Paper No.

6571

FOUNDA'I'JONVIBRATIONS
bY
Tso Kung Hsieh, Ph.D., A.M.I.C.E.
Chief Designer, Planning Department, George Wimpey & Co. Ltd.
For written discussion

SYNOPSIS
The Paper presents a method of calculating the amplitudes of vibration of a
foundation restingdirectly on soil.Sucha foundationhas six degreesof
freedom.Usuallythese are coupledtogether. Equationsare presented for
analysing these coupled modes of vibration in quantitative detail.

INTRODUCIION
INdesigning a foundation to carry a static load, it is now generally appreciated
that it is not only necessaryto ensure that the ground will carry the loading with-
out the risk of shear failure, but it is also essential to estimate the amount of
displacement expected under the loading. This is to ensure that the structure
carried by the foundation cansafely be subjected to the displacement. Such an
estimation of the displacement is just as essential, if not more so, for a foundation
carrying a vibratory machine. The satisfactory operation andefficient perform-
ance of the machinery depend almost entirelyupon the magnitude of oscillation
of the foundation. Further, the amplitude of oscillation influences the loading
intensity on the foundation, and therefore this must be estimated to ensure
safety against shear failure of the ground.
2. A rigidblock foundation restingdirectly in contact with the ground
possesses six degrees of freedom. It may oscillate in its translational modes;
vertically, laterally or longitudinally, or it may oscillate in its rotational modes;
rocking, pitching or yawing. Usually, these oscillations are coupled together.
To make a complete analysis,it is necessary to establish six equations of motion
from which a completesolution may be obtained.
3. A review of the current literature on this subject revealed the need of a
complete theory for dealing with such problems. It is hoped that the present
study may partly fulfil1 this need and that in particular it may stimulate further
researches in that direction.

REVIEW OF PREVIOUS WORK

4. Empirical rulesof thumb have been and arestill being used for the design
of machine foundations. The emphasis has usually been on the provision of a
sizable weightin an attempt to reduce
Delivered the amount of oscillation.
by ICEVirtualLibrary.com to: The weight of
21 1
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
212 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS
the foundation is related to the weight of machine, the weight of moving parts,
the speed of machine, the horse-power of machine, either individually or a
combination of two or more. In ensuring the safety against shearfailure of the
ground, an allowance ismade for a dynamicload (usually as a percentageof the
static load) in addition to thestatic load acting on the foundation.
5. Most of the foundations designed in accordance with these rules havebe-
haved satisfactorily, and indeed under certain conditions, there is a theoretical
basis for some of the rules to justify the procedure. But where these conditions
are not fulfilled, failures must be expected and some are reportedin the literature.
6. Most trouble in machine foundations canbe traced to resonance when one
of the natural frequencies ofthe system coincides withthe operating frequency.
Therefore the determination of the natural frequencies of the foundation-soil
system is of major importance to the design of foundations. Thus by far the
majority of the research work on this subject has been centred around the
studies of natural frequencies ofthe foundation-soil system and the greater part
on the theoretical and experimental studiesof the frequency in the vertical direc-
tion. A considerablevolumeof literature is available.Briefly the theory is
based upon the concept that the foundation andsoil forms a simple mass-spring-
dashpot system. The mass of the system is consideredto be the sum of the mass
of the machine, with the foundation and a certain portion of the soil which is
considered to be oscillating with the foundation. Rules for estimating the portion
of the soil have been suggested by Lorenzl, Crockett and Hammond2, Pauw3,
and Balakrishna Rao and Nagaraj4. The spring constant is usually taken as the
modulus of subgrade reaction.Methodshavebeenproposed by Pauw3 for
taking into account theeffect of sizeand shapeof the contact area andthe effect
of increasing soil modulus with depth. Non-linear behaviour of the soil has also
been considereds. Damping is generally considered to be caused mainly by the
viscous resistanceof the soil to rapid deformation. Ehlers6 suggested that addi-
tional damping may come from the loss of energy by waves radiating into the
soil. However no theory is known to be availablefor the evaluation of damping
effect.
7. It is known from the results of mechanical vibration analysis of a simple
mass-spring-dashpot system, that the magnitude of damping does not seriously
affect the natural frequency of the system. Consequently, all efforts have been
centred around the evaluation of the mass and spring factors. With damping
ignored, it is not possible to estimate the amplitude of vibration near the re-
sonance frequencies.
8. The simple mass-spring model is useful in bringing out the salient factors
affecting the natural frequencies of the foundation. It is, nevertheless an over-
simplification if it is to be usedfor accurate quantitative studies.
9. A foundation subjected to periodic forces from the machines sends out
stress waves which are propagated throughthe ground. These waves are waves
of expansionand distortion propagated throughthe interior of the soil mass, and
Rayleigh waves propagated nearthe surface of the ground. They set the whole
soil mass into motion and in theory this region of motion extends to infinity. To
analyse the vibration of the foundation, therefore, the ground andthe foundation
must be treated as one system. This has been done byReissner7, Quidan*,
Sungg, and Amold, Bycroft, and Warburton'o. All investigatorsconsidered
1-10 The references are listed on pp. 224-225.
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
TSO KUNG HSIEH ON FOUNDATlON VIBRATIONS 213
circular foundations on semi-inlinite
a elastic body. Reissner considered uniform
loading over the contact area of the foundation andground. Quinlan andSung
considered rigid plate and parabolic type of contact loading. Richart9 in the
discussion to Sung’s paper suggested the use of an “effective radius” to cover the
various typesof loading. h o l d , Bycroft, and Warburtonanalysed translational
vibration in the horizontal direction, and rotational vibration about the vertical
and horizontal axis. They have also considered vibration of foundations on a
stratified subgrade.
10. These theoriesare exact as long as the ground canbe consideredas a senli-
infinite, isotropic, linearly elastic body without viscous damping. Reasonably
good agreements have been obtained with test results. The theories dealt with
one of the six modes of vibration at a time and therefore are limited to cases
where the six modes are de-coupled. Usually, some of the modes are coupled
and therefore these theories cannot be used in their present form. Their exten-
sion to cover allthe six coupled modesof vibration would be rather complicated.
The following presentsa different approach to the solution of the problem.
SOME BASIC EQUATIONS
11. The wave equation governing the propagation of waves in a prismatical
bar is:11*12

