CORROSION MECHANISMS, PROTECTIVE COATINGS & MATERIAL FAILURE

“All are from the dust, and to dust all return.”
— Ecclesiastes 3:20

With enough time, water, temperature change, ice, and wind, a mountain becomes rocks. Those rocks become gravel, sand, silt, and clay. Moisture, fungi, microorganisms, insects, sunlight, and chemical reactions turn a fallen tree into soil and return its materials to the surrounding ecosystem. Heat, pressure, and time turn buried algae, plankton, and plants into crude oil and natural gas.

In days, humans do what nature can take thousands or millions of years to accomplish. We break molecular chains, create cross-links, smelt ores, fire minerals, and cast, roll, shape, and forge materials into zinc, copper, concrete, steel, and aluminum—and then into tools, girders, aircraft skins, pipelines, semiconductors, and printed circuits. But for thousands of years, the same forces that turn a mountain into sand have also been breaking down human inventions. Even the word corrosion comes from the Latin corrodere: to gnaw away, consume, or wear away.

Material Production

Nature does not give us bridge girders, water pipes, aircraft panels, or circuit boards. It gives us rock, ore, sand, clay, salt, trees, and the remains of living things. We separate those materials, heat them, dissolve them, strip oxygen from them, force atoms together, and pull molecules apart. Then we roll, cast, draw, forge, cure, and machine what remains.

The finished product may look simple. The process used to make it is not. That process also helps determine how the material will eventually fail.

A Steel Bridge Girder

A steel bridge girder begins as iron-bearing rock. Most of the iron in a rock is bound to oxygen in minerals such as hematite and magnetite. These minerals formed as iron reacted with oxygen through geological processes over immense spans of time. Making the steel for the girder begins by reversing part of that reaction.

In an integrated steel mill, iron ore enters a blast furnace with coke and limestone. The coke burns and forms carbon monoxide. The carbon monoxide takes oxygen away from the iron oxide. Limestone combines with unwanted minerals and carries them into slag. Molten iron collects below.

That iron contains too much carbon for most uses. In a basic oxygen furnace, oxygen is blown through the metal. Carbon leaves as gas. Silicon, phosphorus, and other impurities enter the slag. Steelmakers then add measured amounts of manganese, chromium, nickel, molybdenum, and other elements to change the strength, toughness, hardness, weldability, and corrosion resistance.

Bright electric arc furnace operating inside the Nucor Steel facility in Arkansas — Electric arc furnace steelmaking at Nucor Steel in Arkansas. **Source:** U.S. Department of Energy. Credit: Chris Yountz.

An electric arc furnace reaches the same molten state differently. Graphite electrodes strike an arc through scrap steel, direct-reduced iron, or a mixture of iron-bearing feedstocks. The charge melts. Its chemistry is corrected. It is poured and cast.

Side-view schematic of an electric arc furnace. Source: Wikimedia Commons. Creator: 0x24a537r9. License: CC BY-SA 3.0.

The steel becomes slabs, blooms, and billets. A rolling mill turns a slab into plate. That plate is cut into flanges and a web, fitted together, and welded into the long I-shaped member that carries a bridge deck. Other rolls and dies turn steel into rail, bar, wire, automobile sheet, pipe, and fasteners. Heat treatment rearranges the internal structure. Forging drives the metal into shape. Welding joins those shapes into bridges, vehicle bodies, tanks, tools, pipelines, ships, and buildings.[1][2]

The finished part looks permanent. It is not. It is iron being held in a useful form by the energy and control humans put into making it.

A 45-ton electric arc furnace melting steel scrap at Metal Ravne in Slovenia; the recording includes the operating sound of the furnace. Source: Wikimedia Commons. Credit: Sounds of Changes / Technical Museum of Slovenia; video recordist Barbara Grilc. License: CC BY 3.0.

A Zinc Roof Panel and Rain Gutter

A zinc roof panel begins underground, most often as zinc sulfide locked inside the mineral sphalerite. The ore is crushed until the zinc-bearing particles are small enough to separate from the surrounding rock. Water, chemicals, and air bubbles carry those particles to the surface in a process called froth flotation, producing a zinc-rich concentrate.

At the refinery, the concentrate is roasted in air. Oxygen joins with the zinc while sulfur leaves as sulfur dioxide, changing zinc sulfide into zinc oxide. Sulfuric acid then dissolves the zinc oxide and turns it into a zinc-sulfate solution. Impurities are removed, and electricity pulls the dissolved zinc out of the solution and deposits it onto metal cathodes. The zinc is stripped from the cathodes, melted, and cast into slabs. This sequence—roasting, leaching, purification, and electrowinning—produces the zinc used in sheet metal and other finished products.[41]

For architectural sheet, zinc is commonly mixed with small amounts of copper and titanium to improve its strength, stiffness, and resistance to movement over time. Rolling mills squeeze the cast slabs into long, thin coils. The coils are cut to width and bent, folded, or roll-formed into standing-seam roof panels, wall cladding, flashing, gutters, and downspouts.

An Aluminum Aircraft Skin

The Bayer process converts bauxite into alumina. Source: Wikimedia Commons. Creators: Andreas Schmidt; English derivative by Hans Erren. License: CC BY-SA 3.0.

An aluminum aircraft skin begins as bauxite, an ore in which aluminum is chemically bound to oxygen. First, crushed bauxite is mixed with hot sodium hydroxide. The aluminum dissolves into the liquid while most of the iron oxide, silica, and other impurities are filtered out. The aluminum is then drawn back out of the liquid as solid aluminum hydroxide. That material is washed and heated to remove water, leaving a dry powder called alumina: aluminum oxide. Then electricity removes the oxygen. In the Hall-Héroult process, alumina dissolves in a molten fluoride bath. A large direct current passes through the cell. Liquid aluminum settles at the bottom while carbon anodes are consumed above it.[5]

Hall-Heroult cell used to reduce alumina to aluminum metal. Source: Wikimedia Commons. Creator: Parcly Taxel. License: CC BY-SA 4.0.

Pure aluminum is soft. Magnesium, silicon, copper, manganese, and zinc are added to make useful alloys. A rolling mill reduces an aluminum slab into long coils of sheet. Aircraft producers stretch, cut, form, drill, rivet, and heat-treat that sheet. The same metal can also be extruded, forged, machined, and welded into window frames, heat exchangers, cans, marine hardware, and electronic housings.

The aluminum in an aircraft skin is different from the aluminum in a cast pump housing or a power-transmission line. Copper gives an aircraft alloy greater strength. Silicon helps molten aluminum flow into a mold. Nearly pure aluminum carries electricity efficiently. All three are called aluminum, but their different compositions cause them to corrode in different ways.

A Copper Water Pipe and Printed Circuit Trace

Molten copper being poured into a row of anode molds inside a copper-processing facility — Molten copper being poured into anode molds. **Source:** U.S. Geological Survey. Mineral Commodity Summaries 2026 cover image.

A copper water pipe and the tiny copper paths on a circuit board begin with the same refined metal. Copper ore is crushed and ground, then mixed with water so air bubbles can separate the copper-bearing particles from the waste rock. The concentrated material is melted, and oxygen is blown through it to remove most of the sulfur and iron. The remaining copper is refined again and cast into thick, impure plates called anodes. Electricity moves the copper from those anodes through a liquid solution and deposits it in nearly pure layers on cathodes.

Other ores are leached with acid. Solvent extraction concentrates the dissolved copper. Electrowinning plates it back into metal. Old wire, pipe, and scrap can be melted and returned to the same production chain.[6]

To make a water pipe, a heated copper billet is pierced through the center, forced into a long tube, pulled through progressively smaller dies, straightened, and cut to length. To make a printed circuit board, thin copper foil is bonded to glass-reinforced resin. A pattern is transferred onto the copper, and the exposed metal is etched away, leaving the narrow paths that carry power and signals. The same refined copper is also drawn into wire, rolled into roofing, and shaped into bus bars, motor windings, transformers, heat exchangers, and electrical contacts.

A copper contact can fail when a corrosion film too thin to see interrupts an electrical signal.

A Reinforced-Concrete Bridge Deck and Building Column

A bridge deck or building column begins with limestone, clay, sand, stone, reinforcing steel, and water. Limestone and clay supply the calcium, silicon, aluminum, and iron needed to make cement. The materials are crushed, blended, and heated in a rotating kiln to about 1,450 degrees Celsius. The heat drives water and carbon dioxide out of the raw materials and rearranges what remains into hard lumps called clinker. Removing that water is an important part of creating dry cement minerals that are ready to react when water is added again. The clinker is cooled and ground with a small amount of gypsum to make portland cement.[7]

To make concrete, water is added back to the cement along with sand and crushed stone or gravel. Some of the water makes the mixture fluid enough to pour, but much of it reacts with the cement and becomes part of the hardened material. Those reactions create new solids that spread around and between the sand and stone, joining the loose particles into a single mass of concrete.

The finished concrete depends on what happened before it hardened. Too much water leaves a more open path through the cement paste. Poor consolidation leaves voids. Poor curing interrupts hydration. Thin cover and cracks shorten the distance water, salt, oxygen, and carbon dioxide must travel to reach the steel.

A Graphics Processing Unit (GPU)

A GPU used in a hyperscale AI data center begins as quartz. Quartz is a mineral made of silicon and oxygen chemically bound together as silicon dioxide. The quartz is crushed and mixed with carbon from coal, coke, or wood before entering an electric furnace. At temperatures approaching 2,000 degrees Celsius, the carbon takes the oxygen from the quartz. The oxygen leaves the furnace with the carbon as gas, and molten silicon collects below. The furnace has transformed quartz into metallurgical-grade silicon, which must be purified much further before it can be used in a processor.

