There is described in EP-A2-0 381 344 a water meter of the feedback fluidic oscillator type suitable tor use as a domestic water supply meter.

As described therein , such a meter comprises a convergent entry leading into a )et-formιng orifice in the form of an elongate slit The slit directs a jet ot water into a divergent main channel which is divided by a spl itter Feedback loops trom the main channel return to opposite sides of the path ot the jet adjacent to the inlet I he j et o t f luid attaches itself to one main channel wall and then the other, repeatedly swi tching back and torth laterally at a f requency dependent upon the rate of flow. Measurements ot the frequency of the oscillations can, therefore, be used as a measure of flow rate As described in EP-A2-0 381 344, such measurement may be done by creating a magnetic field across at least one ot the channels and measuring the frequency ot the resultant potential difference generated in a direction transverse to the flow and the magnetic

Such an oscillating jet type ot fluidic tlowmeter mav be equipped with a l inearisation and temperature compensation scheme which is restricted in its potential complexity bv the limitations of a low power microprocessor based algorithm. For this and other reasons it may be required that the meterfactor response curve ot the fluidic oscillator be as l inear as possible , with variations in metert actor occurring through smooth transitions rather than rapid step changes

Particularly for a water meter, this is di f ficul t to achieve To meet domest ic w atet meter perf ormance speci f ications i he meter is i cquired to

operate over a wide range of flow conditions . I rom tul lv laminar flow at low flowrates , througn transitional conditions and into ful lv turbulent f low at high flow rates . The velocity f low prof i le across the pipe , which changes with those different flow conditions , inf luences the performance ot a fluidic oscillator flowmeter bv altering both the jet flow characteristics and the complex local ! low patterns around the internal geometry of the flowmeter. As a result, the varying f low conditions cause the fluidic oscillator to behave in a non-linear wa\ , with abrupt changes in meterfactor response through the f low range

I he use ol devices to improve How c hat acteπstics i n pipes , not least upstream of the entry to f lowmeters . is well known Such devices may be one of two kinds , commonly referred to as flow conditioners and flow straighteners .

Flow conditioners are conventionally used in fully turbulent flow conditions to restore a standard (flat) velocity t low profile following some disturbance in an inlet pipe. They are installed at a distance upstream from the flowmeter (typically , 5D to 10D from the meter inlet, where D is the diameter of the pipe) to allow a standard velocity How profile and level of turbulence to become re-establ ished before the flow enters the meter Such devices are ordinarily used with meters which are required to operate only at higher Reynolds numbers (i .e turbulent flow conditions) .

The other kind of flow conditioning device commonly used is a flow straightener. Flow straightenet arc used to remove swirl in installations where there are insuf f icient upstream lengths ot sti aight pipe followi ng a bend in the pipework However, thev do not bring the profile i nto symmetrv

For many purposes , meters are required to operate in only one flow regime, that is fully turbulent flow , and under such consistent conditions known conditioning devices can be helpful . However, to achieve the flow range required to meet domestic water metering specif ications , a fluidic water meter must be capable of operating at a range of flow conditions which span trom fully turbulent flow , through a transitional region , and into ful ly laminar flow . The known conditioning devices and the ways which thev are customarily used do not provide a solution to the problem o l i mproving the l inearity of a meter opci ating over such a large turndown range

At low tlowrates the velocitv ot the f luid is not the same at all points across the pipe because ot frictional effects at the pipe wall (leading to a parabolic velocity flow profile) and as the flow velocity increases turbulence starts to develop within the flow so that at sufficiently high Reynolds numbers flow becomes { ul ly turbulent and the flow profile becomes flat across the pipe. The turbulent to laminar flow transition at the inlet of the meter causes the f low patterns within the f luidic oscil lator to vary significantly and result in the Strouhal number changing suddenly with Reynolds number, resulting in substantial non- lineaπty .

It is an object of the present invention to enable an oscillating jet type of fiuidic flowmeter lo maintain substantial l inearity throughout a wide range of inlet flow conditions , including tu l l v developed turbulent f low and f low at low Reynolds numbci s where n is transitional oi laminai

The invention provides , in one o t its aspects , a fluidic flow meter of an oscillating jet type in which a j et ot fluid discharged I rom an elongate jet-forming orifice is caused to oscillate in a plane transversely of the plane of the orifice, the rate ol oscillation of the jet being measurable as an indication of the flowrate through the meter, the flow meter comprising flow conditioning means wh ich is positioned upstream of the jet-forming orifice and is effective under laminar and transitional inlet How conditions to f latten the velocity How profile and to introduce turbulence into the flow whereby turbulent f low conditions are simulated at the orifice .

