Video sequences

See the table of available sequences.

Still photographs

Scattering of ultrasound in water

Video sequence - download

• 1. Ultrasonic waves in water approach an aluminiun cylinder of 20 mm diameter. Individual wavefronts are visible. The pulse train has emerged from an undamped transducer, located outside the field of view, that is driven by a single electrical pulse of duration 100 ns. The resonant frequency of teh transducer is 0.75 MHz. This picture is chosen to be the reference for subsequent time measurements and so the time is shown as 0.0 us. Within the pulse can be seen plane waves and edge waves. The plane waves are seen clearly and appear as about 20 parallel fringes of contrast. The edge waves are in the shape of an expanding toroid, being a diffraction effect from the circular transducer, and appear less clearly as waves at the right and left of the plane-wave pulse at an angle of approximately 30 degrees. The optical contrast method, schlieren, views the wave-field from one side and shows greater sensitivity to features producing greatest contrast perpendicular to the viewing direction. The edge waves do not appear to be as strong as the plane waves from the image but this is misleading. When the edge waves have passed out of the region of the plane waves the beam is said to be outside the near-field.
• 2. Waves collide with the cylinder. Calculation shows the speed fo sound to be 1500 (+/-200) m/s - in good agreement with the accepted value for the speed of sound in water. The elapsed time was measured to be 13.4 us. Incident waves are reflected in a circular pattern. Waves ahead of the incident waves result from transmission through the cylinder. The speed of transmitted waves is measured to be 5000 (+/-1000) m/s - in reasonable agreement with the accepted value for the speed of sound in aluminium. From experience edge waves caused problems for many ultrasonic systems working in liquids, such as flow-meters, inspection systems for oil-well risers and pipelines, inspection systems for water mains. This is because the edge-wave can be reflected back to a transmitter/receiver from the inside surface of a cylindrical pipe before the plane waves. This is invariably due to cylindrical aberration.
• 3. Incident waves pass-by the cylinder after 19.4 us. Transmitted waves emerging from the cylinder include several wavefront dislocations - two examples are shown. Dislocations in wavefronts are similar to dislocations in the atomic planes found in crystalline materials. A Burger's circuit drawn around the tip of a dislocation shows a residual phase. Research at Bristol University has been looking into alternative waves of describing diffraction effects by using the dislocation. Waves diffracted into the shadow of the cylinder by the main pulse of plane waves are probably Mie waves.
• 4. The interference pattern of incident waves, diffracted waves and transmitted waves is shown. Elapsed time is 34.0 us. The motion of dislocations can be followed. The principle of causality can be used to help understand some of the surprising structure of the wave-fronts: for example, we know that the leading wavefronts in the plane waves passing the cylinder (unaffected by its presence) and those reflected by the cylinder and those transmitted through the cylinder must have a causal relationship and occur at the same time. Following along a wavefront, without changing the phase value, these wavefronts must eventually join together. Due to the different speed of sound in the cylinder the only way this can happen is if dislocations are present. Dislocations cause problems for ultrasonic systems that measure small time differences, for example flow-meters, since the arrival of a dislocation at a receiver can cause a zero-crossing detector to make an error of one period - this is usually a very large. But these kinds of errors can be eliminated by repositioning the receiver to avoid the dislocation.

Image processing

Video images of visualised ultrasound can be processed to enhance the image. Various stages of processing are possible including using false-colours to represent gray scale values (as used here), capturing the image with and without the wavefronts present then subracting pixel-by-pixel.

This latter step removes the background of the image and any non-uniformities created by the optical method providing contrast. It is also possible to use signal averaging to improve the signal-to-noise ratio and to amplify the contrast range.

In the figures shown here the image has been captures and subtraction used to give a uniform background. When subtracting it is important to add in half of the dynamic range of the analogue-to-digital converter otherwise the processed image will show rectified values of image.

After processing, the final image looks predominantly yellow - being the colour corresponding to the value of 127 (half the 8-bit ADC range). The wavefronts now look much clearer and further stages of processing are possible.

As part of a research project the Hough transform was used to find the plane waves in the wavefield and a line was passed through the centre of gravity of the field - this is the ray that should emerge from the centre of the transmitter.

The objective of the research was to investigate ways to use visualization and image processing to produce enhanced quality control for ultrasonic transducers.

Synchronising with an oscilloscope

Related video sequence - download

Our visualization system is highly adaptable and can be triggered by other equipment (say, a medical scanner or an inspection system) it can also provide a trigger to other equipment (typically an oscilloscope of waveform capture computer system). It is now possible to see an image of the waves causing and the received signal displayed on an oscilloscope or a processed signal displayed on a computer.

This creates a powerful method for investigating your problems. Its possible to find an ultrasonic wave causing a signal component at the receiver and then run time backwards to find-out what has caused that wave and then perhaps eliminate the cause or make the effect greater. You can make a change to any part of the system, such as: changing the geometry, modifying the transducer, adding something to the propagation path or changing the signal processing and see the effects on the received signal. It is the ultrasonic equivalent of a debugger in a compiler.

