Reactions of flexible bearings

and vibrations. In collaboration

with

Laborelec
 References
These are a few textbooks and papers dealing with flexible couplings.
Calistrat M.M.: Flexible Couplings Their design, Selection and Use, Caroline Publishing,
Houston, 1994.
Mancuso J.R.: Couplings and Joints, Design, Selection and Applications, Marcel Dekker
New Yprk, Basel, 1986.

South D.W. and J.R.Mancuso: Mechanical Power Transmission Components,Marcel

Dekker, New York, Basel, 1994
Neale M., P. Needham and R. Horrell "Couplings and Shaft Alignment" Mechanical
Engineering Publication, London,
Piotrowski, J: Shaft Alignment Handbook, Marcel Dekker, New York, Basel, 1986.
Godin, A, Esco Transmission. ( very helpful, handed out his references)
Tel 02 720 48 80 Fax 02 721 28 27

Randiranarivo, EDF. (mentioned that EDF switched from gear to disk couplings)
Tel 00 33 1 30 87 79
Watremez JM et Faure, Compagnie Messiaen et Durand, (now retired)
Tel 00 33 27 73 53 03 Fax 00 33 27 78 36 09
Heyer,R. and W.Mllers : RckstellKrfte und -momente nachgiebiger Kuppplungen bei
Wellen-verlagerungen, Antriebstechnik 26 (1987) Nr.5 ( Research in THU Bochum).

Yanpolskii et al: Formation of disturbing forces in toothed clutches Russian

Engineering Journal , volume LV No 10, translated from Vestnik Mashinostroeniya, vol 55,
issue 10, 1975, pp.12-15.
Acknowledgments: to all the ULB students (JL Nyst, S. Lchaud, etc.), to P. Ripak for
software support, to the ULB SMA workshop (too many people to thank) and to
representatives of coupling manufacturers like Ir Axel Godin of Esco Transmissions, not to
mention the patience of my Laborelec colleagues.
 Why worry about flexible couplings? (partly adapted from Mancuso)
Flexible couplings are a vital part of a mechanical power transmission system. Unfortunately,
many system designers treat flexible couplings as if they were a piece of hardware.
The amount of time spent in selecting a coupling and determining how it interacts with a system
should be a function not only of the cost of the equipment but also how much downtime it will
take to replace the coupling or repair a failure. In some cases the process may take only a little
time and be based on past experience; however, on sophisticated systems this may take complex
calculations, computer modeling, and possibly even testing. A system designer or coupling user
just cannot put any flexible coupling into a system with the hope that it will work. It is the
system designer or the user's responsibility to select a coupling that will be compatible with the
system.
Flexible coupling manufacturers are not usually system designers; rather, they size, design, and
manufacture couplings to fulfil the requirements supplied to them. Some coupling
manufacturers can offer some assistance based on past exposure and experience. They are the
experts in the design of flexible couplings, and they use their know-how in design, materials, and
manufacturing to supply a standard coupling or a custom-built one if requirements so dictate. A
flexible coupling is usually the least expensive major component of a rotating system. It usually
costs less than 1%, and probably never exceeds 10%, of the total system cost. And yet it can
cause severe financial losses when nor well selected and operated
Flexible couplings are called upon to accommodate for the calculated loads and forces
imposed on them by the equipment and their support structures. Some of these are motions due
to thermal growth deflection of structures due to temperature, torque-load variations, and many
other conditions. In operation, flexible couplings are sometimes called on to accommodate for
the unexpected. Sometimes these may be the forgotten conditions or the conditions unleashed
when the coupling interacts with the system, and this may precipitate coupling or system
problems. It is the coupling selector's responsibility to review the interaction of the coupling
with that system. When the coupling's interaction with the system is forgotten in the selection
process, coupling life could be shortened during operation or a costly failure of coupling or
equipment may occur.

In this research, one not only deals with static reactions developed by flexible couplings
(these have been known for a while) but also with the time-varying reactions causing
vibrations. This is much less known and required special rigs and measurement setups.
This approach identifies which amongst the most popular flexible bearings are the least
prone to generate vibrations.

What are flexible couplings (partly borrowed from Mancuso)

Historically, rotating equipment was first connected by means of rigid flanges Experience
indicated that this method did not accommodate the motions and excursions experienced by the
equipment. One then tried to thin these flanges and allow them to flex. Rigid couplings are used
to connect equipment that experiences very small shaft excursions or with shafts made long and
slender enough so that they can accept the forces and moments produced from the flexing
flanges and shafts (think of STEGs in Electrabel between generator and compressor).
The Three functions of a flexible couplings
Fair enough. They must transmit power. The
maximum rotational speed at which they
operate is important when selecting a flexible
coupling.

