USRE27142E - Taplin ettal fluidic engine fuel control system - Google Patents

Abstract

A FUEL CONTROL SYSTEM FOR METERING FUEL FLOW TO THE COMBUSTION CHAMBERS OF AN ENGINE IS HEREIN DISCLOSED. THE SYSTEM INCORPORATES FLUIDIC ELEMENTS TO CALCULATE OPERATOR DEMAND AND ENGINE DEMAND FOR FUEL AND A PROPORTIONING ELEMENT THEN CONTROLS THE POSITIONING OF A VALVE TO PROVIDE THE AMOUNT OF FUEL CALCULATED TO BE REQUIRED BY THE ENGINE TO SATISFY THE OPERATOR DEMAND.

Description

L. E. TAPLIN ET AL Re. 27,142

June 22, 1971 FLUIDIC ENGINE FUEL CONTROL SYSTEM 7 Sheets-Sheet 25 Original Filed June 25, 1963 EU 779 77/VG June 22, 1971 B, T IJN ETAL Re. 27,142

FLUIDIC ENGINE FUEL CONTROL SYSTEM Original Filod June 25, 1963 7 SheetsSheet 4 FLOW 5" PUMP I- /76 INVENTORS June 22, 1971 5, ETAL Re. 27,142

FLUIDIC ENG-INE FUEL CONTROL SYSTEM Original Filod June 25, 1965 7 Sheets-Sheet 6 INVENTORS June 22, 1971 TAPL|N ETAL Re. 27,142

FLUIDIC ENGINE FUEL CONTROL SYSTEM Orignal Filed June 25, 1963 7 Sheets-Sheet '7 WAL 1-52 1? onrwnsz Je. vwauns E, maunsou JOSEPH F? Mfl0l/BSK/ INVENTORS /9TTOENE Y United States Patent 27,142 FLUIDIC ENGINE FUEL CONTROL SYSTEM Lael B. Taplin, Livonia, Walter F. Datwyler, Jr., Royal Oak, Thomas E. Thompson, Rochester, and Joseph P. Madurski, Royal Oak, Mich., assignors to The Bendix Corporation Original No. 3,392,739, dated July 16, 1968, Ser. No. 290,527, June 25, 1963. Application for reissue June 2, 1969, Ser. No. 835,293

Int. Cl. F02c 9/08 U.S. Cl. 60-39.23 22 Claims Matter enclosed in heavy brackets appears in the original patent but forms 110 part of this reissue specification; matter printed in italics indicates the additions made by reissue.

ABSTRACT OF THE DISCLOSURE A fuel control system for meterng fuel flow to the combuston chambers of an engzne is herein dsclosed. The system ncorporates fluidc elemems t calculate operator demand and engine demand for fuel and a proportom'ng element zhen controls the positionng of a valve to provide the amotmt of fuel calculated z0 be required by the engne t0 satsfy the operator demamd.

The present invention relates to a control described herein in a preferred embodiment as specifically applied to controlling fuel delivery to a gas turbine engine utilizing fluid pulse sensing and computing apparatus.

Recent advances in pure fluid amplifying valves of the strearn interaction type have produced amplifying valves having many inherent advantages when applied to an engine control system and is particularly advantageous for present day high speed aircraft or nuclear applications requiring high reliability in extreme temperature environments or control systemsrequiring tolerance for high radiation levels. Additionally, in such applications weight considerations are of utmost importance whereas said fluid pressure interaction valves are adaptable to be packaged in small structures and as they have no or few moving parts the wear factor is eliminated thus permitting greater freedom in selecting materials on a weight or environment basis.

Accordingly it is an object of the present invention to provide a control utilizing pure fluid amplifiers of the stream iuteraction type.

It is another object if the present invention to adapt s-aid pure fluid .stream interaction amplifiers and principles as useful parameter sensors such as speed, surge and the like.

It is still a further object of the present invention to provide basic computing components operating on a timed-pulse or digital principle to effectively utilize said stream interaction amplifiers to compote control requirements based on sensed and input reference requirenrents.

Other objects and advantages of the present inventon will become apparent on consideration of the description and appended drawings wherein:

FIGURES 1 and 2 are schematic representations of prior art pure fluid amplifiers of the stream interaction yp FIGURE 3 is a functional block diagram of a basic engine speed control loop;

FIGURE 4 is a functional block diagram of an engine control system including the speed control loop of FIG- URE 3 with additional surge limiting functions;

FIGURE 5 is a schematic drawing of a pneumatic taehometer or speed sensor;

FIGURE 6 is a schematic drawing of a pneumate tuning fork oscillator or fixed frequency pulse generator;

Reissued June 22, 1971 FIGURE 7 is a schematic drawing of a pneumatc tuning fork oscillator with mechanically variable frequency pulse output;

FIGURE 8 is a schematic drawing of an error signal computer and digital to analog signal conversion devices;

FIGURE 9 is a schematic drawing of a fuel control subsystem for cor1trolling engine fuel delivery in response to a pneumatic signal;

FIGURE 10 illustrates air flow pattern over an engine compressor blade and impending sta1l detection;

FGURE 11 is a functional block diagram of surge subsystem;

FIGURE 12 is a schematic drawing of a gate valve for seiectively blocking a pulse signal in response to a single blocking input;

FIGURE 13 is a schematic drawing of a gate valve for selectively blocking a pulse signal in response to two (or more than one) blocking signal;

FIGURE 14 is a schematic drawing of a pulse shaper device; and

FIGURE 15 is a drawing of an engine fuel system in corporating the teachings of the present invention.

Two prior art examples of pure fluid amplifiers of the stream interaction type are illustrated schematically in FIGURES 1 and 2.

The jet-on-jet amplifier or valve is shown in FIGURE 1. This concept involves the power jet 2 supplying high pressure fluid which, when unrestrained, strikes the two receiver passages 4 and 6 symmetrically, thus building up equal pressures in each receiver line. If a control jet, such as illustrated by jets 8 or 10, is imposed on the power jet laterally, the combined momentum of the two jets causes a vector change which allows impingement of the combined jet on one hole to be somewhat larger than the other receiver hole depending on the intal momentum of the control jet. The pressure level of the control jet is considerably below that of the power jet and this jet-onjet :amplifier has been demonstrate'd to have power gains ranging from 10:1 to 100:1. Diiferential pressure gains, that is the diterential pressure between receiver passages 4 and 6 ratios to the diflerential pressures between a pair of control jets 8 and 10 can be as high as :1.

