Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
HYDRAULIC ACTUATOR CIRCUIT WITH CHAMBER BYPASS CONTROLLED BY NEGATIVE LOAD PRESSURE DIRECTING VALVE
Document Type and Number:
WIPO Patent Application WO/1987/004762
Kind Code:
A1
Abstract:
The fluid-powered actuator (12) has a piston (24) with unequal opposed surface areas (A1, A2). Main valve (11) controls fluid flow to or from at least one of the chambers (28, 29) when a load opposes the piston movement. A negative load pressure directing valve (33) allows fluid to flow from contracting chamber to expanding chamber when a load acts in the direction of movement. The accumulator (34) supplies and absorbs fluid to account for differential of flow volumes into and out of chambers (28, 29). The apparatus and its method of operation prevent unnecessary supply flow when the negative load is available to move the piston (24).

Inventors:
GARNJOST KENNETH D (US)
Application Number:
PCT/US1987/000190
Publication Date:
August 13, 1987
Filing Date:
January 29, 1987
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
MOOG INC (US)
International Classes:
F15B9/09; F15B13/04; F15B21/08; B64C13/40; (IPC1-7): F15B1/04; F15B11/08; F15B13/04
Foreign References:
US2390716A1945-12-11
US3163981A1965-01-05
US3596561A1971-08-03
US3854382A1974-12-17
US3878765A1975-04-22
US4031813A1977-06-28
US4058139A1977-11-15
US4222409A1980-09-16
Download PDF:
Claims:
Claims
1. An energyconserving actuator mechanism, comprising: a fluidpowered actuator having one member movable relative to another member and separating opposed chambers, said one member having opposed surfaces of unequal area facing into said chambers; a main valve associated with a fluid source and selectively oper¬ able to control the flow of fluid with respect to at least one of said chambers when a force acting on said one member opposes the desired direction of move¬ ment of said one member; a bypass valve selectively operable to cause fluid to flow from the contracting one of said chambers to the expanding other of said chambers when a force acting on said one member aids the desired direction of movement of said one member; and volume adjustment means to supply or absorb, as appropriate, the differential flow between said chambers when said force aids the desired direction of movement of said one member.
2. An energyconserving actuator mechanism as set forth in claim 1 wherein said one member is a piston and said other member is a cylinder.
3. An energyconserving actuator mechanism as set forth in claim 1 wherein said main valve is controlled electrically.
4. An energyconserving actuator mechanism as set forth in claim 1 wherein said main valve controls the flow of fluid with respect to each of said chambers.
5. An energyconserving actuator mechanism as set forth in claim 1 wherein said bypass valve is controlled electrically.
6. An energyconserving actuator mechanism as set forth in claim 1 wherein the operation of said main and bypass valves is mutually exclusive.
7. An energyconserving actuator mechanism as set forth in claim 6 wherein said main valve has a movable main valve element and wherein said bypass valve has a movable bypass valve element.
8. An energyconserving actuator mechanism as set forth in claim 7 wherein the movements of said main and bypass valve elements are controlled independently of one another.
9. An energyconserving actuator mechanism as set forth in claim 7 wherein said main and bypass valve elements are coupled for movement with one another.
10. An energyconserving actuator mechanism as set forth in claim I wherein said volume adjustment means is only operable with said bypass valve.
11. An energyconserving actuator mechanism as set forth in claim 10 wherein said volume adjustment means is operatively arranged to maintain the pressure in such expanding chamber at a predetermined pressure.
12. An energyconserving actuator mechanism as set forth in claim II wherein said predetermined pressure is the null pressure of the larger of said chambers.
13. An energyconserving actuator mechanism as set forth in claim 12 wherein said null pressure (PIN) s substantially determined according to the equation: where Ps is the pressure of said fluid source and r is the ratio of the crosssec tional areas of such one member smallerarea surface to such one member larg¬ erarea surface.
14. An energyconserving actuator mechanism as set forth in claim 11 wherein said volume adjustment means includes an accumulator.
15. An energyconserving actuator mechanism as set forth in claim 11 wherein said volume adjustment means includes a pressure regulator.
16. An energyconserving actuator mechanism as set forth in claim 1 and further comprising a load polarity sensing valve arranged to sense the po¬ larity of a load applied to said actuator one member and operable to permit flow through said main valve only when said load opposes the desired direction of movement of said one member and to permit flow through said bypass valve only when said load aids the desired direction of movement of said one member.
Description:
"HYDRAULIC ACTUATOR CIRCUIT WITH CHAMBER BYPASS CONTROLLED BY NEGATIVE LOAD PRESSURE DIRECTING VALVE"

Cross-Reference to Related Application

This application is a continuation-in-part of pending U.S. Patent Application Serial No. 06/825,136, filed January 31, 1986, and entitled "Energy- Conserving Servomechanisms".

Technical Field

The present invention relates generally to the field of servo- mechanisms and servosystems, and, more particularly, to various improved en¬ ergy-conserving actuator mechanisms, and the methods practiced thereby, in which fluid is selectively caused to flow automatically from a contracting actu¬ ator chamber to an expanding actuator chamber whenever it is desired to move the actuator in the same direction as an "aiding" external load applied to the mechanism.

Background Art

The piston-and-cylinder forms the basis of many fluid-powered actuators. The piston is conventionally mounted within the cylinder for sealed sliding movement therealong, and separates two opposed end chambers. A valve, such as an electrohydraulic servovalve, is typically arranged to control the flows of fluid (i.e., either liquid or gas) with respect to the two chambers. After repeated operation, and without an external load being applied to the act¬ uator, the equilibrium pressure in each actuator chamber will settle at a nomi¬ nal null pressure. If the opposing piston faces have equal areas (see, e.g., U.S. Pat. No. 2,964,059), and if these chambers are alternately provided with supply pressure and vented to return, the null pressure in each chamber (i.e., the pres- sure in each chamber when the valve is at null) may be about one-half of the supply pressure.

On the other hand, if such piston faces have unequal areas (see, e.g., U.S. Pat. No. 3,023,782), the null pressures will necessarily be unequal. For example, if the piston faces have a 2:1 area ratio, the null pressure acting on the larger-area piston face may be about one-third of the supply pressure,

while the pressure acting on the smaller-area piston face may be about two- thirds of the supply pressure.

In either case, if an external load is applied to the actuator rod, the pressure in one chamber will rise above its null pressure to resist such load. Thereafter, if it is desired to move the rod in a direction such that the load op¬ poses the desired direction of rod movement, the valve is operated to permit pressurized fluid to flow from the source to the above-null chamber, while fluid is permitted to flow from the other chamber to the return. If the piston faces have equal areas facing into both chambers, these flows will be equal to one another. However, if such faces have unequal areas, these flows will be un¬ equal, and the flow from the contracting actuator chamber will be different from the flow into the expanding actuator chamber.

If it is desired to move the actuator in the same direction as the applied load, such that the load is "aiding" with respect to the desired direction of actuator movement, the valve is operated so as to permit flow from the con¬ tracting chamber to return, while fresh pressurized fluid is permitted to flow from the source to the expanding actuator chamber. However, such fresh pres¬ surized fluid does no useful work in displacing such an "aiding" load, since it tends to push the actuator piston in the same direction as the force exerted by the load. Hence, such flow of fresh pressurized fluid from the source into the expanding chamber in the case of an "aiding" load, consumes energy unnecessar¬ ily. For example, in an "active suspension" system for a vehicle, where the sup¬ ply pressure is generated by a pump driven by the engine of the vehicle, it would be highly desirable to reduce the amount of energy consumed by the system.

Disclosure of the Invention

The present invention provides, in one aspect, various forms of an improved energy-conserving actuator mechanism. In another aspect, the invention provides an improved method of operating an actuator mechanism such that, when it is desired to displace the actuator in the same direction as an "aiding" load, fluid will flow from the contracting actuator chamber to the expanding actuator chamber without unnecessarily drawing fresh pressurized fluid from the source.

