Corrosion and its prevention

Metallurgical and Materials Engineering (B.E./B.Tech) · Physical metallurgy · free study notes for exams, placements and interviews

In short

Corrosion is the destructive chemical or electrochemical attack of a metal by its environment, which returns the refined metal towards the lower-energy oxide, hydroxide or salt from which it came. It costs India on the order of 4-5% of GDP every year (NCB/CECRI-type estimates vary; treat the figure as an order of magnitude, not an exact number), and dominates the service life of bridges, pipelines, ships, refinery vessels, boilers, rebar in concrete and automobile bodies. It is examined in GATE MT (electrochemical thermodynamics and kinetics, forms of corrosion, protection methods), in university physical-metallurgy and materials-science papers, and in interviews at ONGC, IOCL/BPCL/HPCL, L&T, SAIL, Tata Steel, JSW, NTPC, BHEL, Cochin Shipyard and NPCIL. Corrosion engineering is also a licensed-practice area: NACE/AMPP-certified coating and cathodic-protection inspectors are a standard requirement in Indian oil and gas tenders.

Explained

What corrosion is and the two mechanisms

Thermodynamic driving force: most engineering metals are extracted by reducing ores, so they sit above their oxides in free energy; corrosion is the spontaneous reverse reaction with ΔG < 0.
Dry (chemical) corrosion: direct reaction of metal with gas, with no liquid electrolyte - oxidation, sulphidation, carburisation, hot-gas attack in furnaces, boilers and turbine blades.
Wet (electrochemical) corrosion: needs an electrolyte; it proceeds as two coupled half-reactions at separated sites, like a short-circuited galvanic cell. Over 90% of practical corrosion problems are of this type.
The four essentials of a corrosion cell: anode (oxidation, metal dissolves), cathode (reduction), metallic path (electron conduction), electrolyte (ion conduction). Break any one and corrosion stops - this is the logic behind every prevention method.
Anodic reaction: M → Mⁿ⁺ + ne⁻ (e.g. Fe → Fe²⁺ + 2e⁻).
Cathodic reactions: oxygen absorption in neutral aerated water, O₂ + 2H₂O + 4e⁻ → 4OH⁻; hydrogen evolution in acids, 2H⁺ + 2e⁻ → H₂; metal-ion reduction and metal deposition in special cases.
Rust formation: Fe²⁺ + 2OH⁻ → Fe(OH)₂, further oxidised to hydrated ferric oxide Fe₂O₃·nH₂O (red rust); rust is porous and non-protective, unlike the oxide on Al or Cr.

Dry oxidation and the Pilling-Bedworth ratio

Oxidation starts by chemisorption, then an oxide film grows; further growth requires diffusion of metal cations outward or oxygen anions inward through the film.
Pilling-Bedworth ratio (PBR) = volume of oxide formed / volume of metal consumed. PBR < 1 (Mg, alkali metals) → porous, cracked, non-protective film; PBR ≈ 1-2 (Al, Cr, Ni, Ti, Si) → continuous adherent protective film; PBR > 2 (Fe scale layers, W, Mo) → compressive stress spalls the film off.
Growth laws: linear (x = kt) for porous non-protective films; parabolic (x² = kt) when diffusion-controlled, typical of iron above about 500 °C; logarithmic (x = k log(ct+1)) for thin films at low temperature, typical of Al, Cu, Fe below ~200 °C; cubic for some alloys.
Wagner's theory links the parabolic rate constant to ionic and electronic conductivity of the scale - the scale behaves as a solid electrolyte and oxidation is faster when the oxide is a poor insulator with many defects.
Iron scale above 570 °C is three-layered: FeO (wüstite, next to metal, highly defective and fast-growing), Fe₃O₄ (magnetite), Fe₂O₃ (haematite, outer). Below 570 °C FeO is unstable, so scaling is slower.
Alloying for oxidation resistance: Cr (≥12% for passivity, 18-25% for high-temperature scale), Al and Si form Cr₂O₃/Al₂O₃/SiO₂ films; rare-earth and Y additions improve scale adherence. Basis of stainless steels, Fe-Cr-Al heating elements and Ni-base superalloys.

