FIELD OF INVENTION
[1] The present invention relates to the field of fuel supply circuits for engine, and in particular for turbomachines, and more particularly concerns a method for regulating a power supply circuit and such a power supply circuit, wherein a two-phase flow flows.
[2] The present invention can be used in particular for an airplane turbojet.
STATE OF THE ART
[3] The aircraft engines typically include fuel supply circuits, taking fuel in tanks typically located in the aircraft's wings. During his journey from the tanks to the engine, the fuel first flows through an outlet pipe of the tanks belonging to the plane and then in a pipe belonging to the motor supply circuit. The junction between the two pipes is therefore an interface between the aircraft and the engine. Moreover, the power of these engines circuits include pumping equipment for pressurizing the fuel before it is fed into the combustion chamber. This pumping equipment generally comprises two stages: a BP pump (low pressure), and a pump HP (high pressure). BP pump is usually a centrifugal pump impeller with blades, whose functioning depends heavily on its good fuel. In particular, this type of pump is designed to operate with liquid phase fluids, the presence of gas in the fuel flow may affect the proper functioning of the pump.

[4] However, during the production of an engine, engine manufacturers may not know precisely the conditions to which this engine will be subjected in flight and, in particular, may not know the details of design of reservoirs and pipes fuel the plane. Conversely, manufacturers do not necessarily know the type of engine that will be used on a given flight, several engine models are generally compatible for the same aircraft model. This results

ignorance of existing flow conditions at the interface between the aircraft and the engine. However, depending on the configuration of the aircraft and its operating conditions (geometry pipelines, altitude, fuel type, temperature ...), the flow characteristics at this interface can be disrupted. The consequences may include degassing, or cavitation, and the coupling of the two phenomena that appear when the pressure flow becomes weak. Two-phase flow obtained then comprise microbubbles, bubbles or pockets full of gas which might disturb the operation of the LP pump or damage irreversibly, and thus induce motor malfunctions. As such, Figure 1, showing the flow of a fluid observed on a test bench, cavitation illustrates different modes including:

A cavitation pockets mode (Figure 1A.), Corresponding to a low flow rate and a relatively high flow pressure. This mode is characterized by a relatively static flow, with pockets of cavitations (gas pockets) stable, the latter remaining localized at the interface between the aircraft and the motor (modeled by the neck of a venturi with diverging portions which can be seen as the upper surface of the blades), so they do not move in the pipeline and they do not reach the pump. Generally, this type of cavitation can also be formed on the upper surface of the blades pumps.

A vortex cavitation mode (Figure 1B.), Corresponding to a higher flow rate and lower pressure than in the pockets by cavitation mode. This mode is characterized by a strong vorticity, a dynamic flow where cavitation pockets stand in a synchronized manner at a well defined frequency while moving from top to bottom of the section of the next pipe flow.

[5] These cavitating flow regimes can be detrimental to the proper functioning of the pump, and thus the engine. Indeed, in these types of systems, the pump BP may occasionally be supplied with a essentially gaseous phase in the fluid. The vertical line T in FIG 1B shows a section in which the flow is almost entirely gaseous. This configuration may include

leading au désamorçage the pumps or in public J des vibrations, ainsi drop the apparition of a history of inflation pompage endommager the pumps.

[6] In order to avoid these types of flow regimes, it would be necessary to simulate and thus to correctly specify the fuel supply conditions at the interface between the aircraft and the engine during flight. However, as was mentioned above, engine manufacturers rarely have enough information, particularly as to the exact configuration of the aircraft, to make possible such simulations. Therefore, the current solutions to prevent these flow regimes are content to limit the flight envelope to avoid any risk of too low pressure (favoring the phenomenon of cavitation) at the interface between plane and the motor, or over-configure BP pump, which has the effect of unnecessarily increasing the weight of the aircraft.

[7] There is therefore a need for a method for controlling a power supply circuit and a power circuit, avoiding certain harmful flow regimes the functioning of the engine and which are lacking, at least in part, the disadvantages inherent in the aforementioned known methods.

PRESENTATION DE L'INVENTION

[8] The present disclosure relates to a method of regulating a power supply for a circuit comprising at least a first pump and an upstream pipe leading to the first pump, the method comprising the steps of:

determination, in the upstream pipeline, the gas content of a flow supplied to the first pump;

when the value of the gas content in the upstream pipe, as determined at the determining step is greater than or equal to a predetermined threshold value, changing the rate of flow feeding the first pump.