Another type of all fluid valve or amplifier is the vortex valve as shown in FIGURE 2. Here a supply jet 12 is brought into a cylindrical cavity defined by the wall 14 radially. A control jet 16 is located at right angles to the supply jet and tangential to the wall 14. The control jet need not necessarly be introduced in the same vicinity as the supply jet, but may be tangentially located around the wall 14 in many places. The exit flow is taken out through a hole 17 located in the center of the device. If the control jet alone is turned on, it is seen that a vortex sheet will be established within the cylindrical chamber. Introducing the supply jet alone will cause a radial flow across the cavity to the exit hole 17. With the supply jet flow established, if the control jet flow is increased, a vortex build-up occurs causing a centrifugal pressure to exist at the supply jet port so that increasing the control et fliw causes the actual supply flow to be decreased. The power gain for this device has been established at around 10:1 for gases and approximately 20:1 for liquids. That is, the flow work required at the control jet to appreciably cut down the flow work at the supply jet is approximately 10% of the initial flow work entering the supply et. The incremental gain associated with the device, that is, the change in supply flow for a small change in control jet flow has been reported as high as 100,000: 1.

A considerable amount of work is going on in basic valves of the types described above and similar varieties which are generically referred to herein as stream interaction fluid amplifiers. The species shown in FIG URE 1 is referred to herein as jet-onjet amplifier whereas the FIGURE 2 species is referred to as a vortex amplifier.

Engine control parameters The present invention as disclosed herein is appiied to a gas turbine engine, particularly to fue1 delivery there- :o, however, as will be apparent some of the basic computing and sensing subcomponents will have a broader application and for these subsystems the present descrip- :ion should be considered as illustrative and not restric- The primary quantity involved in the control of a gas :urbine engine is engine speed or speed of the turbine md/or compressor assembly. The desired level of engine ;peed is selected by the pilot and is introduced into the :ngine control system via a throttle normally located in :he cockpit.

Referring to FIGURE 3, there is shown a functional alock diagram of a basic engine speed control loop wheren a pilot control throttle is represented by the block 18. ['hrottle position is transmitted via mechanical linkage epresented by block 20 to an input transducer 22 which s a control device for converting the throttle position sig- 1a1 to a pneumatic pulse signal. The pneumatic signal rom transducer 22 is applied to the error computer secion 24 for comparison with speed pulse feedback signal :upplied by connection 26 to produce an error signal. rror computer 24 additionally converts the digital pulse ogic to a continuous analog signal suitable for positioning ahysical apparatus and thus is also designated as a digital/ malog converter. The output of computer 24 is supplied the fuel control subsystem 28 (fuel valve) to control he rate of fuel delivery to the engine 30 and thus affect ngine speed. Actual engine speed output is supplied by L mechanical rotation through connection 32 to speed tensor 34 where it is converted to a pneumatic pulse sig- 1al to comprise the feedback quantity 26 and which is daptable for comparison with the output of transducer L2. This system comprises a basic speed control loop vhich provides accurate control of engine speed and is unctionally adapted to utilize the pulse information to :ontrol the fuel control subsystem. In operation, the pilot equests his speed by positioning throttle 18. If the re- 1uested speed dilers from the actual speed supplied by feedback connection 26 an error signal is generated in :omputer 24 which is supplied to the fuel subsystem 28 for changing fuel flow and thus speed in a direction to correspond with that requested.

Whereas speed is the primary controlled quantity of .nterest in a gas turbine engine control, practical behavior )f jet engines requires consideration of engine compres- ;or surging. Surging results in reduced compressor effi- :iency and in potential vibration damage to the compressor structure and must be avoided. Surging is in- Fluenced by fuel flow and can be avoided by appropriate .rnitations placed on fuel delivery in response to sensed surge conditions.

FIGURE 4 shows a functional block diagram of an engine control system including the basic speed control loop af FIGURE 3, wherein corresponding blocks beat the ;ame numerals and the engine 30 is breken down into its main subcomponents of a compressor section 36, a comaustion chamber 38, and a turbine 40. A surge control system is added in FIGURE 4. A surge sensor represented 3y block 42 is provided and connected to the compressor 36 by connection 44 for sensing an impeding surge coniition. If surge is imminent, surge sensor 42 which is conaected at 46 to the computer section 24 is operative to 3verride the speed control and avoid surge by reducing :he error signal within computer 24 and thus reducing fuel. Surge prevention functionally serves to override the speed :ontrol loop during periods when dangerous surge con- :litions exist. Considering the auxiliary nature of the surge ;ystem the speed and surge control systems can be dis- :ussed separately. Since the speed control system is primary, it is discussed first.

Speed control loop The basic speed control loop functionally shown in FIGURE 3, consists of 011I basic subcomponents shown in FIGURES 5 through 9. Considering that one prime object of this invention is to utilize stream interaction type fluid amplifiers, a digital approach employing frequency-modulated fluid pulse trans is employed for the following major reasons:

(1) Analog signals are categorically less suitable for signal transmission where non-linearites, distortion, and noise levels are significant.

(2) Though generally having appreciable non-linearity, most known stream interacton amplifiers have fast switching response time which make them ideal for digtal applications when linearity is of decreased importance. In this application where speed and surge are the primary sensed conditions, digital sensors of extreme accuracy have been achieved with very simple and reliable designs.

(3) In general, with a digital approach, reference speed and surge signals can be generated as frequency signals which are strongly resistant to distortion or masking by noise.

Referring to FIGURE 5, there is shown a pneumatic tachometer or speed sensor which functions to convert a mechanical rotational velocity into a pneumatic pulse signal wherein the pulse frequency varies with speed to accomplish the function indicated by block 34 of FIG- URE 3.

The pneumatic tachometer comprises a supply chamber 58 containing a high pressure source, P a connecting transmission line 60 terminating with a jet nozzle 62 and containing a restrictive orifice or bleed member 64 upstream and in series with nozzle 62 to define a chamber C therebetween. An output line 66 having a bleed 67 is connected to lines 60 at chamber C. A rotatable gear 68 has a series of spaced teeth or flapper valves 70 which rotate in a close proxirnity to nozzle 62. In this device, the gear 68 is connected by shaft 72 to the engine to rotate in proportion to turbine speed. As a tooth 70 goes by nozz1e 62 and restricts the flow, a positive pressure pulse is developed within the chamber C and is transmitted through restriction 67 as an output through line 66. The pulse frequency is f=n w (pulses/sec.) where n is the number of teeth and w is the rotation speed of gear 68 in revolutions per second. This pneumatic tachometer converts rotational velocity to a pneumatic pulse rate or frequency to provide an output pressure pulse train as characterized by reference numeral 74 wherein the time duration between adjacent pulses is proportional to engine speed.

As functionally shown in FIGURE 3, blocks 18, 20 and 22 collectively produce a pneumatic pulse output repsenting a desired engine speed value selected by the pilot or operator. This is accomplished in the present invention by providing a fixed reference or known frequency pulse oscillator or generator with an operator adjustable means for varying the reference frequency with speed demand.

Referring to FIGURE 6, there is shown a pneumatic tuning fork oscillator or pulse generator for producing a known and fixed frequency pneumatic pulse train. The timing fork oscillator comprises a timing fork 76 as the frequency determining element having a pair of tines 78, one of which is arranged in close proximity to a nozzle 80. As with the speed sensor, a high pressure, P contained in supply chamber 82 is connected by transmission line 84, having bleed 86, to nozzle and an output line 88 is connected to line 84 intermediate to bleed 86 and nozzle 80. Tine 78 vibrates at its natural frequency alternately opening and cl osing nozzle 80 causing pressure pulses in output line 88 at its natural frequency or a known reference value. This oscillator is self-exciting when air pressure is supplied by nozzle 80 and requires no addtonal initiation or excitation means.

