VIBRATION ISOLATION

e by
F. Kunz
(

Mr Kunz served l:m apprenticeship withE Rothlisberger ofWynigenfrom 1946 to 1950. He then worked
for Alpina Reederei, Basel, on tugs and deep sea cargo vessels in a variety of capacities, gaining a 2nd .
class BOT certificate in 1956. From 1957 to 1960 he was a Second Engineer with Lamport & Holt and he
obtained a 1st class Swiss Certificate in 1960. He then worked in turbocharger and fuel injection
development with Simms Motor Units from 1960 to 1966. During this time he studied for an HNC in
Mechanical Engineering and was also awarded a College Diploma in Mechanical Engineering in 1966. In
1966 Mr Kunz joined Lloyd's Register as an Engineering Surveyor, working in the Engineering Research
Department unti/1970 and subsequently in the Technical Investigation Department. ln 1985 he was
awarded an Open University BA Degree in Mathematical Subjects. He is currently Senior Principal
Surveyor in charge of the department.

0 SYNOPSIS are often of greater interest than the vibration of the machine
itself and it then becomes relevant to study them. Forces trans-
This paper presents some techniques employed in the theoret- mitted by resiliently mounted machines to the foundations and
ical and the experimental analysis of resiliently mounted the surrounding areas are rarely destructive in the sense of
machines. The terminology and body of knowledge surround- inducing structural failure in or around the seating if the ampli-
ing the subject are addressed drawing on practical cases to tudes are acceptable to the machine, but they are frequently
illustrate particular points. of concern with respect to human comfort,
The scope is restricted mainly to the interaction of self Habitability problems on cruise ships or complaints from
generated vibration of machines with their surroundings and close neighbours of power stations or industrial plant are typi-
how undesirable effects are attenuated or detuned by the cal examples.
insertion of resilient elements of various configurations. The control of vibration to protect surrounding areas from
Information on current practice and guidance on the inter- machinery excitation by inserting some form of resilient mount
pretation and evaluation of vibration readings produced by forms the topic of most of the cases quoted in this paper.
modern signal analysis equipment are presented. The signifi- Reference is made to passive isolation systems and a glossary
cance of specialised terms, and acceptability·are discussed. A is included to define some of the specialised terms. Guidance
glossary of terms and bibliography are included. is given on the controversial but all important subject of accept-
ability.
Many machinery vibration isolation systems are highly
specialised devices which, at present, fall outside normal clas-
1. INTRODUCTION sification rules, and are based on design techniques and
calculations of potentially great complexity with their own
The purpose of this paper is to bring to the attention of the terminology, literature and standards of acceptability.
Technical Association: members an overview of modern vibra-
tion isolation technology as applied to machinery. Analysis and
experimental techniques are complex and the extensive termi-
nology .surrounding the topic can confound the non specialist. 2. BASIC THEORETICAL CONCEPTS
In principle vibration isolation falls into two distinct main
groups, one which is intended to limit and control the effects The simplest possible analytical representation of a resiliently
of vibrations on the surrounding areas and the other which mounted machine is given in Ref. 2 which considers it to be a
seeks to prevent excitation in the form of steady state vibration single rigid mass, supported through a massless mounting
or shock impulses from reaching a passive sensitive device. The element of finite stiffness to an infinitely rigid and massive
theory of the two distinct functions turns out to have identical sur{ace.. Thl'! .m;1ss is assumed to be capable of motion in one
results and practical hardware implementation is very similar. direction only relative to the frame of reference under the
Examples of passive vibration and shock isolation, or cush- action of a disturbing force. A viscous damping device is intro-
ioning, to protect a sensitive area abound in living organisms. duced to model the observation that in real systems, vibration
For example, an instructive view of the human organism as a amplitudes reach a limiting amplitude at all frequencies.
mass elastic system with all the attributes of passive isolation This concept is used in the introductory material of most other
against the effects of shock is given in Reference 1 (Ref.l ). As standard texts on the subject and is referred to as a single degree
far as man made devices are concerned, they range from the of freedom system.
humble jogging shoe to the sophisticated suspension systems While this model imposes severe limitations on how well a
of modern road and rail transport and resiliently mounted deck practical case can be represented, it provides good insight into
houses on smaller ships. how the major variables affect the amount of vibration trans-
The effects of a vibrating machine on the surrounding area mitted to the surroundings from a machine.

1"
 Dynamic magnification factor:

.,
This concept leads to the well known equation of motion:

d2 dx ~
m ~ + c ~ + kx = F ( t) (1)
de dt (6)

where: m mass of the machine (kg)

d2x acceleration relative to the frame Phase angle of response:
de of reference (rnls2)
dx (rn!s) 2~Q
velocity 8 =tan - - -
.J
(7)
dt
(1 -d)
c damping factor (Ns/m)
X displacement (m) Where
k stiffness (N/m) (I)
F(t) a time dependent disturbing = frequency ratio
force on the system (N) COn
ro = phase velocity of vibration
If damping is neglected the phase velocity of the motion in
~ = fraction of critical damping ·
radians per second and the frequency in cycles per second at
resonance are given by: F 0 = excitation force (N) at timet= 0

(2) Many machines are subjected to a complicated array of

periodic disturbing forces, the magnitudes of which normally
vary with the speed and/or load on the machine. For example
the effects of mass imbalance increase with the square of the
rotational speed and in internal combustion engines the magni-
(3) tude of disturbing gas forces increases with the load.
f= (::) The effects of excitations are evaluated one at a time calling
on Fourier's theorem that a complex disturbance which is
repetitive in time can be decomposed into a series of harmonic
In common with most introductory texts eq. 1 assumes that
functions of different frequencies and the main task of the
damping is viscous in nature and increases linearly with ~ib_ra­
design optimisation of vibration isolation systems is to avoid
tion velocity. Viscous damping is not always the best descnphon
coincidence of excitation frequencies with the natural frequen-
of damping, but makes eq.1 linear and amenable to a neat
cies of the system. A Campbell Diagram is often used to
analytical solution. . . .. summarise the vibration characteristics of a system and a typi-
In the range of damping values expenenced m res1hently
cal diagram for a 10 MW Diesel alternator set is given in Fig.
mounted machines, damping has only a small effect on the
2. Operation at or near resonance gives rise to magnified vibra-
natural frequency and this remains also true of more complex
tion amplitude and force transmission into the foundation and
systems. For example t~e resonanc~ curve ?f th~ 7.5 MW
the design procedures seek to avoid this type of coincidence.
resiliently mounted medmm speed Dtesel engme dnven alter-
For a single mass, single degree of freedom system and using
nator set shown in Fig. 1 indicates a fraction of critical damping
the terminology of Ref. 2, the ratio of the force transmitted
of approximately 0.05 with a probable effect on natural
across the mounts to the applied disturbing force is termed the
frequency of less than one percent. .
transmissibility and is derived by:
If the disturbing force is a repetitive periodic funct10n of
time the dimensionless steady state response vibration ampli-
tude,of a simple one degree offreedom system is given in Ref. 2 Force transmitted to foundation:
by the well known expression for the particular integral: dx
I\=c~+kx
(8)
Non dimensional displacement amplitude: dt

The ratio of forces below and above the mount is:

~ = RI sin (rot- 8) (4)
xo
~ = Tsin (rot -lfl) (9)
Fo
Non dimensional velocity amplitude:
Where the force transmissibility is given by:

dx (10)
dt = QRI cos (rot- 8) (5)
Fo
(11)

If the excitation is caused by cyclic displacements of the

foundation rather than force acting on the machine then the

2
 Service speed motion transmissibility is identical to the force transmissibility.
~

e 0,60
Run down Non dimensional representations of the amplitude ratio and
transmissibility as a function of frequency for a linear single
degree of freedom system are given in many text books and are
therefore not reproduced here (See Refs. 2 and 5). The simi-
E 0,50 larities of amplitude ratio and transmissibility is recalled and it
E
...!_ is pointed out in passing that these relationships describe both
.±. 0,40 the response of a machine to self-excitation and to that caused
(])
"0
::::J
:t=
by movement of the foundation.
a. 0,30 The forces transmitted across resilient mounts to the foun-
E
<tl dation are in excess of the disturbing force if the frequency ratio
~ is below 1.4, whereas at a ratio of 3, only about 20% of the
~
..c
disturbing force would be transmitted.
5 Refs. 6 and 9 give useful if conservative guidance on the
desired frequency ratio of disturbing over natural frequencies
0,00 for solidly and resiliently mounted reciprocating machines to
200 240 280 320 400 440 480
avoid problems with vibrations transmitted to the foundations.
Engine speed (rev/min) The suggested ratio of 3 to 4 implicitly recommends that
resiliently mounted Diesel alternator sets are operated super-
Fig. 1 Vertical fmt order response of 7.5 MW diesel critically, well above the natural frequencies, and inevitably
alternator set resonances are traversed on the rwi up and coast down, making
-it necessary to consider transient magnified vibration ampli-
tudes. For medium speed engine driven alternator sets in the

lo popular power range 10 to 12 MW, the available mount

stiffnesses make it difficult to obtain this separation as with the
rubber mountings, natural frequencies below 4 Hz become
20 difficult to achieve even if the rubber elements are placed under
shear loading which lowers the stiffness of the mounts. Fig. 1
presents a fairly typical practical case.