[9] In the present specification, the flow can be a flow of a liquid, or a two-phase flow, that is to say a flow of a fluid comprising a liquid phase and a gas phase comprising the vapor of liquid and air initially dissolved

in the liquid, the terms "upstream" or "downstream" are included depending on the direction of fluid flow.

[10] The term "gas concentration", it is understood the total volume proportion of the gaseous phase throughout the fluid on a portion of the upstream pipe. Therefore, the predetermined threshold value corresponds to a predetermined proportion of the gas phase. The threshold value may be determined to correspond to the appearance of a particular flow regime, for example corresponding to a particular mode of cavitation, especially swirled ire. Thus, the determining step detects the occurrence of such a flow system upstream of the first pump.

[11] In addition, by "change of the flow rate" is comprises increasing or decreasing the mass flow of fluid fed to the first pump. Thus, when a cavitation regime vortex was detected at the step of determining the mass flow of fluid supplied to the first pump is changed. This rate change causes a change in the pressure upstream of the first pump, and therefore a flow regime change. With this method it is possible to avoid flow regimes corresponding to a vortex cavitation mode may affect the functioning of the engine, detecting the occurrence of such schemes and by simply changing the flow of the flow without it is necessary, in the case of an aircraft engine, to limit the flight envelope. Furthermore, such a method eliminates the need to oversize the pump, thus making it possible to avoid increasing its weight and bulk.

[12] In some embodiments, when the value of the gas content determined at the determining step is greater than or equal to the predetermined threshold value, the rate of flow feeding the first pump is increased.

[13] In some embodiments, when the value of the gas content determined at the determining step is greater than or equal to the predetermined threshold value, the rate of flow feeding the first pump is increased so as to achieve regime "supercavitation" in the upstream pipeline.

[14] The detection, the step of determining a value of the higher gas content than or equal to the predetermined threshold value is characteristic of the presence of a cavitation mode at risk for the pump, e.g. a swirling fashion cavitation. The cavitation vortex mode is usually the most damaging flow regime at the pump, the proper functioning of the latter and therefore the engine, given its unstable nature. Indeed, in this type of scheme, the LP pump can be timely supplied with a essentially gaseous phase in the fluid. This particular configuration can lead to defusing the pump and damaging it.

[15] However, the increase in flow rate causes a decrease in the pressure of said flow, this pressure reduction being effective to enhance the cavitation phenomenon, already present in the upstream pipe, reaching a cavitation mode said "supercavitation", corresponding to a higher flow rate and a lower flow pressure in the vortex cavitation mode. This rate increase is done so voluntarily achieve regime "supercavitation" in the upstream pipeline. Indeed, as illustrated in Figure 1C, this scheme is characterized by a more localized cavitation and "stationary" dynamic point of view of flow. More specifically, the liquid phase flows in the form of a substantially circular cross-section jet whose diameter remains substantially constant along the pipeline. Thus, the generated gas remains in specific areas of the pipe, into an annular flow between said liquid flow and the pipe wall. Therefore, the gas flow is more stable, and therefore less harmful to the functioning of the first pump.

[16] In some embodiments, the increased rate of flow feeding the first pump is greater than 2%.

[17] This increased mass flow rate of flow of at least 2% from the initial flow allows, in most applications, to switch from a swirling supercavitation regime to regime.

[18] In some embodiments, increasing the rate of flow feeding the first pump is less than 15%, preferably less than 10%, more preferably less than 5%.

[19] Increases in low amplitudes of the flow rate supplying the first pump may well be sufficient, in most applications, to switch vortex regime to regime "supercavitation". too greatly This avoids affect engine operation.

[20] In some embodiments, the circuit comprises a downstream pipe downstream of the first pump, and at least a first bypass duct connected in derivation of the downstream pipe and for taking a certain amount of fluid in the downstream pipe , changing the rate of flow feeding the first pump being achieved by modifying an amount of fluid taken from the downstream pipe via the first bypass channel.

[21] The downstream pipe is the pipe through which the fluid from the first pump flows. The bypass channel may be a channel transversely connected to the pipeline. Means are provided to collect a certain amount of fluid flowing in the downstream pipe via the bypass channel. Modifying the rate of flow feeding the first pump can thus be achieved by a simple modification of the quantity of fluid collected in the downstream pipe via at least the first bypass channel.

