PAUT and TOFD Instruments: How to Choose One Before You Buy

Take a job that comes up all the time on pipeline spreads: girth welds on 12-inch carbon steel pipe, 12.7 mm wall, with a specification that calls for phased array backed by TOFD, recorded data and a stated acceptance criterion. The bids come in with near-identical datasheets, each with its 16:64 or 32:128, a touchscreen and bundled software. The lines that decide whether the instrument can handle that job are elsewhere. You need to know how many acquisition groups run together, whether a conventional channel is free for the TOFD pair, which wedges and which encoded scanner are in the price, how the system is verified before each shift, and who is qualified to sign the report.

For whoever runs an NDT department or a quality office, and for the inspector who will actually use the instrument, the real risk is buying a unit that passes its own acceptance test and then stalls at the first client specification. Five items need checking: channel configuration, the probe and scanner kit, software, instrument verification and training. None of them shows up in the list price. All of them end up in the cost of inspection.

Welds, corrosion or composites: three different purchases under one label

The word ultrasonic covers purchases that have little in common. On welds you are looking for planar and volumetric flaws to locate and size, and the instrument works toward an acceptance criterion written in a standard, while in corrosion mapping what matters is wall loss over large areas, measured at points that must stay comparable from one campaign to the next. On composites and attenuating materials the game shifts to frequency, probe damping and anisotropy. Three different trades, in short. An instrument built for thickness gauging copes poorly with a scan plan on a structural joint; the reverse is wasted money, and a daily nuisance.

What 16:64 or 32:128 really means on a PAUT instrument

The colon figure is the most quoted line on any datasheet, and the least explained. The first number is how many elements the instrument drives at once in transmit and receive, that is the maximum active aperture of a single focal law, the set of delays used to fire the elements. The second is how many elements it can address in total through the multiplexer. A 64-element probe with 0.6 mm pitch on a 16:64 instrument therefore works with a maximum aperture of 9.6 mm, stepped along the roughly 38 mm of the probe.

On the 12.7 mm wall of our example, an aperture of that size is often enough. Things change on heavy walls, where focusing at depth takes more active elements, on the long probes used in corrosion mapping, which cover a wider band per pass, and on matrix probes. That is where 32:128 or higher pays for itself. On 10 mm plate you would be paying for channels that stay switched off.

The label is also silent on two points. The first is how many groups the instrument acquires in one pass: the pipeline job needs at least two phased array probes, one on each side of the weld, plus the TOFD pair, and if the instrument handles one group at a time the scan has to be repeated. The second is conventional channels. TOFD works with two single-element probes, one transmitting and one receiving, and you need to know how the instrument drives them: through dedicated channels or through an adapter on the phased array channels, an option some instruments allow and others do not. For FMC/TFM, finally, what counts is how many elements the instrument receives in parallel and how fast it reconstructs the image, and neither figure appears in the label.

PRF and scan speed: the arithmetic that sets productivity

The productivity of a phased array system comes down to a multiplication worth doing before you sign. With a 1 mm scan increment and a travel speed of 50 mm per second, the instrument has to complete 50 acquisitions per second. If each one includes two sectorial scans of 60 focal laws plus the TOFD pair, that is roughly 6,000 pulses per second, before any signal averaging. When the instrument cannot keep up, the operator slows down. Or accepts missing lines in the data, which no later analysis can rebuild.

Pulse repetition frequency, the PRF, cannot be raised at will either. On heavy walls or low-attenuation materials, a pulse fired too soon after the previous one picks up that pulse's late echoes and produces ghost indications, so the limit is set by the instrument and the part together. The scan speed you can reach with your scan plan, on your wall thickness, is worth more than any maximum PRF in a brochure, and it is the first figure to ask the supplier for.

From ISO 13588 to ISO 19285: what the specification asks of the instrument

Many procedures still in use are written to ISO 17640:2018, which covers manual ultrasonic testing of welds with conventional probes, on ferritic steel 8 mm and thicker. Buying a phased array does not change the procedure by itself. Testing with recorded data has its own standard, ISO 13588:2019, which covers semi-automated and fully automated phased array testing of full-penetration joints in low-alloy or fine-grained steel from 6 mm upward, and defines four testing levels.

ISO 13588 says how the inspection is performed and contains no acceptance criteria. Those are in ISO 19285:2017, which sets three acceptance levels for indications classified to ISO 13588, tied to the quality levels of ISO 5817. TOFD follows the same pattern: technique to ISO 10863:2020, acceptance to ISO 15626:2018. For the buyer the consequence is direct, because the required testing level decides how many probes, angles and set-ups must be acquired, and therefore how many groups and channels are really needed.

TFM went down the same road in 2021, with ISO 23864 for welds from 3.2 mm upward and ISO 23865 for the general rules of FMC and TFM; since then the technique can be specified in a contract rather than agreed case by case. Outside the ISO world the references change. ASTM E2700, in its 2026 revision, covers contact phased array testing of welds between 9 and 200 mm thick, while for pipelines built to API 1104, now in its 22nd edition, the procedures and acceptance criteria for ultrasonic testing sit in the standard itself. Before comparing two bids, find out which of these documents your client will cite.