3. APPLICATION OF THEORY & EXAMPLES

A single mass supported by a mount responds to excitation in
six directions, referred to as degrees of freedom: three transla-
15 tions, in each of the principal axes, and three rotations, about
the same three axes. If the mass were concentrated at a point
which also acts as the mounting point and as the centre
through which forces act and about which moments are
applied, each of these motions would be described by indepen-
dent properties of the machine and the mount in the specific
direction. A set of linearly independent differential equations
of motion could be written, the solutions yielding six modes of
vibration (EigenVectors), each with its own natural frequency
10
f) (Eigen value).
Such a model is flawed and much closer approximations to
reality are provided if account is taken of the physical separa-
tion between parameters. This induces cross coupling between
the different vibration modes and in practice various degrees
of complication are taken into account. Cross coupling occurs
because the mounting points are offset from the centres of
gravity and inertia and the disturbing forces and moments are
generally displaced relative to either of these.
For example, in Diesel engines, the transverse reactions on
the liner walls or cross head guides caused by the obliquity of
· the connecting rods, can excite both linear transverse move-
ment and transverse rocking. Moreover, if the engine is
supported by flexible mounts of equal stiffness, but unequal
loading, then any free residual pitching moment would induce
both vertical linear motion and pitching.
1,38/1 ,36 Hz
. Within reason it is possible to influence cross coupling of
1,07 Hz vibration modes by selective disposition ofthe mounting points.
For example equal loading of mounts placed below the mass
0 100 200 300 400 500 centre of the machine decouples the vertical modes and the
Rev/min popular 45 degree inclined arrangement of mounts reduces the
tendency to lateral rocking and stabilises motion induced in a
Fig. 2 Campbell diagram 10 MW diesel alternator set sea way.

3
 A classical treatment of the general equations which takes

As built

(
account of these factors and describes the motions of a single
rigid body on generalised mounts is given in Ref. 2. In practice
many complex multi-mass systems supported on several.
mounts can be reduced to simpler systems by calculating
equivalent masses, inertias and stiffnesses and the traditional
')
analysis methods are devoted to such calculations. These
models assume that the vibrating system can be reduced to one
or more rigid masses connected by spring stiffnesses and appro-
priate damping.
Where machines cannot be represented realistically by rigid
bodies, finite element models have been constructed and with
these the machines, mounts, foundations and the supporting
ground can be modelled in considerable detail. This approach
Crankshaft becomes sometimes essential in the case oflarge long machines,
rotation particularly slow speed Diesel engines where the assumption
that the engine does not ddorm under the influence of the exci-
tation forces is evidently an oversimplification.
For example the X-mode of engine vibration induces
twisting of the engine about the foundations. While slow speed
Diesel generating sets are not resiliently mounted, they are
supported by the ground which can be modelled as springs and
dash-pots, not dissimilar to resilient mounts. In this way the
analysis of resiliently mounted machines merges logically into
that of the more general case of machinery on elastic founda-
tions.
Diesel engines which operate super-critically on mounts
which are directly connected to the bed plate, derive no signif-
icant strengthening effect from the seating below the mounts.
As the seating does not contribute significantly to the stiffness
of the engine block and bed plate the deflections and stresses
due to all internal forces and moments have to be taken up by
the engine structure itself. Consequently not all engines can be
flexibly mounted directly, without the use of a substructure or
raft, but occasionally problems are experienced even if the
After fitting
engine design takes account of the additional loading. For
balance weights
example Fig. 3 illustrates the first order cyclic deformations of
the bed plate of an in-line 9 cylinder 10.6 MW resiliently
mounted 4-stroke engine which suffered from recurrent crack-
ing of oil sump welds. These areas would be relatively ·
unstressed in rigidly mounted engines and gave no problems in
such applications. The problem was overcome by a modest
I reduction of the internal first order moments by balance
weights and reinforcement of the sump to reduce stress levels.

-~·
Jl The practical difficulties of doing the work in situ enforced the
development of a technique which allowed welding on the sump
to be done from one side only.
~
Cli
C>·
sl 2,0 .----.,~.------r-------,,-------,

I I \
1 I
1,6 1------.lt---',~
-------·
K1 (1/2 & 1st order)
-·-··

I c E
:8 E
Ei1,2
1 ',
I
'
K2 (1st order)

I >
~li
::l '
'
I ~ ~ 0,8 1-~......-.-~.s:--t------

0~--------~--------~------~
600 700 800 900
Rev/min

Fig. 3 Vibration amplitudes at engine flange (first order Fig. 4 Vertical vibration of coupling between engine and

'
rotating vectors) altnemator

4
 Super-critically mounted machines describe considerable Measuring
post positions
excursions during start up and run down as the criticals are
traversed despite the provision of limiting stops and on board
· · ships, movements due to wave induced motion take place.
Transient amplitudes of several millimetres may occur at the
cylinder head level, but are acceptable as far as the machine is
concerned and provided that the various service connections
too can accommodate these deflections. Most ship board auxil-
iary generators and engines are rigidly supported together on
a common raft, which is in turn resiliently mounted, thus avoid- Corner
ing alignment problems between prime mover and driven checking
machine, but if the driven machine is mounted on solid chocks post
(steel or resin) and connected to a resiliently mounted engine,
as is often the case with the larger propulsion machines, a
coupling with high alignment tolerance is essential.
Fig. 4 shows the vertical vibration amplitudes on the input Stool
and output side of the coupling in the free end power take off Flexible
train of a flexibly mounted 2000 kW main engine of a modern mounting
research vessel. During the first few months of service the
engine to alternator alignment changed more than expected
and after a coupling failure it was corrected with some degree
of misgiving which led to long term monitoring of the alignment
in service. Using non-contacting sensors on the coupling and
micrometer readings of the height of the engine chocks, it was
0 found that the mean alignment remained reasonably constant
over the review period, but that significant transient excursions
Fig. 6 Detail of flexible mounting and alignment reference

occurred in service largely due to the effect of the ship's motion

on the soft mounted engine in rough weather. The operating Table 1: Engine to Alternator Coupling Flange
envelope of the alignment is shown in Fig. 5. The initial. align- Alignment
ment problem is likely to have arisen as the result of unequal
creep, probably caused by unequal loading. The simple Vertical Transverse
Condition (mm) (mm)
measurement set up to monitor creep (after the problem had
caused a coupling failure) is shown on Fig. 6 and could have Cold Static (Clock Gauges) +1.94 +0.53
been instigated earlier with negligible cost.
Lest this result be considered untypical the alignment Warm Static +1.25 +1.67
changes measured across the flexible coupling between the 10.8
MW Diesel Alternator set referred to earlier is summarised in Warm Dynamic No Load +0.36 +1.45
Table 1. The results include some operation in rough weather,
but the machine is fitted in a large passenger vessel where Warm Dynamic Maximum Load +0.30 +3.23
stabilisers would have had a moderating effect.
SIGN CONVENTION
Vertical - positive indicates engine below alternator.
Radial alternator coupling alignment Transverse - positive indicates engine offset to exhaust
(in service measurements) side relative to alternator.
All readings in mm,relative to alternator

High
Tolerance for _ ~'!! -1- !9__ 4. NOISE ATTENUATION
occasional .,.. ... - 1 - - .....

"v/ ;
excursions ,. ,. " r ', The foregoing considerations have been directed at the effect
1-8 .... ,
l on the machines of periodic repetitive disturbances in the typi-
I
/ a-6 cal frequency range up to about 100 Hz, but flexible mounts
I MEASURED VALUES also affect the transmission of vibration at higher frequencies
I Mean alignment particularly the transmission of noise into the foundations.
/ Calm
Rough weather Rigorous analysis in the higher frequency range becomes more .
{ weather difficult as the local deformations of attachment points of the
transients
I mounts can become significant and when the wave length of
Portl STBD
r-,-r"T-rT-. the vibration waves becomes similar to the dimensions of the
10 8 6 4 mounts, sudden increases of transmissibility take place.
I
\ Current practice is to derive experimental values for the
\ insertion loss of the mounts which are combined with the other
\
\ transmission path losses of the noise path between the machine
' Realignment
required in question and the space where the noise is perceived. The
general noise calculations used by the Technical Investigation
Department (T.I.D.) are described in Ref.17.
Resilient mounting of ship main propulsion engines offers
Fig. 5 Operating envelope of coupling alignment over a
a potent method of controlling the transmission of structure
3-month period

5
 borne machinery noise at source. A striking example is Diesel- 500
electric propulsion where the constant speed of the machines
allows a reasonable separation between the vibration frequen-
. Cies of the rigid body modes and the excitation frequencies at Co
service speed and benefits are thus derived from attenuating "E
the effects of vibratio'n as well as noise.
~400 0
.r::.
-H
C\i
On offshore platforms where again mainly constant speed
machines are involved, resilient mounting is also widely used.
In both areas of application the stringent living quarter noise
Cll
-o
:E
a. Co
I~ \
limits would be difficult to meet in other ways.
E
ttl
"E
0
.r::. i \
Typical values of audio frequency vibration below resilient
mounts of ship main engines reported by engine builders are
given in References 14, 15 and 16 and are summarised on a
c 300
.Q
~
.0
"5
r
~
·-..;l
\
common basis in Fig. 7. Ref. l3 illustrates a typical insertion 'til
I \ I

.,, /
ell
loss measured on an offshore platform below a small engine. ;
~ 200
Quite often generalised data base values are being used in ~ I
calculations of noise transmission and while this may be the
best that can be achieved, these general curves often misrepre-
Co
"E
0
\
.r::.
sent the insertion loss at discrete sharp resonances of the mount
itself or the local attachment structure. The topic is the subject
of ongoing theoretical and practical research. In the view of 100
co
\ /
the author effort to derive more reliable empirical data and \ ./
more realistic estimates of the accuracies of semi empirical '-''
predictions is likely to be the most cost effective approach.
I
120

110 - 60 70
Propeller rev/min
80

~
100
"· ---
,., / ' '
..........
!'" ...
,
....
...........

