The need for a compact, non-invasive, real-time measurement of the density and speed of sound of a liquid led to the development of an ultrasonic sensing technique based upon the reflection of ultrasound at the solid-fluid interface, as shown in the schematic diagram. The transducer is mounted on the outside of the pipeline or vessel wall and an ultrasonic pulse is reflected at the solid-fluid interface and again reflected at the solid-transducer interface and makes over 15 so-called echoes with the pipeline wall. These echoes are recorded by the same transducer, as shown in illustration. Each time the ultrasound strikes the solid-fluid interface some fraction is reflected and some fraction is transmitted into the fluid. These fractions are dependent upon the properties of the slurry - its density and velocity of sound in the slurry. The many echoes act as an amplifier of this effect, and observation of the echo yields the product of the density of the slurry and the velocity of sound in the slurry. Observation of the time-of-flight across the pipeline leads to a measurement of the velocity. The combination of the two measurements gives the density of the slurry. The uncertainty in the density measurement is about 0.2% to 0.5%.
The self-calibrating feature is a very important one for this sensor, and eliminates the problems caused by small shifts in the electronics. For example, suppose that the pulser voltage to the transducer is reduced by 1%. This results in a similar change in each echo shown in the above diagram. The analysis for determining the density value depends essentially upon dividing the height of each echo by the height of the next successive echo. In this division the fractional change cancels out and the result is the same as that before the pulser voltage change. That is, the sensor is self-calibrating and the value of the density will not be affected by the voltage change.