FEED SYSTEM AND A METHOD OF SUPPRESSING THE POGO EFFECT

Abstract

A feed system for feeding a rocket engine with a liquid propellant includes a feed circuit, and a device to vary a volume of gas in the feed circuit. The device is configured to cause a volume of gas in the feed circuit to vary while the rocket engine is in operation. The device to vary gas volume includes at least one variable-flow-rate gas injector to inject gas into the liquid propellant in the feed circuit. Methods of suppressing a POGO effect are also provided.

Claims

1. A feed system for feeding a rocket engine with a liquid propellant, the system comprising: a feed circuit; and a device to vary a volume of gas in the feed circuit, which device is configured to cause a volume of gas in the feed circuit to vary while the rocket engine is in operation, wherein the device to vary gas volume comprises at least one variable flow-rate gas injector to inject gas into the liquid propellant in the feed circuit.

2. A feed system according to claim 1, further comprising a control unit configured to control the device to vary gas volume.

3. A feed system according to claim 2, further comprising at least one sensor connected to the control unit, and wherein the control unit is configured to control variation in the gas volume as a function of signals sensed by the at least one sensor.

4. A feed system according to claim 3, wherein the at least one sensor comprises an accelerometer.

5. A feed system according to claim 3, wherein the at least one sensor comprises a sensor to sense pressure of the propellant.

6. A feed system according to claim 2, wherein the control unit is configured to control variation of the gas volume as a function of time.

7. A method of suppressing a POGO effect, comprising: varying a volume of gas in a feed circuit of a system to feed a rocket engine with a liquid propellant while the rocket engine is in operation, to control a difference between at least one hydraulic resonant frequency of the feed circuit and at least one mechanical resonant frequency of a structure coupled to the feed circuit, wherein the gas volume is caused to vary by varying a rate at which gas is injected into the liquid propellant in the feed circuit.

8. A method of suppressing the POGO effect according to claim 7, wherein the gas volume varies to keep the difference above a predetermined threshold.

9. A method of suppressing the POGO effect according to claim 7, wherein the gas volume is caused to vary as a function of at least one mechanical oscillation value sensed on the structure.

10. A method of suppressing the POGO effect according to claim 9, further comprising performing spectral analysis on at least one mechanical oscillation to determine the at least one mechanical resonant frequency of the structure.

11. A method of suppressing the POGO effect according to claim 10, wherein a filter algorithm, or an unscented Kalman filter, is applied to at least one sensed mechanical oscillation to determine the at least one mechanical resonant frequency and/or to predict its future variation.

12. A method of suppressing a POGO effect, comprising: performing spectral analysis on at least one mechanical oscillation to determine at least one mechanical resonant frequency of a structure coupled to a feed circuit of a system to feed a rocket engine with a liquid propellant; varying a volume of gas in the feed circuit while the rocket engine is in operation, to control a difference between at least one hydraulic resonant frequency of the feed circuit and the at least one mechanical resonant frequency of the structure coupled to the feed circuit.

13. A method of suppressing the POGO effect according to claim 12, wherein a filter algorithm is applied to the at least one sensed mechanical oscillation to determine the at least one mechanical resonant frequency.

14. A method of suppressing the POGO effect according to claim 12, wherein an unscented Kalman filter is applied to the at least one sensed mechanical oscillation to determine the at least one mechanical resonant frequency and predict its future variation.

15. A method of suppressing the POGO effect according to claim 12, wherein the gas volume varies to keep the difference above a predetermined threshold.

16. A method of suppressing the POGO effect according to claim 12, wherein the variable gas volume is located at least in part in a hydraulic accumulator connected to a duct of the feed circuit.

17. A method of suppressing the POGO effect according to claim 12, wherein the gas volume is caused to vary by varying a rate at which gas is injected into the propellant in the feed circuit.

18. A non-transitory computer readable medium including computer executable instructions for performing a method of suppressing the POGO effect according to claim 12.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The invention can be well understood and its advantages appear better on reading, the following detailed description of embodiments given as non-limiting examples. The description refers to the accompanying drawings, in which:

[0029] FIG. 1 is a diagram based on an analogy between hydraulic circuits and electrical circuits, showing a rocket engine vehicle with a liquid propellant feed system in an embodiment of the invention;

[0030] FIGS. 2A and 2B are cross-sections through a variable-volume accumulator installed in parallel with a feed circuit of the FIG. 1 system;

[0031] FIG. 3 is a cross-section of a variable-volume accumulator in a second embodiment of the invention;

[0032] FIG. 4 is a diagram based on an analogy between hydraulic circuits and electrical circuits, showing a rocket engine with a liquid propellant feed circuit in a third embodiment of the invention;

[0033] FIG. 5 is a diagram of a gas injection point connected to the FIG. 4 circuit;

[0034] FIGS. 6A, 6B, 6C, and 6D are diagrams showing the operation of an unscented Kalman filter algorithm; and

[0035] FIGS. 7A, 7B, 7C, and 7D are graphs showing three possible ways in which the hydraulic resonant frequency of the feed circuit can vary in response to a variation in a mechanical resonant frequency of the FIG. 1 vehicle.

