JPH0511768B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a mass flow meter that measures the mass flow rate of a measurement fluid using Karman vortices.

(Prior Art) It has been known for a long time that when an object is placed in a fluid, vortices are generated alternately and regularly from both rear sides of the object and flow downstream as a vortex train. This vortex street is called a Karman vortex street, and the number of vortices generated per unit time (generation frequency) is proportional to the flow rate of the fluid. Therefore, a vortex generator is placed in the pipe that guides the fluid to be measured, and the vortex generator generates a vortex proportional to the flow velocity. Vortex flowmeters have been put into practical use that measure the flow velocity and flow rate of fluid by detecting it with a sensor and extracting only the frequency of the detection signal. Incidentally, the flow rate that one generally wants to know is often the mass flow rate, not only in processes that involve chemical changes, but also in transactions. Furthermore, when the object to be measured is gas or steam, its density changes greatly depending on temperature and pressure, and even when it is a liquid, its density changes considerably depending on temperature. For this reason, the mass flow rate is measured either by installing it in parallel with a vortex flowmeter to measure temperature and pressure, or by measuring the density with a density meter.
However, using a density meter and a vortex flowmeter is complicated and expensive, and a combination of a temperature or pressure gauge and a vortex flowmeter is difficult.
Not only is it complicated and expensive, but it also has poor accuracy and responsiveness because it is difficult to measure the temperature of the fluid.

By the way, as shown in FIG. 13, when the vortex generator 2 is arranged in the pipe line 1 and the measurement fluid is flowed into the pipe line 1, the average force F _D acting on the vortex generator 2 and the fluctuating lift force F _L and pressure loss ΔP generally have the relationship shown by the following equation.

F _D ∝C _D 1/2ρV ² (1) F _L ∝±C _L 1/2ρV ² (2) ΔP∝C _P 1/2ρV ² (3) where C _D ; drag coefficient C _L ; variable lift coefficient C _P ; Pressure loss coefficient ρ; Density V; Flow velocity (1), (2),
P.79, P.87 21 Kutsajyukovsky's Theorem Written by Ichiro Tani Iwanami Shoten" Also, (3) is the separation of the boundary layer from the vortex generator (generation of Karman vortex)
This is derived from the loss due to

Now, average drag force F _D , fluctuating lift force F _L , pressure loss ΔP
is proportional to ρV ² if drag coefficient C _D , variable lift coefficient C _L , and pressure loss coefficient C _P are constants, so vortex frequency = StV/d = KV (where, St: Strouhal number, d: vortex By dividing equations (1), (2), and (3) by the diameter of the generating body 2 (K=St/d, a constant), ρV can be obtained. The values of the drag coefficient C _D , the variable lift coefficient C _L , and the pressure loss coefficient C _P are shown in Figure 14 (Aerospace Engineering Handbook P205, compiled by the Japan Society of Aeronautics and Astronautics, published by Maruzen Co., Ltd.) for the drag coefficient C _D , for example. Various shapes of vortex generators are required to achieve this.

Furthermore, finding the mass flow rate of ρ・V ² /V=ρV by dividing the ρV ² detector and the V detector is also described in Flow Measurement Handbook, page 344, edited by Hiroo Kawada and published by Nikkan Kogyo Shimbun. This is a technique that has been commonly used in the past.

Known examples of this type of vortex flowmeter include:
There is a West German patent DE3032578C2 entitled "Method and device for dynamically determining the density of a fluid to be measured in a density-independent manner (measuring mass flow rate using a vortex flowmeter)", which uses fluctuating lift as strain in a strain gauge attached to a vortex generator. Some methods detect the torque generated in the vortex generator by a spring, or detect the drag force by installing a pitot tube in front of the vortex generator, and calculate the mass flow rate by dividing it by the vortex frequency.

In addition, in Japan, in Utility Model Application Publication No. 174359/1983, ``Measuring device using Karman vortices'',
A capacitance detection section using a diaphragm is installed upstream and downstream of the vortex generator, and the drag force F _D is calculated from the DC component of the capacitance change.
Some methods calculate the mass flow rate by detecting and dividing the vortex frequency from the AC component.

Furthermore, Japanese Patent Application Laid-Open No. 57-61916 "Measuring Device Using Karman Vortices" shows an example of detecting the fluctuating lift force and dividing it by the vortex frequency.

JP-A No. 57-61916 will be explained below.

FIG. 15 is an explanatory diagram of the structure of this Japanese Patent Application Laid-Open No. 57-61916. In the figure, 1a is a pipe through which the measurement fluid flows, and 2a is a columnar vortex generator inserted perpendicularly into the pipe 1a, both ends of which are fixed to the pipe 1a. The main body 21a of the vortex generator 2a is made of stainless steel or the like, and has a cross-sectional shape, such as a trapezoid, which produces a Karman vortex street in the measured fluid and stably enhances lift changes. Top 22a of vortex generator 2a
is made of stainless steel or the like, has a recess 23a, and is integrally formed with the main body 21a by welding or the like.
Reference numeral 41a denotes an element body consisting of a pressure element, which is sealed to the recess 23a of the vortex generator 2a with an insulating material 3a such as glass, and is formed integrally with the vortex generator.
The element main body 41a has a disk shape and is arranged so that its center coincides with the neutral axis of the vortex generator 2a. Furthermore, as shown in FIG. 16, the element main body 41a is provided with electrodes 42a, 43a, 44a, and 45a symmetrically divided into right and left sides on the front and back sides, respectively, and the portion sandwiched between the electrodes 42a and 43a. A first piezoelectric sensor 46a is formed, and a second piezoelectric sensor 47a is formed at a portion sandwiched between electrodes 44a and 45a.
form. And the first and second piezoelectric sensors 46
So that the charges generated on a and 47a are differential,
Electrodes 42a and 45a and electrodes 43a and 44a are connected, and lead wires 48a and 49a are taken out from electrodes 42a and 44a through insulating material 3a, respectively. 8a is a detection signal processing circuit that converts AC charge q detected by piezoelectric sensors 46a and 47a into AC voltage e. 9a is a comparator for converting the AC voltage e into a pulse signal P of a constant level. 10a is an F/V converter which converts the pulse signal P output from the comparator into a DC voltage _E1 proportional to its frequency. 11a is a rectifying and smoothing circuit that rectifies and smoothes the AC voltage e and converts it into a DC voltage _E2 proportional to its amplitude. 12a
is an arithmetic circuit that performs desired arithmetic operations on the outputs E ₁ and E ₂ of the F/V converter 10a and the rectifying and smoothing circuit 11a,
It is for extracting a signal related to the density or mass flow rate of the fluid at its output.