I t has been found that by causing d isturbances in the inlet How which both flatten the velocity flow profile and introduce turbulence into the flow , resulting in a fully developed turbulent flow being simulated at the jet-forming orifice, the linearity of the meter over a wide f low range (from low Reynolds numbers upwards) can be greatly improved. Furthermore , it has also been found that such conditioning of the inlet flow can also enable the flowmeter to operate at lower flow rates than otherwise , oscillation of the jet being maintained l or longer as the flow rate is reduced ; this can give the meter an increased operational range. A further benefit of the invention can be a greater tolerance ot installation conditions , the operation of the conditioning means resulting in uniformity in the flow to the jet-forming orifice despite variations in the condition of the inlet flow from disturbances upstream.

The conditioning means is prel erabi v positioned immediatel y adjacent to . or within , an entry passage wh ich leads i nto the jet-t ormmg orifice and which in the direction ol How i s convergent in planes perpendicular to the plane of the ori fice . Such an entry passage can have the et fect ot ampl ifying the turbulence e l fcct . i ntroduced the

conditioning means , whereby a fully developed turbulent f low can be simulated at the orifice.

In one effective construction , the conditioning means lorms a plurality of openings which are spaced apart laterally of the general direction of flow and through which the inlet flow passes The openings preferably all lie in a plane perpendicular to the direction of flow, and in one preferred arrangement consist of a central array of relatively small holes encircled by relatively large holes The conditioning means may comprise a body tn the form of a multi-apertured plate which extends as a barrier across the How

In another effective construction , the conditioning means forms bluff elongate blocking elements which are spaced apart laterally of the general direction of flow and which separate elongate passages through which the inlet flow passes . Each flow passage may comprise a row of separate openings or may be in the form of a single elongate slot. In either case , the elongate blocking elements and the flow passages preferably l ie in planes parallel to the plane of the jet-forming orifice. Again , the conditioning means may comprise a body in the form of a multi-apertured plate which extends as a barrier across the f low.

There now follows a detailed description , to be read with reference to the accompanying drawings , of a f luidic f lowmeter and alternative flow conditioners which illustrate the invention by way of example.

In the accompanying drawings

Figure 1 is a view in perspective . I r om an inlet end , of a body of a f luidic flowmeter according to the i nv ention.

Figure 2 is a plan v iew of the meter body ;

Figure 3 is a meterfactor response curve for such a meter without any inlet flow conditioning ;

Figure 4 is an axial view showing the hole pattern of a first example of conditioner disc ;

Figure 5 shows a meterfactor response curve obtained util ising the conditioner disc of Figure 4;

Figure 6 shows a slot pattern of a second example ot conditioning disc; and

Figure 7 shows a meterfactor response curve obtained utilising the slot-type conditioner disc of Figure 6.

The geometry ot an oscillating jet type of fl uidic f lowmeter , for metering domestic water supplies , is shown i n Figures 1 and 2. The meter is of a feedback f luidic oscillator type , which consists of a convergent entry passage 1 which leads into an elongate jel-forming orifice 2 (the orifice being in the form ol a sl it extending in a direction perpendicular to the plan of Figure 2) , a spl itter post 3 , two divergent attachment walls 4, and two feedback channels 5. Such a construction of flowmeter is known from EP-A-0 381 344. The direction o l inlet How is shown by an arrow in Figure 2, the f low entering the entry passage 1 through a condi tioning disc 6 at its mouth , the disc being in the l orm of a multi-apertured plate .

As fluid enters the meter a jet flow is formed by the converging entry passage 1 and orifice 2. The entry passage is convergent only in planes perpendicular to the plane of the orifice. The attachment walls 4 provide forces on the jet flow through the action ot the Coanda effect and the jet becomes caused to bend towards one wall and f low along it. The wal l to which the jet first becomes attached is determined by random f luctuations in the jet which are ampl i fied by the Coanda effect.