where U is the longitudinal displacement alongthe axis X

t is the time

c= 8 is the wave velocity

E is Young’s modulus
p is the mass per unit volume.
12. The general solution of this equation is :
U =f(~-ct)+F(~+ct)
Wheref(x-ct) represents waves travelling in the direction of the x-axis while
F(x+ ct) represents waves travelling in the reverse direction.
13. Consider waves travelling in the direction of the x-axis only, we have:
U = f(x - ct)
au au
z=z
au 8U
-
at =
where E =X - ct, and therefore
au
-=
ax
-__
1 au
cat.
&
Since the stress U acting acrossthe bar is E G ,we obtain the well known relation-
ship
Eau
(T = ---
cat . . . . . . . . (l)
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
214 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS
that is, the stress is linearly proportional to the velocity of the particle. The
waves considered in this case are propagated through the bar without change of
form oramplitude.
14. Next consider the wave equation governing the propagation of waves in
a conical barl2:
- c, a 2 w
at2 ar2
The general solution is:
UT - + +
= f ( r ct) F(r cl)
15. Kolskylzobserved in his book that no generalrelationshipseems to
exist between particle velocity and stress in this case. It will be shown however
that stress is related to theparticle velocity and the particle displacement by an
equation obtainedas follows.
16. Confining ourselves to divergent waves travelling away from the apex of
the cone only, we have:
ur = f (r - ct) = f([)
I1 =
(5)
f-
r

Therefore :

au
and the stress U =E 5 is :

that is, there isalinearrelationshipbetweenstress,particlevelocity and

displacement.
17. It is of interest to note that for a cone the stress is a linear function of
particle velocity and displacement whilefor a bar, it is only a linear function of
the particle velocity. The additional function appears as a consequence of the
attenuation of the waves as they travel awayfrom the apex of the cone.
18. For waves radiated from a source at the surface of a semi-infinite homo-
geneous isotropic elastic body created from a weightless rigid circular disk of
radius r0 under a vertical periodic loading P=Zeiat it has been shown7t 8, 9. 10
that the displacement d of the disk is