[22] In some embodiments, increasing the rate of flow feeding the first pump is carried out by decreasing the quantity of fluid collected in the downstream pipe via at least the first bypass channel.

[23] In some embodiments, determining the flow of gas content is performed by a phase meter adapted to determine the gas content of a two-phase flow, arranged in the upstream pipe.

[24] In some embodiments, the phase meter comprises a plurality of concentric electrodes, the phase meter being configured to, by measuring the electrical capacitance between the concentric electrodes, to determine the gas content in the pipeline. [25] The phase meter may comprise an outer cylinder within which are disposed a plurality of cylindrical and concentric electrodes between them and with the outer cylinder. The fluid flowing in the pipe flows within the outer cylinder, along these electrodes. The electrodes used to measure a capacitance whose value is representative of the fluid gas content flowing into the meter. A phase meter of this type is described in detail in French patent FR 2 978 828. Knowing this parameter in real time has the advantage to modify the flow pattern by changing the flow rate of the flow immediately when the value of the gas concentration reaches the threshold value.

[26] In some embodiments, the change in the quantity of fluid collected downstream of the first pump is performed by regulating the opening of a bleed valve provided on the first bypass channel.

[27] The means for collecting a certain amount of fluid flowing in the downstream pipe via the bypass channel can be a simple sampling valve. The degree of opening of said valve is used to regulate the amount of fluid flowing in the bypass channel, and therefore the quantity of fluid collected in the downstream pipe.

[28] In some embodiments, a control unit compares the value of the gas content determined by the phase meter with the preset critical value and, depending on the result of this comparison, controls a degree of opening of the sampling valve.

[29] The controller may be an electronic control unit ECU 'for Electronic Control Unit). The control unit can change the flow regime autonomously, by controlling the opening of the sampling valve according to a value of the gas content transmitted by the phase meter.

[30] In some embodiments, an outer diameter of the phase meter is equal to an outer diameter of the upstream pipe.

[31] In other words, the outer cylinder has a diameter equal to the outer diameter of the upstream pipe. This will not create a discontinuity in the geometry of the upstream pipeline.

[32] In some embodiments, the leading edges and phase meter of concentric electrodes leaks are optimized so as to limit the head losses in the flow.

[33] The leading edges can for example be beveled to limit losses in the flow.

[34] In some embodiments, the flow is a flow of fuel for a turbomachine comprising a liquid phase and a gaseous phase.

[35] The control process makes it possible to regulate the fuel supply of the engine.

[36] In some embodiments, the power supply circuit includes the first pump and a second pump downstream from the first pump, the first and second pumps being connected by the downstream pipe.

[37] The power supply circuit is part of the turbine engine. The first pump may be a pump LP (low pressure), and the second pump may be a pump HP (high pressure).

[38] In some embodiments, the power circuit comprises at least one heat exchanger disposed between the first pump and the second pump, the first bypass channel being connected in parallel between the first pump and the exchanger, and a second bypass duct connected in parallel on the downstream pipe between the heat exchanger and the second pump.

[39] The heat exchanger may be a heat exchanger oil / fuel or air / fuel. It allows to regulate the fuel flow temperature before being injected into the engine. The second bypass channel can be similar to the first bypass channel. A sampling valve may be provided on the first bypass duct and / or the second bypass channel. The first branch channel and the second bypass channel and allow to withdraw a mixture of fuel respectively upstream and downstream of the exchanger, which enables, after mixing, to regulate the temperature of the sample fuel. These different elements are typically present in feed streams of turbomachines. Thus, under the present disclosure, the first and / or second bypass passages are used to modify the rate of flow in the upstream pipe by changing the quantity of fluid collected in the downstream pipe. This method of sampling and to change the flow regime in the upstream pipe by using elements already traditionally in such power supply circuits. Therefore, it is not necessary to use an additional device for changing the rate of flow feeding the first pump, which allows not to increase the engine.

[40] In some embodiments, the fuel used to power the turbine engine is taken from a tank belonging to a vehicle, preferably an aircraft.

[41] In some embodiments, the fuel taken by the first and / or second bypass channel is routed back to said reservoir.

[42] The fuel thus circulates between the tank and the turbine engine by forming a circulation loop.

[43] In some embodiments, the phase meter is arranged downstream of the interface between the vehicle and the turbomachine.

