The UL File Says V-0 but the Moulded PBT Part Did Not Pass. This discrepancy appears repeatedly on injection moulding production floors when a polybutylene terephthalate compound carries a UL 94 Yellow Card at a defined thickness, yet a finished connector shell, sensor bracket, coil former, or power-tool housing fails an incoming vertical burn test. The laboratory classification is not an unrestricted statement of part-level flame performance. It is a material property obtained from rectangular bars of 125 mm × 13 mm moulded under controlled conditions and tested in the vertical position according to UL 94 / IEC 60695-11-10:2013. The moulded PBT part departs from those conditions in wall thickness, melt thermal history, filler orientation, weld-line population, colour package, moisture content, and surface-to-volume ratio. Each departure is capable of moving a V-0 compound into a configuration that drips flaming particles or exceeds the statutory afterflame window.

For PBT, thermal history is rarely neutral. PBT is a semi-crystalline polyester with a melt temperature near 225 °C. Commercial processing occurs between 250 °C and 270 °C. At the upper boundary, and especially in the presence of moisture above 0.02 wt% measured by ISO 15512:2019, hydrolytic chain scission reduces molecular weight and lowers melt viscosity. The UL test bars are generally moulded from dry resin with short residence time. A production tool with a hot runner, a long nozzle seat, or an oversized barrel can add several minutes of melt residency. The resulting moulded PBT part may therefore have a lower effective molecular weight than the classified resin. The lower viscosity reduces char integrity and promotes dripping, which is the most common reason a PBT article that should be V-0 fails a part-level test. Pre-drying at 120 °C for 4 h using a desiccant drier with a dew point of −30 °C or lower is the standard corrective boundary; hopper dryers operating above 60% relative humidity do not reliably maintain the required low moisture content.

Why Does a Laboratory V-0 Rating Transfer So Poorly to Production PBT Geometries?

The vertical burning classification in UL 94 is thickness-specific and colour-specific. A V-0 rating is valid only for the minimum thickness listed on the Yellow Card and for the tested colour. For a V-0 result, no individual afterflame time may exceed 10 s, the total afterflame time for five specimens may not exceed 50 s, and the afterflame plus afterglow time after the second flame application may not exceed 30 s for any specimen. Flaming drips that ignite the cotton indicator placed 300 mm below the specimen are not permitted. A methane burner with a 20 mm flame is applied for 10 s twice to each bar. These pass-fail thresholds are applied to a prismatic specimen clamped at one end with its longitudinal axis perpendicular to the burner. A moulded PBT part rarely replicates that geometry. A connector housing may have a 0.4 mm snap latch attached to a 2.1 mm main wall. The thin latch loses heat rapidly, has a small pyrolysis volume, and reaches its ignition threshold sooner than the UL bar. The thicker wall may char adequately, but the part fails because the latch drips or reignites after the burner is removed. The Yellow Card may state V-0 at 0.75 mm, but that does not certify the 0.4 mm section unless 0.4 mm is separately listed. Similarly, a natural, unfilled PBT grade with a V-0 card at 1.6 mm cannot automatically support a black, glass-reinforced moulded article at 1.0 mm. The thickness and colour fields on the UL file are boundary conditions, not interchangeability clauses.

When the keyword phrase appears in a non-conforming part, the first technical action is to section the failed moulding and verify the actual thickness at the ignition point. The laboratory bar has uniform thickness and a single edge gate. The moulded part has ribs, bosses, snap features, and local flow fronts. A V-0 classification is not assignable to an arbitrary geometry by inference. The part should be tested at the minimum actual wall thickness, and the burn specimen should be taken from the same flow region that failed. If no standard specimen can be cut because the part is too small or curved, the UL 94 criteria may have no direct legal or engineering meaning; the component should be assessed under the relevant end-product standard, such as IEC 60695-2-11 or IEC 60695-2-12 for glow-wire ignition, or the specific appliance standard in force.

The test bar is not the part: skin-core morphology, weld lines and moulded-in stress

In a standard UL test bar, laminar flow fills a simple rectangular cavity from one end, producing a relatively uniform skin-core structure. A production PBT moulding contains weld lines at pin holes, multiple gates, or flow-front junctions around inserts. Weld lines in glass-fibre-reinforced PBT are planes of reduced fibre interpenetration and local resin enrichment. Flame-retardant dispersion can be depleted along the weld line because the additive package segregates differently from the polymer melt at the weld boundary. During combustion, the weld line becomes a preferential propagation path: it opens under thermal expansion, exposes fresh polyester to the flame, and can produce a local afterflame that exceeds the 10 s limit even when the surrounding material is compliant. The UL specimen moulded from the same resin does not contain that weld line.

Moulded-in stress from unbalanced cooling, highly polished surfaces, and non-uniform ejection also changes the ignition pattern. Differential shrinkage between a thick wall and a thin rib can produce tensile stress localisation at the transition radius. When the part is flame-tested, the stored stress relaxes and distorts the part toward the flame or away from the drip path. This distortion is not represented in the freely shrinking UL bar. Consequently, a part may drip onto the cotton indicator because the geometry deflects toward the burner, reducing the distance between the flame and the polymer surface. The material has not changed; the boundary condition has.