With parenthetical reference to the embodiment shown in Fig. 2a, for illustrative purposes only, the improved mechanism broadly includes: a fluid-powered actuator (e.g., 12) having one member (e.g., 24) movable rela-

tive to another member (e.g., the cylinder) and separating opposed chambers (e.g., 28,29), the member having opposed surfaces of unequal area (e.g., A > A2) facing into the chambers; a main valve (e.g., 11) associated with a fluid source and selectively operable to control the flow of fluid with respect to at least one of the chambers when a load (i.e., a force) acting on the one member opposes the desired direction of movement thereof; a bypass valve (e.g, 33) se¬ lectively operable to cause fluid to flow from the contracting actuator chamber to the expanding actuator chamber when a force acting on the one member aids the desired direction of movement of the actuator one member; and volume adjustment means (e.g., 34) for either supplying or absorbing, as appropriate, the differential of the flows into, and out of, the respective actuator chambers.

As a specific example, the actuator one member may be a piston, and the other member may be a cylinder. The main and bypass valves may be, and preferably are, electrohydraulic servovalves. The volume adjustment means may be an accumulator, a pressure regulator, or the like, and may be selectively operable with the bypass valve. Moreover, the volume adjustment means may be biased to maintain the pressure in the expanding chamber at some predeter¬ mined pressure, such as the null pressure of the larger actuator chamber. How¬ ever, the foregoing are only species examples of what the generic claims are intended to cover. Hence, the appended claims should not be viewed as being limited to any of these species, unless a clear explicit limitation to that effect appears therein.

In another aspect, the invention provides an improved method of operating a fluid-powered actuator (e.g., 12). The actuator has one member (e.g., a piston) movable relative to another member (e.g., a cylinder) and sep¬ arating opposed chambers (e.g., 28,29). The improved method broadly comprises the steps of: selectively controlling (e.g., as by an electrohydraulic servovalve, or the like) the flow of fluid (e.g., a liquid or a gas) from a source with respect to at least one of the chambers whenever a load acting on the actuator opposes the desired direction of movement of the one member; and selectively causing fluid to flow from the contracting actuator chamber to the expanding actuator chamber whenever a load acting on the one member aids the desired direction of movement of the one member; thereby to prevent an unnecessary quantity of fluid from flowing into the expanding actuator chamber in the case of an aid- ing load.

The actuator one member may have equal- or unequal area sur¬ faces facing into the chambers. In the latter case, the flow from the contract-

ing chamber will be different from the required flow into the expanding cham¬ ber. The differential of such flows may be either supplied or absorbed, as ap¬ propriate, by the source. An excess flow may also be vented, as to the return. The polarity of the load may be determined by sensing the relative pressure(s) in the chambers. Such sensed pressure(s) may be compared with the desired direction of actuator movement to determine whether the load is aiding or op¬ posing.

Accordingly, the general object of this invention is to provide an improved energy-conserving actuator mechanism having unequal actuator areas facing into opposed fluid-containing chambers, in which fluid is automati¬ cally caused to flow from the contracting actuator chamber to the expanding actuator chamber whenever a force acting on the actuator aids the desired di¬ rection of movement thereof.

Another object is to provide such an improved actuator mechan- ism, wherein fluid may controllably flow from a source into the expanding actu¬ ator chamber whenever a load acting on the actuator opposes the desired direc¬ tion of movement thereof.

Another object is to provide such an improved actuator mechan¬ ism wherein the pressure of fluid in the expanding actuator chamber is main- tained at a predetermined pressure when a load acting on the actuator aids the desired direction of movement thereof.

Another object is to provide such an improved actuator mechan¬ ism wherein energy is conserved by having to supply only the differential of the flows into, and out of, the respective actuator chambers in the case of an aiding load.

Still another object is to provide an improved method of operat¬ ing a fluid-powered actuator, regardless of whether the actuator has equal- or unequal-area surfaces facing into the opposed chambers.

These and other objects and advantages will become apparent from the foregoing and ongoing written specification, the drawings, and the ap¬ pended claims.

Brief Description of the Drawings

Fig. la is an elementary schematic of a prior art servoactuator having a four-way electrohydraulic servovalve operatively associated with a double-acting hydraulic actuator having different piston areas facing into the

opposing chambers.

Fig. lb is a graph of force vs. pressure for the servoactuator shown in Fig. la, and demonstrates the manner by which the null pressure in each chamber may be estimated. Fig. 2a is a schematic of a first embodiment of an improved ser¬ voactuator, and shows the use of an accumulator to absorb or provide, as appro¬ priate, the differential flow to the expanding actuator chamber in the case of an "aiding" load applied to the actuator rod.

Fig. 2b is a schematic of a pressure regulator, which may be sub- stituted for the accumulator in Fig. 2a.

Fig. 2c is a schematic of another form of an accumulator, which may be substituted for the accumulator in Fig. 2a.

Fig. 2d is a schematic showing another form of bypass valve, which may be substituted for the bypass valve in Fig. 2a. Fig. 3 is a schematic of a second embodiment of an improved servoactuator.

Fig. 4 is a schematic of a third embodiment of an improved ser¬ voactuator.

Mode(s) of Carrying Out the Invention

At the outset, it should be clearly understood that like reference numerals are intended to identify the same structural elements, portions or sur¬ faces consistently throughout the several drawing figures, as such elements, portions or surfaces may be further described or explained by the entire written specification, of which this detailed description is an integral part. Unless oth- erwise indicated, the drawings are intended to be read (e.g., cross-hatching, arrangement of parts, etc.) together with the specification, and are to be con¬ sidered a portion of the entire "written description" of this invention, as requir¬ ed by 35 U.S.C. §112. As used in the following description, the terms "horizon¬ tal", "vertical", "left", "right", "up" and "down", as well as adjectival and adver- bial derivatives thereof (e.g., "horizontally", "rightwardly", "upwardly", etc.), simply refer to the orientation of the illustrated structure as the particular drawing figure faces the reader. Similarly, the terms "inwardly" and "outward¬ ly" refer to the orientation of a surface relative to its axis of elongation, or axis of rotation, as appropriate. Referring now to the drawings, this invention broadly provides

an improved energy-conserving actuator mechanism for use in association with a fluid-powered actuator having a member provided with unequal areas exposed to the pressures in opposing chambers. Several embodiments of the improve¬ ment are disclosed herein. However, before proceeding, it is felt desirable to review the structure and operation of a known servoactuator.

Prior Art Servoactuator (Figs, la-lb)

Fig. la is a schematic of pertinent portions of an elementary prior art servoactuator, generally indicated at 10, such as representatively shown and described in U.S. Patent No. 2,993,343, the aggregate disclosure of which is hereby incorporated by reference. This known servoactuator broadly included a four-way electrohydraulic servovalve 11 and a double-acting fluid- powered actuator 12.

The servovalve had a three-lobed valve spool 13 slidably mounted within the cylindrical bore 14 of a body 15. A driver 16, such as a torque motor, was arranged to selectively displace the spool either leftwardly or rightwardly off the null position shown in Fig. la, as desired. Three axially-spaced annular- ly-segmented grooves extended radially into the body from the bore. The left groove 18 communicated with a fluid return at a return pressure R, the middle groove 19 communicated with a source of pressurized fluid at a supply pressure P s , and the right groove 20 also communicated with the fluid return. When the spool was in its null position relative to the body (as shown in Fig. la), the right marginal end portion of left lobe 21 just covered left groove 18, middle lobe

22 just covered middle groove 19, and the left marginal end portion of right lobe

23 just covered right groove 20. Thus, all supply and return ports were blocked, and there was no flow through the valve.

However, if the spool was shifted leftwardly off-null, fluid could flow from the annular space between lobes 21,22 to return through a now-un¬ covered left return port (i.e., defined by the extent to which left lobe 21 uncov¬ ered left groove 18), while fluid could flow from the source through a now-un- covered right supply port (i.e., defined by the extent to which middle lobe 22 uncovered middle groove 19) to the space between lobes 22,23. Conversely, if the spool was shifted rightwardly off-null, fluid could flow from the source through a now-uncovered left supply port to the chamber between lobes 21,22, while fluid could flow from the space between lobes 22,23 through a now-uncov- ered right return port to the fluid return.