The jet flow travels along the attachment wal l 4 , passing to one s ide of the splitter post 3 and through the meter outlet 7. As the flow is biased to one side, some of the main flow becomes entrained within the adjacent f eedback channel 5 and travels through the channel to emerge next to the orif ice outlet at an angle perpendicular to the jet flow . The f eedback f low becomes entrained in a separation bubble between the orifice and diffuser wall. As the separation bubble becomes larger the jet moves away from the attachment wall until the jet meets the splitter post

3 , whereupon the separation bubble bursts and the jet flow becomes attached to the opposite attachment wall . The feedback process is then repeated through the opposite feedback channel 5. A sustained lateral oscillation of the jet flow becomes developed , the rate of oscillation being dependent upon the How rate .

Sensing of the oscillations within the fluidic water meter is carried out using electromagnetic induction . Two permanent magnets 8 mounted within the attachment walls 4 create a magnetic f ield, orthogonal to the jet flow , which generates an e. m.f. within the l iquid itsel t . The induced voltage is detected using stainless steel electrodes 9 mounted in the top ot the fl ow chamber on each side of the spl itter post 3. Since power is generated bv the electromagnetic i nductive effect, the sensor ltscl l

requires no power. The only power required by the water meter is tor the ampl ification and processing of the generated signal .

A plot of meterfactor (volume How per pulse) against Reynolds number tor the flowmeter operating without the conditioning disc 6 is shown in Figure 3. At Reynolds numbers greater than 2500 the response is relatively flat and the device behaves in a linear manner to within

± 1 . 25% . This corresponds to a ful ly turbulent How regime in the jet.

At Reynolds numbers between 700 and 2500 there are peaks in the meterf actor characteristic. The rapid rates ol c hange metert actor with change in Reynolds number mean that the Huidic osci l lator is not behav ing in a l inear manner, with peak to peak dev iation in excess of ± 5% . At these Reynolds numbers the inlet flow cond ition is transitional.

At Reynolds numbers below 700 the meterfactor characteristic rises steadily with reducing Reynolds number. This is due to v iscous damping reducing oscillation frequency at a greater rate than the change in frequency with Reynolds number. Once Reynolds number is below a minimum value the jet flow becomes stable and oscillations cease altogether.

As the flowrate is reduced the veloci ty flow profile at the inlet becomes increasingly parabol ic and the turbulence levels within the jet flow become increasingly stable. The varying levels of jet stability and altering flow patterns within the oscillator cause variations in rate of change of oscillation frequency with flowrate. This results i n a non- linear meterfactor characteristic when the jet flow condit ion is transitional and laminar, which is undesirable from an operat iona l point ot view

The convergent entry passage 1 (immediately upstream of the jet- forming orifice 2) itself distorts the profile of the flow as the section reduces i n width . The How conditioner 6 is mounted immediately upstream of the convergence, so that the convergence wili amplify , and add to , distortion produced by the conditioner.

Conditioning of the fluidic oscillator inlet flow by means of the conditioning disc 6 increases the l inearity ot the device over the required flow range . This i s achieved by causing f lattening oi the velocity flow profile ( i .e. making it more like that of a tul ly developed turbulent How) and introducing turbulence, so resulting in a smoothing ot the contour ot the meterfactor response by reducing the severity ot the changes in jet flow condition . The transition from one How stage to another is improved by disturbing the jet flow so that turbulent conditions are sustained at lower flowrates, thus reducing the definition of the two distinct flow conditions, over a narrower range than the transition normally occurs .

Improving the meterfactor response is not as simple as merely altering the Reynolds number at which transition from one flow condition to another occurs because there are many other processes occurring within the fluidic oscillator, a l l of which arc dependent to some extent on the condition of the jet flow . These processes include the Coanda effect, the build up and separation of boundary layers along the attachment walls 4 and internal faces of the flow chamber, the velocity flow profile of flow within the feedback channels 5, the entrainment ot feedback flow and the stabi l i ty of the jet. The inlet flow condition must be tuned to match and enhance the operational characteristics of the Huidic osci llator.