The silicon is ground and reacted with hydrogen chloride to create a volatile silicon compound called trichlorosilane. Distillation separates that compound from boron, phosphorus, iron, and other unwanted elements. The purified vapor is then passed over heated silicon rods, where it breaks apart and deposits extremely pure silicon. This conversion from solid silicon to a chemical vapor and back again removes impurities until they can be measured in parts per billion.

The purified silicon is broken into pieces and melted again. A small seed crystal touches the surface and is slowly pulled upward while it rotates. Silicon atoms leaving the liquid arrange themselves around the seed in the same repeating pattern. Over time, a heavy cylinder of single-crystal silicon grows behind it. That cylinder is cut into thin wafers, ground flat, polished, and cleaned until each surface is nearly flawless.

The wafer is then changed hundreds of times. Some areas are deliberately exposed to oxygen to create thin insulating layers. Light-sensitive material is spread across the surface and exposed through a pattern. Chemicals remove selected areas. Boron, phosphorus, and other elements are driven into the silicon to change where it will and will not conduct electricity. Thin layers of copper and other metals are added to connect those regions. Repeating these steps builds tens of billions of microscopic transistors and the metal paths between them.

The wafer is cut into individual GPU dies. Each die is mounted on a package built from copper conductors, solder connections, insulating polymers, and layers of ceramic or glass reinforcement. Thousands of microscopic connections carry power and data between the silicon and the systems around it.

Material Corrosion

Once a steel girder, zinc roof, aluminum aircraft skin, copper pipe, concrete column, or GPU enters service, water, oxygen, salt, pollution, heat, and temperature change begin working on it. Moisture carries dissolved chemicals into cracks, seams, pores, and microscopic spaces. Salt makes that moisture more conductive. Higher temperature or a thermal gradient can increase the severity of exposure, depending on the corrosion mechanism, while contact between different metals can create electrical paths that concentrate the attack. These forces do not have to return a material to its original ore; they only have to change it enough that the bridge, roof, pipe, building, aircraft, or processor can no longer do its job.[8]

How the Steel Bridge Girder Corrodes

Close-up of corroded steel showing thick orange-brown rust scale and dark, fractured corrosion-product layers — Layered rust scale on a corroded steel specimen. The close-up shows the buildup and cracking of corrosion products. **Source:** GPI Labs.

Making the girder required energy to separate iron from the oxygen to which it was bound in iron ore. Corrosion pushes the metal in the opposite direction. When steel is exposed to the right combination of water and oxygen, some of that iron can oxidize and eventually return to compounds such as iron oxides and hydroxides—the materials we recognize as rust.[8]

For that corrosion to proceed, an electrochemical cell must form. It needs three parts: an anode, where iron is lost; a cathode, where another reaction consumes the electrons released by the iron; and an electrolyte, usually a film of water containing dissolved salts and other ions. The steel itself provides the electrical path between the anodic and cathodic areas.

FHWA diagram of an electrochemical corrosion cell showing iron dissolving at the anode, electrons flowing through the steel to the cathode, oxygen reduction at the cathode, and ionic current through the electrolyte — Electrochemical corrosion cell showing iron loss at the anode, electron flow through the steel, and the cathodic reaction that completes the circuit. FHWA illustrates the process using reinforcing steel in concrete; the same basic anode-cathode-electrolyte relationship applies to wet structural steel. **Source:** Federal Highway Administration, Figure A-1, FHWA-RD-01-163 .

At the anode, iron atoms in the steel are oxidized. Each iron atom leaves the metal as a positively charged ferrous ion, Fe²⁺, and releases two electrons into the steel:

Fe → Fe²⁺ + 2e⁻

In the formula, Fe is an iron atom in the girder, Fe²⁺ is that iron after it has entered the electrolyte as a dissolved ion, and 2e⁻ are the two electrons left behind in the steel. This is the actual loss of metal: the anodic area becomes thinner as iron atoms leave the girder.

Those electrons travel through the steel to a cathodic area. In ordinary atmospheric corrosion, oxygen dissolved in the water film reacts there with water and the incoming electrons to form hydroxide ions:

O₂ + 2H₂O + 4e⁻ → 4OH⁻

Here, O₂ is dissolved oxygen, H₂O is water, and 4e⁻ are electrons supplied by the anodic reaction elsewhere on the steel. The products, 4OH⁻, are negatively charged hydroxide ions in the water. The iron ions produced at the anode and the hydroxide formed at the cathode then take part in additional reactions that produce iron hydroxides, oxyhydroxides, and oxides—the corrosion products commonly called rust.

The water film completes the corrosion cell by carrying ionic current between these areas. The dissolved iron produced at the anode reacts further with oxygen, water, and hydroxide to form iron hydroxides, oxyhydroxides, and oxides. Those corrosion products collectively become rust.[8][9]

The anode and cathode do not have to be different pieces of metal. They can develop inches apart—or much closer—on the same girder. A difference in oxygen, moisture, salt concentration, surface condition, or geometry can be enough to make one wet area behave differently from the area beside it.

That is why corrosion on a bridge is rarely uniform. Water reaches the girder through leaking deck joints, failed drains, runoff, rain, condensation, and salt-laden road spray. It collects at stiffeners, flange edges, seams, bolted connections, and debris-covered horizontal surfaces. One area may dry quickly while another remains wet. A crevice may have less access to oxygen than the exposed steel beside it. Each difference can help establish separate anodic and cathodic areas.

Deicing salt makes the situation worse because chloride-containing water is a better electrolyte than relatively clean water. The dissolved ions allow the corrosion cell to carry current more readily, while chloride can also interfere with protective surface films. Salt trapped in a seam or debris pile can be redissolved each time the bridge becomes wet, allowing the same location to be attacked again and again.

Rust does not restore the original steel surface. On ordinary structural steel, the corrosion products are generally porous and irregular. They can retain water and salt, crack, and separate from the surface, leaving fresh steel available for another corrosion cycle. Where rust develops between tightly connected plates, its accumulation can also force those plates apart in the form of damage commonly called pack rust.

The result is structural material loss concentrated at vulnerable locations. A web can thin beneath a leaking joint. Flange edges can lose section. Bolt heads and shanks can be consumed. Connections can separate as corrosion products accumulate between plates. The corrosion cell may begin in a microscopic film of water, but repeated wetting can eventually remove enough steel to reduce the girder's load-carrying capacity.

How the Zinc Roof Panel and Gutter Corrode

The first film of rainwater on a zinc roof creates small corrosion cells across the surface. At anodic areas, zinc atoms are oxidized. Each zinc atom leaves the metal as a positively charged zinc ion and releases two electrons into the remaining metal:

Zn → Zn²⁺ + 2e⁻

In the formula, Zn is a zinc atom in the roof or gutter, Zn²⁺ is that zinc after it has entered the water film as a dissolved ion, and 2e⁻ are the two electrons left behind in the metal. This is the actual loss of zinc from the surface: the anodic area becomes thinner as zinc atoms dissolve into the electrolyte.

The electrons move through the zinc to cathodic areas, where oxygen dissolved in the water consumes them. The zinc ions produced at the anode then react with oxygen, water, and hydroxide to form zinc oxide and zinc hydroxide. Carbon dioxide from the air reacts with those products and gradually helps convert them into a thin layer containing basic zinc carbonate.

That layer is the zinc patina. It is made by corrosion, but once it becomes dense and firmly attached, it separates the metal beneath it from water and oxygen and slows the reaction that created it. Normal cycles of rain and drying allow the patina to develop.

The reaction changes when water becomes trapped beneath a panel, between folded seams, under wet leaves, or in the bottom of a poorly drained gutter. The trapped surface stays wet and receives little fresh air or carbon dioxide. Zinc oxide and zinc hydroxide continue forming, but they do not develop into the same dense carbonate patina. Loose white corrosion products collect on the surface, hold more moisture, and expose new zinc as they wash or flake away. Repeated wetting consumes the sheet until the gutter develops pinholes or the roof panel becomes too thin to remain watertight.

Salt accelerates the same electrochemical process. Chloride makes the water more conductive and interferes with the protective surface. Acidic rain and sulfur-bearing pollutants react with the patina and create zinc compounds that dissolve more easily. Rain then carries those compounds away, exposing fresh zinc and allowing another cycle of corrosion to begin. Roof pitch, drainage, rainfall, airborne salt, pollution, and the length of time the surface remains wet all affect how quickly zinc is lost.[42]

Copper creates a more concentrated attack. Water flowing from a copper pipe, flashing, or roof section can carry copper ions onto the zinc. Some of that copper deposits on the surface and becomes a cathode. The surrounding zinc becomes the anode and dissolves to supply the electrons. Direct contact between copper and zinc creates the same galvanic cell. Most of the roof may continue forming a protective patina while the zinc directly below the copper runoff corrodes deeply and fails. The Metal Construction Association identifies both direct contact and water runoff between dissimilar roofing metals as causes of accelerated corrosion.[43]

How the Aircraft Skin Corrodes

An aircraft skin is thin because every additional pound reduces the weight available for fuel, passengers, or cargo. When aluminum is exposed at a scratch, cut edge, fastener hole, or damaged surface, it reacts with oxygen almost immediately. The reaction forms a thin layer of aluminum oxide that clings tightly to the metal. If the layer is scratched in clean air, exposed aluminum reacts with oxygen and forms another protective film.

