GB2160318A

GB2160318A - Karman vortex flowmeters

Info

Publication number: GB2160318A
Application number: GB08502849A
Authority: GB
Inventors: Noriomi Miyoshi; Michihiko Tsuruoka; Mutsumi Nagumo
Original assignee: Fuji Electric Co Ltd
Current assignee: Fuji Electric Co Ltd
Priority date: 1981-11-10
Filing date: 1985-02-05
Publication date: 1985-12-18
Also published as: GB2112938A; GB2159946B; GB2112938B; GB2160314B; GB8502843D0; GB2159946A; GB2160316A; GB8502847D0; GB8502848D0; GB2160314A; GB2160317A; DE3241988C2; GB2160318B; GB8502849D0; GB2160316B; GB2160317B; DE3241988A1; GB8502845D0

Abstract

A flowmeter for measuring the quantity of air inhaled by an internal combustion engine comprises a signal processing circuit 25 for processing a detected vortex detection signal housed in a container 134 communicating via pipe 130 with an air inlet passage on the downstream side of a throttle valve (106, Fig. 32, not shown), the other side of container 134 being in communication with the atmosphere via pipe 132 so that an air flow will be induced through the container due to negative pressure on the downstream side of the throttle valve so that the components of circuit 25 will be cooled. The cooling air prevents a temperature rise of the elements of circuit 25 when the temperature within the engine compartment rises. <IMAGE>

Description

1

GB2 160318A

1

SPECIFICATION

Karman vortex flowmeters, generators and detectors and associated processing circuits 5 and their application to monitoring internal combustion engines

The present invention relates to a Karman vortex flowmeter particularly one arranged for 10 detecting the frequency of an oscillatory member which is caused to oscillate by oscillatory pressure variations occurring in the vicinities of the opposite sides of a pillar-form vortex generator disposed transversly to the flow of a 15 fluid. More particularly, the present invention relates to improvements in a vortex generator in such a Karman vortex flowmeter, improvements in detector means for detecting the vortices or oscillatory pressure variations pro-20 duced by the vortex generator of the Karman vortex flowmeter and improvements in circuits for processing the signal detected by the vortex detector of the Karman vortex flowmeter. The invention also relates to improve-25 ments in arrangements of such Karman vortex flowmeters for measurement of the quantity of air inhaled by an internal combustion engine of an automobile or the like.

This application has been divided out of 30 Application No. 82.32154.

A Karman vortex flowmeter of the type to which the present invention pertains has been already proposed in Patent Application No. GB 2,103,795 A and is shown in Figs. 1 and 35 2 of the accompanying drawings. Fig. 1 is a front elevation of this known Karman vortex flowmeter, and Fig. 2 is a view of the vortex detecting apparatus of the flowmeter.

Referring to Fig. 1, there are shown a pipe-40 line 1, a vortex generator 2 for generating Karman's vortex streets, openings 3 and 3', a vortex detector 4, and optical fibre 5, and a processing circuit 6 for processing the signal detected. The vortex detecting apparatus com-45 prises the devices 4 to 6. As shown in Fig. 2, the vortex detector 4 is provided with an oscillator chamber 43 having a substantially isosceles triangular cross-section. An oscillatory plate 44, which is caused to oscillate by 50 vortices occurring near the vortex generator 2, is installed within the oscillation chamber 43. Pressure changes due to Karman's vortices are introduced through openings 41 and 42. The signal processing circuit 6 consists of a 55 light emitting device 6a, a light receiving device 6 b and a waveform shaping circuit 6 c.

In the operation of this prior art flowmeter, when Karman's vortex streets are generated in the vicinities of the opposite sides of the 60 vortex generator 2 installed in position within the pipeline 1, pressure changes resulting from the vortices are transmitted through the openings 3,41 or 42 to the oscillatory plate 44 thus displacing the latter. These vortices 65 occur alternatiely in the vicinities of the opposite sides of the vortex generator 2 and cause the plate 44 to oscillate. Light from the light emitting device 6a of the signal processing circuit 6 is directed at the oscillatory plate 44 via the optical fibre 5a and reflected from the surface of the vibratory plate, and is then transmitted to the light receiving device 6b through the optical fibre 5 b. Since the end surfaces of the optical fibres 5a and 5fa are oriented substantially parallel to the major surface of the plate 44 in this construction, the quantity of light entering the light receiving device 6 b varies in dependence on the displacement of the oscillatory plate 44. As such, the light receiving device 6b receives a signal corresponding to one oscillation of the plate 44, thus permitting detection of the oscillation frequency of the vortices.

Generally, flowmeters of this kind have the disadvantage that generation of vortices becomes unstable and weak vortices are produced at lower flow velocities, thus rendering accurate measurement difficult.

The problems of vortex generators will now be discussed with reference to Fig. 3 which is a schematic diagram of one example of prior art flowmeter, in which two pillars K, and K2 are disposed at respective positions spaced in the direction of the flow. Pillar K, on the upstream side is for producing vortices and is triangular in cross-section, while pillar K2 on the downstream side is shaped in cross-section like a plate (with a rectangular cross-section). The triangular pillar has a contour resembling a streamline relative to the flow, and is therefore advantageous in that it does not cause a great pressure loss. However, it does not readily produce vortices and, particularly when the flow velocity is low, measurement is difficult. Also, in such a construction, detection of vortices is carried out on the downstream side of the pillars and, accordingly, if pulses or the like occur in the medium within the pipeline, the resulting variation in flow velocity or pressure constitutes noise, so making accurate detection of vortices impossible.

Also, another flowmeter is known in the vortex generator of which a pillar having a substantially isosceles trapezoidal cross section is so disposed that the base of the trapezium is perpendicular to the direction of flow in order to produce vortices over a relatively wide range of rate or velocity of flow. Unfortunately, it is disadvantageous in that a large pressure loss results, because the surface normal to the flow is flat. Still another prior art vortex generator has a plurality of pillars which are disposed at regular intervals along the flow for producing vortices. This is, however, disadvantageous in that it is complex and expensive in construction.

The problems of vortex detectors will now be discussed. Generally, in a flowmeter of this kind, Karman's vortex streets occurring on the

70

75

80

85

90

95

100

105

110

115

120

125

130

2

GB2 160318A

2

downstream side of a pillar or pillars are very feeble when the velocity of flow is low.

Hence, a highly sensitive detector is required. Detectors using highly sensitive heated wire or 5 ultrasonic waves have the disadvantage that they electrically amlify minute analogue signals and so temperature characteristics and the degree of stability of the detector or detecting circuit considerably affect the mea-10 suring accuracy and range. Accordingly, it is desirable that detectors used for detection of vortices when the flow rate is low be relatively unaffected by these temperature and stability factors, and furthermore that they be highly 15 sensitive.

