Mechanistic corrosion science, engineered for the field
The technology behind Corrosion Prognostics is built from first principles — no empirical look-up tables, no site-specific calibration required.
Three complementary data sources
The atmospheric corrosion model draws on three sources of environmental data synergistically. All three together give the best predictions, but any two are sufficient to generate meaningful results.
Sensor Data
Time of wetness, surface and atmospheric temperature, and relative humidity — measured every minute and specific to the deployment location. Highest resolution and most localised data source.
Meteorological Data
Atmospheric temperature, relative humidity, and precipitation from weather station networks. Widely available at hourly resolution. Requires data workup before use in the model.
Pollutant Deposition Data
Wet deposition kinetics, ionic speciation, and rain concentrations from the National Atmospheric Deposition Program (NADP) — approximately 400 sites updated monthly. Used to generate site-specific pollutant fingerprints.
How the Atmospheric Corrosion Model works
The Atmospheric Corrosion Model (ACM) is a time-based, mechanistic model. The Time of Wetness sensor determines when corrosion events occur; the model then calculates the corrosion that takes place during each wetness event, accounting for droplet salt composition, concentration, surface coverage, and oxygen reduction kinetics.
General corrosion is summed across events to give cumulative weight loss over time. Pitting events are collated into pit size distributions, allowing statistical characterisation of the localised corrosion damage.
The model requires no empirical corrosion data from the site — it can generate predictions for any location for which environmental data is available, making it suitable for pre-design location assessments as well as ongoing condition monitoring.
In one Key West deployment, the model identified that 90% of the total pitting damage over an 18-month period occurred within just 5 days — during a series of high-humidity, high-salt-loading wetness events following an extended dry spell. Without the model’s causal analysis, this insight would have been invisible.
Validated at six sites across the USA and internationally
Model predictions for both steel sectional weight loss and aluminium pit size distributions have been benchmarked against independently measured corrosion coupon data at six deployment sites representing a wide range of climatic and pollutant environments. Corrosion section losses ranged from 5 µm (arid Phoenix) to 50 µm (marine Hawaii) over the measurement period.
Phoenix, Arizona
Low corrosion environment. Sectional losses ~5 µm. Validated steel and Al pit predictions.
Salt Lake City, Utah
Moderate environment. De-icing salt influence in winter months captured by model.
Dayton, Ohio
Losses ~14 µm. Industrial/agricultural pollutant mix. Hill AFB and WPAFB sites included.
Key West, Florida
Highest corrosion severity. Up to 40 µm sectional loss. Largest pit diameters observed (~150 µm Al).
Kaneohe, Hawaii
MCB Hawaii site. High humidity and chloride loading. Pit counts up to ~1,500 per four coupons.
International Site
Additional international validation site included in the full dataset.
Steel coupon — AR site, 18 months
Steel coupon — HI site, 18 months
Aluminum Key West Optical
Aluminum Key West 3D
Aluminium pit analysis used the Keyence VR-5100 3D microscope at sub-micrometre resolution. Coupons were mounted in epoxy, ground flat, and polished to 1 µm before measurement of pit depths, volumes, aspect ratios, and equivalent diameters — providing a rigorous independent benchmark for model predictions.
Future innovations in corrosion monitoring
The current platform is the foundation for a broader roadmap of sensor development and model expansion. All future products follow the same development principle: no commercial release without rigorous validation.
Pitting corrosion sensors
Multiple designs for steel, aluminium, and high-performance alloys — capable of detecting the earliest stages of pit nucleation before structural damage occurs.
Coatings breakdown sensors
Real-time monitoring of protective coating integrity — enabling intervention before substrate corrosion initiates rather than after visible failure.
Under-insulation corrosion
Sensors placed within insulation layers to provide continuous visibility of the corrosion environment — addressing one of the industry’s most hazardous hidden risks (CUI).
Expanded material models
Extension of the predictive modelling framework to cover stress corrosion cracking, crevice corrosion, galvanic attack, and additional engineering alloys.