Corrosion Simulation
Theory: corrosion is an electrochemical cell
Overview
What does a corrosion simulation actually calculate?
It calculates current flowing over metal surfaces. When iron rusts, at some sites iron gives up electrons and becomes ions (anodic reaction), and elsewhere those electrons reduce dissolved oxygen (cathodic reaction). Corrosion is thus a small battery, and the current equals the dissolved metal. In salty water such as seawater, the cathodic reaction is often capped by how fast oxygen reaches the surface. When two metals touch, the lower-potential one becomes the anode, taking on the oxygen reduction on its partner as well, and dissolves quickly. Simulation solves the potential field in the electrolyte together with surface reaction rates to find where and how much metal dissolves.
Current to thickness loss
M is molar mass, n the number of electrons and ρ density; for steel M = 55.85 g/mol, n = 2, ρ = 7.87 g/cm³.
Galvanic coupling under oxygen-supply control
i_L is the oxygen-limited current density (15 µA/cm² here); A_a and A_c are anode and cathode areas. The larger the cathode, the more current the anode must dissolve.
Copper hulls and zinc
In 1824 Davy proposed attaching blocks of iron or zinc to the copper sheathing on Royal Navy hulls to stop its corrosion — sacrificing a lower-potential metal to protect copper, the start of modern cathodic protection. Corrosion stopped, but an unexpected problem appeared: with copper no longer dissolving, the barnacles and weed it had kept away covered the hulls. It is remembered as a case where stopping corrosion also stopped another useful effect.
Worked examples
Example 1: steel–copper contact and area ratio (seawater)
| Copper/steel area | Steel corrosion current | Steel corrosion rate | Potential rise |
|---|---|---|---|
| Steel alone | 15 μA/cm² | 0.174 mm/yr | — |
| 0.1 | 16.5 μA/cm² | 0.19 mm/yr | +0.002 V |
| 1 | 30 μA/cm² | 0.35 mm/yr | +0.018 V |
| 10 | 165 μA/cm² | 1.91 mm/yr | +0.062 V |
| 100 | 1,515 μA/cm² | 17.6 mm/yr | +0.120 V |
Potential rise follows the Tafel slope of steel dissolution (60 mV per decade). A steel plate with copper rivets (ratio 0.1) is barely affected, but steel rivets in a copper plate (ratio 100) dissolve at 17.6 mm/yr.
Example 2: cathodic protection with zinc
| Condition | Protection current | Charge per year | Zinc consumed |
|---|---|---|---|
| Bare steel 100 m² | 15 A | 131,400 Ah | about 198 kg/yr |
| Coated (2% bare) | 0.30 A | 2,628 Ah | about 4.0 kg/yr |
Zinc's theoretical capacity is 820 Ah/kg; 780 Ah/kg in practice with 0.85 utilisation was used.
If a ratio of 100 gives 100 times the corrosion, does copper far from the joint count too?
Good catch. Example 1 is an upper bound assuming every part of the cathode receives electrons equally. In reality current must travel through the electrolyte to reach distant cathode, and its resistance drops potential. In well-conducting seawater, cathode far away still counts, so the area effect shows almost fully. In thin films of rain or condensation, conductivity is poor and only cathode right next to the joint works — but then corrosion concentrates right beside the joint as a deep groove. How much cathode is effective depends on electrolyte conductivity and water thickness. Corrosion simulation therefore solves the Laplace equation for potential in the electrolyte with surface reaction rates as boundary conditions to obtain the current distribution.
Modelling workflow
- Measure polarization curves (potential vs current) of each metal in the service water.
- Set electrolyte conductivity and water-film thickness.
- Solve electrolyte potential with surface polarization curves as boundary conditions.
- Convert computed current to thickness loss with Faraday's law.
- If needed, update geometry and re-solve over time.
“A steel fitting joined to stainless piping”
In a plant's cooling-water piping, only a short carbon-steel fitting inserted in stainless pipe perforated within a few years. Its inner surface was tiny compared with the whole line, and the long stainless surfaces around it acted as cathode. Potential-distribution analysis showed corrosion concentrated at both ends of the fitting at the stainless boundary, matching the actual holes. The fitting was replaced with stainless, and insulating flanges were installed to separate dissimilar metals electrically where unavoidable.
Common mistakes
Mistakes and fixes
| Mistake | Effect | Fix |
|---|---|---|
| Judging by potential difference only | Area effect missed | Evaluate area and current |
| Treating all cathode as effective | Distribution wrong in thin films | Solve electrolyte potential |
| Average loss only | Local perforation missed | Resolve current distribution |
| Polarization curves from another water | Rate badly wrong | Measure in service water |
| Sizing protection on bare area when coated | Zinc oversized | Include coating breakdown |
I'd like to learn related topics.
Related pages include corrosion progression analysis, electroplating, electrochemical machining, water electrolysis and battery degradation.