This invention relates to a method and a device for ultrasound measurement of the rates of flow of flowing media.

According to a first teaching, the invention concerns an ultrasound method of measuring the rate of flow of a flowing medium us'mg a measuring tube and at least two pairs of ultrasound transducers attached to the measuring tube forming measuring paths, in which the rate of flow of the flowing medium is determined from the velocities of the medium along at least two measuring paths. The term medium here includes both liquids and gases.

According to a second teaching, the invention also concerns a device for using the ultrasound method of measuring the rate of flow of a flowing medium with a measuring tube, at least two pairs of ultrasound transducers arranged on the measuring tube forming measuring paths, a transducer determining the velocities of the medium along each measuring path from the signals of the pair of ultrasound transducers and an adder that finds the rate of flow of the medium from the velocities of the medium along the measuring paths.

The known methods and devices for measuring the average velocity or the rate

flow of a flowing medium by ultrasound use a large number of measuring paths, which form the supporting points of a digital integration method that is as optimal as possible.

Here, the integration method is normally determined by the dimensions of geometry of the measuring path or the measuring tube. There are various traditional optimal integration methods by Chebychev, Gauss or Taylor, which are given, for example, in patent applications CH-A-610 038, DE-A-30 38 213 and EP-A-0 125 845. The methods and devices known from these patent applications are dependent on the viscosity of the medium and hence on the Reynolds' number for their precision. For example, please refer to the article "A New Integration Technique For Flowmeters with Chordal Paths" in Flow

Measurement and Instrumentation, Vol. 1, No. 4, July 1990, Pages 216 to 224.

The methods and devices known from patent applications for ultrasound flow measurement have insufficient precision due to their dependence on the viscosity of the medium, since the viscosity can change sharply during measurement, particularly as a result

temperature changes. However, high precision is generally required, particularly when measuring rates of flow of flowing gases, petroleum products and chemical products, for example. A second important problem with the known methods and devices is disturbances in the velocity profile caused by installation effects, which also has a negative influence on precision.

TlTe products specified have extraordinarily high requirements for measurement precision. For example for crude petroleum, in the range of a rate of flow from SO2 to 100% of the nominal rate of flow, the maximum error is 10.15%, and in the range of a rate of flow 10% to 100% of the nominal rate of flow, the maximum error is

the past, this precision could only be guaranteed with turbine meters.

When measuring the rate of flow of a medium flowing in a measuring tube, it is advantageous not to disturb the flow of the medium. At the same time, the point is to obtain the high measurement precision required using a relatively inexpensive device, which also has a long life. It is also advantageous if such a device can be calibrated with water after production and can then be recalibrated with other liquids or even gases during operation by the users, in order to guarantee the precision required.

Accordingly it is an object of the present invention to provide an improved technique for measuring the rates of flow of flowing media ultrasonically.

The task of the invention is, therefore, to eliminate the problems mentioned and to provide a method and a device with which such high precision can be guaranteed.

Another task of the invention consists of making it possible with the method and device of the invention to sharply reduce the influence of the viscosity of the medium.

The task of the invention is also to provide a method and a device that reduce the influence of changes in the flow profile and that ofTer the possibility of self calibration during operation.

Finally, it is also the task of this invention to provide a method and a device that make possible constant determination of viscosity of the medium, hence, in "realtime", and that also make it possible to identify the type of medium, for example the type of flowing crude petroleum, based on the viscosity and the sound velocity and/or sound dissipation.

According to the invention, the tasks listed and inferred above in the first teaching are solved with an ultrasound method of measuring the rate of flow of a flowing medium where the Reynolds' number of the flowing medium is constantly measured and where the value .for the rate of flow is corrected using the value for the Reynolds' number. In one advantageous design, the Reynolds' number is determined using the velocities of the medium along at least two measuring paths. These velocities of the flowing medium on various measuring paths can be determined simultaneously or in sequence.

In the second teaching of the invention, the tasks listed and inferred above are solved with a device for using the ultrasound method of measuring the flow of a flowing medium, characterized by the fact that it has a Reynolds' number meter that constantly determines the Reynolds' number and a flow-rate corrector connected to the adder and the

Reynolds' number meter.

