GB2321705A

GB2321705A - Acoustic measurement of fluid properties

Info

Publication number: GB2321705A
Application number: GB9801438A
Authority: GB
Inventors: Michael Langley Sanderson; Alexander Robert Guilbert
Original assignee: Cranfield University
Current assignee: Cranfield University
Priority date: 1997-01-24
Filing date: 1998-01-26
Publication date: 1998-08-05
Anticipated expiration: 2018-01-26
Also published as: GB9801438D0; GB2321705B; GB9701535D0

Abstract

A method of, and apparatus for, measuring the acoustic impedance of, the density of, or mass of a flowing fluid is disclosed comprising providing a first transducer (1) coupled to a matching block (3) and a second transducer (2) coupled to a second matching block (4), the two blocks being spaced apart. The matching blocks (3,4) may be made of a plastic material. A fluid under test is provided between the blocks (5), and a first ultrasonic signal is generated in a first direction in the first block (6). Resultant signals (9,11,14) are detected by the first transducer and the second transducer. Similarly, a second acoustic signal (15, fig. 2) is generated in the second block by the second transducer and the resultant signals (16,18,21 of fig. 2) are detected by the first transducer and the second transducer (i.e. two directional measurements are made). A combination of the signals from both directions can then be processed to provide a measure of the acoustic impedance/density/mass flow of the fluid independent of any coupling losses which arise between the blocks and the fluid.

Description

IMPROVEMENTS RELATING TO ULTRASONIC MEASUREMENT OF PROPERTIES OF FLUIDS This invention relates to a method and apparatus for performing ultrasonic measurements on fluids, and in particular, it relates to an apparatus for the measurement of the acoustic impedance and/or the density of the fluid, either flowing or static. In one aspect the invention relates to a method of measuring the mass flow rate of a fluid.

The measurement of mass flow as opposed to volumetric flow is of considerable interest in several areas. For example, the rate of flow of energy in fuels is often used as a basis for billing, and it is known that mass flow rate is a better indicator of energy flow rate than volumetric flow rate. For example, the expansion of gases with increasing temperature has a significant effect on energy per unit volume of a gas.

The use of ultrasonic techniques to measure flow rates is well established, although the use of ultrasound to measure mass flow is less known. Usual ways of measuring mass flow rate include Coriolis, thermal, angular momentum transfer techniques, and dual measurement devices (where additional readings are used with a volumetric flow meter to determine mass flow). The applicant is aware of some work in the area of dual measurement of mass flow using ultrasound, and in general the prior art techniques require additional measurement of temperature and pressure of the gas in order to infer density. This is disadvantageous in that as well as the requirement for extra sensing elements in addition to the ultrasound sensors, a prior knowledge of the properties of the gas are sometimes needed, and assumptions as to its properties. Sampling techniques which measure the density of a sample extracted from the flow are also used, but can add significantly to the complexity and cost of the final meter and also to the time taken to make a measurement.

According to a first aspect the invention comprises a method of measuring the acoustic impedance and/or the density of a fluid (or other characteristic dependent upon the acoustic impedance and/or the density, such as mass flow rate) comprising providing a first transducer coupled to a first matching block of predetermined length and a second transducer coupled to a second matching block of predetermined length, the two matching blocks being spaced apart by a predetermined distance, and providing a fluid to be tested between the blocks; and in a first direction mode of operation, generating a first acoustic signal in the first block and detecting the signals generated by the first transducer and by the second transducer resultant from said first signal; in a second direction mode of operation, generating a second acoustic signal in the second block and detecting signals generated by the second transducer and by the first transducer resultant from said second signal; and processing the signals from said first and second modes of operation so as to calculate the acoustic impedance and/or the density of the fluid.

Most preferably, the acoustic signals comprise ultrasound signals.

It will be appreciated that in the following paragraphs, the word ultrasound is intended to cover any acoustic signal.

The bi-directional measurements allow us to eliminate the effect of coupling losses between the first transducer and the first matching block and between the second transducer and the second matching block. The coupling at the interface between a transducer and its matching block introduces a loss factor, f, for signals transmitted across that boundary.