Of the prior art apparatus, one example in which an oscillatory plate is displaced by vortex pressure to facilitate signal processing is disclosed in Japanese Utility Model publica-20 tion No. 21501/1971, where an oscillation chamber is provided within a vortex generator and an oscillator in the form of a plate is fixed by one end to the wall of the chamber. The velocity or rate of flow is derived from the 25 oscillation frequency of the oscillator. This apparatus has the advantage that it is simple in construction, because pressure changes resulting from occurrence of vortices are directly detected as displacement or force. However, 30 as the oscillator carries out bending oscillations with one end fixed, it frequently malfunctions due to external vibration. Particularly when flow velocity is low, pressure changes resulting from the occurrence of vor-35 tices are quite small and so it becomes impossible to discriminate oscillations due to votices from those due to external vibration, so that accurate detection of vortex frequency is no longer feasible.

40 Of conventional vortex detectors, a relatively sensitive vortex detector has an oscillatory member in the form of a plate of light resin arranged so that it can rotate on a rotatable shaft. This vortex detector is disad-45 vantageous in that when the flow velocity is high and vortex pressure is also high, an excessively large displacement or force is applied to the plate, because the plate is displaced in proportion to vortex pressure. As a 50 result, it becomes impossible to detect vortices with accuracy, and the detector may be damaged. An example of an apparatus in which a plate undergoes bending in order to detect vortex pressures is disclosed in Japa-55 nese Patent publication No. 36933/1980. This apparatus is able to detect vortices in a stable manner even when flow velocity is high, but it cannot detect vortices when flow velocity is low, because it makes use of bend-60 ing or flexural oscillation of the plate.

The problems of electrical circuits for processing detected signals will now be discussed. In general, a DC component corresponding to a given quantity of ambient light 65 which is received when an oscillatory plate is at rest in a position of equilibrium is added to an AC component proportional to any change in light quantity caused by oscillation of the oscillatory plate. The resultant composite signal is detected by light receiving device 6b (Fig. 2). Conventionally, in order to shape the waveform of such a signal into a desired form, a method has been proposed in which the signal is simply amplified in alternating form. However, the frequency of vortices varies widely from 10 Hz to 1 KHZ. Also, generally, such a processing circuit has the disadvantage that it has a complex circuit configuration.

It has also been suggested to use means in which the output signal from a vortex detector is compared with a given set value to shape the waveform into a desired form. Since the received light quantity varies, because of contamination of the optics for instance, the DC component of the output signal also varies, but the given set value does not follow this change. Hence, the circuit is not able to detect the vortex frequency.

Karman vortex flowmeters are useful for the measurement of the quantity of air inhaled by an internal combustion engine of an automobile or the like. However, conventional Karman vortex flowmeters have disadvantages in such an application. In nearly all engines of recent automobiles, oil vapour or leakage gas (which is also called "blow-by" gas) produced within the crank case is returned to the inlet passage through an air cleaner or the like for preventing air pollution, If a flowmeter is placed in the inlet passage, the optics become contaminated with long term usage, resulting in a decrease in the detection sensitivity.

A flowmeter used in such an engine is usually readily affected by temperature or electrical noise within the engine compartment. In the past, as disclosed in Japanese Utility Model publication No. 28998/1980, an attempt has been made at cooling the electrical circuit for detecting vortices in a flowmeter using inhaled air to counteract temperature effects. However, when less air is inhaled (as during idling) the cooling effect is decreased and the temperature rise becomes great, so reducing the reliability of the electrical circuit. Further, means for compensating for the temperature rise become very costly. In contrast with this, an arrangement has been proposed in which the electrical circuit for vortex detection is installed at a separate location, for example in the passenger compartment, which provides more favourable conditions (e.g. temperature) than the engine compartment. However, in the course of transmission of a feeble electrical signal to such a separately located electrical circuit, the signal is significantly affected by electrical noise such as arises from the ignition system, thus making accurate measurement indication impossible. /

In a flowmeter as disclosed in Japanese

70

75

80

85

90

95

100

105

110

115

120

125

130

3

GB2160 318A 3

Utility Model publication No. 28998/1980, a processing circuit for processing detected signals is housed in a bypass passage where the circuit is cooled with air flowing through the 5 bypass passage. In fact, when less air is inhaled by the engine, the signal processing circuit can get hotter since at that time a minute quantity of air flows through the bypass passage, because it has a greater flow 10 resistance than the main inlet passage. Thus, the cooling effect deteriorates.

According to this invention there is provided a Karman vortex flowmeter for measuring the quantity of air inhaled by an internal 15 combustion engine of an automobile or the like, said flowmeter comprising a signal processing circuit for processing a detected vortex detection signal, and container housing the signal processing circuit, one side of the 20 container being in communication with an air inlet passage on the downstream side of a throttle valve in the air inlet passage, the other side of the container being in communication with the atmosphere so that an air flow 25 will be induced through the container due to negative pressure on the downstream side of the throttle valve so that the components of said signal processing circuit will be cooled.

Embodiments of this invention will now be 30 described, by way of example, with reference to the accompanying drawings in which:-

Figure 1 is a vertical sectional view of a prior art Karman vortex flowmeter;

Figure 2 is a schematic partially sectional 35 view of part of the flowmeter shown in Fig. 1;

Figure 3 is a schematic horizontal section of part of a prior art flowmeter such as is shown in Fig. 1;

Figure 4 is a horizontal section of a Karman 40 vortex generator for a flowmeter embodying the present invention;

Figures 5 to 10 are graphs of various characteristics of the vortex generator shown in Fig. 4;

45 Figure 11 is a partially sectional view of a vortex pressure detector in a flowmeter embodying this invention and taken in a plane transverse to the flow of fluid;

Figure 12 is a plan view of part of the 50 detector shown in Fig. 11;

Figure 13 is a sectional side elevation of the vortex detector shown in Fig. 11 ;

Figure 14a is a sectional view of a displacement detecting sensor for use in the detector 55 shown in Fig. 11;

Figure 14b is a graph of a characteristic of the operation of the sensor shown in Fig. 14a;

Figure 15 is a view of a different embodiment of displacement detecting sensor; 60 Figure 16 is a sectional side elevation of a further embodiment of vortex pressure detector;

Figure 7 7 is a sectional side elevation on a larger scale of a portion of the detector shown 65 in Fig. 16;

Figure 18 is a sectional view of a still further embodiment of vortex pressure detector taken in a plane transverse to the flow of fluid;

Figure 19 is a circuit diagram of a circuit for processing detected signals from the detector shown in Fig. 18;

Figure 20a is a view of a displacement detecting sensor of the detector shown in Fig. 18;

Figure 20b is a graph of a characteristic of the operation of the sensor shown in Fig. 20a;

Figure 2 7 is a view of a still further embodiment of displacement detecting sensor;

Figures 22a to 22c are graphs of the operating characteristics of an oscillator in a flowmeter embodying this invention;

Figures 23a to 23c are graphs of the characteristics of the output signal from a flowmeter embodying this invention;

Figures 24 and 25 are views similar to Figs. 12 and 1 3 and schematically illustrating a still further embodiment of vortex pressure detector;

Figures 26A and 26B are graphs of waveforms of detected vortices using a further embodiment of flowmeter;

Figures 27 and 28 are circuit diagrams of respective embodiments of circuits for processing detected signals;

Figure 29 is a schematic view of the arrangement of a Karman vortex flowmeter in an automobile engine and embodying this invention;

Figure 30 is a view of a specific embodiment of flowmeter for use in the arrangement shown in Fig. 29;

Figures 31 and 32 are schematic views of further arrangements in each of which a flowmeter embodying this invention is installed in an automobile; and

Figure 33 is a vertical section of the flowmeter shown in Fig. 32.

