Selecting the correct non-destructive testing (NDT) method for a weld inspection is one of the most consequential engineering decisions in quality assurance. The wrong choice can leave critical defects undetected — missed lack-of-fusion flaws that radiography routinely fails to detect, or surface-breaking stress corrosion cracks invisible to a volumetric ultrasonic scan. Every NDT method has a distinct physical basis, a characteristic set of detectable defect types, a required access geometry, material constraints, applicable code requirements, and a specific sensitivity threshold below which defects become undetectable. This tutorial provides the systematic decision framework needed to match the right method — or combination of methods — to the inspection requirement, from the first principles of each technique through to code-specific acceptance criteria and personnel certification requirements.

The Six Principal NDT Methods: Physics and Capabilities

NDT Method Capability Matrix

Y = Generally effective | M = Method-dependent / limited applicability | N = Not applicable | *TOFD has near-surface dead zone (lateral wave region) of typically 6–15 mm depth.

Detailed Method Physics and Selection Factors

Ultrasonic Testing (UT): Angle Beam and Straight Beam

Ultrasonic testing transmits high-frequency sound waves (typically 1–10 MHz for weld inspection) into the material through a piezoelectric transducer coupled to the work surface with a couplant gel. The fundamental detection mechanism is acoustic impedance mismatch: when a sound wave encounters a boundary between two materials with different acoustic impedance (Z = ρ × c, where ρ is density and c is sound velocity), a portion of the energy is reflected. Defects (cracks, voids, inclusions) present such boundaries and produce reflected signals detected by the same (pulse-echo) or a separate (pitch-catch) transducer.

Key UT physical relationships:

Acoustic impedance:
  Z = ρ × c   (kg/m²s = Rayls)
  Z_steel ≈ 46 MRayl,  Z_aluminium ≈ 17 MRayl,  Z_water ≈ 1.5 MRayl

Reflection coefficient (normal incidence):
  R = (Z₂ − Z₁)² / (Z₂ + Z₁)²
  R for steel/air = (46×10⁶ − 0)² / (46×10⁶ + 0)² ≈ 1.0  (total reflection)
  → Cracks (air-filled) reflect essentially 100% of incident energy

Wavelength (governs minimum detectable defect size):
  λ = c / f
  For 5 MHz in steel (c = 5920 m/s): λ = 5920/5×10⁶ = 1.18 mm
  Minimum detectable defect ≈ λ/2 = 0.6 mm at 5 MHz

Snell's law at wedge-material interface:
  sinθ₂ / sinθ₁ = c₂ / c₁
  → For plastic wedge (c₁≈2350 m/s) and steel (c₂≈5920 m/s):
     θ₁ = 24.3° → θ₂ = 45° shear wave in steel  (most common angle for welds)
     θ₁ = 30.9° → θ₂ = 60° shear wave
     θ₁ = 38.5° → θ₂ = 70° shear wave

Skip distance (S) for single-V path:
  S = 2T × tanθ   (T = material thickness, θ = refracted angle)
  → Scan position relative to weld centre must cover this geometry

Radiographic Testing (RT) and Digital Radiography (DR)

RT uses ionising radiation (X-rays from an electrical source, or gamma rays from radioactive isotopes such as Ir-192, Se-75, or Co-60) to expose a film or digital detector positioned behind the component. Defects are detected because they attenuate the radiation differently from the surrounding sound metal: voids and cracks transmit more radiation (appear darker on film); inclusions and tungsten deposits may attenuate more (appear lighter). The fundamental limitation of RT is that it produces a two-dimensional projected image of a three-dimensional volume: a planar defect oriented parallel to the radiation beam may produce little or no image contrast and will be missed.