One especially preferred design of this invention provides that before the

Reynolds' number is determined during operation, an operating flow profile based on the velocity measured is recorded and, in the event that the current flow profile is disturbed by inlet effects or other causes, an arithmetic correction of it is made based on

predetermined undisturbed calibration flow profile.

Now there are many ways of designing and improving the ultrasound method of measuring the rates of flow of flowing media in the invention or the device for using the ultrasound method of measuring the rates of flow of flowing media in the invention. For this, please refer, to the description of preferred embodiments of the invention illustrated in the accompanying drawings, in which:

Fig. 1. is a block diagram of a first embodiment of a device embodying the invention for using an ultrasound method of measuring the rates of flow of flowing media;

Figs. 2a to 2f show a flow diagram with explanations of the processes in the correction of the flow profile using the Fig. 5 device;

Figs. 3a and 3b show graphic examples of corrections in flow profiles with high Reynolds' number and low Reynolds' numbers;

Figs. 4a and 4b show graphs of the improvement in precision when the method in the invention is used for media with different viscosities;

Fig. 5 is a block diagram of a second embodiment of the device in the invention for using an ultrasound method of measuring the rates of flow of flowing media;

Fig. 6 is a block diagram of an example of embodiment for a flow-profile corrector;

Fig. 7 is a block diagram of an embodiment for a Reynolds' number meter;

Fig. 8 is a graph showing an example of an error curve based on empirical data for use in a flow corrector according to the invention; and

Figs. 9a to 9d show the dependence of various velocity ratios on the

Reynolds' number to explain the function of the method and the device in the invention.

With the device used for the ultrasound measurement method in the invention,

least two, but advantageously five, velocities are measured on different measuring paths; the measuring paths are formed by pairs of ultrasound transducers consisting of ultrasound transducers assigned to one another and arranged on different sides of the measuring tube.

Preferably, a flow profile of the medium in a duct connected to the invention device has been formed using inlet and outlet sections and developed as fully as possible. The calibration flow profile mentioned is preferably the best possible approximation of the flow profile in a fixly developed flow. It is known from practice that the velocities on certain measuring paths are less dependent on the Reynolds' number and, on other measuring paths, more dependent on it. The measuring paths less dependent on the

Reynolds' number are those at a distance of one half the radius of the measuring tube to the wall of the measuring tube. On the other hand, the measuring paths more dependent on the

Reynolds' number are, for example, in the middle or near the walls of the measuring tubes.

With the latter measuring paths, the flow profile has a maximum influence on the

Reynolds' number. The device in the invention can also work with more or~less than five measuring paths, but there must be at least one measuring path among them that

relatively less dependent on the Reynolds' number.

Since the Reynolds' number in the device of the invention is measured constantly, this measurement can be used in real time to correct the rate of flow and potentially to determine viscosity and, if necessary, also to identify the medium. This will be explained below.

Preferably, the velocities of the flowing medium measured on the measuring paths are used to determine the Reynolds' number. However, it is also possible to determine the

Reynolds' number in other ways, for example based on measurement of the ultrasound damping. The value for the Reynolds' number found is then used to make a correction in the rate of flow using an error curve. Of course, a value for the volume can also be determined from the average velocity and the rate of flow.

The method and the device for using it will now be explained with reference to

Fig. 1. Five pairs of ultrasound transducers connected to measuring tube 1 and forming measuring paths Ml to MS are connected to a transducer 2, which determines the various velocities of the flowing medium along measuring paths M1 to M5, for example from the running-time differences in the ultrasound signals. These velocities are fed to an adder 3 via various units. explained later, where they are multiplied by corresponding weight factors and then totaled. The average velocity feed to the output of the adder 3, hence the rate of flow per surface area of the cross section of the measuring tube, is applied to a rate of flow corrector 6 for correcting the rate of rate of flow. An error curve shown for example in Fig. 8 based on empirical data is stored in the flow corrector 6, and, besides the Reynolds' number, it contains all other technological tolerances connected with the device according to the invention. These tolerances are carefully measured after production of the device according to the invention that uses the ultrasound method of flow measurement according to the invention. The rate of flow calculated by the transducer 2 is now corrected based on a Reynolds' number determined by a

