ABS Batch-to-Batch Colour Shift: Measuring It and Setting Tolerance is approached as a measurement and process-control problem rather than a visual-approval exercise. Acrylonitrile butadiene styrene (ABS) does not possess a single intrinsic colour; its CIELAB coordinates depend on polymerisation route, stabiliser package, pigment dispersion, regrind fraction, moisture history, and moulding conditions. Mass-polymerised ABS normally starts with lower residual yellowness than emulsion grades because emulsifier and coagulant residues are lower; suspension grades fall between depending on suspending-agent cleanliness. Titanium dioxide addition from 2.0 wt% to 5.0 wt% in white and pastel compounds raises L* and shifts b* negative, but pigment dispersion quality and base resin yellowness produce batch-specific differences of 0.3 ΔE*ab to 1.2 ΔE*ab even when the TiO₂ feed is unchanged. Compounders using twin-screw extruders with screw diameters from 40 mm to 92 mm and L/D ratios of 40:1 to 52:1 record barrel temperature profiles, screw configuration, and feeder accuracy as direct contributors to final colour coordinates. Visual evaluation under uncontrolled lighting cannot resolve shifts below 1.5 ΔE*ab; therefore instrumental methods based on ASTM D2244-22 are required for commercial batch acceptance.

How Is Batch-to-Batch Colour Shift Quantified in ABS Supply Chains?

Colour difference is calculated from spectral reflectance data using a bench-top sphere spectrophotometer with D65 illuminant and 10° standard observer as specified in ASTM D2244-22 and ISO/CIE 11664-4:2008. Specimens are injection-moulded as plaques of identical thickness, commonly 2.0 mm or 3.0 mm, because ABS translucency changes L* and b* with thickness. The spectrophotometer geometry must be fixed: d/8° sphere instruments can report specular component included or excluded, and the choice alters readings on textured and glossy ABS surfaces. For pigment-strength comparisons, SCI is used because it reduces surface-gloss influence; for visual correlation with a grained automotive panel, SCE is often preferred. Aperture size is fixed at 25 mm for homogeneous plaques and reduced to 8 mm only when small defects are intentionally included in the acceptance criteria. Sample backing, standard tile calibration, and UV filter position are fixed because translucency and optical brighteners in ABS grades can shift reflectance under fluorescent light. Conditioning follows ISO 291:2008 at 23 °C ± 2 K and 50 % ± 10 % relative humidity.

Colour difference is reported as ΔE*ab under ASTM D2244-22 or ΔE*00 under ISO/CIE 11664-6:2014. The ΔE*00 metric is preferred for low-chroma ABS colours because ΔE*ab can overstate the perceptual difference near the neutral axis. Yellowness index is calculated using ASTM E313-20 for natural and white grades where positive b* is the primary failure direction. Metamerism is checked by measuring under at least two illuminants, commonly D65/10° and A/10°, when material must pass visual inspection in both daylight and warm indoor lighting. Three plaques per batch and three measurements per plaque are averaged. Short-term instrument repeatability on a ceramic tile should be below 0.03 ΔE*ab; if repeatability exceeds 0.08 ΔE*ab, the aperture or sample presentation is corrected before limits are applied.

In injection-moulding production, the measured colour of the same ABS batch can shift when drying, barrel residence time, shear heating, and hold pressure change. Residual moisture above 0.03 % in ABS pellets is a process risk: it can produce surface splay, lower melt viscosity, and shift b* positive through additive degradation. Desiccant dryers are set to 80 °C to 85 °C for 2 h to 4 h with a dew point below -30 °C; in warehouses above 60 % relative humidity, drying is mandatory rather than optional. Melt temperature at the nozzle is maintained between 220 °C and 260 °C for general-purpose ABS, while the downstream injection-moulding machine is sized so that shot weight uses 40 % to 70 % of barrel capacity to limit residence time. Long holdup above 260 °C accelerates oxidation of the polybutadiene phase; the resulting yellowing is recorded as positive Δb* and reduced ΔL*, with changes of 0.3 to 0.8 ΔE*ab after extended residence. Hold pressure and injection speed alter plaque density, gloss, and surface structure; because d/8° SCE measurements detect these physical changes, only plaques produced under the same moulding protocol are compared.

Pellet colour measurement is sometimes used for rapid incoming inspection, but it introduces error from packing density, translucency, surface shine, and pellet size distribution. A sphere spectrophotometer with a large-area sample cup can rank pellets, but the CIELAB values cannot be directly compared with injection-moulded plaque values because inter-pellet void space and translucency reduce L* and alter a*, b*. For this reason the reference procedure is to compound a standard formulation, mould a plaque of fixed thickness under controlled conditions, and compare the batch plaque against a master plaque. If pellet inspection is retained for screening, the method is validated by plotting pellet ΔE*ab against plaque ΔE*ab for a minimum of 30 lots; if the correlation coefficient falls below 0.90, pellet-only acceptance is not used.

