(19)
(11)EP 2 052 223 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
27.11.2019 Bulletin 2019/48

(21)Application number: 07836452.8

(22)Date of filing:  02.08.2007
(51)International Patent Classification (IPC): 
G01F 25/00(2006.01)
G01F 15/00(2006.01)
(86)International application number:
PCT/US2007/017301
(87)International publication number:
WO 2008/021017 (21.02.2008 Gazette  2008/08)

(54)

FLOW MEASUREMENT DIAGNOSTICS

FLUSSMESSUNGSDIAGNOSEN

DIAGNOSTIC DE MESURE D'ÉCOULEMENT


(84)Designated Contracting States:
CH DE LI

(30)Priority: 14.08.2006 US 503878

(43)Date of publication of application:
29.04.2009 Bulletin 2009/18

(73)Proprietor: Rosemount Inc.
Eden Prairie, MN 55344 (US)

(72)Inventor:
  • WEHRS, David, L.
    Eden Prairie, MN 55346 (US)

(74)Representative: Bohnenberger, Johannes et al
Meissner Bolte Patentanwälte Rechtsanwälte Partnerschaft mbB Widenmayerstraße 47
80538 München
80538 München (DE)


(56)References cited: : 
EP-A- 0 827 096
US-A1- 2005 011 278
WO-A-00/50851
US-A1- 2005 284 237
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    BACKGROUND OF THE INVENTION



    [0001] The present invention relates to fluid process control and monitoring systems. In particular, the present invention relates to diagnostics for such systems.

    [0002] Fluid flow meters are used in industrial process control and monitoring environments to measure fluid flow and provide flow signals for flow indicators and controllers. Inferential flow meters measure fluid flow in a pipe by measuring a pressure drop near a discontinuity within the pipe. The discontinuity (primary element) can be an orifice, a nozzle, a venturi, a pitot tube, a vortex shedding bar, a target or even a simple bend in the pipe. Flow around the discontinuity causes both a pressure drop and increased turbulence. The pressure drop is sensed by a pressure transmitter (secondary element) placed outside the pipe and connected by impulse lines or impulse passageways to the fluid in the pipe. Reliability depends on maintaining a correct calibration. Erosion or buildup of solids on the primary element can change the calibration. Impulse lines can become plugged over time which isolates the pressure transmitter from the process such that the transmitter is no longer able to track the pressure and adversely affects the operation of the transmitter.

    [0003] Disassembly and inspection of the impulse lines is one method used to detect and correct plugging of lines. Another known method for detecting plugging is to periodically add a "check pulse" to the measurement signal from a pressure transmitter. This check pulse causes a control system connected to the transmitter to disturb the flow. If the pressure transmitter fails to accurately sense the flow disturbance, an alarm signal is generated indicating line plugging. Another known method for detecting plugging is sensing of both static and differential pressures. If there is inadequate correlation between oscillations in the static and differential pressures, then an alarm signal is generated indicating line plugging. Still another known method for detecting line plugging is to sense static pressures and pass them through high pass and low pass filters. Noise signals obtained from the filters are compared to a threshold, and if variance in the noise is less than the threshold, then an alarm signal indicates that the line is blocked.

    [0004] WO/0050851 is considered the closest prior art and this describes a fluid flow meter which diagnoses the condition of its primary element or impulse lines connecting to a differential pressure sensor. A difference circuit coupled to the differential pressure sensor has a difference output representing the sensed differential pressure minus a moving average. A calculate circuit receives the difference output and calculates a trained output of historical data obtained during an initial training time. The calculate circuit also calculates a monitor output of current data obtained during monitoring or normal operation of the fluid flow meter. A diagnostic circuit receives the trained output and the monitor output and generates a diagnostic output indicating a current condition of the primary element and impulse lines.

    SUMMARY



    [0005] A flow meter for measuring flow of a process fluid includes a sensor configured to provide a sensor output signal related to flow of the process fluid. Circuitry is configured to determine a statistical parameter related to sensor output signal. A diagnostic output is provided based upon the determined statistical parameter and the sensor output signal. Such a flow meter of the invention is described in independent claim 1. A method of diagnosing operation of a flow meter is described in independent claim 12. Preferred embodiments of the invention are given in the dependent claims.