...... r---- ~'"·

'\.
\
Fig. 8 Fore and aft vibration of instrument console

q>
0

~
m
90

80
---- -~. r---
., ..........
[';-
on the bridge deck. Nevertheless, it was accepted by the owner
as being 'good enough' and no problems were reported in
subsequent service.
-o It is possible for transient vibration response of passive
70 devices to be also magnified in shock situations. For example
decelerations measured on the seats of an experimental free
60 fall life saving capsule were between 40 and 300 percent larger
than those measured on the floor of the capsule itself. While .
50 the decelerations were reasonably acceptable to human occu-
31 ,5 63 125 250 500 1000 2000 4000 8000
pants they were decidedly excessive in some conditions on the
Octave band centre frequency
seats. The seats were.fitted with shock absorbers and incorpo-
rated damping elements, but in the test configuration, when
loaded with a weight of a person, very short tt:ansient deceler-
Fig. 7 Vibration levels on engine seating in audio
frequency range ations, measured in milliseconds but up to 25 times greater than
that due to gravity, were recorded when the capsule was
I. .
dropped into the sea from heights between 25 and 35 metres.
5. PASSIVE VIBRATION & SHOCK Important requirements of resilient mounts used for shock
ISOLATION isolation are firstly the ability to absorb large deflections without
hitting stops, thus causing secondary impacts which could be
Analysis of the transient characteristics of resiliently mounted more severe than the initial event and secondly, damping should
equipment is of great importance in applications on war ships, be as high as possible to limit the 'Q' values below about 3.
road and rail transport, but it is also of concern in seismic work Resilient chocks of floor mounted equipment cabinets
and the construction of laboratories which use sensitive equip- should be arranged to restrict the height of the mass centre
ment. The typical general application is that of the response of above the mounts to avoid large movements at the top. A ratio
resiliently mounted equipment to a sudden shock. The classical of height to mount spacing of < 0.4 should be aimed at.
mathematical approach Involves the solution of the comple- Horizontal stays can be used where this is not practicable.
mentary function of the equations. of motion and in more The entire subject of passive isolation is likely to gain in
modern terms digital numerical techniques in the time domain importance. At the time of writing this paper, Lloyd's Register
involving step by step calculations. Calculations of the transient is a joint sponsor, together with a number of mount manufac-
response of a system to step and pulse functions bear obvious turers, engine builders and defence interests, of a research
similarities to control engineering concepts. . project into the transient and steady state response of resilient
On a practical level Fig. 8 shows the fore and aft vibration mounting of machines and instruments. To date this work has
response of a resiliently mounted !:>ridge console of a VLCC resulted in an algorithm and suite of computer programs suit-
measured in the early 1970s. The same natural frequency was able for many facets of this type of analysis, including
excited by third and fourth harmonics of the propeller blade calculations of transient response to shock and the treatment
rate and the response was in excess of the amplitudes measured of non linear properties of mounts. tJ
6
 6. PRACTICAL ARRANGEMENTS AND
MOUNT PROPERTIES
~
·-~
It is possible to outline only the most commonly encountered 60
methods of supporting equipment resiliently within the context
of this paper. Discounting vehicle suspensions, those most likely
46 50
to be met are various forms of rubber elements and coiled steel
springs and popular methods of arranging the mounting points
of sizeable machinery are shown on Fig. 9. A wide ranging 40
account of modern industrial practice is given in Ref. 5, and the ~
following summarises the author's experience. "0 30
_§
20

10

Deflection (nim) 10,5

Compression Inclined shear Common raft - rarely usea
mounts- mounts- for ship's main engines Stiffness: Based on 08 = 4,4 Kn/mm (5Hz)
typical for common for Based on AB = 8,4 Kn/mm (6,9Hz)
ship's large main
D
., generators
and in large
industrial
engines Fig. 10 Load/deflection characteristics of rubber monnt

plant
where z is the·static deflection of the mount due to the supported
Fig. 9 Typical engine monnting arrangements mass in millimetres.
This simple relationship, frequently presented in the form
of nomographs which bring in the various parameters affecting
Rubber mounts which are often used in preference to metal deflection and allowable working loads, can be used for the
springs, have non linear stiffness properties and possess a selection of mounts in simple non critical applications, but diffi-
certain amount ofhysterisis which in calculations is represented culties can arise with the non-linearities.
by viscous damping. This is illustrated by Fig.lO which indicates that the stiffness
The non-linearities depend on the load on the mount, the of a non linear mount can be defined in different ways. Common
frequency of vibration and temperature and are normally sense suggests that the stiffness at the static loading imposed in
approximated by using different stiffness values for the static service should apply, but in the dynamic situation the stiffness
deflection calculations and dynamic calculations. It is also is increased further by a factor which depends on the grade and
necessary to take into account that the connections of the vari- type of material. Typical ratios of effective static to dynamic
ous services to the machines are in effect additional flexible stiffness are in the range 1.1 to 2.0 for natural rubber. Table 2
mounts which, if they are large compared to the stiffness of the lists some of the most important properties of rubber mounts
mounts, affect the natural frequencies and response ampli- published by different sources.
tudes. For reasons stated earlier, it is desirable to dispose rubber
If machines operate above the natural frequencies it is mounts under machines to achieve equal loading, varying the
vital that the stiffness of these additional connections is not spacing to suit. Where mounting positions are fixed and cannot
overriding that of the resilient mounts. Finding the correct be moved the stiffness of mounts is often adjusted to achieve
dynamic stiffness of flexible mounts for calculations is far from equal static deflections. Failure to achieve balanced static load-
trivial as catalogue values may.be based on static rather than ing can result in unequal dynamic properties and unbalanced
dynamic tests.
Rubber in the natural and synthetic forms is the material
used by preference for mounting Diesel generating sets, but
Table 2: Typical Characteristics of Rubber used for
steel springs have been used for quite large machines and while resilient mounts
they lack the inherent damping of rubber elements they have
the major advantage of linear load/ deflection characteristics. Hardness (Shore) 40-60
Combined rubber/steel spring mounts are available in load
ranges up to about 8 tonnes, ·and some makes· have· the general · Specific Gravity 0.9-1.2
approval of Lloyd's Register.
The non linear load/deflection characteristics of rubber Typical Static Stress 1-1.4 N/mm2
mounts result in a dynamic stiffness of the system which
depends on the deflection thus affecting the natural frequency. Typical Dynamic Stress 0.4-0.8 N/mm2
If damping is neglected in eq. 1 the natural frequency in cycles
per second of a linear system is given by: Static Strain (%) 20

Natural Rubber Synthetic Material

16VITJ Tensile Strength N/mm2 28

Dynamic Magnifier 15-25
7-20
10-15
Stiffness Ratio 1.1-2.0 Up to 40

7
creep with a risk of affecting alignment. This characteristic is 7. EXPERIMENTAL ANALYSIS
of evident importance as creep of 5 to 8% must be expeated,
with harder grades of rubber being more sensitive. Thus the It may be thought that in this era of ever more powerful comput-
ers, and matching analytical software, there should be no need
fJ»
settlementto be expected over the first few months at a coupling
is of the order of several millimetres with all being well and for vibration measurements and all the attendant problems of
could be substantially greater if unequal creep results in tilting. analysis. Similar considerations have, for example, led to a vast
Coiled helical steel springs are statically and dynamically reduction of stress measurements on models of all kinds and
stable and can give very low natural frequencies, but they may their effective demise in the Society. The same fate had been
need additional damping elements to limit the vibration ampli- predicted for vibration tests, but rather than seeing the elimi-
tude of transient criticals. Fig. 11 shows the general disposition nation of tests, it can now be argued that more, rather than less
of a successful spring and dash pot arrangement used to support full scale vibration testing should be done.
an llMW Diesel alternator set. This reflects the recognition that results of calculations are
only as good as the assumption on which they have been based.
Machinery excitation levels are not always as well known as
one would hope, particularly the harmonic components of the
combustion process in the latest types of engines; the stiffness
1. of joints can be an impond,erable and the machinery may sit on
resilient mounts supported by a base which itself has elastic
properties which are not always well known.
Against this background the main purposes of vibration
2. tests are to demonstrate acceptability, to confirm the results of
calculations and to provide factual data for the investigation of
3** the causes of problems.
Within the bounds of economical common sense, there is no Ifill<.
single test program or choice of equipment which is ideal for ~
1'-
4. all three purposes and the selection of suitable equipment is
co ""'"
Q)
(0
co
Q)
1'- not always given the attention it warrants.
E 5. The choice of measuring quantity, type and number of trans-
iQ)
6**
ducers and ,measuring positions, and the required analysis
dictate to a large extent the recording medium, but with poorly
1D equipped firms it is not unknown for tests and analysis to be
c. 7.
E planned around the available equipment rather than the other
0
0 ae way around.
0 To maintain and keep a comprehensive measuring capabil-
0 ity up to date is an expensive undertaking in terms of staff and
0 9** hardware costs and the Society deserves credit for maintaining
a range of modern equipment which is not matched by many
--~ ---~ ·16-e- similar organisations in the marine and industrial field.
For acceptability tests, the transducers are best chosen to
-- --- ----~
~4·- avoid the need for multiple integrations, but the overriding ,
12 •
concern must be to ensure that the instrumentation has an
..\1::
0
0
adequate frequency response and amplitude sensitivity over
iD 13** the range of interest. It is noted that measurement of transmis-
0 sibility as the ratio of two forces is not practiCable, but in typical
0 work, with the types of machines considered in this paper, it is
~0 ~14. rarely of direct interest. Vibration amplitudes below mounts &'\
are normally smaller by ,about one order of magnitude than ~
15. those above. Where transmission of vibration into the
surroundings is of concern the measurements below mounts
16** become of interest and may require the use of transducers with
greater sensitivity than that of the ubiquitous Piezo-electric
17. accelerometers. Displacement amplitudes of the order of a few
microns are typically involved.
The choice of test positions and conditions under which a
machine is to be tested needs careful consideration to ensure
that the test objectives are met within a reasonable time span.
18** While there is normally pressure to keep the amount of testing
down there is an irreducible minimum of positions.
19. If for example the six lowest modes of vibrations of a Diesel
alternator are to be identified reliably, measurements in the
three principal directions are desirable in at least six positions
on the engine and four on the generator, assuming the latter is
· close coupled, but normally additionaJ readings would also be
e Spring
required on the engine mounted ancillary equipment and not
* Spring - Dashpot Unit ' ' least, the turbochargers, exhaust manifold and governor.
For investigation work around 24 triaxial measurements
would be typical on a Diesel alternator set.
Fig. 11 Positions of springs and dashpot nnits Variable speed tests to identify resonances on all these items ~)

