CZ228395A3 - Measuring channel for ultrasound meter of liquid flow - Google Patents

Abstract

The fluid flow measurer includes a duct carrying a fluid. The duct has two transducers spaced apart in it with a measurement portion between them. The measurement portion controls the effects of non-fundamental acoustic modes on a signal received by one transducer. The measurement portion is configured so as to control propagation of non-fundamental modes by decreasing their speed. The measurement portion has an obstruction located in the duct. Fluid flow around the obstruction which is shaped so as to control the effects of non-fundamental modes. The obstruction is an ellipsoid or a shape formed of a core and a hemisphere.

Description

The present invention relates generally to ultrasonic ^ ^ o ^ GPR measuring current * ¹ * - _? / - </ fluid-especially for controlling and / or suppressing the side effects of the acoustic modes of the measurements.

BACKGROUND OF THE INVENTION

It is known that the gas flow rate, as well as the sound velocity in the gas, can be determined by taking two measurements of the ultrasonic pulse transit time, one downstream and one upstream in said current. This is the principle of the operation of an ultrasonic gas meter using a transit time measurement. For this time measurement to be performed with little uncertainty, it is necessary to select the exact pulse sign that can serve as a timestamp. Signal crossing over zero can provide very accurate location over time and is an appropriate time stamp. However, there are several crossings over zero, one of which must be reliably selected as one of the timestamps.

Giant. IE illustrates a typical ultrasonic measurement arrangement in which two sensors 2, 3 are arranged opposite one another in a tube 4 having a cylindrical shape and a circular cross-section with the gas flowing in the direction indicated by the arrow 6.

FIGS. 1A to 1D show the signals that are usually obtained from the ultrasonic pulses transmitted to the circular tube 4. The signal obtained is first shown in FIG.

IA, and continues for some time, as shown in FIG.

IB, 1C and ID, where the marks denote 1x, 2x, 3x and 4x after the initial reception of the start signal.

A specific zero crossing to negative values was selected as the time at which the Time Adjustment is performed. International Application PCT / AU92 / 00314 (WO 93/00569) discloses an electronic fluid flow meter that includes a developed circuitry that selects this particular passage in a two-phase process. Using the signal amplitude, the time preceding the desired pass is selected at which the zero-to-negative pass detector is activated. For this initial moment, reasonable freedom is allowed in setting the time, but it is clear that it must take place before the selected pass through zero but after the previous pass. Otherwise, it would be used instead of that particular timestamp.

It has been found that such an arrangement works satisfactorily because it is based on the amplitude of the signal and how it changes over time. Anything that changes the amplitude of the signal has the ability to interfere with the choice of the start time and hence the choice of the correct crossing over zero. One of the causes of changes in electronic systems is amplification changes. These may be the result of aging, temperature, or other environmental influences. It is common to compensate for such changes with some form of automatic gain control (AGC) that almost eliminates these amplitude changes. However, there are also other causes of amplitude changes that affect individual peak sizes in the signal, although due to AGC the maximum peak size is constant. Further, when the amplitude changes cause an incorrect selection of a zero crossing, the subsequent time setting error will be at least one signal period. This is a serious error because it is a systematic error and the average calculation does not provide an unaffected mean with less uncertainty. These grains 'y & t' and haz.

the existence and dissemination of besides gsueSuaEasfcicx kýclteviéů significantly contributes to us.

</ u. / / V ^ V

SUMMARY OF THE INVENTION

The present invention aims to overcome or solve the above problems by providing a means by which they can be further improves shock resistance * le reduce the influence of 5-1s-jšíeh dR = ^ »^ t ^ íýcK RID ^ p to the sensed sig j © _t Nal, usually its amplitude . _t } - ^ £ * '' 7

In this specification, reference is made to minor acoustical modes. ₄ ·· ^ <C- L · ^ 9 U. /, _{ ^ _Λ nu is a reference to the above-mentioned orders of acoustical modes, and vice versa. J

According to one feature of the invention, the fluid flow measuring system comprises a fluid conduit tube in which two sensors are spaced apart defining a measuring portion for measuring the fluid flow therebetween, the measuring portion being adapted to control the effects of acoustic sub-modes on the signal received by at least one sensor.