Chloride reaches an aircraft as sea salt carried by coastal air, ocean spray, and marine operations. It can also remain in dirt, contaminated water, fingerprints, and deposits trapped during service or maintenance. Humidity, rain, wash water, or condensation dissolves the salt and carries it into seams, lap joints, fastener holes, and scratches. As the water evaporates, the chloride becomes more concentrated.

Close-up of zinc-coated bridge-cable wires with localized black pits interrupting the white zinc-corrosion layer — Localized pitting beneath a stainless-steel strap on zinc-coated bridge-cable wires. **Source:** Federal Highway Administration, Figure 136.

The concentrated chloride attacks weak points in the aluminum oxide film. Once it breaks through, the exposed aluminum beneath it begins dissolving. Aluminum ions react with water inside the opening and make the trapped liquid more acidic. Chloride moves into the pit to balance the electrical charge. The acidic, chloride-rich liquid attacks the aluminum faster, drawing in more chloride and driving the pit deeper. A narrow opening at the surface can hide a cavity extending far into the thin aircraft skin.[11]

The alloy itself can create additional corrosion cells. Copper-rich particles act as small cathodes while the aluminum immediately around them becomes the anode and dissolves. Heat treatment and welding can concentrate alloying elements along grain boundaries. Corrosion then follows those boundaries beneath the surface. As corrosion products grow between the grains, they can lift and separate layers of metal in a form of attack called exfoliation.

Fasteners and adjoining materials create larger galvanic cells. Stainless steel, titanium, copper, or carbon-fiber composite can act as the cathode when connected to wet aluminum. The aluminum becomes the anode and corrodes to supply the electrical current. A small area of exposed aluminum connected to a much larger cathodic surface can suffer severe attack around a fastener or joint.

Most of the oxide-covered skin may still look clean while corrosion drives beneath a fastener, along a grain boundary, or into a chloride-filled pit. The same thin oxide film that protects the rest of the aircraft can conceal how deep the localized damage has become.

How the Copper Pipe and Circuit Trace Corrode

A copper water pipe and a copper circuit trace begin corroding through the same basic reaction. Copper atoms give up electrons and combine with oxygen, water, sulfur, chlorine, or other materials in the surrounding environment. What happens next depends on where the copper is used. A pipe may continue working until corrosion penetrates its wall. A circuit trace can fail after losing only a microscopic amount of metal.

Inside a water pipe, copper first reacts with dissolved oxygen and forms a thin oxide layer. Under the right conditions, that layer becomes dense enough to slow further corrosion. Its stability depends on the water’s pH, alkalinity, dissolved oxygen, disinfectants, chloride, sulfide, ammonia, temperature, and flow rate. A stable film protects the pipe. A porous, damaged, or chemically unstable film exposes the copper beneath it.[12]

Fast-moving water can strip the protective film away and produce erosion-corrosion at bends, restrictions, and other areas of turbulence. Stagnant water and deposits create isolated areas with different amounts of oxygen and different chemistry. Chloride or other contaminants can concentrate inside those areas and drive pits into the pipe wall. Repeated damage and regrowth of the film can produce general thinning, deep pits, or a pinhole leak.

On a circuit board, most copper traces are covered, but copper remains exposed at contacts, component connections, damaged areas, and defects in the protective surface. Humidity or condensation creates a microscopic layer of water across the board. Salts from coastal air, fingerprints, manufacturing residues, or contaminated dust dissolve into that water and make it conductive. Sulfur- and chlorine-bearing gases from polluted air can dissolve into the same moisture and react with the copper.

The resulting copper oxides, chlorides, and sulfides can consume a thin trace or raise the resistance at an electrical connection. When voltage is present, dissolved copper ions can also move across the wet surface and deposit between neighboring conductors. These branching deposits can bridge a gap and create a short circuit.

The pipe must lose enough copper to leak. The circuit trace only has to lose enough copper—or grow enough corrosion product—to interrupt a signal.

How the Bridge Deck and Building Column Corrode

The reactions between cement and water created the solid material holding the bridge deck or building column together. They also left microscopic pores running through the hardened cement paste. Water moves through those pores and through larger cracks, carrying salt, acids, carbon dioxide, and other chemicals into the concrete. Concrete made with too much water, poorly consolidated concrete, and concrete that was allowed to dry before it cured contain more open paths for that movement.

Marine bridge piling with longitudinal cracks, missing concrete and exposed corroded reinforcing steel — Cracking and spalling on a marine bridge piling have exposed corroded reinforcement. **Source:** Federal Highway Administration, Figure 1.

On a bridge deck, rain and melting snow carry deicing salt into the surface. Some of the water soaks into the concrete and freezes when the temperature falls. As water freezes in saturated pores, hydraulic and osmotic pressures can develop within the cement paste. One freeze may cause little visible damage. Repeated freezing and thawing widen the pores, connect small cracks, loosen pieces of aggregate, and break thin layers from the surface. The roughened surface then absorbs more water and salt during the next winter.[52]

Carbon dioxide from the air also enters the pores. It dissolves in the water and reacts with the alkaline compounds created when the cement hardened. The reaction converts some of those compounds into calcium carbonate and lowers the pH of the concrete. This process, called carbonation, moves inward from the surface. The concrete may remain hard, but it gradually loses the alkaline chemistry that protects the reinforcing steel.[53]

Concrete specimen with corrosion and spalling concentrated around a reinforcing bar at the concrete-air interface — Corrosion and spalling around reinforcing steel at a concrete-air interface. **Source:** Federal Highway Administration, Figure 11.

Chloride from deicing salt follows the water toward the reinforcing bar. When enough chloride reaches the steel, it breaks through small areas of the protective film created by the alkaline concrete. The exposed steel begins to corrode. Iron is converted into rust, which occupies more space than the metal it replaced. The hardened concrete restrains that expansion until pressure cracks the concrete along the bar. Water and salt enter through the new crack, more steel corrodes, and sheets of concrete separate and fall from the deck as spalls.[14]

The sand and stone inside the concrete can also take part in the damage. Some aggregates contain forms of silica that react with alkalis from the cement when water is present. The reaction forms a gel that absorbs water and swells. That swelling pushes the cement paste apart and creates networks of cracks. Sulfates carried into a foundation or column from soil or groundwater can react with other cement compounds and form expanding minerals that soften and fracture the concrete from within.[52][53]

How the GPU Corrodes and Fails

A hyperscaler controls the temperature, humidity, and cleanliness of the AI data center, but each GPU still experiences its own changing environment. The processor heats rapidly under an AI training load and cools when that load falls. Server fans pull large volumes of air across air-cooled GPUs. Liquid-cooled systems place cold plates, hoses, seals, and fittings close to the processor. Changes in workload, cooling-water temperature, airflow, or room humidity can move a surface below the dew point and create condensation. A leaking fitting can introduce liquid directly.

The air entering the server carries more than water vapor. Dust arrives from outdoor air, construction, packaging, clothing, and normal maintenance. Coastal air can carry chloride from sea salt. Industrial and urban air can contain sulfur- and chlorine-bearing gases. Facilities using outside-air cooling can bring more of those contaminants into the data center. Ionic residue left during manufacturing, cleaning, or handling adds another source. Dust and residue absorb moisture, creating a thin, conductive liquid across surfaces that otherwise appear dry.

The GPU package is built to keep that environment away from the silicon and its microscopic connections. Repeated heating and cooling make that protection harder to maintain. Silicon, copper, solder, ceramic, glass, and polymer expand by different amounts as their temperatures change. The movement can fatigue solder connections, separate bonded layers, and open microscopic paths along the edge of the package. Moisture and contamination can then reach copper conductors, solder bumps, metal finishes, and other vulnerable features.

Once a conductive moisture film reaches two metals, a corrosion cell can form. Copper, tin, nickel, aluminum, and solder do not all react at the same rate. One becomes the anode and begins dissolving while another supports the cathodic reaction. Chloride increases the conductivity of the water and can break through protective oxides. Sulfur-bearing gases can react with exposed copper or silver. Heat from the operating GPU accelerates these reactions, while repeated condensation supplies new water.

Electrical power gives the dissolved metal a direction to travel. Metal ions leave one energized conductor, move through the moisture film, and deposit near another conductor. The deposit can grow into a branching metallic filament called a dendrite. When the dendrite crosses the microscopic gap between conductors, current follows it and creates a short circuit. In another location, corrosion removes enough metal from a solder bump or copper path to increase resistance, interrupt data, or cut power to part of the processor.

Advanced AI GPUs place the processor and stacks of high-bandwidth memory within the same package. Thousands of microscopic connections carry data and power across extremely short distances. A small amount of moisture, chloride, sulfur, or manufacturing residue can therefore affect many connections within a small area. IEC 60068-2-60 uses controlled mixtures of corrosive gases to apply the relevant stressors from environments in which sulfur- and chlorine-bearing pollutants attack electronic materials.[13]

A GPU does not need orange rust or a visible hole to fail. A corroded solder connection, a thinned copper path, or a dendrite only a few micrometers long can disable the processor. The factory transformed quartz into pure silicon and built billions of transistors around it. The data-center environment only has to allow a small amount of metal to migrate across—or contamination to bridge—one microscopic gap to stop it from working.

Coating Processes

A coating is a deliberately placed interruption.

Steel, aluminum, copper, concrete, and electronics will react when the environment reaches them. A coating makes that path longer, slower, or chemically less favorable. It keeps water away. It slows oxygen. It blocks chloride. It absorbs ultraviolet light. It gives chemicals something else to attack. It changes the metal surface, or it puts a more active metal in front of the one we need to preserve.