8
 would require impractical testing times with single channel tion signals TID also use 'real time analysers' for field balancing
measurements, putting a premium on multi-channel data acqui- of turbo machinery and, on reciprocating engines, they have
sition and fast and repeatable data analysis. On a large been used to deal with first order vibration caused by residual
generating set, even-when using multi-channel equipment, test- mass unbalance (Ref.18);
ing in all relevant conditions is likely to involve a few days rather Modern data analysis techniques are riddled with jargon and
than a few hours if more than simplt< acceptance tests are full of pitfalls for the unwary because of the enormous potential
involved. of processing the signals in different ways. A glossary of
Much of vibration analysis of steady repetitive signals frequent]y used terms is included, but is by no means exhaustive.
consists of a decomposition of the signal into the individual Surveyors are most strongly urged to ask for explanations
vibration components to compare them with acceptability stan- of unfamiliar terms and above all to insist on meaningful anno-
dards, or in investigation work, to identify the modes of tations on the output from the various instruments which are
vibrations and the frequency of the major disturbing forces. used to produce output submitted for consideration.
Until not so long ago, analysis by a number of techniques
described in Ref. 4 was used or signals were digitised on a digi-
tising tablet and subjected to harmonic analysis on a time or
order basis, using a Fast Fourier Transform algorithm.
Such work is extremely time intensive if a comprehensive
analysis is to be presented formally, but using a few channels
of simultaneous analogue information, experienced engine,ers
can reach conclusions on site more quickly and safely than in
any other way, particularly near resonance. Above all, such
assessments are made on the basis of the total vibration signal,
an elementary but important feature which is missed only too
easily with modern types of analysis.
0 The availability of PC computers with multi- channel direct
data input has transformed the scope for rapid and reproducible
analysis, as has PCM (Pulse Coded Modulation) data acquisi-
tion which compresses the data of many channels into a
multiplexed digitised data stream.
TID uses a sixteen channel PCM system which can sample
data at the rate of 80000 samples per second, routinely for the
type of application described in this paper and alternatively PC
computers with interfaces capable also of accepting 16 channels
at an even higher scanning speed. These systems greatly expe-
dite data analysis and presentation of results for reporting.
The task of harmonic analysis has also been eased by the so
called 'real time analysers'. These specialised machines give
very rapid harmonic analysis of signals, usually on one or two
channels, and also determine phase relative to a marker and
are suitable for modal analysis. The reference to 'real time' is Fig. 12 Transverse rocking .mode of 10 MW diesel alternator
strictly misleading as they work by acquiring a signal for a finite set on rubber mounts
length oftime which depends on the frequency range to be anal-
ysed. The stored signal is then subjected to analysis.
The results are displayed on a fluorescent screen in the form
of a plot of the amplitudes ofthe vibration components against 8. ACCEPTABILITY
frequency at each of the discrete frequency increments which
can be resolved. The amplitudes are shown as vertical lines and Acceptability of vibrations of large non reciprocating rotating
the plot represents in the frequency domain the harmonic machines such as motors, fans, pumps and rotary compressors
component of the signal which had been analysed and is is defined in ISO 3945:1977 (Ref. 9) which in turn is based on
referred to as a vibration spectrum. The frequency span over ISO 2372(Ref.10). On this basis resiliently supported machines
which the analysers work ranges from a few cycles per second are assessed more leniently and RMS vibrations up to 7.1 mm/s
to MHz, but for each such span there is a finite frequency reso- are acceptable compared to 4.6 mm/s for rigidly mounted
lution limited by the number of possible lines, commonly 400 machines. The limiting values recommended by these docu-
and increasingly 800, on the screen. Modern real time analysers ments are summarised in the Society's Guidance Notes
have a great many additional functions described in depth in (Ref. 11) on vibrations which also refer to reciprocating
Ref. 3. For example correlation checks have become readily machines and human comfort. Ref. 11 is presented in Appendix
possible and can be used to check how regular or random a A as it may not be available to all surveyors. It should be noted
signal is and how it relates to some other signal which intuitively that the Guidance Notes are not Rule requirements. Further
or otherwise is thought to be related. guidance on vibration levels acceptable for human comfort is
One of the most useful features of the analysers used in TID given in Refs. 7 and 8.
is the ability to display the phase of the the vibration compo- No definite international standard exists at present for the
nents relative to a timing signal taken from the machine under assessment of reciprocating engines and compressors, but a
test and it is the basis of all experimental modal analysis work, rather complicated draft ISO document (Ref.12), which in part
a task which forms an essential part of investigations of prob- takes account of the experience of Lloyd's Register, was issued
lems or confirmatory measurements. in 1989. This proposal limits narrow band·RMs vibration veloc-
A typical result obtained on a 10 MW resiliently mounted ities at service speeds to 28mmls for rigidly mounted engines
Diesel alternator set is shown on Fig. 12. Because of the ease and 45 mm/s for resiliently mounted engines in the frequency
with which these instruments can measure the phase of vibra- range 10-250 Hz. Below 10Hz the suggested limits are based

9
on RMS displacement and are 0.45 mm and 0.71 mm respec- the higher the crest factor the more impulse like is the signal.
tively and above 250 Hz acceleration based criteria are 44.4
mists and 70.4 m/s/s respectively for the rigidly and resiliently Critical damping fraction:
mounted machines. See damping ratio
For the ISO assessment the vibration should be measured
on the engine frame, at the cylinder head level. The similarity Cross correlation:
of form to the curves given in the LR guidance notes on vibra- Defines how a signal relates to another signal. A high correla-
tions will be noted. tion indicates that two signals may be casually related.
It is interesting that the form of the criteria, which in the
case of the LR guide lines are based on a completely empirical .Cross coupling:
analysis of service experience, mirrors the statement made Stimulation of more than one vibration mode by a given
earlier that stiffness controls vibrations at low frequencies and excitation.
mass those at high frequency, in the sense that force equals stiff-
ness times displacement or alternatively mass times Damping:
acceleration. Defines losses in a vibrating system which at resonance deter-
The experience of the author suggests that the draft ISO mine the amplitude of the response and which are normally
document specifies safe levels, but that it is difficult to justify taken to be directly proportional to vibration velocity for ease
the more lenient treatment of resiliently mounted machines as of computation. '
the criteria have been formulated to a large extent to protect
engine mounted ancillary equipment and bolted connections. Damping ratio:
Start up and run down transient vibrations are not covered. Ratio of the actual damping coefficient to the critical damping
They are unlikely to cause problems to the machines them- coefficient which would make the system just non oscillatory.
selves, but the service connections have to be able to accept the
deflections. Displacement:
The instantaneous deviation from the static rest position.

9. TERMINOLOGY Eigen value:

In the context of vibration analysis the angular velocity corre-
Acceleration: sponding to the natural frequency. The name is derived from
The second differential with respect to time of the instanta- the matrix metnod of solving the simultaneous equations which
neous displacement of a vibrating particle from the static rest are used to model multi-mass systems.
position. During vibration, acceleration varies throughout each
cycle reaching maximum values at the positive and negative Eigen vector:
extremities of displacement. The vectorial amplitude corresponding to the Eigen value.

Amplitude: Frequency:
The magnitude of the periodic displacement from the rest posi- Number of cycles completed in one second, normally expressed
tion of a vibrating body in the positive and negative sense. 2* A as Hertz (Hz).
is often quoted and should be correctly defined as the peak to
peak amplitude. Frequency ratio:
The ratio of the frequency of concern to the natural frequency
Auto correlation: of the system.
Indicates how a signal relates to itself over a certain time span.
It is used to check for the presence of a small periodic signal in Logarithmic decrement:
the presence of high random (noise) signals. A pure sine wave Natural logarithm of the ratio of amplitudes of two successive
has a high auto correlation, whereas white noise has no corre- cycles of damped free vibration following a stimulus such as an
lation except at the starting point. See also entry under impact. It can be used to derive system damping.
correlation.
Mode:
Campbell diagram: Describes the vibration form taken at a natural frequency when
A plot of disturbing frequency against rotational speed with particles vibrate in a simple harmonic pattern. It is often defined
superimposed horizontal lines of the natural frequencies of the by quoting the number of nodes, a characteristic deflection
system. The disturbing frequencies occur at integer multiples shape or the natural frequency.
of the fundamental frequency of the disturbance referred to as
orders. The intersections of the lines of disturbing frequencies Natural frequency:
with natural frequencies indicate critical speeds. . Frequency at which a given system vibrates in a pure mode and
takes its maximum amplitude. A system usually has several
Correlation: natural frequencies corresponding to different vibration
Defines how a signal relates to itself or to another signal in modes. Natural frequencies defined in terms of displacement,
the time domain. It is given by the integral of the product of velocity and acceleration differ by insignificant amounts for a
the signals at a given time and at a short interval later. given mode.

