Vortex shedding tutorial - Part 2

In Part 1 of this article on vortex shedding flowmeters we looked at the principles of their operation and the basic construction of the meter. In Part 2 we examine the transducer and the transmitter.

As explained in Part 1, the vortex shedding flowmeter can be described as having two major components: the flowtube and the transmitter.

The flowtube

The flowtube of a vortex flowmeter is made up of three functional parts: the flowmeter body that contains the fluid and acts as a housing for the hydraulic components, the shedder that generates the vortices when the fluid passes by (both described in Part 1) and the sensor(s) that by some means detects the vortices and produces a usable electrical signal.

Sensors When vortex shedding is present, both the pressure and velocity fields in the vicinity of the shedder will oscillate at the vortex shedding frequency. Pressure or velocity sensors are used to transform the oscillating fields to an electrical signal, current or voltage, from which the vortex shedding frequency can be extracted. Figure 1 shows an example of a piezoelectric differential pressure sensor located in the upper portion of the shedder, which converts into an electrical signal the oscillating differential pressure that exists between the two sides of the shedder. Some other sensing means utilised to detect vortex shedding are capacitive, thermal and ultrasonic sensing. Often the flowmeter electronics are mounted remotely from the flowtube, and this may require a local preamplifier to power the sensor or boost its signal strength.

Figure 1: Example of a vortex sensor.

Since the sensor is the mechanical component in a vortex flowmeter most likely to fail, most designs have a provision for replacement of the sensor. In some cases they can be replaced without removing the flowmeter from the pipeline. An example of an isolation manifold for sensor replacement under process conditions is shown in Figure 2.

The transmitter

Vortex flowmeter transmitters can be classified into two broad groups: analog and intelligent (or smart). Communication with the analog transmitter is carried out via local mechanical means, including switches and jumper wires. Communication with the newer intelligent device is carried out via digital electronic techniques. Both types of transmitters are two-wire devices. One of the more important features of the intelligent transmitter is that it allows application-specific information to be loaded into the transmitter. This information is then used internally to automatically tailor the transmitter to the application, including calibration of the 4-20 mA output. Before describing these devices, it will be useful to consider the three most common forms of flow measurement signals provided by vortex transmitters. Measurement output signals The three most common ways for a transmitter to communicate measurement information to the outside world are 4-20 mA, digital and pulse signals.

Figure 2: Isolation manifold for sensor replacement.

The 4-20 mA signal is the DC current flowing in the power leads to the transmitter. This current is directly proportional to the vortex shedding frequency and hence is also linear with flow (see Figure 3). A current of 4 mA corresponds to zero flow and 20 mA to the maximum flowrate, that is, the upper range value (URV) of the meter. A frequency-to-analog converter in the analog meter and a digital-to-analog converter in the intelligent meter produce this output. The digital signal is a digitised numeric value of the measured flowrate in engineering units transmitted over the two wires powering the meter.

Figure 3: 4-20 mA and pulse outputs.

The pulse signal is a squared-up version of the raw vortex signal coming from the sensor (see Figure 4) and is accessible via a pair of terminals inside the transmitter housing. The frequency of the pulse signal is either identical to the vortex shedding frequency (raw pulse) or some multiple thereof (scaled pulse). As discussed in Part 1, in either case the frequency of the pulse signal is linearly proportional to flowrate going from zero to the frequency at the URV, fURV (see Figure 3).

As shown in Figure 3, at a low but non-zero flowrate the frequency and mA signals drop to zero and 4 mA, respectively. The flowrate at which this abrupt change takes place is normally referred to as the low flow cut-in, LFCI or cut-off. The reason for this forced zero is to avoid erroneous flow measurements that can occur at low flow resulting from process noise, including hydrodynamic fluctuations, mechanical vibration and electrical interference. The digital signal also drops to zero below the LFCI flowrate. Analog transmitters

Figure 4: Raw pulse output.

Originally, analog transmitters were constructed totally of analog electronic components. Today, they may be built around a combination of analog and digital electronic components. In either case the measurement output is in the form of a raw pulse and/or a 4-20 mA signal. Depending on the particular transmitter, one or more of the following functions is available for tailoring the device via mechanical means to the specific application:

Signal output selection

If the transmitter provides both raw pulse and 4-20 mA signals, but not simultaneously, a means is available for selecting the one desired.