where W is the circular frequency of the loading

t is the time variable
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS 215
e is the base of Napierian logarithms
i is the imaginary number d-1
ro is the radius of the circular disk
fl and f 2 are functions which depend uponthe Poisson's ratio v of the medium
and a dimensionlessquantity a = wro
G is the shear modulus of the medium
p is the mass density of the medium.
Now
-= -
Lidzw
[ifl -f 2 ] eiot
dt Gro
Therefore

or
19. Thus we get the same linear relationship betweenload, disk velocity and
displacement as that obtained for wave propagation through a conical bar.
The only difference isthat in this instance, the coefficients are not constant, but
vary with the frequency of the loading.
20. The wave equations used for the prismatical and conical bars apply only
for compression waves of length large in comparison with the lateral dimension
of the bars. Under these conditions, the lateral motion of the bar is small and
has been ignored. For a semi-infinite body, its lateral dimension is of the same
order as its longitudinal dimension and therefore its lateral motion is just as
prominentas its longitudinalmotion.Rayleigh waves and shear waves are
present in addition to the compression waves. For a circular vibrator operating
normal to the surface of a semi-infinite medium, a largerpart of its power is in
fact radiated as Rayleigh waves.13
21. If the lateral motion of the bars is taken into consideration, it is found12
that waves are dispersed and consequently the coefficients in equations (1) and
(2) are also expected to vary with the frequency of the loading. Geometrically,
a semi-infinite space is a cone with a large solid angle. The difference between
this and the conical bar can therefore be only in degree and it is thus not un-
expected that the same basic functional relationships exist betweenthe applied
force, the disk velocity and displacement.
22. Arnold, Bycroft, and Warburton'o have shown that the horizontal dis-
placement of a weightless circular disk excited by a horizontal forceZeiot is also
given bythe equation (3) except, of course, the values of the functionsfi and f 2 are
different. Similarly, theangular rotationofthe disk excitedbya couple C= Meiot
about a vertical or horizontal axis is given by

Following the same procedure, we arrive at a similar relationship between the

couple, the angular velocity and rotation of the disk:

Delivered by ICEVirtualLibrary.com to:

IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
216 TSO KUNG HSIEH ON FOUNDATION
VIBRATIONS

23. Equations (4) and (6) are basic equations which we shall find useful in
analysing the vibration of a foundation in all its six degrees of freedom.
24. The foregoing theory is based on the assumption that the semi-infinite
medium is homogeneous, isotropic and elastic. The ground upon which a
machine foundation restsis not such an idealmedium.However,provided
there is no rigid stratum very close to the surface (less than three times the
diameter of the base149 9 , it should provide results of sufficient accuracy for
engineering purposes.

THE FUNCTIONS F1 AND F2

or simplicity, equations (4)and (6) will be re-written:
dA
P = - % ' G r , z F --Gro F+l
dt
. . . . . (4a)

''
c = - d / G p r o 4 ~-dt- G ro3 F1 4 . . . . . (6a)
where

f2
F --
- f12+h2
25. The values of fl and f2 for the various cases can be found in papers by
Amold, Bycroft, and WarburtonlO and Bycroftl6t l7. They are in general very
of the Poisson's ratio of the medium v and a dimensionless
complex functions -
quantity u=-ro,/$. The parameters F1 and F2 may be calculated from the
graphs prepared in these papers. For engineering purposes, it has been found
that sufficient accuracy may be obtained from the following. (Values of Fl and
F2for values of a larger than 1-5or 2.0 are generally not required.)
A. Vertical translation (0 < a < 1-5)
v =0 Fl = 4.0-03a2
F2 = 3.3-tO.4 a
y 21. Fl = 5.3- 1.0 a2
F2 = 4.4+0.8 U
V = 4 F1 = 8.0-2.0 U*
F 2 = 6.9
B. Horizontal translation (0 < a < 2-0)
V = 0 F1 = 4.5-0.2~'
F2 = 2-4+0.3 a
V = F1 = 4 . 8 - 0 . 2 ~ ~
F2 = 2.5 0.3 a+
V = f F1 = 5.3-0.1 a2
F2 2*8+0*4U
C. Rotation about vertical axis (0 < a < 2.0)
V = 0 to 3 F1 = 5.1-0.3 U*
F2 = 0.5 U
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
TSO KUNG HSlEH ON FOUNDATION VIBRATIONS 217
D. Rotation about horizontal axis (0 < a < 1.5)
v =0 Fl = 2.5-0*4aZ
F2 = 0.4 a
26. The above are obtained for contact areas of circular shape and for con-
tact-pressure distributions correspondingto rigid bases subjectedto static loads.
In general, foundationsmay be ofother shapes and the pressure distribution may
be different. Under these conditions, the following modifications may be used
to give sufficiently approximate results.
27. For translational motion, use an equivalent radius which gives the area
of the circle equalto thatof the contact area of the foundation with the ground.
28. For rotational motion, use an equivalent radius which gives the moment
of inertia of the circle equalto that of the contact area of the foundation with the
ground.
29. For pressure distribution other than that corresponding to rigid bases
under static loading, the equivalent radius is to be multiplied by a factor m to
obtain the modified radius ro for the purpose of calculation.
Uniform distribution . . . . . . . . m = 0.78
Parabolic ,, . . . . . . . . m=059
30. The parameter a depends upon the circular frequency of the periodic
forces W, the radius ro of the base and the physical properties G and p of the
medium.
a = wro 46I-