Electrochemical thermodynamics and kinetics

Electrode potential: measured against a reference - SHE (0 V by definition), saturated Cu/CuSO₄ (+0.316 V vs SHE, used in soil for pipelines), Ag/AgCl (seawater), calomel (laboratory).
Nernst equation gives the potential at non-standard activity; it tells which metal is anodic in a given solution.
EMF series ranks pure metals by standard potential; galvanic series ranks real alloys in a real environment (usually seawater) and is what engineers actually use, because passivity changes behaviour - passive 18/8 stainless is noble, the same steel active in a crevice is nearly as base as carbon steel.
Pourbaix diagram (potential vs pH) maps regions of immunity, corrosion and passivity for a metal-water system; for iron it explains why high pH (concrete pore solution, boiler water) protects and low pH attacks.
Kinetics - mixed potential theory: the metal settles at a corrosion potential Ecorr where anodic and cathodic currents are equal; the common value is the corrosion current density icorr, which converts directly to a corrosion rate.
Tafel behaviour: activation-controlled polarisation gives η = ±β log(i/i₀); extrapolating the two Tafel lines to Ecorr gives icorr. Concentration polarisation gives a limiting current when the cathodic reactant (usually dissolved O₂) cannot diffuse fast enough - this is why stirring, flow velocity and aeration raise corrosion rates in neutral water.
Linear polarisation resistance (LPR) uses the slope dE/di near Ecorr (Stern-Geary relation) for non-destructive monitoring of plant corrosion.
Passivity: a few nanometres of adherent oxide (Cr₂O₃ on stainless, Al₂O₃ on aluminium, TiO₂ on titanium) drops the rate by 3-6 orders. Passive films break down above a pitting potential, especially in chloride solutions.

Forms of corrosion (the classical eight, plus high-temperature)

Uniform (general) attack: even thinning; least dangerous because it is predictable and handled by a corrosion allowance (typically 1.5-3 mm in pressure-vessel design).
Galvanic (bimetallic): dissimilar metals in electrical contact; the less noble corrodes. Severity rises with potential difference and with a small anode / large cathode area ratio - a steel bolt in a copper plate fails fast, a copper rivet in a steel plate does not.
Crevice corrosion: stagnant solution in gaps, under gaskets, washers, deposits and lap joints; oxygen depletion inside makes it anodic, chloride concentrates and pH falls (autocatalytic). Avoid by welding instead of riveting, using non-absorbent gaskets and full-penetration joints.
Pitting: localised chloride breakdown of a passive film giving deep narrow holes with little weight loss - the most treacherous form because perforation happens before thickness loss is noticeable. Resisted by Mo additions; ranked by PREN = %Cr + 3.3(%Mo) + 16(%N) (304 ≈ 19, 316 ≈ 25, duplex 2205 ≈ 35, super duplex ≥ 40).
Intergranular corrosion: grain-boundary attack. In austenitic stainless steel, holding or slow-cooling through 450-850 °C precipitates Cr₂₃C₆ at boundaries, leaving chromium-depleted zones - sensitisation; next to a weld this is weld decay. Cured by low-carbon grades (304L/316L, C ≤ 0.03%), stabilising with Ti or Nb (321, 347) or solution annealing ≈1050 °C and quenching. Exfoliation is the analogous layer attack in rolled Al alloys.
Selective leaching (dealloying/parting): one element is removed - dezincification of brass (Zn lost, porous copper left) in Indian plumbing fittings, graphitic corrosion of grey cast iron water mains, denickelification. Controlled by inhibited admiralty/arsenical brass or by DZR brass.
Erosion-corrosion: flow removes the protective film - pump impellers, elbows, condenser tubes. Special cases: impingement attack and cavitation damage in pumps and propellers; fretting at vibrating contacts.
Stress corrosion cracking (SCC): brittle cracking from tensile stress (often residual) plus a specific environment on a susceptible alloy - austenitic stainless in hot chlorides, brass in ammonia (season cracking), carbon steel in caustic (caustic embrittlement) and in nitrates, Al alloys in chlorides. Transgranular or intergranular; no general corrosion visible. Remedies: stress-relieve, change alloy (duplex or ferritic for chlorides), shot-peen for compressive surface stress, remove the species.
Related environmental cracking: hydrogen embrittlement/HIC and sulphide stress cracking in sour (H₂S) service - governed by NACE MR0175/ISO 15156, mandatory in Indian sour oil and gas fields; corrosion fatigue, where a corrosive environment removes the fatigue limit; microbiologically influenced corrosion (MIC) from sulphate-reducing bacteria under pipeline deposits and in stagnant tanks.

