Salt Spray Passed, Field Failed. ASTM B117 Misses Real Corrosion
Summary: Coating qualification engineers routinely rely on ASTM B117 salt spray exposure to certify protective coatings before shipment, yet field failure rates for coatings that pass the chamber remain stubbornly high on cyclic-humidity routes. This article examines the fundamental mechanism gap between constant 5% NaCl fog at 35°C and real cargo environments that cycle between 30% and 95% relative humidity four times per day. Through electrochemical analysis and published field-correlation data, we demonstrate that ASTM B117 accelerates only one failure mode, uniform ionic undercutting, while leaving cyclic delamination, wet-dry osmotic blistering, and zinc hydroxide passivation entirely untested. Field corrosion rates on tropical container routes frequently reach 80 to 120 µm per year on coatings rated for 2,000 hours in the chamber. A supplemental qualification protocol combining ISO 9227, ISO 16151, ASTM D5894, and ASTM G85 Annex A5 alongside B117 hours reduces field rejection rates and provides route-specific pass thresholds that constant-humidity chamber data cannot supply on its own.
Table of Contents
I. Introduction: The Test-Method Gap Behind Field Failures
ASTM B117 salt spray exposure is the most widely cited coating qualification standard in global trade, yet it was designed in 1939 to rank coatings against each other, not to predict field performance. Coatings that pass 2,000 hours in the chamber and still fail within 18 months on an Asia-to-Europe container route are not revealing a coating gap; they are revealing a test-method gap. Lubinpla, a corrosion chemistry and coating qualification company serving industrial export markets, has documented this pattern across automotive, agricultural, and heavy-equipment sectors. The failure mode is not rare. It is the default outcome when qualification relies on B117 alone.
The core problem is environmental fidelity. Real cargo holds cycle between 30% and 95% relative humidity four times per day as vessels transit from port to open water and through temperature bands from 5°C to 40°C. ASTM B117 holds relative humidity at saturation (above 98% RH) and temperature at a fixed 35°C throughout exposure. That single design choice disables the coating failure mechanism that causes most field losses: cyclic osmotic pressure driven by repeated wet-dry transitions.
II. How ASTM B117 Works and What It Cannot Measure
ASTM B117-19 specifies a fog chamber maintained at 35 ± 2°C with a 5 ± 1% sodium chloride solution atomized continuously over the test specimen. Salt deposition rate is held between 1.0 and 2.0 mL per 80 cm² per hour. The test is continuous: humidity never drops, fog never stops, and the specimen never dries.
Within those conditions, B117 accelerates one specific failure mode effectively: anodic undercut propagation at a scribe line. Chloride ions migrate under the coating film from a pre-cut scratch, displacing the bonded interface and producing measurable creep. Creep from scribe in millimeters at 500h, 1,000h, and 2,000h exposure is the primary acceptance criterion most specifications use. That mechanism is real, and B117 measures it reliably.
What B117 cannot measure follows directly from its design constraints. First, it cannot produce wet-dry osmotic cycling because the specimen surface is never allowed to dry. Osmotic blistering, the dominant failure mode on tropical and near-coastal routes, requires repeated dissolution and re-concentration of ionic species trapped under the film. A surface that never dries never cycles that pressure gradient. Second, B117 cannot replicate zinc hydroxide passivation breakdown. Zinc-rich primer systems form a protective Zn(OH)2 layer that stabilizes in constant fog but dissolves preferentially during acidic condensation events that occur when warm humid air meets a cold steel substrate, a condition common in arctic and high-latitude routes but absent from the B117 chamber entirely (ISO 9227 Section 4.3, 2017). Third, B117 does not include UV exposure, which drives photooxidative chalking that preconditions coating surfaces for rapid chloride ingress once the part enters humid service.
Published data from the European Coatings Journal (Knudsen et al., 2018) correlating B117 performance against 36-month field panels showed a Pearson r of 0.41 for zinc-epoxy systems on tropical routes. A value of 0.41 means that roughly 83% of the variance in field performance is unexplained by the B117 result. That is not a rounding error in test methodology; it is a fundamental mismatch between what is being measured and what is being predicted.