    BRIEF DESCRIPTION OF THE DRAWINGS



    [0006] 

    FIG. 1 is a simplified diagram of a process control or monitoring system.

    FIG. 2 is a simplified block diagram showing a cross-sectional view of the transmitter shown in FIG. 1.

    FIG. 3 is a graph of standard deviation of a differential pressure signal versus a differential pressure signal.

    FIG. 4 is a graph of standard deviation of a differential pressure signal versus a differential pressure signal.


    DETAILED DESCRIPTION



    [0007] FIG. 1 is a simplified diagram of a process control or monitoring system 10 in which process piping 12 carries a flow 14 of process fluid. A flow transmitter 16 is configured to sense the flow 14 and provide an output related to the flow 14. In the example shown in FIG. 1, the output is provided to control room 18 over a two wire process control loop 20. Loop 20 can operate in accordance with any protocol. Example standard protocols includes 4-20 mA signals, the HART(R) communication protocol or Fieldbus protocols. However, the present invention is not limited to such communication techniques, including wireless techniques, and can even operate in a stand alone device.

    [0008] Process variable transmitter 16 senses flow using any appropriate technique. Example techniques include measuring a differential pressure across a restriction in the pipe 12, magnetic based technologies, pitot tubes, vibrating sensors, etc.

    [0009] In accordance with the invention, transmitter 16 provides a diagnostic output based upon the flow 14 and noise in a sensor output signal. In one specific configuration, a correlation between standard deviation of the sensor signal and the sensor signal is used for diagnostics. The diagnostic output can be used internally to transmitter 16, or can be provided externally, for example by transmission over loop 20.

    [0010] Figure 2 is a simplified block diagram of transmitter 16 shown in Figure 1. In the diagram of Figure 2, transmitter 16 is illustrated as including a differential pressure sensor. In such a configuration, a difference in pressures P1 and P2 is applied to the transmitter 16 and can be used to infer the rate of flow 14 through process piping 12 shown in Figure 1. An orifice plate 47 is placed in the process piping and includes an orifice 49 formed therein. The orifice plate 47 creates a differential pressure which is the difference between pressures P1 and P2 as the flow 14 is forced to flow through the restricted orifice 49. The invention is not limited to such a flow measurement technique.

    [0011] Pressures P1 and P2 are applied to isolation diaphragms 52 and 54 through process coupling 53 and 55, respectively. The isolation diaphragms 52 and 54 form respective cavities 56 and 58 in the transmitter 16. The cavities 56 and 58 are isolated from the process fluid and contain an isolation fluid. Small capillary tubes 60 and 62 lead from respective cavities 56 and 58 to differential pressure sensor 50. The capillary tubes 60 and 62 transfer the pressure from pressure applied by P1 and P2 to the differential pressure sensor 50 while isolating the pressure sensor 50 from the process fluid.

    [0012] Differential pressure sensor 50 can operate in accordance with any technology. One example technology uses a deflectable diaphragm which forms a variable capacitance. The pressure sensor 50 couples to transmitter circuit 70 and provides a sensor output to transmitter circuit 70. Transmitter circuit 70 can comprise any type of circuit configuration. Typically, a transmitter circuit 70 includes a microprocessor or other digital controller along with a memory 72. The memory 72 stores program instructions, configuration information, temporary variables and the like. Transmitter circuit 70 includes measurement circuit 74 which provides an output related to flow of the process fluid. The flow can be determined using any appropriate technique. Diagnostic circuit 76 is also implemented in transmitter circuit 70. The diagnostic circuit 76 includes circuitry 77 to determine a statistical parameter related to the sensor output signal 78. The statistical parameter can be indicative of noise in the sensor output signal 78 and, in one specific embodiment, the statistical parameter comprises standard deviation. Output circuitry 86 receives the flow signal and provides an output on two wire process control loop 20 in accordance with a desired signal and protocol.