Thus, if a signal when it is in the matching block, just before it is detected by a transducer is X, it will be detected as fX by the transducer.

It is practically impossible to make two identical transducer plus matching block with identical transducer/block boundary losses.

Transducer one will have loss fl, and transducer two loss f2.

We prefer to calculate the acoustic impedance and/or the density of the fluid using a formula which has a numerator and a denominator, the numerator comprising the product of detected signals from the first transducer and detected signals from the second transducer, and the denominator comprising also the product of detected signals from the first transducer and detected signals from the second transducer, and the product that is the numerator having the same number of signals from the first transducer as does the product that is the denominator, and the product that is the numerator having the same number of signals from the second transducer as does the product that is the denominator.

Preferably we detect the amplitude, or something dependent upon the amplitude, of the acoustic signals.

Preferably in the first direction mode of operation we detect at the first transducer a signal (al) reflected from the block boundary that is at an end of the first block remote from the region where the first transducer is coupled to the first block, and detect at the second transducer a signal (bl) that has travelled through the first block, the fluid, and the second block, and also a signal (c1) that has travelled through the first block, the fluid, and the second block, and has been reflected by both the first and second blocks so as to travel through the fluid at least once more in each direction than has the signal of (bl).

Preferably the signal of (bl) has travelled through the fluid only once. Preferably the signal of (cl) has travelled through the fluid only once more in each direction than the signal of (bl).

Preferably the signal of (bl) and the signal of (cl) have travelled through each matching block only once.

Preferably in the second direction mode of operation we detect at the second transducer a signal (a2) that has been reflected from the block/fluid boundary that is at the end of the second block remote from the region where the second transducer is coupled to the second block, and detect at the first transducer a signal (b2) that has travelled through the second block, the fluid, and the first block.

Preferably we also detect a signal (c2) at the first transducer that has travelled through the second block, the fluid, and the first block, and has been reflected by both the second and first blocks so as to travel through the fluid at least once more in each direction than has the signal (b2) .

Preferably the signal of (b2) has travelled through the fluid only once. Preferably the signal of (c2) has travelled through the fluid only once more in each direction than has the signal of (b2).

Preferably the signal of (b2) and have the signal of (c2) travelled through each matching block once only.

Preferably we measure the signals, (al), (bl), (c1), (a2), (b2), (c2) and use a formula to calculate the acoustic impedance and/or the density of the fluid that needs less than all of the values of them.

Preferably we process a selection of the detected signals (al), (bl), (cm), (a2), (b2), (c2), in two different ways which are expected to give the same value, and compare the values obtained by the two processes.

This can be used to check that the system is working properly. If there is a significant discrepancy it can be an indication that something is wrong.

Preferably we process the signals (al), (bl), (c1), (a2), (b2), (c2), according to the formula:

or to the formula

We may take the square root of either of the above.

Preferably the method is a method of measuring the acoustic impedance Zg of a fluid using blocks of acoustic impedance Zm and we process the signals to the formula:

where a is a factor associated with the beam spread in the blocks, and is obtained from a look up table or algorithm.

Preferably the a look up table or algorithm has as its input the speed of sound in the fluid Cg (and possibly the speed of sound in the block Cm). The value of Cg may be calculated by measuring the times that signals are received. The value of Cm may be assumed to be a known value, or measured by the apparatus. If this value of Cm is measured by the apparatus, then preferably it will be measured in a similar way to the value of Cg.

An alternative formula which the method may use is

Alternatively we may measure the density of the fluid (Qg). We Z prefer to use the formula Qg= g Cg We may measure the mass flow rate of a fluid, by measuring the volume flow rate of the fluid and multiplying it by our measured Qg It will, therefore, be readily understood that each of the first and second transducer assemblies behaves as both an emitter and a receiver of ultrasound signals. In one embodiment the same structure may be both an emitter and a receiver at different times. Alternatively the "transducer" may have a separate emitter and receiver.

Measuring the signals in both directions can enable the effects of imperfections at the emitter/detector to matching block interface to be eliminated.

In the manner discussed above we get for each direction of operation three measured amplitude signals. Combining the two sets of three signals provides sufficient data to enable a check on the correct operation of the system to be performed in addition to measurement of gas impedance and/or speed of ultrasound and/or gas density.

