Googlepreview PDF

Ultrasonic Treatment of Light Alloy Melts
Ultrasonic Treatment of Light Alloy Melts
G.I. Eskin
All-Russia Institute of Light Alloys

Moscow, Russia
Boca Raton London New York
CRC Press is an imprint of the

Taylor & Francis Group, an informa business
First published 1998 by Gordon and Breach Science Publishers
Published 2018 by CRC Press

Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 1998 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
ISBN 13: 978-90-5699-042-8 (hbk)
This book contains information obtained from authentic and highly regarded sources.
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, please access
www. copyright.com (http://www.copyright.com/) or contact the Copyright Clearance
Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a
not-for-profit organization that provides licenses and registration for a variety of users. For
organizations that have been granted a photocopy license by the CCC, a separate system of
payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered

trademarks, and are used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at

http://www.crcpress.com
British Library Cataloguing in Publication Data
Eskin, G.I.
Ultrasonic treatment of light alloy melts
1. Light metal alloys 2. Ultrasonics in metallurgy
I. Title
669'.028
CONTENTS
PREFACE ix
INTRODUCTION xi
1 FUNDAMENTALS OF MELT SONICATION 1
1.1 Ultrasound and Low-Frequency Vibration 1
1.2 Propagation of Power Ultrasound in Melts 9
1.3 Acoustic Cavitation in Liquid Metals 18
1.3.1 Cavitating Rigidity of Liquids 19
1.3.2 Cavitation Conditions for Melting Metals 22
1.3.3 Non-metallic Inclusions and Cavitation Nuclei 25
1.3.4 Cavity Dynamics 28
1.3.5 Sonoluminescence in Melts 39
1.3.6 Experimental Study of Cavitation in Light
Alloys 41
1.3.7 Cavitation Field in Melts 51
1.4 Acoustic Flows 55
1.5 The Sonocapillary Effect 60
2 ULTRASONIC DEGASSING OF LIGHT ALLOYS 65
2.1 Cavitation and Degassing Nuclei. The
Degassing Mechanism 65
2.2 Ultrasonic Degassing in a Stationary Volume 87
2.3 Ultrasonic Degassing in Flow 96
2.4 Ultrasonic Homogenization of Light Alloys
as an Alternative to Heat Treatment 101
2.5 Effect of Ultrasonic Refining on the Quality
of Castings and Extrusions 103
3 FINE FILTRATION OF ALUMINUM ALLOYS
IN THE CAVITATION FIELD 111
3.1 The Mechanism of Ultrasonic Filtration of
Melts 111
3.2 Filtration with Screen Filters 116
v
VI Contents
3.3 Filtering through Multilayer Screen Filters 122
3.4 Effect of the Uzfirals Process on the
Properties of Cast and Deformed Metal 132
4 SOLIDIFICATION OF LIGHT ALLOYS IN AN
ULTRASONIC FIELD 135
4.1 Mechanisms of Dynamic Solidification of
Metals and Alloys 135
4.2 Cavitation-Driven Solidification of Light
Alloys 144
4.3 Nondendritic Solidification in an Ultrasonic
Field 166
5 CONTINUOUS CASTING OF LIGHT ALLOYS
IN THE ULTRASONIC FIELD 187
5.1 Nondendritic Crystallization Front 187
5.2 Continuous Ultrasonic Casting of Aluminum
Alloys 195
5.3 Continuous Ultrasonic Casting of Magnesium
Alloys 201
5.4 Effect of Refined Grains on Deformed
Semi-finished Products 206
5.5 Effect of Nondendritic Structure on
Deformation in the Solid-Liquid State 221
6 CONTINUOUS CASTING OF NATURAL
COMPOSITES BASED ON LIGHT ALLOYS 227
6.1 Bulk Solidification in a Field of Ultrasound 227
6.2 Effect of Ultrasound on Solidification of
Intermetallic Compounds 229
6.3 Effect of Ultrasound on the Solidification of
Hypereutectic Silumin 240
7 ULTRASONIC MOLD CASTING OF
ALUMINUM ALLOYS 245
7.1 Arrangements for the Insonation of a
Solidifying Melt in Molds 246
7.2 Effect of Cavitation on Solidifying Melt in
the Mold 249
7.3 Structure and Properties of Shaped Castings 251
Contents vii
8 RAPID ULTRASONIC SOLIDIFICATION
OF ALUMINUM ALLOYS 255
8.1 Basic Sonication Schemes for Rapid
Solidification 257
8.2 Formation of Nondendritic Structure in
Rapid Solidification 259
8.3 The Properties of Deformed Semi-finished
Products from Nondendritic Pellets 261
9 ULTRASONIC ENHANCEMENT OF
ALUMINUM ZONE REFINING 265
9.1 General 265
9.2 Regulations of Ultrasonic Zone Refining 268
9.3 Effect of Ultrasound on Recrystallization in
the Moving Melt 278
9.4 Structure and Characteristics of Deformed
Acoustically Refined Aluminum 281
10 INDUSTRIAL APPLICATIONS 287
10.1 A Historical Overview 287
10.2 Optimal Melt Sonication Schemes 290
10.3 Oscillating Systems 295
10.4 Ultrasonic Thyristor Generators 302
10.5 Control of Melt Sonication Parameters 304
10.6 Ultrasonic Equipment for Continuous and
Mold Casting 309
10.6.1 Ultrasonic Melt Degassing in the Furnace 309
10.6.2 Filtering and Degassing in Continuous
Casting of Aluminum Alloys 310
10.6.3 Ultrasonic Equipment for Light Alloy
Insonation in Molds 311
CONCLUSION 319
REFERENCES 323
INDEX 333
PR EFA CE
This book contains the results of studies we conducted in the field of

applying power ultrasound to light alloys over a period of some forty
years. It would appear to be the first attempt ever made in the liter
ature on the subject to formulate the main regularities of the effect of
ultrasonic treatment on the processes of refining (degassing and filtering)
and solidification (formation of a grain structure), as well as to system
atically trace the hereditary effect of the ultrasonic treatment of melts
on the structure and properties of their ingots and deformed products.
In inoculated light alloy melts, ultrasonic treatment creates a new
type of cast structure with extremely fine nondendritic grains. This ef
fect is produced across a wide range of freezing rates and technologies
from slates to large ingots. This book presents a systematic exposi
tion of the nondendritic crystallization of light alloys in a field of power
ultrasound and the effect of this ultrasound on the properties of cast
and deformed metal.
The main proportion of the author’s studies covered in this book were
conducted in the All-Russia Institute of Light Alloys and at smelters
and metal works involved in the production of deformed semi-finished
products and mold castings from light alloys.
The assistance and support of B.I. Bondarev, Director General of the
All-Russia Institute of Light Alloys in conducting these experiments is
gratefully acknowledged.
It is my pleasure to express special thanks to Professor V.I. Dobatkin
for the many years of his creative cooperation in the studies whose results
are used in this book.
My thanks also go to my colleagues, together with whom I did the
research generalized in this volume. The reader can find their names in
the bibliography.
And last but not least, I should like to thank Dr D.G. Eskin for his
assistance in the preparation of this manuscript.
IX
IN T R O D U C T IO N
Metallurgy and metal science have made considerable achievements in

the development of light alloys for parts and assemblies in many diverse
fields from construction engineering, oil production, and the automotive
industries to aerospace technologies. The requirements imposed on the
quality of products made from these alloys call for constant improvement
in alloy technologies, above all in melting and casting.
Most metallurgical defects revealed at later production stages, for
example, slivers in deformed metal and welded joints, pores and blow
holes in castings, owe their existence primarily to an insufficient
purification of the molten metal from solid and gaseous inclusions. Crack
ing and slivers associated with extrusion, welding, and thermal treat
ment can be substantially reduced by refining the grain structure of the
initial cast form. The grain refinement of castings and ingots has a
decisive effect on the casting process itself. Finally, fine grains and a
high density of cast metal are inherited by the deformed product which
is endowed with the enhanced lifespan required by system designers.
Ultrasonic treatment is a new effective dynamic method for treating
a molten and solidifying metal. At a certain intensity sonication induces
acoustic cavitation in the liquid metal, which facilitates refining by the
removal of hydrogen and oxides. This treatment favorably affects the
quality of mold castings and extrusions prepared for welding. The effect
of developed cavitation on the solidifying metal enables one to radically
change the kinetics of crystallization and to obtain extremely fine grains
comparable in size with the cross section of a dendritic branch. Ingots
and mold castings of aluminum alloys with a highly refined nondendritic
structure demonstrate a higher density and the extremely fine dimen
sions of the secondary phases. These properties improve plasticity in the
cast (annealed) and homogenized state without a loss of strength. The
nondendritic structure of a cast form allows a thixotropic deformation in
the solid-liquid state, which recovers the hidden reserves of plastic flow
and thus sharply reduces the required extrusion pressure.
The idea of improving the quality of cast metal by means of elastic
oscillations, produced by mechanical vibration or shaking applied to
solidifying steel, was suggested by Chernov, the outstanding Russian
metallurgist, in 1878. However, the effect of ultrasonic oscillations on the
structure and properties of materials had not attracted the attention of
xi
xii Introduction
researchers until the 1920s. Boyle (1922) and Taylor and Sproule (1929)
reported the possibility of degassing light metals by ultrasonic oscilla
tions. Wood and Loomis (1927) published the results of their studies
with power ultrasound. They excited a quartz plate from a 2 kW gener
ator at a frequency of 200-500 kHz and studied the effect of these oscilla
tions on such processes as dispersion, emulsification, and degassing. The
first experimentally observed effect of ultrasound on the solidification of
molten metals is usually dated back to 1936 when Sokolov reported his
experiments with sonication of molten zinc, tin, and aluminum. Seeman
(1936) and Schmid and Eret (1937) repeated Sokolov’s experiments on
the ultrasonic crystallization of antimony, cadmium, duralumin, and si-
lumin. In the early 1930s, investigations into the sonication of melts with
elastic oscillations continued in three directions: studies of the effect of
various frequencies with the aim of establishing the mechanism of birth
and growth of crystallization nuclei in supercooled liquids such as melts
and solutions, studies of the structure and properties of pure metals and
alloys subjected to low frequency vibrations, and finally, studies of the
effect of ultrasonic oscillations on molten metals.
In the field of ultrasonic treatment of melts, researchers from the
former Soviet Union occupy a leading position. The investigations of
Ukrainian physicists conducted before W.W. II and studies of the team
headed by Danilov (1956) established the mechanism by which ultra
sound affects the birth and evolution of crystal nuclei in supercooled
melts of betol, piperine, salol, and so on. These groups used the method
of ‘development’ of crystal nuclei invented by Tammann (1931).
Danilov (1948, 1955, 1956) and the group led by Kazachkovskii
(1948) studied the action of impurities on the solidification of metals
in ultrasonic fields. Kapustin (1955) and coworkers conducted extensive
research on the crystallization of some organic substances and metals.
Extensive investigations into the ultrasonic treatment of molten fer
rous and nonferrous metals and alloys were conducted in the 1960s. We
note here the efforts of Teumin (1962), Abramov (1967, 1970, 1972)
and their coworkers in the ultrasonic treatment of melts and the crys
tallization of ferrous metals and alloys, and also the investigations of
Polotskii (1960) and his colleagues into the propagation and absorp
tion of ultrasound in melts and the effect of ultrasonic oscillations on
the solidification of some nonferrous metals. The early 1960s also saw
our first publications on the ultrasonic treatment of molten light alloys
(Eskin, 1960; 1965). A considerable contribution to ultrasonic treatment
was made by Pogodin-Alekseev and his colleagues (1966) who studied
the crystallization of Wood’s alloy, beryllium bronze, and other metal
Introduction xiii
systems. The effect of ultrasound on the crystallization of metals was
studied by Balandin (1973), Gorev (1961), and some other Russian sci
entists. Detailed reviews of these studies may be found in books by
Balandin (1973), Eskin (1965, 1988), and Abramov (1972, 1994).
It should be noted that to a considerable degree ultrasonic metal
lurgy owes its advance to the progress in the physics and technology of
ultrasound achieved as a result of extensive studies conducted by a team
of Russian acousticians, led by Rozenberg, which were generalized in a
three-volume monograph by this author (1967, 1968, 1977).
The aforementioned extensive investigations into the ultrasonic treat
ment of melts were supported by the availability of home-made ultra
sonic equipment for laboratory and industrial applications developed by
a team headed by Kitaigorodskii (Agranat et a/., 1974).
An important contribution came from the research on the founda
tions of ultrasonic technology in metallurgical processes (refinement,
flotation, dispersion, and so on) conducted in the Moscow Institute of
Steel and Alloys. These studies were described in detail in our multiau
thor textbooks (Agranat et a/., 1974; 1987; 1990; 1993).
The fact that a host of publications on the subject have appeared in
recent years is indicative of intensive research in the ultrasonic treatment
of melts. While the reference lists of the famous review by Hiedemann
(1954) and the handbook by Herrmann (1958) included 120 papers pub
lished mainly prior to W.W.II, by the time of the first edition of our
book (Eskin, 1965) we had 260 papers, patents, and inventor’s certific
ates concerned with ultrasonic effects on melts. The next edition of this
book (Eskin, 1988) saw that the number of relevant publications had
increased almost tenfold.
In the United States, ultrasonic metallurgy was extensively
studied at Westinghouse Electric Co., Massachusetts Institute of Tech
nology, and Branson Ultrasonics, to name but a few famous institutions.
They studied the effect of ultrasound on the melting of ferrous metals in
electric arc furnaces with expendable electrodes (Lane and Tiller, 1960),
the effect of ultrasound on the mechanisms of crystallization, ultrasonic
atomization (Grant, 1983), ultrasonic brazing, and the like.
In England (Crawford, 1955), ultrasonic studies were concentrated
in the firm Mullard, which also developed industrial ultrasonic equip
ment. At present, ultrasonic studies on sonochemistry in metallurgy are
intensively conducted at Coventry University (Mason, 1991; 1994; 1995).
In Germany, many years of research in the ultrasonic treatment of
light alloys resulted in pilot plants for melting and casting in the ul
trasonic field. These developments were covered in excellent books by
Bergmann (1954), Matauschek (1958), and Herrmann (1958). Profound
XIV Introduction
studies of the physics and technology of ultrasonic welding, ultrasonic
atomization, and some other applications of ultrasound were conducted
by the research team led by Pohlmann (1974). It is also worth noting
the studies of the group led by Seemann (1966, 1967, 1970, 1976), in
particular, the contribution of his scholar Buxmann (1972) who imple
mented the casting of aluminum ingots in an electromagnetic mold at
Aluswiss (Alesa) in Switzerland and experimented with the ultrasonic
grain refinement of aluminum alloys (Gabathuler and Buxmann, 1993).
Reportedly, Furukawa of Japan is carrying out research on continuous
casting of aluminum alloys with the use of ultrasound. Daihatsu Motor
Co. is studying the applicability of ultrasonic treatment of aluminum
melts to seepage of melts and wetting of powders and fibers of aluminum
oxide (Nakanishi et a/., 1993).
In Bulgaria, research into mold casting of aluminum alloys is con
ducted by a team headed by Angelov (1970). In the Slovak Republic,
studies are focused on the development of industrial ultrasonic equip
ment (Svehla, 1978; 1984) and ultrasonic casting of aluminum alloys
(Kratky, 1979).
Studies of ultrasonic treatment of aluminum alloys were also reported
in China (Lin Chun Mao, 1979) and in Poland (Bondarek et a/., 1981).
Detailed reviews of ultrasonic and vibration treatments of metal were
given by Campbell (1981) and Ambramov (1994).
At present, Russian manufacturers produce reliable ultrasonic equip
ment that has proved its worthiness in the industrial environment. A
substantial instrumental improvement was achieved with a system of
automatic frequency control that steers the generator frequency to the
resonance frequency of the transducer and the oscillatory system. How
ever, long-term continuous operation of ultrasound sources requires flex
ible automatic control systems not only for frequency, but also for the
amplitude of the oscillatory system according to a set program. Thus,
the development of industrial ultrasonic robots for metallurgy is only a
matter of time.
Like vacuum melting, continuous casting in electromagnetic molds
and ultrasonic treatment of melts can be used successfully to improve
the quality of metal produced on an industrial scale.
1
FUNDAMENTALS
OF MELT
SONICATION
1.1 Ultrasound and Low-Frequency Vibration

