A scratch-dig grade is a two-number cosmetic designation: the first number grades scratch visibility, the second gives dig diameter in hundredths of a millimetre. A 60-40 surface is acceptable for general work, while 40-20 is the common general-purpose precision grade. Always pair the grade with a named inspection method or a scattering metric, otherwise the requirement cannot be enforced.
TL;DR:
- Visual comparison depends on the inspector, while ISO 10110 Part 7 supports dimensional or visibility tolerancing; name the standard, edition, and inspection method in drawings.
- RFQs should define the clear aperture, inspection lighting, sampling plan, coating inspection stage, and accumulation limits for scratch length and dig count or total diameter.
- One #50 dig raised scattering across a 20 mm aperture about 2–3 times in testing; scratches scatter most strongly perpendicular to their axes.
- A cosmetic grade does not establish fracture strength; parts under load, pressure, or thermal shock need separate controls, such as edge finishing or proof testing.
Table of Contents
- What do the two numbers in a scratch-dig grade mean?
- How do MIL-PRF-13830B, ISO 10110-7 and ANSI/OEOSC OP1.002 differ?
- How is scratch-dig inspected in practice?
- How do scratches and digs affect optical performance?
- How do you specify scratch-dig in drawings and RFQs?
- How should you combine standards and test data?
- How do scratches and digs affect glass strength and durability?
- What do defense, aerospace and medical sectors typically require?
- What causes scratches and digs, and how can you prevent them?
- What advanced detection technologies improve on visual inspection?
- Our view on scratch-dig and inspection evidence
- Precision Glasses: specification, inspection and supply in one place
- FAQ
- Sources
What do the two numbers in a scratch-dig grade mean?
The first number is a visibility grade for scratches, judged by comparison with reference artefacts. It is not a measurement of the scratch you see. The second number is the maximum dig diameter in hundredths of a millimetre. A dig is a pit or crater, and a scratch is a marking much longer than it is wide. This distinction matters because a comparative grade based on visual standards is often misread as a physical dimension, which leads to over-specification and cost without a matching performance gain.
Common industry practice, as set out in a scratch-dig tutorial from the University of Arizona, maps the numbers as follows:
- Scratch #80, #60, #40: implied maximum widths of 8 µm, 6 µm and 4 µm respectively.
- Dig 50 and dig 20: maximum diameters of 0.50 mm and 0.20 mm respectively.
- Scratch accumulation: the combined length of maximum-width scratches is limited to a fraction of the surface diameter in the tutorial’s circular-element example.
- Dig accumulation: the number and summed diameter of maximum-size digs are limited per observation area, and MIL-PRF-13830B sets out these rules in full.
Overlooking the accumulation rules is a frequent cause of rejected parts, so read them before you read the headline grade.
How do MIL-PRF-13830B, ISO 10110-7 and ANSI/OEOSC OP1.002 differ?
The three documents answer the same question in different ways, and the choice decides how objective your acceptance test can be.
- MIL-PRF-13830B: the long-standing comparative method. An inspector judges the surface against standard artefacts, which makes it quick but dependent on the observer. NIST and related programmes have worked to correlate scatter measurements with trained observers because the scratch standards themselves lack objective reproducibility.
- ISO 10110-7:2017: offers a dimensional method based on defect area and a visibility method based on appearance, with standard drawing notation such as 5/N×A. The standard supports objective, traceable inspection, often implemented through automated imaging.
- ANSI/OEOSC OP1.002-2017: provides a numerical notation for visibility and an alphabetic notation for size, and places the onus on the optical engineer to choose. It also prescribes standard inspection practices to reduce misinterpretation. The standard states that it does not itself address performance impacts, as the OEOSC listing notes.
If you need objectivity or automation, use ISO notation or state the chosen inspection method outright. For a deeper look at MIL-style wording, see our guide to MIL-PRF-13830B clause details and RFQ phrasing.
Pro Tip: Write the standard and the notation into the drawing note, never just “60-40”, so a supplier and an inspector read the same requirement.
How is scratch-dig inspected in practice?
Traditional inspection is visual. An inspector holds the part under controlled lighting, such as a 40 W incandescent source or a comparable lamp, and compares what they see with calibrated artefacts. The result depends on the observer’s eyesight, experience and viewing angle, which is why two inspectors can disagree on the same surface.
Automated systems image the surface and report defect area, counts and positions. That gives repeatability and a record you can audit, and it fits naturally with the dimensional method in ISO 10110-7. Industry commentary points to a strong shift towards objective, automated inspection for this reason. Detection thresholds depend on the optics and resolution of the system, so ask any supplier what defect size their equipment resolves.
When you request an inspection, ask for these deliverables:
- Images of the defects found, with their position on the surface.