    [0013] In the configuration shown in Figure 2, the diagnostic circuit 76 receives the flow signal from measurement circuit 74 along with the sensor output 78 from sensor 50. As used herein, the flow signal and sensor signal can be processed prior to receipt by diagnostic circuit 76. Based upon the flow signal 80 and the sensor output 78, diagnostic circuit 76 provides a diagnostic output 82. The diagnostic output 82 can be used locally by transmitter 16 and can be provided externally, for example by transmission over two wire process control loop 20. The diagnostic output is indicative of a condition of one or more components of transmitter 16, or other components coupled to the industrial process 10 shown in Figure 1. For example, the diagnostic output 82 can provide an indication that one of the process couplings 53, 55 is becoming plugged. This information can be used alone, or in combination with other line plugging diagnostics to identify the plugging of process couplings 53, 55. Another example diagnostic is indicative of wear in the restrictive orifice plate used to create the differential pressure. This wear can cause edge corrosion in the plate resulting in inaccurate flow measurements. Similarly, loss of the isolation fill fluid carried in capillaries 60, 62 and cavities 56, 58 can be detected. The diagnostic output 82 can also be indicative of the composition of the process fluid flowing through pipe 12 shown in Figure 1. Change can be due to different fluids, changes in the components, such as contaminants in the fluid such as due to aeration or the addition of solids, etc.

    [0014] The particular algorithm used to correlate the sensor signal and the flow signal with the diagnostic output can be chosen as appropriate. In one example embodiment, the diagnostic circuitry 76 identifies noise in the sensor output signal and provides the diagnostic output based upon this noise and the measured flow rate. In particular, a relationship between the noise level in the sensor signal and the flow rate during steady state flow conditions is used to provide the diagnostic output 82. The standard deviation is indicative of the noise in the sensor output signal. Figure 3 is a graph of the standard deviation of a differential pressure sensor output versus flow. In this example, the process fluid comprises water at approximately room temperature.

    [0015] Figure 4 is a similar graph using a different size orifice plate to create the differential pressure. In the case of both Figure 3 and Figure 4, there is an approximately linear relationship between the standard deviation of the differential pressure signal and the differential pressure signal itself. For example, as the differential pressure increases by 20% due to a change in flow rate, the standard deviation also increases by approximately 20%. Note that there is some variation in the relationship between standard deviation and differential pressure over the range of flow rates, particularly at low flow rates as the baseline noise becomes a more significant part of the total noise. However, typically the increase or decrease of this standard deviation is less than the increase or decrease in the mean differential pressure value. In some applications, the relationship between standard deviation and the sensor signal can be characterized by a more complex relationship, such as a polynomial with additional terms.

    [0016] The particular relationship between standard deviation of the sensor signal and the sensor signal itself can be determined for the particular flow measurement technology. Once the relationship is determined for nominal operating conditions, the relationship can be monitored to identify a particular diagnostic condition.

    [0017] In one particular embodiment, the signal 80 provided to the diagnostic circuitry 76 is the sensor output signal 78, or is directly related to the sensor output signal 78. The noise determination can be made using any technique and is not limited to the standard deviation discussed above. However, in one specific embodiment, the noise is determined based upon the standard deviation. In another example specific embodiment, the noise signal is compared to the mean of the process variable. Such a configuration is advantageous because the calculations are relatively simple. Typically, the noise of interest is in the range of from approximately 1 Hz to approximately 30 Hz. Noise signals lower than this can be difficult to identify due to slow changes in the process. In a further example technique, the diagnostic circuitry 76 includes a high pass filter 79 to filter the process variable prior to calculation of the standard deviation. Any appropriate technique can be used to determine noise in the sensor signal. Examples include a digital band pass filter that provides an RMS value of the noise that can then be related to the mean sensor signal. Another example technique to determine noise includes using a Fast Fourier Transform (FFT), wavelets, etc.

    [0018] During operation, a nominal relationship between the noise signal and the sensor output signal can be stored in memory. This relationship can be determined empirically by observing operation of the transmitter 16 during normal conditions, or can be determined during manufacture and stored in memory 72. Once the transmitter 16 is placed into normal operation, the diagnostic circuitry 76 calculates the noise in the sensor signal and uses this information along with the sensor output signal and performs a comparison with the relationship stored in memory 72. The relationship stored in memory 72 can be, for example, a polynomial equation, a table of data points, etc. This information can also be updated over process control loop 20 or through other programming techniques.

    [0019] Based upon the diagnostic output 82, the diagnostic information can be transmitted over process control loop 20. In another example, the diagnostic output Is used to provide a local alarm, correct measurements in the process variable output from transmitter 16, used to schedule maintenance, etc. The diagnostic output can also be used in conjunction with other diagnostic techniques to provide more accurate diagnostic information.