In FIGURE 7 there is shown a similar pneumati tuning fork oscillator or pulse generator except that means have been provided to vary the reference frequency value in response to a mechanical position as may be applied by a throttle. The basic tuning fork oscillator may be the same as that shown in FIGURE 6 and bears the identical numerals with the addition of a movable wedge or funing slug 90 elamped between tines 78 and having a positionng rod or connection 92 for adjustng the tuning slug longitudinally of the vibrating tines and varying the effective length thereof. Movernent of slug 92 to vary the eflective length of tines 78 alters the vibrating frequency. For example, commercially available tuning forks so adapted have been found to produce satisfactory frequency variations of the order of 3:1 which is an acceptable range of variation for a plots command signal.

The error computer and converter represented by block 24 of FIGURE 3 performs two basic functions which are as follows:

(1) Comparing actual speed pulse signals produced by sensor 34 With requested speed pulse signals from throttle 18, linkage and transducer 22 to developa speed error signal proportional to the difference between actual and requested speed.

(2) Converts digital pulses to a continuous analog signal adaptable to positionng a fuel valve or other physically movable output member.

The error computer and converter is shown in greater detail in FIGURE 8. This circuit consists of a high pressure supply chamber 90, P supplying pneumatic pressure to main supply jet 92. Immediately downstream of supply jet 92 is a deflection chamber 94 which in turn is connected to receiving chamber 96. Near the opposite or right side wall of receiver chamber 96, as viewed in FIGURE 8, are arranged vent passages 98 and 100 connected to a low pressure region such as the atmosphere so that the general pressure level in receiving chamber 96 is low. A primary vent passage 102 has an opening 104 in the right side wall of chamber 96 in alignment with supply nozzle 92 so as to normally -receive the high pressure stream being ejected by the supply nozzle. The main supply stream passes through primary vent passage 102 through bellows chamber 106 and out vent passage 108 to the atmosphere. Spaced symmetrically on opposite sides of vent opening 104 are a pair of output openings 110 and 112 which are connected to output passages 114 and 116 respectively. Output passage 114 communicates with the interior of a first bellows or pressure responsive member 118 contained in bellows chamber 106 and secured to its upper end wall. Output passage 116 is similarly connected to a second bellows 120 secured to the lower end wall of charnber 106 and aligned with bellows 118.

T he tree ends of bellows 118 and 120 are connected by a rod 122 so that rod position is a function of the diflerence in pressures acting on the opposed bellows. An angularly movable output linkage 124 is pinned to rod 122 at 125 for movement therewith. Output passages 114 and 116 are addtionally connected to feedback passages 126 and 128 respectively, each containing a bleed 130 and 132 and terminating at nozzles 134 and 136 exhausting into deflection chamber 94 on opposed sides thereof and generally transverse to the main supply stream flow from nozzle 92.

T wo control passages 138 and 140 terminate with control jets 142 and 144 respectively into deflection chamber 94 generally transverse to the stream flow from main supply jet 92 on opposed sides thereof.

With no control signal supplied by either control passage 138 or 140, a continuous stream of high pressure air will be' ejected by supply nozzle 92 traversing deflection chamber 94 and receiver chamber 96 where it will irnpinge on primary vent opening 104 and be transmitted through passage 102, chamber 106 and vent 108 to the atmosphere. This flow will have no positionng effect on rod 122. To the extent t causes a pressure increase in bellows chamber 106 this pressure acts equally om opposed bellows 118 and 120 and is balanced out. To the extent scattered portions of the main supply stream do not enter primary vent opening 104, they will be either exhausted through vents 98 and or impinge substantially uniformly on symmetrically spaced outlet ports and 112 causing equal and offsetting pressure increases in bellows 118 and If a pressure pulse train control signal is supplied to control passage 138, designated input A, it will be directed by control jet 142 transversely against the main supply stream and deflect the main stream downwardly in the direction of outlet signal port 112 to produce an amplified pulse output in line 116 and bellows 120. At the conclusion of each control input pulse in passage 138, the main supply stream will snap back to its original alignment with primary vent 104. Consequently, the pulse period of the output signal H corresponds with the pulse period of input signal A. A second control pulse, input B, is applied to control passage where it is similarly operative to deflect the main stream into outlet passage 114. When the pulse frequency of inputs A and B are equal, the main supply stream will be deflected an equal number of times per time unit towards outlet passages 114 and 116 causing an equal and offsetting build-up in pressure within bellows 118 and 120. If, however, there is a disparity in pulse frequency between inputs A and B a pressure unbalance in the bellows is created to position rod 122 and linkage 124. For example, should the pulse repetition frequency of input A be 400 c.p.s. while that of input B is 350 c.p.s., for each second, bellows 120 will receive au average of fifty additioual pulses over bellows 118. Bellows 118 and 120 have a relatively large volume and act as accurnulators or integrators wherein the pressure level in each is a function of the pulse frequency received. Thus in the assumed example, the pressure level in bellows 120 will be greater than that in bellows 118 thus positionng rod 122 upwardly and augularly positionng output linkage 124 clockwise. If the pulse frequency of input B were greater than input A, the pressure in bellows 118 would be the greatest, positionng rod 122 downwardly. Since bellows have their own inherent resistance to deformation the degree rod 122 is positioned, is dependent on the magnitude of pressure difference within the bellows which in turn is dependent on the difference in pulse frequency between inputs. Of course, where the bellows deformation resistance is insuflcient or diaphragms are used, springs may be used to establish proportionality.

Thus the pressure diflerence between bellows 118 and 120 to move rod 122 is proportional to the diierence in frequencies or error between inputs A and B. Moreover the device has converted the digital pulse type information to a continuous analog type signal positionng linkage 124.

Feedback means 126 and 128 have been provided in order that a degree of gain control may be obtained. For example, as a pulse input A is applied it diverts the main supply stream to passage 116 whereby a certain portion of the output pulse, depending on the size of bleed 132, will be directed through feedback flow path 128 out nozzle 136 into deflection chamber 94 in opposition to the input flow from nozzle 142. The feedback flow thus subtracts or opposes the control flow and by varying the size of bleed 132, by replacement, the relative quantities may be established to provide a desired gain. The feedback flow in passage 126 operates in a similar manner in opposition to input B.

When feedback passages 126 and 128 have appreciable volume (which may be intentonally added) pressuriza tion of passages 126 and 128 can be delayed with consequent delay in pressurization of nozzles 134 and 136 respectively. In this feedback eonfiguraton, this delayed pressurization results in a frequency variant gain control which provides an output diflerential pressure across bellows 118 and 120 which is not only proportional to dif- 7 erential input (A-B) but also proportional to the time ate of change of the difierental input (A-B).