In general, the measuring line, st is designed to control the insanity / <'*' - ··. - (LC) (among said sensors, or to reduce their effect on sensor reception, said propagation being suppressed by decreasing speed) of the aforementioned other acoustic modes.

Preferably, the acoustic velocity of the basic sign of No / C ..- ^'77 - / ·' 7 · fy a & b £ e * t. '<^ Ζλ reduced and effects jzedlejších - acousto-ckýeh-vhiů are reduced Snl jfte & J ·· ·· their net contribution to the signal amplitude received by said sensor.

In one embodiment, there is at least one obstacle in the tube that liquid in the tube can bypass, the obstacle having the shape and arrangement in the tube;

Generally, the obstacle in the tube is arranged centrally. Optionally, it can be arranged eccentrically in the tube. Preferably, said barrier in terms of fluid dynamics has a shape suitable for reducing and most preferably minimizing changes in ambient fluid pressure. Generally, this is achieved by arranging said obstacle having a suitable shape and a smooth surface in the measuring portion of the tube so as to minimize or substantially reduce the pressure drop of the fluid in the measuring portion.

In one particular embodiment, the at least one measuring portion section has an enlarged, non-circular cross-section. Typically, this is achieved so that said tube has at least two intersecting wall portions, one of the wall portions being curved and one of the wall portions being non-curving. The curved wall panel may be a partial ellipse, circle, parabola, hyperbola, cycloid, hypocycloid and epicycloid. The intersections of the wall portions may be integral or may be formed of two separate panels. The obstacle, as indicated above, is located in the extended measuring section.

In a typical fluid flow measurement system there is provided a fluid conduit tube provided with spaced apart sensors defining a measuring portion therebetween, the measuring portion including a cylindrical tube and a centrally arranged obstacle therein having a shape suitable for fluid dynamics.

Alternatively, the measuring portion may comprise at least one non-circular section in combination with at least one obstacle having a shape corresponding to the fluid dynamics arranged in the measuring portion, the combination being adapted to suppress the effects of acoustic s at ⁴ d v received by at least one sensor.

Generally, the obstruction is an ellipsoid and is disposed centrally in the tube along the main axis of the sensors so that the cross-section around the tube is annular, the width of the flow portion between the walls of the tube and the obstructions varying along the obstruction.

One or two, three to ten obstacles may be provided in the tube. The inner surface of the tube forming the measuring portion can also be roughened účelemμ to suppress the conduction.

the acoustic modes received by the at least one sensor.

The system described above can be used to perform fluid flow measurements, either for liquids or for gases. Generally, the gas may be a household gas, methane, propane, oxygen, hydrogen, or an industrial gas. preferably, the system may be configured as part of a domestic or industrial gas meter specifically adapted to measure the flow of a so-called natural gas. Applications for liquid flow measurement include liquid hydrocarbon and water flow measurement, but also for ship measurement.

BRIEF DESCRIPTION OF THE DRAWINGS

Exemplary embodiments of the invention are illustrated in the remaining drawings, where in Fig. 2 the acoustic modes are basic and the different order of magnitude present in the ring line, in Fig. 3 the converging current and modes / (0,2) are shown in Fig. 4. Fig. 3 shows the waveform of Fig. 3; Fig. 5 is similar to Fig. 4, but after using one embodiment of the invention; Fig. 8 is a waveform obtained from the embodiment of Fig. 7, Figs. 9A-9D are similar to Figs. IA-ID, but for a tube including an obstacle such as Figs. Figures 6 or 7, Figure 10 shows the effects of changing the temperature of the gas pipeline, Figures 11A and 11B show the signals transmitted in the tube of Figure 10, Figure 12 shows an improvement over Figure 11 by another embodiment of the invention 13A and 13B, respectively The waveforms for temperature fluctuations are compared with and without the obstacle, Fig. 14 shows the transmission of ultrasound by ring technology, Figs. 16A to 16D are views similar to those of FIGS. 1A to 1D but resulting from the combination of FIGS. 6 and 15A.

