You are on page 1of 9

I n t J. M a c h Tool. Des. Res.

V o l 19, p p 33 41, Pergamon Press L t d 1979 Printed in Great Britain

ON THE MECHANICS OF MATERIAL REMOVAL


IN ULTRASONIC MACHINING

G . S. KAINTH,* AMITAV NANDY'I" a n d KULDEEP SINGH:~

(Received 15 March 1978)

Abstract - An analysis of material removal in ultrasonic machining considering direct impact of abrasive grains on
the workpiece is presented. Non-uniformity of abrasive grains is considered by using a probability distribution for
the diameter of the abrasive particles as suggested by Rozenberg. The analysis is applied to calculate material
removal rate for the case of glass using 400 mesh Norbide abrasive and mild steel tool for various values of static
force and amplitude of tool oscillation.

NOMENCLATURE
ao amplitude of vibration
D diameter of tool
d diameter of abrasive particle
d mean diameter of abrasive
dm maximum diameter of abrasive particles of a particular mesh size
do minimum diameter of abrasive particles of a particular mesh
f frequency of vibration
F~ contact load on a grit
e~ static load
e(d) frequency distribution of abrasive particles
H~ hardness of tool
H~ hardness of workpiece
h depth of indentation
Me impulse due to contact with grains
M, impulse due to static load
N total number of abrasive particles in the gap
N, total number of abrasive particles in the gap at time t
q = Hw/H ,; hardness ratio
rw radius of the indentation zone in workpiece
Vo volume fractured per grit
Vd metal removal rate due to all the particles of size d
Vg volume of working gap
total metal removal rate
x distance between tool and work at end of stroke
Xt distance between the tool and work at time 't"
Y~ distance of tool from its mean position at start of impact
W average weight of abrasive particles
w, weight of abrasive in the gap at time t
w~ weight of an abrasive particle of diameter d
depth of indentation in tool
depth of indentation in workpiece
0 angular position of tool
P~ density of abrasive particles
Pl density of fluid carrying abrasive
gO frequency in rad/sec.

1. I N T R O D U C T I O N
ULTRASONIC machining is one of the methods used for machining hard materials, e.g.
tungsten and titanium carbides, die and tool steels, and brittle materials such as germanium,
silicon, ferrites, ceramics, glass, and quartz.
In ultrasonic machining the material is removed primarily by blows from hard abrasive
grains. The abrasive particles are under the control of a tool which is vibrated at an ultrasonic

* Assistant Professor, Department of Mechanical Engineering, I.I.T. Kanpur, India.


f Systems and Planning Division, Bharat Heavy Electricals Ltd., Jhansi, India.
Ph.D. Student, I.I.T. Kanpur, India.

33
34 G.S. KAINTH, AMITAVNANDY and KULDEEPSINGH

frequency with a small amplitude. A continuous flow of abrasive slurry flushes away particles
removed in the process. This also enables fresh abrasive grains to replace the particles which
fracture in the process. As the tool vibrates at a high frequency and a large number of particles
are active at a time, the rate of material removal is sufficient for practical purposes.
Miller [1] in his semi-quantitative discussion on the rate of cutting in ultrasonic machining
assumed that the rate of material removal depends upon work-hardening and plastic
deformation of the workpiece. This implies that his analysis is mostly applicable to plastic
materials, however, most of the materials machined by ultrasonic machining are brittle.
Moreover, he also assumed that all the grains are cubes of the same size and that all the grains
take part in cutting.
Shaw [2] assumed that material is removed primarily by two mechanisms due to
(a) direct impact of the tool on the grains in contact with the workpiece and tool, and
(b) impact of grains accelerated by the vibrating tool.
He showed that only a small fraction of material removed is due to the impact of moving
grains and the bulk of the material is removed by direct impact of the tool. He assumed that
the rate of material removal "v" is proportional to the volume V of material dislodged per
impact, the number of particles N making impact per cycle and the frequency of tool
vibration "f". Assuming the grains to be identical spheres of mean grain diameter d, Shaw
expressed the material removal rate as:

v oc [dh]3/2Nf, (1)

where h is the depth of indentation. The depth of indentation was found by equating the mean
static force to the mean force of impact of the tool on the grains. Assuming that the number of
particles in the working gap is inversely proportional to the square of the diameter of each of
the particles for a tool of fixed area, he [2] gave the following expression for h:

8F.yod_ 11f2 (2)


h = L z K H C ( 1 + q)j ,

where, F s is the static force,


Yo is the amplitude of vibration of the tool,
H is the hardness of the workpiece,
q is the ratio of the hardness of the workpiece to that of the tool,
C is the concentration of the abrasive slurry and
K is a constant of proportionality.
Shaw's conception of the process has been confirmed by experiments using high-speed
cinematography [3]. However, Shaw's analysis does not agree qualitatively with experimen-
tal results. Furthermore, the proportionality constant K can be evaluated by experiments
only.
Dikushin and Barke [4, 5] tried to relate the energy consumed in removing the material
from the workpiece, to the amplitude and force of vibration, using the laws of conservation of
energy and momentum. The theory, however, failed to predict quantitatively the material
removal rate.
Kazantsev [6] measured contact force during machining and concluded that machining
rate varies linearly with static load up to a certain optimum load.
Rozenberg [7] considered a statistical distribution for the abrasive grain size d based on
experimental evidence as:

F(d) = 1.095~- 1- - 1 (3)

where, aY is the mean diameter of the particles in the working gap and
N is the number of particles in the working gap.
Material Removal in Ultrasonic Machining 35

2. T H E O R E T I C A L A N A L Y S I S O F M A T E R I A L R E M O V A L RATE
A theoretical analysis of the material removal rate in ultrasonic machining is made
considering direct impact of abrasive grains on the workpiece. The inhomogeneity in the size
of grains is taken into account by using the statistical distribution given by Rozenberg [7].
The grains are assumed to be spherical.

2.1 Nature of material damage


The mechanism of material removal in ultrasonic machining is shown in Fig. 1. The tool
moves down from its mean position and at some position S it touches the larger grains, which
are forced into the tool and workpiece. As the tool continues to move downwards the force
exerted on the grains increases, as a result of which the grains may fracture. Eventually, the
tool comes to the end of its stroke at E at a distance x from the surface of the work, as shown
in Fig. 1.
Assuming that the abrasive particles are incompressible, the total depth of penetration of
an abrasive particle of diameter d is (Fig. 1):
3, + 6 w = (d - x), (4)
where, x is the distance between the workpiece and the tool,
6, is the depth of indentation in the tool and
6w is the depth of indentation in the workpiece.
The tool motion is sinusoidal with amplitude a 0 such that
y = a0 sin 0. (5)
Even though the force between the tool and the abrasive particles is effective only during a
small portion of the cycle and may alter the free sinusoidal motion, it is assumed that the
motion of the tool remains sinusoidal under the loaded condition.
The contact force F c acts only over the portion of the cycle between 0 s and n/2 as shown in
Fig. 1 causing deformation in the tool and the workpiece. Thus the total depth of indentation
is given by
6~ + 6, = a o - Ys, (6)
where ys is the distance the tool has moved downwards from its mean position at the start of
impact.

MEAN POSITION m
z ANGLE
OF TOOL

F TOOL

1
0o WORK

(b) END OF IMPACT (o) START OF IMPACT

FIG. 1. Mechanism of material removal in ultrasonic machining.


36 G.S. KAINTH, AM1TAVNANDY and KULDEEP SINGH

Workpiec~l/

FIG. 2. Geometry at grit-work interface.

