Defect Guide
Click a defect to expand causes, acceptance criteria (ASTM C1048 / C1036), and corrective actions. Severity reflects impact on function and rework risk.
Repeating wave pattern visible when viewing reflected images (light pole, horizontal line). Distortion runs perpendicular to the direction of glass travel through the furnace. More visible in larger lites and Low-E coated glass.
Peak-to-valley amplitude ≤ 0.8 mm over a 300 mm gauge span. Roller pitch marks ≤ 1.6 mm peak-to-valley over a 50 mm span.
Amplitude > 0.8 mm over 300 mm span. Visible as "washboard" in reflected images. Noticeable from 3 m in architectural glazing.
- A ceramic roller with high TIR (total indicated run-out), the out-of-true portion of the roller pushes the glass upward once per rotation as it rolls through, while the glass is still hot and soft enough near the end of the heating cycle to take the shape
- Roller speed too fast relative to furnace temperature, glass not fully softened
- Furnace temperature too low, insufficient viscosity for roller contact
- Glass thinner than recommended minimum for furnace roller pitch
- Incorrect glass positioning, off-center load distribution
- Reduce roller speed (increase soak time) until glass reaches correct viscosity
- Raise furnace temperature by 3–5°C and retest
- Inspect rollers for roundness; replace rollers showing > 0.1 mm runout
- For thin glass (4–5 mm), verify roller pitch ≤ 100 mm
- Center load on conveyor; avoid side-feeding large lites
Diffuse, milky or opalescent cloudiness on the glass surface. May cover the entire lite or appear as localised patches. Worse in raking light or bright transmitted light. Critically, there are two types that look similar but have completely different causes and outcomes.
Caused by deposits sitting on the glass surface, wash water minerals, cutting fluid residue, separator powder, or lightly adhered roller debris. These have not bonded chemically to the glass and may be removed if caught early.
Caused by the tempering recipe itself. When the furnace temperature is too high or the soak time too long, the glass surface begins to devitrify, a partial crystalline structure forms in the top layer of the glass. This is a change inside the glass, not on it. It cannot be polished, wiped, or cleaned off. Immediate reject, return to furnace settings.
Recipe-related haze also occurs when furnace rollers are too hot and brand trace marks onto the glass, or when combustion atmosphere gases react with the glass surface in gas-fired furnaces. Both are permanent.
If one face of the glass heats faster than the other during the heating cycle (commonly the bottom surface, from conduction off the rollers), the lite bows slightly while still in the furnace. A bowed lite only touches the rollers along a narrow central band, so the full weight of the glass concentrates onto that small area, sometimes hard enough to leave a white haze mark or, in severe cases, a slight surface deformation visible as a "lensing" distortion at an acute viewing angle. This is most common immediately after a roller wash and is worse on Low-E coated glass, since the coating reflects IR away rather than absorbing it, widening the top/bottom temperature gap unless the recipe compensates.
Minor localised haze (< 25 mm diameter) in the peripheral 50 mm zone only, not visible from 3 m in normal illumination.
Any haze in the vision area. Haze covering > 5% of peripheral zone. Any haze visible from 3 m in diffuse light. All devitrification haze regardless of location.
- Hard water mineral deposits (calcium, magnesium scale) not fully rinsed before furnace entry, bake onto surface at temperature
- Contaminated or exhausted wash water / spent detergent leaving film on glass
- Incompatible cutting fluid or separator powder burned in at furnace temperatures
- Dirty rollers transferring debris to glass surface
- Ceramic fiber off-gassing in new or repaired furnace lining, deposits on glass surface
- Edge deletion residue or crayon marks on coated glass not fully removed pre-furnace
- Furnace temperature set too high, surface of glass exceeds softening point and begins to devitrify
- Soak time too long at high temperature, cumulative overheating of glass surface even if peak temp is correct
- Roller surface temperature too high, localised contact haze following roller pitch pattern
- Gas-fired furnace combustion atmosphere issues, sulphur or combustion products reacting with glass surface
- Low-E coated glass run with incorrect recipe, coating chemistry interacts with glass surface at elevated temps
Type 1 (contamination): Light mineral haze caught early may respond to approved glass cleaners. A fine cerium oxide polish (applied with a soft felt pad or micro-mesh cloth, NOT abrasive grit mesh) can sometimes remove light surface deposits without scratching. Always test in a corner first.