In FIGURE 9 there is shown a fuel controlling sub- ;ystem 28 for utilizing the angular displacement of linkge 124 of the error computer as an input signal and netering a rate of fuel delivery in response to this input.

Fuel from a source, not shown, is supplied to a rnain fuel passage 150. Low pressure supply fuel is designated P In the upstream portion of main fuel passage there is disposed a high pressure gear pump 152 for pressurizing Euel to a relatively high value designated P Fuel from pump 152 flows rightwardly through passage 150 through orifice or valve seat 154 where it is metered and continues out passage 150 as metered fuel where it is adapted to be ;upplied to the manifold or fuel delivery nozzles of an engine. Metering orifice or valve seat -154 produces a pressure drop or loss so that metered fuel is at a lesser value than P and is designated P The pressure drop 01' 1ead P P is the metering head.

Meterng head P P is maintained at a constant value by a by-pass valve generally indicated at 156. By-pass valve 156 includes a double ported pressure balanced valve 158 controlling P pressure fuel through valve seats 160 and 162 to by-pass conduit 164 which returns fuel to the inlet or low pressure side of pump 152. Valve 156 is controlled by diaphragm 166 peripherally secured to the control housing and secured at its center to rod 168 on which the double valve portions are mounted. P pres 9ure fuel is supplied to the lower side of diaphragm 166 by condut 170 whereas P pressure fuel is supplied to the upper side of conduit 172 so that metering head P P acts on diaphragm 166. An adjustable head spring 174 provides a relatively constant downwardly or valve closing force on diaphragm 166.

The force balance on diaphragm 1-66 established by spring 174 acting in a valve closing direction and the meten'ng head P P acting in a valve opening direction maintains the head across valve seat 154 at a substantially constant value. Should P P tend to increase, the by-pass valve is moved in an opening direction, by passing more fiuel through conduit 164. This decreases the fue1 flow in conduit 150 downstream of the by-pass valve reducing P P to its selected value. Should P P decrease, the reverse action occurs whereby more fuel flows through :onduit 150 and valve seat 154 raising P P in a correctve direction.

Metering valve 176 is operative with the valve seat 154 to control the etfective area of fuel metering orifice. A hydraulic servo piston 178 is seoured to the end of the metering valve and is slidable in a bore in the control housing to define a first control fluid chamber 180 and a second control fluid chamber 182 on opposed piston sides. P pressure fluid trom main conduit 150 is transmitted via passage 184, servo valve chamber 186, rate bleed 188 in branch passage 190, and passage 192 to first control fluid chamber 180 where it acts on one side of piston 178 having the smaller etective area, tending to move valve 176 in a direction to increase eective area and thus rate of fuel delivery. Pressure in chamber 180 is designated P to distinguish trom P fluid upstream of rate bleed 188. A controllable pressure servo fluid (P is supplied to second control fluid chamber 182, from P fluid source in servo valve chamber 186, through servo orifice 194 and passage 196. Chamber 182 is also connected to a low pressure reservoir P through passage 198 having servo bleed 200. P fuel in chamber 182 acts upwardly on piston 178 over the larger surface of piston 178 and in opposition to P fluid in chamber 180. P pressure is controlled by establishng a control pressure drop through servo orifice 194 by means of the pivot input lever 202 which is seoured to the output linkage 124 of the error computer mechanism of FIGURE 8, a portion of which is re-illustrated in FIGURE 9. T he end of input lever 202 is arranged in close proximity to servo orifice 194 whereby the P P pressure drop is controlled by angular movement of lever 202 which thus acts as a servo control valve. A rate feedback force is supplied to lever 202 by means of bellows 204 in chamber 186 which has its movable end pinned at 206 to lever 202. Bellows 204 is fixed at its other end to the control housing and communicates through passage 208 with passage 1-82 downstream of rate bleed 188.

The servo system for positioning metering valve 176 may be termed an integrating system inasmuch as piston 178 will move a distance proportional to the integral with respect to time of the deviation of input lever 202 from its neutral or null position. Operation of the fuel subsystem is as follows:

At a stable or no-movement condition of piston 178 the fluid pressure forces acting on piston 178 are in halance with P having a pressure value a certain fixed percentage less than P corresponding closely with the area ratio on opposed piston sides. Expressed mathematically, and neglectingfluid pressure end loading on valve 176:

where A180 equals the area of piston 178 communicating with P pressure in chamber 180 and A182 is piston area exposed to P fluid in chamber 182.

Re-expressing the above equation:

where K is a constant of less than one representing the area ratio.

There is one position of input lever 202 termed its null position which will establsh the balancing P /P pressure ratio by controlling the P P pressure drop through servo orifice 194. If lever 202 deviates from ths null posi tion as for example in a direction closer to servo orifice 194, the P P pressure drop is increased thus lowering the value of P pressure. This causes a force unbalance across piston 178 causing it to move downwardly. Deviation away from null position by lever 202 causes P to increase moving piston 178 upwardly.

The rate at which piston 178 moves is controlled by the rate fluid can transfer into or out of chamber 180 through rate bleed 188. When piston 178 is not moving there is no fluid through bleed 188 and P equals P In this condition rate feedback bellows204 is ineffective as equal fluid pressure acts both exteriorly and interiorly of the bellows. As piston 17 8 moves it causes fluid to flow through rate bleed 188 establishing a P P (or P P pressure drop proportional to the rate of piston movement. The pressure drop across bleed 188 is also applied to bellows 204 to produce a feedback force on input lever 202 opposing its movement trom null, proportionl to piston velocity.

By means of the rate feedback force applied to input lever 202 in opposition to its movement, the degree of deviation of lever 202 from its null position is caused to be proportional to piston velocity. This is a characteristic of an integrating servo mechanism since if the rate of movement or velocity of piston 178 is proportional to lever deviation then the total piston displacement becomes the integral of lever deviation taken with respect to time since velocity is a time related quantity.

To summarize briefly the operation of the speed loop and combining the devices of FIGURES 5, 7, 8 and 9, a variable pulse frequency reference indicatng desired engine speed is generated by the pneumatic timing fork oscillator of FIGURE 7. The pulse output line 88 may be connected to input line 138 of the error computer of FIG- URE 8 to comprise input A. The pulse frequency output trom line 66 of the pneumatc tachometer of FIGURE 5 may be connected to line 140 of the error computer to comprise input B. The error computer positions linkage -124 proportionately to the pulse frequency error between the desired speed reference of input A and actual speed reference of input B. Lnkage 124 is directly connected to the input lever 202 of the fuel subsystem of FIGURE 9 to cause correcting movement of fuel valve 176 at a.

rate proportional to speed error. As valve 176 changes fuel delivery to the engine, actual speed will change in a corrective direction in response to a fuel change and thereby alter the speed rotating gear 68 of the pneumatic tachometer of FIGURE in a direction to bring the actual speed reference of input B in balance with the requested speed reference of input A. This integrated system is shown in FIGURE which will be discussed at a later point.