DETAILED DESCRIPTION OF THE INVENTION

[0036] The vehicle 1 shown in FIG. 1 has a propulsion chamber 2 incorporating a combustion chamber and a convergent-divergent nozzle. The vehicle 1 also has a feed circuit 3, 4 for each of two liquid propellants that react chemically with each other and that are fed to the propulsion chamber 2. The first feed circuit 3 is shown in part only. Each feed circuit 3, 4 filled with fluid represents a dynamic system that can be modeled like an electrical circuit having resistors 5, inductors 6, and capacitors 7, and that presents at least one resonant frequency.

[0037] In order to cause at least one resonant frequency of the second feed circuit 4 to vary, the circuit includes in parallel therewith a hydraulic accumulator 8 having a volume of gas that is variable and thus presenting compressibility that is also variable. This accumulator 8, shown in FIGS. 2A and 2B, comprises a tank 9 with a pressurized gas feed point 10 on one side and a connection 11 to a duct 15 of the second feed circuit 4 on an opposite side. At various levels between the point 10 and the connection 11, dip tubes 12, 13 connect the tank 9 with the duct 15. Each dip tube 12, 13 includes a respective valve 14, 16 that is interposed between the tank 9 and the duct 15. Opening and closing the valves 14 and 16 thus makes it possible to vary the liquid level, and thus the volume of gas 17, inside the tank 9, as shown in FIGS. 2A and 2B. In FIG. 2A, the valve 14 of the shorter dip tube 12 is open, while the valve 16 of the dip tube 13 is closed. The free surface of the liquid is thus stabilized at the level of the inlet to the dip tube 12, and the volume of gas 17 together with its compressibility thus remains comparatively limited. In contrast, in FIG. 2B, the valve 14 of the dip tube 12 is closed, and the valve 16 of the dip tube 13 is open. The free surface of the liquid is thus stabilized at the lower level of the inlet to the dip tube 13, and the volume of gas 17 and its compressibility is increased accordingly.

[0038] By varying the effective compressibility of the accumulator 8, it thus becomes possible, even while the rocket engine of the vehicle 1 is in operation, to adapt the hydraulic resonant frequency of the second feed circuit 4 so as to prevent it from coinciding with a variable mechanical resonant frequency of a support structure of the rocket engine. Naturally, in order to achieve this result, it is necessary to have perceptible acceleration in order to separate the heavier liquid from the lighter gas. This hydraulic accumulator 8 of variable gas volume therefore does not operate in the same manner under conditions of microgravity.

[0039] In a second embodiment as shown in FIG. 3, the accumulator 8 likewise has a tank 9 with a pressurized gas feed point 10 on one side and a connection 11 to a duct 15 of the second feed circuit 4 on an opposite side, but it has only one dip tube 12, which tube is however movable in the depth direction of the tank 9 in order to vary the level of the liquid, and thus the gas volume 17 inside the tank 9. This embodiment makes it possible to vary the liquid level continuously and thus to vary the gas volume continuously, and hence varies the compressibility in the accumulator 8 and the hydraulic resonant frequency of the second feed circuit 4.

[0040] A third embodiment is shown in FIG. 4. As in the embodiment of FIG. 1, in this other embodiment, the vehicle 1 likewise has a feed system with a feed circuit 3, 4 for feeding each of two liquid propellants that react chemically with each other and that are fed to a propulsion chamber 2.

[0041] Nevertheless, in this third embodiment, the at least one hydraulic resonant frequency of the second feed circuit 4 is caused to vary by injecting gas at a variable rate into the fluid of the feed circuit 4 by means of a gas injection device 20 connected to the second feed circuit 4. Downstream from this injection point 20, the compressibility of the liquid/gas fluid in the circuit is modified by the compressibility of the injected volume of gas. Consequently, the at least one hydraulic resonant frequency of the feed circuit 4 and also the speed of sound in the circuit 4 are also varied.

[0042] The gas injection device 20 is shown in FIG. 5. It is installed on a duct 15 of the second feed circuit 4 and it comprises an annular chamber 21 around the duct 15, which chamber is connected to a source of pressurized gas (not shown) via three valves 22, 23, and 24, and communicates with the duct 15 via injection orifices 25. The rate at which gas is injected into the duct 15, and thus into the second feed circuit, can thus be varied by opening and closing the valves 22, 23, and 24. Alternatively, or in combination with the above arrangement, such a gas injection device could include a variable-opening valve or a flow rate regulator, thus making it possible to obtain continuous variation in the volume flow rate of the gas that is injected into the duct 15, and thus of the at least one hydraulic resonant frequency.

