Higher Transmittance, Worse Glare: Choosing a Linear Diffuser That Passes UGR

Extruded lighting profile sections in clear, opal and amber material
Clear, opal and amber sections side by side. The surface and material behind the emitter set both the transmission and the luminance you are trying to control.

Abstract: Linear luminaires keep getting narrower and brighter, and the diffuser’s job has shifted from “let light through” to “control where light goes”. Chasing a higher transmittance number can push UGR above the limits in EN 12464-1. This article compares prismatic and opal diffusion, shows how to choose between PC and PMMA by failure mode rather than by datasheet, and lists the four measurements to run before you approve a sample.

Start with the optical target

Office lighting specifications built on EN 12464-1 typically ask for 500 lx on the working plane, UGR ≤ 19 and Ra ≥ 80 for open-plan areas. The number that trips projects up is UGR, and the reason is simple: UGR is calculated from luminance in the observer’s field of view, not from total lumens.

Specifying a clearer diffuser to make the lumens-per-watt figure look better raises surface luminance and can push UGR over the limit. The usual field fix is to add film or a louvre late in the project, which costs several points of efficacy and a week or two of schedule.

Decide the photometric outcome first, and record transmittance as a process parameter rather than as the acceptance criterion.

Decision 1: Choose the light-control method

Three options, each with a different trade-off:

  • Opal / frosted diffusion. The most mature process, the most uniform appearance, and the best at hiding LED hotspots. The cost is scattering: light bounces inside the cover and system efficacy takes a double-digit percentage hit.
  • Micro-prismatic (prismatic) diffusion. Surface micro-structure redistributes light and pushes high-angle output down. At comparable system efficacy it can lower UGR by several points. The exact amount depends on luminaire geometry and distribution, so measure it rather than copying it from a sample chart.
  • Lens arrays. The most precise control, suited to narrow beams and accent lighting; more expensive, and sensitive to LED pitch and placement accuracy.

For offices and education, prismatic diffusion with a medium transmittance is usually the better starting point: get glare inside the limit first, then recover lumens with LED count and drive current.

Decision 2: Pick PC or PMMA by failure mode

Datasheets all look good. What actually decides service life is how the material fails.

PMMA has the highest transmittance (above 90% at 3 mm in practice), the best weathering and yellowing resistance, and a harder surface that resists scuffing. It is brittle, has low notched impact strength, and its heat resistance sits around 80–90 °C. It is also prone to stress cracking when it meets certain adhesives and cleaning agents.

PC has impact strength roughly an order of magnitude higher and heat resistance around 120–130 °C, and it reaches UL94 V-0 without much trouble. It suits outdoor, semi-outdoor and impact-prone applications. Its weakness is yellowing under long-term UV, and yellowing is not cosmetic: it shifts the apparent colour temperature, reduces transmittance and leaves the surface easier to scratch.

Rule of thumb:

  • Indoor, long-term colour stability matters → PMMA.
  • Outdoor, high temperature, impact risk, or a V-0 requirement → PC, and specify a UV-stabilised grade, ideally with a co-extruded UV layer. A post-applied coating is a weak substitute; it chalks and peels.

Decision 3: Write optical consistency into the tolerance box

A single good sample does not prove a good batch. Small drifts in melt temperature, calibration vacuum and haul-off speed change wall thickness and how completely the prism fills, which changes transmittance and distribution in ways nobody can see by eye until twenty fixtures are installed side by side.

Beyond geometry, your drawing should state:

  • a transmittance window per batch (not “≥ some value”), plus a haze range;
  • prism pitch and apex angle tolerances;
  • a ΔE limit for within-batch and batch-to-batch colour;
  • a yellowness index threshold after accelerated weathering (for example, maximum ΔYI after a defined QUV exposure).

Without those lines the supplier can only deliver something that looks about right, and you can only accept it on the same basis.

Reading a UGR limit back to a section

A UGR figure in a tender is a room-level result, not a product property. It depends on the luminaire, the room dimensions, the reflectances and where the observer stands. That is why a section cannot simply be “UGR 19 approved” — but it can be chosen so the fixture has a realistic chance of passing.

What the tender says What it means for the section What to check on the sample
UGR ≤ 19 in an office at 0.75 m working plane Emitting surface luminance has to come down, which means diffusion or a wider distribution Luminance of the lit face at the specified viewing angles, not just total output
UGR ≤ 22 in an industrial hall Less demanding; a prism section with moderate texture usually suffices Whether the emitter image is visible at the mounting height actually used
“Low glare” with no number Not a specification. Ask for the number and the standard before choosing. Nothing — get the number first
UGR stated but mounting height changes on site The result moves with geometry; the section may need to change with it Luminance at the worst-case angle, not the catalogue angle

The transmission trade you are making

Lower luminance comes from spreading the same light over a larger apparent area. That is diffusion, and diffusion costs transmission. The honest way to present the decision is a pair of numbers on the drawing: a minimum transmission at the specified thickness, and a maximum luminance at the specified viewing angle. With both written down, a section can be accepted or rejected on measurement. With only one, the project usually trades away the other without noticing.

The range we work across is 0–97% transmission depending on material and surface, so there is usually a section that satisfies both limits. The failure mode is specifying the top of the transmission range and the bottom of the luminance range at the same time.

Four checks before you approve a sample

  1. Measure the system, not the material. Integrating sphere for lumens and lm/W; goniophotometer or simulation for UGR; surface luminance at the specified viewing angles.
  2. Weather it. QUV or elevated-temperature ageing, then re-measure transmittance and colour shift.
  3. Assemble twenty pieces. Real fixtures, real end caps and clips. Look for light leaks, bow and cap gaps.
  4. Check batch spread, not the average. Three consecutive batches from the same tool, five parts each, measure transmittance and chromaticity. Read the spread, not the mean.

What to put on the drawing

Put the photometric target on the drawing alongside the transmittance number: maintained illuminance, UGR limit, Ra, and how each one gets measured. Efficacy has risen to the point where most linear fixtures reach 500 lx without difficulty, so the requirement that still fails projects is the glare limit.

A useful test when you interview a supplier is to ask how a change in cover material or prism geometry moves UGR. A supplier who can answer that from measurement data is easier to work with than one who only quotes a transmittance figure.

Related guides and products

Scroll to Top