Subsequent motion of bearings and/or

foundations may cause misalignment. The
latter should remain within the limits
prescribed by manufacturers for the selected
coupling.
This is the basic function of flexible vs. rigid
couplings.
Very important when thrust bearings come into
conflict with rigid couplings.
Germans consider this as axial misalignment.
ULB test rigs have not focused on "axial
misalignment".
 Efficiency in Power Transmission
Flexible couplings must couple two pieces of rotating equipment, equipment with shafts, flanges,
or both. These interface connections are numerous.
Flexible couplings must also transmit power efficiently. Usually, the power lost by a flexible
coupling is small, although some couplings are more efficient than others. Power is lost in
friction heat from the sliding and rolling of flexing parts and at high speed, windage and
frictional losses indirectly cause lost efficiency.
Most flexible couplings are better than 99% efficient.

How flexible couplings accommodate Misalignment.

Flexible couplings must accommodate three types of misalignment:
Parallel offset: Axes of connected shafts are
parallel but not in the same straight line. One
also calls it radial misalignment later in the
text.

Angular: Axes of shafts intersect at the center

point of the coupling, but not in the same
straight line.

Combined angular and offset (radial): Axes

of shafts do not intersect at the center point of
the coupling and are not parallel.
 Flexible couplings in operation
Most flexible couplings are designed to accommodate axial movement of equipment or shaft
ends. In some cases (e.g., motors) couplings are required to limit axial float of the equipment
shaft to prevent internal rubbing of a rotating part within its case.
Accommodation of misalignment and end movement must be done without inducing
abnormal loads in the connecting equipment. Generally, machines are set up at installation quite
accurately. There are many things that force equipment to run out of alignment. The thermal
effects of handling hot and cold fluids cause some movement in the vertical and axial direction,
together with differentials of temperature in driver media such as gas and steam. The vertical
motions could be a result of support structure expansions due to temperature differences,
distortion due to solar heating, axial growth or a combination of these. Horizontal motions are
usually caused by piping forces or other structural movements, temperature differentials caused
by poor installation practices, and expansions or contractions caused by changes in temperature
or pressure differentials of the media in the system.
It is a fact of life that machinery appears to live and breathe, move, grow, and change form
and position; this is the reason for using flexible couplings. A flexible coupling is not the
solution to all movement problems that can or could exist in a sloppy system. Using a flexible
coupling in the hope that it will compensate for any and all motion is naive. Flexible couplings
have their limitations. The equipment or system designer must make calculations that will give a
reasonable estimate of the outer boundaries of the anticipated gyrations. Unless those boundaries
are defined, the equipment and system designer may just be transferring equipment failure into
coupling failure.

What is known about flexible coupling reactions

One thing to remember is that when subjected to misalignment and torque, all couplings react
on connected equipment components. Some produce greater reactionary forces than others and
if overlooked, can cause shaft failures, bearing failures, and other failures of equipment
components. They may also excite vibrations with time-varying reactions. This last point has
received less attention in the technical literature.

Rigid couplings produce the greatest reactions.

1. Mechanically flexible couplings such as gear, chain, and grid couplings produce high to
moderate moments and forces on equipment that are a function of torque and
misalignment.
2. Elastomeric couplings produce moderate to low moments and forces that are slightly
dependent on torque.

3. Metallic membrane couplings produce relatively low moments and forces which are
relatively independent of torque.
The most commonly used flexible couplings today are those that produce the greatest flexibility
(misalignment and axial capacity) while producing the lowest external loads on equipment.

In the present research, one compares flexible couplings of the first three types (gear,
elastomeric and disk) in terms of
1. Static reactions in presence of misalignment and torque: these reactions reflect on
the loads of mainly the nearest bearings. The present research also investigates the
relationship between the directions of static reaction and misalignment.

2. Dynamic reactions in presence of misalignment and torque: these reactions cause

vibrations. These reactions have never been precisely stated in the technical
literature.
 Types and inner workings of couplings
There are many types of couplings. They can virtually all be put into two classes, two
disciplines, and four categories. The two classes of couplings are:
1. The rigid coupling
2. The flexible coupling
The two disciplines for the application of flexible couplings are:
(a) The miniature discipline, which covers couplings used for office machines,
servomechanisms, instrumentation, light machinery, and so on).
(b) The industrial discipline, which covers couplings used In the steel industry, the
petrochemical industry, utilities, off-road vehicles, heavy machinery, and so on. Therefore,
one does not focus upon universal joints.
In this research we deal exclusively with industrial couplings.

The four basic categories of flexible industrial couplings are:

1. Mechanically flexible couplings
2. Elastomeric couplings
3. Metallic membrane couplings
4. Miscellaneous couplings
The general operating principles of the four basic categories of industrial couplings are as
follows:

1. Mechanically flexible couplings: In general, these couplings obtain their flexibility from
loose-fitting parts and/or rolling or sliding of mating parts. Therefore, they usually require
lubrication unless one moving part is made of a material that supplies its own lubrication
needs (e.g., a nylon gear coupling). Also included in this category are couplings that use a
combination of loose-fitting parts and/or rolling or sliding, with some flexure of material.

2. Elastomeric couplings: In general, these couplings obtain their flexibility from stretching
or compressing a resilient material (rubber, plastic, etc.). Some sliding or rolling may take
place, but it is usually minimal.