Reynolds' number meter S. The corrected rate of flow given by the rate of flow corrector 6 is then shown by an optional optical indicator device 4. As already mentioned, the device according to the invention can be calibrated with water, and the measurement results f' '~ obtained during calibration can also be transferred to other media, like other liquids and even gases, since the following applies to the Reynolds' number Re:

and where Vw and V~, are the flow velocities of water and a second medium v~ and v,~ are the kinematic viscosities of water and the medium, while D is the diameter of the measuring tube 1. At 20°C, the following applies:

and

This means that a device calibrated with water using the ultrasound method of measuring the rate of flow according to the invention works without problems with the mediurri.au, if the velocity of the air is higher by a factor of 15 than the velocity of the water during calibration.

Before determining the Reynolds' number, it is important to check the symmetry of the flow profile using the velocity ratios or velocity differences. If the actual flow profile is not disturbed or rather fully developed, the measured velocities will be used for further processing. A high symmetry of the flow profile is promoted, for example, by installing a Venturi nozzle in the duct.

Before the device according to the invention is started up by the user, it is calibrated, with water for example. Calibration is done in the range in which the device will later be used, for example in a range for the average velocity of 0.1 m/s to 6 m/s, for a plurality of measuring points, for example for 10%, 20%, 50% and 100% of the maximal average velocity. During this calibration, the velocities of the flowing medium measured for each measuring path are filed in a storage device when the calibration flow profile is not disturbed. This so-called calibration-profile matrix is characteristic for the device using the ultrasound method of rate of flow measurement according to the invention, since this matrix contains all of the mechanical, electronic, acoustic and hydraulic tolerances.

When correcting the symmetry of the current flow profile, two cases that depend on the Reynolds' number must be differentiated. On one hand, the case where one is working with large Reynolds' numbers above roughly 100,000 and, on the other hand, the case where one is working with smaller Reynolds' numbers. In the first case, the calibration-profile matrix for five measuring paths //EPM// takes the following form:

Ylpo~,, w VSp~oo~ ~~ ~YPioosc=

Fps' Y~ps~ ..- YS,psosc ~ ~s ~ ~,YPsosc (Equation 1)

YlPvo~ . ... YSp~os~ _ Gs ~YP~a~ where ~ ~ ~ v

Vlp, . . .VSp are the velocities of the flowing medium along the corresponding measuring paths during the calibration flow profile,

Vp is the corresponding average velocity or rate of flow per measuring tube cross section at the calibration flow profile,

GI; . . . GS are the weight factors assigned to the measuring paths, and 10%. . . .100% are the measuring points in the operating range.

With the device using the method according to the invention mounted and ready for operation, first the operating profile matrix /BPM// is recorded in another calibration process, and takes the following form:

Vlb- ... YSZ, yoo~ $Z'M~ ~ ~,~.~ ... Vbb GZ ~Vbxss

- ~ . ~ . , (Equation 2)

Ylb~o~ _.. VSbI~ Gs myb ;

where

Vlb . . . VSb are the velocities along the corresponding measuring paths with the operating flow profile,

Vb is the corresponding average speed or rate of flow per measuring tube cross section at the operating flow profile, and

Gl, . . . GS are, again, the weight factors of the measuring paths.

For the operating profile matrix /BPM// just introduced, the rates of flow in the measuring tube in a uniformity range, hence for Reynolds' numbers greater than

are set artificially identical to the rates of flow when the calibration flow profile is recorded -- for example using a mobile, calibrated flow-rate generator. In this case, the following is true:

EVbloox = EYpioox (Equation 3)

However, in practice, it is difficult to make Equation 3 come out precisely enough, since frequently the same rates of flow cannot be set exactly. To be able to correct the flow profile anyhow, Equation 3 is put in the following form:

- ~i ~ EVbioo.,~= EYploox ~. (Equation 3a) _- IzLEquation 3a, !3 is an interpolation factor to correct the fact that the same rates of flow cannot generally be set. Equation 3 a is synonymous to

Vlbloo~,c' Gl ' ~i...VSbloo~,c ~ GS - (3II=II~Yrme~~I (Equation 4)

Now a profile determinant /!Pr Det// is introduced, for which the following is true:

IIPrDet~~ = Dp'oov' __ V 1P'ooi. .. YSP'oov. (Equation S)

Db,~.,~ Vlbl~,~. ' l3 ~ VSbloo~. ' l3 where

II Dp,~ II is the profile determinant of the calibration profile matrix, and

II Db~~y. II is the profile determinant of the operating profile matrix.