2.2 Rate of material removal


Figure 2 shows the geometry at the grit-work interface after indentation. The radius r w of
the contact indentation zone in the workpiece assuming small values of (6./d) is written as:
r w = [6~d] wE. (7)
The volume Vo fractured per grit is given by the hemispherical portion [8], shown shaded in
Fig. 2:
Vo = (8)
The depth of indentation 6 . is found by considering the hardness of the tool H t and
workpiece Hw. The brittle fracture hardness is defined by the average contact stress for
fracture,
Fc
H~ = n d - d ' (9)
TL - {d 2 -(2rw)2} l/z]

where Fc is the contact load on the grit.


Substituting the expression for rw from (7) and taking 6 . / d to be small, we get

H~ - or F, = 7tdH~6~. (10)
it 6wd
Similarly, by considering the indentation on tool, it can be shown that
F¢ = 7tdH, 6,. (11)
From equations (10) and (11) we get
6t/6., = Hw/H , = q, (12)
where q is the hardness ratio.
From equations (4) and (12) we can derive
6. = (d - x)/(1 + q). (13)
The material removal rate Vd(mmZ/min) due to all the abrasive grains of diameter d is
written in terms of the volume removed per grit, the number of abrasive particles of that
diameter in the working gap and the frequency f :
I'd = VoF(d)T ,
Material Removalin Ultrasonic Machining 37

where F(d) gives the number of particles of size d from equation (3).
The active grains taking part in material removal process are those having diameters
between x (the distance between the tool and the workpiece) and dm (the maximum diameter
of the abrasive grains). Hence, the total material removal rate v, considering all the effective
particles, is :

(6wd)3'2[1.o95N{1-(d-lf}3]fdd.
,=ff
Substituting for 6w from equation 03) we get:
2.29Nf ; " [ (_~)2~
v - (1 + q)3/2j [(d - x)d] 1 - - 1 dd. (14)

2.3 Averaoe number of particles in working gap


To find out the metal removal rate from equation (14), the average number of particles N
and the distance x needs to be determined. The number of particles N is found from the
weight of abrasive particles in the slurry.
The volume of the working gap containing the abrasive slurry is
V o = xDZxt/4, (15)
where D is the tool diameter and
x t is the distance between the workpiece and the tool at any instant of time t.
If C is the concentration of abrasive particles in the slurry by weight, (C kg of abrasive in
1 kg of liquid) Pa is the density of abrasive and p / i s the density of the liquid used for the
abrasive slurry, then the weight of the abrasive particles in the working gap at any instant of
time will be
C 7zD2xt
(16)

The number of particles in the working gap at any instant of time t is:
N, = W f f W , (17)
where W is the average weight of the abrasive particles.
The weight of an abrasive particle of diameter d is:
W a = (n/6)d3pa .

Hence the average weight of particles in the slurry is:


'"(n/6)d 3p a F(d)
where
W=
fdo
~-dd,

d o is the minimum diameter of the abrasive particles and


d,, is the maximum diameter of the abrasive particles.

(18)

Substituting the values of Wt and W in equation (17),

(19)
Nt- C TcD2(o'~Pa)[flfx(i
l -- C
-~ "
Pt + ~
a. ) 21 3]dd "

oda 1- -1
38 G.S. KAINTH, AMITAVNANDY and KULDEEP SINGH

The working gap xt is given by (Fig. 1):


x t = (x + ao) - ao sin 0. (20)

The average working gap during the period between the instant when the tool touches the
largest abrasive grain and the time when it stops at a distance x from the workpiece is
/~n/2
= 1. [(x + ao) - a o sin O]dO/(n/2 - Os),

= (x + ao) - [ao cos 0s/(n/2 - 0s)], (21)


where
0s = sin- l(ys/ao)
and
Y, = (a0 - d,. + x).
Hence, the average number of particles in the working gap is given by
C /tD 2 d :~ ~a d l (22)
N-(~ +~-~C) ( ~ - ) ( 0 . ~ - ~ p ~ ) ~ ~ . d3 i1 _ (_~_ 1) 2 .

