Patent Application
20080181762
1. Field of the Invention
The present invention relates to a method and device for reducing or eliminating axial thrust, axial oscillations and radial oscillations of the rotor commonly associated with rotary machines. The term “rotary machines” for the purposes of this description includes centrifugal, axial, turbo- and other pumps, compressors, pneumatic and hydraulic turbines and motors, turbine engines, micro-compressors and micro-pumps, MEMS, jet engines and other similar machines. More specifically, the present invention relates to rotary machines having a stationary subdividing disc (subdividing means) located in the cavity between the rotor and the housing for the purpose of changing the nature of the flow dynamics and the pressure distribution along the outside of the rotor (between the stationary subdividing means and the rotor), and creating a hydrostatic/hydrodynamic self-pressurized axial/radial bearing as a functional unit consisting of two elements, the subdividing means and the rotor.
Advanced design features for rotary machines are proposed in the U.S. Pat. No. 6,129,507 by Boris Ganelin. Such design features are described for the front cavity and can be used in any one or several stages of a centrifugal pump or compressor. Such features can also be employed in the rear cavity of a rotary machine. Also, such design features as described in any one of the Figures below may be used in any combination with those of the other Figures as described. The disc-shaped stationary subdividing means in the front cavity (referred to as “subdividing means” throughout this description) and the rotor front portion are generally shown in the Figures as perpendicular to the rotor axis for convenience of presentation, while a conical (or curved) gap formed therebetween is preferred for additional radial control of rotor. The bearing elements (restrictive means areas, dam areas, and pre-dam areas) are shown as flat surfaces in the Figures but it should be understood that they can also be curved, wavy or have conical surfaces to produce alternative hydrodynamic/aerodynamic effects.
Such design featured described herein can also be used independently for the design of a self-pressurized hydrodynamic or aerodynamic bearing with excellent stiffness and damping characteristics, either for controlling axial thrust and/or for maintaining the precise axial/radial positioning of a rotating shaft.
2. Description of the Prior Art
Rotary machines are used in a variety of industries. Centrifugal compressor and pumps, turbo-, gas, and jet engines and pumps, axial flow pumps and hydraulic motors are just some examples of rotary machines. A typical single- or multi-staged rotary pump or compressor contains a generic rotor surrounded by a stationary shroud or housing. A primary working part of the rotor is sometimes also called an impeller which typically contains an arrangement of vanes, discs or other components forming a pumping element that transmits its kinetic rotational energy to the pumping fluid. The rest of the description below refers to the turning part of the rotary machine as a rotor.
One known feature of practically all rotary machines is the presence of the axial force (also known as axial thrust), which impacts the dynamic performance of the rotor. Depending on the rotational speed, rotor diameter, fluid dynamics, angular gap leakage flows and many other parameters, the axial thrust may reach such significant levels so as to present a challenge to reliability of the rotary machines operation. Excessive axial load is especially harmful for the axial thrust bearings. Failure of the axial thrust bearing can cause general failure of the machine. Expensive procedures of bearing replacement comprise a significant part of the overall maintenance of rotary machines, especially turbojet engines and similar machines in which access to the axial bearings is quite difficult. The need therefore exists for a device that would reduce or better yet make insignificant axial thrust in a rotary machine in order to improve its reliability and extend the time between repair services, which is one of the objects of the present invention.
It is also known in the art of rotary machines that the level of axial thrust forces depends on the wear state of the rotor seals of the machine. As the seals wear out, the annular gap leakage flow increases, which unfavorably changes the pressure in the cavities between the rotor and the shroud of the rotary machine and typically causes an increase in the axial thrust. That in turn causes higher yet axial loads on the axial thrust bearings and may bring about their premature failure.