In the conventional example shown in FIG. 15 configured in this way,
When the measurement fluid flows into the pipe line 1a, the vortex generator 2a
is generated as a Karman vortex, and is also subjected to lift changes based on the generation of the vortex. When the vortex generating body 2a receives a lift change, a stress change occurs in the opposite direction across the neutral axis as shown in the figure. This vortex generator 2
The stress change that occurs in a is caused by the insulating material 3a and the vortex generator 2a.
The signal is transmitted to the element body 41a, which is integrally attached to the element body 41a. Therefore, the first and second piezoelectric sensors 46a,
47a, a change in the amount of charge occurs in opposite phase to each other in response to a change in lift force. The amount of electric charge generated in the piezoelectric sensors 46a and 47a is extracted differentially, and an alternating electric charge q is generated between the lead wires 48a and 49a.
occurs. The alternating charge q is converted into an alternating voltage e by the detection signal processing circuit 8a. If the frequency of the AC voltage e is taken out via the comparator 9a and the F/V converter 10a, a voltage _E1 proportional to the vortex frequency, that is, the flow velocity ν, can be obtained as in equation (4), similar to a general vortex flowmeter.

E ₁ = K ₁ V (4) However, K ₁ is a proportional constant On the other hand, if the amplitude of AC voltage e is taken out via the rectifying and smoothing circuit 11a, the output of the rectifying and smoothing circuit 11a
E ₂ is given by the following equation, where ρ is the density of the fluid.

E ₂ = K ₂ 〓V ² (5) However, since K ₂ is a proportionality constant, if the arithmetic circuit 12a performs the calculation E ₂ /E ₁ , the output Eo will be Eo = E ₂ /E ₁ = K ₂ /K ₁ ρV=K ₃ ρV (6). If the cross-sectional area of the pipe 1a is S, then the mass flow rate Qm is given by Qm=ρVs (7), so Eo is Eo=K ₃ /SQm (8), which is a signal proportional to the mass flow rate. .

Furthermore, if the arithmetic circuit 12a squares E ₁ and then divides E ₂ , the output Eo will be Eo=E ₂ /E ₁ ² =K ₂ /K ₁ ² ρ=K ₃ /K ₁ ρ ( 9), and a signal proportional to the density of the fluid can be obtained.

The above idea is based on the assumption that the variable lift coefficient C _L is constant. Similarly, in the West German patent DE3032578 C2, the variable lift coefficient
It is necessary that C _L be constant.

(Problem to be solved by the invention) However, now the fluctuating lift due to the Karman vortex
F _L , drag force F _D and vortex frequency output (volume flow rate) F _V are F _L = ±1/2C _L ρV ² dD (10) F _D = ±1/2C _D ρV ² dD (11) F _V = K ₁ V (12) However, F _L ; Variable lift force F _D ; Drag force F _V ; Vortex frequency output (volume flow rate) C _L ; Variable lift coefficient C _D ; Drag coefficient ρ; Density V; Flow velocity d; Opposed to the flow of the vortex generator diameter D; pipe inner diameter K ₁ ; constant Therefore, F _L /F _V = (1/2C _L d・D/K ₁ )ρV K ₃ in equation (6) is K ₃ = 1/2C _L d・D/K becomes ₁ .

Here, the variable lift coefficient C _L is, for example,
Even when the drag coefficient C _D as shown in FIG. 17 and the Strouhal number St as shown in FIG. 18 are constant, they are not constant as shown in FIG. 19. (Figures 17 to 18 are from Toshio Kobayashi's "Fluid Force Acting on Bluff Body" magazine "Production Research" Vol. 24 No. 1, published by Institute of Industrial Science, University of Tokyo, Figures 1, 7, and 8 is cited.) In the conventional example shown in FIG. 15, the detection signal processing circuit 8
In a, a filter circuit is included to reliably detect the vortex frequency, as used in vortex flowmeters. The raw signal from the sensor is
Generally, high-frequency and low-frequency noises are superimposed due to turbulence or the like existing in the fluid. In particular, high-frequency noise becomes significant at low flow velocities, and low-frequency noise becomes significant at high flow velocities, and a filter circuit is required to remove this noise. If the filter effect of the filter circuit is weakened, it becomes difficult to detect the vortex frequency due to the noise contained in the vortex signal.

On the other hand, the presence of this filter circuit makes the variable lift signal inaccurate.

The present invention solves this problem.

SUMMARY OF THE INVENTION An object of the present invention is to provide an apparatus that can measure density or mass flow rate with high accuracy with a simple configuration and a circuit configuration so that the variable lift coefficient does not fluctuate.