Prevention: material, design, environment, coatings and electrochemical

Material selection: stainless and duplex grades for chlorides, Cu-Ni for seawater condensers, titanium for severe chloride/oxidising service, Monel, Hastelloy, FRP and rubber-lined steel for acids; weathering steel (Cu-Cr-Ni-P, IS 11587 / ASTM A588 type) forms a tight patina and is used bare in bridges and transmission towers in non-coastal, non-industrial atmospheres.
Design rules: avoid dissimilar-metal contact or insulate it (gaskets, sleeves, non-conducting washers); avoid crevices and sharp changes of section; provide drainage - no water traps; make the anode large and the cathode small; allow access for inspection and painting; add a corrosion allowance; avoid galvanic mismatch in fasteners.
Environment modification: deaeration and vacuum degassing of boiler feedwater; chemical oxygen scavengers (hydrazine, now largely replaced by carbohydrazide or DEHA for toxicity reasons); pH control to about 9-9.5 in boiler circuits; dehumidification and VCI (vapour-phase inhibitor) packing for storage.
Inhibitors: anodic (chromate, nitrite, molybdate, phosphate - passivating; dangerous below a critical concentration, which causes pitting), cathodic (zinc salts, polyphosphates, carbonate - raise cathodic polarisation; safer), mixed/organic adsorption (amines, imidazolines, thiourea) used in acid pickling and oilfield pipelines, and vapour-phase inhibitors. Chromates are restricted on toxicity grounds.
Barrier coatings: paints and organic systems (primer with zinc phosphate or zinc-rich epoxy + intermediate + polyurethane/epoxy topcoat); three-layer polyethylene (3LPE) and 3LPP on cross-country pipelines; fusion-bonded epoxy (FBE); coal-tar enamel (older lines); surface preparation to Sa 2½ grade blast cleaning is usually the controlling factor.
Metallic coatings: galvanising (hot-dip zinc - sacrificial, protects even at scratches; mass per IS 4759 for structural steel, IS 4736 for tubes, and IS 12594 for galvanised rebar - confirm the current edition before quoting), tinning (barrier only; a scratch on tinplate makes steel anodic and corrodes faster), electroplating (Cr, Ni, Cd, Zn-Ni), metal spraying, cladding, hot aluminising, sherardising and diffusion coatings.
Conversion and anodic treatments: phosphating (before painting on automobile bodies), chromating/passivation of zinc, anodising of aluminium (10-25 µm sulphuric-acid anodic film, sealed or dyed - standard for window sections in India).
Cathodic protection (CP): make the whole structure a cathode. Sacrificial anode (galvanic) CP uses Zn, Al-Zn-In or Mg anodes - simple, no power, limited driving voltage (good for ship hulls, jetties, small tanks, well-coated short pipelines). Impressed current CP (ICCP) uses a DC rectifier/transformer with inert anodes (MMO-titanium, graphite, high-silicon cast iron) in a groundbed - large driving voltage, adjustable, used on cross-country pipelines and large tank farms. Criterion in common use: structure-to-soil potential of −850 mV or more negative vs Cu/CuSO₄ with current applied (polarised, instant-off); over-protection causes coating disbondment and hydrogen damage. CP is always used together with coating, never instead of it.
Anodic protection: anodically polarise a passivating metal into its passive range with a potentiostat - used for stainless and titanium in strong acid service; useless for non-passivating metals and risky if control is lost.
Rebar and concrete: carbonation and chloride ingress depassivate steel in concrete; control by low water-cement ratio, adequate cover and chloride limits per IS 456:2000 (durability clauses), supplementary cementitious materials, corrosion-inhibiting admixtures, fusion-bonded-epoxy or galvanised rebar, stainless rebar in marine zones, and migrating inhibitors or cathodic protection for repair.
Indian codes and practice (verify current edition before citing): IS 2062 for structural steel grades, IS 456:2000 for RCC durability, IS 1477 (painting of iron and steel) and IS 5905/IS 9954-type guides, IS 15311 and IS 8629-series for protective coatings and CP of buried pipelines, IS 2629/IS 4759/IS 2633 for galvanising practice and testing, OISD-STD-141 and related OISD standards for CP of buried pipelines in the Indian oil and gas sector, PNGRB regulations for pipeline integrity, and NACE/AMPP SP0169 and MR0175/ISO 15156 internationally.

