Clamp-on Domestic Water Meter

Monitoring the use of water in households allows for reductions in demand and facilitates efficient use of resources. Clamp-on ultrasonic flow meters offer several advantages in this application, including reduced installation costs and no interruptions to supply. However, there is no mass-produced clamp-on flow meter for this application due to complications with interpreting the ultrasonic signals, which originate from guided waves in the thin metal pipe walls.

The meter uses the transit time difference principle: two transducers are placed on the pipe such that the resulting ultrasound in the water travels partially with/against the flow direction. One transducer is used as a generator, and the other as a detector, to send ultrasound in an approximately upstream direction. The roles of the transducers are then reversed, and ultrasound is sent in the downstream direction. Ultrasound that travels in the same direction as the flow travels faster, and therefore takes less time, than ultrasound that travels against it. The resulting transit time difference is then used to calculate the fluid flow rate inside the pipe.

An example of an ultrasonic signal from a 15 mm domestic copper pipe is shown below.

An ultrasonic waveform from a guided wave flow meter showing 6 evenly spaced arrivals marked 1 to 6 V*.

One would usually expect to see a single burst of ultrasound transmitted through the water and pipe, but in this case several bursts are observed marked 1-6V*. To use these arrivals for ultrasonic flow measurement, their paths through the flowing water must be understood.

Finite element modelling result of a cross-section through an ultrasonic flow meter, showing how ultrasound propagates through the pipe wall and fluid.

The figure above shows a 2D slice through the flow meter with FEM results overlaid. In panel (a), the ultrasound is generated in the left transducer wedge, which generates a guided wave in the top pipe wall. As the guided wave travels along the pipe, it leaks ultrasound into the fluid inside, which propagates towards the bottom of the pipe as shown in panels (a)-(b). In panel (b), the burst in the water is seen coupling into the bottom pipe wall, where it forms another guided wave. This too leaks ultrasound into the water, where it makes its way back to the top of the pipe as shown in panel (c). This process repeats several times, and each guided wave generated in the top pipe wall eventually travels underneath the reception transducer, where it is detected. Each V* arrival marked on the signal above therefore corresponds to ultrasound which has taken increasing numbers of V-shaped paths through the water.

From a flow measurement perspective, these arrivals behave very similarly to simple V-paths that would be observed on a thick-walled pipe. However, the ultrasonic behaviour in this system is much more complex. In a conventional clamp-on flow meter on a thick-walled pipe, the ultrasound is often considered to travel along a single ray path representing the centre of the ultrasonic beam. In the guided wave meter, the continuous leakage of the guided wave into the water results in a continuum of possible paths, and each arrival can be thought of as arising from ultrasound which has travelled along all possible paths resulting in the same transit time. The figure below shows two such paths which both have the same transit time and would both contribute to the 1V* arrival.

Two cross-sectional slices through a flow meter showing two paths that energy could take which would contribute to the 1V* arrival. In the first example, the ray diagram shows ultrasound traveling down through the water in a V shape directly under the transmitting transducer, and completing the rest of its path the the receiver as a guided wave in the top wall. In the right panel, the ultrasound travels down to the bottom of the pipe, where it travels some distance as a guided wave in the bottom wall before being re-emitted into the water, coupling back into the top wall, then being detected.

Since the arrivals observed later in time have travelled greater distances in the flow, the transit time differences measured within them increase linearly with both the number of V paths they have transited and the flow rate, as shown in the figure below. The result is that it is possible to measure the flow in each of the 6 arrivals independently and combine the results for a more accurate flow measurement.

Two plots showing the relationship between the transit time difference and the flow rate from experimental data. The left plot has flow rate as the horizontal axis and transit time difference as the y axis, and shows 6 lines extending from the origin with linearly increasing gradients, corresponding to the 6 arrivals of ultrasound. The right plot shows the residuals when each of these datasets is fitted with a straight line, which indicates the error on the flow rate measurement.

More information on the guided wave flow meter for domestic copper pipes can be found in our journal articles linked below.

Clamp-on measurements of fluid flow in small-diameter metal pipes using ultrasonic guided waves

Transducer design for clamp-on guided wave flow measurement in thin-walled pipes