(Means for solving the problem) In order to achieve this object, the present application provides a signal conversion circuit that converts a vortex signal from a vortex detection sensor into a signal, and a signal conversion circuit that removes high-frequency and low-frequency noise from the signal from the signal conversion circuit. a first amplifier circuit that amplifies the signal from the filter circuit; a comparator that converts the signal from the first amplifier circuit into a pulse signal; an F/V converter that converts to a DC voltage, a second amplifier circuit that amplifies the signal from the signal conversion circuit, a detection circuit that detects the signal from the second amplifier circuit, and a ripple signal for the signal from the detection circuit. a rectifier circuit that removes the signal from the first amplifier; a divider circuit that divides the signal from the rectifier circuit by the signal from the F/V converter circuit; a gate circuit that sets the output of the divider circuit to 0 when the output voltage becomes 0, the time constant of the rectifier circuit and the time constant of the F/V converter are substantially the same, and the beat frequency of the vortex signal is This is a mass flow meter configured to have a larger size.

(Function) In the above configuration, the vortex signal from the vortex detection sensor is converted into a signal by the signal conversion circuit, the filter circuit reduces high frequency and low frequency noise, the first amplifier circuit amplifies the signal, and the comparator pulses the vortex signal. Convert to signal. Next, an F/V converter converts the pulse signal into a DC voltage proportional to its frequency.

On the other hand, the signal from the signal conversion circuit is amplified by a second amplifier and detected by a detection circuit. Next, a rectifier circuit removes the ripple.

The signal from this rectifier circuit is passed through the divider circuit,
Dividing by the signal from the F/V converter yields an electrical signal corresponding to the mass flow rate.

On the other hand, if the signal from the first amplifier does not reach the set trigger level of the comparator, F/V
The gate circuit sets the output from the divider circuit to 0 based on the signal from the converter.

Examples will be described below.

(Example) FIGS. 1A and B are configuration explanatory diagrams of one embodiment of the present invention, A is a front view, B is a side view, and FIG. 2 is a configuration explanation showing the detector section of FIG. 1 in cross section. It is a diagram.

In the figure, 10 is a vortex flowmeter detector, and 20 is a vortex flowmeter converter.

In the vortex flow meter detector 10, 11 is a pipe through which the fluid to be measured flows, 12 is a cylindrical nozzle provided perpendicularly to the pipe 11, and 13 is a columnar vortex inserted perpendicularly into the pipe 11 through the nozzle 12. The generator is made of stainless steel, etc., and its upper end 131 is a nozzle 12.
The lower end 132 is supported by a plug in the conduit 11 by screws or welding. The fluid-contacting portion 133 of the vortex generator 13 that comes into contact with the fluid to be measured has a cross-sectional shape such as a trapezoid, for example, so as to generate a Karman vortex street in the fluid to be measured and to stably enhance changes in lift, and has a recessed portion on the upper end 131 side. It has 134. 135 is a vortex generator 13 formed by a recess 134.
This is an outer cylindrical part formed in. 14 is a sensor part,
A first stress detection sensor 142 and a second stress detection sensor 141 are press-fixed in the recess 134 of the vortex generator 13 at a constant interval. As shown in FIG.
They are placed on the same side or on both sides of position A where the stress generated at is zero. In the following, a case where they are arranged on the same side will be explained. In the sensor section 14, the underlay 143 made of stainless steel or the like acts as a buffer between the first stress detection sensor 142 and the bottom surface of the recess 134, and compensates for the difficulty in controlling the roughness during processing of the bottom surface of the recess 134. A first spacer 144 made of stainless steel or the like and an insulating plate 14 made of ceramic or the like.
5 and a second spacer 146 made of stainless steel or the like are used to determine the distance between the first stress detection sensor 142 and the second stress sensor 141 and to insulate them. Push rod 1 made of stainless steel, etc.
47 is welded to the upper end 131 of the vortex generator 13 in a state where the sensors 141 and 142 are pressed, and the sensor 14
1,142 is pressed and fixed. Note that the sensor section 14 includes a vortex generator 13, an underlay 143 and a push rod 1.
47 so that they come into contact only at the top. The stress detection sensors 141 and 142 are composed of disk-shaped piezoelectric elements 140, and are arranged so that their centers coincide with the neutral axis of the vortex generator 13. Furthermore, as shown in the perspective view of FIG. 4(A), the piezoelectric element 140 has electrodes 14 symmetrically divided into left and right sides with respect to the flow direction of the measurement fluid (direction of the arrow in the figure) on the front and back sides of the piezoelectric element 140, respectively.
01, 1402, 1403, and 1404 are provided, and as shown in FIG. 4(B), stress ( electrodes 1401, 1402 corresponding to compressive stress and tensile stress)
The polarization is reversed so that the charge generated between the electrodes 1403 and 1404 has the same polarity as the charge generated between the electrodes 1403 and 1404. Therefore, as shown in FIG. 4C, charges of opposite polarity are generated between the two electrodes in response to stresses occurring in the same direction. In addition, the amount of charge generated by stress in the flow direction of the measured fluid is canceled between the electrodes and does not come out, and the amount of charge generated by pipe vibration in the flow direction is also canceled between the electrodes and does not come out. . The second stress detection sensor 141 is located between the electrodes 1401 and 1402 and between the electrodes 1403 and 1403.
Charges 1401 and 1403 are connected to push rod 147 in order to make the sum of charges of the same polarity generated between 1404 and 1404 output charge q ₁ and to cancel charges of opposite polarity.
The electrodes 1402 and 1404 are commonly connected to the vortex generator 13, that is, the reference point via the spacer 1.
46 and are commonly connected to a lead wire 1481. The first stress detection sensor 142 is connected to the electrode 14
Between 01 and 1402 and electrodes 1403 and 1404
Output charge is the sum of charges of the same polarity that occur between
Electrodes 1401 and 14 are used to cancel the charge of opposite polarity to q ₂ and to reverse the polarity from q ₁ .
03 is commonly connected to the lead wire 1 through the spacer 144.
482, and the electrodes 1402 and 1404 are commonly connected to the vortex generator 13, ie, the reference point, via the underlay 143. Lead wire 1481,1
482 is taken out to the outside through the through holes provided in each part of the sensor section 14 and the hermetic seal 148, and connected to the vortex flowmeter converter 20. Note that the recess 134 of the vortex generator 13 and the sensor section 1
A gas with a low dew point is sealed in the area surrounded by 4 to prevent condensation, and the push rod 147 is provided with a communication hole 149 for the sealed gas. 15
is a labyrinth flow path provided at the intersection of the vortex generator 13 and the pipe line 11. Labyrinth channel 15
As shown in FIG.
51 and the enlarging chamber 152 following it form a pair and are constituted by at least one or more pairs.

