Phased Array Ultrasound (PAUT)
Spearheaded by radiotelescopy, sonar and medical industries since the 1950’s, phased array technology (PAUT) has proven to be a powerful technology for high-resolution imaging.
RECAP IN 10 EASY PRINCIPLES
1. A MEANS TO CREATE ACOUSTIC BEAMS
Multi-element technology (a.k.a. phased array) allows creating acoustic beams that will have the same characteristics as a beam produced by the conventional technology (a.k.a. mono-channel or mono-transducer): width, depth of field, half-angle divergence, etc. However, the multi-element technology allows changing the beam properties by fast electronic reprogramming. This is why we call a multi-element system a « beamformer ». We typically change 4 things with electronic means: aperture size, aperture position, transmit angle, and focus depth.
2. ELECTRONIC PROGRAMMING IS BASED ON FOCAL LAWS
A focal law creates a transmitted beam (TX) and a sensitivity field (RX). Each beam will have it’s own focal spot. These spots must coincide in space in order for a signal to be detected, in pulse-echo or pitch-catch. Each focal law will produce an A-scan signal. To create a real-time sectorial (S-scan) or linear (L-scan) view, many focal laws must be executed one after the other, each one juxtaposed to the other, to create a full image thanks to color-coding of the sample amplitude. We can also create focal laws that are for discrete non-juxtaposed beams, just like creating 2 or 3 angles or a TOFD tandem.
3. CONTROL OVER EACH ELEMENT
Minimally, a focal law can control, for each element, the activation in TX, delay in TX, activation in RX, delay in RX, and apodization (gain) in RX. The delays are typically from a few tens to a few thousands of nanoseconds (ns). The multi-element system will round (quantize) each delay with a resolution of 2.5 ns, which is generally good to probes up to 15 MHz. The finer the quantization of delays, the better the focusing. But couplant and other unknowns are causing a small impact on the actual quality of the focusing.
4. DELAY LAW
It is the part of the focal law which concerns delays in TX and in RX. A focal law is often presented under the shape of a centered or off-centered hyperbolic histogram. We can interpret this hyperbole like an electronic lens with a variable concavity, depending of the focusing depth. The idea is to generate a wave front that will propagate and possibly focus at the focal point.
5. THE FOCAL NUMBER (F-NUMBER)
Multi-element focusing is similar to optical focusing. Aperture to focal distance ratio will give an idea of the depth of field. It is mostly important to remember the 2 extremes. 1) Many piezo elements can be used to focalize close to the surface, to provide a very small focal spot, a short depth of field, but high resolution at the spot. Major issue is that we’ll be « blind » at other depths. 2) On the contrary, if we use few elements, we’ll enlarge the beam very much, we’ll lose the resolution but the depth of field will extend and we’ll see most indications at all depths. Ideally, we should use an aperture between F/1 and F/4, where F is the focal distance.
6. L-SCAN VIEW TO GET IT FASTER
L-scan view can be perceived as a B-scan view in real-time. The same image can be produced with a mechanical movement from a mono-transducer. The L-scan is often used at 0 degree to verify corrosion, inclusions, delaminations, etc. Angled L-scans are also used for weld inspection, preferably adjusted to have perpendicular incidence on weld side wall, to maximize sensitivity for lack of fusion. L-scan apertures are typically using 4 to 10 elements, except for special applications.
7. S-SCAN TO ENLARGE THE INSPECTED ZONE
Sometimes, we cannot or do not want to move the transducer. S-scan then becomes a very useful view to enlarge the inspected volume. The S-scan is very useful to inspect welds from low to high angles, in shear wave or longitudinal wave, but require careful probe positioning in order to see lack of side wall fusion. S-scan apertures are always 16 elements (or more), to narrow the beam width and maximize the sensitivity.
8. OTHER POSSIBLE VIEWS
Since multi-element is simply another way to produce A-scans, we can then derive standard views like B-scan, D-scan and C-scan. However, we can do much more than that. We can cumulate the so-called 3D voxels to produce a top view, front view and side view. By analogy, top view becomes similar to a view produced by radiography.
9. MULTI-SCAN OPERATION
It is possible to create many scans, in order to obtain many different images simultaneously (ex: 1x S-scan + 1x L-scan). Focal laws of each scan are juxtaposed in memory and executed one after the other by the beamformer. The computer system will know how to create different images adequately. Each scan can be controlled and calibrated separately. A tangible example is applicable to the weld inspection: L-scan checks delamination in the heat-affected zone (HAZ), while S-scan detects discontinuities in volume.
10. MULTI-CHANNEL PULSER-RECEIVER
For space reasons as well as budget, compact chips are used and offer a worthy compromise in PAUT. However, features such as signal-to-noise ratio, total gain or crosstalk are generally less performing than a channel designed for a mono-channel operation. Nonetheless, it is possible to use multi-element channels as a multitude of mono-channels, with or without LEMO or BNC adaptor. Expectations must then be lowered on performance.