 KABALDIN, MURAV’YEV

lishing in the defect subsystem of an order parameter,

which thanks to the subordination principle performs
the role of an informator and controls the structural
transformations from the chaotic dislocation structure
that forms in the initial loading cycles to the cellular
structure and later to the fragmented structure and the
banded structure with the subsequent forming of the
microcracks.
We developed on the basis of these studies the
mechanism of the influence of the information on the
self-organization process in the defect subsystem and
the appearance of the collective effects. The essence of
(‡) this mechanism is as follows. Upon the forming of a
dislocation, a local internal stress is created and it radi-
ates a signal (which can be recorded as an AE signal),
i.e., it transmits a certain amount of information that is
summed in the course of the forming of the other dislo-
cations. The result is the forming of an internal
mechanical stress field as a consequence of the highly
coordinated interaction of the dislocations, which with
their density p ≈ pcrit, i.e., with the critical parameters in
the subsystem, leads to the collective effects. Thus, the
internal mechanical stress field is both an order param-
eter and an informator: on the one hand, it is created by
the dislocations and, on the other hand, it controls the
dislocations on the basis of the subordination principle,
(b) imparting to the dislocations the information (the sig-
nal) with pcrit for their rearrangement into a structure
Fig. 3. Electron photomicrographs illustrating the structural with more information and, consequently, into the more
changes with cyclic loading of OT-4 alloy specimens: (a) stable (ordered) cellular structure (see Fig. 2c).
slip lines in the grain (×350); (b) stable slip bands (×500).
In the course of the subsequent substructure trans-
formations (nonequilibrium phase changes), the role of
defect subsystem of the materials. It was established the information is retained and even increases with the
that the substructure changes in the fatigue tests are aid of the resulting new dissipative (fragmented) struc-
associated with the chaotic nature of the processes in tures, while the preceding structure, transmitting the
the defect subsystem and are accompanied by change information to the new structure, determines in a signif-
of the shape of the attractor from its limit cycle to an icant measure its effectiveness as a result of self-orga-
n-dimensional torus. The limit cycle corresponds to the nization in the defect subsystem. Consequently, the
chaotization of the dislocation structure (points 1 and 2 defect subsystem is nonlinear, which leads to the peri-
in Fig. 4a) in the initial loading cycles (the initial loss odic, chaotic state in it and the structural transitions to
of structural stability—bifurcation), and thereafter its a more stable state upon breakdown of the preceding
temporal evolution is also realized as a result of the cha- coherent structures. The ability of the defect subsystem
otic nature of the processes in the defect subsystem, to accumulate information in the course of the struc-
which confirms the growth of the dimension of the tural transformations indicates that is has memory. It
attractors from DF ≥ 1 to DF ≤ 3 (points 4, 5, 6 in follows from the foregoing analysis that the informa-
Fig. 4a). The growth of the fractal dimension also indi- tion may be the carrier of the collective self-organiza-
cates that in the defect subsystem there takes place both tion modes.
periodic temporal organization and also periodic spatial We should also note that the dimension of the cells
organization (ordering) of the structure as a result of the in the substructure is a quantity of the order of 2–3 nm.
increase of the number of the degrees of freedom. Therefore we can classify the structural transforma-
The analysis of the electron-microscope studies, the tions in the defect subsystem as nanostructuring.
fractal and wavelet analysis, and the evaluation of the The informational self-organization mechanism
information entropy in the analysis of the AE signals in agrees well with the definition of the fractal as consist-
the course of the fatigue tests showed that the evolution ing of parts that are similar to the whole. In other words,
of the defect structure is associated with the accumula- the fractals have scale invariance. We can presume that
tion of both entropy and information in the system with the information can also act as a carrier of the collective
the aid of the dissipative structures in the cyclically self-organization modes on the various structural
deformed volumes. This is accompanied by the estab- (scale) deformation levels.

