GB2122273A - Minimising the effect of temperature changes upon fluid bearings - Google Patents

Abstract

A hydrostatic journal bearing with two opposed bearing pads in which the typical tendency of the bearing clearance to change due to variation in the viscosity of the operating fluid, brought about by temperature change, is counteracted by causing the sum of the pressures (p1, p2) existing at the two pads (22, 23) to be held to a constant value. Typically this value is equal to that of the pressure (ps) of the supply (27). The supply may be connected to the two pads by way of separate adding valves (40, 29) which may be arranged in parallel with restrictors (41, 28) which are themselves in series with the pressure source (27). Such restrictors may be fixed but may also with advantage be controlled by reference to the pressure existing at their respective pads (22, 23), provided they respond symmetrically to variations of fluid temperature and so give rise to no change in the relative position of the bearing members in response to such variations. Alternatively only a single adding valve may be used, which advantageously may provide a third, reference output pressure (po) supplied to the pads (22, 23) in parallel with their respective basic pressure supplies (p1, p2). <IMAGE>

Description

SPECIFICATION Fluid bearings This invention relates to fluid bearings of the kind in which at least two fluid filled bearing pads mounted in one of the bearing members face in opposed directions and confront the other bearing member so that their opposed reactions against it tend to locate it positively - that is to say, against movement of either sign - in at least one linear direction. Such a fluid bearing will be referred to as a fluid bearing of the kind described.

The invention applies particularly to bearings, for instance journal bearings, in which there is high relative speed between the stationary and moving bearing members. In a typical hydrostatic journal bearing to support a shaft rotating at high speed, opposed bearing pads confronting the shaft are formed at angular intervals and in a common radial plane in the housing surrounding the shaft, and power has to be expended to overcome shearing friction of the bearing fluid in two places in particular. Firstly under the lands, that is to say the border areas immediately surrounding the bearing pads and through which the fluid supplied from those pads must pass as a thin film in order to reach exhaust.Secondly, within the pads themselves where, because of the frictional drag of the moving member, a circulatory motion may be set up so leading to shearing between oppositely moving layers of fluid in contact with each other.

Figure 1 of the accompanying drawings shows a conventional hydrostatic journal bearing in diagrammatic radial section. A horizontal shaft 1 is supported by opposed pads 2,3 formed within a housing 4. Exhaust outlets are indicated at 5. Pads 2, 3 are separately connected to a fluid pressure source 6 by way of resistors 7,8 respectively. Reference p, p1 and p2 represent the fluid pressures at source 6, pad 2 and pad 3 respectively, while Ts, To and TL respectively indicate the fluid temperatures upstream of resistors 7 and 8, in pads 2 and 3 and in the bearing clearances separating lands 9 from shaft 1.The work done to overcome the shearing friction just referred to manifests itself in heat, which will leave Ts substantially unchanged, but will increase To and greatly increase TL. These rises will diminish the viscosity of the fluid within pads 2,3 and under lands 9 so that the resistance offered by the bearing clearances to the escape of fluid from the pads diminishes also. Pressures p1 and p2 therefore fall.This results in a decrease in the design pressure ratios of the bearing, that is to say a decrease in the values of the ratios pl/p, and P2/Ps, so affecting the static and dynamic performance for which the bearing was designed. Afall in P2, in particular may decrease the load carrying capacity of the bearing, and a continued fall will lead to bearing collapse.

It will be apparent that in the known bearing already described, a change in the radial dimensions h1 and h2 of the bearing clearances surrounding between the shaft and the lands of pads 2 and 3 respectively can result either from a change in the vertical load upon the shaft 1, resulting in a change in the resultant vertical force P exerted by the shaft, or from changes in temperatures To and TL as already described. For a simple bearing as shown in Figure 1 the size of the pads and the value of p, will be chosen so that no possible variations in P can result in too great a change in h1 or h2.With the high rotational speeds required of such modern equipment, however, it is harder to keep h1 and h2 as well as pl/p, and pdp, within acceptable limits as To and TL vary for the reasons already described, and the present invention is aimed primarily at removing or at least diminishing the effect that such temperature changes may have upon the clearances and pressure ratios in such a simple bearing.