The challenge of reducing axial thrust has been long recognized by designers of the rotary machines. A variety of concepts have been proposed in the prior art in an attempt to solve this problem. One of the most popular methods of reducing the axial thrust is the use of a balancing disk or drum. It is typically added to the back of the rotor and placed in its own balancing cavity in such a way that one side of the disk is subjected to high fluid pressure in order to compensate for the axial thrust cumulatively developed in all of the prior stages of the machine. Another method for axial thrust compensation is to increase the fluid pressure in the appropriate cavity of the rotary machine to exert higher pressure on the rotor and therefore to compensate for the axial thrust. Examples of such method include creating additional fluid passages to increase pressure in the desired area of the rotary machine. Another simple method to address the problem of axial thrust is the use of so-called swirl brakes, a plurality of stationary ribs, grooves or cavities located along the housing in the cavity adjacent the rotor, designed to increase the pressure in the desired area.
Another yet method of axial thrust reduction is proposed in U.S. Pat. No. 6,129,507 by B. Ganelin, a co-inventor of the present invention, this patent is incorporated herein by reference in its entirety. As described in one embodiment of '507 patent, an annular stationary disc (subdividing means) is placed in the cavity between the rotating rotor and the housing and combined with a system of vanes at the perimeter of the cavity. The effect of such new elements is to completely alter the hydrodynamic nature of the flow regime in such cavity, increasing the pressure therein. This in turn has a beneficial effect of reducing the axial thrust forces generated by the machine.
Without the new elements described in '507 patent, the flow regime in such cavity between the rotor and the shroud is characterized by:
The pressure in the cavity is lower near the hub (at lower radius) due to the presence of a tangential velocity component of the flow. That component is directed to the hub as it is needed to feed the outward radial flow layer adjacent the rotor shroud. This also explains why the pressure near the hub declines as leakage flow increases through the cavity with worn eye seals, given the increased volume of fluid that must be transported from the periphery to the hub.
With the new elements (stationary annular subdividing means with peripheral vanes) of the above referenced embodiment of '507 patent, the flow regime in such cavity is transformed as follows:
only outward flow existing in the annular space between the subdividing means and the rotating rotor,
only inward flow existing in the annular space between the subdividing means and the shroud wall, and
the peripheral vanes accepting leakage fluid entering through the perimeter annular gap and fluid centrifuged out by the rotating rotor, redirecting it toward the hub in the annular space between the subdividing means and the shroud wall.
The entering leakage flow (with tangential component) and fluid centrifuged by the rotor is efficiently redirected by the peripheral vanes into radial inward flow in the segregated annular space behind the subdividing means to freely supply the hub area with fluid, and therefore not requiring a low pressure area at the hub to attract such fluid. That in turn results in a greater pressure near the hub and so less axial thrust is generated by the machine. Given such transformation of flow regime in the annular cavity adjacent the rotating rotor, a number of rotor-dynamic benefits are achieved, including a significant reduction in potentially destabilizing turbulence, lower sensitivity of rotor to potentially destabilizing leakage flow, improved rotor-dynamic characteristics of rotor seals, isolation of the rotor from potentially destabilizing downstream pressure variations entering through the peripheral annular gap, etc.
In one embodiment, the '507 patent teaches how to reduce axial thrust using an annular subdividing means with peripheral vanes in the front cavity of a centrifugal compressor or pump, but given larger forces (integral of pressure multiplied by radially exposed surface area of rotor shroud) imposed on the back shroud of the rotor, residual axial thrust directed toward the front is still typically greater than desired. The need exists therefore for a device to further reduce axial thrust, which is simple in design, easy to install, low in cost, does not require monitoring and control devices to work properly, and is effective in its function over a wide range of operating parameters of the rotary machine, which is one of the objects of the present invention.
Centrifugal compressors and pumps utilize a thrust bearing at one end of the rotor shaft to adsorb residual axial thrust acting on the rotor and to determine the axial position of the rotor. Given the varying forces acting on the rotor during operation over its useful life, the variations in the thickness of the lubricating film of the thrust bearing, the potential wear of the thrust bearings and the various potential bending modes of the rotor itself, the axial position of the rotor during operation will vary over the life of the machine. Such variations in axial position of the rotor impact various operating parameters of the pump or compressor, reducing potential machine efficiency and most likely negatively impacting rotor-dynamic stability. Significant efforts are made by engineers to minimize such variations in axial position of the rotor during operation. The need exists therefore for a device to further reduce these variations in axial position of the rotor over the life of said the rotary machine, which is another object of the present invention.