Valve 11 was called a "four-way" servovalve because there were four active fluid connections: the supply pressure (P s ) connected to body groove 19; the fluid return (R) connected to body grooves 18,20; a first control pressure (C ) available between lobes 22,23; and a second control pressure (C2) available between lobes 21,22. Additional details of such servovalve, such as feedback spring wires and the like, are deemed collateral to a fundamental understanding of the operation of such prior art servoactuator, and have been omitted herein for clarity.

The actuator had a piston 24 slidably mounted within a cylinder 25. An actuator rod 26 had its right marginal end portion secured to the piston, and had an intermediate portion sealingly and slidably penetrating the left end wall of the cylinder. Thus, the piston subdivided the cylinder into a rightward or first chamber 28, and a leftward or second chamber 29. The volumes of these two chambers varied inversely in the sense that as the volume of one chamber increased, the volume of the other chamber decreased. However, the opposing faces of the piston were of unequal area. Specifically, the piston's right face had a relatively-large circular area A j , while its left face had a rela¬ tively-small annular area A2. A conduit 30 communicated the space between the spool middle and right lobes with actuator right chamber 28, and a similar conduit 31 communicated the space between the spool left and middle lobes with actuator left chamber 29. The pressure in first chamber 28 is indicated as being Pj, and the pressure in second chamber 29 is indicated as being P2.

In the absence of any external load applied to rod 26, and ignor¬ ing friction, the piston would be force-balanced and remain at rest when PlAi = P 2 A 2

However, because Ai was greater than A 2 , this necessarily meant that P-^ would be less than P 2 in equilibrium. Areas A and A were fixed or constant, and the above equation could be satisfied for various possible relative magnitudes of Pi and P2. However, repeated leftward and rightward displacement of piston 24 caused the values of Pi and P 2 to vary about certain null pressures in the absence of an external load. These null pressures may be approximated from the graph shown in Fig. lb. In this regard, persons skilled in this art will readily appreciate that pressure Pi in right chamber 28 will act across area Aj to exert a leftward force on the piston, while pressure P 2 in left chamber 29 will act across annular area A 2 to exert an opposing rightward force on the piston. If it is assumed that the fluid return is a sump exposed to atmospheric pres¬ sure, then R = 0 psig [0 Kg/cm 2 ]. Moreover, P and P 2 will vary reciprocally

in the sense that when one is at its maximum, the other is at its minimum. Thus, if the valve spool were to be shifted leftwardly to a hard-over position, such that P = P s , then P 2 = R = 0. Conversely, if the spool were to be shifted rightwardly to a hard-over position such that P 2 = P s , then P = R = 0.

Fig. lb is a plot of force (ordinate) vs. pressure (abscissa) for the opposing fluid forces acting on the piston. Since the leftward force (F ) was equal to the product of Pi and Aj, when P = 0, Fi = 0, this being indicated by point A. However, when Pi = P s , then this leftward force was at its maxi¬ mum (Fι raax ). Or,

F lmax = p s A l , this being indicated by point B. Since Ai was constant, F^ varied linearly with

Pi between these two extremes, and points A and B have been connected by

- a straight-line.

However, when Pi = P s , then P = 0. Conversely, when Pi = 0, then P 2 = P s . If the ratio (r) of area A 2 to A 1 is r = A 2 /Aι then A2 = rA-i.. Hence, when P2 was 0, then the rightward force (F 2 ) acting on the piston was also zero, this being indicated at point C. However, when P 2

= P s , then this rightward force was at its maximum (F max ). Or, F 2max = p 2maχA2 = ^^1

, this being indicated by point D. Here again, since A 2 was constant, F 2 varied linearly with P 2 between these two extremes, and points C and D have also been connected by a straight-line. However, line A-B has a slope of -Ai, and line

C-D has a slope of +A 2 . The slopes of these two curves are opposite one anoth- er because Fj opposes F . Lines A-B and C-D are shown as intersecting at point E. The null pressures (PIN . P 2N) °-f pressures P , P may be determined as follows:

PlN-Al = P2N 2 = P 2NΓAI Or, P 1N = P 2N r

But if,

PlN + p 2N = Ps Then, p 2Nf + P2N = s P 2 N(1 + r) = P s

Or,

P 2N = Ps (1 + r)

But since,

PlN + P2N = Ps Then,

PlN = P S r/(l + r) The values of P N and P 2 N, as a function of the supply pressure (P s ), are sum¬ marized in Table I herebelow for various area ratios (r):

Table I r = A 2 /Aι PIN = P s r/(1 + r) P 2N = P s /U + r)

0.9 0.474P S 0.526P S 0.8 0.444P S 0.556P s

0.7 0.412P S 0.588P S

0.6 0.375P S 0.625P S

0.5 0.333P S 0.667P S

0.4 0.286P S 0.714P S 0.3 0.231P S 0.769P S

0.2 0.167P S 0.833P S

0.1 0.091P S 0.909P S

Thus, this data tabulates the various null pressures (PIN . P2N- -* n the opposed chambers 28,29 for various actuator configurations. For a given actuator con¬ figuration, say, one having a 2:1 ratio between Ai and A2 (i.e., r = 0.5), the null pressure in first chamber 28 would be about 0.333P S , while the null pressure in second chamber 29 would be about 0.667P S . In other words, if P s = 3000 psi [210.93 Kg/cm 2 ], then P 1N = 1000 psi [70.31 Kg/cm 2 ] and P 2 N = 2000 psi [140. 62 Kg/cm 2 ]. If it were now desired to displace the actuator piston leftwardly, spool 13 would be shifted leftwardly. This would cause Pi to increase slightly above its null pressure, and would cause P2 to decrease slightly below its null pressure, such differential pressure therebetween being sufficient to displace the piston leftwardly in the absence of an external load. Such leftward move¬ ment of the piston would continue until spool 13 was returned to its null posi- tion. When this occurred, Pi would drop back down to its null pressure (PI ). and P 2 would rise to its null pressure (P N) to again restore a fluidic force bal¬ ance on the piston.

Suppose now that a rightwardly-acting external force was applied to actuator rod 26 when valve spool 13 was in its null position. Such external load would cause Pi to rise above its null pressure to resist the load. However, because liquid is viewed as being relatively incompressible, the actuator piston would not move and P2 would remain at its null pressure. In short, Pi would

simply rise to whatever pressure was necessary to resist the applied load.

If it were desired to move such load leftwardly, such that the load "opposed" the desired direction of rod movement, spool 13 would be shifted leftwardly off-null to permit fluid to flow from the source through conduit 30 into expanding chamber 28, and to permit fluid to flow from contracting cham¬ ber 29 through conduit 31 to return.

On the other hand, if it were alternatively desired to move the load rightwardly, such that the applied load "aided" the desired direction of act¬ uator movement, then the spool would be shifted rightwardly off-null such that fluid could flow from contracting chamber 28 through conduit 30 to return. However, when this occurred, fresh pressurized fluid would flow from the source through conduit 31 into expanding chamber 29. Such pressurized fluid admitted to the expanding left chamber did no useful work, for displacement of the actuator rod was controlled by metering the flow of fluid from contract- ing first chamber 28 through the right return port (i.e., formed by the extent to which lobe 23 uncovered right groove 20).

Notice that in either case, the flow of fluid from the contracting actuator chamber was unequal to the flow of fluid into the expanding actuator chamber. Thus, if it were desired to move actuator piston 24 rightwardly in the same direction as an "aiding" external load applied to rod 26, then the fiow from right chamber 28 would be greater than the flow into left chamber 29. Conversely, if it were alternatively desired to move piston 24 leftwardly in the same direction as a leftwardly-acting "aiding" load applied to actuator rod 26, then the flow of fluid from left chamber 29 would be less than the flow of fluid into right chamber 28.

The general object of the present invention is to provide an im¬ proved energy-conserving actuator mechanism wherein, in the case of an "aid¬ ing" load applied to the actuator rod, fluid from the contracting actuator cham¬ ber is constrained to flow into the expanding actuator chamber. A volumetric adjustment means, described infra, is used to either supply or absorb, as appro¬ priate, the differential of the flows between these two actuator chambers. Hence, the need for fresh pressurized fluid from the source is held to a practi¬ cal minimum.