Over the past century, the Chernov idea of dynamic
solidification has been used in various procedures for treating
liquid and solidifying metal such as vibration, shaking,
mechanic and electromagnetic stirring, and ultrasonic
sonication. Depending on the particular application, all of these
methods have found industrial use. Among these procedures,
sonication with power ultrasound stands out due to the extent
of its effect on the structure and other properties of casting
metal.
With respect to degassing, filtration, and the solidification
of melts, low-frequency vibration and ultrasonic sonication are
frequently considered as a common method of treatment. As a
result, in the literature the term vibration refers to elastic
oscillations of various frequencies ranging from low-frequency
vibration to ultrasound. Several researchers have even
1
2 Ch. 1. Fundamentals of Melt Sonication
suggested the term ultrasonic vibration, which is totally
incorrect, because ultrasonic oscillations and low-frequency
vibrations differ not only physically, but also in the laws that
govern oscillations propagating through the pool.
In order to implement ultrasonic techniques in industry,
particularly in metallurgy, one should clearly understand the
difference between low-frequency vibrations and ultrasound.
Such an understanding will explain the effects produced by the
ultrasonic sonication of melts and will show how ultrasound
can be used in industrial metallurgical processes.
Under a low-frequency vibration, i.e. shaking, the object
alternates with certain cyclic frequency co = 27tf, velocity v, and
acceleration y. When the motion changes its sense, the velocity
and acceleration rapidly vary as if the object under vibration
were being pushed with the same frequency f. The
corresponding dynamic action on the object (elementary volume
of the melt) of static weight P = mg is described by
11
( . )
During one period T = 1//, the weight of this elementary

volume varies from the minimal value P\ = (g - j)m to the
maximal value P2 = (g + y)m. If g = j then P\ = 0 and P2 = 2m.
Obviously by increasing y, one can increase the pressure in the
phases of rarefaction and compression respectively. In the
general case, this alternating pressure induces wave
propagation in the melt; however, for vibrations at low
frequencies (50-500 Hz), their wavelength will be tens of
meters, which significantly exceeds the feasible size of common
ingots.
In this work we will not consider the effects of impulse
vibration (Efimov and Eldarkhanov, 1995). This direction of
research is rapidly advancing in various fields of engineering.
In recent years, it has become a tool for improving the
structure of solidifying ingots. In the case of impulse
vibration, the pulse repetition rate varies from just a few hertz
to thousands of hertz; however, the frequency range of
powerful disturbances in the pool extends to much higher
1.1. Ultrasound and Low-Frequency Vibration 3
frequencies. With such technology, cavitation processes can hardly
originate in the melt, so the mechanism of structural breakdown
seems to be connected with an active stirring of the metal.
On the contrary, the wavelength becomes shorter if the
melt is acted upon by a continuous harmonic excitation at
higher sonic and, especially, ultrasonic frequencies. In this
case, one would expect a wave to be excited in the liquid or
solidifying metal.
By ultrasound we will mean elastic oscillations with
frequencies above 16000 Hz, in agreement with the common
use of the term. More precisely, acoustic phenomena are
classified with respect to their frequencies as follows:
infrasound ranges up to 16 Hz, audio sound from 16 Hz to
16xl03 Hz, ultrasound from 16xl03 Hz to 1010 Hz, and hyper
sound ranges above 1010 Hz.
Hence, the range of elastic oscillations extends up to
hundreds of gigahertzs, i.e. to the frequencies of thermal
oscillations in atoms. True enough, the production of acoustic
power densities even around 104 W/m2 is unattainable for
these frequencies. At the same time, commercial ultrasound
transducers used in industry at frequencies of 18-20 kHz are
capable of generating power densities about 105-106 W/m2, which
allows the necessary treatments of liquid and solidifying melts.
Physically, elastic oscillations are common in nature
regardless of whether they are ultrasonic, audio, infrasonic, or
hypersonic. In any case, alternating mechanic disturbances
occur in the elastic medium. However, a number of new
phenomena occur when hypersonic frequencies are used instead
of sonic.
Phenomena such as acoustic cavitation or flows occurring
in a sonified pool producing irreversible effects occur mainly
under the effect of high-intensity ultrasound.1 As will be
shown below, the ultrasonic intensity is related to the
oscillation frequency and amplitude by a square law, therefore
the actual ultrasonic intensities significantly exceed audio
intensities.
*In what follows we will mean the intensity as the density of acoustic
energy passing through a given section in one direction, i.e. the
energy density in the travelling acoustic wave.
Ultrasonic testing equipment uses specially reduced levels
of ultrasound; otherwise the tested material may suffer
structural changes and this method will no longer be considered
non-destructive.
A source of oscillations introduced into the pool induces an
ultrasonic field whose characteristics depend on the oscillation
parameters and properties of the pool. A basic parameter is the
propagation velocity of elastic oscillations. This velocity is
governed by the physical properties of the pool's material. At a
given temperature, the velocity [m/s] of sonic (ultrasonic)
longitudinal waves in a solid with density p, Young’s modulus
E , and Poisson’s ratio p is determined by
( 1. 2)
In a liquid, where elastic properties depend upon compression,

the velocity of sound is governed by
(1.3)
where p is the liquid density and /?ac\ is the adiabatic
compressibility. For gases, the molecular motion is related to
the adiabatic exponent y = cp/cv (the ratio of specific heats at
constant pressure and volume), gas pressure P q9 and density p:
(1.4)
Generated at any point in the medium (solid, fluid, or gas),
oscillating disturbances propagate through the medium as
elastic waves of alternating compressions and rarefactions. As
follows from expressions (1.2)-(1.4), in an unbounded medium,
the velocity of elastic waves is independent of frequency and,
up to certain magnitudes, intensity (this relation is referred to
as the linear approximation).
The product of oscillation velocity c and density p is called
the wave, or acoustic, resistance pc of the given medium. It is
equal in magnitude to the pressure of sound in the travelling
wave divided by its oscillation velocity.
An important characteristic of an elastic wave is the
distance between adjacent regions of compression or
rarefaction, i.e. the distance that the wave passes in one period
T = 1//, where f is the frequency of sound. This distance is
called the wavelength A. The wavelength is determined only by
the velocity and frequency of ultrasound:
(1.5)
As mentioned above, ultrasonic oscillations excited in the
medium set up alternating regions of elevated and reduced
pressure (PA) with wavelength A as the distance between
adjacent regions of compression and rarefaction. This pattern
for pressure PA propagates in the medium with the velocity
of sound c, whereas every particle of the medium oscillates
with velocity V and displacement A around an equilibrium
position.
In the one-dimensional case, elastic disturbances
propagating in the x-coordinate are described by the wave
equation
16
( . )
where A is the displacement of particles of the medium and t is

the time variable.
A particular solution to wave equation (1.6) can be
represented in the form of a harmonic wave
(1.7)
where co = 2xf is the cycle frequency and x/c is the phase
factor for particles harmonically oscillating around x.
The oscillation frequency is related to period T by / = 1/T,
and the wavelength A is related to the frequency by A = c/f.
If we differentiate equation (1.6) with respect to time, we
obtain an expression for the particle velocity:
V - A q CO c o s (cot - cox/c). (1.8)
Differentiating equation (1.8) once again, we obtain the
acceleration of an oscillating particle:
j = - A q CO2 sin(&tf - cox/c). (1.9)
If we multiply the velocity of an oscillating particle (1.8)
by the acoustic resistance pc of the medium, we obtain for the
pressure PA in the sound wave
PA - pc - pcAqO) c o s (cot - cox/c).
v (1.10)
Quantities (1.8), (1.9), and (1.10) vary periodically in space
and time. For t = 0, i.e. for a fixed time, they are
V = A q CO c o s cox/ c , (l.n )
j = - A qco 2 sin cox/c , (1.12)

PA = pcA^co cos cox/c . (1.13)
If we let x = 0, i.e. fix a position, then we obtain
V - A q CO cos cot, (1.14)
j = - A q CO2 sin&tf, (1.15)
PA = p c A q CO COSCOt. (1.16)
Respective magnitudes of the oscillating velocity [m/s],
acceleration [m/s2], and sound pressure [Pa] will be given by
v = Aco, j = -Aco , PA = pcAco.
n
A quantity of prime importance for ultrasonic field

characterization is the ultrasonic intensity 7, or power flux N ac
normalized by area S. In the simplest case of a plane wave, the
intensity [W/m2] is given by
(1.17)
Notice that all the above relationships hold only for a
travelling wave. They ignore the boundary conditions that give
rise to diffraction, standing waves, and other phenomena
complicating these expressions.
Below, we list the characteristic intensities [W/m2] in the
vicinity of various sonic (ultrasonic) sources:
Voice................................................... 1.6-10 4
Clarinet............................................... 6.3 • 10*4
Piano................................................... 8.0 • 10'4
Chamber-music orchestra................. 1.6 • 10'2
Symphony orchestra......................... 0.2
Aircraft warning siren..................... 0.3
Airplane.............................................. 1.0
Low-intensity ultrasound................. 1.0 • 104
Middle-intensity ultrasound............ 1.0 • 105
High-intensity ultrasound................ 1.0 • 10Q
There is no way of considering an ultrasonic field without
any idea as to the ultrasonic source, i.e. an oscillating system
that conveys ultrasonic oscillations to the melt.
An actual oscillating system (transducer) is always lossy
because of friction in the surrounding medium, internal
friction within the system itself, and a loss due to the active
work in the treated melt. As a consequence, the oscillation will
decay and the displacement amplitude A will be a decreasing
function of time:
(1.18)
where coq = 2/rfo is the system’s natural angular frequency, and
A q is the initial amplitude. The characteristic 8 proportional to
loss is called attenuation. Physically, it describes the amplitude
reduction rate.
A lossy system without any external input oscillates with
frequency f\,
(1.19)
Consequently, the loss reduces not only the amplitude, but also
the frequency of oscillation.
In order to make up for these effects, we should (1)
add power to the oscillation system from an external
source and (2) maintain the resonant mode by tuning the
source frequency to the natural frequency of the oscillating
system.
If the oscillating system is not loaded and ultrasonic power
remains in the system and is not expended there, then it
produces standing waves.
However, the actual situation is quite different. Every
system has some loss, and our prime interest is with systems
loaded by a melt, which plays the role of an active medium that
consumes a considerable portion of the acoustic power fed to
the system.
Being loaded, the oscillating system produces both standing
and travelling waves. Their relative amplitudes are
conventionally defined by a parameter called the travelling
wave ratio.
The quality of an oscillating system may be characterized,
in some sense, by its ability to produce standing waves, i.e.
oscillations of large amplitude occuring due to resonant
enhancement. Standing wave amplitudes can be more than ten
times higher than those of travelling waves, and the transducer
radiating face will oscillate with the amplitude determined by
standing waves.
In melts, standing waves have a little effect if the mold
geometry does not assist resonance at operating frequency.
The transducer face then emits predominantly travelling
waves propagating toward the melt. The amplitudes,
velocity, acceleration, and sound pressure of these waves are
connected with sound intensity I and frequency f by known
relations:
1.2. Propagation of Power Ultrasound in Melts 9
1 20 )
( .
( 1. 21)
1 22)
( .
(1.23)
From these formulae it follows that to produce a fixed,
frequency-independent intensity in a given medium, the
amplitudes of either the velocity or the sound pressure must
remain constant.
1.2 Propagation of Power Ultrasound in Melts

When considering melt sonication, it is very important to
understand how ultrasonic power is transferred to the melt for
a given source oscillation amplitude.
The conditions for transferring a certain ultrasonic power
to a liquid metal crucially depend on the acoustic, or wave
resistance, or, more generally, on the medium's mechanic
resistance, which is defined in acoustics as the ratio of
instantaneous values of force and velocity. This ratio has active
and reactive components when harmonic oscillations are
characterized by a phase shift.
The reactive component defines the shift of the natural
frequency of the radiating system produced by a load, while the
active component shows irreversible internal loss of energy
(possibly small) and useful radiation toward the load.
The ultrasonic power radiated to a melt is proportional to
the melt wave resistance multiplied by the squared oscillation
velocity, i.e. it depends on both the resistance of the liquid
metal and the oscillating velocity of the source:
(1.24)
When cavitation develops in the melt, the force and
velocity at the source radiating face vary, so that the above
relation describes the actual technological processes of melt
sonication only in the first, or linear, approximation. However,
so long as the standing waves dominate travelling waves, the
oscillating system retains the property of a resonant filter. As
a result, the oscillation processes remain almost harmonic, so
that the average loading resistance over the period remains a
useful characteristic.
In the presence of cavitation, loading resistance R\ is a
function of displacement or velocity of oscillations. As before,
the power can be considered to be proportional to the squared
oscillating velocity, but accurate to an amplitude-dependent
proportionality coefficient R\ which drastically decreases when
cavitation occurs.
In the presence of cavitation, the acoustic resistance of the
melt decreases because the velocity of sound and pressure in
the mold fluid metal are no longer equal to those in the non-
cavitating metal.
Figure 1.1 presents the dimensionless parameter K,
which is the ratio of loading resistance i?j under cavitation
to the loading resistance in the absence of cavitation, versus
the oscillating amplitude of the transducer face at 18 kHz
for water (1) and aluminum melt (2 ). When the face oscillation
amplitude exceeds 0.5 pm for water at 20°C and 2-3 pm
for an aluminum melt, the wave resistance decreases to
values ten times smaller than for sonication without
cavitation.
When acoustic cavitation begins, the acoustic power
transferred to the fluid increases because the reactive
component of the load falls to zero.
Figure 1.2 shows the relation between the amplitude of the
transducer face oscillations and the acoustic power generated
by a commercial ultrasonic equipment and conveyed into an
aluminum melt of technical purity at the resonant frequency
18 kHz.
Figure 1.1 Relative acoustic resistance K versus the

amplitude A of the source radiating face at 18 kHz. (1)
corresponds to water (Kikuchi, 1969 and Panov, 1984);
and (2 ) to an aluminum melt (Eskin, 1988).
As a rule, only longitudinal waves propagate in a liquid

metal. This means that medium particles oscillate along the
direction of wave propagation, thus generating alternate
regions of compression and rarefaction.
Using ultrasound as a treatment in a melt or some other
liquid, one should clearly understand the profile and intensity
distribution of the field in the pool.
For the frequent case of piston oscillations of the radiating
face, the plane wavefront usually suffers distortions.
To appreciate the distribution of an ultrasonic field, it is
important to know Drad//l, the ratio of the radiating face size
to the wavelength. The shape of the face is also important. If
the face size £>rad is small compared to A, i.e. if D rad//l << 1,
then the face may be deemed a source of spherical waves. In
actual conditions, the ratio D rad/A varies from 0.01 to 0.1;
therefore, the wave front is almost plane and the directivity
Figure 1-2 Power W & transferred into the melt versus

the amplitude A of transducer face oscillations at
resonant frequency 18 kHz for three face diameters: (1)
65 mm, (2) 40 mm, and (3) 20 mm.
pattern is narrow. For Drad/Z > 1, the directivity pattern

reduces to a main lobe surrounded by smaller side lobes.
Such directional patterns (see Fig. 1.3) are well known and
widely used in antenna theory and loudspeaker analysis. When
the plane radiating face is large compared with the wavelength,
the ultrasonic beam has an almost cylindrical shape.
Generally, any propagating harmonic field has two
characteristic regions: the near zone and the far zone. Only the
near field is of practical importance when describing the
solidification of a sonicated metal. The near field is generally
very nonuniform. Only a perfect sound source emits waves
with a plane phase front and a uniform amplitude. In reality,
the radiating face should be considered as a set of small
sources that oscillate in phase but with different amplitudes.
These sources emit partial waves that produce a diffusion
pattern in which each point oscillates either more strongly or
more weakly depending on the phases of partial waves arriving
at this point from various elements of the face.
Figure 1.3 Directivity patterns of ultrasonic sources.
Unfortunately, no instrument currently exists to measure

ultrasonic fields in melts of aluminum and other high-melting-
point metals. As a result, all known ultrasonic measurements
have been carried out with water, mercury, and other low-
melting-point materials.
For some metals and their low-melting-point alloys, the
velocity of elastic waves in melts can be measured by the
impulse method. It was found that the velocity of ultrasound
near melting points differs only slightly from that in solid
rods. Table 1.1 gives the velocity of ultrasound and other
characteristics for some molten metals.
Ultrasound loses its oscillation energy in a melt. The
displacement and the intensity of a plane ultrasonic wave
decrease exponentially with the propagation path x:
(1.25)
(1.26)
where a is the sound absorption, or attenuation, factor.
The absorption of ultrasound is associated with the
viscosity and thermal conductivity of the melt and with the
frequency of the ultrasound. Its frequency dependence is
described by a quadratic function; therefore, high frequencies
would be impracticable because of high attenuation.
Table 1.1 The velocity of ultrasound c, melting point, density pt and
adiabatic compressibility for several molten metals
melting test C, A Pud,
Metal point, °C point, °C m/s g/cra3 pPa1
Caesium 28.6 28.5 967±10 1.9 5.81
Rubidium 39.0 39.0 1260±10 1.53 4.27
Mercury -38.87 50.0 1350±10 13.5 0.358
Thallium 303 302 1625±15 11.8 0.34
Bismuth 271 271 1635±15 9.84 0.34
Bismuth 271 287 1663 9.84 0.361
Bismuth 271 300 1661 9.84 0.363
Lead 327.4 327 1730+15 11.3 0.292
Indium 156.2 156 2215120 7.3 0.290
Tin 231.9 232 2270120 7.2 0.279
Tin 231.9 247 2454 7.2 0.239
Tin 231.9 300 2455 7.2 0.240
In ideal liquids, the viscosity and thermal conductivity
suffice to describe the absorption loss of ultrasonic energy, but
in actual melts the effect of impurities should be taken into
account. The interfaces between the liquid and suspended
particles (non-metallic inclusions and supersaturation crystals)
may significantly affect absorption. Figure 1.4 shows the
absorption of ultrasound in water versus the proportion of the
suspended phase and the diameter of kaolin particles and
lycopodium powder.
Bubbles produce a similar effect on the absorption of
ultrasonic energy in the melt. The gaseous bubble-melt
interface acts as a scatterer of ultrasound. It is worth recalling
that gaseous and solid inclusions favor the development of
cavitation which absorbs an additional ultrasonic energy.
In an unbounded medium, ultrasonic oscillations may
propagate over large distances. However, in practice, the
sonicated samples are confined within finite volumes. For
Figure 1.4 Absorption coefficient a of ultrasound

versus (a) proportion C of the suspension and (b)
diameter d of kaolin particles in water. Curve 1 refers to
kaolin, d = 0.2-5 pm, and curve 2 refers to lycopodium
powder, d = 30 pm (Bergmann, 1954).
molten metals these are molds, feeders, and continuous casting

molds. It is only in the last case that the casting may be
treated, to a certain degree, as an unbounded medium. In other
cases, the reflections from the walls (reverberation) accompany
the ultrasonic propagation through a sonicated object, in line
with natural absorption along the path due to various
inclusions.
An incident ultrasonic wave is reflected and refracted at an
interface between two media with different wave resistances.
For normal incidence, only the relative wave resistance of
adjacent media governs the reflection coefficient:
(1.27)
The transmission coefficient