Preferably, each of the ultrasonic signals transmitted by the first and second transducer assemblies is substantially identical.

Most preferably, the ultrasonic signals comprise tone burst signals.

These may be in the form of a train of sinusoidal oscillations. The oscillations can advantageously be at the resonant frequency of the transducers.

In addition to obtaining output signals representative of the amplitude of the portions of the ultrasound signals incident upon the detectors, the time delay between the generation of a selection of the output signals may be calculated. Most preferably, the time delay between the generation of the first signal of a set and the transmission of the ultrasound signal may be calculated. Additionally, the time delay between the generation of signals b and c in a set of measurements may be calculated. These signals can be used to provide a measure of the speed of sound both in the gas and in the matching blocks.

In a refinement of the method, the duration of each of the toneburst signals may be selected so that the output signals produced by the transducers reach a steady-state. This can be achieved if the duration of the tone burst signal exceeds the rise time of the detectors.

In accordance with a second aspect of the invention we provide fluid acoustic impedance (or fluid density, or fluid mass flow rate, or other parameter dependant upon acoustic impedance) measuring apparatus, the apparatus having a first transducer assembly comprising an acoustic emitter and an acoustic detector both coupled to a first matching block, a second transducer assembly comprising an acoustic emitter and an acoustic detector, both coupled to a second matching block, and in which the first and second transducer assemblies are arranged adjacent a space adapted to be filled with fluid to be tested.

Preferably, the emitter and detector of each transducer assembly comprises a single transducer. This may be a piezo-electric transducer, which may emit and/or detect ultrasound.

Preferably the first and second matching blocks are made of the same material. The first and second blocks are preferably integral single pieces of material.

Preferably the first and second blocks each have an elongate central axis, and are aligned on a common central axis. Preferably they have the same length along their central axis. Preferably the blocks have the same geometric shape. Preferably the blocks have a front and rear end faces, the emitter and detector of each block being coupled to it at an end face.

Preferably the blocks are cylindrical. Preferably the end faces of the blocks are flat. Preferably, the end faces of the blocks are perpendicular to a common central axis.

The blocks are preferably mounted on rigid support means so as to ensure that they remain a fixed distance apart.

Preferably control means is provided adapted to cause an ultrasound signal to be generated by the emitters, and to process signals generated by the detectors.

Preferably the control means is adapted to generate an ultrasonic signal that is short enough, relative to the length of the blocks, to avoid overlap between successive detected signals.

Preferably the diameter of a block is at least three times the diameter of the transducer attached to it.

Preferably, the first and second matching blocks have a path length (the distance from the transducer to the opposite end of the block) which is chosen in relation to the speed of ultrasound in the block and the duration of the ultrasound pulses. Also, the ultrasonic impedance of the blocks is preferably substantially different to that of the gas which is to be measured.

Ideally, the length of the block should be chosen so that the first reflected signal from the block/gas interface can be measured without interference from unwanted sources of ultrasound reflections and is chosen so that 2 2 > tsigna 2m where t signal is the duration of the ultrasound burst, 1m is the path length of the block and Cm is the speed of propagation of the ultrasound signal in the block.

Preferably, the matching blocks may be made from a plastic or polymer material of low acoustic impedance.

When the blocks are cylindrical the radius of the blocks may be sufficiently large to prevent unwanted reflections from the blocks circumference interfering with the measurements. By choosing larger diameter blocks, the reflected ultrasound from the block sides arrives later than the desired ultrasound signals and it can be ignored or rejected.

Alternatively it can be said that by choosing larger diameter blocks unwanted modes of propagation of ultrasound within the blocks are substantially reduced and/or totally eliminated.

It will be apparent that the first output signals and the second output signals differ considerably in amplitude (i.e. the first output signals have a much higher amplitude than the second output signals due to attenuation at the block/gas interfaces). The apparatus may therefore comprise amplification means which have a switchable gain. Thus, when a detector is used to measure a first output signal it can have a lower gain than when it is used to measure a second output signal. Alternatively, different amplifier circuits of different gain may be provided to amplify different signals.

