Fundamentals of Mechanical End Face Seals

 

An end face mechanical seal is a device intended to prevent or minimize leakage from a vessel through the clearance around a rotating shaft entering that vessel. Perhaps the most common example is the end face mechanical seal used in the water pump of an automobile engine.  Most pumps used in petroleum refineries, chemical plant and pipelines also use end face mechanical seals.

The simplest possible mechanical end face seal consists of a shoulder on a rotating shaft which rubs against a stationary case. This concept is shown in Figure 3.

As with packing or any bearing material, lubrication and cooling are required to prevent heat buildup and wear. Hydraulic pressure tends to force fluid between the faces and provide a lubricating film but the face separation must be kept very small to minimize leakage.  Cooling is provided by the surrounding liquid.  The conceptual design shown in Figure 3 is very simple, but it demonstrates the basic principle of the end face mechanical seals.  Of course, it has functional drawbacks which must be addressed.

A major shortcoming of the conceptual design shown in Figure 3 is that there will be always be some shaft movement during operation of the equipment. This movement is a result of manufacturing tolerances, vibration, hydraulic forces and wear.  Shaft movement may either decrease face separation, causing gross face contact and wear, or increase face separation, causing increased leakage.

Seal face leakage is governed by many variables, but the dominant variable is face separation. A variation of a few micro-inches (millionths of an inch) in the face separation can cause significant changes in leakage.  Unfortunately, shaft movement can amount to several thousandths of an inch.

A practical approach to overcoming shaft movement is to mount one of the seal faces in a flexible manner so that it can move axially. Obviously a desirable feature, this flexibility is a prerequisite for effective seal design.

Figure 4 shows an improved seal as compared to Figure 3. In Figure 4,  The sealing shoulder on the shaft has been removed and replaced with a component which is not permanently attached to the shaft.  This component is called the primary ring.  The “face” of the primary ring rubs against the mating ring.  Since the primary ring and the shaft are two separate parts, an additional sealing device must be used to prevent leakage between the shaft and primary ring.  The flexibly mounted primary ring can compensate for the small variations in movement on the axial plane.  It can also adjust for seal face wear.  Figure 4 is a very simple mechanical seal but it illustrates the concept used by the majority of mechanical seals.  Of course, some additional components are required to preload the faces, transmit torque and provide ease of installation.

Figure 5 shows a more complete mechanical seal including a replaceable mating ring, “O-ring” gaskets, springs, setscrews and various other hardware. As will be seen later, the design of these components may vary considerably according to the required service for the seal.  In addition, the assembled components themselves may be arranged and oriented in various ways to accomplish varying degrees of sealing, reliability and redundancy.

Leakage

In a mechanical seal, the primary leakage path is between the seal faces.  Naturally, increasing the face separation increases the leakage. In fact, as will be shown later, doubling the face separation can increase the leakage rate by a factor of eight!  This relationship between leakage and face separation provides a powerful incentive for minimizing face separation.  In modern mechanical seals, face separations are so small (on the order of a few microns) that the leakage rate is affected by the surface roughness.  The effective face separation is a combination of the surface roughness of the two mating parts and the fluid film thickness. This concept is illustrated in Figure 6.  The seal manufacturer can control the initial surface finish by lapping and polishing.  A typical seal face is flat to within 23 millionths of an inch.  This degree of flatness is so small that refracted light rays must be used to measure it.  The fluid film is established during initial start-up of the seal by hydraulic forces.

The principle of establishing a fluid film is essential to all seal designs. Most mechanical seals are designed to operate in liquids; these seals require a liquid film.  Designing a seal to operate on a gas film is much more complex.  Whether the film is gas or liquid, it reduces the gross contact between the rotating component and the stationary component.  It also provides lubrication to reduce friction and wear.  Without a stable fluid film, gross rubbing contact could damage the faces.

Classification of Mechanical Seals

Mechanical seals may be classified by their design features and the arrangement of those features. The Design category includes the details and features incorporated into a single primary ring/mating ring pair. The Arrangement category includes the orientation and combination of the primary ring/mating ring pair.  Figure 7 illustrates the classification of mechanical seals.

