Air Handling Duct Fire Resistance with Flame Retarded Low VOC Plastisols

In coil coating lines where galvanised steel strip for air handling ductwork is coated with a low-VOC plastisol, the combined requirements of reaction-to-fire classification and duct fire resistance impose simultaneous limits on plasticizer volatility, flame retardant loading, and film fusion temperature. A plastisol formulated as a dispersion of PVC homopolymer or copolymer paste resins in a phthalate-free or low-volatility plasticizer system is applied by reverse roller coater at dry film thicknesses between 80 µm and 250 µm depending on the duct pressure class and internal abrasion duty. The wet film must gel and fuse in an oven with metal peak temperature held at 180 °C to 200 °C for 60 s to 120 s on a continuous line with zone temperatures from 160 °C to 220 °C. Low-VOC performance is assessed on the cured film using VDA 277 headspace gas chromatography with a target total value below 100 µgC/g, or using chamber emission testing according to ISO 16000-6:2021 with toluene-equivalent TVOC below 0.25 mg/m³ after 28 days under 23 °C and 50 % RH. Fire performance of the coated duct must simultaneously satisfy the reaction-to-fire classification criteria in EN 13501-1:2018 and the fire resistance criteria in EN 1366-1:2014 for ventilation ducts. This dual requirement excludes formulations that use volatile phosphate esters or low molecular weight chlorinated paraffins at high addition levels because the emission budget is consumed during early thermal ageing while the char structure remains insufficient for the 120 min furnace exposure when tested as a fully assembled duct section.

Adhesion of the cured plastisol to galvanised steel is governed by the surface pre-treatment, the peak metal temperature, and the residual volatile content at the end of the fusion oven. A zirconium oxide conversion coating at 50–100 mg/m² is preferred because it avoids the environmental restrictions of chromate pre-treatments under REACH and provides a stable polar interface for the PVC film. The cured coating must maintain adhesion under the helical seam compression of spiral duct forming, which applies local pressures between 5 MPa and 15 MPa; the T-bend adhesion test per ASTM D4145-10 is used as a release criterion, with ratings of 0T to 2T required for strip thicknesses below 0.8 mm. The low-VOC flame-retarded systems described below are formulated as dispersions rather than solutions, and their processability depends on the paste resin particle size, plasticizer solvency, and the particle size distribution of the solid flame retardants. Representative property ranges for two flame-retardant approaches and a low-VOC control are summarised in the first table; these ranges derive from published development data for plastisol formulations and must be verified for a specific duct coating because substrate gauge, oven profile, and topcoat thickness influence the absolute classification.

Flame Retardant Synergy with Antimony Trioxide and Zinc Borate in Duct Coatings

Antimony trioxide functions as a gas-phase radical trap only when a halogen donor is present; in a plastisol film the PVC polymer itself provides the required chlorine, but the diffusion of HCl from the condensed phase into the flame zone is velocity dependent and becomes inefficient at the low heating rates experienced by the unexposed side of a duct. In ISO 5660-1:2015 cone calorimeter testing at an incident heat flux of 50 kW/m², a low-VOC DINCH-plasticized film containing 60 phr plasticizer without flame retardant typically exhibits a peak heat release rate between 250 kW/m² and 290 kW/m², while the addition of 5 phr antimony trioxide and 3 phr zinc borate reduces the peak heat release rate to approximately 160–190 kW/m² and raises the limiting oxygen index from 23–25 % to 29–31 % when measured according to ASTM D2863-19. The synergistic efficiency of antimony trioxide is degraded by calcium carbonate fillers, which sequester evolved hydrogen chloride as calcium chloride, so duct formulations requiring high filler loading for hardness must compensate with a finer antimony trioxide particle size below 1 µm or with a replacement of calcium carbonate by alumina trihydrate. Zinc borate acts as a condensed-phase char promoter, forming a glassy boron-zinc oxide network that reduces afterglow and lowers smoke density; at 3 phr with 5 phr antimony trioxide, the specific optical density Ds max in ASTM E662-21a is reduced from 480–620 to 300–380. However, zinc borate releases water of hydration above 290 °C, which can create pinholing during fusion if the coating exceeds 200 °C for more than 90 s, so the process window for this system must not exceed a metal peak temperature of 195 °C.