Preferably, a single detection circuit can be used with both the first and second detectors. This is possible due to the time delay between receipt of the first and second output signals in a set. In a particularly notable embodiment, an amplifier having a ramped gain can be used.

Initially, the amplifier has a low gain and at a first instant it can be used to measure the first output signal in a set. Then, the gain automatically increases during the time delay between receipt of the first output signal and the second output signal. It is therefore automatically set to a suitable high gain in time for measuring the second output signals.

A common drive circuit for the two emitters can also be provided, switching between the first and second transducers as required.

Preferably the apparatus is adapted to measure the acoustic impedance of a fluid, most preferably of a gas. The apparatus may be adapted to measure the density of the fluid. The apparatus may be adapted to measure the mass flow rate of the fluid.

The apparatus may have display means adapted to display a value for the characteristic measured. Alternatively or additionally it may generate an output signal indicative of the characteristic being measured said characteristic being at least one of: the acoustic impedance, density or mass flow rate of the fluid, or it may display or output more than one of these simultaneously. The user may be able to select which parameter is displayed by operating selection means.

The apparatus may be adapted to perform a self diagnostic check to calculate the measured characteristic twice, preferably using two different sets of detected signals (which can have some detected signals in common) and compare the results to check that they are certain and within an allowable margin of each other.

The apparatus may comprise an add-on unit or sub assembly adapted to convert a volume flow meter or sensor to a mass flow meter or sensor.

Our system has no moving parts, does not need us to measure the temperature or pressure of the fluid (gas), does not have a time delay whilst the temperature or pressure are being measured, does not need calibrating for use with different fluids, and does not have a calibration drift with time.

In one preferred embodiment the invention comprises a mass flow meter, either with the density/acoustic impedance sensor originally built into it, or as an add-on module for connection to a volume flow meter.

Whilst the main field of interest is gases, which have mass flows that can vary widely for the same volume flow, the invention can also be used in relation to liquids.

There will now be described an embodiment of the present invention, by way of example only, with reference to the accompanying drawings in which: Figure 1 shows the path of the ultrasonic signal produced by the first transducer assembly which generates the first set of output signals; Figure 2 shows the path of the ultrasonic signal produced by the second transducer assembly which generates the second set of output signals; and Figure 3 shows the effect of the detector band widths on the output signals generated by a constant amplitude sinusoidal pulse tone signal.

The system illustrated in Figures 1 and 2 is adapted to take two sets of measurements. One set (set A) is made using transducer 1 as a transmitter from which an ultrasonic signal 6 is sent out. This signal comprises a toneburst comprising a set of oscillations which are usually chosen to be at or near to the resonant frequency of the transmitting and receiving transducer 1. The second set of measurements (set B) is made using transducer 2 as a transmitter and we usually use substantially the same ultrasonic signal. The order in which these two sets of measurements are made does not matter. The system is bi-directional in that signals are required going from transducer 1 to transducer 2, and vice versa.

Figure 1 shows the relevant ultrasonic signals that occur within the system when transducer 1 is the transmitter, and figure 2 shows the relevant ultrasonic signals when transducer 2 is the transmitter.

A transducer assembly 100 is shown in Figure 1 comprising a transducer 1 coupled to a cylindrical block 3 of matching material (plastics polymer material in this example), and a transducer 2 coupled to a block 4 of matching material identical with block 3. The blocks 3 and 4 are spaced apart by a known distance lg, and have a known length 1m.

They are fixed to a support (not shown) so that their relative positions cannot change. The assembly 100 is adapted to be introduced into a gas, possibly being adapted to be introduced into a pipe to measure flowing gases. Each block is mounted on a common central axis and has a pair of end faces, the transducers being mounted on the outer faces of the blocks.

The transducers 1 and 2 are piezo-electric transducers capable of both creating an ultrasound signal and detecting one.

The transducers 1 and 2 are connected to electronic signal monitoring and processing means (not shown). Control means is provided to control the operation of the apparatus. Display means, or output means, (not shown) is provided to display the results of a test/measurement.