Therefore, the depth L of the recess 134 is selected so as to satisfy the following equation.

Lα _P =t _K α _K +t _S α _S +t _L α _L +t _q α _q α _P ; Thermal expansion coefficient of the vortex generator 13, α _K : Thermal expansion coefficient of the push rod 147, α _S : Thermal expansion coefficient of the insulating plate 145, α _L ; Thermal expansion coefficient of the stress detection sensors 141, 142; α _q ; Thermal expansion coefficient of the spacers 144, 146 t _K ; Length of the push rod 147; t _S ; Thickness of the insulating plate 145; t _L ; Stress detection sensor 141 , 142, _tq ; thickness of the spacers 144, 146, that is, the vortex generator 13 and the stress detection sensor 141,
142, insulating plate 145, and spacers 144, 146
The depth L of the recess 134 is set so that the difference in thermal expansion between the vortex generator 13 and the push rod 147 can be exactly canceled out.

FIG. 6 is a block diagram of the electrical circuit 30 of FIG. 2 (not shown in FIG. 2).

31 and 32 are charge converters, and AC charges q ₁ and q ₂ detected by stress detection sensors 141 and 142
Convert to AC voltages E ₃₁ and E ₃₂ . 33 is an addition/subtraction circuit which adds the outputs from the charge converters 31 and 32 to obtain an AC voltage E ₃₃ . 34 is a second amplification circuit that amplifies the AC voltage E ₃₃ to generate an AC voltage E ₃₄ .
35 is a detection circuit that detects the AC voltage E ₃₄ and converts it into E ₃₅ . 36 is a rectifier circuit that removes ripples from the AC voltage _E35 . 37 is a filter circuit, AC voltage
Low-frequency or high-frequency noise contained in E ₃₃ is removed to obtain AC voltage E ₃₇ . 38 is a first amplification circuit that amplifies the AC voltage E ₃₇ to obtain an AC voltage E ₃₈ . 39 is the Schmitt trigger circuit, which is an AC voltage
Converts E ₃₈ into a constant level pulse signal P ₃₉ . 4
1 is an F/V converter which converts the pulse signal P ₃₉ into a DC voltage E ₄₁ proportional to its frequency. 42
is a division circuit that performs desired calculations on the outputs E ₄₁ and E 36 of the F/V converter 41 and the rectifier circuit ₃₆ , and outputs a signal related to the density or mass flow rate of the fluid to the output.
Take out E ₄₂ . 43 is a gate circuit, and the AC voltage E ₃₈ from the first amplifier circuit 38 is applied to the Schmitt circuit 3.
When the trigger level no longer reaches the set trigger level of 9, the output of the divider circuit 42 is set to 0. Thus, the time constant of the rectifier circuit 36 and the time constant of the F/V converter 41 are made substantially the same and are configured to be greater than the beat frequency of the vortex signal.

In the above configuration, in order to accurately detect the absolute value of the fluctuating lift signal, a signal conversion circuit (charge converter 3) with flat characteristics within the vortex frequency band is used.
1, 32 and an adder/subtractor circuit 33), the physical quantity is converted into an electrical quantity, and then similarly amplified by a second amplifier circuit 34, which has flat characteristics within the vortex frequency band, and used as an input to a detection circuit 35.

On the other hand, for vortex frequency detection, after passing through a signal conversion circuit (charge converters 31 and 32 and an addition/subtraction circuit 33), a filter circuit 37 and a first amplifier circuit are used to reduce high frequency and low frequency noise contained in the vortex signal. After passing through 38, it is input to a Schmitt circuit (comparator) 39.

As a result of the above, by separating the fluctuating lift signal detection circuit and the vortex frequency detection circuit immediately after the signal conversion circuit, it is possible to accurately measure the absolute value of the fluctuating lift signal and to reliably detect the vortex frequency. I was able to achieve both.

Therefore, it has become possible to accurately detect fluctuating lift in a three-dimensional flow inside a circular tube. From the experimental results, the variable lift force F _L is exactly proportional to the square of the flow velocity V ² , that is, the variable lift coefficient C _L =const was obtained when F _L =±C _L ·1/2ρV ² dD.

Conventionally, the fluctuating lift coefficient C _L of a flat body (vortex generating body) was generally extremely unstable even in two-dimensional flow, and was considered to be a function of the Reynolds number Re, etc. According to the literature mentioned above, the value of the variable lift coefficient C _L varies greatly depending on the experimenter, and even if the Strouhal number St is constant (vortex frequency is proportional to the flow velocity), the value of the variable lift coefficient C _L (vortex strength)
was not considered to be exactly proportional to the square of the flow velocity, V ² . Since the detection sensor that measures the fluctuating lift force uses strain, displacement, and force sensors, AC noise of the fluid, pipe vibration, noise of the measurement system, etc. are superimposed, making separate detection difficult.

However, West German patent DE3032578 C2, U.S. Pat.
No. 61916 does not explicitly state anything about C _L detection, but assumes that C _L is constant and describes a known technique.

On the other hand, as for the average drag force, as shown in literature values, when the Reynolds number Re>2000 or more,
It has been shown that it remains almost constant.
This is because the average drag can be detected as a DC component, so the measurement circuit can completely remove AC noise.
The reason is that accurate measurements can be made without special efforts.