Surge system The condition known as surging in a gas turbine engine having a compressor is related to compressor speed, inlet air velocity, temperature and other variable and parameters. Given enough information on engine conditons, surging may be predicted by computation, and appro priate preventing action can be undertaken. Surge prediction by computation requires several input sensors to gather the required input information, and a computing system to process the information. The fact that the computation cannot be programmed with perfect accuracy requires that an appreciable safety margin must be applied Such that anti-surge action (fuel flow reduction) must be made to take effect safely before the actual surge conditions arise. However, When an impending surge is sensed directly as against being computed, the complexity of a computer and numerous sensors can be eliminated and a more efiicient avoidance of surge is possible. Accordingly, the surge system of the present invention utilizes the principle of direct sensing of impending surge and utlizes all pneumatic components to provide an extremely accurate surge control with a minimum of structural complexity.

Surge sensing The onset of surging results in a separation of air flow from the trailing edge of the upper or 10W pressure side of a compressor stator blade indicated by numeral 210 in FIGURES 10A, 10B and 10C. The separation experienced is not unlike that experienced above the wing of a stalled aircraft, and the words stall and surge" are frequently used interchangeably. Refer to FIGURE 10A wherein dotted lines indicate normal air flow over the convex or low pressure surface of a compressor stator blade having an airfoil shape with an upper convex and a lower concave surface; whereas FIGURE 10B wavy lines over the blade convex surface indicate air flow during surge and indicate that the air flow separates from the =blade surface. Since the separation results ma decrease in lift on the upper trailing edge of the blade, pressure in this region is increased. A pressure sensor, indicated by passage 212 in FIGURE 10B, is located at the blade convex surface in the region of separation near the trailing edge and would sense the separation as a step of inereased pressure. Such a pressure signal is adapted for use to indicate surging and initiate appropriate anti-surge control functions.

Unfortunately, surging results in reduced engine efliciency and potential accelerated wear and damage to the engine. For practical reasons it is undesirable to allow an actual surge condition to develop. A method of predicting surge pre-conditions is required. It has been found that separation is imminent in the surge pre-condition state. Therefore, by inducing a transient surge separation on a properly selected blade the proximity to surge may be sensed without actually encountering surge. In the present nvention separation is induced When surge is mminent by applying a short pneumatic pulse through passage 214 which is located toward the leadng edge of blade 210 in comparison to pressure sensing passage 212 and inflow alignment along a chordal blade section. By supplying a short pulse through passage 214 a short transient separation of air flow will occllr if surge is imminent. This short separation will cause an increase of pressure in the trailing edge region Where it will be sensed by passage 212. When surge is not imminent the initiation pulse supplied by passage 214 will not induce separation.

Referring to FIGURE 11 there is shown a block diagram of the surge subsystem and includes a pneumatic pulse generator 216 for supplying an initiation pulse of known frequency, see FIGURE 6 for fixed frequency oscillator. Pulses generated in pneumatic pulse generator 216 are transmitted by passage 214 to a selected compressor statr blade 210. The magntude of the pulses is made such that separation can be induced only When surge is imminent. When a transient separation occurs indicating surge preconditions, the increased pressure developed downstream is sensed by passage 212 and transmitted to a pulse shaper 218 (FIGURE 14). Pulse shaper 218, as will be more fully described at a later point, is triggered by the sensed pulses in passage 212 to produce output pulses in line 220 of uniform amplitude and duraton but of the same frequency of the sensed pulses in passage 212. Pulses developed in shaper 218 are transmitted to the gate device 222, also to be later described, to close a gate in the speed-command-pulse path (for example in path of input A of FIGURE 8) lowering the engine speed and avoiding stall or surge.

Since in the speed control loop, fuel flow is proportiona1 to the pulse frequency of speed command, uel flow can be reduced by gating a number of pulses of input A derived from speed command oscillator (FIGURE 7). The amount of fuel flow reduction for each surge indication signal is related to the number of pulses gated and thus to the time duraton of the gate pulse. The long6r the gate pulse, the more fuel flow and engine speed is reduced for a gven pre-surge separation indication. The gate-pulse time duraton therefore directly influences the gain of the surge control function. In addition, since the rate of command pulses being gated in proportional to speed command, the speed reduction per surge gate pulse is essentially a percentage of command engine speed. For a given surge gatepulse duraton, the amount of speed reduction for a given surge indication pulse will be greater at high speed command levels than at low.

The components of the surge system consist of the pulse generator 216 and surge sensor previously descrbed and gate valve 22 and pulse shaper 222 described as fol lows.

Gate

The basic surge limiting function operates by sensing an impending surge condition and reducing engine speed to avoid the impending surge.

Since speed command pulses always exist during normal engine running, a method of reducing engine speed is to reduce the number of speed command pulses by gating. A closed gate in a pulse transmission line inhibits the flow of pulses beyond the gate, essentially reducing the speed command signal while the gate is closed. In the surge limiting system of the present invention, a pneumatic pulse of specified time duraton is generated When a surge separation is initiated.

An all pneumatic gate valve of the single input type is shown in FIGURE 12. The gate consists of a high pres sure supply chamber 224 supplying a main supply stream of pneumatic fluid via passage 226 to supply jet 228. Supply jet 228 ejects the main supply stream into receiving chamber 230 which it traverses and flows out aligned vent passage 232 to the atmosphere or the like. An output passage 234 is arranged obliquely with respect to vent passage 232 and contains an output receiving port 236 opening into receiving chamber 230 and spaced offset from the main supply stream traversing chamber 230 so that normally the main supply stream does not enter output passage 234, unless, of course, it is deflected downwardly.

A signal input pulse train is supplied to passage 238 having a control jet 240 opening into receiving chamber 230 generally transverse to the main supply stream and offset therefrom in a direction pposite to that of output port 236. So far descri-bed, the device acts as a simple pneumatic amplifier. If a signal pulse train is applied to passage 238 the main supply stream is deflected down into output receiver port 236 for as long as each signal or control pulse exists.

A gate pulse transmission line 242 is supplied having a control jet 244 opening into receiving chamber 230 generally transverse to the main supply stream and closely aligned with control jet 240 on the opposite side of the receiving chamber. When a gate pulse is supplied to line 242, its momentum and direction are such that it prevents input pulses from deflecting the main supply stream for as long as the gate pulse exists. The gating or blocking of one pneumatic pulse by another is thus achieved.

The basic approach illustrated in FIGURE 12 can be used in a gate With multiple gate signal inputs as shown in FIGURE 13. In this configuration a second gate signal transmission line 246 is added having a control jet 248 generally opposed to the signal input control jet 240. Gate pulses supplied by lines 242 or 246 or both wll serve to block an output signal in response to a signal input in line 238.

Pulse shaper The shaping or reshaping of pulses is required whenever an existing pulse form is not proper for an intended use. Thus Wher1 ether a generated pulse does not have optimum shape because of the characteristic of the pulse generator or when a pulse shape has been distorted because of attenuation over a length of transmission line, it is desred that the pulse be reshaped to be of uniform amplitude and duration. In the present invention logic is transrnitted by means of pulse frequency whereby other pulse Characteristics such as amplitude and duration are to be held uniform so as not to introduce errors when pulse averaging is utilized such as for example in the digtal to analog conversion of the error computer of FIGURE 8.