Barrier coatings work by distance. Water and ions must travel through polymer, glass, ceramic, or metal before reaching the substrate. Inhibitive coatings interfere with reactions at the surface. Sacrificial coatings put zinc or another active metal in electrical contact with steel. Conversion coatings react with the metal and replace an irregular surface with a more useful one.[16]

Most coating systems combine these ideas. A zinc-rich primer sacrifices zinc at a scratch. A thick epoxy makes a long path for water and salt. A polyurethane topcoat takes the sunlight that would damage the epoxy. The steel beneath carries the load. Each layer accepts a different part of the environment.

NASA's corrosion program at Kennedy Space Center treats protective coatings as a primary corrosion-control method because surface preparation, coating selection, and application have to act as one system.[15]

A coating cannot make the environment stop. It can only control how quickly that environment reaches the material and what happens when it gets there.

A Galvanized Guardrail

A galvanized guardrail begins as two materials produced separately: steel strong enough to carry the structural load and zinc used to protect that steel from corrosion.

Zinc commonly begins as sphalerite, an ore in which zinc is chemically bound to sulfur. The ore is crushed and ground into small particles, then mixed with water and chemicals in flotation tanks. Air is blown through the mixture, causing the zinc-bearing mineral particles to attach to bubbles and rise to the surface as a froth, while much of the unwanted rock sinks. The froth is collected to produce a zinc-rich concentrate. The concentrate is then roasted, converting much of the zinc sulfide into zinc oxide. Sulfuric acid dissolves the zinc oxide into solution, and an electric current is used to deposit metallic zinc from that purified solution.

The refined zinc can then be cast, rolled, alloyed with other metals, made into pigments and compounds, or used as a protective coating on steel.

Long steel components being lifted above a molten-zinc bath during hot-dip galvanizing — Steel components leaving a molten-zinc bath during hot-dip galvanizing. **Source:** Wikimedia Commons. Creator: jimpg2_2015. License: CC BY-SA 2.0.

The steel guardrail is cleaned, pickled to remove rust and mill scale, rinsed, fluxed, dried, and immersed in molten zinc. The zinc does not simply cool around the steel like a layer of wax. At galvanizing temperature, iron from the steel reacts with the molten zinc. A series of iron-zinc alloy layers forms at the steel surface, topped by a zinc-rich outer layer.[3][4]

The same process is used to protect transmission towers, light poles, platforms, stairs, fasteners, fencing, reinforcing bar, guardrails, and many other exposed steel structures. The steel provides the strength. The zinc provides the corrosion protection.

An Anodized Aluminum Window Frame

An anodized aluminum window frame has the surface of the aluminum converted into a thicker, more controlled aluminum oxide layer.

The finished aluminum extrusion is first cleaned to remove oil, grease, dirt, and residue from cutting, handling, and fabrication. It may then be placed in an alkaline etching solution that dissolves a very thin layer of aluminum from the surface. This removes minor surface imperfections and leaves a more uniform matte finish.

Etching does not dissolve every element in the aluminum alloy at the same rate. Copper, silicon, iron, and other alloying elements can remain behind as a dark or powdery residue called smut. A separate acid treatment removes that residue and exposes a clean aluminum surface. The part is rinsed between each chemical treatment so material from one bath is not carried into the next.

Diagram of an anodizing cell showing an aluminum anode, cathode, electrolyte, current flow and growth of the oxide layer — Electrolytic anodizing arrangement used to grow an oxide layer on aluminum. **Source:** Wikimedia Commons. Creator: Dipl-ing-metaller. License: CC BY-SA 3.0.

Only after that preparation is complete does the aluminum enter the anodizing bath. In the anodizing tank, the aluminum becomes the anode of an electrical circuit. The part is immersed in an acidic electrolyte, commonly sulfuric acid, and direct current is passed through the bath. Oxygen generated at the aluminum surface reacts with the metal and converts part of that surface into aluminum oxide. The coating grows both into the original metal and outward from it. Unlike paint, it is not simply deposited on top.

The newly formed oxide contains microscopic pores. Those pores make it possible to add color before the surface is sealed. Dyes can be absorbed into the pores, or metal salts can be deposited within them to produce durable architectural colors such as bronze, black, and champagne. Clear anodizing leaves the natural metallic appearance of the aluminum largely visible.

The anodized surface is then sealed, commonly with hot water, steam, or a chemical sealing treatment. Sealing changes the porous oxide structure and closes much of the pathway through which water and contaminants could otherwise enter.

The result is a hard aluminum oxide surface that is integral with the underlying metal. It improves corrosion resistance and wear resistance while preserving the metallic appearance of the aluminum.

The same process is widely used on architectural window frames, curtain-wall systems, storefront framing, railings, doors, signs, lighting components, cookware, electronics housings, and other aluminum products. The aluminum supplies the shape and structural properties. Anodizing deliberately converts its outer surface into a more durable oxide layer.

An Epoxy-Lined Water Pipe and Tank

Water, dissolved oxygen, salts, treatment chemicals, and other contaminants can attack the steel. One common solution is to separate the steel from that environment with an epoxy lining.

Epoxy begins as a liquid resin containing reactive chemical groups. A curing agent is mixed with the resin before application. The two components react with one another, joining relatively small molecules into a tightly connected three-dimensional polymer network. The liquid coating gradually becomes a hard, chemically resistant solid bonded to the pipe wall.

Automated industrial sandblasting system fitted with vertical arms inside a manufacturing facility — Automated abrasive-blasting equipment used to prepare industrial surfaces. **Source:** U.S. Department of Energy. Credit: Jake Stevens.

Before the epoxy is applied, the steel is abrasive blasted. High-velocity abrasive particles remove rust, mill scale, and other contamination while roughening the surface. The liquid epoxy is then mixed in the required proportions and sprayed over the prepared steel. As it cures, the coating hardens into a continuous barrier between the steel and the water flowing through the pipe.

Not every epoxy lining is the same. Changing the resin or curing agent changes properties such as flexibility, chemical resistance, curing temperature, and hardness. Pigments provide color and can contribute other properties. Mineral fillers can increase thickness, hardness, or abrasion resistance. Glass flakes create a more difficult path for water and chemicals trying to move through the coating. Ceramic particles can be added where severe wear is expected.

Epoxy can also be applied in other forms. Fusion-bonded epoxy begins as a dry powder. The steel is heated first, and the powder is sprayed onto the hot surface. It melts, flows together, and chemically cures into a continuous coating. Thick epoxy systems containing aggregate are used on floors and containment areas where additional thickness, wear resistance, or traction is required.

Epoxy systems protect the interiors of water pipes, valves, tanks, wastewater structures, process vessels, and secondary-containment systems. They are also used on buried pipelines, marine equipment, industrial floors, and other surfaces exposed to water, soil, chemicals, or abrasion.

AWWA C210 is one example of a specification governing liquid-epoxy coating systems for steel water pipe and fittings. In service, that lining must do more than survive immersion in water. It must remain attached through coating application, curing, handling, shipping, installation, burial, and years of exposure to the environment it was designed to separate from the steel.[17][44]

A Powder-Coated Electrical Enclosure

Electrical enclosures are typically fabricated, cleaned, pretreated, and coated before switches, wiring, controls, and other electrical components are installed. Areas that must remain electrically conductive or free of coating—such as grounding points, threaded connections, gasket surfaces, and selected mating surfaces—can be masked during the coating process. Powder coating is especially well suited to enclosure production because the dry powder can form a durable, relatively thick film in a single application without the solvent carrier used in conventional wet paint. Properly cured powder coatings can provide strong resistance to abrasion, impact, chipping, chemicals, and corrosion while also allowing oversprayed powder to be recovered and reused in many production systems.

The powder is manufactured by first blending solid resin, curing agents, pigments, fillers, and additives. That mixture is fed through a heated extruder, where heat softens the resin and rotating screws force the ingredients together until they are evenly dispersed. The hot material leaves the extruder as a thin sheet or ribbon. After cooling, it becomes hard and brittle enough to break into chips. Those chips are ground into fine powder, and oversized or undersized particles are removed so the remaining powder will spray and melt more consistently.

Before coating, the steel enclosure is cleaned to remove oil, dirt, rust, and fabrication residue. It may also receive a chemical pretreatment that improves corrosion resistance and gives the powder a better surface to bond to.

At the coating booth, the powder passes through a spray gun that gives the particles an electrical charge. The metal enclosure is electrically grounded. Because the charged powder particles are attracted toward the grounded metal, they collect on the surface instead of simply falling away like ordinary dust. This attraction helps coat vertical surfaces, corners, and other areas before the coating has been melted into place.

The coated enclosure then enters an oven. Heat causes the individual powder particles to soften and melt. They flow together until the separate particles disappear into a continuous film. Continued heating activates the curing reaction, linking the resin molecules into a solid coating that remains after the part cools.

Changing the resin changes the coating. Epoxy powders can provide strong adhesion and chemical resistance but generally have poor resistance to prolonged sunlight. Polyester powders are widely used where exterior weathering matters. Other formulations are designed for flexibility, abrasion resistance, appearance, electrical insulation, or particular chemical environments.

The same process is used on electrical cabinets, appliances, shelving, metal furniture, tools, bicycles, architectural components, machinery, and many other factory-finished metal products. Powder that misses the part can often be collected and reused, reducing material waste.