First Embodiment (Fig. 2a)

Referring now to Fig. 2a, a first embodiment of an improved act-

uator mechanism, generally indicated at 32, is shown as broadly including a main "four-way" electrohydraulic servovalve, again indicated at 11; a double- acting fluid-powered actuator, again indicated at 12; a bypass electrohydraulic servovalve 33; and an accumulator 34. Since servovalve 11 and actuator 24 have been previously de¬ scribed, the same reference numerals have again been used in Fig. 2a to indicate the corresponding structure. However, conduits 30,31 have been omitted.

The bypass servovalve 33 is schematically shown as including a three-lobed valve spool 35 slidably mounted within a cylindrical bore 36 pro¬ vided in a body 38. A driver 39, such as a torque motor, is operatively arranged to selectively displace bypass spool 35 either leftwardly or rightwardly, as de¬ sired, from the null position shown in Fig. 2a. Four axially-spaced annularly- segmented grooves extend radially into the body. When the spool is in its null position (as shown in Fig. 2a), the right marginal end portion of spool left lobe 40 just covers body left groove 43, the left marginal end portion of spool middle lobe 41 just covers next-rightward groove 44, the right marginal end portion of spool middle lobe 41 just covers next-rightward body groove 45, and the left marginal end portion of spool right lobe 42 just covers body right groove 46. Hence, if the bypass spool is shifted rightwardly off-null, the space between lobes 40,41 will communicate with groove 44, and the space between lobes 41,42 will communicate with groove 46. Conversely, if the spool is shifted leftwardly off-null, the space between lobes 40,41 will communicate with groove 43, and the space between lobes 41,42 will communicate with groove 45. Accumulator 34 is shown as having a stepped piston 48 slidably mounted within a complementarily-configured cylindrical bore provided in a body 49. Specifically, piston 48 is sequentially bounded by (from left to right): a relatively-small diameter circular vertical left face 50, an outwardly-facing horizontal cylindrical surface 51, a leftwardly-facing annular vertical surface 52, an outwardly-facing horizontal cylindrical surface 53, and a relatively-large diameter circular vertical right end face 54. The stepped bore is sequentially bounded by (again from left to right): a rightwardly-facing circular vertical left surface 55, an inwardly-facing horizontal cylindrical surface 56, a right¬ wardly-facing annular vertical surface 58, and an inwardly-facing horizontal cylindrical surface 59 continuing rightwardly therefrom to join a leftwardly- facing circular vertical right surface 60. Piston 48 is mounted within the bore such that piston surfaces 51,53 sealingly and slidably engage bore surfaces 56,

59, respectively. The axial length of the piston is less than the axial spacing between bore end walls 55,60. Hence, the piston right end face 54 has a rela¬ tively-large cross-sectional area A3 arranged to face into a right end chamber 61, while the piston left end face 50 has a relatively-small cross-sectional area A4 arranged to face into a left end chamber 62. Here again, the volumes of these two end chambers 61,62 vary inversely in the sense that as one expands, the other contracts proportionately. However, because areas A3 and A4 are unequal, the volume changes of chambers 61,62 per unit of piston displacement, are unequal. The major elements of the improved actuator mechanism 32 are connected as follows. Conduit 63 communicates the space between main spool lobes 22,23 with the annular space between bypass spool lobes 40,41. A branch conduit 64 communicates conduit 63 with actuator right chamber 28. Another branch conduit 65 communicates conduit 63 with the space between bypass spool lobes 41,42. Conduit 66 communicates the space between main spool lobes 21,22 with bypass body groove 43. A first branch conduit 68 communi¬ cates conduit 66 with bypass body groove 46, and a second branch conduit 69 communicates conduit 66 with actuator left chamber 29. Main valve body grooves 18,20 communicate with the fluid return, which is at a return pressure R, via conduits 70,71, respectively. Conduit 72 communicates main valve mid¬ dle groove 19 with the fluid source, which is at supply pressure P s . Conduit 73 communicates with accumulator right end chamber 61. Conduits 74,75 com¬ municate conduit 73 with bypass body grooves 44,45, respectively. Conduit 74 has a check valve 76 therein to only permit unidirectional flow from conduit 73 to body groove 44, while conduit 75 has a check valve 78 therein to only per¬ mit flow from body groove 45 to conduit 73. Conduit 79 continuously admits supply pressure to the accumulator left end chamber 62. Conduit 80 continu¬ ously vents the variable-volume annular chamber 81 between accumulator sur¬ faces 51,52,58,59 to the return. When it is desired to displace the actuator against an "opposing" external load applied to rod 26, the bypass driver 39 moves the bypass valve spool to its null position (as shown in Fig. 2a), thereby preventing any flow through the bypass valve. In this condition, the flows of fluid to and from actu¬ ator chambers 28,29, as appropriate, are controlled by the main valve in the conventional manner. Thus, if the main valve spool is shifted leftwardly off- null, fluid may flow from the source through conduits 72,63,64 into actuator right chamber 28. At the same time, fluid may flow from actuator left cham-

ber 29 through conduits 69,66,70 to return. Alternatively, if the main valve spool were to be shifted rightwardly off-null, fluid may flow from the source through conduits 72,66,69 to enter actuator left chamber 29, and may flow from actuator right chamber 28 through conduits 64,63,71 to return. In the preferred embodiment, the accumulator is deliberately dimensioned and proportioned so that the pressure P3 in accumulator right end chamber 61 will be maintained substantially equal to the null pressure (PIN) in the actuator right chamber 28. Thus,

P 3 = PI = P s r/(1 + r) But for the accumulator piston to be forced-balanced,

P S A 4 = P3A3 If piston right end face 54 is configured such that A3 = Ai - A 2 , then these equations may be solved to yield

A 4 = r(Aι - A 2 )/(l + r) Of course, if A3 = K(Aj - A 2 ) where K is a constant, and if A4 = Kr(Aι - A 2 )/(l + r), then P3 = P - Thus, P3 is biased to equal P - While this is preferred, it is not invariable. By redimensioning the accumulator, P3 could alternatively be biased to some other predetermined pressure.

Assume now that it is desired to move the actuator in the same direction as an external "aiding" load applied to rod 26. In this situation, the main valve spool is first returned to its null position (as shown in Fig. 2a), and flow control is handed over to the bypass valve.

Assume that P and P were initially at their respective null pressures, and that such "aiding" external load acts rightwardly on the actuator rod. The application of this load will cause Pi to increase above its null pres¬ sure PIN- Since the main valve spool is held in its null position, there will be no flow through the main valve. The bypass valve spool may then be shifted leftwardly off-null to permit fluid to flow from contracting right chamber 28 through conduits 64,63,66,69 to enter expanding left chamber 29. Because Ai > A 2 , the flow (Q ) from contracting chamber 28 will exceed the flow (Q 2 ) into expanding chamber 29. However, the excess or differential flow (i.e., 3 = i - Q 2 ), which by definition is greater than PIN. ma y P ass through conduits 75,73 to enter and expand accumulator chamber 61. Thus, in this situation, the dif¬ ferential flow is absorbed by the accumulator. Assume now that the actuator is again in equilibrium, with Pi and P 2 being at their respective null pressures (i.e., Pi = PI . 2 = P 2N)- -^ a leftward external load is now applied to actuator rod 26, P 2 will increase

above its null pressure (i.e., P 2 > P 2 N . while P will remain at its null pressure (i.e., P = PI )- If is now desired to move the actuator rod leftwardly, so that the external load is "aiding" with respect to the desired direction of rod movement, the bypass valve spool is shifted rightwardly off-null. Hence, fluid may flow from now-contracting actuator left chamber 29 through conduits 69, 68,65,63,64 to enter the now-expanding actuator right chamber 28. However, because A < A j , the flow (Q 2 ) from contracting chamber 29 will be less than that needed to fill expanding chamber 28. As the actuator piston begins to move leftwardly, Pi attempts to fall below PIN > and make-up or supplemental fluid, which is charged to be at pressure PIN > ma y flow from accumulator chamber 61 through conduits 73,74,63.64 to enter expanding actuator chamber 28. Thus, in this situation, the differential flow is supplied by the accumulator. Moreover, since the accumulator is configured to maintain P3 = PIN . the pres¬ sure in actuator chamber 28 is also maintained at its null pressure. While the arrangement shown in Fig. 2a is functionally operation¬ al, it does suffer from an inherent defect, namely, that if one attempts to move a unidirectional external load back and forth, such that the load is alternatively "opposing" and "aiding" the accumulator will quickly move to the end of its stroke, and thereafter be incapable of either absorbing or supplying, as appro- priate, the required differential flow. To solve this problem, a pressure regula¬ tor, such as generally indicated at 82 in Fig. 2 b, may be substituted for the ac¬ cumulator 34.