Type 2 (devitrification / furnace): Cannot be removed by any cleaning method. The crystalline change is within the glass structure. Reject and investigate recipe.
Most glass manufacturers explicitly restrict certain cleaning methods and will void warranties if non-approved processes are used. Common restrictions include:
- → Abrasive scrubbing pads, steel wool, or grit mesh cloths, universally prohibited on all coated and tempered glass; scratch the surface or damage coatings permanently
- → Razor blades, many manufacturers (Vitro, Guardian, Cardinal) explicitly prohibit razor blade cleaning on tempered glass; can initiate micro-cracks
- → Ammonia-based cleaners on coated glass, can strip or discolour soft-coat Low-E; check each manufacturer's approved cleaner list
- → High-pressure washing on laminated glass, water penetration at edges can delaminate interlayer
- → Dry wiping, never wipe glass dry; always use a damp cloth or approved cleaner first to lift particles before wiping, otherwise debris scratches the surface
Always consult the specific glass manufacturer's cleaning and maintenance guide (Vitro, Guardian, AGC, Cardinal, Pilkington all publish these) before attempting any haze removal on fabricated or coated glass.
- Clean wash brushes and check nozzle pressure (minimum 1.5 bar on glass surface)
- Replace wash water and verify detergent concentration (typically 0.5–1.0%)
- Install DI or RO water system for the final rinse stage, hard water is the most common cause
- Run blank glass through furnace to clean rollers before production runs
- If recipe-related: reduce furnace set temperature 5°C and/or reduce soak time; run test lites
- Verify cutting oils and separator materials are furnace-compatible (silicone-free)
- For Type 3 (point-loading): keep the glass flat through heating, check for asymmetric heating between top and bottom; if used, sulphur dioxide (SO₂) gas can help by lubricating the glass-to-roller contact, but should only be used when necessary, since residue can damage roller surfaces over time if not cleaned off promptly with plain hot water
Glass curves along one or both dimensions. Overall bow is curvature along the longest dimension. Edge lift is upward curvature at the corners/edges when the centre rests flat. Warp is twist, opposite corners elevated.
Overall bow ≤ 0.5% of lite dimension being measured. Edge lift ≤ 0.5% of the edge length. Use the Bow Calculator tab for exact limits.
Any bow exceeding 0.5% of dimension. Warp making full surface contact with a flat table impossible (twist). Cannot achieve adequate sealant contact in IGU.
Stable bow ("water-shedding" or "water-holding") happens when the upper and lower surfaces of the glass are at slightly different temperatures the moment the glass solidifies in the quench. The hotter surface contracts more as it cools, pulling the glass into a fixed bow that sits the same way every time it's set down. The fix is a quench air balance adjustment, operators sometimes call this "blow into the bow," meaning increase quench pressure on the side the glass is bowing toward.
Unstable bow happens when there's a temperature difference across the face of the glass (not through its thickness) when it solidifies, most common in thinner, more square-shaped lites. An "oil-can" bow (centre cooler than the edges) will flex between up and down if pressed by hand; a "saddle" bow (edges cooler than the centre, opposite edges bowing in opposite directions) has a similarly loose, flippable feel. Oil-can bow is frequently traced to the centre of the roller bed running cooler than the edges after several loads in a row, running the next difficult-size glass right after an idle period (e.g. after a break) often comes out flat because the roller bed has had time to even out.
"S" bow in long, narrow lites can occur when air blows back from the quench into the furnace before the glass is fully inside, chilling the top surface near the furnace exit before the leading edge experiences the same cooling, producing opposite bow directions at the leading and trailing edges. The fix is a draught diverter on the upstream face of the upper quench box, not just a quench air rebalance.