The smooth appearance of the finished coating can hide what lies beneath it. Powder coating does not eliminate the need for proper surface preparation. Oil, corrosion products, weak conversion coatings, or other contamination left on the metal can undermine adhesion and corrosion resistance even when the finished surface looks perfect. ASTM D7803 addresses preparation of hot-dip galvanized surfaces for powder coating because the performance of the finished system still begins with the condition of the surface beneath it.[18]

A Multi-Layer Protective Coating System

Severe environments often demand more than a single coating. A bridge girder, for example, may be abrasive blasted to clean and roughen the steel, coated with a zinc-rich primer, covered with a thick epoxy intermediate coat, and finished with a polyurethane or polysiloxane topcoat.

Each layer has a different job. The zinc-rich primer provides sacrificial protection if water reaches the steel through a scratch or damaged area. The epoxy builds a thick barrier that slows the movement of water, oxygen, and dissolved salts toward the metal. The topcoat protects the layers beneath it from sunlight, weathering, and physical wear while also providing the final color and appearance.

The same layered approach is adapted to other environments. A steel structure may use thermally sprayed zinc or aluminum beneath a sealer. A chemical tank may use a glass-flake epoxy lining to make it harder for liquids to penetrate the coating. A buried pipeline may use fusion-bonded epoxy directly on the steel with an abrasion-resistant outer layer to protect it during handling, installation, and burial. Concrete structures may receive elastomeric membranes that remain flexible as the substrate expands, contracts, or develops small cracks.

These systems work only if the layers function together. The steel or concrete must be properly prepared before the first coating is applied. Each coat must bond to the surface beneath it, cure correctly, and accept the next layer. A durable topcoat cannot compensate for contamination trapped beneath the primer, and additional thickness cannot repair a weak bond between layers.

ISO 12944-6 reflects part of this system approach by defining laboratory performance test methods used to assess protective paint systems. The finished protection is therefore not one film. It is a sequence of materials and interfaces designed to work together for the life of the structure.[19]

Coating Failures

Coatings fail from both sides. Sunlight, abrasion, and chemicals attack the outside. Water, salt, and corrosion products attack the interface underneath. Poor mixing, poor cleaning, wrong thickness, and missed edges can build failure into the film before the coated part leaves the shop.

How the Galvanized Guardrail Corrodes

The galvanized guardrail stands through rain, road salt, stone impact, wet leaves, and winter spray. Zinc protects the steel in two ways.

Close-up of the crystalline spangle pattern on a hot-dip galvanized zinc surface — Crystalline surface pattern on hot-dip galvanized plate. **Source:** Wikimedia Commons. Creator: Arakin. License: CC BY-SA 3.0.

First, the galvanized coating acts as a barrier between the steel and the environment. As zinc reacts with oxygen, water, and carbon dioxide, corrosion products form at the surface. Under suitable atmospheric conditions, oxides, hydroxides, and carbonates build into a relatively stable layer that slows further attack.

Second, zinc can protect exposed steel electrochemically. Zinc is easier to oxidize than iron. When zinc and steel are electrically connected beneath a film of moisture, the zinc tends to become the anode and releases electrons. Nearby exposed steel acts as the cathode and is protected.[10]

That is why a scratch through the coating does not necessarily cause the steel to rust immediately. Zinc surrounding the damaged area can continue to sacrifice itself to protect the exposed steel.

But that protection has limits. Make the bare area too large, keep the surface wet with chloride, expose it to strongly acidic or alkaline conditions, or physically wear away the zinc, and the coating is consumed more quickly. Eventually, if enough zinc is lost, the steel can no longer rely on it for protection. The steel then begins oxidizing on its own.

Galvanizing does not make corrosion disappear. It controls where corrosion happens and which metal is consumed first.

How the Anodized Aluminum Window Frame Fails

The anodized window frame is protected by the aluminum oxide layer created during anodizing. That layer is hard, tightly attached to the metal, and more corrosion resistant than bare aluminum. But it is not indestructible.

Failure can begin during manufacturing. If the aluminum was not cleaned or desmutted properly, the anodized layer may form unevenly. Poor control of bath chemistry, temperature, current, or treatment time can leave the oxide too thin or inconsistent. If the porous anodized surface is not sealed properly, water and dissolved contaminants can penetrate more easily into the coating.

Damage can also occur after the frame leaves the anodizing line. Cutting, drilling, machining, and installation can expose bare aluminum at edges, holes, and fastener locations. Scratches can break through the oxide. At those exposed areas, the aluminum loses the full protection of the anodized surface.

Water alone does not usually cause rapid attack because aluminum naturally forms its own thin oxide film. The greater problem is persistent moisture combined with contaminants, especially chlorides from coastal air, deicing salts, or other sources. Chloride can penetrate weak or damaged areas and promote localized corrosion beneath or through the protective oxide. Small pits may form and deepen while much of the surrounding surface still appears sound.

Strong acids and strong alkaline materials can attack the anodized layer directly. Alkaline cleaners, wet concrete, mortar, masonry runoff, and certain construction chemicals can stain, etch, or dissolve the oxide if they remain in contact with the frame. That is why an anodized surface that performs well outdoors can still be damaged quickly by the wrong chemical exposure during construction or maintenance.

Color can fail separately from corrosion protection. Dyes or deposited coloring materials can fade or change appearance under prolonged ultraviolet exposure or chemical attack even while the underlying anodized layer remains largely intact.

Anodizing protects aluminum by deliberately building a thicker and more controlled version of the oxide that aluminum naturally forms. Failure begins when that manufactured oxide is too thin, poorly sealed, chemically attacked, or physically broken. Once aggressive moisture reaches exposed aluminum, corrosion can begin at the defect even while the rest of the window frame remains protected.

How the Epoxy-Lined Water Pipe and Tank Fail

Microscope images comparing a coated surface before and after water exposure, with localized blisters after testing — Coating surfaces before and after water exposure, showing localized blister formation. This example illustrates the type of moisture-related damage investigated in coatings and linings. **Source:** GPI Labs.

The epoxy lining protects the steel only as long as it remains continuous, chemically resistant, and firmly bonded to the surface beneath it. Failure can begin in the coating itself, at the steel-coating interface, or at a defect created during application or service.

Even a fully cured epoxy is not completely impermeable. Water can slowly enter the polymer and occupy space between its molecular chains. Depending on the formulation, temperature, and exposure, the lining can absorb water, swell, soften, and become more permeable. Prolonged exposure can also remove soluble components or chemically damage the polymer network. A lining that was initially hard and well bonded can gradually lose some of those properties.[20][21]

What is beneath the coating can be just as important as the coating itself. If soluble salts remain on the steel after blasting and cleaning, water that penetrates the lining can be drawn toward those contaminated areas. Moisture accumulates at the interface and can produce blisters that lift the epoxy away from the steel.

A small break in the lining creates another path. A pinhole, thin spot, crack, damaged edge, or other discontinuity—often called a holiday—allows water to reach the steel directly. Corrosion begins at the exposed metal. The corrosion products and chemical changes created beneath the lining can weaken adhesion around the defect, allowing corrosion to spread laterally under coating that may still appear intact from the exposed side.

The lining can also be compromised before it ever enters service. Liquid epoxy depends on the resin and curing agent being mixed in the correct proportions and thoroughly combined. Too much of either component can leave material that never becomes part of the fully cured polymer network. Poor mixing can create soft or weak areas. Low temperatures can slow the curing reaction, while moisture during cure can create surface contamination such as amine blush that interferes with adhesion between coats. If too much time passes before a second coat is applied, the new layer may also have difficulty bonding to the cured surface below it.

Surface preparation remains critical throughout the process. Rust, mill scale, oil, dust, or inadequate abrasive blasting can leave areas where the epoxy never develops a strong bond to the steel. Once water reaches one of those weak interfaces, the lining can blister, peel, or detach even though the epoxy itself may still be chemically sound.

Higher temperature or a temperature difference across the lining can make an exposure more severe, depending on the failure mechanism. A tank may hold hot liquid while its outer wall is exposed to cooler air. A pipeline can experience similar differences between the process fluid and the surrounding soil or atmosphere. That temperature difference can affect how quickly water and chemicals move through the lining and where they accumulate. Repeated heating and cooling can also place stress on the bond because the steel and the epoxy do not expand and contract by exactly the same amount.

For that reason, evaluating a lining is not always as simple as placing a coated panel in a jar of liquid. Immersion and one-sided exposure tests can apply relevant stressors such as continuous liquid contact, vapor space, temperature differences, and other conditions that act on real pipe and tank linings. ASTM D6943 and AMPP TM0174 are examples of methods used to evaluate coating and lining systems under these types of exposures.[22]

An epoxy lining fails when the barrier is penetrated, the polymer is degraded, or the bond to the steel is lost. Often, those mechanisms reinforce one another: water reaches a weak area, corrosion begins beneath the lining, adhesion decreases, and the damaged area grows.

How the Powder-Coated Electrical Enclosure Fails

Polished coating cross section showing substantial film-thickness variation along the substrate — A coating cross section reveals substantial thickness variation over a short distance. Cross-sectional examination can expose uneven film build that is difficult to assess from the outer surface. **Source:** GPI Labs.

A powder-coated electrical enclosure looks simple from the outside, but its shape creates difficult areas to coat. Seams, folded edges, knockouts, corners, flanges, and recessed surfaces all have to receive enough powder, then heat enough for that powder to melt, flow together, and fully cure.

The first problem can occur during spraying. Charged powder is strongly attracted to easy-to-reach grounded surfaces, but the electric field can make it harder for particles to enter deep recesses and tight inside corners. Those areas may receive much less powder than the broad flat faces of the enclosure. Sharp outside edges create a different problem. As the powder melts in the oven, the liquid film tends to pull away from the edge, leaving less coating exactly where protection is often needed most.