Exemplary embodiments of the invention

When a sound pulse is sent through the tube or line, _Λ . excites several acoustic modes that carry a signal. These modes can simply be considered as reflections from the tube wall, or they can be considered as vibrations of the taut skin of the drum, but of course spread through space. Both of these analogies have their uses, but both are limited.

The reflection analogy makes it possible to determine that the velocity of propagation along the length of the tube will be slower for modes that are reflected multiple times. The fastest wave or base wave that is not reflected. The plane waves propagate along the tube at the velocity of sound in the space indicated by c. Other xigas due to their reflection or reflections from the wall The tube extends along the tube in a wider range of speeds from nearly c to zero, although precisely, seeing at zero speed does not propagate. Generally, they are slower and less excited, and therefore have less amplitudes than faster y ± 4. For uich / gy having propagation rates less than c / 5, the amplitude is usually 1 or 2% of the major amplitude. Each iodine, except the plane wave, has a cutoff frequency below which it will not propagate in that particular line. There is a large number of modes above this limit for this frequency and line size.

Although these y-ioys are separate units, they overlap, giving an almost continuous signal. Some of the signals are sensed more strongly than others, and thus the overall signal has a shaky appearance as shown in Figures IA-1D, showing that there are modes on both sides that are considerably larger than modes of the same speed.

In Fig. 2, using a drum skin analogy, there are several representations of some of the faster ones present in the circular tube 4. Vid (0,2) is the vibration of the central portion outside the peripheral phase. This is expected because the sensors 2,3 operate in the central part of the tube 4 where they cause a strong excitation. The mode velocity (0.2) is only slightly slower than the plane wave mode velocity (0.1), so the sensed signal is the sum of the two unless a very long tube is used to allow separation of these modes. For a cylindrical tube, the exact phase relationship of the two signals determines the size of the composite signal being sensed. The relation of the phases depends on the diameter of the tube 4, the length of the tube 4 (between the sensors 2,3) and the velocity of the plane wave.

The sound velocity, the plane wave velocity, is given by tP

C = ----- (2)

After where r is the ratio of specific heats, P is the pressure and _e is the density of the gas. Thus, the velocity will depend on the nature of the gas, and for a particular gas at constant pressure and absolute temperature T, the velocity will be given under conditions of a certain standard temperature, in this case 273 K, by the formula = c

273

T

273

Í2)

Changing the C value changes the phase relationship and can greatly change the size of the composite signal. This amplitude can also be affected by the converging current.

Similarly, FIG. 3 shows transmission to the converging current. The shape of the sensed wave in such a situation is shown in Figure 4, from which it is apparent that the second maximum in the envelope is greater than the first. One possible explanation of this is that the converging velocity field bends the sound from the sensor 3, which would otherwise be lost in the area around the periphery of the smaller diameter section. A similar effect can be seen with hot gas in a cold tube (discussed below). This makes it difficult to detect a specific zero crossing.

As shown in FIG. 6, if there is a central obstruction in the tube 4, the central part at ± a (0.2) is indicated by 5, and other more complex vids having a central component are prevented from spreading. . In such a case, the tube 4 has an annular cross section about the obstacle 10, which varies with respect to the aerodynamic shape of the obstacles 10 along its length. The effect of this is apparent from FIGS. 4 and 5, which show the signal sensed through the tube 4 before and after the addition of the central obstacle. Giant. 4 depicts two waveforms, one at zero current, in which the εe (0.2) is approximately two-thirds of the amplitude of a multi-wave wave, and one in a flow where the amplitude ratio is reversed. In Fig. 5, which illustrates an arrangement with a central obstacle 10, the mode size (0.2) is greatly reduced, the zero current waveform and the flow waveform are very similar.

