20.1 Applications: Matching Method to Product Form and Access

The ASNT NDT Level II specific exam — 40 items, 2 hours, General Industry or Pressure Equipment — lists Applications as an official topic next to Codes and Techniques. The general paper asks whether a method can see a class of discontinuity. The specific paper hands you a product, an access problem, and a plant constraint, then asks which method is legal here. A candidate who recites "ultrasound finds cracks" and then picks radiography on a buried line with no film side will miss Applications items even with a perfect general-exam score.

This section is the job-matching chapter. It does not replace the method-selection logic in the fundamentals chapter. It applies that logic to the six product-and-access pictures the specific exam keeps using: a fillet-weld toe in fatigue, an austenitic stainless weld, an aluminum casting, a clad vessel, a buried pipe, and a painted in-service tank. After the pictures, the same five constraints — access, temperature, cleanliness, magnetism, and radiation control — decide whether the pretty method on the traveler can actually be run.

How an Applications stem is built

A well-written Applications item names four facts and hides a fifth.

1. Product form — fillet weld, butt weld, casting, forging, clad plate, pipe, tank shell.
2. Expected discontinuity — toe fatigue, lack of fusion, shrinkage, disbond, wall loss, cold shut.
3. Access — one face, two faces, excavated OD only, ID only, buried, insulated, painted.
4. Constraint — nonmagnetic alloy, hot surface, dirty or coated surface, no radiation window, no couplant water in the unit.
5. The hidden fact is the governing procedure. If the stem prints a procedure that already chose the method, you apply that procedure. If the stem asks you to choose, you build the choice from the four facts, not from the method you like.

ASME Section V, AWS D1.1, ASTM E1444, ASTM E1417, ASTM E94, and ASTM E164 are industry technique references that encode this matching. They are not unpublished ASNT exam secrets. Attribute the document. Do not invent an ASNT-only application table.

Six jobs, one matching table

Memorize the why in each row, not a slogan. The rest of this section unpacks the six jobs and then the five constraints that knock a method out even when the alloy looks right.

Fillet weld toe fatigue — VT plus MT or PT

Fatigue at a fillet toe is the most common in-service surface crack the specific exam will hand you. Cyclic load plus a stress raiser (the toe, undercut, an abrupt weld profile, a pit) starts a crack at the surface and grows it into the throat or the heat-affected zone. Early fatigue is tight. It is a terrible first radiographic target because the crack plane presents almost no thickness change to the beam.

Start with visual testing. VT finds the toe, the undercut, the paint crack that follows the toe, the arc strike, and the missing blend. It does not, by itself, prove a tight crack is absent. After VT, the surface method is:

• Magnetic particle testing with an alternating-current electromagnetic yoke on ferromagnetic carbon or low-alloy steel. AC concentrates flux at the surface, which is exactly where toe fatigue lives. Two yoke orientations (or a procedure that otherwise covers both directions) are required so a crack parallel to one field still leaks flux in the other.
• Liquid penetrant testing when the steel is painted and the procedure wants PT after local strip, when the fillet is a nonmagnetic alloy, or when the written practice simply names PT. The coating must be addressed first (see the painted-tank job below).

Ultrasound belongs after the surface crack is known, when the procedure wants depth or remaining ligament. It is not the screening tool for twenty feet of fillet. Radiography of a tight toe crack is how a candidate demonstrates that they still think "volumetric" means "sees everything."

Pressure-equipment version of the same job: a carbon-steel nozzle-to-shell fillet after a hydrotest wait. Same VT-plus-AC-yoke picture. The wait exists because hydrogen-assisted cold cracks can appear hours after welding; the method pair does not change.

Austenitic stainless weld — no magnetic particle testing

Type 304 and 316/316L austenitic stainless welds are the specific exam's favorite way to trap a candidate who runs magnetic particle testing on every silver metal. These alloys are not ferromagnetic in the condition that matters. A heat-affected zone can pick up a little ferrite or deformation magnetism; that is not permission to treat the weld as an ASTM E1444 part. Magnetic particle testing is not a valid primary method.

What remains:

• Visual testing of contour, undercut, oxidation, and visible cracks.
• Penetrant testing for surface-breaking cracks, crater cracks, and open lack of fusion. Preclean must not smear the opening shut; the procedure's dwell and removal windows apply.
• Radiographic testing when two-side access exists and the target is volumetric (porosity, slag, tungsten). Tight fusion-line lack of fusion that sits perpendicular to the beam is still a classic radiographic miss.
• Ultrasonic testing for fusion-line and mid-wall planar flaws — but only under a procedure written for austenitic weld metal. Coarse grains scatter shear waves and can skew the beam. Industry practice often moves to a lower frequency, a refracted-longitudinal or dual-element approach, and a calibration that matches the weld, not a carbon-steel IIW block used on autopilot.