When the method according to the.invention is used during operation by the user, the correction is made with the current profile matrix //APM//= /BPM//~//PrDet//:

Ylp~oox ~ ... . YSpioo~ G~ ybgec,ooo YIp ~ Ylb~ ... v5 b VSb~~, G2 ~~gec~~

(Equation 6)

Y1P'ooR yl},~ , YSpioos~ y56,~ Gs ~,'Vbgec,o~

VIb,~~ VSb,~~, where E Vbgec are the corresponding corrected average velocities or flow rates per cross section surface of the measuring tube in the measuring tube with the current flow profile.

In the form shown, Equation 6 applies only to media that behave linearly over the range considered from 10% to 100% of the nominal rate of flow. For nonlinear media, the corrections in the velocities of the medium along the measuring paths are made using the accompanying coefficients from the cafbration profile matrix and the operating profile matrix, for example Vlpso~,~/Vlbsoxfor a velocity of the flowing medium on measuring path Ml of Vlbsox. In the case of nonlinear media, it is also necessary to introduce the coefficients ail, biz. . .his; see also Equation 3a. For nonlinear media, the correction coefficients are also interpolated between values known only discretely. .

After correction of the current flow profile using the calibration profile matrix and the operating profile matrix, the relative error in the average velocities and rates of flow can be calculated with the following equation:

EvbgecF - EYE,oo~~. ~~~

Fe = (Equation 7)

YPioo'~ r=o_aoov.

In summary, the profile matrices mentioned were used for processing as follows.

First, using calibration, the velocities of the medium on the measuring paths and the accompanying average velocities and rates of flow were measured with an undisturbed calibration flow profile and then with an operating flow profile. Then, the ratio between the average velocities or rates of flow were recorded when the calibration flow profiles and the operating flow profiles were found. After that, the current measured velocities of the medium along the measuring paths with the current flow profile are changed

accordance with that ratio. Then, the ratios of the velocities of the medium along the measuring paths with the calibration flow profile and the deviating velocities of the medium along the measuring paths with the operating flow profile are found, and the corresponding velocities of the medium along the measuring paths with the current flow profile are multiplied by those ratios. Of course, if necessary, this correction is made with an interpolation.

If the velocity profile after correction of the flow profile as described above is undisturbed, the Reynolds' number can be determined based on that flow profile.

As akeady mentioned, the correction in the flow profile that was mentioned is carried out dependent on the Reynolds' number. The calibration profile matrix given in

Equation 1 can only be used for large Reynolds' numbers, roughly larger than

because in that case the right side of the Navier-Stokes hydrodynamic base vector equation disappears. _ . ~ + 0 ~ (SZ ~ V ) = Re ~ 2 ~ SZ (Equation 8) where SZ is the rotation of the velocity ~( , which means that SZ = O ~ ~/ and

Re is the Reynolds' number. (See also Equation (41.23) in "The Feynman Lectures on Physics, Reading" by R.

Feynman, R Leighton, M. Sands, Massachusetts, Palo Alto, London, AddisonWesley

Publishing Company, Inc. 1964).

If the Reynolds' numbers are large, from the hydrodynamic Equation 8, the hydrostatic base vector equation follows: + O ~ (SZ ~ V ) = 0 . (Equation 9)

For this case, the properties of the medium, for example the viscosity, were left out of consideration, since their influence is small. This small influence has as its result that the form of the flow profile in the uniformity range for Reynolds' numbers over 100,000 only changes insignificantly.

For the second case of smaller Reynolds' numbers, the influence of the properties of the medium is positively essential, so that it is necessary to use the calibration profile matrix in another form. During calibration, in this case, the viscosity of the medium ( v) and the diameter (D) of the measuring tube are measured, so that for each calibration flow profile, a corresponding Reynolds' profile matrix UPM Rep// in dimensionless form is obtained: t ~ ..