The location of the central obstacle 10 is not critical, but it has been found that the best results are obtained by placing approximately four or five diameters of the tube 4 from the inlet of the tube as shown in Fig. 6. . The optimum diameter of the obstacle 10 is approximately one-half the diameter of the tube 4, but again it is not a critical dimension.

More than one obstacle 10 may be used for further improvement. FIG. 7 shows a tube 13 with two drop-shaped obstacles 11, 12 '. The line 13 is not sensitive to the direction of flow and the nature of the gas with respect to the initial waveform. An example of the sensed signal in the case of air flow (for Fig. 4jf and zero natural gas flow is shown in Fig. 4) The waveform for other combinations of gas type and flow also causes a small change in the sensed waveform.

Giant. 9A to 9D show higher values for a cylindrical tube provided with a central obstacle 10 as shown in FIGS.

Fig. 6. The amplitude of the higher levels decreased and, of course, the behavior of the victt was controlled (0.2).

If the gauge body temperature is considerably lower than the gas temperature, either because the gauge body is very cold, or because the gauge body is at room temperature and the gas is very hot, then νΐ fo, 2) may be more excited than normal. The reason for this is shown in FIG. 10. Here, the hot gas stream 6 is in contact with the cold wall of the tube 4, thereby forming a cold gas layer 7. Energy that would normally reflect from tube 4 and thus actually be wasted, as shown by arrow 9, may enter tube protože »because it is refracted by the cold gas layer 7 as indicated by arrow 8. From equation 2 above, the velocity The noise in the gas is less when the gas is cold than when the gas is hot. The beam of sound entering the region of the cold gas is refracted, as shown in FIG. 10, in the same way as the beam of light is reflected in an optically dense medium. The effect on the sensed signal is apparent from FIGS. HA and 11B for the hot gas flowing through the cold tube.

Giant. 11A shows upstream transmission and FIG. 11B shows downstream transmission. Compared to the zero current waveforms at uniform temperature, the mode (0.2) decreased in Fig. 11A and increased in Fig. 11B.

Assuming that the above explanation is generally correct, this effect can be reduced by removing the cold gas layer 7 on the meter wall. This can be accomplished by means of a gas mixing apparatus which is mounted at the entrance to the tube 4 to introduce a swirling motion into the gas stream, thereby inducing wall mixing. The effect of this apparatus can be seen in the comparison of Figs. 11B and 12. In Fig. 12, the vortex apparatus was used, but otherwise the experimental setup was identical to that of Fig. 11B. The reduction of the víP (0.2) is obvious.

The central obstacle is very effective for suppressing defect ^ (0.2) from this source. This is illustrated in Figures 13A and 13B, where Figure 13 is for a tube without a central obstruction. Room temperature gas flows through the tube and the walls are cooled with refrigerant. Again, the highlight of the cap (0.2) is evident. Giant. 13 is for the same conditions except that a central obstacle has been attached. The waveform is now normal in the sense that it is very similar to the waveform for the same tube having room temperature (Fig. 5). očZu.

Although the discussion has so far been discussed in & (0.2), there are many other modes. These generally have less amplitude than v ^ and (0.2), but under certain circumstances they can cause difficulties by interfering with the propagation of the main acoustic wave system. This is particularly the case with the use of ring technology, which is described in detail in the above-mentioned International Application No. PCT / AU92 / 00314.

In this technology, when a signal is detected, a new pulse is immediately transmitted. Due to the very long docmite of higher order modes on the signal, this retransmission reaches the receiver at a time when the previous signal has still had a significant amplitude. As a result, there is an overlap of the grain signatures with higher vtd * fractions from the previously transmitted pulses. These higher values will add to the bull wave and produce the resulting signal, which is the signal on which the timing of the next transmission is based.

