Screen Damage Modes: Distinguishing Production Defects From Field Damage

A returned display arrives with a fault and an argument attached. The supplier sees damage that looks like handling; the customer sees a unit that failed in normal use. Both are sometimes right, and the evidence that decides the question is usually present on the unit already, if anyone looks at the right features before the panel is cleaned, re-fitted or discarded.

Part of the Quality, Testing & Compliance guide — 33 articles on defect classes, inspection, qualification and certification.

What is the difference between a production defect and field damage?

When it formed, and whether use could have formed it.

A production defect exists before installation and traces to a process, a material or a component. Field damage forms afterwards, from impact, abrasion, chemicals, heat, moisture or electrical stress. The two overlap in appearance, so the discriminator is not how bad the fault looks but where it sits, what its edges look like and whether the failure mode is reachable by the conditions the product actually met.

The production defect families

Process, material and component faults.

Production defects include pixel and line defects from driver or bond faults, mura and uniformity variation from the backlight or cell, bonded-stack bubbles from lamination, contamination sealed inside the stack, and dimensional or cosmetic faults on the glass and the bezel. Their common property is that they are distributed across a batch in a pattern set by the process, which is why a single-unit assessment is weaker evidence than a sample drawn from the same lot. The controls that reduce them are the substance of quality assurance and factory compliance, and the automated inspection that catches them is described in automated optical inspection for LCD shipments and, for a stricter customer base, automated optical inspection for automotive displays.

The field damage families

Impact, abrasion, chemistry, heat and moisture.

Field damage clusters around the point of contact. A single impact produces a starburst crack with radiating lines and often a damaged bezel nearby. Repeated abrasion produces a field of fine scratches with a directional pattern. A chemical attack dulls or crazes the surface in a shape that follows a wipe. Heat leaves discolouration near the hottest component. Moisture leaves a corrosion or tide mark at an edge or a connector, and a crack that follows a mounting screw shows that the enclosure, not the panel, applied the load.

Reading the evidence: morphology and location

Where and what shape, before what colour.

A defect’s position is the first discriminator. Faults at the glass edge, under a gasket or at a screw boss point to assembly or enclosure stress; faults in the middle of the active area point to the panel or the stack; faults at a connector point to handling. The shape is the second: a sharp, branching crack with an impact centre is mechanical, while a gradual dulling or a patch of fine scratches is wear. Documenting both before cleaning the unit is essential, because wiping a display destroys the fingerprint evidence that would have settled the question.

Time as evidence

A defect seen at incoming inspection is a supplier problem.

The single most useful control is a documented incoming inspection, because a defect recorded on arrival cannot be attributed to the field. Where that record does not exist, the fault history does the same work: a fault that appears within hours of installation, appears on several units from one batch or appears before the product ever ran is a production defect by inference. A fault that appears after a specific event, a specific operator or a specific installation is field damage until proven otherwise.

Scratch versus abrasive wear

One event or many.

A scratch has a start, a direction and usually a terminus, and it often has raised material at its edges. Abrasive wear is a field of shallow marks with a common orientation and no single origin. The distinction changes the answer: a single scratch is an incident, while a wear pattern is a design issue about the surface treatment, the cleaning method or the user’s tools. Hardness of the cover glass and the coating set how quickly a given abrasion regime produces visible damage, and the optical consequences of surface treatment are measured in the way described for brightness, contrast and uniformity.

Dust, fungus and moisture in the stack

Inside the stack is a process fault; on the surface is not.

Particles sealed between the layers cannot have arrived from the field, so contamination inside a bonded stack is a production defect with no ambiguity. Particles on the outer surface are not a defect at all. Fungus growth requires moisture and organic material and points to a sealing or storage failure rather than a panel fault, and it usually appears in the same corner where the gasket is compromised. Moisture between layers produces a tide mark or gradual haze that changes with temperature, and the edge where it enters identifies the sealing defect.

Delamination, bubbles and edge ingress

Progress along an edge is an ingress story.

A bubble in the middle of a bonded area that appears after thermal cycling is a lamination fault; delamination that begins at a cut edge and spreads inwards is an ingress or adhesion fault made worse by the environment. The two are separated by the direction of growth and by the presence of a seal. Because both reduce contrast and can eventually affect touch, the acceptance criteria should be written before the units are built, and the adhesive families and their environmental behaviour are documented by materials suppliers such as DELO and the glass suppliers such as AGC.

Packaging and transport as a damage source

Many field failures were damaged before the field.

Shock during transport produces corner damage, cracks at the mounting points and connector damage, all of which appear later as field faults. A packaging specification with a drop test and a shock recorder, and an inspection step that photographs each unit’s corner condition on arrival, converts an unfalsifiable claim into a documented one. The same logic applies to internal transport within a factory, where an unprotected panel moved between stations collects the marks that are later blamed on the customer.

Writing an inspection standard both sides accept

Six clauses remove most disputes.

• The inspection conditions: illuminance, distance, viewing angle and time allowed.
• The defect classes and the size or count allowed for each, with a zone map.
• The measurement method for any defect that is judged, not counted.
• The sampling plan for incoming inspection and the record to be kept.
• The handling and cleaning instructions that preserve evidence.
• The escalation path when a claim is disputed, including who pays for analysis.

With those clauses agreed, a return becomes a classification rather than a negotiation. The underlying measurement expectations are set by the CIE for anything judged in colour, the traceability language of NIST for instruments, IPC for assembly workmanship and IEC 62368-1 for the safety context of the equipment, with EMC obligations under the EU EMC Directive. Where the display is part of a machine rather than a monitor, the same classification applies to the differences set out in industrial versus consumer monitors.