[0043] Both the variable gas volume hydraulic accumulator 8 in the first embodiment and the variable flow rate gas injection device 20 of the second embodiment can be connected equally well to a control unit 30 for controlling them by means of a variable setpoint that is issued by the control unit to the accumulator 8 and/or to the gas injection device 20. If the way the mechanical resonant frequency varies is known in advance, as a result of simulations and/or tests that have already been performed, this setpoint may be preprogrammed merely as a function of time. Nevertheless, it is also possible, and indeed preferable under certain circumstances, to cause this setpoint to vary in response to signals that are received in real time or almost in real time. For example, as shown in FIGS. 1 or 3, the vehicle 1 may include at least one accelerometer 31 and a propellant pressure sensor 32 in the circuit 4. The accelerometer 31 is connected to the control unit 30 in order to send signals thereto representative of the mechanical behavior of the structure of the vehicle 1, and the pressure sensor 32 is also connected thereto in order to send signals representative of the hydraulic behavior of the circuit 4.

[0044] These signals are processed in the control unit 30 in order to extract the mechanical and hydraulic resonant frequencies by spectrum analysis. Filter algorithms, such as for example the unscented Kalman filter algorithm as described in The unscented Kalman filter for nonlinear estimation, Proceedings of Symposium 2000 on Adaptive Systems for Signal Processing, Communication and Control (AS-SPCC), IEEE, Lake Louise, Alberta, Canada, October 2000, may be used, not only to filter noise from the signals, but even in predictive manner in order to forecast short-term variation in the resonant frequencies of the modes of oscillation, and to anticipate them in the way the hydraulic resonant frequency is controlled. The control unit may be programmed to initialize such a filter algorithm close to an expected mechanical resonant frequency, thereby making it possible subsequently to track this frequency in flight.

[0045] In a dynamic system such as a vehicle 1, it can be assumed that there exists a Markov sequence of latent states x.sub.t that vary in time in application of a function F. These latent states are observed indirectly by sensors giving measured states y.sub.t as obtained via a measurement function G. Thus, x.sub.t and y.sub.t can be expressed by the following formulas:

x.sub.t=F(x.sub.t1)+

y.sub.t=G(x.sub.t)+

[0046] The values and represent respectively the noise inherent to the system and measurement noise, and both of them present Gaussian distributions.

[0047] The object of a filter algorithm is to infer the state of the dynamic system from noisy values as measured by sensors. A Kalman system provides an inference that is fast and accurate for systems that are linear. It is nevertheless not directly applicable to systems that are non-linear, and the present application is potentially classifiable as a non-linear system. Among various alternatives for adapting the Kalman filter algorithm to non-linear systems, there is known in particular the unscented Kalman filter (UKF). This algorithm propagates several estimates of x.sub.t through the functions F and G and reconstructs a Gaussian distribution, assuming that the propagated values come from a linear system. The positions of these estimates for x.sub.t are referred to as sigma points, and they are calculated from an initial average and variance with an approximation scheme referred to as an unscented transformation.

[0048] In FIG. 6A, a first step is shown in which the initial sigma points x.sub.0.sup.0, x.sub.0.sup.1, x.sub.0.sup.2, x.sub.0.sup.3, x.sub.0.sup.4 are calculated by such an unscented transformation starting from a mean m.sub.0 and a variance P.sub.0 taken into consideration for the latent state x.sub.0 based on a first set of measurements y.sub.0 at the initial moment t=t.sub.0. Thereafter, in a prediction step, shown in FIG. 6B, estimated positions X.sub.0.sup.0, X.sub.0.sup.1, X.sub.0.sup.2, X.sub.0.sup.3, X.sub.0.sup.4 for the sigma points corresponding to the following sampling instant (t=t.sub.1) are predicted by applying the prediction step of the Kalman filter algorithm to the initial signal points X.sub.0.sup.0, X.sub.0.sup.1, X.sub.0.sup.2, X.sub.0.sup.3, X.sub.0.sup.4. In the following step of updating, as shown in FIG. 6C, the actual sigma points X.sub.1.sup.0, X.sub.1.sup.1, X.sub.1.sup.2, X.sub.1.sup.3, X.sub.1.sup.4 are calculated on the basis of the previous sampling at t=t.sub.1. The differences between the positions X.sub.0.sup.0, X.sub.0.sup.1, X.sub.0.sup.2, X.sub.0.sup.3, X.sub.0.sup.4 as predicted on the basis of the initial sigma points X.sub.0.sup.0, X.sub.0.sup.1, X.sub.0.sup.2, X.sub.0.sup.3, X.sub.0.sup.4 and the positions X.sub.1.sup.0, X.sub.1.sup.1, X.sub.1.sup.2, X.sub.1.sup.3, X.sub.1.sup.4 as actually calculated on the basis of the new sampling make it possible to obtain information about the function f representing variation of the latent state x.sub.t over time. In the following step, shown in FIG. 6D, the new mean m.sub.1 and the new variance P.sub.1 are calculated on the basis of the new sigma points X.sub.1.sup.0, X.sub.1.sup.1, X.sub.1.sup.2, X.sub.1.sup.3, X.sub.1.sup.4. This algorithm is recursive, and each step starting with the prediction step is repeated for each new sampling.