In the following paragraphs, a description of how the measurements are made is provided. It refers to obtaining measurement set A with the equivalent parts for set B in brackets.

As shown in Figure 1 (or Figure 2 for set B) the transmitted signal 6 (15) travels through matching material 3 (4) and reflects off its front face 7 (8) which is in contact with the gas 5 being measured. A proportion of this signal 6 (15) is reflected back from face 7 (or face 8) as signal 9 (16) to the transmitting transducer 1 (2). The amplitude of the electrical signal generated across transducer 1 (2) by this signal is measured as V1R(V2R). A proportion of signal 9 (16) is transmitted into the gas 5 as signal 10 (17). This travels across the gas 5 until it reaches the front face 8(7) of the matching material of the other block. A proportion of this signal 10(17) is transmitted directly into matching material 4 (3) as signal 11 (18) and reaches the receiving transducer 2 (1).

The amplitude of the electrical signal generated across transducer 2 (1) by this signal is measured as V2T1 (VITI). A proportion of signal 10 (17) is reflected back into the gas 5 at interface 8 (7) as signal 12 (19) and travels back across the gas 5 a second time to interface 7 (8) where a proportion is reflected from the face 7 (8) back into the gas 5 as signal 13 (20). This travels a third time across the gas 5 to interface 8 (7) where a proportion is transmitted into the matching material 4 (3) as signal 14 (21). This travels across the matching material 4 (3) to transducer 2 (1). The amplitude of the electrical signal generated across transducer 2 (1) by this signal is measured as V2Tz (ViT2). Thus, six amplitudes (VIR V2R V2Tl, V1TI V2T2 .VlT2) are measured. Previous systems have only measured three (either VIR V2Tl, V2T2, or V2R, VITI VIT2).

The time tm between transmission of the initial signal 6 (15) and the receipt of the reflected signal 9 (16) is also measured, as is the time tg between receipt of signal 11(18) and receipt of signal 14 (21). These times are used in the beam spread compensation techniques described below, and tg is also used in the computation of gas density from gas acoustic impedance.

The length lm of the matching blocks 3 and 4 and the width lg of the gas gap 5 need to be large enough to eliminate overlap between the signals arriving at the ultrasonic transducers. The size restrictions are also dependent on the duration of the ultrasonic signals. It is preferred that the ultrasonic signals be sufficiently long to ensure that when they are received by the appropriate ultrasonic transducer (1 or 2), they reach a 'steady state' which enables accurate measurement of signal amplitudes If V2R V2Tl, VITI V2T2 ViT2). If this does not happen then the amplitudes measured may be dependent on the bandwidth of the ultrasonic transducers (either acting as an emitter or a detector). This is shown in Figure 3.

Figure 3(a) shows a typical input waveform, (i.e. the input to the ultrasonic emitter or the ultrasonic signal incident upon a detector) and Figure 3(b) shows the resulting output signal generated by the transducer (either as an emitter in the first case or as a detector in the second case).

It is clear that between time to and tl, the signal is increasing until between t1 and t2 a steady state is reached.

A 'steady state' occurs when the amplitudes of successive cycles do not vary significantly. The bandwidth of the transducers will result in a time these before this occurs, and a time tfali at the end of the toneburst during which time the signal falls to an insignificant level. A certain period of time tsteady at steady state is also required to make the measurement. Thus the total duration signal of the ultrasonic signal is such that tsignal > trise + tsteady + tfall. The duration of all the times is also dependent on the frequency of the oscillations. For the measurements of VIR and V2R to be successful, it is preferable that 2lm > tsignal Cm where Cm is the speed of sound in the matching blocks 3 and 4.

Also to ensure that signals VITI and V1T2 (and signals V2Tl and V2T2) do not interfere is necessary that 2 g > tsignal Cg where Cg is the speed of sound in the gas 5.

Finally, to ensure that there is no interference between reflected ultrasonic signals other than those described above it is necessary that 2 im 's > 2 t Cm Cg siglS/ In this embodiment, the following values were selected: 500kHz for the frequency of the cycles in the ultrasonic signal (and also the resonant frequency of the ultrasonic transducers) 100mm for im 30mm for 1g For ease of computation of attenuation correction factors, the transducers and matching blocks are circular in cross-section, and the transducers and matching blocks all have the same centreline (see aligned on the same axis). The matching layers have a diameter that is considerably greater than the transducers. A transducer diameter of 16mm and a matching block diameter of 50mm have been found to be suitable.

