Material testing and non-destructive testing

In short

Material testing measures the mechanical properties of a material — strength, ductility, hardness, toughness, fatigue and creep life — on a sacrificial specimen, while non-destructive testing (NDT) finds flaws in a finished component without damaging it. Together they answer the two questions every manufacturer asks: *is the material right* and *is this particular part sound*. Every pressure vessel, boiler, pipeline, weld, casting, forging, aircraft part and railway axle in India is released only after specified tests under IS, ASME or API codes. The topic carries marks in GATE (Materials/Manufacturing), is a standing interview area for L&T, BHEL, HAL, ISRO, BARC, Tata Steel, EIL, Reliance, Welspun and every third-party inspection agency (TÜV, Bureau Veritas, SGS, RINA), and an ISNT/NCB Level II card in UT or RT is a direct employable qualification.

Explained

1. Why we test, and the two families

• Destructive testing (DT): the specimen is loaded to failure or permanently deformed. Gives *property values* (σy, UTS, E, hardness, impact energy, fatigue limit) used for design and material acceptance. Tested on samples, coupons or sacrificial production parts.
• Non-destructive testing (NDT): the part remains fit for service. Gives *flaw information* (cracks, porosity, slag, lack of fusion, laminations, wall loss) — not property values.
• Why both: DT proves the material grade; NDT proves this one casting/weld has no defect. A plate of IS 2062 E250 can pass tensile test and still carry a lamination.
• Defect vs discontinuity: a *discontinuity* is any interruption in the structure; it becomes a *defect* only when it exceeds the acceptance criteria of the governing code. NDT finds discontinuities; the code decides.
• Destructive tests on welds (IS 3600 series): transverse tensile, root/face/side bend, nick-break, macro-etch, fillet fracture, used in welder and WPS qualification (ASME Sec IX).

2. Tensile test and the stress-strain curve

• Standard: IS 1608 (Part 1), harmonised with ISO 6892-1; ASTM E8/E8M internationally. Round specimen with gauge length Lo = 5.65√So (proportional specimen, ≈ 5d) on a UTM (screw-driven or servo-hydraulic), extension measured by extensometer.
• Curve regions (mild steel): linear elastic (Hooke's law, slope = Young's modulus E) → limit of proportionality → elastic limit → upper and lower yield point (only in low-carbon steel, caused by carbon/nitrogen atmospheres pinning dislocations — Lüders bands) → uniform plastic deformation and strain hardening up to maximum load → necking (plastic instability) → fracture.
• UTS is the maximum load ÷ original area, so the engineering curve falls after necking; the true stress–true strain curve rises continuously because the real area shrinks.
• 0.2% proof stress: materials without a yield point (Al alloys, stainless steel, copper) use the stress at 0.2% permanent strain — a line parallel to the elastic slope offset by 0.002.
• Ductility measures: % elongation (depends on gauge length — always quote it) and % reduction in area (gauge-length independent, better measure).
• Toughness = area under the stress-strain curve (energy/volume); resilience = area under the elastic part.
• Fracture appearance: ductile metals give a cup-and-cone fracture with a fibrous centre and shear lips; brittle fracture is flat, bright and granular, perpendicular to the load.
• Compression test (concrete, cast iron, bearing metals): ductile metals barrel because of platen friction, brittle materials shear at about 45°; torsion test gives shear modulus G and shear strength; bend test (IS 1599) checks ductility and weld soundness by bending through 180° around a former.