Before proceeding to a statement of the present invention and a description of apparatus according to it, reference should be made to another known type of fluid bearing to differentiate it from the present invention. UK Patent No. 1027395 shows examples of such a bearing and one of the Figures (Figure 4) of the specification of that patent, which is a diagrammatic section through a longitudinally-sliding bearing and associated fluid circuit components, is now essentially reproduced as Figure 2 of the present application. In the bearing shown in this Figure two opposed pads 105 and 106, formed in the stationary member 104 of the bearing, confront the horizontally sliding moving member 101 and hold it at constant height in the vertical plane. Pad 105 is connected to pressure source 208 by way of conduit 209, a special regulating valve 210 and conduit 205.The resistance offered by valve 210 to the passage of fluid through it varies automatically, and in an inverse sense, relative to the instantaneous pressure within pad 105. Reference 108 indicates an "adding valve" comprising opposed pistons 118 and 119 mounted on a common shaft 120. When fluid at the pressure of source 208 is supplied to inlet 112, and fluid at the pressure of pad 105 supplied to inlet 123, piston 118 - 120 takes up a position such that the pressure of fluid leaving outlet 113 is equal to the difference between the pressures at inlets 112 and 123. Outlet 113 is connected to pad 106.

The text of UK Patent No. 1027395 indicates that the vertical location of the horizontally sliding member 101 relative to stationary member 104 is a function of the hydraulic resistance which the upper and lower vertical clearances between the two members offer to the escaping flow of fluid from pads 105, 106 respectively. The Patent teaches that within a given working range a regulating valve such as item 210 will respond to a change in the relative vertical load between members 101, 104 by adjusting the fluid pressures within pads 105, 106 so that the flow resistances of the upper and lower clearances remain unchanged; the relative vertical positions of the two members therefore remain unchanged also. As the Patent also explains, the effective function of adding valve 108 is to nearly double the working range, so that pressure in pads 105, 106 can each vary within the range 0.1 to 0.8 times the supply pressure p, to counteract the tendency of the two members to change their relative vertical position in response to change of load. But since for a given pad configuration the resistance to the escaping flow of fluid from that pad is:

(where K is a constant, h is the vertical dimension of the clearance beneath the lands of the pad, and jlL is the viscosity of the fluid within that clearance) it follows that for the resistance to remain constant then the viscosity uL must remain constant also.The truth is however that because viscosity is temperaturedependent this will only happen if the temperature of the fluid within the clearance also remains constant. In practice it would, in the apparatus shown in Figure 4 of UK Patent No. 1027395, because that Figure illustrates a bearing in which onlyslightrelativevelocityisto be expected between members 101 and 104 and therefore negligible generation of heat. In such conditions the stated pressure relationship and pads 105 and 106 and with it the object of the invention of UK No. 1027395, namely constant gaps in the clearances adjacent pads 105 and 106 despite variations in bearing load, would have been achieved.However should there have been rapid relative movement between members 101 and 104, as would have been the case if member 101 had been a fast spinning shaft within a journal bearing, then the fluid temperature within pads 105, 106 and especially underneath the surrounding lands would have risen to a value far higher than that of the field at the source 208 or elsewhere in the circuit. In such circumstances, irrespective of whether the bearing load had changed or not, change in temperature within the bearing would have caused the relative positions of the members 101, 104 to change because the supplies of fluid to the pads 105, 106 are asymmetrical, the supply to pad 105 containing the resistance provided by valve 210 but the supply to pad 106 containing no such resistance.If H is the clearance at pad 105, valve 210 ensures that > L/H3 remains constant provided the load between members 101 and 104 remains within the working range of the bearing.

Therefore H cannot remain constant unless also remains constant, and therefore when there is a change in the temperature TL of the fluid in the clearance the dimension H (and the corresponding dimension for lower pad 106) will not stay constant but will instead change to a value that sets up a new equiiibrium position satisfying the new pressures in pads 105, 106 resulting from the change in uL The invention is defined by the claims and will now be described by way of example, with reference to the Figures 3 to 5 of the accompanying drawings which are diagrammatic radial sections through a two-pad hydrostatic journal bearing.