The gas-phase contribution of antimony trioxide also carries an operational boundary in duct applications: the antimony chloride formed in the flame zone condenses on cool duct surfaces during intermittent fire tests, and this residue can compromise electrical continuity testing for leakage classification. For air handling units with internal filter housings, formulations using zinc stannate at 2 phr plus alumina trihydrate at 30 phr avoid antimony condensation and achieve char expansion ratios between 5:1 and 10:1, which is beneficial when the duct is intended to meet the insulation criterion of EN 1366-1:2014 by restricting the unexposed face temperature rise to 180 °C above the initial temperature. The char expansion, however, is shear-history dependent: if the plastisol is applied by high-pressure airless spray with a tip pressure above 200 bar, the platelet orientation of the zinc borate particles is modified and the resultant char tensile strength may fall below the minimum required to resist the furnace pressure differential of 300 Pa used in fire resistance tests. Therefore, spray-applied duct liners require a separate qualification for each gun configuration and filter size, and published data for this specific application remains limited.

During isothermal ageing at duct operating temperatures between 50 °C and 80 °C, the residual low molecular weight fraction of a plasticizer migrates along the concentration gradient to the film–air interface and is removed by the air stream as a condensable aerosol. In phthalate-free low-VOC plastisols, the plasticizer system is typically based on diisononyl cyclohexane-1,2-dicarboxylate (DINCH) with a molecular weight of 424.7 g/mol and a vapour pressure below 1.0×10⁻⁵ Pa at 25 °C, or on dioctyl terephthalate (DEHT) with a molecular weight of 390.6 g/mol; both exhibit lower emission rates than di(2-ethylhexyl) phthalate, but their higher molar volumes reduce plastisol solvency and require longer maturation at 25 °C for 24 h before coating to reach a stable viscosity plateau. Emission testing of the cured film according to ISO 16000-6:2021 in a 0.5 m³ chamber at 23 °C and 50 % RH with an air change rate of 0.5 h⁻¹ typically yields a toluene-equivalent TVOC concentration below 0.25 mg/m³ at 28 days for formulations with a total extractable plasticizer content below 15 % of film mass. The migration kinetics follow a Fickian diffusion model only until the surface concentration builds to a monolayer; above this, the emission rate becomes controlled by the air film mass transfer coefficient. For a 100 µm film fused at 190 °C, the apparent plasticizer diffusion coefficient at 60 °C is in the range of 5×10⁻¹⁴ m²/s to 5×10⁻¹³ m²/s, but the presence of epoxidized soybean oil at 3 phr can reduce this by a factor of two due to increased polymer chain entanglement. The low-VOC requirement does not end at the plasticizer; volatile stabilizer solvents, emulsifiers from the PVC paste resin, and moisture adsorbed on alumina trihydrate all contribute to the VDA 277 headspace value, and a filler pre-drying step at 110 °C for 4 h is mandatory when relative humidity during storage exceeds 60 %.

What Limits Coating Thickness Uniformity on Perforated Duct Sections?

Perforated duct sections intended as pre-coated plenum panels introduce a discontinuous substrate profile that disrupts the hydrodynamic pressure profile under a knife-over-roll coater. The low-VOC flame-retarded plastisol must exhibit a yield stress above 25 Pa and a thixotropic index between 2.5 and 3.5 when measured as the ratio of Brookfield RVT spindle 6 torque at 2 rpm and 20 rpm at 25 °C; below this yield stress, edge tear at the perforation perimeter results in a dry-film thickness variation from 40 µm to 180 µm on a 200 µm nominal coating. The solid flame retardants, especially alumina trihydrate with a median particle size of 8–12 µm, increase packing density at the perforation edges, creating a local viscosity increase that cannot be corrected by line speed alone. Production-scale equipment behavior documented on coil coating lines shows that a reverse roll coater with a metering roll gap of 0.15 mm and an application roll speed ratio of 0.6:1 to 0.8:1 produces a more uniform film on perforated galvanised steel than a knife-over-roll coater, but the application roll must be a closed-cell elastomer with Shore A hardness between 70 and 80 to maintain contact across the holes without excessive pick-up. The viscosity must be kept below 4500 mPa·s at 20 rpm; higher values trap air in the perforations and form surface blisters during fusion, which subsequently become localised ignition sites under EN ISO 11925-2:2020 single-flame source testing because the film is thinner and the flame retardant concentration is locally depleted. Addition of fumed silica at 1.5 phr restores the required yield stress but increases the low-shear viscosity to more than 12000 mPa·s, so the formulation must be degassed under vacuum at 20–50 mbar for at least 15 min before application to prevent pinholes. The coating thickness on perforated sections is therefore a process-dependent parameter, and qualification specimens must be taken from the actual production line rather than from laboratory drawdown bars.