2.4 Proximity of the tool from the workpiece


The distance x of the tool from the workpiece can be determined by equating the impulse
due to the static force F s over the entire cycle, to the impulse due to the variable contact force
Fc acting over the contact portion of the cycle [8].
The impulse due to the static force F s during the entire cycle is given as:
M s = 2nFJog. (23)
The contact force at a time t is

F, = rtdHw6~,F(d)dd, (24)
t

where 6~t is the indentation in the workpiece at a time t for a particle of size d and is given by
6w, = (d - x,)/(l + q). (25)
Hence, the contact force is given as:
" d(d - x,)
F. =
ffi
rrHw 1 + ~ F(d)dd. (26)

The impulse due to the contact force during the contact period from 0s to M2 is
Me = - - 1 f,q2~f, nH w
d(d_xO
-
1.095N • [ 1 - (~ - 1
)2~ dd.dO. (27)
to ao. , l+q d
Equating the static force impulse M s to the contact force impulse Me, we get

2nF,to = 10.95nN+Hw(1q)dto j~0,/2~xa,'(dZ-dxO'[ 1 - ( ~ - 1 ) 2 ~ ddd0. (28)

Substituting for xf from equation (20), the integral in equation (28) can be solved to give an
expression in terms of x, din, and d.
Equations (22) and (28) can then be solved for N and x to calculate the machining rate from
equation (14).
Material Removalin Ultrasonic Machining 39

3. THEORETICAL RESULTS AND DISCUSSION


The theoretical analysis is applied to study the effect of static load, amplitude of tool
vibration and the mean size of abrasive particles. The machining rate is found out by solving
equations (22) and (28) for N and x and substituting the values in equation (14). A computer
programme is developed to obtain theoretical results for machining glass with a mild steel
tool of 12.7 mm diameter at a frequency of 25.5 kc/sec using 400 mesh boron carbide
abrasive in water as a slurry. The following material properties were used: hardness of
glass -470 kg/mm 2, hardness of mild steel tool--150 kg/mm 2.

3.1 Effect of static load and amplitude of vibration


The machining rate is calculated for different values of static load and amplitude. The
results are shown in Figs. 3 and 4. The curves show that the machining rate increases with an
increase in the static load and amplitude of tool vibration. Thus the analysis predicts an
infinite increase in machining rate with static load. However, in actual practice the machining
rate first increases with static load and after reaching an optimum value, it falls with a further
increase in static load [-7].
The effect of amplitude on machining rate (Fig. 4) agrees qualitatively with the
experimental results of Nepiras as reported by Rozenberg [7]. However, an approximate
comparison with experimental values quoted by Rozenberg [7] shows that the theoretical
machining rate is an order higher than the practical value of about 32 mm3/min for soda
glass.

3.2 Effect of size of abrasive grains


The machining rate is calculated for various values of mean size of abrasive grains for an
amplitude of 0.0625 mm and a static load of 0.93 kg. The results plotted in Fig. 5 predict a
linear increase in machining rate with an increase in abrasive size whereas the experimental
work of Fukumote quoted by Rozenberg [7] shows an optimum value of grain size which
depends upon the amplitude of the tool oscillation.

4. DISCUSSION AND CONCLUSIONS


The present theory incorporates the non-uniformity of abrasive grains and enables

.c

2
A
c
~2
E
v 2

ill:
I
t.9
z
_z
-r
c.)

2 4 6 8 XO.I
STATIC LOAD (Kg)
FIG. 3. Variationof machiningrate with static load.
40 G. S. KAINTH, A M I T A V N A N D Y and K U L D E E P SINGH

t~

to
E
E
taJ 2
cc

Z
Z
"I"
<J

I E ,3 ~ :~ b f xu.OI

AMPLITUDE (ram)

FIG. 4. Variation of machining rate with amplitude.