RT key parameters:

Linear attenuation coefficient μ:
  I = I₀ × exp(−μt)   (radiation intensity through thickness t)
  μ_steel ≈ 24 cm⁻¹ (for 200 kV X-rays)

Inherent sensitivity (minimum detectable defect):
  2% of wall thickness (minimum) per ASME Sec V Art. 2 / ASTM E94
  → 2% of 25 mm = 0.5 mm minimum detectable through-wall extent
  → Planar defects must be oriented within ±15° of beam direction to be detected

Image quality indicators (IQI/Penetrameters):
  Wire type (EN 462-1 / ASTM E747): wire diameters from W1 (0.032mm) to W19 (3.2mm)
  Essential wire = minimum visible wire defining radiographic sensitivity
  Hole type (ASME BPVC): 2% thickness penetrameter with 2T or 4T holes

Isotope selection vs. material thickness:
  Se-75 (300 keV):      4–40 mm steel
  Ir-192 (612 keV avg): 12–100 mm steel (most common pipeline/vessel work)
  Co-60 (1.25 MeV):     50–200 mm steel (heavy sections)

Magnetic Particle Inspection (MT)

MT is applicable only to ferromagnetic materials (carbon steel, low-alloy steel, some duplex stainless steels) and detects surface-breaking and near-surface defects by the principle of flux leakage. When a magnetic field is applied to a ferromagnetic component, a discontinuity at or near the surface disrupts the flux path and causes flux to leak from the component surface at the discontinuity. Magnetic particles (dry powder or fluorescent wet particles in suspension) are attracted to the flux leakage region, forming a visible indication of the defect location. The technique is highly sensitive for surface-breaking cracks — detectable even when the crack is tightly closed — provided the magnetic field direction is within 45–90° of the crack orientation. At least two magnetisation directions 90° apart are required for complete coverage.

Dye Penetrant Testing (PT)

PT exploits capillary action to detect surface-breaking defects in any non-porous material regardless of magnetic properties. The process sequence (ASME Sec V Art. 6) is: (1) surface cleaning; (2) penetrant application and dwell time (10–30 minutes for visible dye, 5–10 minutes for fluorescent); (3) excess penetrant removal; (4) developer application; (5) inspection under white light (visible) or UV light (fluorescent); (6) post-inspection cleaning. PT is quantitatively less sensitive than MT for near-surface defects but is universally applicable to all non-porous materials including austenitic stainless steels, aluminium, titanium, and ceramics, where MT cannot be used.

Phased Array Ultrasonic Testing (PAUT)

PAUT uses an array of piezoelectric elements (typically 16–128 elements) that can be individually activated with controlled time delays to electronically steer and focus the ultrasonic beam without moving the probe. This enables real-time generation of multiple beam angles (a sectorial or “S-scan” sweeping from e.g. 40°–70°) in a single probe position, providing simultaneous coverage of the full weld cross-section volume. Combined with encoder-controlled linear scanning, PAUT produces a full volumetric record of the weld that can be post-processed, analysed off-line, and archived for future comparison — replacing the subjective strip-chart record of conventional manual UT.

PAUT beam steering (phased array principle):

Time delay for steering angle θ in an array with element pitch d:
  Δt_n = n × d × sinθ / c   (n = element number from 1 to N)

Focal depth F (for focusing):
  F controls depth of maximum amplitude response
  F = N × d² / (4λ) (approx. for unfocused → focused in Fresnel zone)

Advantages over conventional UT:
  1. Single probe position covers multiple angles → faster scanning
  2. S-scan image provides defect sizing directly in 2D cross-section view
  3. Full encoded waveform data stored → reproducible, auditable record
  4. Beam focusing improves resolution at depth
  5. Eliminates need for multiple probe changeovers during manual UT

Required qualification (ASME):
  - Procedure qualification per Appendix VI to Article 4 (Section V)
  - PDI (Performance Demonstration Initiative) qualification for critical applications
  - Level II minimum for examination; Level III for procedure development

Time-of-Flight Diffraction (TOFD)

TOFD is a specialised UT technique using two probes straddling the weld: one transmitter (Tx) and one receiver (Rx), both inclined at an angle to maximise lateral-wave coverage and diffraction signal detection. Unlike pulse-echo UT, TOFD detects diffracted signals from defect tips rather than reflected signals from defect faces, making it essentially orientation-independent for linear defects. The through-wall height of a defect is measured directly from the time-of-flight difference between the upper and lower tip diffraction signals:

TOFD defect sizing:

Depth to upper tip:
  d_upper = √[(t_upper × c/2)² − (s/2)²]

Depth to lower tip:
  d_lower = √[(t_lower × c/2)² − (s/2)²]

Defect height:
  h = d_lower − d_upper

Where:
  t_upper = time of arrival of upper tip diffraction signal
  t_lower = time of arrival of lower tip diffraction signal
  c = longitudinal wave velocity ≈ 5920 m/s (steel)
  s = probe separation (Tx-Rx centre-to-centre distance)

Sizing accuracy:  ±1–2 mm through-wall height (vs. ±3–5 mm for conventional UT)

TOFD dead zones:
  Near-surface dead zone (lateral wave region): ~3–15 mm depth from surface
  → PAUT or MT/PT required for surface zone coverage
  Near-backwall dead zone: ~3–8 mm from backwall
  → Additional backwall scans or PAUT required

TOFD + PAUT combination is the standard practice for:
  - ASME Code Case 2235 (alternative to RT for fusion welds)
  - EN ISO 13588 (advanced UT of austenitic materials)
  - API 1104 Section 9 alternative examination

Code Requirements and Acceptance Criteria

ASME Code Case 2235: Alternative to RT Using UT

ASME Code Case 2235 (and its subsequent revisions) allows the use of ultrasonic examination as an alternative to radiographic examination for fusion welds in pressure vessels under Section VIII, provided: (1) the UT procedure is qualified by demonstrating detection of representative defects in a test block representative of the production weld geometry; (2) the examination uses encoded scanning (mechanised) with full data recording; (3) the procedure has been qualified by a PDI (Performance Demonstration Initiative) blind test demonstrating the required probability of detection. Code Case 2235 was a pivotal development that enabled the transition from RT to PAUT+TOFD in heavy-wall pressure vessel fabrication, where RT is impractical for thick sections and provides no sizing capability for planar defects.

PAUT vs. TOFD vs. RT: A Direct Comparison

NDT Personnel Certification and Quality System Requirements

NDT examination results are only as valid as the personnel performing and interpreting them. All major codes specify minimum certification levels for NDT personnel based on the SNT-TC-1A (ASNT) framework, ASME CP-189, or EN ISO 9712 depending on jurisdiction. The three levels define progressively increasing capability and authority:

NDT Certification Level Summary (SNT-TC-1A / ASME CP-189):

Level I:
  Qualified to perform specific calibrated NDT operations
  Must work under direct supervision of Level II or III
  Cannot independently interpret or accept results
  Training: 40 hours minimum (UT), examination, practical test
  Experience: 400 hours minimum on-the-job experience

Level II:
  Qualified to set up, calibrate, and perform examinations independently
  Interprets and evaluates indications against applicable acceptance criteria
  Prepares written examination reports
  Can train and supervise Level I
  Training: 80 hours minimum (UT); 40 hours additional for PAUT/TOFD specialty
  Experience: 800 hours minimum; medical vision acuity (Jaeger J1 near-vision)

Level III:
  Qualified to establish procedures, interpret codes, and designate methods
  Approves written examination procedures for use under ASME/API codes
  Qualifies procedures and personnel
  Can prepare and review NDT reports submitted to AIJ (Authorised Inspection Agency)
  Training: Level II experience + Level III general, specific, and practical exams
  No minimum hours specified by SNT-TC-1A, but typically 4+ years Level II experience

For ASME-stamped equipment:
  - NDT procedures must be approved by an ASME-certified Level III
  - AI (Authorised Inspector) from the AIA (Authorised Inspection Agency, typically
    an insurance company holding ASME Certificate of Authorisation) must witness
    or accept examination records
  - NDE manual must document procedure number, revision, and personnel qualification

For the underlying weld metallurgy context that determines which defects are most likely in specific weld joints and procedures, see the HAZ Microstructure and Hydrogen-Induced Cracking articles. The Charpy Impact Test and Hardness Testing Methods describe the destructive testing methods that complement NDT in weld procedure and welder qualification. For the corrosion defects that in-service NDT must detect, see Corrosion Mechanisms and Pitting Corrosion.