[0049] In the control unit, the mechanical and hydraulic resonant frequencies are compared, and by way of example if their difference approaches or crosses a certain threshold, the control unit 30 varies the setpoint that is transmitted to the accumulator 8 and/or to the gas injection device 20.

[0050] FIGS. 7A, 7B, 7C, and 7D show four examples of how a hydraulic resonant frequency 50 can be controlled in response to an increasing mechanical resonant frequency 51. In the first example, shown in FIG. 7A, the hydraulic resonant frequency 50 may be varied continuously so as to maintain a constant difference relative to the mechanical resonant frequency 51. In the second example, shown in FIG. 7B, the hydraulic resonant frequency 50 is varied stepwise so that the difference between the two frequencies 50 and 51 is not less than a given threshold. It may also happen that the hydraulic resonant frequency 50 cannot be varied over a range of frequencies that is great as the range over which the mechanical resonant frequency 51 can be varied. Under such circumstances, it is also possible, as shown in FIG. 7C, to implement an almost instantaneous change from a hydraulic resonant frequency 50 that is well above the mechanical resonant frequency 51 to a hydraulic resonant frequency 50 that is well below the mechanical resonant frequency 51 (or vice versa). Coincidence between the resonant frequencies takes place only momentarily and does not lead to dangerous resonance. Finally, it is also possible to combine gradual variations in the hydraulic resonant frequency 50 with abrupt changes, as shown in FIG. 7D.

[0051] The support structure of the rocket engine may also present a plurality of variable mechanical resonant frequencies, just as each feed circuit may present a plurality of hydraulic resonant frequencies. Under such circumstances, controlling the volume of gas in the feed circuit solely for the purpose of maintaining the difference between the hydraulic resonant frequency and the mechanical resonant frequency to a value greater than a predetermined threshold might not be adequate. In at least one alternative, the volume of gas may be controlled so as to maximize a function of differences between a plurality of pairs respectively of a hydraulic resonant frequency of the feed circuit and of a mechanical resonant frequency of the structure.

[0052] Thus, in a first example in which the feed circuit has two variable hydraulic resonant frequencies, namely a higher hydraulic resonant frequency f.sub.h,high and a lower hydraulic resonant frequency f.sub.h,low, and the structure presents a variable mechanical resonant frequency f.sub.s, the function that is to be maximized R.sub.opt may satisfy the following equation:

[00001]$R_{\mathrm{opt}}=\min \left(.Math.\frac{f_{h,\mathrm{high}}-f_s}{f_s}.Math.,.Math.\frac{f_{h,\mathrm{low}}-f_s}{f_s}.Math.\right)$

[0053] This function may be a function that is weighted with one or more weighting coefficients. Thus, in a second example in which the feed circuit presents two variable hydraulic resonant frequencies, namely a high hydraulic resonant frequency f.sub.h,high and a low hydraulic resonant frequency f.sub.h,low, and the structure presents two mechanical resonant modes, with a first mode mechanical resonant frequency f.sub.s,1 and a second mode mechanical resonant frequency f.sub.s,2, the function R.sub.opt for maximizing may satisfy the following equations:

[00002]$R_{\mathrm{opt},1}=\min \left(.Math.\frac{f_{h,\mathrm{high}}-f_{s,1}}{f_{s,1}}.Math.,.Math.\frac{f_{h,\mathrm{low}}-f_{s,1}}{f_{s,1}}.Math.\right)$$R_{\mathrm{opt},2}=\min \left(.Math.\frac{f_{h,\mathrm{high}}-f_{s,2}}{f_{s,2}}.Math.,.Math.\frac{f_{h,\mathrm{low}}-f_{s,2}}{f_{s,2}}.Math.\right)$$R_{\mathrm{opt}}=\min \left(R_{\mathrm{opt},1},x_{1,2}.Math.R_{\mathrm{opt},2}\right)$

in which x.sub.1,2 represents a weighting coefficient for the second mechanical resonance mode of the structure.

[0054] Although the present invention is described above with reference to specific embodiments, it is clear that other modifications and changes may be made to those embodiments without going beyond the general scope of the invention as defined by the claims. In particular, individual characteristics of the various embodiments shown may be combined in additional embodiments. Consequently, the description and the drawings should be considered in an illustrative sense rather than in a restrictive sense.