Crest factor: Node:

The ratio of the peak amplitude of a complex signal to the root A point of zero dynamic deflection. Each vibration mode has
mean square value. Crest factors of 2 to 3 are commonly met a characteristic pattern of nodes giving rise to definitions such
in engineering situations. The purer the vibration signal the as 'two node mode of vibration at first order frequency'.
closer the ratio of crest factor over RMS approaches 1.414 and
Order: estimated from the ratio of vibration amplitudes above and
The number of times a vibration cycle is completed in the funda- below the mounts.
mental time interval, commonly the time for one shaft
revolution, but, with a four stroke engine, the combustion cycle
i.e. two revolutions.
10. REFERENCES
Peak to peak value:
Twice the peak value. 1 Griffin M.J., Handbook of Human Vibration, 1990
Academic Press Ltd.
Peak value:
The maximum instantaneous amplitude of a signal. In the case 2 Harris M.C. and Crede C.E., Shock and Vibration
of a pure sinusoidal vibration the peak value is 1.414 times the Handbook, 1961, Me Graw-Hill
root mean square value. See also 'Crest factor'.
3 Smith J.D., Vibration Measurement & Analysis, 1989,
Period: Butterworths
The time (normally quoted in seconds) taken to complete one
vibration cycle at a particular frequency. 4 Manley R.G., Waveform Analysis, 1950, Chapman &
Hall
Phase:
The time domain relationship between disturbing force and 5 Macinante J. A., Seismic Mountings for Vibration
response expressed in terms of an angle. Isolation, 1984, John Wiley & Sons.

Q-factor: 6 CP 2012: Part 1:1974, Code of Practice for Foundations

Dynamic amplification factor of system. For a lightly damped for Machinery, British Standards Institution.
system Q is found experimentally by calculating the ratio of the
frequency span of the resonance curve at the half power point 7 ISO 2631: Evaluation of Human Exposure to Whole
(maximum amplitude/1.414) to the system natural frequency. Body Vibration, 1985, International Organisation for
Standardisation.
Resonance:
Coincidence of the frequency of a disturbing force with the 8 BS 6841: 1987 Measurement and evaluation of human
natural frequency of a system. At resonance vibration ampli- exposure to whole body mechanical vibration and
tude progressively builds up over the initial number of cycles repeated shock, 1987, British Standards Institution.
to reach a limiting amplitude which is controlled only by the
damping in the system. 9 ISO 3945: 1977, Mechanical vibrations of large rotating
machines with speed range from 10 to 200 rev/s.
Root Mean Square (R.M.S.):
The square root of the mean of the squares of discrete samples 10 ISO 2372: Mechanical vibration of machines with oper-
taken over a given time interval. It is a measure of the power ating speeds from 10 to 200 rev/s -Basis for specifying
of a signal. Integration times of 1 second and 10 seconds are evaluation standards, 1974, International Organisation
commonly used in analogue meters and it is important to note for Standardisation.
the different averaging effects.
11 Lloyd's Register, Guidance Notes on Acceptable
Spectrum: Vibration Values.
A plot of the amplitudes of individual vibration components of
a signal (Frequency domain plot). 12 ISO/TC/108/SC2/WG5/N34- Measurement and
If the frequency of these is a multiple of a basic interval, such Evaluation of Mechanical Vibrations of Reciprocating
as the combustion cycle of a reciprocating machine or shaft Machines with Power Ratings above 100 kW. 1989,
rotational speed, then the spectrum contains discrete dominant International Organisation for Standardisation.
lines. If the frequency content of a signal is distributed evenly
over the band width of analysis a continuous spectrum is 13 Kunz F: "Vibration Control on Diesel Engines",
obtained, indicating white noise or rapid random impact type Institution of Diesel and Gas Turbine Engineers. (Fig. 9)
characteristics. Seminar "Health and Safety Aspects of Diesel Power
Stations" 1986.
Sub-critical:
Operation below the critical frequency. Normally used for rigi- 14 Jofs K: "Machinery Concepts for Advanced Passenger
didly mounted machines, but notable exceptions are marine Ships", The Institute of Marine Engineers, IMAS 88,
high pressure steam main propulsion turbines which are run The Design and Development of Passenger Ships
generally at super-critical speed at NCR.
15 Ohlers W.: "The Re-Engining of the QE2 -The largest
Super-critical: Diesel Power Station afloat", The Institute of Marine
Operation above the critical frequency, generally aimed at with Engineers, IMAS 88, The Design and Development of
resiliently mounted machines. It results in much quieter Passenger Ships.
running.
16 Hansen L.R.H.: "Propulsion System Installation in
Transmissibility: Passenger Cruising Vessels", The Institute of Marine
Ratio between the exCitation force and the force transmitted Engineers, IMAS 88, The Design and Development of
to the foundation across resilient mounts. If local deflections Passenger Ships.
at the attachment points of the mounts are discounted it can be

11
 APPENDIX A:
17 Kunz F: "Passenger Ship Vibration and Noise" The GUIDANCE NOTES ON ACCEPTABLE
Institute of Marine Engineers, IMAS 88, The Design VIBRATION LEVELS AND THEIR
and Development of Passenger Ships. - ---- MEASUREMENT
18 Kunz F: "Solution of machinery problems by field inves- I Background Information
tigation". Tran.l. Mar. E., Vol. 101, pp43-55.
1. General
1.1 The Vibration Guidance Notes have been compiled
from Lloyd's Register's experience of measuring, interpreting
and assessing shipboard vibration. In the areas covered by the
Guidance Notes, these activities can be divided into two main
categories:

(i) Routine vibration measurement surveys.

(ii) Investigation of vibration problems.

1.2 Where possible, relevant International or National

standards have been used to assess vibration measurements.
Certain existing standards therefore form the basis for the
Guidance Notes, and the proposed assessment criteria embody
the experience gained in their interpretation, adaptation and
extrapolation in relation to shipboard conditions. Reference
has also been made to other reputable published information.

2. Basis of Assessment Criteria

Details of the Standards used as references for the assessment
criteria are given in the following sections.

2.1 Accommodation and Workspaces

Guidance values are based on I.S.O. 6954 "Guidelines for the
overall evaluation of vibration in merchant ships". The limits
of this Standard have been adopted as Fig. A1 of the Guidance
Notes. Amendments and interpretation are intended to facili-
tate:

(i) Simpler analysis of routine survey measurements.

(ii) Evaluation of the sum of two or more frequency
components.
(iii) Differentiation between rest and working areas.

2.2 Local Structural Vibration

Guidance values are empirically based on the Society's experi-
ence of vibration induced structural cracking. The principal
source of the data concerns the vibration of panels relative to
their stiffened boundaries.

Structural fatigue is a complex failure mode and of necessity

the guidance values have a broad stippled zone to cater for the
wide variations encountered with regard to:

(i) Structural configurations and geometry.

(ii) Vibration mode.
(iii) Weld details.
(iv) Environment.

2.3 Machinery Vibration

2.3.1 Rotating Machines

Guidance values are based on I.S.O. 3945 "Mechanical vibra-
tion of large rotating machines with speed range from 10 to 200
revs - Measurement and evaluation of vibration severity
in situ".

12
 This standard has been extrapolated to assess: 2.2 Integrators may be used for conversions of velocity
signals to displacement, or acceleration signals to velocity or
(i) The low frequency vibrations associated with main displacement.
propulsion machinery.
(ii) Auxiliary machinery of lower power than the stipulated 2.3 Filters may be used to restrict the frequency range of
300kW minimum. broadband measurements.
(iii) Rotary pumps and hydraulic machines. This has entailed
an application factor. 2.4 The measuring system should be calibrated in all
vibration units of interest.
It is envisaged that other specific application factors may be
introduced as further experience is accumulated. 2.5 The characteristics of the measuring system shall be
known from calibration with regard to the following:
2.3.2 Reciprocating Machines (i) Frequency response.
The upper boundary of the Guidance values shown in Fig. A3 (ii) Effect of transducer orientation and cable length.
corresponds to Verein Deutscher lngenieure- "Measurement (iii) Temperature.
and Evaluation of Mechanical Vibrations of Reciprocatil)g
Piston Engines and Piston Compressors". 2.6 The specifications' for vibration transducers, filter
The standard has been extrapolated to assess machinery of characteristics, signal conditioning, display and recording
lower power than the stipulated 100kW minimum. equipment and calibration procedures should conform to rele-
The lower boundary of Fig. A3 and the broadband limits vant international standards.
quoted in paragraph 5.3.5.1 are empirically based on measure-
ments taken from troublefree installations. They also take 2.7 Permanent records of vibration measurements may
r account of data published by the Japanese Marine Engineering be in the formof:
U Society.
Reference: "Report on the Survey of Marine Engine and (i) Paper oscillographs.
Machinery Vibrations"- Bulletin of the M.E.S.J. Vol3 No. 3. (ii) Magnetic tapes.
(iii) Plots of vibration spectra from a narrowband frequency
analysis.
II Vibration Levels and Their Measurement
2.8 Narrowband frequency analysis should be capable of
1. Introduction resolving the frequency of individual components to 1% band-
1.1 These guidance notes define the application of width over the relevant frequency range.
proposed criteria for assessing the severity of shipboard vibra-
tion in the following areas: · 2.9 Broadband measurements should cover the specified
range of vibration frequencies for each particular application.
(i) Accommodation and workspaces. (Habitability)
(ii) Local structural vibration. (Risk of cracking)
3 Accommodation and workspaces
(iii) Machinery vibration. (Risk of damage or accelerated
wear) 3.1 Scope
1.2 Vibration velocity amplitude is adopted as the prin- Evaluation of hull and superstructure vibration with regard to
cipal parameter for evaluating shipboard vibration. crew habitability in all normally occupied spaces. The
frequency range considered is 1 to 100 Hertz. The evaluation
1.3 Narrowband frequency analysis is the recommended is not intended to apply to manoeuvring conditions or operation
method of interpreting vibration measurements in relation to in heavy weather.
the guidance limits.
3.2 Measurements
1.4 Some guidance values based on broadband measure-
ments of the overall amplitude of composite frequency 3.2.1 Measurements should be taken on the deck in three
vibration are suggested to simplify routine survey procedures. mutually orthogonal directions: vertical, athwartships and
These methods may require supplementary measurements longitudinal.
of overall displacement and acceleration amplitudes.
3.2.2 Where normal transducer attachment methods are
1.5 Unless specified otherwise vibration amplitude is half impractical, due to the presence of deck coverings, an inertia
the peak to peak value of the vibration. block may be used to locate the transducers.