The use of the conditioning disc 6 generates a disturbance at the meter inlet which is ampli fied as it travels through the convergent entry

passage 1 . This introduces instability into the |et flow where the flow would ordinarily be under fully laminar conditions Instabilities of the jet flow are amplified by the Coanda et fect and help to simulate the flow conditions under which the meter would be operating during the turbulent How regime Maintaining turbulence in the jet f low during transitional and laminar inlet flow conditions reduces the variation in Strouhal number with Reynolds number across the range ot the How transition

The use of the conditioning d isc h dramatical l v improves the l inearity ot the f lowmeter, as can be seen l i om a comparison of

F igures 3 and 7 When fitted with the How conditioner the l inearity of the f luidic oscillator is such that the metertaciυr response mav be used by a simple l inearity and temperature compensation scheme

A first conditioning disc 6 (Figure 4) comprises an array of smaller holes 10 within a surround of larger holes 1 1 . The relative position and inclination of the jet-forming orifice 2 is shown in broken line . The disc is in the form of an apertured plate of 2mm thickness and 22mm diameter having eight 4mm diameter holes 1 1 on a 7mm radius , eight 1 mm diameter holes 10 on a 3mm radius , si xteen 1 mm diameter holes 10 on a 4.25mm radius, and a 1 mm diameter centre hole 1 0.

The meterfactor response (Figure 5) produced by use of the first disc has very smooth changes trom laminar to transitional flow and trom transitional to turbulent jet flow . The magnitude of the minimum peak during the change trom laminar to trans itional conditions is relatively small and it a centre l ine for the metert actor response curve is carefully selected the response is within ± 2% over most of the range , except at very low f lowrates . The shape ot the response is such that a best straight

line model of the curve could satisfactoπlv be used by a simple linearisation and temperature compensation scheme

Conditioner discs with increasing hole si e have been tested. The results ot the hole size tests indicated that smallei holes do improve the transition from laminar to turbulent How conditions but the geometry contour of the hole pattern is more significant to overall meter linearity

Another form ot conditioning disc ϋ which has been tound to produce a meterfactor response with a lelativclv smooth transitional How region comprises a hole pattern forming lateiallv - spaced rows ot separate holes extending parallel to the jet-torming orifice 6 Removal of the material trom between the holes lorms slots 12 as seen in Figure 6 This second conditioner was found to produce an even better result than the first disc (Figure 4). It also gave a lower pressure loss and was less susceptible to particle blockage.

The geometry of the slot-type conditioning disc 6 is shown in Figure 6, having three parallel slots 12 ot 3.2mm width separated by elongate bluff blocking regions 13 ot 32mm width <\gaιn the disc is of 22mm diameter and 2mm thickness.

The meterfactor response plot tor use ot this conditioner is shown in Figure 7. The curve is very flat, with very smooth transitions between the tully laminar and turbulent flow conditions The response is linear over a very wide flow range, from Reynolds numbci 550 upwards, requiring linearity compensation only at verv low flowrates where the meterfactor rises with reducing tlowrate due to viscous damping effects which reduce oscillation frequency

The geometry of the slot-type conditioning disc o is such that vortices are shed from the bluff blocking regions 13 between the slots 12 in the form ot vortices. Each vortex navels through the convergent entry passage, creating severe disturbances at the oπlice 2 ^' Ihe disc affects the profile of the inlet flow in planes perpendicular to the orifice 2 to a greater extent than in planes parallel lo the orifice However, the convergence of the entry passage 1 is itself two-dimensional, because the convergence occurs onlv in planes perpendicular to the orifice The convei gence, therefore, distorts the velocity flow profile in the perpendicular planes, thus amplifying the disturbances caused bv the conditioner

Using any of the torms of conditioning disc 6 described, the transition from one flow stage to another is improved by disturbing the jet flow so that turbulent conditions are sustained at lower flowrates The discs are able to perform at low Reynolds numbers to alter a laminar (semi-parabolic) velocity flow profile so that it is more like a turbulent (flat) velocity profile and to increase turbulence levels so that the overall How condition is like that of fullv developed turbulent flow at higher Reynolds numbers The flattening of the velocity flow profile is achieved by contouring the hole patter of the conditioning disc (Figure 4 - smaller holes 10 in the centre region and larger holes 11 in the outer region) or by suitable selection of the number and spacing ot slots 12 and blocking regions 13 (Figure 6). Correct turbulence levels are achieved by selecting hole si/e or blocking region width, lespectively Turbulent mixing is achieved by shedding vortices and turbulent eddies Flow visualisation of the oscillator inlet flow using the conditioners show turbulent eddies shed at a high rate producing vciv disturbed How stream lines through the

entry passage 1 . The vortices shed from the conditioning device produce much greater mixing of the inlet flow even al low Reynolds numbers .