A first example of a Karman vortex flowmeter embodying this invention will now be described with reference to Figs. 4 to 10.

More specifically, Fig. 4 is a cross sectional view of a vortex generator of the flowmeter, and Figs. 5 to 10 are graphs illustrating the characteristics of the flowmeter.

Referring to Fig. 4, there are shown an upstream pillar 2a having a substantially isosceles triangular cross-section and a downstream pillar 2b having a substantially isosceles trapezoidal cross-section. These pillars are so disposed that the base of the triangular cross-section of the upstream pillar 2a and the base of the trapezoidal cross-section of the downstream pillar 2b are parallel to and adjacent each other with a predetermined spacing of "I" in the direction of the flow, these bases being perpendicular to the flow. Further, the projected widths of the upstream pillar 2a and of the downstream pillar 2b at right angles to the direction of flow, that is, representative

70

75

80

85

90

95

100

105

110

115

120

125

130

4

GB2 160318A

4

widths d, and d2. respectively, are equal. In Fig. 4, the triangular cross-section has a vertical angle of a, and the trapezoidal cross-section has a height of "h". The equal sides 5 of the trapezium form an angle of ft. These values are appropriately determined as described later. The downstream pillar 2b is provided with openings 4a and 4b for conveying the vortex pressures produced on opposite 10 sides of the pillar near the longitudinal edges of the latter.

The results of experiments on the effects of varying the vertical angle a of the isosceles triangular cross-section of the upstream pillar, 15 the height h and the angle ft formed by the equal sides of the isosceles trapezoidal cross-section of the downstream pillar, the spacing / between the pillars and the representative widths d, and d2 of the pillars, upon the 20 generation of vortices and the stability of generation have revealed several points as follows.

1) When the representative width of the upstream pillar is substantially equal to the

25 representative width of the downstream pillar (d, = d2=d), vortices are generated most stably and they can be detected with a good S/N (signal-to-noise ratio).

2) When the vertical angle a of the isos-30 celes triangle of the upper pillar is defined by

90°<a<120°

the pressure loss is small and the detected 35 vortex waveforms are stable. When the vertical angle is less than the above-defined range, noise having frequency components lower than the vortex frequency is produced. Similarly, when the vertical angle exceeds the 40 range, noise is generated and a large pressure loss occurs.

3) When height h of the isosceles trapezium of the downstream pillar is such that h<d/2, and the angle ft formed by the equal

45 sides of the trapezium is such that /?<40°, it is possible to detect vortices even at low flow velocities with good linearity. Specifically, as may be understood from the experimental results shown in Figs. 5 and 6, when /J<40° 50 and h<0.5 d, the Strouhal number ( = vortex frequency X representative width/flow velocity) is substantially constant (i.e. a variation less than ± 3%) and this offers a practical characteristic.

55 4) When the spacing I between the pillars is selected so that it has a value between 0.2d and 0.3d, a wide measuring range and good linearity are obtained. It will be seen that when the spacing is within the range, the 60 velocity at which vortices can be detected can be lowest, as shown in Fig. 7, and linearity at high flow velocities is improved, as indicated in Fig. 8.

Figs. 9 and 10 illustrate characteristics of a 65 flowmeter whose design is based on the above results. In this embodiment, the upstream isosceles triangle has a = 90°, the downstream trapezium has h = d/3 and ft = 40°. It is to be noted that the measurement was carried out with air at atmospheric pressure. Fig. 9 shows that non-linearity is kept to less than ± 3% over a wide range from 1 to 60 m/sec or more of flow velocity, thus offering a practical characteristic. Curve X of Fig. 10 indicates the pressure loss characteristic of a prior art flowmeter in which the downstream pillar has an isosceles trapezoidal cross-section and the representative widths of the pillars are equal to each other, while curve Y of Fig. 10 indicates the pressure loss characteristic of the flowmeter embodying the present invention and having pillars as shown in Big. 4. It can be seen from Fig. 10 that the present embodiment has roughly halved the pressure loss in the case of the prior art flowmeter.

The upstream pillar of the isosceles triangular cross-section and the downstream pillar of isosceles trapezoidal cross-section have the same projected widths transverse to the flow direction and are arranged as a whole generally in the form of a pentagon, though they are separated by a predetermined spacing and optimal numerical ranges and relationships are determined by experiment, whereby the pressure loss is small and a characteristic, superior in linearity to those of conventional flowmeters, can be obtained over a wide measuring range. Further, differing from conventional apparatus in which sensors for detection vortices are disposed on the downstream side of a pair of pillars, the opposite sides of the downstream pillar in the flow are provided with openings for introducing pressure changes due to vortices and the detection of pressure difference is carried out via these openings. As a result, the flowmeter can detect vortices with a good S/N over a wide range of flow velocity without being affected by pressure variation caused by pulsation of flow.

An embodiment of Karman vortex flowmeter will now be described with reference to Figs. 11 to 1 5.

Referring to Fig. 11, there are shown a pipeline 1, a vortex generator 2 for generating Karman's vortices and a vortex detector 4. The vortex generator 2 is constructed as shown in Fig. 4, for example, and is provided with slits 4a and 4b on opposite sides of downstream pillar 2b near the longitudinal edges for conveying changes in vortex pressure.

As shown more clearly in Fig. 12, an oscillator 8 consists of a thin (i.e. thickness about 20 n) metal plate, and includes an oscillatory plate 9 on which vortex pressure acts, a pair of taut bands 10a and 10b which support the plate 9 symmetrically on a central axis passing through the centre of gravity of the plate

70

75

80

85

90

95

100

105

110

115

120

125

130

5

GB2 160318A

5

to allow torsional oscillation of the plate, and a frame 11 constituting the outer fixed ends of the bands. These components are formed simultaneously, by a stamping operation, out 5 of a single metal plate having a substantially uniform thickness.

The oscillatory plate 9 is in a state of mass equilibrium relative to the central axis. The taut bands 10a and 10b are so designed that 10 the torsional spring constant defined by the dimensions of the bands is extremely low to allow a sufficient angular displacement of the oscillatory plate 9 even with changes in vortex pressure which are minute, and so that the 15 resonance frequency is as low as possible, Spaces 11 a and 11 fa are formed by the stamping operation.