3. Hardness, impact, bend, fatigue and creep

• Hardness = resistance to localised plastic indentation. Quick, near-non-destructive, correlates with strength and wear resistance.
  • Brinell (IS 1500): 10 mm hardened steel/tungsten carbide ball, 3000 kgf for steel, 10–15 s dwell; measure indent diameter. Load/D² ratio = 30 (steel), 10 (Cu alloys), 5 (Al), 2.5 (soft metals). Large indent averages over coarse structures — ideal for castings and forgings; not for thin or hardened surfaces.
  • Rockwell (IS 1586): direct dial reading of depth. HRC = 120° diamond brale, 150 kgf major (hardened steel); HRB = 1.588 mm ball, 100 kgf (soft steel, brass). Fast, small indent, used on the shop floor.
  • Vickers (IS 1501): 136° diamond pyramid, one continuous scale for all materials, HV = 1.854F/d²; micro-Vickers and Knoop for coatings, case depth and individual phases.
  • Shore scleroscope / Leeb rebound testers and portable UCI testers are used in-situ on installed plant.
• Impact test — measures toughness at high strain rate with a notch (triaxial stress) present.
  • Charpy V-notch (IS 1757): 10×10×55 mm bar, 2 mm deep 45° V-notch with 0.25 mm root radius, simply supported on a 40 mm span, struck opposite the notch.
  • Izod (IS 1598): 10×10×75 mm bar held as a cantilever, struck on the notch side.
  • Ductile–brittle transition temperature (DBTT): BCC metals (mild steel, ferritic steel) lose toughness sharply below a transition temperature; FCC metals (austenitic stainless, Al, Cu) have no DBTT. This is why low-temperature service steels are Charpy tested at −20 °C, −46 °C etc., and why welded ships and bridges have failed brittly in cold weather.
• Fatigue test: cyclic loading, usually rotating bending (R.R. Moore) or axial. Plot stress amplitude vs cycles to failure = S-N curve. Steels show a true endurance limit at about 10⁶–10⁷ cycles (≈ 0.4–0.5 × UTS); aluminium does not, so a fatigue strength is quoted at 10⁸ cycles. Fatigue fracture shows beach marks / striations from a surface initiation site plus a final overload zone. Surface finish, notches, corrosion and tensile residual stress all cut fatigue life; shot peening and nitriding raise it.
• Creep test: constant load at constant high temperature (> 0.4 Tm); strain–time curve has primary (decelerating), secondary (steady minimum creep rate — the design value) and tertiary (accelerating to rupture) stages. Stress rupture test runs to failure. Matters for boiler tubes, steam turbine blades and gas turbine hot sections.
• Fracture toughness (KIC): ASTM E399 CT or SENB specimen with a fatigue pre-crack; gives the material's resistance to crack propagation under plane strain. Combined with NDT flaw size, it drives fit-for-service / fitness-for-purpose assessment (API 579 / BS 7910).

4. Surface NDT methods

• Visual testing (VT): the first and cheapest method — direct, or with borescopes, mirrors, weld gauges. Finds undercut, overlap, spatter, misalignment, surface cracks. Mandatory before any other NDT; minimum illumination typically 500 lux at the surface (ASME asks ~1000 lux).
• Liquid penetrant testing (LPT / DPT) — IS 3658, ASME Sec V Art. 6:
  1. Pre-clean and dry; 2. apply penetrant and allow dwell (typically 5–30 min); 3. remove excess (water-washable / post-emulsifiable / solvent-removable); 4. apply developer, which blots the trapped penetrant out by capillary action; 5. inspect after development time; 6. post-clean.
  2. Types: visible dye (red indication on white developer, white light) and fluorescent (viewed under UV-A ~365 nm in a darkened booth, far more sensitive).
  3. Only surface-breaking defects; works on any non-porous material including austenitic stainless, aluminium and plastics; fails on porous castings; cannot detect a crack filled by grinding smear or paint.
• Magnetic particle testing (MPT) — IS 3703, ASME Sec V Art. 7:
  • Magnetise the part; a surface or near-surface flaw distorts the flux and creates a leakage field that attracts ferromagnetic particles (dry powder or wet fluorescent suspension), forming a visible indication.
  • Ferromagnetic materials only — never on austenitic stainless, Al, Cu, brass.
  • Circular magnetisation (prods, head shot, central conductor) finds longitudinal flaws; longitudinal magnetisation (coil, yoke) finds transverse flaws — so every part is tested in two directions at 90°. A flaw parallel to the flux gives no indication.
  • Detects slightly subsurface flaws (≈ up to 2–3 mm); demagnetisation afterwards is required for parts that will be machined or run near bearings.
• Eddy current testing (ECT): an AC coil induces circulating currents in a conductive part; cracks, conductivity and permeability changes alter the coil impedance.
  • Electrically conductive materials only, no couplant, no contact, very fast, easily automated.
  • Skin effect limits it to near-surface work: standard depth of penetration δ = 1/√(πfμσ) — lower frequency = deeper.
  • Used for condenser and heat-exchanger tube inspection, aircraft fastener holes, bar and tube in-line inspection, alloy sorting, heat-treat verification and non-conductive coating thickness measurement. Lift-off (probe-to-surface gap) is the main nuisance variable.
  • Magnetic flux leakage (MFL) is the related method for storage-tank floors and pipeline intelligent pigs.