The first portion of the first received signal will be free of any other reception signal, the second received signal will include planar waves from the second transmission and γ 13γ which progress halfway velocity versus the plane wave velocity, c / 2, of the first transmission. The third signal received will be the sum of the plane wave from the third transmission, r3u 'at the rate of c / 2 from the second transmission and the modes at the rate of c / 3 from the first transmission.

Other effects on the fourth, fifth, and other impulses will be apparent to those skilled in the art. This addition process is illustrated in FIG. 12. Further, with reference to FIG. 1, the amplitudes decrease as the mode of order increases so that speeds less than c / 4 have little effect on the received signal.

It can be theoretically shown or experimentally shown that the effect of the current on the received signal is to shift the received signal as a whole along the time axis, so that its shape is essentially retained. The time to reach a particular zero crossing that has been selected as a timestamp will vary with current. The passage time will be longer or shorter depending on whether the signal is transmitted downstream or upstream. Particular portions of the long docmete of the received signal that contribute to the above-described layering process will be those portions of the docmete that are whole multiples of this time after the selected pass through zero. Thus, the exact combination will vary with current.

This addition of other ap & s (higher order), whose phases vary with current, to the plane wave in the ring technology changes the passage time over zero relative to the time that would apply to a single transmission. When the currents are calculated from the passage times obtained from the annular process, higher order modes cause periodic deviations from the linear response as the current varies in the zero current range to the maximum. This problem has been solved by transmitting a reversed pulse once every four transmissions, as described in International Patent Application PCT / AU92 / 00315 (WO 93/00570) entitled Suppression of Vβ when measuring fluid jets from the same Applicant.

However, this does not prevent changes in the amplitude of the received signal. When an inverted pulse is transmitted, the signal from the higher keuunast / modes that was previously added to the amplitude of the plane wave signal is subtracted from the plane wave. Thus, the detection system must be able to reliably detect the correct pass over zero when the amplitudes of the early peaks are in the received s? Aior> h

varied by values corresponding to these for the detection circuit generally easier to operate efficiently when the higher mode magnitudes with speeds of approximately c / 2, c / 3 and so on are as small as possible.

Examining the nature of the modes as shown in FIG. 2 shows that they all have a high degree of symmetry as a result of the circular cross-section of the measuring tube. Thus, in order to minimize ^ drivers / higher orders, it has been observed that it is necessary to break or change the symmetry of the tube as much as possible.

Giant. 15A and 15B show two cross-sections of the measuring tubes, which represent a compromise between the above criteria. Giant. 15A shows a cross-section of a tube 20 that includes, in a set ^1, a semicircular curved portion 21 and two flat faces 22, 23. FIG. 15B depicts a tube 25 having an elliptical portion ^{1.1 and a} curved portion 26 and a single flat face 27. The tubes 20, 25 operate on the principle that orders of magnitude tend to

be reflected from the tube wall, and in both cases the modes reflected from the flat faces 22 and 23 or 27 reflect the curved faces J21 or 26, respectively, which, due to the greater circumferential length (and their surface area), reduce the acoustic úroveňaifrGU , and hence the amplitude of the high order JirtOŮ.

Giant. 16A-16D illustrate the aforementioned deformed cross-sectional tube 20 which has been provided with a single central drop-shaped obstruction. Downstream function has now greatly improved. The use of a spatial obstruction reduces the amplitudes of some of the A ^ geSSteasAds, and also see 0.2). Use of two

-: “Together, Αχ technology provides improved performance for higher modes.

In another configuration, the walls of the measuring tube 4 are roughened by grooves and / or bumps that are comparable in size to half the wave length of the acoustic signal, generally in the range of 0.01 to 8 mm for wider ultrasonic frequencies and preferably in the range of about 0.25 to 2 mm. for frequencies that can be utilized in the apparatus described in the above-mentioned international patent applications. It has been found that such ·· <

the roughening is sufficient to reduce the contribution of higher order acoustic waves, allowing the plane wave mode to dominate at all currents and temperatures without noticeably increasing the frictional resistance to current (pressure drop) within the tube 4.