Referring again to Fig. 11 and also to Fig. 13, a housing 12 which houses the oscillator 20 8 consists of a lower plate 1 3 and an upper plate 14. The plates 13 and 14 are provided with grooves which are substantially the same in shape and disposed opposite each other. The shape of the grooves corresponds to the 25 shape of the oscillator 8.

By stacking the lower plate 13, oscillator 8 and the upper plate 14 on flange 15 of the vortex generator 2 in sequence, the oscillator

8 is supported and, at the same time, an 30 oscillation chamber 16 and chambers 17a and 1 7fa (Fig. 13) for housing the taut bands are formed.

The chamber 1 6 is divided substantially equally into an upper compartment 26 and a 35 lower compartment (19a, 1 9b) by the oscillatory plate 9 of the oscillator 8. The lower compartment bounded by the oscillatory plate

9 and the lower plate 13 is further partitioned into equal sub-chambers 19a and 19fa by a

40 protrusion 18 of the lower plate 1 3 transverse to the axis of oscillation of the plate 9. The sub-chambers 19a and 19faare in communication with the slits 4a and 4fa of the vortex generator 2 through apertures 20a and 20fa, 45 respectively. The protrusion 18 acts to prevent circulation of fluid betwen the sub-cham-bers 19a and 19fa so as to transmit changes in vortex pressure from the slit 4a or 4fa to the oscillatory plate 9 without loss.

50 As long as torsional oscillation of the plate 9 is not hindered, the space between the protrusion 18 and the plate 9 should be as narrow as possible, and is preferably in the order of 0.1 to 0.2 mm, for example. For 55 similar reasons the space above the plate 9 in the chamber 16 is preferably of the same order.

Referring to Fig. 13, there is shown an adjusting screw 21 for applying tension to the 60 taut bands 10a and 10fa, and this screw is disposed on the central axis of the taut band 10b. The screw 21 applies a tension to a portion of the band between the fixed end of the taut band 10b and a protrusion 22 of the 65 lower plate 13 for preventing bending oscillation of the oscillator 8. In fact, as described later in connection with Figs. 16 to 17, the taut band is preferably subjected to pressure by the screw via a spring.

As shown in Fig. 11, reflecting type optical fibres 5 for detecting the angular displacement of the oscillator 8 provide two optical paths 5a and 5fa respectively for transmission and reception. The optical axes are normal to the upper surface of the oscillatory plate 9 and the optical paths open into the compartment 26. That is, the exposed end portions of the optics are totally contained within the compartment 26, so that the optics are protected from coming into direct contact with fluid. The other ends of the optical fibres are provided with a light emitting device 6a and a light receiving device 6 fa, respectively. A detector circuit 6 is shown in Fig. 11 as including the light emitting device 6a, the light receiving device 6 fa and an amplification and waveform shaping circuit (not shown).

In the operation of the apparatus shown in Fig. 11, when a vortex is generated on the side adjacent the slit 4fa of the vortex generator 2, for example, the pressure near the slit 4fa becomes lower than the pressure near the opposite slit 4a. Accordingly, the pressure within the sub-chamber 19fa communicating with the slit 4b becomes lower than the pressure within the sub-chamber 1 9a communicating with the opposite slit 4a. Pressure applied to the upper surface of the oscillatory plate 9 is substantially uniform over the whole surface. With respect to the underside of the plate 9, since the pressure within the sub-chamber 19fa is lower than that within the sub-chamber 19a, a clockwise moment corresponding to the pressure difference acts on the plate 9. The plate 9 is therefore pivoted in the clockwise direction, but the bottom surface and the upper surface of the chamber 16 limit the amplitude of the pivoting movement.

When a vortex is generated on the opposite side of the vortex generator, the pressure within the sub-chamber 1 9a becomes lower than that within the sub-chamber 19b, and the plate 9 is pivoted in the counterclockwise direction. The amplitude of this pivoting movement is similarly limited by the bottom and upper surfaces of the chamber 1 6.

Thus, generation of such a pair of vortices causes a single torsional oscillation of the oscillator 8. As the amplitude is limited by the wair surfaces of the chamber 16, the amplitude can be arranged to be substantially constant irrespective of changes in the vortex pressure levels. Since the plate 9 is essentially balanced about the central pivot axis, inertial forces resulting from external vibrations cancel out about the pivot axis. Hence, no torsional vibration will occur. Further, as a tension is always applied to the taut bands 10a and 10b, the oscillator 8 is inhibited from following external vibrations in the vertical direction.

70

75

80

85

90

95

100

105

110

115

120

125

130

6

GB2 160318A

6

Thus, to this extent, the oscillator is free from external vibration effects. Since the tension applied to the taut bands scarcely affects the torsional spring constant, the resistance to 5 vibration can be improved without decreasing the sensitivity in detecting vortices.

Thus, generation of vortices causes the oscillator 8 to oscillate torsionaliy within the chamber 16. To regularise these oscillations 10 over a wide range of vortex frequency, for example from 10 Hz to 1 KHZ, it is important to transmit changes in vortex pressure directly to the plate 9 with minimal loss. For this purpose, the slits 4a and 4b in the vortex 15 generator 2 are arranged to convey changes in vortex pressure into the sub-chambers 19a and 19 b via the shortest distance, and the protrusion 18 is provided to minimise leakage communication between the sub-chambers 20 19a and 19b. Further, the gap between the periphery of the oscillatory plate 9 and the wall surfaces of the oscillation chamber 16 is so formed that leakage between the compartment 26 and the sub-chamber 19a or 19b is 25 also minimised. In this way the vortex pressure acts on the oscillatory plate with minimal loss, so providing stable detection of vortices.

Determination of the oscillation frequency of the oscillator 8 will now be described with 30 reference to Figs. 11, 14 and 15.

The oscillator frequency is determined by measuring the change in a reflected light quantity from the upper surface of the plate 9 using the optical fibres 5. Specifically, the two 35 optical paths 5a and 5b are arranged in a random manner at the end surface 31 opposite the oscillatory plate 9 but the optical axes are substantially normal to the surface of the plate 9. As the quantity of light reflected from 40 the plate 9 decreases with the angle of rotation of the reflecting surface of the oscillatory plate, as indicated in Fig. 14(b), one oscillation cycle of the oscillator produces two light pulse outputs. Since the displacement ampli-45 tude of the oscillator 8 is substantially constant, the light outputs will also be substantially constant. Hence, it is possible to detect the vortex frequency by means of a simple electrical circuit without the necessity for 50 aligning optical axes. Furthermore, the optics consisting of the optical fibres 5 and the reflecting surface is located within the compartment 26 of the chamber 16 and so does not come into direct contact with fluid, so that 55 contamination of the optics by the fluid is substantially avoided.

