-
1Cognitive / UnderstandingExplain why surface-breaking fatigue cracks are invisible to direct visual inspection and describe the capillary action physics that allow liquid penetrant dye to amplify crack width into a visible indication.
-
2Analytical / EvaluationSelect the correct NDT method for a given material, defect type, and field condition — specifically distinguishing when Magnetic Particle Testing (MT) cannot be used and why Liquid Penetrant Testing (PT) is the correct alternative.
-
3Computational / ApplicationApply the UT wall loss formula to calculate pipe wall thinning percentage from actual thickness measurements and classify each result as acceptable, monitor, or scrap-limit condition.
A conveyor system exhaust fan is flagged for annual mechanical inspection. Visually, the aluminum hub looks clean — no gouges, no obvious cracks, no discoloration. A technician passes a flashlight over every surface. The hub looks fine. He is about to sign off.
He decides to run a dye penetrant check anyway, following protocol for rotating components on overhead duty.
He cleans the surface, applies the red penetrant aerosol, lets it dwell 12 minutes, wipes the excess, and sprays white developer. Within 90 seconds, a bright red bleed-out line appears — running radially from the bore, crossing half the hub width. A fatigue crack. Under the developer, it looks like a red river across the white field.
Engineering confirmed the crack had propagated across 48% of the hub cross-section. Fracture mechanics analysis estimated full hub separation within days of continued operation. The fan would have shattered at 1,450 RPM, sending hub fragments through the fan housing.
Visual inspection missed it entirely. The crack was sub-micron width at the surface — below the resolution limit of the human eye under any lighting. Dye chemistry and capillary action amplified it into something a technician could see and act on.
Non-Destructive Testing (NDT) is the family of inspection methods that detect material defects without altering, cutting, or damaging the component being tested. The part remains in service-ready condition after testing. This distinguishes NDT from destructive testing (tensile testing, sectioning, hardness indentation) which sacrifices the part to gather data.
NDT methods divide into two physics-based categories based on where the target defect lives in the material:
Cracks open to the outer skin of the material. The crack path reaches the surface, allowing chemical or physical interaction with test agents.
Primary method: Liquid Penetrant Testing (PT) — dye enters crack via capillary action
Voids, inclusions, or wall thinning beneath the surface. No surface opening. Cannot be reached by chemistry — must be detected by projecting energy through the material.
Primary method: Ultrasonic Testing (UT) — sound waves time-of-flight through material thickness
Selecting the wrong category of NDT method for the defect type produces a false-clear result — the test runs normally, finds nothing, and the technician incorrectly concludes the component is defect-free. Method selection is the highest-stakes decision in any NDT inspection.
Physics: Capillary action. A liquid with very low surface tension and high wetting ability is drawn into any surface-open discontinuity by capillary pressure — the same mechanism that draws water up a paper towel fiber. The finer the crack gap, the stronger the capillary pull.
Works on any non-porous material — aluminum, steel, titanium, ceramics, plastics. Material does not need to be magnetic.
- Only detects surface-breaking cracks
- Cannot find subsurface voids or wall thinning
- 3-can aerosol kit — highly portable
- Low cost, no calibration equipment required
Physics: Magnetic flux leakage. An electromagnetic yoke induces a strong magnetic field through the ferromagnetic workpiece. Where a crack breaks the internal flux path, field lines are forced to exit the metal surface — creating an external leakage field above the crack. Fine iron powder applied to the surface is attracted to the leakage zone, forming a visible powder indication.
FERROMAGNETIC ONLY — carbon steel, cast iron, low-alloy steel. Will NOT work on aluminum, copper, brass, austenitic (300-series) stainless, or any non-magnetic material.
- Catches both surface AND near-surface cracks
- Must test at two perpendicular (90°) orientations
- Requires demagnetization after inspection
- Cannot inspect while energized — LOTO required
Physics: Acoustic wave time-of-flight. A transducer pressed against the metal surface emits a high-frequency sound pulse into the material. The pulse travels through the metal at a known velocity, reflects off the back wall, and returns to the transducer. The time elapsed converts directly to thickness. A void, crack, or internal flaw creates an early reflection.