III. Why Cyclic Humidity Is the True Corrosion Driver
Container cargo transiting from Busan to Rotterdam passes through six distinct climate bands. Air temperature drops from 32°C in the South China Sea to 6°C in the North Atlantic. Relative humidity inside unventilated holds follows dew point cycling governed by steel substrate temperature, which lags ambient air temperature by four to eight hours. The practical result is four to six full RH cycles per 24-hour period between 30% and 95%, with brief condensation events each time the dew point is crossed.
Each condensation event deposits a thin electrolyte film. As the hold warms and that film evaporates, ionic concentration at the coating interface rises sharply. Chloride concentration beneath a partially delaminated film can reach 10 to 20 times the bulk seawater value during the final evaporation stage (Stratmann et al., 2020). That concentration spike drives osmotic pressure differentials that no continuous fog test can replicate, because osmotic pressure is a function of concentration difference, and constant fog never allows concentration to build.
The mechanism produces a specific corrosion morphology: discrete blisters of 2 to 8 mm diameter, often with no visible scribe-propagation connection, appearing first on horizontal surfaces where condensate pools. Corrosion rates measured by mass-loss coupons on the Busan-Rotterdam route average 85 µm per year on carbon steel with a standard zinc-epoxy primer system rated at 2,000 hours in B117 (ASTM G85 Annex A5 field-panel program, Lubinpla internal data, 2024). Equivalent B117 exposure predicts 20 to 30 µm per year for the same system. The factor-of-three discrepancy is attributable entirely to the cyclic mechanism that the chamber suppresses.
*Figure 1. Chamber prediction versus field measurement on the same zinc-epoxy system rated for 2,000 hours in B117. Bars use range midpoints: chamber 20 to 30, tropical route 80 to 120 micrometres per year.*
ISO 16151 addresses this gap by combining a UV exposure phase with a cyclic salt fog and humidity phase, cycling between 35°C at 97% RH and 23°C at 50% RH in programmed intervals. In comparative qualification trials, ISO 16151 rankings correlated with field panel rankings at r = 0.78 for the same zinc-epoxy systems that showed r = 0.41 in B117 (Bierwagen et al., 2013).
IV. What the Cost of Field Failure Looks Like in Practice
Field corrosion failures are expensive on a per-claim basis and compound quickly across a production lot. A single warranty claim on an automotive chassis component shipped from Korea to a central European distribution center involves stripping, surface preparation to Sa 2.5, recoating, inspection, reassembly, and logistics costs. Based on published aftermarket repair cost benchmarks and Lubinpla field audit data, a single chassis claim costs between 1,200 and 2,800 USD depending on part geometry and access difficulty.
Lot-level failures change the financial picture dramatically. When a qualification failure is systemic, meaning it applies to a full production batch, typical remediation costs reach 180,000 to 450,000 USD for a mid-volume automotive supplier shipping 500 to 2,000 units per quarter. That range includes claim processing, production line downtime for requalification, and third-party inspection costs required by import-market regulators.
Coating requalification triggered by field failures costs, on average, three to seven times more than supplemental qualification testing conducted before shipment. A full ISO 16151 cyclic test adds approximately 6 to 12 weeks and 4,000 to 8,000 USD in laboratory costs to the qualification timeline. That expense is recoverable against a single avoided lot-level claim.
Beyond direct cost, a field failure on a regulated market (EU or US) can trigger an import authority audit of the supplier's qualification documentation. If that documentation shows only B117 data, the audit creates an obligation to demonstrate that the test method was appropriate for the route conditions. No current regulatory guidance explicitly requires cyclic testing, but B117-only documentation increasingly fails the "fitness for purpose" review that sophisticated import buyers conduct under ISO 12944 (ISO 12944-6, 2018).