3. Metallic membrane couplings: In general, the flexibility of these couplings is obtained from
the flexing of thin metallic disks or diaphragms.
4. Miscellaneous couplings: These couplings obtain their flexibility from a combination of the
mechanisms described above or through a unique mechanism like spring couplings
 Gallery of flexible couplings
Yellowed boxes indicate which coupling types were tested in the present research.
Miniature couplings
As a refresher, since they are not part of the research.

Well-known in such applications as Elastomeric like used in the original Bently

incremental encoders, etc. Nevada rotor kit. Provides some damping.

Mechanically flexible couplings

Gear (TAK, Escogear, Flexident, etc.) Grid coupling ("Citron", Flexacier, etc.)

Same remarks as for gear, except that

Less rigid in torsion
Piloting less crucial
Grid segments may excite higher
harmonics (4 1/4 grids excite 100 Hz at
1500 rpm Doel III) tested in research
Severe misalignment can cause grid to cut
through cover.
Very popular in high-power applications
because of compact design. Semi-flexible for
high torques. Lubrication is an issue. Piloting
may cause unbalances.
Extensive tests in research
Chain couplings Universal joint (automotive, etc..)

Known to accommodate high misalignments

like in automotive applications. May cause
No comment. Pins traversing both chains vibrations.
ensure torque transmission.

Elastomeric couplings
Beware of overheating due to material losses through hysteresis. Beware of chemical agents and
material ageing. Lesser ratings.

Blocks (jaws) Tire type.

Tested in research. Static reactions between

disk and gear couplings. Less rigid in torsion.
Less rigid in torsion.

Metallic membrane couplings

Disc More compact (ESCODISK tested in lab)

No lubrication, no piloting causing unbalances.

Hub anchors hexagonal flexible disk in three Lab tests show no dynamic reactions causing
points while the other three interlaced points vibrations even at the third of sixth harmonic
are anchored on spacer. Ditto on the other side of the rotational speed despite quasi hexagonal
of the spacer. Long spacer here to shape of flexible disks. For the same rating less
accommodate higher misalignment grades to compact than gear couplings. Misalignment
avoid excessive disk flexing. Length of pacer is can be checked from the shape of flexed disks.
no cure to axial misalignment.
Tested in research.
Diaphragm couplings

More bending resistance than in disks. Theoretically would not cause higher harmonic excitation
than disks since the flexible elements are not anchored in a limited number of points.
Other couplings
Quite a variety.

Spring couplings Pin and bush

Quite flexible in torsion.

The elastomer sleeve takes over the role of
disks and diaphragms without comparable
torsional rigidity.
Slider block couplings Spiral spring coupling
 How flexible are gear couplings?
Ideally one would like such couplings to transmit torque and only torque from one shaft
end to another without adding either shear reactions or bending moments. These reactions
can be either static or dynamic.
This research provides some insight how flexible couplings contribute to vibrations and
bearing loads. It focuses on gear couplings (type TAK), but also examines competitors like
disk and elastomeric couplings.
Static reactions
Static reactions of flexible couplings may over- or unload nearby bearings whose selection
is often based on the weight of the rotor they support. It may happen that the static
reaction of a flexible coupling nears the value of the rotor weight!
Never think that gear couplings do not react on the shaft lines they connect! This is
illustrated below for increasing torques and a large misalignment
You may align shaft ends properly at first.
Due to foundation sagging, it may happen
that one gets a rather severe radial
From experimental rig. misalignment. You can measure it at light
loads via such standard methods as the
reverse dial or with more advanced methods
like OPTALIGN.
Trouble is when the torque transmitted by
the gear coupling increases to the coupling
nominal rating. See what happens! At high
values of torque, the gear coupling behaves
like a rigid coupling. It bends the shaft ends
From experimental rig. to bring them in line and causes extra
bearing loads. Is that the kind of flexible
coupling one generally thinks of?
Dynamic reactions
Dynamic reactions from flexible couplings cause vibrations. The most common cause for
vibrations originates from rotor unbalances. Balancing norms specify the amount of
residual unbalance that is admissible on rotors. These depend on specific balancing grades
for different types of applications, the rotor weight and maximum rotational speed.
To a given admissible residual unbalance corresponds a centrifugal force. The dynamic
reactions of the couplings become a nuisance when they exceed the contribution of these
centrifugal forces.

This is shown to be the case with gear couplings, much less with other types of couplings
like disc and elastomeric one.
 Test rig for flexible couplings

Flexible couplings like gear and grid ones that rely on sliding surfaces to accommodate
misalignment exhibit torque-dependent reactions. Therefore, the rig must be able to
produce torques in the range of the coupling nominal values.
At high speeds this corresponds to huge power transfers. The power delivered by the
coupling is indeed the product of the torque transmitted and the rotational speed in
rad/sec. Assume one wants to transmit a 1600 NXM torque at 1500 rpm. You need ca 250
KW motors. The rig is then costly.