Figure 7.  Classification of mechanical seals.

Classification by Design

All mechanical seals contain both rotating elements and stationary elements which include five basic components:

1. Seal ring (sometimes called primary ring)
2. Mating ring
3. Secondary sealing elements
4. Springs
5. Hardware

The Design classification considers the details which enter into the features of these components. Some examples of these features are balance, face treatment, rotating element, springs, secondary sealing elements and drive mechanism.  In general, these design features are not completely independent; that is, emphasis of a particular feature may also influence other features.  For example,  selection of a particular secondary sealing element may influence the shape of the primary ring.

Primary Rings

By definition, the primary ring is the flexible member of the mechanical seal. The design of the primary ring must allow for minimizing distortion and maximizing heat transfer while considering the secondary sealing element, drive mechanism, spring and ease of assembly.  Many primary rings contain the seal face diameters, although this is not a requirement of the primary ring.  The primary ring always contains the balance diameter.

Balance.   The term “balance” is frequently referred to as the relationship of hydraulic forces on a seal but it is actually a geometric ratio.  Balance ratio is defined as the ratio of the hydraulic closing area to the hydraulic opening area. This ratio is customarily expressed in a percentage.

Figure 8 illustrates the concept of balance. In a seal, hydraulic pressure acts on the back of the primary ring; the resulting force pushes the faces together. This force is called the closing force and this area the closing area.  Similarly, any pressure between the seal faces creates an opening force which tends to separate the faces.  Therefore, the face area is also called the opening area.  The balance ratio is simply the ratio of the closing area to the opening area.

As shown in Figure 8, the area above the seal face outside diameter is disregarded when the closing area is computed. This area is not considered because the pressure is the same all around it;  consequently the contribution of the resultant of the hydraulic forces on this area is zero.

When the closing area is reduced, the closing force is reduced proportionally; this feature can be used to advantage when designing a seal. However, for a seal shape such as shown in Figure 8, the closing area will always be greater than the opening area. In order to make the closing area less than the opening area, the shape can be changed as shown in Figure 9.

The seal shown in Figure 8 is said to be an “unbalanced” seal. Its balance ratio is greater than 100% because of the necessary clearance underneath the mating ring.  Typical balance ratios for unbalanced seals range from 120 to 150%.  The seal shown in Figure 9 is said to be a “balanced” seal; its balance ratio is less than 100%.  The balance ratio of “balanced” seals is typically from 65% to 90%.

The distinction between “balanced” seals and “unbalanced” seals is simply that balanced seals have a balance ratio less than 100%.  A seal with 99% balance ratio is balanced, a seal with 101% balance ratio is unbalanced.

For a given pressure, balanced seals have less face load than unbalanced seals. Therefore, balanced seals are normally used in higher pressures than unbalanced seals.

Primary ring shape.  The shape of the primary ring may vary considerably according to the incorporation of various design features.  In fact, the shape of the primary ring is often the most distinct identifying characteristic of a seal.  Figure 10 shows four examples of typical primary ring shapes.

Figure 10a represents a primary ring associated with elastomeric bellows seals. This primary ring has been optimized to take advantage of the elastomeric bellows and a large, single spring.

Figure 10b represents a primary ring with an inserted seal face. Insert faces must be designed with care because temperature differentials can cause differential expansion between the adaptor and the primary ring.  Insert designs can also have problems associated with mechanical stress and distortion.

Figure 10c and 10d show how the shape of the primary ring is influenced by the secondary sealing element. Figure 10c is a primary ring designed to work with a wedge.  Figure 10d is designed to work with an O-ring.  Also, Figure 10c shows an unbalanced shape while Figure 10d is a balanced shape.

Face Treatment.  The most common seal face design is a plain, flat surface but there are many special treatments designed for specific applications.  Figure 11 shows some of the more common face treatments.  The plain, flat face is most common.  In general, face treatments are a means of modifying the pressure distribution between the seal faces.  The most common objective is to increase the opening force and thereby reduce the magnitude of the mechanical contact.  Face treatments may be considered to produce hydrostatic or hydrodynamic forces.