In FIGURE 14 there is shown a pulse shaper for utilizing an existing signal pulse to trigger the generaton of a new output pulse of determined amplitude and dura tion but having the Same frequency as the triggering pulse. T he pulse shaper includes a high pressure supply chamber 250, P connected by transmission line 252 to main supply jet 254. The high pressure supply stream ejected from jet 254 traverses receiving chamber 256 and normally flow out vent passage 258 to the atmosphere or the lke. A triggering pulse which is a distorted signal pulse is supplied by passage 260 havng a control iet 262 ejecting into receiving chamber 256 generally transverse to the main supply stream. Output passage 264 is arranged generally obliquely to vent passage 258 and has a receiving port 266 opening into receiving chamber 256 slghtly offset from the main supply stream in a direction opposite that of control jet 262. By design, a small depression or volume 268 is formed in the sidewall of output passage 264 downstream of receiving port 266. A triggering pulse supplied by passage 260 is ejected by control jet 262 and causes an upward deflection of the main supply stream from vent passage 258 to output passage 264. As the main supply stream flows out output passage 264 its rapid velocity aspirates Fluid from the volume or region provided by depression 268 causing a low pressure region which holds the main supply stream in its deflected condition even after the triggerng pulse has stopped. Thus once deflected, the supply stream due to passage design has the capability of attaching itself to the wall of output passage 264 somewhat analogous to an electrical push-pull switch having a holding coil whereby when once actuated holds in its actuated state until a deactvating signal is supplied. It has been found that the ability of the main stream to attach itself to a wall in a deflected condition requires generally intermediate main stream velocities. If velocity 12 is either excessive or too low, a suflicient degree of turbulence does not exist required to aspirate flud in region 268.

The reset signal for restoring the main supply stream from its deflected state back to vent passage 258 is supplied by a feedback circuit comprised of passage 270 opening into output passage 264 at a spaced distance downstream of region 268. Passage 270 contains a restriction 272 and exhausts into chamber 274. A movable piston 276, which may equivalently be an adjustable diaphragm or bellows, forms one wall of chamber 274 to provide a means for adjustng the chamber volume. Feedback passage 278 having a restriction 280 connects chamber 274 with feedback control jet 282 which ejects into receiving chamber 256 generally transverse to the main supply stream.

The pulse generation and shaping operation proceeds as follows. A11 acceptable trigger pulse is received in passage 260, is ejected trom control jet 262 and deflects the main stream flow into output passage 264. A triggering pulse is acceptable if the rnagntude is sufiicient or large enough to initiate main stream deflection, the detailed form of trigger pulse is not important and generally is anticipated to be in a considerably distortcd or dcgenerated condition. Once deflected the main stream attaches tself to the output passage wall by means of the aspiratng eifect on region 268. Part of the main stream flow is diverted through passage 270 into the feedback path comprised of restrictions 272 and 280 and the volurne of chamber 274. A time delay in the feedback path is induced by the time required to fill the volume of chamber 274 and build the pressure in the feedback line to a sufliciently high value whereby when ejected by feedback control jet 282 it restores or resets the deflected main stream to its original flow path out vent 258, thus stopping the flow Out output passage 264. The output pulse thus produced, started when the trigger pulse was received and 1asted until the delayed feedback signal reset the flow. The output pulse length is proportional to the feedback delay, which may be varied by adjustment of the volume in chamber 274 by means of piston 278.

Referring back to the surge system of FIGURE 11, as a surge separation signal is generated as stall is approached, this separation signal is transrnitted to pulse shaper 218 to provide a trigger pulse input. Pulse shaper 218 supplies an amplified output pulse train to line 220 wherein the pulses have uniform amplitude and duration, but have the same frequency as the separation signal input pulse. The pulse shaper pulse output is then fed to gate val-ve 222 to provide a blocking gate signal input to reduce the speed command signal and thus reduce engine speed. A fully integrated system showing the specific interconnectiom is illustrated in FIGURE 15 and will be later described.

System The overall control system schematic is shown in FIG- URE 15 and illustrates an integrated pulse control system utilizing the components heretofore described to perform the functions discussed.

A gas turbine engine generally indicated by numeral 360 consists of an air intake secton 362; compressor 364; flame tube combustors 366 receiving air trom compressor 364 and fuel frorn manifold 368 through nozzles 370; a turbine 372 drivingly connected to compressor 364 by shaft 374; and a tailpipe secton 376.

Fuel is supplied to manifold 368 from a fuel subsystem 28 corresponding to that shown in FIGURE 9 through metered fuel passage 150. The controlling signal supplied to fuel subsystem 28 is obtained from error and converter device 24 (shown in detail in FIGURE 8), which in turn receives inputs in transmission lines 138 and 140.

As prevously stated the speed control loop is the basic control. Speed command pulse signal is provided by a varable frequency tuning fork oscillator 400 (see FIGURE 7) having a pulse frequency established by tuning slug 402 which is adjustable by throttle lever 404. The pulse train in output line 406 therefore has a frequency representing speed demand, Depending on the length of line 406 and the resulting pulse attenuaton, one or more pulse shapers 408, correspondng to that shown in FIGURE 14, is arranged in the line to restore pulse shape and strength While maintaining speed command frequency. Line 406 terminates at gate valve 410 (FIGURES 12 and 13) which acts as a simple amplfier When no blocking or gating signals are applied thus permitting the passage of the speed command signal te passage 138 and provide one control input to error computer 24.

An actual engine speed pulse signal is produced by tachometer 412 (FIGURE which is driven in proportiou to engine speed by connection 414 to provide a pulse train in passage .416 proportional to engine speed. The actual speed pulse train in passage 416 is fed to pulse shaper 418 to establish uniform pulse shape and is then transmitted through passage 420 to gate valve 422. When no blocking signal is applied to gate valve 422 the actual speed pulse train is amplified and transmitted to passage 140 as a second input to error computer 24 opposing the speed demand signal in passage 138. When actual engine speed equals that requested by positioning throttle 404 the pulse frequencies in passages 138 and 140 Will be balanced and no error signal is produced thus maintaining fuel delivery at its existing rate. If throttle 404 is adjusted, however, to call for either an increase or decrease in engine speed, the frequency of the speed command pulse train will be changed causing an unbal ance in speed demand and actual speed pulse frequencies thus inducing an error. The error operates to postion lever 202 of the fuel subsystem calling for a corrective change in fuel delivery which in turn induces a speed change in the engine. The engine speed change in turn alters the pulse frequency output of tachometer 412 hringing the actual speed frequency signal back into balance with the demand signal.