When EN 1366-1 Fire Exposure Extends Beyond 120 Minutes

Under EN 1366-1:2014, a ventilation duct is mounted in a furnace and subjected to the standard time–temperature curve defined in ISO 834-1:1999, where the furnace temperature is expressed as T = 20 + 345 log10(8t + 1); after 120 min, the furnace temperature reaches approximately 1049 °C. The coated duct must maintain integrity and, where required, thermal insulation for the classification period, with the unexposed side temperature rise limited to 180 °C above ambient or, for some configurations, 140 °C when measured with fixed thermocouples under the insulation criterion. A low-VOC flame-retarded plastisol film of 200 µm thickness is not itself the primary fire barrier; instead it protects the galvanised steel substrate from destabilisation and reduces surface flame spread so that the assembled duct can achieve an EI 120 S classification under EN 13501-3:2005+A1:2009. The critical threshold in extended exposure occurs when the char layer separates from the steel at the zinc–iron alloy interface, an event governed by the differential thermal expansion between the char, the intermetallic layer, and the steel. In laboratory furnace tests, adhesion loss is observed when the steel substrate reaches 550 °C to 600 °C, which corresponds to a furnace exposure time of 90 min to 120 min depending on the duct gauge and insulation. Formulations containing resorcinol bis(diphenyl phosphate) at 12 phr reduce the char separation by plasticising the char at high temperature, but the phosphate ester can re-volatilise during furnace exposure and contribute to flaming droplets if the film has not been fully fused; therefore the d0 classification per EN 13501-1:2018 requires a fusion temperature of at least 185 °C and a residual volatile content below 0.5 wt% measured by thermogravimetric analysis at 200 °C for 10 min. When a duct is exposed to fire from both sides, the internal coated surface forms an intumescent char that can reduce the free cross-section and increase pressure drop; this must be accounted for during the design of smoke extraction ducts where the free area must remain above the minimum specified in the building code.

Extended exposure beyond 120 min imposes fatigue-like thermal cycling on the char because the furnace control thermocouples cause momentary temperature oscillations of ±10 °C around the setpoint; these oscillations can induce microcracking in the char if the film contains more than 35 phr of particulate flame retardant. High alumina trihydrate contents of 40 phr produce a thick insulating char but the char compressive strength declines below 0.2 MPa after 150 min, whereas zinc stannate-modified formulations retain between 0.4 MPa and 0.6 MPa under the same conditions. The selection of flame retardants for extended exposure therefore involves a trade-off between char expansion and char mechanical integrity; published data for plastisol-coated ducts tested beyond 120 min is limited, and fire resistance classification should be verified on the full duct section rather than on small-scale film specimens.

On a spiral duct fabrication line where pre-coated galvanised strip is roll-formed into a helical seam, the cured low-VOC plastisol must withstand a bend radius equal to the strip thickness plus 0.8 mm without microcracking because the helical seam is compressed by the forming rolls under a pressure between 5 MPa and 15 MPa. The T-bend adhesion test per ASTM D4145-10 is used to evaluate the coating on the strip before roll-forming, with a rating of 0T to 2T required for duct wall thicknesses below 0.8 mm; failure at the seam is observed as a continuous crack along the lockseam, which becomes a preferential path for flame spread in the UL 181 flame penetration test. Pre-treatment of the galvanised substrate with a thin chromate-free conversion coating based on zirconium oxide at a coating weight of 50–100 mg/m² is necessary because the PVC film adhesion is otherwise reduced by the zinc oxide patina formed during storage at relative humidity above 60 %. In production, the most common failure mode is not combustion but delamination during under-fire pressure pulses, where the duct is subjected to a pressure differential of ±500 Pa during the test; the pressure pulse lifts the coating from the substrate at the cut edges and exposes the steel to direct furnace gases. To control this, the duct fabrication must include an edge sealing step using the same plastisol at a wet-film thickness of 100 µm over the cut ends, followed by a localised infrared cure at 160 °C for 30 s. Batch-to-batch variance in PVC paste resin particle size from 0.2 µm to 2.0 µm changes the plastisol yield stress and edge coverage; therefore a minimum 48 h maturation at 25 °C after mixing is required before viscosity release to the coater.