Measurement and testing

Weight-loss coupons: simplest; gives average rate in mpy or mm/year over the exposure period.
Electrochemical: LPR probes and corrosion-rate meters for on-line plant monitoring; potentiodynamic polarisation for pitting and passivation potentials; EIS for coating quality.
Standard laboratory tests: salt-spray/neutral salt fog (ASTM B117, used for coating qualification in automotive tenders), cyclic corrosion tests, ferric chloride pitting test (ASTM G48) for CPT, Huey and Strauss tests (ASTM A262 practices) for sensitisation in stainless steel, U-bend and C-ring SCC tests, intergranular attack by oxalic-acid etch screening.
Field inspection: ultrasonic thickness survey, radiography, intelligent pigging (MFL and UT) of pipelines, holiday detection of coatings, close-interval potential survey (CIPS) and DCVG for CP and coating faults.

Key points

Corrosion - environment-driven degradation of metal returning it to a compound; ΔG for the reaction is negative, so it is spontaneous.
Wet corrosion - electrochemical, needs an electrolyte and separated anodic and cathodic sites; dry corrosion - direct gas-metal reaction, diffusion-controlled through a scale.
Anode corrodes, cathode is protected - every protection scheme either makes the metal a cathode, isolates the electrolyte, or stops electron/ion flow.
PBR - oxide/metal volume ratio; 1 < PBR < 2 gives protection, <1 porous, >2 spalling.
Parabolic law x² = kt means a protective diffusion-controlled scale; linear growth means no protection.
Galvanic series - ranks alloys in a specific electrolyte; more useful in design than the standard EMF series.
Area effect - a small anode with a large cathode gives very high anodic current density and rapid failure.
Pourbaix diagram - E vs pH map of immunity, corrosion and passivity; justifies pH control in boilers and the alkalinity of concrete.
icorr - corrosion current density; Faraday's law converts it directly into mm/year.
Tafel extrapolation - linear polarisation segments extended to Ecorr give icorr and hence the rate.
Passivity - thin adherent oxide (Cr₂O₃, Al₂O₃, TiO₂) that cuts the rate by orders of magnitude; chlorides break it down at the pitting potential.
PREN = %Cr + 3.3%Mo + 16%N - quick pitting-resistance ranking of stainless steels.
Sensitisation - Cr₂₃C₆ at grain boundaries from 450-850 °C exposure; cured by L grades, Ti/Nb stabilisation or solution annealing.
SCC - susceptible alloy + tensile stress + specific environment; cracks with no visible general corrosion.
Galvanising beats tinning on a scratch - zinc is anodic to steel and protects it; tin is cathodic and accelerates attack.
CP criterion - commonly −850 mV (polarised) vs Cu/CuSO₄ for buried steel; coating plus CP, not CP alone.
Sacrificial vs ICCP - no power and self-regulating but low driving voltage, versus powered, adjustable and suited to long pipelines.
Anodic inhibitors are dangerous when under-dosed - partial passivation concentrates attack into pits.