(1.28)
determines the amplitude of a wave passing from one medium
into another. Expressions (1.27) and (1.28) clearly demonstrate
that R and D depend only on the relative wave resistance of the
media.
Figure 1.5 shows the reflection and transmission
coefficients versus the relative wave resistances of the adjacent
media. We note that these relations do not include diffraction
phenomena and hence are mainly suitable for high-frequency
ultrasound in the megahertz range. At frequencies used in
metallurgy (18-25 kHz), the ultrasonic wavelength is
comparable with the casting dimensions, so that the
contribution of the diffraction phenomena becomes
considerable, and formulae (1.27) and (1.28) may be used only
as provisional estimators.
Directed to a melting metal, ultrasonic oscillations excite
standing and travelling waves. Standing waves occur as a
result of the interference between the incident and reflected
waves, provided that these are totally reflected at the
Figure 1.5 Reflection coefficient R and transmission

coefficient D versus relative wave resistance p\C\!P2C2*
(Bergmann, 1954).
interface. Accordingly, the sonicated medium has points (or
surfaces, in the case of a volume wave), where particles remain
stable in their positions, whilst the sound pressure is doubled,
similar to an oscillating string with fixed ends. Conversely,
with other surfaces, the pressure vanishes and the displacement
amplitude doubles.
The basic condition for standing waves to occur is clearly
the resonance condition, which means that the ultrasonic wave
path must be exactly (2n - l)A/2, where n is an integer.
If the propagation path either encompasses a non-integer
number of half-wavelengths, or is very large compared to the
wavelength, or if total internal reflection is not satisfied, then
the field has a travelling component.
We will describe the ultrasonic field using the wave
rigidity of the field = Pcco- This concept provides a
physical insight into the oscillatory motion of the melt with a
standing wave mode. We note in passing that this is an ideal
situation because usually no standing waves occur in the melt.
When a liquid metal makes contact with a surface of higher
wave rigidity (melt-mold wall interface), the displacement
phase changes by 180°. Therefore, a node of displacement and
an antinode of pressure occur on the wall. Conversely, if the
melt interfaces with a medium of smaller wave rigidity (e.g.
melt-air interface), the pressure vanishes at the reflecting
surface. In this case, the melt-air interface contains an
antinode of displacement and a node of pressure. It is due to
this fact that ripples and even eruptions (for power ultrasound)
are visible on the free surface of liquid metal sonicated from
the bottom.
Parameters of eruptions depend considerably on the liquid
(melt) density and surface tension. Kornfeld and Molokhova
(1950) showed in their experiments that the product of the
surface tension and the height of the jet is a constant, which
depends only on the acoustic power. Using high-speed
photography, Rosenberg and Makarov (1957) determined that
eruptions are mainly governed by the radiation pressure.
In practice, any kind of sonication, whether applied to a
melt solidified in a mold (solidificator), or to the melt within a
feeder or a mixer crucible, produces a combination of standing
and travelling waves. Perfectly stiff and soft media do not
exist in nature, therefore a portion of the ultrasonic energy
passes through the interface into the adjacent media, thus
decreasing the displacement amplitude of the reflected wave
below the level of the incident amplitude. As a result, a "non-
ideal" distribution is settled over the volume, with no point at
which the amplitude vanishes.
Considering actual sonicated commercial melts containing
suspensions of insoluble non-metallic inclusions and gaseous
bubbles, we may assume that the ultrasonic field is diffuse. In
other words, this field is characterized by random and variable
distributions of pressure and oscillatory displacements. The
cavitation processes are an additional point in favor of the
diffusion model for a sound field in the melt.
1.3 Acoustic Cavitation in Liquid Metals

Among the physical effects accompanying the propagation of
highpower ultrasonic oscillations in melts, cavitation, or the
formation of cavities filled mainly with gases dissolved in
liquid metal, seems to be the most important. Formed by the
tensile stresses characteristic of the half-period of rarefaction,
these cavities continue to grow by inertia until they collapse
under the action of compressing stresses during the
compression half-period, thus producing high-intensity shock
waves in the fluid.
Consequently, in an external ultrasonic field, gas cavities
are produced at the weakest points of the melt during intervals
of reduced pressure, and collapse during intervals of increased
pressure.
Sonication driven cavitation induces many physical and
chemical processes that strongly affect the melt. These are
primarily processes related to the refinement of liquid metals
from solid and gaseous inclusions and the activation (wetting)
of solid inclusions.
In order to produce acoustic cavitation, one needs to
introduce some energy into the melt. As a result, the melt
1.3. Acoustic Cavitation in Liquid Metals 19
becomes heated and some acoustic streaming occurs within the
sonicated volume.
1.3.1 Cavitating Rigidity of Liquids
An intense alternating pressure applied to a liquid, for example
by power ultrasound, results in disruptions in the liquid.
A statistical consideration of cavitating processes in the
time domain shows that various types of cavities, with
different behavior, originate at points of disruption under the
action of the ultrasonic field. One group of cavities pulsates
and their vapor-gas content remains invariable. A second
group intensively grows because of tensile stresses and one-way
diffusion from the liquid to the cavity. A third group collapses
under the action of compression stresses in the sound wave,
producing a multitude of very small bubbles and giant local
pressures near collapse points.
The sound pressure must exceed a certain level in order to
initiate cavitation in liquid. This cavitation threshold is a
measure of the cavitation capability of liquids.
According to the kinetic theory of fluids, liquids of pure
substances have extremely high cavitation thresholds
corresponding to impressive tensile stresses—in the region of
tens of megapascals. As an example, the fracture strength of
pure water is given by
(1.29)
where a is the surface tension [N/m] and R is the
intermodular distance [nm]. For water with cr= 0.079 N/m
and R = 20 nm, this formula gives a fracture strength of about
103 MPa. This value decreases by an order of magnitude if we
assume that vapor bubbles occur spontaneously, due to thermal
fluctuations. This situation is described by the equation
suggested by Zel’dovich (1942)
(1.30)
where Pv is the vapor pressure in the bubble [Pa] and T is the
absolute temperature [K].
Even with the assumption that vapor nuclei occur in the
liquid, the fracture strength decreases only to 100 MPa, which
remains much higher than the actual cavitation threshold
obtained for water in distilled, normal, and air-supersaturated
states for various frequencies. These findings were obtained by
several researchers, whose data were collected and discussed by
Sirotyuk (1968). Figure 1.6 shows that at frequencies up to 1.0
MHz used in industry, the fresh tap water and air-
supersaturated water have cavitation thresholds that do not
exceed 0.1-1.0 MPa, which is three to four orders of
magnitude below the calculated value.
According to current views on the cavitation threshold, the
tensile stress-induced disruptions in liquids are not governed
by molecular forces, but rather by the presence of nuclei of
other phases. Vapor and gaseous bubbles, solid gas-adsorbing
suspensions, and hydrophobic inclusions may serve as such
nuclei. Cavitation nuclei can be stimulated by upsetting the
thermodynamic equilibrium within the surrounding medium,
by fluctuations in the sonicated liquid, and by radiation acting
on the liquid.
Because the cavitation capability of common liquids is
governed by weak points in the form of impurities, one should
Figure 1.6 Cavitation threshold of water versus

ultrasonic frequency. (Sirotyuk, 1968).
consider basic models for cavitation nuclei in order to choose

an appropriate model for melting metals of metallurgical
purity. Soluble impurities are of little interest to our
discussion because they affect the cavitation process indirectly,
through variations in the surface tension of the liquid.
Nevertheless, for alloys of the solid solution type, e.g. for Al-
Mg alloys, the surface tension decreases by almost 30% due to
either magnesium or surface-active additions, such as sodium
or bismuth being incorporated into the alloy.
The cavitation capability of a liquid is basically governed
by insoluble additions that can occur in any of three possible
states. Liquid insoluble additions seem unlikely to increase the
capability because the molecular adherence of the main liquids
is quite strong. The same inference is true for solid additions if
they are well wetted in the studied liquid.
Frenkel (1945) proved theoretically and Harvey, McElory,
and Whitley (1949) showed experimentally that an increase in
the cavitation threshold is most likely to be governed by the
hydrophobic particles which have small fractures filled with
undissolved gas (Fig. 1.7).
Figure 1.7 Model of a cavitation nucleus (Frenkel, 1945;

Harvey, 1939) in the form of a nonwettable (hydrophobic)
solid particle.
The adhesion of a liquid to the surface of a solid, or the
wettability is characterized by
(1.31)
where <j\ ?2 is the difference of surface energies at the solid-
vapor (gas) boundary, 02,0 is the difference at the vapor-liquid
boundary, and j 1(0 is the difference at the solid-liquid
boundary.
For an unwettable solid particle with a small fracture,
which may contain a gaseous (vapor) bubble at the mouth, we
obtain
(1.32)
where 6 is the contact angle of the bubble with the surface.
From (1.31) we have
(1.33)
The vanishing contact angle 6 = 0 corresponds to a perfectly
plane bubble, e.g. to a surface that is absolutely unwettable in
the liquid.
Frenkel (1945) emphasized that whether these suspended
particles could be cavitation nuclei does not depend on their
dimensions but on their physical and chemical properties,
primarily, adhesion Aa.
From the above it follows that in fluids with a suspension
of weakly wettable particles of complex microstructure (with
fractures), cavitation induced by tensile acoustic stresses could
start from a gaseous (vapor) formation at the mouth of a
fracture.
1.3.2 Cavitation Conditions for Melting Metals
Estimates and measurements of the cavitation rigidity of liquid
metals demonstrate an elevated cavitation threshold as
compared to that of water. The reason is that most of the
material characteristics assume higher values in molten metals.
Gaseous and non-metallic solid inclusions existing in specific
forms in the metals may be an additional cause.
Equations (1.29) and (1.30) for the cavitation threshold
relate the cavitation capability, the surface tension at the
liquid-gas interface, and the initial bubble radius. In addition,
the capability of a liquid is significantly affected by the
concentration of gas and solid unwettable particles. Obviously,
the viscosity of a liquid also markedly affects its capability. As
an example (Frenkel, 1945), the initial radius R q of a cavity
produced in a liquid by heterophase fluctuation of a gaseous
nucleus of diameter A (1 nm) is related to viscosity 77 by
(1.34)
where n is the concentration of nuclei proportional to the
number of gas molecules dissolved in unit volume (1 cm3) of
the liquid, k is the Boltzman constant, and T is the absolute
temperature of the liquid.
The cavitation capability of liquids was extensively studied
by Il’ichev (1968) in wide limits of material characteristics.
Figure 1.8 shows how the surface tension affects the
cavitation rigidity of water solutions sonicated at frequencies
of 8-61 kHz. It is seen that the experimental cavitation
threshold decreases linearly with the surface tension.
Figure 1.8 also presents the cavitation rigidity of water
solutions and glycerin as functions of viscosity. At higher
Figure 1.8 Cavitation rigidity of water versus (a)

surface tension and (b) viscosity at (1) kHz, (2)
28 kHz, (3) 35 kHz, and (4) 61 kHz (Il’ichev, 1968).
8
Figure 1.9 Cavitation rigidity of water versus the AL O

0 3
concentration. Curves 1, 2, and 3 correspond to distilled
water with gas contents of 16, and 82% respectively;
20
and curve 4 corresponds to tap water with % gas.
8 6
(Il’ichev (1968).
viscosities 77, the cavitation threshold Pc increases. Figure 1.9