At lower speeds, one can recover huge torques from faster-rotating motors through
reductors. In the test rig above, back-to-back 1.1 KW electrical motors rotating at roughly
1500 rpm with 1.1 kW power ratings and connected to the input shafts of 260:1 reductors
can produce 1800 NXM at the level of the gear coupling being tested. At the coupling level,
the rotational speed is approximately 5.7 rpm. It can vary around this value depending on
the slip of the induction motor that depends on the power it delivers or retrieves.
Such an approach has many advantages and one drawback.
The advantages are:
lower electrical powers to handle and thus reduced expenses.
one can afford embarking lots of sensors like strain gauges on shaft ends and
proximity sensors playing the role of electronic clocks. Slip ring contacts can power
these sensors and retrieve their signals without expensive multi-channel telemetry
systems.
measuring strains on shaft ends unambiguously yields the coupling reactions.
proximity sensors unambiguously measure on-line misalignments of the shaft ends
 near the coupling, both its static component and its variations.
one need not bother about reactions caused by high centrifugal forces and thus
focus on the reactions of the coupling alone.
The drawback is the oil films may not get renewed as well as at higher speeds. This
concerns only flexible couplings relying on lubricants. This could somewhat influence
coupling reactions via somewhat higher friction coefficients between teeth. To this aim,
tests were conducted on the low-speed test rig with EHP greases, which basically yielded
the same reactions.

Measuring coupling misalignments and reactions in real-time on the

rig.

Top left: Measurement setup to get misalignment from embarked eddy probes acting like
electronic clocks in reverse dial method.
Top right: embarked strain gauges with conditioning to measure bending moment along
the shaft end and finally coupling reactions.

All embarked sensors (5 strain gauge bridge amplifiers and 4 eddy probes are powered via
the slip rings. The same slip rings transmit the signals back to a National AT MIO E-10
data acquisition card. A phase reference one top/rev synchronizes the measurements on the
card to track the angular position of the shaft at all instants, assuming (true here) that the
speed does not vary during a revolution.
Bottom: the type of analysis of signals typically used in this report. More on this below.

Types of analyses
Frequency analysis of coupling reactions: shear and bending moment and ditto for
misalignment: radial and angular, as shown in the diagrams below. How to interpret them
them?
The abscissa is graduated from -5 to 5. When a bar-graph shows a big values at, say +2,
for the shear (radial) reaction of the coupling, then this reaction (and resulting bearing
load) is important and rotate at twice the rotational frequency in the same direction as the
shaft revolutions. If this happens at -2, it tells that it rotates counter the shaft at twice its
frequency. At zero abscissa, one gets the static reactions and misalignments.
Spectra are obtained for two different values of the torque transmitted by the gear
coupling. These are shown in these so-called double-sided (full in Bently Nevada
terminology) frequency spectra for two different values of the torques
At full torque, the dc (static) contribution amounts to 416 KgF or 4160 N.
There is also a significant 35 KgF or 350 N contribution at the harmonic +2 of the
rotational frequency. This corresponds to a force rotating in the same direction as the shaft
at double the speed thereof. Testing another gear coupling with mixes of torques and
misalignments confirms this trend. Based on two types of gear couplings (FEXIDENT and
ESCOGEAR) one can thus claim there is very little hope that radial vibrations caused by
these reactions could indicate in which directions the misalignment occurs.
Results can be compared below where the reaction is somewhat higher due to an initially
larger radial misalignment at no load.

The initial coupling misalignment is rather

severe: ca 75/100 mm radial and small
angular.

At small torque values (reverse -410 N X

M), one retrieves these values with a small
variation at twice the shaft rev in the
direction of shaft revs.
The static radial reaction is 158 daN and 10
daN at H+2. They would be almost zero at
zero torque.
 At full torque, the dc (static) contribution
amounts to 416 KgF or 4160 N! Not pleasant
for bearings
At the same time, the radial misalignment at
the coupling level almost vanishes. Shaft
ends bend as monitored by strain gauges.
The coupling has become rigid!!
There is also a significant 35 KgF or 350 N
contribution at the harmonic +2 of the
rotational frequency. This corresponds to a
force rotating in the same direction as the
shaft at double the speed thereof. All
subsequent tests confirm this. There thus
little hope that radial vibrations caused by
these reactions would indicate in which
direction a misalignment occurs.

One has performed these analyses for all types of misalignments and torques on two
different gear couplings: FLEXIDENT and ESCOGEAR with basically the same results
and the same conclusion as above: vibrations is of no help to locate the direction of the
misalignment.

Before moving to the summary of these results, here is again another virtual instrument to
understand the situation. It features special data processing techniques involving full FFT
of strain orbits obtained from embarked strain gauges as well as misalignment orbits
reconstructed from embarked electronic clocks (eddy probes).