[44] In other words, the phase meter is arranged in the upstream pipe between the interface between the vehicle and the turbine engine and the first pump, that is to say at the entrance of the upstream pipe. This position of the phase meter allows it to measure the gas content directly downstream of the interface, which can detect the flow disturbances generated in the interface, and the possible presence of vortex cavitation.

[45] In some embodiments, the threshold value of the gas content is between 50% and 80%.

[46] A value of gas content within this range of values ​​may be characteristic of the presence of a vortex cavitation regime in the flow.

[47] In some embodiments, changing the rate of flow feeding the first pump is performed when the measurement phases of measuring a variation of at least 5%, preferably 10%, more preferably 15%, in the gas content in less than 1 second.

[48] ​​The stable cavitation patterns, such as cavitation pockets or "supercavitation" have a substantially constant gas content over a given section. Therefore, the detection by the phase measuring of an excessive variation of the gas content may be characteristic of the presence of an unstable flow regime.

[49] In some embodiments, the low-pressure pump is configured in accordance with a predetermined flow fashion. For example, the low pressure pump can be set for flows of which the gas content may be up to 45%. In this case, being able to detect the presence of vortex cavitation, and so modify the flow regime, allows to maintain the gas content of the flow to 45% lower values. It is thus not necessary to use a low pressure pump oversized, configured for all types of flows. The total mass of the power circuit can be reduced.

[50] The present disclosure also relates to a supply circuit comprising at least one pump and an upstream pipe leading to the pump, a phase meter disposed in the upstream pipe, a flow control device and a computing unit, said calculating unit configured to, when the value of the gas content in the upstream pipe is greater than or equal to a predetermined threshold value, controlling the flow rate adjusting device so as to modify the rate of flow to the pump. The technical advantages associated with the use of such a supply circuit are similar to those related to the method for regulating a power supply circuit described above. Furthermore, the features described with reference to this method are transposed, alone or in combination, to the supply circuit.

BRIEF DESCRIPTION OF DRAWINGS

[51] The invention and its advantages will be better understood from reading the detailed description below of an exemplary embodiment of the invention given as non-limiting. The description refers to the appended figures, in which:

- Figure 1A shows, modeled manner, a pipe wherein a fluid flows in a pocket cavitation mode;

- Figure 1B shows, modeled manner, a pipe wherein a fluid flows in a vortex cavitation mode;

- Figure 1C shows, in a modeled manner, a pipe wherein a fluid flows through a means of supercavitation;

- Figure 2 shows an aircraft comprising a supply circuit according to the present disclosure;

- Figure 3 schematically shows such a power supply circuit;

- Figure 4 shows a phase meter.

DETAILED DESCRIPTION OF EXAMPLES OF REALIZATION

[52] Figure 2 shows an aircraft 9 comprising a motor 1 and a reservoir 2 arranged in a wing of the aircraft 9. The motor 1 comprises a supply circuit 10, the latter being powered by the collected fuel in the tank 2. during its passage from the reservoir 2 to the engine 1, the fuel first flows through a first tank line 2a from the tank 2 and 9 belonging to the aircraft, then in a duct 10a belonging to the supply circuit 10 of the motor 1. the junction between these two pipes 2a and 10a is an interface I between the aircraft 9 and the engine 1.

[53] The power supply circuit 10 is shown schematically in Figure 3. It comprises a pumping equipment for pressurizing the fuel prior to its feed into the combustion chamber 20. The pumping equipment includes a first pump 12 (low pump pressure), and a second pump 14 (high pressure pump). The arrows in FIG 3 show the direction of flow of fuel. 10a is a driving upstream pipe, wherein the fuel flows from the tank 2 and leading to the first pump 12. The first pump 12 leads to a downstream pipe 13, wherein the fuel flows to the second pump 14, to which the downstream pipe 13 is connected. The fuel leaving the second pump 14 is then fed into a metering device 19, then into the combustion chamber 20 of the engine 1. A heat exchanger 16 is disposed on the downstream pipe 13 between the first pump 12 and second pump 14 . It allows to regulate the temperature of the flow by exchanging heat between the fluid flowing in the downstream pipe 13, and the fluid collected by the doser 19 downstream of the second pump 14, and supplied to the heat exchanger 16 via a metering pipe 19a.

[54] The power supply circuit 10 further comprises a first bypass channel 13a, 13b and a second bypass channel. The first and second bypass channel 13a and 13b make it possible to collect a certain amount of fuel in the downstream pipe 13. The first bypass channel 13a is connected in parallel on the downstream pipe 13 between the first pump 12 and the exchanger heat 16. the second 13b bypass channel is connected in parallel on the downstream pipe 13 between the heat exchanger 16 and the second pump 14.