gives an idea of how the cavitation threshold of distilled water
is affected by the gas (air) content and the concentration of
20-30-pm solid unwettable particles of aluminum oxide. It is
clear that solid impurities considerably vary the cavitation
threshold only in the gas-free water (curve 1). At a higher
concentration of gas in water, the effect of the impurities
becomes negligible (curves 2-4).
Abramov (1972, 1974a) studied the cavitation threshold of
low-melting-point metals (Bi, Sn, In, Pb, and Cd). For these
metals, he estimated the cavitation threshold at frequency 2 0
kHz to be 0.5-1.0 MPa assuming the bubble radius i ?0 = 0.01-
100 pm. Abramov and Astashkin (1974b) found experimentally
that at 20 kHz tin is characterized by a cavitation threshold of
0.3-0 .6 MPa, which agrees with the theoretical estimates.
It is remarkable that experimental estimates obtained by
these workers demonstrate that for all investigated melts
whose wave resistances pqCq are nearly equal, the cavitation
threshold depends linearly on surface tension, in agreement
with results reported by Il’ichev (1968) for water. It is obvious
that for aluminum, magnum, steel, and cast-iron melts that
have stronger surface tensions, a further increase in the
cavitation threshold must occur. It should be noted that
theoretical and experimental estimates do not take into account
the non-metallic inclusions which were shown to considerably
affect the formation of cavitation nuclei.
1.3.3 Non-metallic Inclusions and Cavitation Nuclei
Liquid metals may hardly be considered to be perfect liquids.
Prominent metallurgists (Chernov et al., 1956) pointed out that
liquid metals contain many insoluble additions. These solid
non-metallic inclusions mainly govern the cavitation capability
of metal when liquid and structure when solid.
Gas dissolved in liquid metals hampers analysis of the
cavitation threshold as a function of insoluble additions. It
turns out that the conditions for cavitation nucleation could be
obtained in a simpler way if we considered the combined the
melt-non-metallic inclusion-gas system.
Because we are mainly concerned with light metals, we will
analyze the cavitation capability of aluminum melt and
consider the liquid aluminum-aluminum oxide-hydrogen
system.
Aluminum interacts almost exclusively with hydrogen and
oxygen. Oxygen is present only in the form of aluminum oxide,
A1203, which is a stable chemical substance resistant to
thermal dissociation. Aluminum oxide is normally present in
the melt mainly in the form of a suspension of 1 .0 pm particles
rather than as rare particles which are either drawn into the
metal from the surface or are added with the charge of large
particles. In commercial aluminum and aluminum alloys,
aluminum oxide is contaminated by other oxides of additions
accompanying aluminum production (such as Fe, Mg, Cu, Ti,
and Si) and also by the oxides of alloying and modifying
components, in particular, transient metals with unfilled d-
shells.
Klyachko (1941) was among the first to establish that in
aluminum heated above 500°C, hydrogen is released from water
vapor adsorbed at the surface. This reaction consists of the
dissociation of water vapor, the formation of aluminum oxide,
and the release of hydrogen.
In molten aluminum, hydrogen occurs predominantly in the
dissolved form; there is no experimental evidence for its
occurrence in the form of free bubbles. Klyachko (1941),
Al’tman (1965), Sharov and Pimenov (1970), Makarov (1983)
and some other workers hypothesized about the relation
between dissolved hydrogen and oxide inclusions. An
experimental result in favor of this relation was obtained by
Pimenov and Demenkov (1972) who demonstrated that the
suspension of aluminum oxide slows down the diffusion of
hydrogen from the melt.
Below, we highlight some ideas underlying the interaction
between hydrogen and aluminum oxide in liquid aluminum.
Al’tman (1965) theorized that this type of interaction
forms certain complexes such as [A120 3]X= Hy, whereas
Pimenov (1972) assumed the finely divided suspension of
aluminum oxide to be a colloid system. Aluminum oxide is an
active adsorbent. Its surface is poorly wetted in liquid
aluminum and adsorbs dissolved hydrogen in the form of ions
(protons). As a result, the surface of these particles assumes a
positive charge. This charge is balanced out by an electron gas
which creates a diffusion layer around the particle. This layer
will also attract protons (but far less efficiently), which are
screened by the next, weaker layer of the electron gas, and so
on.
According to this model, each particle of aluminum oxide is
surrounded by an ion solvate envelope that consists of protons
and electron gas and is called the micelle. Micelles produce a
non-uniform distribution of hydrogen in the melt. The closer a
layer is to the micelle, the higher the concentration of
hydrogen. Commercial aluminum contains about 0.005%
aluminum oxide, and this small amount controls about 5% of
all hydrogen in the melt.
The ability of oxides to adsorb hydrogen was first found at
500°C with synthetic aluminum oxide. Kunin et al. (1971)
investigated the adsorption of hydrogen by aluminum oxide for
temperatures of 300-800°C and high pressures, conditions
which are very close to those present in melts. Under these
conditions, the activity of aluminum oxide with respect to
hydrogen may be caused by altered association and capillary
effects. According to this hypothesis, a great number of
adsorbed layers can occur in capillaries under the influence of
two or more walls. These layers present a gradual transition to
capillary condensation. Because actual oxide particles
suspended in the melt are characterized by a tree-like capillary
system, one can assume that the hydrogen-aluminum oxide
interaction would be even more active than that between
hydrogen and synthetic aluminum oxide. Arbuzova et al. (1971)
showed that hydrogen adsorption can be enhanced by a factor
of ten when an oxide with an iron impurity replaces chemically
pure aluminum oxide. Consequently, molten metals essentially
differ from water and water solutions in cavitation nucleation
in that the aluminum oxide available in the melt adsorbs
dissolved hydrogen through chemosorption.
In modelling acoustic cavitation, it is important to estimate
the pressure that arises in the melt when the cavity collapses.
This problem is poorly analyzed (Flynn, 1964). As far as we
know, only Abramov and Astashkin (1974b) have estimated the
pressure in the shock wave for several liquid metals by
numerically solving the Kirckwood-Bethe-Gilmor equation at a
frequency of 20 kHz for various initial dimensions of bubbles
(from 5 pm to 10 pm) and sound pressure amplitude of 0.3
MPa.
Below we give the relation between the pressure in a shock
wave and cavity parameters for various melts (Rq = 6.5 pm,
PA = 0.3 MPa):
Sn Bi Cd Pb In
R m & x/ Ro 11.9 10.5 10.41 9.9 8.8
^ m in /^ 0 0.0226 0.02332 0.025 0.0247 0.031
^max» km/s 5.8 4.07 4.38 3.62 3.54
P i nax* GPa 1.16 1.07 9.87 9.5 6.5
In order to estimate the effect of the physical and chemical
material properties of a liquid melt on the shock wave
pressure, the calculations were carried out for various
viscosities from 0.0017 to 0.007 Pas, keeping the surface
tension constant. The estimated shock wave pressure varied
very little, e.g. for bismuth, from 1.07 to 1.02 GPa. Variations
in surface tension had a more significant effect. As an
example, an increase of the surface tension of bismuth by a
factor of 3 to 4 increased the shock wave pressure ten-fold.
1.3.4 Cavity Dynamics
The evolution of cavities in water and water solutions has been
given adequate consideration (Flynn, 1964).
The dynamic behavior of a single vapor-gas cavity in an
incompressible liquid is described (neglecting gas diffusion to
the cavity) by the Noltlingk-Neppiras equation (Noltlingk and
Neppiras, 1950)
(1.35)
Here, R is the radius of the cavity, R0 is the initial radius of
the cavity, a is the surface tension of the melt, Pv is the vapor
pressure, p is the viscosity of the melt, p is the melt density,
PA is the sound pressure, P0 is the static pressure, and co = 2 7rf
is the circular frequency.
For correlation analysis, Eskin, Ioffe, and Shvetsov (1972)
suggested the dimensionless variables r = cot and r = R /R q
which can be used to obtain the set of equations
(1.36)
where
Table 1.2 Material constants of aluminum and water
Constant Aluminum Water

(700°C) (20°C)
Density p, kg/m 3 2350 1000
Surface tension cr, N/m 0.860 0.079

Viscosity rp MPa s 1.0 1.0
Pressure Pq, MPa 0.1 0.1
Frequency f, kHz 18 18
Vapor pressure Pv, kPa 0 2.2
We used the Runge-Kutta method to solve this set

subjected to the initial conditions v(0) = 0 and r(0 ) = 1 .
Calculations were carried out for an aluminum melt at a
temperature of + 40°C and for water at 20°C. Table 1.2
summarizes the material constants used in these calculations.
We investigated cavities with initial radii R q varying
between 1 pm and 1 0 0 pm, the minimum radius (critical radius
Rcr) being determined from the stability condition, and the
maximal initial radius from the formula derived by Minnaert
(1933).
Figures 1 . 1 0 - 1 . 1 2 show the relative radius R /R q and gas
pressure Pg calculated for various acoustic pressures. The
families of curves differ in the cavity’s initial radius and the
intensity of the external sound field, and illustrate how the
cavity evolves during 1-3 periods T of the sound wave.
From curves in Figures 1.10-1.12 it follows that if the
sound pressure is small enough (PA < Pc, where Pc is the
cavitation threshold), the cavities pulsate and do not collapse
during this time. The pressure in gaseous bubbles varies very
little. As the sound pressure PA increases to values above
1 MPa and becomes close to Pc, the majority of cavities with
Ro > Rcr behave like a typical cavitation bubble collapsing at
the end of the period. With a further increase in sound
pressure, i.e. for PA » Pc, cavitation develops to a stage
Figure 1.10 Evolution of cavities with initial radii (a)

Bo = 100 jim and (b) Rq = 50 pm in aluminum melt for
acoustic pressures PA: (1) 0 . 1, (2) 0 . 2 , (3) 0 . 6 , (4 ) 1. 0 ,
(5) 2.0, (6 ) 3.0, and (7) 10.0 MPa.
where all cavities expand during one or two periods of the

sonicating wave and then collapse.
When a bubble expands to ten or a hundred times above its
initial dimension, for PA > 2 MPa, the pressure in the cavity
falls to 100-133 Pa. The more intense the sound pressure, the
faster the cavities collapse.
Experiments by Schmid (1959) using high-speed
photography of unstable cavities and bubbles from evolution to
collapse, showed that bubbles may remain spherical, with a
smooth and glossy surface only during the first rarefaction
cycle. During a further rarefaction, the bubbles vary in shape:
they become rough and after the collapse disintegrate into a
multitude of small fragments. Sirotyuk (1968) observed a
similar pattern of cavitation bubble evolution in the field of a
powerful focusing equipment.
Figure 1.11 Evolution of cavities with initial radii (a)

Rq — 10 pm and (6) Rq = 1 pm in the aluminum melt for
acoustic pressures PA: (2) 0.2, (2) 1.0, (3) 3.0, ( 4 ) 5.0, (5)
10.0 MPa
This analysis of bubble dynamics did not take into account

the diffusion of gas dissolved in the liquid toward the cavity.
This fact allows us to assume that the collapse mode is
accompanied by the formation of a multitude of new cavitation
nuclei. The bubble can survive due to gas diffusion in melts
with low saturated vapor pressure or due to vaporization from
the bubble walls in liquids with high saturated vapor pressure.
In order to estimate the effect of surface tension on size
and pressure in cavitation bubbles, we compared molten
aluminum (cr=0.860 N/m), magnesium (cr= 0.50 N/m), and
water (a = 0.079 N/m) for R0 = 1 pm and PA = 10 MPa (see
Fig. 1.12). Despite a significant difference in properties (the
surface tension of molten aluminum is more than ten times
Figure 1.12 Effect of surface-tension on the evolution of

a cavity with initial radius Rq = 1 pm at PA = 10 MPa in
(i) aluminum melt (a = 0.86 N /m ), (2) magnesium melt
(cr= 0.5 0 N /m ), and water (<j= 0.079 N /m ).
that of water), conditions for the one-way gas diffusion from

liquids to cavities do occur after the bubble expansion.
A fter sonication the gas diffusion through bubble walls in
the liquid is governed by the following effects. Firstly, in the
rarefaction phase, the area of a pulsating bubble is many times
greater than its area in the compression phase; therefore, the
gas diffusion flow toward the bubble during the rarefaction
phase exceeds the gas diffusion flow from the bubble during
the compression phase. Secondly, the diffusion is governed by
the diffusion layer around the bubble. During the rarefaction
phase, its thickness increases and the concentration gradient
decreases. Conversely the layer becomes thinner and the
concentration gradient increases during the compression phase.
This dynamic provides an additional increase in the amount of
gas passing to the bubble.
Thus the pulsating cavity enlarges due to the one-way gas
diffusion toward the cavity.
The model of bubble evolution allowing for gas diffusion is
a rather complex set of equations. Eskin, Ioffe, and Shvetsov
(1974) demonstrated that the Noltingk-Neppiras equation may
also provide a solution of practical significance if one replaces
the terms denoting various pressures (excluding sound
pressure) with time-dependent pressure P(t) of a gas in the
bubble:
(1.37)
Obviously, the pressure P(t) will not be a simple analytic
function of radius if diffusion is taken into account.
The mass of gas in the bubble M = (4/3)7rR3mn, where n is
the gas number density and m is the molecular mass of the gas,
obeys the equation
(1.38)
where i(t) is the instantaneous density of gas flow through the
walls.
If we assume that the gas in the bubble is ideal, we can use
the equation of state, P = nkT, to relate the gas pressure P to
the number density n and the temperature T, where k is the
Boltzmann constant. If we also assume that the temperature of
the melt is kept constant during the cavitation (isothermal
conditions), equation (1.38) assumes the form
(1.39)
The gas flow in the bubble may be governed by various

processes, the most usual being diffusion. The diffusion in a
pulsating bubble is described by the convection-diffusion
equation, which, in the case of spherical symmetry, has the
following form
(1.40)
where C is the gas concentration in the melt, r is the radial
coordinate measured from the bubble center, vr is the radial
velocity of the fluid, and D is the diffusion coefficient.
The gas flow density i(t) is given by
Thus, one should simultaneously solve equations (1.37),

(1.39) , and (1.40) to describe the bubble evolution with gas
diffused in the bubble.
The bubble dynamic equation (1.37) and the convection-
diffusion equation (1.40) are very difficult to solve
analytically, and no suitable numerical solution procedure
exists.
We have simplified the problem by using the relation for
the gas flow to the cavitating bubble proposed by Boguslavskii
(1967). Boguslavskii assumed that the convective gas diffusion
occurs in a very thin liquid layer next to the bubble. Under
these conditions, the layer can be considered plane and equation
(1.40) assumes a simpler form:
(1.41)
where y is the coordinate perpendicular to the bubble surface.

The initial and boundary conditions are as follows: C = Cq
at t = 0 , C = 0 at y = 0 (t > 0 ), and C -> C0 for y -> oo (t > 0 ).
Deriving (1.41) we assumed that the bubble surface moves with
a uniform velocity.
Equation (1.41) allows for an analytical solution, which
yields the gas flow density through the bubble boundary as
(1.42)
where C q is the equilibrium gas concentration independent of

the applied tensile stress.
If we rewrite equations (1.37) and (1.39) in terms of the
bubble area S = 4 nR2 and use the Boguslavskii relation for the
total gas flow I in the bubble,
where Z0 = 0.8PA, then the set of equations describing the

bubble dynamics, allowing for gas diffusion in the bubble,
takes the form
(1.43)
This system yields a numerical solution for initial values

S(0 ) = S0, P(0 ) = P09 and S(0 ) = 0 .
It should be noted that other approximate solutions, for
example, the solution obtained by Eller and Flynn (1965) on the
condition that 1 / R0^D / (27tf) « 1, fail to describe the
cavitation bubble evolution in the melt allowing for diffusion.
This can be attributed to at least two factors: one is connected
with low ultrasonic frequencies (about 18 kHz) that are of
interest to metallurgists, and the other with the diffusion
coefficient D = 1 cm2/s in the molten aluminum-hydrogen
system, which significantly exceeds the diffusion coefficient in
the water-air system.
We carried out computations for an aluminum melt with
hydrogen concentration C0 = 0.2 cm3/ 1 0 0 g and diffusion
coefficient D = 1 cm2/s. The initial cavity radius R q was varied
in such a way as to model cavitation nuclei in the form of
Figure 1.13 Relative radius R/R q and pressure P in

cavitating bubbles (a) without cavitation (P^ = 0.2 MPa)
and (b) at incipient cavitation (PA =1. 0 MPa) for: (I) R q =
100 pm, (II) Rq = 50 pm, and (III) Rq = 100 pm. (1)
accounting for gas diffusion in the bubble and ( ) without
2
diffusion.
0 . 3 - 1 . 5 pm unwettable particles of aluminum oxide powder

(Pimenov and Demenkov, 1972) and l-100pm hydrogen lenses
emerging around these nuclei.
Figures 1.13 and 1.14 illustrate the variation of relative
radius R /R q and gas pressure Pg in bubbles according to
equations (1.43) and (1.35), the first of which takes diffusion
into account.
Figure 1.14 Effect of hydrogen diffusion for

sound pressures of (a) 5 MPa and (b) 10 MPa in
the presence of developed cavitation for: (I) R q =
100 pm, (II) R q = 10 pm, and (III) R q = 10 pm. (1)
accounting for gas diffusion in the bubble and ( )
2
without diffusion.
For acoustic pressures PA « 1 MPa, diffusion introduces

only small variations in the evolution, whereas for PA > 1 MPa,
it contributes significantly to the gas pressure in the bubble.
We trace this evolution using a bubble with radius R0 =
10 pm as an example. At PA = 0.2 MPa, the bubble pulsates in
a nonlinear mode. The gas diffusion in the bubble increases the
gas pressure by about two orders of magnitude, keeping the
relative radius intact. However, increasing the sound pressure
to 1 MPa drastically changes the situation: one-way diffusion
increases the gas pressure in the bubble by about four orders of
magnitude. An additional increase in the gas pressure occurs at
PA = 10 MPa.
A reduction of the bubble size to R0 = 1 pm also increases
the effect.
We now determine the mass of gas (hydrogen) transferred
from liquid metal to a cavitating bubble. An estimate can be
obtained from approximate relation
(1.44)
with graphical integration of the curves in Figures. 1.13 and

1.14. Here, P is the gas pressure allowing for diffusion, Px is
the gas pressure without the effect of diffusion, m is the
molecular mass of the gas, and R is the average bubble radius
during a rarefaction stage. Calculated hydrogen masses are
summarized in Table 1.3.
Table 1.3 Calculated hydrogen masses in pg.
i?0, (im Pa , MPa
0.2 1.0 5.0 10.0
1 4.0 4.6 300 2000
10 0.4 4.6 300 1000
1000 0.0004 0.7 200 650