- An automated report or measurement data file, not only a pass or fail.
- A statement of inspection conditions: lighting type, viewing distance and method.
- The standard and edition against which the part was judged.
Where the inspection depends on a person, the evidence should name the person’s qualification or training. Short courses such as SC01 Understanding Scratch Dig exist precisely because these specifications are so often misunderstood by engineers and inspectors.
How do scratches and digs affect optical performance?
Scratches and digs scatter light, and they do it differently. A scratch scatters directionally, with the peak perpendicular to the scratch axis. A dig scatters more uniformly. That difference decides whether a given grade matters in your system, since stray light in an imaging path behaves very differently from a loss in a beam delivery path.
Laboratory BTDF measurements quantify the effect. According to a measured scattering study, a single #50 dig can raise average scattering over a 20 mm aperture by roughly 2–3×, while scratches produce strongly directional increases, with #80 or #60 scratches raising perpendicular-direction scattering by factors of tens in the measured examples.
A single #50 dig can raise average scattering over a 20 mm aperture by roughly 2–3×, according to measured BTDF data. The cosmetic number does not tell you this; only a scatter measurement does.
The practical conclusion is simple. In applications sensitive to stray light, require a scattering, total integrated scatter or microroughness metric alongside the scratch-dig limit, rather than relying on the cosmetic grade alone. Photonics industry coverage reiterates that 40-20 is widely treated as a standard precision level, with tighter grades used where stray light, laser damage risk or imaging contrast demand it.
How do you specify scratch-dig in drawings and RFQs?
Start from the drawing notation. ISO 10110-7 lets you state whether dimensional or visibility tolerancing applies, so say which one. Our guide to decoding ISO 10110 drawing notation shows how those entries read on a drawing.
Define the clear aperture explicitly, because a grade applied to the wrong zone either wastes money or misses the area that matters. Coating defects count against scratch-dig, so state whether the grade applies before or after coating. You can tighten acceptance around the functional zone and relax it toward the edge.
Include these points in every RFQ:
- Standard and edition (for example ISO 10110-7:2017) and the notation used.
- Clear aperture diameter and any zone with tighter limits.
- Inspection method and lighting conditions.
- Sampling plan: every part, or a defined sample.
- Evidence required: images, automated data file, or both.
- Accumulation limits for scratch length and dig count or total diameter.
Pro Tip: State whether coating defects are counted, and at which stage the inspection takes place.
For ready-made tolerance wording, see our RFQ-ready optical window tolerances.
How should you combine standards and test data?
Treat the standards and the measurement data as two halves of one requirement. ISO 10110-7 gives you routes to objective, traceable inspection, while MIL-PRF-13830B remains comparative. The BTDF findings show why the cosmetic grade alone cannot predict stray light: dig and scratch effects differ in size and direction, and the measured increases can be large.
The rule for sensitive systems is to require both a cosmetic scratch-dig limit and a scattering, total integrated scatter or microroughness threshold. For less sensitive applications, a clearly stated grade with a named inspection method is enough.
How do scratches and digs affect glass strength and durability?
Surface flaws are where glass fails. Brittle materials break from the tip of the worst flaw under tension, so a scratch or dig acts as a stress concentrator. The scratch-dig grade is a cosmetic and optical designation and does not give you a strength value. A part can meet 40-20 and still carry a flaw that governs its fracture load.
This matters most for parts under mechanical load, thermal shock or pressure, such as windows in housings, protective covers and sensor ports. Orientation matters as well. A scratch running across the direction of tensile stress is more damaging than one running along it.
For these parts, specify strength-relevant controls alongside the cosmetic grade: edge finish, strengthening process and handling rules. Chemical strengthening puts the surface in compression, which raises the threshold at which a flaw propagates, and our guide to chemically strengthened glass explains how that works for engineers. If your part is exposed to harsh service, our note on how to optimise glass durability covers the wider picture. Where strength is critical, request a proof test or a defined fracture-strength check rather than inferring strength from cosmetic quality.
What do defense, aerospace and medical sectors typically require?
Each sector leans on the standards differently, and the contract usually decides which one binds. Defense work has long referenced MIL-PRF-13830B, so drawings often carry a scratch-dig grade in that notation, together with the inspection and accumulation rules in the specification. Where a programme moves to objective acceptance, ISO 10110-7 notation or ANSI/OEOSC OP1.002 is used instead.
Aerospace and instrument programmes tend to add traceability and documented inspection evidence. In practice that means batch records, inspection images and a stated standard edition on the certificate. Where a sensor or laser path is involved, stray light or damage risk usually pushes the requirement from 60-40 towards 40-20 or tighter, and a scattering limit may be added.
