Clear Boro Quality: Simax, Schott Duran, and Generic Import Clear Compared

Short answer: Nearly all clear borosilicate sold for flameworking is “glass 3.3” — nominally the same COE 33 family — so Simax (Kavalier), Schott Duran, Pyrex-family glass, and generic import clear normally weld to each other without fitting problems. What separates the name brands is not a different glass but published, verifiable quality control: Kavalier and Schott publish technical datasheets, conform to ISO 3585, and (in Schott’s case) publish explicit defect tolerances for airlines, stones, and geometry. Generic import clear is usually nominally the same 3.3 glass and many workers use it happily — the practical difference is batch-to-batch consistency and how much time you spend flame-polishing out haze, chasing bubbles, and grading rod before it ever becomes a piece.

Clear is the glass you use the most of and blame the least, so it pays to be pickiest about it. This guide covers what the standards actually guarantee, the defect vocabulary that matters at the bench, and how to evaluate a clear you’ve never used. (For where boro sits against soft glass and quartz in the first place, see soft glass vs boro vs quartz.)

What COE 33 actually guarantees — ISO 3585 in plain English

“Borosilicate 3.3” is a defined material, not a marketing phrase. ISO 3585 requires a mean linear thermal expansion coefficient of (3.3 ± 0.1) × 10⁻⁶ K⁻¹ measured from 20–300 °C, plus minimum chemical durability: acid resistance under 100 µg Na₂O per dm² (per ISO 1776) and alkali resistance of class ISO 695-A2 or better. Any glass legitimately sold as 3.3 borosilicate sits inside that expansion band.

That’s what the standard guarantees. Here’s what it doesn’t guarantee: freedom from bubbles and airlines, absence of striae or cord, dimensional consistency of rod and tubing, or how the surface behaves after twenty minutes in your flame. Two glasses can both pass ISO 3585 and still work very differently. That gap — between “compatible” and “pleasant to use” — is where the brand differences live.

The big names: Simax (Kavalier) and Schott Duran

Simax is made by Kavalier in Sázava, Czech Republic — a glassworks founded in 1837 by František Kavalier, focused from the beginning on pharmacy and laboratory glass. The Simax brand itself dates to the early 1950s, developed under glass technologist Dr. M. B. Volf, and Kavalier now exports to more than 90 countries. Per the official Kavalier TDS, Simax has a COE of (3.3 ± 0.1) × 10⁻⁶ K⁻¹ (20–300 °C, ISO 7991), density of 2.23 ± 0.02 g/cm³, and a composition of roughly 80.3% SiO₂, 13.0% B₂O₃, 2.4% Al₂O₃, and 4.3% Na₂O+K₂O — essentially the same family as Schott’s Duran 8330. It conforms to ISO 3585.

Schott Duran is borosilicate 3.3 meeting ISO 3585:1998 and ASTM E438 Type I. Schott states that Duran tubing and rod are produced to very tight geometric tolerances with high optical quality, and its datasheet defines the working behavior by standardized viscosity points — with a maximum permissible operating temperature around 500 °C and softening beginning above roughly 525 °C (these are service-temperature figures for finished ware, not flameworking temperatures).

The point isn’t that these two glasses are exotic — it’s that both makers publish what they’re selling. You can pull the TDS, see the expansion band, the composition, and the standard it’s certified to, and expect the next case to match the last one.

The defect vocabulary: airlines, knots, stones, striae

Drawn glass has a standard defect taxonomy, documented in tubing-industry technical literature (VitroCom’s “Glass defects made visible” paper is a good plain-language reference):

This vocabulary matters because it’s exactly what the top-tier makers quantify. Schott publishes explicit tubing defect specifications — airlines are limited by aggregate length per 10 m of tubing and by width, stones and knots are measured by core size — and for its pharmaceutical FIOLAX tubing Schott moved to 100% inspection with zero-defect specs for airlines of 0.08 mm width and up, cracks, and similar flaws. Flameworking rod isn’t held to pharma specs, but a manufacturer that measures and publishes defect tolerances on one line tends to run disciplined QC across the melt. Most generic import clear ships with no published defect spec at all — which doesn’t mean it’s bad, only that you’re the inspection department.

Scum and haze: glass problem or flame problem?

The most common complaint about “cheap clear” is scum — a grey or white haze that develops as you work. Before blaming the glass, know that haze has several distinct causes, and most of them are technique:

• Flame chemistry. A dirty or reducing flame deposits and reduces material on the surface. This is the first thing to rule out — see flame chemistry: neutral, oxidizing, reducing.
• Devitrification. Crystallization of the glass surface, presenting as a white/grey hazy, scummy, or chalky texture. It nucleates when glass is held too hot for too long, or when surface contamination provides nucleation sites.
• Surface contamination. Fingerprints, dust, marker ink, or shelf grime give devit somewhere to start.
• Alkali volatilization. Extended working drives alkali out of the surface, changing its chemistry and finish.

Because technique is a confounder in every one of these, a haze problem cannot be automatically pinned on the brand. The fair test is side by side: same torch, same flame setting, same working time, two rods. If one clear scums noticeably sooner under identical treatment, that’s real information about the glass. If both haze, look at your flame and your prep first. (Color-specific scumming and boiling are their own topic — see boro color troubleshooting.)

Generic import clear: what actually varies

The honest framing on import clear is this: claims that it’s universally dirtier or scummier are community lore, not manufacturer-verified data. Import clear is nominally COE 33, it welds to Simax and Duran, and plenty of production workers run it by the case without drama.