Formulas and examples

ΔG = −nFE — spontaneity of a corrosion cell; n = electrons transferred, F = 96,485 C/mol, E = cell potential (V); ΔG in J/mol
E = E° − (RT/nF) ln(1/a(Mⁿ⁺)) — Nernst equation; at 25 °C, E = E° + (0.0592/n) log a(Mⁿ⁺) volts
w = (I t M)/(n F) — Faraday's law; w = mass dissolved (g), I = current (A), t = time (s), M = atomic mass (g/mol)
CR (mm/year) = (87.6 × W)/(ρ A t) — W = mass loss (mg), ρ = density (g/cm³), A = area (cm²), t = time (hours)
CR (mm/year) = 3.27 × 10⁻³ × icorr × EW / ρ — icorr in µA/cm², EW = equivalent weight = M/n (g), ρ in g/cm³
η = β log(i/i₀) — Tafel equation; η = overpotential (V), β = Tafel slope (V/decade), i₀ = exchange current density
Rp = βaβc / [2.303 icorr (βa + βc)] — Stern-Geary relation; Rp = polarisation resistance (Ω·cm²)
PBR = (M_oxide × ρ_metal) / (n × M_metal × ρ_oxide) — n = metal atoms per oxide formula unit
PREN = %Cr + 3.3(%Mo) + 16(%N) — pitting resistance equivalent number (dimensionless)
x² = k t (parabolic), x = k t (linear), x = k log(ct + 1) (logarithmic) — x = scale thickness, t = time

Example 1 - corrosion rate from weight loss. A mild steel coupon 50 mm × 25 mm × 3 mm is exposed in cooling water for 720 hours and loses 0.84 g.

Total area = 2(5.0 × 2.5) + 2(5.0 × 0.3) + 2(2.5 × 0.3) = 25.0 + 3.0 + 1.5 = 29.5 cm².
W = 840 mg, ρ = 7.87 g/cm³, t = 720 h.
CR = (87.6 × 840)/(7.87 × 29.5 × 720) = 73,584/167,159 ≈ 0.44 mm/year.
Judgement: above about 0.25 mm/year the water treatment is inadequate; typical target for treated cooling water is < 0.125 mm/year (≈5 mpy), so inhibitor dosing needs review.

Example 2 - rate from corrosion current density. An LPR probe on a carbon steel line reads icorr = 25 µA/cm². Take Fe → Fe²⁺, M = 55.85, n = 2, ρ = 7.87 g/cm³.

EW = 55.85/2 = 27.93 g.
CR = 3.27 × 10⁻³ × 25 × 27.93/7.87 = 3.27 × 10⁻³ × 698.25/7.87 = 2.2833/7.87 ≈ 0.29 mm/year.
For a 6 mm wall with a 3 mm corrosion allowance, remaining life ≈ 3/0.29 ≈ 10 years if the rate stays constant.

Example 3 - Faraday's law on a sacrificial anode. A 5 kg magnesium anode delivers an average 0.10 A. M(Mg) = 24.3, n = 2.

Mass consumed per year = (I t M)/(nF) = (0.10 × 365 × 24 × 3600 × 24.3)/(2 × 96,485) = (0.10 × 31,536,000 × 24.3)/192,970 = 76,632,480/192,970 ≈ 397 g/year at 100% efficiency.
Magnesium anodes are typically about 50% efficient in soil → ≈ 794 g/year. With 85% utilisation, usable mass = 4.25 kg → life ≈ 4,250/794 ≈ 5.4 years.

Example 4 - Pilling-Bedworth ratio for aluminium. 2Al + 3/2 O₂ → Al₂O₃; M(Al₂O₃) = 101.96, ρ(Al₂O₃) = 3.97 g/cm³; M(Al) = 26.98, ρ(Al) = 2.70 g/cm³, n = 2 Al atoms per formula unit.

PBR = (101.96 × 2.70)/(2 × 26.98 × 3.97) = 275.29/214.22 ≈ 1.29.
Between 1 and 2 → continuous, adherent, protective film. This is why aluminium needs no paint for atmospheric service and why anodising (thickening the same film) works.

Example 5 - galvanic area effect. Two seawater joints carry the same total galvanic current of 0.5 mA: (a) a steel bolt of 2 cm² exposed area in a copper plate, (b) a copper bolt of 2 cm² in a steel plate of 200 cm².

(a) anodic current density on steel = 0.5 mA/2 cm² = 0.25 mA/cm² → by the icorr formula, 3.27 × 10⁻³ × 250 × 27.93/7.87 ≈ 2.9 mm/year - the bolt fails in a couple of years.
(b) the steel plate is the anode: 0.5 mA/200 cm² = 0.0025 mA/cm² = 2.5 µA/cm² → ≈ 0.029 mm/year - negligible.
Rule confirmed: never put a small anode with a large cathode.

Feature	Sacrificial anode CP	Impressed current CP
Power source	None - galvanic couple	External DC rectifier
Driving voltage	Low (0.2-1.0 V)	High and adjustable (up to tens of V)
Anode material	Zn, Al-Zn-In, Mg	MMO-Ti, graphite, Si-Fe (inert)
Best for	Ship hulls, jetties, short/well-coated lines, tank internals	Cross-country pipelines, tank farms, bare or poorly coated steel
Risk of over-protection	Very low (self-limiting)	Real - coating disbondment, hydrogen damage
Running cost	Anode replacement	Electricity + monitoring

Form of corrosion	Signature	Typical fix
Galvanic	Attack at the junction, less noble metal	Insulate, match metals, large anode area
Crevice	Attack under gasket, deposit, lap joint	Weld joints, non-absorbent gaskets, drain
Pitting	Deep holes, tiny weight loss	Higher-Mo/PREN alloy, reduce chlorides
Intergranular	Grain-boundary attack near welds	304L/316L, 321/347, solution anneal
Dealloying	Porous, colour change (brass → copper)	Inhibited or DZR brass
Erosion-corrosion	Grooves, horseshoe pits facing flow	Lower velocity, harder/liner material
SCC	Brittle branched cracks, no thinning	Stress relief, alloy change, shot peening

Where it is used

Oil, gas and petrochemicals: ONGC, IOCL, BPCL, HPCL, GAIL - 3LPE coatings plus ICCP on cross-country pipelines, sour-service material selection to NACE MR0175, inhibitor injection, intelligent pigging and corrosion-loop inspection under PNGRB and OISD requirements.
Marine and defence: Cochin Shipyard, Mazagon Dock, Indian Navy - hull painting systems, Zn/Al sacrificial anodes, Cu-Ni condenser tubing, offshore platform splash-zone protection with Monel sheathing.
Infrastructure and construction: rebar durability under IS 456:2000, FBE and galvanised rebar in coastal projects, galvanised transmission towers and highway guardrails, weathering-steel and painted bridges; corrosion is a routine clause in L&T, NHAI and railway bridge specifications.
Power and process plant: NTPC, BHEL, NPCIL - boiler feedwater deaeration and pH control, condenser tube selection, flow-accelerated corrosion in feedwater lines, stress-relief of stainless piping welds.
Automotive and consumer durables: Maruti, Tata Motors, Bajaj - zinc phosphating plus cathodic electrodeposition (ED) primer, galvannealed sheet, salt-spray warranty testing to ASTM B117.
Job roles: corrosion/materials engineer, coating and CP inspector (AMPP/NACE Level 2-3 certification is commonly tendered), welding and NDT inspector, plant integrity and inspection engineer, water-treatment chemist, failure-analysis metallurgist at NML, CECRI and in-house labs.

Interview questions

Q: Why is corrosion mostly electrochemical rather than chemical? A: In the presence of an electrolyte the anodic and cathodic reactions can occur at separate sites with electrons flowing through the metal, which is far faster and more common than direct gas-metal reaction.
Q: A scratch on galvanised steel does not rust, but a scratch on tinplate does. Why? A: Zinc is anodic to iron, so it keeps corroding and cathodically protects the exposed steel; tin is cathodic to iron, so the small exposed steel area becomes the anode and corrodes rapidly.
Q: What does the Pilling-Bedworth ratio tell you? A: Whether the oxide formed is bulky enough to cover the metal but not so bulky that it cracks - roughly 1 to 2 gives a protective adherent film, below 1 is porous, above 2 spalls off.
Q: Why is pitting more dangerous than uniform corrosion? A: It perforates a wall with almost no measurable weight loss or thickness change, so it is missed by routine thickness surveys and causes sudden leaks.
Q: What is sensitisation and how do you avoid it in a stainless steel weld? A: Chromium carbide precipitation at grain boundaries between about 450 and 850 °C leaves chromium-depleted zones open to intergranular attack; use 304L/316L, Ti/Nb-stabilised 321/347, low heat input, or solution anneal and quench after welding.
Q: Give the three conditions required for stress corrosion cracking. A: A susceptible alloy, a sustained tensile stress (often residual welding stress) and a specific aggressive species - for example austenitic stainless steel, residual stress and hot chlorides.
Q: Why is the galvanic series preferred to the EMF series in design? A: The EMF series lists pure metals at standard conditions; the galvanic series ranks actual alloys in the actual electrolyte and accounts for passivity, which can reverse the ranking.
Q: What is the usual cathodic protection criterion for a buried steel pipeline? A: A polarised (instant-off) structure-to-soil potential of −850 mV or more negative with respect to a saturated copper/copper-sulphate reference electrode, avoiding over-protection that disbonds coating.
Q: Why can an anodic inhibitor make corrosion worse? A: Below the critical concentration it passivates only part of the surface; the remaining bare area becomes a small anode against a large cathode, giving severe pitting.
Q: Why does steel in concrete normally not corrode? A: The alkaline pore solution (pH about 12.5-13) keeps iron in its passive region on the Pourbaix diagram; corrosion starts only when carbonation lowers the pH or chlorides break the passive film.
Q: How would you distinguish crevice corrosion from pitting in a failed flange? A: Crevice attack is located at the shielded gap under the gasket or joint face and is broad and shallow-to-deep within that geometry; pitting occurs on freely exposed passive surfaces as discrete deep holes.
Q: What does a PREN of 35 versus 19 mean practically? A: Duplex 2205 (≈35) tolerates far higher chloride content and temperature before pitting than 304 (≈19), which is why 304 is unsuitable for seawater or coastal splash exposure.

Common mistakes

Saying the cathode corrodes, or writing the anodic reaction as a reduction - the anode always oxidises and dissolves.
Quoting the EMF series to decide galvanic behaviour of alloys in a real solution, ignoring passivity reversals such as passive stainless being noble.
Forgetting the small-anode/large-cathode area effect, and selecting stainless or copper fasteners in a steel structure.
Mishandling units in the rate formulas: mg vs g, hours vs years, cm² vs m², and mixing mpy with mm/year (1 mm/year ≈ 39.4 mpy).
Treating corrosion rate from a weight-loss coupon as valid for localised attack - it averages away pits and crevices.
Assuming cathodic protection replaces coating, or that more negative potential is always better - over-protection disbonds coating and causes hydrogen damage.
Confusing sensitisation (carbide precipitation, intergranular) with SCC (stress-assisted cracking) or with weld dilution problems, and quoting the wrong remedy.