Property Loss Across ABS Reprocessing Cycles: What Each Pass Costs

Injection-grade acrylonitrile butadiene styrene is a two-phase material comprising a rigid styrene-acrylonitrile matrix and grafted polybutadiene rubber domains that is rarely processed only once. A moulded component rejected for knit-line flash may be granulated, blended with virgin resin, re-dried, and re-moulded; trim from a second pass may then re-enter a compounding extruder with L/D ratio between 32:1 and 40:1. Each pass applies a new thermal and shear history to the same polymer population. The measurable penalty is dominated by two competing reactions: thermo-oxidative chain scission in the styrene-acrylonitrile continuous phase, which decreases molecular weight and increases melt flow, and allylic oxidation of the butadiene phase, which raises gel content and destroys low-temperature ductility. The cost of each pass is not merely the granulation and drying energy; it is the reduction in regrind value, the narrowing of the injection-moulding process window, and the threshold at which the material can no longer meet the original specification.

Thermo-oxidative chain scission in the styrene-acrylonitrile matrix

Thermal failure in reprocessed ABS does not begin with visible yellowing; it begins with radical propagation in the SAN continuous phase. At barrel set points of 210–240°C, a twin-screw extruder with L/D 36:1 and screw speed 300 rpm can produce local melt temperatures of 250–260°C because of viscous dissipation. Under these conditions, β-scission at tertiary carbon centres on styrene repeat units reduces the number-average molecular weight of the SAN phase. Gel permeation chromatography values of Mn below 50,000 g/mol are associated with a drop in tensile yield strength from 44 MPa to below 40 MPa when measured according to ISO 527-2:2012. Simultaneously, the polybutadiene phase undergoes allylic hydrogen abstraction at 1,4-cis and 1,2-vinyl unsaturation sites. Hydroperoxide decomposition generates carbonyl and hydroxyl species, raising the rubber phase glass transition temperature from approximately −80°C toward −60°C as crosslink density increases. This dual mechanism explains why reprocessing loss cannot be corrected by adding a single stabilizer: the SAN phase requires chain-extension or molecular weight rebuilding, while the butadiene phase requires radical scavenging and hydroperoxide decomposition.

Oxidation kinetics in the butadiene phase follow a free-radical chain mechanism with initiation at allylic C–H bonds. The apparent activation energy for thermal degradation of ABS in air has been reported in the range of 180–220 kJ/mol in published thermogravimetric studies conducted under ASTM E1641-18 dynamic heating. Consumption of hindered phenolic antioxidants during a single pass can exceed 15–20% of the original stabilizer loading because the stabilizer is sacrificed to terminate peroxy radicals. Phosphite secondary antioxidants are consumed by hydroperoxide decomposition. When the phenolic antioxidant is exhausted, oxidation shifts to autocatalytic acceleration and yellowing measured by ASTM E313-20 increases by 2–4 yellow-index units per pass. This consumption contributes to the cost of each pass because re-stabilization requires compounding with a masterbatch, and overstabilization can cause plate-out on mould surfaces.

At a regrind fraction of 20 wt%, the notched Izod impact value measured at 23°C according to ISO 180/1A:2019 can fall from a virgin range of 22–26 kJ/m² to 15–18 kJ/m² after the first pass. By the third pass, the same material typically records 12–15 kJ/m² or lower, while elongation at break measured according to ISO 527-2:2012 drops from 20–30% to 6–10%. The cost of this loss appears first in downgauging. A moulder producing an enclosure specified at ≥15 kJ/m² can no longer use second-pass regrind above 20 wt% without falling below the permissible lower control limit. The material value sacrificed per pass is therefore not the granulation cost alone; it is the difference between the virgin resin price and the lower-value application where regrind can still be used. For parts with thin-wall sections below 2.0 mm, the permissible regrind fraction can be as low as 10 wt% because weld-line strength retention, measured as the ratio of welded to unwelded tensile strength, can fall below 0.55 after only two passes. Regranulation of sprues and runners typically consumes 0.10–0.25 kWh/kg, and drying adds 0.08–0.15 kWh/kg; these costs are small relative to the resin value loss but are not zero.

The relationship between regrind content and impact loss is not linear. Adding 10 wt% of second-pass regrind often retains 90–95% of the virgin notched Izod, while adding 30 wt% can reduce it by 15–25%. At 50 wt% regrind, the retention may fall below 65%, depending on the original molecular weight of the SAN phase and the severity of the previous pass. This non-linearity is especially important in multi-cavity tools where regrind content varies across shots due to feeder segregation. A twin-screw extruder with gravimetric feeding at 0.5% accuracy can still produce transient regrind-rich zones if the regrind flake size distribution is broader than 2–8 mm. The result is batch-to-batch variation that is measurable as a wider standard deviation in notched Izod, not just a lower mean value.

Representative property drift for injection-grade ABS processed on a co-rotating twin-screw extruder with L/D 36:1, barrel profile 210–240°C, screw speed 300 rpm, and residence time 30–40 s is shown below.

What Does Each Pass Cost in Drying and Injection Moulding Process Capability?

The drying penalty is often underestimated because flake from a granulator with a screen size of 6 mm has higher surface area than virgin pellets and absorbs moisture more rapidly in a humid environment. ABS must be pre-dried to a residual moisture content of ≤0.05% before processing; failure to do so produces splay and silver streaks on moulded surfaces. Desiccant dryers with a dew point of −30°C to −40°C and a drying temperature of 80°C for 2–4 h are required for most injection-grade compounds. Each additional pass increases the drying burden because the material has already been exposed to atmospheric humidity in the granulation and handling stages. The energy input for drying may add 0.08–0.15 kWh/kg per pass.