In the detector, the optical axes of the optical fibres 5 are substantially normal to the surface of the plate 9. Instead, it is possible to 60 angle the optical axes of the fibres 5 to the normal to the surface of the plate 9 by the maximum angular amplitude 0m of the oscillations of the plate 9, as shown in Fig. 1 5. This configuration is advantageous in that a minute 65 angular displacement can be detected with a good sensitivity and an output signal of substantially sinusoidal waveform is derived to facilitate signal processing as can be seen from Fig. 14(b), because a substantially constant-gradient portion of the characteristic curve is used. Of course, the peak of the reflected light quantity is shifted to the right in Fig. 14(b).

It should be noted that this embodiment of vortex pressure detector is not limited to the optical fibre construction, but any suitable construction can be employed as long as it can effectively detect changes in the quantity of reflected light.

Thus, in contrast to the prior art described earlier, the oscillator is balanced about its axis of oscillation which passes through the centre of gravity of the oscillator being supported by the taut bands on the axis so as to allow torsional oscillation about the axis. Further, a tension force is applied to the bands to inhibit bending vibrations, so that the oscillator is substantially unaffected by external vibrations. Furthermore, the detector has the advantage that although it has a good resistance to vibration, it is capable of detecting vortices with an enhanced sensitivity, in spite of the fact that the torsional spring constant is small.

Also, the chamber for housing the oscillatory plate is effectively divided into two substantially equal compartments by the oscillatory plate. One of the compartments is further partitioned symmetrically into two sub-chambers which are provided with openings for conveying changes in vortex pressure into the sub-chambers to cause the oscillatory plate to oscillate. The amplitude of this oscillation is limited by the walls opposite the oscillatory plate so that oscillation can be obtained having a substantially constant amplitude which is independent of vortex pressure. The other compartment of the oscillation chamber is provided with optics for detecting the displacement of the oscillatory plate which itself separates the two compartments. As a result, the optics of the detector hardly come into direct contact with the fluid and so it is possible to detect vortices while avoiding serious adverse effects of contamination by the fluid. The optics comprise two light paths each opening at one end into the appropriate compartment opposite the oscillatory plate, the light emitting and receiving devices being disposed at the respective other ends of the light paths, so that oscillation of the oscillatory plate causes changes in the quantity of light reflected from the plate.

Referring again to Fig. 1 3, the adjusting screw 21 applies a tension to the taut band 10b. If the material of the oscillator 8 differs from the material of housing 12 (that is, of the lower plate 1 3 and upper plate 14), then heat will give rise to a significant difference between the tensions in the taut bands 10a and 10b because of the difference between

70

75

80

85

90

95

100

105

110

115

120

125

130

7

GB2 160318A

7

their thermal expansions. Also, even if the material of the oscillator 8 is the same as that of the housing 12, when a transient thermal change takes place, there occurs a tempera-5 ture difference between the oscillator 8 and the housing 12. The result is that tensions in the bands differ greatly, whereby a decrease in the resistance to vibration and damage to the taut bands may occur. These difficulties 10 can be overcome by employing counter-measures such as will now be described in detail with reference to Figs. 16 and 17.

Referring to Figs. 16 and 1 7, there is shown a tension applying mechanism 30 for 1 5 applying a tension to the taut bands 10a and 10b, including a compression spring 31, a cap 32 and a piston 33 for applying pressure to the taut band 10b. The piston is made of a light resin and is in the form of a hollow 20 cylinder whose closed bottom is provided with a rectangular protrusion 35. When the spring is compressed, the protrusion is guided by a rectangular guide aperture 34 formed in the upper plate 14 and arranged to prevent the 25 piston 33 from rotating. The piston applies a tension force to a portion of the band 10b between the fixed end and a protrusion 22 formed by the lower plate 1 3 to inhibit bending vibration of the oscillator 8. The spring 31 30 is housed within the piston 33, and the cap 32 prevents it from escaping.

When a rapid temperature change occurs a temperature difference arises transiently between the oscillator 8 and the housing 12 35 (that is, the upper plate 14 and lower plate 1 3). Therefore, due to the difference between their thermal expansions (even though they consist of the same material) the lengths of the taut bands become different from each 40 other. However, the compression spring 31, having a small spring constant, applies a tension to the taut bands so that there is a bend in the strip of material forming the bands as described above, and therefore the 45 bands can absorb variations in length resulting from thermal expansion. Also, as the tensions are kept substantially constant, damage to the taut bands and decrease in the resistance to effects which would otherwise be 50 caused by external vibrations are prevented.

In the detector shown in Fig. 11, the light emitting and receiving devices are connected to ends of the optical fibres remote from the other ends which are opposite the oscillatory 55 plate. However, as will be described below, the light emitting and receiving devices may be adjacent and opposite the oscillatory plate.

Such an embodiment will now be described with reference to Figs. 18 to 21. 60 Referring to Fig. 18, a light emitting device 23 and a light receiving device 24 are so disposed that they are adjacent and facing the upper surface of the oscillatory plate 9 and their optical axes pass through the central axis 65 of rotation of the plate 9. A detection circuit

25, which energises the devices 23 and 24 and processes the output signals, includes (Fig. 19) a coupling circuit 27a for deriving only AC components from the output signal from the light receiving device 24 and comparator 27b for pulsing this signal.

Referring to Figs. 18 to-21, determination of the number of oscillations or the oscillation frequency of the oscillator 8 will be described. In this case, determination of the oscillation frequency is carried out by detecting changes in the quantity of reflected light from the upper surface of the plate 9 using the devices 23 and 24. Since the quantity of light reflected from the plate 9 decreases with increasing angle of rotation of the reflecting surface of the plate, as will be clear from Fig. 20(b), the output signal from the light receiving device 24 changes twice during each oscillation cycle of the oscillator. As the displacement of the oscillator 8 is substantially constant, the maximum light output during a cycle is also substantially constant, and therefore an output of the order of 1 volt can be derived from the device 24 even at low flow velocities. Consequently, by utilising only a simple circuit configuration which forms the output into a desired waveform by means of the comparator 27b the vortex frequency can be determined.

The optics consisting of light emitting and receiving devices and the light reflecting surface are housed within the compartment 26 of the oscillation chamber 16 and do not come into direct contact with the fluid from the pipe-line, so avoiding problems of contamination by the fluid.

In the above example of detector devices 23 and 24 are disposed so that the maximum quantity of light is obtained when the plate 9 is in the horizontal position. However, as shown in Fig. 21, they may be pre-biased by the maximum angular amplitude 6m of the oscillatory plate 9. This method utilises the substantially constant-gradient portion of the curve in Fig. 20(b), and therefore even a minute angular displacement can be detected with good sensitivity and low distortion, a substantially sinusoidal output being obtained, whereby signal processing is facilitated, in this case, one pulse output is obtained during each oscillation of the oscillatory plate.

The construction containing the light emitting and receiving devices is not limited to this example, but any suitable means may be arranged in the compartment 26 to detect changes in the quantity of reflected light.