    [0020] Although the present invention has been described with reference to preferred embodiments, workers skilled in the art will recognize that changes may be made in form and detail without departing from the scope of the invention as defined in the attached claims. Any type of flow measurement technology can be used in conjunction with transmitter 16 including magnetic flow meters, vortex or corioles based meters, etc. For example, sensor 50 can comprise a magnetic sensor for a magnetic flow meter, a vibration based sensor, etc.


    Claims

    1. A flow meter for measuring flow (14) of a process fluid, comprising:

    a sensor (50) configured to provide a sensor output signal (78) related to flow (14) of the process fluid;

    circuitry (77) configured to determine a statistical parameter related to the sensor output signal (78); and

    diagnostic circuitry (76) configured to provide a diagnostic output (82) based upon the current value of the determined statistical parameter and the current value of the sensor output signal (78), wherein the apparatus includes a memory (72) configured to store a nominal relationship between the determined statistical parameter and the sensor output signal (78), the nominal relationship having been determined empirically by observing operation of the flow meter during normal conditions over a range of process fluid flow values, wherein the diagnostic output (82) is based upon a comparison between the current value of the statistical parameter and the current value of the sensor output signal (78) with the stored nominal relationship.


     
    2. The apparatus of claim 1, wherein the diagnostic circuitry determines noise in the sensor output signal.
     
    3. The apparatus of claim 1, wherein the diagnostic circuitry determines noise during substantially steady state flow conditions.
     
    4. The apparatus of claim 1, wherein the sensor comprises one of:

    a differential pressure sensor;

    a magnetic sensor or a magnetic flow meter, or a vibration sensor of a vortex or corioles flow meter.


     
    5. The apparatus of claim 1, wherein the nominal relationship is determined either during manufacture of the flow meter, or during operation of the flow meter.
     
    6. The apparatus of claim 1, including transmitting the sensor output over a process control loop.
     
    7. The apparatus of claim 1, wherein the sensor output is related to clogging of a process coupling.
     
    8. The apparatus of claim 1, wherein the diagnostic output is related to either a condition of an orifice plate, or to a composition of the process fluid.
     
    9. The apparatus of claim 1, wherein the sensor output is indicative of a condition of a process control loop.
     
    10. The apparatus of claim 1, wherein the diagnostic circuitry determines noise in the sensor output signal using a band pass filter, or a high pass filter.
     
    11. The apparatus of claim 1, wherein the diagnostic circuitry determines noise in the sensor output signal by either determining a root mean squared (RMS) value of the sensor output signal, or by using a Fourier transform of the sensor signal.
     
    12. A method of diagnosing operation of a flow meter of the type used to measure flow (14) of process fluid, the method comprising:

    obtaining a signal (78) related to flow (14) of the process fluid;

    determining a statistical parameter related to the signal (78) related to flow (14) of the process fluid; and

    providing a diagnostic output which is based upon the comparison of the current value of the signal (78) related to flow (14) of the process fluid and the current value of the determined statistical parameter with a stored nominal relationship, wherein the nominal relationship is a relationship between the determined statistical parameter and the signal (78) related to the flow (14) of the process fluid, the nominal relationship having been determined empirically by observing operation of the flow meter during normal conditions over a range of process fluid flow values.


     
    13. The method of claim 12, wherein the statistical parameter is related to noise in the sensor output signal.
     
    14. The method of claim 12, wherein the signal is related to a differential pressure sensor.
     


    Ansprüche

    1. Durchflussmessgerät zum Messen des Durchflusses (14) eines Prozessfluids, umfassend:

    einen Sensor (50), der dazu ausgelegt ist, ein auf den Durchfluss (14) des Prozessfluids bezogenes Sensorausgangssignal (78) bereitzustellen;

    eine Schaltung (77), die dazu ausgelegt ist, einen auf das Sensorausgangssignal (78) bezogenen statistischen Parameter zu bestimmen; und