31. This may be simplifiedas follows. The velocity of the shear wave=

27r
the period of the wave =W , and the length of the shear wave =L = - U:'
-.
Therefore, a = 2 a -r0
L
Thus a is a linear function of the ratio between the size of the base ro and the
length of the shear wave of the medium.
32. For the sake of clarity, equations (4a)and (6a) will be simplified and sub-
divided into the different cases as follows:
Vertical translation:
dA
p = - R, 2- K,A,
dt
Horizontal translation:

Rotation aboutvertical axis

CV= - H v d4
-dt
'-Sv&
Rotation abouthorizontal axis

Delivered by ICEVirtualLibrary.com to:

IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
21 8 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS

THE EQUATIONS OF MOTION

33. Let the co-ordinate axis pass through the combined centre of gravity
“G” of the foundation and the machines (Fig. 1). Let the co-ordinates of the
centroid of the contact surface of the foundation andthe ground be B,, By, and
B,.
34. Let the translational displacement of the centre of gravity G be denoted
by Ug, rg,rg and the rotational displacement be denoted by e,,e,, and 0,.
ogrepresents the displacement of G in the x-direction (positive in the direction
of increasing X etc.) and 0, is the rotation of the foundation about
x-axis (positive
clockwise when facingin direction of increasingX , etc.)

B ’

t
Z Z
FIG.1 .-FOUNDATION
C~~RDINATES

35. The translational displacement of B is therefore; (assuming that all

displacements are small) :
ob ug+B,ez-B~e,
E

+
pb = 7 B,e,- B&
p b dg+BxOy-B,,8, l. . . . . .(7)

36. The rotation of the contact surface about the co-ordinate axis is the same
as that of the foundation i.e. Ox, Oy and 8,.
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS 219
37. The ground reaction therefore consists of three forces:

acting at B in the direction of the co-ordinate axisand three couples:

dez
c, = - H v --s,e,
dt
about the three co-ordinate axes.
38. Let the mass of the foundation and machine be M and its moment of
inertia about the co-ordinate axis be Z,, Z,, and Zz.
39. Let the unbalanced forces of the machine be resolved into three forces
acting at G : (Q,, Qy, Q3 in the direction of the co-ordinate axis and three
couples (Tx,T,, TA about the three axes.
40. The six equations of motion are therefore as follows:

d2rg
M- = Qr+Py
dtz
M d2=@g
- df.2 Q,+’,
dZB,
. . . . (10)
Z-
dtz
= Tx+C,+B,Py-B,P,

d’e,
Z y z = Ty+C,+BxP,-B,Px

J,
d20,
-
dt2
= T, -+ C, f ByPx- BxPy
J
41. Expressing Px,pY,P,. C,, C,,and C, in terms of Og,Vg,Pg,e,, By, and ez
by using equations (7), (S), and (g), we obtain six second-order ordinary simul-
taneous differential equations from which the displacements og,Vg,FVg, ex, ey,
and B, can be obtained.
42. Taking Q,, Q,, Q,, T,, T,, and T, as zero, we obtain fromthese equations
the six natural frequencies of the machine-foundation-soil system. It is noted
that in its general form,all the six degrees of freedom are coupled.
43. Usually the machine-foundationsystemhastwoplanes ofsymmetry
(X-z plane and y-z plane) so that B, and By are both zero. Under these condi-
tions, equation (10) can be considerably simplified.