Next, a gate circuit 43 is provided which sets the output of the divider circuit 42 to 0 when the signal from the first amplifier circuit 38 no longer reaches the set trigger level of the Schmitt circuit 39. When the vortex signal becomes small and does not reach the set trigger level of the Schmitts circuit 39, the Schmitts circuit 39 outputs a pulse.
_P39 is no longer output, and as a result, the output of the F/V converter 41 becomes 0. On the other hand, the output voltage E ₃₆ of the rectifier circuit 36 of the variable lift detection circuit does not become zero due to noise in the flow of the measurement fluid, circuit noise, and the like. Therefore, the denominator input of the division circuit 42 is 0 and the numerator input is not 0, resulting in a large error. The gate circuit 43 uses the output voltage of the F/V converter as the gate signal voltage, and performs zero cut to eliminate this error. Output in Figure 7
The relationship between Eo and ρV is shown.

Note that the set trigger level of the Schmitt circuit 39 is set to S/N=1 or a little higher than that. When S/N<1, even though it is possible to detect the vortex frequency by the filter circuit 37,
This is because a large error will occur in the measurement of the fluctuating lift force.

Furthermore, since the time constants of the rectifier circuit 36 and the F/V converter 41 are almost the same and are configured to be larger than the beat frequency of the vortex signal,
It is possible to obtain a device that can accurately follow fluctuations in the measured flow rate of the measured fluid.

Note that the detection circuit 35 may be a half-wave rectifier or a double-wave rectifier. If the noise superimposed on the input signal E ₃₄ of the detection circuit 35 has an S/N>1, the influence of the noise is reduced. For example, as shown in FIG. 8A, the detected waveform E ₃₅ of an input signal E ₃₄ in which high-frequency noise is superimposed on a low-frequency signal is a waveform in which noise is superimposed, as shown in FIG. 8(B). However, for example, with respect to illustrations A and B in FIG. 8B, the noise is averaged and the noise is reduced overall.
Further, if synchronous rectification is performed at the vortex frequency, noise can be almost completely removed.

Now, the stability of vortex shedding (stability of lift force) is poor even when the straight pipe length is sufficient, and the fluctuation is several tens of percent.
There are more than that. Therefore, the amplitude of the output of the detection circuit 35 fluctuates considerably, and even if the time constant of the rectifier circuit 36 is sufficient for the frequency of the vortex, the ripple is large and impractical. The time constant must be considered with respect to the fluctuation frequency of the amplitude of the output of the detection circuit 35.
Judging from the water test results, this value is
5 seconds or more is required.

On the other hand, in order to improve the stability of the lift force, it is necessary to stabilize the flow, and it is effective to install a flow straightening device such as a flow straightening pipe or a perforated plate upstream of the vortex generator 13.

Furthermore, the vortex generator 13 and the stress detection sensor 141,
142, insulating plate 145, and spacers 144, 146
The depth L of the recess 134 is set so that the difference in thermal expansion between the vortex generator 13 and the push rod 147 can be exactly canceled out. The compressive force applied to the stress detection sensors 141 and 142 does not change due to a change in temperature, and is maintained in the initial compressed state initially applied by the push rod 147. Therefore, it is possible to measure the fluid to be measured up to a high temperature range. It is particularly effective for measuring variable lift that is affected in absolute value.

In addition, the vortex generator 13 and the stress detection sensor 14
1,142, insulating plate 145 and spacer 144,1
By making the coefficient of thermal expansion substantially equal to that of 46, it is possible to obtain a product in which the initial pressing force does not change even if the temperature changes.

In addition, the initial pressing force is determined by the stress detection sensor 14.
If the sensitivity to the alternating stress σ _S of 1,142 is sufficiently saturated, it is possible to obtain a device in which a slight change in the pressing force does not cause a large output error. As a specific example, for example, 3Kg/
Apply an initial pressing force of cm ² or more.

Further, in the embodiment shown in FIG. 1, as shown in FIG. 9, the variable lift force F _L is expressed by the following formula.

F _L = ±C _L 1/2ρV ²・d・D (14) However, C _L : Fluctuation lift coefficient ρ : Density V : Flow velocity d : Diameter facing the flow of the vortex generator D : Pipe inner diameter Therefore, fluctuation Vortex generator 13 due to lift force F _L
The signal bending moment Ms is Ms = F _L・D/8・³ (2 ³ −D ² −6 ² x + D ² x) (15); Length in the longitudinal direction of the vortex generator x; Distance from the upper fixed end The bending moment Ms' of the sensor part 14 is
_{Since the bending} stiffness _{is the ratio of the bending stiffness to}
)(16) I ₀ ; Moment of inertia of the sensor section 14 E ₀ ; Young's modulus I ₁ ; Moment of inertia of the outer cylinder section 135 E ₁ ; According to the Young's modulus, the stress in the sensor section 14 is σ _S = Ms'/I ₀ y (17) y: Distance from the neutral axis Stress detection sensors 141, 14 made of piezoelectric elements
The amount of charge Q generated in 2 is: Q=∫ _s σ _S d ₁₁ ds=∫ _s Ms'/I ₀ yd ₁₁ ds (18) S: Area of piezoelectric element 140 d ₁₁ ; Piezoelectric constant Now, Young's modulus changes depending on temperature. However, the bending moment (5-6%/200℃ for stainless steel)
As is clear from equation (16), Ms' is E ₁ /E ₀ , so if the rate of temperature change in the Young's modulus of the sensor section 14 and the Young's modulus of the outer cylinder section 135 is equal, then Ms' will be equal to the temperature. It is constant regardless.

That is, the piezoelectric element 140 and the insulating plate 145 are thin, and the push rod 147 and electrodes 1401, 1402, and 14 occupy the majority of the entire sensor section 14.
03, 1404, and the spacers 144, 146 are of the same type as the metal material constituting the outer cylinder portion 135, E ₁ /E ₀ remains a constant value even if the temperature changes.