The behavior of a cavitating bubble and sound pressure-
dependent hydrogen one-way diffusion in the cavity suggest
that at 18 kHz, the cavitation threshold Pc of liquid aluminum
and magnesium is around 1 MPa. This conclusion immediately
follows from the above data: for PA > Pc, the hydrogen mass in
bubbles with P0 = 1 M-m increases by two orders of magnitude.
A correlation between the calculated data and actual
fracture strength of a liquid metal may be established by
measuring of cavitation threshold in aluminum and its alloys.
1.3.5 Sonoluminescence in Melts
Abramov and Astashkin (1974b) estimated the shock wave
pressure after the cavity collapse in low-melting-point metals
(Sn, Bi, Cd, Pb, and In) and found 9.8 down to 1.0 GPa. Many
researchers relate the collapse to intense cumulative jets with
velocities as high as hundreds of meters per second rather than
to a shock wave. The physics of fluids at the moment when a
cavity collapses is the subject considered up until now. It is
known that cavitation starts in liquid simultaneously with
sonoluminescence (Harvey, 1939). Noltingk and Neppiras
(1950) and Suslick (1990) associate this glowing with increased
temperature (up to 1000°C) in cavities at the point of collapse,
and Frenkel (1945) and Margulis (1984, 1996) associate it with
electric discharge. Crum (1995) provided an additional insight
into this problem. Until recently, it was not clear whether
sonoluminescence occur in molten metals.
Margulis (1996) proved that sonoluminescence is a reliable
indicator of cavitation threshold.
In practice, this method can be used to diagnose incipient
cavitation at earlier stages. Most sonoluminescence studies
dealt with water and other transparent liquids until Kuttruff
(1962) applied this method to cavitation in mercury. Margulis
et al. (1987) applied a modified sonoluminescence method to the
study of cavitation processes in liquid metals.
Sonoluminescence reliably indicated the conditions under which
cavitation starts in a Sn melt sonicated at 22 kHz with an
ultrasonic intensity of 104 W/m2.
Figure 1.15a presents a diagram of relevant experimental
equipment. A corundum crucible (1) with an optical quartz
base (2) was placed in a furnace (3). The horn (4) of a
magnetostrictive radiator (5) connected to an ultrasonic
generator (3) was immersed in the melt from above. One end of a
250mm light guide (7) was attached to the crucible bottom, and
the other end was connected to a photoelectron multiplier (8 ).
Figure 1.15 (a) Setup for studying sonoluminescence in

low-melting-point alloys such as Wood’s and (b) effect of
ultrasound on sonoluminescence.
Sonoluminescence was registered with a photon-counting

system based on a scintillation spectrometer (13).
Simultaneously, cavitation noise was detected with a
hydrophone (0 ) connected to a selective amplifier (10), recorder
(11), and oscilloscope (12). We experimented with (analytical
grade) Wood’s alloy, indium, and (analytical grade) tin without
additional refining.
Experiments showed that a reasonably strong
sonoluminescence occurs in these liquid metals. Its intensity is
by one or two orders of magnitude higher compared to that in
concentrated water solutions of KC1. Glow was easily detected
even without cooling the photoelectron multiplier. The
scintillation rate was as high as 1 0 4- 1 0 5 per second.
Liquid metals are opaque in the visible range; therefore, the
photomultiplier receives light only from the cavity bubbles that
were pushed to the crucible bottom by sound pressure and
emitted light on the internal side of the quartz (2). The
scintillation rate N obtained at 22 kHz with ultrasonic
intensity slowly increased (Fig. 1.15b). This reveals a temporal
characteristic of cavitation phenomena in water and other liquids.
It should be noted that while early stages of cavitation in liquids
(melts) may be associated with thermal and electrical processes
caused by cavity collapse, stable cavitation should be associated
with the heating of the liquid; we will prove this statement later.
1.3.6 Experimental Study of Cavitation in Light Alloys
The cavitation capability was estimated experimentally mainly
for water and water solutions because of the wide application
range of these media in various ultrasonic technologies and the
relative simplicity of the experimental techniques. For liquid
metals and, especially, aluminum and its alloys, the conditions
for cavitation are difficult to estimate quantitatively because
of the high chemical reactivity of melts.
We used cavitation-noise monitoring as a base for
measuring the cavitation threshold (Eskin, 1965). When
cavitation begins, numerous bubbles of various sizes originate
and collapse in the melt, adding to the acoustic signal not only
a main frequency, but also its harmonics, subharmonics, and
incoherent noise. To study the cavitation rigidity of light-alloy
melts, we used the experimental scheme shown in Fig. 1.16a.
In this arrangement based on a commercial generator, we can
continuously adjust the output voltage (from 0 to 500 V) at the
transducer and tune the system resonance frequency.
To feed ultrasound into the melt (up to 0.5 kg), we used a
system of a commercial magnetostrictive transducer, a half
wave conic waveguide made of structural steel, and a
cylindrical radiator made of a high-melting-point material
which slowly dissolves in the aluminum melt.
The acoustical contact between the radiator face and the
melt was improved by metallizing the radiator in the melt
during a preliminary l-2min sonication. The need for the
radiator face to be well wetted by the melt was noted by
Seeman (1972), GrieBhammer (1967), Abramov (1972), and
Eskin (1960).
Figure 1.16 (a) Arrangement for studying the conditions

for cavitation in liquid metals and ( b) characteristic
spectrograms for the main frequency signal (I) before
cavitation, (II) at the cavitation threshold, and (III) in
developed cavitation. (1 -3 ) ultrasonic generators, (4)
magnetostrictive transducer, (5) frequency meter, (d)
tube voltmeter, (7) oscilloscope, (3) cavitometer, (0)
temperature indicator, (10) waveguide sensor, (11)
crucible filled with melt, (12) ultrasonic radiator, (13)
amplitude sensor, (14) voltage regulator of the electrical
furnace, (15) hydrogen analyzer based on the first bubble
method, and (16-17) units for impulse ultrasonic
treatment.
In order to quantitatively estimate the cavitation threshold,

one needs to know the acoustic power transferred to the liquid
metal. Abramov (1972) measured this power indirectly, from
the response of oscillating systems to loading, using the
Teumine method of transferred power2. We measured this
power directly, using a calorimetric method (Eskin, 1988).
2 See Chapter 10.

The measured acoustic power data were related to the
ultrasonic radiator face amplitudes obtained with feedback
sensors hooked up to a tube voltmeter. In subsequent
experiments, we used the voltmeter indications to detect the
acoustic power level at which cavitation started. Acoustic
power was increased in steps of 0.5-1.0 W and each
measurement was repeated at least three times.
Before cavitation starts the sound pressure PA can be
determined from an obvious relation
(1.45)
where N &c is the measured acoustic power, ppc0 is the wave
resistance of the pre-cavitation liquid (melt), and S is the
radiator face area. The table below gives the acoustic power
versus the sound pressure amplitude in commercial aluminum
before cavitation starts:
Nac, W 1.5 3.3 5.0 6.5 7.7 8.4
PA, MPa 0.37 0.55 0.68 0.77 0.88 0.96
Before we began measuring oscillation amplitudes by
feedback voltage, we calibrated the system at no-load, allowing
it to operate in air, with a contact-free vibrometer and a
microscope.
We used a waveguide tester and an oscilloscope to
determine when cavitation started. The waveguide tester with
resonant frequency 200 kHz (significantly above the carrier
18 kHz) was a conic rod of the industrially pure titanium
capped with a piezoceramic receiver of barium titanate. Its
signal was displayed on an oscillograph, so that cavitation
could be observed on the screen as distortions around the main
signal. Oscillograms of developed cavitation differ markedly
from oscillograms of the threshold mode when cavitation
starts.
Spectral cavitometers (electronic instruments showing
relations between the main cw signal, harmonics,
subharmonics, and incoherent noise sensed by the tester in the
cavitation zone) provide important information. Starting from
the assumption that the conversion of the cw signal into all
other oscillations is governed by cavitation, Akulichev and
Makarov (1971) suggested a method for estimating cavitation
and an instrument to measure the dimensionless quantity
(1.46)
where Un and E n are the cavitation noise voltage and energy
respectively, and Ut and Et are the voltage and energy of the
main tone signal; overbars denote averaging.
Unfortunately, because of small non-linear effects in the
oscillating system, such measurements yield non-zero values of
Q even in the absence of cavitation. For this reason, the above
algorithm is unsuitable for studying cavitation thresholds.
Considerable progress was made by Domarkas et al. (1974) who
developed comb filters which separated the incoherent
cavitation noise from all discrete acoustic components.
In the final procedure for estimating the degree of
cavitation with a cavitometer, as shown in Fig. 1.17(a), the
total noise received by the waveguide tester was separated into
two signals: incoherent cavitation noise Un and the sum of the
main tone signal Ut and its harmonics Ug. Squared and
integrated over the observation time, these signals were
transformed into electrical parameters proportional to their
energies. Then the energy of the incoherent cavitation noise E n
was divided by Eg + Et and the result
(1.47)
was displayed by an indicator. In addition one of the
cavitometers used in this arrangement gave a light indication
of the cavitation threshold.
Because the waveguide tester handles discrete and
incoherent components over a relatively wide frequency range,
we also analyzed spectrograms (0-500 kHz) that were obtained
with a recording spectrum analyzer.
Figure 1.17b shows examples of such records for developed
cavitation in various aluminum melts.
Figure 1.17 (a) Block-diagram of cavitometers and (b)

spectrograms of the cavitation noise in various aluminum
melts upon sonication at 18 kHz with amplitude
A = 20 pm at 720°C. (I) noise level for AI O concentration
2 3
from 0.002 to 0.1 wt %; (II) industrially pure Wood’s
alloy ( ) before and (2) after fine filtration.
1
We investigated the cavitation capability as a function of

the metal content using various grades of aluminum and
A1 6 %Mg 0.6% Mn alloy. The composition and characteristics
of these metals are given in Table 1.4.
Table 1.4 Physical and chemical characteristics of investigated

metals
AHoys and AI2O3 Hydrog<3i\ Surface

Metal content
impuirities, wt % (wt) content tension,
cra3/ 1 0 0 1 N/m
Mg Fe Si
A99 (99.9%) 0.003 0 . 0 1 0. 01 l-lO-4 0.2 0.87
A85 (99.85%) 0.1 0.1 510 4 0.25 0.87
A7 (99.7%) 0.15 0.15 510 4 0.3 0.87
A1 6 %Mg 6.0 0.15 0.15 0.7 0.56 0.60
The hydrogen content of aluminum and its alloys was
maintained by processing the melt with a dump asbestos plug.
This parameter was tested by the first bubble method and by
vacuum extraction from a solid sample.
The melt content of solid non-metallic inclusions was varied
with an intense sonication of the melt followed by admixing
synthetic aluminum oxide ((X-AI2O3) with particle size under
1 pm to the melt. The bromine-methanol method was used to
test the content of these inclusions.
Particles of AI2O3 are unwettable in aluminum melts and,
under general conditions, do not mix with the melt and remain
on the surface. To overcome this difficulty, we injected a
heated weighted sample of oxide into the cavitation region near
the oscillating radiator face. This procedure is more efficient if
a focusing radiator is used. In prolonged sonication, the
content of dispersed aluminum oxide can be enhanced from
0.005 wt % to 1.0 wt %. The duration of sonication with
acoustic power below 0.5 kW is as follows:
d, pm 80-100 1 0 - 2 0 1 - 1 0 0 . 1 - 1.0 0 .0 1 - 0.1
r, min 1-2 3-5 8-10 12-15 30-45
It should be noted that prolonged sonication of melts
containing suspended AI2O3 may cause partial dispersion
(erosion) of oxide particles, thus producing abundant interfaces
in the form of Frenkel-Harvey cavitation nuclei. Indeed,
assuming the mass of the melt m\ = 0.5 kg, aluminum oxide
content 0 . 0 1 wt %, the size of oxide particles 2 r = 1.0 pm, and
oxide density y — 3.0 g/cm3, we have:
mass of a single particle: m 2 = 4/37rrsy = 0.810 12 g
total AI2O3 mass in melt: (0.01% = 0.05 g
number of AI2O3 particles: m i/ m 2 = 610iO
When the aluminum oxide concentration increases to 0.1 wt %
and 1 wt %, the number of particles in the melt increases to
1 0 12 and 1 0 13, thus considerably increasing the number of
cavitation nuclei.
Using this procedure, we measured the cavitation threshold
in commercial A7 (1070) aluminum by sequentially varying the
contents of hydrogen and aluminum oxide. The results are
shown in Fig. 1.18. An increased hydrogen content in the melt
lowers the cavitation threshold only at small aluminum oxide
concentrations ranging from 0.0001 wt % to 0.005 wt %. A
Figure 1.18 (a) Cavitation threshold versus the

aluminum oxide concentration in A7 (1070) aluminium
for various hydrogen contents and (b) the degree of
cavitation development versus the ultrasonic
displacement amplitude A for various oxygen contents
(2) 0.03%, (2) 0.008%, (3) 0.005%, and (4) 0.004%.
further decrease in the cavitation threshold is associated
mainly with a lower content of solid non-metallic aluminum
oxide: as the AI2O3 concentration increases from 0.005 to
0.1 wt%, the cavitation threshold decreases by 47%, from
0.8 MPa to 0.55 MPa.
Strictly speaking, equation (1.45) reflects the ultrasonic
energy transfer to a non-cavitating liquid. Therefore, in order
to use it for cavitation threshold estimation, one should
simultaneously measure the radiator face amplitude with a
waveguide tester to see whether or not the initial cavitation
phenomena have begun. Under experimental conditions, a 10-
15mm gap exists between the radiator face and the tester.
Accordingly, the cavitation conditions near the tester surface
differ from those near the radiator face. Therefore, thresholds
calculated with equation (1.45) should be considered as the
upper limit. For the lower limit of cavitation thresholds,
Abramov (1972) recommended taking the wave resistance
reduced by a factor of 3-4. In this case, the cavitation
threshold increases by a factor of 1 .5-2.0.
Figure 1.19a shows the cavitation threshold versus
temperature and melt viscosity for A1 6 %Mg alloy. Near the
liquidus temperature, say at 655°C, the viscosity decreases and
the cavitation threshold increases to 0.8-1.0 MPa, whereas for
overheated melts, at 720°C, the viscosity remains stable at
10-3 Pa s and the cavitation threshold is equal to 0.6-
0.65 MPa, independent of the composition of Al-Mg alloy.
A study of the cavitation capability of Al-Mg alloys with
up to 8 wt % Mg is also of interest. As might be expected,
with an increase in the concentration of magnesium, the
surface tension decreases (see Fig. 1.19b) and the cavitation
threshold falls from 1 MPa for A99 aluminum to 0.75 MPa for
alloys with maximum magnesium content. Al-Mg alloys are
characterized by an enhanced hydrogen solubility: up to
0.6 cm3/ 100 g for 6 wt % Mg. The effect of the hydrogen
content of A1 6 %Mg alloy on the cavitation threshold was also
estimated. Measurements showed in Fig. 1.19(c) demonstrate
that in this alloy, as in pure aluminum, the cavitation
threshold does not really depend on the hydrogen content.
Figure 1.19 Cavitation threshold in Al-Mg alloys versus

(a) temperature, (b) Mg content, and (c) hydrogen
content.
We have already pointed out that hydrogen adsorption can

be considerably enhanced when analytical grade aluminum
oxide is replaced with aluminum oxide contaminated with
transient metals such as Ti, Zr, and Fe. This effect seems to
enhance the nonuniformity in the dissolved hydrogen
distribution and to increase the likelihood of finding enhanced
hydrogen concentrations near liquid metal-solid non-metallic
inclusion interfaces. In turn, the cavitation capability of an
aluminum melt decreases as the amount of transient-metal
inclusions grows.
Figure 1.20 shows the effect of melt temperature and
zirconium content on the cavitation threshold of A99
aluminum. It is apparent that the cavitation threshold
Figure 1.20 The cavitation threshold of A99 (1090)

aluminium melt versus temperature for various
zyrconium contents: ( 1) Al, (2) A1 0.05% Zr, (3)
A1 0.1% Zr, (4) Al 0.2% Zr, and (5) Al 0.3% Zr.
decreases drastically with an increase in transient-metal

content and a decrease in melt temperature. This dependence
on solid unwettable oxide particles, which are the major
additions in aluminum melts and its alloys, means that the
cavitation method can be recommended for on-line control of
liquid metal contamination.
Figure 1.21 shows the cavitation activity of pure aluminum
A5 (1050) and aluminum alloys3 1161 (2324) and 1973 (7475)
melts versus the oxygen content. It is clearly seen that this
technique of measuring the cavitation activity can be used as a
procedure of estimating liquid metals for oxide inclusions.
3 Henceforth we show the US alloy designation in parenthesis.