5. Volumetric NDT: ultrasonic and radiography

• Ultrasonic testing (UT) — IS 3664 practice (contact and immersion pulse-echo), IS 4260 for butt welds in ferritic steel, ASME Sec V Art. 4:
  • A piezoelectric crystal (PZT, quartz) pulsed by a short voltage spike sends 0.5–15 MHz sound into the part; echoes from the back wall and from flaws return to the same crystal (pulse-echo) and are displayed as an A-scan (amplitude vs time/distance). Through-transmission and pitch-catch use separate probes.
  • Couplant (oil, grease, gel, water) is essential — an air gap reflects ~100% of the sound.
  • Normal beam probes find laminations and measure thickness; angle beam shear-wave probes (45°, 60°, 70°) are used for welds, scanning in half-skip and full-skip; TOFD and phased array (PAUT) give accurate through-wall sizing and imaging and are now accepted in place of radiography in many codes.
  • Calibration on IIW V1 and miniature V2 blocks, and on step wedges and side-drilled-hole reference blocks; a DAC or DGS curve corrects for attenuation with distance.
  • Strengths: large thickness capacity, single-sided access, instant result, no radiation hazard, good at planar flaws (cracks, lack of fusion). Limits: operator-dependent, needs smooth surface, poor on coarse-grained austenitic and cast structures, a dead zone near the surface, and flaws lying parallel to the beam may be missed.
• Radiographic testing (RT) — IS 2595 practice, IS 1182 for fusion-welded butt joints in steel plate, ASME Sec V Art. 2:
  • X-rays (tube, adjustable kV) or gamma rays (Ir-192, Co-60 for thick steel, Se-75 for thin) pass through the part; differential absorption (I = I₀e^−μx) forms a latent image on film, on a CR imaging plate or on a DR digital panel.
  • A flaw with less material (porosity, slag, lack of penetration) transmits more radiation and appears darker on the radiograph.
  • Film density usually 2.0–4.0; image quality is proved by an IQI/penetrameter (wire type to DIN/ASTM, or hole type), with 2% sensitivity typically required.
  • Strengths: permanent record, excellent for volumetric flaws, interpretable by a third party. Limits: radiation hazard and area cordoning, high cost and slow, two-sided access needed, and poor at tight planar cracks that lie across the beam.
• Other methods: acoustic emission (AE) listens to stress waves from growing cracks during a proof or hydro test (used on pressure vessels and bridges); infrared thermography for delamination, insulation and electrical hot spots; leak testing (bubble, pressure-decay, helium mass spectrometer); in-situ metallographic replication for creep damage in boiler and refinery plant; holography/shearography for composites; computed tomography (CT) for castings and additively manufactured parts.

