Phased array has been the default answer to almost every ultrasonic question for about a decade now, and that is a problem — not because it is a bad technique, but because "we'll PAUT it" has quietly become a substitute for deciding what the inspection is actually for.
Both methods put ultrasound into a component and interpret what comes back. The difference is in how the beam is produced, what record survives the inspection, and what that costs you. Here is how we choose between them, including the cases where conventional ultrasonics is still the right call.
What actually differs
Conventional ultrasonic testing uses a single-element probe. The crystal fires, the beam goes in at a fixed angle determined by the wedge, and the operator moves the probe across the surface to interrogate the volume. What comes back is an A-scan: an amplitude-against-time trace that the technician reads live.
A phased array probe replaces that single element with an array of small ones — commonly 16, 32 or 64 — fired in a controlled sequence. By delaying each element's pulse by fractions of a microsecond, the wavefronts constructively interfere into a beam that can be steered through a range of angles and focused at a chosen depth, electronically, without moving the probe.
That is the whole technical difference. Everything else follows from it.
Coverage and speed
A single-element probe examines the volume the operator sweeps it across, at the one angle the wedge provides. Full coverage of a weld means multiple passes at multiple angles, and the completeness of that coverage depends on operator technique.
A phased array probe performs a sectorial scan — sweeping electronically through a range of angles, typically 40 to 70 degrees, at every position along the weld. One pass interrogates the volume from many angles at once. On a long run of similar welds, the throughput difference is substantial, and it grows with wall thickness because a thick section needs more conventional passes to cover.
The speed advantage is real but it is often overstated on small jobs. Setup for PAUT — wedge selection, focal law calculation, calibration on a representative block, encoder setup — is materially longer than picking up a shear wave probe and a calibration block. On a handful of small-bore welds, conventional is often on and off site before phased array has finished calibrating.
The record is the real difference
This is the argument that matters most and gets discussed least.
A conventional manual UT inspection produces a report. The evidence is the technician's interpretation of traces that no longer exist. If someone disputes a call six months later, or a failure investigation wants to know whether a particular flaw was present at fabrication, there is nothing to re-examine.
An encoded phased array scan produces a data file. Position is recorded against the weld, every A-scan is retained, and the result can be reopened, re-analysed, re-gated and reviewed by a third party years afterwards. You can hand it to a client's Level 3. You can compare it to a scan taken at the next outage and look at whether something has grown.
For safety-critical or high-consequence work, that auditability is usually the reason to specify PAUT, not the coverage. It changes the inspection from a professional opinion into evidence.
Note the qualifier though: encoded. Phased array used free-hand, without an encoder, produces pretty images and no positional record. It is faster than conventional and better at visualising, but it does not give you the archive. If you are paying for PAUT because you want the data, specify encoded scanning explicitly.
Detection and sizing
The ability to steer the beam means badly oriented flaws are more likely to be found. A planar defect that reflects poorly at 45 degrees may return a strong signal at 60, and a sectorial scan looks at both without the operator having to anticipate which is needed. Lack of side-wall fusion in a narrow-gap weld is the classic case — a defect almost parallel to the fusion face, which conventional UT at a single angle can miss and which shows clearly across a swept range.
Sizing is where phased array pulls ahead most clearly. Amplitude-based sizing with a single element depends on the flaw's orientation and reflectivity; a small well-oriented flaw and a large badly-oriented one can return similar amplitudes. Phased array supports tip-diffraction and image-based sizing that measure through-wall extent more directly. When your acceptance criteria are height-based — which fracture-mechanics-driven criteria usually are — that difference decides whether a component is accepted or cut out.
Conventional UT remains entirely capable of detection. The gap is not that it misses things a competent technician would find. The gap is in how confidently the thing found can be characterised and measured.
Where conventional is still the right answer
We specify conventional ultrasonics regularly, and not because of cost.
Thickness measurement and corrosion mapping at a point. A compression probe reading remaining wall thickness needs no beam steering. Phased array adds nothing except setup time.
Simple geometry, thin sections, straightforward acceptance criteria. A 10 mm butt weld with amplitude-based acceptance is well within conventional capability, and the phased array setup will cost more than the inspection saves.
Access-restricted work. Phased array wedges are larger than single-element ones. Where there is 40 mm of clearance beside a nozzle, a small conventional probe gets in and an array does not.
Small quantities on short notice. Mobilisation and setup dominate the cost on a two-weld callout. Conventional gets a decision faster.
Where the code says so. Some fabrication codes and client specifications still name conventional techniques, and using a different method requires a documented equivalence or qualification. That is a conversation worth having, but it is a conversation, not a substitution you make on the day.
What both share
Neither method is better than the person operating it. Both depend on a technician certified to ISO 9712 or PCN at the appropriate level, working to a written procedure, on equipment calibrated against a representative block. Phased array in particular is unforgiving of poor focal law setup — a scan plan that does not actually cover the volume produces a confident, complete-looking image of the wrong thing.
Both also depend on couplant, surface condition and access. A weld cap that has not been dressed limits what either method can do, and no amount of electronics compensates for a probe that cannot reach the volume of interest.
What about TOFD?
Time of Flight Diffraction comes up in most conversations about phased array, usually as though it were a third option. It is better understood as a complement.
TOFD uses a pair of probes either side of the weld, one transmitting and one receiving, and works on the weak diffracted signals from the tips of a flaw rather than on reflected amplitude. Because it measures tip positions directly, it is exceptionally good at through-wall sizing and at detecting flaw growth between inspections.
Its weaknesses are specific and well known. It has dead zones near the surfaces — a shallow near-surface band and a smaller one at the back wall — where flaws are difficult or impossible to resolve. And the signals are low amplitude, so interpretation is more specialised.
Because those dead zones are precisely where a pulse-echo technique performs well, TOFD and phased array are commonly specified together: PAUT covering the near-surface volume and providing the imaging, TOFD providing accurate through-wall sizing in the body of the weld.
So the question is rarely "PAUT or TOFD". It is whether the sizing accuracy justifies adding TOFD to a phased array scope — and on thick-section, fitness-for-service or in-service monitoring work, it usually does.
Choosing
The decision usually resolves on four questions.
Do you need a permanent, re-examinable record? If yes, encoded phased array. This is the deciding factor more often than any other.
Are your acceptance criteria height-based? If yes, phased array — its sizing is what those criteria assume.
Is the flaw type you are worried about badly oriented? Narrow-gap welds, side-wall fusion, service-induced cracking from an unpredictable direction — phased array.
Is it a short job, simple geometry, tight access, or thickness measurement? Conventional, and don't feel short-changed about it.
Volume matters too. Across a long production run the phased array setup cost amortises and the throughput advantage compounds. Across three welds it does not.
The honest summary
Phased array is the better technique for most weld inspection where a data record has value, and it is genuinely superior at sizing. Conventional ultrasonics is not obsolete, is often the faster and more sensible answer on small or awkward work, and remains fully capable in the hands of a certified technician.
The wrong reason to choose either is that it is what the contractor happens to own. Ask what the inspection is protecting against, what evidence you need to keep, and how the result will be judged — and the method usually picks itself.
If you have a specific scope and are unsure which way it falls, our Level 3 team will tell you straight, including when the cheaper option is the right one.