Appropriate surface treatment of the roughened surface to provide suppression of higher order modes can be accomplished by casting into a suitable mold or by forming a wall of a helical groove tube approaching the length of the acoustic half-plane t,

ny.

The effect of a particular higher mode is reduced because the energy it contains diffuses in the finite time. That is, if it is reflected from a perfectly cylindrical surface, all the faces reach the sensor simultaneously and its full effect is sensed. Roughening or warping of the surface causes part of the front of the mode to travel along a somewhat longer path, thereby reducing the total instantaneous contribution. Excessive roughening of the tube surface can greatly increase the pressure drop despite decreasing the effect of higher order modes.

From the foregoing, it will be appreciated that the addition of an obstruction, a change in the cross-section of the tube 2 of regular shape, and / or roughening of the surface or the formation of grooves on the tube surface between the sensors causes a reduction in the propagation rate of high order acoustic modes without decreasing the propagation rate of the basic mode.

An important advantage of the mode suppression technology described herein is that it enables the electronic gas meter, which includes at least one described arrangement, to be calibrated by air before being inserted into the gas pipeline. With traditional prior art arrangements, this technology has led to incorrect calibration due to the mode response of ultrasound in a combustible gas, for example, a natural gas other than air. In such cases, the different response time results in a different detection time of the received ultrasonic RF pulse. Since the mode response is substantially reduced according to the principles described herein, the detection times for the gases are identical.

The foregoing describes only a few embodiments of the invention, and modifications can be made to those skilled in the art within the scope of the invention.

Claims (12)

A measuring channel for an ultrasonic fluid flow meter, comprising a fluid conduit tube in which two spaced and facing sensors are provided, between which a fluid flow measuring portion is located, characterized in that in the measuring portion (4) at least one obstacle (10, 11, 12) for controlling the effects of higher harmonic acoustic modes is located centrally with respect to its cross-section. Measuring channel according to claim 1, characterized in that each obstacle (10, 11, 12) is dynamically shaped to reduce the change in pressure of the liquid around it. Measuring channel according to claim 2, characterized in that each obstacle (10, 11, 12) has a smooth surface in the measuring portion (4). Measuring channel according to claim 2, characterized in that the obstacle (10) is an ellipsoid. Fifth Measuring channel according to claim 2, yzn mares I C s s e _{Z [T} he that obstacle (11,12) is made \ cone and a hemisphere / spójetíé.svQu circular basic ^ ^^ ^ rT? 'τ; · 'T' ?. Measuring channel according to any one of claims 1 to 5, characterized in that the at least one section of the measuring part (4) has an extended non-circular cross-sectional shape for controlling the effects of higher harmonic acoustic modes. Measuring channel according to claim 6, characterized in that the tube (1) in the section of the measuring part (4) has at least two connecting wall parts (21,22,23, 26, 27), at least one of the parts (21,26) The wall is curved and at least one of the remaining wall portions (22, 23, 27) is flat. The measuring channel of claim 7, wherein the curved wall portions (21, 26) have a shape selected from the group consisting of an ellipse, a circle portion, a parabola, a hyperbola, a cycloid, a hypocycloid, and an epicycloid, and a non-curved portion (22). 23,27) walls are straight. Measuring channel according to claim 8, characterized in that the connection between the wall portions (21, 22, 23, 26, 27) is formed integrally and continuously. Measuring channel according to any one of claims 6 to 9, characterized in that the tube (1) has an internal surface with irregularities in the sections of the measuring portion (4). Measuring channel according to claim 10, characterized in that the irregularities comprise grooves or roughening. Measuring channel according to claim 11, characterized in that the obstruction (10) has the shape of an ellipsoid arranged between the sensors (2,3) centrally in the measuring portion (4) about the longitudinal axis of the tube (1) where with the obstacle (10) there is an annulus between the wall of the tube (1) and the surface of the obstacle (10). _Λ /