Delivered by ICEVirtualLibrary.com to:

IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
220 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS

46. All the six degrees of freedom are therefore independent of one another.
Under thesespecialconditions,eachmode of oscillationmay be treated
separately. Theoretical solutions of the problem have so far been confined to
these special cases. It can be easily verified that the above equations give the
same results as thoseobtained by others101 161 17.
47. For example, consider oscillation in the vertical direction only. This is
governed by equation (13). For a periodic loading of QL=Zeiwtwe have:

The amplitude of oscillation A can be easily shownto be:

L
A =
[(K,-w~M)~+(wR,)~]"~~
From equations 4a and 4b, we have:
R, = l / F r 0 2 F 2 = 1 / q r o 2-
f2
f12+fZ2
. . . . (19)

K , = GroF1 = Gro -
-f1
. . . . (20)
fl2+fZ2
Substituting these into equation (18), and letting
b = -M
pro3
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
TSO KUNG HSIEH ON FOUNDATION VIBRATIONS 22 1

This is exactlythe same as that obtained by others71 91 10.

*l

48. It may be noted that equations (13), (l@, (lla), (l&), (14a), and (15a)
are all similar in form. Further, they are similar to the equation used by many
previous workers based upon the mass, spring and dashpot analogy. However
there are fundamentaldifferences.
49. Firstly, whereas the analogy used previously was entirely empirical, the
present equations are obtained from theoretical considerations. The constants
R,,K,, etc. are now clearly defined and are given by equations (4a), (4b) and
others. It should be noted that these constants are functions of the properties
of the soil (G and p), the dimension ofthe foundation (ro) and the frequency of the
immessed vibration(.W ,) .
50.Secondly,termslike R, werepreviouslyconsidered to be a factor
due to viscousdamping. As no dampingwasassumed in the derivation
of the present equations, these factors appear as a resultof loss of energy through
radiation of waves into the soil. These factors are now quantitatively defined,
and consequently, it becomes possible to estimate the amplitude of vibration of
the foundation-soil systemat any frequency including resonance.
51. The equations of motion obtained in thissection are comparatively
simple and may be solved analyticallyor very easilyand conveniently by the use
of analogue computers. They provide the means of analysing the oscillation of
a foundation in detail which is essentialfor the rational design of the foundation.

APPLICATION TO VERTICAL VIBRATION OF FOUNDATION

52. In many cases, vibration in the vertical direction is de-coupled from the
other modes of vibration. Under these conditions, vertical vibrations may be
considered independently. Most previous researches have been largely concerned
with only this aspectof the problem. The majority were confinedto the natural
frequency of the foundation.
53. The vertical oscillationof a foundation under a periodic forceis governed
by equation (17),

The amplitude of oscillation A, is given byequation (1 8) :

z
A = [(K,--2M)2+ (WRv)2]04
54. The resonance frequency W O and the amplitude at resonance A0 may be
calculated exactly from this equation. An approximate but simple solution is
given by:
K
Wo2 = 2
M
. .
. . . . .
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
222 TSO KUNG HSIEH ON FOUNDATION VLBRATIONS

for the resonance frequency and

. . (23)
for the amplitudeat resonance. Theseare good approximationsfor large valuesof

55. Equation (22) gives (consider the case of v=*)

w02=-=-=
K , GroFl Gro(8- 2 a 0 2 )
M M M

-
M
After rearranging, we have:
8Gro
wo2 M+2pro3 . . . . . . . (24)
56. Now, under static vertical loading, the spring constant of a rigid circular
base restingon the surface of semi-infinite elastic mediumis
. . (25)
For V = & , the value of K,, is equal to 8Gro. Substituting this into equation (24),
we have :
w02 = - K,,
-
M + 2pro3
. . . . . (26a) .
Similarly, it can be shown that, for v = + ,
K,,
M + pr03 . . . . . . . (26b)
wo2 = ~

and for v =0,

WO2
K,,
= M+0.Spro3 * . ' WC)
These equations show that the resonancefrequency of the foundation-soil
system may be calculated on the basis of a simplemass spring system. The spring
constant is given byequation (25). The mass isthe sum of that of the foundation
and machine with a certain amount of the soil which is a function of Poisson's
ratio of the soil, the mass densityof the soil and the size of foundation base.
57. The above therefore providesthe theoretical basis for a conception which
has been much used for qualitative discussions. The portion of the soil to be
taken into calculation is now quantitatively defined. The spring constant to be
used is also clarified. The analysis however reveals the approximate nature of
such a conception.
M
Substituting b = -into equation(24), we have:
pro3

Delivered by ICEVirtualLibrary.com to:

IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
 TSO KUNO HSIEH ON FOUNDATION VIBRATIONS 223
or

where a0 =woro
JPE is the value of a at resonance frequency w0.