1.6 If vibration severity is marginal from the results of 3.2.3 The severity criteria are intended to apply to the posi-
broadband measurements, narrowband frequency analysis tion of maximum vibration within the space under
should be used. Records of the vibration measurements should consideration, and vibrations in each of the three directions
be taken in sucli cases. should be assessed independently.

2. Measurements and analysis 3.3 Test Conditions

2.1 Measurements should be made' with an electronic 3.3.1 Measurements should be taken with the vessel
system employing transducers which generate signals propor- proceeding ahead at a constant speed and course in a depth of
tional to velocity or acceleration. water not less than five times the draught.

13
 3.3.2 Tests should be conducted in sea conditions not 5 Machinery Vibration
greater than State 3.
5.1 General -~)
3.3.3 The vessel should be at a displacement and trim corre-
5.1.1 Measurements should be taken with the IIlaclime at
sponding to a normal condition.
normal operating temperatures.
3.3.4 Rudder angle variations should be limited to ± 2o of
5.1.2 The assessment criteria apply to all operating speeds
the amidships position.
and loads provided the running conditions are stable.

3.4 Assessment 5.1.3 With the exception of main propulsion machinery,

broadband measurements are valid if the vibration severity
3.4.1 The following severity classification applies t~ bro~d­
with the machine operating is greater than three times the back-
band measurements of the maximum overall VIbratiOn
ground value with the machine stopped. Otherwise
amplitude (peak) for vibrations within the frequency range 1
narrowband analysis is required.
to 100 Hertz.
Values are the maximum repetitive amplitudes which occur
5.1.4 Narrowband severity should be assessed on the ampli-
during a minimum measuring period of 30 seconds.
tudes of the frequency components generated by the machine
itself.
3.4.1.1 Vibration levels are excessive if:

(i) Velocity exceeds 15mms·l. 5.2 Rotating Machines

and
(ii) Acceleration exceeds 0.3ms·2. 5.2.1 Scope

3.4.1.2 Vibration levels are normal if:

Assessment cif vibration severity for rotating machines above
10kW.
G»
(i) Velocity is below 6mms·I.
5.2.2 The severity of rotating machinery vibration is
or
defined by the root mean square (r.m.s.) value of velocity for
(ii) Acceleration is below 0.15ms-2.
frequencies up to 1000 Hertz.
3.4.2 For intermediate vibration levels severity should be
5.2.3 Measurements are to be taken at or adjacent to each
assessed from a narrowband frequency analysis of the measure-
main bearing in two orthogonal radial directions and in the axial
ment in relation to the guidance values shown in Fig. Al.
direction referred to the shaft axis. Care niust be taken that the
measurements do not include local resonance effects. ·
3.4.2.1 Values are the maximum amplitudes of ~ndivid~al
components which occur during a minimum measunng penod
5.2.4 The assessment categories given in Table A1 apply
of 30 seconds.
independently to each measuring direction.
3.4.2.2 No single component should exceed the "upper
5.2.5 The distinction between "rigid" and "flexible"
boundary".
supports is defined as follows:
3.4.2.3 For sleeping and rest areas only one component may
(i) Rigid Support- Natural frequency of the predo~~ant
be within the stippled zone.
vibration response mode is greater than the excitatiOn
frequency.
4. Local Structural Vibration (ii) Flexible Support- Natural frequency of the .ti\
predominant vibration response mode is less than the ~J
4.1 Scope excitation frequency.
Evaluation of structural vibration with regard to the risk of
Where the support stiffness cannot be classified the assessment
cracking. The assessment criteria are restricted t? locally
should be based on the "rigid support" values.
magnified vibrations of steel structural_eleme~ts relative to the
boundaries. The frequency range considered 1s 5 to 300 Hertz.
5.2.6 For main propulsion machinery narrowband
frequency analysis will be required if the broadband vibration
4.2 The assessment levels are shown in Fig. A2 ap.d apply
velocity exceeds 4mm-I r.m.s.
to unidirectional narrowband measurements taken at the posi-
tion of maximum vibration amplitude.
5.2.7 For bearings associated ·with rotary pumps and
hydraulic machinery the values given in Table A1 may be multi-
4.3 It is recommended that the overall vibration ampli-
plied by an application factor of 1.5.
tudes (peak) for vibrations in the frequency range 5 to 300 Hertz
are less than each of the following values:
5.3 Reciprocating Machines
(i) Displacement 0.25mm.
(ii) Velocity 30mm-I. 5.3.1 Scope
(iii) Acceleration 20ms·2.
Assessment of vibration severity on rigidly and flexibility
4.4 ·Alternatively, structural vibrations may be assessed mounted reciprocating machines above 20kW. The frequency
range considered is 2 to 300 Hertz.
by dynamic strain measurements in conjunction with relevant
material fatigue data. !) )

14
i
! !
5.3.2 The severity of reciprocating machinery vibrations is
defined by the peak vibration amplitudes.

5.3.3 Measurements are to be taken on the machine frame

(entablature) in three mutually orthogonal directions referred
to the shaft axis. The assessment criteria should be applied to
the position and direction of the maximum vibration amplitude.

5.3.4 The assessment levels given in Fig. A3 apply indepen-

dently to each measurement. Displacement limits below 15
Hertz only apply to rigidly mounted machines.

5.3.5 Assessment
5.3.5.1 Overall peak amplitudes for vibrations in the
frequency range 2 to 300 Hertz should be less than each of the
following values:

(i) Displacement 0.4mm

(ii) Velocity 25mm-t.
(iii) Acceleration 40ms·2.

5.3.5.2 The assessment of higher vibration levels should be

based on a narrowband frequency analysis_ referred to in
Fig. A3.

Table A 1: Vibration assessment - rotating machines

Vibration Severity
MMS-1 R.M.S.
Assessment
Category Rigid Supports Flexible Supports

Good
2 3
Satisfactory
5 7
Unsatisfactory
11 18
Excessive

15
'!!