[55] A sampling device 18 is provided on the first and second 13a and 13b pass channels. The sampling device 18 comprises a first bleed valve 18a provided on the first bypass channel 13a and a second 18b sampling valve 13b provided on the second bypass channel. The degree of opening of the bleed valves 18a and 18b makes it possible to regulate the amount of fuel flowing in the first and second bypass channel 13a and 13b, and therefore the quantity of fluid collected in the downstream pipe 13. The device sampling 18 is also connected to a return pipe 10b, wherein the fuel flows have been taken within the downstream pipe 13, towards the reservoir 2. the return pipe 10b is itself connected to a second tank line 2b , at the interface I. the fuel from the supply circuit 10 flows to the reservoir 2 via the second tank line 2b.

[56] A phase meter 30 is arranged in the upstream pipe 10a, downstream of the interface I. As shown in Figure 4, the phase meter 30 includes a cylindrical casing 30a, within which is arranged a plurality of electrodes 30b, 30c, 30d, 30e cylindrical and concentric with each other and with the cylindrical casing 30a. The fluid flowing in the upstream pipe 10a flows into the phase meter 30, along these electrodes. The electrodes used to measure a capacitance whose value is representative of the gas content of the fluid flowing in the measuring phase 30.

[57] A computing unit 40 is connected to the phase meter 30 and the pickup device 18. The control unit 40 may be of FADEC (for "Full Authority Digital Engine Control" in English). The gas content of the fluid flowing in the upstream pipe 10a, measured by the measuring phase 30, is transmitted to the calculation unit 40. Based on this value of the gas content, the computing unit 40 control an opening degree of the first and second bleed valves 18a, 18b, by the method described below.

[58] A threshold value of the gas content, characteristic of the appearance of a vortex cavitation regime, is predetermined. In this example, the predetermined threshold value corresponds to a gas content of 10%. When the calculation unit 40 determines a value of the gas content in the upstream pipe 10a, measured by the measuring phase 30 is greater than the predetermined threshold value, the computing unit 40 infers the presence of a vortex cavitation regime in the downstream pipe 13, and therefore control an opening degree of the first and second bleed valves 18a, 18b.

[59] In the example above, the existence of a predetermined threshold value used to deduce the presence of a vortex cavitation regime. However, other means can be implemented. For example, the detection of cavitation regime vortex in the downstream pipe 13 can be characterized as the 30 meter measurement phases a variation of at least 5% of the gas content in less than 1 second.

[60] The partial closure of the first and second bleed valves 18a and 18b, reduces the quantity of fuel taken from the downstream pipe 13 via the first and second 13a and 13b pass channels. This reduction in the quantity of fuel taken from the downstream pipe 13 leads to an increase of the fuel flow rate in the upstream pipe 10a.

[61] In this regard, major aerospace pump manufacturers often make the assumption that the entire amount of gas has been compressed when passing through the blades of the inductor and the pulley (first pump 12 in this example). But this may not be the case, as evidence of cavitation can be found on the floor of the main HP pump (second pump 14 in this example). This indicates that a certain amount of gas is not compressed at the output of the first pump 12 and therefore it is found in the downstream pipe 13. This is even more true because the "supercavitation" is considered a state gas content of saturation and as mentioned above, the total volume proportion of the gaseous phase throughout the fluid on a portion of the duct increases more. Therefore the total mass of fluid in the pipe is weighted by the masses of the liquid phase and the gas phase, respectively. The volume of the pipe is constant in an engine configuration, so the total mass of the fluid is weighted by the relative densities of each phase. The density of kerosene at room temperature is of the order of 780 kg / m3 and that of its vapor is of the order of 4.5 kg / m3, which gives a ratio of about 170 between the two values knowing that the acceptable gas content according to standard ARP492C of the low pressure pump (first pump 12 in this example) is 45%. On the other hand, the need for mass flow rate of the first pump in a rotation speed, altitude and temperature remains constant. In addition, it is impossible to filter the fluid collected in the downstream pipe 13, which means the return of a two-phase flow to the tank 2 and hence, the decrease in the total mass of the fluid upstream of the second pump 14 . the two-phase mixture which is found in the tank 2 is thus sucked again by the first pump 12. for this, the sampling device 18 must be in its