- Uneven heating, top/bottom temperature differential in furnace
- Quench nozzle imbalance, more cooling on one face than the other
- Glass not centred on rollers, one edge heats faster
- Coating on one side (Low-E) absorbing IR differently, requires recipe adjustment
- Glass loaded at an angle, non-parallel travel through rollers
- Too-slow conveyor speed allowing one face to over-heat
- Cold centre section of the roller bed (oil-can bow) or poor air release from the centre of the quench (saddle bow)
- Measure top/bottom furnace temperature with thermocouple, balance within ±3°C
- Check all quench nozzles for blockage; verify quench pressure uniformity across width
- For coated glass, place coating face down (tin side up) to balance IR absorption
- Increase quench pressure slightly to "snap" the glass flat faster
- Check conveyor tracking; ensure glass enters parallel to roller axis
- For oil-can bow on occasional difficult sizes, run after an idle/break period or reduce furnace temperature and extend soak time to even out the roller bed
A localized downward droop or kink right at the leading or trailing edge of the lite (relative to furnace travel direction), rather than a smooth overall bow. Most visible as a visible "PLV" (peripheral lift variation) reading right at the edge when checked with a straightedge or gauge.
Edge kink happens only in the last moments of heating, while the glass is still hot and pliable. As the leading or trailing edge of the lite leaves the support of one roller and travels the unsupported gap to the next roller, its own unsupported weight causes it to droop slightly before the next roller catches it. Faster roller-bed speed and a tighter roller pitch (rollers spaced closer together) both reduce the unsupported gap and therefore reduce edge kink. The hotter and softer the glass at that point in the cycle, the more pronounced the droop.
PLV reading at the edge within the facility's internal tolerance, commonly referenced around 0.008 in (0.2 mm), though this is an industry-practice figure rather than a single published ASTM value. Confirm your project's specific requirement.
Visible droop at the leading/trailing edge exceeding facility tolerance, or interfering with IGU sealant contact or frame fit.
- Increase roller bed speed through the final section of the furnace, where the glass is hottest
- Reduce roller pitch (closer roller spacing) if the furnace design allows it
- Process the glass slightly cooler, trades off against increased quench breakage risk, so adjust carefully
- Verify load orientation, running the longer dimension parallel to travel reduces the unsupported span at the leading/trailing edges
Iridescent, rainbow-coloured patches or a repeating dotted pattern ("quench marks" or "leopard spots") visible when glass is viewed in polarised light, e.g. through polarised sunglasses, or in specific sky conditions. Most visible in tinted and Low-E glass.
Anisotropy is an inherent property of all thermally tempered glass. It is NOT a defect under ASTM C1048 or EN 12150. Visible in polarised light only, not grounds for rejection.
Visible to naked eye (without polarised glasses) in diffuse daylight. "Quench spots" visible as grid pattern in normal viewing. If client disputes, document ASTM position before tempering order.
- High quench pressure creates more stress birefringence, more visible anisotropy
- Quench nozzle spacing creates a visible grid pattern in higher-stress glass
- Thinner glass requires higher quench pressure → more anisotropy
- Low-E and tinted glass show patterns more clearly due to colour contrast
- Use heat strengthened (Kind HS) instead of FT where structural requirements allow, lower stress, less anisotropy
- Select glass with lower iron content (extra-clear), patterns less visible
- Specify diagonal quench nozzle arrays, breaks up grid pattern
- Document and set client expectations before tempering, include ASTM C1048 language in spec
Conchoidal fractures (shell-shaped chips) along the cut edge or seamed edge. May be microscopic (micro-chips not visible at 0.5 m) or large conchoidal breaks. Critical because edge damage can propagate into spontaneous breakage under thermal stress after tempering.
Chips ≤ 1.6 mm deep penetrating less than half the glass thickness, seam/grind before tempering. Seaming radius ≥ 0.3 mm per ASTM C1036.
Any edge chip > 1.6 mm depth, or any chip penetrating into the tempered tension zone (mid-thickness). Chips after tempering = reject; cannot repair.