In event speed change is so rapid so as to approach closely a compressor stall condition, the impending stall sensing system becomes operative to automatically reduce the speed command signal and avoid stall! The surge system more fully described in connection with FIGURES 10 and 11 includes a fixed frequency pmumatic oscillator 424 to supply an initation pulse to passage 426 which is applied to the leading edge of a compressor stator blade 210. More .than -one stall sensing system may be used if desired to cover more of the compressor geometry. If stall is imminent, separation of air flow over the blade occurs producing a separation indicating pulse in passage 428. The separation pulse is fed to pulse shaper providing a uniform pulse shape output in passage 432 corresponding in frequency to that of the oscillator 424 durng separation. The stall separation signal in passage 432 is in turn supplied to gate valve 410 as a blocking or gating signal and is operative to reduce the speed command pulse frequency in passage 138 thus calling for a reduction in engine speed which in turn avoids stall. The degree of reduction in engine speed called for is adjusted by changes in the volume chamber in the pulse shaper 480. A larger volume stretches the pulse thus blocking gate 410 for a longer period of time. When stall is not imrninent, no separation signal is produced and speed command pulses are not gated by the stall system.

It will be understood that various portions of the invention described herein may be utilized separately of other components, or may be oombined with other and diierent components without departing from the teachings con tained herein.

We claim:

1. An engine fuel control comprising: a [pneumatic] fluid pulse generator producing a selectable frequency [pneumatic] output pulse train signal, a throttle member positionable to indicate engine speed demand, said throttle mernber connected to said [pneumatic] fluid pulse generator to vary the frequency of said output pulse train signal in response to speed demand, a [pneumatic] tachometer adapted to be driven in relation to engine speed operative to produce an actual speed [pneumatic] output pulse train signal having a frequency that varies with engine speed, pulse signal error computer means connected to said pulse generator and said [pneumatic] tachometer operative to produce an error output signal that varies with the difierence in frequency between said selectable frequency [pneumatic] output pulse train signal and said actual speed [pneumatic] output pulse train signal, engine fuel control means connected to said error computer means operative to control engine fuel delivery in response to said error output signal.

[2. An engine fuel control comprising: a pneumatic fluid pulse generator producing a selectable frequency pneumatic output pulse train signal; said pneumatic generator including a tuning fork to establsh said selectable -frequency in response to the natural frequency of said tuning fork; said tuning fork including a pair of tines and a movable tuning slug clamped between said tines; said tuning slug being movable to vary the natural frequency of said tuning fork and thereby vary said selectable frequency signal; an engine control throttle member connected to said tuning slug operative to vary said selectable frequency signal in response to throttle demand; a pneu matic tachometer adapted to be driven in relation to engine speed operative to produce an actual speed pneumatic output pulse train signal having a frequency that varies with engine speed, pulse signal error computer means connected to said pulse generator and said pneumatic tachometer operative to produce an error output signal that varies with the dfference in frequency between said selectable frequency pneumatic output pulse train signal and said actual speed pneumatic output pulse train signal, engine fuel control means connected to said error computer means operative to control engine fuel delivery in response to said error output signal] [3. An engine speed control comprising: a pneumatic fluid pulse generator producing a selectable frequency pneumatic output pulse train signal; said pneumatic generator including a tuning fork to establish said selectable frequency in response to the natural frequency of said tuning fork; said tuning fork including a pair of tines and a movable tuning slug clamped between said tines; said tuning slug being movable to vary the natural frequency of said tuning fork and thereby vary said selectable frequency signal; an engine control throttle member connected to said tuning slug operative to vary said selectable frequency signal in response to throttle demand; a pneumatic tachometer including a rotatable gear adapted to be driven in relation to engine speed, said rotatable gear having a tooth projection formed thereon operative to 1nduce a pulse train signal having a frequency that varies with the speed of rotation of said gear, pulse signal error computer means connected to said pulse generator and said pneumatic tachometer operative to produce an error output signal that varies with the difierence in frequency between selectable frequency pneumatic output pulse train signal and said actual speed pneumatic output pulse train signal, and means for controlling engine speed connected to said error computer means operative to control engine speed in response to said error output signal.]

4. An engine speed control comprsing: a control throttle member; a [Pneumatic] fluid pulse signal generator connected to said throttle member operative to produce a selectable frequency [pneumatic1 pulse output signal that varies with throttle demand setting; a pneumatic] tachometer responsive to engine speed operative to produce an actual speed [pneumatic] output signal having a frequency that varies with engine speed; error computer means having a high pressure [pneumatc] source, a main supply jet connected to said source and ejecting a man fluid stream, and first and second output receiver openings disposed on opposed sides of said main fluid stream at a spaced distance trom said main supply jet for recevng said rnaiu supply stream when undeflected; a first passage iuter-connecting said [pneumatc] fluid pulse signal generator and said error computer; said first passage including a first control jet within said error computer arranged ntermediate said main jet and said first and second output receiver openings and transverse to said main stream for deflecting said rnan stream towards said first output receiver opening on ejecting a pulse received from said [pneumatic] generator; a second passage nterconnecting said [pneumatc] tachometer and said error computer; said second passage including a second control jet withn said error computer arranged intermediate said main jet and said first and second output receiver openngs and transverse t said main stream; said second control jet arranged on an opposed side of said main stream frorn said first control jet operative to deflect said main stream towards said second output receiver opening on ejecting a pulse received frorn said [pneumatic] tachometer; pressure responsive means connected to said first and second output receiver openings, said pressure responsive means being movable in proporton to frequency difference of pulses receved in said first and second receiver openings, and engine speed control means connected to said pressure responsive means adapted to control engine speed in response to movement of said pressure responsive means.

5. An engine speed control comprising: a manually adjustable [pneumatic] fluid pulse generator producing a selectable refcrence frequency output pulse train signal, a [pneumaticj] tachometer adapted to be driven in proportion to engine speed operative to produce an actual speed [pneumatic] output pulse train signal having a frequency that varies with engine speed, and engine speed control means connected to said pulse generator and said [pneumatic] tachometer operative to control engine speed in response to the error between said reference frequency and said frequency that vares with engine speed.

6. The fuel control as claim in claim 1 wherein: said generator ncludes a timing fork to establish said selectable frequency in response to the natural jrequency of said inning f0rk, said timing fork including a pair of tnes and a movable timing slag clamped between said tines, said timing slag movable to vary the natural frequency of said tuning fork and thereby vary said selectable frequency signal, and said throttle member comprises an engine control throttle member connected t0 said timing slag operative t0 vary said selectable frequency in response to throt tle demand.

7. T he iuel control as claimea' in claim 6 wherein: said tachometer inclua'es a rotable gear adapted to be driven in relation to engine speed, said rotatable gear havng a tooth projection formed thereon operatve to induce a pulse train signal having a frequency that varies with the speed of rotation of said gear.

8. A fuel system for engines comprising: a discharge means adapted t0 clischarge fuel; a first fluid amplifier having a main supply jet for ejecting a fluid stream, a: pair of receiver passages adaptea' to receive said flud stream, and control port means f0r receiving a fluid signal having a characteristic responsive to an engine demand jor fuel, for controllably varying which one of said pair of receiver passages will receive the fluid stream flow for establshing a controlled fluid sgnal; and means responsive to the controlled fluid signal f0r controlling fuel flow through said diseharge means.

