In brief
- FMC records everything — every element transmits, every element receives
- TFM focuses at every pixel, not only along predefined focal laws
- Highest resolution of any ultrasonic technique in current industrial use
- Wave mode selection is the hard part — and where results are won or lost
- Not a replacement for PAUT — slower, more demanding on surface and access
Total focusing method (TFM) is a way of processing ultrasonic data that produces a sharper image than conventional phased array, by focusing the beam at every single point of the image rather than at a limited set of predetermined depths.
It is not a different physical method. The sound, the probe and the material behave exactly as in phased array testing. What changes is how the data is collected and what the instrument does with it afterwards.
Full matrix capture: recording everything first
Conventional PAUT fires groups of elements together, with delays chosen in advance to steer and focus the beam in a particular way. Those delay patterns — focal laws — are decided before the inspection, and the image is limited by them.
Full matrix capture (FMC) works differently. Each element fires individually, and all elements receive the resulting signal. With a 64-element probe that yields 64 × 64 = 4,096 individual A-scans per position — the complete acoustic response of the region, with no prior assumption about where to focus.
TFM then reconstructs an image from that matrix. For every pixel in the region of interest, the algorithm calculates the travel time from each transmitter to that point and back to each receiver, sums the corresponding signals, and assigns the result to the pixel.
The consequence: every point in the image is a focal point. There is no near field or far field trade-off, no depth at which resolution degrades because the focus was set elsewhere.
Wave modes — the part that requires expertise
TFM does not simply “focus better”. It reconstructs the image along a chosen propagation path, and the choice matters enormously.
A pulse can reach a discontinuity directly, or after reflecting from the back wall; it can travel as a transverse wave throughout, or convert to longitudinal at a boundary. Each combination is a distinct wave set — commonly written as TT, TTT, LL, TTTT and so on, describing the sequence of legs and modes.
Different wave sets favour different defects. A vertical crack near the root may be almost invisible in one mode and obvious in another. Selecting the wrong wave set produces a clean, high-resolution image of nothing useful.
This is why TFM demands a written procedure and a qualified technique, not just better equipment. A high-resolution image is persuasive precisely because it looks authoritative — which makes an incorrectly configured TFM scan more dangerous than a poor conventional one.
Amplitude fidelity
Because TFM is a computed image, the reconstruction can distort amplitude — and amplitude drives sizing and acceptance.
Amplitude fidelity measures how much the reconstruction alters the true amplitude of a signal, and standards impose a limit on it. It depends on the pixel grid resolution: too coarse a grid and real amplitude variation falls between pixels. Demonstrating amplitude fidelity within the permitted limit is part of qualifying a TFM setup, and it is a reasonable question to ask any contractor offering the technique.
What TFM detects best
- Cracks with complex or unfavourable orientation, where a single fixed focal law struggles
- Lack of fusion and lack of penetration in geometries that conventional PAUT images poorly
- Small discontinuities requiring resolution beyond standard phased array
- High temperature hydrogen attack (HTHA) — micro-fissuring in refinery equipment, where TFM’s resolution is a genuine advantage over conventional techniques
- Laminations and hydrogen-induced damage in pressure equipment
- Thin-walled components — ISO 23864 addresses thicknesses from about 3.2 mm
Limitations
Computing load. Reconstructing every pixel from a full matrix is demanding. Frame rates are lower than PAUT, which limits scanning speed and makes TFM less suited to long weld runs where PAUT covers the ground faster.
Data volume. FMC files are large. Archiving, transfer and re-analysis need planning, particularly across a long project.
Surface preparation is stricter than PAUT. The technique is sensitive to roughness, coatings and deposits, and normally requires the surface to be ground. Where preparation is impossible, TFM is not available.
Access. Like all contact ultrasonic techniques, TFM needs a scanning surface with room for the probe and wedge.
Coarse grain still limits it. Higher resolution does not overcome attenuation and scatter in austenitic or coarse cast structures. TFM helps, but does not make the material transparent.
Procedure qualification. Given the influence of wave mode selection and pixel grid, most specifications require demonstration on a representative mock-up with known defects before results are accepted
TFM or PAUT?
| TFM | PAUT |
|---|
| Resolution | Highest available | High |
| Focus | Every point of the image | Predefined focal laws |
| Scanning speed | Lower | Higher |
| Long weld runs | Less suitable | Well suited |
| Complex or skewed defects | Superior | Good |
| Surface preparation | Stricter | Standard grinding |
| Data volume | Large | Moderate |
| Setup expertise | Wave mode selection critical | Focal law setup |
The practical division: PAUT covers ground, TFM resolves detail. Many scopes use PAUT for the bulk of the weld and TFM where an indication needs characterising, or where a specific damage mechanism such as HTHA is the concern.
Standards
| Scope | Standard |
|---|
| Automated FMC/TFM testing of welds | ISO 23864 |
| General use of FMC/TFM techniques | ISO 23865 |
| PAUT of welds — techniques and levels | ISO 13588 |
| PAUT acceptance levels | ISO 19285 |
| Equipment characterisation | ISO 18563 |
| General ultrasonic principles | ISO 16810 |
| Personnel certification | ISO 9712 |
| US practice | ASME BPVC Section V, Article 4 |
TFM is a technique within the UT method under ISO 9712, not a separate certified method. Evaluation requires UT Level 2 as a minimum; procedure approval requires Level 3.
Frequently asked questions
What is the difference between TFM and PAUT? PAUT builds an image from a limited set of predefined focal laws, so focus is optimal only at chosen depths. TFM reconstructs from full matrix capture data and focuses at every point of the image, giving higher resolution throughout — at the cost of speed and processing load.
What is full matrix capture? An acquisition scheme in which each probe element transmits individually while all elements receive, recording the complete acoustic response of the region. TFM is the post-processing applied to that data.
Can TFM replace radiography? For planar defects, frequently yes, with no radiation controls required. Radiography remains preferable for fine volumetric porosity, and some specifications still mandate it.
Does TFM require special surface preparation? Yes, and stricter than conventional PAUT. The surface must be smooth and clean for reliable coupling; grinding is usually required.
What thickness range does TFM cover? ISO 23864 addresses components from about 3.2 mm upward. Practical limits at the upper end come from attenuation and processing load rather than the technique itself.
Why does wave mode selection matter so much? Because TFM reconstructs the image along a chosen propagation path. The wrong wave set produces a high-resolution image in which the defect of interest simply does not appear — and the image looks convincing regardless.
Need TFM inspection to ISO 23864? We operate FMC/TFM-capable instrumentation and hold ISO 9712 Level 3 in UT — we write and qualify the technique, including wave set selection, rather than running a default setup. → TFM inspection services · Request a quote
Reviewed by Bartosz Żuberek — Managing Director, ISO 9712 Level 3