As a result, the amount of generated charge Q expressed by equation (18) is independent of temperature changes in Young's modulus. Note that the moment of inertia of area I ₀ and I ₁ are constants determined by the shape and dimensions.

This is an advantage of the so-called stress detection method of the embodiment shown in FIG. 2, and unlike the case where the amount of strain or displacement is measured, the sensitivity does not change with respect to temperature changes, that is, changes in Young's modulus.

Next, the entire pipe line vibrates due to the influence of vibration noise propagating through the pipe line, for example, vibration noise caused by opening and closing of pumps, compressors, dampers, etc. Due to this vibration, an alternating bending moment M _N based on the mass distribution of the vortex generator 13 acts on the vortex generator 13 in the direction in which the aforementioned alternating force F acts.
Due to this alternating bending moment M _N , the vortex generator 1
3 is detected by the stress detection sensors 141 and 142.
detected as noise.

FIG. 3 shows this bending moment M _N , where Ms is the alternating bending moment (object to be measured) caused by vortex generation.

In this application, stress detection sensors 14 are installed on one side of the position A where the bending moment M _N becomes zero.
1,142, for example, at a certain instant, the stress detection sensor 141 detects the positive stress due to external vibration noise, the stress detection sensor 142 detects the positive stress, and the addition/subtraction circuit 33 adds the positive stress. , I tried to actively cancel it. However,
One of the inputs of the addition/subtraction circuit 33 is configured to be variable, so that the magnitude of the noise of the second input processing circuit can be easily adjusted to match the magnitude of the noise of the first input processing circuit. can. Therefore, the addition/subtraction circuit 33 can completely cancel the noise. As a result, it is not necessary to strictly arrange the stress detection sensors 141 and 142 at respective positions where the positive amount of the bending moment M _N is equal to the positive amount of the bending moment M N , and The arrangement becomes easy and the device can be manufactured at low cost.

Now, if we express this in a mathematical formula, the charges q ₁ and q ₂ generated in the stress detection sensors 141 and 142 are the signal charges caused by the vortex generation and the noise charges caused by pipe vibration, etc., superimposed, and the amplitude of the signal charges is S ₁ , S ₂ ,
Letting the amplitude of the noise charge be N ₁ and N ₂ , it is given by the following equation.

q ₁ =S ₁ sinωt+N ₁ sinω′t (19) q ₂ =S ₂ sinωt+N ₂ sinω′t (20) where, ω: Signal angular frequency ω′: Noise angular frequency (18)−λ×(19) Then, q _put = (S ₁ − λS ₂ ) sinωt + (N ₁ − λN ₂ ) sinω′t(21) N ₁ = λN ₂ , and q _put = (S ₁ − λS ₂ ) sinωt (22) Become.

Therefore, if a small gap is maintained between the circumferential side of the recess 134 of the vortex generator 13 and the stress detection sensors 141, 142, the insulating plate 145, and the spacers 144, 146, the bending moment of the signal can be reduced. The bending moment M _N of Ms and noise is stabilized.

In addition, as shown in Fig. 3, vibration noise is represented by the vector sum of N _A caused by the own weight of the vortex generator 13 and N _B caused by the mounting case mounted on the top of the vortex generator 13. Detection sensor 14
The position of 1,142 is N _1A = λN _2A (23) N _1B = λN _2B (24) However, N _1A ; Vibration noise due to the weight of the vortex generator detected by the stress detection sensor 141 N _2A ; Vibration noise due to the weight of the detected vortex generator N _1B ; Vibration noise due to the tower case detected by the stress detection sensor 141 N _2B ; Determined to be the vibration noise due to the tower case detected by the stress detection sensor 142 Then, the noise can be easily removed.

In the case of the vortex flowmeter, since the output is a frequency, changes in the temperature sensitivity of the piezoelectric elements used in the stress detection sensors 141 and 142 (temperature changes in the piezoelectric constant) are not a problem, but in the case of the present application. teeth,
Since the absolute value of the fluctuating lift force is required, piezoelectric elements with a smaller temperature coefficient of piezoelectric constant than lead titanate, lead titanate-lead zirconate (PZT), or lithium niobate (LiNbO ₃ ), such as crystal X-cuts, are used. It is desirable to select

In addition, in an actual piping line, the design λ and 0.05 to 0.1
Δλ/λ may vary by ±5%.
To _make Δq/q less than _{± 1} % with respect to _the change in
/S ₁ −λS ₂ = −ΔλS ₂ /S ₁ −λS ₂ = −Δλ/S ₁ /S ₂ −λ
(25) S ₁ /S ₂ ≧6 (26) should be selected as the condition that λ≒1, Δλ=±0.05 and equation (20) is less than ±0.01.

In other words, regarding the installation positions of the two stress detection sensors, it is necessary to satisfy equations (26), (23), and (24) almost simultaneously, but it is not that difficult. Equation (26) is satisfied at the point where is 0, and then the stress detection sensor 142 is determined to satisfy Equations (23) and (24).

From the calculation and experimental results, the signal bending moment Ms due to the fluctuating lift of the sensor section 14 and the noise bending moment M _N due to vibration are shown as shown in FIG.

The bending moment M _B is a nearly flat straight line, while the bending moment M _A is a curve with a valley in the center, so M _1A /M _2A = M _1B /M _2B = λ
It is easy to find the position of the stress detection sensor 142.