    eine Diagnoseschaltung (76), die dazu ausgelegt ist, einen Diagnoseausgang (82) auf Grundlage des aktuellen Werts des bestimmten statistischen Parameters und des aktuellen Werts des Sensorausgangssignals (78) bereitzustellen, wobei die Vorrichtung einen Speicher (72) enthält, der dazu ausgelegt ist, ein nominales Verhältnis zwischen dem bestimmten statistischen Parameter und dem Sensorausgangssignal (78) zu speichern, wobei das nominale Verhältnis empirisch bestimmt wurde, indem ein Funktionsablauf des Durchflussmessgeräts während normaler Bedingungen über einen Bereich von Prozessfluiddurchflusswerten beobachtet wurde, wobei der Diagnoseausgang (82) auf einem Vergleich zwischen dem aktuellen Wert des statistischen Parameters und dem aktuellen Wert des Sensorausgangssignals (78) mit dem gespeicherten nominalen Verhältnis beruht.


     
    2. Vorrichtung nach Anspruch 1, wobei die Diagnoseschaltung ein Rauschen in dem Sensorausgangssignal bestimmt.
     
    3. Vorrichtung nach Anspruch 1, wobei die Diagnoseschaltung ein Rauschen im Wesentlichen während Durchflussgleichgewichtsbedingungen bestimmt.
     
    4. Vorrichtung nach Anspruch 1, wobei der Sensor eines der folgenden Elemente umfasst:

    einen Differenzdrucksensor;

    einen Magnetsensor oder ein Magnetdurchflussmessgerät, oder einen Schwingungssensor eines Wirbel- oder Coriolis-Durchflussmessgeräts.


     
    5. Vorrichtung nach Anspruch 1, wobei das nominale Verhältnis entweder während der Herstellung des Durchflussmessgeräts oder während des Betriebs des Durchflussmessgeräts bestimmt wird.
     
    6. Vorrichtung nach Anspruch 1, beinhaltend, den Sensorausgang über eine Prozesssteuerschleife zu übertragen.
     
    7. Vorrichtung nach Anspruch 1, wobei der Sensorausgang auf eine Verstopfung einer Prozesskupplung bezogen ist.
     
    8. Vorrichtung nach Anspruch 1, wobei der Diagnoseausgang entweder auf einen Zustand einer Messblende oder auf eine Zusammensetzung des Prozessfluids bezogen ist.
     
    9. Vorrichtung nach Anspruch 1, wobei der Sensorausgang für einen Zustand einer Prozesssteuerschleife bezeichnend ist.
     
    10. Vorrichtung nach Anspruch 1, wobei die Diagnoseschaltung ein Rauschen im Sensorausgangssignal unter Verwendung eines Bandpassfilters oder eines Hochpassfilters bestimmt.
     
    11. Vorrichtung nach Anspruch 1, wobei die Diagnoseschaltung ein Rauschen im Sensorausgangssignal dadurch bestimmt, dass sie entweder einen Quadratmittel-(RMS)-Wert des Sensorausgangssignals bestimmt oder eine Fourier-Transformation des Sensorsignals verwendet.
     
    12. Verfahren zum Diagnostizieren des Funktionsablaufs eines Durchflussmessgeräts des Typs, der dazu verwendet wird, den Durchfluss (14) eines Prozessfluids zu messen, wobei das Verfahren umfasst:

    Einholen eines auf den Durchfluss (14) des Prozessfluids bezogenen Signals (78);

    Bestimmen eines statistischen Parameters, der sich auf das auf den Durchfluss (14) des Prozessfluids bezogene Signal (78) bezieht; und

    Bereitstellen eines Diagnoseausgangs, der auf dem Vergleich des aktuellen Werts des auf den Durchfluss (14) des Prozessfluids bezogenen Signals (78) und den aktuellen Wert des bestimmten statistischen Parameters mit einem gespeicherten nominalen Verhältnis beruht, wobei es sich bei dem nominalen Verhältnis um ein Verhältnis zwischen dem bestimmten statistischen Parameter und dem auf den Durchfluss (14) des Prozessfluids bezogenen Signal (78) handelt, wobei das nominale Verhältnis empirisch bestimmt wurde, indem ein Funktionsablauf des Durchflussmessgeräts während normaler Bedingungen über einen Bereich von Prozessfluiddurchflusswerten beobachtet wurde.