If the stem says "316L fillet, AC yoke, dry powder," the application is already wrong. The correct rescue is PT (surface) or a qualified UT/RT technique (volumetric), not a heavier yoke.

Aluminum casting — penetrant and radiography

Aluminum castings are nonmagnetic. Magnetic particle testing is out before you unpack the kit. The discontinuities that matter are inherent to the pour:

• Gas porosity and shrinkage are volumetric. Radiography is the map. Shrinkage is irregular and dendritic in the last region to freeze; gas pores are rounded.
• Cold shuts and hot tears often break the surface. Visual testing and penetrant testing are the surface pair.
• Inclusions (oxide, sand) may be internal or surface-connected. RT finds many internal inclusions; PT finds the open ones.

Ultrasound can find large internal cavities when the as-cast surface can be coupled and the grain is not a scatter fog. Many production aluminum castings are still radiographic jobs for that reason. Do not pick UT as the default volumetric method just because the general exam taught you that ultrasound finds voids. Access, surface, and grain decide.

A General Industry stem that says "aluminum pump housing, irregular cavity in the heavy section, fluorescent penetrant clean on the OD" is asking you to recognize that a clean PT does not clear internal shrinkage. Call for RT (or a qualified UT), not for a second penetrant brand.

Clad vessel — ultrasonic bond testing

Cladding — roll-bonded, explosion-bonded, or weld overlay treated as a bond — exists so a corrosion-resistant layer stays attached to a cheaper substrate. The Applications question is not "is there a crack in the weld cap." It is "is the clad still bonded."

Pulse-echo ultrasonic testing from the accessible face is the industrial answer:

• A good bond lets sound cross the interface. The A-scan shows a substrate back-wall (or a weak interface when impedances match).
• An air-gap disbond reflects at the failed interface. The interface echo grows; the far-side back-wall drops or vanishes. A thin clad layer often rings — multiples at the clad thickness are one unbond, not six defects.

Through-transmission is the two-side version used more on sandwich and honeycomb than on a heavy vessel wall. Kissing bonds (intimate contact, little strength) can look almost bonded; conventional UT under-calls them. That limitation is why the Level II reports the ultrasonic response, not a strength number.

Penetrant or magnetic particle testing see a disbond only when it breaks a surface — a blister edge, a clad stop, a cut face. Radiography of a bond line that lies parallel to the film presents almost no thickness change. Do not pick RT as the clad-bond method.

Pressure-equipment flavor: a strip-clad or overlay-clad channel on a heat exchanger. One-side UT from the clad face, procedure-named frequency (higher frequency to resolve a thin clad), and a calibration that includes a known unbond or a back-wall from the substrate.

Buried pipe — limited radiography, ultrasonic from the OD

A buried line does not give you a film side on the ID and does not give you a 360° source run unless someone excavates a bell hole and the radiation plan allows it. Applications items that say "buried," "under paving," or "in the ditch" are access items first.

What usually works:

• Ultrasonic thickness from the excavated OD for corrosion and erosion. Grid the area the procedure names. This is remaining-wall measurement, not a crack hunt.
• Angle-beam ultrasonic testing of a girth weld from the OD after the coating is cut back and the scanning surface is prepared. You still need the procedure's angles, skips, and calibration.
• Magnetic particle or penetrant testing only on the excavated, prepared OD surface. They do not examine the ID or the soil-covered arc.

Radiography is limited, not forbidden. A double-wall technique can work on pipe when you can place a source and a detector (or film) and control the radiation area in the excavation. You cannot claim a full volumetric RT examination of a line you cannot reach. If the stem says the trench cannot be occupied and the unit cannot be shut down for a boundary, RT is the wrong application even if the weld would radiograph beautifully in a shop.

Painted in-service tank — do not PT or MT until the coating is addressed

Paint, epoxy lining, fireproofing, and insulation are not "thin dirt." They block capillary paths (penetrant never reaches the crack) and they stand off particles from a leakage field (magnetic particle indications weaken or vanish). A Level II who dusts dry powder onto intact epoxy and signs "no relevant indications" has not examined the steel.