As the ultrasonic signals travel through the system they spread out.

This results in a loss of signal and hence a change in the amplitude of the signals received by the ultrasonic transducers. The degree of beam spread (and signal loss) is determined by the physical size of the measurement system, the operating frequency, and the speed of sound in the matching blocks and the gas. Only the speeds of sound can vary in the current device (due to changes in the gas (the device can be used in different gases) and variations in temperature of the matching blocks and the gas).

The following system is used to compensate for these effects.

A lookup table is stored in memory of the device. This table is computed from standard equations determining beam spread (either analytically or numerically) and converted to provide a multiplicative factor a. Two variables control the value taken from the table by the system. These variables are either the speeds of sound Cm and C g or the times tm and t5. The speeds of sound Cm and Cg can be derived from the times tm and tg using the following formulae.

2l 2l Cm = 2lm C g = tm tg The values used to access the table will be those closest to the values measured by the system. Some form of interpolation (linear or otherwise) may be used to improve the value of a provided by the table.

The following formulae can be used to compute the acoustic impedance of the gas Zg.

or

where Zm is the acoustic impedance of the matching block. This can either be stored as a predetermined number or preferably computed as lm Zm = 2Qm tm where Qm is the density of a matching blocks. Use of this second equation allows for compensation for variations in temperature which can alter the 21 speed of sound in the matching blocks (since Cm = tm Use of tm to correct em is also proposed (the speed of sound in the block is dependent upon its temperature and the degree of thermal expansion (and hence degree of density change) is also dependent upon temperature so we could have a look up table or algorithm to identify Qm from a measurement of tm.

The speed of sound in the gas can be determined from Cg = 2lg fg and the density of the gas Qg can be determined as Zg Qg = Since there are two equations available to measure zgs a comparison of these results from them gives a measure of the quality of the result obtained. Since a and Zm are the same in the two equations it is only necessary to compare the terms

The degree to which these agree with each other is used as a measure of the degree of confidence in the measurement (a validity check). If they are significantly different the result can be discarded or a warning generated. This can also be used as a monitor of the system to determine if anything has failed.

Thus we need to know VIR V2R, VITI V2T1 and one of V2T2 or VITZ in order to know Qg< between them (using the timer mentioned in 1 above or another timer).

4. Repeating items 1,2 and 3 but switching the roles of the transducers and measuring V2R, VITI and VlT2.

5. Computing the relevant values.

6. Going to step 1.

An alternative strategy permits the addition of a step 2a which waits for the system to ring down and repeats step 1 before proceeding to step 3 i.e. signals are taken from two different waves generated by different activation of the transducer 1.

It will also be understood that we may also repeat steps 1 to 3 several times and average the results to obtain better values for VIR, V2Tl and V2T2. (The same averaging would be done in step 4).

By using bi-directional tests we can eliminate factors associated with the effects of the boundary between the transducers and their matching blocks.

The embodiment of the invention show in the drawings has several points worth noting: 1. A bi-directional measurement mode (using ultrasonic transducers 1 and 2 as both transmitters and receivers).

2. Correct sizing of the measurement system.

3. A beam spread compensation algorithm based on the speeds of sound in the matching blocks 3 and 4 and the gas 5.

4. Formulae for the computation of gas acoustic impedance and density (which automatically compensate for variations in the performance of transducers 1 and 2 and their coupling to matching blocks 3 and 4) 5. A system to provide a measure of the quality of the results from the device using extra readings resulting from use of the system in a bidirectional mode, and 6. Switched receiver gain system and measurement strategy.