Figure 1.21 Cavitation activity of the A5 (1050)

aluminium melt and D16 (2024) and 1973 (7475) alloy
melts versus the oxygen content.
1.3.7 Cavitation Field in Melts

As already mentioned, in actual liquids (melts) a multitude
of cavitation bubbles originate even at relatively low acoustic
pressures. These bubbles occupy a volume that will be called
the cavitation region. In order to characterize the evolution of
cavitation, Rosenberg (1968) introduced the cavitation index K
as the ratio of the integral volume AV of all bubbles in the
maximum rarefaction stage to the total volume V, where the
cavitation is studied:
K = AV/V. (1-48)
Cavitation is accompanied by a loss of sound intensity along
the path. If this loss is significant even for paths several
centimeters long, the intensity rapidly falls to the cavitation
threshold and cavitation ceases.
In physical terms, the cavitation index must range within
the limits 0 < K < 1. The lower limit describes the absence of
cavitation and relates to an actual mode, whereas the upper
limit relates to a hypothetical mode, in which the liquid is
completely expelled by bubbles in the rarefaction phase.
Physically, replacing a portion of liquid with bubbles means a
variation of the average density and sound velocity of liquid,
i.e. its wave resistance.
Akulichev (1968) considered the effect of wave resistance
on the cavitation conditions and found that the time-averaged
wave resistance p ccc of a cavitating liquid can be represented
in terms of the wave resistance p $Cq of a non-cavitating liquid
and the time-averaged cavitation index K - 0.1K as
(1.49)
where p^ and are the density and compressibility of the

vapor-gas mix in the bubble and fio is the compressibility of
the liquid. Assuming p^/po << 1, Pb/Po * and K « 1, we
1 0 4
obtain
(1.50)
For 0 < K < 0.1, we can easily correlate the cavitation index
with the rate at which the wave resistance decreases when
cavitation begins in any liquid:
The minimum wave resistance in cavitation was found to be

p ccc = 0.3ppco (Rosenberg, 1968) and p ccc = 0.25poC0
(Kikuchi, 1969) for water, p ccc = 0.25ppc0 for a tin melt
(Abramov and Astashkin, 1974), and p ccc = 0.1/^C q for molten
solder (Abramov and Astashkin, 1974).
Rosenberg (1968) was the first to establish that any
cavitating liquid is a nonlinear medium, so that its average
wave resistance is governed by the ultrasonic power N ac
transferred to the liquid and the oscillation velocity amplitude
(caA)2 on the radiating face of area S:
(1.51)
Therefore, if we know the transferred ultrasonic power, the
oscillation velocity, and the area of the radiating face, we can
estimate pccc .
The acoustic power Nac transferred to the melt was shown
in Fig. 1.2 as a function of the radiating face amplitude A at
18 kHz. This power was obtained by measuring the melt
calorimetry until thermal equilibrium is reached. This method
allowed us to estimate the ultrasonic power transferred during
melt sonication from direct temperature measurements in the
melt heated by ultrasound. Table 1.5 lists average wave
resistances in cavitation calculated from these experimental
data.
Table 1.5 Average wave resistance of an aluminum melt under
sonication at 18 kHz by a radiator of 20mm face diameter.
A, A co9 N ac, >

P c^ c Cavitation
|im m/s W kg-m_2s_1
2 0.228 2.2 17.7 absent,
PoCo = 17 kg m V 1
5 0.56 8.6 2.6 incipient
10 1.29 40 2.4 developed
15 1.69 100 2.7 ditto
20 2.26 200 3.0 ditto
30 3.39 400 2.7 ditto
40 5.52 1000 2.6 ditto
As follows from Table 1.5, the wave resistance drastically

falls (by approximately a factor of 7) when cavitation starts.
However, with a further increase of the radiator face
Figure 1.22 Appearance of a 100pm titanium foil (a)

with evidence of cavitation erosion (x 1) and (b) a single
cavitation break (x 300). The foil remained in the
aluminum melt with developed cavitation for 15 min.
amplitude, it remains almost independent of additional acoustic

power transferred to the melt.
In order to design an ultrasonic procedure for treating the
melt, one should know the geometry and size of the cavitation
region. This geometry can be readily obtained by fixing a
cavitation erosion pattern of thin foil, or measuring the loss of
1.4. Acoustic Flows 55
mass of special samples immersed into the liquid. When
cavitation sets in, the cavitation region enlarges up to a
volume of size a ranging approximately from A/4 to A/2. For
example, this size is 20-40 mm for water and 40-60 mm for
aluminum melts.
We studied cavitation in molten aluminum and its alloys
using 20-500}j.m foil of high-melting-point metals such as Ti,
Nb, Mo, and W. Foil samples were fixed in hard frames and
placed into the melt, below the radiator so that the radiator
axis was in the frame plane. Several experiments were
conducted with the foil set perpendicular to the radiator axis.
Usually, foil made, say, of titanium can remain in a non-
sonicated aluminum melt for an hour without wetting followed
by metallization. When sonication is begun, a high-strength
diffusion layer occurs on the foil surface in 1-3 minutes, and
characteristic punctures, which are evidence of cavitation
erosion, become discernible within 5-15 min (Fig. 1.22).
The degree of cavitation was judged from the number of
punctures in the foil. If the punctures were clustered close
together or continuous areas were eroded from the foil, then
distortions occurred in the quantitative estimations of the
cavitation field. It should be noted that we were never able to
discern punctures until an aluminum layer had been stripped
from the foil surface with 1 0 % caustic soda solution applied
for 2-3 hours.
Our investigation showed that tungsten foil suffered more
damage than molybdenum foil, and the latter undergoes
cavitation diffusion more readily then titanium and, especially,
niobium foil. We obtained similar results when choosing
material for the radiator (Eskin, 1960, 1965, and 1988).
1.4 Acoustic Flows

Sonication generates directed hydrodynamic flows in melts.
These flows are usually called acoustic streams. They occur
both in the bulk of the liquid and near the walls, particles and
other objects within oscillating ultrasonic fields.
The origin of streams relates to the ultrasonic momentum
acquired by the liquid when it absorbs the wave. Therefore, the
velocity of acoustic streams increases with the ultrasonic
intensity and the sound absorption.
Three types of acoustic stream are generally considered
(Zarembo, 1968). The first type is defined as flows in viscous
boundary layers near the solid boundaries or at the phase
interface. Flows of this type will be considered in more detail
in Chapter 9, which is devoted to the zone refining of
aluminum, where flows at the solidification front can be
revealed and studied at ultrasonic intensities much below the
cavitation threshold.
A theory of streaming in boundary layers was developed by
Shlikhting, who stated that boundary layers in which turbulent
boundary flows occur, have a thickness of 5 = (2 v/co)1^2 where
v is the kinetic viscosity and co is the angular frequency of
sound. Therefore, the characteristic scale of these boundary
layers is small compared to the wavelength.
Shlikhting flows may be induced by oscillations of
cavitation bubbles near a solid surface. They are significant in
mass and heat transport. Shlikhting flows are faster than other
types of flow of equal intensity.
The second type of acoustic streaming originates in the
field of standing waves. This is a turbulent motion exterior to
the boundary layer. The characteristic scale of this motion
essentially exceeds the thickness of the boundary layer.
Finally, the third type of acoustic streaming relates to the
absorption of wave momentum in a free bulk of liquid. Here,
the scale of turbulent mass transfer is determined by the
geometry of the bath in which this transfer occurs. These
large-scale flows are slow: their velocity is lower than the
oscillation velocity on the sound radiating face.
It should be noted that when cavitation sets in, the velocity
of acoustic flows can significantly increase. Panov (1984)
reported that streaming velocity may be as high as 2-3 m/s in
the cavitation region.
Little evidence exists for the conditions under which
acoustic flows occur and develop in molten metals and alloys.
This is due to methodological difficulties and, in particular,
Figure 1.23 Large-scale acoustic flows in a liquid bath

of an ingot (270 mm in diameter) of (a) A 7 (1070)
aluminum and ( b) 1960 (7055) Al-Zn-Mg-Cu-Zr alloys: (I)
without and (II) with sonication.
the shielding action of acoustic (ultrasonic) cavitation, which

occurs in melts simultaneously with intense acoustic flows and
at the same ultrasonic intensities. However, these regularities
refer to a lesser degree to flows occurring near and within the
acoustic boundary layer.
Studying the effects of sonication on the temperature
distribution in a liquid bath of an aluminum ingot during
continuous casting, Dobatkin, Eskin, and Borovikova (1973)
found that sonication produces closed acoustic streams with a
loop size comparable with the wavelength (Fig. 1.23, section
II). Acoustic flows homogeneously distribute liquid metal,
superheated above the liquidus temperature, over the bath and
direct the metal to the solidification front. Thus, under
developed cavitation, the sonication of melts develops active
acoustic flows, which not only equalize the bath’ s temperature
and increase its bath volume, but also smooth out the
solidification surface.4
4 The effects of sonication on the temperature field in a liquid bath of

an ingot are described in more detail in Chapter 4.
Without sonication, the motion of the melt in the liquid
bath of an ingot exhibits an entirely different behavior (Fig.
1.23, section I): metal moves from the solidification surface to
the bath. The motion consists of convective flows caused by the
temperature gradient on the surface, in the bath and near the
solidification front.
Abramov et al. (1979) studied acoustic flows in supercooled
melts of organic materials. In this study, sonication was
carried out at frequencies of 45-250 kHz in order to minimize
the effects of cavitation. The model material was chosen so as
to satisfy its similarity in physical and chemical properties to
metallic melts, for which a = Lme\t/(RT0) = 2, where Lmeit is
the melting heat, R is the gas constant, and T0 is the melting
point. The tested material, consisting of naphthalene (35%)
and camphor (65%) had a = 3.5.
The initial substances (naphthalene and camphor) were
refined several times in order to enhance their cavitation
capability and so ensure supercooling of the melt down to 4-
5°C. Photography was used to determine the velocity and scale
of flows. Figure 1.24 shows streaming flows in the supercooled
melt (a) 10 s, (5) 20 s, and (c) 30 s after the onset of sonication
at 139 kHz with oscillation velocity 5 cm/s. Within a few
seconds of sonication with an amplitude above the cavitation
Figure 1.24 Dynamics of acoustic streams in a

supercooled (AT = 4°C) model material (a) 10 s, (b) 20 s,
and (c) 30 s after the onset of sonication at 139 kHz with
oscillation velocity at 5 cm/s. (Abramov et al., 1979)
threshold, solidification nuclei arose near cavitation bubbles.
Analyzing the motion of bubbles and crystals with photograms,
this research team revealed temporal variations in the
geometry of acoustic flows. At the initial stage of sonication,
when the solid phase exists in a small proportion, flows with
the scale L = A prevail, but larger flows with the scale
L = (2.5-3.0)A appear later. An increase in the solid phase
proportion results in the suppression of small-scale flows and
thr development of large-scale flows. With a further increase
in the solid phase proportion, additional flows occur near the
radiating face, in parallel with large- and small-scale flows.
According to Abramov et al. (1979), the flow development
pattern remains stable at sonication frequencies from 45 kHz
to 250 kHz. For frequencies of 45 kHz and 139 kHz, the
maximum velocity omax of the flow, its relative scale L/A, the
acoustical Mach number M = A /c , and the ratio of the flow
velocity to the oscillation velocity were all estimated:
These results indicate that very high velocities are not

characteristic of acoustic flows. The threshold for flow
development almost coincides with the cavitation threshold.
Finally, acoustic flows intensively interact with convective
flows near the solidification front5 if the oscillation velocity
vosc increases considerably (over 8 cm/s for the tested
material). This interaction can equalize the temperature field
in melts, transport solid particles and thus facilitate grain
refinement in solidifying ingots.
These investigations show that the crucial factors
determining the nature, velocity, and scale of streaming are
the ultrasonic intensity (uosc) and the melt temperature
governing its viscosity.
5 We mean the ultrasonic action on the solidification front in a
solidified liquid.
1.5 The Sonocapillary Effect
Konovalov and Germanovich (1962) found that sonicated
liquids penetrate capillary channels with a higher velocity and
to a greater depth. This effect was first studied in metal
cutting, where cooling oil penetrates into a region near the
cutting tool. It appears that liquid fills capillaries faster and
more deeply if it was intensively sonicated just before
penetration. This phenomenon was called the ultrasonic
capillary effect, or the sonocapillary effect.
Many researchers studied the sonocapillary effect in the 60s
and 70s. Experimental data were generalized in the monograph
by Prokhorenko, Dezhkunov, and Konovalov (1981). Among
numerous treatments that explain the effect of ultrasound on
the level of liquids in capillaries, we note the work of
Kitaigorodskii and Drozhalova (1977), who were the first to
prove both theoretically and experimentally the crucial role of
cavitation in the sonocapillary effect. Other ultrasonic
phenomena, such as variations in menisci and the vibration of
capillary walls, may enhance the motion of liquids in
capillaries, but they are far from being crucial factors.
For a number of metallurgical procedures, such as melt
refining, wetting of solid inclusions, and filtering at the
solidification front, the importance of sonocapillary phenomena
cannot be overestimated. As applied to the behavior of liquid
and solidified metal subjected to power ultrasound, the
sonocapillary effect needs further investigation. However, decisive
evidence exists that its role in some procedures is significant.
Considering the effect of cavitation on the level of liquid in
capillaries in a shock-wave induced by cavitation collapse,
Kitaigorodskii and Drozhalova (1977) assumed that: (1 ) a
cavity occurs, grows, and collapses near a capillary repeatedly
in each oscillation period T\ (2) the capillary walls do not affect
the cavity in collapse, so that bubbles remain spherical until
the end of the collapse phase; (3) a shock wave generated by the
collapse acts on the liquid column in the capillary with the
force
(1.52)
1.5. The Sonocapillary Effect 61
where P is the pressure in the shock wave and S is the cross-
section area of the capillary; (4) F acts during a very short
time ( r = 1 0 ps) during the collapse phase, and the liquid
moves then by inertia throughout the rest of the oscillation
period (T - r); and (5) the capillary liquid is treated as a point
of mass m.
They used the momentum conservation law in the simplest
form:
(1.53)
where F is the vector sum of two opposite forces: 2^, which
forces the liquid to penetrate the capillary after cavity collapse
and Ffr, the force of viscosous friction which opposes
penetration and is given by Poiseuille’s formula
(1.54)
where rj is the friction coefficient, H is the liquid column in

the capillary, r is the capillary radius, and v is the velocity of
the liquid.
Recognizing that over a period T the liquid mass increment
in the capillary is several times smaller than the initial liquid
mass, we assume that the mass m = pSH is invariable (p is the
density of the liquid). Then, equations (1.53) and (1.54) may be
reworked to the form:
(1.55)
or
where k = 4 rjH/r'2. Solving equation (1.55) in the interval

0 < t < t, we find the velocity of liquid elevation after a single
pressure pulse:
(1.56)
where t is time, and rrii-i is the mass of liquid in the capillary

at the moment when the ith pulse is applied.
For values of P = 0.46 MPa, r = 2.3 x 10“8 s, and r = 200
pm, which are typical in experiments, the first-order term of
solution expansion will suffice. Then
(1.57)
According to Kitaigorodskii and Drozhalova (1977), the

liquid’s velocity peaks during time r and then moves by inertia
until the moment of the next cavity collapse. In this case, the
increment of the liquid level during one period T will be
(1.58)
so that the resulting level due to the sonocapillary effect will
be
(1.59)
where H q is the column height determined by the surface

tension and n is the number of periods (oscillations) at the
moment of reading.
Kitaigorodskii and Drozhalova (1977) assumed that during
the interval r the liquid in the capillary is subjected to pressure
P equal to Pmax at cavity collapse.
Figure 1.25 presents calculated and measured data
borrowed from this paper. These results indicate a fairly close
agreement with the theory of sonocapillary effect. However,
the maximum pressure Pmax occurs only when the cavity
achieves its minimal radius f?min, which was only 9.0 pm in the
experimental conditions. The maximum radius of such a bubble
is about 0 . 1 - 0. 2 mm, and this fact suggests that in a capillary
1.5. The Sonocapillary Effect 63
Figure 1.25 Elevation velocity V and height H of water

in a glass capillary versus the duration of sonication at
18 kHz: (1) calculated H and (3) measured H; (2)
calculated V and (4) measured V. (Kitaigorodskii and
Drozhalova, 1977)
of radius below 0.5 mm, only one bubble will “work” to raise
the height of the liquid. Clearly, the pressure Pmax and the
shock wave itself will act only if the front of this wave has
arrived at points a distance r away from the cavity center. In
this case, because of spherical divergence, the pressure in the
shock wave decreases by a factor of Rmin/r, and the force
acting on the liquid in a capillary decreases in the same
proportion. The reduction of force F\ = PS will in fact be still
more sensitive because the pressure falls rapidly with the
increasing distance from the cavity center.
Prokhorenko, Dezhkunov, and Konovalov (1981) took into
account these disagreements between theoretical and measured
data and proposed a sonocapillary theory. They included
asymmetry in the boundary conditions for a cavity in collapse,
when it loses its spherical shape and ejects a cumulative jet of
liquid. This cumulative jet is assumed to be responsible for the
elevation of liquid level in capillaries.
These researchers analyzed various situations featuring
collapsing cavities. If a capillary is distant from the radiating
face and the influence of this rigid face on cavity collapse may
be neglected, then for a sufficiently narrow capillary compared
to the cavity size, the jet will be directed into the capillary end
and will raise the liquid by AH. Repeated with a frequency
determined by the probability of cavity occurrence near the
capillary, the cavity collapse and jets will favor the increment
AH and, consequently, will increase the height and the velocity
of liquid elevation, resulting in the sonocapillary effect.
In the interval between two successive jets reaching the
capillary, the liquid can escape from the capillary and decrease
in the column. Prokhorenko, Dezhkunov, and Konovalov (1981)
theoretically and experimentally analyzed cavity collapse near a
capillary and confirmed the asymmetry of collapsing cavities
with cumulative jets.
Hydrodynamic phenomena that play a crucial role in cavity
collapse provide a better insight into phenomena associated
with the sonocapillary effect.
Whatever the conditions that favor an anomalous rise of
liquid in the capillary, the sonocapillary effect accompanies all
phenomena involved in ultrasonic treatment of molten metals.
These are such phenomena as the wetting and activation of
solid non-metallic additions in liquid metals, and the fine
filtration of melts with porous filters. In both processes,
ultrasonic cavitation and the sonocapillary effect accompanied
by cumulative jets provide a quick mass transport into slots
and cracks on the surface of unwettable solid additions, or into
capillary channels of the fine filter.
REFERENCES
Abramov, O.V. (1972) Kristallizatsiya metallov v ultrazvukovom pole