     
    13. Verfahren nach Anspruch 12, wobei der statistische Parameter auf ein Rauschen im Sensorausgangssignal bezogen ist.
     
    14. Verfahren nach Anspruch 12, wobei das Signal auf einen Differenzdrucksensor bezogen ist.
     


    Revendications

    1. Débitmètre destiné à mesurer un flux (14) d'un fluide de processus, comprenant :

    un capteur (50) configuré pour fournir un signal de sortie de capteur (78) relatif au flux (14) du fluide de processus ;

    un circuit (77) configuré pour déterminer un paramètre statistique relatif au signal de sortie de capteur (78) ; et

    un circuit de diagnostic (76) configuré pour fournir une sortie de diagnostic (82) sur la base de la valeur actuelle du paramètre statistique déterminé et de la valeur actuelle du signal de sortie de capteur (78), sachant que l'appareil inclut une mémoire (72) configurée pour stocker une relation nominale entre le paramètre statistique déterminé et le signal de sortie de capteur (78), la relation nominale ayant été déterminée empiriquement en observant le fonctionnement du débitmètre en conditions normales sur une plage de valeurs de flux de fluide de processus, sachant que la sortie de diagnostic (82) est basée sur une comparaison entre la valeur actuelle du paramètre statistique et la valeur actuelle du signal de sortie de capteur (78) avec la relation nominale stockée.


     
    2. L'appareil de la revendication 1, sachant que le circuit de diagnostic détermine du bruit dans le signal de sortie de capteur.
     
    3. L'appareil de la revendication 1, sachant que le circuit de diagnostic détermine du bruit pendant des conditions de flux sensiblement équilibrées.
     
    4. L'appareil de la revendication 1, sachant que le capteur comprend l'un de :

    un capteur de pression différentielle ;

    un capteur magnétique ou un débitmètre magnétique, ou un capteur de vibration d'un débitmètre à vortex ou de Coriolis.


     
    5. L'appareil de la revendication 1, sachant que la relation nominale est déterminée soit pendant la fabrication du débitmètre, soit pendant le fonctionnement du débitmètre.
     
    6. L'appareil de la revendication 1, incluant la transmission de la sortie de capteur via une boucle de commande de processus.
     
    7. L'appareil de la revendication 1, sachant que la sortie de capteur est relative au colmatage d'un accouplement de processus.
     
    8. L'appareil de la revendication 1, sachant que la sortie de diagnostic est relative soit à un état d'une plaque à orifice, soit à une composition du fluide de processus.
     
    9. L'appareil de la revendication 1, sachant que la sortie de capteur est indicatrice d'un état d'une boucle de commande de processus.
     
    10. L'appareil de la revendication 1, sachant que le circuit de diagnostic détermine du bruit dans le signal de sortie de capteur moyennant un filtre passe-bande, ou un filtre passe-haut.
     
    11. L'appareil de la revendication 1, sachant que le circuit de diagnostic détermine du bruit dans le signal de sortie de capteur soit en déterminant une valeur moyenne quadratique (RMS) du signal de sortie de capteur, soit en utilisant une transformée de Fourier du signal de capteur.
     
    12. Procédé de diagnostic du fonctionnement d'un débitmètre du type utilisé pour mesurer un flux (14) d'un fluide de processus, le procédé comprenant :

    l'obtention d'un signal (78) relatif au flux (14) du fluide de processus ;

    la détermination d'un paramètre statistique relatif au signal (78) relatif au flux (14) du fluide de processus ; et

    la fourniture d'une sortie de diagnostic qui est basée sur la comparaison de la valeur actuelle du signal (78) relatif au flux (14) du fluide de processus et la valeur actuelle du paramètre statistique déterminé avec une relation nominale stockée, sachant que la relation nominale est une relation entre le paramètre statistique déterminé et le signal (78) relatif au flux (14) du fluide de processus, la relation nominale ayant été déterminée empiriquement en observant le fonctionnement du débitmètre en conditions normales sur une plage de valeurs de flux de fluide de processus.


     
    13. Le procédé de la revendication 12, sachant que le paramètre statistique est relatif à du bruit dans le signal de sortie de capteur.
     
    14. Le procédé de la revendication 12, sachant que le signal est relatif à un capteur de pression différentielle.
     




    Drawing

















    Cited references

    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description