Rc1.ioosz

b ~ Itc

_ '. ~ : . _ Ro~aos~. ~ ~

(Equation

vP,oos~

P~oosr.

psosc ~=

Pioaasc ~ YPioox

_' = - V~pias~ ~ VSp,oss

Pioo~.... ~ypioo~

It can be inferred from Equation 10 that with the calibration done

the start for

each Reynolds' number, a calibration profile for the device for using the ultrasound method of rates of flow measurement in the invention can be stored in dimensionless form with compensation for tolerances (see Fig. 2a). Dimensionless means that the current velocities measured V 1, . . . VS are divided by the average velocity, hence the rate of flow per cross section surface area of the measuring tube at maximum rate of flow during calibration in the. installed state of the device in the invention, so that

VkT = Vi/E VK",aX.

While carrying out the calibration in the installed state, the velocities on the measuring paths and the average velocity, hence the rate of flow per cross sectional surface of the measuring tube, of the operating flow profile deviating from the calibration flow profile are determined according to Equation 2 as follows:

Vlc;~t-G, ... YSC,~",~-G5~=~~,'Vk~~ (Equation 11)

From this follows lastly the dimensionless operating profile matrix, which looks like this:

Vlk"~ . G~ ... yes. G

(Equation 12)

The dimensionless operating profile matrix is shown in Fig. 2b.

Based on velocities Vlc, ... VSC of the flowing medium measured during . operation on measuring paths Ml to M5, the Reynolds' number is determined in zero approximation Reo using equations later explained. For this Reynolds' number in zero approximation Reo, using Equation 10 for an identical Reynolds' number of the calibration

Rep, the velocities on the measuring paths, which can be shown in analytic form as functions, are determined from the calibration flow profiles (see Fig. 2c).

From these velocities on the measuring paths, the average velocity Vpgm is determined at the same time. This profile is then compared with the current flow profile (see Fig. 2d), in which the average velocity found Vgen (in zero approximation n=0, hence VgemO) is compared with the average velocity Vpgm of the calibration flow profile (dVgem =

Vgen - Vpgm). If the difference between the average velocities is greater than a certain maximum value E, then, in a subsequent iteration process, a smaller difference is assumed, for example Vgen (n+1) = Vgerrin + dVgem/2. From the new average velocity, the

Reynolds' number Rel is determined in the first approximation from the equation Rel =

Vgeml ~D/ v. Using this Reynolds' number in first approximation, from the stored calibration profile matrix, a new average velocity is found, which is then, in turn, compared (see Figs. 2e and 2f). If the difference found dVgem is smaller than the maximum value given E, the last value found for the Reynolds' number is used to correct the rate of flow. Improved precision is guaranteed using the iteration process described.

Figures 3a and 3b show examples of corrected dimensionless flow profiles for large Reynolds' numbers (Fig. 3a) and small Reynolds' numbers (Fig. 3b). In both figures, a is the calibration flow profile, b the disturbed operating flow profile and c the corrected operating flow profile.

After the correction of the flow profile described, the last value found for the

Reynolds' number is forwarded to the rate of flow corrector 6 to correct the rate of flow.

The whole method takes place in real time.

Figures 4a and 4b show examples of graphs of the increased precision with the invention. Fig. 4a shows, for an example of an embodiment with five measurement paths, the percentage of errors with three different medium viscosities (20 cSt, 40 cSt and SO cSt) as a correlation of the velocity in m/s with a state-of the-art method (see waveforms a, b and c) and with the device according to the invention using the ultrasound method of measuring the rate of flow of flowing media (see waveforms d, a and f). Here it is clear that the percentage of error in values, for the most part over 0.5% with the state-of the-art methods, is reduced by the method according to the invention to values under 0.2% for all three media.

Fig. 4b shows, for the same three media with different viscosities, the percentage of errors for the same measurement results that are shown in Fig. 4e, but now not as a function the velocity, but as a function of the Reynolds' number, again before and after correction. Here, what is striking is that all three waveforms a, b and c basically coincide when shown as a function of the Reynolds' number. Here again, it is clear that the precision is decisively improved with the ultrasound rate of flow measurement method according to the invention.

Depending on whether the flow profile has a turbulent or a laminar character, the

Reynolds, number is determined as follows:

For a flow profile with a turbulent character, the Reynolds' number is found from the velocity ratios or differences in measuring paths 2 and 4 (V2 + V4) and measuring paths 1 and S (Vl + Vs).