?oo.
~~'-'
100
,.... ~~
·~I ,~

~c;
·~~
~
~ &'S

->-
c.J
0
<:))/

Q)
>
c:
-0

,_
( 'C

..c
>
10

100 1000
CPM

Fig. Al Vibration Levels -Accommodation and Workspaces

16
 I
CZi.
~

-• -:::...~
. '"':)~
E1 100 t;J')'
--mt
,E

(oo
-·
r
( c.: "?)
'-----

--
>
( .)
0
~~
~~~
~~

CD
>
c:: "'
-
0
«S
'-
..0
·-
>
10

Fig. A2 Local Vibration Levels - Steel Structures

11
·~,100

-
......
(.J
0
ClJ
>
c:
0
......
cc
~

:e 10
>

1~~~~~~~~~~~~r-~~~
2 10 100 300
Fr§Huency {Hertz} ~U\~

100 1000 10000

CPM

Fig. A3 Vibration Levels - Reciprocating Machines

18
Lloyd's Register Technical Association

Discussion
on the paper

VIBRATION ISOLATION
by

F. Kunz

Paper No. 1. Session 1990-91

 Lloyd's Register Technical Association

Discussion
on the paper

VIBRATION ISOLATION
by

F. Kunz

Paper No. 1. Session 1990-91

ERRATA
DISCUSSION: From Mr P. Koller: Page 1

Replace second sentence of the fifth paragraph with the following:

In the geared turbine department where I worked a velocity of
4.5mm/s was considered a normal limit for shop trials with a
maximum of 7.Omm/s in certain circumstances.

Replace third sentence of the eighth paragraph with the following:

Limits are clearly specified and include mechanical and electrical
runout which should not exceed 6Jlm

Replace the sixth sentence of the eighth paragraph with the

following:
The latest draft is based on limits for shaft amplitudes DR
31.5Jlm to 200Jlm peak to peak, or housing velocity VR 31.5mm/s
to 20mm/s r.m.s. with subjective ratings dependent on
transmitted power from the highest 'A' grade-naval applications
to 'D' grade mill drives.
 Discussion on the Paper
VIBRATION ISOLATION

• by
F. Kunz

DISCUSSION unfiltered vibration movements or a range of filtered

measurements may be used with common acceptance
From Mr P. Koller: limits. In general, if narrow band frequency measurements
are made higher overall vibration would be permitted than
May I open by congratulating Mr. Kunz on the preparation if broad band measurements are taken. Would Mr. Kunz
and presentation of what I am sure will be a most useful have any views on this?
reference document for all LR surveyors. And finally, in case anybody doesn't know what the
It seems to me that the paper has probably expanded a Rules say on this complex matter I quote Part 5, Chapter 1,
little from the original intended theme suggested by the Paragraph 5.2.1. "The trials are to demonstrate that any
title. I believe that this was a good opportunity taken, and vibration which may occur within the operating speed
for field surveyors the coverage of vibration measurement range is acceptable".
and the associated acceptance levels will be of particular
interest.
The inclusion of Appendix 'A' I feel is a long overdue From Mr R.V. Pomeroy:
general distribution of this T.I.D. document, and I make no

• apologies for directing my comments to this subject.

All engineers in the field of rotating machinery acquire
experience on vibration limits, and develop a basis for
judging acceptance or non acceptance.
I recall for example being a fresh graduate mechanical
engineer at 'Metro-Vick' s', Trafford Park. In the geared
Mr Kunz has produced another fine paper for the LRTA,
following on from his papers on gearing and crankshafts.
The breadth of these topics clearly demonstrates the wide
field of activity encountered in Technical Investigations and
the order of the papers shows the changing emphasis in the
work of the department. Over the last two decades there
turbine department where I worked a velocity of has been a shift away from experimental stress analysis
4.5mm/sec was considered a normal limit for shop trials towards the measurement of noise and vibration.
with a maximum of 7.0mm/ sec in certain circumstances. The Society has traditionally been associated with the
My first contact with an LR surveyor was at this time assessment of safety and the strength of machinery
during shop trials of a cross compound steam turbine set components. However the paper gives several examples
coupled to its reduction gear. All the carefully prepared where the expertise of TID surveyors has been called on to
vibration measurements were to my surprise brushed aside tackle vibration problems which did not involve integrity
by the surveyor who simply produced a three pence piece but were principally concerned with habitability and health
(a coin no longer in circulation with flats on the and safety.
circumference). This was carefully balanced at strategic The author presents a general introduction to a
locations on the set during the full ahead r.p.m. phase; the complex subject very clearly in his first few pages. In
acceptance criteria being that the balance remained Figure 4 he shows a Campbell Diagram. The form of
undisturbed. Should the piece have fallen on its side the presentation attributed to Campbell shows the practical
32000 shp set obviously stood a good chance of being relationship between speed dependent exciting frequencies
rejected by Lloyd's. I had a nervous few minutes, but the and speed independent responses. It does however take no
test was passed and an unimpressed LR surveyor was taken account of the magnitude of the exciting force or the degree

•
off for his lunch. of damping associated with the responses. Is there a simple
Some years later contact with MOD (N) uncovered a way of including some indication of the likely importance
slightly more advanced approach where a limit for the peak of the various intercepts?
velocity at rotational frequency of 3.8mm/ s was specified. A large part of the paper refers to the use of isolating
More recently working for LR abroad I found that mountings, generally involving rubber. Rubber is a material
VD1 2056 was widely used and accepted. Perhaps Mr. Kunz which has non linear properties which may vary
could explain how this document fits into the reference he considerably between specimens. Is the author aware of any
quotes. problems which have resulted from inconsistency of
Mention should also be made of API 613 & 617 which material property? Due to the non linear load-extension
European manufacturers of centrifugal compressors and relationship the initial set-up is important. Have problems
speed increasing gears were confronted with in the been encountered due to improper control of setting up or
seventies. Proximity probes are required to measure due to changes following initial creep?
unfiltered shaft vibration adjacent to bearings. Limits are There was a period when there was a great deal of
clearly specified and include mechanical and electrical interest in the use of tuned foundations. Steel foundations
runout which should not exceed 6 m. Has Mr. Kunz ever for land-based turbo-generators were used successfully and
experienced difficulty with electrical runout when taking it was suggested that similar arrangements could be
measurements with proximity probes? There is work at employed elsewhere, particularly to replace massive
present within ISO to develop a DIS document 8579-2 concrete foundation blocks. In view of his involvement with
"Determination of mechanical vibration of gear units land-based diesel generators is the author aware of any
during acceptance test". The latest draft is based on limits such application or any interest in pursuing such a
for shaft amplitudes DR 31.5 m to 200 m peak to peak, or solution?
housing velocity VR 3.15 m/s to 20 m/s r.m.s. with For many years anti-vibration mounts were
subjective ratings dependent on transmitted power from the recommended to be installed so that the rubber was always

•
highest 'A' grade-naval applications to 'D' grade mill loaded in compression. Generally this resulted in
drives. A controversy exists however, because either orthogonal mountings and the transverse elements

1
precompressed. As shown in Figure 6 it is now usual to find 10. Good workmanship.
angled mounts where it would seem that the magnitude of However to combat high frequency vibrations and
shearing movement would be increased, even assuming structural borne noise in particular, some auxiliary
that tension could be avoided. Is the author aware of any machines, depending on their location, and
cases where damage in shear of the rubber-to-metal bond resiliently mounted. •
has occurred? Furthermore, it has now become common
Finally I would like to thank the author for a paper practice for the medium speed machinery, i.e.
which distils so much valuable experience. diesel-mechanical and diesel-electric installations,
of cruise ships and luxury ferries to be resiliently
mounted. The chief reason for resorting to elastic
From Mr f. Powell: mounting arrangements is to reduce the noise
levels in the cabins thereby increasing passenger
The author is to be thanked for giving an interesting comfort.
account of vibration theory and the need for elastic With respect to the installations embracing
mounting arrangements. My comments encompass reduction gear boxes, it is claimed that through
preventive measures to tackle problems at source, which appropriate elastic mounting arrangements the
may be of interest. structural-borne noises from the engine can be
The importance of vibration on a ship can hardly be reduced to, or even below the level of that arising
exaggerated, for the presence of vibration even of small from the gear box. Apparently the structural-borne
amplitude may spoil an otherwise excellent design, noise by the gear box is taken as an acceptable
especially if the frequency is high. The forms of vibration limit alleviating the need for the 2-S engines to be
arising from the complex machinery arrangements and resiliently mounted.
their interaction with the hull are diverse, requiring detailed Could the author comment on the above and
analysis and application of sophisticated measuring also the life span and frequency of replacement of
equipment for verification purposes. the elastic mounts under normal service
conditions.
The main areas of concern are: In the section headed "Application of Theory
& Examples" Mr Kunz states that 'Diesel engines •
i) Habitability:
which .... The problem was overcome by a modest
a major factor in commercial success or failure of a
reduction of the internal first order moments by
passenger ship.
balance weights and reinforcements of the sump to
reduce stress levels'. Would the author agree that
ii) Structural failure:
this is clearly a case of lack of communications
fatigue problems usually associated with welds.
between the machinery installation designers and
engine builders or inadvertently a case of inherent
iii) Safe functioning of instruments of all kinds:
structural deficiencies in the bedplate.
interference with the growing number of sophisticated Furthermore, since the firing sequence has not
electronic equipment for automation, microbased
been stated could the author explain whether the
systems, Navigational aids, etc.
reduction in the internal couples did result in an
The foregoing dictate proper consideration of the
increase in the external moments and if so what
vibration aspects of the machinery installation and the
countermeasures were taken.
hull response at the earliest stage of the design process.
Fortunately the principal sources of vibrations are well
known, these being main and auxiliary machinery
From Mr J.T. Stansfeld:
together with the action of the sea and propeller.
Nevertheless providing problems are identified at the
I appreciate that for the purposes of this paper you have
outset, appropriate steps to reduce the possibility of
restricted the discussion to the design of mounts to prevent
vibration can effectively be taken. However this will
transmission of vibration from the machine to its
require close collaboration between naval architects,
surroundings. In your experience, have problems arisen
machinery installation and engine builders.
where low frequency excitation from external sources e.g. •
The following well established and feasible design
an adjacent engine, hull girder modes etc. has been
measures, some listed below, should ensure satisfactory
transmitted through isolation mounts and caused
vibration levels from a projected installation,
unacceptable displacement of the machine? If so is it
alleviating the need for elastic mounting arrangements.
practically possible to design against this?
I believe it is common practice in naval vessels where
1. Correct location of the engine(s).
structure borne noise is critical to have double isolation of
2. Correct choice of engine and propeller speeds.
machinery, i.e. the machine is isolated from a raft which in
3. Correct balancing of main and auxiliary
turn is isolated from the hull. Are you aware of any similar
machinery. installations in the merchant fleet - research ships and
4. Introduction of engine top bracing.
passenger vessels would seem to be candidates?
5. Care in dealing with the axial vibration of the shaft
