← Back to the main guide Aerospace and airfield topic guide

Environmental Reliability, Failure & Maintenance

Thermal, mechanical, chemical, weathering, contamination, coating, inspection, cleaning, and maintenance risks for aviation glass.

Engineering illustration

Environmental risk and acceptance matrix

A flat witness coupon, a finished glass component and a complete equipment configuration answer different questions. Select the specimen level that reproduces the relevant surface, edge, mounting, seal and functional optical path.

Environmental and Mechanical Reliability

Reliability failures usually come from interactions rather than one weak material property. A temperature gradient acts through part thickness and mounting restraint; vibration becomes damaging near a structural resonance; salt or deicing fluid becomes more serious when a coating edge or seal is already vulnerable. This chapter explains those mechanisms first, then shows how a representative component or assembly test can answer the actual service question.

Thermal shock and temperature extremes

Thermal reliability is an assembly problem. Glass composition and CTE matter, but finished thickness, curvature, edges, residual stress, coating absorption, housing stiffness, seal behavior, heater location and transition rate determine the actual risk.

SCHOTT's BOROFLOAT thermal-shock data are a useful material-screening example: the reported 5% fracture-percentile result varies with sheet thickness and a defined water-quench method. It does not rate a molded prism, coated dome or mounted aerospace optic.

For covered airfield fixtures, FAA AC 150/5345-46F provides complete-fixture operating and storage conditions and includes a specific inset-light surface-temperature test. Its 160 °C limit belongs to that configured fixture test; it is not a universal glass temperature rating. Record temperatures at the source, filter, glass, frame and seal, then inspect cracking, coating, leakage, photometry and color in the relevant hot, cold and recovery states.

Mounted thermal-cycle concept

Impact, static load and vibration

Energy alone is not enough. A blunt pendulum, steel ball, hailstone, and sharp runway fragment can carry the same joules while producing very different contact stress. Support, strike location, temperature, edge condition, and mounting change the result as much as the material name.

FAA assembly example. AC 150/5345-46F requires a Type L-850 complete light to withstand repeated steel-ball impact at 29.5 ft·lbf (40 J). This is useful context for inset-light design, but it is not a generic 40 J rating for every loose prism or aircraft-light cover.

A meaningful component specification records:

• impactor mass, material, nose shape, energy or velocity;
• strike location, temperature, support, mounting, and number of impacts;
• production-representative edges, coatings, and prior conditioning;
• acceptance—no breakage, penetration, fragment release, leakage, optical damage, or functional loss.

The result is transferable only to configurations that preserve those conditions.

Common mistake: copying the FAA 40 J complete-fixture impact value onto a loose runway-light prism drawing. The housing, support ring, gasket, impactor geometry, and strike location are part of the test result.

The three loads answer different questions:

Static load

A relatively slow force reveals deformation, broad contact stress, and whether the housing carries load around the glass.

Impact

A short energy pulse creates local peaks and stress waves; flaw location and contact shape dominate.

Vibration

Repeated excitation reveals resonance, loosening, fretting, seal movement, and glass contact that may not appear in one load event.

For scale, FAA AC 150/5345-46F uses up to 15 g from 20–2,000 Hz for complete in-pavement fixtures, alongside the separate static, 11,000 lb shear, and 40 J impact requirements described above. Passing one exposure does not predict the others.

Testing the actual housing, seals, fasteners, source, and optic is important because a sequence can create hidden damage: vibration may loosen a retainer, impact may chip an edge, and a later leakage or thermal test may reveal the functional consequence.

Failure clue: a lens that survives a single impact but develops chips after vibration often has intermittent contact with a retainer. Increasing thickness may not help until that contact is removed.

Weathering, abrasion and water ingress

Outdoor exposure must reproduce the actual surface, coating, texture and cleaning route. A hard glass substrate can retain optical function while a softer coating, printed edge or seal interface fails first.

Remove loose hard particles before wiping; otherwise maintenance can convert removable dust into permanent abrasion. Water can enter through seals, fasteners, cable penetrations, porous housings or pressure breathing even when the glass remains intact. Replacing a fogged lens without correcting the ingress path commonly produces a repeat failure.