- Dull or chipped scoring wheel, score doesn't penetrate full depth
- Inadequate scoring pressure for glass thickness
- Break-out done too quickly or at wrong angle after scoring
- Handling damage, lites contacting each other or rack posts
- Seaming wheel worn, inadequate chamfer on edge
- Holes or notches drilled too close to edge before tempering
- Replace scoring wheel at first sign of dullness (typically 500–1000 m of score)
- Calibrate scoring pressure per glass thickness (typically 30–50 N for 6 mm)
- Inspect all edges pre-furnace; seam and re-inspect any suspicious areas
- Install foam or rubber between glass lites during storage and transport
- Minimum edge-to-hole distance: 2× glass thickness, per ASTM C1048
Linear abrasion marks on glass surfaces. Faint scratches visible only in raking light ("sleeks") vs. deeper scratches visible in normal transmitted light. On coated glass, scratches may appear as bright silver lines where coating is removed.
Scratches not visible from 3 m in normal diffuse light. Width ≤ 0.5 mm, length ≤ 75 mm in peripheral zone only (50 mm from edge). None in vision zone.
Any scratch visible from 3 m in the vision area. Any scratch > 0.5 mm wide anywhere. Coating damage visible from 3 m. Pattern of parallel scratches (washer tracks).
- Dirty washer brushes, hard particles embedded in bristles
- Glass-to-glass contact during handling or storage
- Abrasive particles on conveyor rolls or suction cups
- Incompatible coated side facing abrasive surface
- Hard water scale on washer nozzles scratching surface
- Replace washer brushes on schedule; rinse brushes before startup each shift
- Inspect all rubber/foam spacers; replace when hardened or contaminated
- For coated glass, handle coating face inward; never place face-down without protection
- Clean suction cups daily; inspect for embedded particles
Localised milky or iridescent patches, usually on the tin side (bottom surface of float glass). Appears as "staining" that cannot be wiped off, it is in the glass, not on it. Often correlates with storage marks or moisture exposure of packs.
- Tin side exposed to moisture during storage, tin oxide reacts with condensation
- Interleaf powder (calcium carbonate) absorbing moisture and holding it against tin side
- Temperature cycling causing condensation inside glass packs
- Outdoor storage without weatherproof covers
- Store glass indoors or in sealed, climate-controlled environment
- Ensure packs are angled ≥ 3° so water drains rather than pooling between lites
- Use appropriate quantity of separator powder, excess powder traps moisture
- Inspect incoming glass stock before cutting; quarantine affected packs
- White haze in the vision zone is typically reject; peripheral zone per ASTM C1036 q3
- Nickel sulfide inclusions (NiS), small crystalline impurities that expand post-temper, causing delayed breakage weeks or months after installation
- Edge damage (chips, micro-cracks from handling)
- Thermal stress (inadequate edge clearance, shade line from blinds)
- Over-tempering, surface compression too high for glass dimensions
- Frame contact, glass touching metal frame without adequate clearance/setting block
NiS breakage produces a characteristic "butterfly" pattern: two large pieces originating from a small central origin (1–3 mm), with a rough "mirror zone" around the inclusion. The breakage origin will show a tiny, shiny, spherical inclusion if examined under magnification.
- Specify heat soak tested (HST) glass per EN 14179 for facades and critical glazing, forces latent NiS to break in the oven, not on the building
- Ensure minimum edge clearance ≥ 3 mm (typically 5 mm) around all edges
- Use setting blocks at ¼ points; never allow glass to rest on frame bottom
- Shade analysis for large south-facing lites with internal blinds
Bow & warp acceptance
Bow and warp are now recorded and scored per batch in the Tempering QC Log, against the ASTM C1048 overall-bow allowable (0.1% of the piece length), alongside fragmentation, roller wave, surface compression, and the visual check.
Score bow / warp in the Tempering QC Log
Enter the piece length and measured bow (mm or inches) and it scores each batch against the ASTM C1048 allowable and your internal limit, in one signed shift report.