The more significant process cost is melt viscosity drift. Melt flow index measured according to ISO 1133-1:2022 at 220°C under 10 kg load typically increases from 12 g/10 min in virgin material to 16 g/10 min after two passes and 19–23 g/10 min after four passes for a general-purpose injection grade. In injection moulding, this viscosity drop narrows the processing window to approximately ±5°C of the barrel set point, because higher melt flow causes flash in vented cavities and lower melt flow causes short shots in multi-cavity tools. Shot-weight variability on a 1,200 kN clamping force press can increase from ±0.5% with virgin material to ±1.5–3.0% with high-regrind mixtures, which directly raises scrap rates and reduces the value of the pass.

In injection moulding, a high-MFI reprocessed ABS may fill the cavity at lower injection pressure, which can appear beneficial. However, the lower melt strength increases the risk of jetting, gate blush, and warpage due to reduced packing pressure transmission. Screw recovery time changes because the viscosity drop lowers back-pressure requirements. A processor that does not adjust screw cushion and decompression settings may see drool at the nozzle and inconsistent shot weight. The pass cost therefore includes the trial time required to re-establish a robust process window. On a 1,200 kN press with a 32 mm screw, switching from virgin to 30 wt% regrind with an MFI increase of 6 g/10 min can require reducing barrel set points by 5–10°C and increasing screw back-pressure by 0.5–1.0 MPa. These adjustments are not universal; the specific response depends on the charpy notch sensitivity and the gate geometry, so published data for this specific configuration is limited.

When Regrind Fraction Crosses 30 wt%, Low-Temperature Ductility Shows a Cliff-Edge

For injection-grade ABS containing 15–20 wt% polybutadiene, the room-temperature Izod loss during initial passes is relatively gradual, but low-temperature ductility fails more abruptly. Notched Izod at −30°C according to ISO 180/1A:2019 can lose 40–60% of the virgin value by the third pass, while the same material may retain 80% of its room-temperature value at 20 wt% regrind. This ductile-to-brittle transition is driven by crosslinked polybutadiene domains that can no longer cavitate and trigger shear banding in the SAN matrix. Weld lines become the dominant fracture initiation site. On injection-moulded plaques with a thickness of 2.0 mm, weld-line tensile strength falls below 45% of the unwelded value when regrind content exceeds 30 wt%, as measured by ISO 527-2:2012 specimens cut perpendicular to the weld line. The cost consequence is a step change in reject risk. Components that pass drop tests or falling-dart impact according to ASTM D3763 with virgin material can split at temperatures near 0°C when regrind exceeds this threshold. For flame-retardant ABS, the risk is compounded by migration of brominated flame retardants and loss of UL 94 V-0 performance after two passes, which requires re-testing and can disqualify the regrind stream entirely.

Low-temperature ductility is also influenced by rubber particle size and grafting efficiency retained after grinding. Reprocessing can shift the apparent rubber particle size distribution by breaking down agglomerates and by creating gel particles that act as stress concentrators. In a falling-dart impact test conducted according to ASTM D3763 on 2.0 mm discs, the peak force for virgin ABS may lie in the range of 350–500 N. After three passes with 30 wt% regrind, the failure mode changes from ductile drawing to brittle cracking at 150–250 N. Parts that are loaded below the ductile-to-brittle transition temperature are then exposed to field failures that cannot be detected by room-temperature QC checks alone. This is a hidden cost per pass: a material may pass standard room-temperature lot acceptance tests while losing the low-temperature safety margin that the original part relied upon.

Comparative retention data for general-purpose injection-grade ABS after two passes are summarised below. Values are relative to the virgin compound of the same lot.

Regrind Fraction, Weld-Line Retention, and Shot-Weight Stability on a 1,200 kN Press

On production-scale injection-moulding lines, the cost of reprocessing is most visible in shot-weight stability and weld-line performance rather than in a single mechanical test. A 1,200 kN press running a four-cavity tool with hot-tip gates and a 32 mm screw may accept virgin ABS with a cushion of 3–5 mm. When the same press is fed a blend containing 30 wt% regrind, the reduced melt viscosity can pull the cushion below 2 mm during the transition from injection to packing, producing intermittent sink marks and dimensional variation. The corrective action is not simply to increase shot size; it requires re-profiling the injection velocity, raising the hold pressure, and sometimes lowering the nozzle temperature by 5–10°C. These adjustments represent a pass cost in machine time and rejected parts, especially when the regrind fraction drifts because of granulator flake segregation or manual addition of runners.

Vacuum degassing at −0.08 MPa on a co-rotating twin-screw compounding line can remove volatile oxidation products, but it cannot reverse crosslinking in the butadiene phase or rebuild the molecular weight of the SAN matrix. Re-compounding with 0.1–0.3 wt% of a phenol/phosphite stabilizer package and 1–2 wt% of an ethylene-ethyl acrylate impact modifier can partially offset first-pass losses, but published data for this specific configuration is limited, and the cost of additive replenishment per pass must be compared against the degraded resin value. A regrind stream that retains only 10 kJ/m² notched Izod and an MFI above 25 g/10 min is often unsuitable for injection moulding at wall thicknesses below 2.5 mm; it may be downguaged to non-structural spacers, packaging cores, or filled compounds. Each pass therefore carries a compounding cost, a testing burden, and a reduction in accessible application tier. The technical limit for continued reprocessing is not a fixed cycle number; it is the point at which the material’s residual impact, weld-line, and melt-viscosity properties fall below the lower control limits of the intended moulding cell.