Tolerance Setting and Instrument Agreement Across Production Sites

A batch tolerance is not a universal constant; it is derived from process capability and end-use appearance limits. For light-coloured ABS used in appliance housings and automotive interior trim, a common acceptance limit is ΔE*ab ≤ 0.5 against the master plaque, with individual coordinate limits of ΔL* ≤ ±0.3, Δa* ≤ ±0.3, and Δb* ≤ ±0.4, because a balanced gray can tolerate less b* drift than a saturated colour. Dark saturated compounds may use ΔE*ab ≤ 0.8, while high-gloss black ABS for visible electronic surfaces may require ΔE*00 ≤ 0.4. The tolerance must include inter-instrument agreement: two calibrated sphere spectrophotometers at different plants can differ by 0.2 ΔE*ab to 0.4 ΔE*ab on the same plaque, so a 0.5 ΔE*ab commercial limit leaves only 0.1 ΔE*ab to 0.3 ΔE*ab for actual batch shift. Process capability is assessed from at least 30 consecutive batches; if the standard deviation of ΔE*ab is 0.15, a ±3σ internal control limit is 0.45 ΔE*ab, which supports a specification of 0.5 ΔE*ab with a Cpk of 1.33 only when the mean is centred.

Texture and gloss are fixed in the boundary standard because L* shifts by more than 0.5 between high-gloss and matte surfaces even with identical chemistry. A gloss meter at 60° according to ASTM D2457-21 is used to separate physical surface changes from pigment shifts. Visual arbitration following ASTM D1729-16 uses D65 daylight and a warm source such as A or TL84; a batch that passes ΔE*ab ≤ 0.5 under D65/10° may fail visually under A/10° if the pigment system is metameric. Thermochromism is another source of apparent batch shift: moulded plaques measured before complete cooling or immediately after exposure to warm machine lamps can show lower L* or higher b* than plaques conditioned at 23 °C. ABS plaques should be cooled to room temperature for at least 15 min after demoulding and conditioned according to ISO 291:2008 before colour coordinates are recorded.

Independent of resin lot, colour masterbatch feed variation is a direct source of batch-to-batch shift. Gravimetric feeders on compounding lines maintain short-term feed-rate stability of ±0.5 % to ±1.0 %; at a 3 wt% masterbatch letdown this translates into pigment-concentration fluctuations that are measurable in L* and b*. Carrier resin viscosity mismatch between the masterbatch and ABS base resin reduces distributive mixing in the extruder and produces undispersed titanium dioxide or carbon black agglomerates. Filter-pressure tests on a 5 μm to 15 μm screen pack or optical microscopy of compounded film detect these defects; integrated sphere reflectance alone may under-report speck defects because the aperture averages them. For dark ABS, carbon black dispersion is controlled by sieve residue and pressure-rise viscosity; for white ABS, TiO₂ dispersion is assessed by gouge or drawdown tests. If the masterbatch carrier has a melt flow rate more than 10 g/10 min higher than the base ABS under ISO 1133-1:2022 at 220 °C and 10 kg, local viscosity dilution can shift surface appearance without a corresponding change in total pigment content.

When Recycled ABS Streams Narrow the Acceptable Pigment Tolerance

Recycled ABS narrows the colour tolerance because the butadiene phase has already consumed part of its antioxidant package and because mixing of different base colours introduces uncontrolled pigments. A regrind fraction of 10 wt% to 20 wt% in a light-grey ABS can raise b* by 0.4 to 1.0 and reduce L* by 0.2 to 0.5 before masterbatch correction. In dark colours, the same regrind fraction may increase a* or b* variability because minor colour history differences are not hidden by high pigment loading. When the acceptance limit is ΔE*ab ≤ 0.5, the recycle fraction should be controlled within ±0.2 wt% and the feedstock should be sorted by melt flow and initial colour before blending. Thermal stabiliser adjustment is required for higher recycle levels; without it, reprocessing at 240 °C to 260 °C produces yellowing measurable as positive Δb* and higher yellowness index under ASTM E313-20. For high-gloss ABS with a colour tolerance of ΔE*00 ≤ 0.4, the melt-temperature window around the validated set point is often held within ±5 °C because the butadiene oxidation rate is temperature-sensitive and rapidly shifts b*. For light-coloured medical or electrical enclosure grades, recycled content is often capped below 5 wt% because the tolerance headroom is insufficient for lot-to-lot colour variation; published data for this specific configuration is limited, but production-scale laboratories use this cap as a conservative boundary.