Continuous closed-cell polyolefin foam production by peroxide crosslinking and chemical expansion is operated on tandem extrusion lines comprising a first-stage counter-rotating twin-screw extruder with a screw diameter of 90 mm and an L/D ratio of 32:1, followed by a melt pump and a second-stage single-screw cooling extruder. Low-density polyethylene with density 0.918–0.925 g/cm³ and melt flow index 0.25–2.0 dg/min at 190°C/2.16 kg according to ASTM D1238 is dry-blended with dicumyl peroxide and azodicarbonamide before entering the feed throat. The first extruder disperses the peroxide and blowing agent while limiting the melt temperature to 150–160°C; the second extruder reduces the melt temperature to 105–120°C and delivers the compound to a flat die maintained at 4–12 MPa. Crosslink density in the extrudate is quantified by gel fraction using ASTM D2765-16 Method B, in which specimens are extracted in boiling xylene for 12 h and the insoluble fraction is reported as a percentage. Closed cell volume fraction is measured by gas pycnometry per ISO 4590:2016, and foam apparent density is determined by ISO 845:2006. The expansion ratio is calculated as the quotient of solid polymer density and foam apparent density; a solid LDPE density of 0.920 g/cm³ and a foam apparent density of 33 kg/m³ yield an expansion ratio of approximately 27:1. Industrial LDPE foam formulations producing stable closed cell structures typically require gel fraction in the range of 30–60%, below which cell walls rupture during bubble growth and above which the crosslinked matrix suppresses expansion. This window is not a single fixed value but shifts with melt flow index, peroxide type, blowing agent gas yield, die geometry, and downstream cooling conditions.

What Rheological and Kinetic Boundaries Define Peroxide Crosslinking During Foam Extrusion?

Peroxide crosslinking of LDPE in continuous foam extrusion is governed by the radical half-life of the selected dialkyl peroxide and the residence time distribution across the barrel zones. Commercial dicumyl peroxide with assay between 92% and 99% exhibits a half-life of approximately 10 h at 117°C, 1 h at 135°C, and 1 min at 171°C; these values vary with solvent carrier and peroxide purity. In a tandem line the first-stage barrel zones are set between 120°C and 160°C to achieve uniform resin melting and peroxide dispersion without premature scorch, while the second-stage cooling extruder lowers melt temperature to 105–120°C to prevent uncontrolled bubble growth before the die. Crosslink density is not directly proportional to peroxide loading because radical termination and chain scission compete with macroradical recombination, and undecomposed peroxide can continue to generate radicals during bubble growth after the die. Rheological characterization by small-amplitude oscillatory shear at 130°C shows that the low-frequency storage modulus increases with gel fraction: foam-grade LDPE with gel fraction 20% may display a storage modulus below 10 kPa at 0.1 rad/s, while gel fraction 50% can raise the low-frequency storage modulus above 50 kPa. Capillary rheometry at extrusion-grade shear rates of 100–1000 s^-1 reveals a smaller relative increase in shear viscosity, confirming that crosslink density has a stronger effect on elongational viscosity and strain hardening than on shear thinning behavior. The practical processing window is bounded by die-face production data showing that peroxide additions above 1.5 phr tend to produce die-lip plate-out, gel particles on the foam surface, and die pressure fluctuations exceeding ±0.5 MPa. These instability indicators correlate with non-uniform cell sizes and closed cell content below 85%, even when the average gel fraction remains within the target range.

Below a gel fraction of 25%, the extensional viscosity at bubble growth strain rates of 0.1–5 s^-1 is insufficient to withstand the outward stress, and adjacent cells coalesce into open-cell structures with measured closed cell content below 80% by ISO 4590:2016. The resulting foam may show an attractively low apparent density but also lower compressive resistance, higher water absorption, and greater surface irregularity. In the intermediate gel fraction range of 35–55%, cell walls retain enough elastic recovery to resist capillary forces during gas diffusion and quenching, producing cells with average diameters of 0.2–0.8 mm and closed cell content above 90%. At gel fractions above 70%, the crosslinked matrix behaves as a high-modulus elastomer, and the work required to inflate each bubble increases sufficiently to suppress the expansion ratio. A representative LDPE foam formulation with 0.6 phr dicumyl peroxide and 12 phr azodicarbonamide may produce an apparent density of 33 kg/m³, corresponding to an expansion ratio of approximately 27:1 based on a solid density of 0.920 g/cm³; increasing the peroxide to 1.2 phr can shift the density to 48 kg/m³, or an expansion ratio of about 19:1, despite the same blowing agent loading. This non-monotonic behavior is consistent with the view that crosslink density raises the minimum cell wall thickness that can survive gas expansion without tearing, while simultaneously lowering gas diffusivity through the polymer phase and increasing the elastic resistance to inflation. The table below summarizes representative formulation-dependent property shifts observed in LDPE foam extrusion development work.