RUSSIAN ENGINEERING RESEARCH Vol. 27 No. 8 2007

 INFORMATION MODELS OF SELF-ORGANIZATION 517

DF
3.0 (a)
5 6
2.5
4
2.0
3
1.5 1 2

1.0
0 10 20 30 40 50 60
N, 104 cycles
Structural change attractors
1 2 3 4 5 6

Su, nat
5.5
(b)
5.0

4.5

4.0

3.5

3.0

2.5
0 1 2 3 4 5 6
N, 104 cycles

Fig. 4. Dependences of fractal dimension DF (a) and information entropy Su (b) on the number N of loading cycles of the OT-4
alloy: cyclic load σcy ≈ 282 MPa, frequency f = 22.5 Hz.

These studies also made it possible to suggest the result of the exchange of information and the coordi-
informational mechanism of the genesis of the fatigue nated action between the structural elements (the grains
microcracks as a result of the sequential, self-consistent and the grain junctions). In accordance with the subor-
(cooperative) structural changes on the various infor- dination principle, the system transitions to a new and
mation levels. On the lowest level, the transition of the more stable state with the forming of dissipative struc-
chaotic dislocation structure to the cellular and frag- tures (microcracks). This period is defined by the
mented structure is controlled by the internal mechani- degree of the strain hardening and the degree of the
cal stress field, which is the order parameter and the residual plasticity capacity of the material.
informator. Upon the formation of the slip bands in the The developed informational mechanism of the self-
favorably oriented grains and the transition to the organization process in the defect subsystem also made
higher structural deformation level, the accumulated it possible to model the AE signals on the basis of the
information is transferred to the adjacent grains or to neural networks and thereby predict the fatigue failure on
the near-surface deformation zone with the forming of the basis of the information criterion (the Shannon infor-
the stable slip bands in the grains or the creation of mation entropy). Figure 5 presents the dependence of Su
microcracks at the junctures of the grains, i.e., as a on the number N of the loading cycles in modeling the