For a flow profile with laminar character, the Reynolds' number is found from the velocity ratios or differences in the velocities on measuring paths 2 and

Y4) and measuring path 3 (V3).

The Reynolds' number can thus be found based on velocity ratios (case a) for the velocities on the measuring paths and also based on velocity differences (case b) for the velocities on the measuring paths both for flow profiles with turbulent characters and for flow profiles with laminar characters.

For case a, where the Reynolds' number is found based on the velocity ratios, there is a flow profile with laminar character under the following condition: (V2 + V4)/V3 < 1,9 (Equation 13)

Inversely, a flow profile has a turbulent character when the following applies: (V2 + V4)/V3 > 1,9 (Equation 14)

The following equations for determining the Reynolds' number were found empirically.

For a flow profile with laminar character, the following applies to the

Reynolds' number: _ Red ' 1910g((YZ -t- V,~) ! Y3)s -- 6020Q(Yi +~V,~) VY3 + ~~OQ (Equation

In contrast, for a flow profile with a turbulent character, for the Reynolds' number, if it is smaller than 30,0000, the following applies:

Re, ~ 6500+39000 (5,14(Vz +Y~)ItV -t-Ys)-5,22) (Equation 16)

For a Reynolds' number > 20,000, with a flow profile with a turbulent character, the following is true:

Re, - 5080000((Y2 -~ V,~) ! (Y + Y3))2 ~ ~ ~

-I08600000(V1+V2)I(Y-1-Vj) i 5$33000 (Equation 17)

For case b, in which the Reynolds' number is determined based on velocity differences, the flow profile has a laminar character, if the following is true: (V2-~-V,~)-1,9Y3 <~a (Equation 18)

Inversely, there is a flow profile with a turbulent character if (VZ-~V')--1'gV >0 (Equation 19)

If the flow profile has a laminar character, the following now applies:

Re,PAI((Vi-~-V,,)--~~-Y3)j2)i'~'~(C~z"fY4)"~V-t-Vs))~2~~ (Equation20)

In contrast, the following is true for a flow profile with a turbulent character and

Reynolds' numbers smaller than 30,000:

Vi'fy~)"t~i.'t'Ys)12)s'f~stw~'~'vs)-tY d-Ys))~2-t; fn (equation 21)

Finally, for Reynolds' numbers greater than 20,000 and flow profiles with turbulent-character, the following is true: =A3(~V2'~''~~)-~~i-;-Ys)l2)2-i-Bj(('~2+V~)-('V,-F.VS))/2-1-~ equation)

The coefficients A1 to A3, B1 to B3 and Cl to C3 in Equations 20 to 22 are found empirically.

As already described, for the purpose of smooth functioning of the flow corrector during operation by the user, the current flow profile is controlled for deviations from the calibration flow profiles or asymmetries. Referring to Figs. 5 and 6, this control is done using a profile meter 7 and a profile corrector 9, connected between the transducer 2 and the adder 3. The profile meter 7 compares the velocities on the measuring paths and if there are profile deviations or a defective sensor, it gives a special signal from its output 23 to the profile corrector 9 and to an alarm 8. If in operation, due to installation or inlet effects like curvatures and comparable disturbances, a disturbed flow profile occurs, the deviation in this disturbed flow profile from the calibration flow profiles or the asymmetries in the disturbed flow profile can basically be eliminated by the profile corrector 9. This profile corrector 9 works on the basis of Equations 1 to 12.

A switch 11 shown in Fig. 6, forming a unit inside the profile corrector 9, has three settings: a setting a for flow profile calibration, a setting b for flow profile adjustment and setting c for flow profile monitoring.

Switch 11 is in setting a if the device according to the invention using an ultrasound method of measuring the rates of flow of flowing media is calibrated with an undisturbed reference flow profile. In this setting, the calibration profile matrix //EPM// is stored in the calibration flow profile memory 12 (see also Equation 1).