Referring now to Fig. 2b, pressure regulator 82 is shown as hav¬ ing a two-lobed valve spool 83 slidably mounted within a stepped cylindrical bore provided in a body 84. The spool is sequentially bounded by (from left to right): a relatively-small circular vertical left end face 85, an outwardly-facing horizontal cylindrical surface 86, a leftwardly-facing annular vertical surface 88, an outwardly-facing horizontal cylindrical surface 89, a rightwardly-facing annular vertical surface 90, an outwardly-facing horizontal cylindrical surface 91, a leftwardly-facing annular vertical surface 92, and an outwardly-facing horizontal cylindrical surface 93 continuing rightwardly therefrom to join a rel¬ atively-large area circular vertical right end face 94. Surfaces 88,89,90 and 92,93,94 define the two axially-spaced lobes of the spool. An L-shaped hole 95 communicates the annular space between these lobes with right end face 94.

The bore is sequentially bounded by (again from left to right): a rightwardly-facing circular vertical surface 96, an inwardly-facing horizontal

cylindrical surface 98, a rightwardly-facing annular vertical surface 99, an in¬ wardly-facing horizontal cylindrical surface 100, and a leftwardly-facing verti¬ cal circular surface 101. Spool surface 86 sealingly and slidably engages bore surface 98, and spool surfaces 89,93 sealingly and slidably engage bore surface 100. The axial length of the spool is again less than the axial spacing between bore end walls 96,101. Thus, the spool is slidably mounted within the bore. Moreover, the spool left end face 85, which is of a relatively-small cross-sec¬ tional area A4, faces into a left spool end chamber 102, while the spool right end face 94, which is of a relatively-large cross-sectional area A3, faces into a right spool end chamber 103.

Two axially-spaced annularly-segmented grooves 104,105 extend radially into the body from bore surface 100. When the spool is in its null posi-

tion (as shown in Fig. 2b), the right marginal end portion of the spool left lobe just covers left body groove 104, while the left marginal end portion of the spool right lobe just covers right body groove 105. A conduit 106 provides sup¬ ply pressure P s to left spool end chamber 102. Another conduit 108 communi¬ cates conduit 106 with right body groove 105. Conduit 109 communicates body groove 104 with the fluid return. Conduit 110 communicates conduit 109 with the variable-volume annular chamber 111 defined between surfaces 86,88,99, 100. Because of communicating hole 95, the pressure P3 in spool right end chamber 103 is the same as the pressure in the annular space between the two lobes. If the regulator is substituted for the accumulator, conduit 73 may com¬ municate with right end chamber 103.

Pressure regulator 82 may be readily dimensioned and propor- tioned to maintain P3 at some predetermined pressure, such as PIN- For the spool to be forced-balanced,

P S A 4 = P3A3 But if it is desired that P3 = PIN. and if the spool right end face 94 is config¬ ured such that A3 = A - A 2 , then P S A4 = PlN( i - A 2 )

But since PI = Ps^/d + r),

P S A 4 = P s r(Aι - A 2 )/(l + r) Solving for A4,

A 4 = r(Aι - A 2 )/(l + r) Of course, if A3 = K(Aι - A 2 ) where K is a constant, then the regulator spool will remain balanced if A4 = Kr(A - A 2 )/(l + r).

If P3 falls below PIN, the regulator spool will be shifted right-

wardly off-null to admit fluid from the source through conduit 108 and hole 95 to increase P3, until it again equals P N- Conversely, if P3 attempts to rise above PIN. the spool will be shifted leftwardly off-null to permit fluid to flow from chamber 103 through hole 95 and conduit 109 to return, until P3 again equals PI * Thus, the pressure regulator 82 is appropriately dimensioned and proportioned to maintain P3 = PIN 5 or whatever other predetermined pressure is desired. Hence, if this pressure regulator is substituted for accumulator 34 in Fig. 2a, then Pi will be continuously maintained at its null pressure (i.e., P N)- This, of course, will be independent of the "end-of -stroke" problem in- herent in accumulator 34.

Referring; now to Fig. 2c, a modified accumulator, generally in¬ dicated at 112, might alternatively be substituted for accumulator 34 or pres¬ sure regulator 82. This modified accumulator is shown as having a stepped pis¬ ton 113 slidably mounted within a complementarily-configured chamber provid- ed in a body 114. Specifically, piston 113 has (from left to right): a circular vertical left end face 115, an outwardly-facing horizontal cylindrical surface 116, a leftwardly-facing annular vertical surface 118, and an outwardly-facing horizontal cylindrical surface 119 continuing rightwardly therefrom to join an annular vertical right end face 120. An annular recess, indicated at 121, ex- tends radially into the piston from surface 116 adjacent left end face 115. " An L-shaped hole, indicated at 122, communicates recess 121 with piston right end face 120.

The body chamber is shown as being sequentially bounded by (again from left to right in Fig. 2c): a rightwardly-facing circular vertical left wall 123, an inwardly-facing horizontal cylindrical surface 124, a rightwardly- facing annular vertical surface 125, and an inwardly-facing horizontal cylindri¬ cal surface 126 continuing rightwardly therefrom to join a circular vertical right end face 128. The piston is mounted within this chamber such that piston surfaces 116,119 sealingly and slidably engage chamber surfaces 124,126, re- spectively. An annular groove, indicated at 129, extends radially into the body from chamber surface 124 adjacent left end wall 123. The position of body groove 119 relative to end wall 123 is determined so as to cooperate with the position of piston groove 121 relative to end face 115. A conduit 130 communi¬ cates body groove 129 with the fluid return, which is again indicated as being at return pressure R. Another conduit 131 communicates a suitable source of pressurized fluid, indicated as being at supply pressure P s , with the variable- volume annular chamber 132 defined between surfaces 116,118,125,126.

The piston left end face 115 faces into a chamber 133 which is continuously vented to return via conduit 130. Supply pressure, P s , is continu¬ ously present in intermediate chamber 132 and acts against piston surface 118, which has a cross-sectional area A4. Conduit 73 in Fig. 2a communicates with piston right end chamber 134. Piston right end face 120, which is of cross-sec¬ tional area A3, faces into chamber 134.

Accumulator 112 is so dimensioned and proportioned as to main¬ tain the pressure P3 in chamber 134 at a desired predetermined pressure. For example, if it is desired to maintain P3 at the null pressure of actuator chamber 28 (i.e., P = P 1N ), and if A3 = A x - A 2 , then because PIN = P s r ( + r ) :

A 4 = r(Aι - A 2 )/(l + r) Of course, if A3 = K(Aι - A 2 ) where K is a constant, then the accumulator pis¬ ton will be force-balanced if A4 = Kr(Aι - A 2 )/(l + r). Thus, the accumulator is biased to maintain P3 = P N- If the accumulator is required to absorb the differential flow from chamber 28 to chamber 29, the accumulator piston will move leftwardly relative to body 114. As it approaches the end of its stroke, piston groove 121 will communicate with body groove 130, thereby allowing fluid in chamber 134 to flow through hole 122 and conduit 130 to return. Hence, P3 will fall until it is again substantially equal to P N- Conversely, if the accumulator is re¬ quired to supply the differential flow from chamber 29 to chamber 28, piston 113 will move rightwardly relative to the body. As it approaches the end of its stroke, piston groove 121 will communicate with chamber 132, and fluid may flow from the source through conduit 131, chamber 132 and conduit 122, into chamber 134. Thus, accumulator 112 does not have the "end-of-stroke" limita¬ tions of accumulator 34.