We often use Matlab for prototyping our signal processing because it too allows changes to be made in the processing. The flexibility at each stage gives an extremely powerful way to design or debug ultrasonic systems. our philosophy of design is that the:

• transducer
• propogation path
• receiver
• signal processing

form a system that can only be developed together. It is as essential to see the ultrasonic waves as it is to see the signal after different stages of processing.

The image included here shows the signal received by a membrane hydrophone (the thick black line avove the signal trace box with a vertical pin to show the position of the sensitive region) from the waves shown visualized. The pulse of waves is travelling from top to bottom of the picture. The trigger point is the left of the oscilloscope trace; there is no signal processing. We nearly always choose to use the light flash as the trigger event; so, in this case, there is a delay between that event and the arrival of a signal pulse on the oscilloscope trace - corresponding to the time for the waves to travel through water to the membrane.

If we visualize the waves when they are just touching the membrane then the pulse will be at the trigger event (far left in this case). This condition allows us to label the first pulse on the oscilloscope as "main pulse". If we now visualize at later times we will see a pulse come up through the membrane this is reflected from the wall of the tank used for visualization. We can now label this as the "pulse from the wall". Visualization also shows that the pulse from the wall includes reverberations from the wall of the tank (the wall thickness can be measured very accurately from this pulse).

Whilst this example appears simple enough other simple geometries can create oscilloscope signals that, in isloation, defy this kind of interpretation. But with visualization the interpretation is nearly always possible.

For example, the sketch shown here is a little simpler to understand than the visualization image. It shows a random arrangement of scattering cylinders with two showing what happens to plane ultrasonic waves entering from the top (see the first section above here on scattering of ultrasound). This is based upon experiments carried out in support of developing inspection systems for concrete.

The scattering is repeated for every cylinder and by the time the pulse reaches the bottom of the field of view it has become severely disrupted but visualization still shows a pulse of energy with waves of the same fundamental wavelength.

A receiver, similar to the plane transmitter used to create the waves, registers

• no signal when the waves travel through the model of random scatterers
• strong signal when at the same position without the model of random scatterers.

Yet in both cases visualization showed that approximately the same amount of energy was arriving.

The reason that the receiver registered no signal with the model of random scatterers is that the phase across the receiver was disrupted with a random spatial distribution. Because the receiver integrates the phase it produces a zero signal.

Armed with this information it is possible to start to design inspection systems for concrete. This was an important result for the research and development programme.

Focus of waves

Many ultrasonic inspection systems use focussing ultrasonic waves. When pipes are inspected focussing can occur. Spherical aberration causes plane waves to be focussed to the caustic shown here.

Observe that the plane waves are converted into disrupted cylindrical waves after passing through the focal point - a coherent receiver will not work well with this kind of signal. A spatially matched transducer will work much better.

Visualization can be used to investigate many effects associated with inspecting pipes, risers and pipelines.

For example:

• What happens to waves launched from the axis of the pipe when they return to the receiver/transducer?
• What is the sensitivity to misalignment of the transducer from the axis?
• What is the effect of any window materials on the propagation?
• Can edge waves get back to the transducer before the main plane pulse and cause false readings?
• Can reverberations in the pipe wall be used to measure the thickness ?
• What diameter of transducer can be used so that phase disruption does not lead to a complete loss of signal?
• When must an array of transducers be used instead of a single coherent device?
• What kind of signal processing works best under these conditions?
• What happens when the surface is corroded?
• what kind of transducer should be used to detect corrosion?

Detecting signals with a hydrophone

Video sequence - download

In virtually all ultrasonic inspection system (other than visualization systems) a transmitter is a source of waves and a receiver collects the waves and converts them into an electrical signal. it is this electrical signal that is processed to extract information about the way the wave has travelled.

It's quite common to use the same transducer to transmit and receive but it's also common to use different transducers. It is this latter case that is shown in the next two visualized images.The signal from the receiver has been displayed on an oscilloscope and this is superimposed on the field of view of the receiver (a hydrophone) and the ultrasonic waves. The image is synchronised with the oscilloscope - using the time when the stroboscope light is flashed (an event lasting 30 ns) as the trigger source. The oscilloscope is set-up so that the trigger event is towards the left hand side of the screen but not the left-most point.

As in the earlier example we can see the power of this mode of synchronisation: waves seen to be touching the hydrophone are positioned close to or at the trigger point. Observe that in this case the time for the waves to travel through some opaque surface layer of the transducer appears as a delay in the signal a the trigger point. Clearly, there is scope to perform some reverse engineering on the competition's trandsucers without opening them up!

At Cambridge Ultrasonics we term this mode of synchronised use of the visualization system and an oscilloscope (or PC with data capture) as "watching the signals enter the transducer". By adjusting the position of the waves in relation to the hydrophone it's possible to label each wavelet as coming from a particular source: main reflected pulse, first reverberation in the reflecting plate, second reverberation, unwanted reflection from the side of the water tank, reflection from the water surface, internal reverberation in the oil between the transducer and its housing assembly, reverberation from the wall of the housing, etc...This can of information is absolutely invaluable for the mechanical debugging of transducers - trying to devise signal processing algorithms to overcome mechanical design problems is a waste of time and money! With visualization both mechanical debugging and signal processing development can proceed very much faster.