Three methods cover almost every industrial need: the Taber abrasion test (ASTM D4060, JIS R3221) for coated and bulk glass, the oscillating sand method under ISO 6370-2 for comparative surface wear, and instrumented scratch testing for thin films and chemically strengthened glass. Coatings and ion-exchange treatments change how glass fails under abrasion, so the right method depends on the failure mode you actually expect in service, not just habit or convenience.
TL;DR:
- The choice of abrasion test depends on the failure mode expected in service, with Taber best for frictional wear on coated or bulk glass.
- ISO 6370-2 specifies sand abrasion conditions for outdoor applications, with about ±5% uncertainty, while ASTM D4060 and JIS R3221 govern Taber testing with different cycle and load parameters.
- Proper sample preparation, including cleaning, oven-drying, and desiccation, is crucial to obtaining reproducible and accurate abrasion data.
- Instrumented scratch testing provides detailed failure onset information for thin films and strengthened glass, but requires multiple repeats and careful interpretation.
- Standards promote fixing abrasive media, cycle counts, and loads, preventing unreliable results from improvisation or inconsistent procedures.
Table of Contents
- Choosing the right abrasion testing glass method for the job
- Which standards govern glass abrasion testing
- Preparing and weighing samples for repeatable results
- Running and interpreting the Taber abrasion test
- Oscillating sand testing under ISO 6370-2
- Instrumented scratch testing for coatings and strengthened glass
- Processing wear data and reporting uncertainty properly
- Reading abrasion failure modes correctly
- Reducing variance and catching bad data before it ships
- How Precision Glasses applies these methods in production
- Why standards discipline beats improvisation in abrasion testing
- Sources
Choosing the right abrasion testing glass method for the job
Rotating-wheel testing, commonly called the Taber method, drags abrasive wheels across a rotating specimen to simulate frictional wear from repeated contact, such as touchscreens or dashboard covers rubbed by hands and sleeves. It works well as a fast comparative screen for coated glass because it produces measurable mass loss or optical haze within a few hundred cycles.
Oscillating sand and sand-drop tests instead simulate particle erosion, the kind of abrasion caused by windblown grit on architectural glazing or solar panel covers exposed outdoors for years. Instrumented scratch testing sits apart from both. Rather than accumulating damage over cycles, it applies a single controlled scratch and records force, depth and cracking onset in real time, which makes it the method of choice for thin films and chemically strengthened glass where mechanistic detail matters more than bulk wear rate.
- Rotating-wheel (Taber): frictional wear screening, ideal for coated display and dashboard glass.
- Oscillating sand / sand-drop: particle erosion, suited to outdoor and environmental exposure testing.
- Instrumented scratch: onset-load and hardness data for thin films and strengthened substrates.
Method selection should mirror the abrasive mechanism the component meets in service. No single test replicates every wear condition, which is precisely why standards bodies maintain separate protocols rather than one universal procedure.
Which standards govern glass abrasion testing
ISO 6370-2 is the reference standard for the oscillating sand method, specifying abrasive media, load, and a comparative measurement protocol with a reported uncertainty of approximately ±5%. ASTM D4060 and JIS R3221 both govern Taber-style rotating-wheel testing, though wheel grade, load and cycle count differ between the two documents, so specify which standard governs a given contract.
- ISO 6370-2: oscillating sand, three defined test periods, reference float-glass plates.
- ASTM D4060: Taber wheel abrasion, mass loss per 1,000 cycles, common in coatings and flooring.
- JIS R3221: Japanese equivalent Taber configuration, often specified alongside ASTM D4060 for export components.
Standards fix the abrasive grade, ball or wheel dimensions, and cycle timing. What they leave discretionary, such as specimen thickness or mounting fixture design, still needs documenting in your own test report for the result to be reproducible elsewhere.
Preparing and weighing samples for repeatable results
Poor sample prep is the single most common cause of scattered abrasion data, and it is entirely avoidable with a fixed sequence.
- Clean each specimen with distilled water, then rinse with ethanol at 96 to 98% concentration to remove residues that would otherwise register as false mass loss.
- Dry samples in an oven at 120°C ±5°C, a step ISO 6370-2 requires to strip residual moisture before weighing.