Accelerated UV or weathering exposure supports a defined comparison or qualification condition. It should not be converted directly into a universal number of field years without a validated acceleration model.

Environmental exposure paths

Fluids, salt fog and coating durability

Use the fluid, concentration, temperature, contact pattern, wiping, rinse and recovery condition expected in service. One long immersion can miss damage caused by repeated short maintenance cycles.

A flat witness coupon can show basic film stability but cannot reproduce curved-part deposition, recessed surfaces, edge coverage, metal retainers, galvanic couples or seal contact. Identify the specimen and configuration in every report. If a coating fails, compare a protected-surface location, revised stack, different cleaner or uncoated route against lifecycle and functional requirements rather than approving an appearance-only repair.

How is long-term reliability demonstrated?

Reliability evidence becomes stronger as the specimen represents more of the production configuration. Each level answers a different question.

A reliability plan records the tested configuration and drawing revisions, sample count and production representativeness, test sequence and preconditioning, optical and leakage data before and after exposure, acceptance limits, anomaly and retest rules, and batch and change traceability. Track functional drift—not merely survival.

Use a justified sequence where service stresses interact. Vibration may relax a retainer, thermal cycling may pump moisture past the seal, and a later leakage or optical test may reveal the combined weakness. Separate pristine specimens can each pass one exposure while never reproducing that sequence. Accelerated cycles must not be converted into calendar life without a validated model.

Evidence hierarchy

Failure Modes, Inspection and Maintenance

A scratch, chip, haze patch, or color change is not equally important everywhere on a glass component. Its significance depends on location, growth, optical zone, sealing function, and the light or sensor task. The aim of inspection is therefore to understand what the indication means, not merely to count defects. These questions link visible symptoms to likely causes, functional risk, cleaning limits, and replacement decisions.

Visible defects and disposition matrix

Do not create universal crack, chip or scratch limits from this guide. Measurable acceptance and repair limits belong in the controlled drawing, aircraft AMM/CMM, equipment manual or approved maintenance data.

Clean only by an approved method before classifying a surface indication. Record photographs with scale, optical zone, loaded edge, seal land, operating history and environmental condition. Never polish a suspected crack or blend a chip unless approved repair data define the method, remaining dimensions and reinspection. Repeated damage at one clock position should trigger inspection for a burr, distorted retainer, hard contact, uneven gasket compression or packaging problem.

Condensation and color-shift diagnosis

Classify the symptom before selecting corrective action.

A heater clearing the window does not prove that the enclosure is dry, and cleaning the surface does not prove that the underlying coating remains within specification. Use controlled substitution only when the design permits it, and do not pair an off-nominal filter with a drifting source to create an uncontrolled matched set.

Failure diagnosis comparison

Post-impact inspection and root-cause analysis

Follow the aircraft maintenance manual, airport procedure or equipment manual for the event. An intact transparent surface does not prove that the reflector, housing, gasket, fasteners, wiring or optical alignment behind it remain serviceable.

1. Contain and isolate. Protect personnel and equipment, quarantine affected parts and follow event-specific operational instructions.
2. Preserve the as-found state. Photograph installation, fragments, contacts, seals, contamination, orientation and associated hardware before cleaning or disassembly.
3. Build failure hypotheses. Reconstruct part identity, lot, hours, maintenance, weather, impact and assembly history.
4. Examine the evidence. Use fracture, dimensional, material, coating and assembly inspection to test plausible causes rather than assuming the bulk glass was weak.
5. Confirm root cause and contributors. Separate the initiating event from point contact, damaged edges, seal load, thermal state or process variation.
6. Define corrective action. Address the relevant drawing, edge process, mold, coating, gasket, torque, packaging, training or inspection control.
7. Verify on a representative configuration. Repeat the critical exposure and functional acceptance with production-representative hardware.
8. Monitor later lots and field units. Track recurrence, effectiveness and any configuration-dependent pattern.

Do not clean, polish, separate or mix fragments before photography and records are complete. Sending only broken glass to a laboratory without the housing, seals, retention condition and installed orientation is usually insufficient to identify root cause. Return-to-service remains a complete-equipment decision under the applicable approved maintenance data.