Works on most solid metals. Requires couplant gel between the transducer face and the metal — any air gap reflects 100% of the sound wave, producing a false reading. Standard for pipe wall thickness checks.
- Detects internal voids and wall thinning
- Single-side access — no need to reach back wall
- Handheld gauge, field-portable
- Couplant gel mandatory — no dry contact reads
Physics: Persistence of vision / synchronization. A xenon or LED strobe is tuned to fire at exactly the same frequency as a rotating or reciprocating machine component. When synchronized, the component appears frozen or in slow motion despite running at full speed. Technicians can observe belt wear, coupling misalignment, timing, and cyclic defects without stopping the machine.
Works on any material, any speed. No surface prep required. Used for running-state inspection only — not for dimensional or crack detection.
- Inspects running machinery without shutdown
- Detects belt slip, timing drift, unbalance signatures
- Xenon or LED unit — variable frequency range
- Full physical guard clearance MANDATORY — see safety
tnominal — Original specified wall thickness (from pipe schedule or engineering drawing), inches
tactual — Measured wall thickness at current inspection point, inches (from UT gauge)
Wall Loss % — Percentage of wall thickness lost to corrosion, erosion, or mechanical damage
Use this table to determine the correct NDT method for a given inspection task. Material constraint is the first filter — applying MT to aluminum produces a false-clear result every time, regardless of technique quality.
| Method | Material Constraints | Target Defects | Equipment Used |
|---|---|---|---|
| Liquid Penetrant (PT) | Any non-porous material — aluminum, steel, titanium, ceramics | Surface cracks, weld porosity, fatigue cracks open to surface | Aerosol 3-can kit (cleaner / penetrant / developer) |
| Magnetic Particle (MT) | Ferromagnetic ONLY — carbon steel, cast iron, low-alloy steel | Surface and near-surface cracks, weld toe cracks | Electromagnetic yoke + iron powder or wet suspension |
| Ultrasonic (UT) | Most solid metals with accessible surface | Internal voids, pipe wall thinning, laminations, back-wall flaws | Handheld UT gauge + couplant gel + calibration block |
| Industrial Stroboscope | Any material — inspects while running at speed | Cyclic defects, belt slip, coupling timing, imbalance signatures | Xenon or LED strobe unit, variable frequency |
Error 1: Over-Cleaning After Dwell (PT)
After applying red penetrant and completing the dwell period, some technicians spray solvent directly onto the inspection surface to remove excess dye. This washes the dye OUT of the crack, destroying the indication before developer can draw it up. The developer then shows nothing. False-clear result.
Correct technique: Remove excess dye with a dry cloth only — wipe gently in one direction. If the surface is heavily contaminated, a cloth lightly dampened with solvent (not saturated) may be used for one wipe only. Never spray solvent directly onto the inspection face after dye has been applied.
Error 2: Parallel Yoke Orientation (MT)
Magnetic particle testing only detects cracks that are oriented perpendicular to the magnetic flux direction. If the yoke legs are positioned parallel to a crack — meaning the flux runs alongside the crack length — the crack causes zero flux leakage and the iron powder shows no indication. The inspector walks away with a clean report on a cracked part.
Correct technique: Every MT inspection requires a minimum of two yoke orientations at 90° to each other — one pass oriented lengthwise along the weld, one pass across it. This ensures any crack orientation produces flux leakage in at least one pass.
Error 3: UT Dry Contact — No Couplant (UT)
A transducer pressed against dry, rough metal produces near-total sound reflection at the contact interface — air trapped in surface roughness gaps reflects the ultrasonic pulse before it enters the material. The gauge reads a false near-surface echo, producing meaningless or absent back-wall readings. Technicians may misinterpret this as a thick reading, or the gauge may freeze on an error code.
Correct technique: Apply a bead of UT couplant gel (ultrasonic gel, petroleum jelly, or water for non-porous surfaces) to the transducer face or inspection area before contact. The gel fills all surface roughness gaps, allowing acoustic energy to transfer efficiently from transducer into metal.