What you genuinely give up is the paper trail and the consistency it represents. Kavalier and Schott publish TDS documents, certify to ISO 3585, and control their melts batch to batch. A generic import may be excellent — and the next pallet may not match it. The costs of an inconsistent clear are quiet ones: time spent flame-polishing haze, rod culled for airlines and stones, a stone-seeded crack discovered after annealing. If your time is worth anything, “cheap clear” is only cheap when the batch is good — and you can’t know that until you’ve worked it. For hobby quantities the price gap is small; for production, the calculation depends on how much grading and rework a given batch demands.

Cross-brand compatibility: mixing the 3.3 family

Published expansion values for the major 33-expansion borosilicates cluster tightly: Pyrex 7740 ~32.5, Duran 8330 ~33.0, Kimax KG-33 ~32.0, Simax ~33.0 (all × 10⁻⁷/°C). All of these sit inside the ISO 3585 tolerance band, which is why cross-brand welds normally fit — the family is functionally equivalent for fusing and joining, and boro’s low expansion already makes it more forgiving of thermal shock than soft glass.

Two cautions. First, published values differ slightly by source and measurement convention, so don’t treat the deltas between brands as precise. Second, “inside the band” is not “identical”: in extreme cases — heavy-wall seals, large-diameter joins, thick sculptural masses — the small residual mismatch plus ordinary annealing stress can still show strain. If a cross-brand piece cracks, work through why did my glass crack before assuming incompatibility; annealing is the more common culprit.

Rod vs tubing: different jobs, different defect risks

Rod and tubing fail differently. In rod, defects live in the body of the glass: striae and cord show as grain in thick solid work, stones sit buried until they seed a crack, and airlines streak through what should be optically clean mass. In tubing, geometry joins the list — wall-thickness variation and out-of-round tubing make even heating and clean seals harder, and airlines in a thin wall can open into actual leaks or surface lines. This is why Schott’s tubing specs quantify both defects and dimensional tolerance. When you evaluate a new clear, evaluate the form you’ll actually buy: a maker’s rod and tubing can behave differently, and tubing is less forgiving of sloppy drawing.

How to evaluate a clear you haven’t used

A repeatable acceptance test costs one evening and a few rods:

1. Inspect cold. Sight down each rod or tube against a dark background in good light. Count airlines, seeds, and visible stones or knots. Check tubing for wall uniformity.
2. Melt test. Work a rod in a known-neutral flame for an extended period. Watch for early haze, boiling, or scum against a reference rod of your usual clear worked identically.
3. Weld test. Fuse the new clear to your standard clear, work the joint, and anneal the test piece properly (see annealing schedules for glass).
4. Polariscope check. A polariscope — a crossed-polarizer strain viewer — is the standard nondestructive way to reveal residual stress in glass. It does double duty here: it verifies your annealing and reveals strain concentrated at the test weld if the two clears aren’t as compatible as their labels claim. A clean weld under crossed polarizers is the closest thing to proof you can get at the studio level.

If a candidate clear passes all four against your incumbent, use it with confidence — whatever the label says.

Key takeaways

• All legitimate 3.3 borosilicate sits in the ISO 3585 band of (3.3 ± 0.1) × 10⁻⁶ K⁻¹, so Simax, Duran, Pyrex-family, and import clear normally weld together.
• The brand difference is published QC, not a different glass: Kavalier and Schott publish TDS documents and (Schott) explicit defect tolerances; most generic imports publish nothing.
• Know the defect vocabulary — airlines, seeds, knots, stones, striae — because it’s what you’re screening for when no one else has.
• Haze is not automatically the glass. Flame chemistry, devitrification, contamination, and alkali loss all cause it; test side by side before blaming a brand.
• Cheap clear is only cheap when the batch is good; inconsistency costs cleanup time, culled rod, and post-anneal surprises.
• Evaluate any new clear with a cold inspection, melt test, weld test, and polariscope check before committing to volume.

Sources

• Kavalier, “SIMAX glass — Technical Data Sheet (2025)” — https://www.kavalier.cz/wp-content/uploads/2025/07/TDS-SIMAX-glass-2025.pdf
• Kavalier, company history — https://kavalier.cz/en/company/history.html
• Schott, “DURAN — technical details” — https://www.schott.com/en-us/products/duran-p1000368/technical-details
• Schott, “DURAN tubing datasheet” — https://www.schott.com/en-gb/products/duran/-/media/Project/OnEx/Products/D/DURAN-glass-tubing/Downloads/datasheet/schott-tubing-datasheet-duran-english.pdf?rev=b0f5172192f04795a66f4492f5a4dab2
• VitroCom, “Learning from mistakes — glass defects made visible” — https://www.vitrocom.com/wp-content/uploads/2023/04/Glass-Defects-Made-Visible.pdf
• ISO 3585 — Borosilicate glass 3.3 — Properties — https://www.iso.org/standard/24774.html
• Wikipedia, “Borosilicate glass” — https://en.wikipedia.org/wiki/Borosilicate_glass
• Adams & Chittenden, material properties — https://adamschittenden.com/technical/material-properties

Editor’s note: published COE values vary slightly by source and measurement convention (roughly 32.0–33.0 × 10⁻⁷/°C across Kimax, Pyrex 7740, Duran, and Simax); all fall within the ISO 3585 tolerance. Temperature figures quoted from the Schott datasheet are service limits for finished ware, not working-flame temperatures. Observations about import clear reflect community experience, not manufacturer-verified data — always follow the manufacturer’s documentation for the specific glass you buy.