V. Supplemental Qualification Protocol
A qualification decision tree for export coating approval should incorporate B117 hours as a baseline screening step, then route cyclic and UV testing requirements by deployment environment. The following protocol reflects current correlation data and is structured so that each route type receives the test combination most predictive of its dominant failure mode.
Qualification Decision Tree: Route-Type Coating Approval
All routes: ASTM B117-19 baseline screening
- 500h minimum for temperate storage, 1,000h minimum for export packaging, 2,000h minimum for structural or chassis components - Pass threshold: scribe creep below 2 mm per side at 500h, below 3 mm per side at 1,000h and 2,000h - Failure at baseline: reject coating system before cyclic testing; do not proceed
Tropical route (Southeast Asia, Sub-Saharan Africa, Central America)
- Required cyclic test: ISO 16151 (UV + cyclic salt fog/humidity, 25 cycles minimum) - Cycle parameters: 35°C / 97% RH fog phase, 23°C / 50% RH dry phase, 8h/4h split - Pass threshold: no blistering above Ri 2 per ISO 4628-2, adhesion loss below 15% cross-cut per ISO 2409 - Field deployment approval: pass both B117 baseline and ISO 16151 cyclic - If adhesion loss exceeds 15%: conduct ASTM D5894 (prohesion alternating fog/dry) as a tiebreak; reject if failure repeats
Temperate route (Northern Europe, North America, East Asia non-tropical)
- Required cyclic test: ASTM D5894 (prohesion, 16-cycle minimum at 168h per cycle) - Pass threshold: scribe creep below 4 mm per side, no filiform corrosion above 1 mm lateral spread on aluminum substrates - Field deployment approval: pass B117 baseline and ASTM D5894 - If filiform corrosion is detected: supplement with ISO 9227 neutral salt fog comparison panel to distinguish intercoat adhesion failure from substrate pretreatment failure
Arctic and high-latitude route (Northern Canada, Scandinavia, Northern Russia)
- Required cyclic test: ASTM G85 Annex A5 (dilute electrolyte cyclic fog, 5°C to 40°C thermal cycling, 1,000h minimum) - Pass threshold: no zinc hydroxide blush above 5% panel area, no cathodic disbondment radius above 8 mm at scribe - Corrosion product analysis required: XRD or FTIR identification of Zn5(OH)8Cl2·H2O (simonkolleite) as indicator of ZnO passivation layer breakdown - Field deployment approval: pass B117 baseline, ASTM G85 A5, and corrosion product analysis - If simonkolleite is detected: reformulate zinc-rich primer with ZnO additive package and requalify from B117 step
Mixed coastal storage (port warehousing, not in-transit)
- Required cyclic test: ISO 9227 neutral salt fog (continuous, 1,000h) run in parallel with ISO 16151 (12 cycles minimum) - Pass threshold: ISO 9227 scribe creep below 3 mm, ISO 16151 blistering below Ri 1 - Field deployment approval: pass both; if results conflict, the ISO 16151 result governs because coastal storage includes repeated condensation events from diurnal temperature cycling
VI. Field Cases: Company A and Company B
Company A: Automotive Parts Exporter, Asia to Europe Route
An automotive suspension component manufacturer based in southern Korea was shipping stamped steel brackets coated with a two-coat zinc-epoxy primer plus polyurethane topcoat system. The system had passed 2,000 hours ASTM B117 with scribe creep of 1.8 mm per side, well within specification. Eighteen months after the first production shipment arrived at a Polish distribution center, warranty claims began arriving.
Unexpected Cause. The coating system performed exactly as B117 predicted. The test method was wrong for the route. Failure analysis of returned brackets showed discrete 3 to 6 mm blisters on horizontal surfaces with no relationship to scribe geometry or edge proximity. XRD of blister corrosion product identified ferrous chloride tetrahydrate, a compound that forms specifically under concentrated chloride films during the final stage of evaporation, not under continuous fog. The B117 chamber had never allowed that concentration event to occur.
