In chemically blown ethylene-vinyl acetate foams, the lower density limit is conventionally treated as a function of blowing agent dose, but at vinyl acetate contents of 17–18 wt% the controlling variables shift toward melt viscoelasticity, crosslinking kinetics, and gas retention. EVA copolymers in this composition window exhibit residual polyethylene crystallinity below the level required to maintain cell wall integrity during expansion with azodicarbonamide, while simultaneously retaining sufficient melt elasticity for processing on single-screw extruders, twin-screw compounding lines, and compression moulding presses. The lower density limit is operationally defined as the minimum apparent density obtained before the open-cell volume fraction exceeds 10% as determined by ISO 4590:2016, or before visible surface collapse and internal void channels appear. The apparent density of a foam measured according to ISO 845:2006 or ASTM D792-20 therefore cannot be reduced indefinitely by increasing blowing agent concentration alone; instead, the practical lower density limit arises from the point at which bubble growth transitions from biaxial extensional deformation to cell wall rupture and gross gas escape through the melt. On production lines, this transition is observed as a density plateau or an increase in density at elevated blowing agent loadings, accompanied by surface blowholes, internal channeling, and post-expansion shrinkage.

Why Does the 17–18 wt% Vinyl Acetate Window Constrain Lower Density Limits in ADC-Foamed EVA?

For EVA containing 17–18 wt% vinyl acetate, differential scanning calorimetry according to ISO 11357-3:2018 typically resolves a broad crystalline melting endotherm between 82 °C and 89 °C with a heat of fusion in the range 38–55 J/g, which is substantially lower than the 90–120 J/g heat of fusion recorded for unbranched LDPE under the same heating rate of 10 K/min. This crystallite population functions as a physical crosslink network only at temperatures below the melting onset; once the blowing and crosslinking reaction mixture reaches 160–175 °C, the crystallites are molten and the material is a viscoelastic fluid whose shear viscosity follows a shear-thinning curve measured by capillary rheometry under ISO 11443:2021. At a characteristic shear rate of 100 s⁻¹ and 170 °C, EVA with 18 wt% VA and a melt flow index of 2.5 g/10 min (ISO 1133-1:2022, 190 °C/2.16 kg) can exhibit an apparent viscosity below 1.0 kPa·s, while a low-MFI grade of 0.7 g/10 min remains above 2.5 kPa·s under identical conditions. This viscosity distinction is decisive for density reduction because the expanding gas volume produces biaxial extensional deformation rates in the cell wall that frequently exceed the linear viscoelastic region; melt strength, not shear viscosity alone, governs whether a nucleated cell survives to the final density. The same thermal data explain why lower VA grades are stiffer during expansion but less able to accept high filler or plasticiser loadings, whereas higher VA grades above 24 wt% eliminate the crystalline network completely and require higher crosslink density to stabilize thin cell faces. Compounding of these copolymers on a 40:1 L/D co-rotating twin-screw extruder with barrel temperatures of 130–155 °C, screw speed 300–500 min⁻¹, and a pellet melt temperature below 160 °C is required to disperse azodicarbonamide without premature decomposition. The ADC particle size distribution, typically with a median diameter between 5 µm and 12 µm, controls the number of potential gas-release nuclei; finer grades increase bubble count but can intensify thermal decomposition exotherms near the screw tip if local viscous dissipation raises the melt above 170 °C. Hydrotalcite, zinc stearate, and zinc oxide are commonly used as activators to reduce ADC decomposition onset to 155–165 °C, but at the expense of a narrower processing window because the time to 90% gas release may shorten to less than 4 min at 170 °C. For 17–18 wt% VA EVA, predrying at 70–80 °C for 4 h is prescribed when ambient relative humidity exceeds 60%, as residual surface moisture hydrolyzes acetic acid from the vinyl acetate ester during long residence-time processing and contributes to plate-out on downstream calender rolls. High-shear dispersion of the blowing agent into the melt is therefore a direct boundary condition for reproducible lower density limits, because agglomerates larger than 20 µm produce sporadic large voids that collapse preferentially when internal gas pressure is high.