The Theorist's corner (not to display on the web page)

For data processing buffs, this kind of analysis is rather elegant. In two different cross-
sections of an overhung shaft end next to the coupling, one installs pairs of embarked
strain gauges measuring bending in two perpendicular directions thus rotating with the
shaft. At time zero these directions coincide with the horizontal and vertical respectively.
One performs the sampling at equal rotation interval over a shaft rev or a power of 2
number of revs. Then one combines the bridges responses as "strain orbits" to which one
applies double-sided Fourier Transform. Quite a twisted way to use these! Then one
applies the conventional material strength equations of beams (overhung shaft ends are
none others than rotating beams) to these double-sided Fourier Transforms in order to
retrieve the coupling reactions spatially and in the frequency domain. This requires a
simple coordinate transform from a rotating to a fixed coordinates in space.
In the same vein, one can decompose local misalignment spatially and in the frequency
domain by carefully examining the kinematics of the reverse dial method and what it
corresponds to in terms of the clock readings mimicked by the embarked eddy probes..
There is a report with all equations available that mix motion, material strength and spatial-
frequency analyses.
 Measured gear coupling reactions and local misalignments
The following virtual instruments visualize in real-time what happens when applying an
increasing torque applied to a misaligned gear coupling.

Small torque

Virtual instrument describing the relative positions of the shaft ends (in red Rx and Ry
radial
misalignemnt and Thetax and y for angular misalignment),
the shear coupling reactions in daN (Tx and Ty)
and the bending moment reactions (Mx and My) in daNXM.
The torque transmitted is -480 NXM.
The misalignment is almost the same as with zero torque.
 Large torque.

Same as above except that the torque is at the nominal rating of the coupling
(1800 NXM) and the reactions are way up whereas the local misalignment of the
shaft ends near the coupling approaches zero.
As a result, the coupling behaves like a rigid one. The shaft bends (only tips of the shaft
ends are shown in the above vi.
Gear coupling reactions and gear coupling misalignments vs. torque.
The following diagrams were obtained on the test rig for gear couplings. Others follow for disc
couplings and couplings with elastomer blocks.

Reactions vs. torques

Increasing the torque causes the gear coupling radial reaction to rise substantially up to
2000 N for an initial 38/100 mm radial misalignment. This trend holds on up to 900 NXM.
From there on, the reactions level off because the local coupling misalignment is almost
zero (cf. below). At high values of torques, the gear coupling behaves like a rigid one
bending the shaft ends and loading bearings.

One notices that the coupling reaction exhibits a component perpendicular to the original
misalignment. This is due to friction. The component increases with torque, then levels off
at ca 415 NXM and finally stabilizes at 250 N when the coupling gets self-aligned (see
below
 Coupling misalignment vs. torque

The initial misalignment at zero torque is 37/100 mm radial horizontal. The final coupling
local misalignment hovers around -3/100 mm at high values of torque. Since bearings do
not move, this means that overhung shaft ends bend. In between, some radial vertical
misalignment builds up. It is due to the friction forces developing between the coupling
teeth. A good model for gear coupling can be found in "Couplings and Shaft Alignment"
by M. Neale, P. Needham and R. Horrell , Mechanical Engineering Publication, London.
Based on this model, an average friction coefficient between (crowned) tooth profiles is
around 0.12. This can be obtained by examining the reactions at the crossing point of the
clearance curve vs. torque at zero horizontal clearance.
Note that Neales's model tells that angular misalignments cause little gear coupling
reaction.
 Dynamic (variable) reactions of a gear coupling and vibrations
If flexible couplings produce variable reactions, they also produce vibrations. It was
interesting to evaluate how bad these can be compared to acceptable unbalances as
prescribes by ISO norm 1941B for rigid rotors. Flexible couplings are usually associated
with such rotors and often with roller-element bearings. So we tried to figure out what
would happen when an induction motor delivers a torque through the same coupling as in
the rig.
Let the rotor of a 250 KW 1500 rpm induction motor delivers its torque through the gear
coupling. The torque is 1600 NXM. Such rotors have rotors weighing
5000N.approximately weighs 5000 N. It typically loads its two bearings evenly, i.e. 2500N.
With an initial 37/10mm radial vertical misalignment, the radial reaction of the coupling
reaches 2000N and loads the nearest bearing even more. As a result the bearing may be
totally unloaded. The coupling reaction literally lifts the rotor in its bearings.. For larger
initial misalignments it gets worse. This is reported in numerous textbooks. Bearing loads
usually affect vibrations, especially when these are thick-film sleeve bearings.
What has is lesser known is how the coupling reactions affect vibrations.

Excitation from residual unbalances according to ISO 1941B vs. gear

coupling variable reactions

The above chart was generated from ISO1941B grade 6.3 balancing norm for rigid rotors.
How to interpret this? It plots the rotor weights vs. the maximum operational speed. Level
lines are labeled in centrifugal forces derived from the ISO admissible unbalances (in
daN).
Select 500 Kg for the mass of the induction motor and 1500 rpm as max speed. The
admissible centrifugal force is 230 N and splits on both bearings. In most reasonable
machine construction, this will never cause excessive vibrations which will remain well
within ISO2372 vibration norms. Vibrations caused by that unbalance will be at 25 Hz.
Now compare the forces with the levels reached with the test rig for the dynamic reactions
of a misaligned gear coupling. At twice the rotating frequency, one obtained up to 350 N.
This force loads the neighboring bearing even more.
The score is:

Coupling reaction: More then 350 N on the bearing closest the coupling at 50 Hz

Residual unbalance: 115 N on the same bearing at 25 Hz

Winner: the coupling reactions! They may cause real vibration problems. This was never
reported elsewhere.