Hydrostatic forces do not depend on the rotational speed whereas hydrodynamic forces vary with the rotational speed.

The simplest face treatment is a plain face that is not flat. This design produces forces that are primarily hydrostatic. Figure 11b shows an example of a face that is lapped so that it tends to touch the mating ring at the inside diameter.  This means that the leakage path is converging.  Although this is a simple concept, it is actually difficult to lap the required taper with the desired accuracy.  Too little taper will allow some rubbing to occur at the ID which will change the taper.  Too much taper will cause gross separation of the faces with the resulting high leakage.

Figure 11c shows another type of hydrostatic face treatment — the hydropadded face. Hydropads are recesses which are machined into one face — usually the face with the softer material.  After machining the recesses, the remainder of the face is lapped flat.  The size, quantity and location of the hydropads is such that the faces still rub; therefore, leakage is low.  The depth of the hydropads is sufficient to allow for long life in spite of the rubbing.  Hydropads actually show some slight hydrodynamic effects, especially in viscous liquids.

Spiral grooves, shown in Figure 11d, produce hydrodynamic forces and generally are used to cause non-contact operation. Spiral grooves also produce a pumping effect.  The grooves may be oriented so that pumping is either with or against the pressure differential.  With a suitable design, the pumping effect can overcome the hydrostatic leakage effect.

Rotating Element.  Although most illustrations have shown the primary ring to be rotating with the shaft, either the primary ring or the mating ring may be used as the rotating element.  Seals with rotating primary rings are said to be “rotating” seals; seals with stationary primary rings are said to be “stationary” seals.  Because the springs are always associated with the primary rings, sometimes the distinction is made as “rotating springs” versus “stationary springs”.

For convenience, rotating seals are used in most equipment. Pump shafts are already made of a comparatively high grade material and manufactured to close tolerances. This makes pump shafts well suited for rotating seal applications.   Assembly of rotating seals can generally be done directly on the shaft or by using a relatively simple sleeves.  Figures 5, 8 and 9 all show rotating seals with stationary mating rings.

Stationary seals have some advantages over rotating seals. In small, mass produced seals for modest services, the entire seal may be placed in a package which minimizes shaft and housing requirements for the equipment.  Figure 13 shows a low cost stationary seal.   Stationary seals are also used to advantage in large sizes or at high rotational speeds.  Above 5,000 to 6,000 fpm, a rotating primary ring (which is flexible, by definition) may require dynamic balancing (for rotational imbalance) in order to operate in a stable mode.  A stationary primary ring does not require this balancing.  On the other hand, the stationary seal does require a close bore tolerance.  This close bore tolerance is usually a second manufacturing operation on most equipment.  Stationary seals sometimes incorporate special design features such as auxiliary liners, sleeves or adapters to help retain the mechanical advantages of the seal.  Figure 14 shows a stationary metal bellows seal used in high temperature centrifugal pumps.

Mating Rings

By definition, the mating ring is the non-flexible member of the mechanical seal. The design of the mating ring must allow for minimizing distortion and maximizing heat transfer while considering ease of assembly and the static secondary sealing element.  The mating ring can contain the seal face diameters, although this is not a requirement of the mating ring.  To minimize primary ring motion, the mating ring must be mounted solidly and should form a perpendicular plane for the primary ring to run against.  Figure 15 shows some typical mating rings.

O-Ring. Figure 15a-c shows a mating ring which uses an O-ring. This is a simple arrangement but is limited to the temperature and chemical compatibility range of the elastomer.

Floating.  Figure 15d is the floating ring type.  This design offers the flexibility of using teflon, grafoil, or an O-ring (Figure 15c) for the secondary sealing item.  If teflon or grafoil is used, a pin should be added to prevent rotation due to the low coefficient of friction of these materials.

Cup Mount.  Figure 15e, the cup-mounted construction, offers a low-cost arrangement which can be used with surface finishes over 63 rms.  It is limited on high temperature applications and to lower pressure ranges because of the insulation effect of the cup.