58. Using the above values and noting that R,= l/qro2F2=6-9d q ro2
for the case of v =f,we obtain from equation (23):

59. If the periodicforce Z varieswith the frequencysuch that Z = E W ~

(unbalanced forcesof machines are mostly of this nature), then, at resonance, we
have, from equations (29) and (27):
0.41 E
.
A 0 - . , , / g - ~ , , ~ ~ .3 . . e . . (30a)
or

60. Let W=Mg be the weightof the foundation and machine, we have,
from equation(30b) :
0.41bg E
W = --
d b + 2 ~ .~ . . . . . . (31a)

The above equations apply to the case of v = 3. For v=+, we have:

and forv = 0 we have :

0.61 bg E
= 0.24 + db+0.5 . . (31c)

61. These equationsgive directlythe maximum amplitudeof oscillation of the

foundation. Alternatively, they may be used for calculating the minimum weight
of foundation required to limit the maximum amplitude of oscillation at re-
sonance to a specified value.
62. It is to be noted in this connexion that althoughthe quantity b represents
M
the ratio -, it is a factor which represents more closelythe ratio of the height
pro3
to the width of the foundation, andmay perhaps be better referredto as a shape
factor. For a cylindrical foundation built of a material with the same mass
density as the soil, and assuming that the mass of the machine compared with
that of the foundation is small enough to be ignored, we have:

and therefore b represents accurately the ratio of 2n times the height of the
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
224 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS
foundation to its base width of 2ro. For a foundation different from a cylin-
drical shape and of mass density differentfrom the soil, it may be consideredas
h, such that p7rro2h,=M. In the general case,
an equivalent cylinder with a height
we have, therefore:
b = -T h e
r0
. . . . . . . . (33)
and b is thus independent of the actual mass and size of the foundation.
63. Equations (31) seem to provide a theoretical basis for the generally used
rule of thumb that the weight of a foundation supportinga vibratory machine
should be relatedto the weight of machine, or the weight ofthe rotating part of the
machine, or the horse power of the machine, etc. This is justifiedif the quantity
E is directly related to these factors. The result is of course only an approxima-
tion. It is however, a very simple relationship. Further, it is independent of the
shear modulus of the soil. It can therefore provide a very useful guide for
preliminary design.
64. Other generalizationsmay also beobtained.Howeverthese and their
extension to other modes of vibration may usefully be leftfor another paper.

CONCLUSION
65. A rigidblock foundation restingdirectly in contact with the ground
possesses six degrees of freedom. To ensure the satisfactory functioning of the
foundation, itis necessary to analyse quantitativelythe amplitudes of oscillation
in all six modes. Many foundation troubles are the results of vibrations in-
volving horizontal translation and rocking motion at the same time. These two
modes of oscillation are usuallycoupledtogether. A reviewof the current
literature revealed the need of a complete theoryfor solving such problems.
66. From a consideration of the propagation of divergent waves through a
homogeneous isotropic elasticmedium, a basicrelationshipisestablished
between the externally applied periodic load on a rigid circular disk and its
corresponding displacement and velocity. This forms the basis for formulating
the six equations of motion of the foundation fromwhich a complete analysis of
the various modesof oscillation may be obtained.
67. Fuither studies are necessary to investigate the influence of non-linear
and dissipative properties of the soil. Much experimental work is also required
to determine the shearmodulus and Poisson’s ratio of the soil and the distribu-
tion of the contact pressure between the foundation andthe ground. However,
sufficient is now known such that if conservative valuesof the shear modulus,G,
the Poisson’s ratio v, and the equivalent radius of the foundation ro are used,
calculationsbased upon these equationsshould giveresultsacceptable for
engineering purposes.