That means an enclosure can have an acceptable average coating thickness while still containing thin areas at corners, seams, and edges. Those local weak spots can become the first places where water reaches the metal.

Curing creates another possible failure. The oven heats the enclosure, but the coating does not cure simply because the air reached the specified temperature. The metal itself has to reach the required temperature and remain there long enough for the resin to complete its curing reaction. A heavy section can heat more slowly than a thin panel, so different parts of the same assembly may not cure at the same rate.

If the powder is under-cured, the finished film can remain soft, weak, or more vulnerable to chemicals and moisture. Excessive heat or excessive time in the oven can discolor some coatings and can reduce flexibility or otherwise damage the polymer.

Damage can also occur after curing. Punching, bending, fastener installation, impact, abrasion, or repeated thermal expansion can crack or chip the finished film. Once bare metal is exposed, moisture and salts can reach the steel.

What happens next depends heavily on the surface preparation beneath the powder. If the steel was properly cleaned and pretreated, corrosion may remain localized for some time. If oil, rust, weak conversion coating, or other contamination was trapped beneath the film, water can spread along that weak interface. Corrosion then advances underneath a coating that may still look intact nearby.

That is why the appearance of a powder-coated enclosure can be misleading. A smooth, glossy surface does not prove that every recess was coated, every edge has enough thickness, every section fully cured, or the coating is firmly bonded to clean metal beneath it.

Powder coating fails when coverage, cure, adhesion, or the finished film is compromised. The visible rust may appear at an edge or seam, but the weakness may have begun much earlier in the spray booth, pretreatment line, or curing oven.

How the Multi-Layer Protective Coating System Fails

A multi-layer coating system gains protection by giving different jobs to different materials. It also creates multiple places where failure can begin: at the steel surface, within a coating layer, or at the bond between two coatings.

Five scribed coated-steel panels at increasing exposure times, with rust creepage and blistering spreading outward from the scribe — Progressive corrosion and coating deterioration spreading from a scribe during accelerated laboratory testing at 0, 1,080, 2,160, 3,240 and 4,320 hours. **Source:** Federal Highway Administration, Figure 27.

Failure at the steel surface usually points back to preparation or primer application. If rust, soluble salts, oil, dust, or weak mill scale remain beneath the primer, the first layer may never develop a strong bond. Moisture that eventually reaches that interface can then spread beneath the coating and lift the entire system away from the steel.

Failure can also occur between coats. The zinc-rich primer may remain firmly attached to the steel while the epoxy separates from it. The epoxy may remain sound while the topcoat peels away. Contamination, incorrect recoat timing, incompatible materials, or a surface that has weathered too long before the next coat is applied can all weaken the bond between layers.

The individual coatings can fail as well. A thick epoxy applied too heavily may trap solvent or air and develop pores or blisters. An under-cured film may remain soft or chemically weak. A brittle layer may crack when the structure flexes. Sunlight can degrade an exposed epoxy before the topcoat is applied. Repeated heating and cooling can also place stress on the system because the steel, primer, epoxy, and topcoat do not all expand and contract at exactly the same rate.

Once a crack, pinhole, damaged edge, or other holiday reaches the steel, the environment has a direct path through the system. On a bridge, water and chloride can enter at that point and corrosion can spread beneath surrounding coating. On a buried pipeline, the same defect can interact with the cathodic-protection system.

Cathodic protection supplies current to the exposed steel and suppresses corrosion at the holiday. But the cathodic reaction also changes the chemistry immediately around the defect and can create strongly alkaline conditions. Those conditions can weaken adhesion between the coating and the steel. The coating then begins to separate outward from the holiday even though the cathodic-protection system is doing its intended job of protecting the exposed metal. This mechanism is known as cathodic disbondment. ASTM G8, ASTM G42, ASTM D6676/D6676M, and AMPP TM0115 use different temperatures, electrical arrangements, and heat-flow conditions to evaluate this type of failure.[22]

Detached green coating chips with a magnified rear surface and a polished cross section — Detached coating chips, their rear surface under magnification, and a polished cross section. Examining both the separation surface and the coating layers helps locate where failure occurred. **Source:** GPI Labs.

Where the system separates often helps identify what went wrong. Bare steel beneath a detached coating suggests a problem at the steel-primer interface. A primer still firmly attached to the steel with the next coat missing points toward an intercoat adhesion problem. Clean separation between two coating layers can indicate contamination, recoat timing, or incompatibility. A tear through the body of the coating rather than along an interface suggests that the coating itself was weaker than the bond holding it in place.

A multi-layer system therefore does not fail simply because “the coating came off.” The location and shape of the failure reveal whether the weakness began in surface preparation, application, curing, compatibility between layers, environmental attack, or the coating material itself.

Corrosion Testing

From Failure Mechanism to Test Method

The failure mechanisms above do not happen because a coating reaches a certain age. They happen because the material is exposed to combinations of moisture, salt, temperature, chemicals, electrical potential, sunlight, abrasion, and repeated wetting and drying.

That distinction matters when identifying and applying the stressors associated with the observed failure.

A useful corrosion test does not simply make the environment more aggressive and wait for damage. It attempts to control the conditions that drive the failure mechanism. If chloride and repeated drying are important, the test needs salt exposure and dry-off. If condensation causes the problem, the specimen needs a controlled condensing environment. If a buried pipeline coating is failing around holidays under cathodic protection, salt fog alone will not recreate the same stressors or failure mechanism.

Laboratory corrosion testing therefore begins by asking what conditions caused the material or coating system to fail, then selecting or developing an exposure that applies the relevant conditions in a controlled and repeatable way.

Applying Relevant Stressors in a Controlled Environment

Front view of a cyclic corrosion chamber with solution reservoir, transparent specimen area and control section — Cyclic corrosion chamber used in FHWA bridge-steel research. **Source:** Federal Highway Administration, Figure 30.

Corrosion testing does not turn 1,000 chamber hours into a fixed number of years outdoors. It does something more useful: it places materials and coating systems in controlled environments designed to apply the stressors associated with specific corrosion mechanisms more quickly and repeatably than waiting for failures to develop naturally.

The goal is not simply to make a specimen rust faster. Good corrosion science starts by identifying the weakest points in a material, coating system, joint, edge, defect, interface, or production process and then applying the environmental stresses most likely to expose those weaknesses. Salt deposition, moisture, condensation, drying, temperature, chemical exposure, electrical potential, corrosive gases, and repeated transitions between them can each attack a different part of the system. By controlling those conditions and repeating them in a defined way, accelerated testing can reveal where failure begins, how it develops, and which materials, pretreatments, coating systems, application processes, or production lots resist those failure mechanisms most effectively.

ASTM B117 is most useful when continuous salt deposition is itself a meaningful part of the damage mechanism, or when a specification, qualification program, or production-control system needs a highly repeatable way to compare coatings, pretreatments, materials, or manufacturing lots. In those roles, salt fog can expose defects such as poor coating coverage, inadequate pretreatment, porosity, edge weakness, or loss of barrier protection. But if the product actually fails because of wet/dry cycling, condensation, temperature changes, freezing, galvanic effects, chemical exposure, corrosive gases, or some other service condition, continuously spraying salt may accelerate the wrong mechanism. In those cases, a better test is one that deliberately reproduces and stresses the conditions that drive the real failure rather than simply extending the number of salt-fog hours.[23]

Open cyclic corrosion chamber containing rows of coated test panels and adjustable spray heads — Coated panels and spray headers inside the same cyclic corrosion chamber. **Source:** Federal Highway Administration, Figure 31.

Cyclic methods incorporate more of the conditions that occur outside the chamber. ASTM G85 changes the salt solution or exposure sequence. SAE J2334, GMW 14872, Ford CETP 00.00-L-467, VDA 233-102, and ISO 11997-3 move specimens through combinations of salt application, humidity, drying, temperature change, and, in some methods, freezing or immersion.[24][46][47][48][49][50] These transitions matter because a vehicle, bridge component, fastener, or outdoor enclosure normally becomes wet, contaminated, dries, cools, heats, and becomes wet again rather than remaining continuously inside a salt fog.

Other products require completely different exposures. ASTM D870 keeps coated specimens in water.[25] ASTM D2247 creates continuous condensation.[26] ASTM G210 exposes wastewater coatings to an acidic saline liquid beneath a hot, humid headspace.[45] Cathodic-disbondment testing combines an electrolyte with an imposed electrical potential to study coating separation around a defect.[51] IEC 60068-2-60 exposes electronics to carefully controlled concentrations of corrosive gases.[13]

By controlling these conditions, accelerated testing can expose weaknesses that might otherwise take months or years to appear. Poor pretreatment may produce early underfilm corrosion. An inadequately sealed anodized surface may pit in chloride exposure. An epoxy lining may blister, soften, or lose adhesion. A powder coating may fail first at thin edges and recesses. A multi-layer pipeline coating may disbond around a holiday under cathodic protection.

Producing the failure is only part of the work. The specimen then has to be examined to determine what actually failed. ASTM D610 can be used to rate rusting. ASTM D714 evaluates blistering. ASTM D1654 measures corrosion and coating deterioration around a scribe. ASTM D3359 evaluates adhesion by tape testing, while ASTM D4541 measures pull-off strength and helps identify whether failure occurred at the substrate, between coating layers, or within the coating itself.[36][37][38][39][40]

The Auto Technology corrosion laboratory can perform these exposures and evaluations when a company needs independent testing, qualification work, comparative material studies, failure investigation, or development work without operating the equipment itself. A customer can expose multiple coating systems, pretreatments, materials, or process variations under the same controlled conditions and compare how quickly failure begins, how it develops, and where the system ultimately breaks down.