Open the Tempering QC Log →Tempering Recipe Reference
Approximate starting-point ranges for horizontal roller furnace operation. These are not fixed values, every furnace brand, age, and condition produces different results for the same settings. Always validate with your equipment manual and a fragmentation test.
| Thickness | Furnace temp (approx °C) | Heat time (approx s/mm) | Total cycle (approx) | Quench pressure (approx) | Quench time (approx) | Key watch-outs |
|---|
Environmental & Seasonal Factors
Shop conditions have a real impact on furnace performance. The same recipe set in January may over- or under-temper glass in July if not adjusted. Track these variables in your furnace log.
Cold weather / winter shop
Glass entering the furnace is colder (may be near 0°C if stored outside overnight). The furnace must work harder to bring it to target temperature. Increase heat time 3–8% or raise furnace set-point 3–5°C in cold conditions. Allow longer warm-up time for the furnace at start of shift.
Hot weather / summer shop
Glass enters warmer (25–35°C vs. typical 15–20°C baseline). Less heat energy needed to reach softening point. Reduce heat time ~3–5% or monitor fragmentation counts closely, over-tempered glass shows very fine dust-like fragments. Quench air is also warmer, slightly reducing quench efficiency.
High humidity (> 65%)
Humid air holds more moisture, which can affect glass surface at high temperature and increase mineral deposit risk from wash water. Increase furnace temp 2–3°C to compensate. Check wash water quality and DI filter more frequently in humid seasons.
Low humidity / dry conditions (< 30%)
Very dry shop air increases static on glass surfaces, attracting dust before furnace entry. Increases risk of surface contamination haze. Ensure washer is functioning correctly and glass enters furnace promptly after washing. No significant recipe change needed, mainly a cleanliness issue.
High altitude locations
Lower air density at altitude (e.g., Calgary at 1,048 m, Denver at 1,609 m) means less convective heat transfer in the furnace and reduced quench efficiency. Increase heat time 5–10% and increase quench pressure 10–15% compared to sea-level baselines. This effect is significant above ~1,000 m elevation.
Cold glass from outdoor storage
Glass stored outdoors overnight in winter can be at −10°C to −20°C. This is the most extreme cold-start condition. Bring glass indoors at least 2–4 hours before processing to allow it to reach shop temperature. Running very cold glass risks thermal shock at furnace entry and will require significantly longer heat times, do not try to compensate with recipe alone.
Furnace warm-up state
A furnace that has just been started (first 30–60 min of the day) has cold rollers and uneven zone temperatures. Run 2–3 blank lites or scrap pieces before any production glass. Recipe performance will stabilise as rollers and refractory reach thermal equilibrium.
After a furnace shutdown or power interruption
If the furnace was off for > 1 hour, assume full warm-up procedure is needed. Rollers, refractory, and quench components all need to stabilise. Run test lites with fragmentation checks before resuming production, even if the furnace shows correct set-point temperatures.
Glass Type Modifiers
Apply these as relative adjustments from your established baseline for clear float glass at the same thickness.
Soft-coat Low-E (sputter)
Coating absorbs IR more efficiently than bare glass. Reduce furnace temp 5–10°C or reduce heat time 5–8%. Place coated side facing down in the furnace. Test for bow, coating-side-down often reduces bow on coated glass.
Hard-coat Low-E (CVD / pyrolytic)
Less IR effect than soft-coat. Reduce heat time ~3–5%. Harder coating is more scratch-resistant through furnace. Monitor for haze, some hard coats are sensitive to roller temperature at the contact point.
Body-tinted glass
Absorbs more heat in the glass body than clear glass. Reduce furnace temp 5–8°C or reduce heat time ~5%. Bronze and grey tints absorb more than blue tints. Edge heating lag can cause edge stress cracks if soak is too long.
Reflective coated (sputtered)
High reflectivity reduces IR absorption significantly. Increase heat time 8–15%. Use high-convection furnace section if available. Risk of coating damage if rollers are contaminated, clean furnace thoroughly before run.