6. Choosing a method, codes, personnel and safety

• Selection logic: surface-breaking crack on steel → MPT; on stainless/aluminium → LPT or ECT; internal planar flaw → UT; internal volumetric flaw → RT; tubes in a heat exchanger → ECT; an active growing crack in service → AE.
• Indian and international codes: IS 2825 (unfired pressure vessels), IS 800 and IS 2062 (structural steel), IS 1239/IS 3589 (pipes), ASME Sec V (NDE methods), Sec VIII Div 1/2 (vessels), Sec IX (welding qualification), API 650/653 (tanks), API 1104 (pipelines), AWS D1.1 (structural welding), IS 7318/IS 9595 (welding procedure).
• Personnel certification: ISNT through its National Certification Board (NCB) issues Level I, II and III in India; ASNT SNT-TC-1A is the employer-based recommended practice; ISO 9712 is the third-party central scheme; IS 13805 covers qualification and certification. Level I performs, Level II sets up and interprets, Level III writes procedures and approves.
• Radiation safety (statutory): industrial radiography is licensed by the AERB under the SHANTI Act 2025 (which replaced the Atomic Energy Act 1962) and the Atomic Energy (Radiation Protection) Rules 2004 (framed under the 1962 Act; check for rules notified under the new Act); sources are registered on the e-LORA portal; a certified Radiological Safety Officer (RSO) is mandatory, workers wear TLD/OSL badges monitored by BARC-approved labs, and site work needs cordoning, warning signs and survey meters. The occupational dose limit is 20 mSv per year averaged over 5 consecutive years, with a maximum of 30 mSv in any single year — verify the current AERB figure before quoting it professionally.
• Lab accreditation: test houses are accredited by NABL to ISO/IEC 17025; UTMs, hardness testers and impact machines need periodic calibration traceable to national standards (NPL India).

Key points

• Destructive vs non-destructive - DT gives properties from a sacrificial sample; NDT gives flaw data from the actual part.
• Discontinuity vs defect - a discontinuity becomes a defect only when it exceeds code acceptance limits.
• Engineering vs true stress - engineering uses the original area and falls after necking; true stress uses the instantaneous area and keeps rising.
• 0.2% proof stress - substitute for yield strength in materials with no distinct yield point.
• Ductility - % elongation depends on gauge length; % reduction in area does not.
• Brinell - 10 mm ball, 3000 kgf for steel, averages over coarse structures; Rockwell - direct depth reading, fast; Vickers - 136° diamond, one scale for everything.
• Charpy is simply supported, Izod is a cantilever - and the notch faces away from the striker in Charpy, towards it in Izod.
• DBTT - BCC steels turn brittle below a transition temperature; FCC metals do not.
• Endurance limit - steels have one (≈ 0.5 UTS); aluminium alloys do not.
• Creep - three stages; the secondary (minimum) creep rate governs design life above ~0.4 Tm.
• LPT - surface-breaking flaws only, any non-porous material, needs dwell and developer time.
• MPT - ferromagnetic materials only; magnetise in two perpendicular directions; demagnetise afterwards.
• ECT - conductive materials, near-surface only (skin effect), best for tubes and sorting; lift-off is the main variable.
• UT - best for planar flaws, single-sided access, needs couplant; poor on coarse austenitic and cast structures.
• RT - best for volumetric flaws, gives a permanent record, but misses tight cracks and carries a radiation licence burden.
• Certification and statute - ISNT NCB / ASNT SNT-TC-1A / ISO 9712 for personnel; AERB licence, RSO and TLD badges for radiography.

Formulas and examples

Example 1 - Tensile test on a mild steel bar. Specimen Ø12.5 mm, L₀ = 62.5 mm. Yield load 48 kN, maximum load 72 kN, final gauge length 78.2 mm, neck diameter 8.4 mm; in the elastic range a 24 kN load gave 0.061 mm extension.

• A₀ = 0.7854 × 12.5² = 122.72 mm² (check: 5.65√122.72 = 62.6 ≈ 62.5 mm, so it is a proportional specimen).
• σy = 48,000/122.72 = 391 MPa; UTS = 72,000/122.72 = 587 MPa.
• E: σ = 24,000/122.72 = 195.6 MPa, e = 0.061/62.5 = 9.76 × 10⁻⁴ → E = 195.6/9.76 × 10⁻⁴ ≈ 200 GPa.
• % elongation = (78.2 − 62.5)/62.5 × 100 = 25.1%.
• Af = 0.7854 × 8.4² = 55.42 mm² → % RA = (122.72 − 55.42)/122.72 × 100 = 54.8%.
• Modulus of resilience = 391²/(2 × 200,000) = 0.382 MPa = 0.382 MJ/m³. The values clear IS 2062 E350 (yield ≥ 350, UTS ≥ 490 MPa) with good ductility.

Example 2 - Brinell hardness and strength estimate. 3000 kgf on a 10 mm ball gives a 4.0 mm indent.