That information can then be used to change the product before the same mechanism appears in service: improve cleaning or pretreatment, increase coating thickness at vulnerable areas, change a resin or curing system, improve sealing, modify application or cure conditions, redesign an edge or drainage path, or select a coating system better suited to the actual environment.

The value of that work is not simply that corrosion occurred faster. It is that the specimens, exposure, defects, environmental conditions, and measurements were controlled well enough to learn something from the failure.

Coated test panels partially immersed in water-filled beakers — Coated panels during water immersion testing. **Source:** GPI Labs.

Scribed coated panels showing corrosion and coating deterioration after cyclic weathering exposure — Scribed panels after cyclic weathering exposure. **Source:** GPI Labs.

Close-up of corrosion products and coating deterioration along an intentional scribe — Close-up of corrosion and coating deterioration around an intentional scribe. **Source:** GPI Labs.

Test the Environment the Product Will Actually Face

The correct corrosion test begins with the product and its service environment, not with a chamber.

A galvanized guardrail exposed to road salt and repeated wetting and drying does not experience the same conditions as an epoxy lining holding warm wastewater. A buried pipeline coating under cathodic protection does not fail by the same mechanism as a circuit board exposed to sulfur- and chlorine-containing gases. An anodized architectural frame near the coast does not see the same exposure as a continuously immersed water-pipe lining.

The test environment should therefore be selected to apply the stressors that drive the mechanisms that matter to that product. That may mean continuous salt fog, cyclic salt and dry-off, controlled humidity, condensation, immersion, corrosive gases, acidic solutions, freezing, electrical potential, or a combination of several conditions.

Auto Technology Company builds different corrosion chambers because different products fail by different mechanisms.

M Series and C Series chambers are used for continuous salt-spray exposures such as ASTM B117 and ISO 9227. They can also support high-humidity exposures within the limits of each chamber configuration. These exposures are useful for accelerating coating breakdown, corrosion at scratches and edges, galvanic attack, underfilm corrosion, and loss of protection after the coating or metallic layer has been breached.[23][33][34][54]

Auto Technology M Series salt-fog and humidity corrosion test chamber with transparent chamber cover and separate controls — Auto Technology M Series chamber for continuous salt-fog exposures and supported humidity conditions. **Source:** Auto Technology Company — M Series chamber.

Introduction to Auto Technology chambers used for ASTM B117 and ISO 9227 salt-fog testing. Source: Auto Technology Company — ASTM B117 and ISO 9227 chamber video.

A Series chambers add cyclic operation for programs that alternate between salt exposure, humidity, and other environmental stages. Those transitions allow the test to apply repeated wetting, salt deposition, moisture retention, and drying rather than holding the specimen continuously wet.[33][34]

X Series chambers are designed for more complex cyclic corrosion programs such as SAE J2334, GMW 14872, Ford CETP 00.00-L-467, VDA 233-102, and ISO 11997-3. Depending on the method and chamber configuration, they can control salt application, humidity, dry-off, temperature changes, direct solution spray, immersion, and low-temperature stages. These cycles are used to impose the conditions that drive corrosion at coating defects, underfilm creep, galvanic corrosion, repeated wet-dry attack, chloride concentration during drying, and damage associated with temperature and moisture cycling.[35]

Auto Technology X Series precision cyclic corrosion chamber with a large transparent cover and integrated control cabinet — Auto Technology X Series precision cyclic corrosion chamber. **Source:** Auto Technology Company — X Series chamber.

Direct-spray hardware operating inside an Auto Technology cyclic corrosion chamber. Source: Auto Technology Company — direct-spray video.

Automated fill and drain sequence for an X Series immersion cycle. Source: Auto Technology Company — immersion-cycle video.

Mixed Flowing Gas testing addresses a different corrosion problem than salt fog. Electrical and electronic equipment often operates indoors where there is no continuous salt spray, yet very low concentrations of sulfur-bearing gases, chlorine compounds, nitrogen oxides, and other airborne contaminants can react with copper, silver, contact finishes, solderable surfaces, connectors, and exposed conductors. In these applications, the important question is not whether a component develops obvious red rust, but whether thin corrosion films, tarnish, pore corrosion, corrosion-product migration, or contamination increase contact resistance, create intermittent connections, degrade signal integrity, or eventually cause loss of function.

Methods such as IEC 60068-2-60 and ASTM B827 use controlled gas concentrations together with defined temperature, humidity, airflow, and exposure time to reproduce or amplify these types of indoor atmospheric corrosion conditions. That matters because electronics can experience very different corrosive environments depending on where they are installed. A connector inside a sealed cabinet may see a relatively benign atmosphere, while equipment exposed to industrial process gases, wastewater treatment environments, pulp and paper operations, chemical facilities, polluted urban air, data centers with contaminant ingress, or poorly filtered ventilation can experience substantially greater atmospheric corrosivity.

Environmental classification systems such as the Battelle and related gaseous-corrosivity classifications are useful because they describe these environments in terms of how aggressively the atmosphere attacks sensitive metals. Copper and silver coupons, for example, can be used as indicators of environmental severity even when the surrounding equipment shows little visible corrosion. Mixed Flowing Gas chambers allow those severity levels to be recreated or accelerated under controlled laboratory conditions so engineers can compare contact finishes, connector designs, protective coatings, conformal coatings, housings, seals, materials, and complete assemblies against the type of atmospheric exposure they are actually expected to encounter.

The chamber exposure therefore has to be selected for the failure mechanism being investigated. Salt fog is appropriate when deposited chloride and sustained wetting are important. Mixed Flowing Gas is more appropriate when low-concentration atmospheric pollutants, humidity, and sensitive electrical surfaces drive the failure. The purpose is not simply to make electronics corrode faster, but to reproduce the chemical environment that attacks the weakest functional parts of the system and determine whether corrosion progresses far enough to affect electrical performance.

Mixed flowing gas testing for corrosion-sensitive electrical and electronic materials. Source: Auto Technology Company — mixed flowing gas video.

Other corrosion mechanisms require dedicated or custom equipment. Immersion and controlled-humidity systems can evaluate water absorption, blistering, adhesion loss, and chemical attack in coatings and linings. Cathodic-disbondment equipment combines an electrolyte, an intentional coating defect, temperature, and electrical potential to study coating separation around holidays in buried pipeline systems. Custom chambers can apply controlled corrosive environments around specimens or assemblies too large or unusual for standard equipment.

The chamber therefore follows the failure mechanism and the test method. Salt fog, condensation, wet-dry cycling, immersion, corrosive gas exposure, cathodic protection, freezing, and temperature transitions do not produce the same corrosion process, and they should not be treated as interchangeable ways of making a specimen fail faster.

The Auto Technology corrosion laboratory can run established corrosion specifications, compare materials and coatings, investigate failures, and help determine which environmental conditions are actually driving damage. This allows manufacturers to test a product or process before purchasing equipment or to use the lab for specialized exposures they do not routinely perform in-house.

When a customer specification requires ASTM B117, SAE J2334, IEC 60068-2-60, or another defined method, the laboratory reproduces that required exposure. When the problem is a field failure or an unknown corrosion mechanism, the test can instead be built around the suspected cause. Salt concentration, humidity, temperature, wet and dry periods, immersion, corrosive gases, specimen orientation, or other variables can be changed deliberately to determine what produces the failure and what prevents it.

Once the important exposure is understood and the test method is established, Auto Technology can build equipment to apply those environmental conditions repeatedly in the manufacturer's own facility.

The chamber design then follows the test. Specimen size and loading determine working volume. The method determines spray delivery, solution handling, temperature, humidity, drying, cooling, immersion, drainage, transitions, controls, and data collection. Large or unusual products may require custom fixturing or custom chamber geometry.

The purpose of accelerated corrosion testing is not to guarantee that a product will never fail. It is to expose weaknesses sooner, understand why they occur, compare possible corrections under controlled conditions, and reduce the chance of discovering the same problem years later in the field.

ATC can provide either side of that process: the corrosion laboratory to develop, apply, and evaluate the exposure, and the corrosion chamber to bring the proven test capability into production, quality control, or an internal materials laboratory.