Extra-clear / low-iron
Transmits more IR, slightly faster heating. Reduce heat time ~3–5% vs. standard clear. Any remaining anisotropy will be more visible against the neutral background. Usually no quench change needed.
Large format (> 2.5 m)
Edges cool faster while the centre is still heating → bow risk. Increase top-zone temperature 3–5°C to balance heating. Use staged quench if available. Load parallel to furnace centre-line; avoid off-centre positioning.
Drilled / notched glass
Holes and notches are stress concentrators. All drilling must be completed before tempering. Reduce quench pressure 5–10% to lower stress concentration risk. Min hole diameter = glass thickness; distance hole-to-edge ≥ 2× thickness.
Laminate interlayer pre-assembled
Never temper glass with interlayer already applied, PVB and SGP cannot withstand furnace temperatures. Temper glass first, then laminate. If received pre-laminated for heat strengthening only, verify with your laminate supplier, most interlayers have strict temperature limits.
Fragmentation Test Reference
ASTM C1048 requires destructive fragmentation testing to verify degree of temper. Strike near the centre of the lite with a pointed tool. Count particles in any 50 × 50 mm square, avoiding 25 mm from edges.
| Glass type | Min particles / 50×50 mm | Max particle size | Pattern notes |
|---|---|---|---|
| Kind FT, Fully Tempered | ≥ 40 particles | ≤ 6500 mm² (≈ 80×80 mm) | Should shatter into small, relatively blunt-edged pieces. "Dice" pattern throughout. |
| Kind HS, Heat Strengthened | Not specified by particle count | No maximum defined | Breaks into larger, more irregular pieces similar to annealed, but stays in frame. Does NOT dice like FT. |
| Fail, under-tempered FT | < 40 particles | Large shards present | Large angular pieces indicate insufficient quench. Increase quench pressure and retest. |
| Fail, over-quenched | Very fine dust, > 100 /50mm² | Many particles < 3 mm | Excessive surface compression. Increase glass exit temperature or reduce quench. Risk of spontaneous breakage. |
Breakage Pattern Analysis
The breakage pattern of tempered glass almost always reveals the cause. Examine the origin and fragment shape to diagnose correctly before writing a warranty claim or adjusting the tempering recipe.
Small "dice" fragments, entire lite
Classic fully-tempered breakage. Origin at edge or mid-lite. If fragmentation count is correct (≥ 40/50mm²), this is normal temper performance, not a defect. Investigation: was there an initiating chip or impact? Check origin.
Origin at a single point, radial cracks from one spot on the face
Hard object impact at that point. Spider-web crack pattern radiating from a cone-shaped depression. Look for: corner of a fitting, dropped tool, stone chip. Not a manufacturing defect unless glass was under-tempered (large shards present).
Origin at edge or corner, triangular cracks spreading inward
Initiated by edge chip or micro-crack before or after tempering. Pattern: two large primary cracks from the edge origin, then dicing. Look for: handling damage, frame contact, inadequate edge clearance. Common in improperly set frames.
Origin at edge, single long crack runs across full width first
Thermal stress breakage. A long, nearly straight crack initiates at the edge (often at a shade line) before the dicing pattern. Cause: excessive temperature differential across the glass, shade line from blind, solar film on part of the glass, HVAC duct blowing on glass, or insufficient edge clearance in frame.
"Butterfly" pattern, two large pieces from tiny central origin
Nickel sulfide spontaneous breakage. The origin is a small (1–3 mm) shiny spherical inclusion in the body of the glass. The pattern: two primary wing-shaped fragments radiating from the butterfly origin point. Verify with magnification. Solution: specify heat soak test (HST) for future orders.
Large angular shards, not diced, fails fragmentation count
Glass was insufficiently quenched. Surface compression too low. Large, irregular, sharp shards, similar to annealed glass failure but worse (large under tension). Action: increase quench pressure, reduce furnace exit temperature, run fragmentation test before returning to production.
Origin at edge of a drilled hole or notch
Stress concentration at hole edge initiated the break. Common causes: hole drilled post-temper (never acceptable), hole too close to edge (< 2× thickness), hole diameter too small (< glass thickness), sharp notch without radius. All drilling and notching must occur before tempering.