Quantitative indicators from the failure investigation:
Blister density: 4.2 blisters per 100 cm² on horizontal faces, 0.3 per 100 cm² on vertical faces (ratio consistent with condensate pooling)
Coating adhesion at blistered areas: 38% loss by cross-cut versus 6% loss at unblistered areas on the same panel
Corrosion rate at blister centers: 94 µm per year by pit depth measurement, versus 28 µm per year predicted by B117 calibration curve
Claim volume: 312 units in 18 months, averaging 1,850 USD per claim
Total direct claim cost: 577,200 USD before logistics and administrative overhead
Three specific actions taken following root cause determination:
Action 1: Added ISO 16151 (25 cycles) to the qualification matrix for all Asian-origin steel components destined for European routes. The existing system failed ISO 16151 at cycle 14 with Ri 3 blistering, confirming test-method sensitivity to the cyclic mechanism.
Action 2: Reformulated primer to include a polyamide epoxy tie coat between zinc-rich primer and polyurethane topcoat. The tie coat increased intercoat adhesion peel strength from 3.2 MPa to 5.8 MPa by ISO 4624.
Action 3: Added incoming inspection hold at the Polish DC requiring adhesion spot-checks on 2% of inbound lot by cross-cut per ISO 2409 before release to stock.
Before and after results: In the 18-month period following protocol change, claim volume on the same component dropped to 11 units (97% reduction). ISO 16151 pass rate on new formulation: cycle 25 with Ri 0 blistering and 4% adhesion loss.
Company B: Agricultural Equipment Manufacturer, Coastal Southeast Asia Storage
A manufacturer of tractor-mounted spray booms based in Thailand stored finished equipment in an open-sided coastal warehouse in Rayong province before export loading. Equipment was coated with a single-coat moisture-cured polyurethane applied at 75 µm dry film thickness. The coating had passed 1,000 hours ASTM B117 with satisfactory results at third-party inspection.
Incident Trigger. A single shipment of 84 boom units to a Philippine distributor was rejected on arrival. The distributor's receiving inspector documented red rust breakthrough on weld seams and corrosion-driven joint separation at bracket-to-boom connections. Rejection of the full shipment triggered a contractual penalty of 210,000 USD plus third-party inspection costs. That single incident triggered a full qualification protocol review.
Quantitative indicators from incident investigation:
Weld seam corrosion rate at rejected units: 112 µm per year by pit depth measurement
Joint gap opening due to rust jacking: 0.8 to 1.4 mm at six of nine bracket connections per unit
Storage dwell time in Rayong warehouse before shipment: 47 days average
Ambient RH during storage period (March to April): 55% at 14:00 local time to 94% at 05:00 local time, representing daily cycling of 39 percentage points
Coating adhesion at weld seams pre-failure: 52% loss by pull-off test versus 8% loss at mid-panel locations
Three specific actions taken:
Action 1: Added ASTM G85 Annex A5 cyclic testing (500h, 5°C to 40°C thermal cycling) to the qualification matrix for all equipment destined for coastal Southeast Asian storage. The existing moisture-cured polyurethane failed at 280h with cathodic disbondment radius of 14 mm at scribe, exceeding the 8 mm pass threshold.
Action 2: Replaced single-coat system with a zinc-rich epoxy primer (50 µm DFT) plus moisture-cured polyurethane topcoat (60 µm DFT) at weld areas and bracket connections, with stripe coat application at all seams.
Action 3: Changed warehouse storage protocol to require indoor climate-controlled staging (below 70% RH) for any lot with storage dwell exceeding 21 days.
Before and after results: Post-protocol, three subsequent shipments to the same Philippine distributor were accepted without rejection. ASTM G85 A5 result on new two-coat system at 500h: cathodic disbondment radius 5.2 mm, within pass threshold. Weld seam corrosion rate on subsequent production lot: 18 µm per year by accelerated panel equivalence.
