Extend gear coupling life

Extended service of gear couplings can Since maintenance is a disadvantage of gear couplings
compared with "dry" couplings, why are they 30 pop-
be achieved by understanding the ular? The principal reasons are: Gear couplings can
principles of wear, how lubricants function transmit more pmver per pound, and per inch of diam-
eter than any other type of coupling: Gear couplings
and proper maintenance are forgiving; even when improperly installed, or mis-
handled, they continue to perform better than most
other types of couplings; gear couplings are reliable and
Michael M. Calistrat, Koppers Co., Inc. safe. Their main cause of failure is teeth wear, which is
Baltimore, Md. gradual rather than abrupt. Even when the teeth wear
out completely the only effect is loss of power without
PROPER LUBRICANT selection greatly increases the life
the coupling Hying apart.
of gear couplings. Once properly installed periodic lubri-
Couplings in general can be divided into low speed, or
cation is the only maintenance required. In order to
standard couplings, and high speed, or high performance
select proper lubricants the principles of gear wear and
couplings. Maintenance requirements of the two types
lubricant properties must be understood.
are very different. This article will describe maintenance
of the standard type.

Standard couplings can be defined as the ones "off

the shelf," not made to order. They generally operate at
"motor speeds" and are seldom dynamically balanced.
The gear coupling was invented by Gustave Fast of
Baltimore, Md. in 1920. His basic design is still used
today. Many of the original dimensions for sleeve flanges
and bolt sizes became standards of the industry. As a
consequence, couplings made today by any manufacturer
can be bolted, by half-coupling, with other manufac-
turer's couplings. Standard couplings are sized for the
nominal shaft diameter in Y2-inch increments. A size
1Y2 (or 15 in some cases) is designed to accept a 1.5
in. shaft, however most hubs can be overbored as much
as Y2-in. The size of a coupling also represents, with
few exceptions, the pitch radius of the gear mesh.

Why do gear couplings require lubrication? Al-

though standard couplings are made in a large variety
of constructions, they can be divided into two categories
depending on the method used to accommodate mis-
alignment: through flexing or through sliding. Gear
couplings fall into the second category. Fig. 1 shows a
section of a gear coupling in which the misalignment is
intentionally exaggerated. Let us follow the path of the
hub tooth "A" as the coupling makes one revolution.
Without misalignment the tooth will stay in the plane
C-C. Because of the misalignment the tooth is to the left
of the plane C-C in the upper position, but after one
half revolution it moves to the right of plane C-C in the
Fig. 1-Misalign&d gear coupling. position marked Al. After the coupling completes the

Page 1 of 5
revolution the tooth assumes again the POSltlO;) A.
10
Although the hub tooth remains constantly engaged with
the same sleeve tooth, it moves back and forth (',pr 8

the surface of the mating tooth. The frequency pi this

oscillatory motion is the same as the shaft's speed. The 6
amplitude of the motion (i.e. the total travel of the hub 5
tooth from one extreme to the other) i~ very small and
4
is a function of the pitch diameter and the angle of
misalignmen t:
3
a = P.D. (Tan a) (1)
The velocity of the oscillatory motion is a function of
the amplitude and of the shaft rpm.

= N ( ) = (N) P.D. (T )
s:
VB 7r
60 a
IT
60 ana (2)
19
To prevent the wear that this rubbing of one tooth 1