Probes, wedges and scanners: the real quote is in the kit

On its own, the instrument inspects nothing. Probe and wedge decide which angles can be reached and which zone is covered, and on pipe the wedge has to be contoured to the outside diameter, or coupling is lost right at the edges. The pipeline job needs two probes with their wedges, a TOFD pair with its own, a scanner that holds the probes at the right offset from the weld, and an encoder that records position around the circumference. Each item has a price and a lead time, and a bid that quotes the bare instrument cannot be set beside one that quotes the system.

In corrosion mapping the weight shifts to scanner and encoder. Every C-scan point has to be tied to a known position, because degradation is measured by comparing successive campaigns on the same grid. ISO 16809:2025 sets the principles of ultrasonic thickness measurement, by contact and immersion; ISO 16810:2024 sets the general principles of testing, describes conventional probes and warns that array techniques may need additional verification. A repeatability figure quoted without calibration procedure, block and temperature tells little to someone comparing two maps taken two years apart.

Composites and attenuating materials: claims are proven on the part

In composite laminates the high frequency that gives resolution in steel is absorbed quickly, so you often have to step down in frequency and choose high-damping probes with short pulses. Anisotropy makes the beam path less predictable; curved surfaces call for dedicated wedges or immersion. On these geometries FMC/TFM promises a lot. ISO 23865 itself, however, considers homogeneous, isotropic materials such as low-alloy steels and aerospace-grade aluminum and titanium alloys, and offers only recommendations for others. On a composite, no datasheet replaces a test on a sample with known flaws, ideally your own.

Software and data: where time goes after the scan

Software weighs on inspection cost at least as much as hardware. What counts is the time it takes to build a scan plan and check its coverage, how quickly the analyst measures and classifies indications, and whether the report carries everything the standard requires. Two checks prevent surprises after purchase. The first concerns PC-based analysis: included, or under a separate license, and for how many seats. The second concerns the format in which raw data are exported, and who will be able to open them in ten years, when the plant asks for a comparison with today's campaign.

Instrument verification: what to ask for under the ISO 18563 series

A phased array instrument is verified at three stages, and the ISO 18563 series keeps them clearly apart. Part 1, 2022 edition, defines the functional characteristics of multichannel instruments and the methods for measuring them between 0.5 and 10 MHz, and partly applies to FMC and TFM instruments as well. Part 2, from 2024, covers characterization tests on probes at the end of manufacture. Part 3, also from 2024, verifies the complete system, instrument plus probe, before testing or to rule out degradation over time, and states that these checks do not prove suitability for a specific application: they prove that the system operates correctly with the settings used.

Turned into a purchase order, that means asking for an instrument verification certificate to Part 1 on delivery, a characterization certificate for each probe to Part 2, the blocks and instructions for periodic checks under Part 3, and a contact in Italy for the annual verification. These lines belong in the order, not in a request after delivery. On the ASTM side, guide E2491-23 plays a similar role, with one difference worth knowing: it collects methods for evaluating performance and sets no acceptance criteria, which remain to be agreed with whoever commissions the inspection.

Who signs the report: ISO 9712 and phased array training

No instrument makes up for an unprepared operator. When an inspection carries contractual weight, the construction code, product standard or specification almost always calls for personnel certified to ISO 9712:2021, which allocates responsibilities precisely. Level 2 selects the technique, sets up and verifies the equipment, interprets and evaluates results against the applicable standards, and translates codes and procedures into written instructions, while Level 3 establishes and validates procedures, interprets standards and specifications, and designates the methods to be used.

ISO 9712 certifies by method, and the method is UT. Phased array and TOFD also require technique-specific training, which technique standards and specifications often ask to be documented, plus a period of hands-on practice with the software before set-up times come down to production levels. A well-sized purchase budgets for that path along with the instrument case.

Four buying mistakes NDT departments keep repeating

The first is comparing bids on the channel label, ignoring groups, conventional channels, probes and wedges: two 16:64 instruments can behave very differently on the same joint. The second is buying a weld inspection system to do what is in fact daily thickness gauging, or the other way round. The third concerns encoders and scanners, left until last and found to be indispensable at the first specification that asks for recorded data. The fourth costs more than the others because it is invisible: leaving analysis software, periodic verification and training out of the budget.

Then there is the case where buying makes no sense at all. If phased array inspections are occasional, a qualified outside service costs less than an idle instrument and an operator whose certification has to be kept current. If the job is measuring remaining wall at predefined points, a thickness gauge will do. Phased array pays for itself when volumes, acceptance criteria and the demand for recorded data are already in your contracts, or about to be.

The practical rule comes down to one line: the right configuration follows from the inspection to be performed and the standard that will judge it, and it is proven on the part before you sign. For a first look at techniques, probes and configurations, see the page on industrial ultrasonic testing instruments for PAUT and TOFD. If you already have material, thicknesses, geometry and governing standard to hand, send them over: within 24 hours you receive a proposed configuration with a tailored quote, and we can arrange a test on your real component.