Origin at a small circular pit on the surface
A hard particle (weld spatter, gravel, metal swarf) struck the glass surface. The origin shows a Hertzian cone-crack (circular ring crack) around a pit. Common on job sites. Not a manufacturing defect. Document with photos for liability records before moving the glass.
Breakage Documentation Checklist
Photograph and record these details before disturbing a breakage. This information is needed for warranty claims, insurance, and recipe correction.
| # | What to record | Why it matters |
|---|---|---|
| 1 | Overall breakage pattern photo (full lite, wide shot) | Shows number of cracks, quadrant distribution |
| 2 | Close-up of origin point (macro photo) | Identifies initiating defect type |
| 3 | Origin location (mm from each edge) | Helps distinguish edge, face, or hole origin |
| 4 | Fragmentation count, 50×50 mm at 3 locations | Confirms whether glass met ASTM C1048 |
| 5 | Furnace run number / date / shift / operator | Traceability for recipe audit |
| 6 | Glass thickness, size, type (FT/HS), coating | Recipe and standard reference |
| 7 | Installation conditions (frame type, edge clearance, setting blocks) | Separates manufacturing from installation causes |
| 8 | Time from installation to breakage | NiS failures typically occur weeks to months post-install |
ANSI Z97.1 Particle Weight Test
This test is now available as a free standalone tool, no Pro subscription required.
The full ANSI Z97.1 Class A calculator, procedure guide, pass/fail verdict, SGCC comparison, and safety glazing location reference, is available to everyone at no cost.
Open ANSI Z97.1 Fragment Test →ANSI Z97.1 (Safety Glazing Materials Used in Buildings) is the primary safety glazing standard in North America. For Class A (fully tempered glass), it requires a particle weight test: you break a sample, identify the 10 largest fragments by visual inspection, place all 10 on a scale together, and record the combined weight. That single number is compared to the standard's limit.
This is fundamentally different from ASTM C1048, which counts how many particles fit in a 50 × 50 mm square. The ANSI test measures fragment mass, heavier fragments mean larger, more dangerous pieces regardless of how many there are. Both tests serve different purposes and are often required together.
The limit scales with sample area. The standard reference limit (4.0 g) applies to the standard 12 × 12 in (305 × 305 mm) test specimen. If your sample is larger or smaller, the limit adjusts proportionally, the calculator below handles this automatically.
Where to collect fragments from, impact location & collection area
This is one of the most commonly misunderstood parts of the test. The collection area and impact point are defined by the standard, getting these wrong invalidates the result.
Strike the glass at a point approximately 1 inch (25 mm) from the geometric centre of the sample, on the surface. Use a pointed steel centre punch or a sharp hardened point. Do not strike at the edge or corner.
Collect fragments from the entire broken sample, all fragments within the original glass boundary (including edge fragments). Do not limit collection to just the area around the impact point. Lay the sample on paper or plastic film before breaking, trace the outline, and collect all pieces that remain within the traced perimeter.
- → Collecting only fragments near the impact crater, you must collect from the whole lite
- → Striking at an edge or corner, this initiates a different stress pattern and produces larger fragments near the break origin
- → Touching or disturbing fragments before waiting for all cracking to complete (wait minimum 5 minutes)
- → Using a sample cut from the edge zone of a lite, the standard requires a representative full-thickness sample with seamed edges
Test procedure, step by step
PPE required: safety glasses and cut-resistant gloves. Perform on a stable flat surface.
ANSI Z97.1 Class A Calculator
Enter your sample dimensions and combined fragment weight. The limit is automatically scaled to your sample area, 4.0 g per 929 cm² (12 × 12 in).