Correct application:

1. Visually assess the coating: cracks in the paint that follow a toe, blisters, rust bleed, missing lining.
2. Locally remove the coating over the examination area to bare metal, to the remaining thickness the magnetic-particle or penetrant procedure allows, or as the written procedure states.
3. Run PT or MT on the prepared metal. Restore the coating as the owner requires.
4. Use a coating-tolerant technique only when the governing procedure already qualifies it — some AC-yoke magnetic-particle procedures allow a stated maximum coating thickness after a demonstration. That is a procedure fact, not a field improvisation.

Temperature stacks on top of the coating problem. A tank still in hot service can sit above the penetrant or magnetic-particle process window. Couplant can flash off. The correct Applications answer is "cool it, strip it, or change methods," not "work faster so the penetrant does not dry."

The five constraints that kill a method

Even when the alloy and the discontinuity class look right, one constraint can remove a method. Specific-exam writers love to hide the constraint in a subordinate clause.

Access

Ask "where can I put a probe, a yoke, a film cassette, and my eyes?" before you name a method.

Temperature

Process windows live in the procedure and in the method standards, not in tribal memory. Typical industry pictures — always overridden by the named procedure — include a penetrant part-temperature band in ASTM E1417 (commonly taught around 40–125 °F / 4–52 °C), magnetic-particle media that clump or lose mobility when they are too cold or too hot, ultrasonic couplants that freeze or boil, and film or detector limits. Visual testing on a glowing or steaming surface is not a controlled VT examination. If the stem says the vessel is still at 300 °F, do not pick a standard fluorescent penetrant process.

Cleanliness

Penetrant testing demands a clean, dry, open surface. Oil, scale, paint, and machining smear close the capillary. Magnetic particle testing needs a surface clean enough for particles to collect at leakage fields; heavy scale and thick coatings are disqualifiers unless the procedure says otherwise. Ultrasonic testing needs a coupling surface — loose paint and deep pitting lift the probe. Radiography is the most surface-tolerant of the five, but scatter from heavy scale still hurts contrast. Visual testing needs the surface actually visible: no mud, no glare, no steam.

Magnetism

Magnetic particle testing requires a ferromagnetic examination volume. Carbon steel, most low-alloy steels, many 400-series stainless steels, and cast irons can work. Austenitic stainless, aluminum, copper alloys, titanium, nickel alloys, and nonmetals do not. Residual magnetism after a direct-current shot is a downstream constraint: it can deflect an arc on a subsequent weld or upset a nearby instrument. Demagnetize when the procedure requires it. Magnetism is also why you do not "just try a yoke" on 316L to see if anything lights up.

Radiation control

Radiography is not only a geometry problem. It is a time, distance, shielding, and boundary problem under 10 CFR 20 (U.S. NRC licensees) or the Agreement-State equivalent, plus the employer's radiation-protection program. A live unit, a public sidewalk next to the tank, a trench you cannot evacuate, or a shift that will not grant a shot window can make RT the wrong application even when two-side access exists on paper. The specific exam will offer radiography as the tempting "complete" method. Decline it when the stem has already told you the boundary cannot be set.

General Industry versus Pressure Equipment

The method-matching logic does not change when the candidate sits the Pressure Equipment sector instead of General Industry. The product names change.

• General Industry stems talk about structural fillets, machine housings, aluminum castings, crane welds, and storage tanks under API or owner specifications.
• Pressure Equipment stems talk about boilers, pressure vessels, and process piping under ASME Section V techniques, construction codes such as ASME Section VIII or B31.3, and in-service documents such as the National Board Inspection Code (NBIC) or API 510/570/653.

You still do not magnetic-particle test 316L. You still do not penetrant-test through epoxy. You still do not radiograph a buried line with no film side. You still ultrasonically interrogate a clad bond from the accessible face. Sector language is the costume. Physics and access are the exam.

How to answer the Applications item

Read the stem once for the product and once for the constraint. Cross off any method the material forbids (MT on austenitic or aluminum). Cross off any method the access forbids (conventional RT without a detector side; PT/MT on unprepared coating; ID methods you cannot reach). Cross off any method the discontinuity class wastes (RT of tight fatigue; PT of mid-wall shrinkage). What remains is the application. If two methods remain, the stem's procedure, production-versus-in-service setting, or safety clause will break the tie.

The Level II who treats Applications as a matching exercise — not as a loyalty test to a favorite method — is the candidate this 40-item paper is written to pass.