In addition, centrifugal compressors and pumps also utilize radial bearings at both ends of the shaft to support the rotor in the radial direction. Thus, given that the radial and axial forces acting on the rotor are generated mid-span (on impellers and its sealing elements), and such forces are compensated for at a location distant from where they are generated, the need exists for a device to counteract/correct any destabilizing forces near the place where they are generated to reduce the amplitude of radial and axial vibrations of the rotor to therefore improve rotor-dynamic stability, to allow closer tolerance seals, to improve efficiency and to improve reliability of the machine. This is yet another object of the present invention.
As discussed in Rotor Dynamics of Centrifugal Compressors in Rotating Stall in Orbit (2001) by Donald E. Bently et. al., most publications relating to high pressure pumps and compressors report two types of rotor vibrational behavior:
high eccentricity and rotor first natural frequency re-excitation, and
sub-synchronous forward precession with rotative speed-dependent frequency.
The former is usually referred to as whip-type behavior, and is normally associated with balance pistons, fluid-film bearings, and labyrinth seals. The latter is called whirl-dependent behavior and can be associated either with fluid-film bearings/seals or with rotating stall (appearance of a low sub-synchronous frequency component in the rotor vibrational spectrum). The motion describing the behavior of the rotor when its geometrical center does not coincide with its center of gravity is called whirl. Precession is the other oscillatory type of motion, which is caused by misalignment of the principal axis of inertia of the rotor disk and the axis of the shaft.
Fluid-induced instability can occur whenever a fluid, either liquid or gas, is trapped in a gap between two concentric cylinders, and one is rotating relative to the other. The situation exists when any part of a rotor is completely surrounded by fluid trapped between the rotor and the stator, for example in fully lubricated (360° lubricated) fluid-film bearings, around impellers in pumps, or in seals. Fluid-induced instability typically manifests itself as a large-amplitude, usually sub-synchronous vibration of a rotor, and it can cause rotor-to-stator rubs on seals, bearings, impellers, or other rotor and stator parts. The vibration can also produce large-amplitude alternating stresses in the rotor, creating a fatigue environment that can result in a shaft crack. Fluid-induced instability is a potentially damaging operating condition that must be avoided.
In The Death of Whirl and Whip, Use of Externally Pressurized Bearings and Seals for Control of Whirl and Whip Instability, published by the Bently Pressurized Bearing Company, reference is made to an equation to estimate the Threshold of Instability, Ω:
Ω=(1/λ)*√{square root over (K/M)}
where λ is the fluid circumferential velocity ratio (a measure of fluid circulation around the rotor, and is indicative of the damping of the system), K is the rotor system spring stiffness and M is the rotor system mass. As presented, if the rotor speed is less than Ω, then the rotor system will be stable. Thus, Ω is indicative of the maximum anticipated operating speed to ensure stability.
Based on the above equation, the Threshold of Instability can be increased by either increasing λ or decreasing K. The value of λ can be influenced by the geometry of the bearing or seal, the rate of end leakage out of the bearing or seal, the eccentricity ratio in the bearing system or seal, and the presence of any pre- or anti-swirl that may exist in the fluid. Fluid-induced instability originating in fluid-film bearings is commonly controlled by bearing designs that break up circumferential flow. Examples of such bearings include tilting pad, lemon bore, elliptical, and pressure dam bearings. λ can also be controlled by anti-swirl injection of fluid into the offending bearing or seal.