The displacement detectors shown in Figs. 18 to 21 have the advantages that they are not seriously affected by electromagnetic noise and they produce large electric outputs directly, thus simplifying the signal processing circuit. Furthermore, the stability and reliability with respect to temperature changes are improved, and the detectors are economical in

70

75

80

85

90

95

100

105

110

115

120

125

130

8

GB2160318A

8

construction.

Beneraily, an oscillator (such as is shown in Fig. 12, for example) oscillates in response to the frequency of the vortices. It is now pro-5 posed that the fundamental oscillation frequency defined by the moment of inertia of the oscillatory plate (e.g. as indicated by numeral 9 in Fig. 12) and the torsional spring constant of the taut bands (e.g. as indicated 10 by numerals 10a and 10b in Fig. 12) is made substantially equal to the lowest vortex frequency to be measured, by appropriate selection of the length and width of the bands and the dimensions of the oscillatory plate and 15 other characteristics.

When a disturbing force of a given amplitude is applied to such an oscillatory system, the system offers a frequency response such as is shown in Fig. 22 (A) and is well known 20 in the field of oscillation and vibration. It can be said that in Fig. 22(A) the amplitude increases roughly in proportion to the angular frequency w and reaches a peak at resonance angular frequency wn, and then decreases in 25 inverse proportion to the square of the applied angular frequency, though the characteristic is affected by any damping present. In this figure, the ordinate represents amplitude ratio (dynamic amplitude X/static amplitude ast), 30 while the abscissa represents angular frequency ratio co/io„ which is a dimensionless quantity. On the other hand, it has been found experimentally that the magnitude of change in vortex pressure increases substan-35 tially in proportion to the square of the flow velocity, as shown in Fig. 22(B). As the vortex frequency is proportional to the flow velocity, the magnitude of change in vortex pressure is also found to increase in proportion to the 40 square of the vortex frequency.

Since the natural angular frequency wn of the oscillator is made equal to the lowest vortex frequency to be determined, the oscillator will be subject to an oscillatory force which 45 increases in proportion to the square of the frequency. Thus, the amplitude of the oscillatory plate of the oscillator is substantially constant, as shown in Fig. 22(C).

Thus, by lowering the resonance frequency 50 of the oscillator so that it conforms with the characteristic change in vortex pressure, a large displacement can be obtained at low flow velocities (i.e. at low frequencies) and excessively large displacement is prevented at 55 high flow velocities (i.e. at high frequencies). In order to avoid instability at resonance, the oscillatory system is preferably damped to an appropriate extent.

The frequency response of the oscillatory 60 system and change in vortex pressure with frequency have a complementray relationship whereby a substantially constant amplitude of displacement can be obtained over the whole range of vortex frequencies to be measured. 65 This has the practical advantages that the sensitivity at low flow velocities increases and, at the same time, excessively large displacement is prevented at high flow velocities, thus ensuring stable oscillation and avoiding damage to the apparatus.

In the steady state, a signal detected by a light receiving device has a waveform which is symmetrical relative to a constant DC component corresponding to the equilibrium position when the oscillatory plate is at rest, as shown in Fig. 23(A). However, in a transient state where flow increases abruptly, a resonance occurs at a natural frequency dependent on the mass of the oscillatory plate and the torsional spring constant of the taut bands. For example, if the flow changes abruptly as shown in Fig. 23(A), the waveform is disordered, and it becomes impossible to detect vortices. It is to be noted that a resonance point exists in the frequency characteristic of the oscillator as shown in broken line in Fig. 22(A).

Referring to Fig. 24, there are shown an oscillatory plate 90 and taut bands 100a and 100b. Damping material 36, such as a rubbery viscoelastic material, is applied on the taut bands 100a and 100b. Experiment has revealed that the frequency characteristic of torsional oscillation of the oscillator having such damped taut bands has no definite resonance point in the way indicated in solid line in Fig. 22(A). As a result, even when a transient change occurs, the oscillator oscillates in full response to changes in vortex pressure. Further, as shown in Fig. 23(B), disorder of the waveform can be prevented when an abrupt change in flow takes place so avoiding the occurrence of a region where it is impossible to detect vortices. Therefore resonance is prevented in the oscillator 8 at least within the range of vortex frequency to be determined.

Fig. 25 is a schematic diagram of another embodiment. This alternative embodiment to that shown in Fig. 24 differs from the latter in that chambers 1 7a and 17b, which are formed in the vortex detector 4 to accommodate the taut bands 10a and 10b, respectively, are completely filled with damping material 37 rather than arranging for the damping material to be applied on the bands 10a and 10b. This construction suppresses resonance of torsional oscillation of the oscillator and inhibits bending vibrations in vertical and horizontal directions, thus countering the effects of external vibrations. In this embodiment, the damping material 37 may be a silicone rubber, for example.

The damping arrangement need not be limited to either of these forms, but rather a plurality of damping means may be combined to prevent resonance of torsional oscillation of the oscillator.

Thus, by damping the taut bands supporting the oscillator, undue interference between

70

75

80

85

90

95

100

105

110

115

120

125

130

g

GB2 160318A

g vortex frequency and the resonance frequency of the oscillator is avoided even when flow changes abruptly, thus allowing accurate measurement during a transient flow change.

5 Referring to Fig. 27, there is shown a signal processing circuit comprising a light emitting device 61, a light receiving device 62, buffer amplifiers 631 and 633, an integrator circuit 632, a comparator circuit 634, 10 resistors R, to Re, a capacitor C, and a diode Di-

The output signal from the light receiving device 62 is applied directly to the comparator 634. The signal is also applied to the 15 integrator circuit 632 through the buffer amplifier 631, and the signal is applied to the comparator circuit 634 through the buffer amplifier 633 after having been integrated by the integrator 632. The comparator circuit 20 634 compares these signals and when they are identical it produces an output signal.

More specifically, the output from the light receiving device 62 has a sinusoidal waveform which is substantially symmetrical rela-25 tive to DC voltage E0 corresponding to a constant quantity of light obtained in the equilibrium state when the oscillatory plate is at rest as indicated in Fig. 26(A) and by graph a in Fig. 26(B). On the other hand, the signal 30 derived from the integrator circuit 632 is the same as the DC signal E0. as shown in broken line b in Fig. 26(B). Consequently, it is possible to obtain a rectangular waveform output responsive to the vortex frequency by compar-35 ing these signals and appropriately shaping the resultant waveform in the comparator circuit 634.

Since light emitting devices, light receiving devices, optical fibres and so on have gener-40 ally different characteristics, the DC voltage E0 corresponding to the equilibrium state when the oscillatory plate is at rest varies among vortex detectors and the output signal will change relative to the DC voltage E0 as indi-45 cated by a' in Fig. 26(B). However, this output signal a' is compared with DC voltage b', which has been obtained by integrating the alternating voltage a' so that the waveform is then altered and therefore it is pos-50 sible to detect the vortex frequency precisely even if the DC voltage E0 varies.