Gear Coupling Piloting and unbalances

The following remarks come from Mancuso [] and could not reproduced on a slow-speed rig.

Piloting is defined as the method used to ensure that two connected components rotate
around the same axis. The simplest example is mounting a disk with a hole in the center on
a shaft. The interface between the hole (bore) and the shaft is the pilot of the assembly. If
the bore is not concentric with the disk's outside diameter, then the disk will not rotate
around its axis of gravity (center of gravity, or CG), and an unbalance is created.
Therefore, a good pilot ensures that the two connected components rotate around a
centerline which is at the same time the axis of gravity.

For machinery users it seems obvious that rotors must be balanced after disassembly and
subsequent re-assembly, particularly because rotors are seldom disassembled, and when
they are, the labor involved represents a major expense. The same machinery users, on the
other hand, expect flexible couplings to remain in balance even after they are removed
from machines, and then reinstalled. This is why piloting of coupling components requires
special attention, especially with gear and grid couplings.
Gear couplings have a "floating assembly" which can move axially over the components
which are attached to the shafts: the "spacer" or 'hub". This floating assembly must be
free to move axially, but at the same time should have no radial play over the shaft-
mounted components.
Loose supports are a particular case that applies to gear-type couplings. The figure below
shows the piloting area between the hub and its sleeve. A clearance always exists at this
pilot because:
The heat transferred to the hub from the shaft, particularly in turbine applications,
causes the hub to expand more than the sleeves (which benefit from outside cooling).
Manufacturers provide a clearance at the pilot, to avoid the locking of the coupling, as
caused by the difference in thermal expansion.
Centrifugal acceleration has an effect on sleeves than on hubs, particularly because
hubs are pressed on their shafts, and the two as a unitized body. Because of larger
 growth of the sleeve diameter, the clearance in the pilot area increases as a function of
speed.
The wear at the hub tooth tip creates a clearance at the pilot.

Even with a clearance at the hub to sleeve pilot, gear

couplings still operate smoothly at high speeds because of
the piloting effect created by torque being transmitted
through the involute profile. However, the involute
profile pilots the sleeve to the hub only if torque is
transmitted from the moment the coupling starts to
rotate (which is usually the case). If torque is applied
after operating speed is achieved, gear couplings can
cause severe vibration in the system. Formulae can be
used to determine the torque required for the involute
profile to re-center the coupling. Gear couplings should
not be used in applications where machines operate
without torque.
 Gear couplings and lubrication.
Gear coupling must be properly lubricated. One must avoid sludge formation. When lubricated
with grease at high rotational frequencies, the lubricant trapped in the grease substrate may be
centrifuged to the spacer rim and no longer lubricates gear profiles.
For a proper lubrication, gear couplings welcome some misalignment to ensure some relative
motion between gear profiles and, therefore, some renewal of the oil film. The following pictures
tell the whole story and was highly appreciated in March 98 by the management of Burgo ( a
SCOPE client) when explaining why a gear coupling had exploded in their plant.

Why do gear coupling need some misalignment for a proper

lubrication?

Lubrication Conditions
The lubrication conditions of a coupling vary from application to application. The
correct lubricant for a particular application can be selected only if the particular
conditions of that application are understood.
Lubrication conditions are influenced by:
Motions inside a coupling
Operating misalignment
Rotating speed
Viscosity of the lubricant
Note: Gear type couplings were used in the following discussion, as gear couplings are typical
of other lubricated couplings.

Motions inside a coupling

Couplings that accommodate misalignment through sliding need lubrication to prevent the
wear of the sliding surfaces. The lubrication conditions in couplings are unique and need
clarifications. The following figure illustrates how a misaligned gear coupling operates.
The hub teeth are placed in the plane B-B, which is at an angle to the sleeve teeth, plane A-A.
This angle is the angle of misalignment. Because of the angle between the two planes, the hub
tooth shown is displaced to the left by a distance +L. After half a revolution the same hub tooth
moves to the right of the plane B-B by the same distance, which is now -L. Hence, each hub
tooth slides the distance 2L in half a revolution and returns to the original position after one
revolution is completed. The motion of a hub tooth is oscillatory; the oscillation's frequency is
equal to the operating speed, and the amplitude is 2L. In addition to the sliding motion
described, hub teeth also have a rocking motion.

