Florin Turcu on making the most of the total focusing method
The total focusing method (TFM) is suitable for ultrasonic inspection of components across a range of different industries. When using the TFM however, the quality of results depends greatly on the setup used. Selecting the most suitable scan plan increases probability of detection, and the use of intelligent modelling tools during setup both saves time and increases chances of success.
In nondestructive testing, the TFM is a way to improve the detection capabilities of linear pulse-echo probes. It does this by using all the available probe elements to focus the beam electronically in all points of a given region of interest. The technique may in some cases improve detection sensitivity and resolution inside a predefined area.
In many ways the TFM obeys the same principles as conventional ultrasonic testing (UT) and phased array (PA). For example, electronic focusing with PA converges multiple wave fronts of the individual probe elements in a small area: the focal point. Such convergence is only possible within the near field of the active aperture created by the number of elements used on the PA probe. When using the TFM, these same limitations apply, while the active aperture involves the entire PA probe.
The frequency of a probe, as well as the size and number of elements, influences the setup and data quality – and there are important trade-offs to consider. With a higher frequency and a larger active aperture, for example, you can focus further away from the probe and have a larger focused region for improved TFM imaging. However, this setup will have a negative impact on the near-surface resolution.
Because of the many variables involved, relying on trial and error to determine the ideal probe for TFM setups is impractical. Instead, the use of intelligent modelling tools is essential for fast TFM inspection configuration.
One tool that can make it easy to select the right scan plan when setting up the TFM and increase probability of detection is the acoustic influence map (AIM). AIMs for different combinations of wave propagation paths (wave sets) can be generated on the Olympus OmniScan X3 flaw detector to provide a prediction of the acoustic beam coverage for different wave sets so inspectors can optimise their TFM scan plans.
From the AIM, a sensitivity index can also be calculated automatically. The sensitivity index is a measure of the maximum amplitude of reflected ultrasonic field, before normalisation, displayed in arbitrary units. It is proportional to the predicted voltage amplitude at reception. If you think of the AIM as a heat map showing where the amplitude response is the strongest, the sensitivity index is the maximum temperature.
Comparing AIMs and sensitivity indices gives a good indication of which configuration (probe, wedge, reflector shape and angle, wave set, etc) will perform best for an inspection.
Case study: inspecting a surface-breaking crack
Typical flaws in welds that can arise during the structure lifetime (figure 1), include volumetric flaws like porosity and slag (since manufacturing) or planar flaws like lack of side wall fusion (manufacturing flaw) or cracks (surface-breaking or not).
Surface-breaking cracks or cracks not connected to the weld root or backwall are typical flaws that can be difficult to detect using standard phased array or angle beam techniques. The reason being that it is impossible to generate a beam perpendicular to a flaw with vertical orientation.
The example in figure 2 looks at a 25mm-thick welded plate with a surface-breaking crack and the objective is to select the optimal TFM mode for inspection. Self-tandem modes are considered, because such sound paths are the only ones that can be reflected from a vertical flaw. They are also capable of producing true-to-geometry TFM images.
Inspecting a surface-breaking crack requires the best possible acoustic coverage near the upper surface of the bar. In this example, the TL-T and TT-TTT (5T) modes are likely to be the most suitable. The sensitivity index of the 5T mode 1.83, whereas the TL-T mode only reaches 0.41. This means that the 5T mode will have a signal-to-noise ratio that is around 4.5 times higher than the TL-T mode. Figure 3 shows TFM images acquired using these two modes. Note the higher background noise within the TL-T image correlated with the lower sensitivity index.
During inspection the gain values for the 5T and TL-T modes were set to 16 dB and 35 dB respectively, so that the peak amplitude values for both modes are 80%. As a result, the 5T mode has an acoustic sensitivity that is around 8 times better than the TL-T mode (19 dB), which can explain the lower noise level in the 5T TFM image.
The results also show that the surface-breaking nature of a vertical crack is also more accurately represented in the 5T mode – something that can be predicted by looking back at the AIMs in figure 2. AIMs are generated by assuming an ideal planar reflector, which explains the difference in acoustic sensitivity between the AIM prediction and the value measured experimentally.
Acoustic influence maps (AIMs) can model the acoustic coverage of different TFM modes on the Olympus OmniScan X3, and the sensitivity index helps to predict their relative acoustic sensitivities. This simplifies the process of selecting most suitable mode to three simple steps.
The operator needs to generate AIMs for all TFM modes that are relevant to the inspection configuration, select the TFM modes for which the AIM provides adequate acoustic coverage in the region of interest, and finally select the mode(s) with the highest AIM sensitivity index. Following these steps provides a fast way to maximise probability of detection in TFM ultrasonic inspections.
Florin Turcu is with Olympus Europa