• √(100 − 16) = 9.1652; D − 9.1652 = 0.8348; π × 10 × 0.8348 = 26.23.
• BHN = 6000/26.23 = 229 kgf/mm².
• Estimated UTS ≈ 3.45 × 229 = 790 MPa (empirical for steels only, ±10%).
• A Vickers check with 30 kgf giving a 0.500 mm diagonal: HV = 1.854 × 30/0.25 = 222 — consistent, as HV ≈ HB below ~300.

Example 3 - Charpy impact energy. Pendulum mass 20 kg, effective radius 0.75 m, released at α = 140°, swings up to β = 95°.

• W = 20 × 9.81 = 196.2 N; cos 140° = −0.7660, cos 95° = −0.0872; difference = 0.6788.
• E = 196.2 × 0.75 × 0.6788 = 99.9 ≈ 100 J.
• Impact strength = 100 J / (10 × 8 mm below notch = 80 mm²) = 1.25 J/mm² — fully ductile for a structural steel at room temperature; the same steel at −40 °C might give only 10–15 J.

Example 4 - Ultrasonic flaw location and acceptance. Normal-beam probe on a 100 mm steel plate (v = 5,920 m/s); a flaw echo appears at 25 µs round trip.

• Path = 5,920 × 25 × 10⁻⁶ = 0.148 m → flaw depth = 0.148/2 = 74 mm from the scanning surface.
• Back-wall echo would appear at 2 × 0.100/5,920 = 33.8 µs, confirming calibration.
• Why couplant is needed: Z(steel) = 7,800 × 5,920 = 46.2 MRayl, Z(air) ≈ 0.0004 MRayl → R ≈ 100% reflected. With water (Z = 1.48 MRayl): R = [(1.48 − 46.2)/(1.48 + 46.2)]² = 0.88, so 12% still enters — enough to work.
• Fitness check: if the flaw is a surface crack of depth a with Y = 1.12, σ = 300 MPa and K_IC = 50 MPa√m, a_c = (1/π)(50/(1.12 × 300))² = (1/π)(0.1488)² = 7.05 mm — a crack deeper than ~7 mm would be unstable.

Example 5 - Radiographic geometry and sensitivity. X-ray focal spot 2.5 mm, source-to-film distance 600 mm, specimen thickness 25 mm, film in contact with the back face.

• d (source to source-side surface) = 600 − 25 = 575 mm; t = 25 mm.
• Ug = 2.5 × 25/575 = 0.109 mm — acceptable (ASME limits Ug to 0.51 mm for material under 50 mm / 2 in. thick).
• If the thinnest visible IQI wire is 0.20 mm, sensitivity = 0.20/25 × 100 = 0.8%, better than the usual 2% requirement.
• Attenuation: for 25 mm steel with μ = 0.06 mm⁻¹, I/I₀ = e^{−0.06 × 25} = e^−1.5 = 0.223, i.e. 22% transmitted; HVL = 0.693/0.06 = 11.6 mm.

Where it is used

• Power and process plant: BHEL, NTPC, L&T Heavy Engineering and EIL-designed refineries — RT and UT of boiler drums, headers and pressure-vessel welds to IS 2825 and ASME Sec VIII, with creep and replication testing on aged boiler tubes.
• Oil, gas and pipelines: GAIL, IOCL, Welspun and Jindal SAW — automated UT and RT on longitudinal pipe seams, in-line MFL pigs, API 1104 girth-weld acceptance, tank floor MFL to API 653.
• Automotive and forging: Bharat Forge, Maruti, Tata Motors — hardness and case-depth checks, fatigue testing of crankshafts and suspension parts, MPT of forged axles and connecting rods.
• Aerospace, defence and space: HAL, ISRO (VSSC/LPSC), DRDO and BARC — fluorescent penetrant, eddy current on fastener holes, phased-array UT and CT on castings, composite inspection by thermography and shearography.
• Steel, railways and infrastructure: Tata Steel, SAIL and JSW mills (tensile, Charpy, hardness per IS 2062); Indian Railways ultrasonic rail flaw detection cars and periodic axle testing; UT and MPT on bridge and metro structural welds.
• Job roles: NDT technician/engineer (ISNT NCB Level II in UT/RT/PT/MT is the entry ticket, with RT work requiring AERB-linked radiation training), materials/metallurgical lab engineer in a NABL-accredited lab, QA/QC inspector, welding inspector (CSWIP/AWS CWI), third-party inspection with TÜV, BV, SGS or RITES.