Claims

1. A method of measuring the acoustic impedance and/or the density of a fluid comprising providing a first transducer coupled to a first matching block of predetermined length, providing a second transducer coupled to a second matching block of predetermined length, the two matching blocks being spaced apart by a predetermined distance, providing a fluid to be tested between the blocks; and in a first direction mode of operation, generating a first acoustic signal in the first block and detecting the signals generated by the first transducer and by the second transducer resultant from said first signal; in a second direction mode of operation, generating a second acoustic signal in the second block and detecting the signals generated by the second transducer and by the first transducer resultant from said second signal; and processing the signals from said first and second modes of operation so as to calculate the acoustic impedance and/or the density of the fluid.

2. A method according to claim 1 in which the acoustic signals comprise ultrasound signals.

3. A method according to claim 1 or claim 2 in which the acoustic impedance and/or the density of the fluid is calculated using a formula which has a numerator and a denominator, the numerator comprising the product of detected signals from the first transducer and detected signals from the second transducer, and the denominator comprising also the product of detected signals from the first transducer and detected signals from the second transducer, and the product that is the numerator having the same number of signals from the first transducer as does the product that is the denominator, and the product that is the numerator having the same number of signals from the second transducer as does the product that is the denominator.

4. A method according to any preceding claim in which the signals are indicative of the amplitude of the acoustic signals.

5. A method according to any preceding claim in which in the first direction mode of operation the first transducer is adapted to detect a signal (al) reflected from the block boundary that is at an end of the first block remote from the region where the first transducer is coupled to the first block, the second transducer is adapted to detect a signal (bl) that has travelled through the first block, the fluid, and the second block, and also detects a signal (cl) that has travelled through the first block, the fluid, and the second block, and has been reflected by both the first and second blocks so as to travel through the fluid at least once more in each direction that has the signal of (bl).

6. A method according to claim 5 in which the signal of (bl) has travelled through the fluid only once.

7. A method according to claim 5 or claim 6 in which the signal of (c1) has travelled through the fluid only once more in each direction than the signal of (bl).

8. A method according to any one of claims 5,6 or 7 in which the signal of (bl) and the signal of (c1) have travelled through each matching block only once.

9. A method according to any preceding claim in which in the second direction mode of operation a signal (a2) is detected at the second transducer that has been reflected from the block/fluid boundary that is at the end of the second block remote from the region where the second transducer is coupled to the second block, and at the first transducer a signal (b2) is detected that has travelled through the second block, the fluid, and the first block.

10. A method according to claim 9 in which a signal (c2) is detected at the first transducer that has travelled through the second block, the fluid, and the first block, which has been reflected by both the second and first blocks so as to travel through the fluid at least once more in each direction than has the signal (b2).

11. A method according to claim 9 or claim 10 in which the signal of (b2) has travelled through the fluid only once.

12. A method according to claim 9, 10 or 11 in which the signal of (c2) has travelled through the fluid only once more in each direction than has the signal of (b2).

13. A method according to claim 12 in which the signal of (b2) and the signal of (c2) have travelled through each matching block once only.

14. A method according to any preceding claim comprising the steps of measuring 6 signals, (al), (bl), (cl), (a2), (b2), (c2) and calculating the acoustic impedance and/or the density of the fluid with a formula that needs less than all of the six signals.

15. A method according to claim 14 comprising the steps of processing a selection of the detected signals (al), (bl), (cl), (a2), (b2), (c2), in two different ways which are expected to give the same value, and comparing the values obtained by the two processes.

16. A method according to any one of claims 14 or 15 in which the signals (al), (bl), (cl), (a2), (b2), (c2), are processed according to the formula:

or to the formula

17. A method according to claim 16 comprising the further step of taking the square root of the formula.

18. A method according to claim 14 in which the signals are processed according to the formula:

where a is a factor associated with the beam spread in the blocks, Zg is the acoustic impedance of the fluid and Zm is the acoustic impedance of the blocks, and o is obtained from a look-up table or algorithm.

19. A method according to claim 18 in which the a look up table or algorithm has as its input the speed of sound in the fluid Cg and optionally the speed of sound in the block Cm.

20. A method according to claim 19 in which the value of Cg is calculated by measuring the times that signals are received.

21. A method according to claim 14 in which the signals are processed according to the formula:

where oc is a factor associated with the beam spread in the blocks, Zg is the acoustic impedance of the fluid and Zm is the acoustic impedance of the blocks.

