Why Glass Beads Decide Whether a Road Marking Works at Night

A road marking has two jobs. In daylight it has to be seen. At night it has to return light to the driver who sent it. The second job is the hard one, and it is not done by the paint. It is done by glass beads.

This is not a supplier talking up its own product. It is the conclusion of a peer-reviewed review published in the Journal of Guangdong University of Technology (2021, Vol. 38), funded in part by China’s National Natural Science Foundation. The paper surveyed marking materials, additives and standards and reached a plain finding: glass beads directly determine the retroreflective performance of a road marking, and their effect is stronger than that of titanium dioxide.

Start with the stakes, not the product

The same review sets out why retroreflectivity is treated as a safety-critical property rather than a cosmetic one:

Over 40% of traffic accident deaths in China occur at night, according to the sources cited in the paper.
In the United States, highway accident casualty rates between 19:00 and 07:00 from 2010 to 2017 ran at 1.5 times the rate during other hours.
The paper cites an authoritative statistic putting the correlation between marking retroreflectivity and traffic accidents at 70%.
For every 10 to 100 mcd improvement in a marking’s retroreflective luminance coefficient, traffic risk falls by 0.9% to 8.6%.
Better marking retroreflectivity is credited with improving transport efficiency by more than 30%.

Those figures come from a literature review quoting earlier studies, so they should be read as research estimates rather than regulatory thresholds. But the direction is unambiguous, and it explains why retroreflectivity is one of the first things checked — and one of the most common reasons a striping job fails acceptance.

The mechanism: refraction, reflection, refraction

A glass bead is a tiny lens with a mirror behind it. Light from a headlamp enters the sphere, bends (refracts) as it passes through the curved front surface, travels to the back of the bead, reflects off the reflective layer formed at the bead’s base, and bends again on the way out — travelling back along roughly the same path it arrived on. This is retroreflection, sometimes called regression reflection.

Two properties make it work, and both are non-negotiable:

Transparency. The bead must let light through. An opaque or cloudy bead blocks part of the incoming light before it can ever be returned.
Roundness. Only a near-perfect sphere bends the ray cleanly to the back of the bead and back out to the driver’s eye. A badly formed bead scatters light in irregular directions inside itself, so the ray leaves pointing somewhere other than the driver’s line of sight.

This is why the standards write roundness requirements into law. Under GB/T 24722-2020 (China’s national standard for glass beads used in pavement markings), low-refractive-index beads of Type 1 and Type 2 must have a roundness of not less than 80% — and for Type 1 beads specifically, the fraction falling in the 850 µm to 600 µm size band must reach at least 70% roundness. For medium- and high-refractive-index beads, the standard switches to a defect measure: no more than 20% defective beads by percentage.

Why beads beat titanium dioxide

Titanium dioxide is the other additive that gets discussed in marking formulations, and it does help — but by a different route. TiO₂ raises the whiteness of the marking, which raises its overall light reflectance. The review notes that brightness factor tracks TiO₂ content fairly linearly, and that light reflectance ratio rises clearly as the dosage increases.

Then it stops working. Once TiO₂ exceeds about 7% of the formulation, chromaticity difference, brightness factor and reflectance ratio all plateau. Over-bright markings also soil faster, which drags retroreflectivity down. Because TiO₂ is expensive, real-world marking coatings typically use only 3% to 5%.

Glass beads have no such ceiling — and the review states plainly that their effect on retroreflective performance is superior to that of titanium dioxide. If you want the marking to work at night, the beads are where the leverage is.

Not all beads are equal: the refractive index ladder

GB/T 24722-2020 sorts beads into three refractive index (RI) classes:

Class	Refractive index	What it means in practice
Low	1.50 ≤ RI < 1.70	The workhorse. Standard soda-lime beads, the majority of the market.
Medium	1.70 ≤ RI < 1.90	Coated beads typically land here. Better performance in wet conditions.
High	RI ≥ 1.90	All-weather and wet-night beads. The closer RI gets to 1.93, the better the regression reflection.

The physics behind the ladder: retroreflection coefficient is governed by the angle over which the returned beam spreads. A higher refractive index narrows that spread, concentrating the light so the driver’s eye receives more of it. In theory lower spread is better, but cost intervenes — the review notes that for pavement markings, RI ≥ 1.70 is generally sufficient, with optimum performance around RI = 1.93.

The trade-off is real. American practice, as documented by the Texas Department of Transportation, puts most roadway striping beads at RI 1.50, made from recycled windowpane glass. The 1.65 and 1.90 grades are made from virgin glass, cost more, and may be slightly less durable than 1.50 beads. Some agencies blend high and low index beads to balance cost against performance.

Where the standards point now

Two Chinese standards matter to anyone specifying or supplying beads today.

GB/T 24722-2020 governs the beads themselves: appearance (colourless, white or pale yellow; clean; transparent spherical bodies under a microscope with no obvious bubbles or inclusions), roundness, particle size distribution, density (2.4 to 4.6 g/cm³), water resistance, magnetic particle content (≤0.1%), and — new in this edition — limits on lead (≤200 mg/kg), arsenic (≤200 mg/kg) and antimony (≤200 mg/kg).

That last group is worth pausing on. Heavy metal limits are a 2020 addition, and they are exactly the kind of thing European and Middle Eastern buyers now ask for in documentation. A supplier who can produce those numbers without being chased has an advantage.

GB/T 16311-2024, effective 1 March 2025, governs the finished marking. For dry, non-rainy markings using Type I reflective marking, the initial retroreflective luminance coefficient must be at least 150 mcd·m⁻²·lx⁻¹ for white and 100 mcd·m⁻²·lx⁻¹ for yellow. Measurement must be taken between 48 hours and 30 days after striping to count as the “initial” value.

The 2024 revision also adds something suppliers should know about: it introduces testing of the finished, applied marking for internal bead content, total organic content and heavy metals. In other words, the bead quality has to survive into the laid line, not just into the sack that arrived at the port.

What this means if you are buying beads

Three practical consequences follow from all of the above.

First, ask for the RI class explicitly. “Glass beads” is not a specification. State the class (1.50, 1.70 or 1.90+) and the application (drop-on or premix), because the two carry different requirements and different sieve tables.

Second, match bead size to the marking, not to habit. A thin line cannot hold a coarse bead — there is no depth for it to embed into, so traffic knocks it out. A thick line buried under fine beads reflects poorly because the beads sit below the surface. The mesh number has to follow the application.

Third, treat documentation as part of the product. Roundness, sieve distribution, water resistance, magnetic particle content and heavy metals are all testable, and under GB/T 16311-2024 the finished line is testable too. A batch that arrives with a report attached saves everyone a dispute at the point of acceptance.

If you are working out which grade fits an upcoming job, send us the marking type, the application thickness and the destination country. We will match it against the relevant sieve table and tell you what we can hold on a repeat order — before you commit to a container.