Clamped-In.  Figure 15f represents a clamped-in mating ring design. A clamped-in design uses a series of gaskets to prevent leakage between the mating ring and the pump casing or gland.  This design has a wide temperature range since it can be used with spiral-wound metallic gaskets.

Secondary Sealing Elements

The secondary sealing elements are gaskets which provide sealing between the primary ring and shaft (or housing) and the mating ring and shaft (or housing). They are called secondary sealing elements because their leakage path should be secondary to the seal face leakage.  Loading by hydraulics or mechanical force makes the secondary seal tight in its confined area.  The secondary sealing element for the mating ring is always static axially (although it may be rotating).  Secondary sealing elements for the primary ring are described as being either pusher or non-pusher in the axial direction. The term pusher is applied to secondary seals that must be pushed back and forth by the movement of the shaft or primary ring.  A non-pusher secondary seal is a static seal for the primary ring.

Pusher type secondary seals have the disadvantage of damaging the surface to which they must seal. This damage, called fretting, is caused by the cyclic movement of the secondary sealing element as the shaft rotates.  In contrast to the pusher design, a non-pusher secondary sealing element could not cause any fretting.

On the other hand, this rubbing and dragging effect is also an advantage of the pusher design because it adds damping and therefore stability to the seal. This damping can make a significant difference in performance for some services.

Figure 16 shows examples of pusher, non-pusher and static secondary seals. The pusher design may use O-rings, wedges, etc.  The non-pusher is always some sort of bellows with a static section.  Mating rings use various static gasketing and O-ring designs.

Bellows.  Figure 16a is a full convolution elastomeric bellows.  It offers the greatest possible flexibility to the front section of the primary ring.  The front section of this bellows has minimum contact with the shaft or sleeve, thus minimizing wear and hang-up.  The large tail section provides a considerable sealing area to compensate for imperfections in the shaft.  There are also are variations of the full convolution design  that use a half convolution.   Its ability to accept axial motion is reduced, due to the one-half convolution design.

Figure 16b is a bellows made out of teflon or glass-filled teflon combinations. This seal which is designed for the extremes of corrosive environments, provides the advantages of the non- contacting bellows convolution.  Because it is made out of teflon, support rings or a drive collar is required to attach the tail section or the bellows to the shaft.  Due to the requirements of flexibility, the convolutions must be considerably larger in cross-section than the typical elastomeric bellows design.

Figure 16c is an all-metal bellows style seal. It has the inherent advantage of flexibility associated with bellows design for high temperature applications.  Due to its all-metal construction, it offers considerable freedom of design since it is not restricted to the temperature and chemical limits of elastomers.  Metal bellows are constructed of individually welded metal leaves approximately .004/.012 thick.  A large number of leaves are required to provide the maximum amount of flexibility associated with other styles of bellows seals.  The mechanical closing force that is provided by the spring on other seal designs is accomplished by stressing the bellows from free height in this design.  Metal bellows seals are inherently balanced because of the manner in which the bellows becomes distorted when pressurized.

“V” Rings, “U” Cups and Wedges.  Figure 16d is a wedge. Wedges are typically made of TFE material. They are considerably more flexible than the “V” ring or U-cup arrangement because they operate on the ball-and socket principal. Special manufacturing fits are not a requirement for wedges because of the shallow angle of contact between the primary ring and the shaft.  However, it does require a polished surface for effective sealing (32 rms with polish).

In addition to wedges, there are U cups and V rings.  Their  construction as a secondary seal offers limited amounts of flexibility and requires extremely close tolerances on the cross-section fits.  The “V” rings are generally manufactured in TFE or TFE-filled material, requiring pre-loading of the rings to activate the point contact on the lips.  Because of the “V” ring design, it can be used at high pressures but it is the least flexible, and most rigid of the secondary seal designs. Because of it’s construction, the “V” ring cannot flex to compensate for motion and it must be pushed along the shaft to take up wear at the seal faces.  Shaft and primary ring finish must be highly polished (15 rms) in order for this type of secondary seal to operate leak-free.

O-Rings.  Figure 16e is the O-ring.  This is by far the simplest and most popular secondary sealing element.  It has been used successfully over a wide temperature range and in a variety of fluids.  O-rings are considered to be self-energizing seals and do not require much mechanical preloading.  This feature allows O-rings to be used at very high pressures.  O-rings are offered in a complete range of chemical resistant and general service compounds.  Buna-N, Neoprene, Ethylene-Propylene, Viton, and Kalrez are typical materials selected for a variety of service conditions.  On the other hand, Teflon does not make a good O-ring material, particularly if the O-ring is to be dynamic.

There are encapsulated O-rings. This is an attempt to obtain the chemical resistance offered by Teflon and the flexibility of the elastomeric O-ring.  Unfortunately, the result is dominated by the hardness of the Teflon outer shell.  The recommended surface finish of the shaft/sleeve is 15 rms.  The encapsulated “O” ring is sensitive to temperature fluctuations.  Because of these limitations, encapsulated O-rings are usually recommended only for static sealing.

Springs

In every mechanical seal there is always a need for keeping the faces closed in the absence of hydraulic pressure.   Generally, a mechanical device in some form of spring is used.  Figure 17 shows some of the different types of springs used in mechanical seals.

Single spring.  A single spring seal has the advantage of comparatively heavy cross section coil which can withstand a higher degree of corrosion.  Another advantage is that single springs do not get clogged by viscous liquids.  The disadvantage of a single spring is that it does provide uniform loading characteristics for the faces.  Also, centrifugal forces may tend to unwind the coils.  Single springs also tend to require more axial space and a specific spring size is required for each seal size.

Multiple springs.  Multiple springs are usually smaller than single springs and provide a more uniform load at the faces.  The same spring size can be used with many seal sizes by simply changing the number of springs that are used. Multiple springs resist unwinding from centrifugal force to a much higher degree than a single coil spring since the forces act differently.  The most obvious disadvantage of small springs is the small cross section wire.  This makes the smaller springs subject to corrosion and clogging.

Wave spring.  The next form of spring generally considered is the wave type, simple described as a washer into which waves have been formed to provide a given amount of mechanical loading . The main reason for using this type of spring is that it requires even less axial space than the multiple spring design.  On the other hand, special tooling must be made for best manufacturing results.  Further, the tempering required on this design limits materials to those which are not as corrosion resistant as the high grade stainless and Hastelloy groups.  Also, when using wave springs, a greater change in loading for a given deflection must be tolerated.  That is, a great deal of force loss or force gain, with comparatively small axial movement must be expected.

Belleville washer.  The Belleville washer is a very stiff spring; in fact, the normal problem with the Belleville washer is that the spring rate is to high.  To reduce the spring rate, the washers are stacked.

Metal bellows.  A metal bellows is actually a combination of a spring and secondary sealing element.  Welded edge metal bellows resemble a series of Belleville washers.  Formed bellows may be used to reduce the quantity of welding; however, a formed bellows has a much higher spring rate than a welded bellows.  The bellows thickness must be selected for resistance to pressure without an excessive spring rate.  The welding technique and bellows shape must be selected for maximum fatigue life.

Hardware

The term “hardware” is used to describe the various devices which hold the other components together in the desired relationship.  For example, a retainer might be used to package the primary ring, secondary sealing element and springs into a single unit.  Another example of hardware is the drive mechanism which is necessary to prevent axial and rotational slippage of the seal on the shaft.

A drive mechanism is required because of the torque created between the seal faces. Both static and dynamic drives are required. The static drive is only required to hold an axial position and transmit torque.  The dynamic drive must transmit torque and allow for the axial flexibility of the primary ring.  Figure 18 shows some variations of drive mechanisms.  Dedicated devices made of strong materials are said to be positive drives and are generally preferred in the heavy duty seals.

Dent drive.   The dent drive is an effective drive generally utilized because of its simplicity, comparative economy and because it can be made with simple tooling.

Key drive.  This is one of the more rugged forms of drive. In high pressure, comparatively large size units, the ruggedness of this type of drive is in keeping with the balance of the sturdy features incorporated in seals for high pressure applications.

Set screws.  This simple method of driving takes advantage of a common hardware item.  Set screws are probably the most common type of static drive mechanism.  They are not recommended for dynamic drive because the roughness of the threads can restrict axial flexibility of the primary ring.

Pins.  Other positive drive designs include slotted pins, dowels or split roll pins.  In the mating ring, they are probably one of the simplest methods of insurance to eliminate the possibility of spinning.

Snap ring.  Snap ring drives are sometimes utilized when axial space limitations prevent the use of other types of drive. These are generally limited to light duty services without corrosion.

Slot.  This is another of the more rugged methods of engaging metal to metal, especially when the slots and mating ears are used in multiple forms.  It is one of the best methods to assure positive engagement between two parts that have relative axial motion between them.  Its basic design also allows for the high degree of flexibility generally incorporated in the primary ring.

Elastomer Drive.  This is not a positive drive method and is generally limited to light duty seals; however, it is extremely simple, economical and effective for some services.

Spring drive.  Another effective drive mechanism for light duty seals is to incorporate the drive into the springs.  Care must be taken in this design as to the direction of rotation, corrosion rate and spring rate.

Classification by Configuration

Although all end face mechanical seals must contain the five elements described above, those functional elements may be arranged or oriented in many different ways. Several dimensional and functional standards exist, such as API Standard 682 – Shaft Sealing Systems for Centrifugal and Rotary Pumps, which describes the configurations for used in Oil & Gas applications. Even though the scope of API 682 is somewhat limited, it may be extended to describe end face mechanical seals in general.

Configuration refers to the number and orientation of the components in the end face mechanical seal assembly. For example, springs may be rotating or stationary.  Single or multiple pairs of sealing faces may be used.  For multiple seals, the individual pairs of sealing faces may be similarly oriented or opposed.  Containment devices such as bushings may or may not be used as part of the configuration.

Single seals.  The vast majority of all seals fall into the single seal category. In this category the seal can be mounted so that it is inside the process liquid or outside the process liquid.  Inside mounted seals are easier to cool and generally can seal higher pressures.  Also the direction of leakage is opposed by centrifugal force.  Inside mounted is by far the most popular, see Figure 19.

An outside seal is shown in Figure 20.  Outside seals have minimal contact with the process liquid.  This is an advantage in sealing corrosive liquids providing that overheating is not a problem.  A disadvantage of outside seals is that the direction of leakage is the same as centrifugal force.

Multiple seals.  Multiple seal arrangements can provide environmental controls and/or redundancy.  The most common types of multiple seal arrangements were previously called double and tandem but are called Arrangement 2 and Arrangement 3 by API 682.  The double seal emphasizes environmental controls while the tandem seal emphasizes redundancy.

The principal purpose of the tandem seal is to provide redundancy. The two seals are oriented so that the outer seal can accomplish the sealing task if the inner seal fails. A tandem arrangement is shown in Figure 21.  Because of this redundancy, the outer seal is sometimes called the “backup” or “safety” seal.  Tandem seals use a buffer fluid to lubricate and cool the outer seal.  The inner seal operates in the process liquid and is cooled by that liquid.  Any leakage from the inner seal must be contained by the outer seal and buffer system.  The buffer fluid is normally at a pressure less than the stuffing box pressure.  In order to keep the buffer pressure low, the buffer system is vented continuously.

Tandem seals (API Arrangement 2) are also used in “stages” (perhaps this is the origin of the name “tandem”)  when process pressures are extremely high. When tandems are used in stages, the buffer system is pressurized to some intermediate pressure, typically half the stuffing box pressure.

The double seal (API Arrangement 3) uses two primary ring/mating ring pairs oriented so that a pressurized barrier liquid is maintained between them as shown in Figure 22. On double seals, the inner seal seals between the process liquid and the barrier liquid.  Because the barrier system is at a greater pressure, any leakage is barrier liquid.  The outer seal seals between the barrier liquid and the atmosphere.  In most cases the barrier liquid is circulated and cooled to prevent heat buildup.