I I I I
ol W o r k - piece:- gloss I
x
°- Tool - diameter :- 12.7 m m ' I ,!
71 Frequency :- 25.5 Kc/sec| / ,
- - C o n c e n t r a t i o n :- O. 1 6 8 J ~ -J ] --
Abrosive:-boron c o r b i d e / / /~ /
A m p l i t u d e : - 0.0625 mm ~ J
6 - - Stotic Iood :- 0.93 Kg - ~ - ~ ~ .--
E ! , ,

z
"1-

, 6 8 Io
i 12 14 xo.o~
MEAN SIZE OF ABRASIVE PARTICLES(ram)
I I I I I I I
500 320 240 180 15o 120 I00
CORRESPONDING MESH SIZE

FIG. 5. Variation of machining rate with mean size of abrasive particles (mesh size).

quantitative evaluation of machining rate using material properties of tool and workpiece
whereas this is not possible with the earlier theories.
A limited comparison with the experimental results in the existing literature shows that the
proposed theory predicts machining rates which are an order higher than the practical
values. Even though the theory does predict an increase in machining rate with increase in
static load, tool amplitude and size of abrasive grain, the theory does not predict a drop in
machining rate after optimum values of static load and abrasive size are reached.
Material Removal in Ultrasonic Machining 41

The disagreement may be due to the following reasons:


1. It has been assumed that the amplitude of vibration remains constant with an increase
in static load. Preliminary experiments on a 200 W Cavitron Drilling Machine shows that as
the static load is increased, the tool stops vibrating at higher loads. Even though it is possible
to estimate the amplitude of tool vibration before the machining is done, it has not been
possible to measure the amplitude during the machining operation. The problem of
amplitude measurement is complicated by the fact that the vibration needs to be measured at
the tip of the tool at a very high frequency of 25 kc/sec. Hence it is expected that an increase in
the resistive load at the tip of the tool would decrease the amplitude of tool vibration. It can
be seen from Figs. 3 and 4 that a decrease in amplitude with an increase of static load would
cause a fall in the machining rate as shown by the dotted curve in Fig. 3. The nature of the
curve would depend upon the manner of amplitude variation with static load. As such, in
order to explain the nature of variation of machining rate with static load, it is essential to
study the behaviour of a particular machine on which experiments are conducted.
2. A preliminary study made by continuous impacting of abrasive particles by a vibrating
tool, under the conditions encountered in ultrasonic machining, has shown that the abrasive
grains fracture during machining, thereby decreasing the effective number of particles under
the tool. A deeper study of the phenomenon of fracture of abrasive grains is necessary.
Further work is being done to introduce probabilistic fracture of abrasive grains and
workpiece into the analysis.
3. It is assumed that each abrasive grain removes the hemispherical portion from the
workpiece. The assumption needs to be modified as after initial machining of the flat surface,
abrasive grains encounter a pitted surface created by a preceding impact.
4. Abrasive grains have been assumed to be spherical, whereas actually the grains are of an
irregular shape.
The present analysis takes into account the size distribution of abrasive grains and makes
it possible to evaluate machining rate qualitatively. However, in order to improve agreement
with known experimental results, it is necessary to carry out further investigations on the
lines suggested above.

Acknowledoements - The authors are thankful to Mr. Uma Raman Pandey for typing the manuscript of this paper.

REFERENCES
[1] G. E. MILLER,J. Appl. Phys. 28, 149 (1957).
[2] M. C. SHAW, Microtechnic 10, 165 (1956).
[3] L. D. ROZENaERGand V. F. KAZENTSEV,Docklady Aklad, Nauk SSSR 124, 79 (1959).
[4] V. I. DIKtJSmN and V. N. BARr,E, Stanki i Instrument 5, l0 (1958).
[5] V. N. BARKE,Dissertation, Moscow Machine Tool Institute (1958).
[6] V. F. KAZANTSEV,Machines and Tooling 34, 14 (1963).
[7] L D. ROZEr~BERG,Ultrasonic Cutting, Consultants Bureau, New York, 1964.
[8] G. NISHIMURA,Y. JIMBOand S. SHIMAKAWA,J. Fac. Engng Tokyo Univ. 24, 65 (1955)i

You might also like