against the other generates, gear couplings must be .8

lubricated.
.6
How does the lubricant work? The forces between
5
the mating teeth as well as the oscillatory motion tend
to wipe off the lubricant from the contact surfaces. How- .4
ever, when one tooth slides in one direction the lubricant
has a chance to wet the area left uncovered behind the .3
tooth. To do that the lubricant has to work against two
elements: its own viscosity, and the very short time avail-
able during one cycle. The teeth would operate dry if 3 4 5 6 8 10 20
it wasn't for the high centrifugal forces created by the Pitch dia., inches
couplings' rotational speed. This forces the lubricant Fig. 2-Centrifugal forces in couplings.
between the teeth insuring good lubrication.
It is customary to measure the effect of centrifugal
forces with a factor G which indicates the apparent
6
increase in an object's weight when the object rotates
around an axis. \
G (3)
inside a coupling we can write
G (4 )
2
The graph from Fig. 2 helps in finding the G level in
couplings. For example, a Size 4 coupling, with an B-in. 5,000 10.000 15.000
pitch diameter, operating at 3,500 rpm has a G of 1,390 RPM
or, as it is common to say, has 1,390 G's. This large Fig. 3-Gear coupling wear rates as a function of speed.
apparent increase in weight forces the lubricant to flow
rapidly and replenish the voids left behind by the moving
teeth. The influence of centrifugal forces on the quality
of lubrication was confirmed in lab tests. As Fig. 3 shows, of gear couplings; the Fast type, with an all metal seal,
the wear rate of a coupling decreases rapidly as the speed and the elastomer seal type. When the couplings rotate,
increases up to a point where no further benefit is ob- centrifugal forces cause the lubricant to form an annulus
tained from an increase in the G leveL inside the sleeve. The inside diameter of the lubricant
annulus is always the seal diameter because none of the
Oil versus grease-which is better? To compare coupling seals can retain more lubricant when under
oils with greases we have to look at two of their features: high G's. The seal of the Fast type is one of the few so
lubrication properties and convenience. One must first designed that, under static condition it will retain enough
understand that greases are nothing but oils to which a oil to form an adequate annulus when the coupling is
thickening agent was added. The vast majority of greases rotating.
(about 97 percent of total U.S. production) use a Oils cannot be used with elastomer seal couplings
metallic soap as the thickening agent. Generally, these (Fig. 5) mainly because they cannot retain enough lubri-
soaps have no lubricating properties, hence a grease is no cant under static conditions. Either oil or greases can be
better than the oil used in blending it. The biggest ad- used in Fast type couplings. The question of which is
vantage of greases is that they are easier to seal, thus better applies only to the Fast couplings. The answer
they stay where the lubricant is needed. depends on whether a good grease is available. If it is,
Figs. 4 and 5 illustrate two of the most popular type then a grease is preferable simply because more lubricant
Page 2 of 5
EXTEND GEAR COUPLING LIFE creases in the same ratio that the speed decreases, the
centrifugal forces decrease with the square of the speed,
can be packed in the coupling, even though a small p.lll as shown in Equation -4. Thus, the forces that make the
of the lubricant is a thickener. If a good grease is 11,,1 grease flow decrease more rapidly than the time avail-
available, then any high viscosity oil is preferable. Wily able increases. At some speeds the coupling could operate
high viscosity? Studies performed in the lab with without sufficient lubrication, even when it is full of