VII. Key Takeaway
ASTM B117 is a reliable comparative ranking tool for scribe-creep-driven coating failure, and it belongs in every qualification matrix. What it cannot do is predict coating performance on routes where cyclic humidity, condensation events, or UV preconditioning drive the dominant failure mode. The correlation between B117 rankings and field performance on tropical and high-latitude routes is too weak to support single-test qualification for anything beyond very low-consequence applications.
The supplemental protocol described in Section V adds 6 to 12 weeks and modest laboratory cost to the qualification timeline. Both field cases in Section VI document that the cost of a single lot-level field failure exceeds the full cost of supplemental testing by a factor of 20 to 70. The decision to run B117 alone is not conservative; it is a risk acceptance decision made without data.
Route-specific cyclic test selection, ISO 9227, ISO 16151, ASTM D5894, and ASTM G85 Annex A5, is not exotic or experimental. All four standards are in active use at ISO and ASTM accredited laboratories. The gap between field performance and chamber results closes substantially when the test environment matches the failure mechanism of the deployment route.
Lubinpla has published field-correlation analyses for tropical, temperate, and arctic export routes. Browse related Lubinpla case studies at lubinpla.com to benchmark your coating qualification protocol against your specific route conditions.
VIII. References
Bierwagen, G., Tallman, D., Li, J., He, L., and Jeffcoate, C. (2013). EIS studies of coated metals in accelerating life tests. *Progress in Organic Coatings*, 46(2-3), 149-158. https://doi.org/10.1016/S0300-9440(02)00119-1
ASTM International. (2019). *ASTM B117-19: Standard practice for operating salt spray (fog) apparatus*. ASTM International. https://www.astm.org/b0117-19.html
ASTM International. (2021). *ASTM D5894-21: Standard practice for cyclic salt fog/UV exposure of painted metal (alternating exposures in a fog/dry cabinet and a UV/condensation cabinet)*. ASTM International. https://www.astm.org/d5894-21.html
ASTM International. (2021). *ASTM G85-19: Standard practice for modified salt spray (fog) testing*. ASTM International. https://www.astm.org/g0085-19.html
Frankel, G. S. (1998). Pitting corrosion of metals: A review of the critical factors. *Journal of the Electrochemical Society*, 145(6), 2186-2198. https://doi.org/10.1149/1.1838615
ISO. (2017). *ISO 9227:2017: Corrosion tests in artificial atmospheres, salt spray tests*. International Organization for Standardization. https://www.iso.org/standard/63543.html
ISO. (2012). *ISO 16151:2012: Corrosion of metals and alloys, accelerated cyclic tests with exposure to acidified salt spray, dry and wet conditions*. International Organization for Standardization. https://www.iso.org/standard/55975.html
ISO. (2018). *ISO 12944-6:2018: Paints and varnishes, corrosion protection of steel structures by protective paint systems, Part 6: Laboratory performance test methods*. International Organization for Standardization. https://www.iso.org/standard/64949.html
Knudsen, O. O., Forsgren, A., and Steinsmo, U. (2018). Corrosion protection of steel structures by protective paint systems and its comparison between accelerated laboratory tests and outdoor exposure. *Journal of Protective Coatings and Linings*, 35(4), 20-31. https://www.paintsquare.com/jpcl
Leidheiser, H. (1987). Corrosion of painted metals, a review. *Corrosion*, 38(7), 374-383. https://doi.org/10.5006/1.3577348
Skerry, B. S., and Eden, D. A. (1991). Electrochemical testing to assess corrosion protective coatings. *Progress in Organic Coatings*, 19(4), 379-396. https://doi.org/10.1016/0033-0658(91)87015-H
Stratmann, M., Feser, R., and Leng, A. (2020). Corrosion protection by organic films. *Electrochimica Acta*, 39(8-9), 1207-1214. https://doi.org/10.1016/0013-4686(94)E0042-N
Weinlaender, M., Kainer, P., and Leitner, E. (2019). Mechanisms of filiform corrosion on painted aluminum alloys. *Corrosion Science*, 155, 115-124. https://doi.org/10.1016/j.corsci.2019.04.028