Still waters run deep

Citation

Loughlin, C. (2002), "Still waters run deep", Sensor Review, Vol. 22 No. 3. https://doi.org/10.1108/sr.2002.08722caa.001

Still waters run deep

Still waters run deep

Our themes for this issue are flow, distance and level measurement. In many ways these may seem like very different subjects for inclusion in the one issue, but they share a great deal of common ground.

The basic sensing mechanisms used in these applications do indeed vary considerably with only marginal overlap, especially with flow measurement applications. However, where they do share common challenges is that in all three cases it is very difficult to know what you are actually measuring.

For example, in flow measurement of a liquid through a pipe what you really need to know is how much liquid is going past point A. The only sure way of doing this would be to cut the pipe at that point and measure the volume of liquid that comes out over a given period. Even this is flawed as the approach will almost certainly affect the rate of flow due to the decrease in pressure at the outlet. The biggest problem with flow measurement is that it varies over the cross section of the pipe. Skin effects and turbulence all serve to confuse the issue (see Benhadj and Ovazzane's article, pp. 223-31).

Similarly distance and level have their own uncertainties. Levels are confused by ripples, surface attraction, foaming and thermal gradients; while distance should be OK provided you are absolutely certain of the points that you are measuring the distance between, and that the method is precisely defined. For example, a laser can give a very accurate reading provided you are sure that the spot size does not encompass surface irregularities and that diffraction and thermal effects have not distorted the desired straight line measurement (see the Tutorial by Saadat and Cretin, pp. 199-206).

With all three application areas it is only the intelligent processing of the data provided by the sensors that enables anything better than rudimentary precision to be obtained. Often it is the combination of multiple sensors that provides the additional data. Sometimes these sensors can be all of the same type, with the final result being simply an average of their signals. In other cases sensors with widely different properties and strength and weaknesses may be combined. For example, in the article by Turner on "Future proof signal conditioning" (see pp. 207-12), a flow velocity signal (e.g. from a paddlewheel) is combined with a liquid level sensor (e.g. ultrasonic distance) to produce a measure of flow rate that is corrected for liquid depth.

The use of multiple sensors is also a very powerful method for determining system integrity. If your ultrasonic wind sensor, rotating cup anemometer and wind sock all indicate a Force 10, then you probably do not need to poke your head out of the window to see that you should stay indoors.

Clive Loughlin

Call for papers

SR 23:1 Gas discharges + thermal imagingCopy in deadline: 30 September 2002Thermal imaging sensors and systems. Analysis and visualisation of gas discharges, including arcs and high and low pressure plasma.

SR 23:2 Radar + sonarCopy in deadline: 18 November 2002Sensors, systems and applications for radar and sonar. Concentrating on new applications in industry and service sectors.

SR 23:3 Machine vision + laser scannersCopy in deadline: 2 March 2003Latest developments in machine vision and laser technology. Including 3D inspection and inspection of partially occluded parts.

SR 23:4 Biosensor + biometricsCopy in deadline: 18 May 2003Personal identification and security applications. Biosensors and chemical analysis.