Pipe Wall Loss Analysis — 5-Node Inspection
IE-4-14-01The following 10-step sequence is the standard LEO field protocol for liquid penetrant weld inspection. Each step has a specific technical reason — skipping or abbreviating any step introduces a known failure mode.
- PPE On. Nitrile gloves, safety glasses, and a half-face respirator with organic vapor cartridges. Penetrant aerosols contain acetone and other VOCs. Provide adequate ventilation.
- Surface Preparation. Clean the weld and surrounding heat-affected zone with solvent cleaner applied to a clean rag. Follow with wire brush (stainless bristle, not carbon steel which deposits iron). Objective: bare metal, free of scale, paint, grease, and moisture. The surface must be completely dry before applying penetrant — moisture in the crack prevents dye entry.
- Apply Penetrant Uniformly. Hold aerosol can 8–12 inches from surface. Apply a wet, even coat of red penetrant across the entire inspection area plus 2 inches margin on all sides. The entire zone must be wetted — missed spots produce false-clear.
- Dwell — 10 to 15 minutes. Do not touch, move, or disturb the part during dwell. The penetrant must remain wet on the surface throughout the dwell period — if it dries, wipe off and re-apply. Longer dwell times (up to 30 min) are acceptable and increase sensitivity for fine cracks.
- Remove Bulk Excess — Dry Cloth First. Wipe the surface with a clean, dry, lint-free cloth. Use single light passes — do not scrub. Remove as much surface dye as possible before any solvent contact. This step is the most commonly botched in the field.
- Final Surface Clean — One Lightly-Dampened Wipe. Fold a clean cloth, apply a small amount of solvent cleaner to the cloth (not the surface), and wipe the inspection area once in one direction. Do not re-wipe, do not saturate. The surface should appear clean with no red residue. If red remains, one additional dry wipe is acceptable.
- Apply White Developer. Hold can 8–12 inches from surface. Apply a thin, even, uniform white coat. Avoid puddling — thick developer masks indications. The goal is a translucent white film, not an opaque heavy coat.
- Observe — 5 to 10 minutes. Watch the developer surface closely during the development period. Dots = porosity (trapped gas voids in weld). Lines = cracks or lack of fusion (planar defects). A rapidly bleeding, growing indication in the first 60 seconds is a large, open defect requiring immediate escalation.
- Photograph and Document. Photograph every indication with a scale ruler in frame. Log indication type, location, orientation, and size in the inspection portal before leaving the area. Memory is not an acceptable substitute for field documentation.
- Post-Inspection Treatment. Remove developer and residual penetrant with solvent wipe. Apply rust-preventative oil to bare ferrous metal surfaces to prevent flash corrosion. Log inspection result — clean or indication found — in the maintenance portal with photograph attached.
The stroboscope makes moving machine parts appear completely motionless. A shaft spinning at 1,750 RPM, a fan running at 1,200 RPM, and a belt moving at 3,000 FPM are all capable of producing a "frozen" visual image that is indistinguishable from an actually stopped machine.
NEVER reach past a machine guard, insert hands near a component, or attempt to touch, wipe, or measure a part based on its visual appearance under strobe illumination. The component is at full operating velocity behind the "frozen" image. Contact will cause immediate severe injury.
Maintain full physical distance from all rotating and reciprocating parts during stroboscopic inspection. All guards must remain fully in place throughout the inspection. Stroboscope inspection is a visual-only observation technique — no physical contact with the machine is permitted while the strobe is active or the machine is energized.
- Liquid penetrant inspection reveals a crack indication that crosses the full throat of a structural weld — full cross-section involvement requires engineering fitness-for-service evaluation before any continued operation.
- Ultrasonic thickness measurement shows wall loss at or above 25% of the nominal wall at any single node — this is the LEO scrap limit threshold. Isolate the pipe or component immediately. Do not bypass or cap and continue — place out-of-service tag and notify supervisor before leaving the area.
When in doubt about whether an indication meets the reject threshold, the answer is escalate. The cost of a false reject is a maintenance delay. The cost of a missed critical indication is a catastrophic failure event. These are not comparable outcomes.