Fig. 2d depicts a modified form of the bypass valve in association with actuator 12. This modified bypass valve, generally indicated at 135, is shown as including a three-lobed valve spool 136 slidably mounted within a cy- lindrical bore provided in a body 138. Spool 136 is shown as having left, middle and right lobes 139,140,141, respectively. The body bore is shown as including an inwardly-facing horizontal cylindrical surface 142, and an annular vertical right end wall 143. A projection 144 extends axially and leftwardly from end wall 143 to provide an abutment stop for limiting further rightward movement of the spool relative to the body. Relative movement between the spool and body is selectively controlled by a driver 145.

Two axially-spaced annularly-segmented grooves 146,148 extend

radially into the body from bore surface 142. When the right end face of the spool abuts projection 144 (as shown in Fig. 2d), the right marginal end portion of left lobe 139 just covers body left groove 146, and the right marginal end portion of middle lobe 140 just covers body right groove 148. Conduit 73 co - municates body groove 148 with either accumulator 34, pressure regulator 82, accumulator 112, or some other device or mechanism for supplying or absorbing, as appropriate, fluid at a predetermined pressure. Conduit 149 communicates body groove 146 with actuator left chamber 29. Conduit 150 continuously com¬ municates the annular space between spool lobes 139,140 with actuator right chamber 28. Conduit 151 continuously communicates the annular space be¬ tween spool lobes 140,141 with conduit 150.

For the purposes of illustration, assume that the predetermined pressure in conduit 73 is the null pressure of chamber 28 (i.e., PIN)- When the bypass spool is moved rightwardly to abut projection 144, lobes 139,140 will cover and close body grooves 146,148, respectively, and there will be no flow through the bypass valve. If it is now desired to permit the actuator rod to move in the same direction as an "aiding" external load applied thereto, driver 145 is selectively operated to displace the bypass valve spool leftwardly relative to the body, while the main valve (not shown) is returned to null. Such displace- ment of the bypass spool causes lobes 139,140 to partially uncover body slot 146,148, and to allow flow through the valve. Specifically, fluid may flow from actuator left chamber 29 through conduit 149, a port defined by the extent to which lobe 139 uncovers body groove 146, and conduit 150, to enter actuator right chamber 28. At the same time, conduit 73 will communicate with conduit 150 through a port defined by the extent to which lobe 140 uncovers body groove 148, and conduit 151.

Hence, if the actuator right chamber is contracting, fluid may flow into the expanding actuator left chamber via conduits 150,149, and may also flow to conduit 73 via conduit 151. In this situation, since the flow (Qi) from actuator chamber 28 will be greater than the flow (Q_) into actuator chamber 29, the differential therebetween (i.e., 3 = Qi - Q 2 ) will flow into conduit 73 and be absorbed by the associated accumulator or regulator. Con¬ versely, if the actuator left chamber is contracting, the flow (Q 2 ) from chamber 29 will be less than the flow ( 3) into right chamber, and the differential of such flows will be supplied by the associated accumulator or regulator.

When the external load is "opposing", the driver returns the by¬ pass valve to spool to the position shown in Fig. 2d, and flow control is handed

over to the main valve, as heretofore described.

Thus, in this first embodiment, when it is desired to move the actuator in the same direction as an "aiding" external load applied thereto, fluid is selectively caused to flow from the contracting actuator chamber to the ex- panding actuator chamber. However, because the opposing faces of the actua¬ tor piston are of unequal area, the flow from the contracting actuator chamber will not equal the flow into the expanding actuator chamber. To this end, the volume adjustment means (e.g., an accumulator or pressure regulator) functions to absorb the excess flow, or to supply a make-up or supplemental flow, as ap- propriate. Moreover, in either case, the differential flow is preferably pressur¬ ized to the magnitude of the null pressure (i.e., P N) °f the larger-area actuator chamber.

Second Embodiment (Fig. 3)

Referring now to Fig. 3, a second embodiment of the improved energy-conserving servoactuator, generally indicated at 152, is shown as broadly including actuator 12, previously described; a main servovalve 153; a bypass valve 154; a load polarity sensing valve 155; and pressure regulator 82, previous¬ ly described.

Since actuator 12 and pressure regulator 82 have been described infra, the reader is referred to the foregoing specification for a description of the structure and operation of such components. In Fig. 3, the position of the actuator has been reversed end-for-end.

The main valve 153 is an electrohydraulic servovalve having a five-lobed spool 156 slidably mounted within a cylindrical bore 158 provided in a body 159. The five axially-spaced lobes of spool 156 are individually indi¬ cated at 160-164, respectively. A driver 165, such as a torque motor, is opera- tively arranged to selectively displace spool 156 either leftwardly or rightward¬ ly, as desired, relative to cylinder bore 158. Six axially-spaced annularly-seg- mented grooves are shown as extending radially into the main valve body from bore surface 158. When spool 156 is in its null position, as shown in Fig. 3, the right margin of left spool 160 just covers left body groove 166; the left margin of next-rightward spool 161 just covers body groove 168; the right margin of spool 161 just covers body groove 169; the left margin of lobe 163 just covers body groove 170; the right margin of lobe 163 just covers body groove 171; and the left margin of rightwardmost lobe 164 just covers body groove 172.

The bypass valve 154 is shown as including a two-lobed spool 173 slidably mounted within a cylindrical bore 174 provided in a body 175. The left and right lobes of spool 173 are indicated at 176 and 178, respectively. Two axially-spaced annularly-segmented grooves extend radially into the body from bore surface 174. When the bypass spool is in its null position, as shown in Fig. 3, the right margin of left lobe 176 just covers left body groove 179, and the left margin of right spool lobe 178 just covers right groove 180. In this embod¬ iment, the bypass spool is indicated (via the dashed line) as being coupled to, and movable with, the main valve spool 156. The polarity sensing valve 155 is shown as including a four-lobed valve spool 181 slidably mounted within a stepped cylindrical bore provided in a body 182. The four lobes of spool 181 are individually identified by the refer-

* ence numerals 183,184,185, and 186, respectively. Lobes 183-185 are of the same relatively-large diameter, while rightwardmost lobe 186 is of a smaller diameter. The chamber within which sensing valve spool 181 is mounted, is shown as sequentially including (from left to right): a rightwardly-facing circu¬ lar vertical left end wall 188, an inwardly-facing horizontal cylindrical surface 189, a leftwardly-facing annular vertical surface 190, and an inwardly-facing horizontal cylindrical surface 191 continuing rightwardly therefrom to join a leftwardly-facing circular vertical right end wall 192. Left and right centering springs 193,194 are arranged in the left and right spool end chambers 195,196, respectively, and function to bias the sensing valve spool 181 to the centered or null position shown in Fig. 3. The left end face 198 of the sensing valve spool has a relatively-large cross-sectional area A5, and the right end face 199 of this spool is indicated as having a smaller area Ag. The ratio of areas A5 to Aβ is in proportion to the ratio of Ai to A2. Or,

A 5 /A 6 = K(Aι/A 2 ) where K is a constant. Six axially-spaced annularly-segmented grooves extend radially into the sensing valve body from surfaces 189,191. Thus, when the sensing valve spool 181 is in its null position, the left marginal end portion of lobe 183 just covers body groove 200; the right marginal end portion of lobe 183 partially covers, but does not close, body groove 201; the right marginal end portion of lobe 184 partially covers, but does not close, body groove 202; the left marginal end portion of lobe 185 partially covers, but does not close, body groove 203; the left marginal end portion of right lobe 186 partially covers, but does not close, body groove 204; and the right marginal end portion of lobe 186 just covers body groove 205.

