Seals

Stationary vs. Rotating Seal Designs 11-09

One of the most difficult concepts to teach in writing is the difference in operation between stationary and rotating mechanical seals. It’s like trying to write a set of instructions on to how to tie your shoe laces, easy to demonstrate, but difficult to explain.

Figure #1 is a cross section of a typical rotating seal. It’s called a rotating seal because the spring loaded face rotates with the shaft. If the spring loaded face did not rotate this seal would be called a stationary seal. You can see a cross section of a stationary seal in Figure #2.

FIGURE # 1 THE ROTATING SEAL
A Stationary face
B Rotating face
C Sleeve
D Gland
E Stuffing box
1. Stationary face
2. Dynamic Teflon® wedge
3. Sleeve gasket
4. Sealing interface
5. Gland gasket
FIGURE # 2 THE STATIONARY SEAL
A Stationary face
B Rotating face
C Sleeve
D Gland
E Stuffing box
1. Dynamic O-ring
2. Static O-ring
4. Sealing interface
5. Gland gasket

In these drawings it is easy to place the mechanical seal rotating parts perpendicular, or square to the shaft, but in practice it is just about impossible to do so.

Please look at Figure #1 again:

We would like to have rotating face “B” perpendicular, or square to shaft “c”, but that is impossible because rotating face “B” is being pushed against stationary face “A” that is positioned in gland “D”, and gland “D” is not square to anything.

There are a few reasons why gland “D’ is not square to the shaft:

  • You cannot tighten several gland bolts through a gasket and get any kind of squareness.
  • The gland is manufactured from a casting that is not perpendicular, or square to anything.
  • The face of the stuffing box is not manufactured square to the shaft. Most of the time it is a rough casting. Remember that the pump was designed for packing, there was no need to make this surface machined square to anything.
  • Assume the seal assembly was done with dial indicators to ensure squareness, the minute the pump comes up to temperature thermal growth will alter the careful setting you made.
  • There are some additional causes of this non-squareness to the shaft:
    • Misalignment between the pump and its driver.
    • Operating the pump off of its Best Efficiency Point (B.E.P.)
    • Thermal expansion at the wet end of the pump.
    • Pulley driven pump designs.
    • Dynamic unbalance of the rotating parts.
    • Bent shafts.
    • Pipe strain.

The cocking of the gland and stationary face means that the springs will be loaded unevenly and will have to move back and forth with shaft rotation.

The springs and rotating face “B” actually move back and forth twice per revolution of the shaft, and at 1750 rpm this would be 3500 times per minute, or just about 60 times per second (try and move one of your fingers back and forth 60 times per second to see how fast that really is). Needless to say any interference with this movement can lead to a premature seal failure:

  • The shaft tolerance and finish become critical because the Teflon® wedge has to slide back and forth with this movement.
    • Depending upon the amount of cocking, this sliding will lead to shaft fretting or damage in a short period of time.
  • Spring loaded Teflon® sometimes tends to stick to the shaft or sleeve if the sleeve outside diameter tolerance is on the high side, or if the shaft finish is not smooth enough.
  • The springs can break if they experience too much flexing. They will work harden and fatigue prematurely.
  • Centrifugal force can move the rotating face square to the shaft, opening the lapped faces. This happens at about 5000 fpm. surface speed or 25 m/second.
  • The seal faces can open if the springs fill with solids. There are multiple reasons why they would clog:
    • The pumped product can solidify with a change in temperature. This is not often a reversible process.
    • The product can crystallize with a change in temperature. This is normally a reversible process.
    • The product can become viscous with a change of temperature, or sometimes from agitation. Usually not reversible
    • Dirt or solids in the product can clog the springs.
    • Some fluids like hard water or hot petroleum can, and will build a hard film on the springs and sliding components.

Many of these solidification problems are experienced when the pump is shut down and subject to temperature changes in the stuffing box area. This can cause frequent seal failures when the pump is first started and lasting until the solidified product reverses back to its liquid state, which may be never.

The stationary version of the seal has none of these problems.

Look again at Figure #2. The rotating face is held square to the shaft by a clamped surface. This reference remains even if the pump experiences deflection from operating off the Best Efficiency Point (B.E.P.), pipe strain, or misalignment between the pump and its driver.

When the gland (D) is tightened to the face of the stuffing box it will cock for the same reasons that it did with the rotating version of the seal, but unlike the rotating version, the springs will not move back and forth twice per revolution of the shaft because they are not rotating with the shaft.

If the gland were severely cocked it would cause an uneven wear of the seal faces, but no back and forth movement that can be interfered with, causing the seal faces to open and leak.

  • This is the same type seal that is used in the pulley driven water pump of your automobile. And as you are well aware the radial thrusting caused by the pulley drive mechanism has little to no affect on the seal performance.
  • High speed pumps such as the Sundyne design use this type of seal to prevent the faces from opening as a result of the centrifugal forces generated at the high shaft speeds.
  • The main shaft seals of our atomic submarines use large size stationary seals to compensate for the terrible misalignment problems found in these applications.

Seal manufacturers can supply you with both stationary and rotary versions of the mechanical seal in solid, split and bellows designs. To ensure squareness to the rotating shaft they require positioning the rotating portion of the seal against a shaft shoulder or a clamped reference shoulder that has been installed on the shaft. This clamping arrangement accounts for the higher cost associated with stationary seals.

It’s no contest, choose stationary every time.

Take a look at the drawing Figure #3. The Cartridge Stationary Seal

Consumers like the many advantages of cartridge seals, so it was natural to want to mount the stationary seal on a cartridge to get these benefits.

Unfortunately many consumers got a surprise instead of the hoped for advantages. It turns out that mounting stationary seals on a cartridge might not be such a good idea after all.

FIGURE #3 THE STATIONARY CARTRIDGE SEAL
When you tighten the set screws on the cartridge sleeve, the sleeve will move away from the shaft and cock the rotating face.

Cartridge seals present a special installation problem. When you tighten the set screws on the cartridge, the clearance between it and the shaft will cause the sleeve to raise higher on one side, causing the rotating face to no longer be square to the shaft.

The affect of this is that the stationary unit will now move back and forth twice per revolution just like the rotating seal. The only real advantage of this design in a cartridge version is that because the dynamic elastomer is in the stationary face, there will be no fretting or damage to the expensive shaft or pump sleeve. You will, however, frett the barrel on the stationary face .

In paper 11-10 I will discuss this problem in detail and show you some possible solutions, but mean while time keep in mind that mounting a stationary seal on a cartridge usually is not a good idea.

If you would like this paper to end on a positive note, then be aware that split seals are now available in the stationary configuration which should make them a first choice in many applications.

® Dupont Dow elastomer

Category

Posted

  • On February 18, 2018