(Crystallization of Metals in Ultrasonic Field),
Moscow: Metallurgiya.
Abramov, O.V. (1994) Ultrasound in Liquid and Solid State , London:
CRC Press.
Abramov, O.V. and Astashkin, Yu.S. (1974a) in: Primenenie novykh
fizicheskikh protsessov dlya intensifikatsii
metallurgicheskikh protsessov (Application of New
Physical Processes to Intensification of
Metallurgical Processes), Moscow: Metallurgiya, p.
156.
Abramov, O.V. and Astashkin, Yu. S. (1974b) Ibid., p. 167.
Abramov, O.V., Astashkin, Yu.S., and Stepanov, V.S. (1979) Akust.
Zh , 25(2), 180-186 [Sov. Phys. Acoust., 25(2), 97-
101 ].
Abramov, O.V., Khorbenko, I. G., and Svehla, S. (1984)
Ultrazvukovaya obrabotka materialov (Ultrasonic
Treatment of Materials), Moscow: Mashinostroenie.
Abramov, O.V. and Teumin, I. I. (1970) in: Fizicheskie osnovy
ultrazvukovoi tekhnologii (Foundations of
Ultrasonic Technology), Ser. Physics and
Technology of Power Ultrasound, Ed. L.D.
Rozenberg, Moscow: Nauka, p. 428.
Abramov, O.V., Teumin, 1.1., Filonenko, V.A., and Eskin, G.I. (1967)
Akust. Zh., 14(2), 161-169 [Sov. Phys. Acoust.,
14(2), 141-149].
Agranat, B.A., Bashkirov B.I., Kitaigorodskii Yu.I., Khavskii, N.N.,
and Eskin, G.I. (1974) Ultrazvukovaya
tekhnologiya (Ultrasonic Technology), Moscow:
Metallurgiya.
Agranat, B.A., Dubrovin, M.N., Khavskii, N.N., and Eskin, G.I.
(1987,) Osnovy fiziki i tekhniki ultrazvuka (Physics
and Technology of Ultrasound), Moscow: Vysshaya
Shkoia; [translation into Spanish, Moscow: Mir
Publishers, 1990; translation into Japanese: Nisso
Tsusinsya, 1993].
Akulichev, V.A. (1968) in: Moshchnye ultrazvukovye polya (Power
Ultrasonic Fields) Ser. Physics and Technology of
Power Ultrasound , Ed by L.D. Rozenberg, p. 129
Akulichev, V.A. and Makarov, L.O. (1971) USSR Inventor’s
Certificate No. 310128.
323
324 References
Al’tman, M.B. (1965) Ne metallicheskie vklyucheniya v aluminievykh
splavakh (Non-metallic Inclusions in Aluminum
Alloys), Moscow: Metallurgiya.
Artman, M.B., Eskin, G. I., Gotsev, I.S., and Stromskaya, N.P. (1981)
in: Legirovanie i obrabotka legkikh splavov
(Alloying and Treatment of Light Alloys), Moscow:
Nauka, p. 30.
Artman, M.B., Slotin, V.I., Eskin, G.I., and Stromskaya, N.P. (1959)
Izv.Akad. Nauk SSSR , OTN, Met. Toplivo, 3. 25.
Al’tman, M.B., Slotin, V.I., Eskin, G.I., and Vinogradova, D.V. (1958)
Izv.Akad. Nauk SSSR , OTN, Met. Toplivo, 9, 25.
Angelov, G. (1970) Izv. Vissho Mash.-Elektrotekn. Inst. (Bulg.), 24, 1,
143.
Ansyutina, A.E., Gur’ev, I.I., and Eskin, G.I. (1970) Izv. Akad. Nauk
SSSR, Met., 4, 76.
Apelian, D. and Mithavasa, B., (1980) J. Met., 9, 14.
Arbuzova, L.A., Slovetskaya, K.I., Rubinshtein, A.M., Kunin, L.L.,
and Danilkin, V.A. (1971) ) l 2 v. Akad. Nauk SSSR ,
Khimiya,, 1, 169
Baikov, A.A. (1948) Collected Works, Moscow: AN SSSR Press, vol. 2
(in Russian).
Balandin, G. F. (1973) Formirovanie kristallicheskogo stroeniya otlivok
(Formation of Crystalline Structure of Ingots),
Moscow: Mashinostroyenie.
Baum, B.A. (1979) Metallicheskie zhidhosti (Metal Fluids), Moscow:
Nauka.
Bergmann, L (1954) Ultraschall und seine Anwendung in Wissenschaft
und Technik, Stuttgart, S. Hirzel.
Bochvar, A.A. (1956) Metallovedenie (Metal Science), Moscow:
Metallurgizdat.
Boguslavsky, Yu.Ya. (1967) Akust. Zh., 13(1), 23-27 [Sov. Phys.
A co u st 13(1), 18-22].
Bondarev, B.I. (1973) Plavka i lifyo deformiruevykh magnievykh
splavov (Melting and Casting of Deformed
Magnesium Alloys), Moscow: Metallurgiya.
Bonderek, Z., et al. (1981) Przeglad nodlewnictwa, 7, 229.
Borisov, B.T. (1987) Teoriya dvukhfaznoi zony metallicheskogo slitka
(Theory of the Two-Phase Zone in Metal Ingots),
Boyle, R.W. (1922) Trans. Roy. Soc. Canada, 3(16), 157.
Bradfield, C. (1950) Proc. Phys. Soc., 65(5), 365.
Bretschneider, E. (1959) German Pat. 1053789.
Burton, J.A., Prim, R.C., and Shuchter,W.P. (1953) J. Chem. Phys.,
21, 1987.
Buxmann, K. (1972) Z. Metallkd., 63(9), 516.
Campbell, I. (1981) Intern. Met. Rev., 2, 71.
Chalmers, B. (1964) Principles of Solidification, New York: John
Wiley.
References 325
Chernov, D.K. (1878) Report in the Imperial Russian-Technical
Society, St. Petersburg.
Chernyshov, I.A. (1953) Liteinoe Proizv., 10.
Chersky, N.A. (1977) DokLAkad. Nauk SSSR, 1, 201.
Chukhrov, M.V., Borovikova, S.I., and Teumin, I.I. (1963) in:
Issledovanie Splavov Tsvetnykh Metallov (Study of
Alloys of Nonferrous Metals), no. 4, Moscow: Izd.
AN SSSR, p. 141.
Chvorinov, N.N. (1954) Krystalisace a nestejnorodost oceli, Praha:
Nakladatelstvi Czeskoslovenske Akademie Ved (in
Czech).
Crawford, A.E. (1955) Ultrasonic Engineering with References to High
Power Applications, London: Butterworths.
Crum, A. (1995) Ultrason. Sonochem., 2(2), 147.
Danilov, V.I. (1956) Stroenie i kristallizatsiya zhidkostei (Structure
and Solidification of Liquids), Kiev: Izd. AN
USSR.
Danilov, V.I. and Chedzhenov, G.Ch. (1955) in: Problemy
metallovedeniya i fiziki metallov (Problems in
Metal Science and Physics of Metals), Issue no. 4
of Collected Papers of Central Research Institute
of Pure Materials (TsIIChM), p. 152.
Dobatkin, V.I. (1948) Nepreryvnoe lit'yo i liteinye svostva splavov
(Continuous Casting and Properties of Cast
Alloys), Moscow: Oborongiz.
Dobatkin, V.I. (1960) Slitki aluminievykh splavov (Ingots of
Aluminum Alloys), Moscow: Metallurgizdat.
Dobatkin, V.I., Belov, A.F., Eskin, G.I., Borovikova, S.I., and Gol’der,
Yu.G. (1984) Scientific discovery “Regularity of
crystallization of metal materials”, No. 271,
Vestnik Akad. Nauk SSSR, 1, 137.
Dobatkin, V.I. and Elagin, V.I. (1980) Granuliruemye Aluminievye
Splavy (Pelletized Aluminum Alloys), Moscow:
Metallurgiya.
Dobatkin, V.I. and Eskin, G.I. (1985) in Metallovedenie aluminievykh
splavov (Metal Science of Aluminum Alloys),
Moscow: Nauka, p. 163.
Dobatkin, V.I. and Eskin, G.I. (1986) in: Vozdeistvie moshchnogo
ultrazvuka na mezhfaznuyu poverkhnost metallov
(Effect of Power Ultrasound on Interphase Surface
of Metals), Moscow: Nauka, p. 6.
Dobatkin, V.I. and Eskin, G.I. (1990) Proc. 2nd Int. Conf. on
Aluminum Alloys, Beijing (China), 278.
Dobatkin, V.I. and Eskin, G.I. (1991a) in: SpetsiaVnye sposoby lit'ya
(Special Casting Methods) Ed. V.A. Efimov,
Moscow: Mashinostroyeniye.
Dobatkin, V.I. and Eskin, G.I. (1991b) Tsvet. Met., 2, 64
Dobatkin, V.I. and Eskin, G.I. (1992) Protsessy Lit., 1, 38.
326 References
Dobatkin, V.I. and Eskin, G.I. (1995) Izv. Russ. Akad. Nauk, Met., 4,
36.
Dobatkin, V.I. and Eskin G.I. (1996a) Tsvet. Met., 2.
Dobatkin, V.I. and Eskin, G.I. (1996b) Proc. 4th Int. Conf. on Semi-
Solid Proc. of Alloys and Comp., Sheffield, UK.
Dobatkin, V.I., Eskin, G.I., and Borovikova, S.I. (1973) Fizikokhim.
Obrabot. Mater., 6, 37.
Dobatkin, V.I., Eskin, G.I., Ber, L.B., Borovikova, S.I., Gol’der, Yu.G.
and Klimovich, L.G. (1983) Izv. Akad. Nauk SSSR,
Met., 2, 130.
Dobatkin, V.I., Eskin, G.I., Borovikova, S.I. et al. (1986b) Patent USA
No. 4564059, B22D 27/02.
Dobatkin, V.I., Eskin, G.I., Borovikova, S.I., and Gol’der, Yu.G.
(1976b) in: Obrabotka legkikh i zharoprochnykh
splavov (Treatment of Light and Heat-Resistant
Alloys), Moscow: Nauka, p. 151.
Dobatkin, V.I., Gabidulin, R.M., Kolachev, B. A., and Makarov, G.S.
(1976a) Gasy i okisly v aluminievykh deformirue-
mykh splavakh (Gases and Oxides in Deformed
Aluminum Alloys), Moscow: Metallurgiya.
Domarkas, V.I., Patashus P., and Sayuuskas, S.I. (1974/ USSR
Inventor’s Certificate No. 545923.
Efimov, V.A. (1976) Razlivka i kristallizatsiya stali (Pouring and
Solidification of Steel), Moscow: Metalurgiya.
Efimov, V.A. and El’darkhanov, A.C. (1995) Fizicheskie metody
vozdeistviya na protsessy zatverdevaniya (Physical
Effects on Solidification), Moscow: Metallurgy.
Eisenreich, H. (1950) Die Technik, 5, 310.
Elagin, V.I., Zakharov, V.V., and Rostova, T. D. (1982) Tsvet. Met.,
12, 96.
Eller, A. and Flynn, H.G. (1965) J. Acoust. Soc. Amer., 37, 3, 493.
Eskin, G.I. (1960) Ultrazvuk in metallurgy (Ultrasound in
Metallurgy), Moscow: Metallurgizdat.
Eskin, G.I. (1965) Ultrazvukovaya obrabotka rasplavlennogo aluminiya
(Ultrasonic Treatment of Molten Aluminum),
Eskin, G.I. (1988/ Ultrazvukovaya obrabotka rasplavlennogo
aluminiya (Ultrasonic Treatment of Molten
Aluminum), 2nd ed., Moscow: Metallurgiya.
Eskin, G.I. (1992) Obrabotka i kontrol kachestva tsvetnykh metallov
ultrazvukom (Treatment and Quality Control of
Nonferrous Metals with Ultrasound), Moscow:
Metallurgiya.
Eskin, G.I. (1993) Proc. of Int. Symp. on Light Materials in
Transportation Systems, Kyongju, Korea, p. 585.
Eskin, G.I. (1994a) Proc. of Int. Conf. on High Temp. Capillarity,
Bratislava, Slovak Republic, p. 321.
Eskin, G.I. (1994b) Ultrason. Sonochem., 1(1), 59.
References 327
Eskin, G.I. (1994c) in: Tekhnologiya obrabotki lyogkikh i spetsiaVnykh
splavov (Technology of Treating of Light and
Special Alloys), Moscow: Metallurgiya, p. 136.
Eskin, G.I. (1995a) Ultrason. Sonochem. 2(2), 137.
Eskin, G.I. (1995b,) in: Metallovedeniye, lit'yo i obrabotka lyogkikh
splavov (Metal Science, Casting and Treatment of
Light Alloys), Moscow, VILS, p. 167.
Eskin, G.I. (1995e) Proc. 4th Intern. Conf. on Powder Metallurgy in
Aerospace, Defense and Demanding Applications,
Anaheim (USA).
Eskin, G.I. (1995f) Proc. of 14th Int. Conf. on Utilization of Ultrasonic
Methods in Condensed Matter, Zilina, Slovak
Republic, p. 12.
Eskin, G.I. (1996) in: Advances in Sonochemistry, vol 4, Ed. T.J.
Mason, Greenwich: JAI Press, pp. 101-159.
Eskin, G.I. and Fridlyander, I.N. (1961) Izv. Akad. Nauk SSSR, OTNt
Met. Topi., 5, 110.
Eskin, G.I., Gur’ev, 1.1., and Soluyanov, Yu.F. (1981) Tsvet. Met., 1,
68.
Eskin, G.I., Ioffe, A.I., and Shvetsov, P.N. (1972) Izv. Akad Nauk
SSSR , Met., 6, 69.
Eskin, G.I., Ioffe, A.I., and Shvetsov, P.N. (1974) Tekhnol. Legk.
Splav., 1, 8.
Eskin, G.I., Kudryashov, V.G., Shvetsov, P.N., Kyzminskya, S.K. ,
Skotnikov, I.A., and Petrov, A.D. (1990) Izv.
Akad. Nauk SSSR, Met., 1, 53.
Eskin, G.I., Makarov, G.S., and Pimenov, Yu.P. (1995d,) Advanced
Performance Materials, 2, 43.
Eskin, G.I., Makarov, G.S., Pimenov, Yu.P., Khavskii, N.N. and
Yalfimov, V.I. (1995c) Izv. Vyssh. Uchebn. Zaved.,
Ser. Tsvet. Met., 2, 14.
Eskin, G.I., Neshpor, G.S., Kopeliovich, B.A., Skotnikov, I.A., and
Petrov, A.D. (1984) Probl. Prochn., 2, 25.
Eskin, G.I. and Pimenov, Yu.P. (1996) Tekhnol. Legk. Splav., 2, 52.
Eskin, G.I., Pimenov, Yu.P., Belova, M.V., and Skotnikov, I.A. (1992)
Proc. of 14th Int. Cryogenic Materials Conf., Kiev,
Ukraine, p. 81.
Eskin, G.I., Pimenov, Yu.P., Makarov, G.S. , Belova, M.A., and
Burov, V.P. (1989) Izv. Akad. Nauk SSSR, Met., 4,
87.
Eskin, G.I. and Ryzhenkova, M. P. (1975) in: Sovremennye tsvetnye
splavy i progressivnye metody litya (Modern
Nonferrous Alloys and Advanced Casting
Technologies), Moscow: MDTNP, p. 147.
Eskin, G.I., Slotin, V.I., and Katsman S. Sh. (1967) Tochnoe IWyo
detalei aviatsionnykh agregatov iz aluminievykh
splavov (Precise Casting of Aviation Parts from
Aluminum Alloys), Moscow: Mashinostroyeniye.
328 References
Ferraro, R.I. and McLellen, R.I. (1979) Metall. Trans. A, 10A(11),
1699.
Flemings, M.C, (1991) Metall. Trans. B , 22B(6), 269.
Flemings, M.C. (1974) Solidification Processing, New York: McGraw-
Hill.
Flynn, H.G. (1964) Physics of Acoustic Cavitation in Liquids, in:
Physical Acoustics, ed. by W. Mason, vol. 1 pt. B,
New York: Academic Press, pp. 58-172.
Frenkel, Ya.I. (1945) Kinetic heskay a teoriya zhidkosti (Theory of
Fluid), Moscow: Izd. AN SSSR.
Gabathuler, J.-P. and Buxmann, K. (1993) Patent Shweiz. CH 682402,
C22F, 3/02.
Getselev, Z.N., Balakhontsev, G.A., Kvasov, F.I., Cherepok, G. V.,
Varga, 1.1., and Martynov, G.D. (1983)
Nepreryvnoye lit*yo u elektromagnitnyi kristalliza-
tor (Continuous Casting in an Electromagnetic
Mold), Moscow: Metallurgiya.
Gibbs, I.W. (1928) Coll. Works, 252.
Gorev, K. V. and Shevchyuk, L.A. (1961) Litein. Proizv., 4, 17.
Grant, N.J. (1983) J. Metals, 1, 20.
Griephammer, A. (1967) Ultrasonics, 5(4), 229.
Gumenyuk, V.S. (1958) USSR Inventor’s Certificate No. 116596, 31c,
Gur’ev, 1.1., Eskin,15oi-
G.I. and Ansyutina, A.E. (1978) in: Magnievye
splavy (Magnesium Alloys), Moscow, Nauka,
p. 136.
Gur’ev, 1.1., Eskin, G.I., Karan, A.B., Ansyutina, A.E. and Osokina,
T.N. (1975) Tekhnol. Legk. Splavov, 12, 25.
Harvey, E.W. (1939) Amer. Chem. Soc., 61, 2392.
Harvey, E.W., McElory, W.D., and Whiteley, A.N. (1947/ J. Appl.
Phys., 18, 162.
Hem, S.L. (1967) Ultrasonics, 5, 4, 202.
Herrmann, E. (1958) Handhuch des Stranggieflens, Dusseldorf:
Aluminum Verlag.
Herrmann, E. and Hoffmann, D. (1980) Handbook on Continuous
Casting, Dusseldorf: Aluminium Verlag.
Hiedemann, E.A. (1954/ J.Acoust. Soc. Amer., 26, 5.
Il’ichev, V.I. (1968) Akust. Zh., 14(2), 163-169 [Soy. Phys. Acoust.,
14(2), 135-140].
Kapustin, A.P. (1962) Vliyanie ultrazvuka na kinetiku kristallizatsii
(Effect of Ultrasound on the Kinetics of
Solidification), Moscow: Izd. AN SSSR.
Kapustina, O.A. (1970) in: Fizicheskie osnouy ultrazvukovoi
tekhnologii (Foundations of Ultrasound
Technology), Ser. Physics and Technology of Power
Ultrasound, ed. by. L.D. Rozenberg, Moscow:
Nauka, p. 253.
Kattamis, T.Z., Holmberg, U.T., and Flemings, M.C. (1967) J. Inst.
Met., 95(11), 343.
References 329
Kattamis, T.Z. and Skolianos, St. (1985,) Proc. 5th Int. Conf. on
Rapidly Quenched Metals, Amsterdam: North-
Holland, 1, 51.
Kattamis, T.Z. and Williamson, R.B. (1968) J. Inst. Met., 96(8), 251.
Kazachkovskii, O.D. (1948) Sbornih nauchnykh rabot laboratorii
matallofiziki (Collected Papers of the Laboratory
of Metal Physics), Ed. V.I. Danilov, Kiev: Izd. AN
USSR, p. 76.
Kikuchi, Y. (1969) Ultrasonic Transducers, Tokyo Corona Publ. Co.
Kitaigorodskii, Yu.A. and Drozhalova, B.I. (1977) Primenenie
ultrazvuka v metallurgii (Collected papers of
Moscow Inst, of Steel and Alloys)., No. 90,
Moscow: Metallurgiya, p. 12.
Klyachko, Yu.A. (1941) Zh. Prikl. Khim., 14(3), 342.
Konovalov, E.G. and Germanovich, I.K. (1962,) Dokl. Akad. Nauk
BSSR , 6(8), 492.
Kornfeld, M. (1950) Uprugost i prochnost zhidkostei (Elasticity and
Strength of Fluids), Moscow: GITTL.
Kratky, J. (1979) Trend VUMA, 10(1), 28.
Kruger, F. (1936) Glastechn., 16, 233.
Kunin, L L., Pimenov, Yu.P., Arbusova, L.A., and Danilkin, V.A.
(1971) in: Fizicheskaya khimiya poverkhnostnykh
yavlenii v rasplavakh (Physical Chemistry of
Surface Phenomena in Melts), Kiev: Naukova
Dumka, p. 99.
Kurdyumov, A.V., Inkin, S.V., and Chulkov, V.S. (1980) Flusovaya
obrabotka i filtrovanie aluminievykh rasplavov
(Flux Treatment and Filtering of Aluminum
Melts), Moscow: Metallurgiya.
Kuttruff, H. (1962) Acustica, 12, AB 231.
Lane, D.H. (1965; US Patent 3193889.
Lane, D.H., Cunningham, J.W., and Tiller, W.A. (1960) Trans.
Metall. Soc. AIME , 218(12), 985.
Lane, D.H. and Tiller, W.A. (1960) Trans. Metall. Soc. AIME ,
218(12), 991.
Levi, L.I., Vygovskii E.V., and Karaseva, D.T. (1968) Izu. Vyssh.
Uchebn. Zaved.t Chern. Met., 5, 5.
Lin Chun Mao (1979) Uli, 8, 5, 467 (in Chinese).
Lindstrome, O. (1955) J.Acoust. Soc. Amer., 27(4), 654.
Lorimer, J.P. and Mason, T.J. (1995; Ultrason. Sonochem., 2(2), 79.
Makarov, G.S. (1983) Rafinirovanie aluminievykh splavov gazami (Gas
Refining of Aluminum Alloys), Moscow:
Metallurgiya.
Makarov, G.S., Khodokov, P.E., Valova, G.A. and Ponomarev, V.D.
(1977; Tekhnol. Legk. Splav., 11, 53.
Makarov, G.S., Pimenov, Yu.P., and Eskin, G.I. (1995) in:
Metallovedenie, lit’yo i obrabotka splavov (Casting
and Treatment of Alloys), Moscow: VILS, p. 187.
330 References
Margulis, M.A. (1996) Sonochemistry and Cavitation , Amsterdam:
Gordon and Breach.
Margulis, M.A., Grundel, L.M., Shvetsov, P. N., and Eskin, G.I.
(1987) DoklAhad. Nauk SSSR , 295(5), 1170.
Maslan, L.B., Levi, L.I., and Knizhnin, G.S. (1966) Litein. Proizv ., 5,
37.
Mason, T.J. (1991) Practical Sonochemistry—a User*s Guide to
Applications in Chemistry and Chemical
Engineering , Chichester: Ellis Horwood.
Mason, T.J., Ahmad, J., Bacon, M., and Sterritt, A. (1994),
M etallurgia , 61, 224.
Matauschek, J. (1959) EinfUhrung in die Ultraschalltechnik , Berlin:
Verlag Technik.
Matyia, H., Giessen, B.C., and Grant, N.I. (1968) J. In st . M et ., 96(1),
30.
Minnaert, M. (1933) Philos. M ag., 16(7), 235.
Morris, D.G. (1982) M et. Set, 16(10), 457.
Mortensen, A. and Jin, I. (1992) Solidification processing of Metal
Matrix Comp., Int. Mater. Rev., 37(3), 101.
Nakanishi, H, Tsanekawa, Y., and Mohri, N. (1993,) Met. Abstr. Light
Met. Alloys, Osaka, 26, 141.
Noltingk, B.E. and Neppiras, E.A. (1950) Proc. Phys. Soc ., 63, 674.
Novikov, 1.1. (1966) Goryachelomkost tsvetnykh metallov i splavov
(Hot-Shortness of Nonferrous Metals and Alloys),
Moscow: Nauka.
Novitskii, B.G. (1983) Primenenie akusticheskikh kolebanii v khimiko-
tekhnologicheskikh protsessakh (Acoustic
Oscillations in Chemical and Technological
Processes), Moscow: Khimiya.
Panov, A.P. (1984) Ultrazvukovaya ochistka pretsizionnykh detalei
(Ultrasonic Refinement of Precision Parts),
Moscow: Mashinostroenie.
Patterson, R.I., Cox, A.R. and Van Reath, E.C. (1980) J. M et., 32(9),
34.
Petrov, D.A. (1947) Zh. Fiz. Khim ., 21, 12.
Pfann, W.G. (1958) Zone M elting, New York: John Wiley.
Pikunov, M.V. (1979) Tekhnol. Legk . Splavov, 8, 17.
Pimenov, Yu.P. and Demenkov, A.I. (1972) Tekhnol. Legk. Splavov, 6,
33.
Pogodin-Alekseyev, G.I. and Gavrilov, V.M. (1966) Izv. Akad. N auk ,
Ser. M et., 1, 27.
Pohlman, R., Heisler, K. and Cichos, M. (1974) Ultrasonics, 1, 11.
Polotskii, I.G. (1960) Primenenie ultrazvukovykh kolebanii dlya
issledovaniya svoistv , kontrolya kachestva i
obrabotki metallov i splavov, Kiev: Izd. AN USSR.
Popel, P.S. and Eskin, G.I. (1995) Proc. 14th Int. Conf. on Utilization
of Ultrasonic Methods in Condensed M atter,
Zilina, Slovak Republic, p. 45.
References 331
Popel, P.S., Pavlov, V.A., Konovalov, G.E., et al. (1985) Tsvet. Met.,
7, 68.
Prokhorenko, P.P., Dezhkunov, N.V. and Konovalov, G.E. (1981)
Ultrazvukovoi kapillyarnyi effekt (Ultrasonic
Capillary Effect), Minsk: Nauka i Tekhnika.
Richards, W.T. (1938) J. Amer. Chem. Soc., 54, 479.
Rozenberg, L.D. (1968) Cavitation region, in: Moshchnye
ultrazvukovuye polya (Power Ultrasonic Fields)
Ser. Physics and Technology of Power Ultrasound,
Ed. L.D. Rozenberg, Moscow: Nauka, 221.
Rozenberg, L.D. and Makarov, L.O. (1957) ffAH CCCP, 114(2), 275.
Rutter, J.W. (1958) Liquid Metals and Solidification, Cleveland:
ASM.
Samoilovich, Yu.A. (1977) Formirovanie slitka (Formation of an
Ingot), Moscow: Metallurgiya.
Schmid, I. (1959) Acustica, 9, 321.
Schmid, G. and Ehret, L. (1937) Zt. Elehtrochem., 43, 869.
Seemann, H.J. (1936) Metallforschung mit Ultraschall, Metallwirt.,
15, 1067.
Seemann, H.J. and Buxmann, K. (1972) Metallwiss Techn., 1, 1.
Seemann, H.J. and Dorr, W. (1976) Met. Techn., 1, 1029.
Seemann, H.J. and Menzel, H. (1947) Zt. Metall., 1, 3/4, 39.
Seemann, H.J. and Pretor , K.G. (1966) Zt. Metall., 57(5), 347.
Seemann, H.J., Staats, H., and Pretor, K.V. (1967) Archiv fiir
Eisenhut., 38(4), 257.
Sergeev, S.V. (1952) Fiziko-khimicheskie svoistva zhidkikh metallov
(Physical and Chemical Properties of Liquid
Metals), Moscow: Oborongiz.
Sharov, M.V. and Pimenov, Yu.P. (1970) in: Metallovedenie splavov
legkikh metallov (Metal Science of Light Alloys),
Moscow: Nauka, p. 80.
Sinyavskii, V.S., Val’kov, V.D. and Kalinin, V.D. (1984) Korroziya i
zashchita aluminievykh splavov (Corrosion and
Protection of Aluminum Alloys), Moscow:
Metallurgiya.
Sirotyuk, M.G. (1968) in: Moshchnye ultrazvukovye polya (Power
Ultrasonic Fields), Ser. Physics and Technology of
Power Ultrasound, Ed. L.D. Rozenberg, Moscow:
Nauka, p. 167.
Slotin, V.I. and Eskin, G.I. (1957) Izv. Akad. Nauk SSSR, OTN, Ser.
Met. Topi., 9, 33.
Sokolov, S.Ya. (1936) Zh. Tekh. Fiz., 3, 176.
Surensen Ch. (1936) Ann. Phys., 5, 26.
Suslick, K.S. (1990) Sonochem. Science, 247, 1439.
Svehla, S (1978) Trend Vuma, 9, 38.
Tammann, G. (1931) Zt. Anorg. Algem. Chem., 200, 57.
Taylor, G.B. and Sproule, D.O. (1929) Trans. Roy. Soc. Canada, 3(23),
91.
332 References
Tetyukhin, V.V., Kurilov, V.A., and Denisov, D.R. .(1977) Titan.
Metallovedenie i tekhnologiya (Titanium: Science
and Technology), Moscow: VILS, p. 93.
Teumin, I.I. (1959) Ultrazvuhovye holebatelnye ststemy (Ultrasonic
Oscillatory Systems), Moscow: Mashgiz.
Teumin, I.I. (1962a) Akust. Zh ., , no. 3.
Teumin, I.I. (1962b) Tehhn. Inform. Bull. OKB-UVU, no. 1(19), 9.
8
Veinic, A.I. (1959,} Teploobmen mezhdu slitkami i izlozhnitsami (Heat