standard couplings indicated that the most influential grease. This condition becomes particularly critical when
stiff greases (NGLT No.2 or No.3) are used, because
parameter for coupling wear is the oil viscosity. The
graph (Fig. 6) shows that the higher the oil viscosity, they flow less readily than a No. 1 or No. °
grease. It
was determined experimentally that a No.2 (or stiffer)
the smaller the wear-rate. \\1hen greases are considered
the consistency of the grease should not be confused with grease should not be used when the coupling operates
the viscosity of the base oil. The same grease consistency below 10 G's, i.e. when the speed is below:
can be obtained with a wide variety of oils of different Nt" == 850 V (P.D.) (Remember: P.D. is approximately
viscosities. twice the coupling size)
(5)
Greases and centrifugal forces. Let us flrst see what
happens when the centrifugal force is very small, which Now let us see what happens under normal operating
can be the case either at low speeds or when small conditions. It is known that greases have a tendency to
couplings are used. Because the grease is subjected to a partially lose their oils, a phenomenon known as "bleed-
small centrifugal force it needs more time to flow into the ing." There are three primary forces responsible for oil
voids between the teeth. Fortunately, at lower speeds separation from greases 3 : syneresis, gravity and external
mo.re time is available. While the time available in- pressures. In couplings the apparent gravity generated by
centrifugal forces is so high that the other forces re-
sponsible for bleeding become insignificant.
Centrifugal forces have a larger effect on thickeners
from the grease than 011 the oils because thickeners have
a higher specific gravity. Hence, the thickeners are forced
away from the axis of rotation, and wiII eventually coat
the inside surface of the coupling sleeves. In observing
Figs. 4 and 5 one can see that the gear mesh (the hub
and sleeve teeth) is located on the sleeve's inside
diameter.
Hence, the soap separated from the grease will sur-
round the coupling's teeth and WILL PREVENT THE
OIL FROM LUBRICATING THE TEETH SUR-
FACES. In other words, although the coupling is full of
grease it can be starved of lubricant. Understanding
this phenomenon is very important for proper coupling
maintenance.
7'
For many years it was assumed that the amount of oil
~
<,.,,')/ •
separation was a function of the G level only. Conse-
quently, some greases that were not acceptable for high
Fig, 4-AII metal seal coupling. speed operations were considered acceptable for standard
couplings. Recent studies 4,5 have shown that rather than
being a function of the G level, the maximum oil sep-
aration from a grease is mainly a function of time. This
means that a grease will bleed a given percentage of oil
at any G level, if it is subjected to these centrifugal forces
for a long enough period of time. We nse a standard test
method 4 for evaluating greases: a sample is subjected
to a high centrifugal force (36,000 G's) in a laboratory
centrifuge, and the oil separation is closely monitored.
The test is interrupted when the separation stabilizes.
The greases are then given a (K36) rating which is a
fraction. The numerator represents the maximum oil
separation in volumetric percentage, and the denominator
represents the time required to attain that separation.
For example, if a grease has K36 == 70/35 it has a maxi-
mum oil separation of 70 percent and it needs 35 hours
of operation at 36,000 G's to attain this separation. When
the grease is subjected to a smaller centrifugal force, the
time required for the maximum separation can be cal-
culated with the experimental formula:

Fig. 5-Elastomer seal coupling. t = 2,400,000 T /GL4 (6)

Page 3 of 5
where T is the number of hours used to determine the ,

K36 factor. 15

Which grease is best? Coupling manufacturers rely

both on lab tests and field experience when making grease >, 10
recommendations. Depending on the application some "
u
(3
greases are better than others. Just because a particular ~
Wear rate
grease is not on a list of recommended lubricants does
5
not necessarily mean that it is not suitable.
There are four basic parameters that can be used for
evaluating greases for gear couplings:
60 90 120 150 180
A grease should have more than 92 percent oil. It was
shown that most greases will eventually separate into oil Viscosity, SSU at 21 QCF
and soap. The more soap a grease has, the higher the Fig. 6-Effect of lubricant viscosity on coupling wear.
chances of the teeth being starved of lubricant. The 92
percent level is a desirable number rather than an exclud- gear coupling is completely broken-in in three to five days.
ing one, but it is not unrealistic; there are many greases Breaking-in of a coupling creates a special problem for
that have more than 92 percent oil. the lubricant: high temperatures. Rapid wear of the teeth
that are originally in contact generates heat at a larger
A grease should have a base oil with a high viscosity. rate than the coupling can normally dissipate. Temper-
It was shown (Fig. 6) that the higher the viscosity of the ature at the surface of the sleeve can reach as much as
oil, the lower the wear-rate of the coupling. Oil viscosity 120 0 F above ambient. If the ambient temperature is
should not be confused with the grease's consistency. To high the grease in the coupling could operate close to, or
blend a given grease, let's say a NLG 1 No.1, oils with even above its dropping point. F.or this reason greases
viscosities as low as 200 or as high as 4,000 SSU at with a low dropping point should be avoided. It is also
100 0 F can be used. If the oil is lighter more thickener possible that during breaking-in the flange bolts will lose
is used in order to attain the desired grease consistency. some of their pretorque. If the equipment can be stopped
An oil viscosity higher than 900 SSU at 100° F is de- approximately a week after start-up the coupling should
sirable.
be inspected and. the bolts retightened. Unless there is
A grease should resist centrifugal separation. Very few evidence of grease loss between the flanges there is no
companies are equipped to determine this characteristic, need for relubrication at that time.
however K36 values are available for many greases. For
a good choice a grease should not only bleed as little oil Maintenance procedure. In the order of increasing
as possible, but should also bleed the oil slowly. complexity, maintenance can consist of: relubrication,
relubrication with realignment and coupling replacement.
A gear coupling grease should not have a consistency
higher than NLGI No. 1. It was shown that a more Even the best coupling greases cannot perform too
consistent grease could cause problems at low G's. In long; besides the inevitable soap separation, oils oxidize
order to avoid misapplications one should not use No. 2 and lose some of their lubricating properties. The possible
or No.3 greases. loss of lubricant should also be considered. Many users
Some good quality greases incorporate special additives do not install the paper gasket or a-ring between the
or fillers. Experience has shown that EP additives are flanges, which is a mistake. Manufacturers recommend
beneficial only in highly loaded couplings, and in all relubrication at either six months or every year. Actually,
couplings during the breaking-in period. Under normal depending on the grease used and the G level in the
conditions couplings do not operate under high con- coupling, relubrication periods could be longer, or shorter.
tact pressure. Lead, molybdenum disulfide and other high Independent of how often the coupling is lubricated, it
density compounds will separate immediately inside the should first be opened and as much of the old grease as
coupling, and as such are beneficial only lD very low possible should be removed. Even if cleaning is sometimes
speed, high load applications. difficult it should be done because it prolongs coupling
life and prevents unscheduled equipment breakdown.
Coupling's breaking-in. Usually the gear mesh of a Here are a few rules to be observed when opening a
coupling has 60 teeth at the hub and as many at the coupling: Remove and save the bolts and nuts. In most
sleeve. The number can vary slightly depending on the cases the bolts have special body tolerances and should
coupling's size. Because of the many teeth, and depending not be replaced with "similar" bolts. Lock-nuts should be
on the precision of manufacture, more or less of the replaced after six installations, or when they are loose on
matching teeth will be in contact when the coupling first the bolts. Do not separate the sleeves by forcing a chisel
starts transmitting the torque, even if the coupling is per- or screwdriver between the flanges. By damaging the
fectly aligned. The teeth that do transmit the torque are flange surfaces a good seal can no longer be made and
subjected to larger than normal forces, so they wear rela- one risks losing the lubricant when the coupling is re-
tively fast. However, as these teeth wear, the other teeth started. Remove as much of the old grease as possible.
that originally were not in contact start sharing the load. When reassembling the coupling use a new gasket if
Eventually all the teeth are in contact and the coupling the old one is damaged. Proper baIt torqueing is very
is broken-in. The duration of the breaking-in depends on important not only to form a good seal at the flanges, but
the load, the quality of manufacturing and the materials also for proper torque transmission. One should keep in
used. Experience and laboratory tests show that a standard mind that in a properly installed coupling the torque is
Page 4 of 5
EXTEND GEAR COUPLING LIFE If a long spacer is used between the sleeves, and if the
spacer is not provided with plates at both ends, a full disk
gasket and back-up metal plate should be used between
each sleeve and the spacer. The gasket should be installed
betvJeen the sleeve and back-up plate rather than the
other way. Without a back-up plate the disk gasket will
most likely be ruptured by the pressure developed by the
centrifugal forces. When spacers are used it is important
that each half-coupling be lubricated individually.