REFERENCES
1. H. Lorenz, “New results of dynamic investigations of foundation soils”. 2. Ver.
deutscher Ing., vol. 78. No. 12, March 24, 1934, pp. 379-385.
Delivered by ICEVirtualLibrary.com to:
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
 Tso KUNO HSIEH ON FOUNDATION VIBRATIONS 225
2. J. H. A. Crockett and R. E. R. Hammond, “Natural oscillation of ground and
industrialfoundations”.Proc. 2 Int.Conf. Soil Mech.Found.Eng, vol. 3,
1948, pp. 88-93.
3. A.Pauw, “A dynamicanalogyforfoundation-soil systems”. Symposium on
Dynamic Testing of Soils. &er. Soc. Test. Mat. S&al Technical @ublication
No. 156, 1953, pp. 90-1 12.
4. H.A.BalakrishnaRaoand C . N. Nagaraj, “A new method for predicting the
naturalfrequency of foundation-soilsystem”.StructuralEngineer,October,
1960, pp. 310-316.
5. H. Lorem, “Elasticityanddampeningeffectsofoscillatingbodieson soil”;
Symposium on Dynamic Testing of Soils, Amer. Soc. Test. Mat. Special Tech-
nical Publication No. 156,1953, pp. 113-123.
6. G. EhIers, “The soil as spring in oscillating systems”. Beton und Eisen, vol. 41,
1942, pp. 197-203.
7. E. Reissner, “Stationary, axialiy symmetric oscillations on a homogeneous elastic
half space, caused by a vibrating body”. In-ieur Archiv, vol. 7,1936, pp. 381-
396.
8. P. M. Quinlan, “The elastic theory of soil dynamics”. Symposium on Dynamic
Testing of Soils, Amer. Soc. Test Mat. Special Technical Publication, No. 156,
1953, PP. 3-34.
9. T. Y.Sung, “Vibrations in semi-infinite solids due to periodic surface loading”.
Symposium on Dynamic Testing of Soils, Amer. Soc. Test. Mat. Special Tech-
nical Publication No. 156, 1953, pp. 35-68 incl. disc.
10. R. N. Amold, G. N. Bycroft, andG. B. Warburton, “Forced vibrations of a body on
aninfiniteelasticsolid”. J. AppliedMechanics,vol. 22, 1955, pp. 391400;
bound in Trans. Amer. Soc. Mech. En-, vol. 77, Sept. 1955.
11. S. Timoshenko,and J. N. Goodier. “Theory ofelasticity”. McGraw-Hill, 1951.
12. H. Kolsky, “Stress Waves in Solids”. Clamdon Press, Oxford, 1953.
13. G. F. Miller and H. Pursey, “On the partitionof energy between elastic waves in a
semi-infinite solid”. Proc. Roy. SOC., A, vol. 233, 1955, pp. 55-69.
14. G. B. Warburton, “Forced vibration of a body on a stratum of soil’’. 9th Int.
Cong. Applied Mechanics, vol. 7, Brussels, 1956, pp. 137-142.
15. G. B. Warburton, “Forced vibration of a body on an elastic stratum”. J. Applied
Mechanics, vol. 24,1957, pp. 55-58; bound in Trans. Amer. Soc., Mech. Engrs,
vol. 79, 1957.
16. G. N. Bycroft,“Forcedvibrationsof a rigidcircularplateon a semi-infinite
elasticspaceand on an elasticstratum”. Phil. Trans. Roy.SOC., A No. 948, vol.
248,1956, pp. 327-368,
17. G. N. Bycroft, “Machine foundation vibration”. Proc. Instn Mech. Engrs, vol.173,
No. 18,1959, pp. 469473.

The Paper which was receivedon 2 May, 1961, is accompanied by one drawing
from which the Figure Delivered
in the textbyhas been prepared.
ICEVirtualLibrary.com to:
16
IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05
226 TSO KUNG HSIEH ON FOUNDATION VIBRATIONS

Written discussion on this Paper should be forwarded to reach the Institution

before 15 August,1962and will be published in or after December,1962.
Contributions shouldnot exceed 1,200 words.-SEC.

Delivered by ICEVirtualLibrary.com to:

IP: 147.188.128.74
On: Sat, 13 Apr 2013 09:29:05