Further, although a labyrinth channel 15 is provided at the intersection of the vortex generator 13 and the pipe line 11, if there is no labyrinth channel 15, if the fluid to be measured is steam or the like, immediately after the start, Steam flows from the pipe 11 into the gap between the pipe 11 and the top side of the vortex generator 13, and the steam flows into the outer cylindrical part 135 of the vortex generator 13.
When it becomes a drain, the volume decreases significantly (the pressure decreases significantly), and new steam immediately flows in. By repeating this process,
The temperature of the outer cylinder part 135 of the vortex generator 13 rises rapidly,
A large temperature difference occurs between the outer cylindrical portion 135 of the vortex generator 13 and the inner cylindrical portion composed of the stress detection sensors 141, 142, the underlay 143, the push rod 147, the insulating plate 145, and the spacers 144, 146, and the vortex Thermal expansion of the outer cylinder portion 135 of the generator 13 increases,
The surface pressure on the stress detection sensors 141, 142 decreases, and the sensitivity of the stress detection sensors 141, 142 decreases. When the labiliance channel 15 is provided, the measurement fluid is repeatedly compressed and expanded at each stage, so the pressure drop is reduced and the amount of steam inflow can be reduced, so the sensitivity of the stress detection sensors 141 and 142 is reduced. can be prevented. A decrease in the sensitivity of the stress detection sensors 141 and 142 is not fatal in detecting the vortex frequency because it is sufficient to measure the frequency, but in detecting the lift signal (ρV ² ), the decrease in the absolute value of the lift signal This results in a large error.

Note that with fluids other than steam, even if heat is applied, there is almost no change in the state, and the fluid remains in the gap between the pipe line 11 and the top side of the vortex generator 13, and the flow of measurement fluid is small, causing no problems. Not.

Note that the heat conduction of the fluid from other than the gap has almost the same temperature speed in the inner cylinder and the outer cylinder, and by making the temperature expansion of the inner cylinder and the outer cylinder almost equal, There is no pressure drop.

FIG. 10 shows a specific configuration of the electric circuit shown in FIG. 6.

In the figure, OP1 and OP2 are operational amplifiers that constitute charge converters 31 and 32, and the first and second
The charges generated by the stress detection sensors 141 and 142 q ₁ ,
Convert q ₂ to a low impedance alternating voltage. Here, capacitor C ₁ = C ₂ , resistor R ₁ = R ₂ , and C ₁ and
The constants are selected so that the corner frequency _c of the high-pass filter composed of R ₁ , C ₂ and R ₂ is sufficiently smaller than the lower limit of the vortex signal frequency. OP3 is an operational amplifier that constitutes the adder/subtracter circuit 33 and removes vibration noise. Adjust the constant λ with variable resistor _R4 .
OP4 is an amplifier constituting the second amplifier circuit 34;
OP5 and OP6 are operational amplifiers forming detection rectifier circuits 35 and 36. C ₃ and C ₄ are coupling capacitors whose constants are selected so as to have a sufficiently small impedance with respect to the eddy signal frequency. R ₅ and C ₅ are time constants of the rectifier circuit, and are selected to have values equal to the time constants of R ₁₇ and C ₁₂ of the rectifier circuit of the F/V converter, and a value sufficient to reduce ripple. OP7 is an operational amplifier for obtaining a predetermined input voltage to the divider circuit 42 by adjusting the resistor _R6 .
OP8 and OP9 are operational amplifiers configured with a filter circuit 37 and a first amplifier circuit 38, which remove high frequency and low frequency noise contained in the output of the adder/subtractor circuit 33, and convert the eddy signal into an input signal to the Schmitt trigger circuit 39 at the next stage. It is amplified up to. _C6 , _C7 , _C8 ,
R ₇ , R ₈ , R ₉ are capacitors and resistors that make up the bandpass filter, C ₉ , C ₁₀ , C ₁₁ , R ₁₀ , R ₁₂ , R ₁₃
are also the capacitor and resistor that make up the bandpass filter. D ₁ and D ₂ are Zener diodes for canceling the low-pass filter composed of C ₁₁ and R ₁₃ and limiting the amplitude when the signal level becomes large. _OP10 is an operational amplifier that constitutes the Schmitt circuit 39, and the width of hysteresis is determined by _R14 , _R15 , and _R16 . OP ₁₁ is F/V converter 41
This is an operational amplifier that amplifies the output of the divider circuit 42 to a predetermined input voltage of the divider circuit 42, and is adjusted by _R18 . OP12 is a comparator, which is an operational amplifier that obtains a control voltage for a switch circuit SW that turns on and off the output of the divider circuit 42 based on a reference voltage E set as a zero-cut setting voltage.

In the block diagram of FIG. 6, the filter circuit 37 and the first amplifier circuit 38 are shown separated by function, but in FIG. 10, the operational amplifier OP8,
An active filter using OP9 is shown. Furthermore, the gate circuit in FIG.
In the figure, comparator OP12 and switch circuit SW
It consists of FIG. 11 shows an example of an output waveform when an actual flow test of water at a flow rate of 0.4 m/s is performed using the circuit shown in FIG. 10.

As shown in FIG. 11, the output waveform (A) of the charge converter 31 (point A in FIG. 10) includes a vortex signal and pipe vibration noise, and the charge converter 31
Pipe vibration noise appears in the output waveform (B) (point B in Fig. 10). The output waveform (C) of the adder/subtractor 33 (point C in Figure 10) is a waveform with pipe vibration noise removed, and the input waveform (D) of the Schmitts circuit 39 (point D in Figure 10) is a high-level waveform. The waveform shows a significantly improved S/N ratio due to the cut filter effect.

This waveform example is at a low flow rate,
At high flow speeds, low-frequency noise is superimposed on the vortex signal, or a beat waveform appears on the vortex signal itself, and the S/N ratio can be improved mainly due to the low-cut filter effect.

Further, as a result of an actual flow rate test using water, as shown in FIG. 12, it was confirmed that the output of the rectifier circuit 36 was proportional to the square of the flow velocity within an error range of ±1%. It has also been confirmed that the lifting force of air and water is proportional to the density ρ.

