SUBJECT: Some common misconceptions about
mechanical seals 11-3
- Two hard faces are a sensible choice if there are dirt or
solids in the product you are pumping.
- ans. Seal faces are lapped to
less than three light bands (less than one micron) of flatness.
Dirt and solids cannot penetrate these faces unless they open.
The trick to sealing solids and slurry is to keep the lapped
seal faces together.
- Dual seals are a good choice for a slurry application.
- ans. Putting a clean liquid
between two seals is not going to stop solids from clogging the
inner seal. Since the barrier fluid is at a higher pressure
than the stuffing box pressure you will probably end up
diluting your product.
- Putting the seal outside the stuffing box can keep the springs
and other parts from clogging in an abrasive slurry.
- ans. As the seal faces wear
the seal is going to have to move into the slurry that will
restrict its movement. It is the same problem you face with
many of the dual seal applications used to seal dirt and
- You should not use ceramic seal faces in a mechanical seal.
They will crack when subjected to temperature transients.
- ans. Space vehicles are
covered with ceramic so they can take temperature transients,
its just a matter of which ceramic you are using.
- Modern seal designs are made to fit A.N.S.I. and I.S.O. pump
designs without having to make any modifications to the seal or
- ans. The seals will fit, but
they will not have enough outside diameter clearance for proper
operation. The stuffing box bore should be enlarged.
- Seal faces have to be lubricated.
- ans. Not necessarily. Carbon
graphite is a natural lubricant. Electric motors have used
carbon/graphite brushes for years that do not use any external
- Vibration analysis is a good technique for predicting seal
- ans. Vibration analysis
requires that you know the frequency of the piece of hardware
you are analyzing. This is easy for bearings that are always
made out of the same material, always in the same basic medium
and vary little in shape. Seals come in a variety of shapes and
materials and run in all sorts of mediums.
- Oil is a good barrier fluid to use between dual mechanical
- ans. Actually it is one of the
worse. It has too low a specific heat number and it is not a
very good conductor of heat compared to other liquids.
- Teflon is a universal elastomer. It makes sense to use it in
- ans. Teflon is not an
elastomer because it does not have a memory. To use it in a
mechanical seal you must spring load it to the shaft and that
is never a good idea because you will end up with expensive
shaft damage (fretting). O.E.M. suppliers use Teflon because
they are not sure where the pump is going to be used.
- Shrinking a carbon seal face into a metal holder is an
acceptable manufacturing technique.
- ans. It really is a bad one.
The out of roundness tolerance of the metal holder will clash
with the out of roundness tolerance of the carbon, causing high
loading at several points on the carbon outside diameter. The
carbon should be pressed into the metal holder allowing it to
shear and conform to the metal out of roundness.
- It is a good engineering practice to glue the O-rings in a
split mechanical seal design.
- ans. The glue will create a
hard spot that will give you a leakage problem.
- You should connect the flush connection to the top of gland.
- ans. It should be connected to
the bottom of the gland or stuffing box. This will allow the
flushing fluid to fill the box prior to spilling over the end
restriction in the stuffing box. American prints show the top
half of the drawing, that is why this error is so frequently
- The elastomer Viton is acceptable in water.
- ans. It is a worse choice. The
proper material for water is ethylene propylene. Some specific
grades of Viton can be used in cold water , but none of them
are good for hot water. Viton is cured in sulfur and what ever
attacks the cure attacks the compound. Needless to say sulfur
and water are not a good combination.
- Split seals leak
- ans. It all depends upon your
definition of leakage. If you are talking "fugitive emissions"
that are measured at parts per million you can build a case for
leakage, but if you mean "no visible leakage" then split seals
should be as leak free as any other mechanical seal
manufactured from the same materials.
- You should put a lubricant on seal faces when you install
- ans. It's not a good idea to
put anything on the lapped faces. The trick is to keep the
lapped faces together.
- No one can predict seal life.
- ans. That is a fact, but we
know how long seals should last. They should run leak free
until the sacrificial carbon wears down. (90% of mechanical
seals fail long before that happens).
- In most seal applications the carbon is running on a hard
- ans. The graphite comes out of
the carbon /graphite face and deposits on the hard face. You
can easily see the black mark made by the graphite. The seal
face you are actually running is carbon on graphite. The hard
face is just some place to put the graphite. This is the reason
the seal faces can run dry.
- It is good engineering practice to put a stationary seal ( the
type where the springs do not rotate with the shaft) on a
- ans. Tightening the cartridge
sleeve set screws will pull the cartridge sleeve to one side,
causing the rotating face to no longer be perpendicular or
square to the rotating shaft. This squareness to the shaft is
essential to the performance of any stationary seal
- If you are installing a mechanical seal in a vertical
centrifugal pump, you use the same procedure as installing a seal
in the horizontal version.
- ans. Vertical pumps trap air
in the stuffing box. You will have to install some type of vent
above the seal faces and dynamic elastomer to avoid "dry
running" in these locations.
- PV factors are a legitimate way of predicting seal
- ans. Carbon/graphite seal
faces are sensitive to pressure(P), but not to velocity(V) so
PV has limited value.
- The three hundred series of stainless steel is a good choice
for seal metal components.
- ans. That is true for most of
the metal components, but not for springs or metal bellows. The
three hundred series is sensitive to chloride corrosion
problems in these locations.
- You must not use ceramic as a seal hard face because it will
crack with a rapid temperature change.
- ans. Some ceramics have this
problem, and you should not use them in your designs. The
ceramic called "silicone carbide" is a good choice as a hard
face and does not have the "cold shocking" problem.
- The metal bellows seal should be your first choice for a hot
- ans. the metal bellows seal is
always a good choice in hot fluid to eliminate the temperature
sensitive elastomer or O-ring, but it is not effective in hot
petroleum applications because of "coking problems. In these
applications you have to cool the stuffing box area to prevent
the oil from forming coke solids on the seal moving parts and
- The mechanical seal should be positioned against a shaft or
sleeve shoulder to insure the correct face load and then set
screwed to the shaft or sleeve..
- ans. To compensate for shaft
axial growth or open impeller adjustment, the seal must be
positioned on an adjustable cartridge sleeve.
For information about my CD
with over 600 Seal & Pump Subjects
Link to the Mc Nally home