Conduit 206 provides fluid at supply pressure P s to the annular chamber between sensing valve lobes 184,185. The space between lobes 185,186 communicates with the fluid return via conduit 208. Conduit 209 communicates the space between lobes 183,184 with conduit 208, and thence to return. Con- duit 210 communicates the sensing valve left spool end chamber 195 with the space between main valve lobes 160,161. Conduit 211 communicates sensing valve groove 200 with main valve groove 169. Conduit 212 communicates sens¬ ing valve groove 201 with main valve groove 166. Conduit 213 communicates sensing valve groove 202 with main valve groove 171. Conduit 214 communi- cates sensing valve groove 203 with main valve groove 168. Conduit 215 com¬ municates sensing valve groove 204 with main valve groove 172. Conduit 216 communicates sensing valve groove 205 with main valve groove 170. Conduit 218 communicates the sensing valve right spool end chamber 196 with the space between main valve lobes 163,164. Conduit 219 communicates the space be- tween main valve lobes 163,164 with actuator chamber 29. Conduit 220 com¬ municates the space between main valve lobes 160,161 with actuator chamber 28. Conduit 221 communicates the space between main valve lobes 161,162 with conduit 219. Conduit 222 communicates the space between main valve lobes 162,163 with conduit 220. Conduit 223 communicates conduit 220 with the space between bypass lobes 176,178. Conduit 224 communicates with pres¬ sure regulator right end chamber 103. Conduit 225 communicates bypass valve groove 179 with conduit 224. Conduit 225 has a check valve 226 therein to only permit unidirectional flow from groove 179 to conduit 224. Conduit 228 com¬ municates bypass valve groove 180 with conduit 224. Conduit 228 has a check valve 229 therein to only permit unidirectional flow from conduit 224 to groove 180.

As indicated heretofore, the pressure regulator 82 is so dimen¬ sioned and configured that the pressure P3 in chamber 103 is biased to substan¬ tially equal to the null pressure P of actuator chamber 28. Assume initially that the pressures Pι,P 2 are at their respective null pressures PIN. P 2N- respectively, and, further, that no external load is ap¬ plied to actuator rod 26. In this condition, the actuator piston 24 will be force- balanced (i.e., PINAI = P2NA2) J and will remain at rest. Pressure P 2 is trans¬ mitted through conduits 219,218 and exists in sensing valve right end chamber 196. Similarly, pressure Pi is transmitted via conduits 220,210 and exists in the sensing valve left end chamber 195. However, because the left and right end faces of the sensing valve spool are proportional to actuator areas Aι,A 2 ,

respectively, when Pi = PIN and P 2 = P 2 N> the sensing valve spool will remain in its null position. Assume also that the main valve spool 156 and the coupled bypass valve spool 173 are in their respective null positions.

Assume now that a rightward external load is applied to actuator rod 26. Such load will cause P 2 to increase above P 2 N- AS this occurs, the pressure in sensing valve right end chamber 196 will also rise above P N > and will drive the sensing valve spool leftwardly so that grooves 201,202 will be ful¬ ly uncovered, while groove 204 will be fully covered. In this displaced condition of the sensing valve spool, body groove 171 will communicate via conduits 213, 206 with the fluid source, while flow through conduit 214 will be blocked.

If it is now desired to move the actuator rod leftwardly so that the load "opposes" the desired direction of actuator movement, main valve spool 156 is shifted leftwardly off-null. This permits fluid to flow from the source through conduits 206,213 and 219 to enter actuator chamber 29. At the same time, fluid may flow from actuator chamber 28 through conduits 220,212,209 and 208 to return. Thus, fluid will flow into actuator chamber 29 and from act¬ uator 28, to selectively displace the "opposing" load until the main valve spool is returned to its null position.

If it were alternatively now desired to move the load rightward- ly, so that the applied load would be "aiding" with respect to the desired direc¬ tion of actuator movement, main valve spool 156 would be shifted rightwardly off-null. In this condition, fluid is prevented from flowing from the source to the expanding actuator chamber 29 because conduit 214 is blocked by sensing valve spool lobe 185. Hence, fluid in contracting chamber 29 would be con- strained to flow through conduits 219,218,216,222 and 220 to enter expanding actuator chamber 28. However, because A < Ai the flow (Q ) from con¬ tracting chamber 29 will be less than that ( i) needed to fill expanding cham¬ ber 28. However, such rightward displacement of the main valve spool has also displaced the bypass spool rightwardly, thereby permitting fluid to flow from pressure regulator chamber 103 through conduits 224,225,223 and 220 to enter expanding actuator chamber 28. The flow (Q3) supplied by regulator 82 would equal the needed differential (Qi - Q 2 ).

If the direction of the load were to be reversed, the above-indi¬ cated operation would be modified accordingly. If it were desired to move the actuator piston leftwardly in the same direction as such an "aiding" external load, the excess flow from contracting actuator chamber 28 would be absorbed by the pressure regulator, as previously described.

Thus, in this second embodiment, the sensing valve functions to sense the polarity of an imbalance-causing increased pressure in either actuator chamber, and to selectively enable flow of fluid from the source to the higher- pressure actuator chamber in order that the actuator rod may be moved in a direction which is "opposing" with respect to the direction of the applied load. At the same time, such polar displacement of the sensing valve spool also func¬ tions to block such flow of supply pressure if the load is "aiding" with respect to the desired direction of actuator movement. In this situation, fluid is con¬ strained to flow from the contracting actuator chamber to the expanding actua- tor chamber, while pressure regulator 82 functions to either absorb or supply, as appropriate, the differential flow needed to fill the expanding actuator chamber.

If desired, accumulator 34 or accumulator 112 may be substitut¬ ed for pressure regulator 82.

Third Embodiment (Fig. 4)

Referring now to Fig. 4, a third embodiment of the improved energy-conserving servoactuator, generally indicated at 230, is shown as broadly including actuator 12, previously described; a main servovalve 231; a load polar¬ ity sensing valve 232; and pressure regulator 82, previously described. Here again, since actuator 12 and pressure regulator 82 have been previously described, the reader is referred to the foregoing specification for a description of the structure and operation of such components.

In this embodiment, the main servovalve 231 is shown as broadly including a four-lobed valve spool 233 slidably mounted within the cylindrical bore 234 of a body 235. The individual lobes of spool 233 are indicated at 236 and 238-240, respectively. A driver 241, such as a torque motor or the like, is operatively arranged to selectively displace spool 233 either leftwardly or rightwardly, as desired, relative to bore surface 234.

Six axially-spaced annularly-segmented grooves are shown as extending radially into the main valve body 235 from bore surface 234. When the main valve spool 233 is in its null position (as shown in Fig. 4), the right marginal end portion of lobe 236 just covers groove 242; the left marginal end portion of lobe 238 just covers groove 243; the right marginal end portion of lobe 238 just covers groove 244; the left marginal end portion of lobe 239 just covers groove 245; the right marginal end portion of lobe 239 just covers groove

246; and the left marginal end portion of lobe 240 just covers groove 248.

The sensing valve 232 is shown as including a four-lobed valve spool 249 slidably mounted within a stepped cylindrical chamber provided in a body 250. The individual lobes of spool 249 are indicated at 251-254, respec- tively. A piston-like stem 255 extends axially rightwardly from the right end face of rightwardmost lobe 254. Stem 255 has a circular vertical right end face 256 of a cross-sectional area A7, which is proportional to actuator piston area A . Thus,

A 7 = KA 2 where K is a constant. The leftwardmost lobe 251 has a circular vertical left end face 258 of area Ag. Area Ag is proportional to area A « Thus,

A 8 = KAi where K is the same constant.

The cylindrical chamber in which sensing valve spool 249 is slid- ably mounted, includes (from left to right): a rightwardly-facing circular verti¬ cal left end wall 259, an inwardly-facing horizontal cylindrical surface 260, a leftwardly-facing annular vertical surface 261, and an inwardly-facing horizon¬ tal cylindrical surface 262 continuing rightwardly therefrom to join a circular vertical right end wall 263. Spool 249 is slidably mounted within this enclosed bore such that the cylindrical surfaces of the four lobes slidably engage bore surface 260, with the cylindrical surface of stem 255 sealingly and slidably en¬ gaging bore surface 262. Thus, the sensing valve spool left end face 258 is ar¬ ranged to face into a left spool end chamber 264, while its right end face 256 is arranged to face into a right spool end chamber 265. Springs 266,268 are ar- ranged in chambers 264,265, respectively, and bias the spool to a centered or null position relative to the body. Five axially-spaced annularly-segmented grooves extend radially into sensing valve body 250 from bore surface 260, while another annularly-segmented groove extends radially into the body from bore surface 262. Thus, when spool 249 is in its null position (as shown in Fig. 4), the left marginal end portion of lobe 251 just covers body groove 269; the right marginal end portion of lobe 251 partially covers, but does not close, groove 270; the right marginal end portion of lobe 252 partially covers, but does not close, groove 271; the left marginal end portion of lobe 253 partially covers, but does not close, groove 272; the left marginal end portion of lobe 254 partial- ly covers, but does not close, groove 273; and the right marginal end portion of stem 255 just covers groove 274.