Influence of misalignment
Lubricant can penetrate between teeth as it follows the oscillatory motion, and in particular
during the time the hub teeth are tilted in respect to the sleeve teeth. Therefore, in order to be
lubricated the coupling should be somewhat misaligned. The need for lubrication increases as
the misalignment becomes larger; fortunately, as the misalignment increases so does the amount
that the hub teeth are lifted off the sleeve teeth. Hence, lubrication conditions improve as the
misalignment becomes larger.

From Neale et al
For gear couplings, the coupling alignment capability limits are as explained in Section 3.2.2.
The allowable angular displacement is inversely proportional to speed, and is limited by tooth
wear; axial displacement is limited by the length of the teeth. A typical minimum axial
displacement is 6 mm, as required in API Standard 67 1, although coupling axial capability need
never be a constraint when using gear couplings. In contrast to diaphragm and membrane
couplings, the ability of gear couplings to cope with angular displacement is quite independent of
their ability to cope with axial displacement. The necessary minimum angular displacement of
0.00075 radians, to maintain adequate lubrication, seems worth reiterating here. There may be
variations on the same theme depending on gear profile crowning and the lubricant and the
rotational speed.

Influence of speed
Another factor that improves the lubrication conditions is the centrifugal acceleration. The
time that the hub tooth is lifted off the sleeve tooth, i.e.,. the time available for the lubricant to
penetrate between the teeth becomes shorter. as the speed increases. On the other hand, the
centrifugal acceleration increases with the square of the speed, hence the force that pushes the
lubricant between the teeth increases rapidly with speed. The lubrication conditions, as related to
wear, improve as the speed increases; this fact was demonstrated by actual tests.
Couplings that operate at very low speeds subject their lubricant to low centrifugal acceleration,
therefore have poor lubrication conditions. If a rather consistent grease would be used at low
speeds, couplings would wear-out rapidly. Actual tests have shown that lubrication conditions
become very poor whenever the centrifugal acceleration is lower than 10 Gs. To compensate for
low centrifugal acceleration, slow-turning couplings must be lubricated with progressively
thinner lubricants.

In the test performed in the ULB, some poor lubrication causes an important reaction of the
coupling perpendicular to the original misalignment according to a model from Neale. This
component would decrease at higher speeds. It turned out that after using improved EHP
lubricant this cross-reaction remained high. For financial reasons it was not possible to test
couplings at nominal torque ratings and higher speeds. It could be partly possible now since the
recent revamping of high-speed rig operating from 0 to 3000 rpm up to 125 KW.

Some misalignment is OK for lubrication. Too much causes other

problems.

No comment. Respect tolerances for maximum misalignment with the manufacturer.

Incidentally they mention relative low admissible values to prevent sloppy alignment jobs.
 A case study from Bently Nevada ("Orbit" September 96)
Following a lubrication breakdown due to sludge, a gear coupling ceased to be flexible. The
spacer was transmitted the misalignment and endured considerable alternate bending with
a peak in load concentration areas like fillets. Vibrations increased as a result with banana-
shaped vibration orbits typical of overloaded sleeve bearings. They were lucky to remove
the unit from operation and notice what had happened to the coupling. A little later it
could explode.

What is sludge? What is a banana-shaped orbit?

Prevented by a good oil flow.

Mainly occurs close to overloaded sleeve
(thick-film) bearings. Not applicable directly
for roller-element bearings.
 Reactions of other types of flexible couplings
After testing gear couplings from two different manufacturers without noticing any
substantial difference of reactions, one moved to two other types of couplings: disc and
elastomer.

Disc coupling (ESCODISC) Elastomer

Reactions of disc couplings vs. torque

 Reactions of elastomer couplings vs. torque

Comparison with gear couplings

Disk couplings behave extremely well and their reactions are quasi zero for all values of
torques. Elastomer couplings exhibit torque-dependent reactions that are nevertheless
much less than in gear couplings.
The variable reactions of both couplings that would cause vibrations are always negligible.
 General conclusions
In selecting couplings according to their reactions both static (bearing steady load) and
dynamic (vibrations), the clear winner is the disc couplings. Other considerations come in
when making the final decision. These are summarized hereafter and based on specialized
textbooks.
Bear in mind that some so-called "flexible couplings" are not that flexible.

Couplings advantages Drawbacks

type

Gear Compact, long experience. Gear Lubrication and strong torque-

profiles wear progressively. dependent reactions
Very good to accommodate end Piloting and unbalance. Generates
motion (axial misalignment) reactions at twice the rotational
Poorly operates at very small
Easy to mount. misalignments and high torque or at
zero torque.
Some backlash possible

Grid Basically the same as gear. More flexible in torsion. Strong

misalignment may damage cover
(grid cuts through the cover).
Depending on how one splits the
grid (2X1/2 1X1/4, etc.), one can
excite vibrations at harmonics of the
rotation speed.

Disk No lubrication. Very little static and May suddenly break.

dynamic reactions when misaligned.
More sensitive to axial
Easy to visualize misalignment misalignment (end motion)
through disk flexure pattern.
Axially flexible (axial vibration
Very rigid in torsion. No piloting modes). Not investigated in
and subsequent unbalance. research.
The winner for vibration behavior. Rather bulky.