When Blowing Gas Diffusion Overtakes Melt Strength at the Foam Cell Wall

Based on bubble dynamics in non-Newtonian polymer melts, the lower density limit is reached when the gas escape flux through the cell wall equals the gas inflow from adjacent decomposing blowing agent particles. The permeability of an 18 wt% VA EVA melt to nitrogen and carbon dioxide is higher than that of a 9 wt% VA grade at equivalent MFI because the polar vinyl acetate comonomer disrupts the close packing of crystalline lamellae, increasing fractional free volume and reducing barrier properties. Cell wall survival is therefore dependent on transient extensional viscosity and strain-hardening behavior, which can be evaluated in uniaxial extension using a filament stretching rheometer or estimated from dynamic oscillatory data generated under ISO 6721-10:2015. When the storage modulus G′ at 170 °C and 1 rad/s falls below approximately 1×10⁴ Pa before the peroxide crosslinking network develops, cells coalesce rapidly and the foam density does not fall below 0.20–0.25 g/cm³ even when ADC dose is increased beyond 6 phr. Addition of trimethylolpropane trimethacrylate at 0.3–0.8 phr or triallyl cyanurate at 0.2–0.5 phr increases crosslink density during the blowing window and permits expansion to 0.12–0.16 g/cm³, but excessive co-agent concentration above 1.0 phr can embrittle cell struts and reduce tensile properties below the minimum required by ISO 1798:2008. This property cliff-edge is a central reason that the practical density window for 17–18 wt% VA foam is narrow: formulation changes of as little as 0.2 phr co-agent or 0.5 phr ADC can shift the lower density by 0.03–0.05 g/cm³ or produce open-cell collapse. At the same time, the blowing gas composition generated by ADC is not pure nitrogen; it contains nitrogen, carbon monoxide, carbon dioxide, and ammonia in proportions that vary with decomposition temperature and activator chemistry. Ammonia is highly soluble in polar EVA and can plasticize the cell wall, while carbon dioxide has a lower diffusivity than nitrogen but a higher solubility; these transport differences mean that density reduction is accompanied by a shift in the gas composition retained within the closed cells. If the foam is quenched before the internal gas mixture reaches equilibrium with the surrounding atmosphere, subsequent cell shrinkage can produce an apparent density increase of 0.02–0.04 g/cm³ when measured after 24 h ageing under ISO 845:2006 conditions. This post-expansion densification is accelerated by high residual ammonia and by storage above 40 °C, and it represents a practical lower density boundary that cannot be corrected by further blowing agent addition. Published data for this specific configuration is limited, but supplier technical bulletins commonly report that the lowest stable density is obtained only after a conditioning period that allows the internal gas composition to reach a steady state, meaning that in-process density measurements alone overestimate the true lower limit. On production-scale compression moulding lines, lower density limit failures are observed as pre-cure in the central cavity, surface blowholes, and density gradients exceeding ±0.02 g/cm³ across a 300 mm × 300 mm plaque. A 0.5 MN hydraulic press with electrically heated platens and a clamp surface temperature uniformity of ±3 °C across the mould face is required to hold the compound at 165–172 °C for 7–12 min. If the platens deviate above 175 °C, peroxide decomposition generates crosslinks before ADC releases gas, trapping nitrogen in a highly viscous matrix and producing density above 0.20 g/cm³ with hard, glassy cell walls. Injection moulding grades containing 17–18 wt% VA and MFI above 8 g/10 min are limited to short flow paths because the shear heating in a 20:1 L/D reciprocating screw can raise melt temperature to the blowing agent activation range inside the barrel, causing gas phase separation ahead of the check ring and intermittent shot-to-shot density variation of ±0.05 g/cm³ or more. The same grades exhibit surface delamination when the mould cavity pressure drops too early, particularly in thick sections exceeding 6 mm, because the gas expansion rate exceeds the rate at which the outer skin can deform without rupturing.

Lower Density Limit Comparisons Across Azodicarbonamide and Endothermic Systems

Comparative evaluation of ADC and endothermic chemical blowing agents in the same 17–18 wt% VA base resin shows that the lower density limit also depends on gas yield per unit mass and decomposition thermal signature. ADC releases approximately 220 mL/g of gas at standard temperature and pressure, whereas an endothermic citric acid/sodium bicarbonate system may release only 120–140 mL/g; however, the endothermic system produces carbon dioxide and water vapor, both of which have different solubility and diffusivity in EVA than the nitrogen-rich gas mixture from ADC. The following representative data set, compiled from supplier technical bulletins and normalized test plaques, illustrates the plateau behavior described above.

Code	VA (wt%)	MFI (ISO 1133-1:2022)	Blowing agent	Blowing agent dose (phr)	Co-agent dose (phr)	Cure schedule	Lower density (ISO 845:2006)
A-17	17	1.8	ADC activated	4	0.5 TMPTMA	165 °C/8 min	0.17 g/cm³
B-18	18	2.5	ADC activated	5	0.5 TMPTMA	165 °C/8 min	0.14 g/cm³
C-18	18	0.7	ADC activated	6	0.8 TMPTMA	170 °C/10 min	0.11 g/cm³
D-17	17	2.5	ADC activated	5	none	165 °C/8 min	0.22 g/cm³
E-18	18	2.5	Endothermic HCO3/citric acid	6	0.5 TMPTMA	160 °C/8 min	0.19 g/cm³