While the invention applies to journal bearings with more than two pads, in which case more regulating valves would typically be required to maintain the appropriate relationship between pad pressure, Figure 3 shows a typical and simple bearing in which a shaft 20, rotatable about a horizontal axis, is supported within a housing 21 by opposed pads 22, 23 surrounded by lands 24, 25 respectively. Fluid under pressure supplied to pads 22, 23 escapes to exhausts 26 by way of the clearances between the shaft 20 and lands 24, 25 respectively. The fluid is supplied to both pads from a common source 27, from which it reaches pad 23 by way of fluid conduit means 23a including a fixed restrictor 28 which may for instance be of either viscous or orifice type.Fluid reaches pad 22, however, by way of fluid conduit means 22a which include no such resistor but which do include an adding valve 29 of known type, similarto valve 108 of Figure 2 and comprising two pistons 30,31 mounted on a common shaft 32. Sealing "0" rings 33 are carried on the rims of pistons 30 and 31, allowing these pistons limited sliding travel over the inner wall of cylindrical housing 34 without breaking sealing contact with it. Movement of the piston assembly 30 - 32 varies the dimensions of the clearance between piston 30 and a fixed ring 35 within the valve. If pressures p, p1 and p2 exist at source 27 and pads 22 and 23 respectively, it will be seen that each of these three pressures is associated with a different section of the interior of adding valve 29.Because the right-hand end of the valve is vented at 36 and the piston assembly 30 - 32 is free to move in response to the pressures to which it is exposed, it will be seen that pressure P1, as supplied from the valve to pad 22, must always be equal to the difference between supply pressure Ps and pressure P2 existing in pad 23, on the downstream side of fixed restrictor 28.

If pads 22 and 23 are of equal area so that their combined area is A, and if the resultant downward force exerted by shaft 20 is P, then the equation (P2 - P1) = P/A (ii) must be fulfilled. Since valve 28 and the supplies of fluid to and from it ensure also that p1 + P2== p5= constant (iii) the pressure p1 must always equal 1/2(p5 - P/A) (iv) and pressure P2 must always equal (p5 + P/A) (v) Because the temperature existing at pads 22, 23 has a like effect upon the resistance offered by conduits 22a and 23a - the former contains no resistor and the latter the fixed resistor 28 - functions (iv) and (v) are independent of the fluid temperature existing in the vicinity of pads 22 and 23 and of any difference between those temperatures and the temperature at source 27. Therefore if P is a positive quantity, the dimension of the clearance between the shaft and lands 24,25 will be independent of fluid temperature, theoretically up to the limiting situation where p2 or p1 becomes zero. Contrast this with the behaviour of the bearing in Figure4 of U K Patent No. 1027395. Because temperature change of the fluid under the lands has different effects upon the fluid supplies to pads 105 and 106, it will lead to change in the clearance gaps adjacent these two pads.

Functions (iv) and (v) are of course not independent of variations in P. Variations in this quantity will therefore result in variations of the clearances beneath lands 25 and 24, even though fluid temperature changes will not. The bearing must therefore be designed so that the changes in clearance brought about by any possible variations in load lie within acceptable limits.

Figure 4 shows a refined version of the bearing of Figure 3 in which parts that appear also in Figure 3 are indicated by the same reference numerals. In Figure 4 the conduit 22a supplying pad 22 also includes a second adding valve 40, the connections of valve 29 to source 27 are rearranged, and a second restrictor 41 is associated with valve 40 just as restrictor 28 is with valve 29, although one of restrictors 28,41 could be omitted, without departing from the invention. The use of two adding valves may improve the performance of the bearing generally, particularly as to stiffness and as to maintaining the sum of p1 and p2 accurately equal to p5.

Stiffness may be improved still further if restrictors 41, 28 are variable instead of constant, each restrictor being linked (by means shown diagrammatically at 46) to its respective pad 22, so that the setting of the restrictor reflects the pressure existing at the pad, provided the two resistors respond symmetrically to changes in the temperature of the fluid in the vicinity of pads 22 and 23. The restrictors may, for example, each be of the diaphragm-controlled kind that is described in UK Patent No 1027395, and that would typically be used for the valve 210 shown in Figure 2 of the present specification.

The alternative version of the invention shown in Figure 5 has many similarities with that shown in Figure 3, and parts that work essentially similarly in the two Figures are indicated by the same reference numberals.