t,-r_-+- 1-
At least at the first maintenance stop the original align-
ment should be checked to verify if either machine moved
on its base. :Many good articles, and good instructions on
alignment are available.
If the coupling must be removed, but will be reused,
the hub should not be pulled by the sleeve as permanent
damage at the sleeve could occur. If the hub does not have
I puller holes a plate that does not touch the sleeve should
be used behind the hub.
Many questions have been asked about the use of heat
for hub removal. Heat is beneficial only if the hub can
be heated faster than the shaft, which is seldom the case.
A welding torch will most likely damage both the cou-
Fig. 7 - Permanently sealed gear coupling. pling and the shaft. To avoid difficulties in hub removal
the best procedure is to use an anti-seize compound on
transmitted from one flange to the other mainly through the shaft when the hub is first installed. If an anti-seize
friction; only a small part of the torque is transmitted compound is not available, a grease with a zinc-oxide
through the bolts. If the bolts are not properly tightened filler can be used successfully.
then they are subjected to shear forces that can damage
the bolts and elongate the flange holes. Bolt torqueing A maintenance free gear coupling. In some appli-
values are available from coupling and fastener manu- cations the equipment to which a coupling is installed
facturers and depend on coupling size, type and manufac- cannot be stopped when relubrication is scheduled. While
turer. this causes no problems in lubricating the equipment's
The lubricant capacity can vary anywhere from a few bearings, it prevents coupling lubrication. It is then pos-
ounces to ovel' 25 lbs. In order not to waste lubricant the sible that the coupling is not serviced until the next
manufacturer's recommendations should be followed. If scheduled maintenance, or the next one, or never.
this informatio!1. is not readily available it is safe to remove A coupling G which incorporates the advantages of a
two opposite lube plugs, rotate the coupling until the lube gear coupling, but does not require periodic maintenance
holes are about 45 degrees to the vertical, and pump is shown in Fig. 7.
lubricant through the upper hole until it comes out Each half-coupling is double sealed; the lubricant will
through the bottom one. In no case should grease be not be lost or contaminated when the two flanges are
pumped into a coupling if a second lube plug was not separated.
removed. The pressure developed by a grease gun is very
The coupling has a larger lubricant capacity than the
high and a closed coupling can be damaged. If the cou- Fast's and does not require relubrication for a period
pling is new, or if an old coupling was completely cleaned generally longer than the one of the connected machinery.
of the old grease, it is very important to hand pack the A few prototypes have been in continuous and sLiccessful
teeth with some lubricant to insure proper coupling per- operation in selected petrochemical plants for about three
formance at start-up. There is no danger in overfilling years.
a coupling with lubricant!
NOMENCLATURE
a ::= amplitude of axial motion, inches
About the author P.D. ::= pitch diameter of the gear mesh, inches
a angle of misalignment, degrees
MICHAEL M. CALISTRAT is manager, V, ::= axial sliding velocity, inches/second
Power Transmission Development Sec- N == rotational speed, rpm
tions, Research and Development De- R radius df rotation around an axis, inches
partment, Koppers Co., Inc., Baltimore, N, ~" critical speed for some greases, rpm
Md. In this capacity, he is responsible t time required for centrifugal separation, hours
for improving the existing product line
and developing new flexible coupling LITERATURE CITED
1 \Vright) John, "Which shaft coupling is, best. . lubricated or nOn~
products. Mr. Calistrat gradua.ted from lubricated?" Hydrocarbon Processing} April 1'975.
the University of Bucharest with an :: Calistrat, Michael, ('What causes wear in gcar~type couplings?" Hydro-
carbon Processing, January 1975.
M.S. degree in mechanical engineering. :t Ca.lhoun, S. F., "Fundamental Aspects of Grease Bleeding." NLGI
He is currently chairman of the Shaft SPOkCSlnUll, January 1966.

Couplina and Clutches Subcommittee of ASME. Mr. Calistrat • Calistrat, ~fjchacl. "Grease Separation Under Centrifugal Forces:) AS~{E
Paper No. 75-PTG-3, ASME-ASLE Lubrication Conference, October 1975.
holds 14 patents and has authored and presented a number /j; C.lapp, A. ~f" "Fundamentals of Lubrication Relating to Operation and
of technical papers. J\{aintenancc of Turbomachincry," First Turbomachinery Symposium, Texas
A & M University, October 1972.
• U.S. Patent No. 3,953,986, May 1976. •

Page 5 of 5