Conduit 275 communicates the space between sensing valve lobes

252,253 with a source of pressurized fluid at supply pressure P s . Conduit 276 communicates the space between sensing valve lobes 251,252 with the fluid re¬ turn at return pressure R. Conduit 278 communicates the space between sens¬ ing valve lobes 253,254 with the fluid return. Conduit 279 communicates the annular chamber 280 between lobe 254 and bore surface 261, with conduit 278, and thence to return. Conduit 281 communicates sensing groove 269 with main valve groove 246. Conduit 282 communicates sensing groove 270 with main valve groove 242. Conduit 283 communicates sensing valve groove 271 with conduit 281. Conduit 284 communicates sensing valve groove 272 with sensing valve groove 274. Conduit 285 communicates sensing valve groove 273 with main valve groove 248. Conduit 286 communicates conduit 284 with main valve groove 245. Conduit 288 communicates the space between main valve lobes 236,238 with actuator left chamber 28. Conduit 289 communicates the space between main valve lobes 239,240 with actuator right chamber 29. Conduit 290 communicates conduit 288 with the space between main valve lobes 236,238. Conduit 291 communicates conduit 288 with the sensing valve left end chamber 264. Conduit 292 communicates conduit 289 with the sensing valve right end chamber 265. Conduit 293 communicates with chamber 103 of pressure regula¬ tor 82. Conduit 294 communicates conduit 293 with main valve groove 243. Conduit 294 has a check valve 295 therein to only permit unidirectional flow from conduit 293 to body groove 243. Conduit 296 communicates conduit 293 with body groove 244. Conduit 296 has a check valve 298 therein, which is ar¬ ranged to only permit unidirectional flow from body groove 244 to conduit 293. Conduit 299 communicates the space between main valve lobes 238,239 with conduit 288.

Assume that the sensing valve spool and the main valve spool are initially in their respective null positions, that the pressures in actuator chambers 28,29 are at their respective null pressures (i.e., P = PIN > P2 = P 2N)» and that the pressure P3 in regulator is configured to maintain P3 equal to PI - Assume now that a rightward external load is applied to actuator rod 26. This load will cause P 2 to rise above P2N- However, Pi will remain at its null pressure. Since P2 > P2N > sensing valve spool 249 will be driven left¬ wardly to fully close grooves 272,273, while fully opening grooves 270,271, and partially opening groove 274. If it were now desired to move actuator rod 26 leftwardly, such that the applied load would be "opposing" with respect to the desired direction of rod movement, main valve spool 233 is shifted leftwardly off-null. Hence, fluid may flow from the source through conduits 275,283,281

and 289 to enter expanding actuator right chamber 29. At the same time, fluid may flow from contracting actuator left chamber 28 through conduits 288,290 and 276 to return.

On the other hand, if it were alternatively desired to move actu- ator rod 26 rightwardly so that the applied load would be "aiding" with respect to the desired direction of rod movement, main valve spool 233 would be shifted rightwardly off-null. In this situation, fluid from contracting actuator right chamber 29 would be constrained to flow through conduits 289,292,284,286,299 and 288 to enter expanding actuator left chamber 28. However, because A2 < A , the flow (Q2) from actuator chamber 29 will be less than (Qi) that needed to flow into actuator chamber 28. To supply this deficiency, pressure regulator 82 provides a differential flow of make-up fluid, at pressure PIN > from chamber 103 through conduits 293,294,290 and 288 to actuator chamber 28.

Assume now that a leftward external load is applied to actuator rod 26 when the sensing and main valve spools are in their respective null posi¬ tions. In this situation, such load will cause Pi to rise above its null pressure (i.e., Pi > PI ) , while P2 will remain at its null pressure (i.e., P2 = P2N)- Hence, the sensing valve spool 249 will be shifted rightwardly off-null to close grooves 270,271, while fully opening grooves 272,273, and partially uncovering groove 269. If it is now desired to move this load rightwardly, such that the applied load is "opposing" with respect to the desired direction of rod move¬ ment, main valve spool 233 is shifted rightwardly off-null. Hence, fluid may flow from the source through conduits 275,284,286,299 and 288 to enter expand¬ ing actuator left chamber 28. At the same time, fluid may flow from contract- ing actuator right chamber 29 through conduits 289,285 and 278 to return.

If it were alternatively desired to move actuator rod 26 left¬ wardly, such that the applied load would be "aiding" with respect to the desired direction of actuator movement, the main valve spool 233 would be shifted left¬ wardly off-null. In this situation, fluid would be constrained to flow from con- tracting actuator left chamber 28 through conduits 288,291,281 and 289 to enter expanding actuator right chamber 29. At the same time, however, the con¬ tracting actuator chamber 28 will communicate with pressure regulator chamber 103 via conduits 288,290,296 and 293. Hence, some of the flow ( i) from the larger-area actuator chamber will flow into the smaller-area actuator chamber (Q2) . with the difference between these two flows (i.e., Q3 = Qi - Q2) flowing into and being absorbed by regulator chamber 103.

Thus, in this embodiment, when it is desired to move the actua-

tor rod in the same direction as an "aiding" external load, fluid is constrained to flow from the contracting actuator chamber to the expanding actuator cham¬ ber, with the differential flow therebetween being either absorbed or provided, as appropriate, by the pressure regulator.

Modifications

The present invention contemplates that many changes and modi¬ fications may be made. The common thread to all disclosed embodiments is that, in the case of an "aiding" load, fluid is constrained to flow from one actua¬ tor chamber to the other actuator chamber, with the differential between these flows (i.e., out of one chamber, and into the other) being either absorbed or pro¬ vided, as appropriate, by some additional device. The various embodiments have been shown and described to illustrate the breadth of this invention.

Therefore, in use, the various forms of the inventive apparatus perform an improved method of operating a fluid-powered actuator having one member movable relative to another member and separating opposed chambers. The actuator one member may have equal- or unequal-area surfaces facing into the chambers.

The improved method broadly comprises the steps of: selectively controlling the flow of fluid from a source to, or from, at least one of the chambers whenever a load acting on the actuator opposes the desired direction of movement of the one member; and selectively causing fluid to flow directly from the contracting actuator chamber to the expanding actuator chamber whenever a load acting on the one member aids the desired direction of move¬ ment of the one member. The method may further include the step of: supplying supple¬ mental fluid from a source to the expanding chamber in the case of an "aiding" load if the volumetric change in the contracting chamber is less than the volu¬ metric change of the expanding chamber, thereby to prevent an unnecessary quantity of fluid from flowing into the expanding actuator chamber in the case of an aiding load. If desired, the method may also include the step of: remov¬ ing or absorbing excess fluid flowing from the contracting chamber in the case of an "aiding" load if the volumetric change of the contracting chamber is greater than the volumetric change of the expanding chamber.

As indicated above, the actuator one member may, for example, be a piston having either equal- or unequal-area surfaces facing into the re-

spective chambers. If the flow from one chamber differs from the flow into the other chamber, the method may include the further step of selectively sup¬ plying or absorbing, as appropriate, the differential flow. Excess flow may be vented or returned to the source as desired. The polarity of the load may be determined by monitoring at least one of the chamber pressures, and comparing such sensed pressure(s) with the desired direction of actuator movement. In this manner, the load may be determined to be either aiding or opposing. Such pressures may be sensed hydro-mechanically, as by the use of a sensing or throt¬ tling valve, or electrically. The various accumulators (i.e., 112, 34) and the pressure regula¬ tor (i.e., 82), while presently preferred, are only species examples of mechan¬ isms for supplying and/or absorbing the differential flow. Accordingly, the "vol¬ ume adjustment means", as used in the appended claims, should not be limited to these specific examples, unless an express limitation to that effect appears expressly therein.

Accordingly, while several preferred embodiments have been shown and described, and various changes and modifications thereof discussed, persons skilled in this art will readily appreciate that various additional changes and modifications may be made without departing from the spirit of this inven- tion, as defined and differentiated by the following claims.