Quick answer: Smoke is measured by three unrelated methods, and passing one tells you nothing about the others. Cable uses IEC 61034 — a 3 m³ chamber, pass is ≥60% light transmittance. Building materials use the smoke index inside ASTM E84, capped at 450 for Class A. Rail uses ISO 5659-2 with Dsmax limits and a separate toxicity number. A compound optimised for one can fail another outright.
Smoke is also the clause that most often fails an otherwise good formulation. Flame ratings are what everyone specifies first; smoke is what quietly decides whether a halogenated system can be used at all.
Three Methods, Three Different Questions
| Standard | What It Measures | Pass Criterion |
|---|---|---|
| IEC 61034 (EN 50268) — cable | Cable burned inside a 3 m × 3 m × 3 m chamber, light beam across it | ≥ 60% light transmittance to be called low smoke |
| ASTM E84 building products | Smoke Developed Index, measured in the same Steiner tunnel run as flame spread | SDI ≤ 450 for Class A — alongside FSI 0–25 |
| ASTM E662 (NBS chamber) | Specific optical density of smoke from a solid specimen | Limit set by the product standard citing it, not by E662 itself |
| ISO 5659-2 rail, via EN 45545-2 | Dsmax at 25 kW/m², plus a separate toxicity index | Tiered by hazard level — see the table below |
The important consequence: these are not convertible. A cable compound that passes IEC 61034 comfortably has told you nothing about how it would score in a Steiner tunnel, because one burns a cable in a room-sized box and the other burns a lining along a 25-foot ceiling. Anyone offering you a smoke figure without naming the method has given you a number without a unit.
Rail Is the Strictest, and It Adds Toxicity
EN 45545-2 is where smoke requirements get genuinely hard, because it grades by hazard level and pairs every smoke limit with a toxicity limit. These are the R22 and R23 figures:
| Requirement | HL1 | HL2 | HL3 |
|---|---|---|---|
| R22 — Dsmax | < 600 | < 300 | < 150 |
| R22 — toxicity (CITNLP) | < 1.2 | < 0.9 | < 0.75 |
| R23 — Dsmax | — | < 600 | < 300 |
| R23 — toxicity | — | < 1.8 | < 1.5 |
Rail also specifies LOI above 32 at HL3, which is one of the few places a limiting oxygen index figure is genuinely contractual rather than a development metric.
Why Halogenated Systems Struggle Here
This is the mechanism behind most smoke failures, and it is worth understanding rather than memorising.
Halogenated flame retardants work in the gas phase — they interrupt the combustion reaction. Interrupting combustion means incomplete combustion, and incomplete combustion is exactly what produces soot. The property that makes the system efficient on flame rating is the same property that produces smoke. You cannot tune one without moving the other.
Mineral hydroxides do the opposite. ATH and MDH absorb heat and release water vapour, diluting the smoke rather than adding to it, and leaving an inert oxide residue. That is the entire reason LSZH exists as a material class — not policy, but the physics of what each mechanism leaves behind.
We have seen halogenated panels post an E84 flame spread index of 15 — comfortably inside Class A — and then lose the rating on the smoke index alone. If your specification has both numbers, screen smoke early rather than after booking tunnel time.
What Actually Reduces Smoke
| Route | Effect on Smoke | Cost |
|---|---|---|
| Switch to mineral hydroxide system | Largest single improvement. Water vapour instead of soot | 55–65% loading, mechanical loss, reformulation |
| Add zinc borate | Promotes char, suppresses smoke, and cuts antimony at the same time | 2–10 phr. Cheapest meaningful step |
| Add zinc stannate | Strong smoke suppression at low dose; removes antimony entirely | 1–5 phr, higher price per kilo |
| Reduce antimony trioxide | Antimony contributes to smoke; cutting it usually helps | Requires re-testing the flame rating |
What does not work: adding a smoke suppressant on top of a heavily halogenated system and hoping to keep both numbers. If your smoke limit is tight and the flame package is gas-phase, you are usually changing the package rather than adjusting it.
What We Would Actually Advise
Find out which method your clause names, before anything else. “Low smoke” without a standard reference is marketing. IEC 61034, E84’s smoke index and ISO 5659-2 are different tests with different pass criteria, and a supplier’s figure from one is not evidence for another.
If your specification has both flame and smoke clauses, screen smoke first. It is the harder of the two to fix, and finding out late costs a full reformulation rather than a dosage adjustment.
If you are already halogenated and the smoke number is close, try zinc borate before rebuilding the formulation. It is the cheapest step that moves smoke, and it reduces antimony as a side effect. Full mineral conversion is the answer when the gap is large, not when it is marginal.
And where we would tell you not to bother: if your product has no smoke clause and no plan to enter a market that has one, optimising smoke is spending money on a number nobody will ask you for. Plenty of industrial and outdoor cable is perfectly well served by PVC.
FAQ
What is the pass criterion for IEC 61034?
Minimum 60% light transmittance measured across a 3 m × 3 m × 3 m chamber while the cable sample burns. Below that, the cable does not qualify as low smoke under this standard. It is a cable test — run on finished cable, not on compound — so the certificate belongs to the cable maker rather than to the material supplier.
Is low smoke the same as halogen-free?
No, though they usually travel together. Low smoke is a measured performance result; halogen-free is a composition limit. Low-smoke PVC exists and is still halogenated, which is why cable specifications name both requirements separately when they want both — see LSZH vs PVC.
Why does my compound pass the flame test but fail on smoke?
Most likely because the flame package works in the gas phase. Interrupting combustion produces incomplete combustion, and incomplete combustion produces soot — the efficiency and the smoke come from the same mechanism. Reducing antimony or adding a zinc-based smoke suppressant is the first step; switching to a mineral system is the reliable one.
What is Dsmax?
Maximum specific optical density of smoke, measured to ISO 5659-2. EN 45545-2 sets it by hazard level: under R22 the limits are 600 for HL1, 300 for HL2 and 150 for HL3, each paired with a separate toxicity limit. Rail is the only common application that regulates smoke toxicity as a distinct number.
Do smoke suppressants affect the flame rating?
They can, in both directions. Zinc borate and zinc stannate promote char, which often helps the flame rating as well as smoke. But any change to a synergist balance is a formulation change — re-run UL94 or the relevant cable test rather than assuming the rating carries over.
Can you supply material that meets a specific smoke standard?
We supply the flame retardant layer and can propose a system aimed at your smoke clause, with pre-screening arranged before you commit. What we cannot do is certify your finished product — IEC 61034 and EN 45545 are tested on finished cable or components, and that approval stays with you.
Final Thoughts
Smoke is where most flame retardant projects actually get decided, and it is the clause people check last. Name the standard first, screen it early, and treat the flame rating and the smoke number as one problem rather than two — because in a gas-phase system they are physically linked. Send us your smoke clause with the flame requirement and the base polymer, and we will say which route reaches both, including when the honest answer is that your current package cannot. Start from LSZH cable systems for cable work, or the chemistry overview if you are still choosing a mechanism.
—— Rectivas Materials 团队 技术工程师老陈