Non-destructive testing of structures

In short

Non-destructive testing (NDT) assesses the strength, quality, uniformity and deterioration of an existing structure without cutting or damaging it, using rebound, ultrasonic, electrical, electrochemical, radar and radiographic methods. It is the backbone of structural audits, repair and retrofit design, forensic investigation after fire or earthquake, and quality acceptance where cube results are in dispute. In India the methods are codified in IS 516 Part 5 (Section 1 UPV and Section 4 rebound hammer, which superseded IS 13311 Parts 1 and 2; Section 2 half-cell potential, Section 3 carbonation depth), IS 2770, IS 14858, SP 25 and the IRC guidelines for bridge inspection; municipal bye-laws in Mumbai, Delhi and other cities make periodic structural audit of old buildings mandatory. Civil engineering exams (GATE, ESE, university "Materials and Construction" and "Repair and Rehabilitation" papers) ask for method principles, grading tables and interpretation. Employers test it in interviews for site QA/QC, third-party inspection and structural-audit roles.

Explained

What NDT is, and why cubes are not enough

• NDT: testing that leaves the member fit for service — no measurable loss of strength or section. Partially destructive (semi-destructive): small, repairable damage — core cutting, pull-out, pull-off, Windsor probe, chemical sampling.
• Need: cube strength describes a sample cast on site, not the concrete *in* the member, which depends on placing, compaction, curing and age; and for an old structure no cubes exist.
• Purposes: (i) estimate in-situ strength and uniformity, (ii) locate honeycombing, voids, cracks and delamination, (iii) find reinforcement position, size and cover, (iv) diagnose corrosion and chemical attack, (v) check repairs and grouting, (vi) load-rating and residual-life assessment.
• Golden rule: no single NDT gives strength directly. Each measures a *property correlated* with strength; results are combined (combined/multi-method approach) and calibrated against cores from the same concrete.

Surface hardness: rebound hammer (IS 516 Part 5 Sec 4:2020; earlier IS 13311 Part 2)

• Principle: a spring-driven mass of fixed energy (2.207 N·m for the standard N-type) strikes a plunger held against the surface; the rebound distance, read as rebound number (RN) on a 10-100 scale, increases with surface hardness and hence with compressive strength.
• Procedure: grind off plaster/loose mortar to expose a smooth, dry, rigid surface; IS 516 Part 5/Sec 4 takes six readings around each observation point, impacts at least 25 mm apart and 25 mm from edges, removes outliers per IS/ISO 16269 (Part 4) and averages the rest (ASTM C805 instead takes 10 readings, discards any differing from the average by more than 6 units, and rejects the set if more than two do); convert using the correlation curve.
• Influences: moisture (wet surface lowers RN), carbonation (raises RN, over-estimating strength in old concrete), aggregate type, surface finish, member size/rigidity (thin members must be propped), and hammer orientation — horizontal against a vertical face is the reference; readings taken upwards or downwards need the correction curve.
• Depth of influence only about 25-30 mm, so the result reflects the skin, not the core.
• Indicative quality grading (old IS 13311 Part 2 table; IS 516 Part 5/Sec 4 gives none, only site correlation): average RN > 40 very good hard layer; 30-40 good layer; 20-30 fair; < 20 poor concrete; 0 (no rebound) delaminated.
• Calibration: against a steel test anvil of about 16 kg mass (standard reading about 80 ± 2) before each series; a site-specific correlation on cubes/cores of the same concrete is essential if strength is to be quoted.

Ultrasonic pulse velocity (IS 516 Part 5 Sec 1; earlier IS 13311 Part 1)

• Principle: a transducer sends a longitudinal (P-wave) pulse, typically 50-54 kHz for normal concrete (lower, 20-25 kHz, for thick/mass concrete; higher, 100-150 kHz, for thin members); the receiver picks it up and the instrument reads transit time in microseconds. V = path length / transit time.
• Arrangements: direct (opposite faces, most reliable), semi-direct (adjacent faces) and indirect/surface (same face, requires path-time regression to remove the coupling offset). Indirect velocity is roughly 5-20% lower than direct.
• Interpretation (former IS 13311 Part 1 grading for dry, normal-weight concrete): > 4.5 km/s excellent; 3.5-4.5 km/s good; 3.0-3.5 km/s medium; < 3.0 km/s doubtful — then confirm with cores. IS 516 Part 5/Sec 1:2018 revised the bands (commonly reproduced as > 4.40 excellent, 3.75-4.40 good, 3.00-3.75 doubtful, < 3.00 poor) — read the table in the standard before quoting.
• Uses: uniformity and comparative survey, detecting cracks, voids and honeycombing (pulse detours round a void → longer transit time), measuring depth of a surface crack, monitoring strength gain, checking grouting and fire/frost damage, and estimating dynamic modulus.
• Limitations: steel reinforcement parallel to the path raises velocity (pulse travels faster in steel ≈ 5.9 km/s) so paths must avoid bars; moisture raises velocity by up to 2% or so; air voids, aggregate content and path length all affect it; velocity is insensitive at high strengths — a small velocity change maps onto a large strength change, so UPV alone must never be converted to a strength number for acceptance.
• Good couplant (grease, petroleum jelly, soap gel) and firm contact are essential; air gaps cause complete loss of signal.

Reinforcement, cover and corrosion assessment

• Cover meter / rebar locator (profometer): electromagnetic induction (eddy-current/pulse-induction) finds bar position, spacing, orientation and cover, and estimates diameter; accuracy of the order of ±2-5 mm at normal covers, poorer where bars are congested or in two layers. Used to verify cover against IS 456 (nominal cover 20-75 mm by exposure class) before coring or drilling.
• Half-cell (open-circuit) potential, ASTM C876: a Cu/CuSO₄ reference electrode on a wetted surface is connected through a high-impedance voltmeter to the reinforcement; more negative potential means higher corrosion probability. Guide values (CSE): more positive than −200 mV → < 10% probability; −200 to −350 mV → uncertain; more negative than −350 mV → > 90% probability; below about −500 mV suggests severe/active corrosion. It gives probability, not rate; needs electrical continuity of the steel and fails on dry, coated or saturated concrete.
• Resistivity (four-probe Wenner): concrete resistivity indicates how easily corrosion current flows — above about 20 kΩ·cm corrosion is unlikely/slow, 10-20 kΩ·cm moderate, below 10 kΩ·cm high corrosion rate. Combined with half-cell for a fuller picture.
• Linear polarisation resistance (LPR): gives corrosion current density i_corr (µA/cm²) and hence a section-loss *rate*; < 0.1 passive, 0.1-0.5 low, 0.5-1.0 moderate, > 1.0 high.
• Carbonation depth: freshly broken/drilled surface sprayed with phenolphthalein (1% in ethanol) — pink where pH > ~9 (uncarbonated), colourless where carbonated. Semi-destructive but routine; carbonation front advancing past the cover de-passivates the steel.
• Chemical tests on drilled powder: chloride content (acid-soluble, IS 456 Table 7 — 0.6 kg/m³ of concrete for RCC/plain concrete with embedded metal, 0.4 kg/m³ for prestressed and steam-cured) and sulphate content (water-soluble, as SO₃, limit 4% by mass of cement). Also pH, cement content.

Other methods: coring, penetration, pull-out, impact echo, GPR, thermography, radiography, acoustic emission, load test

• Core test (IS 516 Part 4, acceptance under IS 456 cl. 17.4): the reference method. Diameter preferably ≥ 3× maximum aggregate size (100 mm typical, 150 mm best); preferred length-to-diameter ratio 2, which gives the standard cylinder strength — multiply by about 1.25 for the equivalent cube strength; capped and tested. Corrections applied for l/d ratio, presence of reinforcement and curing condition. IS 456 criterion: concrete in the member is acceptable if the average equivalent cube strength of three cores is at least 85% of the specified grade and no individual core is below 75%.
• Penetration resistance (Windsor probe, ASTM C803): a powder-driven probe is fired into the surface; exposed length correlates with strength. Semi-destructive, affected strongly by aggregate hardness.
• Pull-out (LOK test, ASTM C900) and pull-off: force to pull an embedded insert or a bonded disc; the best-correlated strength estimate after coring; also used to check repair-mortar bond.
• Impact echo / impulse response: a small impact generates a stress wave; the resonance of the reflected wave gives thickness or locates delamination, voids and debonding in slabs, pavements and tunnel linings.
• Ground-penetrating radar (GPR): electromagnetic pulses (about 1-2.6 GHz antennae) reflect from interfaces — maps rebar layout, slab thickness, ducts, voids, moisture; fast and continuous but cannot give strength and resolves poorly in wet/saline concrete and under dense top steel.
• Infrared thermography: surface temperature differences during heating/cooling reveal subsurface delamination, debonded tiles and leakage; non-contact, good for large façades and bridge decks.
• Radiography/radiometry (gamma or X-ray): direct image of voids, tendon ducts and grouting in prestressed members; excellent but slow, costly and subject to AERB radiation-safety licensing in India.
• Acoustic emission: listens to the stress waves from growing cracks and wire breaks — used for monitoring bridges, prestressed tanks and vessels under load.
• Pile integrity testing: low-strain PIT/sonic echo (hammer blow at pile head, reflection shows necking, bulging, length, discontinuity), cross-hole sonic logging (CSL) in access tubes, and high-strain dynamic load test (PDA, ASTM D4945; excluded from the scope of IS 2911 Part 4, which covers static load tests) for capacity.
• Permeability/absorption: initial surface absorption, RCPT (ASTM C1202, coulombs), water-permeability DIN 1048 — durability indicators for the cover concrete.
• In-situ load test (IS 456 cl. 17.6): when NDT is inconclusive, apply the full dead load plus 1.25× imposed load for 24 h. If the maximum deflection during the 24 h under load is less than 40 l²/D mm (l in m, D in mm), recovery need not be measured; otherwise at least 75% of the deflection must recover within 24 h of removing the imposed load, failing which the test may be repeated after 72 h — recovery below 80% on the repeat means the structure is unacceptable.

Planning a structural audit and the Indian code framework

1. Desk study: drawings, as-built records, mix records, repair and loading history, age, exposure.
2. Visual inspection and distress mapping: cracks (type, width, pattern), spalling, rust stains, efflorescence, seepage, deflection, tilt. Crack width by crack-width gauge/microscope; movement by tell-tale (Demec) gauges over time.
3. Non-destructive survey: rebound hammer + UPV grid for strength/uniformity, cover meter for steel, half-cell and resistivity where corrosion is suspected.
4. Semi-destructive confirmation: cores, carbonation depth, chloride/sulphate, pull-out in zones flagged weak.
5. Analysis and report: distress diagnosis, in-situ strength assignment, residual life, repair/retrofit recommendation, cost estimate; load test only if required.
6. Codes and practice: IS 516 Part 5 (NDT — Section 1 ultrasonic pulse velocity, Section 2 half-cell potential, Section 3 carbonation depth, Section 4 rebound hammer; Sections 1 and 4 replaced IS 13311 Parts 1 and 2), IS 516 Part 4 (cores), IS 456:2000 (acceptance, cover, chloride/sulphate limits, load test), IS 2770 (pull-out of bond test specimens), IS 14858 (compression testing machine), IS 13920 (ductile detailing — read with retrofit), IS 2911 Part 4 (pile load test), SP 25 (causes and prevention of cracks in buildings), IRC SP 35, SP 40 and SP 51/52 (bridge inspection, rehabilitation and load testing), and NBC 2016 for maintenance. ASTM C805 (rebound), C597 (UPV), C876 (half-cell), C900 (pull-out), C1202 (RCPT) and BS 1881 parts are commonly cited; some section numbers and limits have been revised, so the current amendment should be confirmed before quoting in a report.

Key points

• NDT purpose - assess in-situ quality, uniformity and deterioration without impairing the member's service capacity.
• No NDT gives strength directly - every method measures a correlated property and must be calibrated against cores.
• Rebound hammer - surface hardness only, 25-30 mm deep; six readings per observation point in IS 516 Part 5/Sec 4 (10 per area in ASTM C805), outliers discarded.
• Orientation correction - rebound numbers taken upward/downward differ from the horizontal reference and need the correction curve.
• Carbonation inflates rebound number - old carbonated concrete reads harder than it is.
• UPV - transit-time measurement at about 50-54 kHz; direct > semi-direct > indirect in reliability.
• UPV grading - > 4.5 excellent, 3.5-4.5 good, 3.0-3.5 medium, < 3.0 doubtful (km/s, dry normal concrete; former IS 13311 table, revised in IS 516 Part 5/Sec 1).
• Voids and cracks lengthen the UPV path and so lower the apparent velocity; reinforcement along the path raises it.
• Combined method - rebound (surface) plus UPV (interior) gives a far better strength estimate than either alone (SonReb).
• Core test is the benchmark: average of three cores ≥ 85% of grade and no core < 75% (IS 456 cl. 17.4).
• Core geometry - diameter ≥ 3× nominal maximum aggregate; l/d = 2 preferred, × ≈ 1.25 for equivalent cube strength, with factors for other ratios.
• Cover meter - electromagnetic location of bars, spacing and cover; accuracy falls where steel is congested.
• Half-cell potential - more negative than −350 mV (Cu/CuSO₄) implies > 90% probability of corrosion; probability, not rate.
• Resistivity - below 10 kΩ·cm means high corrosion risk; above 20 kΩ·cm low risk.
• Phenolphthalein - pink = uncarbonated (pH > 9), colourless = carbonated.
• Chloride limits (IS 456) - 0.6 kg/m³ for RCC, 0.4 kg/m³ for prestressed/steam-cured; sulphate 4% (as SO₃) by mass of cement.
• Load test (IS 456 cl. 17.6) - DL + 1.25× LL for 24 h; deflection below 40 l²/D mm needs no recovery check, otherwise ≥ 75% recovery (≥ 80% on a repeat test after 72 h).
• Pile integrity - low-strain sonic echo for defects and length; CSL for bored piles; PDA for capacity.

Formulas and examples

• Example 1 — UPV and quality grading. A direct reading across a 450 mm thick wall gives 118 µs.
  • V = 450 mm / 118 µs = 3.81 mm/µs = 3.81 km/s.
  • This is "good" concrete under both the old (3.5-4.5 km/s) and revised (3.75-4.40 km/s) bands — acceptable uniformity, no coring triggered on this count.
  • A second grid point on the same wall reads 160 µs → V = 450/160 = 2.81 km/s, i.e. "doubtful" on the old table ("poor" on the revised bands); the 26% drop over a few hundred millimetres signals a void or honeycomb, so core that location.
• Example 2 — dynamic modulus. Same concrete: V = 3.81 km/s = 3,810 m/s, ρ = 2,400 kg/m³, ν = 0.2.
  • (1 + ν)(1 − 2ν)/(1 − ν) = (1.2 × 0.6)/0.8 = 0.90.
  • E_d = 2,400 × 3,810² × 0.90 = 2,400 × 1.4516 × 10⁷ × 0.90 = 3.14 × 10¹⁰ N/m² ≈ 31.4 GPa.
  • Static modulus is lower; IS 456 gives Ec = 5,000√fck, so for M25, Ec = 25 GPa — a dynamic value about 20-30% higher is normal.
• Example 3 — core test acceptance (M25 slab). Three 100 mm cores, l/d = 2 after trimming, measured core strengths 21.0, 19.2 and 23.4 MPa; the horizontally-drilled conversion factor at l/d = 2 is 1.25.
  • Equivalent cube strengths: 26.25, 24.00 and 29.25 MPa.
  • Average = (26.25 + 24.00 + 29.25)/3 = 26.50 MPa; required 85% of 25 = 21.25 MPa → satisfied.
  • Lowest core = 24.00 MPa; required 75% of 25 = 18.75 MPa → satisfied. Concrete accepted under IS 456 cl. 17.4.
• Example 4 — l/d correction. A core could only be trimmed to l/d = 1.2 and crushed at 24.0 MPa.
  • f_cube = 24.0 × 2.5/(1.5 + 1/1.2) = 24.0 × 2.5/2.3333 = 24.0 × 1.0714 = 25.7 MPa equivalent cube strength.
  • Short cores read *high* in cylinder terms because of end restraint, so the factor is smaller than the 1.25 that applies at l/d = 2; different codes give slightly different factors, so state the source used.
• Example 5 — load test on a slab. Simply supported slab, effective span l = 4.0 m, overall depth D = 150 mm, measured maximum deflection under DL + 1.25 LL after 24 h = 3.8 mm, residual deflection 24 h after removal = 0.8 mm.
  • Threshold = 40 l²/D = 40 × 16/150 = 4.27 mm > 3.8 mm → deflection criterion satisfied, so cl. 17.6 does not even require the recovery check.
  • Recovery = (3.8 − 0.8)/3.8 = 78.9%, above 75% in any case → slab passes the load test.

Where it is used

• Structural audit of ageing buildings: mandatory periodic audit under municipal bye-laws (for example MCGM's requirement for buildings older than 30 years, and similar rules in Delhi, Thane and Chennai) — rebound, UPV, cover, half-cell, carbonation and cores, by licensed structural auditors.
• Bridges and highways: NHAI/IRC bridge inspection and rating; UPV and impact echo on decks and piers, CSL and PIT on bored piles, thermography and GPR on decks — used by L&T, AFCONS, Tata Projects and third-party agencies.
• Quality disputes on site: when cubes fail or curing was poor, NDT plus cores decides acceptance or demolition under IS 456 cl. 17; routine on metro, power-plant and industrial contracts (DMRC, NTPC, NPCIL).
• Forensic and post-disaster investigation: fire-damaged buildings (UPV and residual-strength mapping), earthquake and blast damage, collapse enquiries, insurance claims.
• Repair and retrofit industry: pre-repair diagnosis and post-repair verification of grouting, jacketing and FRP bonding (pull-off tests) — firms such as Sika, Fosroc, BASF/MBCC applicators.
• Job roles: QA/QC engineer, NDT technician (ASNT/ISNT Level I-III for weld and metal NDT), structural auditor, third-party inspection engineer with TUV, SGS, Bureau Veritas, and researchers in CSIR-CBRI, CSIR-SERC and NCCBM.

Interview questions

• Q: Why is NDT needed when cube results are available? A: Cubes represent a sample cast and cured separately; NDT assesses the concrete actually in the member, including placing and curing defects, and is the only option for existing structures with no cube records.
• Q: Can a rebound hammer give the compressive strength of concrete? A: Only an indicative value for the top 25-30 mm; a reliable strength needs a site-specific correlation with cores or cubes of the same concrete.
• Q: Why does carbonation affect rebound readings? A: Carbonated skin is harder and denser, so the rebound number rises and strength is over-estimated in old concrete.
• Q: Which UPV arrangement is most reliable and why? A: Direct transmission on opposite faces, because the full pulse energy travels the shortest known path with no surface-layer bias.
• Q: What does a sudden drop in pulse velocity at one grid point mean? A: The pulse has detoured around a void, honeycomb or crack, lengthening the travel path — the location should be confirmed by coring or impact echo.
• Q: Why must UPV paths avoid reinforcement? A: Pulse velocity in steel is about 5.9 km/s against 3.5-4.5 km/s in concrete, so bars parallel to the path give a falsely high reading.
• Q: State the IS 456 core acceptance criteria. A: The average equivalent cube strength of three cores must be at least 85% of the specified grade and no individual core less than 75%.
• Q: What does a half-cell potential of −420 mV against Cu/CuSO₄ indicate? A: More than 90% probability of active corrosion of the reinforcement; it indicates probability, not the rate, which needs resistivity or LPR.
• Q: How do you measure carbonation depth on site? A: Spray 1% phenolphthalein in ethanol on a freshly broken or drilled surface — pink shows uncarbonated concrete (pH above about 9), colourless shows carbonation; measure the colourless depth from the surface.
• Q: Difference between low-strain pile integrity testing and a dynamic load test? A: Low-strain PIT uses a light hammer blow and reflected waves to find defects and length; a high-strain dynamic test (PDA) uses a heavy drop to estimate pile capacity.
• Q: When would you recommend a load test instead of more NDT? A: When NDT and cores leave doubt about the member's actual load-carrying capacity; IS 456 cl. 17.6 prescribes DL + 1.25 LL for 24 h with deflection and recovery limits.
• Q: Which NDT would you use to check grouting of a prestressing duct? A: Radiography or impact echo/ultrasonic tomography; GPR can locate the duct but cannot reliably confirm grout filling.

Common mistakes

• Quoting a strength from the rebound hammer or UPV alone without core calibration, and ignoring the method's ±20-25% scatter.
• Forgetting the orientation correction for the rebound hammer, or testing on plaster, wet or rough surfaces instead of a ground, dry, rigid one.
• Using too few rebound readings, or averaging in outliers instead of discarding outliers (IS/ISO 16269-4 in IS 516; ±6 units in ASTM C805).
• Running UPV along a reinforcement bar, or dividing by the wrong path length (using member thickness where the path is diagonal).
• Reading half-cell potentials as corrosion *rate*, or testing without ensuring electrical continuity of the steel.
• Coring without a cover-meter scan — cutting main reinforcement, or taking l/d < 1 cores and ignoring the correction factor.
• Mixing units: µs with ms, mm/µs with m/s, kΩ·cm with Ω·m — and confusing dynamic with static modulus.

Related topics

• Durability of concrete · Concrete technology and mix design · Repair and rehabilitation of structures · Concrete structures design (IS 456) · Building materials: cement, aggregates and bricks · Construction quality control and site supervision