Interview questions

• Q: What is the difference between engineering stress and true stress? A: Engineering stress uses the original cross-section, so it drops after necking; true stress uses the instantaneous area and keeps increasing until fracture.
• Q: Why is 0.2% proof stress used instead of yield strength for aluminium or stainless steel? A: These materials have no distinct yield point, so the stress producing 0.2% permanent strain is defined as the practical yield value.
• Q: Which gives a better measure of ductility — % elongation or % reduction in area? A: Reduction in area, because elongation depends on the gauge length chosen and is distorted by localised necking.
• Q: Why can't magnetic particle testing be used on austenitic stainless steel? A: It is non-ferromagnetic, so it cannot be magnetised and no leakage field forms at a flaw; use penetrant or eddy current instead.
• Q: Why is a magnetic particle test done in two perpendicular directions? A: A discontinuity lying parallel to the flux lines produces little or no leakage field; the flaw must lie roughly across the flux, so both circular and longitudinal magnetisation are applied.
• Q: When would you choose UT over RT for a weld? A: For thick sections, planar flaws such as lack of fusion and cracks, single-sided access, instant results and where radiation safety or cordoning is impractical; RT is better for porosity and slag and when a permanent image is required.
• Q: Why is couplant essential in contact ultrasonic testing? A: The huge impedance mismatch between steel and air reflects almost all the sound at the interface; a liquid couplant lets the beam enter the part.
• Q: What is the difference between a discontinuity and a defect? A: A discontinuity is any interruption in the material; it is a defect only if it exceeds the acceptance criteria of the applicable code.
• Q: What is DBTT and why does it matter in design? A: The ductile-to-brittle transition temperature, below which BCC steels fail with little energy absorption; low-temperature service steels must be Charpy qualified at the minimum design metal temperature.
• Q: Why do steels have an endurance limit but aluminium alloys do not? A: Interstitial solutes pin dislocations in steel, giving a stress below which fatigue cracks do not initiate; aluminium shows continuously falling S-N behaviour, so a fatigue strength at 10⁸ cycles is quoted instead.
• Q: A penetrant test on a ground weld shows nothing, but a crack is suspected. What do you do? A: Grinding smears metal over the crack mouth; etch or lightly re-clean the surface and retest, or use magnetic particle or ultrasonic testing which are less affected by smearing.
• Q: What statutory approvals are needed to do site radiography in India? A: An AERB licence under the Atomic Energy (Radiation Protection) Rules 2004, source registration on e-LORA, a certified RSO, TLD/OSL monitoring of workers, and area cordoning with survey meters and warning signs.

Common mistakes

• Dividing the maximum load by the necked area, or the fracture load by the original area, when computing UTS.
• Quoting % elongation without the gauge length, or comparing values measured on different gauge lengths.
• Mixing up Charpy and Izod — support condition, specimen length and notch orientation are all different.
• Treating hardness-to-UTS conversion (≈ 3.45 × BHN) as exact, or applying it to non-ferrous metals and hardened cases.
• Claiming LPT or MPT can find internal defects — LPT needs the flaw open to the surface; MPT reaches only a few millimetres.
• Forgetting the ×2 in ultrasonic pulse-echo distance (the pulse travels there and back), or using longitudinal velocity for an angle-beam shear-wave calculation.
• Mixing units: kgf/mm² with MPa (1 kgf/mm² ≈ 9.81 MPa), mm with m in √(πa), or µs with s in UT timing.

Related topics

Engineering materials and heat treatment · Welding processes and weld defects · Fracture mechanics and failure analysis · Metrology and quality control · Strength of materials: stress, strain and failure theories · Quality management, NABL accreditation and inspection codes