4-20 mA calibration

Use of the 4-20 mA signal requires that 20 mA correspond to the desired URV. This is accomplished by inputting, via a signal or pulse generator, a periodic signal with a frequency corresponding to the upper range frequency (URF) and adjusting the output current until it reads 20 mA. The URF, which is the frequency corresponding to the vortex shedding frequency at the desired URV, is calculated using the equation URF = k X URV.

In order to achieve the accuracy specified by the manufacturer, the constant k used in the above calculation must be corrected for process temperature and piping effects according to the manufacturer’s instructions. The temperature effect is a result of thermal expansion of the flowtube, and is described by:

where a is the thermal expansion coefficient of the flowtube material and T₀ is the fluid temperature at which the meter was calibrated. If the shedder and meter body materials are different a must be replaced by (2a₁+a₂)/3, where a₁ is the thermal expansion coefficient of the meter body material and a₂ that of the shedder.

Piping disturbances also affect the K-factor because they alter the flow profile within the flowtube. This will be discussed in more detail in Part 3 when discussing adjacent piping.

Low flow cut-in

For optimum measurement performance the low flow cut-in should be set to fit the specific application. The goal is to set it as low as possible, while at the same time avoiding an erroneous flow measurement output.

Filter settings

To reduce noise present on the signal from the sensor, electronic filters are built into the transmitter. Normally, means are provided for adjusting these filters - that is setting the frequencies at which they become active. By attenuating frequencies outside the range of the vortex shedding frequency, which varies from one application to another, better measurement performance is achieved.

Intelligent transmitters

Intelligent transmitters, which are microprocessor-based digital electronic devices, have measurement outputs which usually include two or more of the following: raw pulse, scaled pulse, 4-20 mA and digital. With regard to the digital output, there is at present no single, universally accepted protocol for digital communication; however, a number of proprietary and non-proprietary protocols exist.

The presence of a microprocessor in the intelligent transmitter allows for improved functionality and features compared to the analog transmitter, including:

• elimination of the need for 4-20 mA calibration
• automatic setting of low flow cut-in
• automatic setting of filters
• adaptive filtering (active filters that track the vortex frequency)
• digital signal conditioning
• K-factor correction for temperature and piping disturbances and correction for non-linearity of K-factor curve including the pronounced non-linearity at low Reynolds numbers.
• integral flow totalisation
• digital measurement output in desired engineering units

Configuring or tailoring the transmitter to a specific application is carried out by one or more of the following digital communicators:

• Local configurator - a communicator built into a transmitter that has a display and keypad
• Handheld terminal - a palm-sized digital device programmed for configuration purposes
• PC configurator - a personal computer containing configuration software
• System configurator - a digital measurement and control system with embedded configuration software

Using one of these configurators, the set of parameters that define the configuration can be modified to fit the application in question. The details of this data set vary depending on the specific transmitter; however, the general categories of information listed below apply.

• Flowtube parameters (eg, tube bore, K-factor, serial number)
• User identification parameters (eg, tag number, location)
• Transmitter options (eg, measurement units, function selections)
• Process fluid parameters (eg, fluid density and viscosity, process temperature)
• Application parameters (eg, K-factor corrections, URV, LFCI level)
• Output options (eg, measurement output modes, damping, fail-safe state)

In Part 3

In Part 3 of this article we will examine application considerations when deploying a vortex flowmeter, as well as recent new advances.

 

By Wade Mattar and James Vignos, PhD, Foxboro, Invensys

 

Additional reading RW Miller, Flow Measurement Engineering Handbook, 3rd edition, McGraw-Hill, 1996, chapter 14. WC Gotthardt, ‘Oscillatory Flowmeters’, Practical Guides for Measurement and Control: Flow Measurement, editor DW Spitzer, Instrument Society of America, 1991, chapter 12. JP DeCarlo, Fundamentals of Flow Measurement, Instrument Society of America, 1984, chapter 8. American Society of Mechanical Engineers, Measurement of Fluid Flow in Closed Conduits Using Vortex Flowmeters, 1998.