What gets checked first on site

A weld can look fine from two metres away and still fail once someone gets a torch, a gauge, or an NDT kit on it. That is why the answer to What non-destructive testing methods are typically used during an on-site welding assessment? is usually less glamorous than people expect: start with what you can see, then only escalate when the joint, the spec, or the service conditions demand it.

On site, the methods you actually see most often are:

Visual weld inspection
Dye penetrant testing for surface-breaking flaws in non-magnetic materials
Magnetic particle testing for surface and near-surface flaws in ferromagnetic steel
Ultrasonic testing welding for internal flaws, thickness checks, and some geometry-driven assessments
Radiographic testing welds where the joint, access, and hold points make film or digital imaging the better call

That is the practical answer to What non-destructive testing methods are typically used during an on-site welding assessment? The method is chosen by defect type, material, access, and what the client actually needs documented. Not by habit.

Key takeaway: the right NDT method is the one that can actually see the defect you are worried about, in the access you have, without turning the job into a delay machine.

Visual inspection is not the “easy” option

A lot of people treat visual inspection like a box tick. That is usually where the trouble starts.

Visual weld inspection is the first gate because it tells you whether the weld is even worth a more expensive test. You are looking for profile, undercut, overlap, craters, root reinforcement, spatter that hides the toe, obvious porosity, arc strikes, distortion, and whether the weld prep matches the WPS or job requirement. If the weld is still buried under slag, scale, paint, or mill finish, you do not have a clean starting point for any other method.

On a messy site, visual often decides the sequence. If you cannot clean the cap properly, you may not get meaningful dye penetrant or magnetic particle results. If the joint is buried in a frame, against a wall, or boxed in by pipework, you may not get a usable line of sight for a decent visual call, which is when the job starts to move towards UT or RT.

That is the first practical answer to What non-destructive testing methods are typically used during an on-site welding assessment? Visual is always first, but it is not always enough.

The first method when access is awkward

When access is limited or the joint geometry is awkward, the first method I reach for is usually visual inspection, then I decide whether the next step is surface testing or volumetric testing.

If the joint is open enough to clean and the concern is cracking at the surface, dye penetrant or mag particle is the next move. If the defect could be internal, or the geometry makes surface methods a waste of time, I move to ultrasonic testing welding or, less often on site, radiographic testing welds.

The method gets ruled out fast when:

the surface cannot be cleaned well enough
the weld is too rough or too porous for a reliable surface test
the material is non-magnetic, which rules out magnetic particle testing
access is too cramped for probe placement, film positioning, or source control
the defect you are chasing is internal, not surface-breaking

That is the bit apprentices usually miss. The question is not “which test is best?” It is “which test can still produce a result worth signing my name to?”

Where surface methods stop making sense

Magnetic particle and dye penetrant testing are brilliant until the surface stops cooperating. Then they become false economy.

Dye penetrant is unreliable if the surface is dirty, porous, heavily rough-ground, or covered in paint, oil, rust, or slag residues. It is also a poor choice if you need to get through the job quickly and the joint is still hot, wet, or contaminated from the environment. Penetrant will happily highlight every bit of surface grime if prep is poor, and that creates a mess of false calls.

Mag particle has its own limits. It only works on ferromagnetic material, so stainless and aluminium are out. It also struggles if the surface is too rough, if the geometry creates magnetic leakage issues, or if access is so poor that you cannot get the yoke or prods where they need to be.

When surface methods are not worth the effort, you switch to ultrasonic testing welding for internal flaws or to RT if the geometry and site controls support it. That is why the answer to What non-destructive testing methods are typically used during an on-site welding assessment? is never just “the same one every time”.

The schedule killer is rarely the scan itself

The part that blows out the schedule most is usually access and hold points, not the actual scan.

Surface prep takes longer than people budget for, especially on painted structural steel, hot-dip galvanised material, or welds covered in spatter and grinding marks. But the real delay is when the inspection has to wait on:

scaffolding or EWPs
permit-to-work windows
client hold points
shutdown access
calibration verification
sign-off from someone who is not physically on the job

Calibration matters, but it is rarely the biggest delay unless the gear has not been prepped properly. The jobs that drag are the ones where the inspection team is ready and the weld is not accessible, or the weld is ready and the client is not.

If you are coordinating an on-site welding assessment, sequence the work so the welds are cleaned, exposed, and cooled to the right condition before the inspector arrives. Otherwise you pay twice, once in labour and once in downtime.

UT versus RT in the real world

Ultrasonic testing welding and radiographic testing welds both find internal issues, but they fail in different ways on site. That is why the “best” method depends on the job, not the brochure.

Ultrasonic testing welding

UT is usually the better call when:

access is limited to one side
the material is thick enough to give a meaningful return
you need a faster field method without film handling
the weld can be coupled properly and the geometry is known

UT struggles with coarse-grained material, awkward profiles, and geometry that scatters the beam. A badly crowned cap, a tight corner, or a complex fabrication can make interpretation ugly fast. If the operator cannot get a clean sound path, the result can look like a defect when it is really a setup issue.