9. The system as claimed in claim 8 wheren said means for controlling fuel flow comprise. a second flud arnplfer having an inlet port connected 10 a source of pressarized fluid for ssiu'ng a main fluid stream front said inlet port, a pair of outlet ports adapted to receive said man fluid stream and a control port connected 0 one of said first amplfier receiver passages whereby said main fluid stream is divided between said pair of outlet ports in accora'ance with the characterstcs of said controllea' flud signal; and valve means responsive to the divided maz'n fluid stream flow through said second amplifier outlet ports for controlling fuel flow through said discharge means.

10. T he system as claimed in claim 9 wherein said pressurized fluid is air and said valve means include means responsive to the proportion of fluid flow in said outlet passages to thereby control fluid flow through said nozZle.

11. An engine fael system comprising: a discharge means adapted t0 discharge fuel; a fluid amplifier having a fluid nteractian region, a fluid inlet passage opening into said region and. outlet means opening from said region; a

source of fluid connected to said nlet passage t0 issue a main fluid stream along a predetermined path within said region; a control flud means in communicatz'on with said amplifier providing fluid the characteristics of which provide a fluid sgnal indicative of an engine demand for fuel and whch causes a variation in the path of said main fluid stream within said nteraction region; said outlet means adapted to receive said main flaid stream and establish a controlled fluid condition which varies in accordance with variations in the path of the main fluid stream, a source of fuel and passage means interconnecting said source with said discharge means and including means responsive 20 said controlled fluid condition for controllng fuel flow through said discharge means in accordance with said controlled fluid condition.

12. The system as claimed in claim) 11 wheren said outlet means comprise a plurality of oatlet passages and divider means to direct the flud stream ssued from said inlet port into said outlet passages in accordance with the variatioris in the path of the fluid stream ana thereby establish a fluid condition in at least one of said outlet passages whch varies in accordance with the control fluid signal.

13. T he system according to claim 12 wherein said plurality of outlet passages comprise a pair of ouflet passages.

14. The system as claimed in claim 11 wherein said fluid amplifier means further incltrde a control passage opening nto said interaction region and wherein said control flud means is connected ta said control passage.

15. T he system as claimed in claim 12 including further: means f0r controlling the rate at whch said control fluid charaeteristics are applied t0 said man flud stream.

16. T he system as claimed in claim 11 wherein said control flua' means is communcated to said amplz'fier by connecting means comprising passage means adapted to lransmit a fluid signal and means responsive to an engine demand for fuel for controlling the transmission of the fluid signal herethrough.

17. The system as claimed in claim 16 wherein said engine responsive means include a second flud am plfier.

18. The system as claim in claim 11 wherein said con trol flud means comprise first ana' second control fluid sources, passage means connecting said first and second source to said interaction region; said first and second sources provia'ing fluid the characteristcs of which combine to produce said engine fuel demand signal.

19. The system as claimed in claim 1l wheren said fuel flow controlling means comprise valve means associated with said passage means for controllng fuel flow therethrough and means for operating said valve means in response to variations in the controlled flud condition.

20. The system as claimed in claim 12 wheren said fuel flow controlling means comprise valve means associated with said passage means for controlling fuel flow therethrough and linkage means responsive to the proportional fluid flow in two of said pluralty of outlet passages f0r controlling said valve means.

21. The method of operating an engine fuelsystem hav ing flaid amplfier means ncluding an interaction region, an inlet port opening into said interaction region and outlet passage means opening from said nteraction regon comprisng: issuing a stream of fluid from sad inlet port along a predetermned path within sad regon; sensng a fluid signal indicative of an engne demand for fuel; amplifyng sazd flud signal by utlizng sazd flud signal t caztse a variatz'on in the path of said flud stream within said interaction regian and receivng sad flud stream in saia' outlet passage means and establshing a flud condition in sad outlet passage means whch carries in accordance with variatons in sad fluid signal; and supplyng fuel to the engne in accordance wth said flud candirion.

22. An engine fuel system camprsing: a fuel pump having an outlet adapted zo delver fuel under pressure to an engz'ne; a by-pass passage means connected t0 saz'd outlet t0 divert a portion of fuel flow from the engirze; a valve means controllng fuel flow through sad by-pass passage means; a fluid amplfier havz'ng an inlet port and a pair of outlets; means to direct a flow of fluid through sazd amplifier from sad inlet t0 said outlets; sad amplz'fier further includng a control part in communication with a source of control fluid and havng means to divia'e the flud flow jrom said inlet port between sad outlets in accordance with the characterstics of the contr0llng fluid; and linkage means responsive to the diflerential fluid fl0w in said outlets to position said valve to thereby control fuel flow to the engne from maximum to minimum flow.

23. An engine fuel control comprisng. tachometer means adapted to be driven in proportion t0 engz'ne speed operatve to produce an output pulse train fluz'd signal indcative of engne speed, means for generating a flud signal indicative of the engne demand for fuel, means for c0mbning sad speed signal and saa' fuel demand sgnal to produce a fuel delivery flud signal, and control means responsve t0 said fuel delvery signal operative to control fuel delvery to the engine.

24. The system as clamed in claim 23 wherein saz'd control means comprse fuel supply means, fuel discharge means connected by condut means to said fuel supply means operative to dscharge fuel and valve means interposed between sad fuel supply means and said fuel discharge means operatz've t0 control fuel flow through said conduzt means.

References Cited The followng references, cted by the Examiner, are of record in the patented file of ths patent or the original patent.

UNITED STATES PATENTS 1,624,093 4/1927 Davs 13757 2,392,262 1/1946 Ramsey 13747 2765,800 10/1956 Drake 137-26 2,916,040 12/1959 Frick 13747 2829,662 4/1958 Carey 13736 2879,467 3/1959 Stern 235-201 2960,097 11/1960 Scheffler 137-82 2,981,271 4/1961 Cowles 13725 3,002,349 10/1961 Arnett -3928 3,006144 10/1961 Arnett 603928 3016066 1/1962 Warren 13781.5 3,053,276 9/1962 Woodward 13781.5 3,099,995 8/ 1963 Raufenbarth 147-82 3,124999 3/ 1964 Woodward 91-3 3,131601 5/1964 Cmran 91-3 3,181546 5/1965 B00the 13781.5 3,191611 6/1965 Bauer 137--81.5 3,191860 6/1965 Wadey 13781.5X 3212261 10/ 1965 Rose 6039.28 3228,408 1/ 1966 Young 6039.28X 3233,522 2/1966 Stern 13781.5X 3,246,682 4/1966 McCombs 6039.28X 3,260,271 7/ 1966 Katz 1378 1.5X 3,260,456 7/1966 -B0othe 13781.5X 3,275,015 9/1966 Meier 13781.5

CLARENCE R. GORDON, Primary Examiner U.S. C1. X.R. 137-81.5

Images