new position, so as to limit the two-phase flow taken. Another effect present in fuel systems is increasing because of kerosene temperature of the compressibility due to the passage of liquid into the first pump 12. More specifically, the temperature of the return flow to the tank 2 is weighted by a mixture of fuel "hot", with the "cold" fuel. The presence of gas at the outlet of the first pump 12, reduces the amount of liquid present, thereby increasing the average temperature of fuel taken. Therefore, the return to the tank 2 also results in an increase in the average fuel temperature. But this increase has a beneficial effect because cavitation is delayed because of the latent heat of the fluid because it is an endothermic phenomenon (the transformation from liquid to vapor requires energy and is taken to the liquid creating a local cooling in the formed vapor-liquid pocket). In the case of kerosene, the saturation vapor pressure increases to a higher temperature, which delays the onset of cavitation. Therefore, the first pump 12 will draw more liquid as two-phase mixture.

[62] In the present example, the control unit 40 controls an opening degree of the first and second bleed valves 18a, 18b to a 5% increase in throughput in the upstream pipe 10a. This increase in flow allows to pass from a cavitation mode vortex in the upstream pipe 10a (shown in Figure 1B), in a mode of "supercavitation" (shown in Figure 1C). by "supercavitation" means a flow mode, more stable and more regular, and more favorable to the functioning of the first pump 12 a vortex cavitation mode and thus reduces the risk of damaging the latter .

[63] Although the invention has been described with reference to specific embodiments, it is obvious that modifications and changes can be made to these examples without departing from the general scope of the invention as defined by the revendications. In particular, individual characteristics of the various embodiments shown / mentioned can be combined in additional embodiments. Therefore, the description and drawings should be considered illustrative rather than restrictive sense.

[64] It is also clear that all the features described with reference to a method are transposed, alone or in combination, to a device, and vice versa, all the features described with reference to a device can be transposed, alone or in combination, to a method.

CLAIMS
1. A method of controlling a power supply circuit (10) comprising at least a first pump (12) and a upstream pipe (10a) leading to the first pump (12), the method comprising the steps of:

- determining, in the upstream pipe (10a) of the gas content of a flow supplied to the first pump (12);

- when the value of the gas content in the upstream pipe (10a), determined at the determining step is greater than or equal to a predetermined threshold value, changing the rate of flow feeding the first pump (12) .

2. The method of claim 1 wherein when the value of the determined gas content in the determining step is greater than or equal to the predetermined threshold value, the flow rate supplied to the first pump (12) is increased so as to obtain a supercavitation regime in the upstream pipe (10a).

3. The method of claim 1 or 2, wherein the increase in the rate of flow feeding the first pump (12) is greater than 2% and less than 15%, preferably greater than 2% and less than 10% , more preferably greater than 2% and less than 5%.

4. A method according to any one of claims 1 to 3, wherein the power circuit (10) comprises a downstream pipe (13) downstream of the first pump (12) and at least a first bypass channel ( 13a) connected in derivation of the downstream pipe (13) and for taking a certain amount of fluid in the downstream pipe (13), and

wherein modifying the rate of flow feeding the first pump (12) is achieved by modifying an amount of

fluid taken from the downstream pipe (13) at least via the first bypass channel (13a).

5. The method of claim 4, wherein the rate of flow feeding the first pump (12) is increased by decreasing the quantity of fluid collected in the downstream pipe (13) at least via the first bypass channel (13a) .

6. A method according to any one of claims 1 to 5, wherein the determination of the flow of gas content is performed by a phase meter (30) adapted to determine the gas content of a two-phase flow, disposed in the upstream pipe (10a).

7. A method according to any one of claims 1 to 6, wherein the predetermined threshold value of the gas content is between 50% and 80%.

8. The method of claim 6, wherein modifying the rate of flow feeding the first pump (12) is performed when the phase meter (30) measures a variation of at least 5%, preferably 10%, more preferably 15%, in the gas content of less than 1 second.

9. A power supply circuit (10) having at least one pump (12) and a upstream pipe (10a) leading to the pump (12), a phase meter (30) disposed in the upstream pipe (10a), means flow control (18) and a computing unit (40), said calculating unit (40) configured to, when a value of the gas content measured by the phase meter (30) in the upstream pipe ( 10a) is greater than or equal to a predetermined threshold value, controlling the flow control device (18) so as to modify the rate of flow to the pump (12).