Medical device programmes add quality system expectations, so inspection method, sampling and record retention matter as much as the grade itself. The practical advice is the same across all three: read the contract flow-down, name the standard and edition, and write the inspection method into the requirement. We do not assume a grade from the sector alone, because the right limit follows the function of the surface. Our optical components work relates to these choices.
What causes scratches and digs, and how can you prevent them?
Most defects come from three places: the polishing line, the handling chain and the cleaning step. The scratch types named in the tutorial show the pattern. A block reek is a chain-like scratch from polishing, a runner-cut is a curved scratch from grinding, a sleek is a hairline scratch, and a crush or rub is a series of small scratches from mishandling.
Prevention follows the cause:
- Grinding and polishing: control abrasive size, filter slurries and clean tooling between stages, so a stray coarse particle does not drag across a finished surface.
- Edge finishing: chamfer and finish edges before final polish, since chips at the edge shed particles that later scratch the face.
- Handling: use gloves or finger cots, separate parts with interleaf material and avoid stacking glass on glass.
- Cleaning: wipe only with suitable materials, never dry-wipe a dusty surface, and clean in a controlled environment.
- Packaging: individual cavities or protective film prevent transit damage.
When defects appear in a batch, trace them to the process step before tightening the grade, because a stricter limit does not fix a process fault. Our troubleshooting guide for glass defects covers that diagnostic route, and our glass machining overview explains where each process step can introduce damage. Equipment tolerance also plays a role in how consistently a line holds quality, as this partner article on selecting equipment tolerances discusses.
What advanced detection technologies improve on visual inspection?
Visual inspection answers one question: does this surface look as good as the artefact? Advanced systems answer a different one: what defects exist, how large are they, and where are they? Automated optical systems capture high-resolution images of the surface under controlled illumination and use software to find, size and count defects. Dark-field illumination makes scratches and digs bright against a dark background, which helps resolve fine defects that an eye would miss or judge inconsistently.
The advantages over traditional inspection are practical. Results are repeatable between operators and shifts. Each part carries a stored image record, which supports traceability and dispute resolution. Defect area and count come out as numbers, which maps directly onto the dimensional method in ISO 10110-7.
Scatter-based measurement goes a step further by linking surface condition to light behaviour, which is the quantity that matters when stray light drives the requirement. That is the reason NIST-type programmes try to correlate objective scatter data with trained observers.
Automated inspection does not replace judgement. You still choose the standard, the limits and the zones. But it turns an opinion into a record, and it lets you write an acceptance criterion that both sides can verify.
Our view on scratch-dig and inspection evidence
We provide specification advice and review RFQ wording before parts go into production. The same omissions turn up repeatedly: no illumination conditions, an unclear clear aperture, no statement of whether coating defects count, and a grade with no named inspection method. Each is a one-line fix, and each prevents a rejected batch. If you want a second pair of eyes on your tolerancing or an inspection scope for a specific part, ask us for a review.
— Alexandra
Precision Glasses: specification, inspection and supply in one place
We support you from tolerancing through to delivery. Our team fabricates optical components and can review your RFQ wording against the points above. Explore our optical components and services, then send us your drawing for a tolerancing review or a quote that includes inspection deliverables.
FAQ
How do you read a 60-40 scratch-dig grade?
The 60 is a scratch visibility grade, commonly mapped to an implied maximum width of 6 µm, and the 40 is a maximum dig diameter of 0.40 mm. The scratch grade is a comparison with reference artefacts, not a direct measurement. Accumulation rules limit the total scratch length and the number of maximum-size digs.
Are coating defects counted against scratch-dig?
Yes, coating defects are normally counted, so state whether the grade applies before or after coating. Naming the inspection stage in the RFQ avoids disputes. Our guide to optical coatings explains where coating defects arise.
When should you require scattering measurements as well as scratch-dig?
Require them when stray light, imaging contrast or laser damage risk drives the application. A measured study shows that a single #50 dig can raise average scattering by roughly 2–3×, so the cosmetic grade alone does not predict optical behaviour. For less sensitive parts, a stated grade and inspection method is enough.
What sampling plan is acceptable for scratch-dig inspection?
It depends on the criticality of the part. Inspect every part for critical or high-value optics, and use a defined sample for large volumes of lower-risk parts. State the sampling plan in the RFQ so supplier and buyer agree before production.
Sources
- ISO 10110-7:2017 – Optics and photonics — Preparation of drawings for optical elements and systems — Part 7: Surface imperfections
- ANSI/OEOSC OP1.002-2017: Optics and Electro-Optical Instruments – Optical Elements and Assemblies – Surface Imperfections – The ANSI Blog
- Osti
- SPECIFICATION STANDARDS