Compression set measured according to ISO 1856:2018 at 23 °C after 50% deflection for 24 h tends to rise when density falls below 0.13 g/cm³, because the same crosslink network that stabilizes thin cell faces is insufficient to prevent irreversible buckling under load. Tensile strength and elongation at break measured with ISO 1798:2008 also decline steeply below 0.12 g/cm³; published data for this specific configuration is limited, but available supplier curves indicate that elongation at break may fall below 200% when cell wall thicknesses approach the submicron range. This property cliff is why grades with 17–18 wt% VA are frequently specified for densities between 0.14 g/cm³ and 0.20 g/cm³ in gasketing, footwear midsoles, and marine flotation closed-cell foam, while lower-density open-cell structures are more reliably obtained with higher VA contents above 24 wt% or with physical foaming technologies. The lower density limit therefore should be understood not as a single material constant but as a formulation-specific point at which mechanical properties fall below the acceptance criteria of the intended application. Regulatory compliance for a closed-cell EVA foam at 17–18 wt% VA depends on the end-use category and regional jurisdiction. In food-contact applications, the base resin must meet FDA 21 CFR 177.1350 for EVA copolymers, and finished articles intended for EU food contact must satisfy overall migration limits under EU 10/2011 with simulant-specific testing according to EN 1186 methods. In electrical and electronic enclosures, the activated foam may require screening for lead, mercury, cadmium, hexavalent chromium, and brominated species under RoHS 2011/65/EU Annex II, although inspection is typically performed by X-ray fluorescence screening according to IEC 62321-3-1:2013. The migration kinetics of additives from EVA into food simulants follow Fickian diffusion with coefficients that increase with vinyl acetate content and temperature; thus 17–18 wt% VA grades occupy an intermediate migration-risk position compared with low-VA polyolefin resins. The following compliance matrix consolidates the central standards and test conditions used for this class of foam.

Requirement	Standard or regulation	Test condition	Typical acceptance boundary for 17–18 wt% VA foam
Apparent density	ISO 845:2006	Conditioned 24 h at 23 °C/50% RH	0.14–0.20 g/cm³ for technical foam
Tensile strength	ISO 1798:2008	Dumbbell specimen, 50 mm/min	Minimum 1.0 MPa at 0.15 g/cm³ typically
Compression set	ISO 1856:2018	50% deflection, 23 °C/24 h	Maximum 30% for sealing applications
Food contact resin	FDA 21 CFR 177.1350	Resin specification	VA content and extractives within codified limits
EU food contact article	EU 10/2011	Overall migration by EN 1186	≤ 10 mg/dm² overall migration
Restricted substances	RoHS 2011/65/EU Annex II	IEC 62321-3-1:2013 XRF screening	Below threshold concentrations per Directive

Thermal Degradation and Crosslinking Boundaries Appear When the Cure Outpaces Gas Release

The overlapping thermal windows of ADC decomposition and dicumyl peroxide crosslinking create the most critical processing constraint for 17–18 wt% VA foam. Dicumyl peroxide at typical loading of 0.6–1.2 phr has a half-life at 170 °C in the range of 1–2 min depending on the exact assay and carrier, while activated ADC under the same conditions reaches 90% gas release within 4–7 min. Vulcanisation kinetics measured on a moving die rheometer under ISO 6502-2:2018 or with a rotorless shear rheometer under ASTM D6601-19 show that the degree of crosslinking at the end of the blowing plateau must exceed approximately 70% of the maximum torque to prevent cell collapse during cooling. If crosslinking lags behind gas release, the density lower limit is not determined by gas availability but by the survival of the foam structure after the press is opened or the die exit is cleared. Conversely, acetyl residues in the EVA chain begin to eliminate acetic acid at sustained temperatures above 200 °C; therefore, any attempt to accelerate the cure by raising temperature above 185 °C introduces discolouration, odour, and acid-catalyzed chain scission that degrade the closed-cell membrane and increase open-cell content. The practical processing window for crosslinked EVA foam at 17–18 wt% VA is therefore limited to approximately 160–175 °C, a span of only 15 °C, and within that span a change of ±5 °C can shift the lower density limit by 0.02–0.04 g/cm³ on compression moulding tooling with nonuniform heat transfer. In a 45 mm single-screw sheet extrusion line producing 2.5 mm unskinned EVA foam tape from an 18 wt% VA grade with MFI 2.5 g/10 min, barrel temperatures are typically profiled from 110 °C at the feed throat to 160 °C at the die, with a gear pump maintaining inlet pressure below 8 MPa and die melt temperature below 168 °C. When the same compound is run on a 27 mm co-rotating twin-screw line with a slit die, the higher specific energy input can shift the lower density limit upward by 0.03 g/cm³ because localized viscous heating at the screw elements introduces premature gas generation before the die land. Operators using compression moulding report a batch-to-batch density range of ±0.015 g/cm³ when MFI moves by 0.2 g/10 min, which is within the normal lot-to-lot variation of EVA copolymers from different polymerisation trains. The practical lower density boundary for 17–18 wt% VA in chemically blown foam therefore remains between 0.11 g/cm³ and 0.17 g/cm³ on production equipment, with the exact value set by the crosslink-to-blowing ratio, moisture content, and the thermal uniformity of the specific tooling.