- Desiccate before weighing on a calibrated balance capable of 0.2 mg repeatability, the tolerance the standard demands for credible comparative data.
- Check specimen geometry and fixturing, particularly for thin or flexible glass, where membrane stresses can distort readings if the sample isn’t rigidly supported.
Pro Tip: Skip the desiccation step and your mass-loss figures will be dominated by ambient humidity, not abrasion. Build a 30-minute desiccator hold into your protocol as standard, not as an optional extra.
Thin glass panels behave as components rather than monolithic slabs under load, and large deflections can introduce membrane stresses that standard clamping ignores. Best-practice guidance for LCD glass testing recommends strain gauging or non-linear analysis for these cases, a caution equally relevant to abrasion fixturing.
Running and interpreting the Taber abrasion test
A typical Taber setup mounts the specimen on a rotating turntable beneath two abrading wheels, often CS-17 grade for coated glass, under a specified applied load. Cycle counts are chosen to match the coating’s expected service life or a standard’s default interval, then results are read out in several ways:
- Cycles to failure: the cycle count at which a defined optical change (haze or transmission loss, ΔT) occurs.
- Peel rate per cycle: mass or coating thickness lost per cycle, the standard metric for thin films deposited onto glass.
- Optical degradation: transmission or haze measurements taken at intervals rather than only at the endpoint.
Thin coatings frequently fail by peeling rather than uniform wear, which limits how well Taber data alone predicts long-term durability. Coatings with higher adhesion strength, greater hardness and lower friction all show reduced peel rate, and hardness tends to dominate that relationship. Where peeling dominates, pair Taber results with adhesion testing to separate wear resistance from bond failure.
Oscillating sand testing under ISO 6370-2
The oscillating sand method drops or oscillates abrasive media across the specimen surface for three consecutive 30-minute periods, a structure fixed by ISO 6370-2 to allow direct comparison between labs. Abrasive selection changes the severity of the test: sanidine (grade P100) represents moderate abrasion conditions, while fused aluminium oxide (grade P80) produces harsher, faster wear.
| Abrasive medium | Grade | Severity |
|---|---|---|
| Sanidine | P100 | Moderate abrasion, general comparative testing |
| Fused aluminium oxide | P80 | Harsher abrasion, accelerated wear screening |
Relative wear (w) is calculated against reference float-glass plates run alongside the test batch, which corrects for day-to-day variation in abrasive feed rate or machine wear. Report results as relative mass loss against the reference plate rather than an absolute figure, and quote the standard’s own comparative uncertainty of roughly ±5% alongside every result.
Instrumented scratch testing for coatings and strengthened glass
Instrumented scratch tests apply either a constant load or a progressively ramped load through a defined indenter, typically a diamond or hard-metal tip with a specified radius, drawn across the surface at a fixed velocity over a set track length. Multiple repeats per specimen are standard practice, since scratch response varies more between runs than Taber mass-loss data typically does.
- Lateral force and indenter displacement: recorded continuously to build a force-versus-position profile.
- Scratch hardness (HS): derived from applied load and residual track width.
- Onset loads: the specific load at which microcracking or microabrasion first appears, a critical acceptance threshold for optical components.
Statistic callout: sodium/potassium ion exchange introduces surface compressive stresses of 200 to 340 MPa, which raises scratch hardness and surface Young’s modulus, but the same treatment can shift microcracking onset to lower normal loads in some configurations, an effect worth checking before assuming strengthening always improves scratch resistance outright.
Pairing scratch data with post-mortem fractography or profilometry gives a fuller picture of which damage mechanism actually governs failure, rather than relying on a single force threshold.
Processing wear data and reporting uncertainty properly
Raw mass-loss and force readings only become useful once processed against a consistent formula and reported with honest uncertainty bounds.
- Calculate relative mass loss as the ratio of specimen mass loss to reference-plate mass loss, following the ISO 6370-2 approach, rather than reporting absolute grams alone.
- Derive peel rate per cycle by dividing total coating mass or thickness lost by the number of cycles completed before the defined failure point.
- Run at least three replicates per condition and report the mean alongside the standard deviation, not just a single figure.
- Apply Weibull statistics when abrasion outcomes link to strength distributions, since glass strength data is notoriously non-normal and a simple average can mislead.