Transfer from Ingots to Lingots), Moscow:
Metallurgizdat.
Volmer, M. and Weber, W. (1926) Zt. Phys. Chem.t 119, 277.
Willard, G.W. (1953) J.Acoust. Soc. Amer., 25(4), 277.
Winegard, W.A. (1964) An Introduction to the Solidification of
Metals, London: Inst, of Metals.
Wood, R.W. and Loomis, A. L (1927) Phil. Mag.t 7(4), 417.
Zarembo, L.K. (1968) in: Moshchnye ultrazvuhovye polya (Power
Ultrasonic Fields), Ser. Physics and Technology of
Power Ultrasound, Ed. L.D. Rozenberg, Moscow:
Nauka, p. 87
Zeldovich, Ya.B. (1942) Zh. Eksp. Tekh. Fiz., 12(11-12), 525.

Googlepreview PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Googlepreview PDF

Uploaded by

Copyright:

Available Formats

Ultrasonic Treatment of Light Alloy Melts

Ultrasonic Treatment of Light Alloy Melts

All-Russia Institute of Light Alloys

Boca Raton London New York

CRC Press is an imprint of the

Published 2018 by CRC Press

© 1998 by Taylor & Francis Group, LLC

No claim to original U.S. Government works

ISBN 13: 978-90-5699-042-8 (hbk)

Trademark Notice: Product or corporate names may be trademarks or registered

Visit the Taylor & Francis Web site at

and the CRC Press Web site at

British Library Cataloguing in Publication Data

This book contains the results of studies we conducted in the field of

Metallurgy and metal science have made considerable achievements in

1.1 Ultrasound and Low-Frequency Vibration

During one period T = 1//, the weight of this elementary

In a liquid, where elastic properties depend upon compression,

where A is the displacement of particles of the medium and t is

j = - A qco 2 sin cox/c , (1.12)

A quantity of prime importance for ultrasonic field

1.2 Propagation of Power Ultrasound in Melts

Figure 1.1 Relative acoustic resistance K versus the

As a rule, only longitudinal waves propagate in a liquid

Figure 1-2 Power W & transferred into the melt versus

pattern is narrow. For Drad/Z > 1, the directivity pattern

Figure 1.3 Directivity patterns of ultrasonic sources.

Unfortunately, no instrument currently exists to measure

Figure 1.4 Absorption coefficient a of ultrasound

molten metals these are molds, feeders, and continuous casting

The transmission coefficient

Figure 1.5 Reflection coefficient R and transmission

1.3 Acoustic Cavitation in Liquid Metals

Figure 1.6 Cavitation threshold of water versus

consider basic models for cavitation nuclei in order to choose

Figure 1.7 Model of a cavitation nucleus (Frenkel, 1945;

Figure 1.8 Cavitation rigidity of water versus (a)

Figure 1.9 Cavitation rigidity of water versus the AL O

viscosities 77, the cavitation threshold Pc increases. Figure 1.9

Table 1.2 Material constants of aluminum and water

Constant Aluminum Water

Surface tension cr, N/m 0.860 0.079

We used the Runge-Kutta method to solve this set

Figure 1.10 Evolution of cavities with initial radii (a)

where all cavities expand during one or two periods of the

Figure 1.11 Evolution of cavities with initial radii (a)

This analysis of bubble dynamics did not take into account

Figure 1.12 Effect of surface-tension on the evolution of

that of water), conditions for the one-way gas diffusion from

The gas flow in the bubble may be governed by various

Thus, one should simultaneously solve equations (1.37),

where y is the coordinate perpendicular to the bubble surface.

where C q is the equilibrium gas concentration independent of

where Z0 = 0.8PA, then the set of equations describing the

This system yields a numerical solution for initial values

Figure 1.13 Relative radius R/R q and pressure P in

0 . 3 - 1 . 5 pm unwettable particles of aluminum oxide powder

Figure 1.14 Effect of hydrogen diffusion for

For acoustic pressures PA « 1 MPa, diffusion introduces

with graphical integration of the curves in Figures. 1.13 and

1 4.0 4.6 300 2000

10 0.4 4.6 300 1000

1000 0.0004 0.7 200 650