Melt residency, regrind ratio and screw recovery

Flame-retardant PBT compounds that are stable under short-residence laboratory moulding can degrade in a production machine with a 24:1 L/D single screw, a shut-off nozzle, and a hot-runner manifold. The effective melt residency in a hot-runner tool may reach 4–6 min during interruptions, compared with less than 60 s for the UL bar mould. If the hot-runner manifold is held above 270 °C, brominated flame-retardant systems can begin to release hydrogen bromide, which then catalyses further hydrolysis of the PBT ester linkage. The melt darkens, the molecular weight drops, and the moulded part fails drip testing at a thickness where the raw resin passes. The same resin lot may be fully compliant on the UL card and non-compliant in the production article because the thermal histories differ.

Regrind content is another controlled but frequently overlooked variable. A moulded PBT part containing 25 wt% or more of in-house regrind is not represented by the UL Yellow Card unless the card explicitly lists regrind content. Each heat history consumes some of the flame-retardant synergist and reduces the molecular weight of the polyester. Differential scanning calorimetry may show only minor changes, but dynamic viscosity measured by ISO 1133-1:2022 can show a melt flow rate increase of several grams per 10 min. The lower viscosity allows the part to slump away from the flame and release burning droplets. Production trials on a 75-ton all-electric injection moulding machine have shown that a 9 s increase in cycle interruption can raise the hot-runner gate temperature by 30 °C and produce a visible V-0 failure at a weld line, while the same tool with a shorter interruption does not fail. The UL file remains unchanged because the UL file does not experience the hot-runner interval.

When glass fibre content exceeds 30 weight per cent, oxygen index and ignition behaviour diverge from unfilled PBT datasheets

Unfilled PBT exhibits a limiting oxygen index in the 20–22% O₂ range when measured by ISO 4589-2:2017, which is below the oxygen concentration of room air and therefore consistent with self-sustained combustion. Halogenated glass-reinforced PBT compounds are formulated to raise the oxygen index above 28%, but the gas-flame response of a glass-reinforced moulded part is not governed solely by the polymer matrix. Glass fibres oriented parallel to the flow front can create capillary paths that transport low-viscosity pyrolysis products toward the flame. In a 30 wt% short-glass PBT article, the burning behaviour measured on actual moulded plaques may diverge from the UL bar because the plaque gate location and the developing fibre mat differ from the single edge-gated UL specimen. At 45 wt% glass loading, the PBT matrix fraction is low. If the flame retardant is introduced as a masterbatch blended at the press, distributive mixing may be inadequate in the low-matrix-volume melt. The resultant glass-rich weld lines can ignite quickly when a small flame impinges on an exposed fibre bundle. The same material may pass V-0 at 0.75 mm on the UL card and fail an internal vertical burn on a 0.4 mm rib containing 45 wt% fibre. The failure is not an error in the file; it is a different composite architecture.

Polybutylene terephthalate compounds containing brominated polystyrene or brominated epoxy plus antimony trioxide depend on a narrow synergist ratio to develop the V-0 char. The ratio of halogen to antimony is typically maintained near 3:1. In a moulded part, the surface skin can become depleted or enriched in the synergist during long hold times and high shear. If the part surface is tested without machining, the surface layer may behave differently from the interior. The UL bar is usually tested as moulded, but its simpler flow path gives a more uniform synergist distribution than a complex moulding with local shear spikes. Published data for this specific configuration is limited; the UL Yellow Card remains the only reliable material-level classification, and it applies only to the tested colour, thickness, and moulding conditions represented in the test programme.

Pigments are not inert dilution. A carbon black masterbatch at 0.5 wt% or higher can alter thermal conductivity and surface radiation absorption. Organic red or yellow colorants can influence the radical pool during combustion, changing afterflame time by several seconds. The Yellow Card is colour-specific. A V-0 classification for natural or black PBT cannot be extended to a custom orange moulded part without an additional test. The same restriction applies to external mould-release agents and process aids. Zinc stearate loadings above 0.1 wt% can interfere with halogen-antimony flame-retardant chemistry by forming volatile zinc halides and destabilising the char layer. A moulded PBT part that is coated with a parting agent during ejection may therefore fail a part-level burn test even though the resin is V-0 at the same nominal thickness.

From the nozzle seat to the flame front: a thermal history audit

When a moulded PBT part fails despite a V-0 UL file, the first audit is not the resin certificate but the thermal and mechanical path from pellet drying to part ejection. Critical measurements include moisture content at the feed throat by ISO 15512:2019, melt temperature at the nozzle with a needle pyrometer, hot-runner manifold temperature over a 10 min interruption, screw recovery time, back pressure, cushion size, and part weight. In one production line using a 100-ton hydraulic press and a 2-cavity hot-runner tool, an increase in cycle interruption from 6 s to 9 s was sufficient to push the gate melt temperature from 265 °C to 295 °C and produce a visible darkening of the PBT melt. The affected parts failed at a weld line, while the same parts moulded with the shorter interruption did not fail. The UL file did not change; the processing window did. The distinction between a V-0 material and a passing moulded PBT part is therefore a manufacturing boundary, not a raw-material property.

Drying, barrel temperature profile, back pressure, screw rotation speed, hot-runner set points, gate design, weld-line location, regrind fraction, and local wall thickness are all part-level flammability variables. They must be controlled and recorded together with the UL 94 rating. A part drawing that cites “PBT V-0” without the associated thickness, colour, regrind allowance, and processing limits does not describe a flame-retardant article; it describes an impossible extrapolation from a test specimen to a production geometry. The UL file may say V-0, but the moulded PBT part behaves according to the conditions under which it was actually produced.