In the above embodiment, a so-called stress detection method was used to detect the alternating stress of a vortex generator with one end fixed and one end supported, but an alternating vortex generator with one end fixed and one end free It may also be one that detects stress.

In addition, other force detection methods that detect fluctuating lift,
For example, in terms of displacement, strain, torque, etc.
By temperature-compensating changes in the sensor's mechanical constants (Young's modulus, spring constant) due to temperature changes, a mass flowmeter using Karman vortices can be constructed.

(Effects of the Invention) As explained above, the present application provides a signal conversion circuit that converts the vortex signal from the vortex detection sensor into a signal, and a signal conversion circuit that converts the vortex signal from the signal conversion circuit into a signal from the filter circuit that reduces high frequency and low frequency noise of the signal from the signal conversion circuit. a first amplifier circuit that amplifies the signal, a comparator that converts the signal from the first amplifier circuit into a pulse signal, an F/V converter that converts the signal from the comparator into a DC voltage proportional to its frequency, and a signal conversion circuit. a second amplifier circuit that amplifies the signal from the second amplifier circuit; a detection circuit that detects the signal from the second amplifier circuit; a time constant circuit that removes ripples from the signal from the detection circuit; It is equipped with a division circuit that divides by the signal from the V converter circuit, and a gate circuit that sets the output of the division circuit to 0 when the signal from the first amplifier no longer reaches the set trigger level of the comparator. Since we constructed a mass flowmeter in which the time constants of the time constant circuit and the F/V converter are approximately the same and are greater than the beat frequency of the vortex signal, the variable lift signal detection circuit and the vortex frequency detection circuit By separating the signals immediately after the signal conversion circuit, it is possible to both accurately measure the absolute value of the fluctuating lift signal and reliably detect the vortex frequency.

In addition, a gate circuit is provided that sets the output of the divider circuit to 0 when the signal from the first amplifier no longer reaches the set trigger level of the comparator, which prevents large errors from occurring near zero flow rate. It can be prevented.

In addition, since the time constants of the time constant circuit and the F/V converter are made almost the same and larger than the beat frequency of the vortex signal, it is possible to accurately follow fluctuations in mass flow rate. can get.

As described above, according to the present invention, it is possible to realize an apparatus that can accurately measure density or mass flow rate with a simple configuration.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram of the configuration of one embodiment of the present invention.
is a front view, B is a side view, FIG. 2 is a configuration explanatory diagram showing a cross section of the detector section of FIG. 1, FIG. 3 is an explanatory diagram of the operation of FIG. 2, and FIGS. Figure 2 is an explanatory diagram of the main parts, Figure 6 is the electrical circuit diagram of Figure 1, and Figure 7 is an explanatory diagram of the main parts.
Figure 8 is an explanatory diagram of the operation of Figure 6, and Figure 9 is an illustration of the operation of Figure 2.
Figure 10 is an explanatory diagram of the specific configuration of Figure 6, and Figure 11 is an illustration of each part of the circuit in Figure 10 when an actual flow test of water at a flow rate of 0.4 m/s is performed. Output waveform diagram, Figure 12 is an output characteristic diagram of the rectifier circuit in Figure 10, Figure 13 is an explanatory diagram of lift, etc. when a vortex generator is placed in the pipe, and Figure 14 is an explanatory diagram of the drag coefficient. , FIG. 15 is an explanatory diagram of the configuration of a conventional example commonly used, and FIG.
Figure 17 is a diagram to explain the characteristics of the drag coefficient, Figure 18 is a diagram to explain the characteristics of Strouhal number,
FIG. 9 is an explanatory diagram of the lift coefficient. 10...Vortex flow rate detector, 11...Pipe line, 12...
... Nozzle, 13 ... Vortex generator, 131 ... Upper end,
132... Lower end, 133... Part in contact with fluid, 134
... recessed part, 135 ... outer cylinder part, 14 ... sensor section, 140 ... piezoelectric element, 141 ... second stress detection sensor, 142 ... first stress detection sensor,
143...Underlayment, 144...First spacer, 1
45... Insulating plate, 146... Second spacer, 1
47...Push rod, 148...Heart mesh seal, 1481, 1482...Lead wire, 149...
...Communication hole, 1401, 1402, 1403, 14
04...electrode, 15...labiliance channel, 15
1... Fin, 152... Expansion chamber, 20... Vortex flow meter converter, 30... Electric circuit, 31, 32...
Charge converter, 33... Addition/subtraction circuit, 34
...Second amplifier circuit, 35...Detection circuit, 36...
Rectifier circuit, 37... filter circuit, 38... first
Amplifier circuit, 39... Schmitt trigger circuit, 41
...F/V converter, 42...Division circuit, 43
...gate circuit.

Claims (1)

[Claims] 1. A signal conversion circuit that converts the vortex signal from the vortex detection sensor into a signal, a filter circuit that reduces high frequency and low frequency noise of the signal from the signal conversion circuit, and a first amplifier circuit that amplifies the signal from the filter circuit. a comparator that converts the signal from the first amplifier circuit into a pulse signal; an F/V converter that converts the signal from the comparator into a DC voltage proportional to its frequency; and amplifies the signal from the signal conversion circuit. a second amplifier circuit for detecting a signal from the second amplifier circuit, a detection circuit for detecting a signal from the second amplifier circuit, and a rectifier circuit for removing ripples from the signal from the detection circuit;
a division circuit that divides the signal from the rectifier circuit by the signal from the F/V converter circuit; and a division circuit that divides the signal from the first amplifier by the signal from the F/V converter circuit; a gate circuit that sets the output of the circuit to 0, and the time constant of the rectifier circuit and the time constant of the rectifier circuit of the F/V converter are made substantially the same and larger than the beat frequency of the vortex signal. Mass flow meter.