ANSI Z97.1 vs ASTM C1048, what each test proves
These standards are complementary. One failing does not mean the other will, both must be tested independently.
| Aspect | ANSI Z97.1, Particle Weight | ASTM C1048, Particle Count |
|---|---|---|
| What it measures | Combined weight of 10 largest fragments from whole sample | Number of particles in any 50 × 50 mm square |
| Pass criterion | ≤ 4.0 g (scaled to sample area) | ≥ 40 particles per 50 × 50 mm |
| Impact point | ~25 mm from geometric centre of sample | Near centre of lite |
| Collection area | All fragments within entire sample boundary | Count within a 50 × 50 mm square only |
| Primary purpose | Safety glazing certification, injury risk from large shards | Manufacturing QC, confirms adequate degree of temper |
| Required for | Safety glazing locations: doors, showers, overhead, railings | All projects specifying tempered glass to ASTM C1048 |
| Can one replace the other? | No : different purpose. A sample can pass ASTM count but fail ANSI weight (few large fragments), or vice versa. Perform both where safety glazing compliance is needed. | |
SGCC certification & CPSC 16 CFR 1201, how they relate to ANSI Z97.1
SGCC (Safety Glazing Certification Council) is a third-party product certification body, not a test standard. SGCC-listed products are tested to both ANSI Z97.1 and CPSC 16 CFR 1201. These are two separate standards with completely different test methods and sample sizes, they must not be confused.
The 34 × 76 in (864 × 1930 mm) sample commonly associated with SGCC and 16 CFR 1201 is for the impactor (pendulum bag drop) test, not for the ANSI Z97.1 particle weight test. You should not use this calculator with a 34 × 76 in sample and expect a 16 CFR 1201 result. The two tests are fundamentally different and cannot substitute for each other.
| Attribute | ANSI Z97.1 Class A (particle weight test) |
CPSC 16 CFR 1201 (impact test, SGCC requires this too) |
|---|---|---|
| Test type | Particle weight, weigh the 10 largest fragments after breaking | Impact, swing a 100 lb (45 kg) leather bag impactor into the glass |
| Sample size | Standard: 12 × 12 in (305 × 305 mm) Limit scales proportionally for other sizes |
Cat. I: min 34 × 76 in (864 × 1930 mm) Cat. II: min 34 × 76 in Fixed minimum, not scaled |
| Pass criterion | Combined weight of 10 largest fragments ≤ 4.0 g (for 929 cm² sample) | No single fragment may penetrate the test frame after impact; sharp edges & fragment size evaluated |
| Who requires it | IBC / IRC safety glazing code, can be satisfied by ANSI Z97.1 or 16 CFR 1201 | CPSC federal requirement for architectural glazing sold in the US; SGCC listing requires passing this |
| Categories | Class A (FT only), Class B (impact) | Category I (lower energy, doors ≤ 9 ft²), Category II (higher energy, all other hazardous locations & larger doors) |
- → SGCC certification requires passing both the ANSI Z97.1 particle weight test and the CPSC 16 CFR 1201 impact test.
- → The particle weight calculator on this page applies only to the ANSI Z97.1 portion of the SGCC test. It tells you nothing about 16 CFR 1201 impact performance.
- → If a 34 × 76 in sample is being cut, it is almost certainly for the 16 CFR 1201 impact test, not for weighing fragments.
- → For official SGCC product listing, glass must be submitted to an SGCC-accredited laboratory. This calculator supports in-house QC verification, not third-party certification.
Where is ANSI Z97.1 safety glazing required?
Under IBC, IRC, and most Canadian building codes. Always confirm with the applicable local code edition and the Authority Having Jurisdiction (AHJ).
- All glass in swinging, sliding, bifold, and storm doors
- Sidelites within 24 in of a door and ≤ 60 in from floor
- Shower and bathtub enclosures, all glass
- Glazing within 18 in of floor (stairways, ramps, landings)
- Overhead glazing, skylights and sloped glazing
- Gymnasiums, arenas, schools, and assembly areas
- Glass railings, guards, and balustrades
- Glazing adjacent to pools, hot tubs, and spas
- Fixed windows > 18 in above floor, away from doors
- Decorative glass not in a hazardous location
- Spandrel glass in non-accessible locations
- Some commercial curtain wall (verify with AHJ)