On page 5 you describe alignment problems arising
system. from creep in the rubber elements in the mounts. After the
6. Correct choice of the shafting system (under I
initial settling in period does the creep rate change and does
overcritical running).
it limit the life of this type of vibration mount. Is there
7. Care in dealing with the lateral shaft vibration and
significant change in mount performance over its life?
shaft whirling.
8. Correct static/ dynamic balancing of the propeller.
9. Minimising the propeller impulses due to wake
variation having a frequency equal to the number
of blades multiplied by rpm and ensuring that the
blade frequency and whirling speed of the tail
shaft stay well apart.

2
•
 Vibration measurements based on shaft run out are
AUTHOR'S REPLY commonly measured with non-contacting proximity probes
which work on an eddy current principle. Mr. Koller
To Mr. P. Koller

•
suggests correctly that irregularities of residual magnetism
in shafts can induce spurious signals. The magmtudes are
Mr. Koller's old style three pence piece used to be well
not easy to separate rigorously from me~hanical ?ut of
known as a vibration checker, possibly more in an anecdotal
roundness of the journal or orbital motwn w1thm the
context than as a qualitative tool, but I can well believe that
bearing as values are typically of the or~er of micromete.rs.
it was used sometimes as a survey instrument. I first came
Thus they are likely to be of concern m connectwn w1th
across it, sat in pride of place, on a Wankel rotary engine
high speed shafts where permissible amplitudes are of the
which ran exceptionally smoothly, if noisly. It passed the
same order of magnitude. The last time the author had to
three pence piece test with flying colours, but this was an
address the problem, it involved the journals of .a 3~0MW
essentially useless bit of test data. The use of three pence
turbine set running at 3000 RPM and the combmatwn of
coins as survey tools is unsound because the v1bratwn
magnetic effect and out of roundness indicated on the
levels which cause the coin to fall over are well above the
turning gear was small relative to the dynamiC shaft orb1ts
modern acceptance levels, as indicated (see.attached graph)
which approached the bearing clearance at synchronous
which shows the result of controlled expenments. Insp1red
by Mr. Koller's remarks, an unworn 1967 three pence piece speed due to the proximity of a whirlin? critical. .
Regarding Mr. Koller's final pomt, 1t 1s considered
was shaken on a vibration table. The results on the f1gure
reasonable to base the level of assessment criteria on narrow
show the response to vertical vibration and, somewhat
band data, because that is how tests to set limits would be
unexpectedly, the response to transverse vibrat~on was
conducted. It follows that the use of a filtered signal allows
similar. After the experiment a plethora of other coms have
a higher acceptance limit than when .using broad band
been suggested as being more applicable, some in earnest
measurements, but this places uncertamty on the s1de of
and some in jest. It must be accepted, unfortu~ately, that th.e
safety.
technique is unsound as a go/no-go dev1ce because 1t
suffers from excessive scatter which prevents proper

•
calibration even if coin wear is disregarded. Surveyors are
To Mr. V Pomeroy:
advised to put their trust in objective, quantifiable
measurements and to insist on evidence of valid calibration.
Mr. Pomeroy's kind remarks about TID are appreciated, but
VDI 2056 is a respected German national standard
I would like to point out that it is not strictly correct to
which defines the acceptability levels of rotating machines
suggest that the emphasis of work has moved awa:y from
in terms of the root mean square value of a complex
full scale dynamic stress measurements. There has mdeed
vibration. It is based on vibration velocity as the
been a reduction of static stress measurements such as
characteristic quantity, as opposed to displacement or
pressure vessel tests, but at the time of writing a number of
acceleration which are used by some, mainly older
large scale stress measurement projects on dyr:amically
standards which attempt to link the onset of problems
loaded structures are in progress and are provmg to be
empirically to vibrations measured at the bearing housings.
challenging and interesting tasks. The Society can put a
The general VDI approach has been incorporated into the
considerable number of simultaneous channels of data
more recent ISO document 3945:1977, which is listed in the
acquisition in the field and has the computing resources
paper.
and staff capable of handling the vast quantities of data
which are gathered within a short time dunn!? ?ynamlc
structural tests. A recent case involved the acqms1t1on and
processing of 250 Mega bytes of information. . .
70
F Turning to the subject matter of the paper, 1t 1s agreed
that the Campbell diagram lacks indication of the
importance of the intercepts in terms of likely magnitu?e of
Vibration level to cause response. Vibration amplitudes depend. on .the prox1m1ty to
60
\ three pence coin to fall resonance, the magnitude of the exCltatwn, mertla and

\ damping in the system and has to be evaluated fr?m these

• (f)
2
a: 50
(jJ
f'
_s
Q)
b
\.\
variables. Some types of excitation are strongly directional
and at some risk can be ignored for particular vibration
modes. For example simple vertical motion of the
foundation is unlikely to induce the rotational vibration
modes, but this would not be the case for axially distributed
mass imbalance which could give rise to an external couple.
u
~40 If a greater depth of insight is required than given by simple
Q_ 0
E inspection, various calculation methods are avallable to
m quantify the response as outlined in the paper and th1s
c
0
0'\ 0
would normally be done for important machines.
~30

"·"·
..0 Rubber mounts in the 45 degree configuration shown
:> in the paper have been used by MAN for the 58/64 medium
speed engines fitted on the Cunard and P&O passenger
ships and to the best of the author's knowledge have g1ven
20 ........... 0 no trouble.
'0 .............. The mounts are arranged where they are unlikely to be
exposed to oil and have protective plastic sheet covers. The
0...........~. Society's data base information specifically covers flexible
10 mounts for the last four years and indicates only two cases
where repairs were required.
t- _ _ _ ISO 3945 :.Jjl~"Sat1sfact<2:L _ _ _ _

• 10 30 50 70 90
Frequency (Hz)
110 130 150

3
To Mr. f. Powell: vibration modes on a resiliently mounted machine under
investigation. This is desirable to help in calculations when
Mr. Powell's interesting contribution lists the major means changes are required.
of alleviating the need for elastic mounting arrangements. TID has been involved fairly recently with two cascs on
In merchant shipping, as opposed to war ships, the choice different ships where resiliently mounted generating sets •
between elastic mounting of machines and traditional vibrated severely due to moderate vibrations at propelll'r
chocking should be looked at on the basis of economics, in blade passing frequency of the deck on which they were
terms of building and long-term maintenance costs and the mounted. In both cases the vertical translational vibration
effect of noise and vibration on the machines themselves, on mode was involved and the stiffness of the rubber mounts
the goods carried and the crew. On this basis it will be implied from the measurements was substantially higher
found that there are clear cases where solidly mounted than that assumed in the design calculations.
medium speed Diesel engines, to cite but one of the items Separation of the natural frequency from the excitation
mentioned by Mr. Powell, will restrict the sensible choice of frequency by changing the stiffness of the mounting system
the layout of public spaces and passenger cabins in large is the solution in such cases and reverting to Mr. Stanfeld's
cruise ships and that there will be a significant risk of noise point, external excitation should be taken into account at
and vibration complaints in certain areas. It is agreed that the design stage. To do so effectively requires reliable
solidly mounted engines are giving good service in large design information, in particular regarding effective
ferries where passenger spaces are geographically more dynamic mount stiffness.
remote from the machinery by virtue of car decks and the Attenuation of structureborne machinery noise at
type of service is less demanding with respect to noise and source by double isolation of the machines is a complex and
vibration. The need to use expert advice and analysis at the costly solution and the author has no direct experience on
design stage cannot be overstressed. the ships visited in the course of TID investigations. In
Regarding Mr. Powell's request for more information naval application the objective is likely to be to minimise
on the effect of fitting balance weights, the firing order of the noise radiated from the hull to the water rather than
the engine described in the paper was such that external structure borne noise control within the ship and this
moments were inherently balanced in the initial condition, constraint does not apply to normal merchant vessels. On
which is of greater importance than internal moments when passenger vessels the current practice is overwhelmingly to •
considering vibrations of the engines as a rigid body on the use one set of mounts under machines and this controls
resilient mounts. It is not uncommon to strike compromises structure borne noise adequately in most cases. Small high
in engineering, not least in the dynamic balance powered passenger vessels such as high speed catamarans
characteristics of engines, but in the case in question the intended for long sea passages may call for more complex
characteristics of the resilient mounts were such that there systems, but the need for this would have to be studied on a
was no significant change of external engine vibration as a case by case basis.
result of introducing additional balance weights at service Mr. Stansfeld questions alignment and creep. Unlike
speed, or need to introduce counter measures. creep of steel at elevated temperatures, creep of rubber
With the benefit of hindsight it can be seen that the mounts diminishes after a few months in service to
bedplate and entablature of a 9-cylinder in-line engine are stabilise, and with reasonable loading it does not limit the
likely to deform more under internal moments when the life of the mount. It is normal to make a settling allowance
engine is resiliently mounted than on solid chocks. for this, and this must be based on the mount
In principle this could have been taken into account in manufacturer's experience. Exposure to oil is detrimental to
the design of the bedplate and engine entablature, but the some types of rubber, for example nitrile rubber swells in
depth of analysis is considerable and such calculations are petroleum products. Thus protective sheets placed over
often more successful in explaining unexpected factors than mounts to keep them clean must be kept in good order. If
preventing them. there is a doubt on the mount performance the most
effective checks, apart from visual inspection to check for
signs of tears and cracks, would be to compare free length
To Mr. f. Stansfeld: with the specification and to check the hardness.

•
Vibration of resiliently mounted machines can be, and
sometimes is, excited by the foundations and the
mathematics indicate that the same natural frequencies and
mode shapes are involved. The physical manifestation is
that a flexibility mounted machine can vibrate even when
stopped and the movement below the mounts is
dynamically magnified if the excitation frequency coincides
with a natural frequency of the machine on its mounts. It
may appear to be stating the obvious that to recognise this
symptom, which requires no measurement at all, is an
important first step in diagnosing the problem and finding
a cure. Some investigators lose sight of the simple fact that
to measure on a stopped engine may provide the
information necessary to define a problem.
In merchant ship applications, only the higher hull
girder vibration modes or after end structural modes can
excite Diesel generating sets on flexible supports if they are
tuned to run super-critical. With rubber mounts it is
difficult to reach natural frequencies much below 8 Hz and
this filters out the first few hull girder modes and first and
second order frequency main engine excitation by slow
speed engines. Investigations on ships are greatly eased by
the fact that in most cases the excitation frequency can be

•
varied and generally, using a modicum of resourcefulness, it
should be possible to identify more than one of the