Similarly, when the DC voltage E0 in the equilibrium state of the oscillatory plate drops due to a decrease in light quantity caused by 55 contamination of the optics including the reflecting surface of the oscillatory plate and the optical fibres, a rectangular output responsive to the vortex frequency can be derived. It is to be noted that the time constant of the 60 integrator circuit 632 is preferably as large as possible. If the time constant is small, ripples are large at low vortex frequencies, which renders the difference between the output signal a from the light receiving device 61 65 and the signal b derived from the integrator circuit 632 small, whereby the apparatus is more affected by noise. Consequently, it is important that the time constant is higher than the lowest vortex frequency. Also, it is desired to provide a hysteresis characteristic, to the comparator 634, by means of resistances R5 and Re as shown in Fig. 27 for example, to counter effects of noise.

Fig. 28 is a circuit diagram of a modified form of the embodiment shown in Fig. 27. Referring to Fig. 28, there is shown an amplifier circuit 635 having a differential amplifier 6352 and an amplifier 6353. A variable resistor element 6 3 54 whose resistance decreases when the applied voltage increase consists of a CdS photocoupler, for example, and acts to control the gain of the amplifier 635. Output A from the buffer amplifier 633 is connected to the resistor element 6354 through a buffer amplifier 635,. This embodiment differs from that shown in Fig. 27 in that the voltage difference between the output signal a from the light receiving device and the DC voltage b, which is derived by passing the signal a through the integrator circuit 632, is amplified and, concurrently, the gain of the amplifier 6353 is automatically controlled by the amplitude of the voltage E0(A) of said DC signal b.

It is now assumed that the light quantity entering the light receiving device 62 has decreased due to contamination of the optics, when the output signal from the device 62 changes from a to a', as shown in Fig. 26(B). Specifically the DC voltage corresponding to the light quantity obtained in the equilibrium state when the oscillatory plate is at rest decreases from E0 to E'0. and the AC voltage corresponding to change in the light quantity caused by oscillation of the oscillatory plate also decreases from a to a'. However, the differential amplifier 6352 can detect the vortex frequency with precision even when DC voltage E0 varies, because it amplifies either the difference between the output signal a from the light receiving device and the DC signal b, which has been derived in the integrator circuit 632, or the difference between a' and b'.

Also, if the DC voltage drops, the resistance of the variable resistor element 6354 increases, resulting in increase in the gain of the amplifier 6353. Hence, decrease in the AC voltage responsive to the vortex frequency can be compensated. It is possible to control the gain of the amplifier 6353 by appropriately selecting the values of the resistors R,, and R12 and the resistor element 6354. Thus, even if the light quantity decreases, it is possible to detect the vortex frequency in a stable manner.

It will be understood that the output signal from the light receiving device 62 is applied to the integrator circuit 632 always via the buffer amplifier 631 in this embodiment.

70

75

80

85

90

gs

100

105

110

115

120

125

130

10

GB2 160318A

10

However, the amplifier 631 can be omitted if the impedance defined by the resistance R4 and the capacitor C, is made sufficiently greater than the load resistance R2 of the light 5 receiving device.

While in these embodiments displacement of the oscillator is detected by means of a change in the quantity of reflected light, changes in transmitted light may be used 10 instead.

Thus with these embodiments, even when the DC (integration) signal component varies due to variations in the state of the optics including the light emitting and receiving de-15 vices, e.g. by contamination of the optics, a rectangular output signal responsive to the vortex frequency can be derived. This gives the practical advantages that individual detector sensors can be adjusted freely and that the 20 flowmeter can operate relatively independently of contamination. Further, when the waveform of the output signal from the displacement detector is altered, the difference between this output signal and the integrated DC signal is 25 amplified and the amplification gain is automatically controlled according to the magnitude of the DC signal, whereby decrease in light quantity attributed to contamination of the optics can be compensated by the circuit, 30 leading to enhanced reliability.

Embodiments of Karman vortex flowmeter adapted for measuring the quantity of air inhaled by the internal combustion engine of an automobile or the like will now be de-35 scribed.

A first such embodiment of Karman vortex flowmeter is described with reference to Figs. 29 and 30.

Referring to Fig. 29 an engine 101 and an 40 inlet passage 102 are shown, the passage comprising an air cleaner 103, filtering element 104, an inlet tube 105 and a throttle valve 106. A pipe-line 108 forms a part of the inlet tube 105, and the flow within the 45 pipe-line 108 is stabilised by a flow-rectifying device 109. A Karman vortex flowmeter 107 consists of a vortex generator 2, inserted in the pipe-line 108, for generating vortices and a vortex detector 4 which converts pressure 50 changes due to generated vortices into light signals for producing a light pulse output signal responsive to the vortex frequency.

An electronic circuit 6 for converting the light signal from the vortex detector 4 into an 55 electrical signal is connected to the vortex detector 4 through optical fibres 5a and 5b. As shown in Fig. 30, the vortex detector 4 has apertures 20a and 20b, chamber 16 communicating with the apertures and oscilla-60 tory plate 9 which oscillates torsionally within the chamber. The apertures are in communication with the respective openings 4a and 4b, which are formed in opposite sides of the generator 2 parallel to the flow, to convey 65 vortex pressure changes.

Respective ends of the light fibres 5a and 5b are so disposed that their optical axes cross the axis of rotation of the oscillator at a predetermined angle, the other ends being provided with light emitting device 6a and light receiving device 6b for detecting the light reflected from the oscillator. These components are accommodated in a housing 120. A purging device 122 comprises a restrictor 124 and a filter 123, and the chamber 16 is in communication with the atmosphere via the restrictor and the filter through an opening 121, which is formed in the housing 120 in the vicinity of the optics comprising the oscillator and the optical fibres.

In the operation of this embodiment when air flows through the inlet passage 102 shown in Fig. 29 and a vortex is generated on the same side of the vortex generator 2 as the opening 4a shown in Fig. 30, the pressure on the side of the opening 4a becomes lower than the pressure on the side of the opening 4b, resulting in a pressure difference between the openings. When two such vortices are generated in succession on opposite sides of the generator, the pressure differences cause the oscillator 9 to effect an oscillation cycle as in other embodiments described above!

Change in the quantity of reflected light resulting from the oscillation is detected by the light receiving device 6b through the optical fibres 5b and used to determine the vortex frequency and the quantity of air inhaled can then be derived from the vortex frequency.

The pressure in the chamber 26 is lower than atmospheric pressure, because at the openings 4a and 4b the pressure is lower than atmospheric pressure, the vortex generator 2 narrowing the inlet flow passages to the openings.

Accordingly, whenever the automobile is driven and air is inhaled, the prewsure in the chamber 26 is negative. In other words, unless the engine backfires air is always drawn in by it, and therefore the inside of the pipeline 108 is in a state of negative pressure.

This gives rise to a supply of clean air via the purging device 122 to clean the optics so as to reduce contamination of the optics. As the purging air flows into the inlet passage 102 through the apertures 20a, 20b and the openings 4a, 4b, this quantity of clean air will not be measured. However, the restrictor 124 limits the quantity to less than 0.1 % of the quantity of inhaled air. Hence, neither the measuring accuracy nor vortex generation is significantly affected thereby.