If the device according to the invention using the ultrasound method of measuring the rate of flow of flowing media is installed at the user, the rate of flow in the duct connected tothe device according to the invention is basically set at the maximum possible rate of flow during operation. In this case, switch 11 is in setting b. In this setting, the operation profile matrix lBPMII is stored in an operating profile memory 13 (see also

Equation 2). Next, a profile comparer 14 determines the profile determinants //PrDet// based ors_Equations 3, 4 and 5.

Under normal conditions, switch 11 is in setting c during operation, so that the information on the velocities on the measuring paths is forwarded from the switch 11 directly to a profile transducer 15, which works according to Equation 6. At the output of the profile transducer 15, in principle, an undistrubed and corrected flow profile is available. Based on this flow profile, the Reynolds' number is determined in the Reynolds' number meter 5 (Fig. 5), which is then made avialable to the flow corrector 6.

This flow corrector 6 works with an error curve that.also takes into account technological tolerances of the device. Fig. 8 shows an example of such an error curve, wherein a maximum error of 0.15% is guaranteed by using this error curve.

If there are changes in the line connecting to the measuring tube or other hydraulic transitional processes, for example, if a control valve closes, the flow profile changes very quickly. This change is controlled by the profile meter 7, and if the change is significant, it gives a signal via output 23 to the alarm 8 and the profile corrector 9 (see also Fig: 6). In the profile corrector 9, the switch 11 is then switched from setting c to setting b. In this setting, the operating flow profile provided by the operating profile memory

compared with the calibration flow profile from the calibration flow profile memory 12. If there is too great a deviation between these two flow profiles, a feedback signal is given to switch 11 via the feedback output 22 (Fig. 6.), whereupon an operating flow profile is filed again in the operating flow profile memory 13. This happens until there

corrected operating flow profile in real time, which is then fed back to the transducer 15 by switch 11 in setting c.

The values for the velocities at the output of the profile meter 7 are fed to the

Reynolds' number meter 5 as well as to the adder 3 (see also Figs. 7 and 8). A turbulent laminar switch 16 in this Reynolds' number meter 5 works based on Equations 13 and 14 or 18 and 19. This turbulent laminar switch 16 is connected to a laminar flow meter 17, a turbulent flow meter 18 and a transitional low meter 19, wherein these flow meters 17, 18 and 19 work based on Equations 15, 16 and 17 or 20, 21 and 22. The values at the outputs of these flow meters 17, 18 and 19 for the Reynolds' number are then fed to an output operation amplifier 20. ~ ~ .

In Figs. 9a to 9d, the ratios on which the function of the Reynolds' number meter 5 is based are shown graphically as an example. Fig. 9a shows a graph of the ratio (V2 +

V4/"V3) as a function of the Reynolds' number, which is traced in millions, whose course determines the action of the turbulent laminar switch 16. Fig. 9b shows a graph of the

Reynolds' number as a function of the ratio (V2 + V4fV3) whose curve determines the action of the laminar flow meter 17. Fig. 9b shows both experimentally determined measurement data and theoretical data. Figs. 9c and 9d show the dependence of the

Reynolds' number, which is traced in increments of a thousand, on the ratio (V2+V~/(V4+V5), whose curve determines the processing in the turbulent flow meter 18

Figs. 9c and 9d show the connections mentioned both for measurement data, with oil and water as the flowing media, and also for theoretically determined data. For 9c, it is true that the Reynolds' number is roughly smaller than 30,000, while for Fig. 9d it is true that the Reynolds' number is roughly larger than 20,000.

The value determined in real time at the output of the output operation amplifier 20 for the Reynolds' number is fed to a viscosity meter 10 (Fig. 5) as well as to the flow corrector 6. This viscosity meter 10 determines the viscosity of the medium, based on the

Reynolds' number, the average velocity, hence the rate of flow per cross sectional surface area of the measuring tube and the diameter of the measuring tube 1.

The viscosity value at the output of the viscosity meter 10 is sent on first

display device 4 and then to a medium identifier 24. This medium identifier 24 is also provided with the ultrasound velocity determined by the transducer 2 within the medium and/or the ultrasound damping of the medium. ~. Based on the viscosity of the medium, the ultrasound velocity in the medium and/or the ultrasound damping of the medium, the medium identifier 24-determines the type of medium, for example, the type of crude petroleum, by making a comparison with data stored for known media.

We Claim: