GB2246629A - Measuring fluid-borne vibrations in pipes - Google Patents

Abstract

Determination of fluid borne vibrations in pipes non-intrusively involving measuring the amplitude of pipe vibration as a function of the electrical response of piezoelectric material deformed by the vibration. Conveniently the piezoelectric material is provided as a wire which is wound around the pipe diameter an integral number of times for accurate determination of axisymmetric n=0 vibrational modes. A suitable piezoelectric material is the biaxially orientated semicrystalline polyvinylidene difluoride polymer. In order to enable accurate measurement the piezoelectric material is held in place against the pipe with double-sided adhesive tape. The technique can be used to measure vibrations in oil, water, gas and motor car pipe installations as well as others. <IMAGE>

Description

FLUID FLUID BORNE VIBRATIONS IN PIPES The e present invention relates to a method for measuring fluid-borne vibrations in pipes and specifically the "n=O" axisyitirietric or breathing type mode of vibrations.

Pipewcrks can form one of the more important vibration and noise sources in equipment through the transportation of unwanted sound energy frommachinery via fluid-filled pipes. For effective noise or vibration control of piping systems, it is of prime importance to be able to predict the distribution of energy of the shell-fluid system for various configurations. In particular, it is desirable to know whether the majority of the energy of vibration is located in the pipe wall or in the contained fluid. This depends on the type of excitation of the piping system and the physical parameters of the shell and the contained fluid.

In this regard it has been found that: (i) at low frequencies, the energy is concentrated in the shell wall for structural excitation. For acoustic (or fluiJ) excitation, however, the energy is predominantly in the fluid at low frequencies; (ii) at higher frequencies, the energy distribution gets more convoluted and the energy may be either in the fluid field or in the shell wall.

Pressure measurements inside fluid-filled pipes are therefore of importance in calculating the fluid borne power transmitted in practical pipe installations.

Techniques for measuring fluid pressures in pipes are conventionally intrusive thus requiring a hole to be made in a pipe for the insertion of a pressure measuring gauge. This represents an impractical approach which renders the pipe unusable if the pressure measuring gauge is removed and the hole not appropriately plugged. Furthermore whilst hydrophones and conventional pressure transducers have been used for pressure measurement they suffer in different extents to flow noise.

The inventor has discovered that for a vertical perspex pipe filled with fluid the surface strain of the pipe wall is proportional to the internal fluid pressure when the fluid is excited. This is found to be true at least below the ring frequency of the fluid-fiiled pipe.

Consequently there is a requirement for a non-intrusive and easily adopted method for measuring fluid-borne vibrations in pipes by determining pipe wall surface strain.

Accordingly there is provided a method for the non-intrusive measurement of fluid-borne vibrations in a pipe which comprises measuring the amplitude of pipe vibration as a function of the electrical response of piezoelectric material deformed by the vibration.

As the piezoelectric material is stressed it acts as a transducer with the distortion which it incurrs being converted to an electric signal proportional to that degree of stressing.

The present invention has the advantages of being simpler non-intrusive and it does not suffer from fluid flow noise to the same extent as conventional pressure sensing devices. Conveniently the piezoelectric material is provided as a wire.

Whilst the piezoelectric material may be used to measure pipe stresses and hence vibrations in various configurations it is necessary when measuring axismmmetric n=O or breathing type modes that the material is wound around the diameter of pipe a number of times. Whilst the amplitude of reading and hence the sensitivity increases with the number of turns of piezoelectric material around the pipe diameter it is found that if a suitable piezoelectric material is used such as the biaxially oriented semicrystalline polyvinylidenedifluoride (PVDF) polymer that the number of turns necessary to achieve sensible results can be as low as 1.Consequently where PVDF is used there is in practice little reason for using more than 5 turns and 2 or 3 turns for many cases will yield good readings with economy of piezoelectric matrerial used.

This is probably due to the fact that PVDF polymers tend to exhibit a stronger piezoelectric effect that other piezoelectric polymers used at present and are therefore particularly useful as pressure transducers. Furthermore PVDF polymer is manufactured in the form of a thin, coaxial wire with an outer protective coating.

It should be noted that for accuracy of readings, especially when detering axisymlretric n=O vibration modes, that the turns of piezoelectric material applied should be integral. If there are more or less than integral turns of material about a pipes circumference then this will give rise to readings which include circumferential waves of higher (asymmetric) orders. Where an integral number of turns of piezoelectric material is used any overall stretching of the pipes circumference which arises mainly from axisymmetric n=O waves is detected.However as circumferential waves of higher (asymmetric) orders will possess both positive and negative lobes the overall effect is cancellation and the method of the present invention may therefore be made essentially insensitive to these modes of vibration.

To enable accurate measurements of pipe vibration it is necessary the piezoelectric materials deformation be representative of the pipe vibration.

Accordingly it is necessary to ensure that the piezoelectric material is properly fixed to the pipe with no looseness whilst at the same time not being over stressed.

To hold the piezoelectric material in place on a pipe it has been found preferable to use an adhesive tape. This is especially so for circumferential arrangements of piezoelectric wire used in measurement of axisymnetric vibrations. It has also been found that aluminium or a similar type of adhesive tape has the further advantage of acting as an electromagnetic screen and therefore acts in the prevention of spurious induced currents in the piezoelectric material. A yet further advantage is that, in the case of PVDF piezoelectric wire at least, the adhesive tape can be removed without damage to the piezoelectric wire allowing its reuse.

Conveniently the degree of deformation of the piezoelectric material is determined by the extent of electric charge generated. Conveniantly a charge amplified is used although the measurement of voltage is also possible. The deformations in piezoelectric wire may be calibrated using a sensitivity formula relating the radial wall displacement W to the internal fluid pressure P for a thin-walled cylindrical shell containing fluid.

As the invention in the form required for n=O structual wave measurement also shows some small sensitivity of the order of 4% for liquid compressional waves it is possible for an empty pipe to make measurements in relation to compressional waves.

The present invention gives good agreement with hydrophone measurements to 1000Hz. It also provides useable assessments of vibration up to the breathing or ring frequency. At higher frequencies other wave motions are found to become significant.

It is envisaged that the present invention can find utility in various industries especially those relating to oil, water, gas and motor car exhausts. It may also be adapted for high temperature applications.

The invention will now be described by way of example only and with reference to the accompanying Drawings of which: Figure 1 shows a sectioned fluid containing pipe with a wire of piezoelectric polyvinylidene-difluoride (PVDF) polymer wrapped an integral number of times around the pipes circumference; Figure 2 shows (a) the mode shapes for the n=0 "breathing" and (b) the n=l "structural bending" mode; Figure 3 shows an experimental test rig designed and constructed to investigate PVDF wire; Figure 4 shows a dispersion plot for the water filled perspex pipe of Figure 3; Figures 5 (i-v) show the average of the (i) transfer, (ii) coherence, (iii) phase, (iv) hydrophone and (v) PVDF output signals between the range 01000Hz; Figures 6(a) and 6(b) compare hydrophone measurments with those using PVDF wires over the frequency ranges of 0-2000Hz respectively.

With reference to Figures 1 to 6 measurement of structural vibrations in a pipe (1) as shown in Figure 1 caused by fluid (2) are determined ty wire (3) made of PVDF polymer which is wound around peripheral circumference of an integral number of turns. The vibrations cause the wire to be deformed and a charge is developed which is proportional to the extent of wire deformation.

Figures 2(a) and (b) show two types of vibration which can occur in a fluid filled pipe. Figure 2(a) shows the n=O breathing type mode. PVDF wire wound around the peripheral circumference of a pipe undergoing the breathing type mode of vibration will be deformed and a charge developed in response to the extent of deformation.

Figure 2b demonstrates the n=l structural bending mode. For a PVDF wire wound around the circumference of a pipe undergoing this type of vibration there is no net charge developed by deformation of the pipe and hence PVDF wire wound around the pipe is insensitive to the structural bending mode n=l.

An experimental apparatus set-up to as shown in figure 3 consists of a vertically hung water-filled (transparent) perspex pipe 3 (1). The pipe has a radius of 69nix, a thickness of 6mm and a height of 1.9m. The material properties are given in Table 1.

1 Table 1 Table 1 Material Young's Modulus Poisson's Density Free Wave (N/m2) ratio (kg/m3) speed (sus) perspex 2.18 x 109 0.40 1100 1536 water - - 1000 1500 A 3 (2) ruler is placed through apperatures near the top of the pipe which are above the water level and such that it cLosses the centre point of the pipe 3(1). Suspended into the water of the pipe from the centre point by the ruler is a hydrophone 3(3).

Towards the lower end of the pipe and at the same level as the hydrophone there is wound around the pipes outer circumference an integral number of times a PVDF wire 3 (4) which is held in place by aluminium coated adhesive tape 3 (5).

Below this and positioned inside the pipe is yet more PVDF wire 3(6).

Aluminium rings 3(7) positioned at the base of the pipe hold a rubber diaphragm 3 (8) in place across the base of the pipe to prevent water leakage.

A piston 3 (9) made of a Honeycomb carbon fibre material has its head in coaxial alignment with the pipe and in contact with the flexible seal. The piston itself is displaced away on electromagnet assembly 3(10) by virtue of a flexible rubber ring 3(11) which fits between the electromagnet assembly and the base of the piston.

The water inside the pipe was excited by the piston at the base of the pipe.

The Honeycomb carbon fibre construction of the piston ensured a high resonant frequency for producing place effecture waves up the pipe.

The apparatus's construction enabled measurements of the following parameters: (i) The transfer function between the pressure using the PVDF wire around the pipe wall and the hydrophone pressure sensor in the fluid with the PVDF wire at the same vertical level; (ii) The axial wavenumber in the fluid; (iii) The radial pressure distribution in the fluid.

The PVDF wire was attached to the pipe using the following procedure. A single layer of double sided self adhesive tape (not shown) was wrapped around the pipe. The PVDF was wrapped an exact integral number of turns around the pipe on top of the self adhesive tape. The end of the wire was cut to the correct length and the end sealed with silicon rubber. This prevented moisture causing an electrical short circuit between the electrodes (not shown). Finally self adhesive aluminium tape was firmly wrapped around the PVDF wire, to totally eliminate any flapping of the wire, which would give erronious signals.

The extension (extn) in the PVDF wire around the pipe wall is related to the radial wall displacement by: w = extn a 2haN where w is the radial wall displacement a is the pipe radius, and N is the integral number of turns of PVDF wire The pressure sensitivity is given by:

where E is Young's modulus h is the shell wall thickness and U is Poisson's ratio The output charge qp from the PVDF wire is related to the extension by: = = extn.s where s is the charge sensitivity of the PVDF wire in (pc/extn). Therefore the pressure sensitivity is given by;

Figure 4 demonstrates that for a dispersion plot of the water filled perspex pipe that the experimental curve approximates to the theoretical curve at low frequencies.

The transfer function between pressure using the PVDF wire and the hydrophone pressure sensors were taken using a random signal generator.

Figure 5 shows the average of the (i) transfer, (ii) coherence, (iii) phase, (iv) hydrophone and (v) PVDF output signals between the range 0-1000Hz. The comparative pressure signals in Figures 5 (iv) and (v) are almost identical demonstrating good agreement between hydrophone and PVDF measurement. The division of these two plots in figure 5 (i) takes a value close to unity once a calibration factor has been applied. There are deviations from unity, which occur at positions of low pressure, and correspondingly low coherency between the two signals. Figure 5 (iii) shows that the signals of the PVDF and hydrophone are in phase.

Figure 6a shows a comparative plot of frequency response 0-1000Hz for a hydrophone and PVDF wire and demonstrates over the frequency range good agreement.

Similarly Figure 6b shows that the frequency response for hydrophone and PVDF sensors over 0-2000Hz demonstrate good agreement between the sensors.

Claims (14)

CMS

1. Amethod for the non-intrusive measurement of fluid-borne vibrations in a pipe which comprises measuring the amplitude of pipe vibration as a function of the electrical response of piezoelectric material deformed by the vibration.

2. Amethod for the non-intrusive measurement of fluid-borne vibrations in a pipe as claimed in claim 1 wherein the deformation occurring in the piezoelectric material is converted to an electric signal proportional to the degree of stressing.

3. Amethod for the non-intrusive measurement of fluidborne vibrations in a pipe as claimed in claim 1 or 2 wherein the piezoelectric material is provided as a wire.

4. Amethod for the non-intrusive measurement of fluidborne vibrations in a pipe as claimed in ary one of the previous claims wherein the piezoelectric material is wound around the diameter of pipe a number of times.

5. Amethod for the non-intrusive measurement of fluid-borne vibrations in a pipe as claimed in claim 4 wherein the sensitivity increases with the number of turns of piezoelectric material around the pipe diameter.

6. Amethod for the non-intrusive measurement of fluid-borne vibrations in a pipe as claimed in claim 4 or 5 wherein there are 2 or 3 turns of piezoelectric material about the pipe diameter.

7. A method for the non-intrusive measurement of fluid-borne vibrations in a pipe as claimed in in one of the previous claims wherein the piezoelectric material is biaxially oriented semicrystalline polyvinylidenedifluoride (PVDF) polymer.

8. Amethod for the non-intrusive measurement of fluidborne vibrations in a pipe as claimed in any one of the claims 4 to 7 wherein the number of turns of piezoelectric material applied is integral.

9. A method for the non-intrusive measurement of fluibborne vibrations in a pipe as claimed in any one of the previous claims wherein the piezoelectric material is fixed to the pipe with no looseness present whilst at the same time not being over stressed.

10. A method for the non-intrusive measurement of fluid-borne vibrations in a pipe as claimed in claim 9 wherein the piezoelectric material is held in place on the pipe with an adhesive tape.

11. A method for the non-intrusive measurement of fluibborne vibrations in a pipe as claimed in claim 11 wherein the adhesive tape is an aluminium or similar type of adhesive tape which is capable of acting as an electromagnetic screen.

12. A method for the non-intrusive measurement of fluid-borne vibrations in a pipe as claimed in any one of the previous claims wherein the degree of deformation of the piezoelectric material is determined by the electric charge generated.

13. A method for the non-intrusive measurement of fluid-borne vibrations in a pipe as claimed in claim 12 wherein a charge amplifier is used in measuring the charge generated.

14. A method for the non-intrusive measurement of fluibborne vibrations in a pipe specifically as herein described and with reference to the accaripanying drawings.