Low-VOC Plastisol Viscosity Depression and Pot-Life Stability

Viscosity control in flame-retarded low-VOC plastisols is complicated by the competing effects of phosphate ester plasticizers, which act as viscosity depressants, and particulate flame retardants, which increase low-shear viscosity through particle packing and polymer bridging. At 25 °C, a DINCH-based plastisol containing 45 phr DINCH, 12 phr resorcinol bis(diphenyl phosphate), 5 phr antimony trioxide, and 3 phr zinc borate typically exhibits a Brookfield RVT spindle 5 viscosity of 2900–3600 mPa·s at 20 rpm, but the viscosity may rise by 25–40 % after 24 h if the formulation is held at 30 °C without continuous slow agitation. This pot-life stability is governed by the dissolution of partially solvated PVC particles and by the release of moisture from the flame retardant surfaces; alumina trihydrate with a median particle size of 10 µm can retain 0.2–0.5 wt% moisture even after normal storage, and this moisture hydrolyzes the phosphate ester in the presence of zinc borate, producing acidic species that accelerate the gelation rate. To maintain a stable pressure in the coating pump, the formulation must be recirculated at a low shear rate below 5 s⁻¹ through a jacketed stainless steel tank held at 23 °C to 25 °C; air must be excluded by a nitrogen blanket, and the mixing blade must be a low-rpm axial-flow impeller to avoid high-shear dispersion of the PVC resin after maturation. The addition of 3 phr epoxidized soybean oil and 2 phr liquid calcium-zinc stabiliser buffers the acidic degradation products and maintains the viscosity increase below 20 % over 24 h. Solvent-based viscosity depressants are incompatible with the low-VOC requirement, but a high-boiling aliphatic hydrocarbon at 2 phr can temporarily reduce viscosity; its boiling point above 180 °C means that the first oven zone must provide an adequate flash-off section at 140 °C for at least 20 s before gelation, otherwise the solvent is trapped in the fused film and raises the VDA 277 value above the 100 µgC/g threshold. In dual-component mixing systems, the flame retardant slurry and the plastisol base must be combined under vacuum because air entrainment of 2–4 vol% increases the measured viscosity by up to 40 % and produces microvoids that reduce the film density below the required 1.25 g/cm³, lowering the char structural integrity during fire testing.

The compliance matrix for air-handling duct flame-retarded low-VOC plastisol is presented below to align laboratory screening, production release, and full-scale duct certification. The acceptance ranges are expressed as target values that must be verified on the actual production-coated strip because the fire classification of a duct assembly is not determined by the coating alone.

Compliance with the above matrix does not guarantee fire resistance of the fully assembled duct because the seam geometry, reinforcement rings, and gasket materials alter the failure mode. In a duct with flanges and transverse joints, the gasket is often a silicone rubber or EPDM element that must be protected from direct flame; the plastisol coating must therefore extend over the flange face by at least 20 mm to delay gasket ignition. The use of intumescent mastic at joints is permitted only if the mastic has a classification compatible with the duct and does not emit ammonia or amines, because amine-based additives cause premature crosslinking of the plastisol and form yellowing at the fusion stage. The pressure classification of the duct is tested separately according to EN 1507:2006 or equivalent national standards, and the coating should not bridge the seam so thickly that the leakage class changes. Published data for the combined effect of low-VOC plasticizer migration and fire resistance on air handling duct coatings is limited; therefore each installation requires a full-scale furnace test on a representative duct section fabricated from the production-coated strip.