In these embodiments, the openings for introducing changes in vortex pressure are formed on opposite sides of the vortex generator 10. However, such openings may be formed at other locations in the generator such as on the downstream side of it, as long as the openings can detect vortices. As the pressure'within the inlet passage 102 is

70

75

80

85

90

95

100

105

110

115

120

125

130

GB2 160318A

always lower than atmospheric pressure, air outside the pipe-line can be used for purging. In modified embodiments transmitted light may be detected, and a light emitting device 5 and a light receiving device may be directly installed without using optical fibres. The oscillation detector may comprise other forms of optical detecting means.

A further embodiment of Karman vortex 10 flowmeter will now be described with reference to Fig. 31, which illustrates the general arrangement in which a flowmeter is mounted in an automobile. Referring to Fig. 31, there are shown engine 101 and inlet passage 102, 15 which includes air cleaner 103, filtering element 104, inlet tube 105 and throttle valve 106. Vortex flowmeter 107 mounted in the inlet passage 102 comprises pipe-line 108, which forms a part of the inlet tube 105, 20 flow-rectifying device 109 for stabilising the flow within the pipe-line 108 and vortex detector 4. This detector 4 is mounted in the pipe-line and acts to convert vortex pressure changes into light signals for producing light 25 pulses in dependence on vortex frequency. These components are all installed in the engine compartment. Electrical circuit 6 for converting the light signals from the vortex detector 4 into electrical signals is installed in 30 the passenger compartment which provides a better environment for electrical circuitry, that is, it is less affected by temperature, electrical noise etc. than the engine compartment, and the circuit 6 is connected to the vortex detec-35 tor 4 of the flowmeter 107 mounted within the engine compartment through optical fibres 5.

In this arrangement, detected vortex signals are transmitted in the form of light signals 40 outside the engine compartment through the optical fibres, so that the apparatus is little affected by the temperature and electromagnetic noise conditions in the engine compartment.

45 Although the light emitting device is installed in the engine compartment, devices of this kind are not easily affected by electromagnetic noise. Further, they are energised by direct current and consequently it is sufficient 50 to stabilise the power supply. Thus, such devices can be relatively freely used in the engine compartment.

Referring to Fig. 32, there are shown an engine 101 and inlet passage 102, compris-55 ing an air cleaner 103, filtering element 104, inlet tube 105 and throttle valve 106. Vortex flowmeter 107 installed within the inlet passage 102 consists principally of pipe-line 108 forming a part of the inlet tube 105, flow-60 rectifying device 109 for stabilising the flow within the pipe-line 108, vortex generator 2 mounted within the pipe-line for generating vortices and vortex detector 4. This mechanism converts vortex pressure changes into 65 light signals and produces a light signal output dependent on the vortex frequency. Also shown are pipe 130 (described below) and restrictor 131.

In order to introduce air into a circuit box 134 (Fig. 33) for cooling the components of a signal processing circuit 25 mounted on a printed circuit board, this flowmeter is so designed that the pipe 130 for conveying inlet negative pressure in the engine is connected to the circuit box 134 via the restrictor 131, as shown in Figs. 32 and 33. An inlet pipe 132 for introducing air into the circuit box opens to the atmosphere via air filter 133. Thus, a flow of air through air filter 1 33, pipe 1 32, box 134, pipe 1 30 and inlet passage 102 is induced due to negative pressure within the engine. Consequently, when temperature within the engine compartment increases, the temperature rise of the circuit elements of the signal processing circuit 25 is inhibited, because cooling air is always flowing through the circuit box 134.

When the engine is being idled after a high speed drive, the temperature within the engine compartment tends to be highest. Also in this case, due to large inlet negative pressure within the engine, air flows through the circuit box 1 34 to cool it sufficiently.

When the throttle valve of the engine is fully open the inlet negative pressure is small and flow of air into the circuit box 1 34 decreases, but in this case a large quantity of air is flowing through the pipe-line, so that the temperature within the circuit box 1 34 is kept within a reasonable range. The flow of cooling air is arranged to be less than 0.1 % of the idling inlet flow, and therefore the accuracy of the flowmeter is not significantly affected by such cooling air flow.

The passage for introducing the cooling air may have an inlet opening as indicated in broken line in Fig. 33, into the inlet pipe-line such that the air filter 1 33 is not needed. Particularly when such opening is formed on the downstream side of the vortex generator 2 and on the upstream side of the throttle valve 106, the quantity of air inhaled is measured before it reaches the opening, so that the measuring accuracy is not affected by such cooling air flow.

Thus the embodiments described above can provide a Karman vortex flowmeter equipped with a vortex generator which can generate relatively strong and regular vortices even when flow velocity is low, produces a relatively small pressure loss and is less affected by disturbances in the fluid than prior art flowmeters. .

The embodiments described above can also provide a Karman vortex flowmeter equipped with a detector which is relatively unaffected by external vibrations even at low flow velocities and is able precisely to detect only oscillations resulting from vortices.

The embodiments also provide a Karman

70

75

80

85

90

95

100

105

110

115

120

125

130

12

GB2 160318A

12

vortex flowmeter capable of detecting vortices stably over a wide range of flow velocity and a Karman vortex flowmeter which can detect vortices with precision even during a transient 5 state. The embodiments also provide a signal processing circuit whose accuracy is relatively unaffected by changes in light quantity resulting from variations in state of the light receiving devices.

10 Finally the embodiments provide a Karman vortex flowmeter which is used for measurement of the quantity of air inhaled by an internal combustion engine of an automobile or the like, is capable of avoiding or prevent-1 5 ing contamination of the optics in a simple manner and is safeguarded against temperature conditions and electrical noise.

Claims

20 1. A Karman vortex flowmeter for measuring the quantity of air inhaled by an internal combustion engine of an automobile or the like, said flowmeter comprising a signal processing circuit for processing a detected vor-25 tex detection signal, and a container housing the signal processing circuit, one side of the container being in communication with an air inlet passage on the downstream side of a throttle valve in the air inlet passage, the 30 other side of the container being in communication with the atmosphere so that an air flow will be induced through the container due to negative pressure on the downstream side of the throttle valve so that the components of 35 said signal processing circuit will be cooled.
2. A flowmeter according to claim 1, wherein air for cooling the components of the signal processing circuit is introduced into the container from an inlet pipe-line which is

40 connected to said air inlet passage on the downstream side of a vortex generator and on the upstream side of the throttle valve.
3. A flowmeter according to claim 1 or claim 2, wherein the components of the signal

45 processing circuit are mounted on a printed circuit board.

Printed in the United Kingdom for

Her Majesty's Stationery Office. Dd 8818935, 1985, 4235. Published at The Patent Office. 25 Southampton Buildings, London. WC2A 1AY. from which copies may he obtained.