Elastomer Compromise between dry and Beware internal heating due to

lubricated couplings in term of hysteretic losses and chemical
reaction and end motion. agents and aging. Not recommended
for large ratings. Etc.
Some may be difficult to mount, e.g.
when inserting elastomer blocks one
must move hubs axially.
 Conclusions
Tests on gear couplings indicate that the flexibility of some so-called flexible couplings is a
myth. When misaligned and transmitting torques close to their maximum rating, such
couplings may alter the bearing loads considerably and raise vibration levels higher than
those caused by admissible residual unbalances according to balancing norms. In other
words, when misalignment looms, balancing may never succeed to bring vibration levels to
admissible levels.
Examining vibrations with the most advanced techniques is unlikely to indicate in which
direction a flexible coupling is misaligned. Variable reactions from couplings rotate without
exhibiting any preferential component that could be linked with the direction of
misalignment.
This study confirms for the case of shaft lines with flexible couplings what many
practitioners and vibration specialists have long claimed about the incidence of
misalignments on vibrations: more than 50 % of unexplained vibrations in rotating
machinery originate from misalignments.
It does it with hard figures based on loads and not displacements and vibrations that
depend on the frequency response of the structure at hand. It establishes a methodology to
evaluate coupling reactions and characterize them. The approach relies on somewhat
advanced data processing techniques.

Applying these techniques to disk and elastomeric couplings, one establishes that disk (and
probably diaphragm) couplings exhibit the least reactions (both dynamic and static) in
presence of torque and radial and/or angular misalignments. Based on these facts, one
recommend using them when feasible. Read the next section from Calistrat to get a more
balanced conclusion based on other considerations.

Retrofitting lubricated (e.g. gear) with dry couplings (Calistrat +

comments)
Many coupling users have replaced their lubricated couplings with "dry" couplings; many
others are planning to do it. That is the case in EDF for many types of auxiliary equipmen
(according to Randiranarivo).
When retrofitting with dry couplings, a few items must be clarified:
The main reason users consider replacing lubricated couplings with dry is the ability of dry
couplings to operate long periods of time without maintenance (lubrication, or cleaning the
couplings of sludge accumulation).

A bad lubrication of gear couplings may lead to catastrophic failures (Burgo incident in
Belgium, and "Orbit" September 96 from Bently Nevada)
At the time that dry couplings started replacing gear couplings in special purpose
applications, a few engineers believed that the service life of dry couplings was indefinite, at
least as long as the alignment stays within acceptable limits. Now that dry couplings have a
"long " history, this belief has proved to be only partially true. Concurrently with the
availability of new type dry couplings, the frequency of gear coupling problems started to
decrease. What apparently happened is users learned that if they are as careful installing
gear couplings as they must be when installing dry couplings, many of the notorious gear
coupling problems tend to dwindle away.
Perhaps the most important thing that was learned from failures of dry couplings is that,
most of the time, they occur without any warning, and they have, in a few cases, caused
severe damages to the connected machines.
That is the reason why coupling manufacturers design disk coupling with a backup inactive
gear type coupling for gas turbine shaft lines. The latter acts as an ultimate protection. The
disk coupling normally carries the load with the outstanding insensitivity of its reactions to
misalignments (consult ESCO transmissions for some advice in Belgium: they are home-
grown coupling manufacturers. NDLR Rclame paye en know-how et rien
d'autreremarque supprimer du document officiel mais introduite en vue d'touffer toute
rumeur).
Lubricated couplings, and particularly gear-type couplings, are significantly superior to all
other type of couplings when it comes to power density (amount of torque that can be
transmitted per pound or inch of coupling), axial travel ability, and safety in operation if
well lubricated. In some applications where space is limited, gear couplings can simply not
be replaced by other types without great financial outlays.
The above statements are in no way intended to discount the many advantages of dry
couplings. Calistrat believes that in considering the many types of couplings available
today, a user must be knowledgeable on the subject of couplings, and must use objective
reasoning in the selection of the type best suited for a given application.
Refer to Calistrat for a guide to best select the type of couplings, knowing that many
advances have occurred in the design of dry couplings like disk.
For the vibration specialist, disk coupling is a master choice for quiet operation, as
compared to gear couplings for the following reasons:
They do not cause unbalances due to piloting and variable torque.
They do not generate variable reactions at twice the rotating frequency that depend
on torque.
They do not load the neighboring bearings when misaligned and transmitting high
torques. With sleeve bearings disk couplings ensure some vibration stability.

Disk couplings have advantages for maintenance:

Disk couplings need no lubrication.

They need not be misaligned to operate properly at high torques and they behave
well at zero torque.
But as shown above, other considerations may temper this choice:
Gear couplings usually fail more progressively, though their spacer may sometimes
explode with very poor lubrication.
 They better accommodate end motion due to conflicting thrust bearings.
They are very compact.

This document was created by G. D'Ans.