- State whether a figure is absolute (grams, microns) or comparative (relative to a reference plate), because mixing the two within one report is one of the most common sources of confusion between labs.
Reading abrasion failure modes correctly
Abrasion damage on coated or strengthened glass rarely looks the same twice. The visual signature tells you which mechanism is at work. Peeling shows as flaking or delamination along scratch edges, usually pointing to weak adhesion rather than poor bulk hardness. Microabrasion appears as diffuse haze without discrete cracks, a sign of gradual surface removal. Microcracking shows as sharp, often branching lines at or beyond a specific load threshold, the kind of damage ion-exchanged glass can develop at lower loads despite its higher surface hardness.
- Peeling: low adhesion strength, visible flaking, often independent of substrate hardness.
- Microabrasion: gradual haze, correlates with coating hardness and friction coefficient.
- Microcracking: discrete cracks at a defined onset load, sensitive to residual stress from ion exchange.
Pro Tip: Don’t set acceptance criteria purely on peel rate if microcracking onset is the more likely field failure mode. Match the acceptance metric to the actual abrasive exposure the component will see, not the easiest number to measure.
Reducing variance and catching bad data before it ships
Abrading wheels wear unevenly and drift out of specification faster than most labs expect, so schedule routine recalibration of wheels and balances rather than waiting for anomalous results to trigger a check. Inspect specimens for edge chipping or handling damage before testing, and discard any sample where damage falls outside the intended test area.
- Recalibrate abrading wheels and weighing balances on a fixed schedule, not reactively.
- Reject specimens with pre-existing edge chips or handling marks before the test begins.
- Combine Taber, instrumented scratch, and optical or profilometry inspection for a rounded verdict rather than trusting one number.
How Precision Glasses applies these methods in production
Laboratory and fabrication teams design abrasion protocols around the coating chemistry and substrate a project actually requires, rather than defaulting to a single standard test. That means selecting between Taber, oscillating sand, and instrumented scratch based on the abrasive exposure a component will meet in aerospace, medical, or optical service, and reporting results with traceability back to batch records.
Bespoke protocols matter most for chemically strengthened glass and thin coated components, where a standard cycle count rarely maps directly onto the acceptance criteria a defence or medical device contract demands.
If your specification calls for traceable abrasion data alongside custom fabrication, explore Precision Glasses’ quality and testing processes or view our sectors to see where these protocols are already applied.
Why standards discipline beats improvisation in abrasion testing
The temptation in most labs is to run whatever abrasion rig is already on the bench and report a number, but that habit produces data nobody outside the building can trust. Standards like ISO 6370-2 and ASTM D4060 exist precisely because comparative wear figures mean nothing without a fixed abrasive grade, load, and cycle structure behind them.
What gets underestimated is how much ion exchange complicates the picture. Chemically strengthened glass often reads as tougher under scratch testing, yet the same compressive stress that raises hardness can pull the microcracking threshold down. Treating “strengthened” as automatically “more abrasion resistant” without checking onset loads is a genuine trap, and one that shows up in field failures more often than test reports.
The gap between a Taber pass and a real-world lifetime prediction is where most specification arguments happen. A coating can survive thousands of Taber cycles and still fail early in the field if the dominant field mechanism is particle erosion rather than frictional wear. Matching the test to the actual service exposure, not the test that is quickest to run, is the discipline that separates useful abrasion data from a number that merely looks rigorous on paper.
— Alexandra
Sources
ISO 6370-2 remains the definitive reference for oscillating sand abrasion, including abrasive media, reference plates, and uncertainty guidance. The Taber film study in Wear explains peel-rate mechanics for coated glass in detail. The chemically strengthened glass hardness study clarifies ion-exchange effects on scratch thresholds, and Corning’s LCD testing guidance covers thin-glass handling that abrasion labs routinely overlook. For abrasive media handling and disposal equipment, Expleco’s glass processing systems are worth reviewing alongside your test-lab setup.
- SIST EN ISO 6370-2:2021 – Abrasion Resistance Test for Vitreous
- Abrasion of thin films deposited onto glass by the Taber test
- Hardness and scratch resistance of chemically strengthened alkali‐borosilicate thin glass