22. A method according to any one of claims 18 to 21 comprising the further steps of measuring the mass flow rate of the fluid by measuring the volume flow rate of the fluid and multiplying it by a measured value of the fluid density Qg calculated from the formula; Zg Qg = Cg

23. A method according to any preceding claim in which each of the ultrasonic signals transmitted by the first and second transducer assemblies is substantially identical.

24. A method according to any preceding claim in which the signals comprise tone burst signals of ultrasound in the form of a train of sinusoidal oscillations at the resonant frequency of the transducers.

25. A method according to any preceding claim comprising the further step of calculating the time delay between the generation of a selection of the output signals.

26. A method according to claim 25 in which the duration of each of the toneburst signals is selected so that the output signals produced by the transducers reach a steady-state.

27. A method of measuring the acoustic impedance and or density of a fluid substantially as described herein with reference to the accompanying figures.

28. A fluid acoustic impedance measuring apparatus, the apparatus having a first transducer assembly comprising an acoustic emitter and an acoustic detector both coupled to a first matching block, a second transducer assembly comprising an acoustic emitter and an acoustic detector, both coupled to a second matching block, and in which the first and second transducer assemblies are arranged adjacent a space adapted to be filled with fluid to be tested.

29. An apparatus according to claim 28 in which the emitter and detector of each transducer assembly comprises a single transducer.

30. An apparatus according to claim 29 in which each transducer assembly comprises a piezo-electric transducer.

31. An apparatus according to claim 28, 29 or 30 in which the first and second matching blocks are made of the same material.

32. An apparatus according to any one of claims 28 to 31 in which the first and second blocks each have an elongate central axis, and are aligned on a common central axis.

33. An apparatus according to any one of claims 28 to 32 in which the first and second matching blocks have the same length along their central axis.

34. An apparatus according to any one of claims 28-33 in which the blocks have the same geometric shape.

35. An apparatus according to claim 34 in which the blocks have a front and rear end faces, the emitter and detector of each block being coupled to it at an end face.

36. An apparatus according to any one of claims 28-35 in which the blocks are mounted on rigid support means so as to ensure that they remain a fixed distance apart.

37. An apparatus according to any one of claims 28-36 in which a control means is provided adapted to cause an ultrasound signal to be generated by the emitters, and to process signals generated by the detectors.

38. An apparatus according to claim 37 in which the control means is adapted to generate an ultrasonic signal that is short enough, relative to the length of the blocks, to avoid overlap between successive detected signals.

39. An apparatus according to any one of claim 28-38 in which the diameter of a block is at least three times the diameter of the transducer attached to it.

40. An apparatus according to any one of claims 28-39 in which the ultrasonic impedance of the blocks is substantially different to that of the gas which is to be measured.

41. An apparatus according to any one of claims 28-40 in which the equation 2 C signal Cm is satisfied where tsignai is the duration of the ultrasound burst, Im is the path length of the block and Cm is the speed of propagation of the ultrasound signal in the block.

42. An apparatus according to any one of claims 28-41 in which the matching blocks are made from a polymer material of low acoustic impedance.

43. An apparatus according to any one of claims 28-42 which further comprises amplification means which have a switchable gain.

44. An apparatus according to anyone of claims 28-42 in which different amplifier circuits of different gain are provided to amplify different signals.

45. An apparatus according to any one of claims 28-44 in which a single detection circuit is used with both the first and second detectors.

46. An apparatus according to any one of claims 28-42 which further comprises an amplifier having a ramped gain.

47. An apparatus according to any one of claims 28-46 in which a common drive circuit for the two emitters is provided.

48. An apparatus according to any one of claims 28-47 adapted to measure at least one of the acoustic impedance of a fluid and the density of the fluid and the mass flow rate of the fluid.

49. An apparatus according to any one of claims 28-48 further adapted to perform a self diagnostic check to calculate the measured characteristic twice, preferably using two different sets of detected signals (which can have some detected signals in common) and compare the results to check that they are certain and within an allowable margin of each other.

50. An apparatus according to any one of claims 28-49 which comprises an add-on unit adapted to convert a volume flow meter to a mass flow meter.