Fluid-induced instability can also be reduced or eliminated by increasing the rotor spring stiffness, K. This effort is complicated by the fact that K actually consists of two springs in series, the shaft spring, KS, and the bearing spring, KB. For these two springs connected in series, the stiffness of the combination is given by the following expression:
For any series combination of springs, the stiffness of the combination is always less than the stiffness of the weakest spring. The weak spring controls the combination stiffness. For example, assume that KB is significantly smaller than KS. Thus, KS is much larger than KB, and so the middle equation can be used (KB controls combination stiffness). As KS becomes relatively large, K becomes approximately equal to KB. For this case, the system stiffness, K, can never be higher than KB; in practice it will always be less. A similar argument can be used with the rightmost equation when KB is relatively large compared to KS; the system stiffness will always be lower than KB.
Stiffness of the bearing, KB, is significantly affected by the level of eccentricity of the axis of rotor relative to the axis of the bearing. Assuming that the source of rotor instability is a plain, cylindrical, hydrodynamic bearing, for example an internally pressurized bearing, when the journal is close to the center of the bearing (the eccentricity ratio is small), the bearing stiffness is much lower than the shaft stiffness. In this case, the ratio KB/KS is small, and so the combination stiffness is a little less than KB. In other words, at low eccentricity ratios, the bearing stiffness is the weak stiffness and so it controls the combination stiffness.
On the other hand, when the journal is close to the bearing wall (the eccentricity ratio is near 1), the bearing stiffness is typically much larger than the rotor shaft stiffness. Because of this, the ratio KS/KB is small. Therefore, the rightmost equation above indicates that the combination stiffness is a little less than KS. Thus, at high eccentricity ratios, the shaft stiffness is the weak stiffness, and so it controls the combination stiffness.
Fluid-induced instability begins with the rotor operating relatively close to the center of the bearing. The whirl vibration is usually associated with a rigid body mode of the rotor system. During whirl, the rotor system vibrates at a natural frequency that is controlled by the softer bearing spring stiffness.
Whip is an instability vibration that locks to a more or less constant frequency. The whip vibration is usually associated with a bending mode of the rotor system. In this situation, the journal bearing operates at a high eccentricity ratio, and KB is much larger than KS. So KS is the weakest spring in the system, and it controls the natural frequency of the instability vibration.
To summarize, at low eccentricity ratios, the bearing stiffness controls the rotor system stiffness. Therefore, any changes in bearing stiffness will show up immediately as changes in the overall rotor system spring stiffness, K. On the other hand, at very high eccentricity ratios, the constant shaft stiffness is in control, and the overall rotor system spring stiffness will be approximately independent of changes in bearing stiffness.
The Bently Pressurized Bearing Company suggests using externally pressurized bearings to selectively control bearing stiffness, in an effort to increase rotor combination stiffness. In whirl, the bearing stiffness is the weak stiffness (controlling element) of the system, and so by increasing the externally supplied pressure in the desired bearing (and in the desired radial direction), the bearing stiffness KB increases, and therefore increasing system spring stiffness, K. It is suggested that whirl can be eliminated in this fashion. In whip, the bearing stiffness KB is very high, and the shaft stiffness KS is the weak spring in the system, so increasing bearing stiffness will have no effect on the overall system spring stiffness, K (combination stiffness). Instead, it is suggested to position the Bently externally pressurized bearing mid-span on the rotor to directly increase the stiffness of the shaft, thereby again making the end bearing stiffness the weakest spring (and so the controlling spring), which is the preferred operating mode for stability. The resulting effect is to increase the Threshold of Instability, Q. A major drawback is that this bearing design is externally pressurized, resulting in higher efficiency losses, added complexity, increased cost and lower reliability.
In another example, U.S. Pat. No. 4,243,274 describes a hydrodynamic bearing capable of transmitting radial, thrust and moment loads between an inner load applying member rotatably connected to the bearing utilizing a pair of cylindrical groups of bearing pads about a longitudinal axis of rotation. The pads have movable face portions with compound curved bearing surfaces symmetrically disposed about and along the longitudinal axis. The curved surfaces are mating with similar curved bearing surfaces on a load applying member. The face portions of the bearing pads are supported so that they are swingable about “swing points” located between the axis of rotation of the bearing and the face portions thereof. The bearing pads are operating under the combined influences of friction and load forces exerted thereagainst by the load applying member, so that through hydrodynamic action wedge-shaped lubricant films are generated between the relatively moving bearing surfaces to maintain the surfaces apart while motion is occurring. While U.S. Pat. No. 4,243,274 teaches a hydrodynamic thrust/journal bearing along with the radial control benefits provided by an angular/conical/curved annular gap, it does not benefit from hydrostatic action and its dimensions do not lend to its application in the rotor side cavity area of rotary machines.
In rotary machines, bearings supporting the rotor shaft in the radial direction are placed near the ends of the shaft, and while it is unusual to position bearings mid-span on the shaft, radial stiffness and damping effects provided by some advanced inter-stage shaft seal designs are viewed as helpful in reducing such radial deflection of the rotor during operation. Minimizing the extent of radial deflection (minimum orbit) of the rotating rotor is a consistent goal of engineers. Minimizing the orbit may enable higher rotational speeds to improve productivity, to reduce potential for damage caused by rotor-dynamic instability, to allow smaller clearance seals, to improve efficiency, to improve reliability, etc. The need exists therefore for a device to further reduce said radial deflection (orbit) of the rotor in order to improve the performance of rotary machines, which is yet another object of the present invention.
In addition to the general use in centrifugal pumps, compressors and other turbo machines, the present invention is particularly useful in rotary machines used for water and air supply, for oil and natural gas recovery, refinement and transport, in chemical and food processing industry, for power plants including nuclear power plants, for turbine engines and particularly jet engines as well as in a number of other applications.
A more complete appreciation of the subject matter of the present invention and its various advantages can be realized by reference to the following detailed description which reference is made to the accompanying drawings in which:
A detailed description of the present invention follows with reference to the accompanying drawings in which like elements are indicated by like reference numerals.
In
An important feature shown in
Importantly, an additional peripheral restrictive means (7) is attached (or formed therewith) at the peripheral portion of the disk forming the subdividing means (4) on the side facing the rotating rotor (2). Such peripheral restrictive means (7) functions as a sealing dam for the self-pressurizing hydrodynamic bearing, producing a localized increase in pressure at the front edge (upstream edge) of restrictive means (7), also producing lift and therefore helping to prevent direct contact with the rotating rotor (2). The restrictive means (7) may alternately be placed on the rotating surface of the rotor as well, given similar peripheral radial placement. More than one (or a series of many) restrictive means (7) may be placed on the subdividing means (4) (or rotating rotor (2)) to increase hydrodynamic lift capacity and stability.
Hydrodynamic thrust bearings are known for their simplicity and excellent stiffness and damping characteristics, allowing for precise axial positioning and high rotational speeds. The restoring forces between the two opposing faces increase as the opposing faces approach, preventing therefore their direct contact. Damping characteristics may be modified by arranging the subdividing means (4) (and correspondingly its opposing rotor face) at an angle greater (or less) than 90° to the shaft axis (conical or knee-shaped front rotor). All design elements used with hydrodynamic bearings are potentially beneficial in improving rotor-dynamic stability for designs of the type described here in
Other design elements common for hydrodynamic bearings are potentially beneficial for application with the present invention. In the ring area on the surface of the subdividing means (4) adjacent to ring area of restrictive means (7) and having smaller radius, thin radial slots (such as Rayleigh steps), or spiral grooves, wavy surface, etc. generally referred to herein as radial ribs can be cut into the surface or otherwise formed within the subdividing means (4). Alternatively, protruding radial ribs directed towards axis or canted at an angle may be formed such that the outward radial flow is conditioned by these grooves or ribs immediately prior to passing over the restrictive means (7) to improve lift characteristics. The groove depth is preferably about the same as the height of the restrictive means (7), or smaller (except in cryogenic conditions, where it should be larger given the lower fluid viscosity). The radial length of such smaller radius ring area may be increased (extend further toward the hub) to increase film stiffness. Given the same radial placement, such grooves (and ribs) can be located on the opposing face of the rotor (2) instead of only on the subdividing means (4). Such radial ribs as Rayleigh steps, spiral grooves, wavy surface, protruding ribs, etc. may also be formed into the radial face of the restrictive means (7) that is opposite the front rotor (2). The inner radial edge plane of restrictive means (7) may be perpendicular to subdividing means (4), at an angle or contoured to provide more desirable lift characteristics. The restrictive means (7) may preferentially be made using a softer material (to abrade sacrificially) than the opposing rotor.