However, the adding valve 47 of Figure 5 differs from the valve 29 of Figure 3 by being divided into two halves by a fixed and impervious annular plate 48, the central hole of which is filled by a sealing gland 49 through which the shaft 32 passes. As well as delivering output pressures p1 and p2 (the sum of which equals p5) like the valve 29 of Figure 3, valve 47 also has two further outlets 50, 51 connected to that part of the valve lying between the plate 48 and the fixed ring 35. The pressure which exists within this compartment, and which is fed to lines 52 and 53 by way of outlets 50 and 51, therefore bears upon neither of pistons 30,31 and does not upset the equation p1 + P2 = p5.It acts however as a stabilising, reference pressure by being supplied to pad 22 by way of line 52 and restrictor 41 in parallel with the basic pressure p1 which reaches that pad by way of line 54, and to pad 23 by way of line 53 and restrictor 28 in parallel with basic pressure P2 which reaches that pad by way of line 55. Conduit means 22a thus now comprises parallel lines 52 and 54, and conduit means 23a comprises parallel lines 53 and 55. Foryetfurther improvement in performance restrictors 41 and 28 may be variable (provided they respond similarly to temperature changes) and may be regulated (by connections 46) by the pressures existing at pads 22 and 23, in the manner already described with reference to Figure 4.

Like the construction shown in Figure 3, that shown in Figure 5 uses only a single adding valve; however, it has the extra advantage that the arrangements of the fluid supplies to the two pads 22, 23 are essentially symmetrical, each such supply being a similar combination of basic and reference pressures (by way of lines 54, 52 and 55, 53 respectively), whereas in Figure 3 pad 22 is fed only with pressure p1 from valve 29 whereas the supply of pressure p2to pad 23 is in parallel with a direct supply from source 27 by way of restrictor 28.

This difference means in practice that the construction of Figure 5 is less susceptible than that of Figure 3 to eccentricity, especially at start-up and in other light load conditions. The performances of the constructions shown in Figures 4 and 5 are similar, the main differences being that the construction of Figure 4 achieves low pressure drop (because each summing valve 40, 29 is in parallel with its associated resistor 41, 28) but at the price of using two summing valves, whereas the construction of Figure 5 uses onlythe single summing valve 47 but the connection between the valve and each pad includes a resistor (41, 28) in series with the valve, which tends to create a higher pressure drop.

Claims (9)

1. Afluid bearing of the kind described including fluid conduits (22a, 23a) connecting the bearing pads (22, 23) to a common fluid source (27), in which at least one of the fluid conduits contains restrictor means (28,41), and including pressure-summing means (29, 40, 47), characterised in that the restrictor means (28, 41) are so arranged that their respective effect upon each of the fluid conduits (22a, 23a) is equally dependent upon the temperature of the fluid in the vicinity of the pads (22,23) and in that the pressure-summing means (29,40,47) are operative to cause the sum of the pressures (pal, p2) existing at the pads to maintain a constant value independent of variation of the temperature of the fluid in the vicinity of the pads.

2. Afluid bearing according to Claim 1 characterisedin that the pressure-summing means include at least one adding value operative to equate the sume of the pad pressures (p1, p2) to the pressure (p5) of the source (27) at all times.

3. A fluid bearing according to Claim 2, characterisedin that the restrictor means comprise a fixed restrictor (28) in one of the fluid conduits (23a) and the other fluid conduit (22a) communicates with its associated bearing pad (22) byway of the pressure-summing means (29).

4. A fluid bearing according to Claim 2, characterisedin that the restrictor means comprise first and second restrictors (41,28), and in that the pressure-summing means comprises a first adding valve (40) arranged in parallel with the first restrictor (41) within one of the fluid conduits (22a) and a second adding value (29) arranged in parallel with the second restrictor (28) in the other fluid conduit (23a).

5. A fluid bearing according to Claim 4 characterisedin that the restrictors (41, 28) are variable and are controlled respectively by reference to the pressures existing at their associated pads (22, 23).

6. A fluid bearing according to Claim 1, characterisedin that the pressure-summing means comprises a single adding valve delivering first and second output pressures (p1, p2) respectively to the two fluid conduits (22a, 23a) and so to the two pads (22, 23).

7. A fluid bearing according to Claim 6, characterisedin that the restrictor means comprises a first restrictor (41) in the first fluid conduit (22a) and a second restrictor (28) in the second fluid conduit (23a) and in that the valve (47, Figure 5) also delivers a third, reference output pressure (p0) which is fed through the first restrictor (41) to the first pad (22) in parallel with the first output pressure (p) and which is fed through the second restrictor (28) to the second pad (23) in parallel with the second output pressure (p2).

8. A fluid bearing according to Claim 7, characterisedin that each restrictor (41,28) is variable and is controlled respectively by reference to the pressure existing at the associated pad (22, 23).

9. Afluid bearing according to Claim 1, substantially as described with reference to the accompanying drawings.