Notice that Figure 21 and 22 are identical.   The only differences are the design of the faces and balance diameter and these details are not shown.  This is somewhat controversial and many would describe both Figure 21 and 22 as being “tandem” because of the orientation of the components.    Traditionally, a “double” seal would have the primary rings in a back-to-back orientation; however, for purposes of operation, either face-to-back (shown) or back-to-back may be used.

Figure 23 illustrates face-to-back, back-to-back and face-to-face orientations.  Although Figure 23 shows rotating primary rings, the same concept and names applied to stationary primary rings.

The biggest problem with the double seals is maintaining the barrier liquid at a higher pressure than the stuffing box. It is generally recommended that a 20 psi or 10% differential in favor of the barrier be maintained at all times.

In the past, double seals were described as primary ring/mating ring pairs which faced in opposite directions.   A classic double seal is shown in Figure 24.  This is usually true but is not a requirement of the double seal arrangement.  Similarly, a tandem arrangement was frequently described as seal pairs facing the same direction, see Figure 21.  Again, this is not a requirement.  In a tandem arrangement, the buffer fluid is at the lower pressure and is continuously contaminated by leakage of the product across the primary seal.  In a double seal arrangement, the barrier fluid is at the higher pressure and the product fluid is continuously contaminated by leakage of the barrier fluid across the primary seal.

Double seals (API Arrangement 3) are used with API Piping Plans 53A, 53B, 53C or 54.   Tandem seals (API Arrangement 2)  are used with API Piping Plans 52 and 55.

Adaptive Hardware

The term “adaptive hardware” is applied to hardware designed to simplify the incorporation of the primary ring and mating ring into the equipment which requires the mechanical seal. The most common examples of adaptive hardware are the sleeve and gland.

When the components are pre-assembled onto a sleeve and gland plate, the complete assembly is called a cartridge seal. This complete assembly can be easily slid onto the shaft and bolted in place thus reducing the potential for installation errors. Some cartridge seals use regular component seal parts whereas other cartridge seals might use specific purpose parts. API 682 specifies that only cartridge seals are acceptable to the standard.

Gland Plates

Figure 23 shows several common gland plate types.

Figure 23a shows the plain integral type gland plate which, if lapped on the sealing surface and gasketed perfectly to the face of the box without distortion, can serve as a sealing face against the rotating seal head. It is an item utilized on many compressor installations, showing that when applied correctly, even a gas such as Freon can adequately be sealed with this type of design.  It becomes a desirable feature, especially where axial room is at a premium.  It must be considered only where practical.  If a relatively large bolt circle in relation to the shaft size is inherent in the unit, this design may prove uneconomical.

Figure 23b is a plain end plate with a pressed-in seat element. Some seal manufacturers advocate this design and it can be utilized where its stresses, replacement features and economics can be tolerated.

Figure 21c shows a so-called plain gland plate but with an O-ring seat added. Where the use of the integral gland plate is not practical from the standpoint of replacement cost or because the seat material itself must be of a higher grade material than it is practical for the endplate, this type of design becomes useful.

Figure 23d shows the clamped in seat with a gasketed flange to overcome the inadequacies of the pressed-in element.

Up to now these gland plate designs have been illustrated without any auxiliary holes. Counterparts of these designs which employ a flushing hole in the gland plates are also shown.

Still another auxiliary feature is termed the quench type gland plate. This gland type incorporates an entrance for liquid behind the seal face on the atmosphere side.  Two other desirable features are to be gained from this quench gland-one is the safety provided by the throttling effect on the escaping liquid if the seal fails; the other relates to toxic liquids.  If the liquid handled is toxic-instead of using the quench gland with a secondary liquid, the liquid can be vented to a safe area.

Sleeves

Sleeves are employed as an adapter from the seal requirements to the shaft requirements of the equipment. For example, the seal may require different diameters, as in a balanced seal.  Figure 24 shows the two most common types of shaft sleeve.

Figure 24a shows the common “hook” sleeve which requires a locating step in the shaft. This is a common sleeve used in API Process Pumps.

Figure 24b is the “cartridge” or “package” sleeve.  This design allows the complete seal, gland and sleeve to be assembled before it is installed in the pump.