Additionally, to increase lift in the region near the periphery of the rotating rotor, the gap between the rotating rotor (2) and the subdividing means (4) may converge slightly with increasing radius. Benefits include improved rotor-dynamic stability, improving reliability.
Given a very small gap (<100 microns) between the rotating rotor (2) and the subdividing means (4), and the significant surface area of the rotating rotor, it is possible to utilize more aggressive lift mechanisms (deeper Rayleigh Steps, spiral grooves, wavy surface, etc.) over a greater area of the subdividing means or rotor to produce additional axial thrust forces, further increasing its load capacity as a self-pressurizing hydrodynamic thrust bearing.
When using a semi-rigid material to make the subdividing means, and its close proximity to the rotating rotor, there is a further potential to provide damping to the rotating rotor through the deflection of (and adsorption by) the semi-rigid subdividing means (4) in response to pressure waves (adsorbing wave energy).
Many design elements of
Many design elements of
Many design elements of
Two additional restrictive means areas are formed on the ring piece (10). A first (axial) restrictive means area is formed between an outer axial face (12) of the rotating rotor (2) and an opposing inner axial face (11) on the subdividing means (4), forming a self-pressurizing hydrodynamic radial journal bearing. A second (radial) restrictive means area is formed between an outer radial face (14) on rotating rotor (2) acting as another dam and an inner radial face 15 of the subdividing means (4), forming an axially-oriented self-pressurizing hydrodynamic thrust bearing. Preferably, to improve axial stiffness, the gap between the face (14) and it opposing face (15) is the same as (or near the same as) the gap between restrictive means (7″) and its opposing face of the subdividing means (4). Preferably, to alter stiffness and damping characteristics, Rayleigh steps (or spiral or radial vanes, or wavy surface, etc.) are cut into the surfaces of restrictive means areas (11) and (15), or their opposing faces as described above. The peripheral surface of subdividing means (4) together with ring piece (10) can be flat (perpendicular to the main flow) as shown by the black line in the drawing, or an additional rounded protruding ring element as shown in the drawing can be formed to improve flow dynamics and to ensure that all of the flow enters the peripheral vanes (8).
In the system depicted on
In this arrangement in
A number of benefits are gained using the proposed arrangement of self-pressurizing hydrodynamic bearing surfaces between the rotating rotor and the subdividing means (4). First, there is the addition of radial control components. There is the hydrostatic radial bearing at restrictive means (12) with opposing face (11), in effect acting as a radial bearing between the main bearings (at the ends of the shaft), thereby providing a means to significantly reduce the orbit of radial oscillations (radial deflections) and to improve radial damping. Another radial control component is added through the use of a conical annular gap between the front rotor (2) and the subdividing means (4), and given the large surface area of this annular gap and its narrowness, the magnitude of this radial component will be substantial. Given the large size/diameter of such radial bearing/rotor surface and the large volume of fluid pumped through the series of annular gaps and dams (the bearing system), and the resulting high stiffness and damping characteristics, such radial bearing capability will result in a significant increase in the first critical speed of the rotor. This is especially beneficial in centrifugal machines with multiple stages utilizing the radial bearing design features suggested in
Second, due to the tortuous path taken by the fluid (a 90° redirection) to restrictive means (12), and then another 90% redirection to restrictive means (14), higher pressure is maintained further along the length of each dam surface (peripheral restrictive means), providing more restoring force (and stiffness) at each dam. Such tortuous path also increases the “squeeze effect” (producing higher pressure at each dam, especially radial dam (12), increasing fluid stiffness) occurring when the opposing surfaces are suddenly forced closer together, therefore protecting the opposing faces against direct contact. As described in an article by Wang (2003), Mixed Lubrication of Coupled Journal-Thrust-Bearing Systems Including Mass Conserving Cavitation, when a journal bearing and a thrust bearing are hydrodynamically coupled, an intensification of the hydrodynamic pressure exists in both bearings, with experimental tests indicating increases in load carrying capacity of 75% and 150% for the journal bearing and the thrust bearing, respectively. In addition, as is known in the art, a controlled eccentricity misalignment angle (non-coincident axis/center of shaft and bearing) improves the load carrying capacity of both the journal and thrust bearings. Wang reported that such effect has an even greater effect on the load carrying capability of the thrust bearing in a hydrodynamically coupled bearing system including that described in the present invention.