RUSSIAN ENGINEERING RESEARCH Vol. 27 No. 8 2007

518 KABALDIN, MURAV’YEV

Su, nat 3. Lokhachev, V.A., Panin, V.Ye., et al. Kooperativnye

5.0 deformatsionyye protsessy i lokalizatsiya deformatsii
(Cooperative Deformation Processes and Deformation
4.5 Localization), Kiev: Naukova Dumka, 1989.
Su = 4.25
4. Vstovskii, G.V., Kolmakov, A.G., and Bunin, I.Zh.,
4.0 Vvedenie v multifraktal’nuyu parametrizatsiyu struktur
materialov (Introduction to Multifractal Parametrization
3.5 of the Material Structure), Moscow-Izhevsk: Nauchn.-
3.0 Izd. Tsentr “Regulyarnaya i Khaoticheskaya Dinamika”,
Experiment 2001.
2.5 5. Terentyev, V.F., Evolution of Structure under Fatigue of
Prediction
Metal as a Result of Self-Organization of Dissipative
2.0 Structures, in: Sinergetika i ustalostnoye razrushenie
0 1 2 3 4 5 6 7 8 metallov (Synergetic and Fatigue Fracture of Metals),
N, 104 cycles Nauka, 1989, pp. 78–87.
6. Terentyev, V.F., Ustalostnaya prochnost’ metallov i
Fig. 5. Dependences of information entropy Su on the splavov (Fatigue Strength of Metals and Alloys), Mos-
number N of loading cycles in experimental studies and in cow: Imtermed Engineering, 2002.
neuronet modeling of the AE signals (σcy ≈ 282 MPa, f =
22.5 Hz). 7. Kolmakov, A.G., Use of the Systems Approach Concepts
in Studying the Structure and Nature of the Plastic
Deformation and Failure of Metals, Metally, 2004, no. 4,
pp. 98–107.
AE signal, and also the results of the experimental stud-
8. Panin, V.E., Fundamentals of Physical Mesomechanics,
ies of the OT-4 titanium alloy specimens. The studies Fiz. Mezomekh., 1998, vol. 1, no. 1, pp. 61–81.
showed that the error of the prediction of the fatigue fail- 9. Koneva, N.A., Lychagin, D.V., Grishkina, L.I., and
ure does not exceed 15 to 20%, which is quite accept- Kozlov, E.V., Accumulation of Defects, Accumulated
able. Consequently, having the results of the fatigue tests Elastic Energy, and Self-Organization of Substructure,
with certain loading conditions (stress and frequency) in: Fizicheskie aspekty prognozirovaniya razrusheniya I
along with the recording of the AE signals, we can sim- deformirovaniya geterogennykh materialov (Physical
ulate on their basis the new AE signal with the loading Aspects of Forecasting of Fracture and Deformation of
conditions that will ensure the required service life. Heterogenic Materials), Leningrad, 1987, pp. 20–35.
10. Koneva, N.A., Teplikova, A.A., Sosnin, O.V., et al., Dis-
Thus, the developed new criteria for evaluating the location Substructures and Their Transformation With
structural state of the materials and the information Fatigue Loading, Izv. Vyssh. Ucheb. Zaved. Fiz., 2002,
mechanism of the self-organization process in the no. 3, pp. 87–98.
metallic materials make it possible to predict their 11. Rybin, V.V., Bol’shoie plasticheskie deformatsii i
fatigue strength in real time. The studies showed that razrushenie metallov (Large Plastic Deformations and
the proposed self-organization mechanism makes it Fracture of Metals), Moscow: Metallurgiya, 1986.
possible to resolve with success another urgent problem 12. Trefilov, V.I., Mil’man, Yu.V., and Firsov, S.A., Fizi-
of materials science—the development of new materi- cheskie osnovy prochnosti tugoplavkikh metallov (Phys-
als with specified properties. ical Basics of the Strength of Refractory Metals), Kiev:
Naukova Dumka, 1975.
An important direction in materials science is the 13. Haken, H., Information and Self-Organization : A Mac-
hardening of the materials (parts) to reduce the local- roscopic Approach to Complex Systems, New York:
ization of the deformations with cyclic loading that lead Springer-Verlag, 2006.
to the forming of microcracks. In this connection the 14. Braginskii, A.P., On Forecasting of Structural Restruc-
possibility for rational increase of the resistance of the turing in Materials Using Specific Features of Collective
materials to failure in the plastic deformation processes Behavior of Acoustic Emission Sources, in: Fizicheskie
(extrusion, rolling, forging, and so on) becomes evi- aspekty prognozirovaniya razrusheniya i deformiro-
dent. Specifically, by imparting to the structure of the vaniya geterogennykh materialov (Physical Aspects of
material in the hardening process the information that is Forecasting of Fracture and Deformation of Heterogenic
needed for its transition to a more stable state. Materials), Leningrad, 1987, pp. 55–75.
15. Vorobyev, V.I. and Gribunin,V.G., Teoriya i praktika
veivlet-preobrazovaniya (Theory and Practice of Wave-
REFERENCES let Transform), St. Petersburg, VUS, 1999.
16. Dobeshi, I., Ten Lectures on Wavelets, Philadelphia:
1. Ivanova, V.S., Oksagoev, A.A., Zakirnichnaya, M.M., SIAM, 1992.
and Prutskov, M.E., Optimizing the Structure of 17. Davydov, A.V., Veivletnye preobrazovaniya signalov:
Machinebuilding Materials, Metal. Mashinostr., 2002, Tematicheskie lektsii (Wavelet Transforms of Signals:
no. 6 (9), pp. 18–24. Topical Lectures), Yekaterinburg: UGGU, IGiG, and
2. Nicolis, G.and Prigogine, I., Self-Organization in Non- GIN, Fund of Electronic Documents, 2004.
equilibrium Systems: From Dissipative Structures to 18. Kabaldin, Yu.G. and Murav’yev, S.N., Predicting the
Order through Fluctuations, New York: John Wiley, Fatigue Strength of Materials on the Basis of Artificial
1977. Intelligence, Metal. Mashinostr., 2004, no. 6, pp. 29–39.

RUSSIAN ENGINEERING RESEARCH Vol. 27 No. 8 2007