Radiographic testing welds

RT is still useful when you need a permanent image and the joint suits it. But on site, it brings more baggage. You need source control, exclusion zones, radiation safety, and enough room to position the source and detector properly. Thin material can also be awkward because the image can be too forgiving or too noisy, depending on the setup.

RT often forces the job to be re-sequenced around the inspection. Production stops while the area is cleared, the exposure is done, and the film or digital image is processed and reviewed. That is why high-production work often leans towards UT, unless the spec or the client specifically wants RT.

If you are asking What non-destructive testing methods are typically used during an on-site welding assessment?, the honest answer is that UT and RT are the heavier tools. They are not interchangeable, and they are not always practical on a live site.

The most expensive planning mistake

The costliest mistake is assuming one method can cover every defect type.

It cannot.

A surface crack, lack of fusion, porosity, slag inclusion, lamellar tearing, and root mismatch do not all show up the same way. A weld can pass visual and still fail UT. It can look ugly and still be acceptable. It can fail dye penetrant because of surface contamination, not because the weld is cracked. That is where people burn money, by booking the wrong method first and then paying for a second inspection after the first one proves useless.

The second expensive mistake is booking NDT before the weld is actually ready. If the cap still needs cleanup, if the joint has not cooled, or if access gear has not been arranged, the inspector ends up waiting. That delay is usually more expensive than the test itself.

For sites in Laverton North, VIC, that kind of sequencing problem is common on busy fabrication floors. In Wingfield, SA, where supply and fulfilment is the focus rather than on-site work, the same lesson still applies to planning material flow and keeping consumables ready before the inspection window opens.

When visual alone is enough

Visual alone can be enough when the weld is low-risk, the acceptance criteria are simple, and the joint is fully accessible. That usually means:

the weld is exposed cleanly
the profile is easy to assess
the spec only requires visual acceptance
there is no reason to suspect subsurface defects
the service conditions are not demanding extra verification

If the weld is in service conditions where you cannot clean, grind, or fully expose the joint, visual becomes a screening tool, not the final answer. Then you escalate based on what the joint can physically support. If the surface can be cleaned enough for a readable result, go to dye penetrant or mag particle. If it cannot, and the concern is internal integrity, go to UT. If the spec requires a record and the setup allows it, RT may be the better fit.

That is the real workflow behind What non-destructive testing methods are typically used during an on-site welding assessment? Visual first, then the least disruptive method that can still prove the point.

The setup mistakes that create fake defects

A lot of “defects” are really testing problems.

With ultrasonic testing welding, the common setup mistakes are:

wrong calibration block or stale calibration
poor couplant coverage
incorrect probe angle
scanning from the wrong side of the weld
not accounting for weld crown, root shape, or material thickness

Those errors can create echoes that look like discontinuities when the issue is really the beam path or coupling. A good operator catches that by checking calibration on a known reference, confirming the weld geometry, and repeating the scan from a second angle before calling a defect.

With radiographic testing welds, the usual traps are:

incorrect source-to-film distance
poor film placement or digital detector alignment
exposure settings that are too hot or too flat
scatter from nearby steel or fixtures
not marking the weld properly before exposure

That can produce density changes, blur, or artefacts that look like porosity or lack of fusion. Experienced teams verify the setup against the technique sheet, check image quality indicators, and compare the indication against weld location and geometry before anyone signs off.

The calls that turn into arguments later

The findings that cause the most back-and-forth are the ones that are real, but not automatically rejectable.

Typical examples are:

minor surface porosity outside the acceptance limit
rounded indications that are actually acceptable under the code
incomplete fill that is cosmetic, not structural
geometry-related echoes on UT that mimic a flaw
film artefacts on RT that sit right where a weld indication would be

This is where documentation matters more than opinion. Record the method used, calibration status, equipment ID, weld location, acceptance standard, the exact indication size or position, and whether the indication was repeatable from another angle or setup. If the result is borderline, note why it was accepted or rejected against the relevant code or spec. That saves a lot of grief later when the welder, inspector, and client all remember the same weld differently.

A practical way to plan the inspection

If you are booking or preparing an on-site welding assessment, do this in order:

Confirm the acceptance standard before the inspector arrives.
Make sure the weld is clean enough for visual inspection.
Decide whether the likely defect is surface-breaking or internal.
Check access, power, lighting, and exclusion zones.
Choose the least disruptive method that can still prove the point.
Have the hold points and sign-off people lined up before the test starts.

That sequence answers the real version of What non-destructive testing methods are typically used during an on-site welding assessment? It is not about using every method. It is about using the right one, once, and getting a result that stands up.

If you are training for tickets or renewing them, WeldCo’s Welding Courses are a practical way to get sharper on inspection expectations, not just arc time. And if you are managing the workshop side of the job, the TradeStore Online Ordering Portal helps trade account customers keep consumables, PPE, abrasives, and machine stock organised with contract pricing and saved favourites, which matters more than people think when an inspection window is tight.

The weld does not care how busy the site is. The inspection still has to be set up properly.

What NDT Methods Are Used in On-Site Welding?