Third, the design shown in the Figures converts the rotating front rotor portion with subdividing means in the front cavity into a self-pressurizing axial-thrust bearing having high stiffness and damping characteristics, resulting in more-precise axial positioning (operates within a more narrow envelope) of the rotor. Using the subdividing means with peripheral vanes according to '507 patent, axial thrust can be reduced. The axial thrust does not increase as the eye seals wear off, so for the useful life of a machine residual axial thrust is within a relatively narrow range. That in turn allows minimizing the energy-draining hydrodynamic elements of the present invention (no need to design them to accommodate increased levels of thrust with worn seals). Particularly with the added axial stiffness provided by the self-pressurized bearing of the present invention, axial travel and vibration orbits will be further reduced.
Many design elements of
One purpose of ring element 16 is to direct the returning flow (the second fluid flow, such flow between said subdividing means 4 and said casing 1 and moving toward the center), whereby such flow feeds the entrance to the annular space between rotating front shroud 2 and subdividing means 4. When such returning flow reaches the center region of the front cavity, it is deflected by diagonal face 17 and radial face 18 of ring element 16 and directed toward the annular space between rotating front shroud 2 and subdividing means 4. In effect, the peripheral vanes and annular space between the subdividing means 4 and casing 1, combined with ring element 16, function similar to a conventional interstage return channel of a multistage compressor/pump (but feeding the annular space between rotor shroud 2 and subdividing means 4 vs. feeding the main flow inlet to the impeller). Such diagonal face 17 and radial face 18 may be constructed as one element or as a combination of a number of elements, and may together be formed in other profile designs in efforts to alter flow characteristics, such as more-rounded contouring.
Preferably, a small annular gap is formed between end face 20 of ring element 16 and its opposing face 19 on the rotating impeller 2, functioning as a seal to inhibit leakage to suction. Such small annular gap acts in tandem with the existing eye seal (labyrinth, honeycomb, etc.), in effect forming the first stage of a (now) two-stage seal. Such seal faces are shown as flat annular faces at 90° to the rotating axis of the rotor, but other designs can also be implemented, such as 1) a curved/contoured surface that follows the contour of the existing design of its opposing face, the neck area of the impeller front shroud, 2) the faces at a different such angle to make the leakage path to suction more tortuous, and 3) other seal interface designs well known in the art, such as where one of the two faces is a labyrinth-, honeycomb-(, etc.) type seal, circumferential grooves, pump-out grooves or vanes opposing leakage flow, or where the two opposing faces follow each other in a step profile, similar to faces 7″, 12 and 14 of the impeller shroud 2 with their opposing faces of the subdividing means, to provide a more tortuous path to impede leakage flow.
Although the present invention is described for a specific radial flow centrifugal pump or compressor, it is not limited thereto. Numerous variations and modifications would be readily appreciated by those skilled in the art and are intended to be included in the scope of the invention, which is restricted only by the following claims.