Peroxide Concentration Gradients Induce Skin-Core Morphology Variation in Thick Extruded Foam Boards

Thick crosslinked polyolefin foam boards with thickness above 20 mm exhibit spatial variation in gel fraction because the thermal history of the extruded profile is non-uniform. The foam surface, exposed to cooling water at 15–25°C immediately after the die, freezes before full peroxide decomposition can occur, while the core retains heat from azodicarbonamide decomposition and exothermic crosslinking reactions. Thermocouple data from production lines indicate core temperatures can remain above 150°C for several minutes after skin solidification. As a result, the skin gel fraction may be 5–15 percentage points lower than the core gel fraction. This gradient affects the closed cell expansion ratio across the board thickness: core zones with higher crosslink density and slower cooling show larger cells, whereas skin zones with lower crosslink density may display collapsed cells, microvoids, or open-cell surface layers. Scanning electron micrographs of polished cross-sections from boards produced with 0.8 phr dicumyl peroxide and 15 phr azodicarbonamide show a skin layer of 0.5–1.0 mm containing flattened cells with aspect ratio below 0.4, while the core exhibits polyhedral closed cells with aspect ratios between 0.7 and 1.0. Corrective measures on production lines include reducing board thickness to below 10 mm per pass, using a water spray with temperature controlled at 25°C and variable spray density, and replacing part of the dicumyl peroxide with a higher-temperature peroxide that has a half-life of 1 min at 185°C to delay crosslinking near the skin. Failure to control this gradient may result in a bimodal cell size distribution and a reduction in overall closed cell content below 85%, detectable by gas pycnometry and ultrasonic thickness profiling.

Polypropylene formulations present a different crosslink density response because tertiary carbon radicals undergo β-scission during peroxide attack, reducing molecular weight before crosslinking becomes dominant. The foam extrusion of high melt strength polypropylene with dicumyl peroxide typically requires coagents such as trimethylolpropane triacrylate at 1–3 phr to improve crosslink efficiency, but the resulting gel fraction rarely exceeds 35% in continuous processes. The foam expansion ratios are generally lower than LDPE systems, with apparent densities in the range 50–150 kg/m³ and closed cell contents above 85%. The processing window is narrower because lightly crosslinked polypropylene shows a steep loss of elongational viscosity with increasing melt temperature; barrel temperature variation greater than ±5°C in the second-stage cooling extruder can produce density fluctuations exceeding ±10 kg/m³ across a roll. The use of higher melt strength polypropylene grades with long-chain branching reduces the dependence on peroxide crosslinking for cell stabilization but increases die pressure and may require die gaps above 0.5 mm to avoid melt fracture. In these systems, the closed cell expansion ratio is governed less by gel fraction alone and more by the branching distribution and the strain-hardening parameter measured by extensional rheometry at 170°C.

Process control in peroxide-crosslinked foam extrusion is dominated by the thermal overlap between azodicarbonamide decomposition and peroxide reaction. Azodicarbonamide decomposes exothermically between 195°C and 210°C, releasing approximately 220 mL/g of gas composed primarily of nitrogen, carbon monoxide, carbon dioxide, and ammonia, while dicumyl peroxide begins significant crosslinking above 150°C. In a tandem line with first-stage barrel profile 120°C, 140°C, 160°C, 170°C, and adapter at 175°C, the blowing agent may begin to decompose in the first extruder, causing premature foaming and erratic melt pressure. Production data indicate that when the first-stage discharge temperature exceeds 180°C, the incidence of melt pressure fluctuation above ±0.7 MPa increases, and the resulting foam exhibits non-uniform densities with apparent density variation of ±5 kg/m³ across roll width. To decouple the reactions, processors select blowing agents with higher decomposition onset, use cooling screws with high surface area, or employ direct gas injection after the crosslinking zone. The die pressure must be maintained above the solubility pressure of the decomposition gases, typically 4–10 MPa, to prevent pre-foaming inside the die land. Pressure transducers with full-scale accuracy of 0.1% are installed at the screen changer and die entry; a pressure drop below 3.5 MPa at the die generally correlates with open-cell content above 20% in the final product. These pressure limits are empirical values derived from production-scale tandem foam extrusion lines processing LDPE with melt flow indices between 0.5 dg/min and 2.0 dg/min.

When Sodium Bicarbonate-Based Endothermic Blowing Agents Replace Azodicarbonamide in Crosslinked Polyolefin Foam

Endothermic blowing agent systems based on sodium bicarbonate and citric acid are used in crosslinked polyolefin foam when lower gas pressure generation and finer cell structure are required. The decomposition of sodium bicarbonate begins near 140°C and produces carbon dioxide and water vapor with a total gas yield of approximately 120–150 mL/g, which is about 55–68% of the gas yield of azodicarbonamide. Because carbon dioxide has higher solubility in LDPE than nitrogen and a lower diffusion coefficient, the supersaturation at the die is lower, and the maximum achievable expansion ratio is reduced. Typical formulations with 0.6 phr dicumyl peroxide and 10–15 phr endothermic blowing agent produce foam apparent densities of 40–90 kg/m³, corresponding to expansion ratios of 10:1 to 23:1, compared with 25–45 kg/m³ for equivalent exothermic systems. The lower exotherm also changes crosslink density development because the cooling profile in the second-stage extruder is more stable, and the skin-core gel fraction gradient is reduced. However, water released by sodium bicarbonate can hydrolyze residual peroxide and reduce crosslinking efficiency; pre-drying of the compound is required at 60°C for 4 h when storage humidity exceeds 60% RH. Endothermic foams show lower residual blowing agent odor and are preferred for applications requiring compliance with European Union REACH restrictions on azodicarbonamide decomposition products. The trade-off is a narrower expansion ratio operating window and a stronger sensitivity to peroxide concentration because carbon dioxide plasticizes the polyolefin, reducing the storage modulus at low frequency by as much as 20% at 5 phr sodium bicarbonate loading.

Closed Cell Expansion Ratio Measurement and Standardized Test Protocols

Closed cell content and expansion ratio are not sufficient to qualify crosslinked polyolefin foam for industrial applications; mechanical and thermal aging properties must be correlated with crosslink density using standardized test methods. The apparent density of extruded foam is determined by mass and dimensional measurement under ISO 845:2006, with specimen conditioning at 23°C ± 2°C and 50% ± 5% RH for at least 16 h. Tensile strength and elongation at break are measured per ISO 1798:2008 for flexible cellular polymeric materials, with specimen thickness compensating for foam compressibility. Compression deflection at 25% strain is reported per ISO 844:2014, and dynamic fatigue resistance is assessed by ASTM D3574-17 test I3 using a constant force pounding apparatus. Gel fraction is measured using ASTM D2765-16 Method B, and crosslink density is calculated from solvent swelling using the Flory-Rehner equation with xylene at 110°C and an LDPE-solvent interaction parameter between 0.33 and 0.36. The table below summarizes the test methods used to correlate crosslink density with closed cell expansion ratio and application performance for extruded polyolefin foam.

At the production scale, operational boundaries for peroxide-crosslinked polyolefin foam extrusion are defined by raw material quality and additive compatibility. Polyolefin granules containing phenolic antioxidants can consume free radicals and reduce the measured gel fraction at a given peroxide loading; formulations containing amine-based antistatic additives or hindered amine light stabilizers may cause premature crosslinking in the feed throat because radical-scavenging by-products accumulate in recycled material. Regrind content above 20% by weight increases the risk of gel formation, melt pressure fluctuation, and surface defects unless the regrind is ground to a particle size below 2 mm and dried to moisture content below 0.05%. The extrusion line should be purged with a low-viscosity polyolefin after each run to remove residual peroxide from dead zones in the die and screen changer; failure to purge can result in black specks and localized crosslinked lumps in the next production run. Published data for the exact quantitative relation between gel fraction and expansion ratio in all polyolefin grades is limited, but the ranges provided here represent reproducible industrial operating windows for LDPE-based formulations. The combined use of ASTM D2765-16 gel fraction, ISO 4590:2016 closed cell content, and ISO 845:2006 apparent density provides a sufficient basis for process capability studies on tandem foam extrusion lines.