Discuss a corrosion test chamber with Auto Technology: https://autotechnology.com/contact-us/

References

World Steel Association, "What Is Steel?" https://worldsteel.org/about-steel/what-is-steel/
U.S. Department of Energy, "Iron and Steel Manufacturing." https://www.energy.gov/cmei/ito/iron-and-steel-manufacturing
American Galvanizers Association, "Batch Hot-Dip Galvanizing." https://galvanizeit.org/corrosion/corrosion-protection/zinc-coatings/batch-hot-dip-galvanizing
ASTM International, ASTM A123/A123M-24, Standard Specification for Zinc (Hot-Dip Galvanized) Coatings on Iron and Steel Products. https://store.astm.org/a0123_a0123m-24.html
International Aluminium Institute, "Primary Production." https://alustory.international-aluminium.org/primary-production/
U.S. Geological Survey, Copper - A Metal for the Ages. https://pubs.usgs.gov/fs/2009/3031/
American Cement Association, "How Cement Is Made." https://www.cement.org/cement-concrete/how-cement-is-made/
Association for Materials Protection and Performance, "What Is Corrosion?" https://www.ampp.org/technical-research/what-is-corrosion
Dwivedi, D., Lepkova, K., and Becker, T., "Carbon Steel Corrosion: A Review of Key Surface Properties and Characterization Methods," RSC Advances 7 (2017): 4580-4610. https://doi.org/10.1039/C6RA25094G
American Galvanizers Association, "How Does Galvanizing Protect Steel from Corrosion?" https://galvanizeit.org/inspection-course/course-navigation/frequently-asked-questions
Parangusan, H., et al., "A Review of Passivity Breakdown on Metal Surfaces: Influence of Chloride- and Sulfide-Ion Concentrations, Temperature, and pH," Emergent Materials 4 (2021): 1187-1203. https://doi.org/10.1007/s42247-021-00194-6
Singley, J. E., Beaudet, B. A., and Markey, P. H., Corrosion Manual for Internal Corrosion of Water Distribution Systems, EPA 570/9-84-001 (1984). https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=10003FIW.TXT
International Electrotechnical Commission, IEC 60068-2-60:2015, Environmental Testing - Part 2-60: Tests - Test Ke: Flowing Mixed Gas Corrosion Test. https://webstore.iec.ch/en/publication/22710
Ali, M., et al., "A Review on Chloride-Induced Corrosion in Reinforced Concrete," Heliyon 10 (2024). https://pmc.ncbi.nlm.nih.gov/articles/PMC11580005/
NASA Kennedy Space Center Corrosion Technology Laboratory, "Coatings Data" and NASA-STD-5008. https://public.ksc.nasa.gov/corrosion/coatings-data/
Mills, D. J., and Jamali, S. S., "Advances in Corrosion Protection by Organic Coatings," Progress in Organic Coatings 95 (2016): 26-37. https://nectar.northampton.ac.uk/id/eprint/9268/
GPI Labs, "Protective Coating & Lining Testing for Water and Wastewater Infrastructure." https://gpilabs.com/services/water-wastewater-infrastructure-testing/
ASTM International, ASTM D7803-25, Standard Practice for Preparation of Zinc (Hot-Dip Galvanized) Coated Iron and Steel Product and Hardware Surfaces for Powder Coating. https://store.astm.org/d7803-25.html
International Organization for Standardization, ISO 12944-6:2018, Paints and Varnishes - Corrosion Protection of Steel Structures by Protective Paint Systems - Part 6: Laboratory Performance Test Methods. https://www.iso.org/standard/51378.html
Sabet-Bokati, K., and Plucknett, K., "Water-Induced Failure in Polymer Coatings: Mechanisms, Impacts and Mitigation Strategies - A Comprehensive Review," Polymer Degradation and Stability 230 (2024): 111058. https://doi.org/10.1016/j.polymdegradstab.2024.111058
Zargarnezhad, H., Asselin, E., Wong, D., and Lam, C. N. C., "A Critical Review of the Time-Dependent Performance of Polymeric Pipeline Coatings: Focus on Hydration of Epoxy-Based Coatings," Polymers 13, no. 9 (2021): 1517. https://doi.org/10.3390/polym13091517
GPI Labs, "Pipeline Materials, Coating & Corrosion Testing Services." https://gpilabs.com/services/pipeline-materials-coating-corrosion-testing/
ASTM International, ASTM B117-26, Standard Practice for Operating Salt Spray (Fog) Apparatus. https://store.astm.org/b0117-26.html
ASTM International, ASTM G85-19, Standard Practice for Modified Salt Spray (Fog) Testing. https://store.astm.org/g0085-19.html
ASTM International, ASTM D870-25, Standard Practice for Testing Water Resistance of Coatings Using Water Immersion. https://store.astm.org/d0870-25.html
ASTM International, ASTM D2247-25, Standard Practice for Testing Water Resistance of Coatings in 100% Relative Humidity. https://store.astm.org/d2247-25.html
Lutze, F. W., Smith, K. A., Mason, R., et al., "Update on the Developments of the SAE J2334 Laboratory Cyclic Corrosion Test," SAE Technical Paper 2003-01-1234 (2003). https://doi.org/10.4271/2003-01-1234
Courval, G., Smith, K. A., Meade, C., et al., "Development of an Improved Cosmetic Corrosion Test by the Automotive and Aluminum Industries for Finished Aluminum Autobody Panels," SAE Technical Paper 2003-01-1235 (2003). https://doi.org/10.4271/2003-01-1235
Bovard, F., Tardiff, J., Jafolla, T., et al., "Development of an Improved Cosmetic Corrosion Test for Finished Aluminum Autobody Panels," SAE Technical Paper 2007-01-0417 (2007). https://doi.org/10.4271/2007-01-0417
Blankenship, K., and Manuel, J., "Two-Coat Inorganic Coating System for Steel Bridges," Transportation Research Record 2679, no. 5 (2025): 914-923. Data collection credited to GPI Labs. https://doi.org/10.1177/03611981241292586
Arukula, R., et al., "Corrosion Resistant Coating Fabrication Through Synergies Between SiOC Conversion and Iron Oxidation at High Temperatures," npj Materials Degradation 9, article 65 (2025). Steel panels supplied by GPI Laboratories, Inc. https://doi.org/10.1038/s41529-025-00584-9
O'Donoghue, M., et al., "The Performance Effect of Coating Rusty Steel With an Epoxy-Phenalkamine Coating," CORROSION 2018, Paper No. NACE-2018-11067. Author affiliation includes GPI Laboratories, Inc. https://onepetro.org/NACECORR/proceedings/CORR18/All-CORR18/NACE-2018-11067/126058
Auto Technology Company, "Industrial Corrosion Test Chambers & Systems." https://autotechnology.com/corrosion-chambers/
Auto Technology Company, "Corrosion Test Chamber Selection Guide." https://autotechnology.com/corrosion-test-chamber-selection-guide/
Auto Technology Company, "X Series Cyclic Corrosion Test Chambers." https://autotechnology.com/cyclic-corrosion-chambers/
ASTM International, ASTM D610-25, Standard Practice for Evaluating Degree of Rusting on Painted Steel Surfaces. https://store.astm.org/d0610-25.html
ASTM International, ASTM D714-25, Standard Test Method for Evaluating Degree of Blistering of Paints. https://store.astm.org/d0714-25.html
ASTM International, ASTM D1654-24e1, Standard Test Method for Evaluation of Painted or Coated Specimens Subjected to Corrosive Environments. https://store.astm.org/d1654-24e01.html
ASTM International, ASTM D3359-23, Standard Test Methods for Rating Adhesion by Tape Test. https://store.astm.org/d3359-23.html
ASTM International, ASTM D4541-22, Standard Test Method for Pull-Off Strength of Coatings Using Portable Adhesion Testers. https://store.astm.org/d4541-22.html
U.S. Department of Energy, Lead and Zinc, Energy and Environmental Profile of the U.S. Mining Industry. https://www.energy.gov/sites/prod/files/2013/11/f4/lead_zinc.pdf
Galster, S., and Helmreich, B., "Copper and Zinc as Roofing Materials - A Review on the Occurrence and Mitigation Measures of Runoff Pollution," Water 14, no. 3 (2022): 291. https://doi.org/10.3390/w14030291
Metal Construction Association, "Dissimilar Metal Corrosion in Roofing Installations." https://www.metalconstruction.org/view/download.php/online-education/education-materials/roofing-educational-materials/dissimilar-metal-corrosion-in-roofing-installations
American Water Works Association, AWWA C210-24, Liquid-Epoxy Coatings and Linings for Steel Water Pipe and Fittings. https://store.awwa.org/AWWA-C210-24-Liquid-Epoxy-Coatings-and-Linings-for-Steel-Water-Pipe-and-Fittings
ASTM International, ASTM G210-13(2023), Standard Practice for Operating the Severe Wastewater Analysis Testing Apparatus. https://store.astm.org/g0210-13r23.html
SAE International, SAE J2334_201604, Laboratory Cyclic Corrosion Test. https://saemobilus.sae.org/standards/j2334_201604-laboratory-cyclic-corrosion-test
General Motors, GMW14872, Cyclic Corrosion Laboratory Test. https://standards.globalspec.com/std/14576230/gmw14872
Auto Technology Company, "Ford L-467 Cyclic Corrosion Test Chamber." https://autotechnology.com/product/ford-l-467-cyclic-corrosion-test-chamber/
Verband der Automobilindustrie, VDA 233-102:2013, Cyclic Corrosion Testing of Materials and Components in Automotive Construction. https://webshop.vda.de/VDA/en/vda-233-102-06-2013
International Organization for Standardization, ISO 11997-3:2022, Paints and Varnishes - Determination of Resistance to Cyclic Corrosion Conditions - Part 3: Testing of Coating Systems on Materials and Components in Automotive Construction. https://www.iso.org/standard/80654.html
ASTM International, ASTM G8-24, Standard Test Methods for Cathodic Disbonding of Coated Steel. https://store.astm.org/g0008-24.html
Federal Highway Administration, "Appendix A: Distress Mechanisms," Guidelines for Detection, Analysis, and Treatment of Materials-Related Distress in Concrete Pavements, FHWA-RD-01-163. https://www.fhwa.dot.gov/publications/research/infrastructure/pavements/pccp/01163/app.cfm
Pommersheim, J. M., and Clifton, J. R., Models of Transport Processes in Concrete, NISTIR 4405 (1990). https://nvlpubs.nist.gov/nistpubs/Legacy/IR/nistir4405.pdf
International Organization for Standardization, ISO 9227:2022, Corrosion Tests in Artificial Atmospheres - Salt Spray Tests. https://www.iso.org/standard/81744.html

Discuss a Corrosion Test Chamber